138mm Ém

‘Émile Benveniste was a giant whose influence has been felt ILE

across semiotics and linguistics. Yet, as John E. Joseph says in the Be

Introduction to his much anticipated magisterial translation of N

Benveniste’s final lectures, many have “seen his work referred to ve reverentially, but have not necessarily read it themselves”. Ranging NI s

across language, writing and general semiology, the sixteen lectures T presented here, along with notes for a seventeenth, will serve as a e coruscating introduction for the uninitiated Anglophone and as a reminder of the greatness of Benveniste for the already converted.’ Collège Lectures: Last Last Lectures: Paul Cobley, Middlesex University London Benveniste’s lectures had a shaping influence on a generation of scholars that Collège de includes Barthes, Deleuze, Bourdieu, Derrida, Kristeva and Todorov. Here, for the first time, these lectures are made available in English for a new generation of linguists and philosophers of language. This book includes the full course of lectures that Benveniste gave in the Collège de France on the Rue des Écoles in France Paris between December 1968 and December 1969. Benveniste’s work as offered here presents the first serious attempt at reconciling the sign theories of Saussure and Peirce and draws together language, writing and society into a comprehensive 1968 and theory of signifying. Benveniste’s philosophy of language considers key concepts

such as utterance, enunciation, speaker, discourse and subjectivity and, as such, is de central to the areas of discourse analysis, text linguistics, pragmatics, semantics, France 1968 France conversational analysis, stylistics and semiotics. 1969 Key Features: • Introduction from editors Jean-Claude Coquet and Irène Fenoglio ÉmILE • New introduction by the translator John Joseph • Preface by Julia Kristeva • Includes Benveniste’s course of lectures BeNveNIsTe

• Afterword by Tzvetan Todorov and

ÉmIle BeNveNIsTe (1902–1976) was the pre-eminent linguist in France for 1969 three decades beginning in the late 1930s. He worked mainly on Indo-European historical linguistics, but became widely known as a theoretician through the two volumes of his Problems in General Linguistics (1966, 1974) and Dictionary of Indo-european Concepts and Society (1969). This book contains the final lectures he gave before a stroke in December 1969 paralysed and silenced him.

ISBN 978-1-4744-3990-9 EDITED by JeAN-ClAuDe CoqueT anD IrèNe FeNoglIo edinburghuniversitypress.com TrANslATeD by Cover image: www.shutterstock.com Cover design: www.richardbudddesign.co.uk JohN e. JosePh

15mm spine Last Lectures

Last Lectures

Collège de France 1968 and 1969

Émile Benveniste

Edited by Jean-Claude Coquet and Irène Fenoglio

Translated by John E. Joseph Edinburgh University Press is one of the leading university presses in the UK. We publish academic books and journals in our selected subject areas across the humanities and social sciences, combining cutting-edge scholarship with high editorial and production values to produce academic works of lasting importance. For more information visit our website: edinburghuniversitypress.com

Originally published as Dernières leçons, Collège de France 1968 et 1969, © Émile Benveniste, 1968, 1969, © Seuil/Gallimard, 2012 © Editorial matter and organisation, Jean-Claude Coquet and Irène Fenoglio, 2012 © Preface, Julia Kristeva, 2012 © English translation, John E. Joseph, 2019

Edinburgh University Press Ltd The Tun – Holyrood Road, 12(2f) Jackson’s Entry, Edinburgh EH8 8PJ

Typeset in 11/13 Adobe Sabon by Servis Filmsetting Ltd, Stockport, Cheshire, printed and bound in Great Britain.

A CIP record for this book is available from the British Library

ISBN 978 1 4744 3990 9 (hardback) ISBN 978 1 4744 3992 3 (webready PDF) ISBN 978 1 4744 3991 6 (paperback) ISBN 978 1 4744 3993 0 (epub)

The right of Émile Benveniste to be identified as the author of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1988, and the Copyright and Related Rights Regulations 2003 (SI No. 2498). Contents

Biographical Information vi Editors’ Acknowledgements viii Biographical Timeline ix

Preface: Émile Benveniste, a Linguist Who Neither Says Nor Hides, but Signifies 1 Julia Kristeva Translator’s Introduction 31 John E. Joseph Editors’ Introduction 61 Jean-Claude Coquet and Irène Fenoglio

1 Semiology 74 2 Languages and Writing 91 3 Final Lecture, Final Notes 121

Annex 1: Bio-bibliography of Émile Benveniste 128 Georges Redard Annex 2: The Émile Benveniste Papers 157 Émilie Brunet Afterword: Émile Benveniste, a Scholar’s Fate 163 Tzvetan Todorov

Name Index 179 Subject Index 182

v Biographical Information

Émile Benveniste (1902–1976) was the pre-eminent linguist in France for three decades beginning in the late 1930s. He worked mainly on Indo-European historical linguistics, but became widely known as a theoretician through the two volumes of his Problems in General Linguistics (1966, 1974) and Dictionary of Indo-European Concepts and Society (1969). This book contains the final lectures he gave before a stroke in December 1969 paralysed and silenced him.

Julia Kristeva, author of many academic books and novels, has been a leading figure in semiotics since the 1960s. She is Professor Emeritus in the University of Paris Diderot, and in 2004 was awarded the Holberg International Memorial Prize for her innovative explorations of questions on the intersection of language, culture and literature.

Georges Redard (1922–2005), a specialist in the languages of Iran and Afghanistan, was professor and dean in the Universities of Neuchâtel and Berne, where he also served as rector.

Tzvetan Todorov (1939–2017), a prominent figure in French literary studies, was a Director of Research in the Centre Nationale de Recherche Scientifique and visiting professor at Yale, Harvard and other top international universities. The Académie Française awarded him the Prix Maujean (1989), the Prix La Bruyère (2001) and the Prix de la Critique (2011).

Jean-Claude Coquet is Professor Emeritus of Linguistics and Semiotics in the Université de Paris 8.

vi Biographical Information vii

Irène Fenoglio directs the Linguistics section of the Institut des Textes et Manuscrits Modernes of the Centre Nationale de Recherche Scientifique.

John E. Joseph is Professor of Applied Linguistics in the University of Edinburgh. Editors’ Acknowledgements

This book would not have been possible without the excep- tional welcome we received at the Bibliothèque Nationale de France, first of all from Monique Cohen, the director of what was then called the Department of Oriental Manuscripts, the department which received the Benveniste bequest; then from Thierry Delcourt, who became director of the Department of Manuscripts when the various departments were unified; and from Anne-Sophie Delhaye, Adjunct Director of the Department. It was in this department that Émilie Brunet had the responsibil- ity for this archive, and we thank her for her collaboration. The renowned linguists Jacqueline Authier-Revuz and Claudine Normand offered us their notes taken during the linguist’s last lectures in the Collège de France to supplement those taken by Jean-Claude Coquet. This precious transmission permitted us to establish a continuity in the text of the course, overcoming the discontinuity in Émile Benveniste’s own notes. Finally, this edition of Émile Benveniste’s last lectures has benefitted from the rigorous work of transcribing manuscripts carried out by Arlette Attali and Valentina Chepiga. We are indebted to them for this long and meticulous undertaking.

viii Biographical Timeline

Émile Benveniste, 1902–1976 1902 (27 May) Birth at Aleppo (Syria, Ottoman Empire), with the name Ezra Benveniste. His father, Mathatias Benveniste (born in Smyrna in 1863), and his mother, born Maria Malkenson in Vilna (Russia, now Vilnius, Lithuania), are school inspectors of the Universal Israelite Alliance (Alliance Israélite Universelle, AIU). A brother, Henri (born Hillel Benveniste at Jaffa in 1901), deported to Auschwitz and murdered there in 1942. A sister, Carmelia (born in 1904 in Aleppo), died in 1979.

1913 Arrives in Paris to undertake his studies in the ‘little semi- nary’ of the Rabbinical School, 9 rue Vauquelin. Studies funded by the AIU. His parents are working in Samokov, Bulgaria.

1918 Receives baccalaureate degree, with poor results (‘mention passable’), including (according to legend) a particularly low score (1) in languages. (October) Letter from his mother to the President of the AIU asking for the whereabouts of her son, who has quit the Rabbinical School. E. B. looks for work as a teaching assistant in a lycée. Enrols in the École Pratique des Hautes Études.

1919 (21 April) His mother, Maria Benveniste, dies in Samokov, Bulgaria. E. B. had probably not seen her again since moving to Paris in 1913.

ix x Last Lectures: Collège de France 1968 and 1969

1919–20 Completes the licence ès lettres (first university degree).

1920 Receives the diplôme d’études supérieures for his thesis The sigmatic futures and subjunctives of Old Latin, supervised by Joseph Vendryes (1875–1960).

1921 (3 May) Granted right of abode in France with legal rights.

1922 Enrols in the École des langues orientales (School of Oriental Languages). Together with his father, brother and sister, settles in Montmorency, a suburb ten miles north of Paris. Receives the agrégation de grammaire (teaching qualifica- tion), ranked ninth in the national competition.

1922–4 Teaches at the Collège Sévigné in Paris.

1924 (9 October) Becomes a naturalised French citizen. Changes his first name from Ezra to Émile.

1924–5 Spends eighteen months in Poona (south-east of Mumbai), British India, as tutor to the children of the Tata family, famous Indian industrialists.

1925 Co-signs three articles in L’Humanité: with Henri Barbusse, ‘Appel aux travailleurs intellectuels: oui ou non, con- damnez-vous la guerre?’ (Call to intellectual workers: yes or no, do you condemn the war?);1 with friends in the Surrealist group (Louis Aragon, André Breton and Paul Éluard), ‘La Révolution d’abord et toujours’ (Revolution first and forever) and ‘“Clarté”, “Philosophies”, “La Révolution surréaliste” solidaire du Comité Central d’Action contre la guerre du Rif’ (‘Clarity’, ‘Philosophies’, ‘The Surrealist Revolution’ in solidarity with the Central Action Committee against the Rif War).

1925 (July) Signs the Manifesto of Intellectuals against the Rif War. Biographical Timeline xi

1926 (May)–1927 (November) Military service as foot soldier in Morocco, despite his opposition to the Rif War.

1927–69 Director of Studies at the École Pratique des Hautes Études.

1935 Completes his doctoral theses Origines de la formation des noms en indo-européen (Origins of noun formation in Indo- European, primary thesis) and Les infinitifs avestiques (The Avestan infinitives, secondary thesis), and publishes them (both Paris: Maisonneuve, 1935).

1936 (26 February) Defends his theses on 26 February 1936, after which he is awarded the degree docteur ès lettres.

1937 Succeeds Antoine Meillet (1866–1936) in the Chair of Comparative Grammar of the Collège de France.

1940–1 Prisoner of war. Escapes, live clandestinely in France and takes refuge in Switzerland thanks to the help of Father Jean de Menasce (1902–1973), a polymath whose fields of interest include Zoroastrianism and Iranian studies, and who gets Benveniste a job as a librarian in the Université de Fribourg.

1942 (23 September) His brother Henri is arrested outside his home in Paris and deported to Auschwitz in Convoy 36, never to return.2

1956 Becomes Secretary of the Société de linguistique de Paris. (December) Suffers first heart attack.

1959–70 Secretary of the Société de linguistique de Paris.

1960 Elected Member of the Institute, Académie des Inscriptions et Belles-Lettres. Author of the linguistics section of the Business Report of the National Centre for Scientific Research (CNRS). xii Last Lectures: Collège de France 1968 and 1969

1963 Director of the Institute of Iranian Studies of the Université de Paris.

1964 Director of the Revue d’études arméniennes (Journal of Armenian studies).

1968 (25 August–1 September) President of the International Symposium of Semiotics at Warsaw.

1969 First President of the newly created International Association for Semiotic Studies. (6 December) Stroke which leaves him permanently paralysed and unable to speak.

1976 (3 October) Émile Benveniste dies at Versailles, where he is buried in the Cimetière des Gonards.

Notes 1 [Translator’s note (Tr.): The Rif War (1920–7) was fought in Morocco, initially between the colonial power Spain and Berber tribes in the Rif mountains. France joined the war on the side of Spain in 1925.] 2 [Tr.: There is some confusion about the arrest of Henri Benveniste. The French text calls him a victim of the Vel’ d’Hiv Roundup, a joint operation by French and German police on 16–17 July 1942 in which over 13,000 Jews were arrested in Paris and confined in the Vélodrome d’Hiver (Winter Velodrome) before being shipped to concentration camps. But this is followed by the information about the arrest on 23 September outside his home and the precise number of the convoy, details which lend credence to the second account.] Preface: Émile Benveniste, a Linguist Who Neither Says Nor Hides, but Signifies Julia Kristeva

What makes a great linguist? Great linguists are distinguished by how, with their knowledge and analysis of languages, they discover properties of language through which they interpret and change speakers’ ‘being in the world’. I hazard this defini- tion so as to put Émile Benveniste (1902–1976) into the per- spective of certain of his predecessors whose work, although meticulous and cold in appearance, nonetheless contributed to and accelerated some of the most decisive steps in the human adventure. Think of the humanists and grammarians of the sixteenth century such as Scaliger and Ramus, whose analysis of the relationship between language and thinking, from Latin to the modern languages, helped to lay the ground for the development of national languages; of Lancelot and Arnauld, whose Grammaire générale et raisonnée (General and rational grammar, 1660), even more than their Logique de Port-Royal (Port-Royal logic, 1662), by introducing the notion of ‘sign’, by trying to determine ‘what is mental in language’ and by basing their judgement on ‘grammatical usage’, inscribed the Cartesian subject into the language’s syntax; of nineteenth-century ‘his- toricism’ and the comparative philology of Franz Bopp and Rasmus Rask, then finally Humboldt, who, extending Schlegel’s and Herder’s insights into the relationship of Sanskrit with the European languages, confirmed the weight of history in the evolutive activity of language. The tragic conflicts of the twentieth century tend to over- shadow the fact that it was also a period of exceptional explora- tion of how language figures at the heart of the human condition: the central activity is the language which conditions, contains

1 2 Last Lectures: Collège de France 1968 and 1969 and clarifies all human experiences. Phenomenology, formal logic, analytic philosophy, structuralism, generative grammar, the human sciences interrogating in language the meaning of behaviours and institutions – not forgetting psychoanalysis, which annexes sex and trespasses on biology – developed­ alongside an unprecedented explosion of literary forms, artistic avant-gardes and stylistic individualities that rocked the domain of letters. A lucid adventure which, in hindsight, seems to pre- figure the explosion of systems of conventional signs and the tide of new virtual hyperlinks which promise as much freedom as chaos. In the middle of this profusion in which it is fully inscribed, the work of Émile Benveniste – if we take the trouble to tease out how the complexity of his thought resonates with the advances in philosophy and the human sciences and the new forms of art and literature – makes contact with the twenty- first century and its challenges. For it profoundly clarifies the universal properties of languages underlying the creative freedom of the human mind, to which it unceasingly holds a stethoscope. Readers attentive to Benveniste’s trajectory, who do not let their attention get detoured toward a linguistics under pressure to produce technical innovations in a society losing meaning and encircled by political ‘spin’, will discover in his Last Lectures that his ‘general theories’ contribute to probing a deep logic which crosses over to our digital writing. Are they ‘chats’ lacking ‘subjectivity’, or on the contrary, routes to the ‘engenderment’ of new ‘signifiances’? Émile Benveniste was an austere scholar, a very great connois- seur of ancient languages, an expert in comparative grammar, an authority on general linguistics. He knew Sanskrit, Hittite, Tocharian, Hindi, Iranian, Greek, Latin, all the Indo-European languages, and in his fifties he plunged into American Indian languages. Yet his work, of an impressive daring, though restrained and modest on the surface, remains relatively little known, and less understood. Born in Aleppo, Syria, in 1902 to a polyglot Jewish family,1 Ezra Benveniste emigrated to France in 1913, enrolling in the ‘little seminary’2 of the Rabbinical School of France. His exceptional predisposition for languages drew the attention of Preface 3

Sylvain Lévi, who introduced him to the great Antoine Meillet (or possibly it was Salomon Reinach who introduced them).3 Ezra Benveniste entered the École Pratique des Hautes Études (EPHE) in 1918, earned his licence ès lettres the following academic year, obtained the agrégation de grammaire in 1922, after which, as a pure product of the lay teaching of the French Republic, he became a naturalised French citizen in 1924 and chose the name Émile. During these student years he made close links with young philosophers and linguists, students of the École Normale Supérieure, all of them to a greater or lesser degree rebels, libertarians, antimilitarists, even communist sym- pathisers, and he notably crossed paths with the Surrealists. He left for India in 1924 as tutor in a family of wealthy industrial- ists, before having to fulfil his military duties in Morocco in 1926. On his return to France, he became the student of Meillet, whom he succeeded as Professor of Comparative Grammar and Director of Studies in the EPHE, where he exerted a strong influence on his colleagues. He entered the Collège de France in 1937, again succeeding Meillet in the Chair of Comparative Grammar. He was made prisoner of war in 1940–1, succeeded in escaping and taking refuge in Switzerland, at Fribourg (where also resided Balthus, Alberto Giacometti, Pierre Emmanuel and Pierre-Jean Jouve), and so escaped Nazi persecution, but in Paris his flat was looted and his brother Henri was arrested and sent to Auschwitz, to be murdered there in 1942. Along with the greatest names of the Jewish intelligentsia (Benjamin Crémieux, Georges Friedman, Henri Lévy-Bruhl, et al.), he signed the collective letter organised by Marc Bloch dated 31 March 1942 and addressed to the Union générale des israélites de France (UGIF), drawing attention to the Vichy policy making Jews a separate category of citizens, a prelude to their transportation to the death camps.4 After the war, Benveniste returned to teaching in the EPHE and the Collège de France, training several generations of students, carrying out linguistic field research in Iran, Afghanistan and then Alaska, and par- ticipated in numerous international linguistics conferences. He became a member of the Institute (Académie des Inscriptions et Belles-Lettres) in 1960, Director of the Institute of Iranian Studies in 1963, and President of the International Association 4 Last Lectures: Collège de France 1968 and 1969 for Semiotic Studies in 1969. On 6 December 1969 he suffered a stroke that left him handicapped for seven years, until his death, putting an end to his career. This concise biography of an ‘agnostic Israelite’, a French nomad, is above all that of a man who made language the path of a life, and through his work transmitted to us his thinking of this experience. Benveniste left an ‘unfinished’ body of work, it is sometimes said, in a phrase that risks diminishing the importance of the texts. Unfinished, certainly, the stroke having left the man in the intolerable situation of a great linguist deprived of speech and paralysed. But ‘unfinished’ too in an absolutely necessary sense, because such is the experience of language that he lived and theorised in a century during which diverse currents of thought, multiplying routes and interrogations on both the epis- temological and aesthetic levels, imposed on the man anchored in his time the Heraclitean refusal to ‘say’, to construct a closed ‘message’, given definitively in a completed system. At the heart of this burgeoning diversity to which he was always attentive (from comparative philology to Saussure, from structuralism to Chomskyan syntax, from Surrealism to 1960s experimental literature), he practised what can only be called a Benvenistian style of thought, in which morphosyntactic detail joins the permanent interrogation of fundamental categories, linguistic and/or philosophical,5 and which is characterised, beyond the refusal to ‘say’, by an avoidance of the aestheticism that ‘hides’ (though he was once sensitive to it, as witnessed by his literary self-analysis, Eau virile [Virile water]),6 by the wish to ‘signify’ (open up to thinking, problematise, question) and to deter- mine how signifying is engendered in the formal apparatus of language. What is it then to ‘signify’? The metaphysical question led Benveniste to look for a ‘material’ solution, in the very function- ing of language: ‘this signifies’ is synonymous for him with ‘this speaks’, and so it is without recourse to any ‘external’ or ‘tran- scendental reality’, but in the ‘properties’ of language itself, that he prospects and analyses the possibilities of meaning-making which are specific to this ‘signifying organism’ that is the speak- ing human. Preface 5

So the young man born in the heart of the Ottoman Empire, his studies funded by the Universal Israelite Alliance, did not become a rabbi.7 At a point in history when the Six-Day War (1967) and the Yom Kippur War (1973) had not yet aroused in many agnostic Israelites the desire to return to the God of their Fathers, it was through what Heraclitus said of the Lord whose oracle is at Delphi, ‘Oute légei, oute krýptei, alla semaínei’8 (unless it translates the unpronounceable tetragram YHWH: the being identified with what is and will be, with ‘signifiance’) that he resumed his ambition to study the ‘signifying power’ in the properties of language. A path, precisely, which ‘neither says nor hides, but signifies’, and which leads from the study of (explicitly) Presocratic Greece, the Bible and (implicitly) the Gospels, to that of the modern sciences produced by secularisa- tion, and most especially general linguistics, which he set out to modulate in such a way that it could analyse how language is organised so as to create meaning (First Lecture).

Double Signifiance Benveniste thus apprehends ‘meaning’ by abstracting away its philosophical, moral or religious ‘value’. The search for meaning in its linguistic specificity is what ‘will command our discourse about languages’ in the Last Lectures restored here through the efforts of and with an Introduction by Jean-Claude Coquet and Irène Fenoglio. ‘For our part we posit [my italics: J. K.] that a language, in its essential nature which commands all the functions it can assume, is its signifying nature.’ The ‘signifi- ance’ which ‘informs’ the language thus posited is a property that ‘transcends’ ‘any use, particular or general’, or again a ‘characteristic we foreground: a language signifies’. He is speaking on 2 December 1968, seven months after the legendary May ’68. A naïve reader, then as now, is surprised: is this so original? What use is a language if it does not signify something? Of course. But do you know exactly what you mean by ‘signify’? And whether ‘communicate’, ‘mean’, ‘contain a message’ are not confused with ‘signify’? Central to the phi- losophy of language, but as a bearer of ‘truth’, meaning is not really the problem of linguists, Benveniste reminds us. Meaning 6 Last Lectures: Collège de France 1968 and 1969 is left ‘outside linguistics’:9 either ‘discarded’ because suspected of being too subjectivist, fleeting, indescribable by aspects of lin- guistic form; or recognised but ‘reduced’ (Leonard Bloomfield, Zellig Harris) to morphosyntactic structural variants, ‘distri- butional’, in a ‘given corpus’. For Benveniste, on the contrary, ‘signifying’ constitutes an ‘inner principle’ of language (Third Lecture). With this ‘new idea’, he emphasises, ‘we are thrown into a major problem, which embraces linguistics and beyond’. If certain precursors (John Locke, Saussure and Charles Sanders Peirce) showed that we ‘live in a universe of signs’ of which the language is the first, followed by the signs of writing, of recogni- tion, of rallying, etc. (First Lecture), Benveniste intends to show how the formal apparatus of the language enables it not only to ‘denominate’ objects and situations, but above all to ‘generate’ discourses with original significations, individual yet share- able in exchanges with others. Better still is how, not content with self-generation, the organism of a language also generates other sign systems that resemble it or increase its capacities, but amongst which it is the only signifying system capable of furnishing an interpretation. Benveniste’s papers collected in Volume 1 of his Problèmes de linguistique générale (PLG 1, Problems of general linguistics, Paris: Gallimard, 1966), whilst relying on the study of ancient languages and on comparative linguistics, already offered answers to these theoretical questions. A second Benveniste, clarifying and displacing the principal interrogations of his first general linguistics, appears in Volume 2 of Problèmes de linguistique générale (1974), published after he had suffered his stroke, in which are gathered together articles written from 1965 to 1969. A careful reading of these two volumes reveals two major stages in the evolution of his thinking, which readers of the present work need to be aware of in order to grasp fully the innovative approach of the Last Lectures. Starting in the first volume of his master work, the theo- retician proposes a general linguistics which diverges from structural linguistics but also from the generative grammar that dominates the linguistic landscape of the period, and puts forward a linguistics of discourse, based on allocution and dialogue, opening the utterance toward the process of enuncia- Preface 7 tion, subjectivity and intersubjectivity. In the wake of analytic philosophy (performative utterances) but also of Freudian psycho­analysis, Benveniste conceives of subjectivity in enuncia- tion as a more complex emitter than the Cartesian subject, since he broadens it to the ‘intentional’ (borrowed from existential phenomenology). Additionally, and without appearing to, he sketches an opening toward the subject of the ‘unconscious’. Not really ‘structured like a language’ but worked by an (instinctual?) ‘anarchic force’ which language ‘restrains and sublimates’, although through ‘rips’ the force can introduce into the language a ‘new content, that of unconscious motiva- tion and a specific symbolism’, ‘when the power of censorship is suspended’.10 A new dimension of general linguistics according to Benveniste is however revealed in the second volume. In dialogue with Saussure and his conception of signs as the distinctive elements of the linguistic system, Benveniste proposes two types in the signifiance of language: the semiotic and the semantic.11 The semiotic (from semeion ‘sign’, characterised by its ‘­arbitrary’ link – the result of a social convention – between ‘signifier’­ and ‘signified’) is a closed, generic, binary, ­intralinguistic, systematising and institutional meaning, defined by a relationship of ‘paradigm’ and ‘substitution’. The semantic is expressed in the sentence which articulates the ‘signified’ of the sign, or the ‘intended’ (frequent allusions to the phenomeno- logical ‘intention’ of Husserl, whose thought influences certain linguists such as Hendrik Josephus Pos). It is defined by a rela- tionship of ‘connection’ or of ‘syntagm’, in which the ‘sign’ (the semiotic) becomes ‘word’ through the ‘activity of the speaker’. This activity activates the language in the discourse situa- tion addressed by the ‘first person’ (I) to the ‘second person’ (you), the third (he/she) being situated outside discourse. ‘On this semiotic foundation, the language-discourse constructs a semantics of its own, a signification of the intended produced by the syntagmation of words in which each of them retains only a small part of the value that it has as a sign.’12 First formulated in a 1966 paper to the Congress of the Society of French Language Philosophy in Geneva,13 then in his 1968 address to the founding Warsaw Symposium of the 8 Last Lectures: Collège de France 1968 and 1969

International Semiotics Association,14 this dual conception of signifiance opens a new field of research. Benveniste insists on surpassing the Saussurean notion of the sign and of language as system, and underscores their importance, which is simultane- ously intralinguistic – opening a new dimension of signifiance, that of discourse (the semantic), distinct from that of the sign (the semiotic)15 – and translinguistic – elaborating a metase- miotics of texts and works, on the basis of the semantics of enunciation.16 And he gives a more precise idea of the immense perspectives thus opened up: ‘We are utterly at the beginning’, therefore it is still ‘impossible to define in a general way’ where this orientation will lead, which, spanning linguistics, ‘will oblige us to reorganise the apparatus of the human sciences’.17 Benveniste’s Last Lectures pursue this reflection whilst relying on a new continent, that of poetic language, as witnessed by his manuscript notes devoted to Baudelaire,18 which develop the key notions of the lectures whilst relocating them. Between the second volume of the Problems of General Linguistics and the manuscripts devoted to poetic language, the Last Lectures propose first of all to show that ‘to signify’, which constitutes the ‘initial, essential and specific property of a language’, is not enclosed in the sign-units (as conceived by Saussure), but ‘transcends’ the communicative and pragmatic functions of language; and then, secondly, to specify the terms and strategies of this ‘signifiance’ insofar as it is a literally vital ‘experience’ (as he had suggested in ‘La forme et le sens’, PLG 2, p. 217: ‘Well before serving to communicate, language serves for living’). Very logically, Benveniste introduces this reflection through a homage to Saussure and Peirce. He recognises Saussure’s ‘particular importance’ and defines his work as a ‘new moment of analysis’, a ‘fundamental step in the history of thought’ (Lecture of 1 Dec. 1969) where, for the first time, ‘the notion of sign’ and of ‘science of signs’ take shape (Third Lecture). Regarding Peirce, he mentions the ‘universal notion’ of the sign divided into three ‘classes’ and detailed in multiple ‘categories’, based on ‘a triad’ that is again ‘universal’ (Second Lecture). But this lucid acknowledgement of debt to his predecessors offers an occasion to make their limitations uncompromisingly Preface 9 clear. Thus Saussure ‘does not rely on the sign’; leaves open a possible ‘exteriority’ of the sign; does not take up the question of the relations between sign systems and the ‘specificity of the language’, which ‘produces’ (‘engenders’) new sign systems, insofar as it is their only ‘interpretant’; or again ‘does not apply himself to the language as production’ (Final Lecture). Peirce, for his part, does not base his theory on the language, but only on the word; his theory excels in its description of the numer- ous diversities of signs, but it ignores the language, and its logic lacks a systematic organisation of the different types of signs (Third Lecture).19 This inventory contributes to clarifying anew the challenge of Benveniste’s new general linguistics: ‘We need to prolong this reflection beyond the point indicated by Saussure’ (Fourth Lecture). And this, notably, by developing a ‘new relation’, absent in Saussure: the ‘relation of interpretation amongst systems’. The language, precisely – unique within the diversity of signifying systems in that it has the capacity to auto-interpret and to interpret other systems (music, image, kinship) – is ‘the interpreting system’: it ‘furnishes the basis of the relationships which permit the interpreted to develop as a system’. The language is, from this point of view, hierarchically the first of the signifying systems, which maintain amongst themselves a relationship of engenderment (Fifth Lecture).

Writing: Centre and Relay The ‘double signifiance’ of the language, as sketched above, is developed by the lever of writing, which realises and reveals its capacity for ‘production’ and ‘engenderment’. However, and although the term ‘writing’ is at the centre of philosophical and literary creation in France,20 the linguist does not refer to this explicitly, but constructs the concept of it within the frame of his general theory of the signifiance of the language. To distance himself from Saussurean semiology, which, by ‘confus[ing] writing with the alphabet and the language with a modern language’, postulates that writing is ‘subordinate to the language’ (Eighth Lecture), Benveniste interrogates the act of writing, the learning of writing and the types of writing that 10 Last Lectures: Collège de France 1968 and 1969 have developed across history. Taking care however to empha- sise that he is not seeking the ‘origin of writing’, but the various solutions to the ‘graphic representation’ of signifiance (Ninth Lecture). It will first involve putting into question the ‘maximally inti- mate’ relation that the civilisation of the book has elaborated amongst writing, languages, speech and thinking, in other words, disentangling them in order to envisage writing ‘in itself and for itself’ as a particular ‘semiotic system’. Thus decoupled from speech, writing appears as a ‘high level of abstraction’: the writing speaker is extracted from ‘living’ verbal activity (ges- tural, phono-acoustic, connecting with another in a dialogue) and ‘converts’ it into ‘images’, into ‘hand-traced signs’. With consequent losses, certainly, as the image replaces speech as tool of ‘exteriorisation’ and ‘communication’. Yet subtle benefits make up for these losses, even abstracting away the ‘utilitarian’ function of writing (memorialising, trans- mitting, communicating messages). The ‘first great abstraction’, writing, by making language into a ‘distinct reality’, detached from its contextual and circumstantial richness, allows speaker- writers to realise that their language or thinking is made up of ‘words’ represented in material signs, in images. What is more, this ‘iconisation of thinking’ (Eighth Lecture) is the source of a ‘unique experience’ of the ‘speaker with himself’: the speaker ‘becomes aware’ that it is ‘not from pronounced speech, from language in action’, that writing proceeds. ‘Global’, ‘schematic’, ‘non-grammatical’, ‘allusive’, ‘rapid’, ‘incoherent’, this inner language, ‘intelligible to the speaker and to him alone’, con- fronts the speaker with the considerable task of carrying out a ‘conversion operation’ of his thinking into a form intelligible to others. Thus understood, the ‘iconic representation’ constructs speech and writing together: it ‘goes hand in hand with the elab- oration of speech and the acquisition of writing’. At this stage of his theorisation, and contra Saussure, Benveniste remarks that, far from being ‘subordinate’, the iconic sign associates thinking with graphism and with verbalisation: ‘The iconic representation would develop in parallel with the linguistic rep- resentation’, which allows us to glimpse a different relationship Preface 11 between thinking and icon, ‘less literal’ and ‘more global’ than the relationship between thinking and speech (Eighth Lecture). This hypothesis associating writing with ‘inner language’, which will be modified further on, takes up Benveniste’s previous enquiry into the ‘anarchic force’ of the Freudian unconscious.21 Would the ‘inner language’ that writing seeks to ‘represent’ be linked to the ‘failings’, ‘games’, ‘free ramblings’, the origin of which Benveniste, a reader of Freud and the Surrealists, would discover in the unconscious? The concise notes on writing in the Last Lectures recall the linguist’s earlier work, and complete the phenomenological intended which he inserts into the seman- tic of discourse through a ‘motivation’ of a different order. The ‘inner language’ of the speaker-writer would not be limited to the propositionality belonging to the transcendental ego of the conscious and its ‘intention’, but could evoke, in his theory of subjectivity, a diversity of subjective spaces: typologies or topologies of subjectivities in the engenderment of signifiance. Baudelaire’s ‘poetic experience’, we shall see, confirms and specifies this progress. As for the history of writing, it brings a new adjustment of the language/writing relationship, and constitutes a new step in Benveniste’s theory of signifiance. Pictographic writing, a sign of external reality, ‘recites’ a message already constituted by ‘the language of another’ (Ninth Lecture): it ‘does not speak’ in the sense that a speaking lan- guage is a ‘creation’. As far back as we can go in its prehistory, writing ‘describes’ ‘events’: if it is ‘parallel’ to language: it is not its ‘decal’. This observation raises a question that remains in suspense: does the specificity of the pictogram, which ‘recites’ (re-produces) but does not ‘create’ (does not produce), remain mutely latent in every iconisation of language? Isn’t this particu- larity more marked in certain modern writing systems (numeri- cal, for example)? And if so, under what conditions? With what consequences for the subject of the enunciation? Two epoch-making revolutions in the history of writing shed light on the double signifiance of the language. The first lies in the discovery of a graphic mark (graphie) reproducing the phonè in a limited number of signs, which comes down to reproducing, no longer the content of the message as the bearer 12 Last Lectures: Collège de France 1968 and 1969 of events, but the message’s linguistic form. ‘Luck’ befalls China: less because in the monosyllabic Chinese language each semiotic unit is a formal unit which cannot be broken down (a word = a syllable) than through the inventive genius of those who conceived it, who succeeded in attributing a sign (graphè: character) to each ‘signifier’ (phonè) – with ‘keys’ to disambigu- ate homophones (Eleventh Lecture). To become aware of the flux of speech, to break words down, to realise that they are polysyllabic: this process entails a higher- level segmentation. For polysyllabic languages, it will be seg- mentation into syllables, with variants: Sumerian and cuneiform writing where ‘The affiliation is clear between certain images and their referent’; its adaptation to Akkadian (Semitic); the ‘rebus’ method in Egyptian hieroglyphs (a picture = a syllable: /ša/ chat ‘cat’ + /po/ pot ‘pot’ = /ša po/ chapeau ‘hat’) (Tenth Lecture). A ‘decisive step’ in the history of the ‘graphic representations’ of a language is taken with Semitic alphabetic writing systems. Hebrew is a major example, which Benveniste does not develop here particularly, although he recalls its specific organisation: the consonantal schema carries the meaning (the semantic), whilst the grammatical function lies with the vowels. The Greek alphabet, in contrast, breaks down the syllable itself and gives the same status to consonants and vowels. This change reveals the role of the voice in every verbal articulation – ‘The breakdown unit of speech will thus be either a vowel or a segment including a vowel (CV or VC).’ For the linguist, too, the syllable is ‘a sui generis unit’ (Eleventh Lecture) which makes it possible to reproduce the ‘natural articulation of speech’ in writing and to materialise the grammatical relations with which this language makes subjective positions explicit in the act of enunciation. Two types of languages are defined by this metasemiotic treatment of the relation they have to writing: those in which etymology or semantics predominate (Phoenician and Hebrew); and those in which consonants are distinguished from vowels, and where grammatical variations, which often destroy ety- mological relationships, lead to a refinement of the inflectional system (morphological modifications by affixation expressing grammatical categories). Preface 13

A ‘consubstantial’ relationship between writing and language is thus defined and can be expressed in these terms: the types of writing accomplish auto-semiotisation, that is, the becom- ing aware of the language types to which they correspond (‘the writing system has always and everywhere been the instrument that has permitted a language to semiotise itself’, Twelfth Lecture). Together, writing and language constitute different types of signifiance. And since languages understood as experi- ences of enunciation ‘contain’ the referent quite as much as the subjective experiences of speakers in their acts and discursive exchanges (First to Seventh Lectures), these types of writing systems reveal, consolidate and recreate very different ways of being in the world. Thus a rather clear ‘dividing line’ is drawn: to the East (in Mesopotamia, in Egypt and all the way to China) ‘civilisations of the written’ predominate, characterised by the primacy of writing, where the scribe (the ‘sage of calligraphy’ in China) plays a central role in the organisation of society; whereas to the West, in the Indo-European world, a deval- orisation, even a certain ‘disdain’ of writing prevail (in Homer, graphoˉ merely signifies ‘scratch’) (Fourteenth Lecture). This barely sketched-out typology of signifiances across types of writing systems already appears rich in potentialities for research in semantics and in semiology of enunciation. Thus one could envision (Fourteenth Lecture), amongst other routes, determining the semiotic and semantic specificities of biblical texts, and of delving into the subjectivity of its speakers and its intended audience. Or enquiring into the opposition set up by Saint Paul between ‘letter’ and ‘spirit’: must it be understood as a dyad joining, on the one hand, the biblical semiotic (the ‘letter’) which is always already semantic in Hebrew words – through the polysemic imprint of graphism memorising the message or the tradition’s history – and, on the other hand, the discourse of an evangelical subjectivity which is actualised in the time of expression, of appearing and of discursive com- munication – manifested and clarified by the categories and modalities of Greek grammar? How do we understand that with ‘the new notions attached to the written/alphabetic’ there appears ‘lay civilisation’ (Fourteenth Lecture)? Must we deduce from this that the diversity of writing systems (notably by the 14 Last Lectures: Collège de France 1968 and 1969

­intermediary of translations from Latin into the vernacular languages) and other sign systems which widen the auto- semiotisation of language in a secularised civilisation enhance its capacity for engendering meaning, and predispose the sub- jectivities present in it to create new signifying experiences? Or, on the contrary, that a certain ‘laicity’ coming in the wake of Christianity could privilege the semantic of a discourse for com- municants, to the detriment of the semiotic of inner language? Without further exploring these barely sketched programmes, and without going in the direction of linguistic relativism either, but opening perspectives complementary to those proposed by Edward Sapir, Benveniste keeps strictly to the plane of general linguistics and marks a new step in its thinking. In the light of how different types of writing participate in the revelation and development of the double signifiance of languages, the author maintains that writing is not simply parallel to a language (and to language types), but that it prolongs them. Iconisation trig- gers and refines the language’s formalisation, so that writing is progressively lettered. ‘It semiotises everything’: writing is a sign system that ‘may be said to have a much closer resemblance to “inner language” than to the discourse chain’ (Twelfth Lecture). A new characteristic of ‘inner language’ is specified here: ‘before’ even the sacred scribe (who semanticises the language from the outset, by the semantic graphism of Semitic syllabic writing systems; or by inventing Chinese characters, where each signified has its image), it is logically inner language which ‘consecrates’ by formulating ‘myth’. And this ‘inner’ narrativ- ity, this ‘train of ideas’, such as a writing of ‘globality’, tells a ‘whole story’. Is it all a kind of ‘fiction’, about which Husserl said that it constitutes the ‘vital element of phenomenology’? Or is it a Benvenistian variant of Freud’s ‘originary phantasm’, which is given to, and given by, ‘free associations’? Or again, does it have to do with these ‘narrative envelopes’ (much more than with ‘syntactic competences’) that cognitivists suppose are the first holophrases of the child beginning to speak? Certainly, whatever the case, poetic language – ‘internal to language’, ‘created by the choice and marrying of words’,22 and written in condensed metaphorical tales (one thinks of the verses of Baudelaire and Rimbaud: ‘Mother of memories, mistress of Preface 15 mistresses’; ‘Vast as the night and as the light’; ‘Behold the Holy City, seated in the West’) – is a manifestation of it. Benveniste succinctly evokes this line of research, always returning to general linguistics and the signifying function that language fulfils. ‘Every social behaviour’, including relations of production and reproduction, does not pre-exist language, but ‘consists in being determined’. ‘Encircling’ or ‘containing’ the referent, a language ‘carries out a reduction of itself’ and ‘semiotises’ itself: writing being the ‘relay’ which makes this faculty explicit. In sum, writing makes explicit and definitively reinforces the non-instrumental and non-utilitarian nature of a language, which, because of this and more than ever, is neither a tool, nor communication, nor dead letter, but a ‘signify- ing organism’ (Aristotle, in Twelfth Lecture), generating and auto-generating. Having reached this point, Benveniste reverses the initial hypothesis concerning writing. As an ‘operation’ in the ‘lin- guistic process’, writing is ‘the founding act’ which has ‘trans- formed the face of civilisations’, ‘the most profound revolution humanity has known’ (Fourteenth Lecture). This particularity of writing in its relation to a language thus reinforces a final observation: the language and its writing ‘signify in exactly the same way’. Writing transfers signifiance from hearing to vision, it is ‘speech in a secondary form’. Since speech comes first, ‘writing is a transferred speech’. ‘Hand and speech stand together in the invention of writing’, writes Benveniste. The writing/speech relation is equivalent to the relation of heard speech to enunciated speech. Writing reappropriates speech in order to transmit, communicate, but also recognise (this is the semiotic) and understand (this is the semantic). Writing is a stakeholder in the language’s interpretance. This relay of speech fixed in a system of signs remains a system of speech, on condi- tion of the latter being understood as a signifiance susceptible to further engenderments by other sign systems. All the way to online blogs and Twitter . . . It is certainly not by chance if, at the heart of this work in progress on the modalities of the specific signifiance of lan- guage, there intervenes a recollection of Plato’s Philebus: within the variety of human sensations and pleasures, each One is an 16 Last Lectures: Collège de France 1968 and 1969

Infinity, and the only way of opposing the absence of limits in the state of nature is to have recourse to numbers, thanks to which it becomes possible to delimit the units in a hierarchical order, to dissociate and to identify. Like the ‘notes’ in music, the letters in grammar (grammatikē technē) are ‘numbers’: in this sense the activity of the grammarian, who enumerates and organises the semiotic in language, beneath the level of signification, is ‘divine’ (Thirteenth Lecture). Recalling this parallel between the analysis of language and the work of the Egyptian Theuth (Thōth in Greek) who was the first to perceive that vowels are ‘infinitely multiple’, Benveniste appropriates the idea of ‘number’ to articulate the idea of limit, unavoidable in ­linguistics – which is about ‘dissociating and identifying units on several levels’, ‘arriving at numbers/at a limit’ – and the idea of the creation of the world through the Word. But it displaces the onto-theology of transcendent meaning and weaves the conno- tations of this ‘transcendence’ (announced in the First Lecture), always inflected inside language, and continuing to construct itself under the eyes of the reader of these lectures: ‘The man instructed in letters, the grammatikos, is the man instructed in the structure of the language.’ ‘The relationship of the one and the multiple is found simultaneously in knowledge (epistēmē) and in the experience of sensations’ (Thirteenth Lecture). Step by step, Benveniste’s theory thus integrates every refer- ent and, implicitly, the infinity of the res divina – by definition external to the human world – in and through the signifiance of language. On this score he relies on Socrates, as we have seen, to whom could be added the fourth book of the Pentateuch, the Book of Numbers, or the Kabbala, which constructs meaning by enumerating. But more than any other, the fourth Gospel, that of John, seems to be the touchstone for this dual signifi- ance of the language, encompassing its graphic representation, the act of writing and the variants of writing systems, as well as intersubjectivity and the referent: ‘In the Beginning was the Word.’ With this slight difference, that, without ‘beginning’, the ‘divine’ is reabsorbed into the engenderment of the ‘folds’ (Leibniz) of signifiance:23 in the elements and categories of this ‘datum’ that is language. Linguists never seek the ‘truth’ condi- tions of this datum, nor its infinite translinguistic configura- Preface 17 tions, potential and future, but are content to ‘try to recognise its laws’.24

Signifiance and Experience Taken at its ‘fundamental’ level (as distinct from ‘contingent’ empirical languages), once the language system has become ‘signifiance’ it is not simply a complement added to the theory of the Saussurean sign coextensive with the ‘social contract’. In taking up the idea that linguistic structures and social structures are ‘anisomorphic’, Benveniste strives to show that the act of signifying is irreducible to communication and institutions, and that it only transcends the ‘given meaning’ through the ‘activity of the speaker put at the centre’. The notion of ‘enunciation’ understood as an ‘experience’ considerably modifies the object of signifiance and/or of language.25 Far from abandoning the ‘sign’, signifiance includes it in ‘dis- course’ as an intersubjective illocutionary act which transmits ‘ideas’. Signifiance is a syntagmatic organisation comprising the various types of syntactic constructions, and on that account ‘contains’ the ‘referent’ of Saussurean linguistics,26 on condi- tion of enriching it with the ‘unique situation’, the ‘event’ of the enunciation which implies ‘a certain positioning of the speaker’. The ‘experience’ of the subject of the enunciation in the intersubjective situation is what interests the linguist, but as it transverses the ‘formal apparatus’ of the ‘intended’: that is, the ‘instruments of its attainment’ as much as the ‘processes through which linguistic forms are diversified and engendered’. The ‘singular dialectic of subjectivity’, ‘independent of any cul- tural determination’, had certainly been proclaimed previously (PLG 2, p. 68). But through writing, the Last Lectures deepen the ‘engenderment’ of signifiance by displacing subjective expe- rience from a dialogic exchange of I and you toward a topology of the subject of the enunciation that downgrades Descartes’ ego cogito as well as the Husserlian transcendental ego. The terms designating this dynamic of language vary: ‘engen- derment’, but also ‘functioning’, ‘conversion’ of the language into writing and into discourse, ‘diversification’; the language being defined as ‘production’, ‘moving landscape’, ‘place of 18 Last Lectures: Collège de France 1968 and 1969 transformations’. But contrary to the ‘transformations’ which interest generative grammars and for which syntactic catego- ries are immediately given, the ‘engenderment’ of signifiance according to Benveniste is deeply engaged in the coming about of pre- and translinguistic signification, and involves three types of engenderment relationships: a relationship of interpretance (a fundamental property, a language being ‘the only system that can interpret everything’); a relationship of engenderment (between sign systems: from alphabetic writing to Braille); a relationship of homology (with reference to Baudelaire’s ‘cor- respondences’). The final lecture revisits each of these, whilst recalling the necessity of revising the ‘formal categories’ (‘cases’, ‘tenses’, ‘moods’), and posits that ‘the entire inflectional appara- tus is in question here’. The subject of the enunciation must himself be affected by this mobility. In this moving landscape of the language, and with regard to the writing which has contributed to making it appear, reflection was required on the specific experience of writing that ‘poetic language’ represents. In fact, Benveniste, in counterpoint to the structuralist reading of Baudelaire’s ‘Les chats’ by Roman Jakobson and Claude Lévi-Strauss,27 and echoing the indications of the Last Lectures, raises the topic in his manuscript notes on Baudelaire of the same period (1967–9). Closer to ‘inner language’ than to discourse, poetic language requires the analyst to ‘change instruments’, as Rilke put it (noted, we have seen, by the young Benveniste). This ‘different language’ that poetry would constitute necessitates therefore a ‘translinguistics’, because the ‘signifiance of art’ is ‘unconven- tional’ and because its ‘terms’, which proceed from the specific individualities of each writer-subject, are ‘unlimited in number’. Immediately, Benveniste establishes what the individualities of this ‘translinguistics’ are: the poetic message, ‘exactly reversing the properties of communication’,28 speaks an emotion that lan- guage ‘transmits’ but does not ‘describe’.29 Similarly, the referent of poetic language is ‘inside the expression’, whereas in every- day language the object is outside language. It ‘proceeds from the poet’s body’, ‘these are muscular impressions’, Benveniste specifies. Poetic language, ‘sensitive’, ‘is addressed only to the entities which participate in this new community: the poet’s Preface 19 mind, God/Nature, the absent one/the creature of memories and of fiction’. Why does Benveniste choose Baudelaire to illustrate his proposition? Because he brought about the ‘first fissure between poetic language and non-poetic language’, whereas in Mallarmé this break is already consummated.30 Contemporaneously with the Last Lectures, these notes on Baudelaire’s poetic experience connect with Benveniste’s reflec- tions on the ‘anarchic force’ at work in the unconscious and which the language ‘restrains and sublimates’.31 The expression of an ‘instantaneous and elusive subjectivity which forms the condition of dialogue’, this experience participates in the infra- and the supra-linguistic,32 or rather in the translinguistic.33 The translinguistic, which applies to works, will be based on the semantic in the enunciation. These last reflections, attentive to the poetics of ancient India as they appear in the sacred texts thoroughly mastered by Benveniste the Sanskritist, resonated with the end of the 1960s, a time when social and generational upheaval, calling for ‘power to the imagination’, sought secret and innovative logics of meaning and existence in the experience of writing (avant-garde or feminist). In hindsight, and in the absence of any explicit reference to psychosexuality, it is not the Freudian theory of sublimation that this general linguistics of experience and subjectivity brings to mind, but the journey – unnamed – of Martin Heidegger. Indeed, according to Being and Time (1927), language is discourse (Rede) or speech, words having no signification outside the Mitsein of dialogue. It is the responsibility of Dasein to interpret: its localisation in the existential analytic is taken into consideration, to the detriment of language as such. We are overly dependent on certain resonances between this early conception of language in Heidegger and the early general linguistics of Benveniste (PLG 1, 1966), which pro- posed placing the formal apparatus of this language regime – ‘discourse’ and ‘interpretant’ – in society and nature. The Heideggerian approach changes in On the Way to Language (1959), where language is envisaged as ‘the said’, Sage, ‘what is spoken’. Dialogue becomes monologue, without however being solipsistic, but, inasmuch as it is ‘inner discourse’, never 20 Last Lectures: Collège de France 1968 and 1969 propositional, without ‘sound’ or ‘­communication’, its ‘inner thinking’ achieves in silence the mental production of a ‘coming into language’. For Benveniste, writing as graphism and as poetic experience – from Baudelaire to surrealism – seems to cross Heidegger’s definition of ‘language that speaks only and solitarily with itself’, and makes sonority possible. But to dis- tance itself from it immediately, since the allusive remarks of the Last Lectures and the manuscript notes on Baudelaire place this ‘letting go’ which would be the essence of language, deafly threatened with becoming ‘meaningless’ in the later Heidegger, in apposition (more than opposition) to the vigilance of the linguist, for whom ‘discourse includes both the limit and the unlimited’, ‘unity and diversity’ (Thirteenth Lecture). In fact, Benveniste never fails to insist on ‘syntagmation’ – probably ‘reflecting a necessity of our cerebral organisation’34 – which confers on the ‘instrument of language’ its capacity to encode whilst codifying, to limit whilst being limited, and thus to assure the semantic of an intelligible discourse; communica- tive, engaged with reality. He adds however that, parallel to the language, and as its relay, writing as graphic representation and as poetic experience, although closer to ‘inner language’ than to ‘discourse’, does not eliminate its pragmatic virtues. But it risks shifting the boundaries of the language by engendering signifying systems that are singular (the poem) and yet share- able in the ‘interpretance’ of the language itself. Neither insti- tutional tyranny nor dreaming hymn, signifiance as sketched by Benveniste at the end of his career is a space of freedom.

‘Linguistics is universal’35 Today everyone communicates, but rare are those who perceive the consistency and the full extent of language. At the time when Benveniste was giving his Last Lectures, the idea that language determines humans in a different and more profound way than social relationships do was starting to become a dan- gerous way of thinking: a veritable revolt against conventions, the ‘Establishment’, the ‘Police State’, doctrinaire Marxism and communist regimes. In Warsaw, in Italy, in Czechoslovakia, in the Soviet-controlled Baltic republics and elsewhere, semiology Preface 21 was synonymous with freedom of thought. Rather logically, it was in Paris (where French research was exhibiting great dynamism, whether through the Semiology Section of the Social Anthropology Laboratory of the Collège de France, the journal Communications or the publications of Émile Benveniste, Roland Barthes and Algirdas Julien Greimas, amongst others) that the idea took shape of bringing these international cur- rents together. And it was logical too that, under the inspired authority of Roman Jakobson, Benveniste’s presidency was imposed on all. The International Symposium of Semiotics, created in August 1968, was to furnish the foundations of the International Association for Semiotic Studies (AIS), of which Émile Benveniste officially became President in 1969. As a young Bulgarian student benefitting from a French gov- ernment scholarship, I had the privilege, along with the linguist Josette Rey-Debove, of being put in charge of the scientific sec- retariat of the publication ‘Recherches sémiotiques’ (Semiotic inquiry), first under UNESCO’s Social Science Information unit, then under the AIS. This context, following my passionate reading of the first volume of Problems in General Linguistics, gave me the opportunity of forming an exceptional personal bond with Émile Benveniste. Our meetings took place at his home, in the rue Monticelli, near the Porte d’Orléans. Still today I remember his office as a ‘sacred’ place (so it appeared to the timid girl I then was), in which the great scholar, with his smile of vivid intelligence, seemed to guard the secrets of the imme- morial Indo-European and Iranian worlds. It was a rather dark office, where books carpeted the walls and strewed the floor, old library stock of which the odour, mixed with the steam of tea which, along with dry biscuits that we never touched, for me evoked ancient parchment scrolls. The administrative details quickly dispensed with, the professor enquired about my work.36 With an insatiable curiosity, he was as interested in the linguistic and philosophical debates in Eastern Europe (Bakhtin’s ‘dialogism’) as he was in literary innovations (then in full flower, with the ‘Groupe théorique’ of Philippe Sollers’s journal Tel Quel which met at 44 rue de Rennes). During these meetings Benveniste acted as teacher, protector and attentive listener. I recall asking him whether writing was an ‘infra-’ and 22 Last Lectures: Collège de France 1968 and 1969

‘supra-’linguistic process (as he was describing it in connection with dreams) or instead translinguistic; putting it to him that Raymond Roussel’s writing could be defined as a ‘productivity’ defying the ‘product’; discovering, when he was talking to me about Jakobson’s work, the notion of ‘spotha’ (simultaneously ‘sound’ and ‘sense’ and always ‘activity’, according to the Indian grammarians). I remember the professor advising me to read on this subject Madeleine Biardeau’s recent book on language in classical Brahmanism; how, on another occasion, he expressed regret that Harris and Chomsky had founded a general syntax without taking account of the diversity of lan- guages (‘It is regrettable to know only a single language’, he had written to a renowned linguist). Most often, he replied to my questions with terse and tolerably provocative remarks: ‘You know, I’m only interested in little things. The verb “to be”, for example.’ And advising me to consult, after his Problems in General Linguistics, a recent publication on this immense subject in a recent issue of the very serious journal Foundations in Language . . .37 Or else, as an answer to my interrogations, he opened the Sanskrit text of the Rigveda, to translate appro- priate passages for me directly into French. Then, after some semantic or grammatical remarks, he returned to the contents of the ‘tale’ and to the ‘characters’ of this great collection of hymns from ancient India, always with an allusive tone and a hint of irony (concerning Aragon, for instance): ‘Do you think, Madame, that woman is the future of man?’ On another day, when I had just discovered the term ‘senefi- ance’ in the ‘soul’s voyage toward God’ dear to the mediaeval theoreticians of the ‘modi significandi’, I asked him what he thought about it. ‘You read a lot for your age’, he replied. ‘I think that, closer to us, Jean Paulhan’s father used to use this term. People still read in Bulgaria, and in Eastern Europe in general, don’t they? You know that čitati, the Slavic root for “to read”, goes back to the meaning of “to count”, as well as “to respect”.’ I had not thought about it, obviously; I did not know very much. He never told me that his parents had been teachers in Samokov, Bulgaria. Only that I reminded him of his mother: a distant resemblance, I suppose. Preface 23

Husserl’s phenomenology interested him a great deal, and he seemed astonished that I had some modest bits of knowledge of his Ideen. But we never mentioned Heidegger, whom I had only just discovered. In Warsaw, I had brought along Antonin Artaud’s Letters from Rodez. ‘Would you mind lending it to me?’, he asked. Émile Benveniste hid the little book under the symposium handouts and I saw that, with a shy smile on his lips, he per- mitted himself to read it when a speaker or a debate became tedious. Encouraged by this evidence of freedom, and having recently spotted his name alongside those of Artaud, Aragon, Breton, Éluard, Leiris and a whole constellation of intellectu- als, artists and writers who had signed the Surrealist Manifesto ‘Revolution First and Forever’ (1925), during the break I asked our future President, ‘Monsieur, what a joy to discover your name amongst the signers of a Surrealist manifesto.’ ‘An unfortunate coincidence, Madame.’ The smile had vanished, a cold and empty look nailed me to the floor, and I slumped in shame before the group of confer- ence participants around us. A few hours later and with no witnesses, the Professor whispered in my ear, ‘Of course it was I, but it must not be said. You see, now I am in the Collège de France.’ On our return to Paris, he invited me to take tea, this time at a café, the Closerie des lilas. ‘It was here that we used to meet. A violent time, the war. But here too there was bloodshed, within the group itself.’ Seeing my surprise, he added: ‘No, the metaphor is not too strong. I quickly realised that I did not belong here.’ Today I reread the Manifesto.38 Indeed. Benveniste had fled the calls to insubordination, abandoned the bloody Stalinist- Trotskyite revolt (Breton and Aragon), ignored the maddening experience of the poetic infinite – which, released from the social contract, takes away the order of language (Mallarmé: ‘One single guarantee, syntax’) in a vocal explosion (Artaud’s glossolalia) – in order to consecrate himself in a sort of priest- hood to signifiance in the logics of language. Academic conven- tion provided this nomad tempted by the conflagration, this ‘poor linguist scattered in the universe’, with a protection and 24 Last Lectures: Collège de France 1968 and 1969 a necessary brake. But it did not stop him from reaching out to dissident thinking under communism – the end of which, marked by the fall of the Berlin Wall, he did not live to see; but the Warsaw Symposium appears in hindsight as one of its foreshadowings. Nor did it discourage him from listening for the trace of free and creative subjectivity in the duality of signifi- ance: between the nameless experience of ‘inner language’ and the semantic of discourse which is used to communicate and to put in order. I remember our last conversation, the end of November or start of December 1969. He had received my book Sèméiôtikè and, benevolent as always, was hoping to finish reading it and to speak to me about it in detail before the vacation. But soon, suddenly, came the shock: the news of the stroke, the paralysis, the aphasia. The administration of the Collège de France and his colleagues took care of all the customary formalities. At the hospital I found his sister Carmelia, heroic in her devotion and sensitivity, who stayed by his side day after day, right up to the end, in miserable conditions. She spoke to me in particular of Benveniste’s long-time friend Father Jean de Menasce, whom I did not meet, but whose personal experience of recovering from a similar illness gave her total confidence that her brother would do the same. The situation was deplorable: the patient was hospitalised in a ward, where each day he had to tolerate unhealthy company and the unwelcome visits of other patients’ families, without any rehabilitative care. The general impression was that the patient no longer understood speech. ‘But he never reacted much to family news before his stroke, it bored him’, Carmelia Benveniste reminded people. We managed to get the great aphasia specialist François Lhermitte to come and apply his expertise; he asked Benveniste to draw a house. No reaction. Terrified at the thought of the expertise running out, I made an effort to ask the patient myself. He drew the house. A programme of speech therapy was then put in place. The result was judged unconvincing. His faithful disciple Mohammed Djafar Moïnfar and I soon realised that it was impossible to find a better place in a private establish- ment – the absent-minded scholar not having paid his insurance premiums to the Mutuelle générale de l’Éducation nationale, I Preface 25 was told. We thought of asking all his friends for a contribution, to sort out the insurance retroactively, but various administra- tive difficulties stood in the way. Still today, I reproach myself for not having attended his rehabilitation sessions: the affection in which he held me, perhaps, might have made him more coop- erative. An illusion, no doubt, but one I still think about. It had often seemed to me that his foreign students and friends were the most motivated, the most conscious of his distress and of the magnitude of his work. I was persuaded that he was still present intellectually. So one day I asked him to autograph a copy of his first book, The Persian Religion According to the Chief Greek Texts (1929), which I had found in English translation in an Orientalist anti- quarian bookshop. With trembling handwriting, he inscribed his name in large capital letters, É. BENVENISTE, adding the date 23-9-1971, which he immediately corrected to 24-9- 1971: thus he remained present in the interlocutory act, and retained the notion of time. In 1971, the special issue of the journal Langages on ‘The Epistemology of Linguistics’, which I edited, was dedicated to him: ‘Homage to Émile Benveniste’ – which gave him joy. Pierre Nora, the editor of Gallimard’s Bibliothèque des Sciences Humaines, and I brought him as well the second volume of his Problems in General Linguistics. In 1975, a collection edited by Nicolas Ruwet, Jean-Claude Milner and me, with the title Langue, discours, société: Pour Émile Benveniste (Language, discourse, society: For Émile Benveniste) was dedicated to him, published by the Éditions du Seuil. He received it with pleasure. Of course reading these works was exhausting, and no doubt he appreciated their existence even more than the details they contained. After that, alas, the nine hospital transfers he underwent in seven years, my thesis for the doctorat d’État, then motherhood made my visits rarer and rarer. But he did not forget me, and in November 1975 a letter from Carmelia Benveniste informed me that the professor was asking expressly to see me. He still managed to express his wishes, and remembered those whom he wished to see again. During one of these meetings, at the hospital of Créteil, he asked me to come close to his bedside, sat up, held up his index finger and, very shyly, with the same adolescent smile, began 26 Last Lectures: Collège de France 1968 and 1969 to ‘write’ on the blouse that covered my breast. Surprised, as much bowled over as bothered, I did not dare to move and could not divine what he was hoping to write or draw in this strange gesture. I asked him whether he wanted something to drink, read or listen to. He shook his head no, and began once more to trace on my breast these disturbing and indecipherable signs. I finally handed him a sheet of paper and a ballpoint pen. Then, with the same writing in large capitals that he had used to autograph his book for me, he traced: THEO. I hardly knew at the time – was it 1972 or 1973? – that Benveniste had arrived in France as a student of the Rabbinical School. Nor had he spoken to me about the Shoah. I did not have a global vision of his works in general linguistics, the second volume of his Problems in General Linguistics not yet having been put together, and in any case my insufficient knowledge would not have allowed me to assimilate it. But I was persuaded that his verbal paralysis had not completely destroyed his intelligence. This THEO meant something. Today, reading his last writings in the light of his published work, I do not pretend to offer you an interpretation: THEO will remain for me forever enigmatic. I am merely sketching a reading. The chances of our respective personal histories had put me on his route, so that he might recall to me, before dying, a message that he needed to trace on a body: Whatever ‘the semantic’ may be in our discourse (such as we communicate it through dialogues in our temporal existences), the diversity of our languages and the language itself engender this ‘semiotic capacity’ (borne witness to by the unpronouncea- ble graphism /YHWH/, but which the professor had undertaken to analyse with the tools of Greek onto-theology /THEO/ and thanks to its scientific continuations) in the meeting of the ‘inner languages’ of our subjectivities. This ‘original force of the work’ (Seventh Lecture) ‘tran- scends’ (/THEO/) every other property of language, and ‘one does not conceive’ that ‘its principle is found elsewhere than in a language’. ‘I’, every speaking person, consist in this duality, stand at this crossroads. ‘I’, every person, experience this ‘SIGNIFIANCE’ which grasps and interprets history. Preface 27

I shall be grateful to readers of these Last Lectures for adding their own path to this crossroads, to this writing.

Notes 1 His mother, Maria Benveniste (born in Vilnius, now in Lithuania), taught Hebrew, French and Russian at the school of the Universal Israelite Alliance in Samokov, Bulgaria; his father, Mathatias Benveniste (born in Smyrna), spoke Ladino; the linguistic envi- ronment of his early childhood included Turkish, Arabic, Modern Greek, probably Slavic. Many great linguists of the late nine- teenth and early twentieth centuries, of Jewish origin, came to the study of languages through the multilingualism of their family milieu (the brothers James and Arsène Darmesteter, Michel Bréal, Sylvain Lévi). 2 A ‘Talmud Torah’ school intended to give students a grounding in Jewish culture, lead them to the baccalaureate and allow them to prepare for the rabbinate. The students were taught Latin, Greek, Hebrew, German and, with very particular care, French. 3 See Françoise Bader, ‘Sylvain Lévi’, Anamnèse, no. 5: Trois lin- guistes (trop) oubliés (Paris: L’Harmattan, 2010), pp. 141–70. 4 The signers urged the UGIF ‘to maintain as close a union as pos- sible between our French brethren and us [. . .] to attempt nothing [. . .] that might isolate us morally from the national community to which, even after being slapped with this law, we remain faith- ful’. See Marc Bloch, L’étrange défaite (The strange defeat, Paris: Gallimard, 1946), pp. 314–19. 5 The most concrete example of this is his Vocabulaire des insti- tutions indo-européennes, 2 vols (Paris: Minuit, 1969; English version, Dictionary of Indo-European Concepts and Society, trans. Elizabeth Palmer, Chicago: University of Chicago Press, 2016; trans. first published as Indo-European Language and Society, Coral Gables, FL: University of Miami Press, 1973). 6 Echoing Rilke, this condensed and allusive confession expresses the young linguist’s longing for the mother he left behind at the age of eleven, and did not see again before her death when he was seventeen. Sensitive to the ‘latent virile violence’ which attracts him beneath the ‘superficially feminine’ appearances of a mater- nal that is vigorous and ‘robust as a man’, Benveniste composed 28 Last Lectures: Collège de France 1968 and 1969

his self-portrait in the guise of the poets (bachelors?), from Homer (the ‘Old man of the sea’) to Lautréamont (‘Old Ocean, o great bachelor!’). See Philosophies, no. 1, 15 mars 1924, year of the publication of the First Surrealist Manifesto. 7 The Rabbinical School in the rue Vauquelin trained rabbis in Europe for communities in the Middle East and Africa, ‘as teachers were trained for schools’. In a letter of October 1918, his mother wrote that her son Ezra’s ‘situation in the school’ had ‘become untenable’: he was drawn to languages and would do his studies in the humanities (see Françoise Bader, ‘Une anamnèse littéraire d’É. Benveniste’, Incontri Linguistici, no. 22, Rome, 1999, p. 20). 8 Cited by Benveniste in ‘La forme et le sens dans le langage’ (Form and meaning in language), in Le langage II: Sociétés de Philosophie de langue française, Actes du XIIIe Congrès, Genève, 1966 (Neuchâtel: La Baconnière, 1967), pp. 29–40, repr. in PLG 2, pp. 215–38, p. 229. 9 ‘La forme et le sens’, PLG 2, p. 216. 10 ‘Remarques sur la fonction du langage dans la découverte freud- ienne’ (Remarks on the function of language in Freud’s discov- ery), La psychanalyse 1 (1956), 3–16, repr. in PLG 1, pp. 75–87, p. 78. 11 [Tr.: Kristeva emphasises Benveniste’s innovative use of a mas- culine definite article with these normally feminine nouns.La sémiotique is ‘semiotics’; le sémiotique is ‘the semiotic’, i.e. what is semiotic in nature. La sémantique is ‘semantics’; le sémantique is ‘the semantic’.] 12 ‘La forme et le sens’, PLG 2, p. 229. 13 Published as ‘La forme et le sens’. 14 ‘Sémiologie de la langue’ (Semiology of the language), Semiotica 1 (1969), 1–12, 127–35, repr. in PLG 2, pp. 43–66. 15 It would be Antoine Culioli who brought this project to fruition in his ‘theory of enunciative operations’, by studying the activity of language across the diversity of national languages. 16 ‘Sémiologie de la langue’, PLG 2, p. 66. 17 ‘La forme et le sens’, PLG 2, p. 238. 18 These manuscripts in the Benveniste Archive of the Bibliothèque Nationale de France (BnF PAP OR DON 0429, env. 6 à 22) have been published as Émile Benveniste, Baudelaire, ed. Chloé Laplantine (Limoges: Lambert-Lucas, 2011). Preface 29

19 However, Benveniste borrows from the American philosopher the term ‘interpretant’, specifying that he only uses this ‘isolated denomination’ and, above all, in a ‘different’ sense (Fifth Lecture), presumably phenomenological. Peirce’s ‘thirdness’ could however have shored up the structure of the subject of the enunciation (an ‘Oedipal’ structure for Freud) in the semiotic according to Benveniste. 20 With Roland Barthes, Writing Degree Zero (1953), Elements of Semiology (1965); Jacques Derrida, Of Grammatology (1967), Voice and Phenomenon (1967), and in the literary domain, after the nouveau roman, with Philippe Sollers, Drama (1965), Logiques (Logics, 1968), Nombres (Numbers, 1968), Writing and the Experience of Limits (1971). 21 ‘Remarques sur la fonction’. 22 See BnF PAP OR DON 0429, env. 22, f. 8. 23 See Julia Kristeva, ‘L’engendrement de la formule’ (Engenderment of the formula), in Sèméiôtikè: Recherches pour une sémanalyse (Sèméiôtikè: Research for a semanalysis, Paris: Seuil, 1969), p. 290: ‘The numerical function of the signifier’. 24 ‘La forme et le sens’, PLG 2, p. 238. 25 Benveniste, ‘Le langage et l’expérience humaine’ (Language and human experience), Diogène 51 (1965), 3–13, repr. in PLG 2, pp. 67–78, and ‘L’appareil formel de l’énonciation’ (The formal apparatus of enunciation), Langages 5/17 (1970), 12–18, repr. in PLG 2, pp. 79–88. 26 [Tr.: référent ‘referent’ (commonly used to translate Gottlob Frege’s Bedeutung) is not a term used by Saussure, but came into later linguistics and semiotics through C. K. Ogden and I. A. Richards’s The Meaning of Meaning (1923), which introduced it in the course of criticising Saussure for disconnecting the linguis- tic sign from things in the world.] 27 Roman Jakobson and Claude Lévi-Strauss, ‘“Les chats” de Charles Baudelaire’, L’Homme 2/1 (1962), 5–21. 28 BnF PAP OR DON 0429, env. 20, f. 204. 29 Ibid., env. 12, f. 204. 30 Ibid., env. 23, f. 358. 31 ‘Remarques sur la fonction’, PLG 1, p. 78. 32 Ibid., p. 86. 33 ‘Sémiologie de la langue’, PLG 2, p. 66. 30 Last Lectures: Collège de France 1968 and 1969

34 ‘La forme et le sens’, PLG 2, p. 226. 35 Letter from Benveniste to Georges Redard, 17 Oct. 1954: ‘Linguistics is universal, but the poor linguist is scattered in the universe’, cited by Redard, below, p. 129. 36 I was finishing my doctoral thesis (3e cycle), which I defended in June 1968, exceptionally, in my capacity as a foreign student: and I was starting my research on the poetic language of Mallarmé and Lautréamont with a view to a thesis for the doctorat d’État. 37 See Charles H. Kahn, ‘The Greek verb “to be” and the concept of being’, Foundations in Language 2/3 (Aug. 1966), 245–65. 38 ‘We consider bloody Revolution to be the ineluctable vengeance of the humiliated spirit. We [. . .] conceive of it only in its social form [. . .] The idea of revolution is the best and most effective safeguard of the individual.’