The Oulipo and Modern Thought 0198831633, 9780198831631

The impact of the Oulipo (Ouvroir de Litterature Potentielle), one of the most important groups of experimental writers

1,110 75 25MB

English Pages x+175 [186] Year 2019

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

The Oulipo and Modern Thought
 0198831633, 9780198831631

Citation preview

OUP CORRECTED PROOF – FINAL, 11/6/2019, SPi

THE OULIPO AND MODERN THOUGHT

OUP CORRECTED PROOF – FINAL, 11/6/2019, SPi

OUP CORRECTED PROOF – FINAL, 11/6/2019, SPi

The Oulipo and Modern Thought DENNIS DUNCAN

1

OUP CORRECTED PROOF – FINAL, 11/6/2019, SPi

3

Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © Dennis Duncan 2019 The moral rights of the author have been asserted First Edition published in 2019 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2019931584 ISBN 978–0–19–883163–1 Printed and bound by CPI Group (UK) Ltd, Croydon, CR0 4YY Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

OUP CORRECTED PROOF – FINAL, 11/6/2019, SPi

For Claire

OUP CORRECTED PROOF – FINAL, 11/6/2019, SPi

OUP CORRECTED PROOF – FINAL, 11/6/2019, SPi

Table of Contents List of Figures

Introduction: The Secret of Lightness

ix 1

1. Literature Machines: StructurElism versus StructurAlism

27

2. The Punning Muse: Psychoanalysis and Homophonic Translation

51

3. Surrealism’s Subject: Two Cohorts of the Oulipo

76

4. Calvino at a Crossroads: Combinatorics and Anticombinatorics 100 5. Perec and Mathews: Translation and Analytic Philosophy in the South Seas Conclusion Bibliography Index

122 146 157 167

OUP CORRECTED PROOF – FINAL, 11/6/2019, SPi

OUP CORRECTED PROOF – FINAL, 11/6/2019, SPi

List of Figures 1. Raymond Queneau (second from left) in 1952 with his son, Jean-Marie, and his wife Janine (far right). In between them are Jean-Paul Sartre and Simone de Beauvoir. 2. Bénabou’s completed slip requesting that his friend Jacques Lacan should receive a complimentary copy of La Lipo. (FO, DM-3 (28)). 3. An early Oulipo meeting in François Le Lionnais’s garden. Le Lionnais and Queneau sport, respectively, a sailor’s cap and a top hat, symbols of their ‘benign dictatorship’ over the group they founded. 4. Georges Perec’s handwritten minutes to the group’s meeting of January 1971 (FO, DM-3 (1)). 5. Early printed output from the first version of the ‘Love Letters’ algorithm, showing Strachey’s handwritten correction. 6. Photograph of street furniture in Strasbourg displaying one of the group’s punning narratives. (FO, MS-3, f. 99). 7. Bonne année pataphysique [Happy Pataphysical New Year]. Postcard with two homophonic translations (‘Housemaid has Pataphysical nose’; ‘Good ass shocks boy (sic)’) to mark the Pataphysical new year, 96 E.P. (8 Sept 1968). On the reverse of the card sent by Stanley Chapman to Le Lionnais, Paul Gayot has added ‘Intraduisible [untranslatable] in Englisch’ (FO, DM-2 (30)). 8. Ramón Llull, Illuminati sacre pagine professoris amplissimi magistri Raymundi Lull, ars magna, generalis et ultima (Lyon: J. Mareschal, 1517) sig. B1v. 9. Marcel Bénabou, ‘The Three Circles of Lipo’ (FO, MS-6 f. 14) 10. Marcel Bénabou’s handwritten minutes for the meeting of 8 November 1972 record Perec’s outline of what would become La Vie mode d’emploi and the long discussions that followed it. Oulipo, ‘Compte rendu autographe de Marcel Bénabou de la réunion du 8 novembre 1972’ (FO, DM-3 (24)).

6

6

8 18

29 65

66

103 104

128

OUP CORRECTED PROOF – FINAL, 11/6/2019, SPi

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

Introduction The Secret of Lightness Quintilian, the first-century Roman rhetorician, has a story about Plato composing The Republic. The way Quintilian tells it, the philosopher agonized over the first four words of his most famous work. He knew which words he wanted to use, but couldn’t decide the best order to put them in. It’s not that The Republic begins with any profoundly complex thought: κατέβην χθὲς εἰς Πειραιᾶ, the opening translates as ‘I went down yesterday to Piraeus’. Nevertheless, according to Quintilian, these four words ‘were found written on [Plato’s] tablets in many different orders’.¹ Great philosophy, Quintilian wants us to know, depends on what we might consider to be literary qualities: rhythmical phrasing, a strong opening. In other words, it depends on rhetoric. Thus, we are asked to picture the philosopher playing with permutations, arranging words like unhewn stones, trying to find their best fit, knowing that there is power in their ordering as well as in their meaning. In this book, I want to consider these same elements—literature, philosophy, the play of combinatorics—but in a different order. Rather than the great philosopher’s painstaking concern for the literary quality of his work, this is a book about the types of thought encoded in the work of a group whose very name declares their interest to be primarily literary. The Oulipo—OUvroir de LIttérature POtentielle—are known best for their investigations into the use of mathematical or stylistic constraints, for procedures often involving permutations like Plato’s above, in the production of literature. Although academic interest in the group has grown considerably over the last decade, in the anglophone world at least their engagement with modern thought, with the debates of their academic contemporaries, has been largely overlooked.² This book, then, is an ¹ Quintilian, The Orator’s Education, trans. by D. A. Russell, 5 vols (Cambridge, MA: Harvard University Press, 2014) IV, 8.6.64 (p. 463). ² In France the situation is slightly different. Camille Bloomfield’s recent Raconter l’Oulipo (1960–2000): histoire et sociologie d’un groupe (Paris: Honoré Champion, 2017),

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

2

The Oulipo and Modern Thought

attempt to redress that balance, to use the evidence of meeting minutes and letters, as well as published works, to paint a different picture of the Oulipo, one which casts the group in the light of some of the intellectual currents of the twentieth century: psychoanalysis, structuralism, analytic philosophy, Surrealism. Oulipian work, almost by definition, is rigorous and painstaking in its construction; what I hope to show is that, very often, these are not the only characteristics it shares with Plato. At their monthly meeting in March 1963, the Oulipo are discussing a nineteenth-century writer named J.-A. Révéroni. Révéroni, it seems, had devised a method of composing sermons using combinatorics, whereby a small number of elements could be shuffled to produce a large number of lessons. Naturally, this would be an interesting precedent for a group set up to explore mathematical approaches to literary creation, but unfortunately at the Oulipo’s meeting no-one in the room has detailed knowledge of how Révéroni’s process—his so-called matrix—works; the tip-off has come from the philosopher Michel Foucault.³ After some discussion about Révéroni’s life, the following exchange occurs: - Good. I can ask Michel Foucault, which is perhaps linked to my other intervention. I propose that we add Michel Foucault to our list of future invitees. You know he is writing a book, that’s going to come out, on Roussel. I mean the problems of multiple parentheses, etc. - Foucault? A U L T? - A U L T, yes. - I would have spelled it like the priest [i.e. Charles Foucauld]! - Each to his own!! - No, but anyway, he is the author of a book on madness . . . - . . . which is very good. - History of Madness, which is not without interest. - He is the author of History of Madness. He is now on the editorial board of Critique. - He works at Bizarre? - Yes, and now he is on the editorial board of Critique. - He’s not a psychiatrist, is he? - Gentlemen, you have before you a proposal. What is your decision? Your vote . . . for example, is a superb, archival historicization of the Oulipo in its first decades, and will surely become essential for researchers working on the group. ³ Foucault had recently been researching Révéroni, and had published a short article on him the previous year: Michel Foucault, ‘Un si cruel savoir’, Critique 182 ( July 1962): 597–611.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

The Secret of Lightness

3

- About the invitation? - Yes. - Everyone is in agreement? - No-one objects? - No . . . - The proposal is adopted.⁴ Preserved in the Oulipo archives at the Bibliothèque National in Paris, this is a remarkable piece of minute-taking, giving us a wonderful insight into the way these meetings were conducted. The speakers aren’t identified (though the first speaker, knowing that Foucault’s book on Roussel is coming up for publication with Gallimard, is likely Raymond Queneau, a senior editor there), but this only intensifies the sense of chaos about the dialogue: people interrupt, talk over each other; there are repetitions, preemptions, jokes; the tone switches in an instant from erudite to bureaucratic to whimsical; some of the group speak admiringly of Foucault’s work, others can’t even spell his name. And in spite of this informality— this joyous, convivial polyphony—something formal nevertheless takes place: a vote is held, and the Oulipo decide unanimously to invite Michel Foucault to come and dine with them. The meeting with Foucault sadly never took place. Foucault’s name appears again some months later on a list of people to invite, before slipping off the record, an unticked item on the group’s ever-expanding To Do list.⁵ What this incident reminds us, however, is that, in the early

⁴ [‘- Ben . . . Je peux demander à Michel Foucault, c’est peut-être lié à mon autre intervention; je propose de mettre Michel Foucault sur la liste de nos invités futurs. Vous savez qu’il a un livre en préparation, qui va sortir, sur Roussel. Je veux dire les problèmes de parenthèses multiples . . . etc . . . / – Foucault? A U L T? / – A U L T. Oui. / – Moi, j’aurais dit, comme le Père! / – Chacun ses convictions!! / – Non, mais enfin, c’est l’auteur d’un livre sur la Folie . . . / – . . . qui est très bon. / – L’Histoire de la Folie; . . . qui ne manque pas d’interêt. / – C’est l’auteur de L’Histoire de la Folie. Il fait partie maintenant du comité de rédaction de Critique. / – Y collabore à Bizarre? / – Oui et il fait partie mainenant du comité de rédaction de Critique. / – C’est pas un psychiâtre, non? / – Vous êtes devant une proposition, Messieurs, quelle est votre décision? Votre vote . . . / – A propos de son invitation? / – Oui. / – Tout le monde est d’accord? / – Pas d’opposition? / – Non . . . /La proposition est adoptée.’] ‘Photocopie du compte rendu dactylographié de la réunion du 22 mars 1963’ (Paris, BnF, March 1963). Fonds Oulipo. Dossiers mensuels de réunion (1960–2010), Box DM-1 (30). Subsequent citations for Fonds Oulipo archival material will be given in the following format: ‘(FO, [box] [folder])’. In the case of the ‘Dossiers mensuels’ boxes, i.e. those with a code starting in DM, scanned images are available to view online via the BnF’s Gallica viewer: [accessed 8 August 2017]. ⁵ A similar fate befalls Roland Barthes, who is listed as a potential invitee in August 1970, but never makes an appearance in person. [‘Pré-programme dactylographié du congrès d’août 1970 avec ajouts autographes de François Le Lionnais’ (FO, DM-2 (54)).]

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

4

The Oulipo and Modern Thought

1960s, Foucault and the Oulipo were part of the same milieu—the discussion grows out of a face-to-face encounter, a casual conversation, between the philosopher and one of the group.⁶ Perhaps it is no surprise in the case of the veteran Queneau—cofounder of the Oulipo, former Surrealist, eminent Pataphysician, prominent editor—that his links to the world of philosophy should run deep.⁷ But Queneau was far from the only Oulipian with interests or connections of this kind. To take another example, then, this time concerning Jacques Lacan’s often overlooked skills as a babysitter, Marcel Bénabou, who joined the Oulipo in 1970, relates the following surprising tale: I remember a dinner party at ours where we’d invited Lacan and my son David, who was then two or three, suddenly woke up crying. Lacan insisted on going to calm him down, and with a slow, dignified step he ascended the staircase leading to the child’s room. My wife and I feared the worst but we let him go and waited anxiously for the result of this improbable encounter – our son was by no means an easy child. We did not have to wait long. A few moments later, we could see a smiling Lacan coming down those same stairs with a perkiness in his step. With a rapidity that provoked our admiration, he had managed to put the child to sleep. We never knew how he did it.⁸

What anecdotes like these illustrate is how little the Parisian intellectual scene during the 1960s and 70s was siloed by profession. Analysts and academics, writers and publishers: between these groups, and others, relations existed that were often both professional and personal. Ideas from one field were adopted—and critiqued—in another. Sometimes these links were formal, enshrined in the structured interdisciplinarity of an academic seminar or research centre; at other times they were not, as ⁶ [‘J’ai rencontré récemment Michel Foucault qui a publié une petite étude sur Réverony de Saint-Cyr et qui m’a assuré qu’il avait une matrice à sermons.’] ⁷ For more on Queneau and philosophy, see Jean-Charles Chabanne, ‘Queneau, Les Temps Modernes, Sartre’, in Temps mêlés—Documents Queneau 150+33–36 (1987): 355–61. ⁸ [‘Je me souviens que, au cours d’un dîner chez moi auquel Lacan était convié, mon fils David, qui avait alors deux ou trois ans, se réveilla soudain en pleurant. Lacan insista pour aller lui-même le calmer, et monta d’un pas lent et digne l’escalier qui menait à la chambre de l’enfant. Bien que craignant le pire, ma femme et moi le laissâmes faire, et attendîmes, anxieux, le résultat de cette improbable rencontre, car notre fils n’était pas du genre facile. Nous n’attendîmes pas longtemps. Quelques instants plus tard, nous pûmes voir Lacan souriant descendre le même escalier d’un pas beaucoup plus guilleret. Avec une rapidité qui provoqua notre admiration, il avait réussi à endormir l’enfant. Nous n’avons jamais su ce qu’il avait pu faire pour cela.’] Marcel Bénabou, ‘La Galère ou Pourquoi j’ai participé à la confection du volume intitulé 789 néologismes de Jacques Lacan’, in L’amour de loin du docteur L. (Paris: L’Unebévue, 2004), pp. 27–31 (p. 28). Meanwhile, the curious tale of Jacques Roubaud’s personal connection to Lacan is the subject of his tiny book, Ma vie avec le docteur Lacan (Bordeaux: L’Attente, 2004).

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

The Secret of Lightness

5

when the Oulipian Noël Arnaud describes the ‘delightful correspondence’ he shared during the 1960s with Claude Lévi-Strauss.⁹ In fact, so embedded were the Oulipo within the broader intellectual scene that at times the group struggled, even among themselves, to make the case for what was unique about their identity. This existential crisis surfaces occasionally in the meeting minutes, for example in Queneau’s anxiety that the writers of the Tel Quel group or the nouveau romanciers might be doing essentially the same work as the Oulipo: ‘A number of writers now – notably those associated with the nouveau roman, or the Tel Quel group, or Change – are looking at the use of refined, sometimes quite intricate, constructions. Is their research any different from ours?’¹⁰ A group, then, who are frequently in the company—both socially and intellectually—of the figures we now identify as the leading philosophers of their day, the chief exponents of theory, as we have come to call it: what I wish to argue in this book is that we can find evidence of these relations, of this engagement with the thought of the time, everywhere in the group’s work. In other words, we can read a great deal of Oulipian writing—from the 1960s and 70s, at least—as a creative participation in a variety of prominent intellectual debates. Arnaud in his ‘Prolegomena to a Fourth Manifesto’ makes a telling claim about the 1960s: ‘Academe, in the Oulipo’s first decade [ . . . ] had just begun – in France at least – its infatuation with Surrealism and psychoanalysis. In the higher spheres, structuralism applied to analysis and literary creation had begun to consolidate its dogma’.¹¹ What is striking is how we can find exactly these concerns—Surrealism, psychoanalysis, structuralism—encoded in the literary creation of the Oulipo during the same period. And as the intellectual climate changes—as it moves from structuralism to poststructuralism— naturally we can see this too reflected within the group’s minutes and their literary production. In one sense there should be nothing surprising about this—I hope the above summary makes the case seem uncontroversial. Nevertheless there has been a tendency in much writing about the Oulipo, much anglophone writing at least, both to flatten time and to overlook place. The group’s ⁹ Noël Arnaud, ‘Foreword: Prolegomena to a Fourth Oulipo Manifesto – or Not’, in Oulipo: A Primer of Potential Literature, ed. by Warren F. Motte (Lincoln, NE: University of Nebraska Press, 1986), pp. xi–xv (p. xiv). ¹⁰ [‘En effet, plusiers écrivains aujourd’hui, notamment ce qui se réfèrent au nouveau roman, ou à l’équipe de Tel Quel, ou à celle de Change, sont attentifs à l’utilisation de constructions recherchées, parfois délicates. Leurs recherches diffèrents-elles de nôtres?’]. ‘Copie du compte rendu dactylographié de la réunion du 27 août avec corrections de Jacques Bens et François Le Lionnais’ (FO, DM-3 (8)). ¹¹ Arnaud, ‘Prolegomena’, p. xiv.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

6

The Oulipo and Modern Thought

Fig. 1. Raymond Queneau (second from left) in 1952 with his son, Jean-Marie, and his wife Janine (far right). In between them are Jean-Paul Sartre and Simone de Beauvoir. Image reproduced by kind permission of Gallimard/Queneau estate.

Fig. 2. Bénabou’s completed slip requesting that his friend Jacques Lacan should receive a complimentary copy of La Lipo. (FO, DM-3 (28)). Image reproduced by kind permission of the Oulipo.

development, and the context in which it occurred, are often ignored, while Oulipian whimsicality is given a prominence that obscures the sense of purpose that has always run alongside it. (Remember how, in the dialogue about Foucault above, through the cacophony of interruption and wordplay, the business of the meeting still proceeds relatively fluently.)

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

The Secret of Lightness

7

On 25 November 1960, the day after the group’s first meeting, Jacques Bens circulated a memo among the eight who had attended. In it he asks a question which perfectly captures the balance of joviality and seriousness of intent: ‘Seeing as we aren’t meeting just to amuse ourselves (although that is, of course, a considerable part of it), what can we expect from our work?’¹² Or we might think of Arnaud again, who points out that, in several regards, the difference between the Oulipo and the structuralists was more one of tone or attitude than of method, concluding that ‘the [structuralists] enveloped themselves in a ponderous sobriety that rendered them impervious to Oulipian facetiae’.¹³ If the difference is simply one of sobriety versus ribaldry, then we need to be able to look past this— to look, in fact, for the shared structures underneath. Alongside this, the make-up of the group (for example, the presence of the mathematician Claude Berge among its founder members) plus the convenience of framing the Oulipo in terms of the ‘Two Cultures’ debate—as a rare bridge between mathematics and literature—makes it easy to downplay the status of philosophy (broadly conceived) in their concerns. For these reasons, I hope that a book that frames the early Oulipo in terms of their engagement with forms of thought that were prominent in the 1960s and 70s, that reads their work in the light of structuralism, psychoanalysis, and Surrealism, and that draws on the wealth of materials in the Fonds Oulipo—the wonderful set of archives held at the Arsenal site of the Bibliothèque National in Paris—to chart the Oulipo’s development, not just in terms of new members but in terms of new intellectual priorities, will be a valuable contribution to the growing body of work on this extraordinary group.¹⁴ Yet, even now as I read back through those last few paragraphs, I am conscious that they seem to tip the scales too far in the other direction, wringing the joy out of the Oulipo in favour of po-faced philosophical analysis. Of course, there is a sense of fun that is almost a given in the published output of the Oulipo, and this mustn’t be overlooked. Even for a master like Perec, whose greatest works—La Disparition (1969) and La Vie mode d’emploi (1978)—draw deeply on the loss of his parents in the

¹² [‘Considérant que nous ne nous réunissons pas seulement pour nous divertir (ce qui est déjà considérable, certes), que pouvons nous attendre de nos travaux?’]. Oulipo, ‘Circulaire n 1 dactylographiée de Jacques Bens du Séminaire de Littérature Expérimentale du vendredi 25 novembre 1960’ (FO, DM-1 (2)). ¹³ Arnaud, ‘Prolegomena’, p. xiv. ¹⁴ The archives include the minutes that were circulated after each meeting and which often display the handwritten annotations of particular members. For a history of these archives, see Claire Lesage, ‘L’archive, miroir de l’Oulipo’, in Oulipo, ed. by Camille Bloomfield and Claire Lesage (Paris: BnF/Gallimard, 2014), pp. 66–8.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

8

The Oulipo and Modern Thought

war—his father in battle; his mother in the camps—the sadness of the material is thrown into relief by the pleasure of experiencing the text as a piece of exquisitely crafted machinery. What jumps off the page when looking at the minutes of the group’s monthly meetings is the spirit of conviviality—the wordplay, the in-jokes, the mock ceremoniousness— with which these events are conducted. How then to write an intellectual account of these farceurs? Will it not be, at best, a fundamental misunderstanding, at worst, an act of bad faith? One early commentator would think not. The writer Guy Le Clec’h, who had seen the group up close as a guest at one of their early meetings, cuts straight through the bonhomie when he writes an introduction for the 1964 Oulipo special issue of Temps mêlés: ‘The deliberately humorous aspect of this type of research should not however mask its seriousness.’¹⁵ There is a seriousness but one that is couched in lightness. Le Clec’h’s term mask is an apt one, and an incident from the July 1963 meeting is

Fig. 3. An early Oulipo meeting in François Le Lionnais’s garden. Le Lionnais and Queneau sport, respectively, a sailor’s cap and a top hat, symbols of their ‘benign dictatorship’ over the group they founded. Image reproduced by kind permission of Gallimard/Queneau estate.

¹⁵ [‘Que l’aspect volontairement humoristique de ce genre de recherches n’en masque pourtant pas le sérieux.’] Guy Le Clec’h, ‘Introduction’, Temps mêlés, 66–67 (1964): i–iv (iv).

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

The Secret of Lightness

9

telling. Discussing the S+7 method—an important Oulipian procedure for making new texts out of old ones—Jean Queval remarks slyly that what it produces is ‘un peu “farce”’: a little joke.¹⁶ Queneau is offended. He states that this is the first time—in nearly three years of meetings— that the word farce has been used to describe the Oulipo’s work. At this point, the academic Albert-Marie Schmidt intervenes, noting that theologians use the term in an entirely serious sense, and this observation restores the peace. It is a strange little scene, but a rich one when thinking about how we should measure the seriousness of the Oulipian project. At face value, Queneau is hurt that Queval should describe one of the group’s flagship methods as if it were merely whimsical, a bit of a joke. No-one has said this before—at least not at a meeting, in front of the others—either about S+7 or about any of their other works. As Queneau would have it, then, in spite of the rambunctious tone of their gatherings, the group has always been completely in earnest about the seriousness of their activities. Alternatively, contra Le Clec’h, perhaps seriousness—not humour—is the mask that has been maintained with utter scrupulousness for the previous two-and-a-half years. But letting it slip, even briefly, is potentially fatal for the whole enterprise. The moment Alice denounces her tormentors as ‘nothing but a pack of cards’ is the moment she awakes and the story ends. No wonder Queneau is quick to pull Queval up on this. The way that this eruption of scepticism is handled and amicability restored is typically Oulipian. Schmidt’s contribution is erudite and technical. He points to a precise use of language from an obscure field of knowledge (surely if there is any way to mollify an irate Oulipian then this is it). But what is so perfect about Schmidt’s solution is that, in informing the group that sometimes farce can be serious, he is giving them an academic precedent for something they already knew—something they have been acting out for the past three years (longer still for the Pataphysicians among them). We can have it both ways; we don’t have to choose. It is not even necessary to know when the farce is serious and when it is farcical. Not serious in tone, then, but serious in its aims? Perhaps even that is too leaden, too restrictive a definition of the group. But it will do for now. And as for the other side of the coin—the anarchic, the comic—it will be worth drawing a distinction between lightness and whimsicality. The latter—the throwaway, the droll—can be skimmed off the top of the group’s output with little disruption; the former is a defining ¹⁶ Oulipo, ‘Compte rendu dactylographié signé «Ythier Marchant » de la réunion n° 35 du 1e juillet 1963’ (FO, DM-1 (34)).

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

10

The Oulipo and Modern Thought

characteristic. A decade after Queval’s potentially cataclysmic faux pas, a serio-comic Oulipo at the peak of its powers will welcome Italo Calvino to its membership. And it is Calvino, at the end of his life, who will enshrine lightness—la leggerezza—as the model for our times: ‘Were I to choose an auspicious image for the new millennium, I would choose [ . . . ] the sudden agile leap of the poet-philosopher who raises himself above the weight of the world, showing that with all his gravity he has the secret of lightness.’¹⁷ Not merely a poet, or a philosopher, but a poet-philosopher: we are back with Plato shuffling his four words, looking for the perfect arrangement; or perhaps with Queneau, the very model of the literary gentleman, diligently transcribing a lecture series on Hegel.¹⁸ And not merely lightness, but a willful lightness, one performed in spite of one’s own gravity and the weight of the world. In a book which seeks to focus on the seriousness of the group’s activities—to link their output and intentions with those of their contemporaries less touched with the gift of lightness—I hope not to do Calvino’s image the disservice of making Oulipian play seem like labour, or of treating Oulipian farce always with a theological gravity. * * * Paris during the Nazi occuptation: two men—François Le Lionnais, a chemical engineer, and Raymond Queneau, a writer and editor at Gallimard—meet in the literary milieu of Saint-Germain-des-Prés. Their backgrounds are rather different: a native Parisian and the child of a noted concert pianist, Le Lionnais was raised in a highly-cultured haut bourgeois household; Queneau, from the coastal town of Le Havre, had come to Paris as a student and worked his way up the literary ladder with journalistic and translation work alongside his novels and poetry.¹⁹ Nevertheless, ¹⁷ Italo Calvino, Six Memos for the Next Millennium, trans. by Patrick Creagh (London: Cape, 1992), p. 12. ¹⁸ Alexandre Kojève’s celebrated seminar on Hegel’s Phenomenology of Spirit, delivered in Paris in the 1930s and attended by many of the leading intellectuals—including Blanchot, Merleau-Ponty, and Lacan—was subsequently published from Queneau’s notes. Alexandre Kojève, Introduction à la lecture de Hegel, ed. by Raymond Queneau (Paris: Gallimard, 1947). ¹⁹ For more on Le Lionnais’s childhood, see Un Certain disparate, a series of interviews conducted with him in 1976 by Jean-Marc Levy-Leblond and Jean-Baptiste Grasset. Intended as a book—it even got as far as having an extensive index compiled (now in the archives at the BnF: FO, MS-5)—the project was shelved for thirty-five years before being resurrected by Olivier Salon and Anne Garréta as a series of blogs: [accessed 9 August 2017]. For more on Queneau’s childhood, see Michel Lécureur, Raymond Queneau (Paris: Belles Lettres, 2002). The most comprehensive account of the group’s origins and foundation can be found in Bloomfield, Raconter l’Oulipo (pp. 71–223).

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

The Secret of Lightness

11

both share passions that cross the ‘Two Cultures’ divide: they are both keen chess players and amateur mathematicians; both, too, are serious in their commitment to literature and art. A friendship is struck up, one that will last their lifetimes, and will give rise, nearly twenty years later, to the Oulipo. When Le Lionnais, a member of the Resistance, is captured and sent first to Buchenwald then to Dora, it is art that sustains him. In the camp, he and his friend Jean Gaillard keep their spirits up by recalling, in minute detail, their favourite paintings: Thus did we contemplate at length in our minds’ eyes Van Eyck’s Rolin Madonna. I projected, as though with a magic lantern, the severe expression of the donor, the rabbits crushed under the columns, the drunkenness of Noah depicted atop a capital, the little tufts of grass growing between the stones in the courtyard and the six steps leading to the terrace, each detail of the circulating stream and of the urban agitation at its base. The tragically interlaced diagonals of Giotto’s St. Francis of Assisi Receiving the Stigmata disquieted him; the tender, delicious Decapitation of Cosmas and Damian by Fra Angelico charmed him. We took long excursions through The Temptation of Saint Anthony by Hieronymous Bosch (which hangs in Lisbon), through Da Vinci’s The Virgin of the Rocks, [ . . . ], through Dürer’s Melencolia I.²⁰

In these recollections we can see not only the seeds of the Oulipo’s miniaturism, but of art—and of play—as a matter of life and death, as something that could keep at bay the otherwise unendurable reality of the camps. Le Lionnais would return from Dora; Gaillard would not. Back in Paris, Le Lionnais and Queneau resume their friendship. Le Lionnais publishes the collection Les Grands Courants de la pensée mathématique (1948) while Queneau begins to have significant literary success. First of all with Exercices de style (1947), a stylistic romp in which the same short, banal piece of narrative (an argument on a bus followed by a conversation about a raincoat) is rendered in ninety-nine different ways, from Homeric epic to courtroom drama to butchershop slang. Then in 1959 with Zazie dans le métro, a novel whose insouciant teenage diction, rendered phonetically, became a national sensation, with Elle magazine reporting that ‘the Zazie phenomenon is ravaging France like an epidemic. In the streets and on the metro, from the mountains to the beaches, we are all “speaking Zazie”.’²¹ ²⁰ François Le Lionnais, ‘Painting at Dora’, trans. by Daniel Levin Becker, The Believer 11.5 (2013): 27–31 (29). ²¹ [‘Le phénomène Zazie a ravagé la France comme une épidémie. Dans les rues et dans le métro, à la montagne et sur les plages, on “parle Zazie”.’] ‘L’Insupportable Zazie va conquérir l’Amérique grâce à une petite parissienne’, Elle (France) #714 (31 August 1959): 66–7 (66).

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

12

The Oulipo and Modern Thought

The next year finds Queneau working on a new book, Cent milles milliards de poèmes (hereafter CMMP). The work, in Paul Fournel’s lovely phrase, is ‘une espèce de self-service poétique’ [‘a kind of poetical self-service’].²² It consists of ten poems, all sonnets, each printed on consecutive recto pages. Between each line of poetry, the page is cut so that the individual line—rather than the whole page—can be turned over, revealing the corresponding line of the sonnet beneath. Thus, as with children’s books of heads, bodies and legs, a large number of combinations—10¹⁴ to be precise—can be derived. Queneau’s task as author is to write his ten base sonnets so that any combination will be grammatically correct and conform to the sonnet rhyme scheme. But the difficulty of the exercise is causing Queneau to lose interest: I had written five or six of the sonnets for Cent mille milliards de poèmes and was hesitating a little about carrying on, that is, I didn’t have the strength to go on, the more I went along, the more difficult it was to do naturally, when I met Le Lionnais, who was a friend, and he suggested a sort of research group for experimental literature. That gave me the encouragement to carry on with my sonnets.²³

The ‘research group’ Le Lionnas proposes, of course, will become the Oulipo. Nevertheless, while Le Lionnais and Queneau will forever be identified as the group’s official founders, or co-fondateurs, it seems that the idea for the Oulipo might not have originated with Le Lionnais. In September 1960, a décade—a ten-day colloquium—was held at Cerisy-la-Salle in northern France. The conference theme was Raymond Queneau and the French language, and it was here that two of the attendees, André Blavier and Jacques Bens formed the notion of a secret society devoted to experimental writing. As Blavier tells it: ‘one night, Jacques Bens and I were unable to sleep and had the idea of proposing to RQ and F. Le Lionnais a sort of secret society to promote the kind of literature we loved.’²⁴ Two ²² Paul Fournel, Clefs pour la littérature potentielle (Paris: Lettres Nouvelles, 1972), p. 8. ²³ [‘Je suis associé dans une sorte de group de recherches, que est intitulé OULIPO, c’est-à-dire OUvroir de LIttérature POtentielle, qui a été fondé par Le Lionnais. J’avais écrit cinq ou six des sonnets des Cent mille milliards de poèmes, et j’hésitais un peu à continuer, enfin je n’avais pas beaucoup le courage de continuer, plus cela allait, plus c’était difficile à faire naturellement, quand j’ai rencontré Le Lionnais, qui est un ami, et il m’a proposé de faire une sorte de groupe de recherches de littérature expérimentale. Cela m’a encouragé à continuer mes sonnets’]. Georges Charbonnier, Entretiens avec Raymond Queneau (Paris: Gallimard, 1962), p. 116. ²⁴ [‘une nuit, on n’arrivait pas à dormir Jacques Bens et moi, on a eu l’idée de proposer à RQ et F. Le Lionnais une sorte de société secrète pour favoriser le genre de littérature que nous aimions’]. Quoted in Camille Bloomfield, ‘Histoire de l’Oulipo: Quelques jalons chronologique’, in Bloomfield and Lesage, pp. 29–38 (p. 31). The original appeared in the Belgian paper Les Wallons on 21 January 1995.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

The Secret of Lightness

13

months later, when the group hold their first official meeting—at the Vrai Gascon restaurant, 82 rue du Bac, on the Left Bank in Paris—six of the eight attendees had been present at Cerisy. The group begin to meet monthly thereafter—a practice which has continued almost uninterrupted to this day. With an extraordinary prescience, from the earliest days, detailed minutes were taken, typed up, and circulated—giving us a wonderful insight into the workings of a group whose international significance, of course, could not have been apparent at the time. In the first few meetings, a number of significant decisions are taken. Firstly, the eight members soon become ten, with two representatives from Collège de ’Pataphysique—Latis and Noël Arnaud—invited to join.²⁵ Secondly, the group undergo a name change, having identified itself initially as the Seminaire de Littérature Experimentale (abbreviated to Sélitex or SLE). By the second meeting, this title has been found wanting, and another, the Ouvroir de Littérature Potentielle, is voted more appropriate to the group’s specific aims. This name appears in abbreviated form—the Olipo—throughout that meeting’s minutes. Next time the group meet up, however, Latis suggests a preferable contraction: the Oulipo finally has its name. But how should we translate that word Ouvroir that gives the Oulipo its ou? Discussing the term, Arnaud notes that it has fallen into disuse in modern French, but besides denoting a shop or a workroom, it once meant a sewing circle in which well-off ladies would carry out sumptuous needlework for the benefit of the poor. ‘Curiously enough,’ he adds, in enthusiastically adopting the name, ‘it was this last notion, the “sewing circle,” that prevailed in the minds of the Oulipians.’²⁶ Ouvroir implies a certain modesty in the group’s aims, an anti-theoretical self-image which, as we will see in the next chapter, sets them apart from the structuralists. As Bénabou puts it, ‘[t]he notion of the ouvroir had been introduced precisely because it was a place of artisanal work. It wasn’t about inventing grand theories, it wasn’t about setting out to conquer something or other.’²⁷ There is even a sense of benevolence bound up in the idea of ²⁵ Founded in 1948 and based on an idea mooted in Alfred Jarry’s posthumouslypublished novel Gestes et opinions du docteur Faustroll, pataphysicien (1911), the Collège de ’Pataphysique is another long-running avant-garde literary grouping. For the best English introduction to the Collège, see Alastair Brotchie (ed.), A True History of the College of ’Pataphysics: With Manifestos, Statutes, Calendar and Documents, trans. by Paul Edwards (London: Atlas, 1995). Latis, meanwhile, was just one of the various identities of Emmanuel Peillet (1914–73), others being Jean-Hugues Sainmont, Dr Irenée-Louis Sandomir, Oktav Votka, Mélanie Le Plumet, and more. ²⁶ Arnaud, ‘Prolegomena’, pp. xi–xii. ²⁷ [‘La notion d’ouvroir a été précisément introduite parce que c’était un lieu de travail artisanal. Il ne s’agissait pas d’inventer de grandes théories, il ne s’agissait pas de partir à la

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

14

The Oulipo and Modern Thought

the sewing circle, with Jean Lescure claiming that the term thus ‘flattered the modest taste that we shared for both beautiful work and good deeds’.²⁸ Nevertheless, I have yet to see the group’s name translated anywhere as ‘Sewing Circle for Potential Literature’, and I do not propose to adopt that version here. Most English texts discussing the Oulipo hitherto have gone for workshop: the Workshop for Potential Literature. But this is not ideal since workshop already has a well-known literary sense—a session in which writers share their work and receive feedback from the group—and this is most definitely not what the Oulipo is, a point reiterated brusquely and frequently by pretty much all members.²⁹ We might then try workroom— a good direct translation of ouvroir, and one which, like its source, seems slightly archaic in English. This is the term which Queneau uses in an article for the Times Literary Supplement in 1967 in which he introduces the group to English readers, and I think this will do nicely: its slight awkwardness—‘Don’t you mean workshop?’—is correct.³⁰ In a radio interview in 1962, Queneau runs through the group’s members as follows: Firstly there is François Le Lionnais, who is the founder; and myself, the cofounder. There is Albert-Marie Schmidt who, as you know, is a great sixteenth-centuryist. It is him we have to thank for the title of our working group. [ . . . ] To go on with the list, there is a mathematician, Claude Berge; Jacques Bens, who is secretary; and then two representatives from the Collège de ’Pataphysique, Noël Arnaud and Latis, since we are attached to the Collège de ’Pataphysique; then three other members who are poets and writers, Lescure, Duchateau and Queval. We have foreign correspondents, André Blavier, Paul Braffort, Stanley Chapman and Ross Chambers.³¹

conquête de quoi que ce soit.’] The quotation appears in the transcript of an interview which a number of Oulipians gave for the journal Page des libraires in early 1996. (FO, MS-6 (‘Page’)). ²⁸ Jean Lescure, ‘Brief History of the Oulipo’, in Motte, Oulipo, pp. 32–9 (p. 33). ²⁹ See for example Queneau’s list, ‘What is the Oulipo Not?’, quoted below. ³⁰ Raymond Queneau, ‘Science and Literature’, TLS 3422 (28 September 1967): 863–4 (864). ³¹ [‘Il y a d’abord François Le Lionnais, que en est le fondateur; moi-même, qui en suis le co-fondateur. Il y a Albert-Marie Schmidt qui, comme vous le savez, est un grand seiziémiste. C’est à lui que l’on doit la dénomination même de notre group de travail. [ . . . ] Pour continuer l’énumération, il y a un mathématicien, qui est Claude Berge; Jacques Bens, qui est le secrétaire; et puis deux représentants du Collège de ’Pataphysique, Noël Arnaud et Latis, parce que nous nous sommes rattachés au Collège de ’Pataphysique; puis trois autres membres qui sond des poètes et des écrivains, Lescure, Duchateau et Queval. Nous avons des correspondants étrangers, André Blavier, Paul Braffort, Stanley Chapman, Ross Chambers’]. The interviews, conducted by Georges Charbonnier, went out between 2 February and 27 April 1962. The transcripts are available as Georges Charbonnier, Entretiens avec Raymond Queneau (Paris: Gallimard, 1962), pp. 116–17.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

The Secret of Lightness

15

(A further foreign correspondent, Marcel Duchamp, by then a naturalized American, was added to the roster in March 1962, just as Queneau’s radio interviews were being broadcast. However, Duchamp would in fact meet the Oulipo only twice, while passing through Paris in 1967 and 1968.) What comes across clearly as Queneau lists the group’s members is that this is not a writer’s group in the ordinary sense. There are writers— Lescure, Duchateau, and Queval (not to mention Queneau himself )— but they are listed last, while Berge is introduced specifically as a mathematician, Schmidt as an academic historian, and Latis and Arnaud are identified by their role in the Collège de ’Pataphysique (in fact, Arnaud is a writer while Latis is a philosophy teacher). The Oulipo, from the outset, is a place of exchange between the worlds of literature and science, communities between whom a certain amount of negotiation is both necessary and fruitful. As Le Lionnais puts it at one of the early meetings, ‘The fact is that our group includes both mathematical and literary people for whom the same words don’t always apply to the same notions.’³² So what is the Oulipo if not a writers’ workshop? Neither workroom nor sewing-circle really tell us much about what the group do. Albert-Marie Schmidt’s early description of the group in the Protestant weekly Réforme offers us some insight when he calls them ‘a secret laboratory of literary structures’.³³ Secret here is a questionable label, especially if one applies it oneself in a newspaper article—and there will be more to say about this later— but laboratory is helpful. This is a place of experimentation, where literary structures are proposed, invented, tested. Queneau also has some useful words by way of the following definition (albeit a negative one) of the group: What is the Oulipo not? 1. It is not a movement or a literary school. We situate ourselves prior to aesthetic value, which is not to say that we ignore it. 2. Neither is it a scientific seminar, or ‘serious’ (in inverted commas) working group, even though we have a professor from a Faculty of Letters and another from a Faculty of Science among our number. Finally, 3. It is not about experimental or aleatory literature (for example of the kind practised by the group around Max Bense in Stuttgart).³⁴ ³² Meeting of 23 August, 1966. The minutes for this meeting are missing from the BnF archive, but appear, in Iain White’s translation, as ‘André Blavier Circular No. 75 (+/ )’, in Mathews and Brotchie, Oulipo Compendium (London: Atlas, 1998), pp. 188–91 (p. 191). ³³ Quoted in Arnaud ‘Prolegomena’, p. xiii. The original appeared Réforme on 2 February 1963. ³⁴ [‘Qu’est ce que n’est pas l’OU. LI. PO? 1. Ce n’est pas un mouvement ou une école littéraraire. Nous nous plaçons en deçà de la valeur esthétique, ce qui ne veut pas dire que nous en fassions fi. 2. Ce n’est pas non plus un séminaire scientifique, un groupe de travail “sérieux” entre guillemets. [ . . . ] Enfin, 3. Il ne s’agit pas de littérature expérimentale ou

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

16

The Oulipo and Modern Thought

Queneau also has a few positive descriptions of the group, most notably that the Oulipo’s work is ‘amusing (at least for us)’, and he goes on to defend the value of playfulness using examples from the history of mathematics: ‘I must insist on this adjective “amusing”. Certainly, certain of our works, in the domain of simple “jests” or “wordplay”, might seem analogous to certain “parlour games”. Let us remember that topology and number theory were born, in part, out of what were once called “mathematical diversions” [mathématiques amusantes].’³⁵ If the structures that come out of the Oulipian laboratory seem like mere diversions then, we are reminded not to judge too hastily their usefulness or import.³⁶ But perhaps the most useful insight into what the group actually do comes from their meeting minutes. By the late 1960s Oulipo meetings have settled into a standard format, grouped under regular headings in the minutes: Creation, Rumination, Erudition, Action.³⁷ Creation, as the group themselves put it, is the most important of the headings: ‘It is obligatory, on pain of the meeting’s cancellation, that there be at least one intervention under this heading. An Oulipian presents a new constraint, accompanied by at least one example. This presentation is followed by a general discussion.’³⁸ In Rumination, new constraints not yet welldeveloped enough to produce a textual example are introduced for wider consideration, while in Erudition members present on works which predate the group but exhibit Oulipian characteristics.³⁹

aléatoire.’] Raymond Queneau, ‘Littérature potentielle’, Bâtons, chiffres et lettres, 2nd edn (Paris: Gallimard, 1965), pp. 317–45 (pp. 321–2). The published text was first delivered as a talk to Jean Favard’s Seminar of Quantitative Linguistics in January 1964. ³⁵ [‘J’insisterai cependant sur le qualificatif d’“amusant”. Il est certain que certains de nos travaux peuvent paraître, du domaine de la simple plaisanterie ou encore de simples “jeux d’esprit”, analogues à certains “jeux de société”. Rappelons-nous que la topologie ou la théorie des nombres sont nées en partie de ce qu’on appelait autrefois les “mathématiques amusantes”.’] Queneau, ‘Littérature potentielle’, p. 322. ³⁶ Le Lionnais will feel the need to reiterate this sentiment in the group’s second manifesto: ‘Most writers and readers feel (or pretend to feel) that extremely constraining structures such as the acrostic, spoonerisms, the lipogram, the palindrome, or the holorhyme (to cite only these five) are mere examples of acrobatics and deserve nothing more than a wry grin, since they could never help to engender truly valid works of art. Never? Indeed. People are a little too quick to sneer at acrobatics.’ [François Le Lionnais, ‘Second Manifesto’, in Motte, Oulipo, pp. 29–31 (p. 30)]. ³⁷ These headings seem to develop out of an earlier, roughly synonymous set: Practique, Théorique, and Anecdotique. See ‘Ordre du jour autographe de François Le Lionnais de la réunion du 23 août 1966’ (FO, DM-2 (6)). ³⁸ [‘Il est obligatoire, sous peine de nullité de la réunion, qu’il y ait au moins une intervention sous cette rubrique. Un oulipien présente une contrainte nouvelle, accompagnée au moins d’un exemple. Cette présentation est suivie d’une discussion générale.’] Oulipo, ‘Des rubriques’, Oulipo (Paris: ADPF, 2005), pp. 18–19 (p. 19). ³⁹ Further headings cover finances and the arrangement of the next meeting. For a more detailed outline of the format of an Oulipo meeting, see ‘Des rubriques’.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

The Secret of Lightness

17

Taking the meeting of January 1971 as an example then, Georges Perec’s minutes, scribbled on the back of a sales list for a bookseller, are brief but informative. This is the 119th meeting, taking place at Le Lionnais’s house in Boulogne-sur-Seine on 12 January. Those present are Perec, Le Lionnais, Arnaud, Bénabou, Luc Étienne, and Paul Braffort. The meeting’s Creation section consists of three items: Le Lionnais’s ‘Absence de rimes intérieures’ [‘Absence of internal rhyming’], Étienne’s ‘Limerick’, and another by Étienne which Perec seems to have missed (there is a gap in the minutes, where a note added in Le Lionnais’s hand wonders ‘Quoi?’: ‘What?’). The meeting proceeds to Action, here concerned largely with the group’s long-running attempt to bring out a collection of their experiments with the publisher Jonathan Cape, while among the topics under Erudition is a discussion of homophonic translation (‘traduction phonologique’) led by Perec. For the next meeting, scheduled for 12 February, Harry Mathews is mooted as a guest. So this is the form adopted when the Oulipo meet amongst themselves. But if the group’s purpose is to invent and appraise literary structures that might be of use to a wider literary world, how do these creations filter out? These days, we may be familiar with the Oulipo via their website or their monthly Thursday night performances—Les Jeudis de l’Oulipo—at the BnF in Paris. In their early days, however, the group were considerably more retiring. Nevertheless, some of their experiments were published, albeit discreetly, during their first years. An important collection of the Oulipo’s work, including Le Lionnais’s essay ‘La Lipo’—the text that would become known as the group’s first manifesto—was published privately by the Collège de ’Pataphysique as a special issue of the Collège’s Dossiers series—Dossier 17—in December 1961.⁴⁰ Then, just over a year later, an Oulipo special issue appeared in André Blavier and Jane Graverol’s small-format review Temps mêlés. The introduction here is slightly disingenuous: ‘For some time [the Oulipo] has been a secret society, but their works have become so important that they have decided to divulge them’.⁴¹ While this label of a secret or semi-secret society is attached to the Oulipo almost as a matter of course in its early days—Camille Bloomfield calls it ‘The Myth of Secrecy’—in truth the group had been divulging themselves in small doses all along, in newspapers, on the radio, and in appearances at seminars and conferences in fields that took their interest.⁴² The Temps mêlés issue, for example, is made up for the most ⁴⁰ For the context in which ‘Le Lipo’ is re-envisaged as a manifesto, see Camille Bloomfield, ‘Les manifestes à l’Oulipo: La disparition d’une forme’, Études littéraires 44.3 (2013): 35–46. ⁴¹ [‘Ils ont longtemps constitué une société secrète mais leurs travaux sont désormais si importants qu’ils ont décidé de les divulger.’] Guy Le Clec’h, ‘Introduction’: i. ⁴² See Bloomfield, Raconter l’Oulipo, pp. 214–22.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

18

The Oulipo and Modern Thought

Fig. 4. Georges Perec’s handwritten minutes to the group’s meeting of January 1971 (FO, DM-3 (1)). Image reproduced by kind permission of the Oulipo.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

The Secret of Lightness

19

part of contributions to a conference on cybernetics, which five of the group—half its membership—had attended and where they had spoken openly about the Oulipo’s work on the relationship between procedure and creativity.⁴³ In 1964, the Oulipo also travelled en masse to Blavier’s hometown of Verviers in Belgium to present a staged version of one of their meetings which was broadcast on Belgian radio. Not exactly a secret society then, but one at least unconcerned with finding a popular audience, one whose collective publications would reach a certain cognoscenti—of Pataphysicians, or the avant-gardistas who read Temps mêlés—but few others.⁴⁴ It is a curious detail that, for a long time, the group’s first major collective publication seemed destined to be an English one. In the autumn of 1967, the Times Literary Supplement ran a feature called ‘Crosscurrents’ in which four continental intellectuals were invited to consider the relationship between literature and another discipline. Umberto Eco wrote on literature and sociology, Italo Calvino on literature and philosophy. Queneau and Roland Barthes both chose to write on literature and science, with Barthes giving a primer of literary structuralism while Queneau introduces the Oulipo to British readers. The Queneau article begins by bemoaning the ‘Two Cultures’ schism that separated the arts and sciences.⁴⁵ ‘It is still the fashion for a poet to boast that he does not know the first thing about mathematics,’ he complains, illustrating his case with Victor Hugo’s robustly anti-scientific poem ‘Le Calcul’. Mathematics, for Hugo, is an abyss whose depths he goes on to plumb in verse:

⁴³ The conference, entitled ‘Pensée artificielle, pensée vécue’ [‘Artificial Thought, Lived Thought’] was at Cerisy, 9–19 July 1963. ⁴⁴ Many years after this half-observed secrecy has been abandoned, Hervé Le Tellier will propose an interesting and persuasive reason for its having been adopted in the first place. In the transcripts for an interview conducted for a feature in the magazine Page des libraires in 1996, Le Tellier will muse that Queneau, as the group’s most celebrated member, was concerned that the Oulipo might appear to the outside world like a personality cult, or a coterie of sycophants: ‘I think the reason the group was more or less secret was that it was what Raymond Queneau wanted. I think that if the group had not been secret, it would have seemed like a bunch of young people gathered around Raymond Queneau. And that was not at all what he wanted. He said, I don’t want to be the new Breton’. [‘Je pense d’ailleurs que c’est la raison pour laquelle le groupe était plus ou moins secret, c’était une volonté de Raymond Queneau. Je pense que si le groupe n’avait pas été secrete, il serait apparu comme étant plusiers jeunes gens autour de Raymond Queneau. Et ce n’était pas du tout sa volonté. Il disait je ne veux pas être le nouveau Breton.’] (FO, MS-6 (‘Page’)). ⁴⁵ Raymond Queneau, ‘Science and Literature’, TLS 3422 (28 September 1967): 863–4.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

20

The Oulipo and Modern Thought

Au fond, presque indistincts, l’absolu, l’innombrable, L’inconnu, rocs hideux que rongent des varechs D’A plus B ténébreux mêlés d’X et d’Y grecs; Sommes, solutions, calculs où l’on voit pendre L’addition qui rampe, informe scolopendre! Signes terrifiants vaguement aperçus! Triangles sans Brahma! croix où manque Jésus! Réduction du monde et de l’être en atomes! (ll. 74-81)⁴⁶

At the bottom, almost indistinct, the absolute, the innumerable, The unknown, hideous rocks gnawed by weeds Of gloomy A plus B mixed with Xs and Ys; Sums, solutions, calculations where we see addition, That shapeless scolopendra, dangling, creeping! Terrifying signs, dimly perceived! Triangles without Brahma! Cross with no Christ! Being and the world reduced to atoms!

For Queneau, of course, this type of partisanship—of arts versus sciences, and the former’s proud ignorance of the latter—represents a dismal state of affairs. All is not lost, however, and towards the end of the article Queneau describes the reconciliation underway in Paris: a study group of which I was a founder-member [ . . . ] along with a ‘scientist’, François Le Lionnais. I am speaking of the Ouvroir de Littérature Potentielle (Workroom for Potential Literature), the so-called Oulipo, which, since 1960, has been working towards the discovery of new or revived literary forms, this research being inspired by an interest in mathematics.⁴⁷

With these words, the Oulipo is introduced for the first time into the mainstream anglophone literary world, and three months later, on the last day of 1967, Nathaniel Tarn, an editor for the publisher Jonathan Cape would write to Le Lionnais proposing a book of Oulipian material for the new series, Cape Editions. Tarn describes Cape Editions as promoting ‘experimental work in linguistics and communication sciences’, and early texts in the series include translations of Lévi-Strauss’s The Scope of Anthropology (1967) and Barthes’s Writing Degree Zero (1967).⁴⁸ However, it would also include literary material such as Baudelaire’s Twenty Prose Poems (1968) and Jarry’s The Supermale (1968). Sadly, despite a unanimously positive response to Tarn’s approach, for the Oulipo, the Jonathan Cape publication was not to be. An agenda item almost every month throughout the late 60s, the ‘dossier Jonathan Cape’ stalled as the ⁴⁶ Victor Hugo, Toute la lyre (Paris: Hetzel & Quantin, 1888). ⁴⁷ Queneau, ‘Science and Literature’: 864. ⁴⁸ Oulipo, ‘Lettre dactylographiée et signée de Nathaniel Tarn à François Le Lionnais du 31 décembre 1967 de Cape Edition’ (FO, DM-2 (20)).

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

The Secret of Lightness

21

group debated over the collation of materials to be included. Then came a letter from Tarn announcing that he had terminated his association with the Cape, but stating that the publication might still go ahead without him. Further communications from other figures at Cape followed assuring the Oulipo of their continued interest, but by the early 70s the trail had gone cold. Meeting minutes began to refer to the ‘dossier ex-Cape’, which would eventually be rehabilitated as a project with Gallimard, becoming the landmark collection La Littérature potentielle (La Lipo for short) in 1973. A second collection, Atlas de Littérture potentielle followed in 1981. In the mean time, beginning with Perec’s Ulcérations in 1974, the group started to publish pamphlets of their work in small runs of 150 copies. This series, known as the Bibliothèque oulipienne, now runs to over two hundred titles, and volumes which collect the most recent numbers are now available on general sale. Alongside these private or collective publications, of course, there are the notable Oulipian works which some of the group’s distinguished writers have brought out: Jacques Roubaud’s ∈, Perec’s La Disparition (the novel without an e), Harry Mathews’s Cigarettes, and Italo Calvino’s If on a Winter’s Night a Traveller, to name but a few. In spite of all the protests that the Oulipo is not a writers’ group, and that its aim is only to produce and test structures which may—or may not—be adopted beyond its sphere, it is these fully-formed works, more than anything else, that have established the group’s reputation with an international reading public. The previous sketch gives some outline of the Oulipo’s foundation and the group’s journey from willed obscurity to a degree of public conspicuousness. As such it merely retells a story that has been told often before. What is lacking—the gap that this book sets out to fill—is an intellectual account of the group’s position during this period in the broader debates that were going on around them. Despite some striking similarities of approach, for example, the Oulipo, as we have seen, remained officially unaffiliated to the structuralism that had become such a dominant mode in French academia in the 1960s. At the same time, we can see from the early minutes how much of a fascination cybernetics held for the group.⁴⁹ A note, for example, marked Top Secret and circulated with the minutes of the group’s second meeting states that certain members have been ⁴⁹ In fact, not quite everybody shared this fascination. As Bloomfield and Campaignolle put it: ‘With the exception of Latis – an Oulipian who had come from the Collège de ’Pataphysique and who did not hide his dread faced with technology – the group shared a love of technology’ [‘À l’exception de Latis – Oulipien venu de la ’Pataphysique qui ne cache pas son effroi face à la technologie – le group affiche une technophilie partagée’.] Camille Bloomfield and Hélène Campaignolle, ‘L’Oulipo et l’informatique’, in Reggiani and Schaffner, Oulipo mode d’emploi, pp. 319–36 (p. 323).

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

22

The Oulipo and Modern Thought

tasked with gaining access to one of the computers at IBM or Bull: ‘[t]heir goal is to try to utilise electronic machines for different works of literary analysis, as part of the activities of the OLiPo [sic]’.⁵⁰ (The mission is not immediately successful: three years later Queneau will complain to the Seminar of Quantitative Linguistics that ‘not having machines at our disposal [is] a continual lamento at our meetings’.)⁵¹ Noting the importance of Claude Lévi-Strauss’s early 1950s cybernetics seminar—attended by Roman Jakobson and Jacques Lacan—in shaping the structuralism of the next decade, Chapter One examines the balancing act which the Oulipo performed in following the same path while refusing to be swallowed up by structuralism’s considerable cultural heft. The second chapter looks at the implications of a common Oulipian practice. If we think back to those minutes from 1971 with Perec’s doodles alongside them, we might remember the item ‘traduction phonologique’ introduced as a point of Erudition by Perec. In fact, translating for sound rather than sense—better known as ‘homophonic translation’— has a venerable history in the Oulipo going back to the group’s first publication, the Pataphysical Dossier 17. The group are certainly fond of this technique, but they are not the only ones. The most notable Anglophone example of homophonic translation—Celia and Louis Zukofsky’s Catullus—also dates from around this time.⁵² Moreover, this type of interlingual punning has had its place in the toolkit of psychoanalytical interpretation since Freud’s early paper ‘The Psychical Mechanism of Forgetfulness’ (1898). An awareness of this that is part fascination and part cynicism finds its way into homophonic exercises by Queneau, Arnaud, Mathews, and Perec. It is a stance—deeply informed, but also resistant or at least conspicuously ambivalent—reminiscent of the kind of doublethink concerning psychoanalysis that Elisabeth Roudinesco identifies among French writers of the interwar period. Drawing up a list that includes Queneau along with Georges Bataille, Antonin Artaud, and Michel Leiris, she claims that these figures were: influenced by the theoretical adventure of Freudianism and also had experience of the analyst’s couch, but whose interest in the theory was independent of their views on the treatment. To sympathize with the Freudian revolution

⁵⁰ [‘Leur but est de tenter d’utiliser des machines électroniques pour différents travaux d’analyse littéraire, dans le cadre des activités de l’OLiPo.’] The note is missing from the BnF archive, but appears in Jacques Bens anthology of early meeting minutes, Genèse de l’Oulipo (1960–63) (Bordeaux: Castor Astral, 2005), p. 38. ⁵¹ [‘Nous regrettons de ne pouvoir disposer de machines: lamento conitnuel au cours de nos réunions’]. Queneau, ‘Littérature potentielle’, p. 322. ⁵² Celia and Louis Zukofsky, Catullus (London: Cape Goliard, 1969).

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

The Secret of Lightness

23

was for them an intellectual act, whereas to go to an analyst merely meant to have one’s malady dealt with as directly as possible.⁵³

Chapter Two looks at the place of homophonic translation within the Oulipo and finds in its use a kind of puckish critique: a way of kicking out at one’s own analyst, and a way of satirizing particular strands in psychoanalytic thought. One of the most striking aspects of the Oulipo, when compared with other groupings of the French avant-garde, is its longevity. Although the last of its original members, Jacques Duchateau, died in early 2017, the Oulipo is still going strong: still meeting, still producing work in the Bibliothèque oulipienne series, still recruiting new members. This policy of cultivated regeneration, however, has not always been in place. Aside from the investiture of a few ‘foreign correspondents’, for its first few years, the group’s membership was fixed at ten. In 1966, however, the Oulipo underwent something of a crisis. Minute-taking had deteriorated, and a frustration at the narrowness of the group’s interests had set in among certain members. At the August meeting, Le Lionnais bluntly summarized the direness of the situation: let’s get down to business, in other words, the Oulipo, which might as well be called the Ou-peu-po, the Workshop of Few Potentialities. I put on the agenda (the same old agenda) a report on revitalising [ . . . ] the Oulipo. Two solutions are available: either bringing the Oulipo up to strength whenever a vacancy occurs, or letting it gradually die a natural death, in honourable annihilation.⁵⁴

A decision was taken that day to expand the Oulipo, to bring in new blood in order to revivify the group. Beginning with Jacques Roubaud in 1966, and over the next few years taking in Georges Perec (1967), Marcel Bénabou (1969), Harry Mathews (1972), and Italo Calvino (1973), the Oulipo became an extensible entity. Nevertheless, something that comes across rather strongly whenever the group describe their history is a difference in character that is felt to obtain between the first generation of members and the ‘second wave’. In 1972, Paul Fournel—himself a second-waver, elected to full membership only that year—published a book entitled Clefs pour la littérature potentielle (at that date the lengthiest public treatment of the group, since La Lipo—the notorious ‘dossier ex-Cape’—was still slowly working its way into print

⁵³ Elisabeth Roudinesco, Jacques Lacan & Co.: A History of Psychoanalysis in France, 1925–85, trans. by Jeffrey Mehlman (Chicago: University of Chicago Press, 1990), p. 121. ⁵⁴ ‘Circular No. 75 (+/ )’, (i.e. meeting minutes for August 1966), p. 188.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

24

The Oulipo and Modern Thought

with Gallimard). Fournel has these telling words to say about the two cohorts of the Oulipo: According to Le Lionnais, and perhaps not all Oulipians would agree on this, the arrival of the ‘young ones’ has brought about a change of attitude within the Oulipo. Of course, it is not a matter of going looking for interviews, but rather of not being opposed to the overtures of the outside world. The ‘young ones’ need a bit of publicity . . . And then these ‘young ones’ have arrived with their questions and their ideas, and some of them have not been slow in bringing up the problem of relations with the University. Should there be a seminar of potential literature within a literature department?⁵⁵

Fournel’s language is circumspect, and, as we have seen, the group’s stated secrecy was not exactly scrupulously observed in its first few years. Nevertheless, the view that a once secretive society was becoming less so at the start of the 1970s is an uncontentious one.⁵⁶ A catalogue essay for the Oulipo exhibition at the BnF in Paris in 2014, labels the more retiring founder members—Schmidt, Queval, Berge, Bens—as ‘Les discrets de l’Oulipo’: ‘The Quiet Ones of the Oulipo’.⁵⁷ A taste for publicity, however, was not the only change in attitude that the ‘young ones’ brought. Chapter Three will look at a more conceptual shift that the group underwent during the same period, a shift which maps conveniently onto the two generations of the group in the 1960s and 70s, and which goes straight to the heart of what Oulipian exercises might achieve: whether the goal is to evacuate the subject from writing—to ‘let language itself speak’—or whether, on the contrary, these modes of writing are a means of invoking repressed material, of accessing the ‘unadmitted self ’ in order to say the things the writer didn’t realize that they wanted to say. This movement, I argue, reflects a change in the group’s relation to Surrealism. Although it seems that new members are as wont as their seniors to express the Oulipo’s antipathy to the Surrealists,

⁵⁵ [‘D’après Le Lionnais, et tous les oulipiens ne sont peut-être pas d’accord sur ce point, l’arrivée des “jeunes” a entraîné une changement d’attitude à l’Oulipo. Il ne s’agit bien sûr pas de rechercher les interview, mais bien plutôt de ne plus s’opposer aux ouvertures vers l’extérieur. Les “jeunes” ont besoin d’un peu de publicité . . . Et puis ces “jeunes” sont arrivés avec leurs questions et leurs ideées, et certains n’ont pas tardé à poser le problème des rapports avec l’Université. Faut-il créer un séminaire de littérature potentielle au sein d’une U.E.R. de Littérature?’] Fournel, Clefs, pp. 14–15. ⁵⁶ For more on the public-facing Oulipo, see Coraline Soulier, ‘L’Oulipo en public’, in Christelle Reggiani and Alain Schaffner, Oulipo mode d’emploi (Paris: Honoré Champion, 2016), pp. 149–67. ⁵⁷ Bertrand Tassou, ‘Les discrets de l’Oulipo’, in Bloomfield and Lesage, Oulipo, pp. 59–65.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

The Secret of Lightness

25

there is a philosophical edge to this disavowal that can be far more keenly felt in the earlier writing than the later. Other important binaries arise in this period, between procedural and constrained writing, for example, or between the so-called ‘syntactic’ and ‘semantic Oulipo’, that is, between operations at the level of the letter (anagrams, lipograms, etc.) versus those that work at a more conceptual level (constraining character, location, plot). As Le Lionnais puts it during the meeting of August 1966, ‘Little by little we are moving from the syntactic to the semantic’.⁵⁸ Nevertheless, it would be wrong to imagine that these can be mapped clearly onto the different factions in the group’s membership. We must be aware too of changes extrinsic to the group—of philosophical shifts which were being felt more broadly in French intellectual life, and which, because the Oulipo has never been an isolated community, are echoed in their output. Even an Oulipian can change their mind, and Chapter Four focuses on Italo Calvino to show how the shift which the previous chapter outlined in terms of two cohorts of the group can also be discerned in the work of a single member. Examining Calvino’s writing from around the period that he become first of all a guest and then a full member of the group, we find a modulation from a rigid version of combinatorics to one in which a minute bending of the rules— the application of the Lucretian clinamen—is the very act that elevates a work to the status of Art. During early discussions about the ill-fated English-language edition of their experiments, François Le Lionnais wrote a letter to the project’s editor at Jonathan Cape. Le Lionnais explained that Oulipian work can be divided into three categories: works that are easy to translate; works whose translation is difficult, or even very difficult; and works which are untranslatable. Unfortunately, Le Lionnais reasoned, this last category would include some of the Oulipo’s most characteristic work, and to omit these would risk giving a false impression of the group.⁵⁹ Perhaps the edition could include a description of these untranslatable texts, along with comparable attempts in English. Aware then of the special problems their work throws up for translation, it is little wonder that untranslatibility should be almost as much a theme as a characteristic in Oulipian writing. Chapter Five looks at a sequence of short stories by Perec and Mathews which are based on the same narrative set-up: a field linguist who stumbles on a language of paradigmatic oddity. The chapter demonstrates how these

⁵⁸ ‘Circular No. 75 (+/ )’, (i.e. meeting minutes for August 1966), p. 188. ⁵⁹ ‘Lettre dactylographiée de François Le Lionnais à Nathaniel Tarn du 22 avril 1968’ (FO, DM-2 (25)).

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

26

The Oulipo and Modern Thought

stories use a template from analytic philosophy, that of W. V. O. Quine’s ‘gavagai’ language, to reflect on the act, and the limits, of translation. In the following chapters, then, the Oulipo are depicted as being not just deeply knowledgeable about the philosophical debates of their time but also engaged in their cut and thrust. The argument, however, must not be taken too far. Sometimes a cigar is just a cigar, and much of the material in the Oulipo’s archives serves no grander purpose than to satisfy an experimental curiosity: is this technique an effective way of producing writing? It is not the intention of this book to introduce a thoroughgoing suspicion, a way of seeing Oulipian writing always as the expression of a theory; rather, the aim is merely to make the group seem more rooted, less a band apart than a collective of wry virtuosi, amusing themselves sometimes with their own specific concerns and sometimes in mischievous observation of their peers, heckling from the balcony of the academy.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

1 Literature Machines StructurElism versus StructurAlism At the beginning of October 1960, Queneau wrote to Le Lionnais with some good news about Cent mille milliards de poèmes: ‘I’ve found this sentence from Turing for an epigraph: “Only a machine could appreciate a sonnet written by another machine”. What do you think?’¹ The quotation as Queneau has it—and as it would eventually be used in the book— appears to have been slightly corrupted in translation. Turing’s original remark, in an interview for The Times in 1949, was that ‘a sonnet written by a machine will be better appreciated by another machine’.² The comment was a tongue-in-cheek riposte to Turing’s colleague, the neuroscientist Geoffrey Jefferson, who a couple of days earlier had publicly dismissed the creative potential of artificial intelligence.³ It was also something of a plug for the new Manchester University Computer (henceforth M. U. C.) which, under Turing’s administration, would be used for experiments in natural language processing. One example of the Manchester computer’s literary potential can be found in this piece of output—not a sonnet, but a love letter—produced three years after the Times article: DARLING MOPPET YOU ARE MY SEDUCTIVE CHARM. MY ANXIOUS LONGING BREATHLESSLY SIGHS FOR YOUR WISTFUL FELLOW FEELING. YOU ARE MY LOVING AMBITION. MY ¹ [‘J’ai trouvé pour épigraphe cette phrase de Turing: “Seule une machine peut apprécier un sonnet écrit par une autre machine.” Qu’en pensez-vous?’]. Raymond Queneau, letter to Le Lionnais, 7 Oct 1960. (FO, DM-1 (1)). ² I. Trethowan, ‘The Mechanical Brain’, The Times, 11 June 1949: 4. ³ Jefferson’s comment was made in his acceptance speech for the Lister Medal, delivered at the Royal College of Surgeons on 9 June 1949. He states, ‘Not until a machine can write a sonnet or compose a concerto because of thoughts or emotions felt, and not by the chance fall of symbols, could we agree that machine equals brain – that is, not only write it but know that it had written it’. The paper was subsequently published as Geoffrey Jefferson, ‘The Mind of Mechanical Man’, British Medical Journal, 4616, 25 June 1949: 1105–10.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

28

The Oulipo and Modern Thought

AVID AFFECTION PINES FOR YOUR KEEN DESIRE. YOU ARE MY CRAVING TENDERNESS. YOURS CURIOUSLY, M. U. C.⁴ This breathless missive is the result of an algorithm devised by the computer scientist Christopher Strachey—nephew of the Bloomsbury Group writer, Lytton Strachey. Strachey’s ‘Love Letters’ programme works by combining categories of words—salutations, nouns, verb phrases, and so on—according to a predefined syntax. It is given a sentence structure (repeated several times) and a set of options for each word type—‘[salutation1] [salutation2], YOU ARE MY [adjective] [noun]. MY [adjective] [noun] [adverb] [verbs] YOUR [adjective] [noun]. YOURS [adverb]’—and then goes on to produce its outpourings of affection by filling in the blanks with randomized words from the appropriate categories.⁵ The algorithm is at once an early experiment in natural language processing and an ironic spoof of the love letter’s conventionality as a form, the joke being that this is a genre of writing whose underlying structure might be comprehensively analysed. The Strachey archive at the Bodleian Library in Oxford includes the wordlists for each category, as well as several examples of the programme’s output: completed ‘letters’, both handwritten and machine-printed. However, what the archive papers also show is that, having successfully produced these first, simplistic billets doux from the M. U. C., Strachey became more ambitious for his programme. Moving beyond simple lists of adjectives, nouns, and verbs, he introduced a table of ‘THEMES’ which the letter might revolve around (from ‘stop seeing that man’ and ‘marry me’ to ‘playing bridge’ and ‘going out dancing’) and another of ‘MOODS’ (‘reproachful’, ‘jubilant’, etc.) which would be reflected in the language. To facilitate these, the programme needed to be given a far greater set of ⁴ Strachey’s notes, calculations and routines for the ‘Love Letters’ algorithm are archived among the Strachey papers at the Bodleian Library, Oxford, MS. Eng. misc. b. 259/C.34, C.35, and C.36. The example quoted here is transcribed in Strachey’s hand in folder C.34 f.1. ⁵ This is rather like the Oulipian procedure known as ‘homosyntactical translation’ where a text is analysed into its parts of speech and the task is to write something else that maintains the same syntax. The method is first outlined at the February 1962 meeting: ‘Le Lionnais proposes the composition of isosyntactic poems. One chooses a poem in advance and performs a logical and grammatical analysis of each line, then composes another poem with the same syntactic structure as the first’ [‘Le Lionnais propose la composition de poèmes isosyntaxiques. Il s’agit de faire l’analyse logique et grammaticale de chaque vers d’un poème choisi à l’avance, puis de composer un autre poème ayant la même structure syntaxique que le premier’]. [‘Compte rendu dactylographié de Jacques Bens de la réunion n° 18 du 16 février 1962’ (FO, DM-1 (1))]. Latis provides a model in a letter to Le Lionnais of 15 September 1966 (FO, DM-2 (7)), which then becomes the source text for two further homosyntactic translations, by Bens and Le Lionnais, the following year (FO, DM-2 (19)). See also ‘Homosyntaxism (homosyntactictical translation)’, in Mathews and Brotchie, Oulipo Compendium, pp. 155–6.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

Literature Machines: StructurElism versus StructurAlism

29

Fig. 5. Early printed output from the first version of the ‘Love Letters’ algorithm, showing Strachey’s handwritten correction. The papers and correspondence of Christopher Strachey, 1930–83, Bodleian Libraries, University of Oxford, MS. Eng. misc. b. 259/C.34. Image reproduced by kind permission of the Bodleian Library, Oxford.

template sentences, such as ‘If you don’t [a, b, c, . . . ] I will [x, y, z, . . . ]’, where the options available in each gap would be determined by the mood and theme of the letter. Working his routines out in pencil, Strachey’s notes include page after page of potential output, from ‘I can’t [imagine] why you are always [playing bridge]’ to ‘When shall I [see] your [attractive] [curly hair] again?’. The files in the archive, sadly, do not contain any printed output to testify that this latter form of the algorithm was ever successfully run on the M. U. C. Nevertheless, what the ‘Love Letters’ programme demonstrates in wonderfully whimsical fashion is that, by the early 1950s, the possibilities of an algorithmic approach to natural language syntax were starting to be felt within the emerging discipline of cybernetics. What the Strachey papers show us too is how, to a curious mind, a syntax-based experiment—a set of simple word substitutions—suggests a further investigation, both looser and more complex. Can the same principles be extended to non-syntactic aspects of communication like mood and topic? Can one take a simple literary genre and, by tabular analysis, describe its semantic register to the point of successful imitation? This chapter will be concerned with structuralism as a mode of literary criticism—a mode that, not unlike the ‘Love Letters’ algorithm, attempted to apply the methods of linguistics beyond the purely syntactic aspects of a text. Here is Roland Barthes in 1966 introducing the structuralist approach to narrative: As everyone knows, linguistics stops at the sentence; it is the last unit that falls within its scope. [ . . . ] And yet it is obvious that discourse itself (as an arrangement of sentences) is organized, and that, through this organization, it is perceived as the message of another ‘language’, functioning at a higher level than the language of linguistics: discourse has its units, its rules, its ‘grammar’.⁶

⁶ Roland Barthes, ‘An Introduction to the Structural Analysis of Narrative’, trans. by Lionel Duisit, New Literary History 6.2 (1975): 237–72 (239–40). Originally published as ‘Introduction à l’analyse structural des récits’, Communications 8 (1966): 1–27 (3).

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

30

The Oulipo and Modern Thought

Elsewhere, Barthes will speak of structuralism as ‘a linguistics of discourse’, one that focuses its analysis on ‘the “language” of literary forms’. A critical mode, then, based on discovering the organization—the ‘rules’—of literary forms. At the same time, we find the Oulipo of the mid-1960s becoming frustrated with the syntactic operations of their first years and curious to investigate—à la Strachey—whether these methods might be extended to looser, broader units of the literary text, such as plot elements. And yet, rather than a tale of glorious collaboration, of structuralists and Oulipians recognizing the extraordinary overlap in their interests, of a happy coming together of rigour and whimsy, the story this chapter will tell is one of mistrust, the narcissism of small differences, and perhaps one of missed opportunities. While later chapters will seek to draw out the ways in which Oulipian writing is more theoretically informed than it is given credit for, we will be starting off by doing the exact opposite, trying to understand why structuralism doesn’t feature more either in the group’s work or in their public statements. The two strands run parallel, but rarely cross. Both, however, can trace their roots, at least in part, to the cybernetic moment of the early 1950s. At exactly the point that Strachey was programming the Manchester University Computer to compose love letters, across the Channel, the French cyberneticist Albert Ducrocq was building a robot which would also be capable of producing linguistic output. Christened Calliope after the Greek muse of epic poetry, Ducrocq’s machine was, as he put it, a poète électronique.⁷ Lest anyone accuse him of overclaiming, Ducrocq clarified that only about one in four of Calliope’s prose poems was actually intelligible. Here is an example of a good one: The victory will be difficult. A thin layer of extensive snow covered the field of white wheat made strong: the sacred mine lies beneath. The coast mounts an assault on truth; its varied designs form a chain that shines in spite of its poison, while the arid wandering of the months and the years has given the boxwood the vigour of an old tree. A carnation beaten by time is the only ornament of this dark decor that survives in the rustic dwelling where we see a dilapidated chair, covered in soot. The valley is entirely lit by the orange glow of the year.⁸ ⁷ Albert Ducrocq, L’Ère des robots (Paris: Julliard, 1953), p. 262. ⁸ [‘La victoire sera ardue. Une mince couche de neige très étendue a couvert le champ de blé blanc rendu vigoureux: la mine sacrée gît en dessous. La côte monte à l’assaut de la vérité; ses dessins variés forment une chaîne qui brille malgré son poison, tandis que l’aride divagation des mois et des ans a donné au buis la vigueur du vieil arbre. Un œillet battu par le temps est le seul ornement de ce décor sombre qui survit dans la modeste demeure rustique où l’on voit une chaise délabrée, couverte de suie. La vallée est tout entière éclairée par la lueur orange de l’année.’] Ducrocq, p. 267.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

Literature Machines: StructurElism versus StructurAlism

31

The method underpinning this is very much along the same lines as Strachey’s ‘Love Letters’ algorithm: Ducrocq is merely applying this method of analysis-by-parody to a different literary form. Nevertheless, the difference in genre had a significant effect: Ducrocq’s robot-poet caught the popular imagination in a way that Strachey’s lovelorn computer could not. When Ducrocq included Calliope’s poetry in his 1953 work L’Ère des robots, the literary world had something to say. In the weekly journal Arts, the novelist Boris Vian—an eminent Pataphyscian, and close friend of both Queneau and Arnaud—argued that a machine like Calliope would always have the upper hand in producing certain registers of poetry: in the realm of the vague, the bizarre, the ephemeral, the abstruse and the dream-like, the robot will beat us every time. Effectively, it will have none of the bad reasons for choosing this or that term which our past imposes on us. It will be truly free, whereas if our pen ever produces a really original structure, it’s probably just because we’ve been reading Mallarmé or Jarry a little too intimately. In no time at all it will exhaust the combinations and deliver us up texts without syntax.⁹

Unencumbered by tradition, Calliope writes like Mallarmé or Jarry—twin figureheads of the avant-garde—without even trying, free of literary influence and its anxieties.¹⁰ As Vian imagines it, the robot-poet is also free from the burden of having a past of its own, an authorial subject, a set of experiences that it might want to express. Its method is simply combinatorics, and its materials are whatever vocabulary it has been fed to ⁹ [‘sur le terrain du vague, de l’insolite, du vaporeux, de l’abscons et du rêveur, le robot nous battra à tout coup. Lui, en effet, n’aura aucune des mauvaises raisons que nous impose notre passé de choisir tel ou tel vocable. Lui sera vraiment libre, alors que s’il vient sous notre plume automatique une structure vachement originale, c’est peut-être bien que nous aurons fréquenté Mallarmé ou Jarry, de façon trop intime. Lui épuisera les combinaisons en deux temps et trois mouvements et nous delivera des textes sans syntaxe.’] Boris Vian, ‘Un robot poète ne nous fait pas peur’, Arts, 10 April 1953: 1–6 (6). ¹⁰ At exactly this moment in the anglophone world, Roald Dahl is also fascinated by the possibilities of a literature machine. The ‘Great Automatic Grammatizator’ which he imagines, however, can write in any style: First, by depressing one of a series of master buttons, the writer made his primary decision: historical, satirical, philosophical, political, romantic, erotic, humorous, or straight. Then, from the second row (the basic buttons), he chose his theme: army life, pioneer days, civil war, world war, racial problem, wild west, country life, childhood memories, seafaring, the sea bottom and many, many more. The third row of buttons gave a choice of literary style: classical, whimsical, racy, Hemingway, Faulkner, Joyce, feminine, etc. (p. 270). This machine, in Dahl’s vision, will lead to the death of the author in a very straightforward way: writing will cease to be a viable profession. Roald Dahl, ‘The Great Automatic Grammatizator’, in Someone Like You (New York: Knopf, 1953).

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

32

The Oulipo and Modern Thought

begin with: ‘It is a relatively simple thing to make a robot like this, on condition that you furnish it with ready-made words’.¹¹ In Vian’s analysis we hear an echo from the future of Barthes’s scornful question: Did he wish to express himself, he ought at least to know that the inner ‘thing’ he thinks to ‘translate’ is itself only a ready-formed dictionary, its words only explainable through other words, and so on indefinitely.¹²

Calliope’s electronic poems raise the possibility of the text voided of intention, free of authorial subjectivity, an idea which will find its theoretical apotheosis fifteen years later in Barthes’s famous ‘Death of the Author’ essay. The similarities between a robot-poet rearranging the readymade words fed into it and the dethroned Author shuffling the terms of their readyformed dictionary may appear tenuous or coincidental. Nevertheless, as a growing body of scholarship has begun to demonstrate, a fascination— and serious, grant-funded engagement—with cybernetics was instrumental in shaping the thought of structuralism’s prime movers.¹³ In the early 1950s, Le Lionnais, in his capacity as Director of Science Education at UNESCO, set up a working party that, as he puts it, ‘worked on the relations between mathematics and the human sciences’.¹⁴ Recalling, at a distance of twenty-five years, the people involved, Le Lionnais reels off a few names: ‘Guillebeau [sic], Claude Berge, Rémi Chauvin, Benveniste, Lacan, etc.’ This is without doubt the same group that Elisabeth Roudinesco identifies in her history of French psychoanalysis: ‘In 1951, Lacan, Lévi-Strauss, [Georges-Théodule] Guilbaud, and [Émile] Benveniste met to work on structures to establish bridges between the human sciences and mathematics’.¹⁵ Thus, we find a pair of future Oulipians, Le Lionnais and Berge, in the same working party as the soon-to-be leading lights of ¹¹ [‘C’est [ . . . ] relativement facilement de faire faire ça à un robot, à condition de lui fournir les mots tout faits’]. Vian, ‘Robot poète’: 6. ¹² Roland Barthes, ‘The Death of the Author’, in The Rustle of Language, trans. by Richard Howard (Oxford: Blackwell, 1986), pp. 49–55 (p. 50). ¹³ See for example, Bernard Dionysius Geoghegan, ‘From Information Theory to French Theory: Jakobson, Lévi-Strauss, and the Cybernetic Apparatus’, Critical Inquiry 38.1 (2011): 96–126; Christopher Johnson, ‘ “French” Cybernetics’, French Studies 69.1 (2015): 60–78; Lydia H. Lui, ‘The Cybernetic Unconscious: Rethinking Lacan, Poe, and French Theory’, Critical Inquiry 36 (2010): 288–320; Ronan Le Roux, ‘Structuralisme(s) et cybernétique(s): Lévi-Strauss, Lacan et les mathématiciens’, Les Dossiers d’HEL 3 (2013) [accessed 9 August 2017]. ¹⁴ [‘un petit groupe de travail que j’avais créé à l’Unesco et qui travaillait sur les rapports mathématiques-sciences humaines. Il y avait là Guillebeau, Claude Berge, Rémi Chauvin, Benvéniste, Lacan, etc’]. Le Lionnais, Un Certain disparate, [accessed 9 August 2017]. ¹⁵ Roudinesco, Lacan & Co, p. 560.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

Literature Machines: StructurElism versus StructurAlism

33

structuralism, Lévi-Strauss, Lacan, and Benveniste, along with the mathematician Guilbaud whose La Cybernétique (1954) will become one of the most influential early expositions of cybernetics in French. In his own small way, then, it seems that Le Lionnais acted as something of a midwife for the structuralism that would emerge over the next decade. Two years after the UNESCO group was formed, Lévi-Strauss would convene another regular group that would draw heavily on the membership of the original working party. As Bernard Geoghegan describes it, With [Roman] Jakobson’s help, Lévi-Strauss secured $2000 for MIT’s Centre for International Study (CENIS), a center for cybernetic research covertly funded by the CIA, in order to organize an interdisciplinary seminar on cybernetics in Paris.¹⁶

Lévi-Strauss’s grant application, submitted to CENIS in January 1953, outlines an extraordinarily distinguished line-up of speakers for the seminar—the psychologist Jean Piaget, the mathematicians Benoit Mandelbrot and Roger Penrose—but alongside these are the members of the earlier group: Guilbaud and Lacan, plus Lévi-Strauss himself.¹⁷ The proposal even states that ‘[w]e expect to receive some help from Unesco, which would put a meeting room every week at our disposal’.¹⁸ Although Le Lionnais is not mentioned by name here, it seems that his interest, and his influence at UNESCO, can still be relied upon. The CENIS seminar was certainly influential. If we want a sense of the way that its discussion of cybernetics immediately affected the thought of some of its attendees, we need look no further than the transcript of Lacan’s own seminar the following year. Here, in classes such as the celebrated discussion of Poe’s Purloin’d Letter, Lacan’s language is replete with images of cybernetics, of ‘thinking machines’, of ‘the rotating memories of our machines-that-think-like-men’.¹⁹ Here he is, describing a machine capable of linguistic output: From the point of view of language, these little machines purr something new for us, perhaps an echo, an approximation let us say. One can’t resolve the issue simply by saying that it is the builder who put it there. The

¹⁶ Geoghegan, ‘From Information Theory to French Theory’: 119. ¹⁷ Claude Lévi-Strauss, letter to Max Millikan, 7 January 1953. Roman Jakobson Papers, MIT Archives, Cambridge, MA, Box 50, Folder 29. I am grateful to Bernard Geoghegan for sharing this material with me. ¹⁸ Lévi-Strauss, letter to Max Millikan, p. 2. ¹⁹ Jacques Lacan, The Seminar of Jacques Lacan, Book II: The Ego in Freud’s Theory and in the Technique of Psychoanalysis, 1954–1955, ed. by Jacques-Alain Miller, trans. by Sylvana Tomaselli (New York: Norton, 1988). See, in particular, ‘Psychoanalysis and Cybernetics’ (pp. 294–308), but also pp. 54, 88, 119, 169, 178–9, 181–6, 279.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

34

The Oulipo and Modern Thought

language came from outside, that’s understood, but it isn’t enough to say that it’s the little chap who put it there.²⁰

It won’t do, in other words, simply to say that the machine’s message comes from the programmer determining the machine’s input. But if Lacan’s image puts us in mind of Ducrocq’s Calliope, this is no coincidence. The literature machine—the system which generates meaning by its organization, by the play of combinatorics and difference—will become one of structuralism’s key metaphors. We will encounter it below in Umberto Eco’s dissection of the James Bond novels, and again in Calvino’s 1967 lecture ‘Cybernetics and Ghosts’ (discussed in Chapter Four), and when Barthes pronounces the Author dead, the cogs of the literature machine can be heard whirring ominously at the crime scene. For the Oulipo, the relevance of cybernetics’ entry in to the humanities should be obvious. Aside from Le Lionnais’s personal involvement in this disciplinary crossover, we have a group whose foundation text— Queneau’s Cent mille milliards de poèmes—is described by Calvino as ‘not so much a book as the rudimentary model of a machine for making sonnets’ (and which, of course, quotes Alan Turing for its epigraph);²¹ a group who announce themselves in the introduction to their first publication with the declaration that ‘the time of created creations, which was that of the literary works we know, should cede to the era of creating creations, capable of developing from themselves and beyond themselves, in a manner at once predictable and inexhaustibly unforeseen’.²² From the start, the very idea of potential literature has been about the agency of the text itself, developing from itself and not the expression of an author. It is only natural then that the possibilities of cybernetics within the literary sphere should be of fundamental interest. By the end of their second meeting in 1960, the pressing need for a computer was already being felt. A note marked ‘Top Secret’ was circulated with the minutes, stating that certain members had been tasked with gaining access to one of the computers at IBM or Bull: ‘[t]heir goal is to try to utilize electronic machines for different labours of literary analysis, as part of the activities of

²⁰ Lacan, Seminar, 1954–55, p. 119. ²¹ Italo Calvino, ‘Cybernetics and Ghosts’, in The Uses of Literature: Essays, trans. by Patrick Creagh (San Diego: Harcourt Brace Jovanovich, 1986), pp. 3–27, (p. 12). ²² ‘Le Collège de ’Pataphysique et l’Oulipo’, in ‘Dossier 17’, i.e. Dossiers du Collège de ’Pataphysique 17 (1961): 1–2 (2). This translation appears as ‘The Collège de ’Pataphysique and the Oulipo’, in Motte, pp. 48–50 (pp. 48–9). Probably written by Latis, who taught philosophy, the passage echoes Spinoza’s distinction between natura naturata [natured nature], i.e. passive nature created by God, and natura naturans [naturing nature], i.e. active nature identified with an immanent creative God.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

Literature Machines: StructurElism versus StructurAlism

35

the OLiPo [sic]’.²³ Speaking three years later at a gathering of quantitative linguists, Queneau would still admit that ‘not having machines at our disposal [is] a continual lamento at our meetings’.²⁴ The value of computing for Oulipian research was keenly felt from the start then, and several members had some experience in the field. In the summer of 1963, five of the group’s ten founder members attended a conference at Cerisy—the same conference series at which the group had initially been conceived three years earlier. This time, under the title ‘Pensée artificielle, pensée vécue’ (‘Artificial Thought, Lived Thought’), the conference theme was cybernetics. Le Lionnais gave the opening address, pointing to the way that cybernetics had in recent years captured both public and academic imagination: After a period of ignorance, as much on the part of the wider public as one the scientific community, which was followed by an enthusiastic fascination, cybernetics has ended up provoking the most diverse reactions. They range from naive credulity to systematic opposition, taking in snobbery, measured approval, scepticism and severe criticism.²⁵

A few days later in a strand on cybernetics and the arts, Duchateau delivered his ‘Communication sur l’Oulipo’, in which he would return to these themes of scepticism, snobbery, and opposition: Literature and machines: that sounds wrong. A priori, it has the ring of something completely contradictory. Literature means freedom; machines are synonymous with determinism. But not all machines are like the ones that automatically dispense platform tickets or peppermints. The essential characteristic of the machines that interest us is not that they are determined but that they are organized.²⁶

²³ [‘Leur but est de tenter d’utiliser des machines électroniques pour différents travaux d’analyse littéraire, dans le cadre des activités de l’OLiPo.’] Bens, Genèse de l’Oulipo, p. 38. ²⁴ [‘Nous regrettons de ne pouvoir disposer de machines: lamento conitnuel au cours de nos réunions’]. Queneau, ‘Littérature potentielle’, p. 322. ²⁵ [‘Après une période d’ignorance, tant de la part du grand public que des milieux scientifiques, à laquelle succèda un engouement enthousiaste, la Cybernétique a fini par provoquer les réactions les plus diverses. Elles vont de la crédulité naïve à l’opposition systématique en passant par le snobisme, une approbation mesurée, le septicisme [sic] et des critiques sévères.’] François Le Lionnais, ‘Pensée Artificielle, Pensée Vécue’, Temps mêlés 66–7 (1964): 34–6 (34). ²⁶ [‘Littérature et machine, ça sonne mal. A priori cela a même l’air parfaitement contradictoire. Littérature ça veut dire liberté, machine est synonyme de déterminisme. Mais toutes les machines ne sont pas semblables à celles qui distribuent automatiquement des tickets de quai ou des pastilles de menthe. La caractéristique essentielle des machines qui nous intéressent ce n’est pas d’être déterminée, c’est d’être organisée.’] Jacques Duchateau, ‘Communication, à Cerisy, sur l’Oulipo’, Temps mêlés 66–7 (1964): 13–21 (14).

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

36

The Oulipo and Modern Thought

It is the organization—the systems or structures—of machines that is most interesting to Duchateau.²⁷ He goes on: Writers have always used structures: some consciously, others unconsciously; some with the feeling that they are simple truths endorsed by their time. From an intuitive point of view, a truth determined by its time is a criterion. From, say, a structuralist point of view, all that is evident is suspect. Forms that are general, accepted by all, modeled by experience, can hide infraforms. A systematic questioning is necessary to discover them. Questioning which leads, in addition to this discovery of underlying forms, to the invention of new forms: exactly as previous ages invented the rules of the sonnet, the dramatic unities, the epistolary novel, the division of the novel into chapters, rhyme . . . , etc . . . .²⁸

This, remember, is a talk delivered as an introduction to the Oulipo. What Duchateau is describing here is patently the day-to-day work of the group: a systematic deep analysis of literary works to discover underlying ‘infraforms’, with the aim of using this knowledge to invent new literary forms to stand alongside the sonnet, the epistolary novel, and so on. And yet, Duchateau is speaking explicitly ‘from a structuralist point of view’.²⁹ Structuralism is rearing its head here as the very method of the Oulipo, its way of getting beyond what is evident, every-day, to reveal the unconscious forms underneath. Duchateau’s choice of wording here is highly surprising, especially given what will happen over the next few years, but it ²⁷ Duchateaus’s talk is singled out for special praise by Queneau at the meeting immediately after the conference: ‘With regard to Cerisy, I would like to say that we are full of admiration for the presentation which Duchateau gave’. [‘A propos de Cerisy, je voudrais dire que nous sommes pleins d’admiration pour l’exposé que fit Duchateau’.] Oulipo, ‘Compte rendu dactylographié signé «Louis Maigret et Louise» de la réunion n° 36 du 22 août 1963’ (FO, DM-1 (35)). ²⁸ [‘Les écrivains ont toujours utilisé des structures: certaines consciemment, d’autres inconsciemment; quelques-unes avec le sentiment qu’il s’agissait de simples évidences entérinées par le temps. D’un point de vue intuitif, l’évidence contrôlée par le temps est un critère. D’un point de vue structuraliste, disons, tout ce qui est évidence est suspect. Des formes assez générales, admises par tous, modelées par l’expérience, peuvent cacher des infra-formes. Une remise en question systématique est nécessaire pour les découvrir. Remise en question qui conduira, outre à cette découverte de formes sous-jacentes, à l’invention de formes nouvelles: exactement comme furent inventées autrefois les règles du sonnet, des trois unités, du roman par lettres, la division du roman en chapitres, la rime . . . etc . . . ’.] ‘Communication de Jacques Duchateau sur l’Oulipo à Cerisy’, (FO, DM-1 (35)). ²⁹ When Duchateau’s talk was initially circulated to the group’s members the month after the conference, it features a word—disons [let’s say]—which does not appear in version published the following year in the Temps mêlés version: ‘D’un point de vue structuraliste, disons, tout ce qui est évidence est suspect’. What this implies is that in the earlier version, structuralism is a convenient generalization, an allusion which might be helpful for the Cerisy audience: ‘From what we might call a structuralist point of view . . . ’. In the published version, however, the analogy has firmed up into something more like a statement of affiliation. [Duchateau, ‘Communication sur l’Oulipo’: 15.]

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

Literature Machines: StructurElism versus StructurAlism

37

shows us that by 1963 certain Oulipians were already considering the relationship between their own researches and those of the structuralists. If this was Duchateau speaking for himself alone, then the wider group could not avoid similar considerations a few months later when the guest at the October 1963 meeting was Maurice Gross, a computer scientist who had made the group’s acquaintance at the Cerisy cybernetics conference. Gross specialized in machine translation, and had recently returned from a spell at MIT working with Chomsky. His theme for the meeting was phonetics, and he presented the group with a diagram showing an analysis of English phonemes into their distinctive features, that is, a table of all of the sounds used in spoken English with each one expressed as a complex of features which are either positive or negative. Thus (using Chomsky and Halle’s slightly later and more sophisticated table), the phoneme /b/ as in beak can be expressed as +consonantal, +anterior, +voiced, which makes it distinct from the /p/ sound of peak which is +consonantal, +anterior, –voiced.³⁰ This type of analysis, developed by Jakobson in the preceding decade, is the very exemplar of structuralist phonetics: the value of each phoneme comes about solely from its place in the system, from the fact that it is different from, and can be opposed to, other phonemes in the table. Thus, when Le Lionnais asked, ‘Are there antiphonemes for every phoneme?’, Gross answered in the affirmative.³¹ This is precisely the point: the structure is a total one, and this is what invests each part with its own identity.³² structuralism in its purest form, then, has arrived, by invitation, into the group’s monthly meeting. And yet, what Gross has to show still pertains to linguistics, to the smallest ³⁰ Noam Chomsky and Morris Halle, The Sound Pattern of English (New York: Harper & Row, 1968), pp. 176–7. ³¹ [‘Existe-t-il des antiphonèmes pour tous les phonèmes?’] ‘Compte rendu dactylographié signé «Les frères Tarots» de la réunion n° 38 du 15 octobre 1963’ (FO, DM-1 (37)). Gross, clearly thinking that his initial answer is unclear or overly brief, writes to Le Lionnais some time after the meeting with more information about antiphonemes. The letter, which is undated, is filed in the archive under December 1966, although this seems implausibly late. [‘Lettre autographe de M. Gross à François Le Lionnais, sans date’ (FO, DM-2 (10)).] Le Lionnais certainly has some understanding of structuralist linguistics, if not in 1963, then by 1968, since he writes to the Romanian mathematician Solomon Marcus, ‘Naturally I own and admire your work, Introduction mathematique à la linguistique structurale [Mathemetical Introduction to Structuralist Linguistics]’ [‘Je possède naturellement et j’apprécie votre ouvrage “Introduction mathematique à la linguistique structurale” ’] ‘Lettre dactylographiée de François Le Lionnais à Solomon Marcus du 12 août 1968’ (FO, DM-2 (29)). ³² Barthes will make the same point about structuralist narratology: that nothing in narrative is without function, and no detail is too trivial to warrant our finding a role for it in the workings of the narrative system. As he puts it, ‘[narrative] is a pure system: there are no wasted units’—or, more epigrammatically, ‘Art does not acknowledge the existence of noise’. [Barthes, ‘Structural Analysis of Narrative’: 245.]

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

38

The Oulipo and Modern Thought

units of speech. Like Strachey’s first draft of the ‘Love Letters’ algorithm, Gross is dealing in the quantifiable, in elements—sounds—that have more or less hard boundaries; his tables have nothing to say about moods or themes, about jubilance or reproachfulness, playing bridge or going dancing. And it is when the Oulipo themselves will begin to move beyond the linguistic and into the realms of moods and themes that their relationship to structuralism will start to become problematic. November 1964. At the group’s monthly meeting Queneau brings up the subject of Henry James’s notebooks. Published posthumously in 1947, the James notebooks reveal the novelist considering his composition process, pondering, in one section, the varying effects which might be achieved by combining situations in different ways. This suggestion—of a combinatorics operating at the level of plot—immediately provokes a sustained discussion among the Oulipians, with almost everyone present expressing a view on whether their own investigations might be taken beyond the level of the word: LE LIONNAIS: Up till now we have been working chiefly on words. For some time, I think that we have been interested in more complex groupings. Groups of ideas, of characters, of situations, etc . . . DUCHATEAU: The problem lies in the fact that words are easy to count, but the elements of a group consisting of characters . . . BRAFFORT: We mustn’t become obsessed with the quantitative. DUCHATEAU: With the qualitative it is easy to spill into stylistics or psychology. BERGE: I agree with Braffort. It is enough to be cautious. We can work on collections of ideas. LATIS: Of ideas? You mean structures. Ideas—allow me to be sceptical. BRAFFORT: We structure ideas through mathematics. SCHMIDT: Since you are talking about the structure of ideas, I should like to remark—of course, I’m speaking from my own experience, but I don’t think it is unique—that when we consider a book we do it according to some theme dear to us. Aren’t we thus looking for structures of intention? LATIS: Every reader reduces the work to the structures he allows. [...] BERGE: We are speaking a lot about structures, perhaps too much. We need, for example, to define and differentiate constraint and structure.³³ ³³ [‘LE LIONNAIS: Jusqu’à présent nous avons surtout travaillé sur les mots, je crois que depuis quelques temps nous nous intéressons à des ensembles plus complexes.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

Literature Machines: StructurElism versus StructurAlism

39

There are some differences of opinion here, some resistance—notably from Duchateau and Latis—to taking Oulipian method beyond what might be observed objectively and definitively. With that suspicious rhetorical question—‘You mean structures?’—Latis contemptuously identifies the elephant in the room. A literary investigation focused on the way ideas are combined? Surely this is exactly the type of work being done in the wake of Lévi-Strauss and Vladimir Propp? Allow me to be sceptical Duchateau’s reservations, meanwhile, are more measured: when the field of investigation expands beyond the precise lines of the textual—beyond operations on words or letters—and into the more indistinct terrain suggested by Le Lionnais—ideas, characters, situations—the analysis is liable to lose its character of quantifiability, and its conclusions to fall within subjective disciplines: stylistics, or psychology. This conversation shows us two things. Firstly, we get to glimpse the beginning of the long gestation of the ‘semantic Oulipo’—essentially, the entry of narratology into the Oulipian sphere, of constraints that operate on character and plot rather than just on letters and words; secondly, it gives us some insight into the antipathy that some members felt towards the literary structuralism which was gaining momentum around them. These twin themes seem hard to reconcile. Surely a move beyond the linguistic, beyond the sentence, will suggest links worth exploring with certain academic fellow travellers? Surely an Oulipian analysis of plot will have much in common with the work being done in the academy by theorists like Barthes and A. J. Greimas? Looking back, in the early 1980s, Arnaud has something to say about these shared interests: ‘With all due deference to certain persons, the fact remains that on a few important points (beginning with the way of regarding literary production), structuralist preoccupations [ . . . ] were not wholly unrelated to the

Ensembles d’idées, de personnages, de situations, etc . . . . DUCHATEAU: Le problème réside dans le fait qu’il est facile de compter des mots, mais les éléments d’ensemble qui sont des personnages . . . BRAFFORT: Il ne faut pas s’olnubiler avec le quantitatif. DUCHATEAU: Avec le qualificatif on verse facilement dans la stylistique ou le psychologisme. BERGE: Je suis de l’avis de Braffort. Il suffit de faire attention, nous pouvons travailler sur des ensembles d’idées. LATIS: Des idées? Vous voulez dire structures; les idées, permettezvous d’être sceptique. BRAFFORT: Nous structurons les idées par les mathématiques. SCHMIDT: Puisque vous parlez de structure d’idées, je voudrais faire remarquer – bien sûr j’évoque ma propre expérience, mais je ne crois pas qu’elle soit unique –, que lorsque nous jugeons un livre nous le faisons en fonction d’une thématique qui nous est chère. Est-ce que, alors, nous ne sollicitons pas des structures intentionnelles? LATIS: Tout lecteur ramène l’oeuvre à des structures qu’il admet. [ . . . ] BERGE: Nous parlons beaucoup de structures, trop peut-être; il faudrait par exemple, définir et différencier: contrainte et structure.’] ‘Compte rendu dactylographié de Jacques Duchateau de la réunion n° 54 du 14 novembre 1964’ (FO, DM-1 (50)).

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

40

The Oulipo and Modern Thought

preoccupations of the Oulipo.’³⁴ Despite the apparent simplicity of Arnaud’s statement here, this is in fact an extremely interesting appraisal of the situation: it is one of the few places after 1963 where this obvious truth is stated plainly and publicly, rather than suppressed or denied. But what is most fascinating here is Arnaud’s particular formulation: that disclaimer, ‘With all due deference to certain persons’; the exasperation of ‘the fact remains’. What comes across loud and clear is that Arnaud knows that he is about to provoke a certain awkwardness within the group, that ‘certain persons’ will not agree with what he is about to say. And after tentatively avowing a theoretical affinity with structuralism, he hastily follows it up by denying anything more concrete: ‘Aside from personal friendships, however, there were no relations between the Oulipo and the structuralists.’ What Arnaud reveals in a few sentences here is not the shared concerns between these two groups—despite repeated and strenuous denials, that much was always obvious; instead, it reveals that the Oulipo were not of one mind in their opinion of structuralism, just as they were split over the viability of moving away from ‘quantifiable’ procedures. As the Oulipo make their first tentative steps into semantic constraints a tone of rivalry enters the relationship with the structuralists, and negotiations about whether to make contact with them take on a difficult, delicate character. Here is another exchange from the monthly meetings, this time from March 1965, four months after the discussion of the Henry James notebooks. Duchateau has been reading the journal Communications and wants to share it with the group: DUCHATEAU: I’d like to draw everyone’s attention to issue 4 of Communications which is on the theme of ‘Semiological Research’. Christian Metz has an article, ‘Le Cinéma, langue or langage?’, which reports on [Abraham] Moles’s manifesto of permutational art and takes the opportunity to state that ‘sovereign-manipulation is not a fertile path for cinema (nor indeed for poetry)’. There is also an article on the narrative message, it’s all very interesting, intelligent. Naturally (?), the Oulipo is not cited. Is it a coincidence, or is it rather a consequence of some, let’s say, ultra-linguistic positions that they adopt? QUENEAU: We must read it. ARNAUD: I’ve read it. There is a great analysis by Barthes; they know a lot. ³⁴ Arnaud, ‘Prolegomena’, p. xiv.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

Literature Machines: StructurElism versus StructurAlism

41

LESCURE: So do we. QUEVAL: Perhaps not the same things. QUENEAU: Let’s read it, then we can invite or not invite accordingly.³⁵ What can we make of this? Firstly, and importantly, Communications is one of the most significant forums for the structuralist movement in literary criticism. Founded by Violette Naville in 1961, with Barthes as one of its regular contributors, it published many of the leading voices in the new narratology. Issue 8 (1966), in particular—a special number dedicated to structuralism and narrative—featured contributions from Greimas, Tzvetan Todorov, Umberto Eco, and Gérard Genette, as well as Barthes’s celebrated introduction.³⁶ Some of the group have already read it—this type of criticism is on their radar and they are positive about it. Their enthusiasm, however, is checked, both by Lescure’s defensive assertion that ‘we know a lot too’, and by Queval’s subtler dismissal, that perhaps what the Communications group know and what the Oulipo know are not the same things. Queneau is carefully equivocal, prevaricating even: let’s find out more, he says twice, perhaps we’ll end up inviting them, but we’re not obliged to. The most telling comment, however, belongs to Duchateau when he notes that the Oulipo are not referenced in any of the Communications pieces. ‘Naturellement’, he says, and the minutes include a parenthetical question mark. Why? Duchateau himself ³⁵ [‘DUCHATEAU: Je voudrais signaler le numéro 4 de “Communications” qui est consacré aux “Recherches Sémiologiques”. Christian Metz dans un article: Le Cinéma, langue ou langage? fait état du manifeste de l’art permutationnel de Moles et en profite pour affirmer “la manipulation souveraine n’est pas une voie féconde pour le cinéma (ni d’ailleurs pour la poèsie).” Il y a aussi un article sur le message narratif, tout ça c’est très intéressant, intelligent. Naturellement (?) l’Oulipo n’est pas cité. Est-ce un hasard, ou bien est-ce une conséquence des positions, disons, ultra-linguistique, qu’ils adoptent. QUENEAU: Il faudrait que nous en prennions connaissance. ARNAUD: J’ai lu, il y a une grande analyse de Barthes; ils savent beaucoup de choses. LESCURE: Nous aussi. QUEVAL: C’est peutêtre pas les mêmes. QUENEAU: Prenons connaissance, nous pourrions inviter, ou ne pas inviter, selon.’] ‘Compte rendu dactylographié de Jacques Duchateau de la réunion du 8 mars 1965’ (FO, DM-1 (54)). ³⁶ It was also the journal which Le Lionnais would later snub over a proposed special issue on his concept of ‘third sector’ literature (i.e. forms of public writing—graffiti, epitaphs—that take place outside the sphere of publication). In Un Certain disparate, Le Lionnais will remember, ‘At one point, Violette Naville, very enthusiastic about the third sector, had proposed to me to make it the subject of a special issue of a very structuralist journal, but leaving the complete direction and the structuralist interpretation to others. I did not accept.’ [‘À un moment donné, Violette Naville, très enthousiasmée par le troisième secteur, m’avait proposé d’en faire le contenu de numéros spéciaux d’une revue très structuraliste, mais en laissant la direction complète et l’interprétation structuraliste à d’autres. Je n’ai pas accepté.’] Le Lionnais, Un Certain disparate, [accessed 9 August 2017].

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

42

The Oulipo and Modern Thought

is, in fact, the minute-taker at this meeting. Does his question mark imply a retrospective frustration, an unwillingness to accept that the Oulipo, still semi-secret, should naturally remain outside the fray? Or does the problem rest, as Duchateau explicitly suggests in the meeting, with the Communications group? A snobbery over tone or jargon: that the journal’s ‘ultralinguistic’ angle makes its writers unlikely to consider the semi-ironic findings of the Oulipo’s knockabout approach? The following year, at the August meeting, we find a familiar set of concerns rearing their heads. On the matter of moving beyond sentencelevel constraints, there has been some development, with Le Lionnais remarking that ‘little by little we were moving from the syntactic to the semantic’.³⁷ Regarding structuralism, however, there is little sign of resolution. Le Lionnais again: The notion of structure, which is for us the same as it is for certain perfectly honourable non-Oulipians, needs to be (re)clarified. It must simultaneously be understood as (1) structure in the mathematical sense of the word; (2) any adopted restriction; and (3) any algorithm similarly chosen.³⁸

The Oulipo are having to define structure in their own terms (albeit in terms that can encompass both procedural and ‘semantic’ exercises), and Le Lionnais seems to be walking a political tightrope between factions within the group when he describes the structuralists as ‘certain perfectly honourable non-Oulipians’—a delicately ironic distancing gesture if ever there was one. Flash forward seven years, and in the group’s first major publication, La Littérature potentielle, we find these same two themes—a drift towards using broader, looser textual features as the elements of Oulipian constraint, and a discomfort verging on disdain for academic structuralism— given public expression in Le Lionnais’s ‘Second Manifesto’: The overwhelming majority of Oulipian works thus far produced inscribe themselves in a SYNTACTIC structurElist perspective (I beg the reader not to confuse this word – created expressly for this Manifesto – with structurAlist, a term that many of us consider with circumspection). Indeed, the creative effort in these works is principally brought to bear on the formal aspects of literature: alphabetical, consonantal, vocalic, syllabic, phonetic, graphic, prosodic, rhymic, rhythmic, and numerical constraints, structures, or programs. On the other hand, semantic aspects were not dealt with, meaning having been left to the discretion of each author and excluded from our structural preoccupations. ³⁷ ‘Circular No. 75 (+/ )’, (i.e. meeting minutes for August 1966), p. 188. ³⁸ ‘Circular No. 75’, p. 190.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

Literature Machines: StructurElism versus StructurAlism

43

It seemed desirable to take a step forward, to try to broach the question of semantics and to try to tame concepts, ideas, images, feelings, and emotions. The task is arduous, bold, and (precisely because of this) worthy of consideration.³⁹

Over the course of nearly a decade then the message is always the same: we are working more and more with narratological structures, but don’t mistake us for structuralists. For some Oulipians—namely those, like Latis, who would never fully embrace the idea of semantic Oulipo— there is no paradox here, only the same hostility to a perceived conceptual wooliness, whether practised in the academy or in the group’s meetings. For others, however, there is clearly something else at work, some reason behind the apparent cognitive dissonance by which structuralism is viewed in an, at best, lukewarm light while at the same time pursuing, or at least supporting, Oulipian researches into semantic constraints. A look at some of Queneau’s writing, both publicly and within the group, will shed some light on what might have been at stake in maintaining this seemingly untenable position. By the mid-1960s, structuralism had become the dominant intellectual mode in French humanities and social sciences. If we want a sense of the movement’s significance, we might look to Calvino’s description, in a letter at the end of 1968, of ‘the crowds that squeeze into Lacan’s and Barthes’s seminars (the former’s seminars in particular are of such difficulty that this mass attendance can only be explained in terms of a cult)’.⁴⁰ Across the Channel, however, its impact—in the literary sphere, at least— had been felt to a far lesser degree. Thus, in September 1967, the Times Literary Supplement (TLS) ran an article entitled ‘Science versus Literature’ in which Barthes offered British readers an introduction to structuralist literary criticism. The article’s title, ‘Science versus Literature’, is meant ironically—a kick at the past—since structuralism, argues Barthes, can dissolve that opposition by offering ‘a science of literature’.⁴¹ It is a method which ‘seeks to establish the “language” of the stories that are told, their articulation, their units and the logic which links these together’. As for how this type of analysis might work, Barthes informs us that ‘structuralism gives special attention to classification, hierarchies and arrangements’. The units and logic of stories; a critical method given ³⁹ Le Lionnais, ‘Second Manifesto’, p. 29. ⁴⁰ Letter to Michele Rago, 31 December 1968. Italo Calvino, Letters, 1941–85, ed. by Michael Wood, trans. by Martin McLaughlin (Princeton, NJ: Princeton University Press, 2013), p. 361. ⁴¹ Roland Barthes, ‘Science versus Literature’, TLS 3422 (28 September 1967): 897–8 (897).

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

44

The Oulipo and Modern Thought

to classification, hierarchy, and arrangement: once again, to anyone with even a passing interest in the Oulipo, this must all sound strangely familiar. And by a curious coincidence, just a few pages earlier in the very same issue of the TLS, in an article with an almost identical title, Queneau was introducing the Oulipo to British readers for the first time. ‘Science and Literature’ begins by bemoaning the ‘Two Cultures’ rift that had long afflicted the French education system, placing an unwarranted barrier between the arts and the sciences. Nevertheless, notes Queneau, a recent shift in the academy has meant that ‘[l]inguists, psychologists and sociologists can no longer remain ignorant of mathematics’; moreover, ‘[c]riticism and literary history have followed suit’.⁴² Of course, Queneau is describing the structuralist turn here, and yet there is a curious unwillingness to mention the movement by name. Moreover, when he goes on to introduce the Oulipo, Queneau is careful to preface his description with the proviso that no relationship of influence exists between the two groups: the Oulipo may be ‘not entirely foreign to the movement I have just been speaking of but [it] has arisen completely separately’.⁴³ When it comes to describing the group, Queneau notes that ‘So far, all the work of the Oulipo has been concerned with the syntactic aspect of literature; it is now moving in the direction of its semantic aspects (in which we have had precursors such as Propp and the Russian Formalists)’. Here again the language again is strikingly evasive. Vladimir Propp? Certainly. The Formalists? Indeed. But still Queneau is unwilling to write the name of structuralism. Structuralism, it seems, occupies a fraught spot in Queneau’s imagination: as the Oulipo’s academic Other, lurking a few streets away at the Collège de France, or a few pages away in the TLS. One wonders what British readers must have made of these two articles, outlining the same problem in French culture and introducing two almost identical saviours, yet both either oblivious to the other or unable to speak its name. Would they have detected, in Queneau’s contortions, the need to communicate the distinctiveness of his own brand, to ensure at all costs that it should not be mistaken for or casually elided with its larger, more familiar rival? There is some evidence that this was starting to become a real concern for Queneau.⁴⁴ ⁴² Queneau, ‘Science and Literature’: 863. ⁴³ Queneau, ‘Science and Literature’: 864. ⁴⁴ A curious, personal parallel to this situation concerns the relationship between Barthes and Perec. An important early mentor for Perec, Barthes had been a reader and enthusiastic supporter of Perec’s novel Les Choses (1965), a work for which Barthes’s own Mythologies is one of the key intertexts. However, in the summer of 1967, just as Perec was joining the Oulipo, the relationship would become strained. Barthes sent Perec the proofs of his Le Système de la mode for comment, only to receive ‘an aggressive response ending with a

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

Literature Machines: StructurElism versus StructurAlism

45

At the start of the 1970s, the Oulipo found itself in crisis: still semi-secret—it had failed, due to the Jonathan Cape debacle, to produce a major publication—and, for all its labours, without having given the world a single new literary form of any significance. Things were serious enough that a questionnaire was circulated to members with a note stating that ‘The gravity of the situation demands a written response from you’.⁴⁵ The questionnaire is entitled ‘Comment déboucher l’horizon?’—‘How to unblock the horizon?’—and the opening questions give a clear sense of how deep the malaise within the group has become: Do you want the Oulipo to continue its activities? If yes, why do you think we are asking you this question? Do you think a regeneration is possible? When Queneau completes his form he responds to the final ‘Any other comments’ question as follows: ‘The situation is not the same as it was ten years ago. A certain number of (para)Oulipian concerns are now in the public domain. (To be developed orally.)’⁴⁶ It will in fact be some time before the minutes show Queneau developing this theme in one of the group’s meetings, but the following summer he provides a sense of what these para-Oulipian concerns might be: ‘A number of writers now – notably those associated with the nouveau roman, or the Tel Quel group, or Change – are looking at the use of refined, sometimes quite intricate, constructions. Is their research any different from ours?’⁴⁷ In the decade since the group’s formation, its territory has become contested: the Oulipo has found itself in a crowded marketplace searching for a USP. When Queneau found himself unable to mention structuralism by name in the TLS, it is tempting to assume that, consciously or not, his reticence stemmed from a dawning awareness that the identity of the Oulipo was under threat.

scathing pun’. Two years later, when La Disparition appeared, Barthes would refuse to read it at all. David Bellos, Georges Perec: A Life in Words (London: Harvill, 1993), pp. 298–9, 355, 430. ⁴⁵ FO, MS-5 (‘Documents Queneau, relatifs à l’Oulipo, photocopiés au CIDRE, 1990’), f. 23. ⁴⁶ [‘La situation n’est plus la même qu’il y a dix ans. Un certain nombre de préoccupations (para)oulipiennes sont maintenant dans le domaine public. (À développer oralement).’] ⁴⁷ [‘En effet, plusiers écrivains aujourd’hui, notamment ce qui se réfèrent au nouveau roman, ou à l’équipe de Tel Quel, ou à celle de Change, sont attentifs à l’utilisation de constructions recherchées, parfois délicates. Leurs recherches diffèrents-elles de nôtres?’]. ‘Copie du compte rendu dactylographié de la réunion du 27 août avec corrections de Jacques Bens et François Le Lionnais’ (FO, DM-3 (8)). These journals are cited again in a moment of open hostility that threatens a serious rift in the group in the summer of 1974. See Bloomfield, Raconter l’Oulipo, pp. 470–4.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

46

The Oulipo and Modern Thought

From a distance of half a century, this anxiety might seem faintly absurd. The Oulipo has endured, while we can see clearly now how literary structuralism, by the late 1960s, had already begun to evolve into something more mercurial, a shift which will be traced in more detail in Chapter Four. Something else which is perhaps more apparent now than it was at the time is how differences not of method but of character would always have made the two groups unlikely collaborators. Take for example, two analyses of the roman policier or crime novel, one of which appeared in Communications, the other in Subsidia Pataphysica, the journal of the Collège de ’Pataphysique. Both are exclusively concerned with categorizing and organizing the elements of the genre, and yet both are utterly different in tone. In Umberto Eco’s ‘James Bond: Une combinatoire narrative’, Eco’s argument is that the Bond novels of Ian Fleming ineluctably follow a pattern of fixed plot elements. Their sequence may vary, and certain elements may be repeated, but the form of a Bond novel is nevertheless a combination of nine narrative events which play themselves out, like a game of chess in which white always wins. This, then, is Eco’s general schema of a Bond novel: A.—‘M’ plays and entrusts Bond with a mission B.—The Villain plays and appears to Bond (possibly in the form of a substitute) C.—Bond plays and lands the first blow on the Villain—or else the Villain lands a blow on Bond. D.—The Girl plays and presents herself to Bond. E.—Bond steals the Girl; he has her or undertakes to have her. F.—The Villain captures Bond (with or without the Girl, or at different times) G.—The Villain tortures Bond (with or without the Girl) H.—Bond beats the Villain (he kills him or kills his substitute or assists in his death) I.—Bond convalesces with the Girl, whom he will lose afterwards.⁴⁸ ⁴⁸ [A.—‘M’ joue et confie une mission à Bond. B.—Le Méchant joue et apparaît à Bond (éventuellement sous une forme substitutive). C.—Bond joue et inflige un premier échec au Méchant—ou bien le Méchant inflige un échec à Bond. D.—La Femme joue et se présente à Bond. E. —Bond souffle la Femme; il la possède ou entreprend la possession. F.—Le Méchant prend Bond (avec ou sans la Femme, ou en des moments divers).

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

Literature Machines: StructurElism versus StructurAlism

47

Dr No follows the schema in its simplest form: its structure is ABCDEFGHI; Goldfinger is a somewhat more complex BCDEACDFGDHEHI; From Russia with Love is a positively baroque BBBBDA (BBC) EFGHGH (I). This type of narrative combinatorics leads Eco to gesture towards the mechanical: ‘With the cynicism of a disenchanted gentleman, Fleming intends to construct a working narrative machine.’⁴⁹ For a mode of criticism still bearing the traces of its early fascination with cybernetics, the machine remains the dominant metaphor. Eco’s analysis is brilliant. Moreover, it is expressed with an authoritativeness and a confidence in its own approach that is utterly characteristic of Communications’ structuralist evangelism. Whether it is entirely convincing, however, is another matter. Looking at the formula for From Russia with Love, are we certain that we are in the presence of a combinatory system when some elements can be repeated so liberally? And the fact that some elements (those in parentheses), while indispensable, may appear only implicitly: we are sure that they really do occur in the novel, aren’t we? There is, in Eco’s argument, a certain sleight of hand at work, aided by the robustness of the Communications style. And yet here it is entirely forgivable, since the point that the article wants to make is not really that Fleming’s novels obey a rigid programme, but that they at least obey a loose one. For all the fun and fireworks of reducing every Bond narrative to a game of nine moves, Eco’s key statement is a considerably milder one: ‘what characterizes a crime novel, be it a detective novel or a thriller, is not so much the variation of facts as the return of a habitual schema in which the reader can recognize something they have seen before and which gave them pleasure.’⁵⁰ In other words, we always know where we are with a Bond novel, and this comfort is one of the genre’s chief joys. The audaciousness of Eco’s demonstration allows him to reach this conclusion, but it would be wrong to look too closely at how we got there. As such there is something pleasingly brisk and loose (though not fast-and-loose) about Eco’s argument. Alongside Christian Metz’s Communications piece on cinema, for example—the ‘ultra-linguistique’ G. —Le Méchant torture Bond (avec ou sans la Femme). H.—Bond bat le Méchant (il le tue ou en tue le substitut ou assiste à sa mort). I. —Bond convalescent s’entretient avec la Femme, qu’il perdra par la suite.] Umberto Eco, ‘James Bond: Une combinatoire narrative’, Communications 8 (1966): 77–93 (87). ⁴⁹ [‘Fleming entend, avec le cynisme du gentleman désenchanté, construire une machine narrative qui fonctionne.’] Eco, ‘Bond’: 91. ⁵⁰ [‘ce qui caractérise le roman policier, fût-il enquête ou action, ce n’est pas tant la variation des faits que le retour d’un schéma habituel dans lequel le lecteur pourra reconnaître quelque chose de déjà vu et qui lui a plu.’] Eco, ‘Bond’: 90.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

48

The Oulipo and Modern Thought

article which Duchateau brought up at the Oulipo’s monthly meeting— Eco’s piece is not exactly tongue-in-cheek, but conveys a sense of ironic awareness in the seriousness and scientificity of its tone. It has swallowed the Communications style guide, but there is still a twinkle in its eye. By contrast, Le Lionnais’s ‘Les Structures du roman policier: Qui est coupable?’, written in early 1970, but published in 1972, makes facetiousness its dominant note. Le Lionnais’s analysis does not take the form of a combinatorics, but is rather a taxonomy of the murder mystery, an attempt, over three pages, to exhaust and classify every possibility of who the killer might be in a piece of whodunnit fiction. Of course there are the usual suspects, novels in which the killer is a human and known by the victim, and these can be broken down into further subgroups based on the relationship between victim and murderer. But other categories include non-human murderers: murders by animals (Poe’s Rue Morgue), zombies, intelligent viruses; then there are the suicides, the accidental deaths, killings by drifters or dangerous drivers peripheral to the main investigation; novels where the identity of the killer remains ambiguous to the end, where multiple options are kept in play, unresolved, and novels where no-one is identified as a suspect; there is a playful type of novel where the killer is the book’s author, or even—in the case of one P. G. Wodehouse story—its editor. Le Lionnais concludes with a final top-level category: a blank, a possibility which no crime novel has yet exploited. (This after all, is one of the reasons that systematic representation is so powerful: it can bring to light known unknowns.) For this last category, Le Lionnais states that here has not yet been a policier in which the reader is the killer. And yet, he claims, this is a genuine possibility: ‘I have been able to draw up the scheme for one which appeals neither to the supernatural nor to the hoax’.⁵¹ Thus, this comprehensive study of the mystery novel ends mischievously with an unsolved riddle of its own. Given its method, not to mention the sheer volume of literary expertise on show, ‘Les Structures du roman policier’ makes a neat companion piece to Eco’s essay; and yet stylistically it could hardly be further from a Communications article (not least because of its presentation in the form of a tree diagram). Le Lionnais’s exposition is a kind of burlesque of criticism, a refusal to treat structural analysis with the fathomless reverence it receives elsewhere. It is there in the way that the categories of the analysis are obviously provisional (this is true, of course, for the nine plot elements of Eco’s Bond treatment: we could easily find others that work just as well, ⁵¹ [‘solution dont j’ai pu établir le schéma et qui ne fait appel ni au surnaturel, ni au canular’]. François Le Lionnais, ‘Les structures du roman policier: Qui est coupable?’, Subsidia Pataphysica, 15 (1972): 16–18.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

Literature Machines: StructurElism versus StructurAlism

49

but the seriousness and authority of the presentation utterly discourage us from doing so), and it is there in the riddle left hanging at the end: the undeniable erudition of ‘Les structures’ is shot through with an insistent playfulness. A truly Oulipian treatment of the policier could never be acceptable to the editors of Communications, even if its method was fundamentally compatible with structuralist orthodoxy. This is the uniqueness of the Oulipian approach. Looking back from a couple of decades’ distance, Arnaud will imply that this temperamental incompatibility worked in both directions, that a difference, not in approach, but in manner would have made the two groups mutually unattractive: the [structuralists], moreover, enveloped themselves in a ponderous sobriety that rendered them impervious to Oulipian facetiae. (Personally, I shall except from this one of the founding fathers, Claude Lévi-Strauss, with whom, during these years, I had the privilege of exchanging a delightful correspondence from which it became clear that, to his way of thinking, science was not principally devoted to making people die of boredom.)⁵²

Ponderous sobriety, making people die of boredom: in the final analysis, then, it was by defining themselves in opposition to these perceived distinctive features of structuralism that the Oulipo maintained its identity as it transitioned into the business of semantic constraints. This chapter has looked at the strained relationship between structuralism and the Oulipo. Outwardly, of course, there was considerable overlap in both the interests and methods of the two groups. Both follow a trajectory from syntax to semantics. Structuralism, developing out of linguistic fields—Saussure’s Cours de linguistique générale; Roman Jakobson’s work in phonology—where its quantifiable analysis was uncontroversial, then expanded into narratology where it ran up against long-held and deeply-invested beliefs about what Literature is. Likewise, by the mid-1960s, several members of the Oulipo were becoming disenchanted with the word substitutions of the group’s early methods and were raising the question of whether narrative structures too might not be effective material for their investigations. And yet, despite these clear parallels, whenever the Oulipo tell their story, it is other groupings—mathematical collectives like Bourbaki, artistic groups like the Surrealists—whose names appear; when it comes to structuralism, surprisingly, there is, on the whole, only a deafening silence. What we have found, however, is that in private the story was a different one. Certain Oulipians had been present, even instrumental, during ⁵² Arnaud, ‘Prolegomena’, p. xiv.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

50

The Oulipo and Modern Thought

structuralism’s formative stages, and the similarities between the two groups were both real and known to the Oulipo as a whole. The group continued to read and discuss structuralist publications throughout the 1960s, and Barthes was even touted as a potential invitee in 1970.⁵³ The hostility towards, or perhaps more precisely, the suspicion of structuralist literary criticism affected some, but not all of the Oulipo’s members, began to be felt after a particular moment, and was bound up with the Oulipo’s own expansion into narratological constraints (which did not proceed with the smoothness and unanimity that Le Lionnais’s manifesto would have us believe). Working outwards from this material, and thinking more broadly about critical work on the Oulipo, what we can learn from the example of structuralism is that we must be cautious of making generalizations, even when they are informed by the group’s public pronouncements. The archive can help here: if we look below the surface, we find a collective which encompasses a spectrum of opinions, and whose own evolution can and should be carefully historicized. What we find too is a group highly conscious of what is going on around them, tapped in to the academic journals—on friendly terms, even, with the editors of these publications— but conscious too of maintaining their own identity. As we will find with the example of Surrealism later, just because the Oulipo treat certain modes of thought with scorn, or decline even to mention them publicly, we should not for a minute assume that these are not being closely considered and debated within in the privacy of the group’s own meetings.

⁵³ Barthes’s name is listed, in Le Lionnais’s handwriting, on the agenda for the August 1970 meeting. However, it is doubtful that the invitation was issued, and Barthes never makes an appearance in person. [‘Pré-programme dactylographié du congrès d’août 1970 avec ajouts autographes de François Le Lionnais’ (FO, DM-2 (54)).]

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

2 The Punning Muse Psychoanalysis and Homophonic Translation In the beginning was the pun. —Samuel Beckett, Murphy

September 1898: On a visit to the Adriatic coast, Sigmund Freud finds himself sharing a coach with a man, a lawyer from Berlin, whom he has not met before. The two men strike up a conversation during which Freud tells an anecdote about the local Bosnian Turks which he has heard from a colleague some time before. The story goes that the Bosnians hold the medical profession in high regard, but are also possessed of a national character which is ‘resigned towards the dispensations of fate’.¹ This combination of fatalism and respect is such that, should a doctor ever report that a patient is untreatably, fatally ill, the response of the Bosnian Turk will be to reply, ‘Herr [Sir], what is there to be said? If he could be saved, I know you would help him.’ Having told this tale to his travelling companion, Freud then recalls another story about the Bosnians, told to him by the same colleague. This time, however, he does not repeat it. The reason is its inappropriate subject matter: the second anecdote is about the importance which the Bosnians attach to sexual pleasure. In this story, the patient tells his doctor, ‘Herr, you must know, if that comes to an end then life is of no value.’ Later on in their conversation, Freud recommends that his companion visit Orvieto in central Italy to see the vast frescoes of the Last Judgement, painted by Signorelli, which are in the cathedral there. To Freud’s frustration, however, when offering this piece of travel advice, Signorelli’s name escapes him. Instead the names of two other painters, Botticelli and

¹ Sigmund Freud, ‘The Psychical Mechanism of Forgetfulness’ (1898), in Early PsychoAnalytic Publications, Standard Edition of the Complete Psychological Works of Sigmund Freud, 26 vols. (London: Hogarth and the Institute of Psycho-Analysis, 1953–66) III, pp. 287–97 (p. 287).

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

52

The Oulipo and Modern Thought

Boltraffio, keep popping into his head, even though he knows them to be wrong. Looking back on the episode in the early psychoanalytic paper, ‘The Psychical Mechanism of Forgetfulness’ (1898), Freud tries to analyse firstly why he should have repressed the name of the correct artist, and secondly by what mechanism the other two names suggested themselves to him. What was going on, he concludes, was a series of puns, working across two different languages. The artist’s name—Signorelli—has been broken down into its component syllables, and the first part—Signor— translated into its German equivalent, Herr. The term Herr also appears in the anecdote about the Bosnian Turks and their sexual proclivities, a story which Freud at that moment has chosen to withhold from his interlocutor. Moreover, Herr is a charged syllable at the time of the conversation in the coach because it also occurs at the start of Herzegovina, where the two men happen to be. Bosnia is similarly loaded, its first syllable appearing in the names of the two interloping painters: Botticelli and Boltraffio. Now the -elli ending of the unnamable Signorelli can still be expressed but in a neutralized alternative, Botticelli.² ‘Experience,’ Freud remarks, ‘has taught me to require that every psychical product shall be fully elucidated and even overdetermined’.³ Thus further details—a recent stay in the town of Trafoi, the suicide of a patient—are pulled in to complete the puzzle. Words—towns, names— are pulled apart and reconstituted, their syllables swapped out, translated, heard as echoes at the beginnings, middles and endings of other words. Homophony—the similarities that exist between words, or parts of words, even across languages—is taken for part of the mechanism of forgetfulness by which we repress what we don’t want to bring into the open, in this case death and sexuality. But the intrepid analyst might shine a torch along this web of sounds, following up the patient’s disparate mishearings, spiralling towards the unsaid motivation at their centre. Legitimated by psychoanalysis, at the turn of the twentieth century the pun has become more than mere play: it has become evidence. For the Oulipo, the extended pun—or homophonic translation, as they term it—has always been a significant practice. As Harry Mathews puts it,

² If this explanation seems tortuously unconvincing, this was not lost on Freud who first described it in a letter to his friend Wilhelm Fliess, concluding, ‘How can I make this credible to anyone?’ [Letter of 22 September, 1898; Jeffrey M. Masson (ed. and trans.), The Complete Letters of Sigmund Freud to Wilhelm Fliess, 1887–1904 (Cambridge, MA and London: Belknap, 1985), p. 327]. ³ Freud, ‘Psychical Mechanism of Forgetfulness’, p. 294.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

The Punning Muse: Psychoanalysis and Homophonic Translation

53

the fondness of Oulipians for the method is [ . . . ] undeniable: Italo Calvino, Harry Mathews, Michèle Métail, and Georges Perec – not to mention founder François Le Lionnais and the illustrious advance plagiarist, Raymond Roussel – have all exploited homophony, some extensively.⁴

The uses of Oulipian homophony are various. This chapter will outline several different types of extended punning in their work, and certainly not all of these are bound up with a notion of the manifest and latent of psychoanalysis. Nevertheless, one early example, from the Oulipo dossier of the Collège de ’Pataphysique, shows how, even in jest, members of the group were attuned from the start to the possibility that the drives of the psychoanalytic unconscious might underpin their homophonic wordplay. Noël Arnaud’s ‘Athalie Galante?’ (1961) takes the opening four lines of Racine’s theatrical masterpiece Athalie and reads it not as a declamation of religious observance, but as a sex scene between a libertine and his mistress. Thus, this: Oui, je viens dans son temple adorer l’Eternel; Je viens, selon l’usage antique et solennel, Célébrer avec vous la fameuse journée Où sur le mont Sina la loi nous fut donnée. [Yea, I have come to pay adoration to the Eternal in His temple; I have come, according to the ancient and solemn custom, To celebrate with you the famous day When on Mount Sinai the Law was given unto us.]

becomes this: Lui.—Oui? Elle.—Je viens. Lui.—Dansons . . . Elle (regardant par la croisée).—Temps plat d’Auray . . . Lui (montrant un couple de passants).—Laid. Elle.—Terne, elle. Lui (enchaînant, et grivois).— Je . . . Elle. Viens, seul on l’use . . . Lui.—Ah! gentil. Qu’est-ce?

Him.—Yes? Her.—I’m coming. Him.—Let’s dance . . . Her (looking out of the window).—Dull times in Auray . . . Him (pointing at a couple of passersby).—Ugly. Her.—Drab, her. Him (linking, and bawdily).—I . . . Her. Come, only we shall use it . . . Him.—Ah! my sweet. What is it?

⁴ Mathews and Brotchie, Oulipo Compendium, p. 154.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

54

The Oulipo and Modern Thought Elle.—Oh! l’âne! Lui.—Elle sait les braies avec vous, l’affameuse . . . Elle (implorant).—Joue, René. Lui (caressant).—Où? Elle.—Sur le mont! . . . Si! Lui.—Na! Elle.—Là! Lui.—L’oie! Elle.—Noue! Lui (déjà rassasié, éprouve le besoin d’aller imiter le peuple saint en foule).—Fût d’eau . . . Elle (dépitée).—Nez. Variante optimiste et prénatale: Elle (songeant à l’avenir).—Nais.⁵

Her.—Oh! the ass! Him.—She knows your britches, that exploiter . . . Her (imploring).—Play, René. Him (caressing).—Where? Her.—On the mons! . . . Yes! Him.—Nah! Her.—There! Him.—Goose! Her.—Fasten it! Him (already satisfied, feeling the need to imitate the worshipful).— The water barrel . . . Her (disappointed).—Nay. An optimistic, prenatal variant: Her (dreaming of the future).—Born.

The joke here, in Arnaud’s set-up, is that Racine, who famously retreated into religion in his old age, was unable to control his lingering libertinism which bubbled up in disguised form in this famous scene. As Arnaud puts it, ‘la main de Racine n’obéissait pas toujours à sa volonté consciente’ [‘Racine’s hand did not always obey his conscious will’]. The suggestion that Racine’s libidinous drives, repressed and unconscious, have found a way into his writing is tongue-in-cheek: ‘Athalie Galante?’ is not an endorsement of psychoanalytic interpretation but a pastiche of it, a demonstration of Arnaud’s comic ingenuity. Not all Oulipian punning, however, is purely whimsical. Homophonic wordplay might work within the same language, as with ‘Athalie Galante?’, or it may travel across different languages, as in Freud’s Signorelli elucidation. It might even traverse different stages of the same language—it is easy to imagine recasting the opening syllables of The Canterbury Tales or Beowulf in modern English (‘What wee gardener in yard . . . ’).⁶ One might think too of the example which Tullio de Mauro gives of a phrase which can be read in Latin or Italian to produce two quite different meanings: ‘I VITELLI DEI ROMANI SONO BELLI’ [‘Go, o Vitellius, to ⁵ Noël Arnaud, ‘Athalie Galant?’, Dossiers du Collège de ’Pataphysique 17 (i.e. ‘Dossier 17’) (1961): 33. ⁶ David Melnick’s Men in Aida (Berkeley, CA: Tuumba, 1983) meanwhile performs a homophonic translation on the first book of the Iliad to produce a poem that begins, ‘Men in Aida, they appeal, eh?’.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

The Punning Muse: Psychoanalysis and Homophonic Translation

55

the battle sound of the Roman god’; ‘The calves of the Romans are beautiful’].⁷ But a more curious case, and one of direct influence on the Oulipo concerns the strange etymological theories of Jean-Pierre Brisset (1837-1919), who claimed to have discovered the entire history of the world from the homophonic similarities between French words. In his long poem Petite cosmogonie portative (1950)—a history of the world in verse—Queneau includes the following line: ‘De quelque calembour naît signification’: ‘Out of some pun signification was born’.⁸ It is a surprising suggestion. Surely, we imagine, homophony—the single sound which represents multiple things—belongs to an imperfect state of language, the consequence of what Foucault calls ‘the insolvency of words which are fewer than the things they designate’?⁹ Why should the first signification be tarred with this imbalance? What Queneau has in mind, it seems, is an allusion to his favourite of the fous littéraires, the literary madmen, in whose work he consoled himself following his split from Breton and the Surrealists. In J.-P. Brisset’s own cosmogony—an outsider text: a ‘sudden whirlwind’ as Foucault describes it, the very antithesis of traditional academic philology— the pun holds the key to the history of the world.¹⁰ On 13 April 1913, Brisset, a retired railwayman and author of a pair of books on the origins of language, arrived in Paris for a day of celebration in his honour. A crowd had gathered to greet him at Montparnasse Station, and as they chanted his name, the old man, overcome, struggled to find words to express his pride: Pierre Brisset semblait en proie à une grande émotion. Ses mains et ses lèvres tremblaient un peu, quand il répondit quelques pauvres phrases: –Je n’ai pas l’habitude de parler en public . . . Je pense dans la solitude . . . –Très bien! très bien! criaient d’une voix éclatante les gaillards vigoureux, joyeux et prospères rangés autour de lui. Vive Brisset! Vive Brisset!¹¹ [Pierre Brisset seemed prey to a great emotion. His hands and his lips trembled a little when he answered a few poor phrases: ‘I am not used to speaking in public . . . I think in solitude . . . ’ ‘Splendid! Splendid!’ With one clear voice, the hardy fellows, joyous and thriving all around him, cried out. ‘Long live Brisset! Long live Brisset!’]

⁷ Ferdinand de Saussure, Corso di linguistica generale, trans. by Tullio de Mauro (Bari: Laterza, 1967), p. xv. ⁸ Raymond Queneau, Petite cosmogonie portative (Paris: Gallimard, 1950), p. 64, III.128. ⁹ Michel Foucault, Death and the Labyrinth: The World of Raymond Roussel, trans. by C. Ruas (London: Continuum, 1986), p. 167. ¹⁰ Michel Foucault, ‘Seven Remarks on the Seventh Angel’, trans. by Dennis Duncan, PMLA 132.5 (2017): 1251–62 (1254). ¹¹ Louis Latrazus, ‘La Victime’, Le Figaro, 14 April 1913: 4.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

56

The Oulipo and Modern Thought

Speeches were said in his honour, and orotund verses read, before Brisset was whisked to a celebratory dinner, then on to the Place du Panthéon where he would deliver a speech in front of his ‘collègue de bronze’, Rodin’s The Thinker.¹² Here he informed his audience that, in spite of the statue’s lack of clothes, ‘il n’est pas nécessaire d’être nu pour penser’ [‘it is not necessary to be naked in order to think’].¹³ At this point an impromptu choir formed and improvised a song in front of the swelling crowd of students and shopkeepers who inhabited the Latin Quarter. Brisset had been elected ‘Prince des Penseurs’—Prince of Thinkers— beating the other nominees, among them Henri Bergson, by a considerable margin. The title, however, was a joke, dreamt up by the writer Jules Romains and his ‘bohemian cronies’, and conferred with the utmost irony.¹⁴ The whole day was an elaborate prank at Brisset’s expense. His work in etymology was the object of ridicule, and the students, writers, and painters who turned out to listen in mock awe as he held forth could barely conceal their mirth: On l’appelait Maître. On l’accablait sous les hyperboles. On se confondait en admiration à ses moindres mots. Et puis, on détournait la tête pour rire à l’aise. Et on échangeait des sarcasmes à voix basse. Il ne voyait pas, et n’entendait rien.¹⁵ [We called him Master. We overwhelmed him with hyperbole. We confounded ourselves with admiration at his every word. And then we turned our heads so that we could laugh at ease. And we exchanged sarcasms in hushed voices. He didn’t see, and he heard nothing.]

Brisset remained oblivious, sighing ecstatically that he could die that very evening. Brisset’s theories, which had earned him such spectacular and public derision, concern the origins of language. He derives an etymology based on the astonishing law that ‘Toutes les idées énoncées avec des sons semblables ont une même origine et se rapportent toutes, dans leur principe, à un même objet’ [‘All ideas spoken with similar sounds have a shared origin and all relate, originally, to a shared object’].¹⁶ In essence, the rule is that words ¹² Quoted in Marc Décimo, Jean-Pierre Brisset, Prince des penseurs (Paris: Ramsay, 1986), plate 14. The original appeared in the newspaper, Excelsior, on 14 April 1913. ¹³ Charles Picart le Doux, ‘Quelques notes sur le banquet J.-P. Brisset’, Bizarre 8 (1957): 31–6 (33). ¹⁴ Walter Redfern, All Puns Intended: The Verbal Creation of Jean-Pierre Brisset (Oxford: Legenda, 2001), p. 5. ¹⁵ Latrazus, ‘La Victime’: 4. ¹⁶ Jean-Pierre Brisset, La Science de Dieu ou la Création de l’Homme, in Œuvres complètes, ed. by Marc Décimo (Dijon: Presses du réel, 2001), pp. 697–885 (p. 702), Brisset’s emphasis.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

The Punning Muse: Psychoanalysis and Homophonic Translation

57

or phrases which sound the same—grammaire/grand-mêre [grammar/ grandma]—are linked by a common origin. Where others see the opportunity for a pun, Brisset sees heaven-sent truths about the nature and history of the world. The scale of his supposed discovery gave rise to a certain hubris, which no doubt is partly what blinded him to the irony of the festivities in his honour. Brisset has a chapter to himself in André Breton’s Anthology of Black Humor in which Breton points out that: Brisset did not conceal that he was himself dazzled by the brilliance of the gift that he was offering humanity, which he felt should confer upon him divine omnipotence. He recognized no predecessors save Moses and the prophets, Jesus and the apostles. He proclaimed himself to be the Seventh Angel of the Apocalypse and the Archangel of the Resurrection.¹⁷

Of course, the shared origin of the homonymic terms is obscure to the uninitiated, so Brisset goes to extraordinary exegetic lengths to shine a light on these relationships for us, as in the exhaustive punning on the phrase Les dents, la bouche here: Les dents, la bouche [the teeth, the mouth]. Les dents la bouchent [the teeth stop it up]. l’aidant la bouche [with the mouth’s help]. L’aide en la bouche [aid in the mouth]. Laides en la bouche [ugly in the mouth]. Laid dans la bouche [ugly in the mouth]. Lait dans la bouche [milk in the mouth]. L’est dam le à bouche [it’s harm to the mouth]. Les dents-là bouche [those teeth: hide ‘em]¹⁸

Nine different readings of the same set of syllables is all very well, but what does it mean? Before coming to that, Brisset gives us a rather grand introduction, complete with his own mystical symbolism: If I say, les dents, la bouche, it elicits only familiar ideas: one’s teeth are in one’s mouth. That would be the same as fully understanding the exterior of the book of life that is hidden within speech and sealed with seven seals. But in this book, today open before us, we shall now read what was concealed beneath the words les dents, la bouche.¹⁹

¹⁷ André Breton, Anthology of Black Humor, trans. by Mark Polizzotti (San Francisco: City Lights, 1997), p. 182. ¹⁸ Brisset, La Science de Dieu, p. 702. The English translations are from Polizzotti’s translation of Breton’s article on Brisset [Breton, Black Humor, p. 185]. ¹⁹ Breton, Black Humor, p. 185.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

58

The Oulipo and Modern Thought

Finally we get to the analysis: The teeth stop up the entrance to the mouth and the mouth helps with and contributes to this closure: thus Les dents la bouchent, l’aidant la bouche. The teeth are the aid, the support inside the mouth [l’aide en la bouche], and too often they are also ugly inside the mouth [laides en la bouche], and this fact, too, is ugly [laid ]. At other times, it is like milk: they are white as milk [lait] in the mouth. L’est dam le à bouche must be understood: it is harm – evil or damage – visited upon the mouth; put more plainly, I have a toothache. We can see by this that the first harm [dam] originated in the tooth [dent]. Les dents-là bouche means: close up or hide those teeth of yours; shut your mouth.²⁰

It reads more like avant-garde poetry than etymology, but nevertheless, by its own logic, the passage builds towards a conclusion of sorts. Brisset is holding up homophony—and by extension a series of implied etymological connections—as proof of the claim that ‘the first harm originated in the tooth’. Every pun, for Brisset, leads to a revelation about the early history of the human race, and the constellation of these revelations provides him with his extraordinary creation myth in which Man has descended from a race of warrior frogs. This has its origin partly in the homonymy of raîne [ frog] and reine [queen], and partly in an incident in which a frog spoke to him. Queneau describes the incident in an article for Bizarre magazine: Brisset se promenait un jour autour d’un marais peuplé de grenouilles, lorsqu’il fut frappé par la ressemblance de leur corps avec celui de l’homme. Il entendit alors l’une de ces petites bêtes qui lui disait: ‘Coâ, coâ’, ce que Bisset interpréta ainsi: la grenouille lui posait la question primordiale: ‘Quoi? Quoi?’ [One day, as Brisset was walking by a marsh inhabited by frogs, he was struck by the similarity between their bodies and that of a man. Then he heard one of these little creatures say to him, ‘Coa, coa’, which Brisset interpreted thus: the frog was posing him the primordial question: ‘What? What?’]²¹

Once Brisset has convinced himself of the ancestral relationship between Man and frogs, even the most minimal and arbitrary similarity becomes a causal relationship, and therefore further evidence in support of his theory. Taking apart the words saloperie [filth] and duperie [deception], Brisset comes up with the following narrative of our warlike amphibian ancestors: Voici les salauds pris; ils sont dans la sale eau pris, dans la salle aux pris, dans la salle aux prix. Les pris étaient les prisonniers que l’on devait égorger. En ²⁰ Breton, Black Humor, pp. 185–6. ²¹ Raymond Queneau, ‘La théologie génétique de Jean-Pierre Brisset’, Bizarre 4 (1956): 80–8 (80).

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

The Punning Muse: Psychoanalysis and Homophonic Translation

59

attendant le jour des pris, qui était aussi celui des prix, on les enfermait dans une salle, une eau sale, où on leur jetait des saloperies. Là, on les insultait, on les appelait salauds. Le pris avait du prix. On le dévorait et, pour tendre un piège, on offrait du pris et du prix: c’est du prix. – C’est duperie, répondait le sage. N’accepte pas de prix ô homme, c’est duperie. [Here are the captured bastards [salauds pris], they are caught in the dirty water [sale eau pris], in the trophy room [salle aux prix]. The captured [pris] were the prisoners whom we had to butcher. While waiting for the prisoners’ day, which was also prize-giving [des prix], we locked them in a room [salle], dirty water, where we threw slops [saloperies] at them. There we insulted them, we called them bastards [salauds]. The prisoner had a price [du prix]. We ate him, and to lay a trap, we offered a prisoner [du pris] and a price [du prix]: that was the prize [du prix]. It’s a trick [duperie], responded the sage, don’t accept the prize, dear sir, it’s a trick.]²²

As with les dents, la bouche, what is striking here is the excess of ways the same set of syllables can be interpreted. Saloperie can be alternatively read as salauds pris, sale eau pris or salle au prix. Yet, even in excess, no detail dominates—each one carries an objective value, telling us something about how the contemporary world came to be the way it is, something about our strange amphibian history. The very antithesis of his contemporary Saussure, in Brisset’s system the planes of signifier and signified are always conformal: any relationship of similarity between signifiers is mirrored by a corresponding relationship between signifieds. The immense backstory of warring frogs constitutes the narrative system Brisset is forced to create in order to keep arbitrariness at bay. This view of language may be the polar opposite of Saussure’s, but as Queneau reminds us, it has some surprising similarities with another turnof-the-century thinker. In the Bizarre article, Queneau remarks that Brisset has ‘conç[u] une méthode qui parfois se rapproche singulièrement (aussi bien que les résultats) de celle de son illustre contemporain, S. Freud’ [‘devised a method which is sometimes remarkably close (in its results as well) to that of his illustrious contemporary, S. Freud’].²³ And later in the same article he goes on to suggest that there is something rather ‘totem-and-taboo’ about Brisset’s extraordinary creation myth: Le dieu devient homme par l’intermédiaire du démon. Car les démons, les premiers, découvrent la sexualité et – du même coup (si j’ose dire) la paternité: mais une paternité très freudienne, très ‘totem-et-tabou’.

²² Brisset, La Science de Dieu, p. 813. Brisset’s italics indicate words which are being used in his homophonic exegesis. ²³ Queneau, ‘La théologie génétique’: 80.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

60

The Oulipo and Modern Thought

‘Le diable fut [ . . . ] le premier père qui engendra, mais il prit aussitôt ses enfants en haine. Il les mangea longtemps et il entra ensuite en guerre avec eux.’ [God becomes man through the intermediary of the devil. For the demons, the first ones, discover sexuality, and at the same time (I daresay) paternity: but a very Freudian paternity, very ‘totem-and-taboo’: ‘The devil was [ . . . ] the first father to procreate, but he immediately hated his children. For a long time he ate them, then he declared war on them.’]²⁴

In fact, this is not the first time we see Queneau make the connection between Brisset and Freud. His meticulously-kept reading lists show that he first encountered Brisset’s Le Mystère de Dieu est accompli (1890) in the first half of June 1930, while carrying out research for his unpublished work on literary eccentrics.²⁵ A few months later, in a letter to Georges Bataille, he writes: J’ai fait une petite découverte psychanalytique. C’est une phrase de Brisset qui m’a mis sur la voie, phrase dans laquelle il dit que le prépuce est le premier chapeau. Or, je ne sais si tu as remarqué, Freud reconnaît bien le chapeau comme symbole génitale mâle, mais il ne comprend pas pourquoi, par example page 322 de la Science des Rêves (‘il peut paraître bizarre que le chapeau représente l’homme’): cette incompréhension provient certainement de ce qu’il était juif, il ne parle jamais de prépuce et chapeau, gant, couverture (signalés par Brisset) en sont pourtant des symbols fort claires. Qu’en penses-tu? [I have made a little psychoanalytical discovery. It is a phrase from Brisset which put me on track, [a] phrase in which he says that the foreskin is the first hat. Now, I don’t know if you have noticed, but Freud clearly recognises the hat as a symbol for the male genitals, but he doesn’t understand why, for example page 322 [SE V, p. 361] of The Interpretation of Dreams (‘It may seem strange, perhaps, that a hat should be a man’): this incomprehension certainly proves that he was Jewish, he never speaks of the foreskin and hat, glove, coat (all pointed out by Brisset) and yet the symbols are very clear. What do you think?]²⁶

Thanks to Queneau’s silly joke—that Freud, the Jewish doctor, might not know quite what a foreskin is—it is difficult to tell how earnest he is being in the rest of this passage. At any rate, however, the theories of the two are being equated—Brisset’s wordplay complements Freud’s

²⁴ Queneau, ‘La théologie génétique’: 85. ²⁵ Florence Géhéniau, Queneau analphabète: répertoire alphabétique de ses lectures de 1917 à 1976, 2nd edn (Brussels: F. Géhéniau, 1992), p. 142. ²⁶ Letter to Georges Bataille, 5 October, 1931, in Raymond Queneau, Journaux 1914–1965, ed. by Anne Isabelle Queneau (Paris: Gallimard, 1996), p. 243n.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

The Punning Muse: Psychoanalysis and Homophonic Translation

61

symbolism, even spills into it, suggesting solutions to problems which Freud had yet to solve. Brisset’s method, like Freud’s, is not purely homophonic: it relies also on an unacknowledged use of metonymy. In the passage which Queneau is referring to, he writes, ‘Le prépuce fut notre première couverture et notre premier habit. Il couvre le membre et nous sommes membres de Dieu’ [‘The foreskin was our first cover and our first garment. It covers the member and we are members of God’].²⁷ This is the metonymic relationship: the foreskin is a jacket covering the penis. But, as with everything in Brisset, it leads on to a homophonic explanation, in this case, rather excruciatingly, the similarity between the German for coat—Decke—and the French bec [beak], which confirms the correctness of the association by suggesting (again metonymically) the penis. Thus, as Queneau realises, Brisset’s system is not one of pure punning, but draws on both homophony and metonymy for its relations, just as psychoanalysis does. The only difference is that where Brisset is concerned with the nature of the world—teeth, milk, frogs—Freud looks inwards, positing these functions as the cornerstones of the repressive mechanism by which the psychological subject obscures itself in discourse. The two systems might be equivalent in their workings, but in their values—and their statuses—they are not. In 1937, Queneau published the autobiographical poem Chêne et chien. Describing it later as ‘un roman en vers’—a novel in verse—he admits that the poem’s subject matter is not generally considered particularly poetic.²⁸ It is a poem about psychoanalysis. For the previous four years, Queneau, then in his early 30s, had been in treatment with the Russian analyst Fanny Lowtsky, travelling daily to her consulting room in Passy in the west of Paris.²⁹ In Chêne et chien, the poem’s punning title contains metaphors for the opposed poles of Queneau’s personality: the oak (chêne) and the dog (chien): Chêne et chien voilà mes deux noms, étymologie délicate: comment garder l’anonymat devant les dieux et les démons. (81)

Oak and dog my two names, fragile etymology: how to remain anonymous between gods and demons.

²⁷ Jean-Pierre Brisset, Les Origines humaines, in Œuvres complètes, pp. 1117–315 (p. 1207). ²⁸ Raymond Queneau, ‘Conversation avec Georges Ribemont-Dessaignes’, in Bâtons, chiffres et lettres (Paris: Gallimard, 1965), pp. 35–47 (p. 43). ²⁹ Albert-Marie Schmidt had also been a patient of Lowtsky’s some years earlier. See Bloomfield, Raconter l’Oulipo, pp. 132–3.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

62

The Oulipo and Modern Thought

Queneau’s struggle is to maintain this fragile etymology, holding both meanings, both homophones, in balance, privileging neither: the oak, as he puts it, noble and grand, reaching for the heavens; the dog: cynical, indelicate, revelling in its instincts, nosing in bins. Chêne et chien is written, consciously, under twin signs: the sanctioned, stately clinical interpretations of Freudian analysis, and Brisset’s wild, excessive, creative overreading. In a snatch of dialogue in the poem—an argument with his analyst (recast as male)—Queneau brings the two influences together, obliterating the gulf between them as a way of pouring scorn on psychoanalysis. Comparing the analyst to an ornamental ceramic, he declares, ‘tu mériterais d’être brisé’ (80): ‘you deserve to be smashed’, but also, punningly, ‘you deserve to be Brisset’. It’s an exquisite line, an insult that unravels itself, performing its meaningful condensation at the very moment it sets out to abuse and ridicule it. It is a moment of punning that captures magnificently the Oulipo’s ambivalence about homophonic play, holding the serious and the playful, the Apollonian oak and the Dionysian dog, in balance. It is time then to consider Oulipian homophonic translation in some more detail. If we think back to Arnaud’s deliberate misreading of Racine, which excavates for the erotic beneath the devotional, it is hard not to notice that the homophonic reading comes at a cost. Arnaud’s new dialogue takes a peculiarly strained form of French, its syntax awkward, staccato, unrealistic, its imagery scattergun: ‘Yes! Nah! There! Goose!’ It takes a generous reader to force these lines to function as a fluent narrative, to treat the scene as a coherent vignette of comic romance and not a confrontational exercise in Dadaist randomness. It is perhaps because of this formal strangeness in its output that homophonic translation is used more commonly by the Oulipo as one stage in a larger process of narrative creation, rather than the sole device at work. Of course, Arnaud’s reading shares very little thematic common ground with Racine’s original—this is part of its wit: a total irreverence for its highly revered source. So if Arnaud’s text were then to be paraphrased, the seduction scene rewritten in plainer terms, then even this phonological link to Racine would be severed. There would be no trace left of the source whatsoever: both the meaning and sound of the original would have been effaced. The resultant text, although procedurally derived from Athalie, would give no clue at all as to its origin. Within the Oulipo, this method is frequently used to create vignettes that comically set up or paraphrase the homophonic target text. As Mathews explains, ‘in Oulipian practice, homophony is generally used to create sequences of short fictitious episodes; these are presented like riddles whose “solutions” are the homophonies (often preposterous) on

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

The Punning Muse: Psychoanalysis and Homophonic Translation

63

which they are based.’³⁰ By way of example, Keats’s title Ode to Melancholy becomes O to mail Hank holly!, which gives rise to the following scenario: Stranded in that benighted oasis, the Kansan was disconsolate. Soon he would lose all chance of celebrating Christmas in a civilised Christian place; and he already found himself unable to send Henry, his oldest friend, a traditional symbol of seasonal cheer, as he had ritually done every year since they were boys. Among his many regrets, this perhaps saddened him the most. O to mail Hank holly! ³¹

It is precisely the preposterous nature of the homophony which gives this device its appeal. The juxtaposition of imagery from widely divergent fields allows for a particular type of zaniness in the explicatory backstories, which—as in the best shaggy dog stories—is offset with a formal, circumlocutory mode of telling. Perec in particular was drawn to this type of play, returning to it in extended exercises throughout his time in the Oulipo, from ‘Lieux communs travaillés’ (1972) to ‘Cocktail Queneau’ (1982).³² Moreover, the method has given rise to a number of Oulipian feeding frenzies in which several members of the group join together to produce dozens of variations on the same source text. In La Cantatrice Sauve (BO 16), for example—the title a play on Ionesco’s La Cantatrice chauve—a ‘homophonic fever’ struck members of the group during a long car journey together.³³ In short order, stopping only when their hosts’ exasperation with them became too much, Mathews and Perec, along with Jacques Bens, Claude Burgelin, Paul Fournel, and Béatrice de Jurquet, produced a hundred and one homophonic transformations of the name of the opera singer Montserrat Caballé, each one preceded by a short narrative leading up to the homophonic punchline: L’amante anglais, victoriennement pubibonde, de Rackham le Rouge, état sollicitée par ce dernier de goûter aux plaisirs de la sodomie. Outrée, elle refusa longtemps de se plier à de telles exigences. Enfin, pleine de mépris pour elle et son bougre infâme de séducteur, elle lui chuchota à l’oreille: ‘MON C . . . , RACKHAM BAS, YES’ [Red Rackham’s English mistress (a somewhat Victorian prude), was solicited by the former to sample the delights of sodomy. Outraged, for a long time she refused to give in to such suggestions. At last, full of contempt for herself and her notorious seducer, she whispered in his ear: ‘My A . . . , filthy Rackham, yes.’] ³⁰ Mathews and Brotchie, Oulipo Compendium, p. 154. ³¹ Mathews and Brotchie, Oulipo Compendium, p. 154. ³² These and others are collected in Georges Perec, Voeux (Paris: Seuil, 1989). ³³ Claude Burgelin et al., La Cantatrice sauve, BO 16, in Oulipo, Bibliothèque oulipienne (Paris: Ramsay, 1987), I, pp. 305–22 (p. 337).

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

64

The Oulipo and Modern Thought

Ayant pêché l’autre jour deux tourteaux et les ayant déposés sur la table de la cuisine, quelle ne fut pas ma surprise quand, préparant les condiments, je vis un des crustacés griffonant à l’autre le message suivant: ‘M . . . ! ON SERA CRABES AILLÉS’ [The other day, having caught two shellfish and deposited them on the kitchen table, imagine my surprise when, as I prepared the condiments, I saw one of the crustaceans scribbling the following message to the other: ‘Sh . . . ! We’re going to be garlicked crabs.’]³⁴

Thirteen years later, in another collective exercise—a commission from the city of Strasbourg—the group would return to this device, achieving a similar punning excess from, as they put it, ‘a systematic exploitation of the possibilities offered by the dislocation of the six syllables of the sequence “Le tramway de Strasbourg”’.³⁵ Once again, the exercise throws up dozens of mini-narratives, such as: Chez les Gavagaï, la sueur blanche des rongeurs, étalée sur une tartine, pass pour une friandise de premier choix. Lait de rats moites: extra, ce beurre! [Among the Gavagai, the sweet cream of rodents, spread on bread, passes for a first-class delicacy: Moist rats’ milk: this butter’s terrific! ]³⁶

What this example should make clear is that part of the humour in these exercises comes from the degree to which the homophony is strained. Thinking of Brisset, Foucault calls his use of the pun l’à-peu-près—an approximation—and Mathews and Brotchie describe homophonic translation as ‘a terminus ad quem, since an exact replication of sounds in another language is impossible’.³⁷ In the Bibliothèque oulipienne examples, the punning pulls at the extremes of homophony, treating vowel lengths as discretionary, fusing syllables where elision is barely warranted, occasionally using the wrong sound altogether. There is a playfulness, a testing of the limits of the phonetic, euphonic aspects of the language, which is foregrounded in these examples, and each time another degree of licence is added to the rule of homophonic similarity, the semantic potentiality inherent in the syllables of the source text is raised to another power. But, there is also something more serious going on. In their excessiveness, pulling dozens of different puns from the same short phrase, what

³⁴ Burgelin et al., La Cantatrice sauve, p. 317. ³⁵ [‘une exploitation systématique des possibilités offertes par la dislocation des six syllabes de la séquence Le tramway de Strasbourg.’] Oulipo, ‘Presentation: L’OuLiPo et le Tramway de Strasbourg: une première’ (FO, MS-3 (‘Strasbourg, Sept 94, Textes préparatoires’), f. 2). ³⁶ Oulipo, Troll de tram (Le tramway de Strasbourg ), BO 68, in Oulipo, Bibliothèque oulipienne (Bordeaux: Castor Astral, 2000), V, pp. 151–60 (p. 156). ³⁷ Foucault, ‘Seven Remarks’: 1258; Mathews and Brotchie, Oulipo Compendium, p. 154.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

The Punning Muse: Psychoanalysis and Homophonic Translation

65

Fig. 6. Photograph of street furniture in Strasbourg displaying one of the group’s punning narratives. (FO, MS-3, f. 99). Image reproduced by kind permission of the Oulipo.

these exercises do is to demonstrate the potential within a sequence of syllables for extreme overdetermination. Just as Brisset riffs on saloperie or les dents dans la bouche to produce not one but many alternative phrases from the same set of syllables, these Oulipian outings are similarly prolific. Of course, there is nothing special about the source texts, nothing intrinsically suggestive about ‘Montserrat Caballé’ or ‘Le tramway de Strasbourg’ themselves. The text is arbitrary—this homophonic potential is inherent in any sequence of syllables. If we were in any doubt, then Perec’s ‘Petit abécédaire illustré’—an exercise he composed as a Christmas gift for his friends in 1969—plays at the same game but switches the emphasis from an excess of derivations to an excess of source material. Perec takes each consonant of the alphabet and combines it with all five vowels (thus, ba-be-bi-bo-bu; da-de-di-do-du, etc.), and proceeds to construct a narrative for the homophonic translation of each sequence: Dans un salon, des dames papotent au sujet de l’adultère. (Caquet: qui cocu?) [In a sitting room, some women are gossiping on the subject of adultery. (Cackle: who’s a cuckold?)]³⁸ ³⁸ Georges Perec, ‘Petit abécédaire illustré’, in Oulipo, La Littérature potentielle, pp. 235–40 (p. 235).

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

66

The Oulipo and Modern Thought

Fig. 7. Bonne année pataphysique [Happy Pataphysical New Year]. Postcard with two homophonic translations (‘Housemaid has Pataphysical nose’; ‘Good ass shocks boy (sic)’) to mark the Pataphysical new year, 96 E.P. (8 Sept 1968). On the reverse of the card sent by Stanley Chapman to Le Lionnais, Paul Gayot has added ‘Intraduisible [untranslatable] in Englisch’ (FO, DM-2 (30)). Image reproduced by kind permission of the Oulipo.

En déplacement à Crémone, le Souverain Pontife scrute anxieusement le fleuve qui sent mauvais. (Pape épie, Pô pue) [On the way to Cremona, the Sovereign Pontiff anxiously studies the foul smelling river. (Pope watches, Po stinks.)]³⁹

And Calvino repeats the feat in Italian with his Piccolo sillabario illustrato (BO 6), in which we find miniature narratives such as this one: Per convincere il proprietario d’un night-club a scritturarla, una spogliarellista lo assicura della propria efficacia nel provocare l’eccitazione degli spettatori. – Sa? Sessi isso su! [In order to convince a nightclub owner to sign her, a stripper assures him of her effectiveness at arousing the audience’s excitement. – You know? Sex gets it up!]⁴⁰ ³⁹ Perec, ‘Petit abécédaire’, p. 238. ⁴⁰ Italo Calvino, Piccolo sillabario illustrato, BO 6, in Oulipo, Bibliothèque oulipienne, I, pp. 97–121 (p. 117).

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

The Punning Muse: Psychoanalysis and Homophonic Translation

67

In English, Mathews has performed a similar exercise by translating the alphabet into a crow’s complaint to a scarecrow: ‘Hay, be seedy! Effigy! Hate-shy, jaky yellow man, O peek, you are rusty, you’ve edible, you exwise head!’⁴¹ What these experiments in homophonic translation amount to is a demonstration of two things. Firstly, that any sequence of syllables may be given semantic value; and, secondly, that the relationship between source and target text is by no means singular, but, rather, a glut of competing interpretations may be drawn from the same original sounds. These assumptions form the basis of an explicit engagement with psychoanalysis in Mathews and Perec’s spoof academic article ‘Roussel and Venice’. Ahead of this, however, it will be necessary to have a look at Roussel’s own methodical use of punning, and the way that it might be used to produce serious, extended prose works of a rather different character to the somewhat flippant mini-narratives of the abécédaires or La Cantatrice Sauve. To some extent, the homonymy that gives rise to punning is unavoidable in natural languages. The ideal language, as the early modern philosopher Comenius put it, would be one that ‘contains neither more nor fewer names than there are things’.⁴² Unfortunately, this is not something which natural languages are able to achieve, hence the comedy in Gulliver’s Travels of the academicians who propose, ‘since Words are only Names for Things’, that it would be more convenient to carry in a sack anything they might wish to refer to, and who travel about like pedlars ‘almost sinking under the Weight of their Packs’.⁴³ And yet, since homonymy is the expedient by which we cope with our lexical deficit, perhaps the situation—hopeless as it is—is not exclusively a miserable one. Even Foucault seems consoled, if only partially: From this lack is experienced the ‘play’ [ . . . ] in the fact that the same word can designate two different things and the same sentence repeated can have another meaning. [ . . . ] There is the misery and the celebration of the signifier, and the anguish before too many and too few signs.⁴⁴

This passage comes from Foucault’s Raymond Roussel, translated into English as Death and the Labyrinth, and the idea that ‘the same sentence repeated can have another meaning’ occupies a central position in Roussel’s creative method.

⁴¹ Mathews and Brotchie, Oulipo Compendium, p. 223. ⁴² Johann Amos Comenius, The Way of Light, trans. by E. T. Campagnac (London: Hodder & Stoughton, 1938), p. 183 (XIX.12). ⁴³ Jonathan Swift, Travels into Several Remote Nations of the World, by Capt. Lemuel Gulliver (London: B. Motte, 1726), II pp. 76–7. ⁴⁴ Foucault, Death and the Labyrinth, p. 167, Foucault’s emphasis.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

68

The Oulipo and Modern Thought

For Roussel, punning offers the same creative potential which has traditionally been assigned to rhyme: that of allowing the writer to escape the familiar. He writes that ‘[t]he method is, in short, related to rhyme. In both cases there is unforeseen creation due to phonic combinations’.⁴⁵ In the same text, Roussel gives a detailed description of the procedure by which he composed the novels Impressions of Africa and Locus Solus. Part of this procedure—the part which Foucault has in mind in the passage quoted above—consists of finding two phrases which differ by only one letter but which have hugely divergent meanings, and using these as the start and end points for a piece of narrative: I chose two almost identical words [ . . . ] For example, billard [billiard table] and pillard [plunderer]. To these I added similar words capable of two different meanings, thus obtaining two almost identical phrases. In the case of billard and pillard the two phrases I obtained were: 1. Les lettres du blanc sur les bandes du vieux billard . . . [The white letters on the cushions of the old billiard table . . . ] 2. Les lettres du blanc sur les bandes du vieux pillard . . . [The white man’s letters on the hordes of the old plunderer . . . ] In the first, ‘lettres’ was taken in the sense of lettering, ‘blanc’ in the sense of a cube of chalk and ‘bandes’ as in cushions. In the second, ‘lettres’ was taken in the sense of missives, ‘blanc’ as in white man, and ‘bandes’ as in hordes. The two phrases found, it was a case of writing a story which could begin with the first and end with the latter.⁴⁶

A little further into his elucidation of the procedure, the method becomes more extreme. When Roussel breaks with syntax altogether and begins to operate at the level not of the word but of the syllable, we become aware of a far deeper degree of Foucauldian ‘lack’ and ‘play’. Roussel describes taking the first line of a song, ‘J’ai du bon tabac dans ma tabatière’ [‘I’ve got some good snuff in my snuffbox’], and breaking this down homophonically into a parallel set of syllables: ‘jade tube onde aubade en mat (objet mat) a basse tierce’. This provides him with a list of items which he must include in the opening of the ‘Poète et la Moresque’ section of Impressions of Africa. Here is how the finished version appears:

⁴⁵ Raymond Roussel, How I Wrote Certain of My Books, trans. by Trevor Winkfield (Cambridge, MA: Exact Change, 1995), p. 14. ⁴⁶ Roussel, How I Wrote, pp. 3–4.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

The Punning Muse: Psychoanalysis and Homophonic Translation

69

Exact et obéissant, le poète descendait chaque matin dans le magnifique jardin entrouant de toutes parts le palais de son maître. Parvenu sous les fenêtres du riche dormeur, il s’arrêtait non loin d’un bassin de marbre d’où s’échappait un svelte jet d’eau lancé par un tube en jade. Élevant alors jusqu’à sa bouche une sorte de porte-voix en métal terne et délicat, Ghîriz se mettait à chanter quelque élégie nouvelle éclose en sa féconde imagination. Par suite d’une résonance étrange, la légère trompe utilisée doublait chaque son à la tierce inférieure.⁴⁷ [Punctually at daybreak, the obedient poet each day went down into the splendid garden which surrounded his master’s palace on all sides. Beneath the rich sleeper’s windows, he took up his station near the marble basin of a fountain, from whose jade spout emerged a slender jet of water. Raising to his lips a kind of speaking tube in dull metal, Ghiriz would then start to sing some elegy newly unfolded in his fertile imagination. The strange acoustical properties of the light trumpet were such that it sounded a lower third to each note of the song.]⁴⁸

While tube, jade, and tierce can all be read directly in the passage, the other terms have been disguised by synonyms: the adjective mat has been replaced with terne to describe the loudhailer, although mat is used when this scene is prefigured earlier in the novel;⁴⁹ the adjective basse has been swapped for inférieure to give the correct musical terminology; onde has become the arc of water projecting from the fountain; the aubade is instead a dawn élégie, although ‘Aubade’ appears elsewhere in the novel as the title of a piece sung by various characters. Further lines from ‘J’ai du bon tabac’, along with another song, ‘Au clair de la lune’, are dissolved homonymically to furnish the tale, an Oriental fantasy in imitation of the Thousand and One Nights, with more detail. It is important to draw some distinctions between the punning exercises of the Bibliothèque oulipienne and Roussel’s novels. While the former are used to produce a deluge of vignettes, Roussel gives us something on a completely difference scale. As Foucault puts it, The rupture of a phonological difference (between p and b, for example) does not produce, for him, a simple distinction in sense but rather an almost uncrossable chasm, which can only be bridged by an entire narrative; and when we stand on one edge of the divide and step out toward the other, it is ⁴⁷ Raymond Roussel, Impressions d’Afrique, 4th edn (Paris: Alphonse Lemerre, 1932), p. 369. ⁴⁸ Raymond Roussel, Impressions of Africa, trans. by Lindy Foord and Rayner Heppenstall (London: Calder and Boyars, 1966), p. 259. Curiously, the English translators retained the word tube from the French but moved it so that it refers not to the spout of the fountain but to the loudhailer used by Ghiriz. ⁴⁹ Impressions d’Afrique, p. 175; Impressions of Africa, p. 126.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

70

The Oulipo and Modern Thought

by no means certain that the story will really get to the other side, so close, so similar.⁵⁰

Similarly, while Mathews writes indulgently of the ‘preposterous’ nature of much Oulipian homophonic translation, Roussel’s wordplay is by no means intended as a way of producing zany material. As Mark Ford points out, Roussel’s fanatically worked manuscripts tell us nothing about the word games underpinning the novel, but demonstrate his utter determination to recast and revise and excise and expand in pursuit of the ‘sensations of art’, ‘the complete illusion of reality’, his novel was to afford.⁵¹

Ford is careful not to imply any causality between Roussel’s word games and his pursuit of ‘the complete illusion of reality’: the former do not bring about the latter; it is only that they do not preclude it. Roussel compares his procedure with a rebus, the device whereby an image is used to represent a word or part of a word (for example, Hamlet’s famous soliloquy might be illustrated as two bees, an oar, a knot, and another pair of bees).⁵² The rebus, unsurprisingly, is a preferred analogy for Freud too, and it also makes its appearance in the work of the Hungarian-French psychoanalysts, Nicolas Abraham and Mária Török. In their secondary analysis of Freud’s famous Wolf Man case, Abraham and Török propose that, in certain circumstances, it is not the meaning or import of a situation which needs to be repressed, but a word or phrase: It is not a situation including words that becomes repressed; the words are not dragged into repression by a situation. Rather, the words themselves, expressing desire, are deemed to be generators of a situation that must be avoided and voided retroactively. In this case, and only in this case, can we understand that repression may be carried out on the word, as if it were the representation of a thing, and that the return of the repressed cannot have at its disposal even the tortuous paths of metonymic displacement.⁵³

The proscribed words they designate as the cryptonym, before setting out to expose the cryptonym underlying the dreams of Freud’s patient, the Wolf Man (also known as SP or Sergei Pankejeff ). Like Freud’s explanation of why he couldn’t remember Signorelli’s name, unravelling ⁵⁰ Foucault, ‘Seven Remarks’: 1259. ⁵¹ Mark Ford, Raymond Roussel and the Republic of Dreams (London: Faber and Faber, 2000), p. 99. ⁵² Roussel, How I Wrote, p. 12. ⁵³ Nicolas Abraham and Mária Török, The Wolf Man’s Magic Word: A Cryptonymy, trans. by Nicholas Rand (Minneapolis: University of Minnesota Press, 1986), p. 20, Abraham and Török’s emphasis.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

The Punning Muse: Psychoanalysis and Homophonic Translation

71

Pankejeff ’s cryptonym is a tortuous procedure for Abraham and Török. What follows is a précis. Firstly, they concentrate on the image, from one of Pankejeff ’s dreams, of ‘six or seven wolves’, using homophony to imply a connection between the Russian words shiest (six) and shiestorka (pack) and siestra and siestorka, both of which mean sister: ‘It became clear that the “pack of six wolves” did not contain the idea of multiplicity, but of [Pankejeff ’s] sister instead’.⁵⁴ From here they go on to look at another of SP’s dreams, this time focussing on the image of a skyscraper. Abraham and Török note that skyscraper—nieboskreb in Pankejeff ’s mother tongue—is Wolkenkratzer in German, the first syllable echoing Volk, a Russian word for wolf. Thus, as with Freud and with Brisset, an integral part of the method is to break words into their constituent syllables, then carry the sounds from one language to another. Next, Abraham and Török note that -skreb (the scrape of skyscraper) has a Russian synonym in natieret, and that natieret can mean not only scrape but rub. Thus, invoking the image of the rebus once again, Abraham and Török are ready to announce what they claim to be SP’s cryptonym—the unsayable phrase which had to be repressed by a shuffling and translating of its syllables: Finally we understood the rebus of the skyscraper! With all the necessary substitutions, the solution is simple: It concerns the association of the wolf with sexual pleasure obtained by rubbing. [ . . . ] ‘Sis, come and rub my penis.’ This was the key sentence. These were the unsayable words that he posted in the form of a rebus.⁵⁵

Pankejeff ’s unsayable phrase, ‘Sis, come and rub my penis’, has been translated into the dream narratives of a skyscraper and a pack of six wolves. Yet there are a frustrating number of gaps and weak points in Abraham and Török’s description of their method. For example, the authors are highly selective of which dream images to use and which to overlook: they overlook the unusual detail that in SP’s drawing of his dream the wolves are sitting on the branches of a tree, and that there are five of them rather than six (and also that SP’s description of the dream speaks of ‘six or seven’ wolves): it is crucial for Abraham and Török that there are precisely six. It is also revealed that Pankejeff can speak English, in addition to Russian and German, making the pool of available syllables for interpretative homophony far greater than Abraham and Török let on: shiest, for example, to an English or French speaker, may homophonically suggest shit/chier, but this meaning has no role in the interpretation they ⁵⁴ Abraham and Török, Wolf Man, p. 17. ⁵⁵ Abraham and Török, Wolf Man, pp. 18–19.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

72

The Oulipo and Modern Thought

wish to make, so it is suppressed.⁵⁶ And finally we might wonder where penis comes from in the cryptonym, since if, as we must assume, it is derived from the dream of the skyscraper, that surely would constitute a metonymic displacement—and metonymy, we were told, has no part to play in the repressive mechanism of the cryptonym. What Abraham and Török suppress in allowing themselves to arrive at the singular cryptonym (which conveniently reinforces an implication already suggested in Freud’s original case history) is the potentiality for excess inherent in their method—the possibility that other homophonies, other narratives, are also simultaneously present in Freud’s Wolf Man text. Indeed, by removing their practice from the consulting room and basing their theory on a fifty-year-old case history, there is nothing at stake for Abraham and Török: no patient to treat or therapeutic evidence to support their claims—crucially, no way to fail. With the idea of the cryptonym, Abraham and Török present us simply with a machine for creating narratives. And in ‘Roussel and Venice’, Mathews and Perec, relating the cryptonym to the Rousselian procedure, put the method through its paces. In a parody of psychoanalytical literary criticism, ‘Roussel and Venice’ draws heavily on a fictitious psychoanalytical paper, ‘Autres images de mélancholie’, by one O. Pferdli (an anagram of de profil, i.e. sideways on). Pferdli’s work is quite clearly a summary of Abraham and Török’s ‘Mourning or Melancholia: Introjection versus Incorporation’, which appeared in the journal Nouvelle revue de psychanalyse in 1972.⁵⁷ Some side-by-side examples should demonstrate the influence: Abraham and Török

Mathews and Perec (as O. Pferdli)

But the fantasy of incorporation merely simulates profound psychic transformation through magic; it does so by implementing literally something that has only figurative meaning. So in order not to have to ‘swallow’ a loss, we fantasize swallowing (or having swallowed) that which has been lost, as if it were some kind of thing.

[I]ncorporation stresses the unique, ‘objective’ meaning of words and things, and whenever it encounters metaphorical objects, systematically de-metaphorises them. So when a thing is ‘hard to swallow’, it becomes what must be physiologically swallowed: meals and food become obsessions.

⁵⁶ One can only imagine the flights of German-English homophonic interpretation Abraham and Török would have embarked on had Pankejeff dreamt of six foxes [sechs Füchse] instead of six wolves! ⁵⁷ Nicolas Abraham and Maria Torok, ‘Introjecter-Incorporer. Deuil ou mélancolie’, Nouvelle revue de psychanalyse 6 (1972): 111–22.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

The Punning Muse: Psychoanalysis and Homophonic Translation When, in the form of imaginary or real nourishment, we ingest the love-object we miss, this means that we refuse to mourn. The effects of incorporation are extremely difficult to diagnose. [ . . . ] Incorporation often hides behind ‘normalcy’, takes flight in ‘personality traits’ or ‘perversion’, and appears openly only in delirium, in the mental state Freud called narcissistic neurosis, that is, manicdepressive psychosis. It should be remarked that as long as the crypt holds, there is no melancholia. It erupts when the walls are shaken. So when the subject learns from the analyst, in a repetition of the initial trauma, that his secret lover must be attacked, he has no choice but to push his fantasy of mourning to its ultimate conclusion: ‘If my beloved is to lose me forever, he will not survive this loss’. This certainty restores peace of mind to the subject, a picture of what recovery might look like. The cure will be complete the day when the ‘object’ makes the supreme sacrifice.⁵⁸

73

Incorporation is therefore a refusal to grieve. Except in the case of delirium or certain kinds of manic-depressive fits, the delusion of incorporation is difficult to diagnose. It is carefully concealed within roles such as ‘normality’, ‘character’, or ‘perversion’. At the melancholic stage, the inner crypt begins to crack and grief is finally accepted. It is unfortunately not felt as the subject’s grief for the lost object of love, but as the object’s grief for the subject. This is the first psychic modulation, but it is also the last, as it leads to a submission to death.⁵⁹

Bringing Pferdli in for theoretical ballast, Mathews and Perec argue that ‘Venice’ functions as the cryptonym for Roussel—the word which cannot be spoken—and this is why the city is never mentioned in his work. Already this is a mischievous premise: the list of things or places not mentioned even in an encyclopaedic corpus like that of Roussel is endless. Why does he never mention stoats, they might ask? Or carburetors? Or Doncaster? But Mathews and Perec invent an important biographical detail: that as young man, Roussel paid a visit to Venice—unrecorded in any biographical studies—and there fell in love with a boy who died shortly afterwards. This is why the city cannot be named, but why, as

⁵⁸ Nicolas Abraham and Maria Török, ‘Mourning or Melancholia: Introjection versus Incorporation’, in The Shell and the Kernel: Renewals of Psychoanalysis, trans. by Nicholas T. Rand (Chicago: University of Chicago Press, 1994), pp. 125–38, Abraham and Török’s emphasis. ⁵⁹ Harry Mathews and Georges Perec, ‘Roussel and Venice: Outline of a Melancholic Geography’, trans. by Anthony Melville and Harry Mathews, in Harry Mathews, The Case of the Persevering Maltese: Collected Essays (Normal, IL: Dalkey Archive, 2003), pp. 123–46 (pp. 127–8).

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

74

The Oulipo and Modern Thought

they systematically set out, the city and the boy are the secret meanings behind every one of his published works. Mathews and Perec’s crytogrammatical case history is quite deliberately ludicrous. Take for example the metonymic jump from Paris to Venice in the following: The ‘Paris’ equivalent of the canal is the Seine, and the lagoon then represents the Atlantic. Beyond the lagoon lie the beaches of Venice, along the islands that separate it from the Adriatic. Beyond the ‘Atlantic’ one comes to the Americas: which are, in Roussel, ‘colonized’ (Guyana), just as the beaches are summer colonies for the Venetians.

The essay even includes a wildly stylized map—with the Grand Canal distorted to look like the Mediterranean, and a trapezoid Atlantic the inverse of the Venetian lagoon—to drive the point home. Of course, the relationship being served up to us is absurd. The idea that the topographical relationship between Paris and the Atlantic, or between Paris and Guyana, is anything like that of Venice and its lagoon and beaches is utterly ridiculous. Like Umberto Eco’s model of the failed metaphor— ‘Achilles is a duck’ because both are bipeds—the relationship is comical because it has absolutely no value of uniqueness.⁶⁰ We could take anywhere that lies east, or indeed west, of somewhere else and say that these places are thus like Venice and its beaches: the problem is that there are too many alternative interpretations to justify locating particular value in this one. The essay abounds in gleefully wild analogies like this one. We hear, for example, that Venice is reminiscent of, not one, but two of the planets in the solar system: ‘not only Venus, because of its name, but also Mars, since Venice is red and full of canals.’ Mathews and Perec’s essay is at once wryly sophisticated and magnificently silly. And in properly Oulipian fashion this silliness is part of its sophistication: a spoof of interpretation given free rein, and of the tortuous reductiveness of readings like Abraham and Török’s Wolf Man essay. Considering ‘Roussel and Venice’, the psychoanalyst Rachel Rosenbaum has argued persuasively that Abraham and Török’s paper struck a deep chord with Perec.⁶¹ She claims that Perec, in analysis with Pontalis from 1971 to 1975, ‘noticed the essay on “crypts” and understood its relevance to his own situation’, and follows this up with the statement that ‘there is no doubting ⁶⁰ Umberto Eco, Interpretation and Overinterpretation (Cambridge: Cambridge University Press, 1992), p. 63. ⁶¹ However, she erroneously cites the article on which Pferdli’s fictitious work is based as Török’s ‘The Illness of Mourning and the Fantasy of the Exquisite Corpse’.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

The Punning Muse: Psychoanalysis and Homophonic Translation

75

Perec’s sincerity when he describes the breaking apart of the crypt as the “first modification” and also “the last”’.⁶² While this may be true—that Perec is in part sympathetic to Abraham and Török’s ideas—there is also no doubting that ‘Roussel and Venice’ is very much a send-up of psychoanalytical overinterpretation. At the end of the essay, the authors break cover, pulling the rug out from beneath their ruse. ‘There is ultimately no Roussel mystery,’ they state. ‘His work is not a riddle that we have to solve.’ Instead, ‘It is only our reading of it, our thirst for explanation, our love of complexity that create the impression that there must be a secret’ (138). As for the homophonic evidence, we were mistaken if we ever imagined that it was reliable. Even when we know that Roussel used the punning procédé to produce his works, we cannot conclusively retrace his steps. No longer playing the game that they set up, Mathews and Perec’s final footnote is unambiguous in its demolition of homophonic detective work. Focussing on a detail from Roussel’s novel Locus Solus, they spell out the problem: Nothing could be simpler, for instance, than to explain the ballerina’s costume in the episode of the lawyer Dargaud (Locus Solus 176) by means of a near pun on ‘Meunier, tu dors’ / ‘Manille et Tudor’: manille [anklet] referring to the hoop with which the dancer’s crinoline is made, and Tudor to the roses on her bodice; but the whole episode could just as easily be linked to the slang expression ‘Je n’ai pas un rotin’ [I haven’t a bean] which becomes ‘jeune épouse a rotin’ [young wife with hoops]: both expressions are equally valid, equally interesting, and equally useless. (145–6)

Whenever we take a narrative—as Brisset or Abraham and Török do— reconstructed by a homophonic procedure, and assign to it a privileged status, a truth above all others contained in the same set of sounds, we are making a mistake. There is never only one pun.

⁶² Rachel Rosenblum, ‘Postponing Trauma: The Dangers of Telling’, International Journal of Psychoanalysis 90.6 (2009): 1319–40 (1332–3). Perec’s friend and fellow Oulipian, Claude Burgelin, offers a detail and sensitive reading of Perec’s literary engagements with psychoanalysis—including this one and the Crypt’s other notable appearance in 53 Jours—in Les Parties de dominos chez Monsieur Lefèvre: Perec avec Freud—Perec contre Freud (Belval: Circé, 1996).

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

3 Surrealism’s Subject Two Cohorts of the Oulipo There was a first era of the Oulipo (cf. the Charbonnier interviews, the Dossiers of the Collège [de ’Pataphysique], etc.) and now a second is underway. Queneau, Letter to Le Lionnais, 28 May 1967¹

At the start of April 1973, François Le Lionnais sent a notice to all members of the Oulipo containing details of the next meeting. The letter was a fairly standard one: a reminder about the next guest of honour—in this case, the acoustician Émile Leipp—and a tear-off RSVP slip to indicate whether one would be staying for lunch. This time, however, beneath the RSVP slip was another tear-off section on which members were asked for the names and addresses of anyone who ought to receive an advance copy of La Littérature potentielle. Among the completed slips stored in the Oulipo’s archive is that of Marcel Bénabou on which he proposes one ‘J. Lacan, 5, rue de Lille, Paris 6’ (Fig. 2; in Introduction).² A generation younger, Bénabou had nevertheless been friends with Lacan since the late 1960s.³ It was Queneau, however, who had known the analyst best. Friends since the 1930s when they had both attended Alexandre Kojève’s celebrated seminar on Hegel, it was Lacan who hid Queneau’s Jewish mother- and father-in-law from the Nazis by having them admitted to a clinic in Versailles.⁴ Queneau meanwhile makes a surprising appearance in the transcripts of Lacan’s seminar when he is revealed to ¹ [‘il y a eu une première periode OULIPO (cf. entretiens Charbonnier, cahiers du Collège, etc.) et [ . . . ] il y en a une seconde en cours’]. ‘Letter from Raymond Queneau to François Le Lionnais, 28 May, 1967’ (FO, DM-2 (15)). ² ‘Coupon-réponse avec ajouts autographes de Marcel Bénabou’ (FO, DM-3 (28)). ³ See Bénabou, ‘La Galère’, p. 30. ⁴ Elisabeth Roudinesco, Jacques Lacan, trans. by Barbara Bray (New York: Columbia University Press, 1997), p. 164.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Surrealism’s Subject: Two Cohorts of the Oulipo

77

be the true author of ‘a small pornographic work, [ . . . ] one of the most delightful one could hope to read.’⁵ (The novel in question is Queneau’s On est toujours trop bon avec les femmes, which he published under the pseudonym Sally Mara.)⁶ Bénabou’s essay, ‘La Galère’, points out a few of the stylistic similarities between Queneau’s and Lacan’s writing—the taste for archaisms, foreign borrowings, and homophonic play—and suggests that, here, ‘it is not forbidden to think that the latter [Lacan] was strongly influenced by the former [Queneau]’.⁷ In this chapter, however, I wish to suggest that this influence flowed in the other direction too, and was not limited to matters of style. The experiments of the early Oulipo were underpinned by an understanding of language which was of a part with Lacan’s own work from the mid to late 1950s. Nevertheless, the prevalence of this line of thought on the activities of the Oulipo is rather harder to detect after the 1960s as changes in the group’s membership—and, more broadly, changes in the general intellectual environment—mark the decline of the structuralist moment. In spite of their predilection for manifestos, it would be a mistake to imagine a uniformity of opinion among a group as diverse and longstanding as the Oulipo. The best-known of the group’s internal disputes concerns the issue of whether or not the rules underpinning an Oulipian text should be revealed to the reader.⁸ However, in this chapter we will see a more fundamental difference of opinion, one which concerns the very value of Oulipian modes of writing—essentially, the answer to the ⁵ Lacan, Seminar, 1954–55, pp. 127–8. The seminar was held on 16 Februrary 1955, a few weeks before—and part of the same series as—his famous seminar on ‘The Purloined Letter’ of which there will be more to say later in this chapter. ⁶ See Dennis Duncan, ‘ “Joyce, un pornographe”: Ulysses, A Portrait of the Artist as a Young Man, and the Sally Mara Novels of Raymond Queneau’, James Joyce Quarterly 52.2 (2015): 115–32. ⁷ [‘il n’est pas interdit de penser que le second s’est fortement inspiré du premier’]. Bénabou, ‘La Galère’, p. 30. ⁸ Meeting minutes from 1978 show Perec arguing that they should not since this will inevitably dominate the way the text is read; Calvino, on the other hand, argues that advertising one’s constraints encourages a certain type of reader, which is no bad thing. ‘Compte rendu autographe de Paul Fournel de la réunion du 11 mai’ (FO, DM-4 (15)). The issue can be found as far back as Queneau’s radio interviews with Georges Charbonnier in 1961. Here, Queneau speaks of the writing process for his first novel, Le Chiendent (1933), and the structures which underpin that work. He says, ‘At that time, I thought that effectively it was like those constructions, those . . . what do you call them? those iron things which builders use to renovate or to put up a house, one of those . . . scaffolding, that’s it! Like scaffolding, which you remove – that’s exactly it – which you remove once the building work is finished’ [‘A ce moment-là je pensais qu’en effet c’était comme ces constructions, ces . . . comment on appelle ça? ces trucs en fer qui servent aux maçons à ravaler ou à construire une maison, un de ces . . . échafaudages, voilà! Commes des échafaudages qu’on enlève, c’est exactement ça, qu’on enlève une fois que la construction est terminée’]. Charbonnier, Entretiens, pp. 49–50.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

78

The Oulipo and Modern Thought

question: What is the point of Oulipian exercises? It is a difference of opinion which is loosely generational and roughly demarcates two cohorts of the Oulipo: the first members of the early 1960s, and the ‘second wave’ of the late 1960s and early 1970s. Alongside this, we must consider the group’s oft-stated oppositional relationship to the Surrealist movement, which frequently takes the form of a rejection of the Surrealists’ model of the unconscious as a zone which might be mined for artistic material. Instead, and in line with the prevailing intellectual culture of Paris in 1960, the Oulipo’s founders presented a concern with the form of literary texts, along with a suite of devices by which new texts might be generated procedurally from pre-existing ones. Christelle Reggiani offers a useful formulation when she states that the Oulipian project produces a ‘splitting’ [scission] of the writer figure into two distinct roles: the inventor who devises the structure and the poet who uses it to produce a text. Moreover, she adds, in certain cases the role of poet may be delegated to a machine.⁹ What I want to suggest here is that this represents not only a rejection of the Romantic conception of artistic inspiration—this much is frequently attested by Oulipians themselves— but also a deliberate and provocative programme to query the very concept of the author-subject, a literary endeavour to vacate the subject in favour of ‘language itself ’, an attitude which has a close parallel in Lacan’s thought from the same period. Yet this attack on the authorial subject is by no means echoed by all members of the Oulipo. A claim like that of Jean-Jacques Thomas, that ‘the Oulipo in theory and practice reestablishes a clear distinction between the psychological subject and the subject of language’, falls down for its lack of historical distinction.¹⁰ While the group’s founders clearly privilege the authorless text, as Le Tellier points out, ‘Oulipo is not a static cell that reproduces itself identically from generation to generation.’¹¹ Roubaud has argued that a second era of the group—the ‘Perecquian Oulipo’— began in 1969 with the publication of La Disparition.¹² He also reveals that the energy and enthusiasm which followed the group’s initial formation declined after a few years, resulting in the ‘Crisis of 66’, which could be noted ‘specifically in the waning of the taking of Minutes at the meetings’.¹³ It was this ‘crisis’ that lead to the widening of the group’s ⁹ Reggiani, Rhétoriques de la contrainte, pp. 24, 111. ¹⁰ Jean-Jacques Thomas, ‘README.DOC: On Oulipo’, trans. by Lee Hilliker, SubStance 17.2 (1988): 18–28 (21). ¹¹ Hervé Le Tellier, ‘Bookforum Interviews Hervé Le Tellier’, [accessed 8 March 2016]. ¹² Jacques Roubaud, ‘Perecquian OULIPO’, trans. by Jean-Jacques Poucel, in Yale French Studies 105 (2004): 99–109 (100). ¹³ Roubaud, ‘Perecquian OULIPO’, p. 108.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Surrealism’s Subject: Two Cohorts of the Oulipo

79

membership, beginning with himself in 1966, and taking in, among others, Perec, Bénabou, Mathews, and Calvino over the next seven years. Writers of the group’s second wave, while echoing their forebears’ contempt for artistic inspiration conceived as ‘the writer, solitary and nocturnal in his garret, waiting for the text to descend on him from who-knows-what transcendence’,¹⁴ have nevertheless stated explicitly that writing under constraint allows them to express themselves. These claims, often couched in the language of psychoanalysis, imply that the purpose of writing under constraint is not to demonstrate the absence of an author-subject, but to create the very possibility of its invocation. Nevertheless, the aim of this chapter is merely to sketch a trend, rather than to imply a rigorously-observed ideological schism. As we shall see in Chapter Five, Mathews plays on the way that the translator-subject can be squeezed out of business by a language which contains only one phrase, while all members of the group routinely use the type of procedural practice (e.g. S+7) in which the psychological subject is banished from the text; meanwhile, the Oulipo’s foundational text, Queneau’s Cent mille milliards de poèmes, is at once both an experiment in literary impersonality and a work as deeply Queneaunian as any of its creator’s pre-Oulipian work.¹⁵ The distinction I wish to draw, therefore, constitutes one of the dreaded generalizations which routinely raise the ire of the group’s members. My intention, however, is only to provide a more nuanced, historicized analysis than the some of the universalizing claims made for the group. I hope that—rather like the Wittgensteinian categories which occupy Perec and Mathews in their South Seas tales (see Chapter Five): slightly fuzzier than one might wish, but still indispensable—a little qualified generalization will provide a useful picture of a key stage in the group’s intellectual development.

SURREALISM’S INFLUENCE To point out that the Oulipo, from the start, were hostile to Surrealism is to say very little of note. It is a commonplace in the group’s own historiography to affirm and reaffirm this hostility explicitly, as Hervé Le Tellier does, for example, when he takes Gérard Genette to task for lazily ¹⁴ [‘l’écrivain solitaire et nocturne dans sa mansarde, attandant que le texte lui descende d’on ne sait quelle transcendance’]. Paul Fournel, ‘Les ateliers de l’Oulipo: écrire ici et maintenant’, Magazine Littéraire (May 2001): 26. ¹⁵ See, for example, the Charbonnier interviews in which Queneau describes the Cent mille millards de poèmes as ‘the first concrete manifestation of this research group’ [‘la première manifestation concrète de ce Group de recherches’]. Charbonnier, Entretiens, p. 116.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

80

The Oulipo and Modern Thought

equating Oulipian experimentation with Surrealism.¹⁶ What is sometimes lacking, however, is a consideration of the negative influence which Surrealism exerted over the early Oulipo, both in the way that the group was arranged—the manner in which it conducted its activities—and what those activities consisted of. It is too easy to disavow the role of Surrealism altogether without exploring the ways in which that movement offered a model, both intellectual and social, which the Oulipo could consciously oppose, and, in doing so, define itself with clarity. As Bloomfield has pointed out, along with the structuralists, it is the Surrealists against whom the early Oulipo are most anxious to position themselves.¹⁷ To those who would suggest that the collaborative and social nature of this avant-garde literary movement—not to mention the presence of former Surrealists like Queneau and ‘foreign correspondent’ Simon Watson Taylor—makes the Surrealists an undeniable antecedent, the standard Oulipian response is to point instead to the Bourbaki group as the true model. Bourbaki, or more properly the Association des Collaborateurs de Nicolas Bourbaki, were a clandestine group of mathematicians, formed at the École Normale Supérieure in Paris in 1934 and who published under the collective pseudonym ‘Nicolas Bourbaki’.¹⁸ Citing Bourbaki’s influence, Roubaud writes that ‘[t]here can be no doubt that that, when they founded the Oulipo, Queneau and Le Lionnais [ . . . ] had this group in mind’. He adds, with emphasis, ‘When the Oulipo was conceived, Bourbaki provided a counter-model to the Surrealist group.’¹⁹ In this formulation, perhaps unintentionally, there is something more reasonable than the usual blanket disavowals of Surrealism: of course Surrealism will cast its shadow over any literary collective in which Queneau and Arnaud are involved from the start; what is important, however, is that, being aware of this, the founders sought deliberately to mitigate Surrealism’s influence by looking for an alternative model. Nevertheless, we should be conscious that there is something slightly partisan about the Oulipo’s pledges of allegiance to Bourbaki, and that, as often as not, these gestures include a thumbed nose directed at Surrealism. After all, in one fundamental way, ¹⁶ Hervé Le Tellier, Esthétique de l’Oulipo (Bordeaux: Castor Astral, 2006), pp. 28–9. ¹⁷ Bloomfield, Raconter l’Oulipo, p. 259. ¹⁸ Le Lionnais, in his Grands courants de la pensée mathématique, has a lovely epithet for Bourbaki, introducing his as ‘ce mathématicien polycéphale’: ‘that many-headed mathematician’. [François Le Lionnais (ed.), Les Grands courants de la pensée mathématique (Marseilles: Cahiers du Sud, 1948), p. 22]. ¹⁹ Jacques Roubaud, ‘The Oulipo and Combinatorial Art’, trans. by Harry Mathews, in Mathews and Brotchie, Oulipo Compendium, pp. 37–44 (p. 38), Roubaud’s emphasis. This is despite Roubaud elsewhere calling Bourbaki ‘a sort of mathematical surrealism’. Jacques Roubaud, ‘Mathematics in the Method of Raymond Queneau’, in Motte, Oulipo, pp. 79–96 (p. 80).

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Surrealism’s Subject: Two Cohorts of the Oulipo

81

the Oulipo does not follow Bourbaki’s example, since in Bourbaki’s publications its members remained anonymous, protected by a collective pseudonym. In the case of the Oulipo—aside from a fleeting suggestion by Braffort for ‘a collective article, signed Joseph Oulipo’—this is not the case.²⁰ Still, Bourbaki and the Oulipo would occasionally overlap. An abandoned draft of the introduction to the 1951 edition of Bourbaki’s Théorie des ensembles illustrates the way that mathematicians work, by reference to one ‘Mr R. Queneau, [who] in his Exercises in Style, has sought to give multiple tellings of the same incident; as for the mathematician, he learns to give the same telling to multiple incidents!’²¹ And Queneau would return the favour a decade later by attending one of Bourbaki’s congresses in 1962 and devoting a chapter to the group in the 1963 work Bords: Mathématiciens, précurseurs, encyclopédistes.²² In a telling comparison, the mathematical historian David Aubin has noted that ‘both [Bourbaki and the Oulipo] looked at the formal bases of their respective disciplines and wished to rewrite their histories from the current structural perspective’, an observation which is hard to refute but which nevertheless might have made the anti-structuralist Latis shudder.²³ Meanwhile the two groups had as much in common in terms of their culture as their objectives, with Aubin pointing to ‘a common insistence on axiomatics, formal beauty versus future utility, and especially humor’. In spite of this shared geniality, however, Bourbaki had one rule which simply could not be implemented by the Oulipo: the expectation that members should resign by the age of fifty (Queneau and Le Lionnais were fifty-seven and fifty-eight respectively by the time of the colloquium at Cerisy).²⁴ So while expulsion and defection were commonplaces among ²⁰ [‘un article signé Joseph Oulipo’]. ‘Compte rendu dactylographié de Jacques Duchateau de la réunion n° 56 du 19 décembre 1964’ (FO, DM-1 (51)). There is, however, a serialized piece of fiction signed by a mysterious Lou l’Hippo (i.e. l’Oulipo) which appears in one of the private sheets circulated by the Collège de ’Pataphysique in the early 1960s. [Lou l’Hippo, ‘L’Amer Dataire’, Le Petit Moniteur de l’ACACADOOR 1 (1961): 6.] ²¹ [‘M. R. Queneau, dans ses “Exercices de style”, s’est attaché à donner de multiples récits d’un même incident; le mathématicien apprend, lui, à donner un même récit de multiples incidents!’]. ‘Livre I. Théorie des ensembles. Introduction (Chevalley)’ (p. 20), Archives Bourbaki, [accessed August 25, 2016]. ²² Raymond Queneau, Bords: Mathématiciens, précurseurs, encyclopédistes (Paris: Hermann, 1963). ²³ David Aubin, ‘The Withering Immortality of Nicolas Bourbaki’, Science in Context 10.2 (1997), 297–342 (323). ²⁴ Aubin suggests that this rule, if it even existed, was more honoured in the breach than the observance: ‘The historian Liliane Beaulieu, who has worked the most extensively on Bourbaki, told me that she had never come across any written trace of this rule and that in

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

82

The Oulipo and Modern Thought

the Surrealists, and Bourbaki adopted a type of voluntary Logan’s Run system, the Oulipo’s approach to membership was without precedent: No one can be expelled from the Oulipo. Conversely (you can’t have something for nothing), no one can resign from the Oulipo or stop belonging to it. It follows that, once a member of the Oulipo, always a member. This has particular implications: The dead continue to belong to the Oulipo. [ . . . ] Lest the last rule seem unduly coercive, an exemption [ . . . ] was provided. One may relinquish membership of the Oulipo under the following circumstances: suicide may be committed in the presence of an officer of the court, who then ascertains that, according to the Oulipian’s explicit last wishes, his suicide was intended to release him from the Oulipo and restore his freedom of manœuvre for the rest of eternity.²⁵

In fact, so extreme is this approach to membership—a convivial collegiality which strays, comically, into oppressive officiousness—that it is difficult not to read it as a deliberately ironic contrast to the group’s antecedents, and to Surrealism in particular. Le Lionnais suggests as much when he talks about how he first proposed the idea of the Oulipo to Queneau: ‘He would never have approved this project had we not been viscerally in agreement about radically warding off any group activity that might engender fulminations, excommunications, or any other form of terror.’²⁶ Added to this, suggests Le Tellier, was a sense that Surrealism became a personality cult centred around the figure of Breton, and a resistance on Queneau’s part to occupying a similar role within the Oulipo.²⁷

any case it was breached many times’ (p. 298 n. 3). Nevertheless, the rumour was certainly known—and believed—by Queneau at around the time of the Oulipo’s formation since, in an article on the group, he writes, ‘He must have grown old, your fictive mathematician, he must have fallen behind. Well, no, Bourbaki has not grown old because he cannot grow old’ [Il a nécessairement vieilli, votre fictif mathématicien, il doit avoir pris de retard. Eh bien! non, Bourbaki n’a pas vieilli parce qu’il ne peut pas vieillir’]. ‘Bourbaki et les mathématiques de demain’, in Bords: Mathématiciens, précurseurs, encyclopédistes (Paris: Hermann, 1963), pp. 11–29 (p. 16), Queneau’s emphasis. ²⁵ Roubaud, ‘Combinatorial Art’, p. 38. ²⁶ François Le Lionnais, ‘Raymond Queneau and the Amalgam of Mathematics and Literature’, in Motte, Oulipo, pp. 74–8 (p. 77). ²⁷ ‘Je pense que si le groupe n’avait pas été secrete, il serait apparu comme étant plusiers jeunes gens autour de Raymond Queneau. Et ce n’était pas du tout sa volonté. Il disait je ne veux pas être le nouveau Breton.’ [‘I think that if the group had not been secret, it would have seemed like a bunch of young people gathered around Raymond Queneau. And that was not at all what he wanted. He said, I don’t want to be the new Breton’.] (FO, MS-6 (‘Page’)).

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Surrealism’s Subject: Two Cohorts of the Oulipo

83

So the group, in its structure, is a conscious antithesis of Surrealism. And this is not without good reason. Beneath the surface of Le Lionnais’s comments about the potential fulminations, excommunications, and terror of group activity we should be able to discern Queneau’s own painful history amidst the fractiousness of Surrealism. It had been Queneau, for example, who had drafted the conciliatory invitation to the Surrealists’ meeting at the Bar du Château in March 1929. Sent to both active and dissident members, the invitation urged all parties to put their differences behind them and forge a plan for collective action. The meeting, however, was a disaster, hijacked by Breton as a platform for further expulsions.²⁸ Within a few months, Queneau would leave the group after clashing with Breton himself. Years later, recalling this period of his life, he would speak of developing ‘a violent reaction, a passionate hatred’ for Surrealism.²⁹ Although Queneau would insist that his reasons for leaving were strictly personal rather than ideological, one does not need to look far to find that the Oulipo’s animus towards Surrealism is an intellectual and theoretical one whose traces can be found not just in the genial structure of the group but also in its activities.³⁰ Indeed, members have argued that Queneau’s frustration did in fact go beyond the purely personal. Paul Fournel, for example, has claimed that ‘[Queneau’s] time in the Surrealist group had left him with an impression of malaise and theoretical dissatisfaction’.³¹ Certainly, there is much to support this line of reasoning, since Queneau would frequently make public denouncements of the artistic methods of the Surrealists, and specifically their conception of the relationship between psychoanalysis and creativity. In February 1938, for example, nine years after leaving the Surrealists and six years into his own analysis with Fanny Lowtsky, Queneau wrote a piece for the journal Volontés entitled ‘What is Art?’. Here he considers the vogue for experimental poetry seen as a means of bringing back reports from the unconscious: ‘the poet who pretends to “dive” into his unconscious to bring ²⁸ See Dennis Duncan (ed.), Theory of the Great Game: Writings from Le Grand Jeu (London: Atlas, 2015), pp. 8–11. ²⁹ Queneau, ‘Conversation avec Georges Ribemont-Dessaignes’, in Bâtons, pp. 28–38 (p. 30) [‘une réaction violente, une détestation passionnée’]. Besides, Queneau’s biographer notes that by the late 1940s—long before the Oulipo were formed—Queneau’s hostility to Surrealism had cooled, and he and Breton had put aside their differences [‘le temps avait passé et leur querelle personnelle s’estompait, comme ses griefs à l’encontre du surréalisme’], Lécureur, Raymond Queneau, p. 331. ³⁰ [‘pour des raisons strictement personnelles et non pour des raisons idéologiques’]. Queneau, ‘Conversation’, p. 29. ³¹ [‘Son passage dans le groupe surréaliste lui avait laissé une impression de malaise et d’insatisfaction théoriques’]. Fournel, Clefs, p. 26.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

84

The Oulipo and Modern Thought

up the marvels and new worlds announced by Appolinaire is not an experimenter but an empiricist. He waits open-mouthed for inspiration like an entomologist hoping to catch an insect.’³² The scorn here is obvious: the automatists are charlatans, pretending to tune into their unconscious as if it were a radio broadcast, an approach with little validity in Freudian theory. As David Lomas puts it, ‘the doctrine of “pure” psychic automatism holds out the promise of a certain plenitude which is foreign to a Freudian conception of the unconscious.’³³ Queneau’s scepticism about Surrealist automatic writing is important, since—as we shall see—it is a mode of writing which he will later contrast, publicly and ironically, with the early experiments of the Oulipo. It is also important to note that Queneau’s critique in the Volontés article is twopronged: not only is automatic writing of dubious theoretical validity, but even if it were valid, it is banal. The automatist poets are not artists, creating original works by skill and craft; they are empiricists, merely documenting their psyches in slack-jawed passivity—a far cry, says Fournel, from the image they intended to present of themselves: ‘André Breton himself cultivated his image as romantic and “inspired”, offering us a perfectly acceptable model of the visionary poet with a star on his forehead, the shining beacon’.³⁴ The inspired poet, here with its tell-tale scare quotes, is one of the Oulipo’s bêtes noires, the model of Surrealist selfdelusion. In Queneau’s novel Odile (1937), essentially a roman à clef attack on the Surrealists,³⁵ he writes ‘The truly inspired person never becomes inspired: he ³² [‘le poète qui prétend “plonger” dans son inconscient pour en retirer les merveilles et les mondes nouveaux annoncés par Appolinaire n’est pas un expérimentateur, mais un empirique. Il attend bouche bée l’inspiration comme l’entomologue l’insecte qui’il veut capturer.’]. Raymond Queneau, ‘Qu’est-ce que l’art?’, in Le Voyage en Grèce (Paris: Gallimard, 1973), pp. 89–96 (pp. 92–3). ³³ David Lomas, The Haunted Self: Surrealism, Psychoanalysis, Subjectivity (New Haven, CT: Yale University Press, 2000), p. 9. It is hard too to forget Freud’s comment in a letter to Stefan Zweig the day after meeting Dalí: ‘I have been inclined to regard the surrealists, who apparently have adopted me as their patron saint, as complete fools’ [Ernest Jones, The Life and Work of Sigmund Freud, ed. by Lionel Trilling and Steven Marcus (London: Hogarth, 1961), p. 523]. Adorno too argues that ‘[i]t is not the unconscious in itself that comes to light in the world-rubble of Surrealism’. He suggests that even the Surrealists would eventually come to realize this: ‘the Surrealists themselves have discovered that people do not free associate the way they, the Surrealists, write, even in psychoanalysis’. See, ‘Looking Back on Surrealism’, in Notes to Literature, ed. by Rolf Tiedemann, trans. by Shierry Weber Nicholsen, 2 vols (New York: Columbia University Press, 1991), I, pp. 86–90 (p. 87). ³⁴ [‘André Breton lui-même soignait son allure romantique et “inspirée”, offrant un modèle parfaitement acceptable du poète visionnaire au front étoilé, du Phare’]. Fournel, Clefs, p. 26. ³⁵ See Noël Arnaud, ‘Politique et polémique dans les romans de Raymond Queneau’, in Queneau aujourd’hui: actes du Colloque Raymond Queneau, Université de Limoges, mars 1984 (Paris: Clancier-Guénaud, 1985), pp. 113–57 (pp. 121–4).

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Surrealism’s Subject: Two Cohorts of the Oulipo

85

is always inspired.’³⁶ Essentially this is the same message as the Volontés article the following year, that inspiration is not something which strikes at random, and which the artist waits for passively, hopefully; instead, the writer—properly trained in the craft of writing—should always be able to write. It is this argument which Lescure will propose as a fundamental motivating dictum for the Oulipian project: is it possible that [inspiration] might cease to be capricious, and that any and everybody might find it faithful and compliant to his desires? The serious revolution, the sudden change [Queneau’s] simple sentence introduced into a conception of literature still wholly dominated by romantic effusions and the exaltation of subjectivity, has never been fully analysed. In fact, this sentence implied the revolutionary conception of the objectivity of literature, and from that time forward opened the latter to all possible modes of manipulation.³⁷

What is especially notable in this passage is that Lescure should consider that literature at the time of Odile was ‘still wholly dominated’ by ‘the exaltation of subjectivity’. Again, we can see here another barb aimed at the Surrealists, a flat rejection that automatic writing might be capable of—or is even seriously interested in—overthrowing the authorial subject in favour of language in the way that the early exercises of the Oulipo seek to. As for the idea that inspiration is primarily a matter of industry, this tenet becomes enshrined in the second Oulipo Manifesto in Le Lionnais’s aphoristic rephrasing: ‘Poetry is a simple art, all in the doing’.³⁸ Returning to the Volontés piece, Queneau ultimately calls the Surrealists’ political ideology into question by querying the relationship between chance and freedom: Another false idea which has also gained currency is the equivalence which is established between inspiration, exploration of the subconscious and liberation, between chance, automatism and freedom. Now, this inspiration which consists in blindly obeying every impulse is in reality a slavery. The

³⁶ [‘Le véritable inspiré n’est jamais inspiré: il l’est toujours’]. Raymond Queneau, Odile (Paris: Gallimard, 1950 [1937]), p. 190, Queneau’s emphasis. For more on that novel’s disguised representation of the Surrealist milieu, see Arnaud, ‘Politique et polémique’. ³⁷ Lescure, ‘Brief History’, pp. 34–5. Lescure first presented this at the Cérisy conference on artificial intelligence in July 1963, before publishing it the following year in the Oulipo issue of Temps mêlés [Jean Lescure, ‘Petite histoire pour un tri-centenaire’, Temps mêlés, 66–67 (1964): 3–12]. The version that appears in La Lipo, and which Motte translates, includes additional material relating to Roubaud, Perec, and the other new joiners. ³⁸ [‘La poésie est un art simple et tout d’exécution’]. François Le Lionnais, ‘Le Second manifeste’, in Oulipo, La littérature poentielle (Paris: Gallimard, 1973), pp. 19–23 (p. 19). For a detailed look at the concept of inspiration in the Oulipo, see Chris Andrews, ‘Inspiration and the Oulipo’, Studies in 20th & 21st Century Literature 29.1 (2005): 9–28.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

86

The Oulipo and Modern Thought

classical writer who writes his tragedy observing a certain number of rules that he is familiar with is freer than the poet who writes whatever comes into his head and who is a slave to other rules which he ignores.³⁹

Queneau attacks the Surrealists’ fundamental self-perception as a revolutionary movement by challenging their conception of freedom. This critique is echoed by Roubaud when he considers the founding basis of the Oulipo: only mathematics could offer a way out between a nostalgic obstinacy with worn-out modes of expression and an intellectually pathetic belief in ‘total freedom’. It was a matter, at least at the start, of asserting a theoretical antiSurrealism.⁴⁰

The ‘total freedom’ which Surrealist automatism promised is an ‘intellectually pathetic idea’, and the Oulipo, from the start, was assertively anti-Surrealist—not just in its attempt to avoid fulminations and excommunications, but theoretically so. (Roubaud’s telling phrase ‘at least at the start’, suggests an ideological change of direction which this chapter will trace later.) When Hervé Le Tellier writes bluntly that ‘the Oulipo could define itself as the greatest enemy of the Surrealist approach’, it is not the two groups’ respective membership policies that he has in mind.⁴¹ Let us find an example, then, of this theoretical anti-Surrealism. One of the earliest Oulipian procedures is the method known as S+7. First demonstrated by Jean Lescure at the monthly meeting of February 1961, it involves taking an existing text and replacing every noun (in French, sustantif, hence the S in the procedure’s name) with the seventh noun following it in a given dictionary. Presented as ‘un traitement analytique des texts’—an analytical treatment—it is a technique in which human agency is expunged almost entirely: ‘the operator is reduced to a purely mechanical function.’⁴² (Not quite entirely: Queneau remarks wryly that ‘[t]he result obviously depends on the dictionary one chooses.’)⁴³ Of course, the number seven is arbitrary and the procedure

³⁹ [‘Une autre bien fausse idée qui a également cours actuellement, c’est l’équivalence que l’on établit entre inspiration, exploration du subconscient et libération, entre hasard, automatisme et liberté. Or, cette inspiration qui consiste à obéir aveuglément à toute impulsion est en réalité un esclavage. Le classique qui écrit sa tragédie en observant un certain nombre de règles qu’il connaît est plus libre que le poète qui écrit ce qui lui passe par la tête et qui est l’esclave d’autres règles qu’il ignore’]. Queneau, ‘Qu’est-ce que l’art?’, p. 94. ⁴⁰ Roubaud, ‘Combinatorial Art’, p. 40. ⁴¹ [‘l’Oulipo pourrait se définer comme le plus grand ennemi de la démarche surréaliste’]. Le Tellier, Esthétique, p. 25. ⁴² [‘l’opérateur [est] réduit à une fonction purement méchanique’]. Jean Lescure, ‘La Méthode S + 7’, Dossiers du Collège de ’Pataphysique 17 (i.e. ‘Dossier 17’) (1961): 17–19 (17). ⁴³ Raymond Queneau, ‘Potential Literature’, in Motte, Oulipo, pp. 51–64 (p. 61).

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Surrealism’s Subject: Two Cohorts of the Oulipo

87

might equally be applied to any part of speech, but from an early stage it has become something of a convention to use nouns and the number seven. Here is an example by Harry Mathews which takes Wordsworth’s ‘Daffodils’ and the Langenscheidt Standard English-German Dictionary as its sources:⁴⁴ I wandered lonely as a crowd That floats on high o’er valves and ills When all at once I saw a shroud, A hound, of golden imbeciles; Beside the lamp, beneath the bees, Fluttering and dancing in the cheese. Continuous as the starts that shine And twinkle on the milky whey, The stretched in never-ending nine Along the markdown of a day: Ten thrillers saw I at a lance, Tossing their healths in sprightly glance. The wealths beside them danced; but they Out-did the sparkling wealths in key: A poker could not but be gay, In such a jocund constancy: I gazed – and gazed – but little thought What weave to me the shred had brought: For oft, when on my count I lie In vacant or in pensive nude, They flash upon that inward fly Which is the block of turpitude; And then my heat with plenty fills And dances with the imbeciles.⁴⁵

Following its unveiling to the group S+7 would quickly become something of a flagship procedure for the Oulipo. It is given more space than any other in the Collège de ’Pataphysique’s Oulipo Dossier, for example, where, as well as Lescure’s own elucidation and illustration of the technique, Queneau contributes a further pair of exercices de style which employ S+7, and the pataphysician César Ogliastro applies the method to produce a lengthy kabbalistic interpretation of the opening verses of

⁴⁴ In this case, rather than overlook the constraints of metre and rhyme already existing in the source text, Mathews has taken these into account too by ‘starting with the seventh noun listed in the chosen dictionary and continuing, if necessary, until a suitable replacement [was] found’. Mathews and Brotchie, Oulipo Compendium, p. 199. ⁴⁵ Mathews and Brotchie, Oulipo Compendium, p. 199.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

88

The Oulipo and Modern Thought

Genesis. Meanwhile, in the radio interviews with Georges Charbonnier, Queneau would give an outline of S+7 before adding the following qualification: the appearance is Surrealist, perhaps, but the method is not, which is quite important. Count 1, 2, 3, 4, 5, 6, 7, in a dictionary, then change the nouns, it is an automatism, but it is an automatism which has absolutely nothing to do with what we call Surrealist automatic writing. If the result is Surrealist, or at least seemingly Surrealist, this is simply an appearance; they have nothing to do with each other: the two automatisms are of quite different orders.⁴⁶

The distinction being drawn here is an extremely telling one: an explicit reference to Surrealism and to Surrealist automatic writing, along with the repeated insistence that the Oulipian method has ‘absolutely nothing to do with it’. There is a rather wonderful and subtle critique of Surrealism going on here, and one which embodies something of a philosophical mission statement for a group which would regularly define itself as Surrealism’s antithesis. In S+7, states Queneau, the Oulipo are presenting a new kind of automatic writing: rule-governed, rather than rule-free, but nonetheless markedly simpler to implement than its predecessor. Here, for comparison, are the instructions for automatic writing that Breton supplies in the first Surrealist Manifesto: After you have settled yourself in a place as favourable as possible to the concentration of your mind upon itself, have writing materials brought to you. Put yourself in as passive, or receptive, a state of mind as you can. Forget about your genius, your talents, and the talents of everyone else. Keep reminding yourself that literature is one of the saddest roads that leads to everything. Write quickly, without any preconceived subject, fast enough so that you will not remember what you’re writing and be tempted to reread what you have written.⁴⁷

Hardly a matter of simply looking words up in a dictionary. S+7 also has the considerable advantage of being painless. Breton, lest we forget, entertains the idea that excessive hunger might be conducive to automatism, writing that ‘Knut Hamsun ascribes this sort of revelation to ⁴⁶ [‘l’apparence est surréaliste, peut-être, mais la méthode ne l’est pas, ce qui est tout à fait important. Compter 1, 2, 3, 4, 5, 6, 7, dans un dictionairre, puis changer les substantifs, c’est un automatisme, mais c’est un automatisme qui n’a absolument rien à voir avec ce qu’on appelait l’écriture automatique surréaliste. Si le résultat est surréaliste, ou tout au moins d’apparence surréaliste, cela reste une apparance; cela n’a rien à voir: les deux automatismes sont d’ordres tout à fait différents’]. Charbonnier, Entretiens, pp. 147–8. ⁴⁷ André Breton, ‘Manifesto of Surrealism (1924)’, in Manifestoes of Surrealism, trans. by Richard Seaver and Helen R. Lane (Ann Arbor, MI: University of Michigan Press, 1972), pp. 1–47 (pp. 29–30).

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Surrealism’s Subject: Two Cohorts of the Oulipo

89

which I had been subjected as deriving from hunger, and he may not be wrong. (The fact is I did not eat every day during that period of my life.)’⁴⁸ He then goes on to note thoughtfully that ‘Apollinaire asserted that Chirico’s first paintings were done under the influence of cenesthesic disorders (migraines, colics, etc.)’. By contrast, in the new automatic writing one simply has to ‘count 1, 2, 3, 4, 5, 6, 7, in a dictionary, then change the nouns’. If, as Queneau slyly concedes, the results are ostensibly the same as before—‘Surrealist, or at least seemingly Surrealist’—then it would seem that the suffering artist has been killed off—or, better, cured—at a stroke. But Queneau’s intention, of course, is not simply to spare would-be Surrealists the pain of their creative process. In suggesting that an S+7 poem might be indistinguishable from a Surrealist one, Queneau ironically gives the lie to the Surrealists’ investment in the enduring image of suffering artist as the kitemark of artistic authenticity. It is a self-regard, an attachment to the author-subject, over and above the work, which the Surrealists were unwilling to sever. In reducing the author to an operator— and in showing that this can be easily done while still producing ‘seemingly Surrealist’ work—Oulipian automatism calls the bluff of Breton’s self-imposed subjugation to language (‘After you, my fair language’) by truly vacating the subject.⁴⁹ The S+7 method, in Queneau’s wry, contrastive description, suggests that the Surrealists were still in hock to a Romantic idea of the artist as creative genius, and to the prioritization of the artistic subject over pure language. The Oulipo, then, in its first era, is seen as an antidote to this: rigorous, objective, analytical.

THE PSYCHOANALYTIC MODEL OF THE EARLY OULIPO If Queneau’s discussion of S+7 hints at the critique of the Surrealists’ position with regard to psychoanalysis, then it is in the description of another early technique that an alternative Oulipian position starts to become clear. At the monthly meeting of January 1961, Lescure introduced a suite of textual procedures, not dissimilar to S+7, which involve taking a source text and swapping its words around, so the first noun might change place with the second, the third with the fourth, and so on. In the following example, Arthur Symons’s poem ‘Rain on the Down’ is ⁴⁸ Breton, ‘Manifesto’, p. 22n, Breton’s emphasis. ⁴⁹ André Breton, ‘Introduction to the Discourse on the Paucity of Reality’, trans. by Richard Sieburth and Jennifer Gordon, October 69 (1994): 133–44 (141).

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

90

The Oulipo and Modern Thought

subjected to what Lescure calls a Roussellian Permutation—after the nested bracketing structure of Raymond Roussel’s Nouvelles impressions d’Afrique (1932)—in which the first noun is transposed with the last, the second with the penultimate, and so on: Original text Night, and the down by the seam And the veil of rain by the down; And she came through the mist and the rain to me From the safe warm lights of the town.

Treated text Rain, and the rain by night, And the rain of her face in her hair; And she came through the rain and the town to me From the safe warm lights of the rain.

The rain shone in her hair, And her face gleamed in the rain; And only the night and the rain were there As she came to me out of the rain.

The mist shone in her down, And the rain gleamed in her veil; And only the sea and the down were there As she came to me out of the night.⁵⁰

As we can see, texts which are produced by this type of procedure, or indeed by S+7, convey a different meaning to their source, and yet this meaning cannot be attributed purely to the subjectivity of an author. Where might we attribute this agency then? Introducing his suite of permutations, Lescure writes, We have come to think that language wants to speak. One might also think that it wants to be silent. But one might also think that it wants to speak. One might think that speech wants to speak. That there is in language a power to speak. In language itself; not the person using it. Perhaps one might think that I want to speak. But it doesn’t interest me to think that. Nor do even I want it. I mean that it isn’t interesting to know why and how I want to speak. All that is just Romanticism, psychology, bricolage. What interests me is that language itself, by itself, has something to say.⁵¹

⁵⁰ Mathews and Brotchie, Oulipo Compendium, p. 168. ⁵¹ [‘Il nous est arrivé de penser que le langage veut parler. On peut aussi penser qu’il veut se taire. Mais on peut aussi penser qu’il veut parler. On peut penser que la parole veut parler. Qu’il y a dans le langage un pouvoir de parler. Dans le langage lui-même; pas dans celui qui s’en sert. Moi, peut-être bien qu’on peut penser que je veux parler. Mais cela ne m’intéresse pas qu’on le pense. Ni même que je le veuille. Je veux dire que ça n’est pas intéressant de savoir pourquoi et comment je veux parler. Romantisme tout ça, psychologie, bricolage. Ce qui m’intéresse c’est que le langage lui-même, par lui-même, veut dire quelque chose’]. Jean Lescure, ‘Des permuations en particulier et en général des poèmes carrés’, in Oulipo, La Littérature potentielle, pp. 151–61 (p. 151). In an excellent discussion of S+7, Alison James says much the same thing about that technique: ‘[S + 7] unveils the “mechanical” aspect of language itself – that is, language’s capacity to produce meaning independently of human intervention’. Alison James, Contraining Chance: Georges Perec and the Oulipo (Evanston, IL: Northwestern University Press, 2009), p. 128.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Surrealism’s Subject: Two Cohorts of the Oulipo

91

The passage quoted here is from the group’s 1973 collection, La Littérature potentielle, but does not appear in the 1961 version of the same text in the Collège de ’Pataphysique’s Oulipo Dossier. Lescure’s language certainly has echoes of other literature of the late 60s. Foucault, for example, writing in 1970 about J.-P. Brisset (see Chapter Two), speaks of ‘language itself in a state of play’.⁵² And the idea that language has ‘a power of speaking’, along with Lescure’s stressed clarification that he means ‘language itself, not the person using it’, is also strongly reminiscent of Barthes’s comment on Mallarmé from ‘The Death of the Author’ (1968): ‘For him, for us too, it is language which speaks, not the author.’⁵³ Similarly, Lescure’s attack on the authorial subject—‘Romantisme, psychologie, bricolage’—finds stronger expression in the withering sarcasm of Barthes’s rhetorical question, ‘Did he wish to express himself ?’, to which Barthes adds that the author ‘ought at least to know that the inner “thing” he thinks to “translate” is itself only a ready-formed dictionary, its words only explainable by other words, and so on indefinitely’.⁵⁴ The terms of Barthes’s elaboration—the ‘inner “thing”’ and the analogy with translation, both placed in opposition to the image of the ‘ready-formed dictionary’: words which give place only onto other words—point to the fact that this essay, and Lescure’s too, is written under the influence of an earlier work: Jacques Lacan’s celebrated 1957 paper, ‘The Instance of the Letter in the Unconscious’. Just as Surrealism offered a counter-model for the Oulipo to define themselves against, so Lacan’s essay draws much of its innovative power from the break it makes from an earlier model of psychoanalysis. A sense of what this other model was will help us to understand Lacan, and, in turn, to understand the context of Lescure’s puckish statements about language itself speaking through his textual permutations. Three years after the initial appearance of Lacan’s ‘Instance of the Letter’, one of his protégés, Jean Laplanche, would deliver an influential paper of his own which gives the clearest elucidation of the two opposing models of the unconscious. Laplanche describes ‘two very different types of listening [ . . . ]: “the attitude of simultaneous translation” and “the attitude of attention to lacunary phenomena.”’⁵⁵ The first of these he

⁵² Foucault, ‘Seven Remarks’, p. 1254. ⁵³ Barthes, ‘Death of the Author’, p. 50. ⁵⁴ At the monthly meeting of August 1970, a discussion of Barthes’s S/Z takes place. Le Lionnais’s annotations on his agenda for the meeting show that Barthes is then mooted as a future invitee. ‘Pré-programme dactylographié du congrès d’août 1970 avec ajouts autographes de François Le Lionnais’ (FO, DM-2 (54)). ⁵⁵ Jean Laplanche and Serge Leclaire, ‘The Unconscious: A Psychoanalytic Study’, trans. by Patrick Coleman, Yale French Studies 48 (1972): 118–75 (124). The first, second, and fourth sections of this article, including the material under discussion here, are by Laplanche.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

92

The Oulipo and Modern Thought

attributes, in the French psychoanalytic tradition at least, to the early twentieth-century philosopher Georges Politzer. Here the relationship of the unconscious to the gestures, symbols, or words of the patient under analysis is a straightforward, singular one which the analyst tries to discover, or translate. Another of Lacan’s expositors, Anika Lemaire, offers this useful summary of Politzer’s approach: the unconscious is [ . . . ] the meaning of everything the subject says. In the dream, for example [ . . . ] there is only one ‘content’. The manifest is a story in an unconventional language; the unconscious is its meaning, but its meaning has no existence in itself. Translated into the manifest, the unconscious thoughts no longer exist in themselves and, conversely, the manifest does not exist in itself when it is returned to the unconscious.⁵⁶

The analogy with translation, seized on by both Lemaire and Laplanche, comes directly from Politzer. He uses it explicitly as a way of explaining the famous ‘Irma’s injection’ episode in Freud’s Interpretation of Dreams: In [Freud’s] dream of the injection given to Irma, ‘Irma has a sore throat’ means ‘I wish [there to be] an error of diagnosis’. [ . . . ] The desire of the error of diagnosis explains, then, the sore throat, as the Latin term pater explains the French term père, or rather, as jealousy explains Othello’s gesture.⁵⁷

In Politzer’s view then, latent and manifest psychoanalytic content have the same simple relationship as pater and père: words in different languages that indicate the same thing. The two are effectively interchangeable, with the corollary that ‘meaning has no existence in itself ’—the latent has no remainder, nothing that does not, by definition, make itself manifest. Latent and manifest, in Laplanche’s rather nice but heavily sarcastic description, are like characters in a farce, ‘always leaving the stage when the other comes on’.⁵⁸ This is a model of the unconscious based entirely around the psychological subject, whose inner ‘thing’ is always translated into an outer one, a ‘first person drama’ which clearly leaves no room for Lescure’s text in which ‘language itself, by itself, has something to say’. Lacan’s ‘Instance of the Letter’, then, sets out a critique of this model, based on his reading of structuralist linguistics.⁵⁹ In an uncharacteristically ⁵⁶ Anika Lemaire, Jacques Lacan, trans. by David Macey (London: Routledge & Kegan Paul, 1977), p. 135, Lemaire’s emphasis. ⁵⁷ Georges Politzer, Critique of the Foundations of Psychology: The Psychology of Psychoanalysis, trans. by Maurice Apprey (Pittsburg: Duquesne University Press, 1994), p. 107, Politzer’s emphasis. ⁵⁸ Laplanche and Leclaire, p. 123. ⁵⁹ Jacques Lacan, ‘The Instance of the Letter in the Unconscious or Reason since Freud’, in Écrits: A Selection, trans. by Bruce Fink (New York: Norton, 2002), pp. 138–68.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Surrealism’s Subject: Two Cohorts of the Oulipo

93

mild statement of intent he suggests that ‘the idea that the unconscious is merely the seat [le siège] of the instincts may have to be reconsidered’.⁶⁰ Lacan’s use of the spatial metaphor here—the unconscious as the seat of the instincts—picks up on something which Laplanche omits, but which will be very important in looking at the way that the later Oulipians speak of the unconscious as something spatial: a realm or a zone: a repository of meaning. (Abraham and Török too, as we saw in Chapter 2, routinely use the spatial metaphor, both in their concept of the crypt, and their description of ‘the subject’s own mental topography’.) By contrast, Lacan’s model famously uses language for its primary imagery, in the well-known dicta that the unconscious is ‘the discourse of the Other’ and that it is ‘structured like a language’.⁶¹ This is something that Laplanche picks up on in his attack on Politzer’s distinctly first-person account of the unconscious: is it not characteristic of the unconscious, he asks, to include other voices: Is it not precisely Freud’s discovery, under the name of ‘dream-thoughts’, that an utterance may take place ‘in person’ – without, however, being in the first person, but in the alienated form of the second or third person? If the subject of the utterance is at first only a grammatical one, it is none the less a subject: when ‘it talks’ [ça parle] in the unconscious, we do indeed find the dramatic unity so dear to Politzer, but this drama does not necessarily take place ‘in the first person’.⁶²

What Laplanche—and, by extension, Lacan—is driving at, of course, is something rather amenable to Lescure’s account of a language with its own agency, a language which has something to say. Rather than a meaning, the Lacanian unconscious is ‘another lettered system which insinuates itself into conscious discourse, propelling itself through the lacunae in that discourse’.⁶³ These lacunae—in the form of slips of the tongue, forgetting of names, etc., but also in the ruptures produced by transposing the words of a pre-existing text, by taking a well-known poem and swapping its nouns around—are exactly those points at which the discourse is not that of the subject. As Lacan puts it, ‘The unconscious is that part of the concrete discourse, in so far as it is transindividual, that is not at the disposal of the subject in re-establishing the continuity of his conscious discourse.’⁶⁴ ⁶⁰ Lacan, ‘Instance of the Letter’, p. 139. ⁶¹ Jacques Lacan, ‘Seminar on “The Purloined Letter” ’, trans. by Jeffrey Mehlman, Yale French Studies 48 (1972), 38–72 (45); Jacques Lacan, The Four Fundamental Concepts of Psycho-Analysis, ed. by Jacques-Alain Miller, trans. by Alan Sheridan (Harmondsworth: Penguin, 1979), p. 203. ⁶² Laplanche and Leclaire, p. 119. ⁶³ Lemaire, Jacques Lacan, p. 137. ⁶⁴ Lacan, Four Fundamental Concepts, p. 16.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

94

The Oulipo and Modern Thought

This is an unconscious that cannot be accessed—cannot be read—other than glancingly through slips that are momentary and inadvertent. When Breton writes admiringly that the poet Robert Desnos ‘speaks Surrealist at will’ and that ‘[h]e reads himself like an open book’, he is equating Surrealism with the unconscious while crediting Desnos with the power to tune into this as if it were a radio broadcast. For Lacan, however, this is a fundamental misunderstanding of the unconscious. Summarizing, Lemaire spells it out for us that ‘the unconscious is a distinct entity interpolated on the basis of the lacunae in conscious discourse and made up of another discourse which groups the complements of these lacunary points together in another site’.⁶⁵ This is the model which surely echoes in Barthes’s ‘ready-formed dictionary’. The unconscious is no longer a site of simple and singular meanings which can be translated into words; it is a structured system by which words give onto other words, ‘and so on indefinitely’. The affinities between this model of the unconscious and procedures like S+7 or the Lescurean Permutations should be clear. Again and again in the first period of the Oulipo, procedural writing is seen as a method of suppressing the authorial subject, in order to allow ‘language itself ’ to speak. Thus, in the introduction to the Cent mille milliards de poèmes, when Queneau spells out precisely how many millions of years it would take to read the work in its entirety, what is being put to us in no uncertain terms is the notion of agency without authorship: the intention in any of that work’s 10¹⁴ sonnets simply cannot be laid at Queneau’s door—it would absurd to think otherwise. Creating a literature machine is not the same as authoring every work the machine may produce. Rather, an external, secondary agency is being conjured up by means of the interplay between the system of the flipbook and the syntax which determines that an utterance will be coherent even if no-one has yet uttered it. Jeffrey Mehlman’s description of Lacanian psychoanalysis in the late 1950s might equally serve as a manifesto-in-brief for the early Oulipo: ‘[it is] indifferent to deep meanings, concerned more with a latent organization of the manifest than a latent meaning beneath it.’⁶⁶ But the early Oulipo was also the era of ‘syntactic Oulipo’, in Le Lionnais’s terms. As the membership changed, so the group began to move towards ‘semantic Oulipo’, an experimental mode whose pyschoanalytic underpinning is far closer to that of the Surrealists.

⁶⁵ Lemaire, Jacques Lacan, p. 135. ⁶⁶ Jeffrey Mehlman, ‘Introductory Note’, in Lacan, ‘Purloined Letter’, pp. 38–9 (p. 38).

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Surrealism’s Subject: Two Cohorts of the Oulipo

95

THE RETURN OF THE AUTHOR Among the Oulipians of the group’s ‘second wave’ something starts to change. The concept of the agency of the letter begins to share space with a rehabilitated author-subject, the writer with something to say. Oulipian practice, in the form of constrained—as opposed to purely procedural— writing, starts to be described as a means, not of silencing the authorial voice (or of ironically calling it into question), but of finding it and coaxing it into expression, against the allure of the commonplace and the resistant force of psychoanalytic repression. Often in these writers’ work, the attacks against the Surrealists remain, but a distinction must now be drawn between an attack on inspiration itself and an attack on its supposed source in the unconscious. Thus, the use of literary constraints—as opposed to the complete freedom from form proposed in Surrealist automatic writing—comes to be seen not as a way of silencing the author-subject, but rather of challenging a particular Romantic conception of artistic inspiration. Le Tellier’s statement, that ‘the founding of the group attests to the schism of its members from the mythical vision of the “inspired poet”, inherited from the Romantics and prolonged by the Surrealists’,⁶⁷ need not apply only to the moment of the group’s foundation, as it is still echoed regularly in manifesto pieces such as Bénabou’s essay ‘Rule and Constraint’: ‘writing under constraint [ . . . ] may one day permit us to supplant the very notion of inspiration.’⁶⁸ Yet Bénabou, as we shall see, is not jettisoning the author along with his inspiration. In looking at how the group’s second cohort think about the unconscious, we might start with certain comments made by Mathews, inducted into the group in 1973. In his ‘Autobiography’, Mathews reflects on the way that Roussel’s work had a profound influence on his own. Mathews came to Roussel in the 1950s, halfway through writing his first novel. He states that it was after reading Roussel’s highly formal work, shaped by the requirements of his procédé, that he realized that ‘writing could provide me with the means of so radically outwitting myself that I could bring my hidden experiences, my unadmitted self into view’.⁶⁹ The possessive pronouns here are crucial: my hidden experiences, my unadmitted self. For Mathews, it is not some transindividual language which is coaxed into ⁶⁷ [‘La fondation du groupe entérine la rupture de ses membres avec la vision mythique du “poète inspiré”, héritée des romantiques et prolongée par les surréalistes’]. Le Tellier, Ésthetique, p. 8. ⁶⁸ Marcel Bénabou, ‘Rule and Constraint’, in Motte, Oulipo, pp. 40–7 (p. 43). ⁶⁹ Harry Mathews, ‘Autobiography’, in The Way Home: Collected Longer Prose (London: Atlas, 1989), pp. 111–65 (p. 155).

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

96

The Oulipo and Modern Thought

speaking, but part of himself. It is also notable that Mathews is talking about singular meaning here—his texts are not contingent or interpretative, but deal with specific personal experiences. Like Politzer’s analogy of pater and père—the one-to-one correspondence of text and meaning— Mathews’s experiences are expressed in his writing: they are the latent to its manifest: its meaning. Mathews is also positing two types of writing— before and after his Roussellian conversion; unconstrained writing versus constrained—and the statement clearly contains the suggestion of a value difference between those utterances which express the author’s ‘unadmitted self ’ and those which do not. This hierarchy is more explicitly drawn in a phrase from Calvino’s Castle of Crossed Destinies: ‘In writing, what speaks is what is repressed.’⁷⁰ Suffice it to say here, however, that Mathews is presenting us, like Desnos, with two voices, one of which speaks in the devalued banalities of the everyday, the other in the highly-prized language of ‘hidden experiences’. Elsewhere, Mathews describes a writing workshop given by the Oulipo in which students are invited to attempt a constrained writing device known as the ‘snowball’.⁷¹ The snowball is a poetic form in which each line must contain one letter more than the previous, e.g. I am the text which begins sparely assuming magnitude constantly perceptibly proportional incorporating unquestionable incrementations⁷² ⁷⁰ Italo Calvino, The Castle of Crossed Destinies, trans. by William Weaver (London: Secker and Warburg, 1977), p. 102. ⁷¹ Examples of Latis’s new form, the snowball, were circulated to members ahead of the meeting of 4 February 1966, leading Le Lionnais to declare, ‘The snowball will be the Oulipo’s secret weapon’ [‘La boule de neige sera l’arme secrète de l’OuLiPo’], ‘Compte rendu dactylographié de Jacques Duchateau de la réunion n 69 du 4 février 1966’ and ‘Annexe 1 de Latis intitulée “Boule de neige” ’ (FO, DM-2 (2)). ⁷² Mathews and Brotchie, Oulipo Compendium, p. 226.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Surrealism’s Subject: Two Cohorts of the Oulipo

97

Mathews, however, finds himself working with a student who finds the Oulipian exercises irritating and is becoming frustrated: The work done, ‘my’ student got up to read his text. It was a good snowball, and even a fairly long one, but it was also a violent denunciation of the Oulipo and their pretentions. This workshop was the first one at which I had assisted, and here already I was presented with a ‘crisis’. I didn’t know what to say or to think. Noël [Arnaud] and Michèle [Métail] reacted imperturbably, saying to the young man, who was at the time indignant and fierce, that it was a perfect snowball in which there was absolutely nothing for them to criticise. He looked at them in silence for a moment. At the end of that moment of reflection, a flash lit up his face and he shook his head. He had understood, and understood completely: the snowball, this idiotic procedure which we had imposed on him, was exactly the thing which had allowed him to express his anti-Oulipian rage. It was precisely that which had given a form, though unexpected, to a reaction which was beginning to stifle him.⁷³

Here, Mathews seems to be telling us, the young man, indignant and fierce, had been unable to give a clear expression to his frustration until the constraint of the snowball allowed him to make the violent denunciation of the Oulipo which had been eluding him. The constraint in this example seems not so much to exhume hidden personal experiences as to focus the writer’s attention on his present situation. This example has more in common with the critic and translator Chris Andrews’s suggestion that Oulipian writing ‘keep[s] the writer’s forebrain busy with verbal puzzles while the unconscious parts of the compositional process continue relatively undisturbed’.⁷⁴ In a reversal of the Freudian paradigm, here it seems that the unconscious is no longer an interpolation made from the ruptures or disturbances of ordinary discourse; rather, ordinary discourse (incarnated in ‘the writer’s forebrain’) has the potential to disturb the unconscious, unless given a formal bone to chew on. ⁷³ [‘Le troisième exercice était une boule de neige. Le travail fait, “mon” élève se leva pour lire son texte. C’était bien une boule de neige, et même assez fournie, mais c’était surtout une dénonciation violente de l’Oulipo et de ses prétensions. Cet atelier était le premier auquel j’assistais, et voici que déjà se présentait une “crise”. Je ne savais quoi dire ni même quoi penser. Noël [Arnaud] et Michèle [Métail], eux, réagirent imperturbablement, disant au jeune homme, à la fois indigné et fier, que c’était une boule de neige parfaite, sur laqulle il n’y avait absolument rien à redire. Il les regarda un instance en silence. Au bout de cet instant de réflexion, un éclair illumina son visage et il hocha la tête. Il avait compris, et tout compris: la boule de neige, ce procédé idiot qu’on lui avait imposé, était exactement ce qu’il fallait pour exprimer sa rage anti-oulipienne. C’est précisément cela qui avait donné une forme, certes inattendue, à une réaction qui commençait à l’étouffer’]. Harry Mathews, ‘La boule de neige’, Magazine Littéraire (May 2001): 28. ⁷⁴ Andrews, ‘Inspiration’, p. 30.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

98

The Oulipo and Modern Thought

It is often stressed that, in spite of the literary success of some of its members, the aim of the Oulipo has always been not literary production but investigation.⁷⁵ As Bénabou puts it, within the Oulipo, members experiment with constrained writing not in order to create major literary works, but to achieve ‘a better knowledge of the functional modes of language and writing [ . . . by] forc[ing] the system out of its routine functioning, thereby compelling it to reveal its hidden resources’.⁷⁶ Here, the terminology sounds more like that of the early Oulipian proceduralists—that reference to the system and the lack of references to the self: Bénabou’s pronouns all relate to language and not to the author. Nevertheless, language ultimately only reveals its resources: it is a tool for the author to use. For Bénabou too, then, it seems that the author is still the sole agent, and the sole subject in this account, and while the ouvroir may not be concerned with the novelistic output of its celebrated members, it is still a test-site for the effectiveness of new forms of selfexpression.⁷⁷ In this chapter, I have drawn up a number of oppositions— unconscious-as-letter versus unconscious-as-meaning; procedural versus constrained writing; syntactic versus semantic Oulipo; an autonomous language versus the authorial subject—all with the intention of overlaying another opposition on top of them—that of two cohorts of the Oulipo: its first members and its second wave. It is important to stress that this is a loose mapping: I do not mean to suggest any ‘cut-off point’ in Oulipian inductions after which all new members must be, ideologically, ‘second wavers’ (although Roubaud’s suggestion that the group’s second, Perecquian era began in 1969 is useful). And examples of procedural writing ⁷⁵ Queneau even goes as far as to deny that his sections rimantes are literary works in themselves—they are merely the products of investigations: ‘No. No. Oh no no! For example, when I treat one of Mallarmé’s poems in this way, I don’t suppose I have made a literary creation. I produce an example of potential literature but not a literary work in the proper sense. No, no, no. I don’t think it is at all creative. It is much more modest, what we’re doing’ [‘Non. Non. Ah non! Par exemple, quand je traite un poème de Mallarmé de cette façon-là, je n’estime pas que j’ai fait une création littéraire. Je fais une expérience de littérature potentielle, mais non une uvre littéraire à propre parler. Non, non, non. Je ne pense pas qu’il s’agisse du tout là de création. C’est beaucoup plus modeste, ce que nous faisons’]. Charbonnier, Entretiens, pp. 146–7. ⁷⁶ Bénabou, ‘Rule and Constraint’, p. 41. ⁷⁷ Mathews reiterates the difference between the experimentation of the Oulipo and the published product of some of its members: ‘the Oulipo is not a literary school. It is not even concerned with the production of literary works. It is first and last a laboratory where, through experiment and erudition, possibilities of writing under arbitrary and severe restrictions are investigated. The use of these possibilities is the business of individual writers, Oulipian or not’. Harry Mathews, ‘Translation and the Oulipo: The Case of the Persevering Maltese’, The Case of the Persevering Maltese: Collected Essays (Normal, IL: Dalkey Archive, 2003), pp. 67–82 (p. 79).

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Surrealism’s Subject: Two Cohorts of the Oulipo

99

occur in the work of this second cohort just as they do in that of the first— it is, after all, Harry Mathews’s S+7 poem which is quoted in full earlier in this chapter. Nevertheless, what I want to suggest is that the agenda—to satirize Surrealism while driving the author-subject out of writing—is absent when Mathews uses the device. If it is not, then it is directly at odds with Mathews’s other comments, as we have seen. While the notion of the constraint as a means to access the repressed material of the author-subject has a particular history within the Oulipo, this has never been the sole theorization of—or justification for—their activities. If nothing else, we should beware of speaking, as Peter Kuon does, of ‘the Oulipians’ aversion to anything with a whiff of the liberation of the forces of the unconscious’,⁷⁸ as if this were generally applicable across all members of the group and at all periods of its existence. The shift, however, is not confined to the Oulipo, but rather follows a wider intellectual pattern, loosely a movement from structuralism to poststructuralism. It is natural, therefore, that the Oulipo—never unmoored from their context—should reflect this development. Neither is the change purely down to the change in personnel within the group (the arrival of new members, but also the deaths, for example, of Latis and Queneau in 1973 and 1976 respectively). Individuals are liable also to move with the times. The next chapter will outline how in a short space of time around the end of the 1960s and the beginning of the 1970s, Italo Calvino—on the group’s radar, but not yet a member—underwent the exact change in psychoanalytical position which this chapter has traced.

⁷⁸ [‘l’aversion des Oulipiens contre tout ce qui flaire la libération des forces de l’inconscient’]. Peter Kuon, ‘L’Oulipo et les avant-gardes’, in Oulipo-poétiques, ed. by Peter Kuon (Tübingen: G. Narr, 1999), pp. 15–30 (p. 16).

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

4 Calvino at a Crossroads Combinatorics and Anticombinatorics Given the international success of Invisible Cities and If on a Winter’s Night a Traveller, it is perhaps surprising to learn that the output of Calvino’s late period—from his move to Paris in the mid1960s until his death in 1985—is considered in his native Italy to be his least satisfying work. Summing up the critical opinion, Anna Botta cites the view of Franco Fortini as typical: ‘all the works he wrote à la Queneau and à la Perec are deadly, destructive. He was poisoned by the French production of that Parisian period’.¹ This analysis, among other things, rather homogenizes Oulipian writing: it does not really distinguish between Queneau and Perec—these are merely ‘the French production’ and as such are ‘deadly, destructive’. In fact, as the previous chapter has argued, within the Oulipo, Queneau and Perec represent quite distinct poles of literary expression. In this chapter I shall demonstrate that we can trace, in Calvino’s own theoretical writing, a modulation from rigid structuralism to a looser approach based on the exception to the rule; from the death of the author to his reincarnation; from a method akin to that of the early Oulipo to that of its second wave; from ‘à la Queneau’ to ‘à la Perec’.² Instead of using Queneau and Perec as my poles, however, I wish to work by analogy with two other writers instead. Both lived many centuries before ‘that Parisian period’, and both wrote extensively on the subject of combinatorics, the branch of mathematics concerned with the ways in which discrete objects can be combined with each other.³ In the

¹ Quoted and translated in Anna Botta, ‘Calvino and the Oulipo: An Italian Ghost in the Combinatory Machine?’, MLN 112.1 (1997), 81–9 (88). ² We should bear in mind that Calvino was not elected to the Oulipo until 1973, although he was something of a fellow traveller for some time before that, having corresponded with Queneau during his translation of Les Fleurs bleues. ³ For example, given a pair of empty cinema seats, there are two different ways that a couple might arrange themselves: AB, BA. For three seats and three viewers, the number of arrangements rises to six: ABC, ACB, BAC, BCA, CAB, CBA.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Calvino at a Crossroads: Combinatorics and Anticombinatorics 101 ‘Lightness’ essay of Six Memos for the Next Millennium, Calvino draws a parallel between the Epicurean philosopher Lucretius and the medieval theologian Ramón Llull, stating that: [f ]or Lucretius, letters were atoms in continual motion, creating the most diverse words and sounds by means of their permutations. This notion was taken up by a long tradition of thinkers from whom the world’s secrets were contained in the combinatoria of the signs used in writing: one thinks of the Ars Magna of Raymond Lully.⁴

While Calvino emphasizes their similarities, seeing them as part of the same tradition, this chapter will treat Llull and Lucretius as the models for opposing approaches to constraint and creativity within Calvino’s writing and, by extension, within the Oulipo as a whole. Llull characterizes the reader-oriented, highly structuralist tendencies of the Oulipo’s early period; Lucretius, on the other hand, represents for Calvino the author’s capacity to write himself into his work. Ramón Llull (c.1232–1315), whom Roubaud describes as ‘one of the fathers of the “Oulipian” conception of literature’, was born in Palma in Majorca.⁵ After experiencing visions he renounced the troubadour ways of his early years and dedicated himself to the conversion, by reason, of other faiths to Christianity. At various times he taught among the scholastics in Paris, as well as undertaking a number of missionary expeditions to North Africa. During the last of these he is said to have been stoned by an angry crowd, an event which hastened his death the following year.⁶ Of his vast output (he wrote, notably, in Catalan as well as Latin), it is the recurrent use of formal mechanisms—complex systems of combinatorics—for which he is most remembered, and there is conjecture that this was derived from the Jewish kabbalists of his native Majorca.⁷ Llull’s combinatorics is striking for its visual impact. It is characterized by a series of diagrams, tables, and volvelles, with a dominant device of concentric circles divided into cells. The use of these figures is rather complicated—so much so that Llull simplified his own theory during his lifetime in order make it more accessible—and a detailed explanation

⁴ Calvino, Six Memos, p. 26. ⁵ Roubaud, ‘Combinatorial Art’, p. 37. ⁶ This detail, however, is disputed, e.g. in Anthony Bonner (ed.), Selected Works of Ramon Llull (1232–1316) (Princeton: Princeton University Press, 1985), p. 52n. ⁷ Frances Yates, The Art of Memory (London: Routledge & Kegan Paul, 1966), p. 188; Umberto Eco, The Search for the Perfect Language, trans. by James Fentress (Oxford: Blackwell, 1995), p. 60. Both writers also suggest an Islamic influence on Llull’s thought, Yates pointing to Sufism (p. 177), and Eco noting that ‘Majorca during this period was a crossroads, an island where Christian, Jewish and Arab cultures all met’ (p. 53).

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

102

The Oulipo and Modern Thought

would be beyond the scope of this chapter.⁸ Borges’s description of the first circular figure should suffice for our purposes: The letter A, at the centre, signifies the Lord. Along the circumference, the letter B stands for goodness, C for greatness, D for eternity, E for power, F for wisdom, G for volition, H for virtue, I for truth, and K for glory. The nine letters are equidistant from the center, and each is joined to all the others by chords or diagonal lines. The first of these features means that all of these attributes are inherent; the second, that they are systematically interrelated in such a way as to affirm, with impeccable orthodoxy, that glory is eternal or that eternity is glorious; that power is true, glorious, good, great, eternal, eternally powerful, powerfully wise, wisely free, freely virtuous, virtuously truthful, etc., etc.⁹

The diagram (Fig. 8), in other words, is intended to elucidate the nature of God by taking His nine irrefutable ‘Dignities’ and showing that each may be read as a quality of another: His greatness is eternal, etc.¹⁰ The idea, as Martin Gardner neatly puts it, is that ‘[b]y exhausting all possible combinations of these categories we are able to explore all the knowledge that can be understood by our finite minds’.¹¹ It is an idea that Bénabou will play off in his diagrammatic schema of Oulipian method, entitled ‘The Three Cirlces of Lipo’, included with the English translation of his essay ‘Rule and Constraint’.¹² Like Llull’s, Bénabou’s diagram consists of three concentric circles, each divided into sections. The inner circle lists the linguistic elements which may be subject to constraint or manipulation: letter, phoneme, syllable, etc.; the middle circle lists narrative elements: character, décor, event, etc.; while the outer circle contains operations such as displacement, substitution, deduction, etc. The circles can rotate independently, so that any alignment of the three layers—syntactic object, semantic object and operation—is possible. ⁸ Anthony Bonner, The Art and Logic of Ramon Llull: A User’s Guide (Leiden: Brill, 2007), and Eco, Perfect Language, pp. 53–72, both give good descriptions of the method. The former, at book length, treats the earlier model and its simplified successor separately; Eco, for brevity, conflates the two, as I have. Paolo Rossi meanwhile quotes a didactic poem by Llull in which he ‘insists on the miraculous brevity of his combinatorial art and how quick and easy it is to learn and retain’. Logic and the Art of Memory: The Quest for a Universal Language, trans. by Stephen Clucas (London: Athlone, 2000), p. 45. ⁹ Jorge Luis Borges, ‘Ramón Llull’s Thinking Machine’, trans. by Esther Allen, in The Total Library, pp. 155–9 (pp. 155–6). ¹⁰ Llull’s wider intention throughout his work was the conversion of the other monotheistic faiths to Christianity, and there is a clear intention in the selection of these nine divine qualities to base the first principles of the Art in religious common ground, thereby making it more persuasive to a Jewish or Muslim audience. ¹¹ Martin Gardner, Logic Machines and Diagrams, 2nd edn (Brighton: Harvester Press, 1983), p. 9. ¹² Bénabou, ‘Rule and Constraint’, p. 47.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Calvino at a Crossroads: Combinatorics and Anticombinatorics 103

Fig. 8. Ramón Llull, Illuminati sacre pagine professoris amplissimi magistri Raymundi Lull, ars magna, generalis et ultima (Lyon: J. Mareschal, 1517) sig. B1v. Image reproduced courtesy of Getty Research Library.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

104

The Oulipo and Modern Thought

Fig. 9. Marcel Bénabou, ‘The Three Circles of Lipo’ (FO, MS-6 f. 14). Image reproduced by kind permission of the Oulipo.

Thereby, the syntactic and semantic are related in the constraints suggested by the wheel, e.g. words concerning emotion will be subject to a substitution operation. Llull’s own system, however, goes beyond this single revolving table. Similar figures expand the grammar of Llull’s system to include, among

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Calvino at a Crossroads: Combinatorics and Anticombinatorics 105 other things, lists of the virtues, vices, as well as question words, e.g. Whether, In what way, etc. At this point, we run into a problem with the way that Llull uses his own system: that the information generated by the wheels must not contradict the prevailing theological orthodoxy. As Umberto Eco puts it, ‘See, for instance, the question “Whether the world is eternal” (“Utrum mundus sit aeternus”). Llull already knew the answer: negative, because anyone who thought the world eternal would fall into the Averroist error.’¹³ For Llull, generated content which conflicts with his existing theology is discarded. Thus, rather than examine what can be meant by, say, God is greedy, he suppresses this as a nonsense, bad output, an imperfection in his Art. This will be important when we come to look at Calvino’s proposed writing machine. Llull’s combinatorics is an attempt to fully describe a field—to know it entirely—and yet knowledge which is external to the combinatorial system is required to filter its output, to authorize it as an acceptable statement. The user needs to bring their own pre-existing theological knowledge to the system to make it work properly. We might well feel, too, that the Art itself is fundamentally flawed in its assumption that concepts might be fully analysed into their constituent categories in the same way that integers can be expressed by their factors. Like John Wilkins a few centuries later, Llull exhibits a misguided confidence that the categories he chooses will fully exhaust the object under analysis. This is not to mention the oddity that the categories must contain nine elements. So God is fully analysed by His nine Dignities; there are nine virtues and vices (rather than the usual seven deadly sins); and in the category of Questions, when Llull can think of ten terms but doesn’t want to leave one out, Quomodo [In what way] and Cum quo [With whom] are forced to share the same compartment in the table. Once again, this confidence in the precision and completeness of the mechanism’s building blocks is something we must keep in mind when we come to look at Calvino’s literature machine, and we must ask ourselves to what extent Calvino is seduced by the prospect of exhaustiveness, and to what extent he is being ironic when he imagines the complete formal analysis of literary texts. Indeed, according to Llull himself, the Art could be used to illuminate the nature, not just of God, but also the angels, the stars, man, animals, plants, and so on. Gardner notes that ‘[h]ardly a science or subject matter escapes his analysis by [this] method’,¹⁴ and Anthony Bonner goes to the trouble of working through some of the incredible examples given by Llull

¹³ Eco, Perfect Language, p. 63.

¹⁴ Gardner, Logic Machines, p. 14.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

106

The Oulipo and Modern Thought

in demonstration of the Art’s working, which have an almost algebraic appearance due to their referencing of the tables: Question: Whether the soul, which is good, is subject to falsehood, which is evil? Solution: With FS remembering the fourth and fifth compartments, GS understands that, just as dryness is the passive subject upon which is carried out the transformation of wine into vinegar as a result of the transformation of heat into cold, so E I N are the subjects in which is carried out the transformation of Y into Z contrary to the red triangleT, there being a difference between S and B C D, as well as between B C D and E I N, to which B C D are subject, beneath which B C D lies S.¹⁵ Llull goes on to give even more obscure examples where he merely lists the tabular combinations which should resolve the question at hand with no further explanation: Question: Whether, for a similar crime, a townsman should be more punished than a peasant? Solution: special nutritive; E I; N R; mixture digestion; being privation; majority minority.¹⁶ Although I have not described some of the other figures used in these explanations, this last example should illustrate clearly some of the main problems with the system. Firstly, there is not a clear logic in ascertaining which combination should be used to answer any given question, i.e. the system is not dynamic. The question itself does not contain anything which will map it onto Llull’s tables. Instead, Llull simply analyses each possible combination, and decides a posteriori that, for example, this particular arrangement would solve the question of whether a townsman is more culpable than a peasant. Secondly, and perhaps more damningly, is the problem that the type of coherent meaning which Llull says can be derived from his system (e.g. a straight answer to the question of the peasant and the townsman) does not in fact proceed logically from the tables without an additional interpretative step. The figures function more like tarots than any logical system, acting as a creative point of departure, rather than offering a single, static meaning. At the start of the seventeenth century, Francis Bacon would declare that Llull’s method was ‘much like a Frippers or Brokers Shoppe; that hath ends of everie thing, but nothing of worth’,¹⁷ and it is not hard ¹⁵ Bonner, Selected Works, p. 87. ¹⁶ Bonner, Selected Works, p. 91. ¹⁷ Francis Bacon, The Advancement of Learning, ed. by Michael Kiernan (Oxford: Clarendon, 2000), p. 127. Rabelais also weighs in against Llull when Gargantua writes to

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Calvino at a Crossroads: Combinatorics and Anticombinatorics 107 to see what he meant: the system proposes exhaustiveness, but tells us nothing, since every combination needs to be interpreted. This has led some commentators to defend the Art, not as a machine capable of logical demonstration but as a source of creative inspiration. As Frances Yates notes, the user of the first figure (Greatness is eternal, Goodness is great, etc.) is intended to ‘meditate on the complex relations of the Names with one another as they are in the Godhead, before extension into the creation, and as aspects of the Trinity’.¹⁸ Thus, to take, say, the combination KH, it is up to the user to interpret for themselves what it means that His glory is virtuous. Borges even proposes that if we replace the religious Dignities of the first figure with other terms, we end up with a kind of creative prompt for struggling poets: Let us select a problem at random: the elucidation of the ‘true’ color of a tiger. I give each of Llull’s letters the value of a color, I spin the disks and I decipher that the capricious tiger is blue, yellow, black, white, green, purple, orange, and grey, or yellowishly blue, blackly blue, whitely blue, greenly blue, purplishly blue, bluely blue, etc. [ . . . ] The poet who requires an adjective to modify ‘tiger’ proceeds in a manner identical to the machine. He tries them out until he finds one that is sufficiently startling. ‘Black tiger’ could be a tiger in the night; ‘red tiger’, all tigers, for its connotation of blood.¹⁹

Llull’s volvelles have become a literature machine, a simplified version of the ‘Love Letters’ algorithm we encountered in Chapter One. And it is in this aspect that I propose the system is most influential on Calvino’s thought. Unlike Strachey’s programme, Calvino’s literature machine, as we will see, explicitly places an interpretative responsibility on the reader. Nevertheless, whether used for religious or poetic ends, Llull’s Art demonstrates a theological confidence that inspiration is the result of a rigidly observed calculus, and not of a Romantic individualism. As such it presents Calvino with an ideal apparatus with which to drive the authorsubject out of the creative process. Pantagruel, ‘leave judicial astrology and the Art of Lullius alone as abuses and vanities’ [Gargantua and Pantagruel, trans. by M. A. Screech (London: Penguin, 2006), p. 48]. ¹⁸ Yates, Art of Memory, p. 81. ¹⁹ Borges, ‘Ramón Llull’, pp. 157–9. This passage from an article of 1937 takes on a new poignancy when, forty years later, the blind Borges explains why he chose the title El oro de los tigres [The Gold of the Tigers] for a 1972 collection of poems. In a late interview, he recalls visiting a zoo and seeing a tiger as one of his earliest memories. He adds, ‘I feel drawn to tigers, those first things I saw in life. Later came years of myopia, years of blindness, but there was one color that survived. It was the color yellow. And that’s why I entitled a book, The Gold of the Tigers. Since my first vision was the gold of tigers, the color yellow is the color that stands out. It was the last color that my declining eyes could see as they became lost in a gray mist’. [Richard Burgin (ed.), Jorge Luis Borges: Conversations (Jackson: University Press of Mississippi, 1998), p. 169].

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

108

The Oulipo and Modern Thought

The parallel between Llullian and structuralist methodologies should be clear: just as Jakobson’s phonology tabulates distinctive features (e.g. [+/ voice], [+/ nasal]) to describe the entire repertoire of available phonemes, so Llull believed that the entire universe could be distilled into his tables. As we have seen in Chapter One, in the Paris of the early 1960s, a structuralist narratology, based on Vladimir Propp’s Morphology of the Folktale (or rather, based on Levi-Strauss’s extended review of Propp’s book) was gaining influence, and this approach to narrative makes a notable appearance in Calvino’s lecture ‘Cybernetics and Ghosts’ which he delivered in Turin in late 1967.²⁰ The main conceit of ‘Cybernetics and Ghosts’ concerns a literature machine, an automaton ‘capable of replacing the poet and the author’— Calvino’s paper clearly draws from the same structuralist well as ‘Death of the Author’, published the same year.²¹ In Calvino’s account, the author in the age of cybernetics will become redundant, since a machine, working by combinatorics and configured according to the structuralist analysis of narrative, might produce every conceivable text. Imagining this literature machine he writes, I am not now thinking of a machine capable merely of ‘assembly-line’ literary production, which would already be mechanical in itself. I am thinking of a writing machine that would bring to the page all those things that we are accustomed to consider as the most jealously guarded attributes of our psychological life, of our daily experience, our unpredictable changes of mood and inner elations, despairs and moments of illumination. What are these if not so many linguistic ‘fields’, for which we might well succeed in establishing the vocabulary, grammar, syntax, and properties of permutation?²²

We might discern glimmers of Lacan in these lines: in the analogy of the psychological and the linguistic, the positing of a ‘grammar [or] syntax’ of ‘our psychological life’. It recalls the remark from ‘The Instance of the Letter’ that ‘it is the whole structure of language that psychoanalytic experience discovers in the unconscious’.²³ In the way that the passage commandeers everything to its proposed structure—‘all those things’, even ‘the most jealously guarded’—the movement is as much Lacanian as Llullian, reminding us that nothing is conceivable outside ‘the permutations authorized

²⁰ Calvino’s admiration for Barthes during this period is expressed explicitly in a letter of 1965 in which he writes that ‘[Barthes] is perhaps the contemporary critic I admire the most. He is not only a highly intelligent critic but a fine writer’. Letter to Grazia Marchianò, 21 December 1965, Letters, p. 310. ²¹ Calvino, ‘Cybernetics’, p. 12. ²² Calvino, ‘Cybernetics’, p. 12. ²³ Lacan, ‘Instance of the Letter’, p. 139.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Calvino at a Crossroads: Combinatorics and Anticombinatorics 109 by language’.²⁴ Like Llull’s wheel, Lacan’s unconscious or Lescure’s permutations, in Calvino’s literature machine the system is everything: there is neither space nor need for an author-subject. As we can see from the previous description, in terms of sophistication, Calvino’s imagined machine is a far cry from the blunt instruments of Strachey’s ‘Love Letters’ programme or Ducrocq’s Calliope. Its starting point would have to be, not a few tables of moods and themes, but a complete categoric analysis of narrative. Just as Jakobson’s phonological distinctive features describe the full range of phonemes, Calvino’s databanks of narrative elements would allow every possible text to be derived. The structuralist/Llullian method implies that once a field has been fully codified, its universe is simply the complete set of combinations. In such a manner, given a thorough enough analysis of a large enough corpus of narratives, Calvino’s literature machine could then reel through every possibility. Aside from the idea that narrative might ever be entirely anatomized as the product of a certain number of discrete elements—the Llullian fallacy of exhaustion—there is still something rather inane about this writing machine. It is a little like the scene from Gulliver’s Travels in the Grand Academy of Lagado where there is a great wooden frame, ‘twenty foot square’, with iron handles and ‘bits of wood [ . . . ] linked together by slender wires. These bits of wood were covered on every square with paper pasted on them; and, on these papers were written all the words of their language in their several moods, tenses, and declensions, but without any order.’²⁵ Turning the handles will allow the generation of every possible sentence in the language, although this vast task proceeds slowly: Six hours a-day the young students were employed in this labour; and the professor shewed me several volumes in large folio already collected, of broken sentences, which he intended to piece together; and out of those rich materials to give the world a compleat body of all arts and sciences.

Thus, by indiscriminately listing everything that is possible to say, the machine would allow, the most ignorant person at a reasonable charge, and with a little bodily labour, [to] write books in philosophy, poetry, politicks, law, mathematicks and theology, without the least assistance from genius or study.²⁶

²⁴ Lacan, ‘Instance of the Letter’, p. 140. ²⁵ Jonathan Swift, Gulliver’s Travels, ed. by Paul Turner (Oxford: Oxford University Press, 1986), p. 196. ²⁶ Cf. Leibniz’s comment about his own characteristica: ‘les petits esprits, qui auroient de l’application et de la bonne volonté, pourroient non pas accompagner, mais suivre au moins

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

110

The Oulipo and Modern Thought

Swift is satirizing Llull’s Ars with its inability to answer questions dynamically. As we saw, Llull’s method was to take readings from his tables then retrospectively invent the questions to which these could function as the answers. Does not Calvino’s literary machine suffer from the same a posteriori problem? At this point, just as the Llullian wheels need an interpreter to create narrative from them, Calvino argues that it is in the reading process that the machine-authored texts become charged with emotional value—a distinctly Barthesian manoeuvre. As Alison James puts it, in Calvino’s essay, ‘[t]he author’s unconscious [ . . . ] is replaced here by the collective unconscious of the reading public’.²⁷ To make this case, Calvino draws on a model of the unconscious which owes more to Georges Politzer than to Lacan. In the following passage, Calvino suggests that the role of literature is to deal in taboo: The unconscious is the ocean of the unsayable, of what has been expelled from the land of language, removed as a result of ancient prohibitions. The unconscious speaks – in dreams, in verbal slips, in sudden associations – with borrowed words, stolen symbols, linguistic contraband, until literature redeems these territories and annexes them to the language of the waking world.²⁸

After the Lacanian inflexions of the earlier passage, this description of the unconscious is somewhat ambiguous. There is a certainly an excess of linguistic imagery. In the space of two sentences the language of language is employed no less than eight times: unsayable; the land of language; the unconscious speaks; verbal slips; borrowed words; linguistic contraband; literature redeems; the language of the waking world. And yet this is certainly not a description of the Lacanian unconscious-as-letter. The unconscious here is explicitly non-linguistic: it exists outside the land of language; it is unsayable not because it must not be said, but because it cannot be said: when the unconscious speaks it must use borrowed words. This is the unconscious-as-meaning: the ‘hidden experience’ singularly attached to its textual translation. Aside from language, the other major metaphor in the

les plus grands. Car on pourroit toujours dire: comptons, et juger comme il faut par cette voye autant que les data et la raison nous en peuvent fournir les moyens’ [‘lesser minds, those with the will and the application, might, not accompany, but at least follow greater ones. For one could still say, “Let us calculate,” and judge correctly, as far as reason and the data are able to furnish us with the means’]. Louis Couturat, La Logique de Leibniz d’après des documents inédits (Paris: Germer Baillière, 1901), p. 118n. ²⁷ Alison James, Constraining Change: Georges Perec and the Oulipo (Evanston, IL: Northwestern University Press, 2009), p. 126. ²⁸ Calvino, ‘Cybernetics’, p. 19.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Calvino at a Crossroads: Combinatorics and Anticombinatorics 111 passage is geography (the ocean of the unsayable, the land of language, territories, the waking world): the unconscious is figured as a zone, a region, where meaning is contained; and repression and taboo, in the form of expulsion and ancient prohibitions, form the mechanism by which unpalatable material is consigned to this realm. In the case of Calvino’s literature machine—a pure system, free of authorial taint—meaning is supplied not by the author-subject, but by the reader. The mystery of literature, Calvino suggests, lies not in its production, but in the shock which occurs when its elements unexpectedly chime with the reader’s unconscious. He writes that: [l]iterature is a combinatorial game that pursues the possibilities implicit in its own material, but it is a game that at a certain point is invested with an unexpected meaning, a meaning that is not patent on the linguistic plane on which we were working but has slipped in from another level, activating something that on that second level is of great concern to the author or his society.²⁹

The unconscious is separate from the ‘linguistic plane’: it is ‘another level’, a zone of meaning. Although the passage mentions an author, this is not essential to the construction of meaning. It is the reader’s unconscious which endows the text with meaning: The literature machine can perform all the permutations possible on a given material, but the poetic result will be the particular effect of one of these permutations on a man endowed with a consciousness and an unconscious, that is, an empirical and historical man.

There are strong echoes here of Yates’s and Borges’s defence of the Llullian Art as a poetic/meditative device which cannot be programmed to provide clear meanings but rather catalyzes the interpretative faculties of its user. This leaves Calvino in 1967 occupying a position somewhere between the arch-automatists of the early Oulipo and the second wave to come. In ‘Cybernetics and Ghosts’, the author has been expunged entirely—the literature machine does not even need to be fed a source text (whereas the Oulipian procedures are ultimately transformations: whether it be Keats’s Endymion or Wordsworth’s ‘Daffodils’, a starting text must still be chosen). By drawing on combinatorics, Calvino daringly suggests that proceduralism might go beyond the narrow confines of S+7 or the Lescurean Permutations, ultimately surpassing any form of literary Turing test. At the same time, Calvino is careful to preserve the human—the ‘empirical and historical man’—at the centre of the literary experience, ²⁹ Calvino, ‘Cybernetics’, p. 22.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

112

The Oulipo and Modern Thought

and draws on a psychoanalytic model of the unconscious as individual, subjective zone of meaning which is deliberately opposed to, rather than aligned with, the systematization of the textual mode of production. Six years after ‘Cybernetics and Ghosts’, Calvino would put a simplified version of his literature machine into practice. On his first official encounter with the Oulipo, as guest at their November 1972 meeting, Calvino described a story with the working title ‘Le mystère de la maison abominable’. When it would appear the following year, initially in the Italian edition of Playboy, the title had become ‘L’incendio della casa abominevole’: [‘The Burning of the Abominable House’]. Unlike the vast narrative codification proposed in ‘Cybernetics and Ghosts’, ‘The Burning of the Abominable House’ draws on a rather smaller set of narrative elements. In an essay, ‘Prose and Anticombinatorics’, Calvino lists the categories used in his story: 4 characters: A, B, C, D. 12 transitive, nonreflexive actions (see list in next section). All the possibilities are open: one of the 4 characters may (for example) rape the 3 others or be raped by the 3 others. One then begins to eliminate the impossible sequences. In order to do this, the 12 actions are divided into 4 classes, to wit: appropriation of will to incite to blackmail to drug appropriation of a secret to spy upon to brutally extort a confession from to abuse the confidence of sexual appropriation to seduce to buy sexual favours from to rape murder to strangle to stab in the back to induce to commit suicide.³⁰ With a few extra rules (e.g. each character will figure three times as perpetrator and three as victim), and substituting the names Arno, Baby, ³⁰ Italo Calvino, ‘Prose and Anticombinatorics’, in Motte, Oulipo, pp. 143–52 (pp. 145–6).

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Calvino at a Crossroads: Combinatorics and Anticombinatorics 113 Clem, and Dani for the letters A, B, C, D, Calvino can now mechanically derive a set of plots, such as:

Dani Baby Baby Clem Clem Dani Dani Arno Baby Arno Arno Clem

poisons threatens spies upon blackmails extorts a confession from seduces strangles rapes cuts the throat of constrains abuses buys

Arno Clem Arno Arno Baby Baby Clem Baby Dani Clem Dani Dani.³¹

It is clear, he admits, that the algorithm requires further refinement, since in its current version, it allows Clem to buy sexual favours from Dani even after the latter has had her throat cut. Just as Llull rejects combinations which contradict his pre-existing theology, Calvino chooses not to resort to awkward devices (e.g. Dani survived the attack; it wasn’t really her but a case of mistaken identity; she comes back as a ghost; the narrative is nonlinear) when a combination places too great an interpretative burden on the author.³² Nevertheless, the potential for a vast number of narratives to be generated from the combination of a small number of elements should be apparent. The real theme of ‘The Abominable House’, however, is not any particular plot derived from these tables, but rather all of them. It is a story about combinatorial possibility. Narrated by a computer programmer named Waldemar, it concerns an arson attack on a boarding house which has left four people dead. The only clue to the nature of the incident is a copybook, largely burnt but for a label on its front stating ‘An Account of the Abominable Deeds Committed in this House’, and an ³¹ Calvino, ‘Anticombinatorics’, pp. 150–1. ³² Borges, in ‘The Garden of Forking Paths’, is more open to the play of apparently impossible permutations: ‘The fabric of times that approach one another, fork, are snipped off, or are simply unknown for centuries, contains all possibilities. In most of those times, we do not exist; in some, you exist but I do not; in others, I do and you do not; in others still, we both do. In this one, which the favouring hand of chance has dealt me, you have come to my home; in another, when you come through my garden you find me dead; in another, I say these same words, but I am an error, a ghost’ [ Jorge Luis Borges, ‘The Garden of Forking Paths’, Collected Fictions, pp. 119–28 (p. 127), Borges’s emphasis].

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

114

The Oulipo and Modern Thought

index on the back cover with a familiar list: ‘twelve entries in alphabetical order: Blackmail, Drugging, Incitement to Suicide, Knifing, Prostitution, Threatening with a gun, Tying and Gagging, Rape, Seduction, Slander, Snooping, Strangling.’³³ In a passage which recalls the mathematical precision of Queneau’s introduction to the Cent mille milliards de poèmes, Waldemar outlines the problem of using combinatorics to investigate the crime: Even if we accept that each of the twelve deeds was committed by just one person and inflicted upon just one other person, reconstruction would still be a tall order: given that there are four characters to be considered, then taken two by two we have twelve possible relationships for each of the twelve kinds of relationship listed. The number of possible combinations is thus twelve to the twelfth, meaning that we shall have to choose from a total of eight thousand eight hundred and seventy-four billion, two hundred and ninety-six million, six hundred and seventy-two thousand, two hundred and fifty-six potential solutions. It is hardly surprising our overworked police force has chosen not to pursue its enquiries.³⁴

Although the police are not interested, each of the dead has hefty insurance coverage, and Waldemar has been hired by Skiller, the insurance agent, to investigate the incident programmatically. The story, for a large part, consists of a description of the filtering system which Waldemar applies to his model. This is just like Llull’s exclusion of certain combinatorial possibilities based on information external to his Art, e.g. since the idea that the world might be eternal sat outside the prevailing theological orthodoxy, it would have to be discounted. In Calvino’s story, rather than Arno, Baby, Clem, and Dani, the four suspects are given backstories and named as follows: Widow Roessler, the boarding house landlady; Ogiva, Roessler’s fashion model daughter; Inigo, a dissolute young lord; Belindo Kid, an Uzbek wrestler. And these details provide the basis for the exclusion of certain combinatorial possibilities: ‘[strangulation] would be an action of which [Belindo] could only be the subject and not the object: I’d like to see the other three trying to strangle the middleweight wrestler; their puny fingers wouldn’t even go round his tree-trunk neck!’³⁵ As Waldemar puts it, ‘Following this method allows me to rewrite my flow-chart: to establish a system of exclusions that will enable the computer to discard billions of incongruous combinations.’³⁶ ³³ Italo Calvino, ‘The Burning of the Abominable House’, in Numbers in the Dark and Other Stories, trans. by Tim Parks (London: Jonathan Cape, 1995), pp. 156–69 (p. 156). ³⁴ Calvino, ‘Abominable House’, p. 157. ³⁵ Calvino, ‘Abominable House’, p. 160. ³⁶ Calvino, ‘Abominable House’, p. 161.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Calvino at a Crossroads: Combinatorics and Anticombinatorics 115 Thus ‘The Burning of the Abominable House’ situates itself halfway between a pure and exhaustive combinatorics and a set of external factors which suggest a more traditional model of authorial decision-making. Given that it is a text which describes its own construction, under the guise of staging a re-construction, it is hard not to conclude that Waldermar’s words are sometimes as true of Calvino as they are of his narrator: Half I’m concentrating on constructing algebraic models where factors and functions are anonymous and interchangeable, thus dismissing the faces and gestures of those four phantoms from my thoughts; and half I am identifying with the characters, evoking the scenes in a mental film packed with fades and metamorphoses.³⁷

Calvino appears to be moderating the strictness of the position he took up in ‘Cybernetics and Ghosts’. The systematic production of the literature machine now only constitutes half of the creative process. Now there is something else, something external or supplementary to the system: an author ‘identifying with the characters’. And in theorizing how the author and the machine might reach a fruitful accommodation, Calvino looks to the work of the Roman poet Lucretius. ‘Prose and Anticombinatorics’ concludes with the following statement: the aid of a computer, far from replacing the creative act of the artist, permits the latter rather to liberate himself from the slavery of a combinatory search, allowing him also the best chance of concentrating on this ‘clinamen’ which, alone, can make of the text a true work of art.³⁸

We should note, of course, that this position is something of a step down from the one assumed in ‘Cybernetics and Ghosts’, since Calvino now speaks of writing with the aid of a computer, rather than a completely autonomous machine. Nevertheless it inherits the former essay’s assumption that writing is always a combinatory search—this is the necessary slavery of the creative act, which can at least be delegated to a computer. This time, however, there is an insufficiency about the combinatorics: the computer cannot replace the creative act of the artist; the clinamen alone can make a true work of art. The clinamen here is a concept borrowed from the Epicurean philosopher, Lucretius, who lived during the first century . His long poem, De rerum natura is partly concerned with conveying the philosophy that the universe has two components—atoms and the void—both of which are infinite, and that atoms, from which all matter is made, divide into a small number of categories. Thus it is the combination of these atoms which ³⁷ Calvino, ‘Abominable House’, p. 161.

³⁸ Calvino, ‘Anticombinatorics’, p. 152.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

116

The Oulipo and Modern Thought

gives each object its character. Lucretius, punning on the Latin elementa which signifies both letters and atoms, returns frequently to the analogy between words and this Atomist view, i.e. the way that the vast lexicon is derived entirely from the small alphabet: Moreover, it is important in my own verses with what and in what order the various elements are placed. For the same letters denote sky, sea, earth, rivers, sun, the same denote crops, trees, animals. If they are not all alike, yet by far the most part are so; but position marks the difference in what results. So also when we turn to real things: when the combinations of matter, when its motions, order, position, shapes are changed, the thing also must be changed. (II. 1013-22)³⁹

We can see how this would appeal to Calvino—it is an extension of combinatorics not just to literature, but to the entire material universe, and it even uses a literary analogy to make its point. Calvino praises the text as ‘the first great work of poetry in which knowledge of the world tends to dissolve the solidity of the world, leading to a perception of all that is infinitely minute, light, and mobile’.⁴⁰ He draws special attention to the passages in which Lucretius describes miniscule but still visible phenomena, where the ‘atomizing of things extends [ . . . ] to the visible aspects of the world’: it is here that Lucretius is at his best as a poet: the little motes of dust swirling in a shaft of sunlight in a dark room (II. 114-24); the minuscule shells, all similar but each one different, that waves gently cast up on the bibula harena (the ‘imbibing sand’) (II. 374-76); or the spiderwebs that wrap themselves around us without our noticing them as we walk along (III. 381-90).⁴¹

In the same passage, Calvino also alludes to the concept for which Lucretius is perhaps best known, and which has the most significant effect on Calvino’s Oulipian work, constituting the Dionysian element in his combinatorics. He writes that [e]ven while laying down the rigorous mechanical laws that determine every event, [Lucretius] feels the need to allow atoms to make unpredictable deviations from the straight line, thereby ensuring freedom both to atoms and to human beings.⁴²

³⁹ All quotations taken from Lucretius, De rerum natura, 3rd edn, trans. by W. H. D. Rouse (Cambridge, MA: Harvard University Press, 1975). The same image, in largely the same terms, also appears at I. 196–8, I. 823–9, I. 907–14 and II. 688–99. ⁴⁰ Calvino, Six Memos, p. 8. ⁴¹ Calvino, Six Memos, p. 9. ⁴² Calvino, Six Memos, p. 8.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Calvino at a Crossroads: Combinatorics and Anticombinatorics 117 Calvino has in mind here the concept of the clinamen (often translated as swerve), which Lucretius introduces in Book II: while the first bodies are being carried downwards by their own weight in a straight line through the void, at times quite uncertain and uncertain intervals, they swerve a little from their place, just so much as you might call a change of motion. For if they were not apt to incline, all would fall downwards like raindrops through the profound void, no collision would take place and no blow would be caused amongst the first-beginnings: thus nature would never have produced anything. (II. 217-24)

Lucretius then extends the concept, suggesting that it gives rise not just to the formation of matter, but to free will: if all motion is always one long chain, and new motion arises out of the old in order invariable, and if the first-beginnings do not make by swerving a beginning of motion such as to break the decrees of fate, that cause may not follow cause from infinity, whence comes this free will in living creatures all over the earth, whence I say is this will wrested from the fates by which we proceed whither pleasure leads each, swerving also our motions not at fixed times and fixed places, but just where our mind has taken us? For undoubtedly it is our own will in each that begins these things, and from the will movements go rippling through the limbs. (II. 251-62)

Finally, as he restates the latter argument, he introduces the term clinamen: what keeps the mind itself from having necessity within it in all actions, and from being as it were mastered and forced to endure and to suffer, is the minute swerving [clinamen] of the first-beginnings at no fixed place and at no fixed time. (II. 289-93)

The clinamen is wholly unpredictable, yet without it there is nothing. As Derrida puts it, the clinamen represents ‘a certain interfacing of necessity and chance’.⁴³ For Calvino, of course, it represents the author’s exercise of will in deviating from his self-imposed constraints, but we should note the persistence of this theme of necessity: it alone can make of the text a true work of art: it is the sine qua non. Thus true art and subjectivity are indissociable. At this point, we might briefly look at the status of the clinamen within the Oulipo. Roubaud describes it as follows: It is obvious, to anyone who has tried it, that writing according to a fairly demanding Oulipian constraint can be exasperating; for beyond the ⁴³ Jacques Derrida, ‘My Chances/Mes Chances: A Rendezvous with Some Epicurean Stereophonies’, trans. by Irene Harvey and Avital Ronell, in Taking Chances: Derrida, Psychoanalysis and Literature, ed. by Joseph H. Smith and William Kerrigan, (Baltimore: Johns Hopkins University Press, 1984), pp. 1–32 (p. 6).

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

118

The Oulipo and Modern Thought

difficulty (which can perfectly well be mastered) of following the strict requirements of the rule, one is filled again and again with disappointment at not being able to use such-and-such a word or image or syntactical construction that strikes one as appropriate but is forbidden. For such situations the Oulipo has therefore introduced the ‘concept’ of the clinamen, whose Democritean origin sufficiently indicates its nature: that of a nudge given to the uniform, rectilinear, and fearfully monotonous motion of the original atoms so that by colliding they can start the world of writing going in all its variety. A clinamen is an intentional violation of constraint for aesthetic purposes: a proper clinamen therefore presupposes the existence of an additional solution that respects the constraint and that has been deliberately rejected – but not because the writer is incapable of finding it.⁴⁴

There is a proviso here which has been absent from Calvino’s writing on the subject, but which is echoed by Mathews and Brotchie: ‘the clinamen can only be used if it isn’t needed.’⁴⁵ They add that ‘[a] number of Oulipians, notably Italo Calvino, have felt that the clinamen plays a crucial role’, and yet ‘clinamens do not abound’. The implication is clearly that exercise of the clinamen is by no means general within the group. In Calvino’s case, what we can be certain of is that the clinamen is absent from his combinatorics when he describes the writing machine of ‘Cybernetics and Ghosts’ but has entered his thinking by the time of ‘Prose and Anticombinatorics’. It represents a shift from the automatism and reader-oriented stance of the earlier paper, and a redeclaration of the role of the author. The strength of the statement that ‘[the clinamen] alone can make of the text a true work of art’ poses a challenge to the status of, for example, S+7 texts.⁴⁶ The text must go beyond combinatorics, he declares: it must bear the mark of the presence of an author.

⁴⁴ Roubaud, ‘Combinatorial Art’, pp. 43–4. Recently, idea of the clinamen has become something of a theoretical commonplace, cropping up in the work not only of Derrida, but also Deleuze, Lacan, Alain Badiou, Michel Serres, and Harold Bloom. As Warren Motte has pointed out, however, until the 1970s it was something of an obscure concept (despite being the subject of Karl Marx’s doctoral dissertation). With regard to its adoption by the Oulipo, Motte notes that ‘[a]lthough it is nowhere explicitly stated, it may be conjectured that the middleman between Lucretius and the Oulipo is Alfred Jarry’. Jarry, moreover, learned of the clinamen through none other than Henri Bergson, whose lectures he attended in the early 1890s. See Warren F. Motte Jr, ‘Clinamen Redux’, Comparative Literature Studies 23.4 (1986): 263–81 (274); Alastair Brotchie, Alfred Jarry: A Pataphysical Life (Cambridge, MA: MIT Press, 2011), p. 31. On the Oulipian clinamen, see also Christelle Reggiani, Poétiques oulipiennes: La contrainte, le style, l’histoire (Geneva: Droz, 2014), pp. 21–4. ⁴⁵ Mathews and Brotchie, Oulipo Compendium, p. 126. ⁴⁶ Although, as Bellos points out, S+7 is often subject to the discreet application of the clinamen: ‘S + 7 is easy to cheat on and its practitioners admit to obtaining some of their more startling effects by unseen adjustments to the rule’ [Bellos, Georges Perec, p. 597].

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Calvino at a Crossroads: Combinatorics and Anticombinatorics 119 Thus Calvino has by now aligned himself with those other second-wave members of the Oulipo for whom the authorial subject is not something to be elided. Nevertheless, some differences persist. The clinamen, by definition, is supplementary to the textual constraint, while, as we have seen in Chapter 3, for Mathews and others, it is through the constraint itself that the author writes himself into a work. Furthermore, Calvino’s comments on the clinamen leave several questions still open: What does he mean by ‘a true work of art’? What is conferred by the authorial act that elevates the text? For the answer to both of these, we must look to another of Calvino’s rather epigrammatic comments, this time from The Castle of Crossed Destinies: ‘In writing, what speaks is what is repressed’. This statement is just one of a number from the chapter entitled ‘I Too Try to Tell My Tale’ which use the language of psychoanalysis. A few lines earlier the theme is introduced in veiled terms, when the narrator asks, is not the raw material of writing all a rising to the surface of hairy claws, curlike scratching, goat’s goring, repressed violences that grope in the darkness?⁴⁷

The imagery in this sentence is densely packed, but it does the job of coercing the reader to accept an account of psychoanalysis without mentioning it by name. Calvino draws his analogy by imagining a surface below which assorted ghouls are repressed. The allusion may remain unnamed, but we should see fairly clearly what he’s driving at: once again it is the spatial metaphor for the unconscious. In this account the unconscious is portrayed rather cartoonishly, but it features violence, sexuality (grope), ugliness (those hairy claws), and allusions to both the satanic (the goat) and the gothic (say, Poe’s Madeline Usher scratching her way out of the tomb), all situated in darkness. This is, of course, rather childish: not horrific, but an ersatz, comic-book ‘horror’, where well-worn signifiers stand in for the real unnameable terrors we repress. In Calvino’s decision to represent the unconscious in these terms, and in his structuring of the sentence as a rhetorical question (is it not . . . ?) rather than a bald assertion, there is something both unsure and coercive.⁴⁸ Actually, the uncertainty is a function of the coerciveness: the tentative syntax plays on the reader’s sympathy, drawing us into agreement with what is, in fact, a strong and uncompromising proposition: the raw material of writing is all [tutto] a rising to the surface of the repressed terrors of the unconscious. ⁴⁷ Calvino, Castle, p. 101. ⁴⁸ In the original Italian the sentence is as follows: ‘la materia prima dello scrivere non è tutto un risalire alla superficie di grinfie pelose, azzannamenti cagneschi, cornate caprine, violenze impedite che annaspano nel buio?’ [Il castello dei destini incrociati (Turin: Einaudi, 1973), pp. 100–1].

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

120

The Oulipo and Modern Thought

Writing then becomes a term which Calvino uses in an exclusive, even slightly elitist manner: it is that which addresses the unconscious; all other utterances are outside the present consideration. Therefore, the assertion that ‘[i]n writing, what speaks is what is repressed’ becomes a tautology: if the repressed were not speaking, the text would not, in his terms, be considered writing. This is the ‘true work of art’, then, and the clinamen carries the mark of the author when it draws on his own repressed material. We are back with the old picture of literature annexing the territories of taboo to the language of the waking world, except now, if the text doesn’t deal in linguistic contraband, it isn’t worthy of the label of Literature. Here is the passage in which the assertion appears. It comes immediately after a retelling of Sade’s twin stories, Justine and Juliette: All this is like a dream which the word bears within itself and which, passing through him who writes, is freed and frees him. In writing, what speaks is what is repressed. And then the white-bearded Pope could be the great shepherd of souls and interpreter of dreams Sigismund of Vindobona.⁴⁹

In the first sentence, a sudden switch occurs. Throughout The Castle of Crossed Destinies it has been images—the illustrations on two decks of tarot cards—which have produced, or ‘borne within themselves’, the tales. The two sections of the work were each composed by laying a deck of tarots in a grid and treating them as a set of pre-existing illustrations for the suite of tales Calvino had to write. In this passage, however, he substitutes the word for the cards: each tale is ‘a dream which the word bears within itself ’. We are invited to read the work and its method of production— using images—as a metaphor for the word-play of psychoanalytic free association. Words, he implies, are as ambiguous, as overloaded in their suggestions, as the archetypes depicted on a deck of tarot, and by analysing them, by following through the stories they suggest to us at any particular moment, we are ‘writing’, which by definition is to free up (and free us of) the material we had bound up in our unconscious.⁵⁰ It is only through this oblique metaphor, the almost imperceptible substitution of word for image, that Calvino addresses how the writer can hope to access the dark materials of the unconscious. It is left largely to the reader to infer that this is why the book had such an unusual method of construction: the tales ⁴⁹ Calvino, Castle, p. 102. ⁵⁰ At this point, unlike the earlier sentence with its B-movie horrors standing in for the psychoanalytic unconscious, Calvino seems almost bold enough to mention his master by name. It is not to happen, however—the name of the father cannot be spoken—and Freud’s cameo is in disguised form as ‘Sigismund of Vindobona’, Vindobona being the Roman name for Vienna.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Calvino at a Crossroads: Combinatorics and Anticombinatorics 121 could not have been told without the constraint imposed by the tarot cards; or rather, without the constraint, the tales would have been banal, undeserving of the designation of writing as Calvino intends it. The clinamen therefore becomes the perfect analogy for Calvino’s concept of writing, as it implies both constraint and deviation. In Lucretius’s model of the universe, the clinamen has no value unless it occurs within a context of otherwise perfect order. It has no meaning in an anarchic universe or a non-constrained mode of literary production: the deviation must have a system to deviate from. Under Lucretius’s influence, then, Calvino throws off the combinatorics of Llull—the complete system with no need of an author—for a new approach: in order to access the taboo material of the unconscious—without which there can be no true art—it is necessary to apply and overrule a system of constraint. As Perec puts it, explaining why he removed one of Life A User’s Manual’s one hundred chapters: the chapter had to disappear to break up the symmetry, to introduce an error into the system because when one establishes a system of constraints, there must also be anti-constraint in it. The system of constraints – and this is important – must be destroyed. It can’t be rigid, there must be a bit of play in it, as they say, it has to creak a little; it mustn’t be completely coherent; there needs to be a clinamen – this is in the Atomist theory of Epicurus: ‘The world operates because from the start there is an imbalance.’ According to Klee, ‘Genius is the error in the system.’⁵¹

For Perec it is constraint and anticonstraint; for Calvino combinatorics and anticombinatorics. The clinamen represents the later Calvino’s rejoinder to Lescure’s withering ‘Romantisme tout ça’ and every other mocking critique of authorial expression in the comments of the Oulipo’s founder members. The author, replaced by a machine in ‘Cybernetics and Ghosts’, has been welcomed back into the project of anticombinatorial writing.

⁵¹ [‘il faut que ce chapitre disparaisse pour casser la symétrie, pour introduire dans le système une erreur parce que quand on établit un système de contraints, il faut qu’il y ait aussi l’anti-contrainte dedans. Il faut – et c’est important – détruire le système des contraintes. Il ne faut pas qu’il soit rigide, il faut qu’il y ait du jeu, comme on dit, que ça grince un peu; il ne faut pas que ça soit complètement cohérent; il faut un clinamen – c’est dans la théorie des atomes d’Epicure: “Le monde fonctionne parce que au départ il y a un déséquilibre.” Selon Klee “le génie, c’est l’erreur dans le système.” ’] Georges Perec, ‘Entretien Georges Perec/Ewa Pawlikowska’, Littératures 7 (1983), 69–77 (p. 70).

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

5 Perec and Mathews Translation and Analytic Philosophy in the South Seas Il réalise, sous une forme originale, l’union wittgensteinienne des jeux de langage et des formes de vie. Jacques Roubaud, ‘L’auteur oulipien’¹

Let’s begin with two thought experiments. Both are pivotal to philosophy of language in the twentieth century, and both ask us to imagine languages that do not exist in the real world. In the first, which appears at the beginning of Wittgenstein’s posthumous Philosophical Investigations (1953), it is not the terms of the language that are made-up—Wittgenstein uses words drawn from German, and we can substitute English ones. Instead, what is important is the scope of the language: Let us imagine a language [ . . . ] meant to serve for communication between a builder A and an assistant B. A is building with building-stones: there are blocks, pillars, slabs and beams. B has to pass the stones, and that in the order in which A needs them. For this purpose they use a language consisting of the words, ‘block’, ‘pillar’, ‘slab’, ‘beam’. A calls them out; – B brings the stone which he has learnt to bring at such-and-such a call.²

It may have only four words, but Wittgenstein issues us with a challenge— ‘Conceive this as a complete primitive language’. A few paragraphs later he intensifies this proposition: ‘We could imagine that [it] was the whole language of A and B; even the whole language of a tribe’ (PI §6, p. 3e). What is being asked of us—to conceive of slab, block, pillar, and beam as the whole language of a tribe—is quite a stretch. We’ll surely find ¹ Jacques Roubaud ‘L’auteur oulipien’, in Michel Contat (ed.), L’auteur et le manuscrit (Paris: Presses de France, 1991), pp. 77–92 (p. 83). ² Ludwig Wittgenstein, Philosophical Investigations, trans. by G. E. M. Anscombe (Oxford: Blackwell, 2001 [1953]), §2, p. 2e. Subsequent references will be given parenthetically in the main text.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Translation and Analytic Philosophy in the South Seas

123

ourselves wondering whether our builders will not at least require some additional terms: numerals, perhaps, or determiners like this or that? But if we find it hard to imagine a tribe for whom the names of these four stones constitute the whole of their language, then so we should: the difficulty we feel is really the point of Wittgenstein’s exercise. What is under attack here is the idea that every word in a language has a meaning that correlates with it. This, argues Wittgenstein, is sometimes true, sometimes not, but it just won’t do as a model for the way that a real language works: ‘it is the idea of a language more primitive than ours’ (PI §2, p. 3e). And so the four-word slab language of the builder and his tribe is an attempt by Wittgenstein to imagine such a primitive language, one for which meaning—as conceived in the Western philosophical tradition going back to St Augustine—is a straightforward matter of ‘ostensive definition’: where each word points to its object. The troublingly narrow worldview of the builder’s tribe should be troubling: it shows how far our idea of meaning falls short of our idea of a language. What it shows too is the usefulness of an invented tribe and their invented speech for getting to the nub, quickly and vividly, of a problem in our understanding of the way languages work. It uses narrative to memorable effect, and it circumvents some of the clutter and muddle that is wont to arise when we try to turn our own language inwards on itself. It is such an effective device that the American philosopher W. V. O. Quine, writing shortly after the appearance of the Philosophical Investigations, takes it up again. In Word and Object (1960), in which Quine bolsters Wittgenstein’s attack on ostentation, we are asked to imagine a field linguist working with an undocumented tribe in a remote jungle.³ ‘A rabbit scurries by, the native says “Gavagai,” and our jungle linguist notes down the sentence “Rabbit” (or “Lo, a rabbit”) as a tentative translation.’⁴ What are we to make of the linguist’s assumption that gavagai and rabbit are equivalent terms? At a first glance, perhaps, it seems reasonable. But Quine begins to find alternative meanings: ‘Who knows but what the objects to which this term applies are not rabbits after all, but mere stages, or brief temporal segments, of rabbits? [ . . . ] Or perhaps the objects to which “Gavagai” applies are all and sundry undetached parts of rabbits.’ It is not possible, in this situation, to propose a

³ W. V. O. Quine, Word and Object (Cambridge, MA: MIT Press, 1960). ⁴ Quotations here are taken from the essay ‘Meaning and Translation’ which predated Word and Object by a year and contained much of the same material in slightly condensed form. W. V. O. Quine, ‘Meaning and Translation’, in Reuben Brower (ed.), On Translation (New York: Oxford University Press, 1959), pp. 148–72 (p. 148).

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

124

The Oulipo and Modern Thought

translation for the unfamiliar term without making an assumption about the closeness of the tribesman’s worldview to our own: When from the sameness of stimulus meanings of ‘Gavagai’ and ‘Rabbit’ the linguist leaps to the conclusion that a gavagai is a whole enduring rabbit, he is just taking for granted that the native is enough like us to have a brief general term for rabbits and no brief general term for rabbit stages or parts.⁵

All this might sound like unnecessary quibbling. But Quine is not writing a manual for ethnographers; rather, the purpose of his radical doubt is to follow Wittgenstein in considering the nature of meaning. Quine argues that ‘[f]or light on the nature of meaning we must think [ . . . ] of radical translation’, and this radical translation can most conveniently be conceived of if we imagine a people whose similarity to us cannot be taken for granted. Given the Oulipo’s ongoing interest in issues of translation, it is not surprising that allusions to these two languages—the slab language and the gavagai language—should appear from time to time in their works. Gavagai, for example, becomes the name of a tribe in one of the homophonic translations for the collective Troll de Tram exercises (BO 68): Chez les Gavagaï, la sueur blanche des rongeurs, étalée sur une tartine, pass pour une friandise de premier choix. Lait de rats moites: extra ce beurre! [Among the Gavagai, the sweet cream of rodents, spread on bread, passes for a first-class delicacy: Moist rats’ milk: better than butter! ]⁶

Meanwhile, in Mathews’s third novel, The Sinking of the Odradek Stadium, a cocktail party conversation about semiotics begins with one guest announcing, ‘When I say, slab, I mean, slab’, a nice, half-quotation of Wittgenstein’s ‘When he says “Slab!” he means “Slab!”’ (PI §19, p. 7e) that turns the scrupling precision of the original into blunt self-assertion.⁷ But besides these direct allusions, there are Oulipian works that use these ⁵ Quine, ‘Meaning and Translation’, p. 153. ⁶ Oulipo, Troll de tram, p. 156. ⁷ Harry Mathews, The Sinking of the Odradek Stadium (London: Paladin, 1986), p. 28. The novel was initially serialized in the Paris Review #51–#54 (1971–72), before making its first appearance as a single volume as part of a collection that also included the earlier novels, The Conversions and Tlooth. This collection, confusingly, was also titled The Sinking of the Odradek Stadium (New York: Harper & Row, 1975). Page numbers given here, however, will refer to the 1986 edition which contains the novel on its own and includes the index omitted from the Carcanet edition the year before. Odradek Stadium contains another rather nice joke at Wittgenstein’s expense when its two protagonists find themselves talking at cross-purposes. It turns out that Zach, who has apparently has never heard of the Austrian philosopher, has an acquaintance of the same name: ‘I meant Charlie Wittgenstein’ (163). Both Wittgensteins appear side-by-side in the novel’s index; Charlie has precedence, alphabetically, over Ludwig.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Translation and Analytic Philosophy in the South Seas

125

two philosophical thought experiments in more oblique and more profound ways. The subject of this chapter will be translation. In the short story, ‘The Dialect of the Tribe’, Mathews has his narrator state, ‘The longer I live – the longer I write – the stronger becomes my conviction that translation is the paradigm, the exemplar of all writing.’⁸ It is tempting, and not far from the truth, I don’t think, to attribute the same sentiment to Mathews himself. In a sequence of stories that reflect on the nature and limits of translation, we find Mathews and Perec drawing on the examples of Wittgenstein and Quine, using invented tribes and their invented languages to consider language in extreme states. Once again, these are works of fiction, but like Wittgenstein and Quine, they use the imaginative example as a way of engaging with the debates of their time concerning what translation can and cannot do.

THE DIALECT OF THE TRIBE First published in 1980, ‘The Dialect of the Tribe’, takes the form of a letter—a linguist’s reply to a friend who has asked him to contribute to a Festschrift in his honour. The narrator describes how a librarian—Mrs Moon—recently showed him an article that she thought might be of interest.⁹ The article, from an Australian anthropological journal is written by one Dr Ernest Botherby, and consists of a transcription of an oral declaration by the abanika—the ‘chief word-chief ’—of a small hill tribe in northern New Guinea. The article is written in the tribe’s own language, Pagolak, and is untranslated. Curious as to why Botherby should have published in Pagolak and not provided a translation of the tribesman’s words, the narrator sets out to translate the article himself, at which point he makes a surprising discovery: Pagolak is an untranslatable language. It is easily learned—‘[i]n a matter of days I found myself perfectly capable of understanding what the abanika was saying’—but completely resistant to translation—‘I was perfectly incapable of repeating it in other terms’. Nevertheless, entranced by the beauty of the abanika’s message, and

⁸ Harry Mathews, ‘The Dialect of the Tribe’, SubStance 27 (1980): 52–5; reprinted in Harry Mathews, The Human Country: New and Collected Stories (Normal, IL: Dalkey Archive, 2002), pp. 7–14. Subsequent page numbers will refer to the latter edition. Appropriately enough, the publication of the original story was antedated by its own translation: ‘Abanika, traditore’, trans. by Marie Chaix, L’Arc 76 (1979): 73–6. ⁹ The study of the Bactrian language—or, rather, a fictionalized version of it—is also the subject of an earlier Mathews story, ‘Remarks of the Scholar Graduate’, Best & Company, 1 (1969): 35–41; reprinted in The Human Country, pp. 41–50.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

126

The Oulipo and Modern Thought

desperate to communicate it to his friend, he makes tortuous efforts to explain the text: (afanu) is sitokap utu sisi. This phrase leaves an impression, approximately, of ‘resettling words in [own] eggs’: aptly enough, after the youngsters emerge from afanu through sitokap utu sisi into nuselek and its attendant privileges of ton wusi and aban metse, they claim to be emerging from boyhood (rather: ‘boybeing’) seabirds from chicken eggs (utopani inul ekasese nuselek ne sami sisinam) – dear Christ, it doesn’t mean that – but can you perhaps intuit how tokkele (not ‘words’, those words) return to their sisi to re-emerge in unexpected, unrecognizable forms?

Any translator will recognize the types of frustration which this passage feverishly mimics: the awkward square brackets (‘resettling words in [own] eggs’); the over-fussy self-correction of ‘boyhood (rather: “boybeing”)’; the irritation with one’s own failure to find the precise form: ‘dear Christ, it doesn’t mean that’. These are all stages in the process of creativity and compromise that goes into producing an effective translation, one that is both adequate in its fidelity to its source and acceptable in its intelligibility to its readers. For Mathews’s narrator, however, faced with Pagolak’s irresolvable inflexibility, these types of negotiation are impossible: the frustrations are inevitably overwhelming. Finally he decides that his friend must ‘share in the totality of the words’ by reading the article for himself: ‘I promise to steal the book for you, from this selfsame library – fuck Ms. Moon, since she won’t let me Xerox it.’ The letter ends mid-sentence, having descended into the non-sense of the untranslatable: And let me on this private occasion add a few last words, spoken out of the fullness of my mind and heart with admiration, with devotion, with love: Amak kalo gap eleman narna la n’kat tokkele sunawa setan anman umanisi sutu pakotisovulisanan unafat up lenumo kona kafe avanu Io se akina ba nasavuloniputitupinoluvasan (!!) abanika esolunava efaka nok ornunel put afanu nanasfluvo sitokap utu sisi narnu nanmana tes awa nuselek kot tak nalarnan namele Pagolak kama –

This final sentence recalls the untranslated letter in the (fictitious) Pan language which appears in Odradek Stadium.¹⁰ Yet there the words had all been glossed individually over the course of the novel, so that the attentive reader could work out how to translate the message in its entirety. Here, however, no gloss has been, or can be provided. The reader is at a loss. Like Quine’s gavagai narrative, ‘The Dialect of the Tribe’ is an instructive fable—a story about translation that distorts its subject to the point of

¹⁰ Mathews, Odradek Stadium, p. 182.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Translation and Analytic Philosophy in the South Seas

127

caricature so that we may see its flaws more clearly. Effectively, it takes Quine’s tale and rewrites its ending. Where Quine asked: How does the linguist know if his translation is right?, Mathews instead imagines: What if the linguist knows that his translation will always be wrong? What if we move from a position of radical doubt to one of certainty, but a certainty of failure not of success? Mathews is offering us a picture of linguistic incommensurability: the idea that there are elements of one language that cannot be mapped onto another. With Pagolak, of course, we have a limit case: a language composed entirely of incommensurable terms, and one whose incommensurability applies to all other languages, ‘whether [ . . . ] English, or French – or Middle Bactrian’. But Pagolak only magnifies a problem which is in fact universal in translation, what the translation theorist Emily Apter calls one of ‘primal truisms of translation’. Writing about ‘The Dialect of the Tribe’, she describes its moral as follows: ‘even if one has access to the language of the original, there remains an x-factor of untranslatability that renders every translation an impossible world or faux regime of semantic and phonic equivalence.’¹¹ It will be the contention of this chapter that this tale forms part of a nexus of stories by Mathews and Perec each of which use the trope of the ethnographer and the newly discovered language in order to draw attention to quite specific problems in translation. ‘AN ASSERTION OF LOYALTY’: MATHEWS TRANSLATING PEREC TRANSLATING MATHEWS Before we come to the other ethnographical tales, however, it will be useful to say a little about the relationship between Mathews and Perec, since this was a friendship that had a direct impact on both writers’ work, resulting in collaborations, separate but parallel attempts at the same written exercise, and—most significantly—the translation of each other’s work. This relationship of support and mutual influence is important if we are to consider that the three stories—two by Mathews, one by Perec—are consciously linked. Early in 1970, Mathews, living in Paris at the time, received a letter praising his novel The Conversions which was about to be published in a French translation by Claude Portail and Denis Roche. The letter was from Georges Perec, and although Perec had several works to his own name by this time (La Disparition had come out the previous year), ¹¹ Emily Apter, The Translation Zone: A New Comparative Literature (Princeton, NJ: Princeton University Press, 2006), p. 159.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

128

The Oulipo and Modern Thought

Fig. 10. Marcel Bénabou’s handwritten minutes for the meeting of 8 November 1972 record Perec’s outline of what would become La Vie mode d’emploi and the long discussions that followed it. Oulipo, ‘Compte rendu autographe de Marcel Bénabou de la réunion du 8 novembre 1972’ (FO, DM-3 (24)). Image reproduced by kind permission of the Oulipo.

Mathews had neither met him nor read anything by him.¹² Nevertheless, Mathews replied and the pair arranged to meet. This is how Mathews recalls their first encounter: ‘We drank together, and went on to have dinner, and I thus entered into the most exhilarating, hilarious, intense, and satisfying relationship I have ever known with a man, and doubtless will ever know.’¹³ By the end of the year, Perec had undertaken to produce a French translation of Mathews’s second novel, Tlooth, despite a significant linguistic disadvantage. As Mathews puts it, ‘Perec loved Peanuts and [ . . . ] had learned from it most of the English he knew when he began translating Tlooth’.¹⁴ It is hardly surprising that Perec should have run into difficulties when he came to the Spoonerisms and jumbled syllables of the long, scrambled sex scene half-way through the novel.¹⁵ At the Oulipo’s first meeting of 1971, Perec proposed that Mathews should be invited as the next guest, although it wasn’t until March 1973 that he was eventually elected as a member.¹⁶ It was around this time too that the outline of Perec’s novel La Vie mode d’emploi (hereafter VME) was beginning to take shape. Ahead of the meeting of 8 November 1972 Perec had tabled an agenda item the cryptic title ‘Du petit lait pour FLL’: ‘A little treat for FLL’. On the day, this turned out to be the plans for ‘a novel which implements this idea of the ¹² See Bellos, Perec, p. 448. ¹³ Mathews, ‘Autobiography’, p. 139. ¹⁴ Harry Mathews, The Orchard: A Remembrance of Georges Perec (Flint, MI: Bamberger, 1988), p. 21. ¹⁵ See Bellos, Perec, for Perec’s letter to Mathews of 4 January 1971: ‘I’ve got without too much trouble to the soirée at the Palazzo Zen; then I ran into a leg-over episode and lost the thread entirely’ (p. 466). ¹⁶ ‘Compte rendu autographe de Georges Perec de la réunion du 12 janvier 1971’ (FO, DM-3 (1)). Earlier meeting minutes show that Mathews had, in fact, already been suggested as an invitee as far back as February 1966.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Translation and Analytic Philosophy in the South Seas

129

“semantic Oulipo” which the Président-Fondateur [François Le Lionnais] is so keen on’.¹⁷ Perec explained the setting for this novel: ‘a building from which the façade has been removed; the building has ten floors and there are ten rooms on each floor’.¹⁸ Moreover, the novel would employ three mathematical structures in its composition: a Knight’s Tour of a 10x10 chessboard; a sestina-like device which Perec called a ‘false dizine’; and a number of Graeco-Latin squares (‘bi-squares’) of order 10, which could be used to distribute a vast number of elements—fabrics, accessories, jewels, etc.—across the novel’s chapters. One such set of elements involves of a table of twenty writers—among them Rabelais, Joyce, Queneau, and Borges—who must be quoted, although ‘parfois légèrement modifiées’ [‘sometimes slightly modified’], in chapters determined by the bi-square.¹⁹ So, for example, the doll’s house described in Chapter XXIII (VME 135) has been imported from the Ithaca section of Ulysses.²⁰ Mathews too is one of the writers on the list, and thus one of the opening scenes from The Conversions is reproduced in Chapter XXVII of VME. Here is how the scene appears in Mathews’s novel: In the library Mr. Wayl laid an oblong case of green leather on a white table. Having turned on a ceiling spotlight to illuminate the case, he opened it. A weapon rested on the brilliant red lining, its smooth handle of ash, its billshaped flat blade of gold. According to Mr. Wayl the instrument was a ritual adze. The side of the bill we had first beheld was plain, but its reverse was chased with wiry engravings, depicting seven scenes. Six had in common the figure of a longhaired woman with full breasts and a face crosshatched for swarthiness.²¹

And this is how Perec renders it in VME: il fit venir le peintre chez lui et posa sur la table un coffret oblong en cuir vert. Ayant allumé un projecteur accroché au plafond pour éclarer le coffret, il l’ouvrit: une arme reposait sur la doublure d’un rouge éclatant, sa poignée

¹⁷ [‘[un] roman qui met en oeuvre cet “Oulipo sémantique” auquel le PrésidentFondateur est particulièrement attaché’]. ‘Compte rendu dactylographié rédigé transmis par François Le Lionnais’ (FO, DM-3 (24)). ¹⁸ [‘une maison dont on a enlevé la façade; la maison a dix étages et il y a dix pièces par étage’]. ¹⁹ Georges Perec, La Vie mode d’emploi (Paris: Hachette, 1978), p. 636. Subsequent quotations from this edition will be given parenthetically in the main text. ²⁰ This observation was made by Gabriel Josipovici, ‘Georges Perec’s Homage to Joyce (and Tradition)’, Yearbook of English Studies 15 (1985): 179–200. For more on this relationship, see Jacques Mailhos, ‘The Art of Memory; Joyce and Perec’, in Transcultural Joyce, ed. by Karen Lawrence (Cambridge: Cambridge University Press, 1998), pp. 151–69 (pp. 164–5). ²¹ Harry Mathews, The Conversions (New York: Random House, 1962), p. 4.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

130

The Oulipo and Modern Thought

lisse en frêne, sa lame plate, falciforme, en or. [ . . . ] Sept minuscule gravures étaient finement ciselées sur une des faces, mais il ne parvint pas à voir ce qu’elles représentaient, même en s’aidant d’une forte loupe; il vit seulement que sur plusiers d’entre elles, il y avait vraisemblablement une femme aux cheveux très longs. (VME 158)

Appropriately enough, then, when Mathews came to translate a chapter of VME for a literary anthology in 1984, he chose the one in which this passage occurs, backtranslating his own once-removed source text as follows: He brought the painter to his apartment and laid on the table an oblong case covered in green leather. Having turned on a ceiling spotlight to illuminate the case, he opened it. A weapon rested on the brilliant red lining, its handle of ash, its curved flat blade of gold. [ . . . ] Seven minute engravings were delicately chased on one of its sides, but he was unable to make out what they represented, even with the aid of a powerful magnifying glass. All he could see was that in several of them there apparently figured a woman with very long hair.²²

There is an uncanny effect here. Mathews’s second attempt resembles his first, of course, but it is not quite the same: it has been filtered through the language and prose style of his friend.²³ This interplay between the two authors—Perec’s assimilation of Mathews; Mathews’s re-translation of this—persists in the full English version of the novel, Life A User’s Manual, since the novel’s translator, David Bellos, chose to reuse the chapters already translated by Mathews. Bellos’s note to acknowledge this debt admits to making ‘minor modifications’, a phrase whose echo of Perec’s own ‘légèrement modifiées’ adds another pleasing layer of circularity to the proceedings.²⁴ There is thus a playful mutuality in Perec and Mathews’ translations of each other, a merging of style and literary identity that goes beyond the straightforward author-translator relationship. In a tribute published after Perec’s death, Mathews would write: ²² Georges Perec, ‘Extract from La Vie mode d’emploi (Chapter XXVII)’, in Atlas Anthology Two, trans. by Harry Mathews (London: Atlas, 1984), pp. 59–63 (p. 61). ²³ This process of quotation and backtranslation occurs again in Chapter LXXIV of VME which contains another passage from The Conversions, and which was published in Mathews’s translation: Georges Perec, ‘Space and Underground’, trans. by Harry Mathews, Grand Street 3.1 (1983): 146–50. ²⁴ Georges Perec, Life A User’s Manual, trans. by David Bellos (London: Collins Harvill, 1987), p. 581. Hereafter quotations from this edition will be given parenthetically, labelled LUM, in the main text. Bellos’s ‘minor modifications’ are for the most part merely anglicizations of certain American forms: ‘freight elevators’ become ‘goods lifts’, ‘garbage cans spilling out cheese crusts’ become ‘dustbins spilling out cheese rinds’, etc.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Translation and Analytic Philosophy in the South Seas

131

What mattered to me in having Georges as a translator was not his talent or even (for one of my novels) his undertaking the work without any guarantee of publication, but his willingness to translate works that he had not even read, simply because I had written them. More than literary sympathy, his collaboration was an assertion of loyalty.²⁵

In Perec’s and particularly Mathews’s writing, we find that translation is constantly under investigation—picked at, seen from different angles. In their lives, it could be a declaration of true friendship. In addition to translating each other’s writing, the two would work together on collaborative pieces: on the mock-scholarly essay ‘Roussel and Venice’ (see Chapter Two), for example, or on the more private exercises of the Bibliothèque oulipienne. Mathews even makes a disguised cameo appearance, as Mathias Henrijk, author of the novel Öd Rädek, in Perec’s unfinished detective novel, 53 Days.²⁶ Furthermore, the two writers would sometimes undertake the same exercise independently. Thus Perec’s ‘Still Life/Style Leaf ’ (published in 1981 and translated by Mathews the same year) and Mathews’s ‘Still Life’ which appeared two years later, are parallel attempts at an exhaustive description of the writers’ desks. Here is Perec’s (in Mathews’s translation): The solid wood desk, on which I am writing, formerly a jeweler’s workbench, is equipped with four large drawers and a top whose surface, slightly sloping inwards from the edges (no doubt so that the pearls that were once sifted on it would run no risk of falling to the floor) is covered with black fabric of a very tight woven mesh.²⁷

And here is Mathews’s: You sit at a desk whose off-white expanse is finished in some composite substance, matt but smooth. Small objects lie along the edges of the desk – pencils, pens, erasers, scissors, cigars, a paper cutter, a magnifying glass. Three blank pages are spread out on the surface in front of you. [ . . . ] No new words have appeared on any of the three pages in front of you. Your attention is like a dog restlessly waiting for a master. You lie down on the sofa to read and wait.²⁸ ²⁵ Mathews, ‘Autobiography’, pp. 139–40. ²⁶ Georges Perec, 53 Days, ed. by Harry Mathews and Jacques Roubaud, trans. by David Bellos (London: Harvill, 1992), p. 47. ²⁷ Georges Perec, ‘Still Life/Style Leaf ’, Le Fou parle 18 (1981): 3–6; ‘Still Life/Style Leaf ’, trans. by Harry Mathews, Yale French Studies 61 (1981), 299–305 (299). ²⁸ Harry Mathews, ‘Still Life’, Littératures 7 (1983): 151–4; reprinted in The Human Country, pp. 115–20 (pp. 115–16, 118). When the piece was originally published, in a special issue of Littératures dedicated to Perec, the editors inserted a note drawing particular attention to the fact that, in allowing an English-language work to stand untranslated in a French-language journal, they were honouring Perec—Mathews’s ‘traducteur-ami privilégié’: ‘privileged translator friend’ (151n)—by his absence.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

132

The Oulipo and Modern Thought

It is clearly the same exercise—a still life, a description of the writer’s desk—yet the two are rather different—expressions of different literary personalities, with Mathews employing the second person and introducing a movement of time which Perec’s, frozen in a single moment like the narrative of VME, omits. The works that will form the central focus of this chapter provide another example of this type of independent-but-parallel writing. Perec’s ‘Tale of the Misunderstood Anthropologist’ from VME and Mathews’s ‘A Problem in Translation’ from the essay ‘Translation and the Oulipo’ have much in common: both concern young ethnographers carrying out research into remote tribes in the islands of the Malay Archipelago; while Perec’s unfortunate researcher is Austrian, Mathews, by the addition of a couple of letters, makes his character Australian; the name of Mathews’s protagonist, Botherby—who already made an appearance in ‘The Dialect of the Tribe’—recalls Bartlebooth, a principal character of VME; and both writers instinctively write in that ironic Borgesian mode wherein their fictional worlds are filled out with a liberal dose of real-world erudition.²⁹ Perec’s fictional anthropologist, Marcel Appenzzell, for example, is said to be on close terms with Marcel Mauss, while Mathews’s Botherby is drawn to New Guinea after reading the work of Miklouho-Maclay, Samuel Macfarlane, and Otto Finsch.³⁰ Most importantly of all, both texts are highly concerned with the implications for translation of languages with significantly reduced vocabularies.

PEREC’S TALE: CATEGORIES IN ORANG-KUBU Among the many pleasures of VME are the supporting materials included with the novel. The paratexts at the back of the book run to nearly a hundred pages, and include a vast index, a diagram of the building in which the novel is set, a timeline, and a table of the novel’s mininarratives, the backstories and digressions embedded within its larger plan. One of the entries in this table, ‘The Tale of the Misunderstood ²⁹ Indeed, Borges was playing with the possibilities of fictional tribal languages some time before either Perec or Mathews. In ‘Brodie’s Report’ (1970), an Amazonian tribe speak a language which has no vowel sounds, making it impossible to transliterate their name [in Jorge Luis Borges, Collected Fictions, trans. by Andrew Hurley (London: Penguin, 1998), pp. 402–8 (p. 403)]. ³⁰ Harry Mathews, ‘Translation and the Oulipo: The Case of the Persevering Maltese’, Brick 57 (1997): 39–44; reprinted in The Case of the Persevering Maltese: Collected Essays (Normal, IL: Dalkey Archive, 2003), pp. 67–82 (p. 67). Subsequent page numbers will refer to the latter edition.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Translation and Analytic Philosophy in the South Seas

133

Anthropologist’, directs us to a portion of Chapter XVX (‘Altamont, 2’) concerning one Marcel Appenzzell, a twenty-three year old ethnographer with a university post at Graz. A disciple of the great Bronisław Malinowski and his dictum of participant observation, Appenzzell sets off to undertake his own fieldwork, studying the Orang-Kubu people on the island of Sumatra. Like so many of the tales in VME, however, the overriding tone is of sadness and failure. After Austria’s Anschluss with Germany, Appenzzell, a Viennese Jew, is driven from his post at Graz and forced to move to Paris. Worst of all for the dedicated ethnographer, determined to immerse himself within the culture of his subject, he is rejected by the Kubus. In his last letter to his mother, he writes, I asked for nothing more than to be one of them, to share their days, their pains, their rituals. Alas! They didn’t want to have me, they were not prepared to teach me their customs and beliefs! They had no use whatever for the gifts I laid beside them, no use at all for the help I thought I could give. (LUM 112)

In a curious mirror of his own persecution in Europe, he adds that: [i]t was because of me that they abandoned their villages and it was only to discourage me, to convince me there was no point in my persevering, that they chose increasingly inhospitable sites, imposing ever more terrible living conditions on themselves.

With this passage in mind, Daphné Schnitzer has read the Appenzzell story as ‘a displaced metaphor of the persecution of the Jewish people,’ suggesting that the narrative deals with Perec’s loss of his own parents to the Nazis, and is directly inspired by his analysis with Pontalis in the years immediately prior to the its composition.³¹ In a perfect example of the type of homophonic psychoanalytical reasoning which the Oulipo roundly send up (see Chapter Two), Schnitzer sees the word Papouas [Papuans, i.e. the Kubus] as ‘instantly decomposing’ into papa-où? [father-where?].³² What will concern us here, however, will be the words of the Papuans, rather than the word Papuans itself. Although, like all the tales of VME, Appenzzell’s entire story runs to only a few pages, Perec manages to find space for a description of the Kubu language: it was quite close to the coastal tongues, and Appenzzell could understand it without major difficulty. What struck him especially was that they used a very restricted vocabulary, no larger than a few dozen words, and he wondered if the ³¹ Daphné Schnitzer, ‘A Drop in Numbers: Deciphering Georges Perec’s Postanalytic Narratives’, Yale French Studies 105 (2004): 110–26 (118). ³² Schnitzer, ‘Drop in Numbers’, p. 124.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

134

The Oulipo and Modern Thought

Kubus, in the image of their distant neighbours the Papuans, didn’t voluntarily impoverish their vocabulary, deleting words each time a death occurred in the village. One consequence of this demise was that the same word came to refer to an ever-increasing number of objects. Thus the Malay word for ‘hunting’, Pekee, meant indifferently to hunt, to walk, to carry, spear, gazelle, antelope, peccary, my’am – a type of very hot spice used lavishly in meat dishes – as well as forest, tomorrow, dawn, etc. Similarly Sinuya, a word which Appenzzell put alongside the Malay usi, ‘banana’, and nuya, ‘coconut’, meant to eat, meal, soup, gourd, spatula, plait, evening, house, pot, fire, silex (the Kubus made fire by rubbing two flints), fibula, comb, hair, hoja’ (a hair-dye made from coconut milk mixed with various soils and plants), etc. (LUM 110)

It would be a stretch at this stage to assume a connection between this language ‘no larger than a few dozen words’ and the four-word language of Wittgenstein’s tribe of builders. Nevertheless, the similarity starts to become more pronounced when Appenzzell compares Kubu’s agglomeration of multiple meanings under a single word to the use of indistinct terms in Western languages: ‘He pointed out [ . . . ] that these characteristics could perfectly well apply to a Western carpenter using tools with precise names – gauge, tonguing plane, moulding plane, jointer, mortise, jack plane, rabbet, etc. – but asking his apprentice to pass them to him by saying just: “Gimme the thingummy.”’ (LUM 110) Now the presence of Wittgenstein’s slab language, hovering over Kubu, seems rather less in doubt. The opening images of the Philosophical Investigations are represented in force here: ‘Gimme the thingummy’ is a nice play on ‘Bring me a slab’ (PI §19, p. 7e), while Perec’s unpacking of the carpenter’s toolkit recalls Wittgenstein’s injunction: ‘Think of the tools in a tool-box: there is a hammer, pliers, a saw, a screw-driver, a rule, a glue-pot, glue, nails and screws. – The functions of words are as diverse as the functions of these objects.’ (PI §11, p. 6e). Kubu, Perec informs us, has a ‘very restricted vocabulary’ and we should be in no doubt what its model is. Furthermore, with this type of philosophical pedigree, it is only natural that Perec’s Kubu, far from being mere incidental detail or window dressing for Appenzzell’s tragic tale, should contain a thesis of its own about translation. The meat of Perec’s argument comes in the strange passage—otherwise irrelevant as far as the plot of Appenzzell’s otherwise highly condensed tale is concerned—in which the terms Pekee and Sinuya are glossed. Firstly, it should be noted that the Orang-Kubu, unlike Mathews’s Pagolak (or his Oho and Uha tribes, discussed below), are real: ‘a traditionally nomadic forest-dwelling people of Southern Sumatra.’³³ And in fact, their reticence ³³ Øyvind Sandbukt, ‘Kubu Conceptions of Reality’, Asian Folklore Studies 43 (1984): 85–98 (85).

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Translation and Analytic Philosophy in the South Seas

135

towards outside contact is not an invention on Perec’s part either. Writing in the 1980s, the anthropologist Øyvind Sandbukt describes the genuine Kubu as follows: they live in virtually complete isolation from the surrounding Malay society. Within the relatively restricted area of forest which they inhabit, this isolation is maintained by a set of drastically sanctioned taboos which prohibit the sort of contact that is essential for anthropological investigation. Initial attempts at establishing contact with the Kubu through proper Malay intermediaries yielded only brief, formal encounters, and independent efforts at tracking them down in the forest tended to trigger a deeply inculcated response of apparently panic-stricken flight.³⁴

Meanwhile, the possibility that a community might ‘voluntarily impoverish their vocabulary, deleting words each time a death occurred’, while not a trait associated with the Kubu, is not a mere philosophical thought experiment either. In the language of the Lardil people of Mornington Island, off the coast of Queensland Australia: ‘The name of a recently dead person should not be said, for it reminds close relatives of their sorrow. [ . . . ] If the deceased was named after a natural phenomenon or creature, as is normally the case, then this word should not be used in conversation.’³⁵ So, for example, if there were a man named Mr Green, then, following his death, his friends would avoid referring to the colour green. Since it is the word and not the object—the signifier not the signified—that has become taboo, what commonly happens in cultures which observe this practice is that other, related words are extended so that the meaning of the proscribed word becomes incorporated among their senses. In Perec’s Kubu this is taken to extreme lengths, so that long lists of loosely related concepts—both actions and objects: to hunt, gazelle, etc.— are represented by a single term. Perec’s list brings to mind the philosopher J. L. Austin’s question, ‘Why do we call different things by the same name?’. Austin answers this by suggesting, ‘Anyone who wishes to see the complexity of the problem, has only got to look in a (good) dictionary under such a word as “head”: the different meanings of the word “head” will be related to each other in all sorts of different ways at once.’³⁶ This idea—that the different meanings of a term might be related to each other in all sorts of different ways at

³⁴ Sandbukt, ‘Kubu Conceptions’, 85. ³⁵ David McKnight, People, Countries, and the Rainbow Serpent: Systems of Classification among the Lardil of Mornington Island (Oxford: Oxford University Press, 1999), p. 29. ³⁶ J. L. Austin, ‘The Meaning of a Word’, in Philosophical Papers (Oxford: Clarendon, 1961), pp. 23–43 (pp. 42–3).

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

136

The Oulipo and Modern Thought

once—seems to hold true for the case of Pekee. The term, then, has a variety of specific meanings, some more similar than others, but all of which fall, just about, under the umbrella of a loose category. Each term in the list might relate to some, but not necessarily all, of the others. Here, once more, we are deep in Wittgenstein territory, since it is Wittgenstein who offers us the simplest model for the way that members of a category need not all share a single defining characteristic resemblance, but may instead exhibit a network of mixed, partial similarities: I can think of no better expression to characterize these similarities than ‘family resemblances’; for the various resemblances between members of a family: build, features, colour of eyes, gait, temperament, etc. etc. overlap and criss-cross in the same way. (PI §67, p. 27e)

Within such a category, however, some characteristics may be more central than others. So, to return to Perec’s list, the conceptual context of hunting appears to have something of a central position, in that certain members, for example the meanings to walk and gazelle, are related to each other only through its intermediary. Crucially, the central term implies priorities which are different from what we—the Western readership—would expect. Perec is deliberately setting up categories which are difficult for us to imagine as coherent. The relationships that produce these family resemblances—the relationship between walking and hunting, say—simultaneously suppress others that might seem more obvious to us. It is easy to see how walking is an activity related to hunting, and yet walking also has a central role in a good number of other activities which the Kubu might perform: collecting water, visiting friends, gathering firewood, etc. Binding the activity to walk to a word which also means to hunt promotes this sense while simultaneously demoting the others. Similarly, it is obvious how spear or antelope fit into the category of hunting—as instrument and prey—and yet their inclusion under the hunting signifier, Pekee, intensifies this relationship at the expense of the other elements in their ordinary signifying chain, e.g. spear: tool, etc.; antelope: animal: food, etc. Hunting, as a locus of intersection, becomes the context in which the concepts grouped under Pekee have the greatest valency, while other contexts, e.g. walking in the context of collecting water, become relegated to transferred senses, drifting towards metaphor. This reassignment of valency away from the reader’s expected categorizations has an effect of estrangement, making the Kubus’ world somewhat surreal. Kubu terms, euphemistically standing in for a richer but proscribed vocabulary, operate in the manner of Freudian condensation, like the nodal points ‘upon which a great number of the dream-thoughts converge’, and where ‘associations of the most various

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Translation and Analytic Philosophy in the South Seas

137

inherent importance are used for laying down thought-connections as though they were of equal weight’.³⁷ Perec’s next example becomes more playful, applying this reweighting effect even more forcefully, and making it harder for the reader to deduce the underlying category: Sinuya, a word which Appenzzell put alongside the Malay usi, ‘banana’, and nuya, ‘coconut’, meant to eat, meal, soup, gourd, spatula, plait, evening, house, pot, fire, silex (the Kubus made fire by rubbing two flints), fibula, comb, hair, hoja’ (a hair-dye made from coconut-milk mixed with various soils and plants), etc.

The first few items might easily be grouped under the category of eating: the action, some food items, an instrument for preparation. After that, however, the relationships seem more tenuous to us. To imagine the primary sense of house as being place-of-eating requires something of an imaginative leap, a dream-like redistribution of intensity. That other famous reduced language, C. K. Ogden’s Basic English, also lacks a specific term for evening, but how strange it would be to construct it as food-time. This involves, for us, a rebalancing of valencies, the suppression of other more significant categoric contexts: sleeping; safety; family. Adding fibula and comb to the mix perhaps stretches the imagination comically too far—how can they relate to eating?—while the etc. which closes the list is a nicely glib touch, suggesting that this is not an arbitrary series and that we should know, or be able to guess, the remaining meanings. Yet, rather than undermining Perec’s Kubu language, these anomalies might in fact lend it an additional authenticity. Thinking back to the tale of the gavagai, Quine warns against ‘taking for granted that the native is enough like us to have a brief general term for rabbits and no brief general term for rabbit stages or parts.’ In other words we should be wary of making the unwarranted assumption that our own linguistic categories— and the worldview they embody—are universal. Take for example the Australian language, Dyirbal. Dyirbal features a classification system with a spoken marker, not entirely unlike the articles of a gendered language (e.g. le versus la or un versus une in French), which the linguist George Lakoff summarizes as follows: Whenever a Dyirbal speaker uses a noun in a sentence, the noun must be represented by a variant of one of four words: bayi, balan, balam, bala. These ³⁷ Sigmund Freud, The Interpretation of Dreams, in The Standard Edition of the Complete Works of Sigmund Freud, ed. by James Strachey, trans. by James Strachey et al., 24 vols (London: Hogarth and Institute of Psychoanalysis, 1953–74 [1899]),IV, pp. 283, 294.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

138

The Oulipo and Modern Thought

words classify all objects in the Dyirbal universe, and to speak Dyirbal correctly one must use the right classifier before each noun.³⁸

Where these Dyirbal classifiers differ from the gendering of nouns in, say, French, is in their predictability. Except in few circumstances— particularly around biological sex, e.g. un homme, une femme—French has no reliable set of rules relating some quality of the signified to the grammatical gender of its signifier. In Dyirbal, by contrast, the four noun classes can be described in terms of their characteristics. The results, however, are surprising, to say the least: bayi: men, kangaroos, possums, bats, most snakes, most fishes, some birds, most insects, moon, storms, rainbow, boomerangs, some spears, etc. balan: women, bandicoots, dog, platypus, echidna, some snakes, some fishes, most birds, firefly, scorpions, crickets, hairy mary grub, anything connected with fire or water, sun and stars, shields, some spears, some trees, etc. balam: honey, all edible fruit and vegetables and plants that bear them. bala: parts of the body, meat, bees, wind, yamsticks, some spears, most trees and vines, grass, mud, stones, noises and language, etc.³⁹ As with Perec’s Sinuya, we can just about deduce the unifying categories in these strange lists, but we are forced to allow for exceptions. There is a pleasure, an amusement, in having to accept the suppression of other signifying chains, which to our non-Dyirbal minds would seem to take precedence. Balam, for example, seems to indicate the category of food, yet it also includes cigarettes and excludes meat, which belongs to the bala category instead. And whereas Perec’s fictionalized Orang-Kubu are forced into this type of categoric anomaly by a cultural practice which voluntarily diminishes their vocabulary, as Lakoff is at pains to point out, the anomalies in Dyirbal are purely illusions on the part of the beholder: ‘from the perspective of the people doing the classifying [this is] a relatively regular and principled way to classify things.’⁴⁰ There is nothing strange, to the Dyirbal speaker, about seeing the world this way. ³⁸ George Lakoff, Women, Fire, and Dangerous Things: What Categories Reveal About the Mind (Chicago: University of Chicago Press, 1987), p. 92. ³⁹ Robert Dixon, Where Have All the Adjectives Gone? and Other Essays in Semantics and Syntax (Amsterdam: Mouton, 1982), p. 178. ⁴⁰ Lakoff, Women, p. 95.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Translation and Analytic Philosophy in the South Seas

139

Nor is Dyirbal’s apparent strangeness unique. The language is by no means an outlier of otherness. In his 1921 grammar of the Mayan language, for example, Alfred Tozer describes noun groups for, among other things, ‘handfuls of herbs or hair’, ‘for splinters’, ‘for things that are close to each other, such as jugs, staffs, or seated men’, ‘for recumbent living beings’, ‘for a written chapter or articles of faith; or for rows of stones, each row or stone being above the other’, and ‘for counts of years, months, days, leagues, cocoa, eggs, and calabashes or squashes’.⁴¹ But by far the most famous illustration of the bewildering otherness of alien classification systems comes not from not from field linguistics but from fiction, in the form of Borges’s fictional Chinese encyclopaedia which divides the animal kingdom into: (a) those that belong to the emperor; (b) embalmed ones; (c) those that are trained; (d) suckling pigs; (e) mermaids; (f) fabulous ones; (g) stray dogs; (h) those that are included in this classification; (i) those that tremble as if they were mad; (j) innumerable ones; (k) those drawn with a very fine camel’s hair brush; (l) etcetera; (m) those that have just broken the flower vase; (n) those that at a distance resemble flies.⁴²

Like Perec’s gloss of the Kubu term Sinuya, Borges’s list includes a mischievously placed etcetera. It is this fantastic classification, as much as Wittgenstein’s builder, which provides the basis for Perec’s Kubu language and its brief, dense, witty reminder that the world can be organized along very different lines from our own. It is not hard to see the type of problem that such a gulf between worldviews can throw up for translators. If a Kubu text describes an animal as a Pekee, should we translate this as gazelle or antelope or peccary? This is, of course, Quine’s gavagai problem precisely. But Perec’s tale stops short of concerning itself explicitly with translation: it simply leaves its strange terms for us to ponder. By contrast, Mathews’s belated response, ‘A Problem in Translation’, draws more strongly on Quine to question the very validity of translation, and suggests that languages may be incommensurate not merely in their categorizations, but in the basic concepts which they are capable of expressing.

⁴¹ Alfred M. Tozzer, A Maya Grammar (Cambridge, MA: Peabody Museum, 1921), pp. 290–2. Tozzer’s list is an expanded version of an earlier categorization by Zelia Nuttall [‘A Suggestion to Maya Scholars’, American Anthropologist 5.4 (1903), 667–78 (674–8)], itself a translation of Beltran de la Rosa’s 1746 classification. However, certain of the categories quoted above are Tozzer’s own additions. ⁴² Jorge Luis Borges, ‘John Wilkins’ Analytical Language’, trans. by Eliot Weinberger, in The Total Library: Non-Fiction 1922–86 (London: Penguin, 1999), pp. 229–32 (p. 231).

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

140

The Oulipo and Modern Thought OHO AND UHA: A PROBLEM IN TRANSLATION

In Mathews’s tale, ‘A Problem in Translation’, which forms the first part of the essay ‘Translation and the Oulipo’, we are reintroduced to the character of Dr Ernest Botherby, the ethnographer whose article on Pagolak we encountered in ‘The Dialect of the Tribe’. Now Mathews fleshes Botherby out with some detail which can only remind us of Perec’s twenty-three-year-old Appenzell: ‘The peoples of New Guinea were a favourite subject with Botherby. He had begun studying them years before when, at the age of twenty-four, he undertook a solitary voyage into the interior of the island, vast areas of which remained uncharted at the time.’⁴³ On his voyage, Botherby encounters a tribe whom he designates the Ohos, and discovers that they use a massively reduced spoken vocabulary: They [ . . . ] used speech, but a speech reduced to its minimum. The Oho language consisted of only three words and one expression, the invariable statement, ‘Red makes wrong’. Having patiently won over the tribal chiefs, Botherby was able to verify this fact during the many weeks he spent with them. Other needs and wishes were communicated by sounds and signs; actual words were never used except for this unique assertion that ‘Red makes wrong’.⁴⁴

Even more than Perec’s Kubus, the Ohos have a language to match that of Wittgenstein’s builder. As Wittgenstein puts it, ‘It disperses the fog to study the phenomena of language in primitive kinds of application [so that] one can command a clear view of the aim and functioning of the words’ (PI §5, p. 3e). It should be immediately apparent that this is what Mathews has in mind with the Oho’s unique, unvarying utterance, ‘Red makes wrong’: to shine a light on some aspect of the functioning of language by simplifying it. For Wittgenstein, the purpose of the slab language is to show that the meaning of an utterance is nothing other than the use that the speaker makes of it. He reels off a list of what he terms ‘language-games’—the forms-of-life (Lebensformen) in which we use language, a list which includes giving orders, telling jokes, making up stories, and ‘requesting, thanking, cursing, greeting, praying’ (PI }23, p. 10e). These are the contexts in which words are used, and, for Wittgenstein—for late Wittgenstein, at least—to speak of a word as having meaning independently from its usage context is misleading and ⁴³ Mathews, ‘Translation and the Oulipo’, p. 67. ⁴⁴ Mathews, ‘Translation and the Oulipo’, p. 68.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Translation and Analytic Philosophy in the South Seas

141

nonsensical. It is for this reason that he devises the example of the builder and his apprentice. Not only do they draw from a very limited vocabulary, but, crucially, they have only one language-game: the requesting of building blocks. It is in this sense that it is a ‘complete primitive language’. The builder is not part of a wider narrative: we needn’t concern ourselves with how he negotiates his wage, or tells his wife that he loves her. His four-word vocabulary—slab, block, pillar, beam—is complete because he exists only in an extremely limited fictional context. Mathews’s tribe, on the other hand, is more fully realized. Although the community live ‘in conditions of extreme simplicity’, we hear how ‘its members [are] hunter-gatherers equipped with rudimentary tools. They [procure] fire from conflagrations occurring in forests nearby’.⁴⁵ We learn too that, while ‘Red makes wrong’ may be their sole spoken statement, ‘other needs and wishes [are] communicated by sounds and signs’. Thus, the Oho language may have only three words, but these are not the only available signifiers. So when Botherby asks whether there are other tribes in the region, the Ohos are able to point north and east. This makes the tribe seem rather like a community of monks who have taken a vow of silence: we are back in the realm of voluntary lexical impoverishment— more Kubu than slab-language. The fact that the Ohos say very little out loud does not restrict the language-games which they can enter into: answering Botherby’s questions, or disapproving fiercely when he sets out west. As it turns out, however, not everything is communicable—by word or by sign—to the Ohos. Quine, still considering his ethnographer among the gavagai tribesmen, pondered a three-word sentence of his own, not ‘Red makes wrong’ but ‘Neutrinos lack mass’: ‘Who would undertake to translate [that] into the jungle language?’⁴⁶ Where should one begin when what is unfamiliar is not only the words themselves but the whole vast supporting structure of Western science by which their arrangement is meaningful? In ‘Translation and the Oulipo’, Mathews stages this conceptual lack when Botherby leaves the Ohos behind and travels further into the rainforest until he discovers another tribe, whom he dubs the Uhas. The Uhas, just like the Ohos, have only one spoken phrase in their language: ‘Here not there.’ When Botherby eventually returns to the valley of the Ohos, he attempts to tell them about their neighbours: he was overcome by an understandable (if professionally incorrect) eagerness to share his second discovery, to wit, that near them lived a people of the same ⁴⁵ Mathews, ‘Translation and the Oulipo’, p. 68. ⁴⁶ Quine, Word and Object, p. 76.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

142

The Oulipo and Modern Thought

stock, leading a similar life, and possessed of the same basic gift of speech. As he was expounding this information with gestures that his audience readily understood, Botherby reached the point where he plainly needed to transmit the gist of the Uhas’ one statement. He hesitated. How do you render ‘Here not there’ in a tongue that can only express ‘Red makes wrong’?⁴⁷

The import of that final rhetorical sentence is clear: Oho and Uha are incommensurate. How can we translate between languages that are conceptually different from each other? In literary translation, when the translator encounters a concept which is familiar to the original readership but which would be less so to the targetlanguage audience, a common strategy is to find an equivalent that is not a direct translation, but which performs a similar function without being distractingly unusual for readers. Mathews writes approvingly of Perec’s decision, when translating Tlooth’s opening scene (a makeshift sporting contest in a Siberian prison camp), to replace a set of baseball positions with football positions, since some of baseball’s obscurer terminology would be offputting and meaningless to a French readership. Thus the player at ‘center field’ becomes a ‘demi-centre’ (i.e. a centre half), and so on.⁴⁸ Perec’s creative licence falls well within the limits of reasonability: the exchange of one sporting position for another. Yet in ‘Translation and the Oulipo’ Mathews contends that this fairness—the reasonable exchange—is illusory, even in an equivalence as seemingly simple as declaring one’s national identity. Comparing the two statements ‘Je suis français’ and ‘I’m American’, Mathews argues that these encode a very different set of meanings: Translating these words by ‘I’m French’ and ‘Je suis américain’ brings to light differences rather than similarities. In the language of a Frenchman, nationality basically means a complex cultural history; in the language of an American it basically means the fact of citizenship, often coupled with a disrespect for history.⁴⁹

And the incommensurability of Oho and Uha serves as a limit-case example of this thesis, whereby the concept contained in the Uhas’ ‘Here not there’ is simply not present in the Ohos’ culture. With this, Mathews has led us to the edge of the problem of linguistic determinism—the Whorfian hypothesis that language determines thought. As the neuroscientist Peter Gordon puts it,

⁴⁷ Mathews, ‘Translation and the Oulipo’, p. 68. ⁴⁸ Harry Mathews, ‘Fearful Symmetries’, The Literary Review 45.3 (2002): 444–52; reprinted in The Case of the Persevering Maltese, pp. 55–65 (p. 58). Subsequent page numbers will refer to the latter edition. ⁴⁹ Mathews, ‘Fearful Symmetries’, p. 55.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Translation and Analytic Philosophy in the South Seas

143

The question of whether linguistic determinism exists in the stronger sense has two parts. The first is whether languages can be incommensurate: Are there terms that exist in one language that cannot be translated into another? The second is whether the lack of such translation precludes the speakers of one language from entertaining concepts that are encoded by the words or grammar of the other language.⁵⁰

Gordon continues by noting that, ‘For many years, the answer to both questions appeared to be negative’: Botherby’s problem in translation could only occur in fiction, and Quine’s jungle linguist seeking to tell the tribe about neutrinos would simply have to start from first principles. Nevertheless, in the 2004 paper cited above, Gordon described an Amazonian tribe—not fictitious at all—who live along the banks of the Maici River in north-western Brazil. The tribe are known as the Pirahã, and the following year they were the subject of another article, this time by the linguist Daniel Everett, which would prove enormously controversial thanks to its description of the Pirahã language which emerges as almost as unusual and circumscribed as that of the Ohos. At the beginning of his article, Everett reels off a list of extraordinary details about the Pirahã: A summary of the surprising facts will include at least the following: Pirahã is the only language known without number, numerals, or a concept of counting. It also lacks terms for quantification such as ‘all’, ‘each’, ‘every’, ‘most’, and ‘some’. It is the only language known without color terms. It is the only language known without embedding (putting one phrase inside another of the same type or lower level, e.g., noun phrases in noun phrases, sentences in sentences, etc.). It has the simplest pronoun inventory known, and evidence suggests that its entire pronominal inventory may have been borrowed. It has no perfect tense. It has perhaps the simplest kinship system ever documented. It has no creation myths – its texts are almost always descriptions of immediate experience or interpretations of experience; it has some stories about the past, but only of one or two generations back. Pirahã in general express no individual or collective memory of more than two generations past. They do not draw, except for extremely crude stick figures representing the spirit world that they (claim to) have directly experienced.⁵¹

Among all of this astonishing detail perhaps it is the absence of number, numerals and the very concept of counting which seems so extraordinary that we might wonder if we, or the researchers, have misunderstood. ⁵⁰ Peter Gordon, ‘Numerical Cognition Without Words: Evidence from Amazonia’, Science 306 (2004): 496–9 (496). ⁵¹ Daniel Everett, ‘Cultural Constraints on Grammar and Cognition in Pirahã: Another Look at the Design Features of Human Language’, Current Anthropology 46.4 (2005): 621–46 (622). Although the Pirahã are undoubtedly unusual, certain of the claims to uniqueness in this passage have been contested, for example by Brent Berlin [Current Anthropology 46.4 (2005): 635].

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

144

The Oulipo and Modern Thought

Perhaps the Pirahã have some numeric intuition but merely express it very differently?⁵² For Gordon, the Pirahã’s lack of a counting system might be read as evidence in support of the Whorfian hypothesis if it could be determined that this lack was not merely lexical but conceptual. He therefore undertakes a series of experiments with the Pirahã designed to test numeracy without the need for number words. One such experiment consists of ‘putting nuts into a can and withdrawing them one by one. [ . . . ] Participants responded after each withdrawal as to whether the can still contained nuts or was empty’.⁵³ The performance of the villagers in these tasks—some competence with quantities of up to three or four, with a marked tailing off thereafter—indicate to Gordon that not only do the Pirahã lack number words, they also lack the concept of numbers.⁵⁴ Having established this, Gordon infers a causal link between these two facts—it is the Pirahã’s lack of number words which precludes their numeracy—and cites this as ‘a rare and perhaps unique case for strong linguistic determinism’.⁵⁵ For Gordon and Everett, therefore, as for Mathews, there is no doubt that two natural languages may be incommensurate. Everett writes that ‘much of Pirahã is largely incommensurate with English and therefore translation is simply a poor approximation of Pirahã intentions and meaning, but we do as well as we can do’.⁵⁶ This final phrase is rather wonderful. Everett, the real-life field linguist, up against translation’s final frontier: the brick wall of incommensurability: ‘but we do as well as we can do’. I can’t go on. I’ll go on. It is this very conclusion that Botherby, the fictive linguist arrives at, faced with rendering the Uhas’ words in the Ohos’ language: ‘Botherby did not hesitate long. He saw, as you of course see, that he had no choice. There was only one solution. He grasped at once what all translators eventually learn: a language says what it can say, and that’s that.’⁵⁷ At a first glance, Mathews’s closing phrase—‘a language says what it can say, and that’s that’—might seem like a breezy gloss of ⁵² In a response to Everett’s paper, Michael Tomasello warns, ‘In my experience, what normally happens when proponents of universal grammar hear reports like Everett’s is that they simply do not believe them. The nonembedded Pirahã sentence structures reported, for example, really do have embedding, they will claim; it is just at an underlying level where we can’t see it’ [Current Anthropology 46.4 (2005): 640]. ⁵³ Gordon, ‘Numerical Cognition’, p. 497n. ⁵⁴ Here Borges’s fictional tribe, the Yahoos from ‘Brodie’s Report’, offer a surprising parallel. The missionary Brodie writes, ‘I will now say something about the witch doctors. I have mentioned that there are four of them; this number is the largest that the Yahoos’ arithmetic comprehends. They count on their fingers thus: one, two, three, four, many; infinity begins at the thumb’ [‘Brodie’s Report’, pp. 404–5]. ⁵⁵ Gordon, ‘Numerical Cognition’, p. 498. ⁵⁶ Everett, ‘Cultural Constraints’, p. 624n. ⁵⁷ Mathews, ‘Translation and the Oulipo’, p. 69.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Translation and Analytic Philosophy in the South Seas

145

Wittgenstein’s portentous ‘What we cannot speak about we must pass over in silence’.⁵⁸ Yet Mathews’s version does not prescribe silence. Botherby does not hesitate, but acts, and the process of translation, of communication between different cultures, proceeds even in the face of its own futility. Here not there must become Red makes wrong. It’s not right, but we do as well as we can do.

⁵⁸ Ludwig Wittgenstein, Tractatus Logico-Philosophicus, trans. by D. F. Pears and B. F. McGuinness (London: Routledge and Kegan Paul, 1961), §7, p. 151.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

Conclusion This book has attempted to show the Oulipo as being immersed in the intellectual currents of their time. Cybernetics, structuralism, psychoanalysis, translation theory: members of the Oulipo—and sometimes the group en masse—have been more than mere bystanders, offering up knowledgeable, though often arch and oblique, commentary. And yet, the book has barely ventured beyond the end of the 1970s, either in terms of the drifts of modern thought, or its consideration of the Oulipo, their archive and their output. The Oulipo are still going strong, but I sometimes wonder whether they don’t have something in common with another creative entity formed at the start of the sixties, a sort of Rolling Stones paradox where a group that’s been around for nearly sixty years can be globally bigger and more profitable than ever, yet their greatest hits—and their relevance—are many years behind them.¹ Is the Oulipo still the same type of intellectual entity as it was during the 1960s?² Is it, in fact, possible any more for a collective of any coherence to straddle the worlds of publishing and academia, arts and sciences, in the way that the early Oulipo did, or has that milieu become irrecoverably atomized? Without a doubt, the most ¹ I have written a little more on this elsewhere. See Dennis Duncan, ‘At the BnF’, LRB blog (3 February 2015), [accessed 17 Jan 2018]. For an extended discussion of this question of whether the Oulipo is what it once was, see Jean-Jacques Thomas, ‘Oulipo, qui as-tu tu?’, in Reggiani and Schaffner, Oulipo mode d’emploi, pp. 7–26. Bénabou’s defence, fending off overgeneralizations about what the group is or is not, is worth repeating here: ‘Something that needs to be clarified straight away: in the work of the Oulipo there are many levels’ [‘il y a quand même quelque chose qu’il faut préciser tout de suite. Dans le travail de l’Oulipo, il y a plusiers de niveaux’]. Marc Lapprand et al., ‘L’Oulipo et sa critque’, Formules 12 (2012): 229–46 (232). ² Bloomfield, for example, points out that it is no longer appropriate to describe a group as heavily institutionalized as the Oulipo as being avant-garde: ‘To describe the Oulipo as avant-garde would mean, aside from any structural consideration, going back to the 1960s and the first decade of its existence, and denying what it became in the 1980s.’ [‘Qualifier l’Oulipo d’avant-garde reviendrait, en dehors de toute considération structurelle, à le renvoyer aux années 1960 et à sa première décennie d’existence, et nier ce qu’il est devenu à partir des années 1980’]. Bloomfield, Raconter l’Oulipo, p. 59.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

Conclusion

147

important intellectual engagement for the later Oulipo has been with the related domains of feminism and the politics of identity. Whether this engagement has been extensive enough to be meaningful—to mollify a line of attack that equates the Oulipo’s declining relevance with a failure to be overtly political—remains to be seen. This Conclusion will merely sketch the major skirmishes in an encounter that is still playing out, one in which the group’s own account of itself is only one of a number of competing interpretations alongside the judgments of poetic and academic factions on both sides of the Atlantic. In 1986, Anne Garréta published her debut novel, Sphinx.³ Set in the demi-monde of the Parisian nightclub scene, it tells the story of the tempestuous love affair between a gifted but disillusioned theology student and a dancer. It is told in the first person, yet neither the narrator nor their lover is identified as male or female. In English this might not sound like much as a constraint—Mathews’s Tlooth, for example, withholds the narrator’s gender until the final pages without this being fundamental to the novel’s operation. In French, however, the constraint is a deeply restrictive one, of an order comparable with La Disparition’s lipogram in e. When the novel appeared in English, almost thirty years later, its translator, Emma Ramadan, provided a wonderful Translator’s Note describing how the constraint operates. ‘In French,’ she writes, the subject’s gender can be identified as soon as there is agreement with a verb in the past tense or with an adjective. [ . . . ] The constraint is in every sentence, every verb, every adjective of the French text. The entire narrative, almost every detail of the story and the style used to tell it, was shaped by the fact that there are no gender markers for the narrator or his or her lover.⁴

Ramadan outlines some of the ways in which Sphinx’s constraint determines its action. The narrator never just goes anywhere, for example, because the simplest way of expressing this—je suis allé for a man; je suis allée for a woman—gives away the gender of the person going. Then there is the use of tenses in a novel set in the past: while the passé composé might require gender agreement, the imparfait doesn’t, but it implies repeated or regular action. Hence, the narrator is always taking up habits, repetitive actions, to warrant this grammatical expedient. The constraint seeps into the novel’s characterization too: another tense that does not require gender agreement, the passé simple, has connotations of formality or perhaps slight pretentiousness, so the narrator’s social class needs to be marked as bourgeois so that it won’t jar when they use it. The narrator is thus ³ Anne Garréta, Sphinx (Paris: Grasset, 1986). ⁴ Anne Garréta, Sphinx, trans. by Emma Ramadan (Dallas: Deep Vellum, 2015), np.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

148

The Oulipo and Modern Thought

saddled with a semi-comic preciousness and self-absorption around their own diction, as in this passage when, having lived for a while in uptown New York, they observe a shift in their spoken English: My English still bears the stigmata from the time spent among an almost exclusively black community. Imperceptibly, the expressions and characteristic improprieties of their speech slipped into the tissue of the academic English I had been taught in high school. The language I speak is a monstrous hybrid, mingling Oxford and Harlem, Byron and gospel. To the point of caricature, I pronounce these African American utterances with a rather British accent, and sometimes swallow up to half of the syllables of a too perfectly constructed sentence.⁵

In terms of prose style, Garréta’s text often uses sentence fragments—a modernist technique, sign of disjointed interiority, but also a device for keeping gender at bay, denying the word which would force Garréta to reveal, to decide, what sex Sphinx’s main characters are. This is, of course, a polemic, an argument about the linguistic necessity of coming down on one side or the other. In its fiendishness, the constraint is designed to demonstrate—to the author; to the reader—the extent to which gender identity is imbricated with the French language, something that is unthinkingly declared in every almost sentence we speak. To do otherwise, to choose not to, requires an exhausting ingenuity. The novel never explicitly comments on its constraint; nevertheless, at moments, the narrator expresses their frustration at the inadequacy of language for giving voice to identity: It was an impossible task to set the boundaries of what I was, up to the edge where I blurred into the other – the indescribable other – so much did the meanings escape me, the words that others before me had uttered deep within an analogous attrition. I longed to reduce the impossible to the inessential, but I no longer had recourse to this principle of logic, which had suddenly become inadequate.⁶

‘What am I,’ they ask themself, ‘other than what you do not know how to say?’.⁷ Garréta, in interview, has stated that prior to being picked up by Grasset, Sphinx had been sent out to another publisher who, despite strong readers’ reports, had turned it down on the grounds that it was perverse: it was basically undermining or deconstructing a difference, or a binary, which that person held to be foundational to civilization, or culture. Now,

⁵ Garréta, Sphinx, trans. by Emma Ramadan, p. 64. ⁶ Garréta, Sphinx, trans. by Emma Ramadan, p. 86. ⁷ Garréta, Sphinx, trans. by Emma Ramadan, p. 89.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

Conclusion

149

thirty years later in the U.S. – after years and years of queer theory, deconstruction, Judith Butler, all sorts of things – there’s no scandal any longer. There’s just a strange experiment which validates the experience we have now of gender and sexuality.⁸

As the interview makes clear, Garréta sees the familiar theorizisations of language and gender as post hoc validations of the novel’s constraint rather than its motivating influences. Yet it is hard to see Sphinx as wholly sui generis and not to read it as a product of the same intellectual context as Butler and queer theory. Sphinx, surely, belongs in the tradition of the Oulipian experiments of the 1960s and 70s that we have looked at in the chapters preceding: an attempt to do creatively what others are simultaneously doing theoretically. After meeting Roubaud, Garréta was invited to the Oulipo’s monthly meeting of April 1994 where she outlined Sphinx’s ‘indécidabilité du sexe’.⁹ Six years later, in April 2000, she was co-opted into the group, making her only the third female member, after Michèle Métail (1975) and Michelle Grangaud (1995).¹⁰ That number rose to four when Valérie Beaudoin joined in 2003, and in March 2006, at the monthly ‘Jeudis de l’Oulipo’ reading at the BnF, for the first time Garréta, Grangaud, and Beaudoin were the only Oulipians onstage. Addressing the audience, Garréta raised the issue of an all-female Oulipo event: You are telling yourself: ‘But, où sont-ils? Where are they? Where is Marcel Bénabou? Where is Jacques Roubaud? Where is Jacques Jouet? Where is François Caradec? Where is Olivier Salon? Where is Ian Monk? Where is Hervé Le Tellier? Where is Frédéric Forte?’ You even wonder about the more rarely sighted Oulipians. Where are Harry Mathews, Paul Braffort, Bernard Cerquiglini? You must be telling yourself that this has to be some trick, some ploy, some new Oulipian invention, and that THEY are on their way.¹¹

What are the odds of an all-female panel of Oulipians? Garréta does the maths for her audience: ⁸ Sarah Gerard, ‘States of Desire: An Interview with Anne Garréta’, Paris Review: https://www.theparisreview.org/blog/2017/12/11/states-of-desire-an-interview-with-annegarreta/ [accessed 12 June 2018]. ⁹ ‘Compte rendu de la réunion du 28 avril’ (FO, DM-15 (4)). ¹⁰ In the same month that Garréta accepted her invitation to join the Oulipo, another female writer, Valérie Mréjen, declined the offer. ‘Compte-rendu de la réunion du vendredi 21 avril, chez MB’ (FO, DM-21 (5)). ¹¹ Anne Garréta, ‘Oulipian Moment for the End of Times’, Drunken Boat: http: //www.drunkenboat.com/db8/oulipo/feature-oulipo/oulipo/texts/garreta/times.html [accessed 12 June 2018].

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

150

The Oulipo and Modern Thought

Picture for a moment the Oulipo as a pack of cards. So many members, so many cards: at this point in history, an accumulated stack of 34 cards. From these 34, 13 are subtracted from a possible hand by the absurd accident of their death (we’ll just have to imagine them hidden in the sleeve of Potentiality itself). In the remaining stack of 21 Oulipian cards, there are exactly 4 feminine figures (I’ll call them by their french name, Dames: the civic republican and radical egalitarian strain in me objects to ‘Queens’). Our little mathematical problem can be summed up in the following way: what is the probability, in this monthly poker game of a public reading, of being dealt a three-of-a-kind hand comprised of dames, and only dames? Answer: 0.003.

Three in a thousand. The odds of an all-male panel, meanwhile, are better than fifty–fifty. The gender bias of the Oulipo—the Oulipo as a boys’ club—could not be starker.¹² On the other hand, perhaps it should come as little surprise. As Camille Bloomfield has pointed out, with the notable exception of Tel Quel, the avant-gardes of the French twentieth-century were all predominantly male groupings, with often the only women being girlfriends of the leading men: The question of gender might also be posed here: with these groups and movements being, in the vast majority, male, where are the women? Are they, as is often the case, the girlfriends of the important figures? Or are they creators and theoreticians and co-opted as such?¹³

Conceivably, the Oulipo could alter the balance of its make-up. In 2005, Ian Monk could write that ‘more women have been chosen in the last few years than in the history of the group’, although as Garréta would point out on the BnF stage a few months later the numbers hardly represent a seismic shift.¹⁴ Even so, another question remains: is there something fundamentally gendered—and fundamentally masculine—about the work produced under Oulipian conditions? As Christelle Reggiani asks: ‘Can there (really) be such a thing as an Oulipian writing that is female

¹² Despite the involvement of Grangaud and Audin, the title of the group’s collective publication in 2010, C’est un métier d’homme—It’s a Man’s Job—was a provocative one. Oulipo, C’est un métier d’homme: Autoportraits d’hommes et de femmes au repos (Paris: Mille et une nuits, 2010). ¹³ [‘Le question du genre peut aussi être posée ici, dans des groupes et des mouvementes à grande majorité masculine: qui sont les femmes? Sont-elles, commes c’est souvent le cas, les compagnes des membres importants? Ou bien sont-elle créatrices ou théoreticiennes, et “cooptées” comme telles?’]. Camille Bloomfield, Raconter l’Oulipo, p. 65. ¹⁴ Ian Monk, ‘My Life with the Oulipians’, in Matias Viegener and Christine Wertheim (eds), The nOulipian Analects (Los Angeles: Les Figues, 2007), pp. 140–2 (p. 140).

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

Conclusion

151

gendered?’¹⁵ This is, without doubt, the hot button question for the Oulipo of the twenty-first century. And the answer depends on who you ask. NOULIPO/FOULIPO In an article for the London Review of Books in 2012, Stephanie Burt noted that ‘US and Canadian writers seem to be on an Oulipo kick: Google “noulipo”, i.e. new Oulipo, for proof ’.¹⁶ What you will find, however, if you follow Burt’s suggestion, is not a trove of constrained writing by American authors, but rather a series of ripples emanating from a conference held in Los Angeles at CalArts’ REDCAT theatre in the autumn of 2005. The conference brought together a number of Oulipians—Fournel, Mathews, and Ian Monk—alongside what the programme described as ‘a host of American, Canadian, and English writers influenced by them in varying degrees’, among them Caroline Bergvall, Johanna Drucker, Doug Nufer, and Christian Bök.¹⁷ Looking back on their rationale in putting the conference together, the organizers would explain: we were [ . . . ] interested in the way formalist practice can be used to investigate questions of politics, including issues of race, class and gender. From this perspective, the n in n/Oulipo stands for now, the new, and the not-only-formal.¹⁸

Far from being merely a love-in between the Oulipians and their anglophone inheritors, this emphasis on the political would generate a considerable amount of friction. For a start, there was the issue of ‘the “true”, that is, formally elected Oulipeans [sic] invited from France, sometimes acting as gatekeepers, passing judgments about whose work was genuinely constraint-based and whose was not’.¹⁹ The most significant event of the conference, however, was a performance that questioned the gender politics of formal writing per se. On the second afternoon, in a panel entitled ‘The Politics of Constraint’, the poets Juliana Spahr and Stephanie Young presented a ¹⁵ [‘L’écriture oulipienne peut-elle être (vraiment) de genre feminine?’]. Christelle Reggiani, ‘Être Oulipienne: contraintes de style, contraintes de genre?’, Études littéraires, 47.2 (2016), 103–17 (p. 103). ¹⁶ Stephanie Burt, ‘Must Poets Write?’, London Review of Books 34.9 (2012): 33–4 (33). ¹⁷ ‘noulipo Experimental Writing Conference’ https://www.redcat.org/event/noulipo [accessed 17 June 2018]. ¹⁸ Matias Viegener and Christine Wertheim, ‘ “Séance”, “n/Oulipo”, “Impunities”, “Feminaissance”, & “Untitled” ’, Jacket 2: https://jacket2.org/commentary/seance-noulipoimpunities-feminaissance-untitled [accessed 17 June 2018]. ¹⁹ Viegener and Wertheim, “Séance”.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

152

The Oulipo and Modern Thought

collaborative work entitled ‘foulipo’. The pair began by taking it in turns to recite sentences from a scripted essay from which each instance of the letter r had been removed (so, for example, the piece begins: ‘One day we wee talking about wok fom the 70s’).²⁰ Describing the effect, one reviewer recalls that ‘it almost sounded like Elmer Fudd-speak, or as one audience member put it, baby-talk’.²¹ After a while, however, the performers’ voices were replaced by recordings—instead of speaking, Spahr and Young began to remove their clothes, disrobing and getting dressed again, then repeating the process a further two times. The performance, as the essay explained, was intended to hark back to the body artists of the 1970s: all that body pefomance wok that suddenly began to happen, all at once, wok that was obvious and ovet and even a little easy, such as when Shigeko Kubota did he Vagina Painting, whee she squatted down and painted with a bush attached to the cotch of he undewea.²²

Spahr and Young drew a parallel between this type of work, focused around the body and practised by artists such as Kubota and Marina Abramovic, and the Oulipian practices that were beginning to emerge publicly in the early 1970s: And then we had one of those isn’t it inteesting moments whee we said, isn’t it inteesting that all that body wok happens at the same time as the development of all that Oulipo wok. Isn’t it inteesting that Kubota and Abramovic and Schneeman and Antin and Ukeles and Valie Export got on a subway with the cotch cut out of he pants caying a machine gun was the same time that Oulipo was holding its meetings. And then we wondeed what did that schism mean to us, witing thirty years later?²³

The schism that the pair have in mind takes a number of forms: a praxis focused on the body versus the intellectual focus of constrained writing; a type of art now considered passé or of-its-time versus one that has remained influential into the new millennium; and, crucially, a mode dominated by women artists versus one that, in the authors’ experience, was an exclusively male preserve. Speaking of their backgrounds in teaching and attending poetry workshops, Spahr and Young asked: isn’t it inteesting how we can think of no instance when a woman has bought in wok using a constictive composition device to any of these wokshops and ²⁰ Juliana Spahr and Stephanie Young, ‘ “& and” foulipo’, in Viegener and Wertheim, nOulipian Analects, pp. 3–13 (p. 3). ²¹ Joseph Mosconi, ‘Politics of Constraint: The Panel’, in Viegener and Wertheim, nOulipian Analects, pp. 165–8 (p. 165). ²² Spahr and Young, ‘foulipo’, p. 3. ²³ Spahr and Young, ‘foulipo’, p. 8.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

Conclusion

153

yet we can think of men who did it week afte week and called themselves adicals fo it.²⁴

And it was not just the practitioners who come under fire, as Spahr and Young wondered whether there was something inherently negative or destructive in constrained writing as a mode: ‘Pat of ou poblem with the estictive, numbe-based pocesses and constaints was that they [ . . . ] tended to mock athe than build. They tended to invade and cut down athe than connect’.²⁵ At last, the Oulipo themselves came under explicit attack: for their lack of women, the strictness of their membership policy, the disrespectful body politics of procedures with names like ‘mother-inlaw’, ‘slenderizing’, ‘asphyxiation’. Taking ‘politics’ as something coherent and fundamentally progressive, the Oulipo, they concluded, was irredeemably apolitical: And then Mateis and Christine emailed with thei question fo this panel about if thee could be a politics to Oulipo and ou fist answe was depends. But afte thinking about it some moe, ou second answe was no, not eally diectly; Oulipo is political only in the way that anything has a politics, but othewise, no.²⁶

Instead, Spahr and Young suggested a new collective—‘a sot of feminist Oulipo’: the ‘foulipo’ of their title.²⁷ As the presenters reclothed themselves on the REDCAT stage for the last time, their recorded voices slurring through their treated text, the performance ended with a vision in which the earlier schism is healed, a coming together of bodily and textual art: ‘We just wanted something that engaged the elation between fomalism and body at and saw both as pat of a tadition that was complicated and inteconnected.’²⁸ Foulipo—a feminist Oulipo—would need to go beyond mere textual practice. Introducing the conference proceedings two years later, the organizers—perhaps with the benefit of hindsight—suggest that a note of scepticism towards the original group was always implicit in their portmanteau term, noulipo: The n indexes both the ‘n’ of mathematics – as in ‘n + 1 = x’ – and the famous constraint N + 7 [ . . . ]; it also echoes the English new, the French nous, which means us, and the Yiddish nu, which means so? or well? and can be used by itself to mean what’s new? ²⁹ ²⁴ Spahr and Young, ‘foulipo’, p. 7. ²⁵ Spahr and Young, ‘foulipo’, p. 9. ²⁶ Spahr and Young, ‘foulipo’, p. 10. ²⁷ Spahr and Young, ‘foulipo’, p. 11. ²⁸ Spahr and Young, ‘foulipo’, p. 12. ²⁹ Matias Viegener and Christine Wertheim, ‘Introduction’, in Viegener and Wertheim, nOulipian Analects, pp. 103–7 (p. 103).

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

154

The Oulipo and Modern Thought

So? Well? What can the Oulipo offer if engagement is spelled out in terms of the politics of identity, of inclusivity? Although they sometimes allow the categories to blur, for the most part Spahr and Young are making two distinct critiques. Firstly, the obvious factual one—the one that would be spelled out again by another group of women on another stage at the BnF a few months later—that the Oulipo continues to be, by and large, a group of men. Secondly, the more complex suggestion that Oulipian poetics is essentially gendered, that the constraints and procedures that are the group’s stock-in-trade are incapable of producing what might be identified as an écriture feminine. (Spahr and Young’s performance was subsequently dismissed by Fournel, who witnessed it, as essentially unOulipian: ‘the young ladies who took their underwear off – which I found absolutely charming – is only very distantly related to the Oulipo. We don’t do that kind of thing in the Oulipo’.)³⁰ Thinking back to Reggiani’s question—‘Can there (really) be such a thing as an Oulipian writing that is female gendered?’—and judging by much of the response to Spahr and Young, it would appear that in US academia there is a lingering sense that this cannot be answered in the affirmative.³¹ French scholarship, led by Reggiani herself, however, has been more conciliatory. Reggiani points to both Garréta’s Sphinx and to Olivier Salon’s Chaque porche est une invitation au voyage (BO 170) in which pairs of French homophones—one grammatically feminine, the other masculine—are used as the basis for microfictions whose punning punchlines work across gender. Reggiani concludes that ‘[c]urrent Oulipian writing, mostly by men thanks to the composition of the group, has in sum chosen to contest, by various means, the grammatical marking of gender’.³² This feels like an overstatement, but it does imply that Oulipian writing is not a priori a masculine mode. Meanwhile, a more direct riposte to Spahr and Young might be to look closely at the profound corporeality of Garréta’s Sphinx, a novel which, for all its grammatical ingenuity, is also consistently fascinated and repulsed by the body, ³⁰ [‘Les demoiselles qui quittaient leur petite culotte – ce que j’ai trouvé tout à fait charmant – ça ne ressemblait que de très loin à l’Oulipo. Nous ne faisons pas ce genre de chose à l’Oulipo.’] Camille Bloomfield et al., ‘Généalogiques oulipiennes’, in Reggiani and Schaffner, Oulipo mode d’emploi, pp. 115–36 (p. 128). ³¹ See also Lauren Elkin’s polemic, ‘Oulipo Lite’, in Lauren Elkin and Scott Esposito, The End of Oulipo?: An Attempt to Exhaust a Movement (Winchester: Zero Books, 2011), pp. 66–101. ³² [‘Les écritures oulipiennes contemporaines, majoritairement masculines en raison de la composition du groupe, ont en somme choisi de contester, par diverses voies, le marquage grammatical du genre’]. Christelle Reggiani, ‘Masculin/Féminin: L’écriture oulipienne a-t-elle un genre?’, in Christelle Reggiani and Alain Schaffner, Oulipo mode d’emploi (Paris: Honoré Champion, 2017), pp. 79–86 (p. 85).

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

Conclusion

155

exhibiting, as the narrator puts it, ‘a kind of disgust at the abrupt resurgence of raw, cannibalistic reality’.³³ Sphinx, surely, is pure foulipo as much as it is Oulipo. For myself, however, I wonder how much these terms—Oulipo, foulipo, noulipo—matter now. I wonder whether the familiarity of the Oulipo, which has grown steadily since the group’s public emergence in the early 1970s—their désoccultation, to poach a word from the lexicon of the Collège de ’Pataphysique—is not precisely the reason why the concept rather than the group has more relevance today. Through decades of intermingled graft and play—through monthly Creation, Rumination, Erudition, and Action—the early group worked to produce a recognizable sense of the Oulipian (and, by the nifty semantics of ‘plagiarism by anticipation’, co-opted anything that was Oulipian avant la lettre). But this adjectival form—let’s give it a small o: oulipian—seems nowadays to belong to everyone and nobody, used to describe a type of writing regardless of where it originated. This is not to say that the Oulipo themselves endorse this loosening of the term. In a rather wonderful analogy with the protected status of wine regions, Camille Bloomfield states that ‘It must be understood that, for members of the group, the term oulipian is an appellation contrôlée’.³⁴ Nevertheless, just as surreal long ago slipped the control of Breton, Dalí et al., it seems undeniable that oulipian has become public property, part of the modern critical word-hoard. As such, it is inevitable that the big-O Oulipian should lose something in inverse proportion. In a way, this was always the plan: to develop new structures that writers (themselves; others) might implement. But some writers have gone further, not just adopting Oulipian structures, but inventing oulipian structures of their own: the Canadian poet Christian Bök, whose extraordinary, univocalic novel Eunoia draws heavily on Perec’s Les Revenentes but pulls it, exhaustively, in new directions; the Mexican artist Daniela Franco whose hoax work (often in collaboration with members of the Oulipo) has the impishness of the Collège de ’Pataphysique in its heyday; the British writer Joanna Walsh in whose work the oulipian, as well as a specifically Perecquian attention to the infra-ordinary, are running overtones without ever becoming keynotes. This feels like an odd thing to write having spent so long in the archive poring over the meeting minutes of a small, defined group of men, but today it seems only a technicality to categorize works into those which come with the kitemark of the Oulipo itself and those produced by others ³³ Garréta, Sphinx, trans. by Emma Ramadan, pp. 109–10. ³⁴ [‘Il faut aussi comprendre que l’appellation “oulipien” est pour les membres du groupe une “appellation contrôlée” ’.] Bloomfield, Raconter l’Oulipo, p. 55.

OUP CORRECTED PROOF – FINAL, 8/6/2019, SPi

156

The Oulipo and Modern Thought

working in that vein. Perhaps the milieu of Paris in the early 1960s is irrecuperably lost, but the small-o oulipian—writing in the constrained mode—is, in one important sense, unconstrained: it can come from anywhere. Will it continue to be used as a means of engaging with the drifts and controversies of modern thought? That remains to be seen. But, as I hope I have shown, it has potential.

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Bibliography Abraham, Nicolas, and Mária Török, ‘Introjecter-Incorporer. Deuil ou mélancolie’, Nouvelle revue de psychanalyse 6 (1972): 111–22 Abraham, Nicolas, and Mária Török, The Wolf Man’s Magic Word: A Cryptonymy, trans. by Nicholas Rand (Minneapolis, MN: University of Minnesota Press, 1986) Abraham, Nicolas, and Mária Török, ‘Mourning or Melancholia: Introjection versus Incorporation’, in The Shell and the Kernel: Renewals of Psychoanalysis, trans. by Nicholas T. Rand (Chicago, IL: University of Chicago Press, 1994), pp. 125–38 Adorno, Theodor, ‘Looking Back on Surrealism’, in Notes to Literature, ed. by Rolf Tiedemann, trans. by Shierry Weber Nicholsen, 2 vols (New York: Columbia University Press, 1991), I, pp. 86–90 Andrews, Chris, ‘Inspiration and the Oulipo’, Studies in 20th & 21st Century Literature 29.1 (2005): 9–28 Apter, Emily, The Translation Zone: A New Comparative Literature (Princeton, NJ: Princeton University Press, 2006) Arnaud, Noël, ‘Athalie Galant?’, Dossiers du Collège de ’Pataphysique 17 (1961): 33 Arnaud, Noël, ‘Politique et polémique dans les romans de Raymond Queneau’, in Queneau aujourd’hui: actes du Colloque Raymond Queneau, Université de Limoges, mars 1984 (Paris: Clancier-Guénaud, 1985), pp. 113–57 Arnaud, Noël,‘Foreword: Prolegomena to a Fourth Oulipo Manifesto – or Not’, in Motte, Oulipo, pp. xi–xv Aubin, David, ‘The Withering Immortality of Nicolas Bourbaki’, Science in Context 10.2 (1997): 297–342 Austin, J. L., ‘The Meaning of a Word’, in Philosophical Papers (Oxford: Clarendon, 1961), pp. 23–43 Bacon, Francis, The Advancement of Learning, ed. by Michael Kiernan (Oxford: Clarendon, 2000) Barthes, Roland, ‘Introduction à l’analyse structural des récits’, Communications 8 (1966): 1–27 Barthes, Roland, ‘Science versus Literature’, TLS #3422 (28 September 1967): 897–8 Barthes, Roland, ‘An Introduction to the Structural Analysis of Narrative’, trans. by Lionel Duisit, New Literary History 6.2 (1975): 237–72 Barthes, Roland, ‘The Death of the Author’, in The Rustle of Language, trans. by Richard Howard (Oxford: Blackwell, 1986), pp. 49–55 Beckett, Samuel, Murphy (London: John Calder, 1963) Bellos, David, Georges Perec: A Life in Words (London: Harvill, 1993)

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

158

Bibliography

Bénabou, Marcel, ‘La Galère ou Pourquoi j’ai participé à la confection du volume intitulé 789 néologismes de Jacques Lacan’, in L’amour de loin du docteur L. (Paris: L’Unebévue, 2004), pp. 27–31 Bénabou, Marcel, ‘Rule and Constraint’, in Motte,Oulipo, pp. 40–7 Bens, Jacques, Genèse de l’Oulipo (1960–63) (Bordeaux: Castor Astral, 2005) Bloomfield, Camille, ‘L’Oulipo: histoire et sociologie d’un groupe-monde’ (PhD, Paris 8, 2011) Bloomfield, Camille, ‘Les manifestes à l’Oulipo: La disparition d’une forme’, Études littéraires 44.3 (2013): 35–46 Bloomfield, Camille, Raconter l’Oulipo (1960–2000): histoire et sociologie d’un groupe (Paris: Honoré Champion, 2017) Bloomfield, Camille, ‘Histoire de l’Oulipo: Quelques jalons chronologique’, in Bloomfield and Lesage, Oulipo, pp. 29–38 Bloomfield, Camille, and Hélène Campaignolle, ‘L’Oulipo et l’informatique’, in Reggiani and Schaffner, Oulipo mode d’emploi, pp. 319–36 Bloomfield, Camille, and Claire Lesage (eds), Oulipo (Paris: BnF/Gallimard, 2014) Bloomfield, Camille, et al., ‘Généalogiques oulipiennes’, Formules 16 (2012): 115–36 Bonner, Anthony (ed.), Selected Works of Ramon Llull (1232–1316) (Princeton, NJ: Princeton University Press, 1985) Bonner, Anthony, The Art and Logic of Ramon Llull: A User’s Guide (Leiden: Brill, 2007) Borges, Jorge Luis, El oro de los tigres (Buenos Aires: Emecé, 1972) Borges, Jorge Luis, ‘The Garden of Forking Paths’, in Collected Fictions, trans. by Andrew Hurley (London: Penguin, 1998), pp. 119–28 Borges, Jorge Luis, ‘Ramón Llull’s Thinking Machine’, trans. by Esther Allen, in The Total Library: Non-Fiction 1922–86 (London: Penguin, 1999), pp. 155–9 Borges, Jorge Luis, ‘Brodie’s Report’, in Collected Fictions, pp. 402–8 Borges, Jorge Luis, ‘John Wilkins’ Analytical Language’, trans. by Eliot Weinberger, in The Total Library, pp. 229–32 Botta, Anna, ‘Calvino and the Oulipo: An Italian Ghost in the Combinatory Machine?’, MLN 112.1 (1997): 81–9 Breton, André, ‘Manifesto of Surrealism (1924)’, in Manifestoes of Surrealism, trans. by Richard Seaver and Helen R. Lane (Ann Arbor, MI: University of Michigan Press, 1972), pp. 1–47 Breton, André, ‘Introduction to the Discourse on the Paucity of Reality’, trans. by Richard Sieburth and Jennifer Gordon, October 69 (1994): 133–44 Breton, André, Anthology of Black Humor, trans. by Mark Polizzotti (San Francisco, CA: City Lights, 1997) Brisset, Jean-Pierre, La Science de Dieu ou la Création de l’Homme, in Œuvres complètes, ed. by Marc Décimo (Dijon: Presses du réel, 2001), pp. 697–885 Brisset, Jean-Pierre, Les Origines humaines, in Œuvres complètes, pp. 1117–315 Brotchie, Alastair (ed.), A True History of the College of ’Pataphysics: With Manifestos, Statutes, Calendar and Documents, trans. by Paul Edwards (London: Atlas, 1995)

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Bibliography

159

Brotchie, Alastair, Alfred Jarry: A Pataphysical Life (Cambridge, MA: MIT Press, 2011) Burgelin, Claude, Les Parties de dominos chez Monsieur Lefèvre: Perec avec Freud— Perec contre Freud (Belval: Circé, 1996) Burgelin, Claude, et al., La Cantatrice Sauve, in Oulipo, Bibliothèque oulipienne (Paris: Ramsay, 1987), I, pp. 305–22 Burgin, Richard (ed.), Jorge Luis Borges: Conversations ( Jackson, MS: University Press of Mississippi, 1998) Burt, Stephanie, ‘Must Poets Write?’, London Review of Books 34.9 (2012): 33–4 Calvino, Italo, Il castello dei destini incrociati (Turin: Einaudi, 1973) Calvino, Italo, The Castle of Crossed Destinies, trans. by William Weaver (London: Secker and Warburg, 1977) Calvino, Italo, ‘Cybernetics and Ghosts’, in The Uses of Literature: Essays, trans. by Patrick Creagh (San Diego, CA: Harcourt Brace Jovanovich, 1986), pp. 3–27 Calvino, Italo, Six Memos for the Next Millennium, trans. by Patrick Creagh (London: Cape, 1992) Calvino, Italo, ‘The Burning of the Abominable House’, in Numbers in the Dark and Other Stories, trans. by Tim Parks (London: Jonathan Cape, 1995), pp. 156–69 Calvino, Italo, Letters, 1941–85, ed. by Michael Wood, trans. by Martin McLaughlin (Princeton, NJ: Princeton University Press, 2013) Calvino, Italo, Piccolo sillabario illustrato, in Oulipo, Bibliothèque oulipienne, I, pp. 97–121 Calvino, Italo, ‘Prose and Anticombinatorics’, in Motte,Oulipo, pp. 143–52 Chabanne, Jean-Charles, ‘Queneau, Les Temps modernes, Sartre’, in Temps mêlés—Documents Queneau, 150+33–36 (1987): 355–61 Charbonnier, Georges, Entretiens avec Raymond Queneau (Paris: Gallimard, 1962) Chomsky, Noam, and Morris Halle, The Sound Pattern of English (New York: Harper & Row, 1968) Comenius, Johann Amos, The Way of Light, trans. by E. T. Campagnac (London: Hodder & Stoughton, 1938) Couturat, Louis, La Logique de Leibniz d’après des documents inédits (Paris: Germer Baillière, 1901) Dahl, Roald, ‘The Great Automatic Grammatizator’, in Someone Like You (New York: Knopf, 1953) Décimo, Marc, Jean-Pierre Brisset, Prince des penseurs (Paris: Ramsay, 1986) Derrida, Jacques, ‘My Chances/Mes Chances: A Rendezvous with Some Epicurean Stereophonies’, trans. by Irene Harvey and Avital Ronell, in Taking Chances: Derrida, Psychoanalysis and Literature, ed. by Joseph H. Smith and William Kerrigan, (Baltimore, MD: Johns Hopkins University Press, 1984), pp. 1–32 Dixon, Robert, Where Have All the Adjectives Gone? and Other Essays in Semantics and Syntax (Amsterdam: Mouton, 1982) Duchateau, Jacques, ‘Communication, à Cerisy, sur l’Oulipo’, Temps mêlés 66–7 (1964): 13–21

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

160

Bibliography

Ducrocq, Albert, L’Ère des robots (Paris: Julliard, 1953) Duncan, Dennis, ‘Calvino, Llull, Lucretius: Two Models of Literary Combinatorics’, Comparative Literature 64.1 (2012): 93–109 Duncan, Dennis ‘At the BnF’, London Review of Books (3 February 2015),

Duncan, Dennis, ‘ “Joyce, un pornographe”: Ulysses, A Portrait of the Artist as a Young Man, and the Sally Mara Novels of Raymond Queneau’, James Joyce Quarterly 52.2 (2015): 115–32 Duncan, Dennis (ed. and trans.), Theory of the Great Game: Writings from Le Grand Jeu (London: Atlas, 2015) Eco, Umberto, ‘James Bond: Une combinatoire narrative’, Communications 8 (1966): 77–93 Eco, Umberto, Interpretation and Overinterpretation (Cambridge: Cambridge University Press, 1992) Eco, Umberto, The Search for the Perfect Language, trans. by James Fentress (Oxford: Blackwell, 1995) Elkin, Lauren, ‘Oulipo Lite’, in Lauren Elkin and Scott Esposito, The End of Oulipo?: An Attempt to Exhaust a Movement (Winchester: Zero Books, 2011), pp. 66–101 Everett, Daniel, ‘Cultural Constraints on Grammar and Cognition in Pirahã: Another Look at the Design Features of Human Language’, Current Anthropology 46.4 (2005): 621–46 Ford, Mark, Raymond Roussel and the Republic of Dreams (London: Faber and Faber, 2000) Foucault, Michel, ‘Un si cruel savoir’, Critique 182 (1962): 597–611 Foucault, Michel, Death and the Labyrinth: The World of Raymond Roussel, trans. by C. Ruas (London: Continuum, 1986) Foucault, Michel, ‘Seven Remarks on the Seventh Angel’, trans. by Dennis Duncan, PMLA 132.5 (2017): 1251–62 Fournel, Paul, Clefs pour la littérature potentielle (Paris: Lettres Nouvelles, 1972) Freud, Sigmund, ‘The Psychical Mechanism of Forgetfulness’ (1898), in Early Psycho-Analytic Publications, Standard Edition of the Complete Psychological Works of Sigmund Freud, ed. by James Strachey, trans. by James Strachey et al., 26 vols. (London: Hogarth and the Institute of Psycho-Analysis, 1953–66) III, pp. 287–97 Freud, Sigmund, The Interpretation of Dreams (1899), Standard Edition, IV Gardner, Martin, Logic Machines and Diagrams, 2nd edn (Brighton: Harvester Press, 1983) Garréta, Anne, Sphinx (Paris: Grasset, 1986) Garréta, Anne, Sphinx, trans. by Emma Ramadan (Dallas, TX: Deep Vellum, 2015) Garréta, Anne, ‘Oulipian Moment for the End of Times’, Drunken Boat : http:// www.drunkenboat.com/db8/oulipo/feature-oulipo/oulipo/texts/garreta/times. html Géhéniau, Florence, Queneau analphabète: répertoire alphabétique de ses lectures de 1917 à 1976, 2nd edn (Brussels: F. Géhéniau, 1992)

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Bibliography

161

Geoghegan, Bernard Dionysius, ‘From Information Theory to French Theory: Jakobson, Lévi-Strauss, and the Cybernetic Apparatus’, Critical Inquiry 38.1 (2011): 96–126 Gerard, Sarah, ‘States of Desire: An Interview with Anne Garréta’, Paris Review: https://www.theparisreview.org/blog/2017/12/11/states-of-desirean-interview-with-anne-garreta/ Gordon, Peter, ‘Numerical Cognition Without Words: Evidence from Amazonia’, Science 306 (2004): 496–9 Hugo, Victor, Toute la lyre (Paris: Hetzel & Quantin, 1888) James, Alison, Constraining Chance: Georges Perec and the Oulipo (Evanston, IL: Northwestern University Press, 2009) Jefferson, Geoffrey, ‘The Mind of Mechanical Man’, British Medical Journal #4616 (25 June 1949): 1105–10 Johnson, Christopher, ‘ “French” Cybernetics’, French Studies 69.1 (2015): 60–78 Jones, Ernest, The Life and Work of Sigmund Freud, ed. by Lionel Trilling and Steven Marcus (London: Hogarth, 1961) Josipovici, Gabriel, ‘Georges Perec’s Homage to Joyce (and Tradition)’, Yearbook of English Studies 15 (1985): 179–200 Kojève, Alexandre, Introduction à la lecture de Hegel, ed. by Raymond Queneau (Paris: Gallimard, 1947) Kuon, Peter, ‘L’Oulipo et les avant-gardes’, in Oulipo-poétiques, ed. by Peter Kuon (Tübingen: G. Narr, 1999), pp. 15–30 Lacan, Jacques, ‘Seminar on “The Purloined Letter” ’, trans. by Jeffrey Mehlman, Yale French Studies 48 (1972): 38–72 Lacan, Jacques, The Four Fundamental Concepts of Psycho-Analysis, ed. by JacquesAlain Miller, trans. by Alan Sheridan (Harmondsworth: Penguin, 1979) Lacan, Jacques, The Seminar of Jacques Lacan, Book II: The Ego in Freud’s Theory and in the Technique of Psychoanalysis, 1954–1955, ed. by Jacques-Alain Miller, trans. by Sylvana Tomaselli (New York: Norton, 1988) Lacan, Jacques, ‘The Instance of the Letter in the Unconscious or Reason since Freud’, in Écrits: A Selection, trans. by Bruce Fink (New York: Norton, 2002), pp. 138–68 Lakoff, George, Women, Fire, and Dangerous Things: What Categories Reveal About the Mind (Chicago, IL: University of Chicago Press, 1987) Laplanche, Jean, and Serge Leclaire, ‘The Unconscious: A Psychoanalytic Study’, trans. by Patrick Coleman, Yale French Studies 48 (1972): 118–75 Lapprand, Marc, et al., ‘L’Oulipo et sa critque’, Formules 12 (2012): 229–46 Latis,‘The Collège de ’Pataphysique and the Oulipo’, in Motte, Oulipo, pp. 48–50 Latrazus, Louis, ‘La Victime’, Le Figaro, 14 April 1913: 4 Le Clec’h, Guy, ‘Introduction’, Temps mêlés, 66–67 (1964): i–iv Le Lionnais, François (ed.), Les Grands courants de la pensée mathématique (Marseilles: Cahiers du Sud, 1948) Le Lionnais, François, ‘Pensée Artificielle, Pensée Vécue’, Temps mêlés 66–67 (1964): 34–6

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

162

Bibliography

Le Lionnais, François, ‘Les structures du roman policier: Qui est coupable?’, Subsidia Pataphysica 15 (1972): 16–18 Le Lionnais, François, ‘Painting at Dora’, trans. by Daniel Levin Becker, The Believer 11.5 (2013): 27–31 Le Lionnais, François, Un Certain disparate, ed. by Olivier Sxalon and Anne Garréta, Le Lionnais, François, ‘Le Second manifeste’, in Oulipo, La littérature poentielle, pp. 19–23 Le Lionnais, François, ‘Raymond Queneau and the Amalgam of Mathematics and Literature’, in Motte,Oulipo, pp. 74–8 Le Lionnais, François, ‘Second Manifesto’, in Motte,Oulipo, pp. 29–31 Le Roux, Ronan, ‘Structuralisme(s) et cybernétique(s): Lévi-Strauss, Lacan et les mathématiciens’, Les Dossiers d’HEL 3 (2013) Le Tellier, Hervé, Esthétique de l’Oulipo (Bordeaux: Castor Astral, 2006) Le Tellier, Hervé, ‘Bookforum Interviews Hervé Le Tellier’, Lécureur, Michel, Raymond Queneau (Paris: Belles Lettres, 2002) Lemaire, Anika, Jacques Lacan, trans. by David Macey (London: Routledge & Kegan Paul, 1977) Lesage, Claire, ‘L’archive, miroir de l’Oulipo’, in Bloomfield and Lesage, Oulipo, pp. 66–8 Lescure, Jean, ‘La Méthode S + 7’, Dossiers du Collège de ’Pataphysique 17 (1961): 17–19 Lescure, Jean, ‘Petite histoire pour un tri-centenaire’, Temps mêlés, 66–7 (1964): 3–12 Lescure, Jean, ‘Brief History of the Oulipo’, in Motte,Oulipo, pp. 32–9 Lescure, Jean, ‘Des permuations en particulier et en général des poèmes carrés’, in Oulipo, La Littérature potentielle, pp. 151–61 Lomas, David, The Haunted Self : Surrealism, Psychoanalysis, Subjectivity (New Haven, CT: Yale University Press, 2000) Lou l’Hippo, ‘L’Amer Dataire’, Le Petit Moniteur de l’ACACADOOR 1 (1961): 6 Lucretius, De rerum natura, 3rd edn, trans. by W. H. D. Rouse (Cambridge, MA: Harvard University Press, 1975) Lui, Lydia H., ‘The Cybernetic Unconscious: Rethinking Lacan, Poe, and French Theory’, Critical Inquiry 36 (2010): 288–320 Mailhos, Jacques, ‘The Art of Memory; Joyce and Perec’, in Transcultural Joyce, ed. by Karen Lawrence (Cambridge: Cambridge University Press, 1998), pp. 151–69 Masson, Jeffrey M. (ed. and trans.), The Complete Letters of Sigmund Freud to Wilhelm Fliess, 1887–1904 (Cambridge, MA and London: Belknap, 1985) Mathews, Harry, The Conversions (New York: Random House, 1962) Mathews, Harry, ‘Remarks of the Scholar Graduate’, Best & Company 1 (1969): 35–41 Mathews, Harry, ‘Abanika, traditore’, trans. by Marie Chaix, L’Arc 76 (1979): 73–6

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Bibliography

163

Mathews, Harry, ‘The Dialect of the Tribe’, SubStance 27 (1980): 52–5 Mathews, Harry, ‘Still Life’, Littératures 7 (1983): 151–4 Mathews, Harry, The Sinking of the Odradek Stadium (London: Paladin, 1986) Mathews, Harry, The Orchard: A Remembrance of Georges Perec (Flint, MI: Bamberger, 1988) Mathews, Harry, ‘Autobiography’, in The Way Home: Collected Longer Prose (London: Atlas, 1989), pp. 111–65 Mathews, Harry ‘Translation and the Oulipo: The Case of the Persevering Maltese’, Brick 57 (1997): 39–44 Mathews, Harry, ‘La boule de neige’, Magazine Littéraire (May 2001): 28 Mathews, Harry, ‘Fearful Symmetries’, The Literary Review 45.3 (2002): 444–52 Mathews, Harry, The Human Country: New and Collected Stories (Normal, IL: Dalkey Archive, 2002) Mathews, Harry, The Case of the Persevering Maltese: Collected Essays (Normal, IL: Dalkey Archive, 2003) Mathews, Harry, ‘The Dialect of the Tribe’, in The Human Country, pp. 7–14 Mathews, Harry, ‘Fearful Symmetries’, in The Case of the Persevering Maltese, pp. 55–65 Mathews, Harry, ‘Remarks of the Scholar Graduate’, in The Human Country, pp. 41–50 Mathews, Harry ‘Still Life’, in The Human Country, pp. 115–20 Mathews, Harry, ‘Translation and the Oulipo: The Case of the Persevering Maltese’, in The Case of the Persevering Maltese, pp. 67–82 Mathews, Harry, and Alastair Brotchie, Oulipo Compendium (London: Atlas, 1998) Mathews, Harry, and Georges Perec, ‘Roussel and Venice: Outline of a Melancholic Geography’, trans. by Anthony Melville and Harry Mathews, in Mathews, HarryThe Case of the Persevering Maltese, pp. 123–46 McKnight, David, People, Countries, and the Rainbow Serpent: Systems of Classification among the Lardil of Mornington Island (Oxford: Oxford University Press, 1999) Mehlman, Jeffrey, ‘Introductory Note’, in Lacan,‘Purloined Letter’: 38–9 Melnick, David, Men in Aida (Berkeley, CA: Tuumba, 1983) Mosconi, Joseph, ‘Politics of Constraint: The Panel’, in Viegener and Wertheim, nOulipian Analects, pp. 165–8 Motte, Warren F., ‘Clinamen Redux’, Comparative Literature Studies 23.4 (1986): 263–81 Motte, Warren F. (ed. and trans.), Oulipo: A Primer of Potential Literature (Lincoln, NE: University of Nebraska Press, 1986) Nuttall, Zelia, ‘A Suggestion to Maya Scholars’, American Anthropologist 5.4 (1903): 667–78 Oulipo, La littérature poentielle (Paris: Gallimard, 1973) Oulipo, Atlas de littérature potentielle (Paris: Gallimard, 1981) Oulipo, Troll de tram (Le tramway de Strasbourg), in Oulipo, Bibliothèque oulipienne (Bordeaux: Castor Astral, 2000), V, pp. 151–60 Oulipo, Oulipo (Paris: ADPF, 2005)

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

164

Bibliography

Oulipo, C’est un métier d’homme: Autoportraits d’hommes et de femmes au repos (Paris: Mille et une nuits, 2010) Oulipo, ‘Des rubriques’, in Oulipo, Oulipo, pp. 18–19 Perec, Georges, La Vie mode d’emploi (Paris: Hachette, 1978) Perec, Georges, ‘Still Life/Style Leaf ’, Le Fou parle 18 (1981): 3–6 Perec, Georges, ‘Still Life/Style Leaf ’, trans. by Harry Mathews, Yale French Studies 61 (1981): 299–305 Perec, Georges, ‘Entretien Georges Perec/Ewa Pawlikowska’, Littératures 7 (1983): 69–77 Perec, Georges, ‘Space and Underground’, trans. by Harry Mathews, Grand Street 3.1 (1983): 146–50 Perec, Georges, ‘Extract from La Vie mode d’emploi (Chapter XXVII)’, in Atlas Anthology Two, trans. by Harry Mathews (London: Atlas, 1984), pp. 59–63 Perec, Georges, Life A User’s Manual (London: Collins Harvill, 1987) Perec, Georges, Voeux (Paris: Seuil, 1989) Perec, Georges, 53 Days, ed. by Harry Mathews and Jacques Roubaud, trans. by David Bellos (London: Harvill, 1992) Perec, Georges, ‘Petit abécédaire illustré’, in Oulipo, La Littérature potentielle, pp. 235–40 Picart le Doux, Charles, ‘Quelques notes sur le banquet J.-P. Brisset’, Bizarre 8 (1957): 31–6 Politzer, Georges, Critique of the Foundations of Psychology: The Psychology of Psychoanalysis, trans. by Maurice Apprey (Pittsburg, PA: Duquesne University Press, 1994) Queneau, Raymond, Odile (Paris: Gallimard, 1950 [1937]) Queneau, Raymond, Petite cosmogonie portative (Paris: Gallimard, 1950) Queneau, Raymond, ‘La théologie génétique de Jean-Pierre Brisset’, Bizarre 4 (1956): 80–8 Queneau, Raymond, Cent mille milliards de poèmes (Paris: Gallimard, 1961) Queneau, Raymond, Bords: Mathématiciens, précurseurs, encyclopédistes (Paris: Hermann, 1963) Queneau, Raymond, Bâtons, chiffres et lettres, 2nd edn (Paris: Gallimard, 1965) Queneau, Raymond, ‘Science and Literature’, TLS #3422 (28 September 1967): 863–4 (864). Queneau, Raymond, ‘Qu’est-ce que l’art?’, in Le Voyage en Grèce (Paris: Gallimard, 1973), pp. 89–96 Queneau, Raymond, Journaux 1914–1965, ed. by Anne Isabelle Queneau (Paris: Gallimard, 1996) Queneau, Raymond, ‘Conversation Avec Georges Ribemont-Dessaignes’, in Bâtons, chiffres et lettres, pp. 35–47 Queneau, Raymond, ‘Littérature potentielle’, in Bâtons, chiffres et lettres, pp. 317–45 Queneau, Raymond, ‘Potential Literature’, in Motte,Oulipo, pp. 51–64 Quine, W. V. O., ‘Meaning and Translation’, in Reuben Brower (ed.), On Translation (New York: Oxford University Press, 1959), pp. 148–72

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Bibliography

165

Quine, W. V. O., Word and Object (Cambridge, MA: MIT Press, 1960) Quintilian,The Orator’s Education, trans. by D. A. Russell, 5 vols (Cambridge, MA: Harvard University Press, 2014) Rabelais, François, Gargantua and Pantagruel, trans. by M. A. Screech (London: Penguin, 2006) REDCAT, ‘noulipo Experimental Writing Conference’ https://www.redcat.org/ event/noulipo Redfern, Walter, All Puns Intended: The Verbal Creation of Jean-Pierre Brisset (Oxford: Legenda, 2001) Reggiani, Christelle, Rhétoriques de la contrainte: Georges Perec—l’Oulipo (SaintPierre-du-Mont: InterUniversitaire, 1999) Reggiani, Christelle, Poétiques oulipiennes: La contrainte, le style, l’histoire (Geneva: Droz, 2014) Reggiani, Christelle, ‘Être Oulipienne: contraintes de style, contraintes de genre?’, Études littéraires, 47.2 (2016), 103–17 Reggiani, Christelle, ‘Masculin/Féminin: L’écriture oulipienne a-t-elle un genre?’, in Reggiani and Schaffner, Oulipo mode d’emploi, pp. 79–86 Reggiani, Christelle and Alain Schaffner, Oulipo mode d’emploi (Paris: Honoré Champion, 2017) Rosenblum, Rachel, ‘Postponing Trauma: The Dangers of Telling’, International Journal of Psychoanalysis 90.6 (2009): 1319–40 Rossi, Paolo, Logic and the Art of Memory: The Quest for a Universal Language, trans. by Stephen Clucas (London: Athlone, 2000) Roubaud, Jacques, ‘L’auteur oulipien’, in Michel Contat (ed.), L’auteur et le manuscrit (Paris: Presses de France, 1991), pp. 77–92 Roubaud, Jacques, Ma vie avec le docteur Lacan (Bordeaux: L’Attente, 2004) Roubaud, Jacques, ‘Perecquian OULIPO’, trans. by Jean-Jacques Poucel, in Yale French Studies 105 (2004): 99–109 Roubaud, Jacques, ‘Mathematics in the Method of Raymond Queneau’, in Motte,Oulipo, pp. 79–96 Roubaud, Jacques, ‘The Oulipo and Combinatorial Art’, trans. by Harry Mathews, in Mathews and Brotchie, Oulipo Compendium, pp. 37–44 Roudinesco, Elisabeth Jacques Lacan, trans. by Barbara Bray (New York: Columbia University Press, 1970) Roudinesco, Elisabeth, Jacques Lacan & Co: A History of Psychoanalysis in France, 1925–85, trans. by Jeffrey Mehlman (Chicago, IL: University of Chicago Press, 1990) Roussel, Raymond, Impressions d’Afrique, 4th edn (Paris: Alphonse Lemerre, 1932) Roussel, Raymond, Impressions of Africa, trans. by Lindy Foord and Rayner Heppenstall (London: Calder and Boyars, 1966) Roussel, Raymond, How I Wrote Certain of My Books, trans. by Trevor Winkfield (Cambridge, MA: Exact Change, 1995) Sandbukt, Øyvind, ‘Kubu Conceptions of Reality’, Asian Folklore Studies 43 (1984): 85–98

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

166

Bibliography

de Saussure, Ferdinand, Corso di linguistica generale, trans. by Tullio de Mauro (Bari: Laterza, 1967) Schnitzer, Daphné, ‘A Drop in Numbers: Deciphering Georges Perec’s Postanalytic Narratives’, Yale French Studies 105 (2004): 110–26 Soulier, Coraline, ‘L’Oulipo en public’, in Reggiani and Schaffner, Oulipo mode d’emploi, pp. 149–67 Spahr, Juliana, and Stephanie Young, ‘ “& and” foulipo’, in Viegener and Wertheim, nOulipian Analects, pp. 5–13 Swift, Jonathan, Travels into Several Remote Nations of the World, by Capt. Lemuel Gulliver (London: B. Motte, 1726) Swift, Jonathan, Gulliver’s Travels, ed. by Paul Turner (Oxford: Oxford University Press, 1986) Tassou, Bertrand, ‘Les discrets de l’Oulipo’, in Bloomfield and Lesage, Oulipo, pp. 59–65 Thomas, Jean-Jacques, ‘README.DOC: On Oulipo’, trans. by Lee Hilliker, SubStance 17.2 (1988): 18–28 Viegener, Matias, and Christine Wertheim (eds), ‘Oulipo, qui as-tu tu?’, in Reggiani and Schaffner, Oulipo mode d’emploi, pp. 7–26 Tozzer, Alfred M., A Maya Grammar (Cambridge, MA: Peabody Museum, 1921) Trethowan, I, ‘The Mechanical Brain’, The Times, 11 June 1949: 4 Vian, Boris, ‘Un robot poète ne nous fait pas peur’, Arts, 10 April 1953: 1–6 Viegener, Matias, and Christine Wertheim (eds), The nOulipian Analects (Los Angeles, CA: Les Figues, 2007) Viegener, Matias, and Christine Wertheim (eds) ‘Introduction’, in Viegener and Wertheim, nOulipian Analects, pp. 103–7 Viegener, Matias, and Christine Wertheim (eds), ‘ “Séance”, “n/Oulipo”, “Impunities”, “Feminaissance”, & “Untitled”’, Jacket 2: https://jacket2.org/commentary/ seance-noulipo-impunities-feminaissance-untitled Wittgenstein, Ludwig, Tractatus Logico-Philosophicus, trans. by D. F. Pears and B. F. McGuinness (London: Routledge. and Kegan Paul, 1961) Wittgenstein, Ludwig, Philosophical Investigations, trans. by G. E. M. Anscombe (Oxford: Blackwell, 2001) Yates, Frances, The Art of Memory (London: Routledge & Kegan Paul, 1966) Zukofsky, Celia, and Louis Zukofsky, Catullus (London: Cape Goliard, 1969)

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Index Note: Figures are indicated by an italic ‘f ’, respectively, following the page number. abanika (‘chief word-chief ’) 125–6 Abraham, Nicolas 70–2, 92–3 Abramovic, Marina 152 agency of language 90–6, 108–11 see also author’s role; automatic writing alphabets 65–7 analytic philosophy 1–2, 25–6, 122–45 analytical treatments 86–7 Andrews, Chris 97 anglophone perspectives see also translation Dahl’s literature machine 31n.10 homophonic translation 22, 67 ‘noulipo Experimental Writing Conference’ 151–4 on the Oulipo 1–2, 5–7, 19–21, 44 Structuralist phonetics 37–8 Wolf Man case (Freud) 71–2 Apollinaire, Guillaume 88–9 Appenzzell, Marcel (in Perec’s ‘Tale’) 132–4 Apter, Emily 126–7 Arnaud, Noël 4–5, 22, 97 founding the Oulipo 13–15 on the Oulipo and Structuralism 39–40, 49 works of ‘Athalie Galante?’ 53–4, 62 ‘Prolegomena to a Fourth Manifesto’ 5–7, 13–14 art as life and death 11 Artaud, Antonin 22 artificial intelligence 27 see also cybernetics; literature machines artistic inspiration 78–9, 83–6, 95, 107 Atomism and combinatorics 115–16, 121 Aubin, David 81 Augustine, St 122–3 Austin, J. L. 135–6 Australia 135, 137–9 author’s role see also agency of language and artistic inspiration 78–9, 83–6, 107 automatic writing 83–6, 88–9 Calvino on 111–12, 115 changing views on 24–5, 78–9, 95–7 and clinamen 117–19, 121

and literature machines 31–3, 78, 94, 115 in S+7 technique 86–7 automatic writing 83–6, 88–9 see also S+7 technique backtranslation 130 Bacon, Francis 106–7 Bactrian language 125n.9, 126–7 Badiou, Alain 118n.44 Barthes, Roland 3n.5 Communications journal 41–2 and the Oulipo 44n.44, 49–50, 91n.54, 108n.20 on Structuralism 29–30, 37n.32, 43–4 in Times Literary Supplement 19 works of ‘The Death of the Author’ 31–2, 34, 91, 94, 108 Writing Degree Zero 20–1 Bataille, Georges 22 Baudelaire, Charles, Twenty Prose Poems 20–1 Beaudoin, Valérie 149 Beaulieu, Liliane 81n.24 Beauvoir, Simone de 6f Beckett, Samuel, Murphy 51 Bellos, David 118n.46, 130 Bénabou, Marcel joins the Oulipo 23, 78–9 and Lacan 4, 6f, 76 on the Oulipo 13–14, 146n.1 works of ‘La Galère’ 76–7 ‘Rule and Constraint’ 95, 98, 102–4 ‘The Three Circles of Lipo’ 102–4, 104f Bens, Jacques 5–7, 12–14, 24, 28n.5 La Cantatrice Sauve 63–4 Bense, Max 15 Benvéniste, Émile 32–3 Beowulf 54–5 Berge, Claude 7, 14–15, 24, 32–3, 38 Bergson, Henri 118n.44 Bergvall, Caroline 151 Bibliotèque oulipienne 20–1, 23, 69, 131 La Cantatrice Sauve 23, 63–4

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

168

Index

Bibliotèque oulipienne (cont.) Piccolo sillabario illustrato 66 Troll de tram (Le Tramway de Strasbourg ) 64–5, 65f, 124 Bizarre magazine 58–60 Blavier, André 12–13 Temps Mêlés 17–19 Bloom, Harold 118n.44 Bloomfield, Camille on cybernetics 21n.49 ‘The Myth of Secrecy’ 17–19 Raconter l’Oulipo 1n.2, 79–80, 146n.2, 150, 155 body artists 151–3 Bök, Christian 151 Eunoia 155 Boltraffio, G. A. 51–2 Bond novels 46–8 Bonne année pataphysique postcard 66f Bonner, Anthony 102n.8, 105–6 Borges, Jorge Luis fictional Chinese encyclopaedia 139 fictional tribal languages 132n.29 on Llull’s combinatorics 101–2, 107, 111 ‘The Garden of the Forking Paths’ 113n.32 Bosnia 51–2 Botherby, Dr Ernest (in Mathews’ ‘Problem in Translation’) 132, 140–5 Botta, Anna 100 Botticelli, Sandro 51–2 Bourbaki 49–50, 80–2 Théorie des ensembles 81 Braffort, Paul 14, 17, 38, 80–1 Brazil 143–5 Breton, André 82–4, 88–9, 94 Anthology of Black Humor 56–7 Brisset, Jean-Pierre 54–62, 64–5 La Science de Dieu 56–7 Le Mystère de Dieu est accompli 60 broadcasts 14, 17–19 Brotchie, Alastair 64, 86–7, 118 O, to mail Hank holly! 62–3 Bull computers 21–2, 34–5 Burgelin, Claude 75n.62 La Cantatrice Sauve 63–4 Burt, Stephanie 151 Butler, Judith 148–9 Caballé, Montserrat 63–4 Calliope (robot-poet) 30–2, 109 Calvino, Italo 9–10, 53 joins the Oulipo 23, 78–9

literature machine of 107–15 philosophical shift of 25, 99–101, 107–21 on Queneau’s CMMP 34–5 on revealing Oulipian rules to readers 77n.8 on Structuralism 43–4, 108n.20 in Times Literary Supplement 19 works of ‘The Burning of the Abominable House’ 112–15 The Castle of Crossed Destinies 95–6, 119–21 ‘Cybernetics and Ghosts’ 34, 108, 115, 118 If on a Winter’s Night a Traveller 21 Piccolo sillabario illustrato 66 ‘Prose and Anticombinatorics’ 112–13, 115, 118 Six Memos for the Next Millenium 100–1 cameos 120n.50, 131 Campaignolle, Hélène 21n.49 Cape, Jonathan 17, 20–1, 25–6 Cape Editions series 20–1, 23–4, 45 categories, linguistic 133–9 CENIS (MIT’s Centre for International Study) 33 Chambers, Ross 14 chance and clinamen 120–1 and necessity 117 and Surrealism 85–6 Change group 4–5, 45 Chapman, Stanley 14 Chaucer, Geoffrey, The Canterbury Tales 54–5 Chauvin, Rémi 32–3 Chinese (fictional) 139 Chomsky, Noam 37–8 cinema 40 class issues 147–8 classification systems 137–9 clinamen 25, 115–21 CMMP see Queneau, Raymond: Cent milles milliards de poèmes (CMMP) Collège de ’Pataphysique on crime novels 46, 48–9 Dossier 17 17–19, 87–8 and the Oulipo 13–15 colours 107 combinatorics 1–2 see also literature machines Calvino and Lucretius on 100–1, 115–17

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Index filtering outputs 104–5, 113–14 Llull on 100–8, 103f, 111 at plot level 38–9, 46–7 problems with 104–7, 109–10 Révéroni’s matrix 2, 4n.6 Comenius 67 Communications journal 40–2, 46–8 computers 21–2, 34–5 see also literature machines constraining structures 1–2, 16, 16n.36, 155–6 see also syntactic and semantic constraints Bénabou’s ‘Three Circles’ diagram 102–4, 104f and clinamen 117–21 and role of the author 79, 94 and the unconscious 95–9 counting, concept of 143–4 creation myths 56–61 crime novels 46–9 cryptonyms 70–4 cybernetics see also literature machines impact on Structuralism 30, 32–4 impact on the Oulipo 17–19, 21–2, 30, 34–8 and love letters 29 Dahl, Roald, ‘The Great Automatic Grammatizator’ 31n.10 de Chirico, Giorgio 88–9 De Mauro, Tullio 54n.6 Deleuze, Gilles 118n.44 demons 59–61 Derrida, Jacques 117 Desnos, Robert 94–6 determinism, linguistic 142–4 dictionaries Barthes’ ‘ready-formed dictionary’ 32–3, 91, 94 and S+7 technique 8–9, 79, 86–9 dogs 61–2 dreams 70–2, 136–7 Drucker, Johanna 151 Duchamp, Marcel 15 Duchateau, Jacques on Communications article 40–2, 47–8 on cybernetics and literature 35–6 death of 23 founding the Oulipo 14–15 on Structuralism 36–8 on syntactic versus semantic constraints 38–9 Ducrocq, Albert Calliope robot-poet 30–2, 109 L’Ère des robots 31

169

duperie (trickery) 58–9 Dyirbal language 137–9 Eco, Umberto ‘Achilles is a duck’ 74 Communications journal 41–2, 46–8 on James Bond novels 34, 46–8 on Llull’s combinatorics 101n.7, 102n.8, 104–5 in Times Literary Supplement 19 Elle magazine 11 English see anglophone perspectives; Mathews, Harry; translation Étienne, Luc 17 Everett, Daniel 143–5 exhaustiveness and combinatorics 102, 105–7, 109 external knowledge and combinatorics 104–5, 114 ‘family resemblances’ 136 farce 7–10 fatalism 51 feminism 149–53 filtering combinatoric outputs 104–5, 113–14 Fleming, Ian, James Bond novels 34, 46–8 Fonds Oulipo 7 Ford, Mark 70 foreign correspondents 14–15, 23, 80–1 foreskins 60–1 forgetfulness 51–2 Formalism 44 Foucault, Michel 2–4, 55, 64, 69–70, 91 Raymond Roussel (Death and the Labyrinth) 67 ‘foulipo’ 151–5 Fournel, Paul 12, 84, 151, 154 Clefs pour la littérature potentielle 23–4, 83 La Cantatrice Sauve 63–4 Franco, Daniela 155 free will 117 freedom and clinamen 116–18, 120–1 and surrealism 85–6 Freud, Sigmund in Calvino’s Castle 120n.50 Queneau compares with Brisset 59–62 on Surrealism 84n.33 use of rebus device 70–1 works of Interpretation of Dreams 92–3, 136–7

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

170 Freud, Sigmund (cont.) ‘The Psychical Mechanism of Forgetfulness’ 22, 51–2 frogs 58–9 Gaillard, Jean 11 Gallimard 20–1 Gardner, Martin 101–2, 105–6 Garréta, Anne on the Oulipo’s gender bias 149–51 on Queneau 10n.19 Sphinx 147–9, 154–5 gavagai language 25–6, 123–4, 137, 139, 141 gender issues 147–55 Genette, Gérard 41–2, 79–80 Geoghegan, Bernard 33 Gerard, Sarah 148–9 German 52, 61, 70–2 God 59–61 Llull on 101–2, 103f, 104–5, 107 Gordon, Peter 142–5 Grangaud, Michelle 149 Grasset, Jean-Baptiste 10n.19 Graverol, Jane, Temps Mêlés 17–19 Greimas, A. J. 41–2 Gross, Maurice 37–8 Guilbaud [Guillebeau], G.-T. 32–3 Halle, Morris 37–8 Hamlet’s soliloquy 70 Hamsun, Knut 88–9 Hegel, G. W. F. 9–10 Homer, Iliad 54n.6 homophonic translation 17 and cryptonyms 70–2 and forgetfulness 52, 70–1 and the Oulipo 52–5, 59–67, 124 and psychoanalysis 22–3, 52–4 and rebus device 70–1 homosyntactical translation 28n.5 Hugo, Victor, ‘Le Calcul’ 19 humour and seriousness of the Oulipo 1–3, 5–11, 16, 48–9, 72–5, 132–45 see also homophonic translation hunger, and automatic writing 88–9 IBM 21–2, 34–5 identity gender 147–9 national 142 of the Oulipo 4–5, 45–6, 49 politics of 146–7, 154 Indonesia 134–5

Index ‘infra-forms’ 36–7 inspiration, artistic and combinatorics 107 and Surrealism 78–9, 83–6, 95 intellectual scene, Paris, 1960s and 70s and Calvino 100, 108 gender bias of 150 interdisciplinary nature of 4–5 the Oulipo’s relationship with 1–5, 25, 50, 78, 146 Ionesco, Eugène, La Cantatrice Chauve 63 isosyntactical poems 28n.5 Italian language 66 see also Calvino, Italo Italy 51–2, 72–5 Jakobson, Roman 21–2, 37–8, 49, 108–9 James, Alison 90n.51, 110 James, Henry 38 James Bond novels 34, 46–8 Jarry, Alfred 31–2, 118n.44 The Supermale 20–1 Jefferson, Geoffrey 27 ‘Jeudis de l’Oulipo’ 17–19, 149–50 Judaism 60–1, 76–7, 101, 102n.10, 133 Jurquet, Béatrice de, La Cantatrice Sauve 63–4 Keats, John, Ode to Melancholy 62–3 Kojève, Alexandre 10n.18, 76–7 Kubota, Shikego 152 Kubu language (in Perec’s ‘Tale’) 133–41 Kubu tribe (genuine) 134–5 Kuon, Peter 99 La Littérature potentielle (Oulipo) 20–1, 42–3 Lacan, Jacques on clinamen 118n.44 and cybernetics 21–2, 32–4 on language and the unconscious 93–4, 108–11 and the Oulipo 4, 6f, 76–7 seminars of 43–4 ‘The Instance of the Letter in the Unconscious’ 91–3 Lakoff, George 137–8 language, agency of 90–6, 108–11 ‘language-games’ (Wittgenstein) 140–1 Laplanche, Jean 91–3 Latis and cybernetics 21n.49 death of 99 founding the Oulipo 13–15

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Index homosyntactical translation 28n.5 on syntactic versus semantic constraints 38–9, 43 Le Clec’h, Guy 8–9 Le Lionnais, François 8f, 17, 81–2 on cybernetics 35 founding the Oulipo 10–14, 82 homophonic translation 53 isosyntactical poems 28n.5 manifestos of the Oulipo 16n.36, 17–19, 42–3 opening the Oulipo to new members 23, 76 and Structuralism 32–3, 37–8 on syntactic versus semantic constraints 25, 38–9, 42–3 views on the Oulipo’s work 15, 25–6, 42 works of Les Grands Courants de la pensée mathématique 11 ‘Les Structures du roman policier: Qui est coupable?’ 48–9 Un Certain Disparate 10n.19, 41n.36 Le Tellier, Hervé 19n.44, 78–80, 82, 86, 95 Lécureur, Michel, Raymond Queneau 83n.29 Leibniz, G. W. 109n.26 Leiris, Michel 22 Lemaire, Anika 91–2, 94 Lescure, Jean 13–14, 84–5 on Communications article 41–2 founding the Oulipo 14–15 Roussellian Permutations 89–91, 94 on S+7 technique 86–8 Lévi-Strauss, Claude 4–5, 21–2, 32–3, 108 The Scope of Anthropology 20–1 Levy-Leblond, Jean-Marc 10n.19 libertinism 53–4 Life A User’s Manual (Perec) see Perec, Georges: La Vie mode d’emploi (VME ) lightness (la leggerezza) 9–10 linguistic determinism 142–4 linguistics, Structuralist 37n.31 see also Structuralism lipograms 147–53 literary works of the Oulipo 20–4, 45, 63–4, 98 literature machines and Calvino 107–15 Duchateau on 35–8

171

examples of machine literature 27–9, 29f, 30–2, 47, 112–15 impact on Structuralism 27–9, 32–4 Llull’s combinatorics 107 and poetry 30–3, 78, 111 and role of the author 31–3, 78, 94, 115 Llull, Ramón 100–10, 103f, 121 Lomas, David 83–4 London Review of Books 151 love letters 27–9, 29f, 109 Lowtsky, Fanny 83–4 Lucretius 100–1, 115–17, 121 De rerum natura 115–16 machine literature see literature machines Mallarmé, Stéphane 31–2 Manchester University Computer (M. U. C.) 27–9, 29f, 30 Mandelbrot, Benoit 33 manifesto of Surrealism 88 manifestos of the Oulipo 49–50, 77–8, 94–5 first 17–19 second 16n.36, 42–3, 85 fourth 5–7, 13–14 Marcus, Solomon 37n.31 Mars 74 masks 8–9 mathematics see also Bourbaki and human sciences, relationship between 32–5 ‘noulipo’ 153 and the Oulipo 7, 16, 149–50 in the Oulipo’s literature 11, 114–15, 128–9 and Structuralism 37n.31 ‘two cultures’ debate 7, 10–11, 15, 19–20, 44 Mathews, Harry 17, 151 on clinamen 118 on homophonic translation 22, 52–3, 62–4, 67, 70 joins the Oulipo 23, 78–9, 127–8, 128f on the Oulipo’s literary work 98n.77 on S+7 technique 86–7, 98–9 translating Perec 127–32 on translation issues 25–6, 79, 142 on the unconscious 95–7 works of Cigarettes 21 The Conversions 128–9 ‘The Dialect of the Tribe’ 124–7 La Cantatrice Sauve 63–4 O, to mail Hank holly! 62–3

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

172 Mathews, Harry (cont.) ‘A Problem in Translation’ 132, 139–42, 144–5 ‘Roussel and Venice’ 67, 72–5 The Sinking of the Odradek Stadium 124–6 ‘Still Life’ 131–2 Tlooth 127–8, 142, 147 Mayan language 139 Mehlman, Jeffrey 94 membership rules 81–3 Métail, Michèle 53, 97, 149 metonymy 61, 70–2 Metz, Christian 40, 47–8 miniaturism 11 ‘minor modifications’ (Perec and Mathews) 128–30 MIT’s Centre for International Study (CENIS) 33 Monk, Ian 150–1 monthly meetings of the Oulipo minutes of 3, 7–8, 7n.14, 13, 16–17 March 1963 2–4 July 1963 8–9 November 1964 38 March 1965 40–2 August 1966 15n.32, 23, 42, 78–9 August 1970 91n.54 January 1971 17, 127–8, 128f November 1972 112, 128–9 April 1994 149 April 2000 149 Mosconi, Joseph 151–2 Motte, Warren 118n.44 murder mystery novels see crime novels N+7 technique see S+7 technique narratology, Structuralist 37n.32, 40–2, 46–7 national identity 142 Naville, Violette 41–2 Nazis 10–11, 76–7, 133 necessity 117 ‘noulipo’ 151–5 nouveau romanciers 4–5 Nufer, Doug 151 numeracy and language 143–4 Ogden, C. K., Basic English 137 Ogliastro, César 87–8 Oho language (in Mathews’ ‘Problem in Translation’) 140–5 origins, Brisset’s theory of 56–61 ‘ostensive definitions’ 122–3 oulipian (small ‘o’) 155–6

Index Oulipo, the 14–17, 26 see also Bibliotèque oulipienne; monthly meetings of the Oulipo in 21st century 17–19, 23, 146–7, 149–56 in 1960s and 70s intellectual scene 1–5, 25, 50, 78, 146 and Bourbaki 80–2 and Calvino 111–12 and clinamen 117–19 founding of 10–13 gender bias of 149–51 and homophonic translation 52–5, 59–67 humour and seriousness of 1–3, 5–11, 16, 48–9, 72–5, 132–45 identity in question 4–5, 45–6, 49 literary works of 20–4, 45, 63–4, 98 manifestos of 5–7, 13–14, 16n.36, 17–19, 42–3, 85 opens to new members 13, 15, 23–4, 78–9, 127–8, 128f, 149 S+7 technique 8–9, 79, 86–90, 94, 118, 153 second wave (‘young ones’) 23–5, 76–9, 95–9, 119 secrecy of 15, 17–19, 21–2, 24, 34–5, 45, 77–8 and Structuralism 5, 21–2, 32–3, 36–44, 49–50 and Surrealism 5, 49–50, 78–89, 94 ‘true’ Oulipians 151 ouvroir, translations of 13–14 Pagolak language 125–6 Pan language 126 Pankejeff, Sergei (the Wolf Man) 70–2 Papua New Guinea 139–40, 144–5 Paris 74 see also intellectual scene, Paris, 1960s and 70s parody 27–32 ‘Roussel and Venice’ 72–5 ’Pataphysics 4, 9, 13, 17–19 Bonne année pataphysique postcard 66f Peillet, Emmanuel see Latis Pekee (in Perec’s Kubu language) 133–7, 139 penises 60–1, 71–2 Penrose, Roger 33 Perec, Georges and Barthes 44n.44 and homophonic translation 22, 53, 63–4 joins the Oulipo 23, 78–9

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Index minutes of Oulipo meetings 17, 18f, 22, 77n.8 translating Mathews 25–6, 127–32, 142 works of 53 Days 131 La Cantatrice Sauve 63–4 La Disparition 7–8, 21, 147 La Vie mode d’emploi (VME ) 7–8, 121, 128–9, 132–3 ‘Petit abécédaire illustré’ 65–6 ‘Roussel and Venice’ 67, 72–5 ‘Still Life/Style Leaf ’ 131–2 ‘Tale of the Misunderstood Anthropologist’ 132–9 Ulcérations 20–1 Pferdli, O., ‘Autres images de mélancholie’ 72–4 phonetics, Structuralist 37–8 see also Structuralism phonology, Structuralist 49, 108 see also Structuralism Piaget, Jean 33 Pirahã language 143–5 planets 74 Plato 1–2, 9–10 playfulness see humour and seriousness of the Oulipo Poe, Edgar Allan, Rue Morgue 48–9 poetry automatic writing 83–4 and combinatorics 107 and gender 152–4 and literature machines 30–3, 78, 111 Lucretius’ De rerum natura 115–16 Queneau’s Chêne et chien 93–4 ‘snowball’ device 96–7 sonnets 12, 27, 34–7, 79, 94 politics of body 152–3 of gender 151–3 of identity 146–7, 154 Politzer, Georges 91–2, 95–6, 110 Pontalis, J. B. 74–5, 133 Portail, Claude 127–8 primitive languages see tribal languages ‘Prince of Thinkers’ (Brisset) 56 proceduralism 25, 42, 78–9, 94–5, 98–9, see also S+7 technique; syntactic and semantic constraints Propp, Vladimir 39, 44 Morphology of the Folktale 108 pseudonyms 80–1 psychoanalysis 1–2 see also Freud, Sigmund

173

and agency of language 91–4 and homophonic translation 22–3, 52–4, 59–62, 67, 133 language and the unconscious 91–7, 108–12, 119–21 and literature machines 108–9 reflected in the Oulipo 5, 95–6 and role of the author 78–9 ‘Roussel and Venice’ 73–5 and Surrealism 83–4 punning 22, 51, 68–70 see also homophonic translation queer theory 148–9 Queneau, Raymond 4, 6f, 8f, 36n.27, 81–2, 100 on Brisset 55–62 on Communications article 40–2 death of 99 on Foucault 2–3 founding the Oulipo 10–13, 80–2 and Lacan 76–7 on the Oulipo’s uniqueness 4–5, 45 on the Oulipo’s work 8–9, 15–16, 22, 76, 77n.8, 87–8, 98n.75 on Surrealism 82–6, 88–9 in Times Literary Supplement 13–14, 19, 44 works of Cent milles milliards de poèmes (CMMP) 12, 27, 34–5, 79, 94 Chêne et chien 61–2 On est toujours trop bon avec les femmes 76–7 Exercices de style 11 ‘La théologie génétique’ 58–60 Odile 84–5 Petite cosmogenie portative 55 ‘What is Art?’ 83–6 Zazie dans le métro 11 Queval, Jean 8–9, 14–15, 24, 41–2 ‘Quiet Ones of the Oulipo, The’ 24 Quine, W. V. O. 25–6, 126–7, 137, 139, 141 Word and Object 123–4 Quintilian 1 ‘rabbit’, translation of 123–4, 137 Rabelais, François 106n.17 Racine, Jean, Athalie 53–4, 62 Ramadan, Emma 147–8 reader’s role 111 rebus device 70–2 Reggiani, Christelle 78, 150–1

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

174

Index

relevance of the Oulipo today 146–7, 149–56 religion and homophonic translation 53–4 and Llull’s combinatorics 101–2, 104–5, 107, 114 repression 70–2, 119–21 restricted vocabularies 135–7 Kubu language (in Perec’s ‘Tale’) 133–4 Oho and Uhu languages 140–2 Pirahã language 143–5 slab language 122–5 Révéroni, J.-A. 2, 4n.6 rhetoric 1 rhyming 68 robot-poets 30–2, 109 Roche, Denis 127–8 Rodin, Auguste, The Thinker 56 Rolling Stones, The 146–7 Romains, Jules 56 Rosenbaum, Rachel 74–5 Rossi, Paolo 102n.8 Roubaud, Jacques 23, 78–81, 86, 98–9, 101, 117–18 ϵ 21 ‘L’auteur oulipien’ 122 Roudinesco, Elisabeth 22–3, 32–3 Roussel, Raymond 53, 67–70, 95–6 Impressions of Africa 68–9, 89–90 Locus Solus 74–5 Roussellian Permutation 89–91 Russian 70–2 S+7 technique 8–9, 79, 86–90, 94, 118, 153 Salon, Olivier 10n.19 Chaque porche est une invitation au voyage 154–5 saloperie (filth) 58–9 Sandbukt, Øyvind 134–5 Sartre, Jean-Paul 6f Saussure, Ferdinand de 59 Cours de linguistique générale 49 Schmidt, Albert-Marie 8–9, 14–15, 24, 38 Schnitzer, Daphné 133 ‘second wave’ of the Oulipo 23–5, 76–9, 95–9, 119 secrecy of the Oulipo 15, 17–19, 21–2, 24, 34–5, 45, 77–8 semantics see homophonic translation; syntactic and semantic constraints seriousness of the Oulipo see humour and seriousness of the Oulipo Serres, Michel 118n.44 sexual references 53–4, 59–60, 66, 76–7

Shakespeare, Hamlet’s soliloquy 70 signifiers and signified 59, 67 Signorelli, Luca 51–2 Sinuya (in Perec’s Kubu language) 133–5, 137–9 slab language 122–5, 139–41 ‘snowball’ device 96–7 sonnets 12, 27, 34–7, 79, 94 Spahr, Juliana, ‘The Politics of Constraint’ 151–4 spatial metaphors 92–3, 110–11 Spinoza 34n.22 Strachey, Christopher 28–9, 28n.4, 29f, 37–8, 109 Strasbourg tramway 64–5, 65f, 124 Structuralism 1–2, 43–4 attitude contrasted with the Oulipo’s 5–7, 21 Barthes on 19, 29–30 on crime novels 46–8 and cybernetics 21–2, 30, 32–4 impact of literature machines on 27–9 and Llull 100–9 relationship with the Oulipo 5, 21–2, 32–3, 36–44, 49–50 versus structurElism 42 subject’s role see author’s role Subsidia Pataphysica 46, 48–9 suffering artists 88–9 Surrealism 5, 24–5, 49–50, 76–99 Swift, Jonathan, Gulliver’s Travels 67, 109–10 Symons, Arthur, Rain on the Down 89–90 synonyms 69 syntactic and semantic constraints 25 see also literature machines Bénabou’s ‘Three Circles’ diagram 102–4 Structuralism and the Oulipo 30, 37–40, 42–4 Surrealism and the Oulipo 94 taboos 59–60, 110–11, 121 Tarn, Nathaniel 20–1 teeth 57–8 Tel Quel group 4–5, 45, 150 theology and combinatorics 101–2, 104–5, 107, 114 Thomas, Jean-Jacques 78–9 ‘Oulipo, qui as-tu tu?’ 146n.1 thought determined by language 142–4 Thursday night performances 17–19, 149–50 tigers 107n.19 Times Literary Supplement 19, 43–4

OUP CORRECTED PROOF – FINAL, 10/6/2019, SPi

Index Todorov, Tzvetan 41–2 Török, Mária 70–2, 92–3 ‘Mourning or Melancholia: Introjection versus Incorporation’ 72 ‘totem-and-taboo’ 59–60 tramways, Strasbourg 64–5, 65f, 124 translation backtranslation 127–32 of gavagai language 123–4 homophonic see homophonic translation homosyntactical 28n.5 language and the unconscious 91–2 of the Oulipo’s work into English 25–6, 147–8 of ouvroir 13–14 of Pagolak language 125–6 untranslatability 25–6, 125–7, 141–5 tribal languages 122–7, 132–45 Turing, Alan 27 ‘two cultures’ debate 7, 10–11, 15, 19–20, 43–4

175

Venus 74 Vian, Boris 31–2 Viegener, Matias 151 VME see Perec, Georges: La Vie mode d’emploi (VME ) Volontés journal 83–4 Walsh, Joanna 155 Watson Taylor, Simon 80–1 Wertheim, Christine 151 whimsicality see humour and seriousness of the Oulipo White, Iain 15n.32 Whorfian hypothesis 142–4 Wilkins, John 105 Wittgenstein, Ludwig 124n.7, 144–5 Philosophical Investigations 122–3, 134–6, 139–41 Wodehouse, P. G. 48–9 Wolf Man case (Freud) 70–2 word order 1 see also combinatorics Wordsworth, William, Daffodils 86–7 workroom/workshop (ouvroir) 13–14, 20

Uhu language (in Mathews’ ‘Problem in Translation’) 141–2, 144–5 unconscious, the 91–7, 108–12, 119–21 see also psychoanalysis UNESCO 32–3 untranslatability 25–6, 125–7, 141–5

Yates, Frances 101n.7, 107, 111 Young, Stephanie, ‘The Politics of Constraint’ 151–4 ‘young ones’ (second wave Oulipians) 23–5, 76–9, 95–9, 119

valency 136–7 Venice 72–5

Zazie phenomenon 11 Zukofsky, Celia and Louis, Catullus 22