The Life of Texts: Evidence in textual production, transmission and reception 9781350039056, 9781350039087, 9781350039063

Citation preview

The Life of Texts

i

Also published by Bloomsbury Adonis, Carlo Caruso Interpreting Classical Texts, Malcolm Heath From Byzantium to Italy, N. G. Wilson

ii

The Life of Texts Evidence in textual production, transmission and reception Edited by Carlo Caruso

iii

BLOOMSBURY ACADEMIC Bloomsbury Publishing Plc 50 Bedford Square, London, WC 1B 3DP, UK 1385 Broadway, New York, NY 10018, USA BLOOMSBURY, BLOOMSBURY ACADEMIC and the Diana logo are trademarks of Bloomsbury Publishing Plc First published in Great Britain 2019 with a contribution from Dipartimento di Filologia e Critica delle letterature antiche e moderne, University of Siena Copyright © Carlo Caruso and contributors, 2019 Carlo Caruso has asserted his right under the Copyright, Designs and Patents Act, 1988, to be identified as Editor of this work. For legal purposes the acknowledgements in the Foreword on pages xii and xiii constitute an extension of this copyright page. Cover image © Tetra Images/Getty All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage or retrieval system, without prior permission in writing from the publishers. Bloomsbury Publishing Plc does not have any control over, or responsibility for, any third-party websites referred to or in this book. All internet addresses given in this book were correct at the time of going to press. The author and publisher regret any inconvenience caused if addresses have changed or sites have ceased to exist, but can accept no responsibility for any such changes. A catalogue record for this book is available from the British Library. A catalogue record for this book is available from the Library of Congress. ISBN : HB : 978-1-3500-3905-6 ePDF : 978-1-3500-3906-3 eBook: 978-1-3500-3907-0 Typeset by RefineCatch Limited, Bungay, Suffolk To find out more about our authors and books visit www.bloomsbury.com and sign up for our newsletters.

iv

Contents List of illustrations Notes on contributors Foreword

1 2 3 4 5 6 7 8

9

Introduction: Conceiving the Life of Texts Richard Gameson Editing Homer Barbara Graziosi The Canon and the Codex: On the Material Form of the Christian Bible Francis Watson Wandering Nights: Shahrazād’s Mutations Daniel L. Newman A Text in Exile: Dante’s Divine Comedy Annalisa Cipollone Textual Metamorphosis: The Manuscripts of Leonardo da Vinci Carlo Vecce Montaigne: The Life and After-Life of an Unfinished Text John O’Brien Rescuing Shakespeare: King Lear in Its Textual Contexts David Fuller Textual Evidence and Musical Analysis: Once More on the First Movement of Beethoven’s Tempest Sonata, Op. 31, No. 2 Julian Horton Fragments Shored against Ruin: Reassembling The Waste Land Jason Harding

Notes Index of subjects and themes Index of manuscripts and annotated printed copies Index of names

vi ix xii 1 28 48 62 93 115 133 154

176 191 204 239 241 244

v

Illustrations 0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

vi

Laurence Binyon, ‘For the Fallen’. Oxford, Trinity College. Reproduced courtesy of the President and Fellows of Trinity College, Oxford. New Testament (London: Robert Barker, 1640) and Metrical Psalms (London: ‘MF ’, 1641), together in a contemporary embroidered binding. Durham University Library, SB 2537. Reproduced courtesy of Durham University Library. Die heimlich offenbarung iohannis (Albrecht Dürer’s Apocalypse) (Nuremberg: Anton Koberger, 1498), chapter 22. Author’s collection. René Magritte, ‘Les mots et les images’, La Révolution surréaliste (December, 1929). Author’s collection. Courtesy of the Estate of René Magritte. Geoffrey Chaucer, The Canterbury Tales (Westminster: William Caxton, 1483), General Prologue, Wife of Bath and Parson. Reproduced after the facsimile: Geoffrey Chaucer, The Canterbury Tales, with a note by J. A. W. Bennett (London, 1972). Mr William Shakespeares Comedies, Histories and Tragedies published according to the True Originall Copies (London: Isaac Iaggad and Ed. Blount, 1623), pp. 130–1. Durham University Library, Cosin W.2.11. Reproduced courtesy of Durham University Library. Bible with ‘standard’ chapter divisions; Paris, third quarter of the thirteenth century. Durham Cathedral Library, A.II .3, fol. 313r. Reproduced courtesy of the Dean and Chapter of Durham Cathedral. Glossed Psalter, Northern France or England; middle of the twelfth century. Durham Cathedral Library, A.III .9, fol. 51v. Reproduced courtesy of the Dean and Chapter of Durham Cathedral.

2

9

11

12

15

16

17

18

Illustrations

0.9

0.10

1.1 1.2

3.1 3.2

3.3 3.4 4.1 4.2

6.1

6.2

6.3 6.4

Joshua Reynolds, Works . . ., 2nd ed.; 3 vols. (London, 1789), vol. I, title page; annotated by William Blake. (London, British Library, C.45.e.1820.) Reproduced after M. Phillips, William Blake, Apprentice & Master (Oxford, 2014), no. 52. Writing exercise on papyrus (Virgil, Aeneid, Book XI , lines 371–2), second half of the first century ad. (Oxford, Sackler Library, P.Oxy. L. 3554.) Reproduced courtesy of the Egypt Exploration Society and the Oxyrhynchus Papyri Project. Papyrus fragment, second century bc , University of California 2390, P.Tebt. 4. The beginning of Iliad 1, surrounded by scholia in the so-called ‘Venetus A’ manuscript (Codex Marcianus Graecus 822), c. tenth century ad. Getty Images. A. Galland’s journal, Paris, Bibliothèque nationale de France, MS fr. 15277, p. 54. 1001 Nights manuscript (fourteenth century) used by Galland, Paris, Bibliothèque nationale de France, MS arabe 3610, fol. 4v. Beginning of the 70th Night. Earliest 1001 Nights manuscript (ninth century), Chicago University Oriental Institute, MS E17618. Page from 1001 Nights manuscript (sixteenth century), Manchester, John Rylands Library, Arabic MS 706, fol. 206r. Domenico di Michelino, Portrait of Dante (1465). Florence, Santa Maria del Fiore. Wikimedia/Jastrow. Stemma codicum from Dantis Alagherii Comedia, ed. F. Sanguineti. Redrawn from Sismel-Edizioni del Galluzzo: Florence 2001, lxv. Montaigne’s Essais, ‘Of Cato the Younger’ (‘Du jeune Caton’). Reproduced by permission of Montaigne Studies, University of Chicago. Montaigne’s Essais, ‘Of Cato the Younger’ (‘Du jeune Caton’). Reproduced by permission of Montaigne Studies, University of Chicago. Montaigne’s Essais, EB , fol. 47r. Reproduced by permission of Montaigne Studies, University of Chicago. Montaigne’s Essais, ‘Of Presumption’ (‘De la Præsumption’), p. 439. Reproduced by permission of the Librarian,

vii

22

25 32

35 64

67 73 74 95

110

134

135 141

viii

6.5

E8.1 E8.2 E8.3 E8.4

Illustrations

Maynooth University, from the collections of St Patrick’s College, Maynooth. Montaigne’s Essais, ‘Of Presumption’ (‘De la Præsumption’), EB , fol. 292v. Reproduced by permission of Montaigne Studies, University of Chicago. Opening to Beethoven’s Op. 10 No. 1. Opening to Beethoven’s Tempest Sonata. Hepokoski’s analysis of Beethoven’s Tempest Sonata. The first movement of Haydn’s Symphony No. 44 of 1772.

142

146 180 181 188 189

Contributors Annalisa Cipollone teaches Italian Literature at Durham University. She has published extensively on Dante, Petrarch, and more generally on early Italian poetry and its textual transmission, as well as on twentieth-century literature. She has edited Petrarca e Boccaccio: modelli letterari fra Medioevo e Umanesimo (with C. Caruso, 2006), and Il viaggio e le arti: il contesto italiano (with L. Bertolini, 2009). Among her most recent publications are Hell, Heaven and Hope. A journey through life and the afterlife with Dante (Durham: Institute of Advanced Study, 2017) and ‘Parole tra parentesi’, in La filologia in Italia nel Rinascimento (Rome: Edizioni di Storia e Letteratura, 2018), 37–59. David Fuller is Emeritus Professor of English in the University of Durham. He is the author of Blake’s Heroic Argument (1988), James Joyce’s ‘Ulysses’ (1992), Signs of Grace (with David Brown, 1995) and The Life in the Sonnets (2011) in the Bloomsbury series Shakespeare Now!, editor of Tamburlaine the Great for the Clarendon Press complete Marlowe (1998), and William Blake: Selected Poetry and Prose (Longman Annotated Texts, 2000), and co-editor (with Patricia Waugh) of The Arts and Sciences of Criticism (1999), and (with Corinne Saunders) of the medieval poem Pearl, modernized by Victor Watts (2005). His Shakespeare and the Romantics (Oxford Shakespeare Topics) is forthcoming from OUP in 2019. Richard Gameson is Professor of the History of the Book at Durham University. He has published over 100 studies of manuscripts, book illumination and early libraries. His most recent books are The Cambridge History of the Book in Britain I (Cambridge, 2013), The Lindisfarne Gospels: New Perspectives (2017) and The Medieval Manuscripts of Trinity College, Oxford: A Descriptive Catalogue (2018). Barbara Graziosi is Professor of Classics at Princeton. She previously worked as Professor of Classics and Head of Department at Durham University: her contribution to this volume stems from her long-standing work on Homer, viewed here from the perspective of interdisciplinary conversations with colleagues working on other textual traditions in Durham. Her most recent books are The Gods of Olympus: A History (Profile Books, 2013), Homer (Oxford ix

x

Contributors

University Press, 2016) and, edited with Nora Goldschmidt, Tombs of the Ancient Poets: Between Literary Reception and Material Culture (Oxford University Press, 2018). The latter two publications stem from a research project she directed under the aegis of the European Research Council: Living Poets: A New Approach to Ancient Poetry. Jason Harding is Professor of English Studies at Durham University, UK . He is the author or editor of six books on T. S. Eliot, including The New Cambridge Companion to T. S. Eliot (2017), a volume in The Complete Prose of T. S. Eliot (2015) and a study of Eliot’s literary-intellectual review, The Criterion. Julian Horton is Professor of Music at Durham University. He was educated at Trinity College, Cambridge, and has taught at University College Dublin and King’s College, London. He is author of Bruckner’s Symphonies: Analysis, Reception and Cultural Politics (Cambridge, 2004) and Brahms’ Piano Concerto No.  2, Op. 83: Analytical and Contextual Studies (Leuven, 2017), editor of The Cambridge Companion to the Symphony (Cambridge, 2013) and co-editor, with Lorraine Byrne Bodley, of Schubert’s Late Music: History, Theory, Style (Cambridge, 2016) and Rethinking Schubert (Oxford, 2016). He is currently serving his second term as President of the Society for Music Analysis, and is a Council Member of the Royal Musical Association. Daniel L. Newman holds the Chair of Arabic at the University of Durham. His research areas include the nineteenth-century Arab reform movements, the history of translation in the Arab world, and medieval Arabic-Islamic medicine and food culture. He is the author of An Imam in Paris (20112) and The Sultan’s Sex Potions: Arab Aphrodisiacs in the Middle Ages (2014). His next book, Rifa‘a al-Tahtawi: A Nineteenth-Century Egyptian Educationalist and Reformer, will be published by Edinburgh University Press in 2019. John O’Brien is Emeritus Professor of French at the University of Durham. His research focuses principally on Montaigne, on whom he is currently preparing a monograph, ‘Nostre bastiment, et public et privé’: Decline, Disorder and Civil War in Montaigne’s ‘Essais’. His co-authored book, La première circulation de la ‘Servitude volontaire’ en France et au-delà, on a work by Montaigne’s friend La Boétie, will be published in late 2018 by Champion (Paris). Carlo Vecce is Professor of Italian Literature at the Università degli Studi di Napoli ‘L’Orientale’. He has widely published on Italian Humanism and Renaissance, including the standard edition of Iacopo Sannazaro’s Arcadia

Contributors

xi

(2013). For the Commissione Vinciana, of which he is a member, he has edited Leonardo da Vinci’s Libro di Pittura (with C. Pedretti, 1995) and Codex Arundel (1998). His biography of Leonardo (20062) has been translated into numerous languages. He is the author of La biblioteca perduta. I libri di Leonardo (2017). Francis Watson holds a Research Chair in Early Christian Literature at Durham University, having previously taught in Aberdeen (1999–2007) and King’s College London (1984–99). His work focuses primarily on the early reception of Jewish and Christian scriptural texts and on canonical and non-canonical gospel literature. His most recent books are Gospel Writing: A Canonical Perspective (Eerdmans, 2013), Paul and the Hermeneutics of Faith (T&T Clark/Bloomsbury 20152), The Fourfold Gospel (Baker Academic, 2016) and Connecting Gospels: Beyond the Canonical/Non-canonical Divide (Oxford, 2018, co-edited with Sarah Parkhouse). He holds a position as Visiting Professorial Fellow at Australian Catholic University, Melbourne (ACU ), and is director of an international research programme entitled ‘Texts, Traditions, and Early Christian Identities’, based in Melbourne, Durham and Leuven.

Foreword In the winter of 2014, the Institute of Advanced Study at Durham University proposed ‘Evidence’ as the general theme of its annual activities. A group of colleagues from the Faculty of Arts and Humanities concurred that a common reflection on the nature of textual evidence might be considered a stimulating topic, likely to stir the curiosity not only of scholars but of the wider public as well. Once the general title ‘The life of texts. Evidence in textual production, transmission and reception’ had been agreed, a series of public lectures was held at Palace Green Library in the Michaelmas and Epiphany terms of 2015–16. Speakers from subjects as diverse as Arabic, Classics, English, French, History of the Book, Italian, Music and Theology brought their audiences to reflect on a number of compelling case studies, which together encompassed nearly 3,000 years of literary and intellectual history. Textual criticism, ordinarily regarded by many as the epitome of dry-as-dust, turned out to be as alive and thoughtprovoking a discipline as ever, all thanks to the speakers’ consummate scholarship and power of engagement. ‘The life of texts’ had proved an appropriate title for the series. Each lecture was introduced by a chair moderating the lively exchange of questions and answers that concluded each session. Chairs were often persuaded to join speakers for dinner so as to continue the conversation inter pocula laeti, for things thought and said ‘merrily over a glass’ are in no way less interesting, or valuable, than those pronounced in more formal circumstances. Elizabeth Archibald, Patrick Gray, Richard Maber, Michael O’Neill, Alec Ryrie and Marc Schachter are to be thanked for discharging their role as chairs in the most gracious and effective way possible. On one special occasion, following an invitation extended by the Principal of St Cuthbert’s Society Professor Elizabeth Archibald, speakers and chairs were hosted in the college’s SCR for a memorable dinner. Such a congenial atmosphere naturally led to the idea of turning the lecture series into a book. The efficiency and forbearance of the staff at Bloomsbury have been essential for the smooth running of the production process. Likewise, the authors’ commitment has made the editor’s task as painless and straightforward as any editor would wish it to be. xii

Foreword

xiii

Neither the lecture series nor the volume would have materialized without the unstinting support of the Director of Durham’s Institute of Advanced Study, Professor Veronica Strang, and her staff. The Leverhulme Trust generously agreed to fund part of the lecture series, which allowed its inclusion in the activities of a Leverhulme Major Research Fellowship (2013–16). Palace Green Library, the Institute of Medieval and Early Modern Studies, the Durham Italian Seminar and several departments in the Faculty of Arts and Humanities of Durham University backed the project since its earliest stages; further support was subsequently received from the Dipartimento di Filologia e Critica of the Università di Siena. To all of them go the editor’s heartfelt thanks. C. C. Siena, 15 June 2018

xiv

Introduction: Conceiving the Life of Texts Richard Gameson

The best known and most frequently declaimed stanza of all English verse illustrates some of the ways in which texts can approximate a life of their own. First published in The Times in September 1914, widely read in public thereafter, set to music by Edward Elgar, and still broadcast on national media at least once a year, Laurence Binyon’s For the Fallen lines 13–16 have enjoyed unrivalled public exposure in the Anglophone world. They shall grow not old as we that are left grow old: Age shall not weary them, nor the years condemn. At the going down of the sun and in the morning We will remember them.

A first point to note is that these lines are disproportionately famous in relation to the rest of the poem, short though it is: they can probably be quoted by most educated adults in Britain and the Commonwealth who would be unlikely to remember – and might possibly not even recognize – the other twenty-four lines. Second, there is the circumstance that many people quoting from memory typically misrecite the first of the famous lines, substituting the more idiomatic ‘They shall not grow old’ for the poetic and archaic ‘They shall grow not old’ of the original (a nice illustration, incidentally, of the text-critical axiom, lectio difficilior stet – ‘the reading that is more difficult should stand’). Third, there is the interesting fact that the second of the famous lines was initially slightly different. Binyon’s autograph manuscript (Fig. 0.1) shows that he originally wrote ‘Age shall not wither them’ (as opposed to ‘weary them’). That this echoes a line from Shakespeare (Antony and Cleopatra, Act 2, scene 2: ‘Age cannot wither her, nor custom stale her infinite variety’) is unlikely to be coincidence; nor is it surprising that a word should then have been changed to reduce the debt – for the modern mind takes an altogether dimmer view of literary borrowings than did the classical and medieval one. This one short extract thus reveals one text 1

2

Life of Texts

influencing another, the propensity of parts of a text to gain an existence independent of the rest, the ease with which wording is changed in popular recollection, and how literary appropriations are viewed in contrasting ways in different contexts.

Figure 0.1 Laurence Binyon, ‘For the Fallen’. Oxford, Trinity College.

Introduction

3

The rapidity with which Binyon’s lines achieved widespread fame makes a further point: that the diffusion of a text is inextricably linked to the historical circumstances in which it was produced. A major author in classical antiquity might legitimately wonder whether his literary masterpiece would achieve even modest currency; a completely unknown individual today may make an online posting that can gain worldwide celebrity within hours. This stark contrast highlights the extent to which available technologies affect the form, function, circulation and even survival of texts, irrespective of their literary merit and of the weightiness of their content. Moreover, the works of a single author at one particular moment in time can enjoy radically different circulation according to their nature and the form in which they are issued: whereas ‘For the Fallen’ rapidly reached millions, the original edition of Binyon’s Dream-Come-True (1905) comprised a mere 175 copies issued to subscribers by a small private press, while the two enormous folio volumes that he contributed to The Catalogue of the George Eumorfopoulos Collection were even rarer, being specialist studies in a lavish format destined only for wealthy collectors and select institutions; never reissued in any form, they remain Binyon’s most magnificent yet least accessible work. Other than at authorial recitals, audiences generally experience texts through intermediaries of one form or another; and the interfaces between authors and intermediaries on the one hand, and then between intermediaries and audiences on the other, necessarily affect the nature of texts themselves, how they are transmitted and the ways in which they are received. The aim of my comments here is simply to raise awareness of such issues and how they evolved across time, focusing on the practical rather than the theoretical, in order to provide the broadest of contexts for the detailed case studies that follow in the present volume. * To start with an early example, Homer’s Odyssey, standing on the threshold between oral and literary cultures, raises the issue of whether such an epic could have been composed (in a form akin to that in which we now have it) without the tool of writing. Suggestively, the presumed period of its genesis (the eighth century bc ), while hundreds of years after the events to which it relates, is around the time the Greek alphabet became available. Be that as it may, it is then another couple of centuries before there is reasonable evidence for a book culture of sorts in Greece, and the same again before the oldest surviving literary papyri (fourth century bc ). How stable is a text likely to have been in such circumstances? With

4

Life of Texts

the Anglo-Saxon epic Beowulf, known from a single copy transcribed c. ad 1000, the issues of stability in transmission (or otherwise) are even thornier. Did this tale of Scandinavian heroes and monsters originate as a pagan saga that was transmitted orally, only being written down – with a Christian veneer – after the coming of Christianity? Was it instead a nostalgic yet monitory evocation of a heroic past composed much as we have it in a literate context within Christian England? Was it something in between these two extremes, or might it even have been, sequentially, all of these? If comparable uncertainties surround the transmission of these two great works that are now regarded as emblematic of their respective cultures, the demonstrable impact of each in its own context is strikingly different: the Odyssey resonates through Greek and Latin civilization; one now-imperfect copy of Beowulf plus faint verbal echoes of it in other Old English verse are virtually all that attest to knowledge of that poem in early England. Equally, reflecting the contrasting fortunes of the works in nineteenthcentury England, it was ‘odyssey’ not ‘beowulf ’ that was adopted as a noun in modern English. Antique authors, well into the era of physical books (in the form of rolls/ scrolls) were readily aware of how a work could slip out of their control and into circulation at a premature stage – a preliminary draft sent to a friend for criticism, for example, might itself be copied and passed on. Symmachus (d. ad 402), who acknowledged that ‘once a poem has left your hands, you resign all your rights; a speech when published is a free entity’, was himself upbraided by Ausonius (d. c. ad 395) for not keeping to himself a supposedly private copy of the latter’s works. The Academica of Cicero (d. 43 bc ) provides an example of an antique text that was transmitted to posterity – albeit imperfectly – in two quite distinct authorial versions. Numerous medieval writings likewise went through various processes of revision – painstakingly reconstructed by modern scholarship – and could be simultaneously available in variant forms. The several recensions of the letter-collections of St Anselm of Canterbury (d. 1109) and the different versions of the Pseudo-Bonaventura, Stimulus amoris, are cases in point. The phenomenon continues in our own age in relation to works that are reissued more or less revised in different editions (with or without Prefaces that summarize the key changes). The three German editions of Karl Marx’s Das Kapital (volume I) provide one type of example, the expurgated and the first authorial versions of Thomas Hardy’s Tess of the D’Urbervilles (both 1891) another. Now as hitherto, the most recent edition of a work, whatever its merits, is unlikely to be used by a reader who has an earlier one more readily to hand. Current technology has also returned writers to the situation lamented by

Introduction

5

classical authors, namely of a text slipping out of their control before it and they were ready. The ease with which electronic files are shared means not only that many more readers can effortlessly be sent a given draft, but also that they can potentially circulate it themselves, increasing the chances that the wrong version might eventually be ‘canonized’ in one way or another. Authorial manuscripts of all periods naturally command fascination for the insights they offer into the creative process of composition as words and phrases, sometimes images too, were jotted (or sketched), then visibly changed and reordered. The working notes of artists such as Leonardo da Vinci (d. 1519) and Albrecht Dürer (d. 1528), with their marriage of text, diagram and image, and their juxtaposition of theory and practice, can be particularly engaging. The circumstance that Leonardo never reorganized the abundant material in question into the thematic collections or treatises that were apparently envisaged, and that Dürer, abandoning his proposed general treatise on painting in favour of the more focused Underweysung der Messung (1525) and Vier Bücher von Menschlicher Proportion (1528), did so only partially adds to the fascination: for what might have been has so much more potential than what actually is. To what extent subsequent ‘official’ publication can do justice to such material and the form that this can best take will always be matters of debate. Indeed, unfinished texts of any sort pose problems, and the dilemmas of a literary executor – whether to publish nachlass, in what form, how far to revise and to what extent to declare these interventions – are eternal. Should the supposed death-bed wish of Virgil (d. 19 bc ) to burn The Aeneid or of Franz Kafka (d. 1924) to dispose of all his unpublished work (including The Trial and The Castle) have been respected? To the question of whether J. R. R. Tolkien (d. 1973) would have favoured the publication of all the unfinished tales and drafts that have appeared in print since his death, there is a clear answer since his son acknowledged that ‘he himself, peculiarly critical and exacting of his own work, would not have dreamt of allowing even the more completed narratives . . . to appear without much further refinement’. Correspondingly, the decisions and processes behind the texts that result and the extent to which they reflect the intentions of the author can be opaque: what might Lucius Varius Rufus (d. 15 bc ) and Plotius Tucca have contributed to The Aeneid as we have it (assuming that they did indeed revise it for publication)? One thing Varius was specifically said to have done was to invert the order of a couple of the Books. This is germane to another case, The Canterbury Tales, for the two oldest manuscripts of that work, although written by one and the same scribe and within about a decade of Chaucer’s death (1400), present the individual Tales in

6

Life of Texts

different orders. Accounting for this and deciding which ordering might better reflect the poet’s intentions are intractable, if engaging, problems. Returning to authorial manuscripts (latterly typescripts), nothing is seemingly more revealing of the life of a text and the intellect of its creator than witnessing first hand how the work was formed, then revised. The drafts of The Waste Land by T. S. Eliot (d. 1965), replete with interventions by Ezra Pound (d. 1972) and Vivienne Eliot (d. 1947), are a celebrated example. Yet for all that such ‘textarchaeological sites’ reveal, there are further processes – and hence additional compositional fluidity – that they conceal: the initial plan jotted elsewhere; the alternative phrases tried out mentally or orally before only one of them was committed to papyrus, parchment, paper or processor; the unrecorded oral interactions with other minds (discussing ideas and even sentences with spouse, friend, colleague or critic). None of these or similar phases of composition and revision are likely to be apparent from even the earliest surviving draft in one authorial hand. And if letters, reminiscences and other memorabilia – such as those of Eliot relating to The Waste Land or ‘Stories I have Tried to Write’ by M. R. James (d. 1936) – may shed sidelights on the processes, these same sources simultaneously alert us to the additional layers of experience and thought that are irrevocably lost. The hidden processes are especially tantalizing in cases like the Book of Margery Kempe (d. c. 1440), where the physical and spiritual experiences of the illiterate authoress were crystallized into words through the medium of first one, then a second amanuensis, the former supposedly hindered by his limited command of language and poor handwriting, the latter by deficient eyesight. As ever more composition is done electronically, the opportunities for studying drafts become rarer – unless computer scientists track back through the teeming footprints of successive electronic versions to reveal every erratic step from first letter to final version. And here there is a danger of too much detail reducing creativity to a mechanistic listing. Moreover, the way in which a text is corrected by hand – tentatively or vigorously, with one pen or with several implements in distinguishable phases – can itself be revealing of some of the processes in a way that an undifferentiated inventory of every major and minor change enacted on a keyboard cannot be. The most important general message of the word-processor, however, is its advertisement of how fluid for how long is not only the text per se, but also the way in which it is presented. For its layout on the page, the font, even the colour are all subject to transformation at the press of a button. Though easily forgotten, this has always been the case, albeit formerly via more time-consuming procedures.

Introduction

7

How a text is perceived and received – not to mention its chances of survival – are affected by the form in which it is ‘published’. Pliny the Elder (d. ad 79) noted that papyrus could be expected to last for a couple of centuries (see Fig. 0.10). Thus if copies of classical texts were not renewed every 200 years or so, the works in question would eventually vanish. And if no copy of a particular text were lodged in a repository where it was likely to be re-transcribed when necessary – in effect, a public library as opposed to a private household – its fate was even more uncertain. This is something to which we shall return. Furthermore, a papyrus roll had a limited capacity, typically containing one ‘book’ (subdivision) of an historical prose work, up to two or three of a verse epic. If this permitted the use of part of a text unencumbered by the remainder – encouraging the perception of each subdivision as a self-contained work in a series – it equally meant that the survival of one section of an author’s oeuvre, or even of an individual text, in no way guaranteed the preservation of the rest. Hence of the 142 books of Histories (Ab urbe condita libri) written by Titus Livy (d. ad 17), only 1–10 and 21–45 plus a few fragments remain. Correspondingly, the form of the ancient book could influence literary composition – encouraging authors to divide long works into self-contained units (‘Books’) that could be fitted individually or in small groups onto an average scroll, and discouraging the publication of short works that were less than a ‘scroll’s worth’. The poet Martial (d. between ad 101 and 104) commented that padding was sometimes needed to achieve this. (One might compare the situation of works composed in the nineteenth century for serialization in literary magazines or as independent monthly fascicules, that therefore had to be delivered in blocks of a prescribed size, each of which, moreover, would ideally end in such a way as to motivate the reader to buy the next instalment.) When the Hebrew scriptures were translated into Greek, the written form of which tends to take up more space, certain books that in their original language could fit onto a single scroll, now had to be divided into two – whence I–II Samuel, I–II Kings and I–II Chronicles. A broader point here is that, when even a single major composition had to be split across a series of rolls, the concept of a ‘collected edition’ was virtually meaningless: it was physically impossible to achieve. The rise to predominance of the codex (the form of the book still familiar today), a phenomenon of the fourth century ad, redefined what a text could be, as its vehicle became potentially far more capacious. The concept of a canon of scripture, for example, was now more meaningful, because all the relevant texts could actually be ‘fixed’ between a single set of covers. The fourth-century Codex Sinaiticus exemplifies a once-complete Greek Bible, the early eighth-century

8

Life of Texts

Codex Amiatinus, a still-complete Latin one. The replacement of papyrus by parchment as the principal support for writing gave books a hitherto unimaginable longevity. Indeed, one example that dates back to the fourth century ad has just been mentioned. The parchment codex, time-consuming to make but then extremely durable, was ideally suited to the outlook and context of the early medieval ecclesiastical scribe, labouring meticulously for the values of eternity. It is one of the many ironies of cultural history, however, that it was on the pen of this same scribe, with his cloistered Christian values, that the survival of classical literature was also to depend: for, with the disintegration of the infrastructure that had supported the public libraries of the ancient world, if elderly and decaying papyrus books were not recopied onto parchment during the first millennium ad, the works in question slowly but surely turned to dust. That scribes could make errors which, compounded in successive copies, might change or traduce parts of a text, is well known. Less widely appreciated is the circumstance that a thoughtful copyist making intelligent deductions about how to turn a corrupt or lacunose exemplar into a clean and logical transcription might distort the original more than his pedestrian counterpart content either to reproduce whatever was in front of him or simply to leave gaps where he encountered obscurities. In addition, a scribe’s rendering of one text could be affected by his knowledge of another – gospel manuscripts, for example, will commonly err in passages that were used in the liturgy or of which versions appear in more than one gospel, since familiarity with these alternatives could influence the copying of any individual version. As well as increasing the longevity of a literary text, the parchment codex enriched the possibilities for adorning it – illumination within, sumptuous bindings without – thereby enhancing paratextual features and, by extension, the roles that a book could fulfil. There are plentiful cases from the early Middle Ages where visual beauty was almost as important as the textual content in the overall effect and value of a volume; and when, by the later Middle Ages, one finds a bibliophile-connoisseur like Jean, duc de Berry (d. 1416), acquiring and commissioning multiple copies of similar texts (Books of Hours), distinguished by ever more elaborate decoration, it is difficult to resist the conclusion that the appearance had become more important than the written content. Paradoxical examples of the phenomenon are Codices Aurei or Argentei, deluxe manuscripts wherein conspicuous opulence was lavished on the text itself through the use of gold and silver inks, sometimes on purple parchment: for this very apotheosis of the words into glittering expressions of supreme value simultaneously rendered them more difficult to read. Extravagant bindings, such as the jewelled covers

Introduction

9

of the Middle Ages and the ornately gold-tooled or fabric-adorned ones of the sixteenth and seventeenth centuries (Fig. 0.2), are comparably paradoxical, for here the greatest care and resources were expended on the external casing rather than on any aspect of the text itself, and the elaboration and delicacy of such structures could then impose restrictions, direct and indirect, on how one might handle the book and hence read the words within. Preferring a hardback edition to a cheaper paperback version of the same text is a humble modern counterpart of the phenomenon, the Private Press movement – exemplified by the Kelmscott Press of William Morris (d. 1896) and the Eragny Press of Lucien (d. 1944) and Esther (d. 1951) Pissarro – its modern apogee. Binyon’s Dream-Come-True, which included a frontispiece designed by the author and which was noted above for its rarity, was produced by the latter. The most celebrated publication of the former is of course their Chaucer (1896). This grand folio – its type, border ornamentation, titles and initials all designed by William Morris, and containing over eighty illustrations by Edward Coley Burne-Jones – is a justly celebrated example of art and design that transforms encountering that poet’s oeuvre into a multi-sensory experience. No one would choose this edition simply in order to read what Chaucer wrote; and the admiration that has been lavished upon the book since the moment of its

Figure 0.2 New Testament (London: Robert Barker, 1640) and Metrical Psalms (London: ‘MF ’, 1641), together in a contemporary embroidered binding. Durham University Library, SB 2537.

10

Life of Texts

publication appropriately focuses on the finesse of its artistry and craftsmanship, not on the quality of the text – W. B. Yeats, for instance, declared it ‘the most beautiful of printed books’. Be it by illumination, woodcut, metalcut or a modern photographic process, the provision of illustrations on a page creates a dialogue between text and image. By punctuating the text, the illustrations necessarily structure it; by rendering certain sections of it in visual images, they accentuate these parts over others; while their content and artistry will colour the reader’s reaction to the work as a whole. The perils and potential of such artwork were articulated from the perspective of the writer by Anthony Trollope (d. 1882), who complained that ‘An artist will frequently dislike to subordinate his ideas to those of the author and will sometimes be too idle to find out what those ideas are’, while singling out for praise the contributions of John Everett Millais (d. 1896) on the grounds that ‘in every figure that he drew it was his object to promote the views of the writer whose work he had undertaken to illustrate and he never spared himself any pains in studying that work so as to enable him to do so’. Master practitioners can create illustrations of a power to match even the finest words. Those to Dante’s Divine Comedy by Giovanni di Paolo (d. 1482), Sandro Botticelli (d. 1510), William Blake (d. 1827) and Gustave Doré (d. 1883) are cases in point, each one replete with references and resonances that go beyond what is explicit in the text. Yet the contribution of less exalted artists can be just as profound: one thinks of the role of Sidney Paget’s (d. 1908) illustrations for Strand Magazine in determining the sleuthing image of Sherlock Holmes, or of E. H. Shepard’s (d. 1976) drawings for defining the anthropomorphized characters of Kenneth Grahame (d. 1932) and A. A. Milne (d. 1956). And in the pamphlet wars of the Reformation, the title-page imagery could itself be as effective a polemic as the texts that followed – vividly casting (and castigating) Catholic priests and monks as rapacious wolves, for example, or Martin Luther as the seven-headed beast of the Apocalypse. In works such as Dürer’s Apocalypse (1498: Fig. 0.3) and Holbein’s Dance of Death (1538), it is the illustrations, more than the text, that drive the book. The modern graphic novel thus has a long lineage – one, moreover, that stretches back beyond Holbein and Dürer, through the Biblia pauperum of the later Middle Ages, to the heavily illustrated manuscripts of late antiquity. The particular relevance of all such works here is that their synthesis of word and image in the service of narrative and characterization challenges any facile equating of ‘text’ with ‘writing’. Even more disruptive of simple paradigms are the books where, rather than collaborating to tell a tale or convey information, artwork and text are

Introduction

11

Figure 0.3 Die heimlich offenbarung iohannis (Albrecht Dürer’s Apocalypse) (Nuremberg: Anton Koberger, 1498), chapter 22.

in what might be described as apposition, their conceptual relationship ambiguous or even in tension. One thinks, for example, of publications like Vingt-cinq poèmes (1918) and De nos oiseaux (c. 1923), where woodcuts by Hans Arp (d. 1966) with their own artistic agenda are interspersed among poems by Tristan Tzara (Samy Rosenstock, d. 1963). Some of the semiotic implications of these and other paradoxical juxtapositions for texts, images, and their interrelationships were succinctly articulated – in combinations of word and image, appropriately – by René Magritte’s ‘Les Mots et Les Images’ in La Révolution surréaliste 12 (December, 1929), itself a response to the French Surrealists’ prioritizing of words (especially poetry) over images for expressing elevated ideas (Fig. 0.4). As modern publishers appreciate, the design of a book cover and the imagery thereon can be crucial factors in wooing potential readers to identify a title as of interest to them, and even in persuading them to select one in particular over its neighbours on the shelf. The axiom ‘Don’t judge a book by its cover’ has weight precisely because it is all too easy to do just that; and a single image, no matter how masterly, is unlikely to give a reliable impression of all that is within. The witticism ‘Don’t judge a book by its movie’ also has serious traction. This is not only because the requirements of cinema can mean a radically different

12

Life of Texts

Figure 0.4 René Magritte, ‘Les mots et les images’, La Révolution surréaliste (December, 1929).

interpretation of the work in question, but also and more fundamentally because the extended visual experience in question, which necessarily shows rather than evokes people, places and events, not to mention myriad ancillary details, crystallizes open-ended evocations into fixed definitions. These, whatever their merits, can prove ineradicable: can anyone who has seen the film version of Umberto Eco’s Il nome della rosa (The Name of the Rose) then (re)read the doings of William the Englishman without their image of him being coloured by the distinctive features and (Scottish!) accent of Sean Connery? Initially, printing with movable type was not the revolution that it is often assumed to have been. Yes, more copies could be made of a work and – as printers refined their techniques, exploiting the strengths of the new technology as opposed to using it to replicate manuscript modes – for a cheaper unit price. But the resulting texts were no more accurate than their manuscript predecessors (from which they were, in any case, copied – and by the hand of a typesetter who, as he was setting in reverse, was more likely than a scribe to make errors). Individual prints could not necessarily be produced faster than manuscripts (the pecia (‘piece’) system of later medieval Paris, whereby master copies of key texts were available as a series of separate fascicules which could be given to a team of

Introduction

13

scribes who could all work on different parts simultaneously, meant that a manuscript copy only once removed from the exemplar could be produced very rapidly). Nor were all the books issuing from a print run even identical. Decoration and articulation, often still added by hand, varied substantially from one copy to the next. More fundamentally, as proof-reading could only begin once printing was underway and as the initial uncorrected copies of each sheet were often used as well as the subsequent corrected ones, each finished book could in principle be composed of a different selection of corrected and uncorrected pages. The economic model behind many manuscript books produced in the later Middle Ages was fairly straightforward: a commercial scribe would copy a text when a client ordered it, being remunerated immediately before or after that individual job; materials – primarily parchment or paper – could be acquired shortly before they were needed. Printing, by contrast, presupposed an entirely different model (as the many early printers who went bankrupt discovered to their cost): there were substantial outlays – on paper and wages, not to mention the investment in the presses and the metal type – long before there was any prospect of recouping them from a product that could still take weeks and months to produce, let alone to distribute. The need to envisage selling, say, 500 copies in due course in order to cover production costs, and the challenge of getting those copies to the 500 potential clients in question, inevitably made printers cautious about what they would produce. Nor, if a title proved to be a ‘best-seller’, was there an easy way to capitalize on the phenomenon, for (until the invention and commercial application of the stereotype and electrotype processes from c. 1800) there was no way to reprint it without the same typesetting expenses and other overheads as the first time. Books and hence texts became, in consequence, more of a commodity than ever before. The need to market their wares meant that printer-publishers had strong motives to mould works to make them more attractive to the consumer. This could extend beyond elements of presentation and artwork, through the addition of prefaces of various forms (often, in effect, a sales pitch), to editing the work itself in one way or another, as did William Caxton (d. 1491) with Malory’s (d. 1471) Le Morte Darthur – not least in re-presenting its eight ‘Tales’ as twenty-one books. More radically, an author’s collected oeuvre might be enlarged with spurious items to ensure that the edition in question would seem more comprehensive than any rival one. Such seems to have been the policy of Johann Herwagen the Younger (d. 1564) for his edition of Bede’s Opera omnia (Basel, 1563). Many of these strategies survive in modern times in the work of literary editors, graphic designers and marketing managers, not to mention accountants.

14

Life of Texts

That printing gave a text stability is also an illusion – one that is laid bare by contemplation of almost any suite of early editions of a single work. They are likely to differ not only in some or all of format, font, layout, articulation, decoration and ancillary texts, but also in actual readings. If such is true of modern works like The Waste Land (published three times between October and December 1922, first in periodicals on either side of the Atlantic, then in book form, and again as a book the following year, each impression differing in aspects of presentation from the other two), it is all the more so of earlier ones. The first two editions of The Canterbury Tales (c. 1476 and 1484) both published by William Caxton, exemplify the point, being distinct both visually – the second with illustrations (Fig. 0.5), the first without – and textually. Furthermore, the continuing debate about the motives for the second edition and its relationship to the first – despite a Preface by Caxton himself that addresses precisely these issues – highlights the difficulties of identifying the reasons for some such changes. One impulse behind Caxton’s Preface was surely to persuade purchasers of the first edition that they ought now to buy the second on account of its superiority. How many did so is unknowable. What is abundantly clear, however, is that ‘improved’ editions regularly failed to supersede earlier ones, or would do so only very slowly. It was centuries before the Vulgate of St Jerome (d. 420) entirely replaced the multiple Old Latin translations that preceded it. The first edition of Don Quixote (what is now called ‘Part I’) by Miguel de Cervantes Saavedra (d. 1616), published in Madrid at the end of 1604 (though dated 1605), was riddled with errors, and its popularity encouraged the publisher to issue a revised text within a matter of months (Spring 1605). Nevertheless, the next two editions – both produced in Lisbon in 1605 – reprinted the error-ridden 1604 text, not the better 1605 one. A contemporary parallel is the reissuing – in print and online – of inferior nineteenth-century and earlier impressions of literary classics and other works, rather than better editions, for the simple reason that the former are no longer protected by copyright. Even with regard to classical and patristic authors, the editions that are freely available online are generally antiquated ones, whereas the better scholarly ones as also old but still ‘standard’ versions are jealously guarded behind pay-walls erected by publishing houses. The internet has not, as is sometimes claimed, ‘democratized access to texts’; rather, it has democratized access to inferior copies in which no one has a financial interest. Extreme cases both of the contrasts between early publications of a given work and of the prices that books can command are the first witnesses to the oeuvre of William Shakespeare (d. 1616). The various Quarto copies of his plays, on the one hand, and the First Folio edition of 1623, on the other, differ fundamentally not

Introduction

15

Figure 0.5 Geoffrey Chaucer, The Canterbury Tales (Westminster: William Caxton, 1483), General Prologue, Wife of Bath and Parson.

only in format and in the quantity of material that they include, but also in countless readings for the texts they have in common. Attempts to rationalize and reconcile their conflicting witnesses for a given play are often as revealing about modern critical attitudes – and hence one aspect of the life of the text in our own time – as about seventeenth-century circumstances and publishing practices. More generally, the phenomenon reminds us that for a playwright, it is performance rather than print that realizes a text. The extravagant six- and sevenfigure prices that copies of the First Folio now command reflect a unique combination of the author’s status, the authority of the texts, the comparative rarity of the edition (with some 250 copies extant, it is not particularly rare) and the prestige associated with owning what is widely recognized as a milestone of Western civilization (Fig. 0.6). Of printed books, only a Gutenberg Bible – the first of the type (pun intended!) and surviving in much smaller numbers than the First Folio (about fifty copies) – can be relied upon to fetch a consistently higher price. It is another of the many paradoxes of our field that unquestionably the most valuable printed book, as judged by sale-room figures and insurance premiums, is of negligible worth as a witness to the text it contains (an unremarkable Vulgate).

16

Life of Texts

Figure 0.6 Mr William Shakespeares Comedies, Histories and Tragedies published according to the True Originall Copies (London: Isaac Iaggad and Ed. Blount, 1623), 130–1. Durham University Library, Cosin W.2.11.

If the earliest manuscripts and the first printed editions have a cachet by dint of being the oldest available record of a text and of the form in which it was first published, subsequent recensions and editions sometimes introduced valuable features of their own. Copies of writings by the Fathers of the Church that were produced from the thirteenth century onwards tended to incorporate organizational elements such as running headings and paraphs (coloured markers highlighting subdivisions) that, responding to the changing requirements of readers, made them more suitable for reference purposes than earlier manuscripts of the same works, better written though those generally were. Occasionally such added features may themselves become part of the orthodox text. The standardization of Biblical chapters during the thirteenth century (Fig. 0.7) is one such case (recognition of the value of the system being manifested in the time and effort that was then devoted to inserting the new divisions into older copies), the introduction of verses in the French (1553) and Latin (1555) Bibles issued by Robert I Estienne another. In the same way, the subdivisions that the scholar-printer Henri II Estienne (d. 1598) inserted into

Introduction

17

Figure 0.7 Bible with ‘standard’ chapter divisions; Paris, third quarter of the thirteenth century. Durham Cathedral Library, A.II .3, fol. 313r.

his edition of Plato defined a standard system for organizing those texts. A particularly formidable example of the phenomenon was the gradual emergence from informal annotations in the margins of books of the Bible and legal works of formal apparatuses of commentary known as Glosses (Fig. 0.8): these were then laboriously transcribed and subsequently printed as an integral part of the

18

Life of Texts

Figure 0.8 Glossed Psalter, Northern France or England; middle of the twelfth century. Durham Cathedral Library, A.III .9, fol. 51v.

texts in question. Select literary works might be treated in the same way – whence the early editions of Virgil wherein his poetry was surrounded by an extensive apparatus of commentary. Modern equivalents, albeit with a much shorter lifespan, are the study editions bristling with explicatory notes. In all such cases,

Introduction

19

the page presents texts of different statuses in dialogue with each other, permitting the reader to witness and even participate in an interaction of minds. Alongside such influential restructuring and augmenting of texts, one sometimes finds a more fundamental instability – namely the identity of the author. As touched upon in connection with sharp publishing practices, great names attract spuria, both from the desire of readers for a larger canon from a respected figure than is in fact available, and from the temptation to promote lesser works on the reputation of an established authority. The Virgilian apocrypha are one such case, the manifold writings falsely attributed to Bernard of Clairvaux (d. 1153) another – and here it is said that of the works circulating under his name in the Middle Ages, more were actually pseudonymous than genuine. Similarly, in the seventeenth century, ‘A Yorkshire Tragedy’ by Thomas Middleton (d. 1627) appeared under the name of William Shakespeare; while the runaway success of John Bunyan’s The Pilgrim’s Progress (1678) inspired not only The Second Part of the Pilgrim’s Progress by T[homas] S[herman] – to which Bunyan’s own Second Part (1684), with its criticism of those who ‘counterfeit my pilgrim’, was in part a response – but also a spurious Third Part (1693) published five years after Bunyan’s death. Now as then, the name of the author can make a difference to how a text is received and the circulation it is likely to enjoy, as the infamous ‘Hitler Diaries’ demonstrate. Conversely, in relation to works that have proved very popular and, in consequence, highly profitable yet whose creators have inconsiderately proved to be mortal, there is a temptation to engage others to continue the series – as has happened with Ian Fleming’s James Bond and Leslie Charteris’s The Saint. Certain authors have themselves added a dimension of ambiguity by concealing their identities behind pen names. Currer, Ellis and Acton Bell (Charlotte, Emily and Anne Brontë, d. 1855, 1848 and 1849, respectively), George Sand (Amandine Aurore Dupin, d. 1876) and George Eliot (Mary Ann Evans, d. 1880) are famous female examples, Boz (Charles Dickens, d. 1870), Lewis Carroll (Charles Dodgson, d. 1898) and Louis-Ferdinand Céline (Louis-Ferdinand Destouches, d. 1961) well-known male ones. Sieur Louis de Conte and Mark Twain (both Samuel Clemens, d. 1910) and Tristan Ruia, S. Samyro and Tristan Tzara (all Samy Rosenstock, d. 1963) remind us that such alternative identities could be multiple. Impeding the reader from prejudging a work by its author’s name, sex or established reputation, while shielding authors from being branded by their oeuvre (or one part of it) or persecuted on account of it, occasionally even a device of disorientation and jest, the pseudonym casts a curtain of fiction between a text and its consumer – albeit one that can be pulled aside or ripped asunder at any point. When the real identity and circumstances of an author are

20

Life of Texts

irretrievable – be it behind a now impenetrable pseudonym, such as the ‘Alonso Fernández de Avellaneda’ used by whoever published a continuation to the first part of Don Quixote in 1614, or simply lost in the mists of time, as with Homer – a dimension to our understanding of the work vanishes with them. A different sort of curtain between author and reader is translation. If this is superficially an overt and transparent barrier, it can easily become a covert and impermeable one. Translators themselves readily declare that their work necessarily involves interpretation (one need only think of the contrasting overtones of The Waning of the Middle Ages, as opposed to The Autumn of the Middle Ages as renderings of Johan Huizinga’s Herfsttij der Middeleeuwen (1919), the first word of which was a neologism implying literally ‘autumntide’ or ‘harvest-tide’), yet most consumers rapidly forget this. How many Anglophone readers who would unhesitatingly say that they ‘knew’ The Odyssey, The Aeneid, Beowulf or The Divine Comedy really mean that they are acquainted with them through the mediation of Alexander Pope, E. V. Rieu, Seamus Heaney, Dorothy L. Sayers or other translators? The vigour with which certain individuals will defend a particular translation of the Bible, irrespective of any real knowledge of that version’s fidelity or otherwise to the sense, let alone the tone, of the original Hebrew and Greek speaks volumes on the point. The strengths and limitations of Pope’s paraphrase of Homer were succinctly articulated by the great classical scholar, Richard Bentley (d. 1742): ‘A pretty poem, Mr Pope, but you must not call it Homer.’ Moreover, for every translation that has been revised by the author of the original work, there are countless more that have not been so vetted. Nor is it inevitable that an authorial version will be the most influential. Leon Battista Alberti (d. 1472) first composed his treatise on painting, De pictura, in the vernacular, subsequently working up a more polished version of it in Latin; however, it was the Italian rendering of the latter produced over a century later by the prolific translator and editor Cosimo Bartoli (d. 1572) that assured the circulation of the text and, long mistaken for Alberti’s own work, served as the basis for translations into other languages. Nonetheless, some translations are recognized to have independent merits of their own, be it on account of the authority of the translator or the literary qualities of the result: Marsilio Ficino’s (d. 1499) rendering of Plato as presented in the edition of Simon Grynaeus (d. 1541) might exemplify the former, the socalled ‘Authorized’ or ‘King James’ version of the Bible (1611) the latter. And there are plentiful cases where translations have almost certainly been more influential than the original version in the dissemination of the work. The case of Bartoli’s De pictura has just been mentioned, and Seamus Heaney’s Beowulf is a recent example of the same phenomenon.

Introduction

21

Printed copies of texts, no less than manuscripts, remain living organisms to the extent that they receive the annotations of readers. Indeed a few classes of texts, such as Books of Hours (collections of Psalms, prayers and readings structured for daily private devotions), were as, if not more likely to receive annotations if printed than when copied by hand: because printed editions (unlike manuscripts) could not be customized to the individual client’s needs while they were being produced, any desired personalization had to be added by hand subsequently. Annotations of all periods can range from merely flagging passages of interest, through occasional written responses triggered by the text, to concerted dialogue with it. The proscriptions against writing in books that feature in myriad library regulations speak volumes about the widespread impulse of readers to do precisely that – and hence of the power of texts to evoke reactions in writing from those reading them. When they are identifiably from the pen of a prominent figure like Petrarch (d. 1374) or John Dee (d. 1609), such additions command special interest. Indeed, it is almost impossible now to read anything by Sir Joshua Reynolds (d. 1792) without recalling the annotations that were inflicted upon one particular copy of the second edition of his Works (1798), the tone of which is set at the outset, with the vigorous jotting at the head of the title page: ‘This man was Hired to Depress Art. This is the opinion of Will[iam] Blake | my Proofs of this Opinion are given in the following Notes’ (Fig. 0.9). Yet all annotations reflect interactions with the texts in question and are witnesses to their reception. Some additions themselves attract annotations from subsequent hands – a phenomenon which, in addition to its intrinsic interest, has further value as a precious written reminder of the many oral debates about texts that have left little or no record of their existence. The more general point here is that texts relentlessly spawn other texts, as readers comment on them, be it in the margins or elsewhere. Indeed, the body of commentary to which many texts have given rise is truly formidable – to the extent that it can seemingly threaten to submerge the original work. Fortunately, however, critics and commentators have yet to find a way to stop the innocent enthusiast from simply reading the source work unimpeded by their accumulated ‘wisdom’. We are too close to the unfolding digital revolution to be able to evaluate even its short-term implications for the life of texts; nevertheless, several aspects that are not in doubt may still be highlighted. The new media have revolutionized the rapidity with which new texts can be distributed; they have removed the necessity for formal publishers to mediate the process (the private circulation of works that flourished in the Graeco-Roman world and continued in the manuscript era is, in effect, enjoying a renaissance on a hitherto unimaginable scale); and they

22

Life of Texts

Figure 0.9 Joshua Reynolds, Works . . ., 2nd ed.; 3 vols. (London, 1789), vol. I, title page; annotated by William Blake. (London, British Library, C.45.e.1820.)

have enormously enhanced the flexibility of ‘formally produced’ texts, whose appearance as well as their content can be altered with ease by diverse hands in the course of circulation. Simultaneously, the new media have made it much easier to gain access to countless older works, steadily eliminating the concept of ‘out of print’ – never before in the history of humankind have so many texts been so readily available to so many people. They have also provided the means to search huge quantities of textual content in seconds, redefining how such material can be used – and, arguably more radical, separating such use from the process of reading as traditionally conceived. No less important, however, is what the new media have not done. They have not created any more time in which to read the burgeoning quantity of available writing: never before in the history of humankind have so many accessible texts languished unread by so many people. Nor have they provided a medium that is demonstrably durable. It seems unlikely in the extreme that any of the current electronic resources will enjoy the longevity of parchment codices which, if

Introduction

23

treated with respect, can be as good as new after 1,000 years and more. A couple of centuries from now, while medieval manuscripts and books printed on acidfree paper will still be readily available, most of the electronic resources of our own time are likely to have perished or become unusable – just like the papyrus rolls of antiquity. This reminds us of a more general truth as old as literary composition itself: that the survival of texts may, in part, reflect the importance and popularity of the works in question, but it is also dependent on numerous external factors, principal among which is the durability of the materials to which they are entrusted. The poetry of Sappho (fl. seventh century bc ) has largely vanished, not through any lack of admiration (she was, after all, hailed as the tenth Muse), but because of the limitations of the available technology for preserving it. In addition, certain works have had to face the extra challenges of persecution, be it selective censorship or outright destruction: the sixteenth century in particular, with the Reformation on the one hand and the Catholic Church’s Index librorum prohibitorum (1559+) on the other, was a particularly challenging time for many texts. Censorship can, however, backfire, fostering rather than suppressing interest in the writings in question. The formidable demand for the 1960 Penguin paperback edition of Lady Chatterley’s Lover that was created by its obscenity trial is a case in point. In relation to the handwritten codex, the number of surviving copies can be an index of popularity. The more than 800 manuscripts of Dante’s Divine Comedy assuredly attest to real interest at an early date, as too do the more than 100 manuscripts of Bede’s Historia ecclesiastica, or the more than fifty of Chaucer’s Canterbury Tales. But what should one then make of works like Gawain and the Green Knight and Beowulf, each known from but a single copy? To what extent is this a side-effect of the circumstances of composition or transcription, then of accidents of survival; and how far might it reflect contemporary disinterest? To be more specific: did such works lack the occasion to become better known and hence multiplied, or did they simply not engage contemporaries to the extent that the modern critic might expect? Context, opportunity and even luck may be as important as literary merit in achieving distribution – as many contemporary authors who have failed to make the best-seller lists would doubtless agree! While enthusiastic circulation can be taken to reflect real popularity, minimal circulation does not necessarily mean the reverse, because so many imponderables are involved. The contrast between the nearly 200 editions of Virgil that were produced in the first fifty years of printing with movable type and the fewer than twenty of Homer (of which, moreover, only three were in Greek) is an equivocal

24

Life of Texts

guide to the relative popularity of their respective oeuvres during the second half of the fifteenth century because it is skewed by the geography of early printing itself (initiated in Germany), by the fact that Greek type only became available in the 1490s (forty years after Latin type), and by the commercially motivated conservatism of early printers. Equally the elevated circulation figures and personal celebrity enjoyed by writers in the nineteenth century had less to do with the intrinsic merits of their work than with the flourishing then of diverse literary organs, their production made feasible by steam-powered lithographic presses, allied to rising educational opportunities and the founding of increasing numbers of subscription and public libraries – all underpinned by burgeoning urbanization, the abolition of paper duty, the use of grass rather than rags to make it, and the opportunities for cheaper, faster distribution via the railways. Nor, conversely, does popularity necessarily equate with literary merit (as a glance at almost any modern best-seller list will underline). The definitive decoupling of degree of diffusion from quality of text is another dubious achievement of the internet. That said, fluctuations in taste – or, to be more precise, in the supply of, and demand for, a particular genre of writing – are a reality. One may contrast the formidable popularity of Arthurian literature in the later Middle Ages and of detective fiction in our own age with their exiguous existences prior to the twelfth and nineteenth centuries, respectively. Moreover, the shifts and contrasts can be swifter and more localized: what is appreciated at one moment may fall from favour shortly afterwards, and what is devoured within a particular circle may be shunned by others. If Dickens (d. 1870), Dostoevsky (d. 1881) and Zola (d. 1902) retain, after some oscillations, a broad audience (of one sort or another) in the twenty-first century, other major nineteenth-century writers, celebrated in their day, fail to do so. And it is a lamentable fact that the most insightful, acclaimed and accessible academic work is unlikely ever to be consulted by the public at large. More generally, it is sobering to reflect that With each new writing of whatever sort produced, The time for reading through the corpus is reduced.

Citations are a similarly slippery guide to popularity. Do they show that the work was really known, or rather that certain extracts from it had a currency of their own? It is revealing in this connection that one of the ancient authors best represented in terms of manuscript survival was the grammarian Donatus (fl. fourth century ad ), for not only does this advertise the importance of pedagogy per se, it also alerts us to one of its consequences for the lives of texts. What is

Introduction

25

studied at school is often retained in a way that texts read later in life, be it for work or leisure, are not. When antique and medieval authors cite excerpts of classical literature, was it because they really knew the works in question or rather because these were the passages that they had learned as part of their schooling (Fig. 0.10)? What percentage of the people who today could readily quote ‘To be or not to be . . .’ would be able to outline the plot of (say) five other Shakespearean plays, let alone recite lines from Titus Andronicus? And the power of the game that asks participants to identify whether a given quotation is from Shakespeare or the King James Bible arises in part from the fact that the familiarsounding phrases in question will often have been encountered out of context. Yet such ‘indirect’ transmission of a work or extracts from it, through teaching, anthologizing and popular culture of various forms, is part of the life of a text as much as its circulation in toto. Homer’s Odyssey, Virgil’s Aeneid, the Bible, Dante’s Divine Comedy, Shakespeare’s plays, Milton’s Paradise Lost, Karl Marx’s Das Kapital and so on live as energetically – if not more so – outside the context of their canon proper as within it. It is a nice point whether having a few ‘tags’ from a particular work as

Figure 0.10 Writing exercise on papyrus (Virgil, Aeneid, Book XI , lines 371–2), second half of the first century ad. (Oxford, Sackler Library, P.Oxy. L. 3554.)

26

Life of Texts

part of one’s active mental repertoire – the case of ‘For the Fallen’ was highlighted at the outset – means more or less for the life of that text than possessing an unopened copy of the entire opus. The shift in book manufacturing during the twentieth century to providing volumes with cut (as opposed to uncut) pages has removed an invaluable index to the possession, as opposed to the consumption, of literature of all sorts. The potential through electronic media of discovering not only how many people have obtained a copy of a particular work during a given time period (the e-book most frequently downloaded from Project Gutenberg at the time when this chapter was being written was The Adventures of Sherlock Holmes), but how much and what exactly of it they have read offers fascinating possibilities to make up for this (it would be very interesting to have such data for the fourth most popular text on the very same download list – Beowulf ). It would, one suspects, put flesh on the bones of Mark Twain’s jest that ‘A classic is something that everybody wants to have read but nobody wants to read’! In this connection, it is appropriate to highlight the paradoxical nature of libraries as institutions. Although generally considered organs for facilitating the preservation of and access to written (and other) heritage, libraries exclude as well as include. They may exclude readers of one sort or another, be it directly by constitution and regulation or indirectly by location and ambiance; they will exclude written material for reasons ranging from cost, through space, to content; and many will periodically de-accession, even destroy, material alongside continuously accumulating it. There is a perpetual tension between, on the one hand, permitting books to be borrowed – meaning that they will not then be available to other users, and are more exposed to the risk of damage and loss – and, on the other, confining them to the building, which is more likely to guarantee their security and long-term availability but limits when, how and by whom they can be consulted. Concerns over who was admitted and what they should be able to consult, where and for how long are overt in the statutes of medieval libraries, but continue more covertly in many other systems: as late as the second half of the twentieth century, public libraries in Britain had ingenious systems designed to restrict the borrowing of works of popular fiction as opposed to non-fiction. The most elementary – and effective – form of censorship, be it by commission or omission, is simply not acquiring a particular work. * The forces to which texts are subject are thus manifold. Not only do readers excerpt and misremember, but scribes miscopy, typesetters miss-set, editors and publishers emend and reshape, and translators interpret; technologies and

Introduction

27

institutions then influence who will have access to them, in what forms, and for how long. Correspondingly, all efforts to stabilize and control texts are in vain. Each reader and hearer will receive and use them differently, no matter how ‘stable’ the form in which they are transmitted. For every reader brings unique experiences and perceptions to his or her interaction with any work, which, moreover, change with each encounter with it – when one returns to a text on a subsequent occasion, it is inevitably in the light of different experiences (not least prior knowledge of that work itself), creating new patterns in the perceptual kaleidoscope of text, context, experience and intellect. ‘What is a text?’ is a philosophical question which, in these brief reflections, has implicitly been given the pragmatic answer that it is, in essence, what one finds in a society’s books. Put thus bluntly, such a definition is patently simplistic and reductive: for with every human interaction, a text takes on a new lease of life as it is moulded, excerpted and juxtaposed differently in the mind, if not on a page. Yet while readily admitting the conceptual limitations of such a model, the key point in the present context is surely that, inadequate though our bookbased approach may be, it has revealed a field that is multifarious and complex, in which texts can indeed be seen to live – breeding, procreating, aging, sometimes even dying and being reborn – subject to manifold forces, internal and external. These are just the bare bones of a subject whose reality – a lens through which much human thought and interaction is refracted – is almost infinitely varied. The closely focused contributions that follow in the present collection all in their different ways put flesh on the bones of the boundless yet endlessly engaging field that is the life of texts.

1

Editing Homer Barbara Graziosi

Editing the Homeric poems is a demanding task. The poems are long and there are many manuscripts and papyrus scraps to consider; the latter, moreover, keep turning up from the sands of Egypt, from mummy cartonnage and even more often from libraries where they languish unedited for years, decades and even centuries after discovery.1 Above all, however, editing Homer is a demanding task because it raises difficult issues of method, practice and aim – which are related to the fundamental question of what Homeric epic is and should be. Great as they are, the practical difficulties pale into insignificance when compared to the challenge of establishing a workable theoretical framework for the task. What I offer here is a brief outline of some of the main issues an editor of Homer needs to face, starting from uncertainties regarding authorship, methods of composition, ancient performance and ancient editing. I then offer some examples of how those uncertainties play out in practice, in the modern editing of the text. I conclude with some considerations about how we may most usefully approach the Homeric text in future, in order to improve our understanding of its meaning and history. This chapter aims to foster an interdisciplinary discussion of editorial theories and practices. For this reason, I use transliterated Greek and translations throughout, indicating in the footnotes where more specialist guidance can be found. As well as providing some orientation on current approaches, this chapter also aims to inspire further specialist work, by suggesting some lines of enquiry which, in my view, seem promising.

Authorship The first surviving sources that mention Homer by name date to the sixth century bc : from them, we can establish that Homer was thought to be an 28

Editing Homer

29

ancient and authoritative poet – but also that nothing much was known about him. Several cities, for example, claimed to have been his birthplace, but none seems to have commanded universal authority on the matter. In fact, even the name ‘Homer’ was disputed: most ancient authors seem to have used it in straightforward manner, though some insisted that it was only a nickname meaning ‘blind’, or ‘hostage’.2 As so often in Homeric studies, ancient uncertainties still reverberate in current scholarship: ‘Homer’ is not a standard Greek personal name, but it is not an obviously made-up name either, which means that some scholars accept it as the name of a real individual while others suggest that the Iliad and the Odyssey are, in point of fact, anonymous.3 As well as doubts about the name and identity of Homer, the ancient sources suggest uncertainty about which poems, exactly, this great and elusive poet Homer once composed. The authenticity of the Iliad was never questioned in antiquity, but that of the Odyssey sometimes was, and a host of other epic poems – now surviving only in fragments and plot summaries – were sometimes attributed to Homer and sometimes dismissed as inauthentic. In general, it seems that definitions of Homer’s oeuvre narrowed in the course of time. In the sixth and early fifth century bc , Greek authors tended to treat Homer as the author of whole epic sagas, and more specifically a Trojan Cycle revolving around the Trojan War, its origins and its aftermath, and a Theban Cycle dealing with Oedipus, his descendants, and the siege of Thebes. They also attributed to Homer other epic poems, such as The Capture of Oechalia (also surviving in fragments) and several hymns to the gods (which are still extant).4 Already in the second half of the fifth century, we know that Homer was expected to be self-consistent and that contradiction was considered a reason to doubt authenticity. Herodotus, for example, made the following observation about a specific passage in the Cypria, a cyclic poem that focused on the origin and early stages of the Trojan War: These lines and this passage are no small proof that the Cypria is not by Homer, but by someone else; for in the Cypria we are told that, when Paris led Helen away, he arrived from Sparta to Troy in three days, enjoying a fair wind and a smooth sea, whereas in the Iliad Homer says that he wandered off course when he took her away.5

The technical language Herodotus uses, as well as the fact that he offers just one argument (implicitly among several) against authenticity – suggests a background of scholarly debate, which is now lost to us. What deserves attention, as part of that lost context of criticism, is the assumption that Homer never contradicts

30

Life of Texts

himself – unlike, say, the lyric poet Stesichorus (who offered one account of the rape of Helen, and then recanted it in a different poem) and unlike the playwrights, for that matter, who created different (and often contradictory) versions of the same myth year after year. In short, Homer was allowed to believe only one thing about Helen’s journey, and was not supposed to change his mind – not even when composing different poems. We can reconstruct at least two other expectations of Homer in the classical period (fifth and fourth centuries bc ). First, the poet was thought to be a good imitator of character. In Plato’s Ion, for example, a rhapsode (a professional performer of epic) claims that Homer knows ‘the kind of things that a man says, or what a woman might say, or a slave, or a freeman, or someone receiving orders, or someone giving them’.6 In Aristotle’s Poetics, this ability to imitate character is expressed in terms of a distinction between main narrative and character speech: Homer deserves praise for many things but especially for this, that alone of all poets he does not fail to understand what he ought to do himself. The poet should speak as seldom as possible in his own character, since he is not representing the story in that sense. Now the other poets play a part themselves throughout and only occasionally imitate a few things, but Homer after a brief prelude at once brings in a man or a woman or some other person, never without character, but all having characters of their own.7

The emphasis, in Aristotle, is not just on what the poet says, but on what he omits, or lets his characters say for themselves. This ability to select and arrange is the third quality attributed to Homer in the late classical period which I would like to discuss here. Aristotle differentiates between the Iliad and the Cypria not in terms of factual consistency, as Herodotus did in the fifth century, but in terms of design, focus and, explicitly, omission. In his discussion of the Odyssey, he likewise comments on what the poet leaves out: whether out of natural instinct or poetic technique . . . Homer did not put into the Odyssey all that ever happened to Odysseus, his being wounded on Parnassus, for instance, or his feigned madness when the army was assembled – since these events did not necessarily or probably led one to the other; but rather constructed his Odyssey around a single action in our sense of the phrase. And the Iliad the same.8

The three qualities attributed to Homer in classical sources, namely consistency, good imitation of character and the ability to select, are of course crucial skills for the orator – and it can be no coincidence that, in this period, the study of

Editing Homer

31

Homeric epic was considered foundational for a good rhetorical education.9 Still, these three qualities expected of Homer continued to shape ancient approaches to the poems attributed to him even when these became the focus of intense scholarly attention in their own right – rather than as a means of training future public speakers.10 In the Hellenistic period, after the conquests of Alexander the Great and the foundation of Alexandria, the Homeric poems became the object of intense scholarly work in the library attached to the Museum, an institution devoted to the cult of the Muses, as the name suggests. In the third and second centuries bc , scholars working in the library of Alexandria applied ever more stringent criteria in order to establish what was truly Homeric. They analysed in detail the diction and grammar of the Iliad and the Odyssey, on the basis of several different manuscripts they collected in different Greek cities, and kept to their great library.11 They placed a special sign – a long dash called the obelos – next to lines or passages whose authenticity they doubted, and then had full discussions and explanations in separate commentaries called hypomnemata. These commentaries are lost, as are the Alexandrian editions. However, some papyri preserve texts with signs of scholarship on them, and even snippets of commentary (Fig. 1.1). As well as trying to establish exactly what Homer had composed, scholars working in the Alexandrian library speculated about the poet’s ‘character’ (ēthos), ‘habit’ (ethos) and ‘persona’ (prosōpon). Homer, they insisted, had to be ‘clarified with reference to Homer’ (Homēron ex Homērou saphēnizein).12 In short, the process of editing was, in ancient Alexandrian practice, intertwined with the attempt to establish what Homer would have said and, more specifically, with the assumption that he would have been consistent across his oeuvre. When contradictions were detected, scholars sometimes explained them by claiming that they arose out of Homer’s talent for characterization. In fact, they established the ‘solution from character’ (lysis ek tou prosōpou), namely the idea that characters could contradict each other, the poet and even themselves.13 The clearest statement of this ancient approach to textual contradiction is found in a discussion about Homer’s views on wine consumption before battle. In Iliad 6, Hecuba and Hector disagree on the subject: she offers her son some wine, claiming that it will give him the strength required to face the battlefield; he refuses, stating that, on the contrary, wine would weaken him. Here is what Porphyry (third century ad ) had to say on the matter, on the strength of earlier, Alexandrian discussions:

32

Life of Texts

Figure 1.1 In this papyrus fragment dated to the second century bc (University of California 2390, P.Tebt. 4), we see in the margin between the two columns an obelos in correspondence with the first line, and below the letter beta, β, which marks line 200 in book 2 of the Iliad. The sign is only one line off with respect to medieval manuscripts and modern editions.

The question here is why the poet contradicts himself, since, after saying ‘When a man is tired . . . [wine greatly increases his strength]’ he now states ‘Do not . . . [offer me mind-cheering wine and] sap my limbs of strength’. According to many, the solution to this problem is as follows: one thing is Hecuba’s character, who says that wine is useful; and another is Hector’s, who denies that; and it is not surprising if, according to the poet, contradictory things are said by different characters. All the statements he makes in his own character need to be consistent and not self-contradictory; whereas all those he ascribes to characters are not his

Editing Homer

33

own, but are thought to belong to those who speak them; for this reason, he often allows dissonance, as also in these verses.14

The solution from character was sometimes used, in antiquity, to defend or emend the text. So, for example, one ancient scholar, Zenodotus, suggested that Telemachus could never have said of Odysseus’ return ‘it will never happen, not even if the gods want it’, because the statement amounted to blasphemy, i.e. a denial of divine power.15 He therefore corrected the text to ‘this will never happen, unless the gods want it’. His younger colleague Aristarchus, however, disagreed on the matter, arguing that Telemachus had spoken with the typical petulance and easy despair of the young. Telemachus, in short, was too young to be ‘long-suffering’, like his father: ‘he said this exaggerating, according to his character: not understanding this, Zenodotus wrongly corrects to “unless the gods wish it”.’ 16 As well as expectations of consistency on the part of the poet (‘clarify Homer with reference to Homer’), and a keen interest in the complexities of characterization (the ‘solution from character’), ancient scholars made interesting arguments about omission and arrangement, when trying to establish the correct text of the Homeric poems. The most famous example concerns the only reference to the judgement of Paris in Homer, which features in the last book of the Iliad. According to a famous story, Aphrodite, Hera, and Athena approached Paris when he was tending his flocks and asked him who was the most beautiful. When Aphrodite bribed him with the promise of the most beautiful woman on earth, namely Helen, Paris decided to spurn the bribes offered by the two other goddesses, and vote for Aphrodite. That, then, precipitated the Trojan War: Paris eloped with Helen, stealing her from her Greek husband Menelaus and taking her to Troy; the Greeks then set out on an expedition to retrieve her. The Iliad never narrates or even refers to the judgement of Paris, even though Aphrodite sides with Paris and the other Trojans throughout the poem, whereas Hera and Athena are staunch supporters of the Greeks. There is an allusion to this story only at the end of the poem. Aristarchus suspected that the lines were a late addition and placed obeloi next to them. Other ancient scholars disagreed, on arguments that are best preserved for us in the work of the great Byzantine Homeric scholar Eustathius (twelfth century): See how the poet held in store for the end of the Iliad the event that above all caused the Trojan War, keeping the listener in suspense to that degree. Some athetize the passage: if the poet had known about the judgement concerning beauty, they say, he would have mentioned it often. But concerning this we can

34

Life of Texts say that the poet mentioned many other things just once, and they are not athetized.17

Eustathius refers here to the ancient practice of identifying hapax legomena, ‘things the poet said only once’, and points out how frequent these unique expressions and references are. What we see, here, is an argument that combines the dictum ‘clarify Homer from Homer’ with an interest in architecture and arrangement, as the point about ‘suspense’ confirms. Even this brief discussion suggests two general points. First, the scholars working in the library of Alexandria in the third and second century bc shared with earlier readers and audiences of Homeric epic some important assumptions and expectations of the poet: consistency, good imitation of character, and the ability to select and arrange mythical material so as to offer an effective narrative. Second, those assumptions continued to shape a long tradition of exegesis, still in evidence in Eustathius’ commentaries and, indeed, in contemporary scholarship. We have inherited from the Greeks not just the name ‘Homer’ – and several portraits and legends about this figure – but a habit of editing the Iliad and the Odyssey with reference to their imagined author, mindful for example of the difference between the words of the poet and those of his characters, and of the choices made in selecting and arranging mythical material. The connections between ancient editorial work on Homer and modern attempts to restore the text can be pinpointed in detail. The first full manuscripts of the Iliad and the Odyssey which have reached us come from Byzantium and contain, in their margins, copious notes on the texts, called ‘scholia’ (Fig. 1.2): these medieval marginal notes often summarize the work of ancient scholars working in the library of Alexandria and are, for that reason, essential evidence for any modern editor of Homer.18 They contain, for example, variant readings, showing that Alexandrian scholars were themselves working with different manuscripts and versions of the Homeric poems and perhaps making conjectures of their own. Also important for modern editors are extensive Byzantine commentaries, particularly those written by Eustathius, bishop of Thessaloniki, who has already featured in this discussion: these commentaries incorporate a long tradition of commentary and editing, which reaches back to the library of Alexandria and, indeed, further back still. From Byzantium, the manuscripts travelled to Italy.19 When Western readers encountered the Iliad and the Odyssey for the first time they found the poems surprising, especially in relation to what they thought they already knew about Homer from Latin sources, and thus reopened the question of their authorship.

Editing Homer

35

Figure 1.2 The beginning of Iliad 1, surrounded by scholia in the so-called ‘Venetus A’ manuscript (Codex Marcianus Graecus 822), c. tenth century ad.

36

Life of Texts

In the eighteenth century, Giambattista Vico argued for the first time that the Homeric epics could not be the creation of a great poet, but rather stemmed from the collective, popular culture of the ancient Greeks. He insisted that they were too ‘vile, rude, cruel, proud . . . unreasonable, frivolous and light’ to be the product of an exceptional individual, and added that inconsistencies of style and factual detail pointed to collective authorship.20 Some decades later, the German scholar Friedrich August Wolf formulated a similar hypothesis – this time, however, not on the basis of ethical and aesthetic judgements of the kind made by Vico, but specifically on the basis of the scholia found in the Venetus A, which had recently been published.21 Wolf concluded that Homeric epic was the product of ancient editing and revision based on early, oral compositions. Wolf ’s Prolegomena ad Homerum (1795) privileged the history of the text over the identification of an original poet: it may be argued that Wolf admired the work of the Alexandrian scholars more than Homeric epic itself and insisted that modern philologists could do even better than their ancient Alexandrian ‘colleagues’ in identifying different strata of Homeric composition. His formulation of the ‘Homeric Question’ in his Prolegomena defined classics as a modern discipline – not least because he insisted on, and in part demonstrated, the possibility of philological progress. Yet even Wolf could not escape the ancient terms of the debate about Homer. Goethe satirized his work in a biting couplet, in which he suggested that Wolf, or ‘the wolf ’, was tearing Homer apart, so that every ancient city that had once claimed him could now finally have its own piece of the poet.22 Goethe engaged closely with Wolf ’s theories and explored their creative potential, even suggesting that he himself might be the latest composer of Homeric epic, but he eventually gave up on that line of exploration, advocating rather a focus on the artistic integrity of Homeric poetry. Eventually, two approaches to Homeric epic emerged in response to Wolf ’s work: the analysts, who tried to identify different strata of composition, with associated different poets; and the unitarians, who insisted on the artistic unity of the poems and saw them as the creation of one or, in some cases, two poets. Contemporary scholars working on Homer still often insist on a specific view of authorship – for example, they may argue that the poems are the product of oral re-composition in performance over a long period of time; or that the Iliad was composed by one poet at the end of the eighth century.23 Despite ongoing debates, the main point was already articulated by Nietzsche in his inaugural lecture of 1869: ‘Homer as the poet of the Iliad and the Odyssey is not a transmitted, historical fact – but rather an aesthetic judgement.’24 Arguments about the composition of the Iliad and the Odyssey are

Editing Homer

37

therefore intertwined with judgements about their beauty, and assumptions about how that beauty may have been created. We can ask, for example, whether the poems were the gift of a single individual or the result of a more collective effort; how long the process of composition lasted; whether variants represent different versions of a fluid text or attempts to improve or correct an identifiable original version; and, more generally, where composition ends and reception of a stable text begins.

Composition and Performance As must by now be obvious, in order to edit Homeric epic, it is crucial to come to a view about the line between composition and reception. One of the most important features that needs to be taken into consideration when investigating that line is the formulaic nature of Homeric verse. The sea is invariably ‘winecoloured’ in Homer – despite the strangeness of this expression in light of modern perceptions of both wine and sea. Women are always ‘slim-ankled’. War is ‘bad’. Homeric formulae, as these phrases are called, are important evidence for the composition of the Iliad, the Odyssey, and the many other epic poems that were sometimes attributed to Homer in antiquity. The decisive breakthrough concerning their function was made in the 1930s, when two young American scholars, Milman Parry and Albert Lord, set off to record the oral epics performed in what was then Yugoslavia, and more precisely Bosnia.25 They recorded the performances of many illiterate singers, which they then proceeded to study systematically. As a result of this work, they were able to prove that Bosnian singers, or guslari, relied on a complex system of formulae which helped them describe characters, places, actions and situations to a specific epic rhythm – without having to take too long thinking over appropriate turns of phrase in live performance. Parry and Lord were, moreover, able to show that the same techniques were used in the composition of Homeric epic. For example, if an epic singer wanted to say ‘Achilles’, he could choose between different formulae, each of which was designed to take up a different number of measures in the line. Depending on how long he needed to take over it, he could either say ‘Achilles’, or ‘luminous Achilles’, or ‘swift-footed Achilles’, or ‘swift-footed luminous Achilles’, and – by choosing the appropriate formula – reach the end of the line. He would then breathe, organize his thoughts and embark on the next line of poetry. This method of composition in performance helps to explain many features of Homeric epic, including the fact that formulae cluster towards the end of each hexameter line.26

38

Life of Texts

Now, in each instance, the poet chose a specific formula not because Achilles was, in the particular situation described, luminous rather than swift-footed or vice versa, but because he needed a formula with a specific metrical shape. Parry demonstrated the principle of ‘formulaic economy’ for both Homeric and SouthSlavic epic: in each tradition, formulaic systems developed in answer to particular metrical needs. Generally speaking, there was just one formula, or a narrow range of options, for each metrical need. Now, Parry and Lord did not thereby prove that the Iliad and the Odyssey are ‘oral poems’ (after all, what we have are written texts), but rather demonstrated that Homeric epic stemmed from a long tradition of oral composition, and re-composition, in performance. Within that tradition, what was the invention of one singer would be received and modified by the next, so if we ask where composition ends and reception begins it is clear that already the earliest text of the Iliad we can possibly reconstruct receives an earlier tradition of epic songs about Achilles. Here is an example. Achilles is ‘swift-footed’, according to traditional formulae – even though for most of the Iliad he refuses to move. Modern readers, used to poets who choose the uniquely appropriate turn of phrase for the situation at hand, often find the repetitiveness of Homeric epic disappointing: Parry himself suggested that traditional formulae like ‘swift-footed’ meant nothing much, did not attract the attention of ancient audiences, and were perhaps best left untranslated in modern versions of the Iliad. These conclusions have since been exposed as problematic: formulae may not draw attention to themselves, or fit a specific situation, but they do express traditional expectations – of Achilles, for example.27 So, when our hero refuses to move, he is falling short of his very name or, more precisely, of the way he is repeatedly called in the epic tradition. There is more: when Achilles finally gets up and runs towards the end of the Iliad, the poet draws attention to the fact that he is finally behaving as the formulae that describe him suggest he should. And yet, precisely then, Apollo chooses to remind Achilles of his mortal insignificance. He disguises himself as a Trojan, induces Achilles to run after him, and then reveals that the task is futile, because the hero will never catch up with a god, despite his ‘swiftness of foot’.28 We hear, in this, how the poet elaborates creatively on inherited formulas and expectations: Apollo himself comments on their limitations. This kind of creative reworking of traditional material is typical of both Homeric and South-Slavic epic: we thus know that it is independent of literate techniques of composition. For this and other reasons, the role of writing in the composition of the Homeric poems remains an open question. We know that use of the alphabet spread at about the same time as the Iliad and the Odyssey

Editing Homer

39

took shape, but it is hard to determine the significance of this fact. Parry and Lord believed that an oral poet immediately realized the potential of this new technology and dictated the poems to a scribe: this may be so, but one consideration suggests caution. Parry and Lord were clearly modelling this ancient scenario on their own experience in Bosnia – where illiterate bards were asked by Harvard-educated ‘scribes’ to perform long epics, in order to match a Homeric prototype. The situation in ancient Greece must have been very different. Length of composition and complexity of structure did not just arise because a new technology enabled them – the poems must have been created to serve, and be appreciated by, ancient listeners. If no literature at all survived from ancient Greece, we certainly would not expect it to start with two monumental poems, one about the anger of Achilles, the other one about the return of Odysseus, set in the distant past when heroes could throw boulders that ‘two men could not lift, such as they are nowadays’.29 We would rather assume that literature began with shorter compositions destined for specific occasions (for example, wedding songs and funeral laments), and answering specific purposes such as courtship, political propaganda, martial exhortation, dinner entertainment or mourning. We know that those kinds of compositions did exist, and indeed the Iliad and the Odyssey make several references to them. The Homeric epics, however, are considerably longer and more complex than the performances described in the Odyssey, where singers offer extemporized epic entertainment over dinner. They also fail to resemble the performances recorded in Bosnian coffee houses, which were generally much shorter. It is true that, upon request, the guslari could perform epics of roughly the same length as the Iliad and the Odyssey, but their structure was simpler. So, although long oral compositions such as The Wedding Song of Meho Smailagić prove that poems of similar length to the Iliad and Odyssey could be composed without the aid of writing, we still know very little about the social and artistic motivations for the composition of the two monumental Homeric epics. After all, in ancient Greece, poets were not composing long epics in order to satisfy the requests of Harvard-educated scholars. In short, to make progress, we need to investigate the historical context in which Homeric epic developed. First, and at a very general level, we can note that the Homeric poems show awareness of material circumstances not found before the later eighth or early seventh century bc , such as temples and cult statues, narrative art, and knowledge of the world extending from Thrace to Phoenicia and Egypt. This gives a terminus post quem: the poems cannot have been composed much before 700 bc . Their influence on visual art (primarily vase

40

Life of Texts

paintings), the first quotations, the first mentions of the author Homer become visible to us in the late sixth century bc . This gives us a terminus ante quem. We can safely assume that the most important period for the composition and written fixation of the Homeric texts falls between the end of the eighth century and the end of the sixth century bc . Moreover, given that we know of a thinker from southern Italy, Theagenes of Rhegium, who was writing about Homer in the late sixth century bc , we can safely assume that texts of Homer also existed by then.30 For all that written versions of the Iliad and the Odyssey will have circulated in archaic and classical Greece, the fact is that nobody seems to have cared about them – in the same way as, today, people care little for musical scores. Listening, rather than reading, was what mattered for Homeric poetry in ancient Greece and for music today. The Iliad is more than 15,000 lines long, and it is possible to calculate that it must have taken approximately three full days (or nights) to perform from start to finish.31 The Odyssey is almost as long. The structure of the poems, moreover, is carefully articulated – which suggests that these two monumental epics must have been intended for re-performance. The effort required to compose and indeed receive them was not justifiable in terms of a one-off experience. The whole operation must have required commitment and organization – that is to say, some form of institutional support in order to establish, for example, dates for performance, breaks in recitation, and adequate food supplies and facilities for audiences. The earliest known context for the performance of the Homeric poems are city festivals in honour of the gods – most prominently, but by no means exclusively, the Great Panathenaea celebrated every four years in Athens in honour of the patron goddess of the city. According to fourth-century sources, the tyrant Pisistratus, or one of his sons, passed a decree in the sixth century bc , according to which ‘Homer and Homer only’ was to be performed at the Great Panathenaea.32 Even later sources mention a ‘Panathenaic recension’ – that is to say, an official text supposedly established by Pisistratus for approved state recitations.33 The nature and even existence of such a recension remains a controversial subject in Homeric studies, also because some of the ancient sources imagine a process akin to what then actually took place, in the third and second centuries, in the library of Alexandria. Here is an example of this tendency to retro-date ancient editions of Homer. An anonymous commentator on Dionysius Thrax, a pupil of Aristarchus of the second century bc , claims that the first critical edition of Homer was produced in Athens, under Pisistratus. The tyrant, we are told, engaged the services of

Editing Homer

41

seventy-two ‘philologists’ (grammatikoi), each of whom was supposed to work independently and produce his own text. Since each was to be paid by number of lines edited, some slipped into the text a few extra verses to make more money. In later times, the extra lines were identified as spurious, but by then the damage was done: The verses had already won the place of familiarity and habit among readers of Homer. None of this escaped the critics, and so, respecting the force of custom and mental habit, they let these lines stand. But to each of the lines that were suspect, foreign, or unworthy of the poet, they affixed obeloi, to express their judgement that this was the case.34

This is an ancient story of philological one-upmanship: what the early philologists working under Pisistratus corrupted the later philologists working in Alexandria corrected and marked up with the obelos.35 There is no historical basis for this story: these seventy-two philologists working under Pisistratus are an invention (a fiction, in fact, which resembles a story about the first Greek translation of the Bible).36 My point, here, is that this is an invention concerned with the promises of philology – and with its limitations too. Not even the superior Alexandrian philologists could, in fact, free the text of additions because, by then, they had to contend with the expectations of readers. These two points about philological one-upmanship and negotiation with readerly habits remain relevant when it comes to the latest editions of Homer.37

Editing Two major editions of the Iliad appeared at the end of the twentieth century: Helmut van Thiel’s for Olms, and Martin West’s for Teubner.38 Van Thiel had previously edited the Odyssey,39 and West embarked on the Odyssey after completing his Iliad: his edition appeared posthumously too late for proper consideration in this chapter, though an initial consultation suggests that he changed his mind on some of his previous editorial decisions.40 For the sake of simplicity, I make a few observations here on the two editions of the Iliad, comparing aims, methods and results.41 In general, Helmut van Thiel trusts the direct transmission, i.e. the best medieval manuscripts, often called ‘the vulgate’. He takes the position that ancient variants reported in the Homeric scholia are usually ‘suggestions’ of ancient scholars ‘towards the improvement of the text, or . . . deliberations about it’, and that they are therefore of little significance when constituting the text. He also insists that

42

Life of Texts

modern editors should not indulge in conjectures of their own – because the transmitted text is basically as good a text of Homer as we can get. What the editor should do, rather, is represent the vulgate as faithfully and intelligently as possible. This, he concedes, is a modest aim. According to van Thiel, ‘laurels in textual criticism are not to be won from the text of Homer.’42 Martin West took a different view: he insisted on trying to restore what he thought was the original wording of Homer (or rather of ‘the poet of the Iliad’ because he also argued that that poet was not, in fact, called ‘Homer’).43 In order to pursue that goal, he made use of weakly attested ancient variants reported in the scholia and, above all, employed his own critical acumen in order to weed out what he considered to be corruptions and modernizations. The intention to restore the earliest possible text of the Iliad led to radical interventions, such as the placement of the entire Book 10 between square brackets, singling it out as spurious. This in turn has had the consequence that Book 10 is altogether omitted in an influential recent translation of the Iliad by Stephen Mitchell.44 Here already we see how philological ambition can come to clash with readers’ expectations. Book 10 seems indeed to be a relatively late addition to the Iliad, as ancient scholars already suspected. Nevertheless, it belongs to the poem at the very least in the sense that it fits nowhere else: it was created to fit between Book 9 and Book 11, adding an Odyssean night-time adventure to the plot.45 Unsurprisingly, the editions by van Thiel and West have sparked a lively debate.46 One important contributor is Gregory Nagy who, elaborating and modifying the theories of his Harvard predecessor Parry, argues that the Homeric text evolved over a long period of time, from a stage of relative fluidity in the Dark Age to one of relative stability in the Hellenistic period. He consequently advocates an inclusive approach to all variant readings, which he regards as equally valid realizations of what he calls a ‘developing multitext’.47 Three main issues seem to me to emerge from the current debate on how to edit Homer. One is the status of variants – that is to say, alternative readings to those best attested in the manuscript tradition. The second concerns what might count as ‘innovation’. The third relates to the issue of ‘corruption’ of the text – that is to say, possible errors that crept into the tradition and that need correcting on the part of the editor. On the issue of variants, the evidence is scant and difficult to assess. Quotations of Homer in classical authors display some divergences from the vulgate that go beyond single words – though, it must be said, not much beyond single words.48 The problem with that kind of evidence, however, is that manuscripts of ancient authors quoting Homer may, in the course of their transmission, have been

Editing Homer

43

corrected and made to conform with, say, Hellenistic rather than archaic versions of Homeric epic. Early papyri are, also for this reason, more interesting: they differ from the medieval manuscripts in some of their readings (‘horizontal variation’) and in the number of lines (‘vertical variation’) as already demonstrated in the papyrus reproduced above (Fig.  1.1), where line 200 is one line out, compared to the medieval vulgate. The papyri also show that the work done by Alexandrian scholars in the library did have an effect on the stability of the text: the early Ptolemaic papyri are ‘wild’ in comparison with later ones. Still, even the so-called ‘wild papyri’ are not as wild as all that: variations tend to involve single words or at most lines.49 We do not have evidence of a rich and varied multitextual tradition on the scale of, say, the Indian Maha¯bha¯rata, or the different versions of the Babylonian Epic of Gilgamesh which have reached us, or even of what we can reconstruct of the archaic Greek Epic Cycle.50 What we have, in the Hellenistic period, is an essentially stable Homer. The situation before the work done in Alexandria is much harder to assess. There is no evidence about large-scale variation, as already stated; and, even at the level of small-scale variation, it is hard to decide on the nature and status of the variants reported in the scholia and found in the papyri. Do they represent just a small proportion of the variants that were available to the Alexandrian scholars? Are they Alexandrian attempts to improve the text, or are they genuine earlier variants? Sometimes we are told that a variant reading was found in an ancient manuscript, but often we are not, and it is certainly possible that some Alexandrian scholars offered conjectures of their own. In order to address these questions, I take as my sample Iliad 6, for the contingent reason that Johannes Haubold and I have worked extensively on it and have drawn the following conclusion: almost all the variants in that book seem to cater to Hellenistic taste – either they seek to explain the text, or make Homeric language more context-specific, or address perceived lapses of decorum. Of the twenty variants that we have – and this is in any case a rather low number in 529 lines – only two seem to reflect early variation in performance, and they concern just two words.51 The conclusion that we do not have a multitext for this book seems inescapable. Whether variants were conjectured, that is to say made up by the Alexandrian scholars themselves, or found in some earlier versions of the poems, they are preserved only because they helped the editors in Alexandria to deal with that they saw as shortcomings in the Homeric poems. With admirable restraint, they recorded or made up a few alternatives which, in their view, could potentially represent improvements. They did not, however, simply change what must have been, already then, a rather stable text.

44

Life of Texts

In light of this, preferring weakly attested variants in order to improve the text, as West sometimes does, seems problematic. Equally problematic is the assumption that variants represent a vast multiform tradition, as Nagy and his students maintain. It is important to distinguish here a tradition of oral poetry in the hexameter, which we know existed, and multiple, various and vastly different versions of the Iliad and the Odyssey – that is to say, epics about the anger of Achilles or the return of Odysseus. One plausible model may be this: we have a flexible tradition of oral poetry that yields two monumental epics – but as far as those two epics are concerned, we do not have evidence of large-scale variation. We have, in fact, various snippets of evidence from the sixth century onwards, which express anxiety at the possibility that the poems might be altered to suit local interests, for example.52 Given this emerging scenario, it is possible to return to the question of where composition of the Homeric poems ends and where preservation and explanation of their meaning begins. I already pointed out that even the earliest texts we can possibly reconstruct are shaped by attempts to explain and comment on an earlier poetic tradition: the Iliad, as I indicated, riffs on the formula that characterizes its protagonist, ‘swift-footed’ Achilles. Sometimes, the Homeric habit of explaining or commenting on words on the go, in the course of performance, has direct implications for editors of the Iliad. I offer just two examples, drawn from previous work.53 First, a simple case: at Iliad 4.528, we find a strange word for lung, pneumōn, a sort of pun on pneuma, breath, rather than the etymologically correct pleumōn. West prints the ‘correct’ form, but he has almost no textual support for his choice – in fact, only one (learned) papyrus, and two Byzantine scholars; i.e. people who, like West, were determined to correct and improve the Homeric text. So, one straightforward question is whether West is right, or whether editors should follow the ancient transmission and print pneumōn, as van Thiel and others do. Grammatical correctness is an obvious criterion for judging the Homeric text, but the crucial issue here (and an issue which West never confronts directly) concerns the text’s own criteria of correctness – or, to put it differently, what early audiences may have considered acceptable in terms of morphological formation. Were their criteria for what is correct the same as ours? That would be surprising. Assuming that pneumōn is indeed a corruption, or rather an etymologizing, punning version of pleumōn, it is still necessary to decide whether it could and should feature in the Iliad. West rejects this possibility, but, as a matter of fact, there are plenty of other pseudo-etymologies and word-games in Homer – in fact, several scholarly monographs list and explain them in detail.54 So, on the

Editing Homer

45

principle ‘clarify Homer with reference to Homer’, this ‘incorrect’ form should stand, in my view. Here is a second, more interesting example. There are two expressions in Homer which look similar but, according to the medieval manuscripts, were written and (more to the point) pronounced differently: arēiphilos and Dii philos, ‘dear-to-Ares’ spelled as one word, and ‘dear to Zeus’ spelled as two words. West objects to the apparent inconsistency, and writes both expressions as single words. This, however, seems problematic not only because of the preponderant spellings in the tradition, but also because the expressions actually seem to mean different things. The epithet arēiphilos is used in Homer primarily of Menelaos and the Achaeans. Neither is particularly ‘dear to Ares’, who of course fights on the Trojan side. The expression, then, has little narrative force in the poem, and in fact serves as a metrically useful alternative for the common epithet ἀρήϊος, arēios, ‘warlike’ (again most commonly used of Menelaos and the Achaeans). The manuscript spelling arēiphilos, as one word, thus rightly treats philos ‘dear’ as quasi enclitic, soft-pedalling any suggestion of personal affection on the part of the god Ares. The situation is quite different for Dii philos, ‘dear to Zeus’. This expression is used of people who are actually dear to Zeus, primarily Achilles and Hector. It hardly ever occurs in the plural, because Zeus’s affections tend to focus on individuals. There is only one exception, and it confirms the rule: heralds as a group are under Zeus’s special care and therefore ‘dear to Zeus’. The expression Dii philos has clear thematic resonance in the Iliad – it actually means ‘dear to Zeus’ – whereas arēiphilos means not ‘dear to Ares’, but rather ‘warlike’. The different spellings of the manuscript tradition preserve knowledge about how these words were uttered in performance – and, as every actor knows, pronunciation is the beginning of interpretation. So here the manuscripts, and van Thiel, preserve an oral interpretation of the text, an interpretation which West sacrifices in the name of morphological consistency. Much more work on examples of this kind still needs to be done. The point I want to make here is simply that, in terms of aim, such work represents a departure from the philological one-upmanship described above, since it involves understanding, rather than correcting, how Homeric epic was shaped and received in antiquity. Here is one final example, which takes its cue from the ancient ‘solution from character’. An interesting observation – though not one that has, to my knowledge, been made by others – is that most ancient attempts to correct the text of Homer concern character speech.55 The ipsissima verba, the words uttered by the poet himself in his own voice, are, by and large, left alone. Within character speech, moreover, the highest percentage of ancient

46

Life of Texts

attempts to correct the text concerns the words uttered by female characters. Clearly, these attracted extensive debate about appropriateness and correctness. What emerges from this kind of work is not a new edition of Homer (and, in my view, there is little need for one), but a better understanding of ancient ideas about authority – as reflected in both the composition and editing of Homeric epic.

Conclusion The first general conclusion that can, I think, be drawn on the basis of the evidence and observations presented here is that the composition and reception of Homeric epic are intertwined. Explanations of inherited epic words are already contained in the poems themselves: they affect composition, in other words, and not just the ways in which the poems were explained by later scholars. Similar arguments can be made about other aspects of the transmitted texts. The evidence does not, in my view, support a model of multitextuality on a grand scale. What it does suggest is live interpretation in performance, starting from pronunciation and extending to small-scale textual variation. Editors who fail to engage with ancient reception – by asking, for example, what might or might not have sounded grammatical to ancient audiences – miss important evidence for the constitution of the text. Conversely, however, students of reception cannot simply take the latest edition of the Iliad for granted: in fact, they would be led astray if they did so. Although textual critics and students of reception seldom talk to each other, in actual fact reception and transmission must necessarily be considered together.56 Likewise, the position of Homer between oral epic and canonical literature needs to be characterized carefully.57 It seems wrong-headed to try to separate the ‘original contribution’ of one single genius, ‘the poet of the Iliad’, thus discarding or correcting whole sections and even books. It also seems problematic to insist, with equal dogmatic vehemence, that anything goes, in a multiform tradition. Both these approaches are short-cuts: they privilege modern theory over ancient practice. The Iliad and Odyssey have been the focus of intense interest for almost three millennia: wellattested oddities in medieval manuscripts point, quite often, to deep study, rather than straightforward error. Just as it makes little sense to draw a line between composition and reception according to strict principles (such as that of reconstructing the most consistent or earliest possible version of the poems), so it seems problematic to refuse to draw any line at all, again on principle.

Editing Homer

47

As I suggested, not all variants are born equal. It makes sense, therefore, to think carefully and flexibly about that line. If we do so, the result is unlikely to be a radically different text of Homer. What we may achieve, rather, is a better understanding of the specific processes through which the Homeric poems were created, interpreted and transmitted through the centuries. This kind of work reveals a close and productive cooperation – across vast distances in space and time – between poets, performers, audiences, scholars and readers.

2

The Canon and the Codex: On the Material Form of the Christian Bible Francis Watson

‘All scripture is inspired by God and useful for teaching, for reproof, for correction, and for training in righteousness . . .’ (2 Tim. 3.16). This text has featured prominently in modern conservative attempts to formulate an appropriate doctrine of scripture – that is, an account of the role of the canonical texts within the divine economy. The decision to highlight this particular text reflects the assumption that the concept of inspiration lies at the heart of any doctrine of scripture. The text also hints at a pragmatic dimension: the inspiredness of the holy scriptures comes to light as and when they prove fruitful in shaping authentic Christian living. If God is the inspirer of scripture, that does not make God scripture’s author. With the sole exception of the stone tablets inscribed by the divine finger, the biblical God never writes anything. Writing is a human activity, dependent not only on a writer’s skilful manipulation of shared communicative conventions but also on the materiality of pen, ink and the book. There is no reason to suppose that the spirituality and materiality of the biblical text are in tension with one another. Yet it is prima facie likely that the use of the inspired text will be determined in part by its physical format, as roll or codex. That at least is the working hypothesis of this chapter, which takes its cue from another text in 2 Timothy: When you come, bring the cloak I left with Carpus at Troas, and also the books, especially the parchments. 2 Tim. 4.13

Paul and his Books Towards the end of his second letter, Paul (the fictive author) appeals to Timothy to visit him as soon as possible, preferably before winter makes travelling difficult 48

The Canon and the Codex

49

(2 Tim. 4.9, 21).1 Paul himself is in Rome and in chains, nearing the end of his life: ‘The time of my departure has come’ (4.6). Apart from Luke, his closest friends and fellow-workers are all elsewhere (4.9–12, 20). Local Christians have failed to support him, although he passes on their greetings (4.16, 21). Timothy may still be in Ephesus. In the previous letter he was instructed to remain there in order to combat the ‘myths and endless genealogies’ promoted by the teachers of a ‘knowledge falsely so-called’ (1 Tim. 1.3; 6.20). But now he is to travel to Rome to provide the lonely apostle with companionship during his last days. En route he is asked to collect Paul’s cloak and books from Carpus in Troas (2 Tim. 4.13). A reference follows to one ‘Alexander the coppersmith’ who, it is said ‘fiercely opposed our message’ and ‘did me great harm’ (4.14–15). Perhaps we are to imagine that Alexander’s hostility had forced Paul to leave Troas in a hurry, with no time even to collect his precious cloak and books from his lodgings. Now he wants them back. In revisiting Troas to collect them, Timothy should himself beware of the malevolent coppersmith – a dangerous opponent who recalls the figure of Demetrius the Ephesian silversmith, whose business was undermined by the Christian gospel and who mobilized public opinion in defence of the goddess Artemis (Acts 19.23–41). Timothy is to collect ‘the books [ta biblia], especially the parchments [tas membranas]’. Evidently the parchments form a subset within the wider category of the book; these are the texts that Paul is most anxious to retrieve and that Timothy should prioritize as he attempts to transport the entire library from Troas to Rome. How are these ‘parchments’ differentiated from the other books, and why are they so uniquely valuable?2 We might suppose it is their physical composition that differentiates these membranai from the other biblia. The membranai have been manufactured out of animal hides, in distinction perhaps from books written on papyrus (chartēs). The anonymous author of the Johannine epistles speaks of writing on papyrus with ink and pen (2 Jn. 12; 3 Jn. 13). It may be that Paul values his parchment books still more than his papyrus ones. This seems unlikely, however. Parchment is not inherently more valuable than papyrus. In economic terms it has the advantage that it can be manufactured locally and does not need to be imported from Egypt. If parchment seems to us to speak of a medium used only for precious and revered texts, that was not the case in the ancient world. A contrast between two writing media – parchment and papyrus – would be out of place here. An explanation for Paul’s distinction must be sought elsewhere. A fifth-century commentator on this passage, Theodoret of Cyrrhus, notes that the term membrana is a Latin loanword and suggests that the apostle uses it here to refer to scriptural rolls:

50

Life of Texts He calls rolls membranai, for that is the term the Romans use for ‘skins’. In earlier times the divine scripture used to be contained in rolls [heilēta], as indeed they still are among the Jews.3

On Theodoret’s interpretation, Paul is especially anxious to retrieve his ‘parchments’ from Carpus of Troas because these are his copies of scriptural texts, formatted as rolls in the traditional Jewish manner. For Theodoret’s readers, scriptural texts are familiar in the codex format still preserved in the modern printed book. Jews and Christians share much of the same scripture, but in different formats: a copy of the Book of Isaiah intended for Christian use will normally be easy to differentiate from its Jewish counterpart, for one will be a codex, the other a roll.4 Theodoret is aware that the roll format is the ancient one and that the Christian scriptural codex is an innovation, and he therefore assumes that Paul’s highly valued membranai are scriptural rolls that predate the early Christian shift to the codex. If the membranai are rolls, the non-scriptural biblia that Paul also wants back are presumably codices. In Theodoret’s time the codex format is in regular use for all kinds of literature, and there is no difficulty in supposing that the apostle possessed non-scriptural codices as well as traditional scriptural rolls. Theodoret’s explanation of the highly valued membranai fails to account for the use of a Latin loanword to refer to Jewish rolls. The significance of the Latin term is rightly noted by the papyrologists C. H. Roberts and T. C. Skeat in their groundbreaking book, The Birth of the Codex (1987).5 According to Roberts and Skeat, contemporary usage shows that membranai must refer not to Jewish scriptural rolls but to a parchment codex, that is, a book divided into separate sheets bound together at the fold. Paul’s membranai are codices rather than rolls as Theodoret supposed. Indeed, since the plural seems to refer to the separate pieces of skin that constitute the pages of a codex, it may be that Paul has in mind a single codex rather than codices in the plural. His other biblia, differentiated from his codex or codices, must be rolls. Paul is anxious to recover all of his books, but he is especially concerned about the codices – scriptural or otherwise – rather than the rolls. In favouring the codex above the roll, the Paul of the Pastoral Epistles is in line with broader trends of Christian book production during the early centuries. Of extant Greek Old Testament manuscripts thought to predate the fourth century, around forty-three derive from codices, seventeen from rolls.6 Of these seventeen rolls, ten are dated from prior to the Christian era and so are clearly Jewish; some or all of the later rolls may also have a Jewish origin. Christian scribal conventions confirm that almost all the forty-three codices are Christian.

The Canon and the Codex

51

Most seem to have contained a single scriptural book, with Genesis (six codices), Exodus (five), Psalms (eight) and Isaiah (four) proving the most popular. As for the books of the New Testament, there are thirty-nine partially extant second-or third-century codices, but no rolls. The four gospels account for twenty-two of the thirty-nine codices, with one text (P45) containing all four gospels and Acts, others containing Matthew and Luke (P64 + P67+ P4) or Luke and John (P75). Otherwise there are individual codices of Matthew (six), Luke (three) and John (ten). Mark is represented only in the context of a four gospel codex, but some of the surviving fragments of the Egerton Gospel and the (Greek) Gospels of Thomas, Peter, and Mary reflect a codex format. Especially striking is the existence of five codices of the Shepherd of Hermas – an indication of this text’s popularity and proto-canonical status, although other copies exist in roll form. This Christian preference for the codex over the roll may be compared and contrasted with wider trends in contemporary book production. Using information drawn largely from the Leuven Database of Ancient Books, Larry Hurtado has calculated that only 1.5 per cent of the 1044 surviving first-century manuscripts were codices, whereas 77.5 per cent were rolls.7 Indeed the figure for rolls was probably higher, as many of the remaining manuscripts are too fragmentary to determine their original format. Assuming that early Christians were already using codices in the first century, they did so in conscious defiance of the overwhelming cultural preference for the roll. The balance between rolls and codices shifts, however, in subsequent centuries: 74 per cent rolls, 5 per cent codices in the second century, 56 per cent rolls and 21 per cent codices in the third, 38 per cent rolls and 56 per cent codices in the fourth. The rise of Christianity to its fourth-century cultural dominance is coextensive with the rise of the codex. In seeking to explain the remarkable early Christian preference for the (as yet) unfashionable codex, it would be a mistake to over-emphasize its countercultural dimension. The codex was not a Christian invention. In some form or other it was already available to be exploited, and examples of its general use can help to establish a context for its adoption by Christians. Paul was not the only author who valued his membranai.

The Use of the Codex Writing in Rome towards the end of the first century, two Spanish authors have occasion to speak of the codex and its uses: Quintilian the rhetorician and Martial the satirist.

52

Life of Texts

In the tenth of the twelve books of his Institutio Oratoria, Quintilian devotes a chapter to the role of writing in relation to public speaking in the law courts or elsewhere. The public discourse will not itself be read from a text but performed as though extempore, yet careful preparation is needed if the speech is to be clear, coherent and persuasive. Among other matters, Quintilian considers what is the best medium for this preliminary writing: There are certain lesser points – though in study nothing is unimportant – which should not be overlooked. It is best to write on wax tablets [ceris], which make deletion easier, unless indeed poor eyesight makes it necessary to use parchments [membranarum]. While these aid vision, they delay the hand by the constant to and fro of pen to inkwell, breaking up the continuity of thought. Whichever medium is used, however, some pages [tabellae] should be left blank so that additional material can be freely added. For lack of space makes one reluctant to make corrections, and will certainly produce confusion in the old when new material is inserted.8

The cerae are thin rectangles of wood with a margin enclosing a wax-filled hollow in which one writes with a stylus. Letters become visible as the darker wooden background is exposed beneath the lighter wax.9 For better visibility, parchment, pen and ink are also available. The stylus has the advantage over the pen that it allows for continuous writing without having to be constantly recharged with ink; allusion is made here to the technical problem to which the fountain pen was a belated and temporary solution. Wax has the further advantage over ink on parchment that it allows for instant erasure. The ‘pages’ or tabellae are primarily the wax-filled ‘boards’, although the term also covers the analogous pieces of parchment. Quintilian’s recommendation that pages be left blank indicates that the wood-andwax artefact and its parchment equivalent are purchased in made-up rather than loose-leaf form, with a specific number of leaves securely bound together. These objects are notebooks rather than individual writing tablets. New pages cannot easily be inserted to accommodate new matter, so recto pages must be set aside for future additions to their verso opposite numbers. The fact that the key terminology is all plural – cerae, membranae, tabellae – indicates that the singularity and continuity of the roll is the norm and that writing on separate pieces of material is simply a matter of convenience. The key practical advantage of this notebook format is that one always has immediate access to every part of one’s drafts or notes, without having to scroll laboriously from one section to another. In the anecdote that follows, the codex notebook is found in a classroom setting:

The Canon and the Codex

53

I recommend that the wax tablets [ceras] should not be unduly wide, having once had an otherwise able student who wrote at inordinate length because he measured his compositions by counting the lines. Repeated admonitions failed to correct this fault, and it was only put right by replacing his writing tablets [mutatis codicibus].10

A codex or caudex is initially a block of wood, and the term is retained when the block is sawn into tablets, which are then hollowed out to receive the wax. Once again Quintilian uses a plural term (codices), referring to the notebook by way of its leaves or pages. Since tabellae, ‘boards’, can be extended to refer to the pages of a parchment notebook, the same may also be true of codices. Evidently these bound codices are easy to obtain and may be purchased in a variety of sizes and formats. Their use is varied and flexible: they are essential equipment both for the student in the classroom and for the professional politician or lawyer as he prepares his public discourses. In their parchment form, they also have the advantage of closer resemblance to a standard book: one does not unroll them, but they are written in ink on the same material as might be found in a roll. A text inscribed in wood-and-wax tablets would hardly be saleable, but the reading public might be willing to purchase a text in parchment codex format. That at least was the view of Quintilian’s contemporary, the poet Martial, who marketed his own and others’ literary works in codex editions. Introducing his books of epigrams, Martial claims that their codex format makes them more portable than conventional rolls: Qui tecum cupis esse meos ubicumque libellos et comites longae quaeris habere uiae, hos eme, quos artat breuibus membrana tabellis: scrinia da magnis, me manus una capit.11 (If you want my books to be with you everywhere, / And are looking for companions for a long journey, / Buy these, confined within small parchment pages. / Keep your book-boxes for large volumes – a single hand can hold me!)

A parchment or papyrus codex is written on both sides of the page, whereas a roll is written on one side only. A roll must be held in both hands, a codex may be held in just one. Martial claims that his miniature codices make more text available with less effort. He also provides the name and address of the bookseller where they may be obtained. It may have been the same bookseller who published parchment codex editions of classics such as Homer, Cicero, Livy, Virgil and Ovid, for each of which Martial provided an introductory epigram. On Livy he writes:

54

Life of Texts Pellibus exiguis artatur Livius ingens, quem mea non totum bibliotheca capit.12 (Into scant parchment pages great Livy is crammed, / whom, entire, my library could not contain.)

Martial’s Livy edition is perhaps an abridgement or a selection. What is important for our purposes is that the late first-century codex can encompass the entire range from the school exercise-book to editions of literary classics. It is used primarily as a notebook, but it is also possible to publish literary texts in this format when the medium is parchment or papyrus rather than wood and wax. This versatility of the codex may prove to be an important factor in its enthusiastic adoption by Christians. Tracing the Christian use of the codex back to its origins is a necessary preliminary to the more important task of assessing its significance for Christian canonical hermeneutics.

Writing the Gospel Tradition By the end of the first century, it was possible to publish literary texts in codex format. If Christian gospel texts were initially published as codices, that would not itself have been an anomaly. What is anomalous is the exclusive use of the codex. As Harry Gamble notes: [T]here must have been a decisive, precedent-setting development in the publication and circulation of early Christian literature that rapidly established the codex in Christian use, and it is likely that this development had to do with the religious authority accorded to whatever Christian document(s) first came to be known in codex form.13

The problem then is to identify the earliest Christian use of the codex and to explain why it set such a decisive precedent for later Christian text-production. The Gospel of Mark and the Pauline letter collection have each been proposed as possible candidates for this precedent-setting role.14 Yet there is no reason to ascribe the original use of the codex to a specific part of the New Testament, as though there had been no Christian writing before Mark or Paul’s editor. There is good reason to suppose that, long before Mark, collections of Jesus’ sayings were in use as aids to preaching and teaching.15 These should not be identified with the hypothetical ‘Q’ document. The Q hypothesis rightly finds it implausible that Jesus’ sayings were transmitted in purely oral form over four or

The Canon and the Codex

55

more decades, but it is dependent on the questionable assumption that Luke’s Gospel cannot reflect any awareness of Matthew’s. If Luke may plausibly be shown to have drawn, selectively and critically, from Matthew as well as Mark, then Q in its conventional form simply vanishes.16 Evidence for the existence of early sayings collections is to be found elsewhere. While the Gospel of Thomas shows an awareness of passages from Matthew and Luke, its non-narrative format is entirely independent of the canonical narrative gospels.17 Most sayings are introduced and differentiated by the simple formula, ‘Jesus said . . .’ They may be linked by catchwords, but there is little or no sequential development. This text is best seen not as a unique and isolated phenomenon but as a relatively late and theologically developed descendant of primitive Christian collections of Jesus’ sayings. The canonical evangelists will also have drawn from such sayings collections, accommodating their contents to their own continuous narratives. Sayings collections must also underlie texts such as 2 Clement, where Jesus’ sayings occur in distinctive forms and combinations: The Lord said, If you are gathered to my breast but do not obey my commandments, I will throw you out and say to you, Depart from me, I do not know where you are from, you workers of iniquity.18

The saying is obviously related to canonical material in Matthew and Luke, but it is nevertheless quite distinct from it. The introductory formula, ‘the Lord said . . .’, closely resembles the ‘Jesus said’ of the Gospel of Thomas, suggesting that in both cases the ultimate source is an early sayings collection. These collections will have been anonymous, a tradition preserved in the canonical gospels. If they bore a title at all, it may simply have been ‘Gospel’. Thus the author of 2 Clement introduces another of his non-canonical citations with the formula, ‘The Lord says in the gospel . . .’ (8.5). If primitive sayings collections were headed ‘Gospel’, that would account for the opening of Mark: ‘The beginning of the gospel of Jesus Christ’ (Mk. 1.1). While euaggelion is not used at all by Luke and John, it is nevertheless firmly established as the generic term for all written tradition relating to Jesus. If Mark and other gospels incorporate the contents of primitive sayings collections, this suggests a solution to the problem of the Christian preference for the codex. As we have seen, the parchment or papyrus codex of the late first century is beginning to display its potential versatility. It can be used to accommodate literary texts, including new ones such as Martial’s epigram collections – a striking extension of its previous role as a notebook for various forms of personal use. It seems probable that the early Christian adoption of the

56

Life of Texts

codex displays a comparable range, encompassing both literary texts such as full narrative gospels and their simpler antecedents in written collections of Jesus’ sayings. The gospel itself is articulated in Jesus’ words, and these are preserved as gospel precisely as they are written down. A notebook containing Jesus’ sayings could thus become a hallowed object whose codex format was preserved as its contents were assumed into more ambitious literary structures.

Time, Space and the Codex If this account is correct, the Christian codex established itself not because it promoted any specific interpretative aims but because the sanctity of the earliest written tradition was extended to its format. And so the convention arose that gospel writing required the codex – a convention rapidly extended to the ancient scriptures and to other Christian texts. Codex format does, however, possess potential hermeneutical significance, even if that was not the reason for its adoption. A codex still allows one to read its contents from beginning to end, just as one would read a roll. Neither format has any advantage over the other for ‘normal’ linear reading practice. Yet the codex has considerable advantages for reference purposes, as every part of it is always directly accessible to the user.19 In the case of a roll, it is a laborious undertaking to return from near the end of a text to a point near the beginning, perhaps to check that one has remembered something correctly. A codex makes this easy. The relationship between two points in a codex is an immediate one; there is no need to work backwards or forwards through the intervening material. Linear or sequential reading is still straightforward, but one can now also read across the pages, from one point to another that may be distant yet relevant to the matter in hand. In the codex a text still unfolds through time, as with a roll. And yet the text is no longer a purely temporal entity, it is also a space, a container for the play of intertextual connections and significances. This potential for cross-referencing can most effectively be realized when multiple biblical books are included within a single codex. The oldest extant cross-referencing system is the one developed by Eusebius of Caesarea as a framework for his single-volume edition of the four canonical gospels – a framework so popular that it spread rapidly into gospel codices in Latin, Syriac, Coptic, Ethiopic and Armenian.20 Eusebius identifies ten different categories of interrelation between the gospels, ranging from material shared by all four to material unique to each of them. The categories give rise to ten ‘canons’, tables

The Canon and the Codex

57

listing enumerated parallel passages, located at or near the beginning of the volume. Corresponding enumeration is inserted in the margin of each gospel text; section numbers relate not to sense-units per se but to points where one category of interrelation gives way to another. One part of a Matthean passage may be shared with Mark and Luke (Eusebius’s second category, canon II ), another part with Luke alone (canon V), and the Matthean enumeration will change at the point where this relationship changes. Beneath each section number is placed a further number in red ink, indicating the canon table in which the section and its parallels in other gospels is to be found. Thus the Matthean summary of Jesus’ first preaching (‘Repent, for the kingdom of heaven is at hand’) is identified in Eusebius’s margin as section 20 and as belonging to canon VI . Turning to the canon tables at the front of the four gospel codex, we encounter lists of section numbers in parallel columns, one for each of the gospels in question, typically located within an arcade-like structure and separated by pillar. The page containing canon VI consists in four columns: two Matthean listings alternating with two Markan ones. Near the top of the first two columns, section 20 of Matthew is aligned with section 9 of Mark, to which we may now turn to find the parallel version of Jesus’ initial preaching: ‘The time is fulfilled and the kingdom of God is at hand – repent and believe the gospel.’ It is the codex format that makes this cross-referencing system possible. The Markan parallel is many pages distant from its Matthean counterpart, and yet it is immediately available to the reader of Matthew. A linear reading of a four gospel codex is still possible; one may very well read through Matthew, Mark, Luke and John in order and from beginning to an end. This indeed is the standard reading practice in which the book – whether codex or roll – unfolds its contents in temporal sequence. The necessity of a readingthrough-time is acknowledged in Eusebius’s sequential enumeration system, which locates every point in the text on a continuum between its beginning and its end. Yet the primary role of the canon tables is to promote an account of the four gospels as a single space for the play of their complex intertextuality. This point may be further emphasized by the use of architectural imagery: columns with their bases and capitals, the lower arcade and the upper arch that links them, a structure that unites the solid earth with the dome of heaven. To fulfil its intended function, a canon table not only guides the user from one text to its counterparts but is itself a symbolic representation of the canonical gospel construct in its intricate order and harmony. In these pages the user sees the four gospel structure simultaneously and as a whole. This architecture is functional but it is also a work of art. The space that enables this architecture is the space of the codex.

58

Life of Texts

The Scriptures and the Book In his Life of Constantine, Eusebius recounts how the Emperor Constantine commissioned him to prepare ‘fifty parchment copies . . . of the sacred writings that you know to be most urgently necessary for the provision and use of the church and its teaching’.21 The reference here is not to single-volume Bibles, as it often supposed, but to copies of scriptural texts that Eusebius is himself to select as most important for the new churches that Constantine proposes to build in Constantinople. Even in the fourth century, however, the whole of Christian scripture could be accommodated within a single codex, as Codex Vaticanus and Codex Sinaiticus demonstrate. In the one case, each double page spread contains six columns, in the other, eight, mimicking the appearance of a roll while demonstrating the codex’s capacity to absorb almost unlimited quantities of material. In these manuscripts, the plurality of the ‘holy scriptures’ anticipates the singularity of the book we now know as ‘the Bible’. It was the codex that made this singularity possible, long before the invention of printing.22 The question is how this capacious container of the sacred texts shapes the way they are read. Around the middle of the sixth century, the ageing Cassiodorus composed his Institutiones divinarum et saecularium litterarum.23 After a career as a minister at the court of Theodoric the Ostrogoth in Ravenna, Cassiodorus returned south to his country estate at Scyllacium where he founded a monastery noted for its well-stocked fish-ponds and its equally well-stocked library.24 Book one of the Institutions is in effect a systematized and annotated library catalogue, doubling as an educational programme. This library contains eight bookcases with books in Greek in the eighth and Latin works in the other seven.25 Distributed between these seven bookcases are nine codices containing the scriptures of the Old and New Testament.26 The first codex is devoted to the ‘Octateuch’, from Genesis to Ruth. Four books of Kings and two of Chronicles follow in the second, the Psalms in the third, works attributed to Solomon in the fourth: not only Proverbs, Ecclesiastes and the Song of Songs, but also the Wisdom of Solomon and Ecclesiasticus. The fifth codex, containing the prophets, was probably the largest of the nine, closely followed by the first. The sixth codex is devoted to texts that Cassiodorus describes as hagiographa, borrowing the term from Jerome although several of the texts it contains are outside the limits of Jerome’s canon: Tobit, Judith, two books of Maccabees, and 1 Esdras as well as the securely canonical Job, Esther and Ezra. Codices seven, eight and nine contain the four gospels, the apostolic letters, and Acts and Revelation,

The Canon and the Codex

59

respectively – the last is a surprising juxtaposition for which Cassiodorus finds a precedent in Augustine.27 Along with each of these scriptural codices, Cassiodorus has assembled a collection of commentaries, which he compares to Jacob’s Ladder, enabling their readers to scale the scriptural heights.28 The library also contains three further codices that Cassiodorus calls pandects, all-containing: these are copies of scripture in a single volume, one in Greek and therefore located in the eighth bookcase, and two in Latin. The first Latin volume is based on Jerome, especially his translation of the Hebrew Old Testament.29 Cassiodorus was involved in the preparation of this volume, which he sought to make as compact as possible by means of small handwriting. Its Old Testament is said to contain just twenty-two books, corresponding to the number of letters in the Hebrew alphabet; this figure results from Jerome’s restriction to the Hebrew canon and from counting pairs or groups of books as a single book. In contrast, the second Latin volume is described as the ‘larger codex written in clear lettering’ (codex grandior littera clariore conscripto).30 Its Old Testament translation is from the Septuagint and includes forty-four books. This codex grandior contained seventy scriptural books: forty-four in the Old Testament section, twenty-six in the New. The Greek pandect contained seventy-five books.31 The contents of these pandects may differ, but the total figures all have symbolic significance. The forty-nine books of the smaller Latin pandect recall the Jubilee year; the seventy books of the larger one are prefigured in the seventy palm trees in the oasis at Elim (Ex. 15.27); and the seventy-five books of the Greek pandect recall both the number of descendants of Jacob who find refuge in Egypt (Gen. 46.27) and Abraham’s age at the time of his call (Gen. 12.4).32 Alongside these three pandects, the nine codex edition of scripture should also be seen as a singular entity. It contains seventy-one books, the ones contained in Augustine’s canon although not in the same order. If the divine unity is added to seventy-one, we arrive at seventy-two, a number whose symbolic significance Cassiodorus asserts but does not explain.33 There are then four copies of scripture in this library, all different, three in single-volume format and the fourth distributed over nine coordinated codices. These all-encompassing codices are precursors of ‘the Bible’ as we know it, and the question is how they will tend to shape the way the canonical texts are read. In the first instance, a singular ‘Bible’ requires clear decisions about inclusion and exclusion. If holy scripture occurs in the form of multiple individual codices, no sharp differentiation is required from other ancient and edifying texts located perhaps in the same bookcase: we recall those five early copies of the Shepherd of

60

Life of Texts

Hermas, referred to above. Cassiodorus’s embarrassment at the differences between his pandects would not have been necessary if each of the books they contain came in its own codex. A singular Bible has a certain homogenizing effect on the texts it contains. In Cassiodorus’s description of his nine codices, there is a seamless transition from volume six (Old Testament hagiographa such as Job and Esther) to volume seven (the four gospels). The gospels are of course more significant than Job or Esther, yet these are essentially texts of the same kind. The status differential between scriptural texts is reduced. An overwhelming preference for Matthew or John over Mark may be maintained so long as those texts circulate as individual codices, but Mark will not be so easily overlooked once safely installed in a fourgospel or whole-Bible codex. Users of such volumes will experience a pressure to read and familiarize themselves with their entire contents, since a book only partially read represents an unfinished task. A singular Bible means that all scriptural texts are readily accessible. In the early eighth century, Cassiodorus’s pandects were to inspire the monastic scriptoria at Wearmouth-Jarrow to produce three pandects of their own.34 One was intended as a gift to the Pope. (This is now known as Codex Amiatinus, the oldest surviving complete single-volume Bible and the main exemplar for printed editions of the Latin Vulgate from the sixteenth century onwards.)35 The other two pandects were intended to be permanently available for consultation in the Northumbrian monastery’s twin churches of St Peter and St Paul, ‘so that’ (as an anonymous contemporary historian puts it) ‘everyone wishing to read some specific chapter from either Testament would easily find what they wanted’.36 A singular Bible also requires decisions about the order and structuring of the scriptural texts. This is the point that causes Cassiodorus the most difficulty as he contemplates his nine codex edition and his pandects. Should one follow Jerome’s preference for the normal Jewish arrangement, in which the law is followed by the ‘former’ and ‘latter’ prophets, that is, the historical books from Joshua to 2 Kings and the prophetic books from Isaiah to Malachi? The tendency of such an arrangement is to highlight the links between the prophets and Israel’s history, and so to historicize them. Or, following Augustine, should one gather all the historical or narrative books into a single sequence? Books such as Chronicles and Esther may be read differently if one assigns them to the category of history rather than edifying literature. Within the New Testament, locating Acts immediately before Revelation and not as a sequel to the gospels is likely to alter and perhaps downgrade its significance.

The Canon and the Codex

61

Finally, a singular Bible will require a degree of consistency in its format. In the exemplars he uses for his nine codices, Cassiodorus finds some manuscripts that present the text without any subdivisions, others that introduce numbered sections with headings. Cassiodorus’s preference is to gather these headings together and to insert them as a table of contents at the start of a book, providing them himself where they are lacking in his manuscripts.37 A single-volume Bible is subject to extensive editorial interventions of this kind. In Cassiodorus’s great pandects, the codex has developed out of all recognition from the simple notebooks that the fictive Paul of the Pastorals hopes shortly to recover. When the Pastoral Epistles were composed, almost all the scriptural texts that make up those pandects were already in circulation, and canonical or proto-canonical status was already ascribed to many of them. Yet there are profound differences between these texts considered as individual artefacts and as incorporated into an encyclopaedic collection, just as there are profound differences between a text formatted as a codex or a roll. ‘The Bible’ as we know it is the result not just of authorial activity, divinely inspired or otherwise, but also of evolving techniques of book production.

3

Wandering Nights: Shahrazād’s Mutations Daniel L. Newman

In 2010, a group calling themselves ‘Lawyers Without Shackles’ (Muh.āmūn bilā quyūd) filed a complaint with Egypt’s Prosecutor-general when a new edition of the Arabian Nights, known in Arabic as Alf Layla wa Layla (‘Thousand and One Nights’), was published, due to its allegedly ‘obscene’ content. This was not the first time these accusations were levelled, nor is it just in the East that the Nights have caused controversy. When Disney released its animated musical Aladdin (1992), the opening song had to be amended following protest about the racism in some of the lyrics, while the film was accused of pandering to Orientalist stereotypes of Arab-Muslim society. Controversy has beset the Nights for several centuries, as its genesis and early development are still, for the most part, shrouded in mystery. The story is one of manuscripts – real and imaginary, forged and reconstructed – and contains as many, if not more, extraordinary twists and turns as the tales themselves.

A Traveller from the East . . . In 1766, an elderly scribe from Aleppo by the name of H·annā Diyāb decided to chronicle his extraordinary adventures during a journey to the capital of the ‘land of France’ (bilād Faransā), where he met men of great learning: An old man would often visit us. He was in charge of the library of Arabic books. He could read Arabic well, and translated Arabic books into French. At that time, he was translating the book of stories of One Thousand and One Nights. This man sought my help for some things he did not understand, and I explained them to him. There were a number of Nights missing and I told him the stories that I knew. As a result, he was able to complete his book, and was extremely happy with me.1 62

Wandering Nights

63

His account is corroborated by the journal kept by the ‘old man’, who was none other than Antoine Galland (1646–1715). The diary entries for 1709 – the year he was appointed to the chair of Arabic at the Collège de France – make several references to the young visitor from the East he met at the house of fellow Orientalist and traveller Paul Lucas. The first dates to Sunday 17 March (Fig. 3.1): J’allai le matin chez M. Paul Lucas lui reporter les medailles qu’il m’avait confiées 8 jours auparavant: je m’entretins quelque tems avec Hanna Maronite d’Halep: qui outre sa langue qui est l’Arabe, parloit turc, et Provençal, et françois assez passablement.2 (This morning, I visited Mr Paul Lucas to return medals he had given me a week ago; I spoke for a while with Hanna, a Maronite from Aleppo, who, besides his mother tongue Arabic, is able to speak Turkish, Provençal, and French reasonably well.)

One week later (Monday 25 March), ‘M. Hanna’ narrated ‘some very nice Arab tales, which he promised to put down in writing.’3 Unfortunately, none of Diyāb’s transcripts has survived, but Galland took notes during his meetings, and included summaries of some of the tales in his Journal. Diyāb provided a total of sixteen stories, ten of which made it into Galland’s translation, among them such emblematic tales as ‘Aladdin and the Magic Lamp’ and ‘Ali Baba and the Forty Thieves’.4 When Galland met H·annā, eight (of the eventual twelve) volumes of Les mille et une Nuit (sic) had already been completed. The Syrian informant’s tales were included in volumes 9–12 (1712–17), in the order in which they were narrated, but without, however, any mention of the source.5 As no Arabic original has ever been found of the Diyāb stories, they became known as the ‘orphan tales’.6 Let us now turn to the protagonists in the story so far. Trained as a classicist but turning to Oriental languages very early on, Galland first visited the East as secretary to Louis XIV ’s ambassador to the Sublime Porte in 1670. During his stay, he perfected his knowledge of Turkish, Persian and Arabic, and also acquired a number of manuscripts. In 1675, Galland was back in Paris but he was soon sent on another Eastern mission to purchase coins for the Cabinet du Roi. In 1679, he undertook his third, and longest journey to the Ottoman Empire (1679–83). The brief he received from the Minister Colbert, himself, was to learn all he could of the manners and customs of the Turks, but also to acquire artefacts and manuscripts.7 Interestingly enough, there is no mention of the Arabian Nights in any of the Journals from this period.

64

Life of Texts

Figure 3.1 First reference to H·annā Diyāb in A. Galland’s journal, BNF MS fr. 15277, p. 54.

Wandering Nights

65

Galland’s translation was not his first foray into the genre; in order to capitalize on the popularity of fairy tales, he started translating a collection of Persian fables, known as Kalīla wa Dimna, which were (and still are) immensely popular in the Arab world. This was followed by the tale of ‘Sindbad the Sailor’;8 however, as he explained later in his dedication to ‘Madame la marquise d’O, Dame du Palais de Madame la Duchesse de Borgogne’, he discovered that ‘it was excerpted from a wonderful collection (Recueil prodigieux) of similar tales, in multiple volumes, entitled 1001 Nights’.9 As a result, he decided to hold fire and withdraw the book from publication. At this stage, another actor appeared on the stage, François Pétis de la Croix (1653–1713),10 Professeur Ordinaire du Roy en Langue Arabique at the Collège de France, and the King’s Interpreter for Arabic, Persian and Turkish. He, too, had realized the sales potential of Sindbad, and in 1701 completed his translation entitled Histoire arabe de Sindabad le marin.11 When exactly Galland made the connection between the two texts is not known, nor is the basis upon which he made it. It is more than likely that he assumed they were related merely because they were clearly part of the same genre. To this day, no authentic Arabic manuscript of the 1001 Nights containing the story of Sindbad has ever been found. The absence of textual evidence did not stop Galland from inserting Sindbad’s tales (Nights 70–80), and adding night breaks to make them fit into the text. Intriguingly, he never referred to the nine-volume Turkish translation of the Nights (dated 1636), which had been in Paris since 1660.12 In October of 1701, Galland obtained an Arabic copy in three volumes, from Syria, as he explained in a letter of 13 October to Pierre-Daniel Huet (1630–1721), the Bishop of Avranches: Depuis trois ou quatre jours, j’ai appris par la lettre d’un ami de Halep, resident à Paris, qu’il a reçu de son pays, un livre arabe, que je l’avois prié de faire venir. Il est en trois volumes intitulé . . . Les Mille nuits. C’est un recueil de contes, dont on s’entretient en ce pays-là dans les veillées. J’ai prié cet ami de me le garder jusqu’à mon arrivée à Paris, pour le prix de dix écus, à quoi il revient d’achat et de port. Ce sera de quoi me diverter pendant les longues soirées.13 (About three or four days ago, I received a letter from an Aleppine friend, who resides in Paris, in which he stated that he had obtained an Arabic book I had requested from his country. It is in three volumes and entitled ‘The Thousand Nights’. It is a collection of tales, with which people entertain themselves in the evening in those countries. I asked this friend to keep it for me until I arrived in Paris, for the price of ten écus, which represents the cost of acquisition and postage. This will be something to amuse me during the long evenings.)

66

Life of Texts

However, the manuscript was incomplete as it contained a mere 35 stories, spread over 281 nights, and broke off at the beginning of the tale of ‘Qamar alZamān’.14 The dating of the manuscript has been the subject of some debate. The earliest date mentioned in the manuscript, and thus the terminus ante quem, is 3 December 1536 in a reader’s annotation by a Christian owner called Rizq Allāh Ibn Yūh.annā in Tripoli (Lebanon). The manuscript later travelled to Aleppo, where the oldest Muslim reader’s note is dated 18 November 1591, and it remained there until it was sent to Galland’s ‘friend’ in Paris. Where exactly it was written and/or compiled is impossible to say; the language used is of the socalled ‘Middle-Arabic’ variety, a mix of classical Arabic and the vernacular (in this case mainly Syrian Arabic), which was the usual register for this type of literature in the period (Fig. 3.2).15 Over the past centuries, many tentative dates have been put forward, ranging from the (late) thirteenth century, to the fifteenth. The current consensus lies somewhere in the middle, i.e. fourteenth–fifteenth, with numismatic references pointing to the mid-fifteenth century or, in any case, not before 1426.16 Another development occurred in 1702, when Galland allegedly received a fourth volume. However, as this has never been found, it has been suggested that Galland ‘invented’ it to legitimize his translation. The most compelling argument is that other manuscripts of what has become known as the ‘Syrian recension’ of the Nights end at the same juncture.17 Galland’s translation met with overwhelming success and the ensuing ‘Orientomania’ quickly transcended borders, influencing literatures across the European Continent, and embedding itself in the public consciousness. As one observer remarked, it ‘introduced to world literature a collection of tales that in terms of its international repercussion in imagination and creativity is second only to the Bible’.18 Soon after Galland’s 1001 Nights started to appear (1704), others released their own ‘Oriental’ tales, first among them Pétis de la Croix, who published Les Mille et un Jours (1710–12). This collection had its own imaginary pedigree; de la Croix claimed that it was based on a Persian manuscript which he had received from his Persian teacher (‘Moclès’) during a stay in the East. Closer examination revealed that the main source was, in fact, a Turkish translation of a well-known tenth-century Arabic literary anthology, al-Faraj ba‘d al-shidda (‘Joy after Hardship’) by al-Tanūkhī (940–94).19 As each volume garnered more success, Galland came under enormous pressure to produce more tales. He ran out of material from his original manuscript by volume 7, adding Sindbad and a full version of ‘Qamar al-Zamān’

Wandering Nights

67

Figure 3.2 1001 Nights manuscript (fourteenth century) used by Galland, BNF MS arabe 3610, fol. 4v. Beginning of the 70th Night.

68

Life of Texts

from other – unknown – sources. Desperate to satisfy demand, his publisher – without Galland’s permission or knowledge – printed volume 8 (1709), which contained a story (‘Ganem’) translated by Galland (once again from a hitherto unknown manuscript), alongside two stories (‘Zeyn Alasnam’, ‘Codadad’) translated by Pétis de la Croix for inclusion in his Les Mille et Un Jours. And so, Galland’s meeting with Diyāb could not have come at a more propitious moment. Galland’s translation method has attracted its fair share of criticism over the centuries, due to its cavalier use of sources, the unfettered embellishments and emendations to suit the taste of his audience. The often dramatic differences between source and target texts have even led to speculation that he might have used other manuscripts, or that much of his text is a mere fabrication. Galland’s translation not only forged the image of the Nights in European literatures for centuries to come, but his approach also became a model for other translators. In summary, there are a large number of stories for which no Arabic original has survived, or may reasonably be assumed to have been accessible to Galland, or even existed for that matter. Some are based on – now lost – manuscript versions provided by H·annā Diyāb, or on oral transmission. Some have no provenance, even if Galland never gave up hope of finding a ‘complete’ manuscript of the text, which has remained the Holy Grail in Arabian Nights studies to the present day. It was perhaps the search for ‘closure’ that caused him to add the most fanciful of his additions, i.e. the ‘conclusion’ of the tale, which is very rarely found elsewhere – and for good reason; how is one to explain the ‘deus ex machina’ of the king’s children appearing without any prior reference to pregnancy or delivery! Who was Diyāb, and why was he in France? Galland provided us with the basic information: he was a Maronite from Aleppo and had an association with Paul Lucas. Until recently, very little else was known about him, until a manuscript held at the Vatican was identified as his autobiography. As it is dated in 1766 and the author gives his age at the time as 75, we are able to place his birth in 1691. He was, thus, only about 16 or 17 years old when he met Paul Lucas, adventurer extraordinaire, who was on his second journey in the Levant (1704–8) to acquire coins and Oriental curiosités for the Cabinet du Roi (just as Galland had done some forty years earlier). His introduction to the young H·annā could not have been more enticing: If you want to travel the world, you’ll not find anyone better than myself . . . My task . . . is to record everything I see, to look for ancient dates . . ., coins and . . . local plants. . . . If you come with me, I’ll put you in the library of Arabic books;

Wandering Nights

69

you will receive payment from the King and be under his protection for the rest of your life. [My] Minister also asked me to bring back a man who knows Arabic.20

ʿAbd al-Qārī Ant·ūn Yūsuf H·annā Diyāb was born into a well-established merchant family in Aleppo, and had been a novice at a monastery near Tripoli. However, it seems the young man’s calling was weak and he returned home to reflect on his future. It is then that he met Lucas, and embarked upon a journey that would lead to Beirut, Cyprus, Egypt, the Barbary Regencies of Tripoli and Tunis, Livorno, Marseille and Paris, before returning home via Smyrna and Constantinople. Though Diyāb often speaks with affection and admiration of Lucas, the latter, like Galland, made no mention whatsoever in his own travelogue of his young Arab guide, who saved his life on at least one occasion. Galland’s predecessor at the Collège Royal was another Aleppine, But·rus Dīb (1620/2–1709), who was better known in France as Pierre Dipy and was probably the ‘friend’ providing the manuscript of the Nights. After studying in Rome at the Collegium Urbanum (precursor of the Pontifical Urban University), Dipy had arrived in France in 1667, and served as government interpreter before becoming Keeper of Oriental manuscripts. A number of Levantine Christian travellers in the period were men of the cloth, who received their clerical training at the Romebased Maronite seminary (Collegium Maronitarum), which had been set up by Pope Gregory XIII in 1584. It was run by Jesuits and was intended to help strengthen ties with the Maronites, who had accepted Papal authority in 1182. The same Pope also founded a college for Neophytes (1599), that is, converts from Islam and Judaism, as well as the first Arabic press, the Typographia Medicea in order to produce Arabic and Syriac materials for distribution in the Orient. Eastern Arab Christians were considered a particularly susceptible group and a prime target within the Congregatio de Propaganda Fide (Congregation for the Propagation of the Faith), which was founded in 1622 by Pope Gregory XV. Like Dipy, a number of these Arab clerics stayed on in Europe after their studies to work as teachers or librarians. They included Gabriel Sionita (Jibrā’īl al-S·ah·yūnī, 1577– 1648), who contributed to the Paris Polyglot Bible, and Joseph Assemani (Yūsuf al-Simcānī, 1687–1768), the curator of Arabic and Syriac manuscripts at the Vatican library. Other Christian Levantines headed to Europe to escape poverty, lured by the prospect of employment (for instance, as interpreters), or to seek their fortune in trade. In Paris, a number of Syrians could be found in the café trade. How did H·annā know all these stories? The English naturalist and physician Alexander Russell (1715–68) left a detailed account of the dramatic performance

70

Life of Texts

– and impact on audiences – of professional reciters (known as h.akawātī), who ‘in the midst of some interesting adventure, when the expectation of [the] audience is raised to the highest pitch, [break] off abruptly, and … escape from the room’.21 It is very likely that H·annā would have attended such gatherings, which were very popular among Christians and Muslims alike. In fact, they proved a little too popular and a number of Aleppine Catholic Orders prohibited their members from attending such meetings due to their alleged immoral character.22 The stories also reflect the ‘multiconfessional’ audiences and contain Christian, as well as Jewish and Muslim elements. The practice was a common one in other cities, too, and the British Arabist Edward Lane wrote on such gatherings in Egypt, where it was a particular type of reciters specializing in the pre-Islamic Arabian hero ʿAntara’s exploits that would also present the Nights. He remarked that copyists regularly ‘altered and modernized’ the language to suit the intended audiences.23 Before leaving Diyāb, there is a passage in his travelogue that has some relevance to one of his stories. Not long after leaving Aleppo, Lucas and his entourage hit upon a tomb at the top of a mountain: He [Lucas] walked around the tomb, looking for a way in. He only saw a small opening, and asked one of the armed escort to go down through it. None of them did; they said that it could contain a wild animal, like a hyena, leopard, or something. . . . As we were talking, a shepherd walked by, and the officers asked him to go down. . . . The tomb was six feet and one span of the hand deep. The Frenchman said to the shepherd: ‘Go around the tomb and give me everything you find.’ He started to walk around and saw a human skull, which he handed over. It was the size of a large watermelon. The Frenchman told us it was the skull of a man. Then, the shepherd handed him another skull, which was smaller. The Frenchman said it belonged to a woman. He claimed that the tomb was that of rulers of the area. He threw the shepherd a piece of cloth and said: ‘gather everything you find on the ground and give it to me.’ The shepherd proceeded to do so, and among the things he collected we saw a large flat ring. The Frenchman examined it and said that it was rusty, and that there was no clear writing on it. He was not able to identify the metal from which it was made, nor whether it was gold, silver, or something else. He kept it with him. Then he said to the shepherd: ‘Feel around the walls of the tomb.’ As the shepherd did so, he felt a niche inside of which there was a lamp, similar to those of butter vendors, but he could not identify the material . . .24

Someone asked to go down a mysterious tomb to collect precious items, a ring, a lamp – is the uncanny resemblance to elements in the story of Aladdin a mere

Wandering Nights

71

coincidence? Or is this a case of a talented raconteur aiming to please ‘the old man’ by drawing on a pool of stories he has heard, and then like all good h.akawātīs embellishing it with his own imagination and ‘combining the incidents of different tales’?25 In the absence of an Arabic source text, it is as valid a hypothesis as any.

Dawn of the 1001 Nights What is actually known about the composition of the earliest texts of the 1001 Nights? Where and when did they originate? Where should they be situated within Arabic literature? The first reference can be found in the monumental encyclopaedia Murūj aldhahab (‘Meadows of Gold’) by the Baghdad-born historian al-Masʿūdī (first half tenth century). In a section on translations of books from Persia, India and Greece, he mentioned a book called Hezār Afsāneh, ‘which is Persian and translates into Arabic as “One thousand fairytales (khurāfāt)” . . . It is the story of a king, his vizier, his daughter and slave, Shīrāzād and Dīnāzād. This is similar to . . . the book of Sindbad’.26 A few decades later, Ibn al-Nadīm’s ‘Index’ (fihrist) of Arabic books (986) provided additional information, as well as the only literary critique of the Nights until the early modern age: ‘it is truly a coarse book, without warmth in the telling’.27 A number of crucial elements can be deduced from these accounts. Although the tales were traced to a Persian original (with Indian influences), they underwent a number of transformations rather than mere translation into Arabic. A Persian origin is further supported by the names of the characters in the frame-story, even though their status and spelling show some variation: Dīnārzād/Dīnāzād (the Dunyāzād of ‘modern’ texts) isn’t Shahrazād’s (or Shīrāzād) sister, but her slave (al-Masʿūdī), or the head of the household (Ibn alNadīm). The Nights were already considered part of a genre, which included the tale of Sindbad (the Sindbād-nāmeh of Persian literature),28 and fables like Kalīla wa Dimna. The tales were recited in gatherings at courts, and despite their low literary value they merited written preservation because of their cultural significance, which gainsays the hypothesis that the Nights were only produced orally. Their importance is underscored by the fact that the brother of one of Islam’s greatest religious scholars, al-Shāfiʿī, is said to have made a compilation of the Nights, whereas Ibn al-Nadīm reported that he saw several complete copies. Last but not least, he mentions the name of an alleged principal compiler, Abū

72

Life of Texts

ʿAbd Allāh al-Jahshiyārī, who selected tales from ‘books of night-time stories’, but only managed to complete 480 out of 1,000 nights before his death. In 1949, Nadia Abbott discovered what is still the oldest manuscript extract amidst a collection of papyri purchased by the University of Chicago (Fig. 3.3).29 The document consists of two folios with six fragments of a miscellany, including the draft of a letter and some legal testimony, dated 879. As a result, the extracts from the Nights – the title page and a few lines from the opening page – must be older than that, which is in keeping with the type of script used. Like everything else in the history of the Nights, the fragment raises a plethora of fundamental questions, but answers none. The fact that the text is on paper that had become waste already by 879 would exclude Egypt as the origin since papyrus was in much longer use there, and the manufacture of paper was introduced only in the first half of the tenth century. There is a reference to Syria and Bedouins in the text, an owner who had left Antioch (northern Syria), but it was found in Cairo! This is crucial evidence of the fact that already by the mid-ninth century, a version (still ‘one thousand nights’, rather than 1001) was circulating across the eastern Mediterranean. The popularity of the Nights is supported by a number of later sources, such as the Tārīkh Mis.r (‘History of Egypt’, 1160–71) by Ibn Saʿd al-Qurt·ī, which contains the oldest known literary reference to ‘One Thousand and One Nights’ (rather than ‘One Thousand’). In the same period, possibly slightly earlier, the title also appeared in a Jewish bookseller’s records from the Cairo Geniza (c. 1150).30 It is very difficult to make any statements about what happened to the Nights between the tenth and fifteenth centuries. Based on the currently available evidence, it would appear that there was a more or less stable ‘core’, including the frame-story and the first five cycles of the Galland manuscript (‘The Merchant and the Demon’, ‘The Fisherman and the Demon’, ‘The Porter and the Three Ladies’, ‘The Three Apples’ and ‘The Hunchback’s tale’), based on a thematic coherence between the frame-story and the embedded stories. It is to this core, then, that an unstable collection of stories was added through various compilations.31 Research has shown that prior to the Galland manuscript there was probably at least one other text tradition. This hypothesis relies on the inclusion of the tale of S.ūl and Shumūl in two manuscripts, held in Tübingen and Riyad, tentatively dated to the fourteenth and fifteenth centuries, respectively.32 A collection of between fifteen and twenty stories entitled ‘101 Nights’ (Mi’at layla wa layla) constitutes another interesting line of inquiry, especially since an early thirteenth-century copy has recently (2010) come to light.33 As the known manuscripts were all produced in the Muslim West (North Africa, Muslim Spain),

Wandering Nights

73

Figure 3.3 Earliest 1001 Nights manuscript (ninth century), Chicago University Oriental Institute, MS E17618.

it may be (part of) a ‘Western’ variant of the Nights, or even a separate stage in their development. There are similarities in the frame-story and a number of the tales, but its exact status in the history of the Nights remains to be determined. The Vatican library holds a copy of the Galland manuscript from 1001/1592, though the ‘coincidence’ in the Muslim year and the number of Nights has raised some doubts about the veracity of the copyist’s note.34 Two other manuscripts tentatively dated to the sixteenth century are also worthy of mention. One is in Kayseri (Turkey) and is thought to be the oldest to include the conclusion of the Nights,35 whereas the other (John Rylands Library) is of particular significance since it is the only one to include illustrations (twenty-nine in total) (Fig. 3.4).36

74

Life of Texts

Figure 3.4 Page from 1001 Nights manuscript (sixteenth century), Manchester, John Rylands Library, Arabic MS 706, fol. 206r.

Wandering Nights

75

The seventeenth century was a particularly fruitful period with no fewer than thirteen manuscripts of the 1001 Nights – once again all incomplete and with uncertain dating.37 The manuscript brought back by the French consul in Egypt, Benoît de Maillet (1656–1738) in 1702 is significant as it covers 1–905 Nights (with some omissions), which was the highest number until the appearance of complete sets in the nineteenth century.38 Secondly, it cannot be subsumed in the main recension families and may thus afford valuable insight into the development of the cycle since the Galland manuscript. Let us turn to the hypotheses regarding the various interconnecting strands in the genesis of the Nights. Abbott held that the Hezār Afsāneh was translated from Persian into Arabic in the ninth century, with additional Arabic material being added in the initial version of the Nights. As a result, the tales were expanded and/or ‘re-sited’ through the inclusion of settings and characters. As regards the alleged collection by al-Jahshiyārī in the tenth century, its contents are entirely unknown as the text is lost and no other references to it have been found. The next stage would have taken place between the thirteenth and sixteenth centuries, with the inclusion of Egyptian tales. It is possible to identify a number of other antecedents. First among them is the Indian influence in the use of a frame-story and even similarities in some of the tales.39 One should be wary, however, of establishing narrative lineages as a number of common features co-existed in oral stories across the Middle East.40 Unfortunately, no text of the Hezār Afsāneh, or of any of the proposed compilation stages has survived. It is not unlikely that a ‘canonical’ text never existed, and that different compilations were in circulation at any given time. Indeed, the title, itself, may be hyperbolic, implying an ‘openended’ set of stories, rather than a closed cycle of exactly 1001 nights. There is little doubt that the tales were primarily intended for recitation and that they were part of a large tradition of oral folk tales, many of which were written down. The ‘low-brow’ language and intent also explains the near-absence of the text from classical Arabic literature and the low esteem in which it was – and still is – held.

Imaginary Wanderings On the heels of the success of the Galland translation, the Nights began the wander across ever more labyrinthine paths as Orientomania rose to fever pitch, with an added impetus being given by Napoleon’s Egyptian campaign (1798–1801).

76

Life of Texts

The appetite for new stories caused a renewed hunt for manuscripts and the dearth of copies in the East presented lucrative opportunities for copyists – as well as would-be forgers. First among the latter, we find an Aleppine cleric, known as Dom Denis Chavis (Diyūnīsiyūs Shāwīsh), a former student of the St Athanasius Greek School in Constantinople. He first arrived in Paris in the early 1780s and taught at the Bibliothèque du Roi, presumably Arabic to the so-called ‘Jeunes de langue’, young boys destined for a career as dragoman (interpreter) in French consulates in the Ottoman Empire. Chavis claimed to have found a manuscript from Baghdad, which completed that of Galland.41 In addition to continuing the story of ‘Qamar al-Zamān’, it contained the much-sought-after Arabic version of a number of orphan stories, including Aladdin, combined with other material from an eclectic mix of sources. Chavis soon realized that the real potential – and financial gain – lay in a French text, and he approached the editor of Le Cabinet des Fées, Paul Barde. The proposal to render the ‘continuation’ of Galland into French was eagerly accepted and Barde suggested a collaboration with Jacques Cazotte (1719–92). Chavis ceased work on his ‘recension’ (reaching the 631st Night) and refocused his efforts on a French translation. It is interesting to note the similarity in methodology with Galland and Diyāb, as Chavis provided a French gist translation, which Cazotte then transformed into a French literary text. The translation appeared in 1788–9 in Geneva under the title Continuation des Mille et Une Nuits, the editor stating proudly: le reste des manuscrits qui devoient terminer l’ouvrage . . . viennent d’y être apportés par Dom Denis Chavis . . . ce savant arabe a entrepris d’enrichir notre littérature amusante, de cette charmant Suite qu’il a apportée en France.42 (The remainder of the manuscripts that were supposed to complete the work . . . have been provided by Dom Denis Chavis . . . this Arab scholar, who has made it his task to enrich our entertaining literature with this charming sequel, which he has brought with him to France.)

Seven of the stories in the French translation cannot be found in the Chavis recension, or anywhere else for that matter, and must therefore be assumed to be inventions by Cazotte and/or Chavis. Very early on, the Chavis manuscript raised doubts, especially due to the large number of French calques in the Arabic. In fact, the entire copy was a fraud and was essentially a patchwork of ‘backtranslation’ of Galland’s French text. Besides replacing the Sindbad cycle with another story, there are additions from other collections,43 and an Arabic translation of Pétis de la Croix’s Zayn Asnām (from the eighth volume of Galland’s translation). For a while, the duo got away with their deception and the

Wandering Nights

77

‘Continuation’ became a success abroad as well, with no fewer than three English translations between 1792 and 1794. Many of the Chavis stories can also be found in some of the nineteenth-century translations of the Nights. The eighteenth century saw a dramatic increase in the number of manuscripts, with over thirty partial collections. One should be wary to explain this through a sudden boost in interest in the tales in the Arab world as many of the manuscripts were copies made for European travellers. Among them, there is one copied in Egypt in 1764 and owned by Edward Wortley-Montague (1713– 76).44 It stands out not only because of its size (only one of an originally complete set of eight volumes is missing), but also because over thirty stories are found only in this manuscript.45 Although some parts are similar to the Galland manuscript, most of it belongs to a different tradition.46 The manuscript acquired by Alexander Russell in Aleppo merits attention as well. He commented that he only found copies containing 280 Nights, which corroborates the earlier point that Galland’s Syrian recension was already incomplete (and remains so to the present).47 As the Russell manuscript is a copy of the one in the Vatican, it is the final known descendant of the Galland text.48 In the nineteenth century, the story of the Nights would take even more unusual turns, not least because for the first time complete sets were discovered (or copied), and by the end of the century about ten of them were held in European libraries. However, the circumstances in which they came into being often have more in common with detective fiction than with codicology. The man at the centre of the story of the first of these manuscripts was a Syrian expatriate by the name of Mikhā’īl Ibn Niqūlā Ibn Ibrāhīm al-S.abbāgh (c. 1775–1816), better known in France as Michel Sabbagh. Born into a prominent Melkite family in Acre, where his grandfather was personal physician to the ruler, he spent his early childhood in Damascus before moving to Egypt. When the French invaded, he entered the employ of the occupying army as an interpreter. After Napoleon’s defeat, a large number of Christians (both Egyptians and Syrians) who had collaborated with the French accompanied the retreating army to France. Though the community was initially welcomed, their economic situation grew increasingly precarious as the majority could only rely on meagre government pensions. Sabbagh was able to eke out an existence as a scribe and corrector at the royal printing works, and then at the Bibliothèque du Roi where he was entrusted with the restoration and copying of Arabic manuscripts. Money was always tight, as shown by the moving letter he wrote to the King in which he begged for funds, after his forced departure from Egypt when ‘the Muslims pillaged, sold our women and killed my uncle’.49 At one point, the doyen of French

78

Life of Texts

oriental scholarship, Silvestre de Sacy (1758–1838) even intervened on Sabbagh’s behalf, praising the assistance he provided to ‘des travaux relatifs à la langue et à la littérature arabe’ (‘research related to Arabic language and literature’).50 His contribution to the Nights came when he transcribed a manuscript for Jacques Caussin de Perceval (1759–1835), then curator of Arabic manuscripts at the Royal Library as well as incumbent of the chair of Arabic at the Collège de France. The manuscript Sabbagh claimed to have discovered was remarkable in a number of ways.51 It was a complete set, dated 21 October 1703 in Baghdad, and thus predated Galland’s translation. More significantly, its provenance held out the promise of an Iraqi strand in the textual tradition of the Nights, which would be logical as the Abbasid capital was the setting of many of the core stories.52 Besides the tales of the Galland manuscript, Sabbagh’s text contained the additional stories of the mysterious fourth volume, as well as tales also found in the Chavis redaction. All of this, combined with the absence of the Sindbad and Ali Baba cycles, pointed towards the Sabbagh manuscript being the missing link between different recensions. The Sabbagh text was neatly broken down into Nights and included the happy ending so vital to fairytales. The problem was that, as we have seen, there was no fourth volume and Chavis’ text was a counterfeit. Furthermore, the original manuscript was never actually seen by anyone else. Contrary to Chavis, however, Sabbagh was an experienced manuscript copyist and through astute linguistic manipulation was able to hoodwink specialists for almost two centuries. The deception was unravelled only in 1984 by Muhsin Mahdi, who established that Sabbagh copied the framestory and the first 69 Nights from the Galland and Chavis manuscripts, added with stories from the Maillet version and a miscellany of other manuscripts, amending and emendating as he went along. Very much like a painter going through several sketches, Sabbagh honed his forgery in a number of drafts.53 His manuscript hugely impacted – or ‘contaminated’ (to use Mahdi’s term) – the textual history of the Nights as it directly or indirectly served as a source for a number of translations as well as printed editions.54 In the same period, there was an increase in manuscripts of the text in Egypt with, for the first time, complete sets, several of which were brought back to Europe by travellers such as the Swiss John Lewis Burckhardt (1784–1817),55 the German Ulrich Seetzen (1767–1811),56 and the Frenchman Jean-Louis Asselin de Cherville (1772–1822).57 Another major development was the recognition of a new branch of the Nights that was distinct from the Galland text and had emerged in Egypt. The earliest compilation of what would later become known as the Egyptian recension, or the ‘Zotenberg Egyptian Recension’ (ZER ), after the

Wandering Nights

79

scholar who first researched it, can be traced back to the 1770s, when a complete copy was made by an Egyptian shaykh, whose name and motives remain unknown. The oldest reference can be found in Seetzen’s journal for July 1807, which confirmed the theory of a nucleus of tales being added with a variable pool of other stories; Asselin (the then dragoman of the French consulate) discovered that the Arab tales of 1001 Nights were not all made by the author (Verfasser), who only wrote around two hundred Nights, the remaining ones being drawn from a collection of existing stories, which he gathered together. The one who did this was a local shaykh by the name of . . ., who died some 26 years ago and was known by many living shaykhs.58

This rather cryptic statement is generally taken to mean that the anonymous shaykh used a pre-existing core of two hundred Nights to which he added other known tales.59 The first Arabic printed edition was another foreign endeavour, and intended as a teaching tool for British officials studying at the College of Fort William in Calcutta, where it was also printed in two volumes (1814–18). In the Persian preface, the editor ‘Shuekh Uhmud Bin Moohummud Shirwanee Ool Yumunee’ (Ah.mad Ibn Muh.ammad Shīrwānī al-Yamanī), who was an Arabic teacher at the College, explained that the Syrian author wrote the text in a conversational style and informal register for those wishing to learn the language. By accident, rather than by design, the edition relied on a recension close to the Galland manuscript.60 Al-Yamanī randomly selected two hundred nights, but even if it essentially corresponded to the Galland (Syrian) recension, there are some significant differences, such as the addition of the tale of Būrān and al-Ma’mūn’ (Nights 94–100) from ZER , and stories from Langlès’ Les voyages de Sind-Bâd le Marin et La Ruse des Femmes (1814). Al-Yamanī also made considerable editorial changes to the language, classicizing many of the colloquialisms. Due mainly to its limited scope, this Calcutta edition was never widely circulated, though there was a single-volume lithograph reprint in 1829.61 The next decade saw the publication of the first – and only – European edition of the Arabic text, based on another mysterious manuscript, this time from a purportedly North African strand of the Nights. The editor was the German scholar Christian Maximilian Habicht (1775–1839), who arrived in Paris in 1797 as secretary to the Prussian legation and remained until 1807, taking classes at the École des Langues orientales, where his teachers included de Sacy, and the Aleppoborn Melkite Priest Rafā’īl Ant·ūn Zākhūr (1759–1831), known in France as Dom de Monachis, who had been one of the official interpreters to the French occupation

80

Life of Texts

force. In recompense for his services, he was appointed to the chair in dialectal Arabic (1803–16) at the School, where Champollion was among his pupils. During his stay in Paris, Habicht established close relations with members of the Arab expatriate community, among them the Tunisian Jew Mordechai Naggiar (Murdīkhāy al-Najjār), with whom he even shared a flat in Paris.62 Hardly anything is known about this scholar, who was the last of the Arab compilers of the Nights in Europe. According to the American missionary Joseph Greaves, who met him in Tunis in 1825, al-Najjār had contacts with de Sacy and ‘other Oriental scholars.’63 In the 1820s, Habicht, who by then had taken up the chair of Arabic in his native Breslau, began publishing both an Arabic edition (1824) and a German translation of the Nights (1825), both of which have been discredited.64 Habicht’s methodology was very much in line with the Chavis school of ‘textual tradition’, with a fabricated ‘Tunisian’ recension of the Nights provided by al-Najjār serving as the fictional basis for an expanded translation.65 The fraud was unmasked by the American Arabist missionary Duncan B. MacDonald (1909), who proved that the Tunisian manuscript never existed; rather, Habicht’s edition was concocted from the Chavis, Sabbagh and, possibly, Maillet manuscripts, added with some stories ‘originated’ by al-Najjār from a variety of sources. Habicht’s German translation, which contains nearly seventy stories that are not in his Arabic edition, is based on another amalgam of texts, including the translations by Caussin de Perceval (1806) and Jonathan Scott (1811), combined with several Nights (884–1001) from the spurious Tunisian recension, and nearly thirty stories found only in Wortley-Montague. It is unclear whether al-Najjār prepared the fraudulent text at Habicht’s request – and knowledge – or whether he did so without a specific buyer in mind. It is known that Habicht started collecting materials relating to the Nights during his Paris stay and his interest would, of course, have been known to his Arab friends, nearly all of whom were active in what had become a cottage copying industry for 1001 Nights manuscripts to satisfy demand from scholars. When it comes to Habicht’s role, one observer accused him of ‘wilfully creat[ing] a literary myth, and enormously confus[ing] the history of the Nights’.66 Another member on the fringes of the Parisian Arab community in the early part of the century was a certain Jean Varsy, whose name became associated with what was thought to be the oldest Arabic redaction of ‘Ali Baba and the Forty Thieves’. The story surrounding this text is every bit as enthralling as those of Sabbagh and Chavis. In 1860, a Parisian bookseller sold the Bodleian Library a manuscript containing the tales of ‘Hārūn al-Rashīd and The Daughter of Kisrā’,

Wandering Nights

81

and ‘Ali Baba and the Forty Thieves and Morgana the Slave-girl’ (ʿAlī Bābā maʿa ’l-lus.ūs. al-arbaʿīn wa ’l-jāriya Marjāna). It lay unexplored for nearly half a century, before drawing the attention of Duncan MacDonald, who published the text and suggested it was the version given to Galland by Diyāb. There were a few oddities, though, such as the discrepancy between the clearly Christian name of the copyist – Yūh.annā [Ibn Yūsuf Wārisī] – and the use of Islamic religious formulae. Secondly, Galland’s journal did not mention a written version of the text, only a synopsis of what was recounted to him by Diyāb.67 Finally, there was the language, which was a rather peculiar hotchpotch of dialects and literary Arabic, with the odd calque from French. However, Macdonald, and others, excluded that it was a translation of Galland’s text, on the grounds that there were significant differences between the two. Macdonald was able to establish that ‘Yūh.annā Wārisī’ was, in fact, Jean Varsy, and concluded that he, like Galland, had copied a tale he had read or heard.68 For seventy years, this remained the consensus, until Muhsin Mahdi re-examined the text and, based on inconsistencies in both content and language, was convinced that it was, in fact, another fiction, an Arabic ‘recreation’ of Galland’s translation.69 K. Zakharia’s recent researches70 have shed more light on the alleged forger, Jean-Georges Varsy (1774–1859), whose family ran a trading house in Rosetta. Like so many European merchants, they were forced to flee the country after the defeat of the French army, as the local population exacted its revenge on their property. Besides his commercial activities, Varsy was a keen Arabist, manuscript collector and copyist.71 Zakharia makes a compelling case for Varsy’s Arabic ability, which successfully erodes Mahdi’s comments about his linguistic incompetence, while proving that the mixture of Classical Arabic and vernacular in the Ali Baba text is entirely in keeping with the Middle Arabic register of the genre. Although Zakharia’s findings undermined Mahdi’s arguments, they do not, in themselves, refute the forgery claim, rather that it is a more adept one. On the other hand, if one adds the extent and nature of the discrepancies between Galland’s and Varsy’s texts, it does become less likely that Varsy adapted the French text. As a result, the hypothesis that the text was copied from an existing – but hitherto unknown – source cannot be rejected out of hand. After the flurry of copying activity at the start of the century in order to meet the demand by European scholars, the supply of manuscripts of the Nights had dwindled in Cairo by the 1830s. According to Lane, ‘even fragments of them are with difficulty procured; and when a complete copy of “the Thousand and One Nights” is found, the price demanded for it is too great for a reciter to have it in his power to pay.’72 In 1831, another complete set was made for the German vice-

82

Life of Texts

consul in Egypt, Carl Reinhardt, and remains important since, like the WortleyMontague manuscript, it falls outside the Syrian and ZER families.73 The increased demand, and perhaps also interest in the Nights in Egypt, led to the printing of the first complete edition in October 1835, at the government press in Bulaq (then a suburb of Cairo), which had been set up only a decade earlier (1822). The two volumes went on sale for the considerable price of 100 piasters, which was equivalent to the cost of a year’s supply of sugar or rice for the average middle-class Cairene family.74 The Bulaq edition represents the Egyptian recension of the seventeenth and eighteenth centuries, and was based on the text compiled by the above-mentioned anonymous shaykh in the 1770s, though the actual manuscript(s) has not been located. It was edited by Qit·t·a alʿAdawī (d. 1864), a teacher at the newly founded Language School and a revisor at the printing press.75 It is in this edition that the phrase most associated with the Nights in Arabic today – ‘adraka Shahrazād al-s.abāh. wa sakatat ʿan al-kalām al-mubāh. (‘As morning overtook Shahrazād, she ceased her lawful speech’) – made its first appearance. The text gained considerable popularity and a number of other editions followed in Cairo (1862, 1880) and Beirut (1880, 1888–9). The first printed Turkish translation appeared in Istanbul in 1850. Four years after the Bulaq edition, it was in Calcutta that the so-called ‘Vulgate’ of the Arabian Nights was published in four volumes (1839–42), edited by Sir William Hay Macnaghten.76 It was primarily based on a manuscript that originally belonged to Britain’s Consul-General in Egypt, the Egyptologist Henry Salt (1780–1827), or so it was thought. Mahdi’s ingenious detective work revealed that here, too, deceit had been aforethought, and that the Salt manuscript was a composite of a late ZER manuscript, Calcutta I and the Breslau edition.77 As the latter were based on the Galland manuscript, the Syrian and Egyptian strands thus merged for the first time in Calcutta II , after being separated for several hundred years. In a final irony, it is this edition which soon gained ‘canonical’ status as the most complete, authentic and original Nights. From then on, Arabic editions have been reprints of Bulaq or Calcutta II , with varying degrees of editorial additions, omissions and bowdlerization.

Noctes Volant, Scripta Manent As already stated, it was not until the late nineteenth century that the Arabic text of the Nights became the object of serious research with the trailblazing work by Hermann Zotenberg (1887, 1888), who identified an Egyptian recension,

Wandering Nights

83

compiled in the eighteenth century and represented by Bulaq and Calcutta II .78 It is distinct from the Syrian strain through the addition (often from miscellaneous folktales), omission and amendment of tales, as well as a number of other features in terms of genre, language, characterization, cultural origins, and setting. Dating – and authentication – of the manuscripts poses a number of intractable problems; palaeographic evidence is not always conclusive, whereas we have seen that scribal and other metatextual information cannot always be taken at face value. Rather than tracing a single ‘hypotext’, the analysis often has to take into account variable dates of individual tales. There are many stories whose provenance and time of inclusion in the 1001 Nights is unknown. What is more, in cases where tales occur in only one manuscript, one cannot exclude the possibility that these were added by a copyist. Historical details often provide valuable clues; for instance, the reference to coffee and tobacco in the tale of ‘Abū Qīr the Dyer and Abū S.īr the Barber’ (Calcutta II ) precludes a date prior to the sixteenth century when these items were introduced in Egypt. Linguistic features, too, are a highly complex indicator, as dialects develop over time and judgements based on current geographical distribution of features cannot always be extrapolated. There has been interference between versions from different branches, further compounded by emendations at the hands of copyists and editors, who often felt compelled to ‘classicize’ colloquialisms.79 A key question revolves around the comparative analysis of the extant manuscripts that contain stories from the Nights. At present, their number stands at around 105, most of which await further study.80 Only thirteen manuscripts are complete (all of the ZER text), with different parts surviving in the partial copies. The history and development of the text of the Nights do not enable us to make reliable judgements about any collections predating the incomplete Galland manuscript. The fact that a ‘nucleus’ can be found in the major manuscript and printed collections does not provide any evidence regarding some sort of canonical set prior to the fifteenth century, rather of subsequent copying. The Encyclopedia of the Arabian Nights lists 551 stories across the major manuscript, printed and translated collections.81 If we exclude some duplicates across editions, we are left with about 305 main stories,82 with the Bulaq and Calcutta II editions each containing around 260. Even within the same recension there is variation in the stories that are included. A particular feature of texts that are neither in the Syrian nor ZER branches is the number of unique stories, though this affects others as well. For example, over thirty tales are found only in Reinhardt, around ten only in the Maillet manuscript, and over sixty only in the Wortley-Montague and Breslau texts.83

84

Life of Texts

Other differences include the absence of a breakdown into nights, the length of individual nights, the language and content of the stories, as well as the composition of tale cycles. Even within the canonical ‘core’, differences can be quite substantial. Let us take, for example, the famous ‘Tale of The Three Apples’. This has been called the first whodunit in Arabic literature, and involves the discovery of a box containing the dismembered body of a young woman, and the subsequent unmasking of her killer. In the Galland manuscript, it starts in Night 69, though in his translation it only comes in the 90th (after Sindbad’s 7th Journey). The Breslau edition keeps the beginning in the 69th Night, whereas in Calcutta I, it is the 68th, in Calcutta II and Bulaq, the 19th, in Maillet, the 66th, and the 18th in the 1935 Cairo edition.

Translation and Appropriation It was the next journey of the Nights, that of ‘translation’ to the West in the early modern period that would have the widest impact on the status and image in both Western and Arabic literatures. The almost obsessive quest of nineteenthcentury scholars for more Arabian Nights was matched by the insatiable appetite for translations by Western audiences. In 1836, the critic and poet James Leigh Hunt wrote that the Arabian Nights were, quite simply, ‘the most popular book in the world’.84 The complexities of the Arabic source text(s) discussed above naturally impinged on the translations, which further muddied textual waters. In addition to translators basing themselves on material that is now known to be fraudulent, some further corrupted the text by idiosyncratic restructuring of existing elements and/or interpolating others from extraneous sources, both real and invented. The issues with Habicht’s German text were by no means exceptional. Equally whimsical and fraudulent in its sourcing was Joseph Mardrus’ French rendition (1899–1904), which introduced a number of stories previously unrelated to the Nights.85 Edward Lane, for his part, based his English translation (1838–41) on the Bulaq, Calcutta I and Breslau editions, and not only heavily expurgated the text, but also omitted the division into nights and much of the poetry.86 Lane’s translation quickly gained popularity, with new editions appearing in 1859 and 1883. However, it was soon eclipsed by Sir Richard Burton’s (1821–90) version, which contains 417 tales and remains the most extensive version of the 1001 Nights available in English, at least in quantity, if nothing else. Burton used a

Wandering Nights

85

variety of editions (Calcutta II , Breslau,Wortley-Montague) as well as translations (Galland, Pétis de la Croix, Chavis & Cazotte) available to him. The extent of Burton’s reliance on John Payne’s text (based on the Calcutta II and Breslau editions), which appeared shortly before (1882–4), resulted in accusations of plagiarism, which are still hotly debated today.87 The success of the first ten volumes (1885–6) prompted Burton to publish a second set of six supplemental volumes (1887–9). Due to its allegedly lewd content – much of it greatly enhanced by Burton, himself – the text was officially printed in Benares, for the members of the Kama Shastra Society, to avoid prosecution under British obscenity laws. In addition to textual liberties, it was Burton’s salacious pseudoscholarly apparatus that attracted most attention, particularly the ‘Terminal Essay’ and its discussion of the ‘turpiloquium’ in Eastern societies.88 By the first quarter of the twentieth century, the image of the Nights in both France and England was determined by, respectively, Mardrus’ and Burton’s fanciful translations, which were the fountainhead of many of the Orientalist stereotypes and tropes that have been reinforced in popular culture. Mardrus’ text served as the source for the translation by Edward Powys Mathers (1923), whose lack of knowledge of Arabic did not prevent him from making further amendments. Although the most recent translators Malcolm and Ursula Lyons eschewed Burton’s ‘mock-Elizabethan, bogus Oriental style’,89 their source text (Calcutta II ) remains equally flawed. Notable exceptions are the translations of the Galland manuscript into English and German.90 Very little has changed in the meantime in terms of the textual history of the Nights. The Galland manuscript remains the oldest Arabic version to be unaffected by modern fraud and represents at least one medieval recension. The post-nineteenth-century Arabic editions (thirty-eight since 1950)91 often combine multiple unacknowledged recensions with extreme editorial licence. In the modern era, the most important innovation, if one may call it such, in the Arabic textual tradition has been the insertion of European tales, such as the orphan stories, and ‘Europeanized’ versions of others. The naturalization or cultural appropriation – or re-appropriation, to be more exact – of these is perhaps the most remarkable part of the story of the Nights, as they mutated through redaction and translation, and back again. What is more, it is the European invention of the Nights that has penetrated the editions in the Arab world; today most Arabic-speaking children claim Aladdin as their own, and are as familiar with the magic lamp and the flying carpet as their Western counterparts. The shibboleth ‘Open, Sesame!’ is firmly embedded in Arabic cultural idiom (Iftah. yā Simsim!), and even serves as the title of the Arabic

86

Life of Texts

production of the popular television show Sesame Street. And so, in a bizarre case of re-transference, the text of the Nights has returned to its homeland as much a fiction as the stories themselves. This raises the question to what extent the current versions of Arabian Nights can be considered a representation of Arab culture of any age, and about their long-standing use as a historical source or manual to medieval Islamic civilization. Furthermore, in the absence of an authentic medieval recension, and anything remotely resembling a canonical version, where does this leave the ever increasing Arabian Nights scholarship, which is still predominantly the preserve of Western academe? A lot remains to be done, of course, not least in terms of the analysis of extant texts, many of which await further examination. Translation, too, has more to offer, and early Turkish versions hold out the hope of new insights into the textual development and distribution of the Nights. In Western culture, the image is equally distorted, and it is Burton’s prurient monument to Orientalism, which, despite its ludicrous and impenetrable prose, has had the greatest influence on the reception of the Arabian Nights and, by extension, of Arab-Islamic culture. Like all works of world literature, the Arabian Nights have transcended literatures, cultures, and history, and may rightfully be claimed by all. Regardless of the shape and format the Nights mutated into, they have continued eminently to serve the aim with which they were once created – to entertain and divert wherever they may wander.

Appendix I: Main Editions and Translations of the Thousand and One Nights Editions H · ikāyāt mi’at layla min alf layla wa layla, ed. Ah.mad al-Yamanī, 2 vols (Calcutta: Hindoostanee Press, 1814–18). Kitâb Alf layla wa-layla min al-mubtada’ ilâ al-muntahâ/Tausend und Eine Nacht. Arabisch. Nach einer Handschrift aus Tunis, ed. Maximilian Habicht and Heinrich Leberecht Fleischer, 12 vols (Breslau: Ferdinand Hirt, 1824–43). Alf layla wa layla, ed. Muh.ammad al-ʿAdawī, 2 vols (Cairo: Būlāq, 1252/1835). The Alif Laila Or Book of the Thousand Nights and One Night. Commonly known as ‘The Arabian Nights’ Entertainments’, ed. William Hay Macnaghten, 4 vols (Calcutta: W. Thacker & Co, 1839–42). The Thousand and One Nights (Alf Layla wa-Layla) from the Earliest Known Sources, ed. Muhsin Mahdi, 3 vols (Leiden: Brill, 1984–94).

Wandering Nights

87

Translations Les Mille et une nuit. Contes arabes, trans. Antoine Galland, 12 vols (Paris: Barbin, 1704–17). Tausend und Eine Nacht. Arabische Erzählungen. Zum erstenmal aus einer Tunesischen Handschrift ergänzt und vollständig übersezt, ed. and trans. Maximilian Habicht, Friedrich Heinrich von der Hagen and Karl Schall, 15 vols (Breslau: Josef Max und Komp, 1825). Tausend und eine Nacht. Arabische Erzählungen. Zum Erstenmale aus dem ara-bischen Urtext übersetzt, trans. Gustav Weil, 4 vols (Vol. 1, Stuttgart: Verlag der Klassiker, 1838–41; Vols 2–4, 1839–41). The Thousand and One Nights, commonly called in England the Arabian Nights’ Entertainments, trans. Edward William Lane, 3 vols (London: Charles Knight, 1839–41). The Book of the Thousand Nights and One Night; now first completely done into English prose and verse, from the original Arabic, trans. John Payne, 9 vols (London: Villon Society, 1882–4). A Plain and Literal Translation of the Arabian Nights’ Entertainments, Now Entitled the Book of the Thousand Nights and a Night; with Introduction, Explanatory Notes on the Manners and Customs of Moslem Men and a Terminal Essay upon the History of the Nights, trans. Richard Burton, 10 vols, 7 supplementary vols (Benares: Kamashastra Society, 1885–8). Le Livre des Mille Nuits et Une Nuit, trans. Jean-Charles Mardrus, 16 vols (Paris: Éditions de la Revue Blanche, 1899–1904). Die Erzählungen aus den Tausend und ein Nächten: vollständige deutsche Ausgabe in sechs Bänden, zum ersten Mal nach dem arabischen Urtext der Calcuttaer Ausgabe vom Jahre 1839 übertragen, trans. Enno Littmann, 6 vols (Leipzig: Insel, 1921–8). The Book of the Thousand Nights and One Night. Rendered from the literal and complete version of Dr. J. C. Mardrus, and collated with other sources, ed. E. Powys Mathers, 4 vols (London: The Casanova Society, 1923). Neue Erzählungen aus den tausendundein Nächten. Die in anderen Versionen von ‘1001 Nacht’ nicht enthaltenen Geschichten der Wortley-Montague-Handschrift der Oxforder Bodleian Library. Aus dem arabischen Urtext übertragen und erläutert, trans. Felix Tauer (Frankfurt am Main: Insel Verlag, 1995). The Arabian Nights, trans. Husain Haddawy (New York: W. W. Norton, 1995). The Arabian Nights: Tales of 1001 Nights, trans. Malcolm C. Lyons, Robert Irwin and Ursula Lyons, 3 vols (London: Penguin, 2008). Tausendundeine Nacht. Nach der ältesten arabischen Handschrift in der Ausgabe von Muhsin Mahdi erstmals ins Deutsche übertragen, trans. Claudia Ott (Munich: C. H. Beck, 2011). Tausendundeine Nacht. Das glückliche Ende, trans. C. Ott (Munich: C. H. Beck, 2016).

88

Biblioteca Apostolica Vaticana (Vatican) Bibliothèque générale de Philosophie et Lettres (Liège)

Berlin Staatsbibliothek

Bangladesh National Museum Library (Dacca) Bayerische Staatsbibliothek (Munich)

al-Math.af al-Wat.anī (Baghdad) Arcadian Library (London)

al-Azhar (Cairo)

10th c. 14th c.

Extant Arabic manuscripts of 1001 Nights92

782 [264 fols]

15th c.

16th c.

23220 [30 octavo notebooks]

17th c.

Appendix II

X

19th c.

2241C [422 fols] (Continued)

623–6, 4 vols* [498/339/316/340 fols] 627–8, 2 vols 629 Pett. 109 [40 fols] Do183–4, 2 vols [105/17 fols] We.521 [60 fols] We.663 [100 fols] We.662 [132 fols], dated 1759 We.1082 [50 fols] We.601 [197 fols] 778–81 (182/186 fols]

. āSd9483/ʿayn133413 adab, dated 1719 849 [260 fols] 1413

18th c.

89

Bibliothèque universitaire des langues et civilizations (Paris) Bodleian Library (Oxford)

Bibliothèque Nationale et Universitaire (Strasbourg)

Bibliothèque Nationale de France (Paris)

10th c. 14th c. Arabe 3609–11 [70/67/79 fols]

15th c.

16th c. Arabe 3612 [408 fols] Arabe 3619 [176 fols] Arabe 3620 [99 fols] Arabe 3621 [42 fols] Arabe 3635

17th c.

19th c. 3595–7, 3 vols* [298/479/343 fols] 3598–3601, 4 vols* [531/334/367/391 fols] 3602–5, 4 vols* [481/ 408/414/245 fols] 3606–8, 3 vols [548/277/416 fols] 3617 (94 fols] 4675–7, 3 vols [489/305/300 fols] 4678–9, 2 vols 5112 [300 fols] 6919 4278–81, 4 vols* [570/564/608/478 fols], dated 1831–2 4282–3, 2 vols [333/399 fols], dated 1867 4287 [292 fols] 4288 [222 fols] 4289 [20 fols] 495 [(263 fols] Arab f.76 [21 fols]

18th c. Arabe 3613–16, 4 vols [120/130/231/320 fols] Arabe 3617 (94 fols] Arabe 3618 [237 fols], dated 1799 Arabe 3634 [117 fols] Arabe 3644 [137 fols], dated 1784

Or 550–6, dated 1764 Ouseley242 [195 fols)

90

Dār al-Kutub al-Wat.aniyya (Tunis)

Cambridge University St John’s College Chicago University Oriental 17618, Institute dated 879 Dār al-Kutub wa ’l-Wathā’iq al-Qawmiyya (Cairo)

Cambridge University Library

British Library (London)

10th c. 14th c.

15th c.

16th c. ADD (SL )2644 [31 fols] ADD 7404 [186 fols]

17th c.

(Continued)

149 adab m [50 fols] 316 z [272 fols] 12188 z [399 fols] 13523 z, 4 vols* [427/343/342/359 fols], dated 1809 22803 [40 fols]

ADD 7407 [407 fols] DEL AR 1308 [355 fols] ISO ISL .444 [128 fols] ISO ISL .3678 [130 fols] ISO ISL .3722 [149 fols] Or.1595–8, 4 vols* [440/371/333/369 fols], dated 1829 Or.2916–9, 4 vols* [534/399/414/428 fols] Or.4699 [40 fols] Or.5312 [251 fols] Qq106–9*, 4 vols [495/301/300/340 fols] 1432, 2 vols [234/163 fols]

ADD 7405 [246 fols] ADD 7406 [225 fols] ADD 23468 [246 fols] IO ISL 2699 [389 fols] ADD 7412 [232 fols]

Or.1762 Or.1763

19th c.

18th c.

91

Institut français d’Archéologie orientale (Cairo) John Rylands Library (Manchester) Khuda Bakhsh Oriental Public Library (Patna) King Faisal Center for Research and Islamic Studies (Riyad) Leipzig Universitätsbibliothek Library of the Holy Sepulchre (Jerusalem) Maktabat al-Asad alWat.aniyya (Damascus) Oxford University Christ Church College

Edinburgh University Library Geneva (Bibliothèque de Genève) Gotha Forschungsbibliothek

10th c. 14th c.

KFCRIS 2415 [76 fols]

15th c.

16th c.

6728 [42 fols]

647 [40] [229 fols]

Orient A.2637 [100 fols] Orient A.2638 [198 fols], dated 1759

Or.1695 [209 fols]

18th c.

207 [144 fols] 206 [96 fols], dated 1724

646 [706] [263 fols]

17th c.

12423 [40 fols]

636, 2 vols [71/131 fols] 157

2637–40, 4 vols* [449/370/332/368] KFCRIS 1855 [94 fols]

Orient 2632–5*, 4 vols [402/397/416/369 fols] Orient A.2636 [50 fols] Orient A.2639 [60 fols] 18, 4 vols* [515/484/440/430 fols], dated 1809

o.22 a [19 fols]

19th c.

92

Wizārat al-Turāth wa ’l-Thaqāfa (Muscat) Wizārat al-Awqāf al-Mis.riyya (Cairo) Yale University Beinecke Library TOTALS [105]: 1

St Petersburg Institute of Oriental Manuscripts Tübingen University Library

Raşit Efendi Yazma Eser Kütüphanesi (Kayseri) Real Academia de la Historia (Madrid) Royal Asiatic Society (London) St Petersburg Public Library

1

M.a.VI .33 [85 fols]

10th c. 14th c.

3

15th c.

1

614 [152 fols]

16th c.

Gayangos 49

18th c.

13

32

Arabic 272 [221 fols]

M.a.VI .32 M.a.VI .34 [61 fols] [210 fols] M.a.VI .37 [58 fols]

17th c.

54

1668 (184 fols), dated 1813

1362 [433 fols], dated 1863

RAS 44 [168 fols] ANS 355, 3 vols [303/300/310 fols] B 114, 4 vols* [494/338/326/382 fols]

19th c.

4

A Text in Exile: Dante’s Divine Comedy Annalisa Cipollone

Not a single line written by Dante (1265–1321) has survived, not even a signature. The analogous case of Shakespeare, though much later and only slightly more fortunate – six signatures and possibly some three pages from a co-authored manuscript play – is often mentioned as a suitable comparison. Two of the most eminent authors of world literature share an unenviable dearth of autograph material. Dante’s handwriting was last observed in the early fifteenth century when Leonardo Bruni, quoting from a subsequently lost letter, described him as an accomplished practitioner of the pen (‘fu ancora scrittore perfetto’), capable of producing a thin, elongated, very correct script (‘et era la lettera sua magra et lunga et molto corretta’).1 Since that time, scholars have never given up hope of being able to one day recognize Dante’s hand. Their wildest dream has always been the recovery of an autograph exemplar of the Commedia from the dust of some attic, cellar, remote mansion or forgotten library shelf – or even, as a novelist imagined a few years ago, the oblivious mud of the Venetian lagoon.2 That hope should spring eternal in matters of this kind is understandable. Autograph documentation is ordinarily, though not necessarily, regarded as the most valuable currency in the assessment of literary evidence. It may seem strange that no Dante autograph should have survived or reemerged. As a prominent member of the Florentine Commune, his presence at assemblies and the words he pronounced were kept on record in the minutes of Florence and other city councils.3 He operated in the context of what was arguably, in many respects, the most advanced city of the Western world, where levels of literacy and written production were so high that they remained unmatched across Europe until the eighteenth century.4 However, Dante was also an exile for the last eighteen years of his life, the years during which he wrote the Commedia and most of his other works; he suffered the confiscation 93

94

Life of Texts

of his property, received a death sentence in absentia and was possibly made the target of some form of damnatio memoriae, the consequences of which would have been considerable in a city such as Florence, desperately torn by opposing political factions. It was only the posthumous dissemination of the poet’s fame that induced the Florentines to regret their actions against their great citizen. By that point crucial damage had presumably already been done, either by the elimination of archival documents which the authorities might periodically consider to be of reduced or no interest or, most likely, by accidental fires – actions which will have removed most or all residual traces of Dante’s hand, together with any effective means of enabling posterity to recognize it. The transmission of Dante’s literary works is similarly affected by a lack of indisputably authoritative textual witnesses. This has usually been ascribed to the vicissitudes of the author, which contributed to shaping the textual tradition of his oeuvre and notably of his Commedia. The poem’s textual life, in other words – perhaps in a more pronounced manner than in other cases in Western literature – appears to be tied up with the events of its author’s life, which for the most part was the life of an exile. The poem itself can thus be described as a text in exile. As to the significance of exile on its content, it would be pointless to bring owls to Athens and repeat what has already been said in hundreds of contributions. Yet, as I hope to be able to show, one or two aspects still deserve to be revisited, and the notion of the exiled text may help us to reassess the contentious issue of the title of Dante’s poem.

A Very Short Life of Dante Domenico di Michelino’s portrait of Dante in the left nave of Santa Maria del Fiore in Florence (Fig. 4.1) was painted in 1465 to celebrate the second centenary of Dante’s birth. (Centenary celebrations are not, as is often maintained, an eighteenth-century invention.) The painting shows the poet as a pensive, melancholy man holding an open book with a road winding through an unusual landscape behind him. On his right-hand side demonic figures carry damned souls into the belly of the earth; in the background is a depiction of Mount Purgatory with Paradise at the top; and above him is a representation of the cosmos, divided into bands, hosting the planets and the Sun according to the Ptolemaic system. On Dante’s left-hand side is a walled city with a closed gate, to

A Text in Exile: Dante’s Divine Comedy

95

Figure 4.1 Domenico di Michelino, Portrait of Dante (1465). Florence, Santa Maria del Fiore.

which he directs his gaze. This is clearly Florence, easily recognizable from the tower of Palazzo Vecchio, the belfries of Santa Maria del Fiore and the Church of Badia, and Brunelleschi’s dome built a few decades before (1436) over the cathedral where the painting itself is located. The open book shows the first lines of a poem in capital letters: Nel mezzo del cammin di nostra vita . . . This visualized biography captures the salient features of Dante’s life and achievement: his Florentine origin, his exile from his home city and his wanderings through Italy, and his masterpiece, the Commedia, represented by the open book and alluded to in the three realms of the otherworld painted on the poet’s right-hand side. The closed gate in particular reminds the viewer that Dante never re-entered Florence, and the inscription beneath the painting suggests a sense of atonement on the part of the city: Qui coelum cecinit, mediumque imumque tribunal, Lustravitque animo cuncta poeta suo Doctus adest Dantes, sua quem Florentia saepe Sensit consiliis ac pietate patrem.

96

Life of Texts Nil potuit tanto mors saeva nocere poetae Quem vivum virtus, carmen, imago facit. (He who sang of Heaven and of the midway and the abysmal tribunal and all surveyed in spirit is here, Dante the learned poet, whom his Florence often held as father for his wisdom and devotion. Cruel death could bring no harm to a famed poet like him: his worth, his verse, his effigy give him life. trans. E. H. Plumptre, adapted5

Dante Alighieri was indeed a Florentine. While his family did not feature in the highest social ranks,6 it was sufficiently prominent to take advantage of the change in legislation that restored the aristocracy to power in Florence in 1295, provided its members enrolled in one of the major guilds. As a young man with lofty aspirations, Dante cultivated the practice of writing love poetry in praise of a young woman named Beatrice, identified by posterity as the daughter of Folco Portinari. Dante sang of her ethereal beauty and premature death in the Vita nova (‘New Life’), his first important work. On coming of age, Dante was called to participate in the political and military life of his city. He fought on horseback against the Aretines at the battle of Campaldino (1289), and his political career culminated in his election as one of the six Priori – the senior executive body of the Florentine Commune – for two months in the summer of 1300. This office was the origin of all his troubles. In order to oppose the intrusion of Pope Boniface VIII into the politics of the city, Dante and other Florentine representatives were sent on a diplomatic mission to Rome in 1301. The mission was unsuccessful, and in a sudden reversal of fortunes the government of Florence fell into the hands of the enemy faction. According to some of his earliest biographers, Dante learnt in January 1302 on his way back to Florence (or while temporarily in Florence, according to others) that he had been investigated for the crime of barratry (corruption) under false pretences, found guilty, condemned to pay a fine and expelled from Florence for two years in retaliation for his fierce opposition to the Pope. Exile was subsequently extended to capital punishment if Dante was ever found within Florence’s walls, as duly recorded in the register of sentences known as the ‘Libro del Chiodo’ (‘Book of the Nail’).7 From then on, Dante’s life was that of a banished man. Initially, he conspired with fellow citizens to re-enter the city by force; when, however, a military expedition in 1304 ended in disaster, Dante decided to ‘make himself a party by himself ’ (Par. 17.69). He thus began wandering Central and Northern Italy in search of places to reside (North-West Tuscany, Pisa, Verona, and eventually

A Text in Exile: Dante’s Divine Comedy

97

Ravenna), while proving himself to be a powerful lyric poet (Rime) and political thinker (De monarchia), as well as a literary theoretician of self-imposed very high standards – hence, presumably, his failure to complete the two groundbreaking works of a treatise on vernacular language and style (De vulgari eloquentia) and a philosophical commentary on a selection of his doctrinal poetry (Convivio). While an early source claims that Dante had begun his Commedia in Florence where he left the first seven cantos of Inferno (allegedly retrieving them from exile), it was probably around 1306 that he properly embarked on the project of writing a long poem in the vernacular. Prodigiously varied in content, language and style, the poem offered a judgmental vision of the world as seen from the vantage point of the realms of the damned, the purging and the saved souls. As he himself wrote when he had nearly accomplished his great enterprise, the great poem could claim to be ‘shared by heaven and by earth’ (Par. 25.1–2). Sometime between 1316 and 1320, Dante dedicated the last canticle, Paradiso, to the ruler of Verona, Cangrande della Scala, with an epistle explaining how his work should be interpreted. By then Dante had found shelter in Ravenna, where he died on the night of 13–14 September 1321, presumably of malaria contracted while travelling back from a mission to Venice. His demise was so sudden that only a provisional burial could be arranged and, in the general turmoil, no complete text of the Commedia was found – allegedly, only one copy lacking the last thirteen canti of Paradiso; but one night Dante’s son Jacopo had a vision of his father’s ghost telling him to look in a niche in the wall, where the missing canti were duly found. In subsequent months the city’s ruler, Guido Novello da Polenta, was elected Capitano del popolo in Bologna and left Ravenna, from where he – and presumably Dante’s sons – were banned a few months later after a coup orchestrated by Guido’s cousin Ostasio.8

(How Far Can) Literary Texts (Count) as Evidence? The above account is a mixture of documentary evidence, unverified sources and scholarly interpretation. To sieve facts from fiction should be the fundamental task of a scholar, yet this appears to be exceptionally difficult in the case of Dante and his Commedia. Seven centuries stand between us and the events rapidly sketched above. The amount of scholarship produced in the intervening period has an overwhelming effect on whoever dares to approach the matter, and a good deal of it can be as misleading as it is helpful.

98

Life of Texts

Most of the information on Dante summarized above depends on sources which are questionable at best; among them, the name of Giovanni Boccaccio (1313–75) stands supreme. Boccaccio, over and above being one of the greatest storytellers in world literature and a highly influential author in many other literary departments, was the leading Dante scholar of his generation and the first to deserve being called such. However, Boccaccio’s narrative genius must have crept into his account of Dante’s life, since he is responsible for divulging some of the most popular stories about the author he so admired. The partisan point of view that informs his successful biography – eulogy, in fact – of Dante is declared in the longish title of the work’s first redaction: On the birth, life, studies and mores of the most learned Dante Alighieri, the illustrious Florentine poet, and on the works composed by the same (De origine vita studiis et moribus viri clarissimi Dantis Aligerii Florentini poetae illustris et de operibus compositis ab eodem), more commonly cited as A Short Treatise in Praise of Dante (Trattatello in laude di Dante).9 The fabulous story of the missing canti found by Jacopo Alighieri, for example, first occurs in the Trattatello and would require a true leap of faith to be believed. Other of Boccaccio’s statements may look less fanciful, but are by no means more authoritative. That Beatrice be identifiable as the daughter of Folco Portinari is an assertion for which no supporting evidence survives. The claim that Dante had to interrupt his Commedia because he had left the first seven canti of Inferno in Florence is equally unattested anywhere else, and may simply follow from Boccaccio understanding the beginning of Inf. 8 (‘I say, continuing . . .’, ‘Io dico seguitando . . .’) as evidence of an actual interruption in the poem’s composition. Boccaccio also provided a detailed physical description of Dante without ever actually meeting the man.10 It could be claimed that we have Dante’s own words to illuminate his great work. In Paradiso’s dedicatory letter to Cangrande della Scala, Dante offers an explanation of the poem’s title and a fourfold approach for a proper understanding of its contents. Yet, this letter still hangs precariously between those who swear to its authenticity and those who assert the irreducibility of its disappointing content (notably in the second part) to Dante’s mind, to the extent that it has been suspected of being a forgery, at least in part.11 Even those words of Dante’s which are above suspicion may not be as helpful as one would expect. The power of his literary skills is such that the meaning of historical events is transfigured in his readers’ eyes to fit the Commedia’s ideological design and purpose, the extent of which we are still unable to appreciate in its entirety. ‘Dante war ein großer Mystificator’ is Ernst Robert

A Text in Exile: Dante’s Divine Comedy

99

Curtius’s famous verdict on Dante’s reticent and manipulative habits, a verdict reportedly pronounced with an admonishing raised finger.12 By characterizing Dante as a clever conjuror, Curtius implicitly acknowledged the poet’s consummate art in directing the way in which his readers should approach the narrated facts, and react accordingly. The circumstances of Dante’s trial and subsequent exile, later extended to capital punishment for him and his sons, are a good case in point. Dante is clever enough not to make precise statements about it; he merely hints repeatedly at unjust sentencing. Plotting and unfair procedures, of which he presents himself as the sole target, are described as the persecutory reaction of perverted justice to his righteous and uncompromising conduct. In one of his letters (Epist. 3.1) he portrays himself as ‘an undeserving exile’ (exul immeritus). In the Commedia he lets prominent characters confirm the surreptitious circumstances of the accusation, as in the case, for example, of his mentor Brunetto Latini (Inf. 15.55– 69). The heart-rending account of Count Ugolino, unjustly condemned to capital punishment together with his sons and nephews (Inf. 33.37–75, 85–90), is clearly meant to mirror – admittedly in an intensified manner – the equally unfair condemnation of Dante and his sons. Among the poem’s many memorable passages, one is paramount for its vibrant, moving, irresistible eloquence, where the poet’s pride is entwined with his unappeasable desire to return to ‘the fair abode where I used to sleep as a young lamb’, and where ‘the wolves’ who had declared war on him still oppose his legitimate aspiration: Se mai continga che ’l poema sacro al quale ha posto mano e cielo e terra, sì che m’ha fatto per molti anni macro, vinca la crudeltà che fuor mi serra del bello ovile ov’io dormi’ agnello, nimico ai lupi che li danno guerra; con altra voce omai, con altro vello ritornerò poeta, e in sul fonte del mio battesmo prenderò ’l cappello . . . Par. 25.1–913 (If it should happen . . . If this sacred poem – / this work so shared by heaven and by earth / that it has made me lean through these long years – / can ever overcome the cruelty / that bars me from the fair fold where I slept, / a lamb opposed to wolves that war on it, / by then with other voice, with other fleece, / I shall return as poet and put on, / at my baptismal font, the laurel crown . . .). trans. Allen Mandelbaum

100

Life of Texts

How can one resist such eloquence and expressive pathos? Boccaccio did not hesitate to describe Dante as the victim of ‘false charges’.14 Leonardo Bruni similarly harboured no doubt that the measures taken against Dante had been ‘iniquitous and perverted’.15 On the other hand, much as we would like to acknowledge Dante as the victimized champion of liberty and honesty and be outraged by a document condemning him to death, we must consider that the procedure against him was far from unusual in his time.16 Many others were investigated and tried by the same process in an attempt on the part of the Commune to deter and control any malfeasance perpetrated by public officers.17 A closer examination of the document registering Dante’s condemnation, the aforementioned Book of the Nail, helps to illustrate this point. With Dante’s verse resonating in their ears, readers may fail to notice that Dante’s name is in no visible way singled out, but is actually eleventh in a list of fifteen convicted people. No surviving evidence suggests that Dante was the sole target of political envy, as Brunetto Latini is made to say at Inf. 15.55–69; nor was the initial investigation anything other than the norm in Florence. Every person who had served as priore was subjected to the same procedure and ensuing measures, intended to have retroactive force. While it is fair to assume that the charges would often have had a political motivation, the procedure had in effect been devised as a tool to suppress episodes of general misgovernance. The same purpose was served by the decision to drastically limit the time during which the priori were allowed to remain in charge to two months. More can be said about the actual significance of this evidence, as official documentation is also subject to the vagaries of textual transmission. The original document from 1302 is lost; copies of the sentences passed in 1302 have been preserved in two registers, the former compiled c. 1348–54, the latter – the Book of the Nail, a copy of the former – c. 1358, with later additions.18 Giuliano Milani has recently re-examined the process whereby Dante’s condemnation acquired a more distinctive political colouring. In 1302 the sentence was transcribed in a liber condempnationum (‘register of condemnations’), where no distinction was made between charges of differing natures. A few months later, in the heat of the strife between the rulers’ party and that of the exiles, those sentences which could be exploited for political purposes were culled from the original register and collected in a separate volume (also lost). Another selection ensued, which resulted in the two surviving documents (late 1340s and late 1350s) listing the names of the sentenced culprits, including Dante’s. These transcriptions were prompted by a further political motivation: they aimed at determining the Ghibelline faith of politicians active in mid-fourteenth-century

A Text in Exile: Dante’s Divine Comedy

101

Florence by looking at the political convictions of their ancestors.19 By then, some thirty years had elapsed since Dante’s death. Thus, doubts must be cast over Dante’s account of events. To recall Curtius’s admonition, one needs to measure the extent of Dante’s mystifying powers. When this is done, not only does his case lose the aura of uniqueness which his poetical talent had managed to conjure, but the documents attesting to his condemnation must be assessed in light of the changing political and cultural contexts in which their successive versions appeared.

The Textual Tradition of the Commedia As we have seen, textual transmission plays an important role in determining the value of historical documents concerning Dante’s life. Its importance is even greater, as one would expect, in the case of his literary works and the Commedia in particular. In the absence of autograph exemplars, the question is how far one can trust texts that have been transmitted by witnesses of diminished authority. The known copies of the Commedia are at several removes from the original, which means they can only offer as reliable a text as scribal accuracy will allow.20 We have indirect evidence that the Commedia was well known outside Florence during Dante’s lifetime. Francesco da Barberino’s Officiolum (1306–9?), the work long known to have existed but only recently recovered of a Tuscan notary in exile, shows in its illuminations the influence a great poem still in the making could exercise over medieval visual lore.21 Another of Francesco’s works, Documenti d’amore (Teachings on Love, 1313–14), explicitly refers to Dante’s Commedia and emphasizes the presence in it of infernal scenes ‘among many other things’.22 Hunc [Virgilium] Dante Arigherij in quodam suo opere quod dicitur comedia et de infernalibus inter cetera multa tractat, commendat protinus ut magistrum, et certe siquis illud opus bene conspiciat videre poterit ipsum Dantem super ipsum Virgilium vel longo tempore studuisse vel in parvo tempore plurimum profecisse. [My emphasis] (Dante Alighieri, in a work of his entitled comedia dealing with Hell among many other things, praises Virgil as his teacher; in fact, if we closely inspect his work, we can see how he spent a long time studying Virgil, or that he was able to get the most out of him in a very short time.)

102

Life of Texts

Both works testify to the circulation of parts of Dante’s Commedia before its completion. The first two canticles, Inferno and Purgatorio, were presumably allowed to circulate autonomously, although no surviving manuscript shows any detectable sign of this partial diffusion of the text.23 Further proof of an early circulation of the Commedia comes from the pen of a Tuscan notary active in Bologna, Ser Pieri degli Useppi da San Gimignano, who in 1317 copied three lines from Inferno (3.94–6) in the official registers of the Bolognese Commune (Registri della Curia del Podestà). These registers, known as Memoriali bolognesi, represent a unique phenomenon in the history of textual transmission. They came into existence in 1265 on the initiative of two local podestà, Loderingo degli Andalò and Catalano de’ Malavolti; every deed drafted and ratified in Bologna was copied in them.24 In order to avoid interpolations, any unused space left below the transcribed deeds was filled with texts of a disparate nature – mostly popular songs and ditties of which the Memoriali are often sole witnesses, as well as short lyric poems and passages from longer compositions. Dante’s works had come to the attention of the notaries of Bologna as early as 1287, when some of his lyric poems were copied in the registers. This makes the Memoriali the earliest extant witnesses of Dante’s lyric poetry, as well as of the Commedia.25 At this point we encounter a gap in the poem’s tradition. The first generation of manuscripts issued during Dante’s lifetime or directly after his death appears to have been completely destroyed. We can assume that these were exemplars with only a limited number of errors, perhaps still preserving traces of direct authorial intervention. The earliest commentaries appeared in the period immediately following Dante’s death, 1322–4, but only survive in later copies; the snippets of text transcribed at the beginning of each entry overall reveal lections of no great quality.26 The most ancient surviving manuscript of the entire Commedia with an ascertained transcription date (1336) was produced outside Tuscany, in Genoa, by a scribe from Fermo in the Marche on behalf of a jurist from Pavia.27 However, by that point, Tuscany and Florence had already awakened to the need to reclaim primacy in the transmission of the Commedia. As things stand, bar a few residual doubts concerning the hand that traced ‘in August 1335’ (‘dogosto 1335’) in its last folio, MS Florence, Ashburnham 828 (from West Tuscany) can claim chronological priority over the rest of the tradition. Another manuscript, the colophon of which showed it had been commissioned and copied in Florence in the autumn and winter of 1330–1, is unfortunately lost, but the colophon and the manuscript’s peculiar readings were scrupulously transcribed by Luca Martini, a

A Text in Exile: Dante’s Divine Comedy

103

Florentine philologist, over two centuries later.28 In 1337, Francesco di ser Nardo da Barberino in Val di Pesa completed MS Milan, Trivulziano 1080. On account of its codicological, scriptural and morphological homogeneity, MS Trivulziano 1080 can be considered one of the best representatives of the early Florentine group.29 Francesco managed an active scribal workshop in Florence, which played an important role in catering for socially and culturally diverse readerships. Boccaccio’s arrival from Naples in the early 1340s gave Florence further authority as the main centre of reproduction for Dante’s work: three copies of the Commedia in Boccaccio’s own hand survive (MSS Toledo, Zelada 104 6; Florence, Riccardiano 1035; Vatican City, Chigi L VI 213); a fourth and earlier copy from West Tuscany (Vat. Lat. 3199) is that donated by Boccaccio to his friend Petrarch.30 Reproduction, of an unprecedented rate for a vernacular work, was mainly the product of dedicated workshops set up to satisfy the high demand from a rapidly widening readership. It continued without intermission until the invention of the printing press and beyond, with over 800 existing manuscripts of Dante’s Commedia identified to this day – the envy of any best-seller in any age.31 The authority of Florentine and Tuscan Dante manuscripts remained unchallenged for centuries, almost to the present. Apart from the fact that Dante was a Florentine and that it was a common assumption that the language of the best Commedia manuscript should present discernible Florentine features, any critical revision of the text would have been unthinkable without tapping into the city’s virtually inexhaustible repositories. This did not prevent other centres of learning from rising to the role of protagonists in the field of textual studies, especially after the advent of the printing press. The Aldine edition of 1502, edited by the Venetian Pietro Bembo, became the standard Dante text during the Renaissance.32 At the end of the eighteenth century, a new critical text by the Veronese Gian Iacopo Dionisi received the honour of being published by another great printer, Giovan Battista Bodoni of Parma (17951, 17962).33 In 1862 in Berlin, Carl Witte published the first modern critical edition of the Commedia on the basis of four main witnesses after extensive collations of nearly 400 manuscripts.34 Following in Witte’s footsteps, while expanding the range of collations and refining the recension technique, Edward Moore produced the authoritative ‘Oxford Text’ in 1894.35 For all these scholars, despite their non-Florentine origin and place of residence, it was the Florentine and Tuscan traditions that provided the standard against which their texts were to be prepared for the press. Bembo essentially relied on MS Vat. Lat. 3199 (West Tuscany) and the Florentine edition by Cristoforo Landino (1481); both Dionisi and Witte elected as their copy-text – with some opportune ameliorations – MS Laur. 26 sin. 1, a singularly accurate

104

Life of Texts

witness issued in the late fourteenth century from the circle of Coluccio Salutati; Moore essentially introduced improvements to Witte’s text.36 Furthermore, the first attempt at a critical edition of the Commedia was published by the academicians of the Florentine Accademia della Crusca in 1595, a landmark in the history of collaborative scholarship.37 The twentieth-century editorial history of the Commedia confirmed the primacy of Florence. Building on the foundations laid by Michele Barbi, the greatest Dante scholar of the past century, the editions by Giuseppe Vandelli (1921), Mario Casella (1923), Giorgio Petrocchi (1966) and, more radically and controversially, Antonio Lanza (19951, 19962), present texts which are firmly rooted in the Florentine tradition.38 Are there any drawbacks to this situation? Despite the poem’s length (14,233 lines), the tradition of the Commedia is characterized by very limited textual variation. This isn’t necessarily a reassuring fact. The dominance of Florentine manuscripts since the fourteenth century has tended to homogenize the text by ironing out a number of textual ‘difficulties’ using the much-feared practice of contamination. Parodying an old medieval proverb, Paul Maas declared that ‘no specific has yet been discovered against contamination’ (‘Gegen die Kontamination ist noch kein Kraut gewachsen’).39 Contamination appears to be endemic in the Florentine tradition of the Commedia, so much so that standard stemmatic procedures cannot lead to any reliable conclusion. It follows that manuscript recension must go hand in hand with interpretation from the outset, and the textual critic must work through the text line by line, comparatively assessing competing variant readings as they appear. There is no doubt that such a method – which is ultimately the method used by the humanists – offers the chance to ameliorate the text of selected critical passages, or loci critici, which is a valuable outcome in itself and can have great methodological and pedagogic merit.40 The end result, however, is likely to be a hybrid text, vulnerable to the legitimate criticism of those for whom it is an artificial construct that never existed and was therefore never read by anybody.

Competing Variant Readings In an overall stable tradition such as that of the Commedia, critics are generally confronted with textual alterations which can barely be classified as errors. Gaps produced by involuntary omissions of portions of text, for example, are rare in such a close-knit fabric as that of the Commedia, whose metrical structure is itself a guarantee against accidental textual losses.

A Text in Exile: Dante’s Divine Comedy

105

From the surviving evidence, it is clear that most Florentine and Tuscan copyists took an active approach to the transcription of their text; they made a constant effort to produce sense, at least at the level of word and sentence meaning. Since most of those copyists worked in or for highly organized workshops, where multiple copies were simultaneously being prepared and assembled, textual homogenization through contamination became the norm and banalization ensued as a direct consequence. The tendency to remove or remedy difficulties by resorting to ‘smoother’ readings is always a major threat to the survival of the genuine lection, which is often not immediately intelligible in texts of complex and sustained style. The treacherous nature of more readable variants presents them as if they did not compromise the text; if anything, they often seem to improve it by making its meaning more immediately accessible. The three examples that follow show different degrees of certitude which the textual critic may hope to attain in a text founded on the Florentine tradition, as is the one edited by Giorgio Petrocchi in 1966. At Par. 29.100 two readings differ by one single consonant. One text reads mentre (‘while’, ‘as’), the other mente (‘[he] is lying’). In the following quotations the relevant portions of text are in italics. Un dice che la luna si ritorse nella passion di Cristo e s’interpose per che ’l lume del sol giù non si porse, e mente, ché [var. mentre che] la luce si nascose da sé . . . Par. 29.97–101 (One [preacher] says the moon changed place during the Passion of Christ and blocked the light of the sun, and he is lying, because the light vanished of its own accord . . . [var. ‘and while the light vanished of its own accord . . .’])

Both versions may at first glance look acceptable, although the reading mentre (‘while’, ‘as’) dominates the tradition with only few exceptions. Punctuation would have helped, of course, but fourteenth-century poetry manuscripts are very seldom graced with detailed and unequivocal punctuation systems.41 External evidence can, however, direct the reader towards a probable solution. The version with mente (‘he is lying’) entails a powerful statement: a preacher is accused of being a liar. As Paolo Trovato has recently observed, Bruno Nardi demonstrated that Dante is here rephrasing Peter Comestor’s Historia scholastica:

106

Life of Texts

Non fuit eclipsis solis, ut quidam mentiti sunt quod luna et regione fere est ad solem . . .42 (It was not a solar eclipse, as some pretended [mentiti sunt] on the assumption of the moon’s proximity to the sun . . .)

Dante is clearly taking sides on the causes of the wonderful darkness enveloping the land at the time of Christ’s death on the cross (Mt. 27.45, Mk 15.33, Lk. 23.44–5). The alternative variant mentre (‘while’) annuls all this; despite its presence in the vast majority of manuscripts, it is a patent trivialization of the text and therefore bears no authority. Another revealing example of competing variant readings occurs in the final part of the ascension to Mount Purgatory and the Garden of Eden (canto 30). Virgil has reached the end of his journey as Dante’s guide; as a pagan, albeit a righteous one, he cannot enter Paradise proper, much to Dante’s disbelief and momentary despair. Virgil’s role will henceforward be taken by Beatrice. In canto 27 Virgil had already given his pupil warning (although the latter had not seemed to register it) that his mission was about to be accomplished: the will and judgement of Dante-the-character were, according to his guide, both sound and freed of sin. E disse: ‘Il temporal foco e l’etterno veduto hai, figlio; e se’ venuto in parte dov’io per me più oltre non discerno. ... Non aspettar mio dir più né mio cenno; libero, dritto e sano è tuo arbitrio, e fallo fora non fare a suo senno: per ch’io te sovra te [var. me] corono e mitrio.’ Purg. 27.127–42 (He said: ‘My son, you’ve seen the temporary fire / and the eternal fire; you have reached / the place past which my powers cannot see. . . . Await no further word or sign from me: / your will is free, erect, and whole-to act / against that will would be to err: therefore / I crown and mitre you over yourself’ [var. myself ].)

The alternative reading concerns the second pronoun in the final verse: is it ‘you over yourself ’ (‘te sovra te’), which is dominant in the poem’s best tradition, or ‘you over myself ’ (‘te sovra me’)? Given the circumstances, Virgil may well be recognizing Dante’s superiority over himself both as a poet and a Christian – hence the crown and the mitre. On the other hand, if the mitre is supposed to be a symbol of spiritual faith (as some early commentators suggested), wouldn’t the

A Text in Exile: Dante’s Divine Comedy

107

pagan Virgil be an unsuitable authority to bestow such powers?43 The problem was raised as early as the 1320s, when Iacomo della Lana gave it full consideration in his commentary and eventually inclined towards the second solution (‘you over myself ’), which looks more easily and logically graspable than the other.44 Yet, once again, what looks easier is not necessarily true. Firstly, ‘crown and miter’, as Anna Maria Chiavacci Leonardi noted in her commentary, was an integral part of the formula for the crowning of the Emperor. An early thirteenthcentury chronicle describing Otho IV ’s crowning (1209) reports that ‘once he had been crowned Emperor, . . . Otho proceeded crowned and mitered (mitratus et coronatus) to the gates of Rome with the Pope’.45 Since the expression is formulaic, it does not seem necessary to extract from it any recondite meaning in support of one variant. On the other hand, ‘you over yourself ’ vs ‘you over myself ’ raises a different issue, which is easily forgotten when one engages with this type of microsurgery on the text of the Commedia. Were a similar case to be discussed in the context of a less famous poem by a less famous author, wouldn’t one feel rather tempted to regard ‘you over yourself ’ (‘te sovra te’) as a plain error of repetition – one of the most common in scribal practice – which may have hit the tradition at a very early stage? In other words, aren’t we forcing the issue in demanding that Dante constantly prove himself worthy of himself – or more, if needs be – without even allowing him the occasional nod that Horace granted Homer? Everyone knows the dangers of this kind of approach. When put under a magnifier, every debatable reading grows in size and complexity to an extent that is probably disproportionate to its actual weight. To be sure, editors of the Commedia have a number of good reasons for electing ‘te sovra te’ over ‘te sovra me’: its prevalence in the tradition in the first place, but also its being the more difficult, more daring, more intellectually challenging expression, hence a clear case of lectio difficilior; yet the alternative reading cannot be conclusively ruled out. The third example is taken from Canto 24 of Purgatorio. It concerns the only passage in the Commedia where, invited by fellow poet Bonagiunta da Lucca, Dante makes an explicit statement about his personal poetics: E io a lui: ‘I’ mi son un che, quando Amor mi spira, noto, e a quel modo ch’e’ ditta dentro vo significando’. Purg. 24.52–4 (I answered: ‘I am one who, when Love breathes / in me, takes note; what he, within, dictates, / I, in that way, without, would speak and shape.’)

108

Life of Texts

At which Bonagiunta replies: ‘O frate, issa vegg’ io’, diss’elli, ‘il nodo che ’l Notaro e Guittone e me ritenne di qua dal dolce stil novo ch’i’ odo!’ Purg. 24.55–7 (‘O brother, now I see’, he said, ‘the knot / that kept the Notary, Guittone, and me / short of the sweet new manner that I hear!’)

Three of the major Italian lyric poets from the generation before Dante’s – Giacomo da Lentini, Guittone, and Bonagiunta himself (who as a purging soul is bound to be self-abasing) – are here understood as backward practitioners of a manner, or style, incomparable with Dante’s ‘sweet new manner’ (‘dolce stil novo’). This passage does not seem to have attracted much attention before the second half of the nineteenth century. It was at that point that ‘dolce stil novo’ looked like a historiographical category apt for characterizing the lyric poetry of the young Dante and any poet likely to share his ‘sweet new manner’: Guido Guinizzelli, Guido Cavalcanti, Lapo Gianni, among others.46 In spite of not being explicitly mentioned in the passage under examination, these poets have been called ‘stilnovisti’, and ‘stilnovo’ has provided for the past 150 years a handy formula for capturing a crucial transition phase in the periodization of the early Italian lyric. By contrast, the passage in Purgatorio 24 did not generally catch the attention of early commentators; when it did, it appears to have been something of a riddle to them. Particularly disconcerting is the reaction of Dante’s son Pietro, who in his commentary interpreted ‘ch’ i’ odo’ (‘that I hear’) as ‘chiodo’ (‘nail’), and doggedly tried to extract some sense out of it. After quoting Horace, Geoffrey of Vinsauf and other earlier authorities in order to frame the question within some sort of querelle des anciens et des moderns, Pietro added: . . . ‘dulcem et novum clavum’, in hoc alludens vendentibus cartas Bononie et quaternos cum lineis habentibus certos clavos in se secundum maiorem et minorem commensurationem et formam librorum seu voluminum veterum et novorum.47 (. . . ‘a sweet and new nail’, by this hinting at those in Bologna who sell paper and notebooks [bound] with wooden [plates], with different nails hammered through them depending on the size of the books or if they are old or new.)

Insisting on his bizarre explanation, Pietro went from metaphor to metaphor to forge a logical link between his ‘sweet [style and] new nail’ and Dante’s statement of poetics, with little success.48

A Text in Exile: Dante’s Divine Comedy

109

The debate on this passage was reopened at the turn of the new millennium. On the one hand, the ‘stilnovo’ as a credible historiographical category has been challenged as unsustainable;49 on the other, a new critical edition of the Commedia, provided on an entirely new textual basis by Federico Sanguineti in 2001, has subverted the syntax and meaning of the passage in question and concurrently dissolved the presence of ‘stilnovo’ altogether:50 . . . di qua dal dolce stil! e ’l novo ch’io odo! (. . . short of the sweet manner! and the new one I hear!)

Whether rightly or wrongly, in Sanguineti’s edition the ‘sweet manner’ is severed from ‘new’, because the manuscript he singled out as the most authoritative, MS Vat. Urb. 366, compelled him to do so.51 The passage is once again a riddle to be solved. Even Pietro di Dante’s ‘nail’ has been recently rehabilitated in an ingenious attempt to find a possible solution to the riddle.52 It seems difficult, however, not to regard that ‘loose cannon variant’ (‘variante impazzita’), as it has been aptly described, as a nail destined to seal any commentator’s coffin.53

A Northern Tradition: ‘Recentiores Non Deteriores’ La Commedia secondo l’antica vulgata (1966), edited in four volumes by Giorgio Petrocchi, is the impressive outcome of a great enterprise. It was meant to be the closest approximation to a critically established text, so there is some justification for the current reluctance to abandon it as the poem’s reference text. Petrocchi’s operating principle was to reconstruct the poem’s ‘old vulgate text’ on the basis of 27 manuscripts written before c. 1355. These had been selected as witnesses of an earlier and ‘purer’ Florentine tradition before an irreversibly ‘contaminating’ wave, allegedly instigated by the text edited by Boccaccio, confounded the lines of transmission. Nevertheless, both the premises and results of Petrocchi’s operation have revealed methodological weaknesses of some gravity. Not all the selected manuscripts had actually been written before 1355; the premise that Boccaccio had irreversibly contaminated the poem’s text has turned out to be inaccurate; a number of manuscripts written well after 1355 have proved to be more reliable than some included in Petrocchi’s selection.54 Finally and most importantly, the confident establishment of a critical text on the authority of the best Florentine witnesses has been dismissed in favour of an alternative solution. It was the appearance in 2001 of Federico Sanguineti’s already-mentioned edition that changed the scene. The application of stemmatic procedures,

110

Life of Texts

conducted more systematically and on a far wider scale than ever before, allowed Sanguineti to dramatically reduce the relevant witnesses to eight manuscripts and to identify a non-Florentine exemplar as singularly authoritative. Although written in Romagna in 1352 by a local scribe and coloured with a perceptible dialectal coating, the text in MS Vat. Urb. 366 appeared to be superior, all things considered, to those of the witnesses of Florentine or Tuscan origin. According to the stemma codicum that Sanguineti managed to trace, MS Vat. Urb. 366 is shown to descend directly from the archetype, with no, or at least fewer, intermediate passages than those separating the archetype from the other seven manuscripts (Fig. 4.2).55 Dante’s land of exile, Romagna, was thus giving back what looks like the most reliable text of the Commedia. On the face of it, the solution appears to be a

Figure 4.2 Stemma codicum from Dantis Alagherii Comedia, ed. F. Sanguineti (Florence: Sismel-Edizioni del Galluzzo, 2001), lxv.

A Text in Exile: Dante’s Divine Comedy

111

natural one. Ravenna and the Romagna and, above all, Bologna, on account of the prestige of its university, must have been the first stations in the early phase of the transmission. This is confirmed by the precocious local production of commentaries on the text. The very first commentary, limited to Inferno, was written in Bologna in 1322 – only one year after Dante’s death – by one of his sons, Jacopo. A highly influential commentary by a notary from Bologna operating in Venice, Iacomo della Lana, was written between 1324 and 1328. Another Bolognese notary and chancellor of the Commune, Graziolo de’ Bambaglioli, similarly began a commentary on Inferno in Latin in 1324.56 The primacy of the non-Florentine tradition was also confirmed, almost paradoxically, in terms of linguistic accuracy. Sanguineti has shown that, despite its Romagnole colouring, MS Vat. Urb. 366 has retained a handful of original Tuscan forms which the manuscripts of the ‘old vulgate text’, mostly of Tuscan origin, have lost. All the surviving Tuscan and Florentine manuscripts must therefore have been copied from one or two exemplars coming from the region situated north of the Apennines.57 Sanguineti’s edition is also instructive in another sense. It has brought to the fore one of the tenets of modern textual scholarship – recentiores non deteriores; namely, later manuscripts are not necessarily the worse ones. It is easy to forget that the age of manuscripts is no automatic guarantee of the quality of their texts; what really matters is the number and quality of transcriptions that stand between the surviving copy and the lost original.58 Finally, Sanguineti’s new text is founded on one single manuscript. While the other seven manuscripts may occasionally provide a better reading and may legitimately claim priority over Vat. Urb. 366 where this appears to be corrupt, the danger of producing a hybrid text by cherry-picking variants has been reasonably contained. Sanguineti’s edition caused quite a stir on its appearance and instigated a lively debate among textual scholars, a debate which is far from exhausted.59 A famous critic once wrote that books could be divided into two varieties: those which show something and those which give something.60 Whatever one may think of its results, Sanguineti’s edition has fulfilled both demands. It has shown the way in which a more rigorous methodology can help achieve a genuinely critical text, as well as a way in which his results can be improved on. Among those who have taken up the challenge issued by Sanguineti are Paolo Trovato and his team. Their current goal is an edition of the Commedia according to the Northern, trans-Apenninic tradition, which spread in the area of Romagna and Bologna before, or rather without ever crossing the Apennines and reaching Tuscany and Florence (thus escaping re-Florentinization). The full weight of the

112

Life of Texts

argument may remain accessible only to specialists for a while, in consideration of the hair-splitting questions that need to be preliminarily resolved in the course of the recensio.61 This is the main reason I feel it necessary to conclude my sketchy survey at this point. The evolving plan of action, however, is clear, and the necessary consequences must once again be drawn from the fact that the Commedia was written in exile; that its first transmission was that of a text in exile; and that linguistically the poem was a product issued and circulated for some time outside the walls of Florence – no matter how negligible the differences between the ‘Northern’ and ‘Florentine’ traditions may appear to the untutored eye. Exile permeated Dante’s experience in the minutest perceivable details, which demand to be examined as far as the powers of textual scholarship allow.

‘Outside the walls’: The Title of the Commedia One last reflection about Dante’s exile may invite reconsideration of the issue regarding the poem’s title. As the testimony of Francesco da Barberino shows, the title still in use today, Commedia, was known even before the poem had reached completion. The initial manuscript rubrics, whether in the vernacular or in Latin, read ordinarily ‘Comincia la Commedia’ or ‘Incipit Com(o)edia’. Dante himself calls it comedìa twice:62 ma qui tacer nol posso; e per le note di questa comedìa, lettor, ti giuro . . . Inf. 16.127–8 (but here I can’t be silent; and by the notes / of this my comedy, reader, I swear . . .) altro parlando che la mia comedìa cantar non cura . . . Inf. 21.1–2 (talking of things my comedy is not / concerned to sing . . .)

The term commedia, or comedìa, did not bear in Dante’s time the same meaning and associations it bears for us. Ancient comedy and its modern revival is a dramatic genre characterized by humorous tones and depictions of amusing

A Text in Exile: Dante’s Divine Comedy

113

people and incidents performed on stage, with events moving towards happy conclusions and characters ultimately triumphing over adversities. In the Middle Ages, comedy – and tragedy – had different connotations. The dramatic or performing aspect was conspicuously absent, if not lost altogether. The standard works of reference from the time make it clear that the main concerns were subject matter (cheerful), appropriate characters (ordinary people, only occasionally those of loftier status) and style (lowly, everyday language), the progression of the story (from troubled to victorious) and its ultimate aim (to amuse). There is hardly any discussion of performance, although the names of the ancient playwrights, and possibly some of their works, were known to medieval readers. At De vulgari eloquentia 2.4, written before the Commedia, Dante reports the medieval vulgate opinion: ‘By tragedy I mean the higher style, by comedy the lower, and by elegy the style of the unhappy’.63 The Letter to Cangrande proposes an etymological explanation for the term (‘comedy’ resulting from combining comos/villa ‘village’ with oda/cantus ‘song’ – hence a ‘rustic song’) and suggests that comedy, as opposed to tragedy, commences from some ‘difficult situation’ to reach a ‘prosperous’ conclusion (Epist. 13.28–9: assuming, of course, that the letter is trustworthy).64 Both statements repeat commonplace definitions available in standard medieval encyclopaedias (Isidore of Seville’s Etymologies, Papias’s Elementarium Doctrinae Rudimentum, Uguccione da Pisa’s Derivationes, Giovanni Balbi’s Catholicon). Comedy is a fluid term in medieval literary theory. The extant Latin ‘comedies’ produced in the twelfth and thirteenth centuries are still a matter of debate with regard to the genre they should be ascribed to and the mode of their reception. The term had undoubtedly a wider resonance back then, and its semantic boundaries were larger and possibly more blurred than those in use today. A passage by Isidore of Seville (Etym. 8.7.7) has been indicated as revealing in this context.65 When dealing with the ancient writers of comedy, Isidore makes a distinction between the Old Comedians – Plautus, Accius (in fact, a tragedian), and Terence – and the New Comedians – Horace, Persius, and Juvenal among them, whom we regard today as authors of satires. While the Old Comedians, says Isidore, pursued the amusement of their public as their main goal, the New Comedians were rather concerned with the exposure and reprehension of vices. The characterization of Horace at Inf. 4.89 as ‘Horace, satirist’ (‘Orazio satiro’), with no clear indication that Dante ever read Horace’s Satires, shows that a convergence and overlap between the genres of comedy and satire was a standard assumption in Dante’s time.66

114

Life of Texts

Another aspect worth investigating is offered by a text which is rarely considered. In 1278, William of Moerbeke completed, probably at Viterbo, the Latin translation of Aristotle’s Poetics. Only two manuscripts of his version survive, both written in Italy near the end of the thirteenth century: the text’s circulation may not have been very wide.67 All the same, certain passages help reinforce the point about the wider significance of ‘comedy’; as with Poet. 1453a30, for example, where single works narrating tragic events followed by a happy outcome – such as the Odyssey, as Aristotle/Moerbeke points out – are said to deserve the name of comedies rather than tragedies.68 One passage in particular, close to the beginning of the work, deserves greater attention. This is where Aristotle explains the rival positions of Athenians and Megarians as to the origin of the word ‘comedy’ and reports the latter’s etymological derivation from komai ‘districts’, or ‘unwalled villages’, as implicitly distinct from cities proper. This is clearly the source from which the standard interpretation of ‘comedy’ in medieval encyclopaedias is ultimately derived. In Aristotle/Moerbeke, however, one crucial element occurs which medieval encyclopaedists fail to convey. Signum autem: isti [the Megarians] quidem enim ‘komas’ perioikidas (idest ambulatoria circa domos) vocare aiunt, Athenienses autem ‘demos’, ut komodos non a ‘komazare’ dictos, sed ea que per komas (idest vicos) seductione deshonoratos ex municipio . . . Poet. 1448a3569 (An indication [of their using different names] is that these [the Megarians] use the expression komai perioikidai (namely, the villages surrounding the city) – the Athenians, on their part, say demoi – as if comedians were so named not from komazare [‘to revel’], but rather from their being forced to wander from village to village [per komas] because of their ignominious exclusion from the city . . .)

By translating directly from the Aristotelian source, Moerbeke is able to retain the reason underpinning the etymology of ‘comedy’ from komai (‘villages’). This is recognized in the comedians’ forced peregrinations from village to village as a consequence of their dishonourable rejection from the city. One cannot help but wonder whether this passage may have fallen under Dante’s eyes, either within or outside its original context – perhaps as an excerpt, or in the form of a gloss. There may not be an immediate answer to this question. But, if Dante did know the passage, then Commedia would indeed be the most fitting title for a text born and grown in exile, written by an exiled poet ‘because of ignominious exclusion’ from his home city.

5

Textual Metamorphosis: The Manuscripts of Leonardo da Vinci Carlo Vecce

The corpus of manuscripts and drawings by Leonardo da Vinci constitutes one of the most extraordinary intellectual legacies in the history of mankind. It comprises thousands and thousands of sheets produced over a period of almost fifty years, from the early 1470s until the year of Leonardo’s death (1519), presented loose or gathered in quires and in notebooks of various formats (folio, quarto and octavo in both larger and smaller sizes), some of which were assembled by Leonardo himself. Scholars and the general public alike are wont to refer to them as codici – ‘codices’ – but the term is perhaps not entirely appropriate. ‘Codex’ designates the form of book which dominated the transmission and circulation of texts in the centuries from the end of antiquity to the invention of the printing press; it is usually associated with that particular type of manuscript book which has preserved classical and medieval literatures in the libraries of Europe’s monasteries and cathedrals. In addition to this, ‘codex’ often goes hand in hand with ‘code’ in the popular imagination – in Italian the same word, codice, is used for both purposes – suggesting the idea of secret languages produced by encryption. This simple conflation is likely to have fuelled the modern myth of Leonardo as the protagonist of stories of mystery and initiation into arcane practices, an example of which is the worldwide best-seller The Da Vinci Code.1 In order to look at the matter from a scholarly point of view, it is important to reflect on Leonardo’s actual practice. Firstly, he used the word libri (‘books’) for his own notebooks as well as for books by other authors in his personal library which he had read and card-indexed since his youth. Most of these were printed books; only a few were manuscripts. Thus libro designated for Leonardo the physical object in general, which could also be a notebook, large or small, occasionally identified by the use of a letter; alternatively, and more specifically, 115

116

Life of Texts

libro could refer to the project of collecting texts on a certain subject or theme: the lost ‘libro A’; the ‘libro B’ (currently known as Codex E); the ‘libro di mia vocaboli’ (‘book of my words’); the ‘libro delle acque’ (‘book on waters’); the ‘libro di pittura’ (‘book on painting’), to mention a few examples. This is confirmed by a sheet in Codex Madrid II , fol. 3v, which follows a provisional inventory detailing the holdings of Leonardo’s library in Florence at the end of 1503. The inventory is preceded by the indications ‘record of the books I am leaving locked in the cassone’ (‘richordo de’ libri ch’io lascio serrati nel cassone’) and ‘in the chest at the monastery’ (‘in cassa al munistero’), and is followed by a list of libri Leonardo described in terms of quantity and format though omitting their titles: ‘25 small books | 2 large books | 16 very large books | 6 books on parchment | 1 book bound in green chamois | 48’ (‘25 libri picholi | 2 libri magg‹i›ori | 16 libri più grandi | 6 libri in cartapechora | 1 libro con coverta di camoscio verde | 48’). For our purposes, it is important to stress that the works of ancient and modern authors – whom Leonardo called altori – shared the same desk and shelves as Leonardo’s own works. This constant interaction between a reader and his books, and between reading and writing, is reminiscent of analogous cases of illustrious intellectuals such as, for example, Montaigne or Petrarch. On the back flyleaf of a codex comprising Cassiodorus’ De anima and Augustine’s De vera religione (Paris, Bibliothèque nationale de France, MS lat. 2201), Petrarch inscribed the words libri mei with the revealing addition peculiares: namely, ‘my most precious books’, ‘the books [that are] most intimately mine’, the presence of marginalia in them being an eloquent testament to his dialogue with ‘his’ authors.

Size and Quality of the Corpus What do we know about the nature and number of Leonardo’s manuscripts? The list in Codex Madrid II (late 1503) gives a total of fifty items. Comparing this number with the manuscripts known to be in existence today, or previously known, the sad conclusion is that more than half of Leonardo’s ‘books’ are now lost, dispersed in part after his death (1519) and in even greater number after the demise of his last pupil and heir, Francesco Melzi (1570). Coveted by princes and wealthy collectors alike, manuscripts and loose drawings journeyed across Europe to end up in most cases in the great public and private libraries – the Biblioteca Ambrosiana in Milan, the Biblioteca Real in Madrid, and the collections of Thomas Howard Lord Arundel in London, Thomas Coke Lord Leicester at Holkham Hall, and the Princes Trivulzio in Milan. There the books

Textual Metamorphosis in Leonardo’s Manuscripts

117

remained, virtually undisturbed or even ignored, until the end of the eighteenth century. Only the texts on the art of painting enjoyed some degree of popularity after Melzi gathered them under the title Libro di pittura around 1540 (MS Urb. Lat. 1270), and again after the publication of a redacted version entitled Trattato della pittura in Paris in 1651. This is an essential point, and one that should never be forgotten when dealing with any of Leonardo’s writings: not a single text by him was published in his lifetime, so all may be described as private writings. Moreover, their author constantly revised them and produced new transcriptions which amount to as many new versions. Those texts remained on his desk as ‘open works’, prone to undergo a metamorphic process with no perceivable end-point. For these reasons Leonardo’s writings lack that kind of structural organization that is commonly associated with the idea of a finished book. Nearly twenty of Leonardo’s manuscripts survive to this day. Since some notebooks were bound together after their author’s death, the total is in fact just over thirty. In addition, manuscripts such as Codex Atlanticus, Codex Arundel and Codex Windsor are great collections of texts, documents and drawings which originally lay scattered on Leonardo’s desk. Italy holds Codex Atlanticus (Milan, Biblioteca Ambrosiana), the greatest collection of loose leaves and fragments (over a thousand) by Leonardo, assembled by the sculptor Pompeo Leoni at the end of the sixteenth century, Codex Trivultianus (Milan, Castello Sforzesco) and Codex of the Flight of Birds (Turin, Biblioteca Reale). In France, the library of the Institut de France holds the twelve manuscripts removed from the Ambrosiana by Napoleon in 1797, designated by letters of the alphabet, A to M. In England, the three Codices Forster are in London at the Victoria and Albert Museum; an important miscellany of originally loose leaves known as Codex Arundel is now in the British Library; the Royal Library at Windsor preserves a conspicuous collection of drawings and, above all, the sensational notebooks containing Leonardo’s anatomical studies; the two Madrid Codices, I and II , were rediscovered at the Biblioteca Nacional in the 1960s; the former Codex Leicester is now in Seattle in the private collection of Bill Gates. How can one navigate a textual labyrinth of such proportions? Indeed, navigate or ‘surf ’ are contemporary expressions that come to mind. The current familiar notions of ‘internet’ and ‘hypertext’ adapt well to the perusal of texts which do not conform to any visible hierarchy while, at the same time, offering an intricate network of cross references leading from one leaf or ‘codex’ to another. In this respect, none of the editions produced over the past centuries

118

Life of Texts

can be regarded as satisfactory. Conceived as they were in adherence to a model which was (and meant to be) both stable and static, the exceptional complexity and dynamism of Leonardo’s texts were bound to be represented, if at all, by means of serious oversimplifications. Predictably, the aspects which have suffered most from that type of approach are those that did not fit with the then-dominant idea of books as completed and structured works, so that the ‘provisional nature’ and ‘fragmentary state’ of so many of Leonardo’s writings were inevitably perceived as negative features. The problem is already apparent in the first thematic anthology of Leonardian writings, the Libro di pittura that Francesco Melzi put together on his own initiative, albeit following or least remembering – one presumes – some of his late master’s directions. Yet the problem persists even with the most recent critical editions of discrete Leonardo manuscripts, since the printed book format can only imperfectly reflect the nature of texts conceived by their author as ‘open works’. In other words, the classic book format does not enable modern readers to approximate the processes that characterize Leonardo’s ‘mental discourse’ and its deeper significance, for which the constant interrogation of the entire corpus is needed. In many respects, Leonardo’s textuality has a better chance of being understood in our digital age than ever before. This does not mean that the lesson from the great Leonardo scholars of the past century has lost any of its relevance. As scholars of the calibre of Girolamo Calvi, Augusto Marinoni and Carlo Pedretti have shown, the point of departure remains the examination and assessment of Leonardo’s writings with respect to their chronological evolution. The reason for this is simple. The timeline of Leonardo’s activity, as it enfolds and variously manifests in the leaves and manuscripts he progressively drafted, neatly coincides with the timeline punctuated by the circumstances of his life as are known from a variety of sources. Without a rigorous approach to the chronology of his life and writings, different phases of Leonardo’s formation and achievement as a scientist and an artist can easily get confused and mixed up, thus making the advantages of technological progress illusory, and digital navigation a pointless exercise.

A Chronology of Leonardo’s Manuscripts Leonardo’s earliest surviving writings, in a limited number of leaves of Codex Atlanticus and Codex Arundel, are from the years of his apprenticeship in Florence, 1469–82. These are humble testimonies of his work in the bottega – mainly recipes for mixing colours or producing varnishes and finishing oils, often

Textual Metamorphosis in Leonardo’s Manuscripts

119

transcribed from the master’s words or from those ‘books of the art’ (‘libri d’arte’) or ‘books of secrets’ (‘libri di segreti’) so common in that environment. In those sheets one also encounters drawings of various mechanical gears and devices with scanty captions, doodles, personal recollections, dates, and people’s names. Those sheets also contain tokens of Leonardo’s earliest readings, such as Ovid’s Metamorphoses in the vernacular translation by Arrigo de’ Simintendi da Prato, Luca Pulci’s poetical epistles and Petrarch’s Trionfi (Codex Atlanticus, fol. 195v), as well as his first attempts at elaborate literary compositions. One such instance is the description of a sea monster and its cave, a fragmentary conte philosophique narrated in a mixed form encompassing visions, dialogues and first-person recollections. This is one of the earliest of Leonardo’s texts to unveil his conception of life and nature as processes dominated by perpetual movement and transformation (Codex Atlanticus, fol. 715r–v; Codex Arundel, fols. 155r–156v). These early testimonies must be complemented with Leonardo’s first dated leaf displaying a landscape in the valley of the Arno with the caption ‘the day of Saint Mary of the Snow | today 5 August 1473’ (‘dì di santa Maria della neve | adì 5 d’aghossto 1473’, Florence, Museo degli Uffizi, Gabinetto dei Disegni, 8P). In it, Leonardo’s characteristic right to left handwriting form is already discernible, called ‘specular’ because it is more easily legible with the help of a mirror. It is not ciphered writing – only the spontaneous graphic expression of a self-taught, lefthanded boy, who mentally reversed the letters before reproducing them on paper. Leonardo could also write from left to right, as is shown by other leaves from his early years. His writing conforms to the mercantile hand which was typical of the world of notaries and merchants where Leonardo grew up and received the first rudiments of his education. The most ancient Leonardo manuscript is Codex B, a veritable hotchpotch, or zibaldone, of texts and drawings aimed at acquiring and reinforcing his expertise, both theoretical and practical, in the fields of engineering, architecture and military science. Leonardo relied on such expertise to introduce himself to Duke Ludovico Sforza on his arrival in Milan: the letter in which he offers his services survives in a non-autograph copy (Codex Atlanticus, fol. 1082r). The move to Milan constituted a turning point in Leonardo’s development as a writing practitioner. During the years 1485–90, he moved from occasional jotting to systematic reading and transcription of large portions of texts, notably from Roberto Valturio’s De re militari (first published 1472), which Leonardo read in Paolo Ramusio’s vernacular translation published in Verona in 1483. Military matters are paramount in Codex B. Yet one also remarks the presence of notes on elementary geometry which are functional to the study of linear perspective;

120

Life of Texts

drawings of imaginary war machines – such as bombards, catapults and crossbows of gigantic proportions; chariots both armoured and falcated (i.e. with scythes sticking out on both sides); flying and submarine vehicles – which go well beyond the intent of Valturio’s antiquarian erudition; and pages devoted to architecture and town planning in connection with the coeval debate on the ideal city, a subject of great momentousness in Lombardy under the rule of the Sforzas. Leonardo’s most urgent problem appears to have been the acquisition and mastering of the appropriate terminology for each of the disciplines he was tackling, combined with the promotion and ennoblement of his studies through the adoption of models of style derived from the literary culture of the humanists. As this entailed a thorough familiarity with the classical rhetorical tradition, mastering Latin became an obstacle for the self-taught boy raised in a Tuscandominated environment where he had had no chance of learning – let alone mastering – the language of the ancients. In a number of programmatic texts, all entitled Proemio (‘Proem’, Codex Atlanticus, fols. 323r, 327v), Leonardo acknowledged his lack of expertise in litteratura and admitted to not being a litterato, yet at the same time vehemently rejected the charge of being a ‘omo sanza lettere’, an illiterate person. Even though his condition was that of a man not equipped with the necessary skills required to comprehend and translate a Latin text, he was brave enough to undertake a strenuous course of action that would enable him to increase his familiarity with the written word, as well as to expand his personal library significantly. Leonardo tried to remedy his ignorance of Latin by resorting to the practice of listing those terms of Latin origin he found in the books available to him in his study. This was a widespread method in the context of the vernacular culture of the fourteenth and fifteenth centuries, and one recommended by Cristoforo Landino in both the inaugural lecture of his course on Petrarch in the Florentine Studio (1460s) and in the preface to his commentary on Dante’s Commedia (Florence 1481). Landino himself could be said to represent an important example of the fruitful interaction of humanist and vernacular culture; some of his works, such as his Formulario di pìstole and the vernacular translation of Pliny’s Natural History, feature among Leonardo’s books of the Milan years alongside Giovanni Simonetta’s Sforziade. Leonardo’s earliest examples of lists of terms make their first appearance in Codex B (fols. 40r and 100r: names of weapons, culled from Valturio), whereas the most advanced occurrence of this practice is represented by the long lists of terms (over 8,000!) that populate the pages of Codex Trivultianus (1487–90). This latter list begins with a partial transcription from an analogous compilation, Luigi Pulci’s so-called Vocabulista,

Textual Metamorphosis in Leonardo’s Manuscripts

121

and continues with further terms obtained from Valturio; from a vernacular translation of Poggio Bracciolini’s Latin Facetiae (produced in the Po Valley, as the dialectal colouring suggests); from Masuccio Salernitano’s Novellino; and from other sources not yet identified. It may be interesting to observe that, owing to their provenance, none of these texts could have earned the praise Landino had bestowed on the ‘good Tuscan tongue’ (‘buono toscano’) as the only vernacular variety capable of ennobling one’s mother-tongue (‘lingua materna’). Leonardo, however, aimed principally at registering Latinizing terms or any term he identified as characteristic of a lofty, aulic style. Once he was confident he had built firmer foundations, Leonardo began to cultivate an ambition of becoming a writer and even an author – an altore. He therefore set out to plan his own ‘books’. A symptom of this new course is reflected in the proud registration of a start date for the endeavour, which occurs in one of the earliest leaves of his notebooks on anatomy: ‘Today 2nd April 1489 a [new] book entitled On the human figure’ (‘A dì 2 d’aprile 1489 libro titolato de figura umana’, Windsor, Royal Library, fol. 19059r). Leonardo was envisaging for this a treatise in which the physical and anatomical description of the human body would go hand in hand with its visual representation, ideally to be connected with the part of his book ‘On painting’ which is devoted to the human figure. One can recognize here a typical feature of Leonardo’s ‘mental discourse’: no enquiry remains confined within the boundaries of one subject, but rather tends to open up and be cross-fertilized by encounters with other disciplines. The practice and theory of painting are constantly accompanied by his interests in linear perspective and optics, which in turn intersect with the study of anatomy – to the extent that the ‘proems’ for the separate works he was planning on anatomy and linear perspective ended up sharing the same pages. This intellectual and scientific fervour even made Leonardo’s activity as a painter recede into the background. After the Virgin of the Rocks, he completed only two other paintings in those years, the Lady with an Ermine and the Musician. Leonardo began a new manuscript, Codex C, with a solemn note: ‘Today 23rd April 1490 I began this book and started work on the horse again’ (‘Adì 23 d’aprile 1490 cominciai questo libro e ricominciai il cavallo’), which reminds us of the other important work he should have accomplished for Ludovico il Moro, a colossal equestrian monument of Francesco Sforza, the realization of which would be constantly postponed with no appreciable result or progress. Codex C is a manuscript of great dimensions with wide margins; almost a veritable ‘book’, written in a neat hand and carefully drawn, mainly devoted to the study of light and shade, after an initial section on water and on percussion.

122

Life of Texts

Codex A belongs to the subsequent period (1490–2) and is predominantly concerned with the subject of painting, particularly the theme of linear perspective, while in the last quire Leonardo introduces a series of texts entitled Modo di figurare (‘Drawing technique’) intended for the planned book ‘On painting’. The didactic nature of these texts is clear from the use of the second person singular, as if they were addressed to an ideal pupil. Yet the subjects they cover appear as impossible challenges to a painter of the fifteenth century, almost a prelude to the pictorial revolution of the following century and the early modern age more broadly: the dramatic turmoil caused by natural phenomena such as storms and floods; the intermingling of the elements; the dramatic contrasts of light and shade in the darkness of the night; and the irrational vortex of violence in the description of a battle. Leonardo’s attention to the Sforza horse appears to have been revived in the second part of Codex Madrid II , an ‘irregular’ notebook containing notes about the fusion process (1491–3). During the same period Leonardo’s studies on mechanics showed a tendency to evolve towards a coherent whole, which was to have taken the shape of a book on ‘machine elements’ (‘elementi macchinali’), as testified by Codex Madrid I (1493–7). However, a great number of other texts continued to be drafted on loose sheets and quires and were never organized, even provisionally, into a single manuscript. This is the case with the most attractive pages intended for a ‘book of the waters’ (‘libro delle acque’), surviving as loose sheets in Codex Arundel (fols. 57r–59v, 210r, 233r–236v) and Codex Atlanticus (fols. 468r, 1100r). A similar origin characterizes a limited number of short writings of a literary nature which reflect Leonardo’s Tuscan-based education and modest social provenance: proverbs and maxims, citations from authors of poetical works (Dante, Petrarch, Cecco d’Ascoli), and, above all, fables and facetious short stories – Leonardo owned Aesop’s Fables and, as we have seen, Poggio’s Facetiae. An intimate friend of the Pulci Brothers, Benedetto Dei, whom Leonardo had met in Milan, introduced him to Luigi Pulci’s Morgante and Luca Pulci’s Driadeo and Ciriffo Calvaneo; Dei also inspired Leonardo to compose an extraordinary fictional letter, addressed to Dei himself (Codex Atlanticus, fols. 852r e 265v), written in imitation of the popular stories narrated in cantari style, in which Leonardo feigns to send a report from the East about the appearance of a terrifying giant. Elsewhere, Leonardo transcribed from Antonio Pucci’s Historia della reina d’Oriente a stanza describing a giant (Codex I, fol. 139r). The format of Leonardo’s manuscripts adapted to changing circumstances. Finding himself deeply involved in projects on structural engineering and architecture in collaboration with Bramante and Francesco di Giorgio, while at

Textual Metamorphosis in Leonardo’s Manuscripts

123

the same time engaged in frequent missions on behalf of the Duke of Milan, Leonardo devised a type of portable notebook suitable for rapid annotation, calculation and sketching, similar to those in use by coeval engineers. Among Leonardo’s manuscripts this special type is represented by Codex Forster I-1 (1487–90) and Codex Forster III (1490–3). One of their noticeable features is the frequent use of red pencil (sanguigna) over quill and metal tip – which were reserved for the composition of more elaborate texts and transcriptions in fair copy, executed in the study with all the necessary tools at hand (quill, inkpot, ink and so forth) – whereas red pencil was for rapid, hurried note-taking, desultory and temporary. Certain jottings even appear to have been disturbed by the rocking and swinging of a horse or a chariot. The three notebooks constituting Codex H are characterized by their smaller format. They were prevalently used in 1494 during Leonardo’s frequent journeys from Milan to Vigevano, the favourite residence of Ludovico il Moro’s court that year and a special object of attention for ongoing projects concerning city-planning. Taking advantage of the proximity of the river Ticino and his own involvement in complex works aiming at regulating as well as exploiting water flow – watermills, canals (navigli), locks and river mouths, ‘water ladders’ – Leonardo made a series of detailed observations of water movement which he faithfully registered in the codex. After returning to Milan, he utilized the remaining pages for a bestiary, a collection of texts on animals both real and fictitious, which he freely transcribed from the Fiore di virtù, Cecco d’Ascoli’s L’Acerba and Pliny’s Natural History: a repository assembled, presumably, for the invention of devices and allegories and the composition of fables. Over the same period, he made an effort to learn Latin – at the age of forty-two! – with the sole assistance of a popular elementary grammar, Niccolò Perotti’s Rudimenta grammatices, from which he copied out the rules governing declensions and conjugations. While he was painting The Last Supper in the refectory of Santa Maria delle Grazie in Milan, Leonardo aimed to improve the elementary knowledge of arithmetic he had acquired as a boy in the abacus schools (scuole d’abaco). He passionately devoted himself to the study of a recent book he had bought for the remarkable sum of 248 soldi (Codex Atlanticus, fol. 288r): Summa de Arithmetica, Geometria, Proportioni et Proportionalità, published in Venice by the Franciscan friar Luca Pacioli, a pupil of Piero della Francesca (10 November 1494). Long extracts from Pacioli are transcribed into Codex Forster II and interspersed with important figurative notes on the poses of the Apostles in the fresco of The Last Supper, together with some facetiae and prophecies. The second part of the codex, which had originated as a separate notebook, deals with medieval physics

124

Life of Texts

and the measurement of weights; as such, it converges with the material already gathered in Codex Madrid I and adds copious notes for a treatise ‘on weights’ (De ponderibus). Pacioli arrived in Milan in 1496 and became Leonardo’s tutor in all matters geometrical and mathematical. Pacioli introduced Leonardo to Euclides’ Elementa (investigated in Codex M, 1497–8), and Leonardo collaborated on Pacioli’s treatise Divina proportione, completed in 1498, designing the tables of Platonic polyhedra. During these years, apart from geometry, Leonardo explored in dense notes the art of war, as well as optics and technology, mechanics and physics (the fall of bodies and of liquids, and the resistance of the medium), which testify to his reading of the works of Jordanus Nemorarius and Albertus Magnus. The study of Euclides occupies the first part of Codex I, which can be dated to 1497, wherein Leonardo concludes his notes from Book 1 of the Elementa (begun in Codex M) and moves on to consider further propositions from Books 2, 3 and 10. The second part, generally contemporary with the first, includes texts on water and movement, and a further series of extracts and lists of Latin words culled from Perotti’s grammar book. Finally, the manuscript harbours the longest and most homogeneous series of the short texts Leonardo called ‘prophecies’, which are in fact parodic texts in the form of riddles on apocalyptic subjects – a fashionable topic in the last years of the century. In 1499, Milan was conquered by the French, Ludovico il Moro was dethroned and Leonardo was forced to leave the city. This event marked the beginning of his Wanderjahre: Mantua, Venice and Friuli, Florence, Piombino, the Marche and Romagna, and Rome. The portable notebook which followed Leonardo everywhere, faithfully documenting his movements during this period, is Codex L, begun in 1497. It comprises notes on elementary mathematics, the most recent works of art executed in Lombardy, and an account of the author’s services as military engineer to Cesare Borgia (1502–3). Back in Florence, Leonardo began a new composite manuscript, or zibaldone, of sizeable format, Codex Madrid II . In it he principally gathered notes referring to duties he was performing on account of the Florentine Signoria, such as the deviation of the course of the river Arno during the war against Pisa (1503) and the planning of new fortifications at Piombino (1504). But the manuscript also includes disparate matter on applied geometry (regarding the squaring of the circle and further reflections on Euclides); an attempt at stereometry in which he cites a recently published humanist encyclopaedia, Giorgio Valla’s De expetendis et fugiendis rebus (Venice: Aldo Manuzio, 1501); texts on painting, optics and

Textual Metamorphosis in Leonardo’s Manuscripts

125

linear perspective; and passages transcribed from Francesco di Giorgio Martini’s treatise on architecture. Leonardo attached such importance to Martini’s text that he felt the need to consult a different version of it, now in the Laurentian Library in Florence as MS Ashburnham 36 – the only book that has been recognized as belonging to Leonardo’s library, made even more valuable by the presence of his glosses in the margins. The years spent in Florence were particularly fruitful for Leonardo. The first two parts of Codex K date back to those years and include transcriptions from Euclides, Books 2–6. The wonderful Codex of the Flight of Birds from 1505 presents Leonardo’s research into the nature of flight, aimed at the construction of a flying machine. Finally, the first part of Codex Forster is preceded by the title ‘A book on transformation, that is, of one body into another with no diminution or augmentation of matter’, followed by the precise inscription: ‘Begun by me Leonardo da Vinci today 12 July 1505’ (‘Libro titolato de strasformazione cioè d’un corpo ’n un altro sanza diminuzione o accrescimento di materia’; ‘Principiato da me Leonardo da Vinci a dì 12 di luglio 1505’). From 1506 to 1508, while moving between Florence and Milan, Leonardo was busy putting his papers in some order. The thousands of leaves were a labyrinth difficult even for him to orientate himself in. He transcribed texts from one manuscript into another and collected scattered notes to form discourses with greater internal coherence. The small Codex D dates from this period: ten sheets on the human eye, the meeting point of investigations conducted in fields as diverse as painting, optics, linear perspective and anatomy. The study of water and of earth sciences dominates the third part of Codex K and the Codex Leicester: a remarkable attempt at ordering material, formerly dispersed in loose sheets or loose quires, which was now transcribed in fair copy and in a more coherent sequence. (Among the sources was a lost ‘Book A’, which contained miscellaneous notes on hydrodynamics and painting.) The incessant flow of Leonardo’s writings generated by the free rein given to thoughts and investigations was not easily channelled or constrained within tight boundaries. His ambivalent attitude, perpetually oscillating between the need for self-discipline and an unquenchable reluctance to accept constrictions of any kind, is best expressed at the beginning of the first quire of Codex Arundel, wherein a note marks the beginning of transcriptions conducted in Florence in 1508: E questo fia un raccolto sanza ordine tratto di molte carte, le quali io ho qui copiate, sperando poi di metterle per ordine alli lochi loro, secondo le materie di che esse tratteranno. fol. 1r

126

Life of Texts

(This is going to be an unordered harvest extracted of many papers I have copied here, in the hope that at some point I shall give each of them its proper place according to the subject they are concerned with).

His investigations into the natural world moved from water to other elements, notably air and fire in their manifold manifestations: winds, clouds, rain, lightning, celestial phenomena – the same wondrous movement of nature as is manifested in the landscapes of the Mona Lisa and Saint Anne. Codex F, mainly assembled in Milan in 1508–9, deals with the movement of waters but also tackles questions of cosmology in texts of a very disparate nature: an attempt to disprove Plato’s position on the geometrical form of primary elements in nature, and a ‘Praise of the Sun’ (‘Lalde del sole’), similar in tone to Michele Marullo’s Natural Hymns and Gregorio Dati’s Sphere, supporting the doctrine of the excellence and centrality of the sun in the world system. In 1507–8, in the Florentine hospital of Santa Maria Nuova, Leonardo extended his research into the human body by performing anatomical sectioning. His interests had clearly moved beyond the question of the artistic representation of human figures and were progressively branching out in wider directions and into further specialized areas of knowledge. He devised a revolutionary method of representation in his anatomy sheets and quires that used both pictures and words, figurare and descrivere. Around 1510, owing in part to the assistance of Marcantonio Della Torre, a young physician and humanist who had studied Galen and taught at the University of Pavia, Leonardo had almost completed his treatise on anatomy: a complete atlas of the human body, its great tables the equivalent – in his eyes – of the geographical maps in Ptolemy’s cosmography, a comparison instigated by the ancient twin notion of macrocosmos and microcosmos. Codex G documents Leonardo’s move from Milan to Rome (1513) and encompasses texts from 1510 to 1515. After several not always successful attempts at organizing his previous writings, this notebook shows once again an accumulation of notes on different themes: the life of plants, force and percussion, movement and the flight of birds, geometry (with quotations from Archimedes and Vitruvius), water and the sun, optics and linear perspective, and observations on the movement of ships. Leonardo’s most recent surviving manuscript, Codex E, can be dated with certainty to between 1513 and 1514, when the old artist resided in Rome and worked in his laboratory in the Vatican Palazzetto del Belvedere amidst playful mirror effects and capricious inventions, pursuing what his sceptical contemporaries regarded as oddities and strange chimaerae. ‘Yet another of the world’s

Textual Metamorphosis in Leonardo’s Manuscripts

127

foremost painters,’ Baldassare Castiglione observed ironically, ‘is disregarding his art, in which he has very few competitors, and has set out to learn philosophy’ (‘Un altro de’ primi pittori del mondo sprezza quell’arte dove è rarissimo, ed essi posto a imparar filosofia’). In fact, Codex E goes back to earlier studies on mechanical physics, the science of weights, and the flight of birds, while also presenting notes on painting for the purpose of teaching. Leonardo kept writing indefatigably until his final days, mainly on loose sheets now preserved in Codex Atlanticus, Codex Arundel and the Windsor collection. In one of those sheets, a brief note appears to encapsulate the absorbing nature of his endless writing, which only external circumstances and eventually the end of his lifespan would arrest. Coming to the end of a geometrical demonstration, which is left hanging with an eccetera, Leonardo adds the reason for the interruption: ‘perché la minestra si fredda’ (‘for the soup is getting cold’, Codex Arundel, fol. 245r).

The Models for Leonardo’s Manuscripts After surveying, however rapidly, the contents of Leonardo’s manuscripts in their chronological sequence, we can now try to recognize the presence of a few constant features among so many variables. One prominent aspect is the tight bond of both texts and drawings with the material support which has allowed their survival. Conceived in such a way as to be only imaginable, as well as transmissible, in manuscript form, Leonardo’s oeuvre set itself apart from the new linguistic and literary system that was emerging in Europe at the dawn of the early modern age with the new invention of the printing press, which afforded the mechanical reproduction of books in multiple copies. Being at odds with that dominant trend does not detract from the extraordinary methodical consistency with which Leonardo’s writings were deposited on thousands of sheets of paper, layer after layer and virtually without intermission, for a period of nearly fifty years. His manuscript corpus, heterogeneous in appearance, does in fact come across as one immense single work, the structure of which is laid bare in the chronological succession of the texts it contains. For this reason, while they may have never achieved the status of ‘book’ or ‘treatise’ in line with the conventional notion of an accomplished work, Leonardo’s texts cannot be easily disregarded as erratic fragments. The act of writing acquired for him an almost ancestral significance with respect to the representation of perceived reality; it is a tool for communication as well as of

128

Life of Texts

knowledge acquisition. Urged on by his feverish enquiries, it is as if Leonardo could not afford to stop or hesitate for questions of textual organization and contextualization. What mattered for him was the perception of the reality that surrounded him as a world subjected to perpetual mutability, as a continuous quantity in constant flux. His writing habits followed that perception so closely that they eventually identified and coincided with it, and variety and mobility became essential ingredients of his textual performance. In this immense book of his, which acts as a mirror to the greater book of nature, one struggles to find a structured discourse extending over a single page or a folio. Even in the manuscripts showing greater thematic homogeneity, the turning of a page almost systematically entails a change of subject, of chapter, of paragraph. The external limitations imposed by the writing material also determined the nature of the vast majority of Leonardo’s texts: rapid annotation predominantly dictated by a sense of immediacy; relatively short scientific observations and succinct demonstrations of theorems which often remained uncompleted, as the frequent ‘eccetera’s placed at their ends show. Nonetheless, even within a context dominated by temporality, some main thematic lines recognizably develop in the guise of a tree that grows with the passing of time – the tree of knowledge. Leonardo had initially intended to use the power of words to motivate his understanding of painting as a ‘science’ (scientia) in opposition to the traditional system of the liberal arts. This led him to expand and deepen his linguistic skills, to try to learn Latin, and to engage with the arts of trivium, notably grammar and rhetoric. Leonardo’s interest in linear perspective had, on the one hand, a theoretical strand in connection with mathematics and geometry and the arts of quadrivium (including, during his Milan years, his engagement with astronomy and music); on the other hand, it also encouraged his radical reformulation of Leon Battista Alberti’s principles following his reading of Euclides and his new enquiries into the physiology of the human eye and the mechanics of visual perception. His approach to the representation of the human body and the phenomena of movement, percussion and force covered all aspects of scientific study: from anatomy to mechanics, from physics to hydrodynamics. Originally prompted by his project on the ‘book on painting’, these interests evolved over time into autonomous enquiries, while, at the same time, converging towards Leonardo’s practical activities in his double capacity as a structural engineer and an architect. These tendencies impelled Leonardo to produce texts that are mostly registrations of facts and observations, in an attempt to cope with the prodigious accumulation of data derived from both direct experience and the reading of texts.

Textual Metamorphosis in Leonardo’s Manuscripts

129

Leonardo’s writing habits, wonderful as they may seem, should not be regarded as an isolated phenomenon or as the form of expression of a superhuman genius, as one is often tempted to believe. His habits are best described as a form of private writing and, as such, they exhibit an affinity with other coeval instances of writing as the registration of records in the guise of an ‘open book’. The most influential model for this kind of practice was the humanist composite book, or zibaldone, a type of notebook prepared to receive notes of philological and erudite nature in chronological succession, thus mirroring progress in the studying and reading of texts. In its apparent disorder, a humanist zibaldone was meant to provide ‘raw’ material to be subsequently rearranged in different forms and contexts – philological commentaries on ancient texts, miscellaneous collections of castigationes, lectures. The material they customarily collect comes, rather than from physical reality, from the world of ancient books and auctores which humanism had recovered and revived after intense restoration work. The classification, correction and even ‘navigation’ of such texts was made possible by new research tools such as lexica, dictionaries and encyclopaedias: one example is Giorgio Valla’s De expetendis et fugiendis rebus (1501), which Leonardo owned. To be sure, Leonardo was no humanist; he did, however, reveal the same kind of cognitive tension when he combined his direct sperientia of natural phenomena with the need to consult the books of the auctores (whom he calls altori). He also adopted in his manuscripts certain writing and organizational techniques which are characteristic of the humanists’ practice. Another relevant model of the ‘open book’ was common among the structural engineers and authors of works on practical subjects of the Tuscan Quattrocento, such as Lorenzo and Buonaccorso Ghiberti, Giuliano da Sangallo and Francesco di Giorgio Martini. These authors translated into the vernacular and made available to a wider readership the principles of mechanics and technology which Mariano di Taccola had dealt with in his Latin treatise De ingeneis in the early fifteenth century, and which Filippo Brunelleschi had concretely put to the test some time afterwards by raising the dome of the Florentine cathedral. All of these authors shared the same ambition of producing accomplished books in the form of treatises, a good example being Martini’s Trattato, extant in more than one version, which Leonardo owned and read with great attention. Yet their working practice entailed the preliminary drafting of notebooks comprising drawings and texts in a continuous sequence. Martini kept at least two such notebooks, one for drawings of Roman monuments, the other for transcriptions from Taccola (Biblioteca

130

Life of Texts

Apostolica Vaticana, MS Urb. Lat. 1757); Buonaccorso Ghiberti constructed his own zibaldone (Florence, Biblioteca Nazionale Centrale, MS B.R. 228) in order to collect his drawings (his own and those of his predecessors), which were subsequently studied by Leonardo. Fragments from notebooks, quires or even sheets compiled by anonymous engineers, including recipes, ingegni (‘devices’ or ‘machines’) and rough drawings and tracings, often became mixed with Leonardo’s own papers (one such example is in Codex Arundel, fols. 258r–261v). These occasionally reveal an illustrious origin, as shown by the presence of Paolo dal Pozzo Toscanelli’s hand in Codex Atlanticus, fols. 611br and 915ar. Yet, when one considers Leonardo’s social provenance, the model that was arguably most familiar to him from a very early age is without doubt that of the ‘libro di famiglia’, in which fifteenth-century Florentine merchants and middleclass citizens compiled their ricordi. Records of daily events, including contracts, aides-de-memoire, momentous family occurrences such as births or deaths, proverbs, and moral considerations on the meaning of life, were registered in no apparent order. As in Leonardo’s manuscripts, these texts often adapt to the size of the page or folio and are introduced by such titles as memoria or ricordo. It is yet another case of an ‘open’ kind of writing, with no pre-established intention of closure; the last page of such libri di famiglia is the last one written by their owners before their demise. Similarly, Leonardo’s manuscripts accompanied his daily life, assuming the aspect of memoirs or ricordanze: they preserve the names of people met or due to be met, of loans, debts, credits, places, itineraries, memorable events, dates. The entire Leonardian corpus often presents the aspect of a diary. A further aspect of Leonardo’s manuscripts is the strong flavour of orality that permeates his language and style and which emphasizes a tendency to perpetual metamorphosis. Orality dominated the environment in which Leonardo had been raised as a boy in Vinci and subsequently lived in as an apprentice in Florence: fables and stories told by his mother Caterina or his grandfather Antonio; the cantari recited in the village square; the mysteries and nativity plays performed before the village church; later, the practical teachings imparted in the Florentine workshops, such as that of Andrea del Verrocchio, where Leonardo revealed his skills as an accomplished draughtsman. In his notes on the so-called ‘Paragone of the Arts’, which he began drafting in Codex A in 1490 and expanded some ten years later into what became the first part of his Libro di pittura, the comparison between painting and poetry is conducted on the basis of an opposition between two different channels of transmission, the visual and the phonic-aural, hence between the two senses presiding over reception: sight and hearing.

Textual Metamorphosis in Leonardo’s Manuscripts

131

In anthropological terms, Leonardo’s affirmation of the primacy of the eye – ‘the window of the soul’ – has an exceptional historical value. It coincides with that delicate moment of transition, from the so-called Middle Ages to the early modern age, at which visual perception appeared to prevail over aural and olfactory perception in the relationships with and representations of reality. Poetry defined as ‘painting that is heard and not seen’ (‘una pittura che si sente e non si vede’, Libro di pittura, Chapter 20) or as ‘a blind picture’ (‘una pittura cieca’, Chapter  21), has its point of reference in a notion of textuality that is oral and aural rather than written. Indeed, Leonardo’s writing presents the typical syntactic and stylistic characters of orality (parataxis, loose aggregation, repetition, redundancy and emphasis), including frequent addresses to fictional interlocutors. It is possible to imagine Leonardo writing his texts through some sort of self-dictation, either silent or aloud. This would also explain the strong presence of non-standard phonetic traits in his spelling. The influence of orality, which may occasionally represent an advantage but is more frequently a drawback whenever precision, clarity and concision are demanded, is partly overcome by Leonardo’s use of drawings. Word and image materialize and develop jointly in his pages and influence each other, both contributing to structure the insurgence and formation of thought. It is often barely possible to tell which came first, especially in those cases where different versions of the same image and the same text fill a page. Textual variation appears to be Leonardo’s most natural and congenial way of following the train of his thoughts. As such, and for the mental process that it involves, textual variation is akin to the variation of images typical of his drawing practice, which E. H. Gombrich identified as a crucial factor in Leonardo’s method of ‘permutation’. As the image in a drawing is first traced in red (sanguigna) or black pencil and subsequently written over with the pen, so are texts originally jotted down in sanguigna visibly written over in ink, character after character (just as in a palimpsest one can distinguish a lower and an upper script), until it is transcribed into a different book or quire. Overwriting and transcription cannot be described as mechanical copying, as in each phase Leonardo introduced changes that configured his text as an entity in perennial movement. Owing to Leonardo’s disregard for any assumed permanent validity of what is written and thereby for its ‘fixity’, as encapsulated in the old adage verba volant, scripta manent, his page is prepared to receive and preserve all the different phases of textual ‘movement’. The autonomous, generative nature of Leonardo’s script-image equally frees up the process of his scientific enquiry from any obligation to pronounce verdicts of perennial validity.

132

Life of Texts

In fact, by its very mobility, his writing practice approximates the intimate nature of natural phenomena, which are by definition mutable and therefore transient. Once one has acknowledged the sheet of paper as Leonardo’s field of projection of perceived reality, his writing and drawing habits may be said to acquire a special metatextual dimension in respect of configuration and the apportioning of the available space. These procedures naturally involve a visual approach to page layout. Leonardo would soon have become aware of the general problem of text flowing across the page, which for him moves from right to left, as in a book written in Arabic or Hebrew. His unique mise-en-page is far freer than in any coeval example, either handwritten or in print. In general, Leonardo’s script is best characterized as a ‘textual window’ rather than a ‘writing area’. Even when the layout does not look perfect, the text always appears framed in square or rectangular windows with justified margins; and whenever a text does not fill the window’s last line, Leonardo appears to feel a need to end it by tracing a wavy line, in much the same way as a seismograph or an electrocardiogram would do. This habit becomes frequent in manuscripts written after 1500. His terminal line occasionally follows his eccetera – a way of postponing conclusion by reverting to his own ‘mental discourse’, which at this point appears to be at one with his writing, in a continuum both physical and biological. The habit of using eccetera was presumably derived from the practice of notaries and chanceries. Very rarely present in the writings of the pre-1500 period, Leonardo’s eccetera becomes more and more frequent between 1503 and 1508, thus coinciding, as we have observed, with intense ‘re-writing’ campaigns, and is virtually ever-present on every page of his later manuscripts (E, G) and his last sheets. Perhaps the issue can be summarized as follows: Leonardo was unwilling to put an end to whatever he was engaging with. His endless writing is yet another instance of his blurring the boundaries between different forms of expression, language, discipline, and intellectual experience. It also testifies to his reluctance to accept that ultimate act of separation from a text that is implied in its closure, in driving the development of a discourse to conclusion. The biological interruptions and boundaries of his writing appear to be, time after time, a single word, a line, a paragraph, a page, a book; those of the author at work are the extinguishing light of day and that of the oil lamp at night, the bare necessities of living (‘the soup is getting cold’) and the end of his own life.2

6

Montaigne: The Life and After-Life of an Unfinished Text John O’Brien

Montaigne’s Essais are a complex artefact. First published in French in two books in 1580, and reprinted with minor changes in 1582 and 1587, they acquired a third book in 1588. Between 1588 and 1592, the date of his death, Montaigne worked on a further edition of his work, but did not live to see its publication in 1595.1 It was to prove an international best-seller in the early modern period, as indeed it has been ever since, and the question of its life and particularly its after-life is of particular significance. Previous studies have dealt with its afterlife in France and from printed sources.2 Part of the purpose of the present essay is to examine hitherto unknown manuscript and printed sources from outside as well as inside France in order to determine what Montaigne’s Essais meant to some of their first European readers. As a way into this whole question, chapter 37 of book 1 of the Essais, entitled ‘Of Cato the Younger’ (‘Du jeune Caton’), offers some leads (see Figs. 6.1 and 6.2). In the 1580 version, this is a short piece clearly displaying a story of difference articulated in the first line between the essayist and heroic souls, in an age – Montaigne’s own – when virtue is not what it seems. The essay then passes on to the younger Cato, who is the emblem of ‘the splendor of virtue in its native purity’ (‘la splendeur de la vertu en sa pureté naifve’).3 The essayist’s inadequacy in ethical imitation of heroic models is matched by his inability to represent them in his book, and so he turns to five Latin poets who have written about Cato and quotes their characterization of the Roman hero. And in 1580 the essay ends there. By its close, it has become a story about literary imitation: literature is what can reach the heights that the essayist cannot. After 1588 and before his death in 1592, Montaigne further develops these aspects. In those years, the focus of the essay shifts again: a long addition towards the end of the chapter 133

134

Life of Texts

Figures 6.1 and 6.2 ‘Of Cato the Younger’ (‘Du jeune Caton’), with Montaigne’s own handwritten additions and amendments.

turns it into a story about the essayist’s own susceptibility to the ecstasy induced by poetry: ‘[poetry] does not persuade our judgment, it ravishes and overwhelms it. . . . From my earliest childhood poetry has had that power to transpierce and transport me’ (‘Elle ne pratique point nostre jugement: elle le ravit et

Montaigne’s Unfinished Text

135

Figure 6.2 Continued.

ravage. . . . Dés ma premiere enfance, la poësie a eu cela, de me transpercer et transporter’) (F171, V232). Poetry lifts him to the heights that he could not otherwise attain. But these are literary heights, not ethical heights: moral virtue of the degree of Cato’s remains inimitable. Thus whereas the very first

136

Life of Texts

version of the essay strikes the contrast between the essayist and the younger Cato, the final version becomes fully a comment about Montaigne himself. He is now the protagonist in both parts of the essay. In the chapter ‘Of the Education of Children’ (‘De l’institution des enfans’), he says that he speaks of others in order to speak better about himself (F108, V148), and this is exactly exemplified in the final state of ‘Of Cato the Younger’. We shall return to that same idea – speaking of others in order to speak about oneself – in due course. We must first turn, however, to deal with editorial matters, beginning with the extraordinary document whose importance we can see in ‘Of Cato the Younger’, a document usually designated by the letters EB . These two letters refer for Montaigne specialists to an object of almost venerable status in French Renaissance literature – the exemplaire de Bordeaux, a copy of the 1588 edition of Montaigne’s Essais heavily annotated in the author’s own hand.4 Prepared by the essayist between 1588 and 1592 with a view to a new edition of his work, and now preserved, as the name suggests, as a national treasure in the vaults of Mériadeck, the bibliothèque municipale in Bordeaux, EB is a rare survivor in a print world that did not hesitate to destroy an author’s copy once its usefulness had past. It has suffered from the ravages of time; when rebound in the seventeenth or eighteenth centuries, its margins were shaved by the binder’s knife, cutting off the edges of a number of Montaigne’s addenda, and it has taken astute editorial work (usually by recourse to the 1595 edition of the Essais) to restore the missing words and letters. Following Montaigne’s death in 1592, EB was eventually deposited by the essayist’s daughter, Léonor, with the religious order in whose monastery Montaigne had been buried, and was only rediscovered in 1772. Editors from the early nineteenth century onwards recognized the importance of EB . It was, however, the twentieth century that saw in it the true source of Montaigne’s text, an origin whose authority was uncontested. Foremost among a number of important publications in the early 1900s, the so-called Municipal Edition of the Essais, published by Fortunat Strowski, François Gébelin and Pierre Villey between 1906 and 1933, monumentalized the status of EB in five massive volumes that took as their base text the 1588 edition of the Essais, with the variants of the editions from 1580 onwards given at the bottom of each page, and Montaigne’s post-1588 marginalia added in italics to distinguish them from the body of the printed text.5 This enterprise was remarkable not only by its scale, but also by the dramatic nature of the collaboration: François Gébelin, who volunteered to fight in the Great War, was badly wounded in 1915 and sent in much of his contribution from a hospital bed, while Pierre Villey had been blind since the age of four and yet, until his tragic death in a train crash in 1933,

Montaigne’s Unfinished Text

137

continued to produce a large number of scholarly works on Montaigne and editions of the Essais. It was indeed Villey’s more compact edition of the Essais, originally published in 1922–3, but re-edited by V.-L. Saulnier in 1965, that was to have lasting influence as a scholarly edition of Montaigne in the mid-twentieth century. It shared this privilege with the 1962 ‘Pléiade’ edition of Albert Thibaudet, re-edited by Maurice Rat. Both editions exhibit similar editorial features: Montaigne’s addenda had their distinctive spelling and punctuation altered to fit in with the printer’s typographical conventions of the 1588 edition; they were now integrated into the text, not kept separate as in the Municipal edition, and each element of the Essais was labelled A, B or C (or similar) to designate respectively the text of 1580–7, the text of 1588 and Montaigne’s post-1588 addenda. Each chapter was also divided into paragraphs, a feature that, with rare exceptions, was not present in the original editions and frequently distorted the work’s meaning, but had been introduced in the eighteenth century, an innovation of Pierre Coste in his editions of 1724 onwards.6 These editorial principles were widely accepted and were reproduced in cheaper editions of the Essais that were frequently reprinted in the last forty years of the twentieth century. It was in such a composite form that Montaigne was known and used by generations of twentieth-century scholars, teachers, students and readers, even though the text thus established corresponded to none that had appeared in the essayist’s lifetime nor even exactly to any that appeared immediately after his death. By the 1990s, the limitations of this approach to editing the Essais became apparent. In a series of brilliant studies, André Tournon demonstrated that no modern edition had respected one of the most significant features of EB : its unusual and deliberate punctuation.7 His close analysis of EB showed that, over and above the handwritten additions that increased the length of the 1588 edition by approximately a third, Montaigne had made up to 9,000 changes to punctuation, following a principle of ‘staccato language’ (‘langage coupé’) that was not an effect of chance, but part of the essayist’s written instructions to his printer also preserved on EB . With his characteristic combination of precision and acute attention to the workings of the Essais, Tournon re-sensitized us to the newly invigorated rhythms and cadences produced by Montaigne’s segmentation of his work, with their ‘effects of scansion’ as he labelled them, bold articulations and subtle shifts in emphasis resulting in redefined balances and asymmetries. An example from a famous passage in ‘Of Repentance’ (‘Du Repentir’) will show this process at work:

138

Life of Texts

Je ne peints pas l’estre. Je peints le passage: non un passage d’aage en autre . . ., mais de jour en jour, de minute en minute. Il faut accommoder mon histoire à l’heure. [. . .]. C’est un contrerolle de divers et muables accidens et d’imaginations irresoluës et, quand il y eschet, contraires: soit que je sois autre moy-mesme, soit que je saisisse les subjects par autres circonstances et considerations. Tant y a que je me contredits bien à l’adventure, mais la verité, comme disoit Demades, je ne la contredy point. V805 (I do not portray being: I portray passing. Not the passing from one age to another, but from day to day, from minute to minute. My history needs to be adapted to the moment. [. . .]. This is a record of various and changeable occurrences, and of irresolute, and, when it so befalls, contradictory ideas: whether I am different myself, or whether I take hold of my subjects in different circumstances and aspects. So, all in all, I may indeed contradict myself now and again; but truth, as Demades said, I do not contradict.) F611

Thus Villey/Saulnier. Here is EB : Ie ne peints pas l’estre. Ie peints le passage: Non vn passage d’aage en autre . . . : mais de iour en iour, de minute en minute. Il faut accommoder, mon histoire à l’heure. [. . .]. C’est vn contrerolle de diuers & muables accidens, & d’imaginations irresoluës. Et quand il y eschet, contraires: Soit que ie sois autre moymesme: Soit que ie saisisse les subiects, par autres circonstances, & considerations. Tant y a, que ie me contredits bien à l’aduenture, Mais la verité, comme disoit Demades, ie ne la contredy point.

The changes made by Montaigne to his own text – here highlighted for the sake of clarity – transform the balanced periods of the previous version by reinventing these sentences as short clausulae, thus giving greater prominence to the alternatives, contradictions and paradoxes that the essayist is discussing. In 1588, this passage sets out those paradoxes as a fact for the reader’s consideration. After 1588, the re-punctuation enables the reader to witness the means by which the paradoxes are discovered and not just presented, ending in the saying of the Athenian orator Demades whose tensions Montaigne makes his own through the emphatic re-punctuation. In this example, as in countless others, repunctuation slows down the activity of reading, compelling us to pay attention to the grain of the text and the colouring of individual clauses and segments, so that thought in process is revealed by language in motion; the proliferation of capital letters, full stops, commas, colons and semi-colons makes the texture of

Montaigne’s Unfinished Text

139

the passage purposely less smooth, favouring sinewy sentences that modify their data as they proceed. Meaning in the Montaignean sentence is thus created by an opening up of new vistas within the individual phrase or passage. No sentence by Montaigne can be taken for granted as a once-and-for-all product: it contains within itself the potential for its own disruption, as well as for its own forward momentum. The effect of Tournon’s work can be seen not only in his own edition of the Essais, but also in the more recent edition by Emmanuel Naya and others which likewise follows EB .8 By contrast, Jean Céard, the eminent French scholar, published an edition preferring to follow the 1595 text, and the same policy was adopted by the editors of a new ‘Pléiade’ Montaigne in 2007.9 The gulf between the two camps remains as wide as ever. In the meantime, the Villey/Saulnier edition remains the fall-back choice for the French text. This modern attentiveness to what we might call the philological dimension of editing the Essais has in turn highlighted the role of Montaigne’s first editor, Marie de Gournay (1565–1645).10 She was his fille d’alliance (usually translated ‘adoptive daughter’ or sometimes ‘covenant daughter’). The two met when the essayist was in Paris in or around 1588 and he then went to stay with her in her family home in Picardy between June and November 1588. It was clearly mutual admiration at first sight. Gournay was named as Montaigne’s executor along with a minor poet from Bordeaux, Pierre de Brach, but there is no doubt that it was she who had the major influence on the after-life of the Essais. In chapter 2.18, ‘Of giving the lie’ (‘Du Démentir’), Montaigne writes, ‘I have no more made my book than my book has made me, a book consubstantial with its author’ (‘Je n’ay pas plus faict mon livre que mon livre m’a faict, livre consubstantiel à son autheur’) (F504, V665). The same idea of consubstantiality could equally apply to Gournay’s devotion as an editor of the writer she called ‘my father’. For a period of more than forty years, from the 1590s certainly up to the 1630s and perhaps beyond, she worked tirelessly to produce an edition of the Essais with which she finally felt any satisfaction. Her meticulous correction of editions throughout this period strongly recalls Montaigne’s own incessant reworkings of his book; the Essais became for her, as they had for the essayist himself, ‘an integral part of my life’ (‘membre de ma vie’) (F504, V665). This vigilant activity belies the notion that she was careless or inadequate, an unspoken assumption of those who preferred EB to her editions. Critics in the final years of the twentieth century and the start of the twentyfirst reopened the whole question of the nature and purpose of EB and its relationship to the editions of 1595 onwards prepared by Gournay.11 Scholars

140

Life of Texts

such as Strowski, Villey and Saulnier had noticed the discrepancies between EB and the first posthumous edition of the Essais in 1595 published by her.12 Not unreasonably it may seem, all had preferred the evidence of the essayist’s own work when preparing their editions and had taken EB to represent his definitive wishes. However, it is generally accepted that EB was not the only copy bearing Montaigne’s additions and alterations. There would have been at least another, sometimes referred to as either a fair copy or an exemplar, which, it is agreed, was in fact the copy used for the printing of the 1595 edition.13 This second copy is no longer extant, but it is not unreasonable to infer its existence. Indeed, for the more recent French scholars such as Simonin and Balsamo who championed her cause as an editor, the disparate states of the various textual witnesses of which Gournay endeavoured to make sense after Montaigne’s death complicate the stemma of the Essais and relativize the status of EB . In other words, these French scholars took Gournay’s enterprise seriously. Any purposeful critical discussion of the Essais had to be based, they argued, on her 1595 edition; EB was at best a witness to that text, it remained a personal copy on which Montaigne seems to have continued to make additions and alterations. EB , for these scholars, was not in itself a definitive text; at the very most it was a prelude to such a text. Gournay’s own relationship to EB is complex. The usual argument is that she did not see it when preparing the 1595 Essais for publication and certainly the printed version of 1595 does not correspond exactly to EB . This is notably the case for the punctuation, as Tournon has pointed out in some detail. It is sometimes held that Gournay did not have sight of EB until her stay at the chateau de Montaigne from May 1595 to July 1596, that is after the first publication of the posthumous edition. Yet even her 1598 edition of the Essais, which corrects some of the oversights of 1595, still does not entirely replicate EB . This suggests that she had strong reasons for preferring the fair copy she prepared for the posthumous edition, in other words she believed the fair copy best reproduced Montaigne’s wishes for that edition, an edition he did not live to see. However, questions remain. For example, in three places we can plainly see Gournay’s interventions on EB itself.14 Fig.  6.3 shows an example. What are we to make of these interventions? Legros argues that Gournay wrote them under Montaigne’s dictation and notes that in at least one instance the initial letters are in his hand.15 Confirming Simonin’s reasoning that Montaigne was already starting to revise his work with Gournay’s help,16 Legros dates the dictations to 1588 when Montaigne stayed at Gournay’s house in

Montaigne’s Unfinished Text

141

Figure 6.3 EB, fol. 47r. Gournay’s handwriting (right) and Montaigne’s (bottom margin).

Picardy. In that case, Gournay had had sight of EB before the publication of 1595 but may have thought of it as a working copy – a sort of essay about essays, a book of experiments – rather than anything that was itself intended for publication.

142

Life of Texts

Figure 6.4 ‘Of Presumption’ (‘De la Præsumption’), Gournay’s manuscript corrections to p. 439.

One further example of Gournay’s interventions brings us to one of the most significant pages in chapter 2.17, ‘Of Presumption’ (‘De la Præsumption’). Of crucial importance here is a hitherto unrecorded copy of the 1595 Essais held in Maynooth, which is in fact a presentation copy made by Gournay to someone whose identity will be discussed later.17 Figure 6.4 shows the extensive corrections she made to this page. Despite some cropping which occurred during re-binding, her intentions are clear. The page itself has been poorly set and partly smudged, and she has scrupulously inked in many faint letters, including in sentences which she has subsequently crossed out. Yet it is her alterations to the 1595 printed text itself which are highly significant. As printed in the 1595 edition, before correction, the text reads as follows:

Montaigne’s Unfinished Text

143

I’ay pris plaisir à publier en plusieurs lieux, l’esperance que i’ay de Marie de Gournay le Iars ma fille d’alliance: & certes aymée de moy beaucoup plus que paternellement, & enueloppée en ma retraitte & solitude, comme l’vne des meilleures parties de mon propre estre. Ie ne regarde plus qu’elle au monde. Si l’adolescence peut donner presage, cette ame sera quelque iour capable des plus belles choses, & entre autres de la perfection de cette tres-saincte amitié, où nous ne lisons point que son sexe ait peu monter encores: la sincerité & la solidité de ses mœurs y sont desjà battantes, son affection vers moy plus que sur-abondante: & telle en somme qu’il n’y a rien à souhaiter, sinon que l’apprehension qu’elle a de ma fin, par les cinquante & cinq ans ausquels elle m’a rencontré, la travaillast moins cruellement. Le iugement qu’elle fit des premiers Essays, & femme, & en ce siecle, & si ieune, & seule en son quartier, & la vehemence fameuse dont elle m’ayma & me desira long temps sur la seule estime qu’elle en print de moy, auant m’avoir veu, c’est un accident de tres-digne consideration. Les autres vertus ont eu peu ou point de mise en cet aage: mais la vaillance est devenue populaire par noz guerres ciuiles: & en cette partie, il se trouue parmy nous, des ames fermes, iusques à la perfection, & en grand nombre, si que le triage en est impossible à faire. Voila tout ce que i’ay connu, iusques à cette heure, d’extraordinaire grandeur & non commune. V661–2 (I have taken pleasure in making public in several places the hopes I have for Marie de Gournay le Jars, my covenant daughter ( fille d’alliance), whom I love indeed much more than a daughter of my own, and cherish in my retirement and solitude as one of the best parts of my own being. She is the only person I still think about in the world. If youthful promise means anything, her soul will some day be capable of the finest things, among others of perfection in that most sacred kind of friendship which, so we read, her sex has not yet been able to attain. The sincerity and firmness of her character are already sufficient, her affection for me more than superabundant, and such, in short, that it leaves nothing to be desired, unless that her apprehension about my end, in view of my fifty-five years when I met her, would not torment her so cruelly. The judgment she made of the first Essays, she a woman, and in this age, and so young, and alone in her district, and the remarkable eagerness with which she loved me and wanted my friendship for a long time, simply through the esteem she formed for me before she had seen me, is a phenomenon very worthy of consideration. The other virtues are given little or no value nowadays; but valor has become common through our civil wars, and in this respect there are among us souls firm to the point of perfection, and in great numbers, so that a choice is impossible. This is all the extraordinary and uncommon greatness I have known up to this moment.) F502

144

Life of Texts

In the following transcription, Gournay’s deletions are represented by a plain black line, while, for the sake of clarity, insertions, additions and reorganization are in italic; supplied letters and words missing through cropping are in parentheses. Les autres vertus ont eu peu ou point de mise e[n] cet aage, mais la vaillance est deuenuë populaire par nos guerres ciuilles: & en cette partie, il se trouue des ames fermes iusques à [la] perfection & en grand nombre: si que le tria[ge] en est impossible à faire. I’ay pris plaisir à publier en plusieurs lieux, l’esperance que i’ay de Marie de Gournay le Iars ma fille d’alliance: & certes aymée de moy beaucoup plus que paternellement, & enueloppée en ma retraitte & solitude, comme l’vne des meilleures parties de mon propre estre. Ie ne regarde plus qu’elle au monde. Si l’adolescence peut donner presage, cette ame sera quelque iour capable des plus belles choses, & entre autres de la perfection de cette tres-saincte amitié, où nous ne lisons point que son sexe ait peu monter encores: la sincerité & la solidité de ses mœurs y sont desjà battantes, son affection vers moy plus que surabondante: & telle en somme qu’il n’y a rien à souhaiter, sinon que l’apprehension qu’elle a de ma fin, par les cinquante & cinq ans ausquels elle m’a rencontré, la travaillast moins cruellement. Le iugement qu’elle fit des premiers Essays, & femme, & en ce siecle, & si ieune, & seule en son quartier, & la bienueillance q[u’elle] me voüa, vehemence fameuse dont elle m’ayma & me desira long temps sur la seule estime qu’elle en print de moy, long temps a[uant] qu’elle m’eus[t vu,] sont des accide[nts de] tresdigne conside[ration.] auant m’auoir veu, c’est un accident de tres-digne consideration. Les autres vertus ont eu peu ou point de mise en cet aage: mais la vaillance, elle est devenue populaire par noz guerres ciuiles: & en cette partie, il se trouue parmy nous, des ames fermes, iusques à la perfection, & en grand nombre, si que le triage en est impossible à faire. Voila tout ce que i’ay cognu, iusques à cette heure, d’extraordinaire grandeur & non commune.

This would produce the following final text: Les autres vertus ont eu peu ou point de mise e[n] cet aage, mais la vaillance est deuenuë populaire par nos guerres ciuilles: & en cette partie, il se trouue des ames fermes iusques à [la] perfection & en grand nombre: si que le tria[ge] en est impossible à faire. I’ay pris plaisir à publier en plusieurs lieux, l’esperance que i’ay de Marie de Gournay le Iars ma fille d’alliance: & certes aymée de moy plus que paternellement. Si l’adolescence peut donner presage, cette ame sera quelque iour capable des plus belles choses. Le iugement qu’elle fit des premiers Essays, & femme, & en ce siecle, & si ieune, & seule en son quartier, & la bienueillance q[u’elle] me voüa, sur la seule estime qu’elle en print de moy, long temps a[uant]

Montaigne’s Unfinished Text

145

qu’elle m’eus[t vu,] sont des accide[nts de] tresdigne conside[ration.] Voila tout ce que i’ay cognu, iusques à cette heure, d’extraordinaire grandeur & non commune. (The other virtues are given little or no value nowadays; but valor has become common through our civil wars, and in this respect there are among us souls firm to the point of perfection, and in great numbers, so that a choice is impossible. I have taken pleasure in making public in several places the hopes I have for Marie de Gournay le Jars, my covenant daughter ( fille d’alliance), whom I love indeed more than a daughter of my own. If youthful promise means anything, her soul will some day be capable of the finest things. The judgment she made of the first Essays, she a woman, and in this age, and so young, and alone in her district, and the good will she showed me, simply through the esteem she formed for me before she had seen me, are phenomena very worthy of consideration. This is all the extraordinary and uncommon greatness I have known up to this moment.)18

This is a unique manuscript reworking of a famous passage; it has no parallel in the editorial annotations on Montaigne discovered in Gournay’s own hand up to the present day. To deal, first, with the reorganization of the text, the transfer of the penultimate sentence of 1595 to the start of the extract greatly improves the logic and flow of the thought. In the previous sentence, Montaigne has been praising the humane conduct of François de La Noue, one of the most distinguished Protestant commanders during the French Wars of Religion; his further observation about valour, elevated at the expense of the other virtues, now follows on and his comments run together seamlessly. In fact, on EB , his train of thought had already demonstrated this coherence. What had disrupted the coherence in the 1595 printing was the passage praising Gournay, which was inserted between the comment about La Noue and the observation about valour. Moreover, while this praise of Gournay is present in editions of 1595 Essais, its absence in EB , as shown in Fig.  6.5, led to speculation that it may have been interpolated later, after Montaigne’s death, perhaps even by Gournay herself.19 However, as can equally be seen, this page in EB is discoloured and it may be that there was a stuck-on inserted piece of paper which has subsequently disappeared: this appears to have happened in at least one other place in EB .20 There is no incontrovertible evidence that Montaigne’s scrupulous editor added the passage about herself and as we may reasonably surmise, the revised manuscript of the Essais she worked from for publication in 1595 was not identical to the exemplaire de Bordeaux.21 It is perfectly possible that the passage appeared in that revised manuscript, which we no longer possess. At any event,

146

Life of Texts

Figure 6.5 ‘Of Presumption’ (‘De la Præsumption’), EB , fol. 292v, without the additional passage about Gournay.

the quoted passage from 2.17 undergoes radical alteration on the Maynooth copy. Its exalted tone has been considerably muted and its verbal extent radically reduced. The 1595 highlighting of the friendship between Montaigne and Gournay, Gournay as the focus of the essayist’s attention (‘she is the only person I still think about in the world’), her sincerity, her sterling character, her abundant

Montaigne’s Unfinished Text

147

affection for Montaigne, the vehemence with which she loves him. . . all this has vanished, leaving only a new term she inserts on the Maynooth copy, ‘bienueillance’, the ‘good will’ (or ‘regard’) she bore him even before meeting him. It so happens that these manuscript corrections exactly match the revisions to this page published in the 1625 edition of the Essais, which carries the amended text as seen on the Maynooth copy.22 The revised text for this passage of 2.17 was followed subsequently in the 1635 Gournay edition, yet with a further shortening of a key phrase ‘whom I love indeed much more than a daughter of my own’ to ‘whom I love indeed like a daughter of my own’, which completely changes the sense.23 A further fact clinches the parallel with 1625. In that edition, as in the Maynooth copy, a passage has become displaced: The other virtues are given little or no value nowadays; but valor has become common through our civil wars, and in this respect there are among us souls firm to the point of perfection, and in great numbers, so that a choice is impossible.

This passage followed the praise of Gournay in 1595 and subsequent editions. It precedes the praise of Gournay in Maynooth and in the 1625 edition. By 1635, it has been restored to its original place, after the praise of Gournay. This confirms that the appropriate link we should make is between the Maynooth copy and the 1625 edition of the Essais; the likelihood is that Maynooth was corrected to follow the 1625 text. Moreover, it seems certain that Gournay herself did not compose the whole passage she amended, because at the end of her preface to the 1625 edition she states as follows: En ce seul point ay-ie esté hardie, de retrancher quelque chose d’vn passage qui me regarde: à l’exemple de celuy qui mit sa belle maison par terre, affin d’y mettre auec elle l’enuye qu’on luy en portoit. Ioinct que ie veux dementir maintenant & pour l’aduenir, si Dieu prolonge mes années, ceux qui croient; que si ce Liure me loüoit moins, ie le cherirois & seruirois moins aussi.24 (In this one point alone have I made so bold as to cut out something from a passage which concerns me; following the example of one who demolished his fine house so as to do away, along with it, with the envy that was directed against it. In addition, I wish to give the lie, now and for the future, if God extends my years, to those who believe that if this Book praised me less, I would cherish and serve it less also.)25

This statement, which is the counterpart of a parallel personal declaration in the 1595 Essais, appeared for the first time in the preface to the 1625 Essais.26

148

Life of Texts

Combined with all the particularities together, we can be sure that the Maynooth copy has more in common with the 1625 Essais than with the 1635. Gournay’s 1625 edition of Montaigne formed, in fact, part of an editorial campaign which stretched from 1617 to 1635, in the three major printings she oversaw in Paris (1617, 1625 and 1635). In the 1617 edition, she reinstated the long preface found in the 1595 edition, subsequently reworking it in 1625 and 1635, as a sign of her renewed attempts to produce an edition of the Essais with which she could feel some satisfaction. There was a corresponding distinctive change of size from the prevalent octavo editions of the early years of the seventeenth century to quarto (1617 and 1625, recalling the format of the 1588 Essais) and finally folio (1635, echoing the 1595 Essais). Critics have commented on the shifts in emphasis which she makes to the prefaces of the Essais in 1617 and 1625, noting in particular that the total identification between Gournay and Montaigne which has been posited in 1595 disappears in 1625. This description also fits perfectly Gournay’s changes to the final section of ‘Of Presumption’ on the Maynooth copy. The pride and emotion of the 1595 statements are considerably reduced. Once again, critics have observed that Gournay’s development of her own career is not necessarily at odds with her admiration for Montaigne; but that admiration is now recalibrated according to new imperatives. Discussing the changes to the preface of the 1617 edition, which served as a basis for 1625, Mary McKinley, for instance, writes as follows: The 1617 revisions point beyond the preface to the woman whose identity is no longer simply that of Montaigne’s fille d’alliance (adoptive daughter). The preface is one of several texts that a busy literary figure, writer as well as editor, is engaged in producing.27

In this light, it is highly significant that pp. 587–8 in the 1625 edition, and p. 439 in the Maynooth Essais, leave part of one particular sentence unchanged: ‘The judgment she made of the first Essays, she a woman, and in this age, and so young, and alone in her district.’ The emphasis is now firmly on Gournay’s own intellectual qualities rather than simply on the relationship between herself and Montaigne. Each of the expressions in the phrase Gournay purposely leaves unaltered takes on its full importance, all the more so since the phrase echoes the weight she lends the term ‘judgement’ in the 1625 preface: ‘The gift of judgement is the thing that humankind possesses in diverse manner. . . . Judgement alone lifts humans above animals, Socrates over them and the angels over them. . .’28 No less crucial is her standing as a woman in her century: her editorial and literary activities are now specifically related, as McKinley rightly says, to her female

Montaigne’s Unfinished Text

149

identity, her greater self-assurance, and to the publications she undertakes in one of the most intensely productive periods of her life, the 1620s.29 From another inscription on the Maynooth copy, it has been possible to identify the recipient of this presentation volume. It was Henri de Beringhen (1603–92), a court favourite of Louis XIII , who succeeded in his turn to his father’s position of ‘premier valet de chambre’.30 He was exiled from France by Richelieu in 1630, supposedly because he refused to reveal to Richelieu a secret that Louis XIII had entrusted to him. He served abroad with distinction in the armies of Gustavus Adolphus of Sweden and Frederick Henry of Orange before again being embroiled in intrigue in 1637. On that occasion, Henri de Beringhen fled to Brussels, but was recalled to France in 1643, shortly before the death of Louis XIII . In 1645, he was promoted under the regency of Anne of Austria, and through her influence, to the office of the King’s first equerry, as a result of which he was subsequently known at Court as ‘Monsieur le Premier’. Gournay’s connection with Beringhen is, however, less puzzling than might at first sight appear. His aunt, Marie des Loges, held a notable salon in the rue de Tournon in Paris, which Gournay frequented during the 1620s (and perhaps before) and where she met writers such as Malherbe, Voiture, Guez de Balzac and Godeau.31 Furthermore, Marie des Loges was the dedicatee of Gournay’s Defence of Poetry and of the Language of Poets, which appeared in 1626 and subsequently in 1634 and 1641. This dedication thus acts as proof of both a personal and an intellectual connection between Gournay and her literary hostess. The likelihood also is that Gournay met Beringhen at his aunt’s salon, although of course this would not exclude other forms of acquaintance. She describes her gift to him as a ‘present’, the term which also features in the title of her 1634 collection, Counsels or Presents (Les Advis, ou Les Presens), where particular individual treatises are regarded as a gift with their own recipientdedicatee. Her present of the Essais to Beringhen thus parallels her present to his aunt. It may also be no coincidence that in 1622 Gournay dedicated her treatise on the equality of men and women to Anne of Austria, with whom Beringhen was connected. This treatise too was shortly afterwards collected in 1626 and later in 1641. So far, then, the Essais have played a key role in the formation of Gournay’s career as a writer and they do so by offering an important relationship which, as we saw earlier, can be described as a consubstantiality between the author, Montaigne, and one special reader, his covenant daughter. Yet the paradoxical result of this consubstantiality is not greater dependence, but a process by which Gournay comes to know her own independent creative powers, exercise her own

150

Life of Texts

judgement and, from the 1620s onwards, feels able to revise substantially her relationship with her ‘father’. The life of the Essais lies in the capacity it bestows on its readers to reshape themselves, by offering more than intellectual affinity. And consubstantiality is offered between the author of the book and its recipients; it is something the Essais creates with its readers, not something it inherits from its intellectual genes. In the final part of this chapter, we turn to two non-French instances of this phenomenon. The first is the German jurist, Christoph Besold (1577–1638), seven times rector of the University of Tübingen and the author of over 100 works.32 Montaigne is a vibrant presence throughout his writings; the essayist is ‘magnus Montanus’ (the great Montaigne) or ‘Magnus Michaël Montanus’ (the great Michel de Montaigne), ‘Gallorum Socrates’ (the French Socrates), ‘seculi nostri inter Philosophos Phœnix’ (the philosophical phoenix of our age).33 The preface to The Beginning and End of Political Doctrine (Principium et finis politicæ doctrinæ), first published in 1625, shows that process of consubstantiality: Adhuc dum versor in novorum Academicorum Schola: Quæro omnia, dubitans plerunque & mihi diffidens. Sæpè non tàm id sensi, quod dictum à me est: quàm exercere ingenium volui, materiæ difficultate. vide Montaig.lib.2.cap.12. fol.m.467.&c.&.fol.480. Pherecides morti appropinquans, Thaleti sua scripta commendavit: de iis ita sensit, nullam ea certitudinem continere, quâ sibimetipsi satisfacere possint. Idem ego censeo de meis. Acquiesco solùm in perpetuo Ecclesiæ vetustioris consensu: sed eum adhuc indago.34 (While I am still at the school of the New Academics: I search into everything, doubting most things and mistrusting myself. Often, it is not so much that I have believed what I have said as that I wanted to exercise my wits on the difficulty of the material. See Montaigne book 2, chapter 12, page 467 in my copy etc and page 480. Pherecides drawing close to death entrusted his writings to Thales: his opinion of them was they contained no certainty with which he might be satisfied. I think the same about my writings. I assent only to the unbroken agreement of the older Church: but I still inquire into it.)

With its dense echoes and patterning, this extract is a complex exercise not just in miming Montaigne’s textual technique or ventriloquizing his manner, but more particularly in adopting his stance by reflexively meditating on his own intellectual enterprise, all supported by precise page references to his own copy of the Essais.35 To that end, Besold combines and recombines quotations from Montaigne’s longest chapter, ‘The Apology of Raymond Sebond’, as part of his understanding of what Academic Scepticism means to him, but he also brings

Montaigne’s Unfinished Text

151

the first-person pronoun into prominence: that first-person pronoun – a firstperson ending in Latin, of course – now coordinates all the diverse quotations, whose syntax is reworked as necessary to foreground the singular voice of the lawyer-philosopher. It is no longer Cicero or Quintilian or Montaigne who predominates, even though they are plainly visible in this passage,36 but Besold, and the composite passages of which this extract is composed are marshalled into order by the constantly repeated ‘I’ of the writer (‘I am at school’, ‘I search’, ‘I have believed’, ‘I have said’, ‘I wanted’, ‘I think’, ‘I assent, ‘I inquire’). As with Gournay, one ‘I’ (Montaigne’s) incubates another (Besold’s); consubstantiality has most fully performed its task when writers of the succeeding generation simultaneously signal their affinity with the French essayist and yet mark their independence from him. They have most become like Montaigne when they have most become themselves. Another lawyer, this time in mid-seventeenth-century England, offers an (if anything) even more remarkable appreciation of Montaigne. Sir Ralph Bankes of Kingston Lacy was MP for Corfe Castle and had shortly before been a member of Gray’s Inn, a distinction he shared with John Locke, another reader of Montaigne.37 A rear flyleaf of his copy of the Paris 1657 Essais contains his account of Montaigne precisely dated 12 August 1659. Account of the Booke And Author Mich. De Montaigne He was A Man of A profound Judgement and quicke Apprehention, A greate Humorist, and mutch wedded to his owne wayes & fancies. His Wrightings are full of Many Excellent Quotations, And Noe lesse abound with Rarities of his owne growth, his Language is very Apt & significant, (and for the time hee wrote) Elegant, Hee takes A greate freedome in Expressing himselfe, and *gives every thing itts owne Name without disguise, disaproving the Modesty of our Age in Bookes and discourse wch hee calls (Mauvaise Honte) Hee gives A Particular and Minute account of his Mind And Body and decends to Particular and Private actions, Hee was A greate Vser of Woemen and temperate in all other things, A greate Ennemy to Physick and Physitians, showing good reason for itt, and Enioyed his health wthout the Helpe of Either. There was A greate Evenesse and Constancy in his Mind and actions and hee lived A most Happy and pleasant life being alwayes pleased and satisfied wth himselfe, Acheve de lire ce livre att Shipton the 12th of August 1659.38 *Non pudeat dicere, quod non pudeat sentire (One should not be ashamed to say what one is not ashamed to think)

152

Life of Texts

The keynote is struck by the initial mention of judgement and apprehension, but the passage swings between formal characteristics such as judgement and apprehension at the beginning and ‘evenness and constancy’ at the end, and, on the other hand, ‘ways & fancies’ and humours. In this passage in which everything is double and doubled: ‘book and author’, ‘judgement and apprehension’, ‘mind and body’, ‘mind and actions’, quotation is supplanted by the earthy particularity of the body and then itself replaced by ‘Evenesse and Constancy’, which is as much the expression of Montaigne’s ‘freedom’ as are his ‘wayes & fancies’; all leading to ‘A most Happy and pleasant life’. The Stoic tonality of those final lines must not lead us astray: the quirks and particularities remain. Note for example, ‘Hee was A greate vser of Woemen and temperate in all other things’: it is one expression of the contradictory tensions that are Montaigne. The freedom of expression of his book is a freedom to detail his private and particular actions. The writer becomes himself through fusion with his work. This sketch of the book and author is a picture of consubstantiality as one early modern English reader understands it. If it appears a portrait at one remove, an account of someone else, it nonetheless could only be written by someone such as Bankes who has understood what Montaigne is trying to achieve by consubstantiality with himself and what the French writer can offer his readers by way of a model. In that light, the line ‘Hee was A greate vser of Woemen and temperate in all other things’ may seem a far more likely occurrence in the mid-seventeenthcentury English squire than it does in the author he was reading. Looking back, then, to the chapter ‘Of Cato the Younger’, we can see that Montaigne’s contemporary and subsequent readers perform in respect of him what he himself executes in respect of his classical or in some cases modern antecedents. ‘I speak of others only the better to speak of myself ’, was, we recall, the essayist’s dictum and the chapter in which he makes that point, ‘Of the Education of Children’, proved a perennially popular one with early modern readers and not only because it could be applied to the upbringing of the heirs of the aristocracy.39 More particularly, that chapter implements an ethical ideal by which the student does not repeat or mimic her or his reading, but uses the Essais to shape independence of personality and autonomy of judgement. The consubstantiality Montaigne offers with himself is premised on the view that friendship will involve precisely that practice, as we saw in the case of Gournay, Besold and Bankes. All three read Montaigne in the French original – unsurprisingly in Gournay’s case, more interestingly in the other two who thus provide evidence of the influence of his work in the original language outside France. In at least two of the three instances we have considered, that project also

Montaigne’s Unfinished Text

153

relates to the formation of the writer: Gournay in one instance, Besold in another; on the other hand, Besold and Bankes are both lawyers and may well feel an affinity with Montaigne on that score. Thus to speak of others in order to speak of oneself is a flexible principle: it can relate to a child of the mind as much as to a child of the flesh, but in all cases it connects the constitution of the human being with the acquisition of knowledge and the development of judgement and its means of expression. Even removing Montaigne’s sudden death from the picture, the Essais would thus, in an important sense, remain permanently unfinished, not only because they could be constantly added to (an aspect that was clearly anticipated and practised by their author), but because they envisage each reader continuing their project in herself or himself. The after-life of this text lies with its centuries-long generations of readers. For to understand the Essais – not merely to read them, but to understand them – is to undergo a process of change, to be transformed into oneself, not as a fully finished object, but through a process of perpetual becoming. The enthusiasm with which Montaigne’s admirers from the earliest days have embraced this idea remains a testimony to the extraordinary nature of the work which the essayist produced.

7

Rescuing Shakespeare: King Lear in Its Textual Contexts David Fuller

There are two different substantive texts of King Lear: the Quarto of 1608 and the Folio of 1623. The differences between these two texts are not small: the earlier has dozens of mangled lines and downright errors; the later, along with many other differences, omits a major part of one of the most important scenes and much else besides; and the focal point of the drama, the way in which Lear dies, is seriously affected by the choice of text. Characterization, tone and rhythm, local and structural, are affected by how the two texts are combined (if combined they are). The current Oxford edition of individual Shakespeare plays prints the earlier text; the current comparable Cambridge edition prints the later text; and the current Arden edition combines the two, though often in unsatisfactory ways.1 So of the three main scholarly editions of Shakespeare, none gives what is, in the view of textual scholarship up to the 1980s, and in the view of many current scholars, a proper text of one of the greatest works written in English. The reason for this lack of agreement between editions is the revival since the 1980s of the theory that differences between the Quarto and the Folio texts arise not from textual corruption, which needs to be corrected, but from authorial revision, which demands to be respected.2 The issue here is not radically rewritten versions of a work, such as the A, B, C and Z texts of Langland’s Piers Plowman, the three- and four-book Dunciad of Pope, the 1799, 1805 and 1850 texts of Wordsworth’s The Prelude, or the original and ‘New York’ editions of Henry James. With Shakespeare we have to consider texts that contain significant but more minor variants, the grounds of which are often uncertain. Also at stake in the differences between contemporary texts of King Lear are general issues about the transmission and editing of Shakespeare that in different degrees affect all his work. It is from those general issues that a proper understanding of the problems of King Lear needs to begin – those general issues seen in context. 154

Rescuing Shakespeare

155

* Sceptical innovators are vain men, and will gratify themselves at any expense. Truth will not afford sufficient food to their vanity; so they have betaken themselves to error. Truth, Sir, is a cow that will yield such people no more milk, and so they are gone to milk the bull. James Boswell, Life of Samuel Johnson, 21 July 1763

It may be that this melancholy reflection is all too generally applicable to the study of the arts in universities. Certainly there are points of view from which over the past forty years a good deal of the study of Shakespeare’s text has been milking the bull. But Johnson’s example of a vain innovator was David Hume. Great man as Johnson was, a just judgement of Hume was beyond his scope. One difficulty of the current study of Shakespeare’s text is over-confident polemic. In a situation where there are no proofs, where all arguments involve elements that must finally be no more than plausible postulates, it is important to balance conviction based on an impartial view of facts – a view, that is, not driven by a theory of what it wishes to find – with a complementary scepticism. It is also important to recognize – as textual scholars are not always willing to do – that cultivated and informed literary judgement, and practice tested in the theatre, are as much part of the relevant ‘facts’ as bibliographical information. Literary judgement, theatrical practice and bibliographical knowledge: these are the sine qua non. There are sometimes also less large-minded issues shaping contemporary textual study, connected with the ways in which universities, as the institutions in which scholarship earns a living, are dominated by the science models of knowledge which provide their funding, and so have a built-in bias towards overvaluing novelty. Similarly with publishing houses that wish to be up to date, ‘leading the field’. The effects of this are not simple, and do not admit of full discussion here; but new material, especially new material with a technological aspect (a digital edition, accessed from computer, tablet or smartphone), is a clearer product than that most delicious of intellectual fruits: tried and tested views brought to life anew. Both the funding needs of universities and the natural biases of publishers alike tend to undervalue established positions. But it is part of the work of the study of the humanities to keep alive traditions of thought – traditions of thought responsive to real and disinterested contemporary knowledge, but not simply reshaped to suit a dynamic of perpetual innovation. A scholar-critic’s ability to do this is dependent not simply on scholarship but also on intensity of perception and depth of understanding. The scholar who, knowing ‘everything’ (that is, all the inessentials) understands nothing, is a

156

Life of Texts

familiar figure throughout the history of criticism. Pope gives a vivid account in The Dunciad, and Pope’s eighteenth-century forms are permanent types. Living traditions of thought properly understood may act as commentary on and critique of the present – and so they can, and should, with the textual criticism of Shakespeare.

Quartos and the First Folio No play by Shakespeare exists in an authorial manuscript. No play by Shakespeare exists in a contemporary manuscript of any kind.3 For our knowledge of what Shakespeare wrote we are dependent entirely on contemporary printed texts. The First Folio of 1623 (published seven years after Shakespeare’s death) contains the only substantive texts for half of his plays, including some of the most famous: As You Like It, Twelfth Night, Macbeth, Antony and Cleopatra, The Tempest. The texts underlying its production were various, including (transcripts of) authorial copies, copies made by a professional scribe, playhouse ‘prompt books’ and existing printed texts (quartos) corrected by reference to playhouse sources. The various plays published in quartos over a thirty-year period before the Folio sometimes appeared in forms that are substantively different, both from the Folio and from each other: for example, it was not until Q4 of Richard II (1608) that the play was printed with its abdication-cum-deposition scene. In cases of this kind, there is no issue of authorial revision: the play must always have contained the scene; the earlier texts were cut by the censor for their political sensitivity. Not all the reasons for variation between the early texts are, however, so clear. Some of the quartos are thought to be derived from authoritative sources. These include manuscripts written by Shakespeare himself (the author’s socalled ‘foul papers’ – difficult to read; more readily subject to misreading); fair copies written by a professional scribe (easier to read, but likely to include misreadings of its perhaps authorial original – the scribal peculiarities of one such scribe, Ralph Crane, have been much studied); and what are usually called (anachronistically) company ‘prompt books’ (book-keepers’ copies, more likely to represent what happened in the theatre, a better source for stage directions, but perhaps modified by the company, with or without Shakespeare’s involvement and agreement).4 Texts of this kind can be used to correct errors in the Folio. Some quartos, however, have been thought – and by many scholars still are thought – not to derive from authoritative sources. A standard theory of the

Rescuing Shakespeare

157

unreliable texts, the so-called ‘bad’ quartos, throughout most of the twentieth century was that they were memorial reconstructions by one or more actors, and could therefore be used at most tentatively in correcting errors in texts from more reliable sources. These ‘bad’ quartos – re-identified more neutrally as ‘short quartos’ – are now seen by some scholars as potential evidence about earlier authorial states of their texts. What was once held as the epitome of the ‘bad’ quarto is Q1 of Hamlet (1603), in which the most famous speech in English drama begins as follows: To be, or not to be – ay, there’s the point. To die, to sleep – is that all? Ay, all. No, to sleep, to dream – ay, marry, there it goes, For in that dream of death, when we’re awaked And borne before an everlasting judge From whence no passenger ever returned – The undiscovered country, at whose sight The happy smile and the accursed damned. But for this, the joyful hope of this, Who’d bear the scorns and flattery of the world – Scorned by the right rich, the rich cursed of the poor . . .5

This is but distantly derived from Shakespeare – and though some scholars have persuaded themselves that this garbled, arrhythmic and asyntactic nonsense may show how Shakespeare wrote before he had, as Ben Jonson put it, turned and filed his lines, most readers who are not scholars with a theory but lovers of poetry will accept the long-held view that as any kind of Shakespearean version of the speech this is more than improbable. However, it is also now thought by some to have value as a textual witness, not in detail, but because of its place in the dramatic sequence. It is thought by those who favour the theory of Shakespeare the reviser that, while the reporter could not remember accurately more than a few words in sequence, his report may show correctly the speech’s original position in the play, which is quite different from that in the good quarto and the Folio. One fundamental problem of ‘bad’ (short) quartos is to know how they were produced. It is assumed that they were in some sense stolen – and they would be stolen, not from the author, who had sold his rights to the company, but from the playhouse. But how? It is evident from the poor state of the text in the ‘bad’ quarto of Hamlet that it was not derived from a stolen document, so from what was it derived? It is just possible that plays were stolen by being taken down in shorthand by a reporter in the theatre. There is evidence of a kind for this practice

158

Life of Texts

(‘stenography’) in the Prologue to Thomas Heywood’s If you Know Not Me, You Know Nobody (1605), but it is very doubtful that systems of shorthand used by the Elizabethans could have been adequate to the task, and it is hard to imagine how such a theft could be accomplished without detection. The classic theory of ‘bad’ quartos is that they are ‘memorial reconstructions’ by an actor or group of actors, and in the usual analysis one or two parts (those played by the informants) would be remembered accurately, others less so, scenes in which that actor or those actors did not appear least well. The great early twentieth-century Shakespeare bibliographers, A. W. Pollard and W. W. Greg, working in close collaboration, evolved the theory.6 As a clear-cut case, Greg identified the actor of the role of the Host in the ‘bad’ quarto of The Merry Wives of Windsor as the reporter, because the Host’s role is remembered more fully and accurately than others. Though Greg remained tentative about, even sceptical of, his own theory, it achieved a currency which, out of his hands, hardened into orthodoxy and was used to account for textual problems in non-Shakespearean as well as Shakespearean plays. Characteristics of ‘bad’ quartos are: inconsistent action and dialogue; mangled text – sometimes loose paraphrase, sometimes entirely invented, sometimes remembered out of place; and intrusions – which sometimes comically include fragments of other plays in which the reporter(s) may be presumed to have acted. Since, with normal doubling for all but major characters, an Elizabethan actor might be playing up to fifty parts in a month in twenty or more different plays, and might be reconstructing from memory long after performing a play, it is not surprising that memorially reconstructed texts can be gravely and comically inaccurate. Shakespeare’s friends and colleagues, John Heminges and Henry Condell, in their preface to the 1623 Folio, mention ‘stolne, and surreptitious copies, maimed, and deformed by the frauds and stealthes of iniurious impostors’ (sig. A3), and these were classically thought to be the ‘bad’ quartos, as understood by Pollard and Greg; that is, the first quartos of 2 Henry VI (The First Part of the Contention between York and Lancaster, 1594), 3 Henry VI (The True Tragedy of Richard Duke of York, 1595 [actually an octavo]), Romeo and Juliet (1597), Henry V (1600), The Merry Wives of Windsor (1602), Hamlet (1603) and Pericles (1609). However, the whole theory of ‘bad’ quartos and memorial reconstruction has been variously questioned over the same period as the idea of Shakespeare as a reviser has been revived; and some forms of the variation from acknowledged good texts in ‘bad’ (short) quartos can be used to support that theory. It has been proposed, for example, that some ‘bad’ quartos were abridgements designed for performance by a small cast, perhaps when touring – abridgements in which

Rescuing Shakespeare

159

Shakespeare may have had a hand.7 It is also argued that ‘bad’ quartos may reflect more normal practices of cutting to reduce plays as performed to a length of between two and three hours – but there were no intervals in the Elizabethan public theatres and Elizabethan actors may have spoken more rapidly than is now customary. Attempts to estimate what might be performable within the two or three hours supposed available are therefore problematic, though probably not (for example) the full text of Hamlet – from which (though he apparently took so little interest in the publication of his work) it might be concluded that Shakespeare was writing for readers as well as for play-goers.8 Even the most comprehensive and detailed discussion of all the play texts of the period postulated as memorially reconstructed, Shakespearean and nonShakespearean (forty-one in all, including eleven by Shakespeare), acknowledges that ‘it is easier to argue that a playtext is not memorially reconstructed than to prove that it is’, and nevertheless accepts that a strong case can be made for the theory in relation to a limited number of texts.9 Since some of the texts judged not memorially reconstructed in this account have since been reinstated by editors working on their detail in light of the arguments both pro and contra, the issue of ‘bad’ quartos and memorial reconstruction can scarcely be regarded as other than open to competing views.10

Shakespeare the Reviser The evidence about whether or not Shakespeare revised is limited, equivocal and in any decisive form confined to revision during the process of composition. Heminges and Condell, in their preface to the First Folio, reported that ‘wee haue scarse receiued from him a blot in his papers’ (sig. A3). That is, even in the process of initial composition he was not a reviser. Ben Jonson reported similarly: ‘I remember the players have often mentioned it as an honour to Shakespeare, that in his writing, whatsoever he penned, he never blotted out line.’ In his prefatory poem in the First Folio, however, Jonson gave a different view: ‘art doth give the fashion’ to inspiration, and Shakespeare’s lines are ‘well-turnèd and true-filèd’.11 If Hand D of the manuscript of additions made to the play Sir Thomas More by Anthony Munday and Henry Chettle is Shakespeare’s, this is the one fragment of dramatic text that we have in his handwriting, and, contrary to the reports of Heminges and Condell and the players, it shows revision in the process of composition. Here at least, as he worked on his additions, Shakespeare revised.12 But all this tells us nothing about whether Shakespeare went back to some of his

160

Life of Texts

own works years after their first performances to rework them substantially, changing their implications with additions, cuts and other significant reshapings. In line with Jonson’s praise and Sir Thomas More’s additions, a few play texts contain evidence of what may be revisions made during the process of composition. Love’s Labour’s Lost contains two versions of the same speech which are clearly alternatives (4.3.286ff., Q and F). Romeo and Juliet contains a speech in which elements of an earlier form are retained alongside revisions (5.3.101ff., Q2 and F). In A Midsummer Night’s Dream mislineation (5.1.5–83, variously; Q1) may suggest that the compositor was confused by material added in the margin of the manuscript from which he was setting.13 The two reports of Portia’s death in Julius Caesar (4.3) have sometimes been thought of as evidence of revision during the process of composition, though it is also possible to understand them as serving different dramatic functions. The list of plays, apart from King Lear which it is now argued by some scholars Shakespeare revised between earlier and later printings, is under extension.14 Regularly included are Richard III, Hamlet, Troilus and Cressida and Othello. With Richard III the F text contains about 200 lines not in Q1, while Q1 contains about 35 lines not in F.15 With Hamlet, Q2 contains about 230 lines not in F, while F contains about 70 lines not in Q2.16 Troilus and Cressida contains a few lines in Q not in F, about 50 lines in F not in Q, and many substantive (verbal) variants. The Folio text of Othello contains about 160 lines not in Q, including Desdemona’s protestation of her innocence (4.2), her Willow Song (4.3) and Emilia’s speech on marital infidelity (4.3); the Quarto has a few unique lines of its own. With Troilus and Othello, those who see the differences as evidence of revision are not agreed about whether the Q or F texts represent the revised version – and it is sometimes admitted that it is impossible to know whether differences arise from revisions made by Shakespeare or changes made by the company.17 The first quarto of 2 Henry VI (called The First Part of the Contention between York and Lancaster, 1594, about one-third shorter than the F version), and the first published text of 3 Henry VI (The True Tragedy of Richard Duke of York, 1595, an octavo about 800 lines shorter than the F version) were early taken into the reviser theory, but the memorial reconstruction (‘bad quarto’) view of both these early texts has been convincingly defended by recent editors.18 Other plays now taken into the Shakespeare-as-reviser theory by some scholars also include Titus Andronicus (3.2, F only, seen as added later) and Henry V (the shorter Q1 of 1600 seen as an abridgement in which Shakespeare may have had a hand). The fundamental result of all this is that over the past forty years, there has been a gradual but ultimately radical shift destabilizing the texts of Shakespeare.

Rescuing Shakespeare

161

The idea that it is possible, with whatever doubts and adjustments, to envisage an ideal form of each play and therefore to edit the individual work – Hamlet, King Lear – has been given up in favour of the practice of editing only each individual version of the text – Hamlet Q2 or F, King Lear Q1 or F. The grounds of this fundamental change are either the confident but ultimately unprovable postulate that Shakespeare is responsible for the differences, or the less dogmatic but in practice equally disintegrative view that it is impossible to postulate with confidence what the relations between variant texts might be, and that any attempt to do so should therefore be abandoned. A keynote of the editing practices thus evolved is irresolvable multiplicity – highly congenial to the interpretive assumptions off which and into which they feed. With no single work does this shift have greater consequences than with King Lear.

King Lear The fundamental textual situation with King Lear is that Q1, printed in 1608 – and (as all agree) badly printed, by a printer who had never previously printed a play and had inadequate fonts of type for so large a text – contains about 300 lines that do not appear in the Folio. The Folio, well printed in 1623, contains about 100 lines that do not appear in Q1. There are no certain explanations of these differences. Every explanation is a postulate based on bibliographical knowledge and literary judgement – that is, a postulate based on interpretations of multiple differences many of which admit of several explanations. Where many differences admit of several more or less plausible different explanations, it is unlikely that a single master explanation accounts for most. The view that Shakespeare revised the play is one postulate used to explain the differences: there is no evidence for it beyond those differences, which can often be explained on quite other grounds. The text underlying the printed copy for Q1 is not agreed.19 On the contrary, it is one of the most vexed questions in the history of Shakespearean bibliography. It may have been Shakespeare’s manuscript, so authoritative, but difficult to read. It may, however, have been a manuscript with much less authority: it has been thought at various times to be a manuscript derived from dictation (because the text contains errors that most probably result from mishearing); and a manuscript derived from memorial reconstruction by a group of actors – probably a group rather than a single actor, because it is unusually accurate for a single memory (actors of the roles of Goneril and Regan have been suspected – and exonerated).

162

Life of Texts

A second quarto (falsely dated 1608), published in 1619 as one of a group of ten so-called ‘Jaggard-Pavier quartos’ (Jaggard the printer, Pavier the bookseller), was printed from Q1, but also made corrections (and introduced errors), probably from a playhouse source. The text underlying the copy for F is also debated: it has been thought to be a copy of Q1, corrected perhaps by reference to Q2 or to a playhouse manuscript copied by a professional scrivener – authoritative, less liable to misreading, but perhaps incorporating scribal errors; the presence of Q2 underlying the Folio copy is also argued to have been more direct. The situation is made more complex by the fact that Q1 (as was usual) was corrected while it went through the press (so there are variants between extant copies – and there would be other variants in copies no longer extant); and Q2, though based on Q1, was corrected from some other source. With both Q1 and F there are so many elements difficult to explain that every postulate about underlying copy carries with it problems and doubts. Nevertheless, whatever their bases, Q1 and F are the two major witnesses, though Q2 also has some claim to independent authority. There are two simple patterns of difference between the Q1 and F texts: Q1 contains material, generally agreed to be excellent in itself, that the F text does not contain; Q1 contains a greater number of plain errors than F – that is (at the simplest) words and phrases that do not make sense. Those two simple differences – significant material in Q1 not in F; errors in Q1 corrected in F – are the bases for the traditional assumption that the proper base text for an edition of King Lear is the more correct F, incorporating the material found only in Q1, and corrected by reference to Q1 in the relatively small number of cases where F may be supposed corrupted by error (mainly errors made by a compositor in reading manuscript or setting type). Or to put it the other way round: there are too many bad lines in Q1 to believe that it represents the text as Shakespeare wrote it; and there are too many good things missing from F to believe that it represents the text as Shakespeare revised it. The gains of the consolidated text of King Lear – F and Q1 together – are that the play does not, as F does, lose a major part of one of its most original and significant scenes, the mock trial of Goneril (3.6), which is vital in tracking Lear’s progress from the conflicts with his two elder daughters in Act 2 to the more extreme mental derangements of Act  4; and the play does not lose other important material, including a confrontation between Goneril and Albany (4.2), which is both a vivid extension of Goneril’s wickedness and an important scene for Albany, the only figure of goodness in the otherwise variously evil antiLear group. On the other hand, the Folio text does not contain the dozens of

Rescuing Shakespeare

163

botched lines in Q1, which it is impossible for anybody but an out-and-out revision theorist to believe that Shakespeare wrote and which can be confidently corrected by reference to the F text. Anyone who reads Q1 and F together impartially, without looking for arguments in favour of one or the other, will find that at various points each has better readings – better by a variety of criteria; as making better sense; as more appropriate to character, or dramatic situation; as having greater expressive force; as better cohering with the dramatic atmosphere or the themes; or as better showing the beauty and energy of metrical structure for which Shakespeare’s work considered in its entirety shows he aimed. Of course, in relation to all these things choices require making literary judgements: they require the disciplines of literary criticism. There is no substitute, in editing King Lear or anything else, for the combined exercise of bibliographical discipline and literary-critical intelligence. And this must be the free exercise of discipline and intelligence, not intelligence led – that is, misled – by the wish to create a stir; by the attention-drawing shock of the new for its own sake. The reviser-theorists’ main claims about fundamental differences between Q and F include that in F the role of Edgar is exalted, the roles of Kent and Albany are correspondingly diminished, that what is in Q presented as a French invasion in F becomes a civil war, and that the whole presentation is thereby made faster moving and more objective. It is not possible here to discuss in more than outline these supposed Shakespearean changes, but even in outline it is important to bear in mind that the ‘text’ is words presented as embodied sequence with expressive action. The nature and importance of the elements of embodied sequence varies from moment to moment, but is always significant. While scholars discuss the detail of the ‘text’ as printed words, that is not a form in which many Elizabethan play-goers ever thought of Shakespeare’s plays; and even with those that existed as books in Shakespeare’s lifetime, only a minority of play-goers saw them in that form. For Shakespeare’s audience a text was not words read but words heard partially interpreted by visual contextualization. The most familiar form of Shakespeare now, as printed text, was the least familiar to Shakespeare’s contemporaries. Treating Shakespeare as a predominantly printed text distorts how his audiences usually experienced a play. Also, and more simply, each case of supposed authorial revision depends on the cumulative effects of many often small differences, which individually admit of a variety of explanations. A good editor looks both for patterns of textual difference (so as to elicit fundamental reasons underlying separate changes), and at each difference independently, not through a lens of a priori assumptions about its

164

Life of Texts

likely cause – especially not as part of some pattern of causes that the editor is hoping to find. The supposed change to the role of Kent depends substantially on the omission of 4.3, a scene of limited action which can be readily recognized as an obvious possibility for a cut if the book-keeper or other member of the playhouse staff were looking to shorten a long text. It is not necessary to attribute so straightforward an omission to the art of the revising author. Supposed effects of this supposed authorial change include making the moral qualities of Cordelia (honesty, fidelity, love) stand out more clearly by their isolation, reducing the parallel between Kent and Edgar (disguised care of a suffering old man), and making Edgar appear unique so as to emphasize his ‘triumph’.20 But in both texts Edgar fails in his main intention, which is to protect Gloucester; and he says he fails (‘O fault’: 5.3.193). The idea of his ‘triumph’ Edgar would not recognize. No more should the audience. Edgar is an over-confident moralist who believes his father suffered justly on account of his adultery (5.3.171–4). ‘Die for adultery? No’ (4.6.111) is not only a radical Learism. An annotator of Edgar’s severe morality may quote (as the Arden 3 editor does) the Old Testament Apocrypha on the simple and direct relation of sin and punishment (Wisdom 11.16): so the discredited ‘Friends’ of Job think. Like Lear, Christ turned aside the Mosaic punishment of death for adultery (John 8). In King Lear the innocent and the guilty suffer alike. Moralists such as Edgar affirm what they believe should be so in a morally coherent universe, not what is so in the Lear world.21 It can always be observed that a cut text ‘moves with greater swiftness’.22 Could it be otherwise? That greater swiftness is invariably a dramatic virtue, however, is open to question. The significance of Edgar’s narrative of his reunion with Kent (5.3.205–21; Q only) is well realized by Michael Warren in his argument for authorial revision: it brings together for a final emblematic moment the disguised carers, the devoted servants, the two men with whom the action began. His affirmation, however, that these are elements the audience has forgotten, or in which it has no interest, is based purely in the imperatives of revision theory. If a commentator can see the point of the scene, why may not an audience feel it? That the action moves more quickly without Edgar’s passionate report is undoubted. That the play is better without it is open to more than one view. That a director and actor who cannot interest an audience in this great tableau are incompetent is certain. The affirmations of reviser theorists about the superiority of what is simply shorter often have the effect of drawing fresh attention to the greatness of Shakespeare’s art in handling his massive canvas.

Rescuing Shakespeare

165

The issues of most consequence with the supposed refashioning of the role of Albany are F omissions from his confrontation with Goneril (4.2), and the reassignment to Edgar of the play’s final lines. Albany’s confrontation with Goneril (4.2.31–50, 53–9, 62–8; Q only) reveals the strength of both characters. What the audience sees here in Albany builds from reservations expressed earlier with the negotiatory tentativeness proper to a loyal husband (1.4.304–44), but is new in its forcefulness. What the audience sees in Goneril – how she handles Albany’s uncompromising probity – is also new. It pushes her to new depths: ‘Marry, your manhood, mew’ – one of the great lines of the play, magnificent in its corrupt derision. Surely even Shakespeare must have had a pang of regret if some disputable design about Albany persuaded him to sacrifice the thrilling dialogue of which this is the climax. These are much more likely to be playhouse than authorial cuts. But in any case, the fundamental development in Albany that Q more amply displays is present in F, albeit less strikingly. In Q Albany does not become a character who ‘seems to espouse no positively defined ethical standards’.23 On the contrary, his ethical standards of justice and compassion, clearly proclaimed, are just too hopeful for the world he inhabits (4.2.78–80, 94–6; in Q and F). The changed conception of the character for which revision theorists argue is a critical fiction. On the final lines (Q, Albany; F, Edgar), in both texts it is Albany, as the only leader of the victorious army left alive, who assumes powers of direction and decision making. Whether his final speech in F (5.3.297–305) is taken to mean that he restores Edgar and Kent to the titles and power they lost in the course of the play, or (more radically) that he invites them to govern jointly with him, either way, it is Albany who assumes the power to make such a decision. A similar forcefulness is evident in his pre- and post-battle confrontations with Edmund (5.3.59–61, 82–95; broadly the same in both texts). That the last words are spoken in F by Edgar in no way compromises Albany’s assumptions about his power: political arrangements are his prerogative. Whoever speaks last, in both texts Albany asserts himself as the central figure politically. But probably – as elsewhere in Q – the speech prefix assigning these lines to Albany was simply wrong:24 Edgar must (as Kent does) make some response to Albany’s proposal, and ‘we that are young’ comes much more naturally from him. With the final battle as presented in Q (‘foreign invasion’) and in F (‘civil war’), there are indeed systematic differences – differences in the presentation of an initially successful French invasion of Britain that might well result from censorship, and have nothing to do with Shakespeare or even with choices made within the company.25 But even here, the differences, though real, are limited.

166

Life of Texts

That ‘the army of France is landed’ is plainly stated in both texts (3.7.2). In both, in one of Shakespeare’s more obvious biblical allusions, the pro-Lear leader, Cordelia, is implicitly compared to Christ (‘Oh dear father, / It is thy business that I go about’: 4.4.23–4; Luke 2.49) – which in the context of Cordelia as a figure of Goodness and Compassion (4.3.17–23), and as a premonition of her suffering, only perversity can view as ironic. And in both texts Cordelia says, affirming the support of France for this view, ‘No blown ambition doth our arms incite, / But love, dear love, and our ag’d father’s right’ (4.4.27–8). What can this mean but to make clear, in Q as in F, that the aim of French arms is not to annex Britain but to restore Lear as the ruler of an independent country? Probably the greatest single difference between the two texts is the omission from F of the mock trial of Goneril (Q, 3.6.17–56).26 In proposing that this great episode was cut by Shakespeare Roger Warren gives a fine exposition of how it functions in itself and how it relates to the play as a whole: that it prefigures Lear’s more general arraignment of justice as socially partial in his later dialogue with Gloucester (4.6); that the naked beggar as ‘learned justicer’ and the jocoserious Fool as ‘sapient sir’ precisely embody this view of the world turned upside down; and that Tom-Edgar’s obsession with devils and the Fool’s anarchic pleasure in weirdness (semi-inscrutable songs and riddles) fragment and destabilize Lear’s meaningful inversion. That is, they do not keep it simple: they add inversions characteristic of their own meaningful alienation. This is excellent criticism.27 To this need only be added that, in its movement towards madness, the scene is an essential element in the development of Lear from the early part of Act 3 where, though passionate and furious, he is aware of his own interior and exterior state, and learns from his experience as an outcast about social injustice (3.4.28–36: his prayer for the poor) in ways endorsed by the parallel experience of Gloucester (4.1.67–71). Even the grotesquely comic transformation of this – his assumption that Tom-Edgar, since he is destitute, must have had ‘unkind daughters’ (3.4.71) – has a kind of sense: Lear is absorbed in the complete realization of his own agony. As for his desire to strip physically in order to make himself vulnerable to experience unprotected by the usual defence of clothing (3.4.108), terrible as this is, it too shows a kind of wisdom: Lear recognizes that the protected life he led formerly fostered ruinous illusions. This Lear has foreseen madness as a possibility (‘My wits begin to turn’: 3.2.67; ‘that way madness lies’: 3.4.21), but, though teetering on the edge of madness, he remains deeply aware of self and circumstances. The character of Act  4 has a wisdom which appears more entirely through radical derangement of normal sensemaking. The mock trial connects the two.

Rescuing Shakespeare

167

Why should Warren claim that it is impossible for an audience to understand what he so well explains? Of course, in performance any audience will be puzzled by detail, but a good actor can convey fundamental significances without conveying every shade of meaning that can be followed by a reader. No doubt here the puzzles come thick and fast, but so they do in Lear’s (wise) ‘mad’ dialogues with Gloucester, and so they do almost whenever the Fool riddles and whenever Edgar speaks in his persona as Poor Tom. King Lear is full of difficult text. That a director working with good actors cannot make clear the broad meanings, including the anarchic thrill of moment-by-moment threatening to tumble these apart while fundamentally holding them together, assumes that readers and audiences come by their knowledge through the same access to textual detail; but they do not. The reader works out in detail from the words alone what the well-directed actor delivers in essence interpreted by tone and action. Warren’s excellent exposition is actable – and so performance has found. It is part of the platform of the revision theorists to imply that their views are more theatrically conscious, that those who oppose the revision theory with King Lear are excessively literary, bookish. In practice, quite the reverse is true. The actor-director Robert Clare has shown that no major modern stage production from Peter Brook (1962) to Adrian Noble (1993) has followed the (supposedly author-revised) Folio text in omitting the mock trial.28 In his 2007 performances with the RSC Trevor Nunn, though aware of scholarly views of Q and F as supposedly different Shakespearean versions, was ‘unwilling to lose rich and evocative material from either version’,29 and included the trial. Nicholas Hytner, who attempted to use a Folio-only text for his 1990 production with the RSC , incorporated what Peter Holland presents as ‘a few small additions from the Quarto’ – including the mock trial.30 Even a director committed to a Folioonly experiment, that is, apparently found the episode impossible to omit. To this can be added that, despite Shakespeare texts being regularly more extensively cut for films than for theatrical performances, no film of King Lear omits the mock trial either.31 Far from being a bookish preference, this major aspect of the consolidated text appears to be a performance imperative. Revision theory proposes Shakespearean changes to some limited outlines of King Lear. The proposal that Q and F are separate versions of the play also produces significant differences in many areas of the text which, on the view that Q and F are corrupt versions of a single original, appear quite differently. These differences extend into minute detail, but the general issues can be viewed through major examples.

168

Life of Texts

Sometimes the Q and F texts present different versions where either Q or F must be preferred on the ground that the other makes no sense, as at Q 1.4.127–47, F 1.4.129–33. In the abbreviated and incoherent F version the Fool’s answer (his egg joke) does not relate to his question (about the difference between a bitter fool and a sweet fool). There is a probable explanation of the cut. The censor cancelled the missing passage because it refers to the sensitive subject of monopolies: James I’s granting of these, contrary to the Elizabethan Act against them (1601), was unpopular. The Q text is correct; the F text makes no sense; and it was changed for a reason there is little likelihood that Shakespeare sanctioned. Conversely, some vital text is found in F only, as at 4.6.165–70: Lear. . . . Plate sin with gold, And the strong lance of justice hurtless breaks: Arm it in rags, a pigmy’s straw does pierce it. None does offend, none, I say none, I’ll able ’em: Take that of me, my friend, who have the power To seal th’ accuser’s lips.

This is a radical extension of Lear’s wisdom-in-madness arraignment of justice as partial and dependent on social status: ‘None does offend, none, I say none’: no individual can properly be regarded as guilty in a context where all are guilty. Because Lear is preoccupied in parts of this scene with adultery, it has been suggested that he alludes here to Christ’s overthrow of Old Testament law in the story of the woman taken in adultery (John 8), whose accusers Christ deflects by in effect accusing them all of being sinners. With or without this radical resonance, in both texts, Q and F, Lear makes that case specifically: ‘the usurer hangs the cozener’ (4.6.163). Only in F does he make it generally. It is obvious why this startlingly subversive claim that the whole process by which social order is maintained conceals the realities of sin and crime might attract reprobation from the censor. Sometimes Q is correct; sometimes F. Sometimes both texts are required, as in 1.4 (Q, 218–23; F, 203–5). Quarto: Lear. Who is it that can tell me who I am? Lear’s shadow? I would learn that; for, by the marks Of sovereignty, knowledge, and reason, I should be false persuaded I had daughters. Fool. Which they will make an obedient father. Lear. Your name, fair gentlewoman?

Rescuing Shakespeare

169

Folio: Lear. Who is it that can tell me who I am? Fool. Lear’s shadow. Lear. Your name, fair gentlewoman?

Q cannot be right by itself: it has Lear both answer his own question (‘Lear’s shadow’), and respond to the answer he has given himself (‘I would learn that’). In F the Fool answers Lear’s question, and answers it in his usual vein, forcing Lear to recognize the consequences of his errors. But F by itself is not satisfactory either, because it has Lear not engage with the Fool’s answer, as he does elsewhere, and turn inconsequentially straight to Goneril. The consolidated version gives the full text from Q (which has several mistaken speech prefixes), but assigns the speeches as in F. A passage traditionally combining Q and F which figures in revision-theorist arguments as producing a text for which it is supposed there is no evidence is part of the dialogue between Lear and Kent in the stocks. Lear. What’s he that hath so much thy place mistook To set thee here? Kent. It is both he and she, Your son and daughter. Lear. No. Kent. Yes. Lear. No, I say. Kent. I say, yea. Q Lear. No, no, they would not. Kent. Yes, they have.Q Lear. By Jupiter, I swear, no. F Kent. By Juno, I swear, ay.F

Arden 3, 2.2.202–12, showing Q only and F only text by superscript letters

To claim that there is no evidence for this combined text is to insist on an unduly narrow notion of evidence. No early text prints the lines in this form, but a pattern of duplication (in Q) which is balked in the very element where the pattern requires greatest emphasis (at the end) is itself evidence of corruption; while in patterned repetitions accidental omission of one statement-response pairing (in F) is easy to suppose. The effective symmetrical crescendo of the combined texts – ‘yes’ and ‘no’ stated, elaborated, more emphatically elaborated and finally enforced with an oath – is so effective as to confirm the corruption of patterning in Q and the probability of accidental omission in F. As Foakes

170

Life of Texts

comments, the full text ‘echoes Lear’s amazed incredulity in his confrontation with Cordelia . . . and his inability to comprehend any kind of opposition to the power he had taken for granted for so long’ (Arden 3, 239–40). That is, the full text is both locally effective and congruent with the developing presentation of Lear. That this is in part a literary judgement need alarm nobody who recognizes that textual criticism is an art as well as a science. What can be seen here in extended passages is equally true of individual words: both texts are needed because each can be corrected from the other. F can be corrected from Q, as at 1.1.287: Goneril. The observation we have made of it hath not been little (Q; F ‘hath been little’ – which is clearly wrong because the sisters go on to discuss several such observations).

Or 1.4.320: Goneril. Y’are much more a-taxed (Q corrected ‘attaskt’; F ‘at task’) for want of wisdom (‘task’ [as a verb] and ‘tax’ both meaning in the period ‘to blame, censure’).

Or 4.6.83: Lear. No they cannot touch me for coining (Q; F ‘crying’). I am the King himself.

Q can also be corrected from F, as at 2.2.153–4: Kent. Nothing almost sees miracles (F; Q uncorrected ‘my rackles’; Q corrected ‘my wracke’).

Or 3.4.97: Lear. Come, unbutton here (F; Q uncorrected ‘come on bee true’; Q corrected ‘come on’).

Or 4.6.104: Lear. They told me I was everything; ’tis a lie, I am not ague-proof (F; Q ‘argueproof ’).

Or (one of Q’s apparent mishearing errors) 4.6.150: Lear. A dog’s obeyed in office (F; Q ‘a dog so bade in office’).32

Another significant effect of combining Q and F relates to a different kind of textual difference. The Fool and Edgar as Poor Tom are parallel characters: each, though highly vulnerable, honourably cares for and suffers with one of the

Rescuing Shakespeare

171

central old men; and Edgar in his Poor Tom persona notably deploys his own version of the Fool’s crypto-riddling manner. Each has one scene-ending soliloquy of general reflection addressed to the audience, which brings out the parallel between them more forcefully than any other single element of the play. But the Fool’s soliloquy is found only in F (3.2.80–95), and Edgar’s soliloquy only in Q (3.6.95–108). Only the consolidated text emphasizes the parallel between these characters in this way – and Edgar’s soliloquy (on revision theory the speech supposedly cut by Shakespeare) also points the parallel between the two plots with unusual directness (‘he [Lear] childed as I fathered’), endorses Lear’s view that mental suffering is greater than physical (3.4.6–15), and allows Edgar to remind the audience of the very different character concealed by his Poor Tom persona. The speech is vital. Certainly arguments of this kind for the inclusion of Q-only material involve aesthetic judgement. But how does a revision theorist distinguish a cut made by a member of the playhouse staff from a Shakespearean cut other than by making an aesthetic judgement? While there are many words and phrases in which Q and F offer largely indifferent alternatives and cannot be combined, there are also places where both versions are required, as at 1.2.100–105 (F only) and 1.2.131–37 (Q only). These two passages are close to one another in the same scene, and both are about astrological predictions of division and disaster. Is this duplication and redundancy? According to revision theorists, it is: Shakespeare cancelled the second passage when he wrote the first. But there is no redundancy. That the two passages are on the same subject is beside the point. The point lies not in the subject but in the attitude to that subject of the speaker – Gloucester’s an attitude of belief that might be seen as credulous; Edmund’s – completely different: a contrast with his father – an attitude of belief assumed in order to present himself to Edgar as naively unsceptical. The subject of the two passages is the same, but their function in relation to character and situation is entirely different. It is a typical example of the need in textual criticism for critical judgement as well as bibliographical information. There are also occasions where F is to be altogether preferred. The most crucial of these is Lear’s death. Lear. And my poor fool is hanged. No, no life. Why should a dog, a horse, a rat have life, And thou no breath at all? O, thou wilt come no more. Never, never, never. Pray you, undo This button. Thank you, sir. O, O, O, O! Edgar. He faints. My lord, my lord.

172

Life of Texts

Lear. Break, heart, I prithee break. [Dies] Edgar. Look up, my lord. Kent. Vex not his ghost. O, let him pass. He hates him That would upon the rack of this tough world Stretch him out longer. Q, 5.3.297–307 Lear. And my poor fool is hanged. No, no, no life. Why should a dog, a horse, a rat have life, And thou no breath at all? Thou’lt come no more. Never, never, never, never, never. Pray you, undo this button. Thank you, sir. Do you see this? Look on her, look, her lips, Look there, look there. [He dies] Edgar. He faints. My lord, my lord. Kent. Break, heart, I prithee break. Edgar. Look up, my lord. Kent. Vex not his ghost. O, let him pass. He hates him That would upon the rack of this tough world Stretch him out longer. F, 5.3.279–91

The simplest difference here lies in the expressivity of F’s five-times repeated ‘never’, the emphasis given by making from the one word a complete line of iambic pentameter to convey Lear’s desolating grasp of death’s finality. Just as significant are the added lines, which mean that, while it is open to interpretation whether Lear dies with a joyous delusion that Cordelia is still alive, or a despairing conviction that she is dead, he dies wholly concentrated on her – completely absorbed in her state, whatever he takes that to be. Q is surely just wrong: after giving Lear an extended death groan (conventionally represented by a repeated ‘O’), it keeps him alive for two more lines, so that after Edgar’s exclamation, ‘He faints’, he speaks again to desire his own death – ‘Break, heart, I prithee break’. In F, ‘Break, heart, I prithee break’ is a wholly proper response from Kent to having witnessed his beloved master’s death. Is it really possible to imagine that Shakespeare wrote this line for Lear and in revision reassigned it to Kent? It seems that to sustain their view revision theorists will persuade themselves of anything. The difference here between the two texts raises the question of how best to produce a consolidated text. Should an editor, as Foakes often does (Arden 3), print everything – give the death groan from Q but then keep Lear alive, adding the extra lines from F? Or should an editor – as most have done – correct Q by

Rescuing Shakespeare

173

F, omitting the groan, so that Lear dies in response to whatever delusion or revelation he sees on the lips of Cordelia? That is the preference of Duthie and Dover Wilson (Cambridge), of Muir (Arden 2), and of Hunter (Penguin); and it is surely correct. Q1 cries out for correction on one final ground – its many hypermetrical lines: inserted extra syllables that are typical actor additions; missing words that are typical scribal or compositorial errors. W. W. Greg rightly comments on ‘the difficulty of believing that Shakespeare . . . could ever have written the clumsy and fumbling lines we find in Q, or that these could . . . represent a stage in the development of F’.33 The constant ugly and pointless disruption to metrical structure is not at all similar to the musical freedom-within-pattern of the poetry of other late plays (Antony and Cleopatra, The Winter’s Tale), and is an important element in the problems of Q1 ignored by revision theorists. With Q1 metrical structure has in any case often to be an editorial construction, because in parts of Q1 verse is printed as prose. But however the verse structures are hypothecated, there is regularly throughout the ugly and jarring over-spill or hiatus created by purposeless irregularity. Sometimes this is a result of what are most obviously actors’ extra-metrical intrusions: ‘Well’ (1.1.98), ‘Come sir’ (1.4.207; a pre-echo of ll. 223 and 300), ‘Why’ (1.4.213; 3.2.18), ‘or’ (2.2.97), ‘Nay’ (5.3.151); or other kinds of simple actor confusion: as when F’s imperious [Lear] ‘How, how, Cordelia? Mend your speech a little’ becomes Q’s fretful ‘Go to, go to, mend your speech a little’ (where ‘Go to, go to’ is also repeated unmetrically later in the scene [Q, 1.4.222]); or as when ‘Go, go, my people’ (correct at Q 1.4.259) is repeated (in place of F ‘Away, away’) at Q 1.1.276. Sometimes the effect is so unShakespearean that one can only posit actor, scribe or compositor error: as when F’s incisive [Regan] ‘I am made of that self mettle as my sister’ (1.1.68) becomes the hobbling flab of Q’s ‘Sir, I am made / Of the self-same mettle that my sister is’ (1.1.61-2); or, with similar extra-metrical padding, [Lear] ‘Thou shalt find / That I’ll resume the shape which thou dost think / I have cast off for ever. Q Thou shalt, I warrant thee.Q / Goneril. Do you mark that, Qmy lordQ?’ – where, as elsewhere, Foakes (Arden 3) destroys the rhythm of F by including Q intrusions. With extra syllables and with missing syllables, in Q the structure of verse is constantly impaired: ‘Stands still in experience [F, esperance], lives not in fear’ (4.1.4); ‘The good [F, good years] shall devour them, flesh and fell’ (5.3.24). Lines memorable for beauty and power of structural form are corrupted and enfeebled. The effect of this constant collapse of the intensifying structure of verse rhythm cannot be fully appreciated in individual examples: it is necessarily accretive. Anybody to whom this is not a matter of the first importance should

174

Life of Texts

not be editing Shakespeare. As a practising verse dramatist, T. S. Eliot writes about poetry in drama as providing ‘a musical pattern’ that strengthens our ‘excitement . . . with feeling from a deeper and less articulate level’; that ‘to work out a play in verse is to be working like a musician . . . it is to see the whole thing as a musical pattern. . . . Underneath the action . . . there should be a musical pattern which intensifies our excitement by reinforcing it with feeling from a deeper and less articulate level’.34 The whole discussion makes it clear that Eliot’s concern is twofold: with overall form and with versification – the movement of the verse line by line. The essay ends by re-emphasizing the importance of this – ‘the musical pattern, as well as the dramatic pattern’ (Eliot’s emphasis). This is an issue relevant not only to editing but also to how the text is realized in performance. It is characteristic of actors trained in naturalistic styles, for film and television, and with democratic intuitions that suspect anything heightened by stylization of an appeal to ‘elites’, that they will not perform Shakespeare’s poetry in such a way as to retain – through whatever colloquial counterpointing – the aural structures of verse. ‘Turn what he will to verse, his toil is vain: / Actors like this will make it prose again’. Editing without attention to the rhythmic structures of verse supports and encourages this diminution. But, as the great French poet Paul Valéry contends, verse is a form of speech tending towards song.35 The actor performing Shakespeare must find in his or her voice the music that can discover, work with and project the music of the lines. Of course that music has many sources, but fundamental to it is rhythm: no rhythm, no poetry. And the Q text of King Lear, with its hypermetrical hobbling, constantly untunes verse towards prose. As T. S. Eliot makes clear, this is not a matter only of local effects: it is a matter of the emotional temperature throughout and the cumulative effect of the whole. Metrical considerations need to be reinstated among the criteria of editorial choice – though, as with everything else, there can be no simple formula: aesthetic judgement is necessary. There are many significant differences between the Quarto and Folio versions of King Lear – above all the omission from Q of the mock trial of Goneril, the cuts to the Act 4 confrontation between Albany and Goneril, and the omission of 4.3; but there are in the Folio no new elements, no changes of plot or character that affect the fundamental substance of the play, the presentation of the parallel Lear and Gloucester plots. The Folio text quite possibly shows how the play was reshaped for some performances by Shakespeare’s company, perhaps with Shakespeare’s cooperation, perhaps entirely contrary to his wishes. Cuts made by playhouse staff, censorship, mistakes in transcription, compositorial and other printing house error: the possible and probable sources of difference between

Rescuing Shakespeare

175

the Q and F texts are multiple. There is no evidence that they are authorial. The consolidated text, in which generations of editors and critics have found unity and integrity, should not be kept from readers and theatre-goers by an editorial fashion that presents tendentious opinion as fact. If one could survey all the evidence from all viable perspectives, one could not expect to convert revision theorists to this view or change the course of the practice of editing. Criticism and scholarship have their dynamics suited to the times: these currently favour multiplicity and disintegration, which accordingly have their own inertly powerful persuasive force. From this consensus there has been copious well-articulated dissent, but the dissenters have not been in charge of major editing projects. Critique in scholarly publications does not have the same force as deciding how a text will appear in print. One might hope, nevertheless, to show readers and theatre practitioners that revision theory asserts its dogmatism on grounds that are more than questionable. Pace that dogmatism, the consolidated text of King Lear, which theatrical practice shows persuades of itself, should continue to be published, read and serve as the basis for performance. Scholarship and criticism will rediscover this when – as they were from Pope and Johnson to Greg and beyond – the cultural conditions for their doing so are again favourable.

8

Textual Evidence and Musical Analysis: Once More on the First Movement of Beethoven’s ‘Tempest’ Sonata, Op. 31, No. 2 Julian Horton

Introduction: What Is Musical Textuality? At first sight, music might seem an odd subject to include in a book about textuality. Music, after all, is sound, not written language; it is normally disseminated through collaborative or solo performance, not via direct engagement with a text; and although some traditions require music to be written down – Western-European art music is, for instance, predominantly literate – this is hardly a cultural universal. In truth, much of the world’s music-making, today and in the past, has had no need of a textual form, being transmitted pedagogically and disseminated to its public entirely by aural and oral means. With the rise of recording technology, the concept of music as a written document as well as an activity or process seems to be decisively in abeyance, if not decline. Given that digital recording has become not only a method for capturing the performance of a musical text, but also a compositional medium in its own right (as is normal, for example, in popular music), we may well ask how long musical literacy centred on a score and a body of theoretical knowledge with which to decode it can survive. If the idea of music as text has gained currency anywhere, then it is in scholarship on Western art music. The question of music’s textuality is, in this context, ostensibly self-evident: any literate musical culture will generate texts susceptible to scholarly scrutiny, which are, in the first instance, the domain of manuscript studies, editorial scholarship and philology.1 Although there is much to be said about the life of musical manuscripts and editions, I want, in this chapter, to pursue the more elusive sense of music’s textuality: not as a document as such, but as a text that stimulates discourse. Specifically, I want to interrogate what has come to be called the ‘text-event dichotomy’ – the double life music 176

Textual Evidence and Musical Analysis

177

lives as ‘text and act’ – via consideration of the discourse that has grown up around one piece, Beethoven’s Tempest Sonata, Op. 31, No. 2.2 To many, the idea that music leads a parallel textual existence might seem counter-intuitive: after all, for the vast majority of listeners, music is a sensuous phenomenon produced by human action, and certainly not something to be read and interpreted in the manner of a poem or novel. Yet, especially for Western ‘classical’ music,3 the text-event dichotomy remains crucial to musicological debate, and no deep understanding of the art form is possible unless we engage with it. For a long time, the idea prevailed that pieces of classical music constituted ‘works’, the essence of which was enshrined textually in scores, which encoded their biographical, stylistic, formal or historical truth-content. Recently, however, predominantly postmodern intellectual currents, responding to broader intellectual trends in the humanities, the burgeoning plurality of musical experience and the dizzying pace of technological change, have sought to dislodge this attitude. Musicologists have dismissed the idea that the score should constitute a governing authority via the accusation of naïve positivism, emphasizing instead the performativity of analysis and interpretation, the subjectivity of listening, or the critical discourse that music provokes as constructive alternatives. One such argument has been proposed by Nicholas Cook, who takes musical scholarship to task for its general failure to disentangle performance from the prioritization of the score. Taking aim primarily, but by no means exclusively, at the Austrian theorist Heinrich Schenker and his post-war English-language interlocutors, Cook diagnoses what he calls ‘Plato’s curse’: the lingering tendency to assume that performance in the Western classical tradition is essentially a matter of realizing structures enshrined in notation. Cook seeks in contrast to advance the ‘performative turn’ in musical research, by divesting score-based analysis of its authority and shifting attention towards the analysis of performance.4 More radical commentators have challenged the very notion of a musical object. Christopher Small for example argues that music has no objective existence at all; as he bluntly asserts: ‘there is no such thing as music’.5 We imagine that there is thanks to the habit of reifying human action, which creates a musical objectivity that is potent but illusory. Like Cook, Small emphasizes the centrality of performance, resolving the problem of music’s ontology by reimagining it as a verb (‘to music’), and thus as a gerund: we engage in ‘musicking’, which is (usually communal) ritualized action in time. Musical meaning, in these terms, is relational, arising from the set of interactions encompassed by any musical event, and has nothing to do with the content of a score, which is simply one means of facilitating performance: ‘performance does not exist in order to present musical

178

Life of Texts

works’, Small writes; instead, ‘musical works exist in order to give performers something to perform’.6 This mentality is at base sociological, or perhaps ethnographic: music’s condition as human social interaction circumscribes its meaning. In Western classical music, this perspective exposes a troubling ideology, as its relations, enshrined in their most concentrated form in the symphony concert, are revealed as a mechanism for enforcing political power: As a style of musicking that is and has always been cultivated by the holders of power, first in Europe and later in its colonies and outposts, the act of composing, performing, and listening to [classical music] have reflected and shaped the perceptions of those who have held power.7

In this way, the idea of a musical object is more than an aesthetic fiction: it is a foil for the perpetuation of social inequality. Historical musicology has rethought music’s textuality in less aggressive ways, re-orientating scholarship around the discourse comprising music’s historical trace, rather than the analysis of works themselves. The assumption underlying this stance is that ‘the music itself ’ is too evasive a concept to be useful. This ambiguity has been acknowledged at least since Roman Ingarden’s The Work of Music and the Problem of Its Identity, which pointed out that no single version of ‘the work’ has ontological stability. To locate the score as the work’s essence is to neglect performance; but since no two performances of a piece are identical, it has no performative security either.8 For Lydia Goehr, the notion of a musical work only comes about at all as a result of specific historical circumstances, which coalesced at the end of the eighteenth century.9 The idea of music as an autonomous object, which has an existence independent of performance, social function or ideological context, is therefore historically contingent. Faced with such objections, musical scholarship retreats behind discourse: we cannot analyse ‘music’ as a determinate thing, but we can examine discourse about music as it accumulates historically. In a turn that suggests Foucauldian ancestry, music history becomes the history not of musical texts, but of texts about music.10 A subtle variant of this argument has been offered by Roger Parker, whose reconstruction of the discourse about Germanic instrumental music and FrancoItalian opera in 1830s London exposes an epistemic shift in critical attitudes, from an ‘event-based’ economy focused on performance, to a ‘score-based’ economy, which measured performances against abstracted ‘works’ concretized by notation.11 Parker notes that attentive listening reinforced by knowledge of a full score emerged at this time as part of the reception of Beethoven’s instrumental music, contrasting the listening habits accruing to Italian opera or the kinds of

Textual Evidence and Musical Analysis

179

miscellany common to concert programmes of the late eighteenth century.12 The modern notion of a musical text, graspable in its entirety and susceptible to structural analysis, grows from this new culture, and is therefore historically circumscribed. By this view, we think music is an object not because its objectivity is ontologically true, but because a culture of constructing it that way came into being as part of the formation of our modern cultural economy. At present, no version of this argument prevails; rather, we may be justified in writing of an ongoing crisis of textuality in musical scholarship, created by the fact that the musical text has become at once ideologically compromised and pragmatically indispensable.13 Even the most staunchly ethnographic or culturalcritical approaches to Western art music periodically fall back on the score, if only as an absent presence around which discourse or human action circulates.14 Neither does research focusing on reception dispose of the work; after all, the discourse about music must refer in some way to an initial stimulus, whether that is a score or a performance. All of this, moreover, co-exists with analyticaltheoretical research, which maintains the score’s authority as a source of evidence.

Textuality, Autonomy and Beethoven’s Tempest Sonata This essay’s central claim is that a text-based concept of the musical object should remain central to the study of post-Enlightenment classical music, for two principal reasons: first, because the idea that music is thought before it is action is epistemologically crucial to it, and the score is the primary medium of this aesthetic; and second, because, pace Parker, the score was essential to music’s production long before it was complicit with music’s dissemination, which means that a relationship between musical thought and musical text obtains, even if the listening public sees no virtue in accessing it. Few pieces in the classical canon instantiate these points more clearly than Beethoven’s Tempest Sonata. This piece belongs to an august family of works, which have served as touchstones for debates, encompassing analytical, historical, cultural and aesthetic concerns.15 Composed in 1802, the Tempest is invariably regarded as one of the seminal works of Beethoven’s middle period. Its credentials in this regard are normally underwritten by the composer’s reported conversation with the violinist Wenzel Krumpholz, dated to 1801, in which he vowed to ‘strike out on a new path’ (‘einen neuen Weg einschlagen’).16 Whatever the provenance of this comment (and it survives third-hand), the perception has taken hold that a new compositional mentality germinates in this music, essential to which is its status as

180

Life of Texts

a text enshrining novel compositional concepts. At its most ambitious, however, this discourse moves well beyond the analysis of Beethoven’s style, enfolding historical, aesthetic and sometimes cultural-political claims on the largest scale. The first movement’s reputation as formally problematic can be tracked to Adolph Bernhard Marx, who in 1863 detected a ‘puzzle’ (ein Rätzel) for the performer, which is whether we should interpret the movement’s opening twenty-one bars as an introduction or a main theme.17 Marx thought the former; early twentieth-century commentators both supported and disputed his view.18 The problem is perhaps best explained by comparing this passage, quoted in Example 8.1, with an earlier Beethovenian example, the theme that opens the Sonata Op. 10 No. 1, composed in 1798 and given in Example 8.2.

Example 8.1 Opening to Beethoven’s Sonata, Op. 10, No. 1.

Textual Evidence and Musical Analysis

181

Example 8.2 Opening to Beethoven’s Tempest Sonata.

Opus 10 follows a pattern common to thousands of instrumental themes from the second half of the eighteenth century, which music theory calls a sentence.19 Bars 1–4 state core material, called the basic idea, which is then repeated in varied form in bars 5–8, producing the succession statement and response. In this case, as in many others, the variation is harmonic: the statement

182

Life of Texts

emphasizes tonic harmony; the response favours vii°7, which is a common substitute for dominant harmony. Subsequently, the theme has two principal constituents: bars 9–16 comprise a continuation phrase, which extends the music of bars 1–8 without closure; bars 17–221 complete the theme by supplying a perfect authentic cadence. There is one further element, bars 22–31, which reinforces the tonic harmony confirmed by the cadence, and which we call a codetta. As William Caplin explains it, themes organized in this way exhibit four kinds of labour: bars 1–8 initiate the theme, and are presentational; bars 9–16 continue the theme, and are medial; bars 17–221 are cadential and effect closure; and bars 22–31 come ‘after the end’, and are a post-cadential framing function.20 This music conveys no functional ambiguity: it has one collective purpose, which is to serve as the movement’s main theme. By the time we reach bar 31, that task has been completed, and what happens next is functionally different; specifically, bar 32 initiates a transition, which has the job of modulating in order to prepare the arrival of a subordinate theme in a new key. A division of labour like this is basic to the Viennese classical style. Everything you hear has a distinct formal responsibility, which is often unambiguous; and when it has been performed, the music moves to the next task. Music theorists consequently talk about the music’s formal function – the task it executes within the form. The first movement of Op. 10, No. 1 is cast in sonata form, and bars 1–31 function as its main theme. From the outset, Op. 31, No. 2 challenges this musical order. Bars 1–2 seem to possess a clear motivic identity (the ascending arpeggio), but note their slow tempo (Largo) and dominant harmony (this is a first inversion chord of A major, not the tonic D minor), which gives them the character of an extended upbeat. Bars 24–5 respond to this with faster music (Allegro), and also insinuate D minor; yet by bar 6 this momentum has dissipated, and we find ourselves back in a slow tempo (Adagio), and over the dominant again (this time via a half cadence). This is clearly not a slow introduction, in the sense of a slow section preceding the start of a sonata form; yet neither does it possess the kind of stability we hear in Op. 10. Rather, the music seems to vacillate between introductory and presentational functions, between stating a theme and preparing the ground for it. This ambiguity is compounded in bars 7–21. Bars 7–8 are directly analogous to bars 1–2, and bar 84 initiates a return of the material of bar 24. If anything, however, this music is less stable than bars 1–6. The Largo returns over another non-tonic harmony (V of III ) before the tonic has had a chance to settle, and the retrieved Allegro fights to bring us back to D minor via an ascending sequence. The climax at bar 13 constitutes the point at which the

Textual Evidence and Musical Analysis

183

dominant of D minor is recovered in the bass, and the passage ends by resolving cadentially to a root position D minor on the downbeat of bar 21, after which a new section begins by recalling the opening figure, now Allegro and expressing a tonic triad. The tension between introduction and presentation in this music could easily be explained if bar 21 initiated a thematic statement, analogous to bar 1 of Op. 10 No. 1: we would then simply describe bars 1–21 as an introduction. But bar 21 initiates a passage which, from bar 31, moves away from D minor, at which point we realize that a transition towards a subordinate theme is in progress. Despite their instability and introductory character, bars 1–211 have retrospectively to function as a main theme. The German-language discourse on Marx’s ‘puzzle’ culminates in the work of Carl Dahlhaus. For Dahlhaus, bars 1–21 exemplify in nuce an essential characteristic of Beethoven’s new path, which is the emergence of a post-Classical dialectical musical consciousness. As he writes: The ‘theme’ is both an improvisatory introduction and a transitional pattern; instead of being presented in a standard exposition, it dissolves into an ante quem and a post quem: measure 1 is ‘not yet’ and measure 21 is ‘no longer’ the actual exposition, which in Op. 31 No.  2 does not exist. Nowhere does the thematic material take on a basic form; instead, it manifests itself in changing guises according to its location in the formal process, like variations without an explicit theme.21

Whereas the late eighteenth century took for granted music’s sense of formal function – the way the material is organized aligns explicitly with the function it performs – Beethoven makes his material express a dialectical ambiguity. For Dahlhaus, this is more than a revolution in compositional style: it defined what Dahlhaus called the ‘strong’ form of the musical art. By collapsing the theme’s formal function into a dialectic, Beethoven privileged an idea over an inherited convention. The Tempest’s material does not simply inhabit the models bequeathed by Haydn and Mozart, but converts them into a tradition in relation to which a new kind of musical thought unfolds, which emphasizes a concept (the changing function of the opening motive) over the conventions that define a genre (in this case, the syntax of a sonata form). In other words, in Op. 31, No. 2, Beethoven reconceives music as an ‘artwork of ideas’: what is important is not the material itself (which is after all no more than an arpeggiated triad), but the thought that it serves.22 Widening his field of vision, Dahlhaus situated this practice as one half of an historical dialectic, the other half of which was supplied by Franco-Italian opera.

184

Life of Texts

This is the so-called ‘twin-styles’ argument: the early nineteenth century is the ‘age of Beethoven and Rossini’, defined by the dualism of Beethovenian idealism and Rossinian theatricality. The notion of form as text-based idea is central to this distinction. For Dahlhaus, Rossini did not compose melody as the means to an end, as Beethoven did, but as an end in itself. Sustaining an older tradition, Rossini’s melodies are vehicles for performance, and often for specific performers. In other words, he valued melody for its inherent qualities, not its formal potential. In the Tempest, however, the performer’s task is to convey the formal idea. The score is consequently not ‘a mere recipe’, which serves the performer; rather, the performer becomes the servant of the music, which constitutes an ‘inviolable’ text enshrining the composer’s intentions.23 In this view, the Tempest’s cryptic opening arpeggio instantiates a seismic cultural shift, which elevates the integrity of the literate musical artwork above the activity required to disseminate it. Dahlhaus’ most important Anglophone interlocutor is Janet Schmalfeldt, seminally in an article for the Beethoven Forum published in 1995, and more expansively in the 2011 book In the Process of Becoming, which extends Dahlhaus’ dialectical notion of form as a paradigm for understanding music by Beethoven, Schubert, Mendelssohn, Chopin and Schumann.24 Schmalfeldt divests Dahlhaus’ argument of its barely concealed chauvinism (Beethoven’s ‘strong’ Germanic form of the art contrasts Rossini’s antonymically ‘weak’ Mediterranean alternative), rethinks aspects of his analysis, and fills in some of its historical lacunae. She explains the Tempest’s idealism as one expression of the emerging concept of autonomy. Liberated from the constraints of aristocratic, monarchic or clerical authority to which Western music is indentured for much of its early modern history, Beethoven is free to compose autonomous artworks, which reflect on music’s condition as music. Crucially, this does not signal the art form’s retreat into itself, but rather enables a kind of social emancipation. Beethoven’s ‘music about music’ vaults the obstacles placed in its way by political and clerical authority and embraces utopian, world-historical aspirations, expressed most famously in the heroic, per aspera ad astra narratives of the Third, Fifth and Ninth symphonies, but nascent in the Tempest’s germinal dialectic. At the same time, Schmalfeldt also stresses the Tempest’s experiential dimension. The fact that we are forced to rethink the music’s function as we hear it – listening to the transition, we have simultaneously to reimagine the introduction as a main theme – means that we are actively engaged in constructing the form as part of its experience.25 Unlike eighteenth-century sonata forms, which cultivate the impression that what you hear confirms what convention dictates, Beethoven’s new style construes form as a process, creating

Textual Evidence and Musical Analysis

185

the illusion that it is brought into being by the act of listening. Returning to the oft-cited analogy between Beethoven and his exact contemporary Hegel, Schmalfeldt explains this distinction in terms derived from Hegel’s Encyclopaedia Logic: Beethoven rethinks form as the experience of its ‘coming into being’, that is, its ‘becoming’ (werden), in analogy with the condition Hegel posits as the synthesis of ‘Being’ and its antithesis ‘Nothing’.26 This association leads Schmalfeldt to locate Dahlhaus’ argument as a late contribution to the ‘Beethoven-Hegelian tradition’: the lineage of dialectical musical thought emerging with Beethoven and Hegel and extending through A. B. Marx and Theodor Adorno. As she explains: Much may be gained from approaching the works of Adorno and Dahlhaus as manifestations of a long-standing tradition in which Beethoven’s music serves as an agent within the kind of historical process that Hegel’s philosophy predicates. Central to this distinct mode of thought within the larger domain of Beethoven reception is the development in Germany of concepts about form that are chiefly inspired by Beethoven’s music, but also imbued with the spirit of Romanticism and influenced by an idealist epistemology that found its last system builder in Hegel.27

This tradition pivots around Beethoven in various ways. Marx’s theory of musical form drew predominantly on Beethoven’s music, whilst proceeding dialectically in Hegel’s manner. For Marx, formal conventions (Kunstformen) had to be distinguished from music’s inner form (Form), which describes musical content’s generative proclivities. In effect, composers generate Kunstformen from Form, by overcoming the incipient tension between the dialectically opposed forces of rest (Ruhe) and motion (Bewegung), the former embodied in stable thematic presentations (Sätze), the latter in music of transitional character (Gänge).28 Marx offered both a theory of form and a compositional method, which develops Beethoven’s example through Hegel’s philosophy: the Tempest’s generative dialectic blossoms into a kind of Hegelian musical pedagogy. Adorno also stressed the analogy between Beethoven’s music and Hegel’s philosophy, but raised the cultural stakes further, by installing Beethoven’s new path as the fulcrum upon which Western cultural history pivots.29 The Tempest’s collaboration of autonomy and idealism signifies the high watermark of bourgeois culture: the middle style that ensued encodes the possibility of uniting the radical artistic freedom promised by music’s autonomy with a higher ideal of social community, a synthesis Beethoven was later to critique in the fractured music of his late style.30 Where Marx is pedagogical, Adorno is critical/philosophical; and where

186

Life of Texts

Marx emphasizes dialectical synthesis, Adorno is pessimistic. Both however foreground the epochal shift that Op. 31 facilitates. Elaborating on Goehr’s historicization of the work concept, Schmalfeldt furthermore connects the Tempest’s processual character to the emerging notion of ‘absolute music’: the idea, essential to an understanding of musical Romanticism, that, as a vehicle for higher, ‘wordless’ meaning, instrumental music transcended music that sets a sacred, poetic or dramatic text. This idea underwrites both the possibility of autonomy and the perception of musical ‘works’ in the strong sense.31 Beethoven is able to make his Fifth Symphony (for example) convey utopian aspirations in the absence of a poetic text or libretto because of this idea: the genre’s independence from text puts it in range of ideals, the essence of which stands above linguistic expression. With this new status comes a new consciousness of form as an independent category. Eighteenth-century theorists tended to understand form in rhetorical terms, via analogies with oration or the poetics of speech, or else they emphasized the assemblage of small-scale harmonic formulae or ‘schemata’, influentially in the Italian partimento tradition, in which composers built up pieces from the elaboration and concatenation of melodybass patterns.32 The idea that formal types define pieces of music, however, only proliferates by the mid-nineteenth century, influentially in the treatises of Marx and Carl Czerny, which explicitly name classical forms and teach composition in these terms.33 To put this another way: the nineteenth century possessed for the first time a formal self-consciousness, of which the Tempest can be understood as a seminal expression. Because Beethoven composes against the background of an emerging consciousness of form as a concept, the first movement of Op. 31, No. 2 is not only a contribution to the sonata genre, but a reflection on the idea of a sonata form that the genre requires. It is, in brief, a sonata about the idea of a sonata. The notion of becoming reflects this new level of formal self-reflection; Beethoven’s middle style can be understood as an exploration of its formal and expressive potential. Finally, Schmalfeldt also relates these developments to economic circumstances. The rise of the autonomous work is the abstract correlative of music’s economic annexation of the public sphere, as monarchic, aristocratic or clerical control ceded to the emerging musical free market and the aesthetic aspirations of the bourgeoisie. This change also placed a new value on originality, both as an aesthetic imperative (composers had to be novel against the background of tradition) and as a legal mandate, codified in modified laws, which around the turn of the nineteenth century passed copyright ownership from publishers to

Textual Evidence and Musical Analysis

187

composers.34 The Tempest’s germinal dialectic both responds to and typifies a Romantic ‘imperative of originality’, which locates the work in its mercantile as well as aesthetic context.35 Schmalfeldt’s analysis has stimulated multiple responses, notably in essays collected by Pieter Bergé, William Caplin and Jeroen D’hoe in Beethoven’s Tempest Sonata: Perspectives of Analysis and Performance, and in a special issue of Music Theory Online dedicated to her work.36 Caplin’s and James Hepokoski’s contributions in both publications keenly exemplify the work’s continuing capacity to stimulate debate.37 Caplin contests both the claim that bars 1–21 are introductory, and that the music from bar 21 has thematic qualities. He argues that the tendency to hear bars 1–21 as introductory results from ‘a general misunderstanding of the nature of a main theme in the classical repertory’.38 For Caplin, we can hear this passage as thematic on several grounds. Primarily, its tempo is predominantly fast, whereas fully-fledged classical introductions are always slow. In addition, its syntax, although in some respects unorthodox, has many parallels with organizational principles common to late-eighteenthcentury main themes. And crucially, it expresses two properties common to many classical main themes: it is closed with a decisive perfect authentic cadence in the tonic key; and its texture is relatively unstable compared to the transition, which, although it changes key, has a textural stability redolent of hundreds of Classical precedents. In all, Caplin encourages us to distinguish between the syntactical aspects of Classical main themes, which exhibit stability and are ‘tight-knit’, and the ‘statistical parameters’ (texture, rhythm, dynamics), which are invariably unstable and prone to contrast, disjunction and rhythmic variety, in which respect they contrast transitions, which are syntactically ‘loose’ but statistically stable. In these terms, the Tempest’s opening is not classically aberrant. Hepokoski also stresses the classically normative features of the Sonata’s opening, but highlights different associations. For him, the two Largo interjections are introductory gestures incorporated within an overall main-theme presentation: as Example 8.3 clarifies, the Allegro sections comprise the theme as such, to which the slow arpeggiations add prefixes, which resemble expanded upbeats.39 Hepokoski points out that similar gestures appear regularly in the Classical repertoire; an example is found in the first movement of Haydn’s Symphony No. 44 of 1772, quoted in Example 8.4. This movement begins with a call to order, initiated tutti in bare octaves. The first material expressing thematic syntax however appears in bar five; the music from here to bar 12 is organized as

188

Life of Texts

Example 8.3 Hepokoski’s analysis of Beethoven’s Tempest Sonata.

a sentence, as the Example explains, which culminates on a converging half cadence. Anticipating the Tempest, bar 13 then immediately returns to the opening, in the process requiring us to rethink the call to order as part of the theme rather than as prefatory to it.40 If Caplin and Hepokoski are, in their own ways, right, then the consequences for the dialectical reading are profound. Dahlhaus and Schmalfeldt offer futureorientated analyses, which stress the Tempest’s novelty in relation to Classical precedent. Caplin and Hepokoski, in contrast, angle the Sonata towards the eighteenth century. The contextual framework that comes with the Hegelian reading is, if not undone, then held at arm’s length by such orientation, since the claim that Op. 31, No.  2 evidences the new autonomy of the work and its attendant idealism depends very much on the post-Classical character of the first twenty-one bars. But if this music is conventional after all, then the Tempest returns to its eighteenth-century comfort zone, or at least, such novelty as it expresses does not reside in its thematic organization.

Textual Evidence and Musical Analysis

189

Example 8.4 The first movement of Haydn’s Symphony No. 44 of 1772.

Conclusions: In Defence of Musical Textuality Disagreements of this kind lead us back to the debates about textuality with which I began. None of the readings explored here radically misrepresents the textual evidence; and yet shifts of perspective within it have consequences not only for our understanding of the piece, but for constructions of its historical and aesthetic context. Deconstructively minded commentators may see such promiscuity as detrimental to music’s textuality: yet I would argue that plurality is not the same as instability. The Tempest’s capacity to sustain analytical debate arises from the text’s complex relationship with theory and analysis, which seek to describe the building blocks of musical practice and explain their usage in individual works. Understood this way, the Tempest is not problematic as a work, but as an exemplar of theory: we struggle to agree about its organization because the evidence it presents imperfectly fits our standardized models of how Classical instrumental forms should behave. In brief, the Tempest does not problematize the musical

190

Life of Texts

text, so much as the discourse about it. The text nevertheless remains vital to that discourse and by extension to our discourse about Beethoven and music written in the traditions he has come to represent. If this discourse demonstrates anything, it is therefore the folly of jettisoning the musical text in favour of an event-centred attitude towards this repertoire. If Schmalfeldt and Dahlhaus are right, then Op. 31, No. 2 is about the history of ideas at least as much as it is about musical performance. The first movement’s ‘processual’ character and the manipulation of listening that this entails reflect an idea, not a prescription. In other words, music ‘of Beethoven’s type’ is thought before it is text or event; and it is the score, not the performance, which gives that thought its concrete form and allows us to situate it in its philosophical context, which turns out to be the whole history of dialectical philosophy from Hegel to Adorno. In this context, the prioritization of text over act is not a fallacy to be overturned, but the manifestation of a compositional mentality embedded in a major strand of central-European thought. In these terms, the replacement of the musical text wholesale with performance amounts to a kind of historically indifferent essentialism, which fetishizes performance at the expense of compositional technique, music-theoretical discourse, the vast body of textual evidence that literate musical cultures accumulate, and the score-based aesthetic to which many composers subscribed. In the Beethovenian tradition, texts are not enablers of performance, like midwives assisting at a birth, but are integral to the act of composition, essential to its aesthetic and technical understanding, and, to the extent that they are idealized notations of a conceptual essence, anterior to performance, which is the means by which that essence is given sonic representation. However we interpret the Tempest Sonata, the irreducible complicity of what we hear with the literacy that infuses its production and reception is impossible to ignore without doing violence to the artefact and closing off access to its aesthetic and cultural context. That this culture may be politically controversial in the present, as Small insists, is beside the point. That controversy is only intelligible at all in relation to Western musical literacy, and will remain invisible if we disregard the music’s inherent textuality.

9

Fragments Shored against Ruin: Reassembling The Waste Land Jason Harding

‘To attempt to explain to an intelligent person . . . how The Waste Land works,’ remarked Graham Hough, who viewed modernist poetry as an alarming foreign invasion, ‘is to be overcome with embarrassment at having to justify principles so affected, so perverse, so deliberately removed from the ordinary modes of rational communication.’1 The Waste Land demanded new forms of attention and it is easy to forget how it perplexed and vexed many early readers. In 1923, Sir John Squire, editor of London’s leading literary monthly, The London Mercury, complained that what is attempted in The Waste Land ‘is a faithful transcript, after Mr Joyce’s obscurer manner, of the poet’s wandering thoughts in a state of erudite depression. A grunt would serve equally well; what is language but communication, or art but selection and arrangement?’2 If literature is communication, then The Waste Land represents a flamboyant act of presentational impoliteness. After nearly a century of exegesis and explication – the difficulty of the poem packaged into student-friendly critical guidebooks – it is perhaps natural to conclude that the professionalism of an academic specialist is required to explain this poet’s obscurer manner. T. S. Eliot did not think so. In 1925, I. A. Richards, doyen of Cambridge’s fledgling English Faculty, presented a fashionable reading of The Waste Land as a lament for a post-war crisis of civilized values enacting ‘a complete severance between his poetry and all beliefs’.3 A few years later, Eliot offered a tart response in the pages of Wyndham Lewis’s combative avant-garde magazine The Enemy: ‘I cannot for the life of me see the “complete separation” from all belief ’.4 Nor did he believe that academic solutions were better across the Atlantic. In 1937, Cleanth Brooks’s explanation of the poem as an agonized search for faith tracing a schematic Grail quest was met by Eliot’s patient (yet, one suspects, patronizing) letter: 191

192

Life of Texts

Reading your essay made me feel . . . that I have been a great deal more ingenious than I had been aware of, because the conscious problems with which one is concerned in the actual writing are more those of a quasi-musical nature, in the arrangement of metric and pattern than those of a conscious exposition of ideas.5

It is a pity that Brooks, an advocate of New Critical irony, could detect no irony in Eliot’s claim in the notes added to the poem that Jessie Weston’s thesis in From Ritual to Romance (now discredited as scholarship) ‘will elucidate the difficulties of the poem’.6 Matters were not helped when Brooks’s article was reprinted in his influential collection of essays Modern Poetry and the Tradition, and in several anthologies, leading to an epidemic of Grail-themed exam answers in American college classrooms. In 1956, before an audience of 14,000 in a Minnesota basketball arena, Eliot addressed the shortcomings of academic criticism head-on, instancing ‘the remarkable exposition of bogus scholarship’ sparked by his notes to The Waste Land. ‘I have sometimes thought of getting rid of these notes; but now they can never be unstuck’.7 Talk of unsticking the notes to The Waste Land takes us to the heart of those sticky, sometimes intractable, issues addressed in this collection of essays on ‘The Life of Texts’. While it might be supposed that the student of a twentiethcentury poet faces fewer textual problems than those raised by the poetry of Homer, Dante, Shakespeare – where no authorial manuscripts exist, let alone the modern advent of typescripts and carbon copies – this would be to oversimplify. Although both practices flourish on the wise congruence of scrutiny and judgement, the frontiers between textual scholarship and literary criticism are uncertain. None of the issues that I will examine concerning textual variants, corruptions and emendations could decisively settle any overarching thesis regarding the meaning of The Waste Land. However, I do hope to unsettle speculative theories about authorial intention, or critical dogmas about the role of collaboration in the writing of this poem, that have been advanced on the basis of the evidence of a fragmented, incomplete textual record. This is not an admission of a failure to solve the puzzle or ‘elucidate the difficulties of the poem’ since the experience of reading of The Waste Land cannot be reduced to the decryption of a scrambled logic. Rather, the benefit of approaching in a spirit of open-minded enquiry the extant textual materials bearing on the production, transmission and reception of The Waste Land can return us to this arresting, disturbing, elliptical and enigmatic poem with a renewed sense of wonder.

Fragments Shored against Ruin

193

The Drafts of The Waste Land The publication in 1971 of Valerie Eliot’s facsimile and transcript of the original drafts of The Waste Land, including the annotations of Ezra Pound, inaugurated a new era in our understanding of the poem. These manuscripts, long thought lost, were in fact presented in 1922 to John Quinn, a New York lawyer, art collector and patron to a coterie of modernist writers (he defended James Joyce’s Ulysses during a trial for obscenity). On 19 July, Eliot wrote to Quinn: ‘I should like to present you the manuscript of The Waste Land, if you would care to have it – when I say manuscript, I mean that it is partly manuscript and partly typescript, with Ezra’s and my alterations scrawled all over it’.8 On 21 August, Eliot told Quinn: ‘I certainly cannot accept your proposal to purchase the manuscript at your own price, and if you will not accept it in recognition of what you have done for me lately and in the past, it will not be any pleasure to me to sell it to you. I therefore hope that you will accept it.’9 A month later, Eliot told Quinn, ‘I have gathered together all of the manuscript in existence.’10 He posted a packet containing these drafts on 23 October and they arrived at Quinn’s New York office on 13 January 1923. Sadly, Quinn died of intestinal cancer on 28 July 1924 at the age of 54. While his valuable collection of paintings and sculptures were sold at auction in 1927, the manuscripts of The Waste Land were not mentioned in his will. They were bequeathed to his sister, Julia, and following her death in 1934, they were inherited by her daughter, Mary Conroy, and for many years lay in storage cases in an apartment in Hillsborough, California. After a thorough search of these cases, Mrs Conroy rediscovered the manuscripts and on 4 April 1958 they were acquired by the Berg Collection of the New York Public Library for $18,000. The purchase remained private and Eliot was not informed of the sale. In 1946, Eliot wrote to his bibliographer, Donald Gallup, that the manuscripts and typescripts of The Waste Land had ‘disappeared from sight after Mr. Quinn’s death’.11 In 1959, he reported to Donald Hall in an interview that the whereabouts of the original manuscripts were an ‘unsolved mystery’.12 The Curator of the Berg Collection, Dr John D. Gordon, made an attempt to contact Eliot in London to discuss a ‘business matter’ but a meeting was not possible.13 Eliot died in 1965 unaware of their location. Valerie Eliot was eventually informed of the purchase in the summer of 1968 by James W. Henderson, Chief of the New York Research Libraries, but she was asked to observe secrecy until the publication of Ben Reid’s biography The Man from New York: John Quinn and his Friends, which contained a detailed account of the manuscripts. No public announcement of the sale was made until a New York

194

Life of Texts

Public Library press release on 25 October 1968, timed to coincide with the publication of Reid’s biography of Quinn. The draft materials held by the New York Public Library are fascinating documents: a corpus of 54 leaves, divided into two groups, comprising some twenty-six distinct pre-publication documents (although classification is far from straightforward and remains contested). There are six manuscripts and five typescripts (with four carbon copies) of draft materials that are definitely related to the published poem, as well as several miscellaneous drafts which are less closely related to the poem as published and which are only doubtfully classified as ‘original drafts’ of The Waste Land. Some of the typescripts, for example the opening of Part II , ‘A Game of Chess’, are heavily marked with cancellations, alternative readings and annotations in the hands of Pound and Vivien Eliot. In order that this material could be available for study by scholars, photographs of these papers were produced with transcriptions on a facing page. The poet’s widow shaped the reception of The Waste Land drafts in her edition of 1971. Her editorial introduction and explanatory notes highlighted the role of Pound, who contributed a preface, allowing scholars to appreciate for the first time the full force of the dedication to il miglior fabbro (‘the better craftsman’) that Eliot had privately bestowed on Pound and publicly acknowledged in the 1925 publication of the poem in his collected Poems 1909–1925. Many reviewers of the 1971 edition argued that Pound’s excisions liberated the final five-part structure of the poem out of these disorderly drafts. The Waste Land exhibits the experimental poetics of Pound’s own ‘ideogrammatic method’ since it juxtaposes multi-vocal and multilingual fragments without discursive narrative connections. In this sense, the structural dynamics of the poem reflect Pound’s Cantos rather than the smoother transitions between verse paragraphs apparent in the interior monologues of ‘The Love Song of J. Alfred Prufrock’ or ‘Gerontion’. From the drafts, it is clear that Pound counselled Eliot to remove a series of satirical couplets (a pastiche of Pope’s The Rape of the Lock) and a lengthy shipwreck passage (inspired by the Ulysses canto in Dante’s Inferno). Pound scored out 26 lines of the satirical quatrains depicting the loveless sexual encounter of the typist and young clerk, and the rest were interlinked as a continuous passage. Pound persuaded Eliot to discard three miscellaneous poems that may have been intended to feature as lyric interludes between the section Parts. It is not clear from the drafts whether Eliot removed a long opening scene depicting a drunken night out in Boston at Pound’s behest. Moreover, Eliot adopted dozens of minor suggestions for improvements of phrasing from Pound’s pencil and ink annotations on the drafts. In 1938, Eliot publicly credited

Fragments Shored against Ruin

195

Pound with turning ‘The Waste Land from a jumble of good and bad passages into a poem’.14 In a 1946 essay, Eliot declared that if the missing ‘manuscript of a sprawling, chaotic poem called The Waste Land’ were ever discovered, they would constitute ‘irrefutable evidence of Pound’s critical genius’.15 In a 1922 letter to Eliot, Pound wittily claimed in a poetic squib that ‘Ezra performed the caesarean Operation’ on The Waste Land.16 These obstetrics are the most remarkable instance of a creative collaboration between two major poets since Wordsworth and Coleridge laboured on the Lyrical Ballads. Critics were rightfully grateful for the appearance of The Waste Land drafts, but it is important that we are clear-sighted about what the 1971 facsimile edition does not do. Valerie Eliot’s facsimile edition of The Waste Land drafts did not solve the question of the precise chronology of composition of these miscellaneous materials. Lawrence Rainey’s inspection in 2004 of the watermarks on distinct batches of paper used in the manuscripts and typescripts, supported by his identification of the three different typewriters used to produce the drafts, allied to additional evidence gathered from correspondence, does enable a reconstruction of the rough chronology of the writing of the poem. Rainey’s own conclusions were the object of sceptical dissent from Jim McCue in an essayreview published in Essays in Criticism in January 2006, and so I will carefully reassemble the material evidence in detail below without seeking to advance suppositions that are not openly available to inter-subjective scrutiny.

The Waste Land: From Composition to Publication After publication of the drafts, conjectural theories sprang up claiming to establish the order in which the published parts were written. One ingenious American academic, Grover Smith, contended that Part III was the first part to be written. The extant material, however, points towards the unexciting conclusion that the poem was mostly composed seriatim in the order Parts I, II , III , IV, V. Importantly, textual evidence suggests that Eliot only began writing the poem in earnest in 1921. Wyndham Lewis informed his patron Sydney Schiff in early February 1921 that Eliot had begun a new long poem. On 3 April, Eliot told Schiff: ‘My poem has still so much revision to undergo that I do not want to let anyone see it yet, and also I want to get more of it done – it should be much the longest I have ever written.’17 On 9 May, Eliot informed Quinn that he had ‘a long poem in mind and partly on paper which I am wishful to finish’.18 No manuscripts of Part I or II exist but at some point over the summer Eliot produced (on an old

196

Life of Texts

typewriter he had been using since student days) typescript drafts of Parts I and II of a poem then provisionally titled ‘He Do the Police in Different Voices’ (after Dickens), although it should be recalled that this working title was not retained on later drafts. It appears that work on the poem was suspended in June, after the arrival in London of Eliot’s elderly mother and his sister Marian, when the Eliots decamped from their flat in Clarence Gate Gardens to a cramped attic room in Wigmore Street. In September, Eliot suffered an emotional collapse and took three months’ leave from Lloyds Bank in the City of London. In mid-October he went to the seaside resort of Margate in Kent from where he told Schiff: ‘I have done a rough draft of part of part III , but do not know whether it will do, and must wait for Vivien’s opinion as to whether it is printable. I have done this while sitting in a shelter on the front . . . But I have written only some fifty lines, and have read nothing, literally’. The New York Public Library papers contain manuscript drafts of 53 lines in pencil on Hieratica Bond notepaper (the same paper used in the letter to Schiff ) including the poignant lines ‘On Margate Sands. / I can connect / Nothing with nothing’.19 A typescript draft of Part III was produced on Eliot’s return to his London home, on a new Corona typewriter that had been left as a present by his brother Henry, before Eliot travelled to Lausanne (via Paris) where he underwent a rest cure with the psychiatrist Dr Roger Vittoz. Recovering health, Eliot wrote drafts of Part IV and V on quadruled paper in Lausanne on the shore of Lake Geneva (Lac Leman), reflected in another coded yet poignant lament: ‘By the waters of Leman I sat down and wept . . .’20 In early January 1922, Eliot returned to London via Paris. This was another opportunity to discuss the poem with Pound and he typed a fair copy of Parts IV and V on Pound’s typewriter containing a distinctive violet ribbon. Still uncertain of the final form of the poem, Eliot left drafts of all five parts with Pound. They had taken him almost exactly a year to write. In correspondence in late January, Eliot digested Pound’s recommendations on the annotated typescript of the full draft of The Waste Land. Pound had dramatically reshaped the poem to run ‘from April . . . to shantih without break. That is nineteen pages, and let us say the longest poem in the Englisch langwidge. Dont [sic] try to bust all records by prolonging it three pages further.’21 Eliot wondered whether he should lose the whole of Part IV which Pound had extensively cut (‘Perhaps better omit Phlebas also???’22) as well as the epigraph taken from Conrad’s Heart of Darkness. He proposed ‘Gerontion’ as a preface but Pound replied: ‘I do not advise printing Gerontion as preface. . . . I DO advise keeping Phlebas . . . an integral part of the poem.’23 During this period, Eliot and

Fragments Shored against Ruin

197

Pound had opened negotiations for publication of the poem with the New York publisher Horace Liveright, and with Scofield Thayer, editor of the New York monthly magazine The Dial. The story of these tangled complicated negotiations has been told by Lawrence Rainey, including the abortive approach to the publisher of deluxe limited editions Maurice Firuski, and Eliot’s irritation at Thayer’s initial offer of $150 for the poem, until it was finally agreed that he would be awarded the annual Dial prize of $2,000 (a huge sum equivalent to ten months’ salary at Lloyds Bank). In the summer of 1922 Eliot produced a typescript fair copy of the poem for Quinn, so that he could assist with contract negotiations with Liveright – another typescript was apparently made by Quinn’s New York office before Eliot’s copy was passed to Liveright, and a further fair copy typescript was prepared for the editors of The Dial. In September, The Dial (although the editors had been largely unimpressed by the poem, so far as they could comprehend it) agreed to buy 350 of 1000 copies of the Liveright book edition (offered at a discount to new subscribers of the magazine) in order to secure the first US publication. In the meantime, Eliot pressed ahead with plans for the first UK publication of The Waste Land, to be split over the first two issues of his quarterly review The Criterion. At this stage of proceedings, multiple typescripts and proof-copies were circulating in London and New York – a somewhat chaotic state of affairs. Ever since he had arrived at the final form of the poem, Eliot insisted upon ‘certain spacings essential to the sense’.24 In July 1922, he had worried in a letter to Quinn: ‘I only hope the printers are not allowed to bitch the punctuation and the spacing, as that is very important for the sense.’25 Eliot was pleased with the proof produced by Liveright but he complained to the publisher of The Criterion, Richard Cobden-Sanderson, that the proof ‘of the first part of my poem’ was riddled with ‘undesired alterations made by the printers’.26 At the last moment, a decision was taken to publish The Waste Land entire in the first issue of The Criterion which appeared in London on 16 October 1922. A fortnight later the poem was published in New York in the November issue of The Dial. Then, in mid-December, Boni & Liveright published the first book edition of The Waste Land. It contained line numbers (every ten lines), in order to more easily identify the notes that Eliot added the poem upon Liveright’s prompting (this shrewd publisher was concerned that the poem would otherwise be too short for a book volume). The notes were never part of Eliot’s original intention for the poem and although no manuscript or typescript survives to assist with dating it is clear from correspondence that they were not yet written in mid-July and, in fact, not sent to Liveright until August 1922. The first English book edition of The Waste

198

Life of Texts

Land did not appear until 12 September 1923, typeset by Virginia Woolf on the hand-press of the Hogarth Press in an edition of 443 copies. By then, Eliot was telling correspondents: ‘As for The Waste Land, that is a thing of the past so far as I am concerned and I am now feeling toward a new form and style.’27

Reflecting on the Textual Lives of The Waste Land Scholarship has been slow to assess critically the ramifications of textual instability in The Waste Land. In the ‘Contributions to Periodicals’ section of the 1952 first edition of Donald Gallup’s bibliography of the works of T. S. Eliot, it is incorrectly stated that The Waste Land, as it was first printed in The Criterion, was subsequently ‘Reprinted in The Dial’.28 As we have seen, publication of the poem on either side of the Atlantic followed independent paths of typesetting and also of proof-reading. While Eliot wrestled with ‘undesired alterations’ that were introduced into the two-part Criterion proofs of the poem, he did not see any proofs for The Dial, which was probably typeset from a typescript fair copy as setting copy. Even though Eliot thought the proofs of this first book edition were ‘excellent’29 the text published by Liveright – taken as the copy text for Michael North’s 2001 Norton critical edition and Lawrence Rainey’s 2005 Yale University Press annotated edition – displays a number of omissions (for instance, a missing umlaut on the line from Wagner Öd’ und leer das Meer and missing apostrophes in the pub refrain HURRY UP PLEASE ITS TIME ). In 1936, Eliot explained to Djuna Barnes: ‘I have never succeeded in getting a first edition of one of my own books printed without some errors in it, and I sometimes find that when those are corrected new errors appear.’30 It is noticeable that textual corruptions, albeit minor and somewhat understandable given the hectic circumstances of the appearance of Eliot’s poetry in Britain and America, are described by Eliot as ‘errors’ since they deflect from the textual meaning(s) that have been carefully inscribed in the draft manuscripts and typescripts of his poems. A comparison of the four original debut appearances of The Waste Land – Criterion, Dial, Boni & Liveright and Hogarth Press – reveal striking divergences in the use of font, typography, spacing and indentations. For a start, the inescapable challenge of Eliot’s practice of incorporating non-translated fragments from foreign languages throughout The Waste Land is accentuated in The Dial printing by the use of italics; that is to say, these languages look different on the page, indicative of what Christopher Ricks has called the ‘tonal recesses of foreignness’

Fragments Shored against Ruin

199

that reverberate throughout this poem.31 The emphasis that Eliot placed on The Waste Land’s idiosyncratic punctuation and experimental typography is crucial. In 1930, he explained that ‘the declamation, the system of stresses and pauses’ in poetry, as opposed to prose, ‘is partially exhibited by the punctuation and spacing’.32 The mise en page of the poem as it appeared in both the Criterion and Dial, characterized by a brisk run-on between separate Parts to fit the whole poem into one journal issue, is very different from the widely spaced and expansive type of the Boni & Liveright edition with its narrow measure, so that a great number of the lines are turned. If the typographical spacing of the poem is different in periodical and in book versions, it follows that ‘the declamation, the system of stresses and pauses’ in this idiosyncratically punctuated (often due to a lack of punctuation) sequence will also be dramatically altered. In the short ‘Death by Water’ section, the oceanic assonantal roll of the ‘deep sea / swell’ sweeping across the Boni & Liveright book edition has become in The Criterion, due to the officious intervention of British printers, the hyphenated and bounded (by the addition of a comma) rapid rhythmical wave of ‘deep-sea swell, [sic]’ – a minor but still a significant change. Eliot defended his grammatical licence to John Quinn: ‘I hold that the line itself punctuates, and the addition of a comma, in many places, seems to me to over-emphasise the arrest.’33 Here, then, ‘undesired alterations’ inserted by the printers have, as Eliot feared, wrenched the sense and rhythm of the line. Although the Boni & Liveright edition is arguably the most accurate reproduction of Eliot’s final typescripts, Liveright’s commercial imperative to stretch the poem out to maximum page length had an unfortunate effect on the expressive rhythms – at times sonorous; at times angularly abrupt – of The Waste Land. In September 1923, Eliot thanked Virginia Woolf for the attractive layout of the Hogarth Press edition, laboriously printed by hand on high-quality paper. Eliot said that the ‘spacing and paging are beautifully planned . . . far better than the American edition’ although he apologized for ‘my abominable proofreading’.34 Eliot’s inattention had confirmed Woolf ’s occasional incompetence as a compositor, leaving the published text with several mistakes, most notably the surreal aquatic image of a crowd flowing ‘under’ rather than ‘over’ London Bridge, and in the richly allusive baroque diction describing a lady’s sumptuous boudoir, a ‘coloured’ dolphin was repeated by mistake, thus supplanting a more ornate word ‘carvèd’ (the accent emphasizing the mannered ornate diction had been added by hand on a typescript). Collation of the first four editions of The Waste Land demonstrates that these unique variants of substantives and accidentals in the Hogarth Press edition are unauthorized. It is also on the basis of collation

200

Life of Texts

that it is logical to conclude that the Hogarth Press text is derived from the Boni & Liveright edition. We know that Eliot used the Liveright text of The Waste Land as setting copy for the first Faber trade edition of his collected Poems in 1925. Critical reflections, then, on the textual life of The Waste Land must reckon not only with the formidable interpretative challenge presented by its avant-garde techniques of estrangement but ponder the subtleties of textual instability that have been present right from its birth. Taking the Boni & Liveright edition as the most authoritative text, after all, conflicts with a long-established principle of respecting an author’s final intentions. No modern edition could ignore the dedication to Pound that was added to the Faber edition of 1925 and which is now inseparable from the poem. Then there is the question of authorial changes to the notes. Anglican Eliot, no doubt embarrassed at having described St. Paul’s words to the Philippians as ‘feeble’ in comparison to the Wisdom of the East, revised his note to line 433 in 1932 to suggest that ‘“The Peace which passeth understanding” is our equivalent to’ the soothing Sanskrit Shantihs which end the poem without a full stop.35 Eliot’s death in 1965 did not end the disputes over his final authorial intentions. In her 1971 edition, Eliot’s second wife claimed that Eliot’s first wife had insisted that the cryptic line ‘The ivory men make company between us’ was deleted.36 Eliot restored it, in parenthesis, to an autograph manuscript fair copy of The Waste Land made in 1960 to raise funds for the London Library. The 1963 Faber edition of the Collected Poems, the last edition published during Eliot’s lifetime, incorporating minor changes from the 1961 edition, and thus the final form of The Waste Land authorized by the living poet, does not restore this line about ‘ivory men’. Ricks and McCue include the (parenthetical) line in their authorized 2015 edition, but the authority here is questionable since Eliot had removed it from all publications of the poem during his lifetime.37 Students and critics of The Waste Land are often working with subtly different texts. Even unauthorized corruptions, such as the egregious errors displayed in Lawrence Rainey’s 2005 Yale text (as Jim McCue mischievously pointed out in his review, it is ironic that an editor called Rainey mistranscribes a ‘Drip’ as a ‘Drop’ in Part V’s water-dripping song), have become entangled in the textual lives of this most intensively studied of twentieth-century poems, an exemplar of the textual instability of celebrated modernist texts (Joyce’s Ulysses wins the prize for volatility) studiously resisting the idea of a definite or closed interpretation – the Holy Grail for Cleanth Brooks and countless academic explicators. Is it the status of The Waste Land as the modern poem upon which aspiring academic exegetes have cut their teeth that emboldened Christopher Ricks and

Fragments Shored against Ruin

201

Jim McCue to publish a 678 line Editorial Composite of the poem, complete with Commentary and Textual History, in their heavily annotated two-volume The Poems of T. S. Eliot (2015)? This reading text of the drafts is stitched together from manuscripts and typescripts which, as Ricks and McCue openly admit, ‘are not all of the same stage of development (nor all compatible in a reading text), so this composite does not represent the poem at any particular moment’.38 The case for its inclusion appears to be based on the interest of raw materia poetica seen in relation to the realized final form of the poem. The ‘versioning’ of poems has been a fruitful procedure in editions of Romantic poetry since the 1980s. On the one hand, it might be said that the Editorial Composite of The Waste Land drafts sheds a fascinating light on the composition of the poem, illuminating new critical interpretations. ‘The more we know of Eliot, the better’ as Pound put it in his preface to Valerie Eliot’s facsimile edition.39 On the other hand, the authorized and annotated nature of this Editorial Composite, an artificial construct of Ricks and McCue, not of T. S. Eliot, could interpose interpretations on the poem that actively override or counteract what the author had intended. For it cannot be denied that the rationale for this Editorial Composite overrules the author’s own wishes. Eliot sent the drafts to Quinn with the clear instruction: ‘I hope that the portions which I have suppressed will never appear in print.’40 Shortly before his death, Eliot told Daniel H. Woodward: ‘there was a great deal of superfluous matter in The Waste Land which Pound very rightly deleted.’41 Of course, Valerie Eliot’s 1971 edition had given us the drafts in print, but facsimiles with transcripts are a quite different editorial matter from an Editorial Composite that rests upon a mass of implicit and explicit judgements that will be beyond the grasp of the vast majority of users of The Poems of T. S. Eliot to judge for themselves. Consider, for example, one of the superfluous materials, ‘Dirge’, drafted on Hieratica Bond notepaper in Margate in November 1921, but deleted on Pound’s advice, which resurfaced in public view with the drafts of The Waste Land. In T. S. Eliot and Prejudice, Ricks referred to this draft as the ‘ugliest touch of anti-Semitism in Eliot’s poetry’.42 In T. S. Eliot, anti-Semitism, and Literary Form (1995), Anthony Julius went further by characterizing it as distasteful ‘because it combines sophistication and viciousness, its effects artful and repulsive, bewildering in the way it gloats in tranquillity over the dead Jew, his body still house to disease and subject to humiliation’.43 According to Ricks and McCue, ‘Dirge’ was at one time positioned as a lyric interlude between Parts III and IV of The Waste Land. The editors feel that it ‘would have been grimly appropriate’ as a transition between these sections.44 However, they do not reprint it as part of their Editorial Composite on the grounds that Pound’s

202

Life of Texts

extensive cuts to ‘Death by Water’ rendered it redundant. And so, the reading text of the reconstructed draft poem suppresses the anti-Semitic material that the author of the poem wanted suppressed, and yet in doing so suppresses a draft that the editorial principle of reconstructing an intermediate stage of the workin-progress might have included. By contrast, the epigraph from Conrad heads the reconstructed draft, even though there is no evidence that it was written before 1922. The risks of editorial speculation contaminating critical interpretation of the poem are raised by Ricks and McCue themselves when they reject John Haffenden’s contention that the so-called ‘Fresca couplets’ discovered among Vivien Eliot’s papers at the Bodleian Library, are a prior manuscript draft of the pastiche of Pope’s couplets that Pound had objected to as too diffuse to merit publication. On the basis of scrupulous research, Ricks and McCue date this draft to 1924, two years after the publication of the poem, thereby discrediting Haffenden’s attempt to weave this material into speculations about Eliot’s authorial intentions during the writing of The Waste Land. Textual scholarship should unravel the labour of previous editors, perhaps, as David Fuller suggests in this volume with regard to recent editions of King Lear, in an effort to ‘rescue’ the work. Valerie Eliot’s 1971 edition presents The Waste Land as a poem of private anguish. But to my mind, she accords too much prominence to a reported remark by Eliot, quoted by Theodore Spencer during a lecture at Harvard University, and recorded by Eliot’s brother Henry, which is foregrounded as a resonant epigraph: Various critics have done me the honour to interpret the poem in terms of criticism of the contemporary world, have considered it, indeed, as an important bit of social criticism. To me it was only the relief of a personal and wholly insignificant grouse against life; it is just a piece of rhythmical grumbling.45

Eliot’s most chilling remark on The Waste Land was also recorded privately in a late memoir: ‘To her [Vivien] the marriage brought no happiness . . . to me, it brought the state of mind out of which came The Waste Land.’46 The focus of such remarks has led to biographical readings of The Waste Land that detect an animus towards Eliot’s first wife (foregrounded by his second) that may distort a more balanced view of Vivien’s role in the composition of the poem. There is something terribly resonant about Vivien’s words to Sydney Schiff when she held the first copy in her hands: ‘Perhaps not even you can imagine with what emotions I saw The Waste Land go out into the world. . . . it has become a part of me (or I of it) this last year. It was a terrible thing, somehow, when the time came at last for it to be published.’47 Examination of the textual history of The Waste

Fragments Shored against Ruin

203

Land demonstrates that Vivien was a valued commentator on the drafts, if less of a collaborator than Pound, certainly an important contributor to what D. F. McKenzie has called the ‘sociology of the text’ – ‘human motives and interactions which texts involve at every stage of their production, transmission and consumption’ – a process in which compositors, printers, editors and publishers join hands with Eliot, his wife Vivien, and midwife Pound, in the ongoing transformission (to use Randall McLeod’s term) constituting the life of the texts of The Waste Land.48

Notes 1 Editing Homer

1

2 3

4 5 6 7 8 9 10

11

12

I would like to thank my friend and colleague Carlo Caruso for his vision in bringing together a whole group of Durham scholars interested in textual transmission and for his patience in seeing this volume through to completion. A visit to the Ashmolean Museum in Oxford is instructive: vast quantities of papyrus scraps, discovered at the turn of the twentieth century at Oxyrhynchus, Upper Egypt, await patiently in cardboard boxes for scholars to publish first editions. I wonder whether they will all be edited by the turn of the twenty-second century. See B. Graziosi, Inventing Homer: The Early Reception of Epic (Cambridge: Cambridge University Press, 2002), ch. 2. See, for example, M. L. West, ‘The invention of Homer’, Classical Quarterly 49 (1999): 364–82, where West argues that there was one single poet who created the Iliad, but that he was not called Homer. See also M. L. West, Studies in the Text and Transmission of the Iliad (Munich and Leipzig: Saur, 2001). Graziosi, Inventing Homer, ch. 5. Herodotus, Histories 2.117. Plato, Ion 540b. Aristotle, Poetics 1460a. Aristotle, Poetics 1451a. On the place of Homer in the education system of the classical period, see T. Morgan, ‘Literate education in classical Athens’, Classical Quarterly 49 (1999): 46–61. On the origins of ancient criticism, the standard point of reference is A. Ford, The Origins of Criticism: Literary Culture and Poetic Theory in Classical Greece (Princeton: Princeton University Press, 2002). R. Pfeiffer, History of Classical Scholarship: From the Beginnings to the End of the Hellenistic Age (Oxford: Oxford University Press, 1968) is still an important work of reference on Alexandrian scholarship. On the ‘character’, ‘habit’ and ‘persona’ of the poet, as well as on ancient scholarship more generally, essential reading is R. Nünlist, The Ancient Critic at Work: Terms and Concepts of Literary Criticism in Greek Scholia (Cambridge: Cambridge University Press, 2009). On the motto ‘clarify Homer with reference to Homer’, see J. Porter, ‘Hermeneutic lines and circles: Aristarchus and Crates on the exegesis of Homer’, in

204

Notes to pp. 31–36

13

14 15 16

17

18

19

20 21 22 23

24

205

R. Lamberton and J. J. Keaney (eds), Homer’s Ancient Readers: The Hermeneutics of Greek Epic’s Earliest Exegetes (Princeton: Princeton University Press, 1992) and R. Nünlist, ‘What does Ὅμηρον ἐξ Ὁμήρου σαφηνίζειν actually mean?’, Hermes 143 (2015): 385–403. On the ‘solution from character’ and its attribution to the Alexandrian scholar Aristarchus, H. Dachs is still essential reading (Die lysis ek tou prosōpou: Ein exegetischer und kritischer Grundsatz Aristarchs und seine Neuanwendung auf Ilias und Odyssee (Erlangen: Junge & Sohn, 1913)). For a briefer and more recent discussion in English, see Nünlist, The Ancient Critic at Work, ch. 4. Porphyry, Quaestionum homericarum ad Iliadem pertinentium reliquiae, ed. Schrader, 99f., discussing Iliad 6.258–65. The line in question is Odyssey 3.226. The views of Zenodotus and (implicitly) Aristarchus are preserved in Schol. H. M. ad Od. 3.228, ed. F. Pontani (Scholia graeca in Odysseam (Rome: Edizioni di storia e letteratura, 2007–). For discussion, see H. Dachs, Die lysis ek tou prosōpou. Eustathius 1337.29–30, ed. G. Stallbaum (Eustathii archiepiscopi Thessalonicensis commentarii ad Homeri Odysseam ad finem exempli romani editi (Leipzig: Weigel, 1825–6)). On the scholia, E. Dickey offers essential guidance in Ancient Greek Scholarship: A Guide to Finding, Reading, and Understanding Scholia, Commentaries, Lexica, and Grammatical Treatises, from Their Beginnings to the Byzantine Period (Oxford: Oxford University Press, 2007). On the most important Byzantine manuscript of Homer, Marcianus Graecus 822 (now in Venice), see the attractive introduction edited by C. Dué: Recapturing a Homeric Legacy: Images and Insights from the Venetus. A Manuscript of the Iliad (Washington, DC : Center for Hellenic Studies – Cambridge, MA and London: Harvard University Press, 2009). Giambattista Vico, New Science 3.1. See A. Grafton, G. W. Most and J. E. G. Zetzel (eds), F. A. Wolf: Prolegomena to Homer, 1795 (Princeton: Princeton University Press, 1985). Goethe’s couplet on Wolf ’s Homer is reported in J. W. Goethe, Gedenkausgabe der Werke. Briefe und Gespräche, ed. E. Beutler (Zurich: Artemis, 1953), II , 478. On these two views, see, respectively, G. Nagy, Poetry as Performance: Homer and Beyond (Cambridge: Cambridge University Press, 1996) and G. Nagy, Homeric Questions (Austin, TX : University of Texas Press, 1996); also West, ‘The invention of Homer’, and Homeri Ilias, ed. M. L. West, 2 vols (Stuttgart and Leipzig: Teubner, 1998–2000). Nietzsche’s inaugural lecture is published in F. Nietzsche, Werke. Kritische Gesamtausgabe, ed. G. Colli and M. Montinari (Berlin: de Gruyter, 1967–91), II .1 (1982), 247–79.

206

Notes to pp. 37–42

25 The most significant publications are M. Parry, The Making of Homeric Verse. The Collected Papers of Milman Parry, ed. A. Parry (Oxford: Oxford University Press, 1971), and A. Lord, The Singer of Tales, 2nd edition with CD -ROM , ed. S. Mitchell and G. Nagy (Cambridge, MA : Harvard University Press, 2000). On the Bosnian epic tradition, G. Danek is excellent: Bosnische Heldenepen (Klagenfurt: Wieser, 2003). 26 For further discussion, see Homer: Iliad VI, ed. B. Graziosi and J. Haubold (Cambridge: Cambridge University Press, 2010), Introduction. 27 J. M. Foley makes this point with precision and eloquence in Homer’s Traditional Art (University Park: Pennsylvania State University Press, 1999). 28 Iliad 22.1–20. 29 Iliad 5.302–4, 12.445–9; 20.285–7. 30 A good point made by A. C. Cassio in ‘Early editions of the Greek epics and Homeric textual criticism’, in Omero tremila anni dopo, ed. F. Montanari (Rome: Edizioni di Storia e Letteratura, 2002), 105–36. 31 See O. Taplin, Homeric Soundings: The Shaping of the Iliad (Oxford: Oxford University Press, 1992). 32 The main sources are collected and discussed in H. Kotsidou, Die musischen Agone der Panathenäen in archaischer und klassischer Zeit: Eine historisch-archäologische Untersuchung (Munich: Tuduv, 1991). 33 Again discussed in Kotsidou, Die musischen Agone. 34 Schol. Dionysius Thrax 29.16–30.17, ed. Hilgard. 35 On modern-day philological one-upmanship, see H. U. Gumbrecht, The Powers of Philology: Dynamics of Textual Scholarship (Urbana and Chicago, IL : University of Illinois Press, 2002). 36 The story concerning the Septuagint is best preserved in the Letter of Aristeas, second century ad. 37 Compare the latest edition of Homeri Odyssea, ed. M. L. West (Berlin and Boston: de Gruyter, 2017), xvii: ‘hoc igitur tenui, ne eideise edendo lectores horrore afficerem’, i.e. ‘I kept the form eideise in order not to horrify the reader.’ 38 Homeri Ilias, ed. H. van Thiel (Hildesheim: Olms, 1996) and Homeri Ilias, ed. West. 39 Homeri Odyssea, ed. H. van Thiel (Hildesheim: Olms, 1991). 40 See, for example, Homeri Odyssea, ed. West, xx–xxi on heōs and teōs, xxii on genitive vs dative personal pronouns after different forms of the imperative ‘listen!’ and xxiii on the past tense of oida, ‘to know’. 41 These are based on B. Graziosi and J. Haubold, ‘The Homeric text’, in New Approaches to Homer and Hesiod, ed. H. Morales and S. Lindheim, Ramus 44 (2015): 5–28. 42 Homeri Odyssea, ed. van Thiel, p. xxiv. 43 Homeri Ilias, ed. West, and West, ‘The invention of Homer’. 44 Homer’s Iliad, trans. S. Mitchell (New York: Weidenfeld, 2011).

Notes to pp. 42–46

207

45 On Iliad 10, see C. Dué and M. Ebbott, Iliad 10 and the Poetics of Ambush: A Multitext Edition with Essays and Commentary (Cambridge, MA : Harvard University Press, 2010). 46 E.g. R. Janko, ‘West’s Iliad’, Classical Review 50 (2000): 1–4; G. Nagy, Review of Homeri Ilias, ed. West, Bryn Mawr Classical Review (= BMCR) 2000.09.12 and G. Nagy, Review of West, Studies, Gnomon 75 (2003): 481–501; J.-F. Nardelli, Review of West, Studies, BMCR 2001.06.21; A. Rengakos, Review of West, Studies, BMCR 2002.11.15; M. L. West, ‘West on Nagy and Nardelli on West’, BMCR 2001.09.06; M. L. West, ‘West on Rengakos’, BMCR 2002.11.15; M. L. West, ‘Nagy (Gnomon 75, 2003, 481–501) on West’, BMCR 2004.04.17. 47 See, among several other publications, G. Nagy, Poetry as Performance: Homer and Beyond (Cambridge: Cambridge University Press, 1996); G. Nagy, Homeric Questions (Austin, TX : University of Texas Press, 1996) and G. Nagy, Homer’s Text and Language (Urbana and Chicago, IL : University of Illinois Press, 2004). 48 Quotations in Plato are particularly frequent and, on these, J. Labarbe’s L’Homère de Platon (Liège: Bibliothèque de la faculté de Philosophie et Lettres, 1949) remains relevant. 49 See S. West, The Ptolemaic Papyri of Homer (Cologne: Westdeutscher Verlag, 1967); M. Haslam, ‘Homeric papyri and transmission of the text’, in A New Companion to Homer, ed. I. Morris and B. Powell (Leiden: Brill, 1997), pp. 55–100; and G. D. Bird, Multitextuality in the Homeric Iliad: The Witness of the Ptolemaic Papyri (Cambridge, MA : Harvard University Press, 2010). 50 See M. Finkelberg, ‘The Cypria, the Iliad, and the problem of multiformity in oral and written tradition’, Classical Philology 95 (2010): 1–11. 51 See Graziosi and Haubold, ‘The Homeric text’, itself drawing the primary material from Homer: Iliad VI, ed. Graziosi and Haubold. 52 See B. Graziosi, ‘Homeric scholarship in its formative stages’, in The Homeric Epics and the Chinese Book of Songs: Foundation Texts Compared, ed. F.-H. Mutschler (Newcastle upon Tyne: Cambridge Scholars Publishing, 2018). 53 Graziosi and Haubold, ‘The Homeric text’. 54 See M. Leumann, Homerische Wörter (Basel: Reinhardt, 1950); L. Ph. Rank, Etymologiseering en verwante verschijnselen bij Homerus (Assen: Van Gorcum, 1952); A. Dihle, ‘Leumanns Homerische Wörter und die Sprache der mündlichen Dichtung’, Glotta 48 (1970): 1–8; E. Risch, Wortbildung der homerischen Sprache (Berlin and New York: de Gruyter, 1974); and O. Hackstein, Die Sprachform der homerischen Epen. Faktoren morphologischer Variabilität in literarischen Frühformen: Tradition, Sprachwandel, sprachliche Anachronismen (Wiesbaden: Reichert, 2002). 55 This is an insight that Johannes Haubold and I plan to explore further in future work. 56 A point demonstrated by L. Battezzato in Tradizione testuale e ricezione letteraria antica della tragedia greca (Amsterdam: Hakkert, 2003).

208

Notes to pp. 46–53

57 See Homer in the Twentieth Century: Between World Literature and the Western Canon, ed. B. Graziosi and E. Greenwood (Oxford: Oxford University Press, 2007).

2 The Canon and the Codex: On the Material Form of the Christian Bible 1 On the authorship issue, see I. H. Marshall, A Critical and Exegetical Commentary on the Pastoral Epistles, International Critical Commentary (Edinburgh: T. & T. Clark, 1999), 57–92. A possible second-century context in conflict over the Pauline heritage is proposed by D. R. MacDonald, The Legend and the Apostle: The Battle for Paul in Story and Canon (Philadelphia, PA : Westminster, 1983). 2 According to T. C. Skeat, malista here does not mean ‘especially’ but introduces a definition: ‘. . . the books, that is, the parchments’ (‘Especially the Parchments: A Note on 2 Timothy iv.13’, Journal of Theological Studies 30 (1979): 173–7). The use of malista elsewhere in the Pastorals (1 Tim. 4.10, 5.17; Tit. 1.10) provides no clear support for Skeat’s claim. Even if it is correct, the present argument would not be materially affected. 3 Theodoret, Commentarius in omnes sancti Pauli epistolas, in Patrologia Graeca 82.853. 4 In Luke 4.17, where Jesus is said to ‘unroll’ the book of Isaiah (anaptuxas), a textual variant has him simply ‘open’ it (anoixas). The variant replaces a Jewish roll with a Christian codex. 5 C. H. Roberts and T. C. Skeat, The Birth of the Codex (London: OUP for the British Academy, 1987), 15–23. 6 The indicative figures given here are based on the catalogue of biblical papyri in Larry W. Hurtado, The Earliest Christian Artifacts: Manuscripts and Christian Origins (Grand Rapids, MI : Eerdmans, 2006), 209–29. 7 Hurtado, Earliest Christian Artifacts, 44–9, 90–3. 8 Inst. or. 10.3.31–2 (Quintilian, The Orator’s Education, ed. D. A. Russell, 5 vols (Cambridge, MA : Harvard University Press, 2003). Here and elsewhere, translations are my own. On writing tablets and notebooks, see Roberts and Skeat, Birth of the Codex, 15–23; Graham N. Stanton, Jesus and the Gospel (Cambridge: Cambridge University Press, 2004), 173–8. 9 Of the 180 tables of this type found in the Vindolanda excavations near Hadrian’s wall, many ‘contain remains of incised texts, often palimpsests, which have survived where the metal stilus penetrated the wax coating on the wooden surface below’ (Alan K. Bowman, Life and Letters on the Roman Frontier: Vindolanda and its People (London: British Museum, 20033), 8). 10 Inst. or. 10.3.33.

Notes to pp. 53–58

209

11 Martial, Epigrams, 1.2, 2 vols (London and New York: Heinemann and Putnam, 1919). 12 Martial, Epigrams, 14.190 (cf. 14.184, entitled Homerus in pugillaribus membranis; 14.186, Vergilius in membranis; 14.188 [Cicero], 192 [Ovid]). 13 Harry Gamble, Books and Readers in the Early Church: A History of Early Christian Texts (New Haven, CT and London: Yale University Press, 1995), 58. 14 For a survey of scholarly views on this, see Hurtado, Earliest Christian Artifacts, 61–83. 15 See Francis Watson, Gospel Writing: A Canonical Perspective (Grand Rapids, MI : Eerdmans, 2013), 217–85. 16 Watson, Gospel Writing, 117–216; Mark Goodacre, The Case against Q: Studies in Markan Priority and the Synoptic Problem (Harrisburg, PA : TPI , 2002). 17 Thomas’s familiarity especially with Matthew and Luke is ably and independently demonstrated by Simon Gathercole and Mark Goodacre: Simon J. Gathercole, The Composition of the Gospel of Thomas: Original Language and Influences, Society for New Testament Studies Monograph Series (Cambridge: Cambridge University Press, 2012); Mark Goodacre, Thomas and the Gospels: The Case for Thomas’s Familiarity with the Synoptics (Grand Rapids, MI : Eerdmans, 2012). 18 2 Clem. 4.5. 19 Roberts and Skeat are too quickly dismissive of ‘ease of reference’ as a factor in the rise of the codex (Birth of the Codex, 50–1). ‘[E]ven though it may be true that practical advantages alone do not adequately explain the broad adoption of the codex in early Christianity . . . those advantages cannot simply be disregarded’ (Gamble, Books and Readers, 274n). 20 Eusebius’s canon tables and his explanatory letter to Carpianus are printed in successive editions of the Nestle-Aland Novum Testamentum Graece (Stuttgart: Deutsche Bibelgesellschaft, 201228), 89*–94*. Representative and aesthetically attractive canon tables are found in the Lindisfarne Gospels (Latin, eighth century), the Rabbula Gospels (Syriac, sixth century), the Garima Gospels (I, Coptic; II , Ethiopic; both c. 600), and the Zeyt’un Gospels (Armenian, thirteenth century). 21 Eusebius, Vita Constantini, 4.36: πεντήκοντα σωμάτια ἐν διφθέραις . . . τῶν θείων δηλαδὴ γραφῶν ὧν μάλιστα τήν τ’ ἐπισκευὴν καὶ τὴν χρῆσιν τῷ τῆς ἐκκλησίας λόγῳ ἀναγκαίαν εἶναι γινώσκεις (Greek text in Eusebius Werke, Band 1.1, ed. F. Winckelmann, Die Griechischen Christlichen Schriftsteller (Berlin: Akademie Verlag, 19912)). A reference here to single-volume Bibles would deprive the relative clause of any real function. Constantine did not send Eusebius an order for ‘fifty new Bibles’, as David L. Dungan claims (Constantine’s Bible: Politics and the Making of the New Testament (London: SCM Press, 2006), 121). Gamble rightly notes the practical impossibility of any such request (Books and Readers, 79–80). 22 Portable and affordable single-volume Bibles, ‘small enough to comfortably fit into a large pocket or small satchel’, were already in widespread use during the thirteenth

210

23

24 25 26

27 28 29 30 31 32 33 34

35

36 37

Notes to pp. 58–61

century (Eyal Poleg and Laura Light, ‘Introduction’, in Form and Function in the Late Medieval Bible, ed. E. Poleg and L. Light (Leiden: Brill, 2013), 1–7; 2). Cassiodori Senatoris Institutiones, ed. R. A. B. Mynors (Oxford: Clarendon, 1937). An English translation is available which unfortunately tends to translate the key term codices as ‘sections’ (i.e. of the Bible), a concept for which Cassiodorus actually uses the term divisio. Cassiodorus Institutions of Divine and Secular Learning, and, On the Soul, ed. James W. Halporn and Mark Vessey (Liverpool: Liverpool University Press, 2004), 109 and passim. For Cassiodorus’s life and work, see James J. O’Donnell, Cassiodorus (Berkeley, CA : University of California Press, 1979). Inst. 1.8.15, 14.4. The novem codices are introduced in Inst. 1, praef. 8, 10, and the following chapters are devoted to each of them in turn and to their accompanying patristic commentaries (Inst. 1.1–9; cf. also 1.11.3, 13.2). The fifth codex (the prophets) is dealt with out of sequence in the third chapter, probably under the influence of Jerome’s organization of the Old Testament canon (cf. Inst. 1.12.1). Cassiodorus, Inst. 1.9.1, 13.1–2 (Augustine, De Doct. Chr. 2.8.13). Inst. 1, praef. 2. Inst. 1.12.1–3. Inst. 1.14.1–2. Inst. 1.14.4. Inst. 1.12.2; 14.2, 4. Inst. 1.13.2. This is reported in Bede’s History of the Abbots (15), and in a closely related anonymous work of the same title (15, 37). Latin texts in Venerabilis Baedae Historiam ecclesiasticam gentis Anglorum, Historiam abbatum, Epistolam ad Ecgberhtum, una cum Historia abbatum auctore anonymo, ed. C. Plummer, 2 vols (Oxford: Clarendon, 1896), I, 364–87 and 388–404. The Northumbrian pandects contained the ‘new translation’ (Bede, Hist. abb. 15), that is, Jerome’s translation from the Hebrew (Anon., Hist. abb. 37) together with his edition of the gospels. The monastery already possessed a pandect copy of the ‘old translation’, which Abbot Ceolfrith had brought from Rome and which may be related to Cassiodorus’s codex grandior (Bede, Hist. abb. 15). English translations of the two short histories are available in The Age of Bede (Revised Edition), ed. J. F. Webb and D. H. Farmer (London: Penguin Books, 1998). Of the oldest Greek Bibles, the fourth-century Codex Vaticanus has lost its beginning (Gen. 1–46) and end (the Pastorals, Philemon, Revelation). Codex Sinaiticus (also fourth century) survives only in fragments in the books that precede the Psalms. Codex Alexandrinus (fifth century) is more nearly intact, but its four volumes resemble Cassiodorus’s novem codices more closely than his pandects. Anon., Hist. abb. 15. Inst. 1.1.10, 2.13, 5.7, 6.5.

Notes to pp. 62–66

211

3 Wandering Nights: Shahrazād’s Mutations 1 H.. Diyāb, Rih.la, Vatican City, Biblioteca Apostolica Vaticana (= BAV ), MS Sbath 254, fol. 128a. 2 Antoine Galland’s journal (Journal Années 1708–1709), in the Bibliothèque nationale de France (= BNF ), MS fr. 15277, p. 54. 3 Ibid., p. 58. 4 See U. Marzolph, ‘Les Contes de Hanna’, in Les mille et une nuits, ed. É. Bouffard and A. Joyard (Paris: Institut du monde arabe, 2012), 87–91; S. Larzul, ‘Further Considerations on Galland’s Mille et une Nuits: A Study of the Tales Told by Hannâ’, in The Arabian Nights in Transnational Perspective, ed. U. Marzolph (Detroit: Wayne State University Press, 2007), 17–31; R. B. Bottigheimer, ‘East Meets West: Hannā Diyāb and The Thousand and One Nights’, Marvels & Tales 28, no. 2 (2014): 302–24; E. Kallas, ‘Aventures de Hanna Diyab avec Paul Lucas et Antoine Galland (1707– 1710)’, Romano-Arabica 15 (2015): 255–67. 5 A. Galland, Les Mille et une nuit. Contes arabes, 12 vols (Paris: Barbin, 1704–17). 6 The term was coined by M. I. Gerhardt, The Art of Storytelling: A Literary Study of the Thousand and One Nights (Leiden: Brill, 1963), 14. 7 On Galland’s life and works, see R. Schwab, L’auteur des Mille et une nuits: vie d’Antoine Galland (Paris: Mercure de France, 1964); M. Abdel-Halim, Antoine Galland: sa vie et son œuvre (Paris: Nizet, 1964). 8 BNF, MS arabe 3645, Khabar Sindabād al-bah.rī wa Hindabād al-h.ammāl (‘The Story of Sindbad the Sailor and Hindabad the Porter’). 9 Galland, Les Mille et une nuit, I, [4]. 10 L. Valensi, ‘Pétis de la Croix, François’, in Dictionnaire des orientalistes de langue française, ed. F. Pouillon (Paris: Karthala Editions, 2012), 798–9. 11 Munich, Bayerische Staatsbibliothek, Codex Gallicus 799. The text was recently edited by U. Mazolph and A. Chraïbi: Sindbad le Marin. Traduction inédite de 1701 (Paris: Expaces & Signes, 2016). 12 See J. Thomann, ‘Die frühesten türkischen Übersetzungen von Tausendundeiner Nacht und deren Bedeutung für die arabische Textgeschichte’, Asiatische Studien 70, no. 1 (2016): 171–219. 13 Abdel-Halim, Antoine Galland, 414–5. 14 BNF, MSS arabe 3609 (1–67 Nights); arabe 3610 (67–166 Nights); arabe 3611 (167–281 Nights). See The Thousand and One Nights (Alf Layla wa-Layla) from the Earliest Known Sources, ed. M. Mahdi, 3 vols (Leiden: Brill, 1984–94). 15 See J. Lentin, ‘La langue des manuscrits de Galland et la typologie du moyen arabe’, in Les Mille et une Nuits en partage, ed. A. Chraïbi (Paris: Sindbad, 2004) 434–55; H. Grotzfeld, ‘Schriftsprache, Mittelarabisch und Dialekt in 1001 Nacht’, Jerusalem Studies in Arabic and Islam 15 (1992): 171–85; J.-P. Guillaume, Le moyen-arabe,

212

16

17

18

19 20 21

22 23 24 25 26

27

28

Notes to pp. 66–71

langue des Mille et une nuits? (2002): http://www.univ-tours.fr/arabe/Mille et une nuits (accessed 22 August 2017); B. Halflants, Le Conte du Portefaix et des Trois Jeunes Femmes dans le manuscrit de Galland (XIVe-XVe siècles). Édition, traduction et étude du Moyen Arabe d’un conte des Mille et une nuits (Leuven: Peeters, 2007); M. Sartori, ‘La langue des manuscrits grammaticaux arabes médiévaux: entre fushā et āmmiyya’, Romano-Arabica 14 (2014): 301–17. H. Grotzfeld, ‘The Age of the Galland Manuscript of the Nights: Numismatic Evidence for Dating a Manuscript?’, Journal of Arabic and Islamic Studies 1 (1996–7): 50–64; H. and S. Grotzfeld, Die Erzählungen aus ‘Tausendundeiner Nacht’ (Darmstadt: Wissenschaftliche Buchgesellschaft, 1984), 26–7. BAV, MS Vat. Ar. 782; Manchester, John Rylands Library, MS 647; A. Chraïbi, Les Mille et une nuits. Histoire du texte et Classification des contes (Paris: L’Harmattan, 2008), 56; I. Akel, ‘Liste des manuscrits arabes des Nuits’, in Arabic Manuscripts of the ‘Thousand and One Nights’: Presentation and Critical Editions of Four Noteworthy Texts. Observations on Some Osmanli Traditions, ed. A. Chraïbi (Paris: Espaces & Signes, 2016), 71. U. Marzolph and A. Chraïbi, ‘The Hundred and One Nights: A Recently Discovered Old Manuscript’, Zeitschrift der Deutschen Morgenländischen Gesellschaft 162, no. 2 (2012): 299–316 (300). The Arabian Nights Encyclopedia, ed. U. Marzolph and R. Van Leeuwen (Santa Barbara, CA , Denver, CO, Oxford: ABC -CLIO, 2004), 720–1. Diyāb, Rih.la, BAV, MS Sbath 254, fol. 8r. A. Russell, Natural History of Aleppo and parts adjacent, containing a Description of the City, and the principal Natural Productions in its neighbourhood; together with an Account of the Climate, Inhabitants, and Diseases, particularly the Plague; with the methods used by the Europeans for their preservation, 2 vols (London: A. Millarm, 1756 (2nd ed. 1794)), I, 148–50. B. Heyberger, ‘Les Partenaires syriens’, in Bouffard and Joyard (eds), Les mille et une nuits, 71. E. W. Lane, An account of the manners and customs of the Modern Egyptians, 2 vols (London: Charles Knight, 1836), II , 117ff., 130, 150. Diyāb, Rih.la, BAV, MS Sbath 254, fol. 9r–v. Russell, Natural History of Aleppo, I, 149. A. al-Masʿūdī, Murūj al-dhahab wa maʿādin al-jawhar, ed. and trans. C. Barbier de Meynard and P. de Courteille, 9 vols (Paris: Imprimerie nationale: 1861–77), IV (1875), 89–90 (2005: II , 201). M. Ibn al-Nadīm, Kitāb al-fihrist, ed. G. Flúgel, 2 vols (Leipzig: F. C. W. Vogel, 1872), I, 304; B. Dodge, The Fihrist of al-Nadîm: A Tenth Century Survey of Muslim Culture, 2 vols (New York: Columbia University Press, 1970), II , 713–5. The oldest Persian manuscript of this text dates to 1160, though a slightly earlier (1090) Greek version is also extant. The origins of the story probably go back to the

Notes to pp. 71–75

29

30

31

32

33

34

35 36

37

213

famous Middle Egyptian text of The Shipwrecked Sailor. For a discussion of the Persian elements in the 1001 Nights, see U. Marzolph, ‘The Persian Nights: Links Between the Arabian Nights and Iranian Culture’, Fabula, 45, no. 3–4 (2004): 275–93. N. Abbott, ‘A Ninth-Century Fragment of the “Thousand Nights”: New Lights on the Early History of the Arabian Nights’, Journal of Near Eastern Studies 8 (1949): 129–64. S. D. Goitien, ‘The Oldest Documentary Evidence for the Title Alf Laila wa-Laila’, Journal of the American Oriental Society 78 (1958): 301–2 (= The Arabian Nights Reader, ed. U. Marzolph (Detroit: Wayne State University Press, 2006), 83–6). The Arabian Nights Encyclopedia, 372ff.; A. Chraïbi, ‘Qu’est-ce que les Mille et une nuits aujourd’hui? Le livre, l’anthologie et la culture oubliée’, in Bouffard and Joyard (eds), Les mille et une nuits, 33–9. For a discussion, see Akel, ‘Liste’, in Chraïbi (ed.), Arabic Manuscripts, 69–70; Akel and Chraïbi, ‘The Tale of Sûl and Shumûl’, in id., 115–260. The dates are, however, subject to considerable controversy as chronological references in the manuscripts are much later: 1773 (Riyad) and 1843 (Tübingen). Marzolph and Chraïbi, ‘The Hundred and One Nights: A Recently Discovered Old Manuscript’. The Arabic editions available to date are based on later manuscripts, dated to 1190/1776 – ed. M. T.arshūna (Tunis: Dār al-ʿArabiyya li ’l-Kitāb, 1979; 2nd ed. Cologne: al-Jamal, 2005; 3rd ed. Tunis: Bayt al-H . ikma, 2013); French trans. Maurice Gaudefroy-Demombynes, Les Cent et Une Nuits (Paris: Librairie E. Guilmoto, 1911; repr. Paris: Sindbad, 1982) – and 1257/1841 – ed. A. Shuraybit. (Algiers: al-Maktaba al-Wat.aniyya al-Jazā’iriyya, 2005). A German translation of the recently discovered manuscript was made by C. Ott, Hundertundeine Nacht (Zurich: Manesse Verlag, 2012). U. Marzolph, ‘Re-locating the Arabian Nights’, in Philosophy and Arts in the Islamic World. Proceedings of the Eighteenth Congress of the Union Européenne des Arabisants et Islamisants held at the Katholieke Universiteit Leuven (September 3 – September 9, 1996) (Orientalia Lovanensia Analecta 87), ed. U. Vermeulen and D. De Smet (Leuven: Peeters, 1998), 155–63 (158). Tausendundeine Nacht. Das glückliche Ende, trans. C. Ott (Munich: C. H. Beck Verlag, 2016). D. B. MacDonald, ‘The Earlier History of the Arabian Nights’, Journal of the Royal Asiatic Society (1924): 353–97 (393–5); Marzolph, ‘Re-locating the Arabian Nights’, 158; H. and S. Grotzfeld, Die Erzählungen aus ‘Tausendundeiner Nacht’, 41; Akel, ‘Liste’, in Chraïbi (ed.), Arabic Manuscripts, 75–6; Mahdi (ed.), The Thousand and One Nights, II , 298–300. The text covers parts 6–12 of the Nights, as from Night 255. Also see list at the end of this chapter.

214

Notes to pp. 75–78

38 Akel, ‘Liste’, in Chraïbi (ed.), Arabic Manuscripts, 72–3. 39 D. B. MacDonald, ‘Alf laila wa-laila (A.)’, in The Encyclopaedia of Islam. Supplement (Leiden: E. J. Brill, 1938), 17–21. 40 H. El-Shamy, ‘The Oral Connections of the Arabian Nights’, in The Arabian Nights Encyclopedia, 9–13 (10). 41 For a detailed discussion of the background to the Chavis manuscript, see Mahdi (ed.), The Thousand and One Nights, III , 51–61. The manuscript is held at the BNF, MS arabe 3613–4 and 3616. Also see The Arabian Nights Encyclopedia, 520; Akel, ‘Liste’, in Chraïbi (ed.), Arabic Manuscripts, 87–8. 42 Continuation des Mille et une nuits, contes arabes (Geneva: Barde, Manget & compagnie, 1788–9), I, 6. 43 The tale of al-Bunduqānī (the first in the Continuation, I, 11–112) is taken from a volume of miscellaneous stories (BNF, MS arabe 3637, dated 1772); Akel, ‘Liste’, in Chraïbi (ed.), Arabic Manuscripts, 87. 44 H. and S. Grotzfeld, Die Erzählungen aus ‘Tausendundeiner Nacht’, 43–4; Akel, ‘Liste’, in Chraïbi (ed.), Arabic Manuscripts, 87. Edward was the son of Lady Mary Wortley Montagu (1689–1762), known for her letters from Turkey, where she had accompanied her ambassador husband, and for introducing smallpox inoculation to Britain. 45 F. Tauer, Neue Erzählungen aus den tausendundein Nächten. Die in anderen Versionen von ‘1001 Nacht’ nicht enthaltenen Geschichten der Wortley-Montague-Handschrift der Oxforder Bodleian Library. Aus dem arabischen Urtext übertragen und erläutert (Frankfurt am Main: Insel Verlag, 1995). 46 H. Zotenberg, ‘Notice sur quelques manuscrits des Mille et une Nuits et la traduction de Galland’, Notices et extraits des manuscrits de la Bibliothèque nationale 28 (1887): 167–235 (216–7). 47 Russell, Natural History of Aleppo, I, 306. 48 Akel, ‘Liste’, in Chraïbi (ed.), Arabic Manuscripts, 83–4; Mahdi (ed.), The Thousand and One Nights, II , 262–8. The remaining volume (of two) of the Russell manuscript is currently held at the John Rylands Library (No 647 [40]). 49 Service historique de la Défense (SHD ), Paris, ‘XL /37j: Orientaux a Paris, 1789 à 1815: Aumôniers et Interprètes’, dossier ‘Mikha’il Sabbagh’; letter, dated 4 April 1814. 50 Ibid., letter, dated 30 July 1814. 51 Mahdi (ed.), The Thousand and One Nights, III , 61–72. The two-volume manuscript is still at the BNF, MS arabe 4678–9. 52 This was probably also why Chavis chose Baghdad as the origin of his fictitious manuscript. 53 Akel suggested that the first draft is the one in Munich (BS 629), which was followed by the manuscript held in St Petersburg. This was refuted by Lentin, on the basis of differences in handwriting and content. Akel, ‘Liste’, in Chraïbi (ed.), Arabic

Notes to pp. 78–80

54 55 56

57

58

59 60

61 62

63

64

65 66

215

Manuscripts, 91, 100–1; J. Lentin, ‘Comparing the Language of Manuscript Versions of Two Tales: Promise and Limitations’, in id., 353–66 (355). The manuscript submitted to Caussin de Perceval is now at the BNF, MS arabe 4678–9. Mahdi (ed.), The Thousand and One Nights, III : 71–2. Cambridge, Cambridge University Library, MS Qq106–109; Akel, ‘Liste’, in Chraïbi (ed.), Arabic Manuscripts, 92. Besides the complete set (Gotha, MS Orient A 2632–5), he brought back three partial copies (Gotha, MSS Orient A 2637, A 2638, A 2639); Akel, ‘Liste’, in Chraïbi (ed.), Arabic Manuscripts, 80–1, 89–90. The ‘fonds Asselin’ in Paris (BNF, MSS arabe 3595–7, 3598–3601, 3602–5). There is also a partial copy made by Asselin himself (Arabe 3617, 285–326 Nights); Akel, ‘Liste’, in Chraïbi (ed.), Arabic Manuscripts, 97–8. U. Seetzen, Reisen durch Syrien, Palästina, Phönicien, die Transjordan-Länder, Arabia Petraea und Unter-Aegypten, ed. Fr. Kreuser et al., 4 vols (Berlin: G. Reimer, 1854–5), III , 188. Mahdi (ed.), The Thousand and One Nights, III , 99. It was thought to be a copy of the Russell manuscript, made for John Leyde in India in 1811 (British Library, MS IO ISL 2699) until British Library, MS DEL AR 1308 was identified as the source. See Akel, ‘Liste’, in Chraïbi (ed.), Arabic Manuscripts, 82–3, 92–3, 432–4; Mahdi (ed.), The Thousand and One Nights, III , 88–92; The Arabian Nights Encyclopedia, 545. H.ikāyāt mi’at layla min alf layla wa layla, ed. A. al-Yamanī, 2 vols (Calcutta: Hindoostanee Press, 1814–18). See L. Valensi, Mardocée Naggar. Enquête sur un inconnu (Paris: Stock, 2008); I. Coller, Arab France. Islam and the Making of Modern Europe, 1798–1831 (Berkeley/Los Angeles: University of California Press, 2011), 76, 78, 80. W. Jowett, Christian Researches in Syria and the Holy Land in MDCCCXXIII and MDCCCXXIV, in Furtherance of the Objects of the Church Missionary Society: with an Appendix Containing the Journal of Mr. Joseph Greaves on a Visit to the Regency of Tunis (London: L. B. Seeley and J. Hatchard, 1825), 478. Akel, ‘Liste’, in Chraïbi (ed.), Arabic Manuscripts, 434; The Arabian Nights Encyclopedia, 579–80. Only eight volumes were published in Habicht’s lifetime (1824–38), with H. Fleischer editing the last four (1842–3): Kitâb Alf layla wa-layla min al-mubtada’ ilâ al-muntahâ/Tausend und Eine Nacht. Arabisch. Nach einer Handschrift aus Tunis, ed. M. Habicht and H. Leberecht Fleischer, 12 vols (Breslau: Ferdinand Hirt, 1824–43). A copy of Mordechai’s manuscript has survived in Berlin, Staatsbibliothek, MS Do.183–4; Akel, ‘Liste’, in Chraïbi (ed.), Arabic Manuscripts, 88–9. D. B. MacDonald, ‘Maximilian Habicht and His Recension of the Thousand and One Nights’, Journal of the Royal Asiatic Society (1909): 685–704 (686).

216

Notes to pp. 81–83

67 BNF, MS fr. 19277, p. 140; D. B. MacDonald, ‘Further Notes on “Ali Baba and the Forty Thieves” ’, Journal of the Royal Asiatic Society (1913): 41–53 (41–6). 68 D. B. MacDonald, ‘ “Ali Baba and the Forty Thieves” in Arabic from a Bodleian Ms.’, Journal of the Royal Asiatic Society (1910): 327–86; id., ‘Further Notes’. 69 Mahdi (ed.), The Thousand and One Nights, III , 72–86; The Arabian Nights Encyclopedia, 90–1. 70 K. Zakharia, ‘Jean-Georges Varsy et l’Histoire d’Ali Baba: révélations et silences de deux manuscrits récemment découverts’, Arabica 62 (2015): 652–87; id., ‘La version arabe la plus ancienne de l’ “Histoire d’Ali Baba”: si Varsy n’avait pas traduit Galland? Réhabiliter le doute raisonnable’, Arabica 64 (2017): 50–77. 71 BNF, MSS arabe 3602–5. N. Elisséeff, Thèmes et motifs des Mille et une Nuits: Essai de Classification (Beirut: Institut français de Damas, 1949), 6; Akel, ‘Liste’, in Chraïbi (ed.), Arabic Manuscripts, 97. 72 Lane, An account of the manners and customs of the Modern Egyptians, II , 150. 73 The four-volume manuscript is currently in the Strasbourg University Library (Nos 4278–81). See A. Chraïbi, Contes nouveaux des 1001 Nuits: Étude du manuscrit Reinhardt (Paris: Maisonneuve, 1996). 74 T.-X. Bianchi (with the incorrect date of 1836), ‘Catalogue général et detaillé des livres arabes, persans et turcs, imprimés à Boulac en Egypte depuis l’introduction de l’imprimerie dans ce pays’, Journal Asiatique 4, no. 2 (1843): 24–61 (46); Lane, An account of the manners and customs of the Modern Egyptians, II , 381. 75 D. Newman, An Imam in Paris, 2nd rev. ed. (London: Saqi Books, 2011), 48ff.; J. al-Shayyāl, Tārīkh al-tarjama wa ’l-h.araka al-thaqāfiyya f ī ʿas.r Muh.ammad ʿAlī (Cairo: Dār al-Fikr al-ʿArabī, 1951), 39, 149. 76 The Alif Laila Or Book of the Thousand Nights and One Night. Commonly known as ‘The Arabian Nights’ Entertainments’, ed. William Hay Macnaghten, 4 vols (Calcutta: W. Thacker & Co, 1839–42). 77 Mahdi (ed.), The Thousand and One Nights, III , 101–26. 78 H. Zotenberg, ‘Communication relative au texte arabe de quelques contes des Mille et une Nuits’, Journal Asiatique, 8th series, 9 (1887): 300–3; id., ‘Notice’; id., Histoire d’Alâ al-Dîn, ou, La lampe merveilleuse. Texte arabe avec une notice sur quelques manuscrits des Mille et une nuits (Paris: Imprimerie nationale, 1888). 79 Lentin, ‘Comparing the Language’, in Chraïbi (ed.), Arabic Manuscripts, 353–66. 80 See Appendix. This is a number based on Akel, ‘Liste’, in Chraïbi (ed.), Arabic Manuscripts, 65–114. On the manuscript traditions, also see: H. Grotzfeld, ‘The Manuscript Tradition of the Arabian Nights’, in The Arabian Nights Encyclopedia, 17–20; Marzolph, ‘Re-locating the Arabian Nights’; V. Chauvin, Bibliographie des ouvrages arabes ou relatifs aux Arabes, publiés dans l’Europe chrétienne de 1810 à 1885, 12 vols (Liège: H. Vaillant-Carmanne/Leipzig: Otto Harrassowitz, 1892–1922) IV, 197–213.

Notes to pp. 83–93

217

81 This list is based on the Wortley-Montague manuscript; the editions by Habicht and Fleisher (1825–43), Al-ʿAdawī (‘Bulaq’, 1835), Macnaghten (‘Calcutta II ’, 1839–42), Mahdi (1984–94), and on Chraïbi (1996); the translations by Galland (1704–17), Habicht (1825), Weil (1838–41), Lane (1839–41), Payne (1882–4), Burton (1885–8), Mardrus (1899–1904), and Littmann (1921–8). See Appendix I. 82 Chraïbi, Les Mille et une nuits. Histoire du texte et Classification des contes. 83 Data based on Marzolph, ‘Re-locating the Arabian Nights’. 84 London and Westminster Review, XXXIII : 106. 85 S. Larzul, Les traductions françaises des Mille et une Nuits: Etude des versions Galland, Trébutien et Mardrus (Paris: L’Harmattan, 1996), 140–216; R. Irwin, The Arabian Nights: A Companion (London: Allen Lane, 1994), 36–40. 86 The Arabian Nights Encyclopedia, 618–20; Irwin, The Arabian Nights: A Companion, 23–5, 90–1, 112–3. 87 The Arabian Nights Encyclopedia, 504–8; Irwin, The Arabian Nights: A Companion, 28–36. 88 Burton (trans.), A Plain and Literal Translation of the Arabian Nights’ Entertainments, X, 173ff. 89 The Arabian Nights: Tales of 1001 Nights, trans. M. C. Lyons, R. Irwin and U. Lyons, 3 vols (London: Penguin, 2008); B. Lewis, Islam and the West (London: Oxford University Press, 1993), 71. 90 The Arabian Nights, trans. H. Haddawy (New York: W. W. Norton, 1995); Tausendundeine Nacht. Nach der ältesten arabischen Handschrift in der Ausgabe von Muhsin Mahdi erstmals ins Deutsche übertragen, trans. C. Ott (Munich: C.H.Beck, 2011). 91 Most of these have been published in Beirut (twenty-two), and Cairo (ten). See Akel, ‘Liste’, in Chraïbi (ed.), Arabic Manuscripts, 431–91. 92 Based for the most part on Chauvin, Bibliographie des ouvrages arabes, IV, 197–213 and Akel, ‘Liste’, in Chraïbi (ed.), Arabic Manuscripts, 65–114; id., ‘Ahmad al-Rabbât al-Halabî : sa bibliothèque et son rôle dans la réception, diffusion et enrichissement des Mille et une nuits’, unpubl. PhD diss. (2016), INALCO, Paris. The asterisked [*] items refer to complete sets of the Nights.

4 A Text in Exile: Dante’s Divine Comedy 1 L. Bruni, ‘Vita di Dante e del Petrarca’, in Opere letterarie e politiche, ed. P. Viti (Turin: UTET, 1996), 537–60 (548). 2 V. M. Manfredi, L’isola dei morti (Venice: Marsilio, 20042). A book of memoirs by the late Vittore Branca ends with a delicate fantasy about finding the autograph of Dante’s Commedia in the Marciana Library in Venice (V. Branca, Ponte Santa Trinita. Per amore di libertà, per amore di verità (Venice: Marsilio, 1987), 197–9).

218

Notes to pp. 93–98

3 The historic Codice diplomatico dantesco, ed. R. Piattoli (Florence: Gonnelli, 19502; 1st edition 1940) has now been replaced by Codice diplomatico dantesco, ed. T. De Robertis, L. Regnicoli, G. Milani and S. Zamponi, in D. Alighieri, Opere di dubbia attribuzione e altri documenti danteschi, Nuova Edizione Commentata delle Opere di Dante, VII .3 (Rome: Salerno editrice, 2016). 4 R. Davidsohn, Storia di Firenze, trans. G. B. Klein and R. Palmarocchi, 8 vols (Florence: Sansoni, 1972–3; 1st German edition 1896–1927), IV, 172; VII , 211–43; C. T. Davies, ‘Education in Dante’s Florence’, in Dante’s Italy and other Essays (Philadelphia: University of Pennsylvania Press, 1984), 137–65 (formerly Speculum 40 (1965): 415–35). 5 D. Alighieri, The Divina Commedia and Canzoniere. Translated, with notes, studies and estimates, by E. H. Plumptre, Dean of Wells, 5 vols (London: W. Isbister, 1886–7), II , 419. 6 Dante repeatedly refers to his family in the Commedia, especially during his encounters with his mentor Brunetto Latini (Inf. 15.73–8) and his ancestor Cacciaguida (Par. 15.91–6 and 130–48; 16.1–9 and 34–5). In so doing, he portrays a family history of much nobler origins than the surviving historical documents attest: cf. U. Carpi, La nobiltà di Dante, 2 vols (Florence: Polistampa, 2004). 7 Archivio di Stato di Firenze. Il libro del chiodo. A critical edition with a facsimile, ed. F. Klein (Florence: Polistampa, 2004), 14. The Book of the Nail, so called because of the one surviving nail from its original binding, lists the condemnations of those families considered hostile to the Florentine government during the period 1268–1379. 8 Cf. G. Petrocchi, Vita di Dante (Bari: Laterza, 20085; 1st edition 1983), and the relevant entries in Enciclopedia dantesca, 6 vols (Rome: Istituto dell’Enciclopedia Italiana, 19842; 1st edition 1970–78). 9 On the multiple versions of Boccaccio’s biography see P. G. Ricci, ‘Le tre redazioni del Trattatello in laude di Dante’, Studi sul Boccaccio 8 (1974): 197–214. 10 G. Boccaccio, Trattatello in laude di Dante, ed. P. G. Ricci, in Tutte le opere, General Editor V. Branca, 11 vols (Milan: Mondadori, 1974), III , Chapters 26, 5, 21 and 20 respectively. 11 The bibliography on the Letter to Cangrande is extraordinarily copious: see D. Alighieri, Epistola a Cangrande, ed. E. Cecchini (Florence: Giunti, 1995), xxv–li (the editor leans toward the letter’s authenticity: xxv). The letter was recently examined in a British Academy Italian Lecture (C. Ginzburg, ‘Dante’s Epistle to Cangrande and its Two Authors’, Proceedings of the British Academy 139 (2006): 195–216). Two opposing assessments of the question have been recently provided by A. Casadei, ‘Essential Issues Concerning the Epistle to Cangrande’, in Medieval Letters – Between Fiction and Document, ed. E. Bartoli and C. Høgel (Turnhout: Brepols, 2015), 381–92, and S. Bellomo, ‘L’Epistola a Cangrande, dantesca per intero:

Notes to pp. 98–102

12 13

14 15 16

17

18

19 20

21 22

23

219

“a rischio di procurarci un dispiacere” ’, L’Alighieri 45 (2015): 5–19. The jury is still out on this vexed question. As reported by G. Contini, ‘Preliminari sulla lingua del Petrarca’, in Varianti e altra linguistica (Turin: Einaudi, 1970), 169–92 (175). F. Ricciardelli, ‘Lupi e agnelli nel discorso politico dell’Italia comunale’, in The Languages of Political Society. Western Europe, 14th–17th centuries, ed. A. Gamberini, J.-Ph. Genet and A. Zorzi (Rome: Viella, 2011), 269–85. Boccaccio, Trattatello in laude di Dante, Chapter I. Bruni, ‘Vita di Dante’, in Opere letterarie e politiche, 545. See F. Ricciardelli, ‘Confini e bandi. Azione politica a Firenze in età comunale’, in Images and Words in Exile. Avignon and Italy during the First Half of the 14th Century, ed. E. Brilli, L. Fenelli and G. Wohlfs (Florence: Sismel, 2015), 9–21, with ample bibliography. P. Gualtieri, Il Comune di Firenze tra Due e Trecento. Partecipazione politica e assetto istituzionale (Florence: Olschki, 2009), 243–4; Ricciardelli, ‘Confini e bandi’, passim and 21. Florence, Archivio di stato, Registri dell’Archivio delle Riformagioni, Capitoli, Registri, 19 A, fols 2–3 and 9 recto, and Capitani di parte, Numeri Rossi, 20 (= Libro del Chiodo), pp. 3–5, 14–15. Cf. Klein (ed.), Archivio di stato di Firenze. Il libro del chiodo, Introduction (xxxvii–xxxviii) and Text (169–70 [Document 1.2] and 184–5 [Document 1.9]). G. Milani, ‘Appunti per una riconsiderazione del bando di Dante’, Bollettino di Italianistica 8 (2011): 42–70 (43, 49). The standard, masterly survey remains that of G. Folena, ‘La tradizione delle opere di Dante Alighieri’, in Atti del Congresso internazionale di studi dantescbi (FirenzeVerona-Ravenna, 20–27 aprile 1965), 2 vols (Florence: Sansoni, 1965), I, 1–78, to be supplemented with a number of new acquisitions, some of which are recorded and discussed in this chapter. Francesco da Barberino, Officiolum [facsimile]. Edizione Nazionale dei Commenti danteschi. I commenti figurati, 5 (Rome: Salerno editrice, 2015). Id., Documenti d’Amore, ed. F. Egidi, 4 vols (Rome: Società Filologica Romana, 1905–27), II , 275–6; id., I documenti d’Amore – Documenta Amoris. I. Versi volgari e parafrasi latina. II. Glossae, ed. M. Albertazzi, 2 vols (Lavis, Trento: La Finestra, 2011), II , 143. See also A. Petrucci (with F. Petrucci Nardelli), ‘Minima Barberina. I. Note sugli autografi dei Documenti d’amore’, in Miscellanea di studi in onore di Aurelio Roncaglia, 4 vols (Mucchi: Modena, 1989), III , 1005–9 of 1005–14; C. Panzera, ‘Per l’edizione critica dei Documenti d’Amore di Francesco da Barberino’, Studi mediolatini e volgari 40 (1994): 91–118. G. Indizio, ‘Gli argomenti esterni per la pubblicazione dell’Inferno e del Purgatorio’, Studi danteschi 68 (2003): 17–47.

220

Notes to pp. 102–104

24 M. Saccenti, ‘Memoriali bolognesi’, in Enciclopedia dantesca, III , 892–4. 25 Interestingly, Dante placed Loderingo and Catalano in Hell among the Hypocrites (Inf. 23.91–2). Dante condemned them because they sided with the Pope when acting as peacekeepers in Florence in 1266. 26 M. Volpi, ‘Dai lemmi del commento verso il perduto Dante del Lana’, in Nuove prospettive sulla tradizione della ‘Commedia’. Seconda serie (2008–2013), ed. E. Tonello and P. Trovato (Padua: Libreria universitaria, 2013), 47–70. 27 Piacenza, Biblioteca Comunale Passerini-Landi, MS 190. Cf. Facsimile del Codice Landiano. Preface by A. Balsamo. Introduction by G. Bertoni (Florence: Olschki, 1921). 28 Martini’s collation exemplar is a copy of the Commedia (Venice: Aldo Manuzio, 1515), now preserved in Milan, Biblioteca Nazionale Braidense, Aldina AP XVI 25. Cf. D. Alighieri, La Commedia secondo l’antica vulgata, ed. G. Petrocchi, 4 vols (Milan: Mondadori, 1966), I, 76–8; O. Schiavone, ‘Luca Martini filologo dantesco: collazioni, annotazioni e committenze (1543–51)’, in La filologia in Italia nel Rinascimento, ed. C. Caruso and E. Russo (Rome: Edizioni di Storia e Letteratura, 2018), 117–32. On the presence in Florence of the lost manuscript about 1330–1 see G. Inglese, ‘Il codice Alighieri e lo scrittoio del Pievano’, Studi e problemi di critica testuale 78 (2009): 9–11. 29 Il Codice Trivulziano 1080 della Divina Commedia [facsimile]. Cenni storici e descrittivi di L. Rocca (Milan: Hoepli, 1921); Dante, La Commedia secondo l’antica vulgata, ed. Petrocchi, I, 85–6. 30 Boccaccio editore e interprete di Dante, ed. L. Azzetta and A. Mazzucchi (Rome: Salerno Editrice, 2014). 31 M. Roddewig, Dante Alighieri. Die göttliche Komödie. Vergleichende Bestandsaufnahme der Commedia-Handschriften (Stuttgart: Anton Hiersemann Verlag, 1984). 32 Le terze rime di Dante (Venice: Aldo Manuzio, 1502). 33 La Divina Commedia di Dante Allighieri, 3 vols (Parma: Nel Regal Palazzo, 1795). 34 La Divina Commedia di Dante Allighieri ricorretta sopra quattro dei più autorevoli testi a penna da Carlo Witte (Berlin: Decker, 1862). 35 Tutte le opere di Dante Alighieri, rivedute nel testo dal Dr. E. Moore (Oxford: Nella stamperia dell’Università, 1894). 36 On Bembo and sixteenth-century editorial practice see E. Tonello, ‘L’edizione bembina della Commedia’, L’Alighieri 57 (2016): 113–35 (where the misprint ‘Vat. Lat. 3196’ must be read as ‘Vat. Lat. 3197’), and A. E. Mecca, ‘La tradizione a stampa della Commedia: dall’Aldina del Bembo (1502) all’edizione della Crusca (1595)’, Nuova rivista di letteratura italiana 16 (2013): 9–59; on Dionisi, L. Mazzoni, Dante a Verona nel Settecento: studi su Giovanni Iacopo Dionisi (Verona: QuiEdit, 2012); on Witte, G. Folena, ‘La filologia dantesca di Carlo Witte’, in Dante e la cultura tedesca,

Notes to pp. 104–108

37 38

39

40 41

42

43 44 45 46 47

48

221

ed. L. Lazzarini (Padua: Tipi dell’Antoniana, 1967), 109–39 (subsequently in G. Folena, Filologia e umanità (Vicenza: Neri Pozza, 1993), 25–52); on Moore, J. Lindon, ‘Gli apporti del metodo di Edward Moore nei primi decenni della Società Dantesca Italiana’, in La Società Dantesca Italiana 1888–1988, ed. R. Abardo (Milan and Naples: Ricciardi, 1995), 37–53. La Divina Commedia di Dante Alighieri nobile fiorentino ridotta a miglior lezione dagli Accademici della Crusca (Florence: Domenico Manzani, 1595). Cf. Folena, ‘La tradizione delle opere di Dante Alighieri’; P. Trovato, Everything You Always Wanted to Know about Lachmann’s Method (Padua: Libreria universitaria, 2014), 299–314. P. Maas, ‘Textkritik’, in Einleitung in die Altertumswissenschaft, ed. A. Gercke and E. Norden, 2 vols (Leipzig and Berlin: Teubner, 19273), II , 13 (English translation: Textual Criticism, trans. B. Flower (Oxford: Clarendon Press, 1958), 61–2). Maas is rephrasing here Contra vim mortis non crescit herba in hortis (‘No garden herb can contrast the power of death’). Trovato, Everything You Always Wanted to Know, 253–65, offers a very instructive selection of examples. There are exceptions. MS Laur. 26 sin. 1, for example, shows an extraordinarily sophisticated punctuation and notation system, on which see G. Tanturli, ‘Filologia del volgare intorno al Salutati’, in Coluccio Salutati e l’invenzione dell’Umanesimo, ed. C. Bianca (Rome: Edizioni di Storia e Letteratura, 2010), 83–144. B. Nardi, Nel mondo di Dante (Rome: Edizioni di Storia e Letteratura, 1944), 375–6. A thorough discussion of these readings is in Trovato, Everything You Always Wanted to Know, 151–4, with important observations on their distribution in the tradition of the Commedia which may prove the existence of an archetype (an aspect I cannot consider here). Cf. A. Quondam, ‘coronare’, in Enciclopedia dantesca, II , 213–4. Iacomo della Lana, Commento alla ‘Commedia’, ed. M. Volpi with the collaboration of A. Terzi, 3 vols (Rome: Salerno editrice, 2009), II , 1504. D. Alighieri, Commedia, ed. A. M. Chiavacci Leonardi, 3 vols (Milan: Mondadori, 1994), II , 817. E. Bigi, ‘Genesi di un concetto storiografico: Dolce stil novo’, Giornale storico della letteratura italiana 132 (1955): 333–71. Comentum super poema Comedie Dantis. A critical edition of the third and final draft of Pietro Alighieri’s Commentary on Dante’s ‘The Divine Comedy’, ed. M. Chiamenti (Tempe, AR : Center for Medieval and Renaissance Studies, 2002), 521. The same explanation is offered in the so-called ‘Cassino glosses’, which often reproduce the substance of Pietro’s commentary: cf. Il Codice Cassinese della ‘Divina Commedia’, per la prima volta letteralmente messo a stampa per cura

222

49

50

51 52

53 54 55

56

57 58

59

60

Notes to pp. 108–111

dei monaci benedettini della Badia di Monte Cassino (s.l.: Tipografia di Monte Cassino, 1865), 312. A. Cipollone, ‘I quattro sensi della scrittura di Bonagiunta. Ancora sulla tenzone con Guinizzelli’, in Intorno a Guido Guinizzelli. Atti della Giornata di studi, Università di Zurigo, 16 giugno 2000, ed. L. Rossi and S. Alloatti-Boller (Alessandria: Ed. dell’Orso, 2002), 99–135. Dantis Alagherii Comedia, ed. F. Sanguineti (Florence: Sismel-Edizioni del Galluzzo, 2001), ad locum. In the subsequent and complementary Appendice bibliografica 1988–2000 (Florence: Sismel-Edizioni del Galluzzo, 2005), xxi, Sanguineti marked the verse with a crux desperationis. The importance of MS Vat. Urb. 366 had been previously noticed by Petrocchi (cf. Introduction to Dante, La Commedia secondo l’antica vulgata, I, 87–9). See M. Zaccarello, ‘Canto XXIV. Rimatori e poetiche “da l’uno a l’altro stilo” ’, in Cento canti per cento anni. II . Purgatorio, ed. E. Malato and A. Mazzucchi (Rome: Salerno Editrice, 2014), 712–44. Cf. A. Dardi, review of Comentum super poema Comedie Dantis, ed. Chiamenti, in Lingua nostra 68 (2006), 126. See for all M. Boschi Rotiroti, Codicologia trecentesca della ‘Commedia’. Entro e oltre l’antica vulgata (Rome: Viella, 2004). A very balanced and informative survey of the critical debate following Petrocchi’s edition, rich with valuable suggestions and insights, is provided by A. Canova, ‘Il testo della Commedia dopo l’edizione Petrocchi’, in Nuove prospettive sulla tradizione della ‘Commedia’. Seconda serie, 29–45. J. Alighieri, Chiose all’Inferno, ed. S. Bellomo (Padua: Antenore, 1990); Iacomo della Lana, Commento alla ‘Commedia’; G. Bambaglioli, Commento all’Inferno di Dante, ed. L. C. Rossi (Pisa: Scuola Normale Superiore, 1998). Dantis Alagherii Comedia, ed. Sanguineti, lxv–lxix. G. Pasquali, Storia della tradizione e critica del testo (Florence: Le Lettere, 1988; 1st edition 1934), Chapter IV ‘Recentiores non deteriores’, 41–108; S. Timpanaro, La genesi del metodo del Lachmann (Turin: UTET, 2004; 1st edition 1963; English trans. Chicago: University of Chicago Press, 2006). Cf. Canova, ‘Il testo della Commedia’, passim. An important critical survey in Trovato, Everything You Always Wanted to Know, pp. 299–333, where a number of contributions are discussed. G. Inglese’s ‘Per lo stemma della Commedia dantesca. Tentativo di statistica degli errori significativi’, Filologia italiana 4 (2007): 51–72, is arguably the most important contribution to the debate which resists the innovations proposed by Sanguineti. W. Benjamin, review of G. Salles, Le regard. La collection, le musée, la fouille, une journée, l’école (Paris: Librairie Plon, 1939), in id., Gesammelte Schriften, 8 vols (Frankfurt a. M.: Suhrkamp, 1972–89), III , 589 (‘Il y a peut-être deux sortes de livres: ceux qui montrent et ceux qui donnent’).

Notes to pp. 112–114

223

61 Cf. Trovato, Everything You Always Wanted to Know, esp. 326–33. Disagreement on Trovato’s classification of manuscripts is, however, expressed in F. Sanguineti, ‘ “Novissime prospettive” dantesche’, L’Alighieri 43 NS (2014): 107–12. Sanguineti and Eleonisia Mandola have recently proposed a new solution for the poem’s linguistic colouring: Dante, Paradiso I–XVII. Edizione critica alla luce del più antico codice di sicura fiorentinità, ed. E. Mandola, with a ‘Premessa’ by F. Sanguineti (Genoa: Il melangolo, 2018). On the trans-Apenninic tradition of the Commedia see also Nuove prospettive sulla tradizione della Commedia. Una guida filologicolinguistica al poema dantesco, ed. P. Trovato (Florence: Cesati, 2007); P. Pellegrini, Dante tra Romagna e Lombardia: studi di linguistica e di filologia italiana (Padua: Libreria universitaria, 2016). 62 On the stress on the ‘i’ of comedìa, demanded in both cases by prosodic rhythm, see H. A. Kelly, Tragedy and Comedy from Dante to Pseudo-Dante (Berkeley, Los Angeles and London: University of California Press, 1989), 10. English translations tend to render comedìa with the initial letter capitalized as a proper title. 63 ‘Per tragoediam superiorem stylum induimus, per comoediam inferiorem, per elegiam stylum intelligimus miserorum.’ 64 ‘Comedia vero inchoat asperitatem alicuius rei, sed eius materia prospere terminatur’. A. Casadei, who denies the authenticity of the Letter to Cangrande, suggests that Commedia may not be Dante’s own title after all: ‘Il titolo della Commedia e l’Epistola a Cangrande’, in Dante oltre la ‘Commedia’ (Bologna: Il Mulino, 2013), 14–44. 65 Cf. E. J. Webber, ‘Comedy as Satire in Hispano-Arabic Spain’, Hispanic Review 26 (1958): 1–11. 66 G. Brugnoli and R. Mercuri, ‘Orazio’, in Enciclopedia dantesca, IV, 173–80, esp. 179–80; S. Reynolds, ‘Orazio satiro (Inferno IV, 89): Dante, the Roman satirists, and the medieval theory of satire’, and ‘Dante and the medieval theory of satire: A collection of texts’, in ‘Libri poetarum in quattuor species dividuntur’. Essays on Dante and ‘genre’. Supplement to The Italianist 15 (1995), ed. Z. G. Barański, 128–44 and 145–57. 67 Aristoteles Latinus XXXIII . De arte poetica. Translatio Guillelmi de Moerbeka, ed. L. Minio-Paluello (Bruxelles and Paris: Desclée de Brouwer, 1968), xiv–xvi: MSS Eton College 129 (Italy, end of the thirteenth century); Toledo, Bibl. Capitular, 47.10 (Viterbo?, c. 1280). In 1978 Giorgio Agamben made the suggestion that Dante might have had cognizance of Moerbeke’s translation: G. Agamben, ‘Comedìa: la svolta comica di Dante e la concezione della colpa’, Paragone 29, no. 346 (December 1978): 3–27; English translation ‘Comedy’, in id., The End of the Poem: Studies in Poetics, trans. D. Heller-Roazen (Stanford: Stanford University Press, 1999), 1–22 and 135–42: see esp. 9 and 139. 68 Aristoteles Latinus, De arte poetica, ed. Minio-Paluello, 16–17. 69 Id., 5.

224

Notes to pp. 115–132

5 Textual Metamorphosis: The Manuscripts of Leonardo da Vinci 1 The texts and manuscripts of Leonardo are accessible in the critical editions supplied by the Commissione Vinciana and from the portal e-Leo promoted by Romano Nanni at the Biblioteca Leonardiana in Vinci (www.leonardodigitale.com). On Leonardo’s life and works see C. Vecce, Leonardo (Rome: Salerno Editrice, 2006). On the Book of Painting and its reception, see C. Farago, J. Bell and C. Vecce, The Fabrication of Leonardo da Vinci’s ‘Trattato della pittura’, with a scholarly edition of the ‘editio princeps’ (1651) and an annotated English translation (Leiden: Brill, 2018). 2 On Leonardo’s manuscripts and related textual questions: G. Calvi, I manoscritti di Leonardo da Vinci dal punto di vista cronologico storico e biografico, ed. A. Marinoni (Busto Arsizio: Bramante Editrice, 1982; first edition Bologna: Zanichelli, 1925); C. Dionisotti, ‘Leonardo uomo di lettere’, Italia medioevale e umanistica 5 (1962): 183–216; C. Pedretti, Leonardo da Vinci on Painting. A Lost Book (Libro A) Reassembled from the Codex Vaticanus Urbinas 1270 and from the Codex Leicester (Berkeley and Los Angeles, CA: University of California Press, 1964); E. H. Gombrich, ‘Leonardo’s Method for Working out Composition’, in id., Norm and Form (London: Phaidon, 1966), 58–63; A. Corbeau, Les manuscrits de Léonard de Vinci (Caen: Centre régional de documentation pédagogique, 1968); A. Marinoni, ‘L’eredità letteraria’, in Leonardo, ed. L. Reti (Milan: Mondadori, 1974), 56–85; C. Pedretti, ‘Eccetera: perché la minestra si fredda’ (Florence: Giunti Barbèra, 1975); E. H. Gombrich, ‘Leonardo da Vinci’s Method of Analysis and Permutation: the Grotesque Heads’, in id., The Heritage of Apelles (London: Phaidon, 1976), 57–75; C. Pedretti, The Literary Works of Leonardo da Vinci. A Commentary to Jean Paul Richter’s Edition (Oxford: Oxford University Press, 1977); C. Vecce, ‘Scritti di Leonardo da Vinci’, in Letteratura italiana. Le opere, ed. A. Asor Rosa, (Turin: Einaudi, 1993), II, 95–124; R. Zwijnenberg, The Writings and Drawings of Leonardo da Vinci. Order and Chaos in Early Modern Thought (Cambridge: Cambridge University Press, 1999); C. Vecce, ‘Parola e immagine nei manoscritti di Leonardo’, in Percorsi tra parole e immagini (1400–1600), ed. A. Guidotti and M. Rossi (Lucca: Maria Pacini Fazzi editore, 2000), 19–35; id., ‘Word and image in Leonardo’s writings’, in Leonardo da Vinci Master Draftsman, ed. C. C. Bambach (New Haven, CT and London: Yale University Press, 2003), 59–77; C. C. Bambach, Un’eredità difficile: i disegni e i manoscritti di Leonardo tra mito e documento (Florence: Giunti, 2009); C. Vecce, ‘Scrittura, creazione, lavoro intellettuale alla fine del Quattrocento’, in ‘Di mano propria’. Gli autografi dei letterati italiani, ed. G. Baldassarri, M. Motolese, P. Procaccioli and E. Russo (Rome: Salerno Editrice, 2010), 211–40; id., La biblioteca perduta. I libri di Leonardo (Rome: Salerno Editrice, 2017); id., ‘The Fading Evidence of Reality: Leonardo and the End’, Insights 10 (2017): 2–10.

Notes to pp. 133–137

225

6 Montaigne: The Life and After-Life of an Unfinished Text 1 All the editions mentioned here can be accessed most easily on the Bibliothèques Virtuelles Humanistes (BVH , Tours) site: https://montaigne.univ-tours.fr/category/ oeuvres/fac-similes/. The 1595 edition is also available in page-by-page form at the Chicago Montaigne Project: https://www.lib.uchicago.edu/efts/ARTFL/projects/ montaigne. 2 Notably A. Boase, The Fortunes of Montaigne: A History of the ‘Essays’ in France, 1580–1669 (London: Methuen, 1935) and O. Millet, La première réception des ‘Essais’ de Montaigne, 1580–1640 (Paris: Champion, 1995). 3 The Complete Essays of Montaigne, trans. D. Frame (Stanford: Stanford University Press, 1958), 170; Les Essais de Michel de Montaigne, ed. P. Villey and V.-L. Saulnier (Paris: PUF, 1965), 231. References will henceforth be included in the text using the letters F and V, followed by the page number. Frame and Villey/Saulnier remain the most commonly used English translation and French edition; the merits of Villey/ Saulnier are discussed below. 4 The exemplaire de Bordeaux is available at Gallica: http://gallica.bnf.fr/ark:/12148/ bpt6k11718168/f1.item.r= or in page-by-page form at the Chicago Montaigne Project: https://www.lib.uchicago.edu/efts/ARTFL/projects/montaigne. Some of the material that follows in this section is drawn from J. O’Brien, ‘Are We Reading What Montaigne Wrote?’, French Studies 58, no. 4 (2004): 527–32, reproduced by the kind permission of the General Editor of French Studies. 5 Les ‘Essais’ de Michel de Montaigne, publiés d’après l’Exemplaire de Bordeaux, avec les variantes manuscrites & les leçons des plus anciennes impressions, ed. F. Strowski, F. Gébelin and P. Villey, 5 vols (Bordeaux: Pech, 1906–33). 6 Coste’s edition dominated eighteenth-century printings of the Essais. See further T. Sankovitch, ‘ “Un Travail vétilleux . . . fort nécessaire”: The Coste Edition of 1724’, Montaigne Studies 7 (1995): 131–46, and W. Boutcher, The School of Montaigne in Early Modern Europe 2 vols (Oxford: Oxford University Press, 2017), II , 368–71. 7 See A. Tournon, ‘Un langage coupé. . .’, in Writing the Renaissance: Essays on Sixteenth-Century French Literature in Honor of Floyd Gray, ed. R. C. La Charité (Lexington, KY: French Forum Publishers, 1992), 219–31; id., ‘Syntaxe et scansion: l’énergie du “langage coupé” et la censure éditoriale’, in Montaigne et la rhétorique, ed. J. O’Brien, M. Quainton and J. Supple (Paris: Champion, 1995), 117–33; id., ‘La segmentation du texte: usages et singularités’, in Éditer les ‘Essais’ de Montaigne, ed. C. Blum and A. Tournon (Paris: Champion, 1997), 173–83; id., ‘L’inquiétante segmentation des Essais’, Le Discours psychanalytique 18 (1997): 277–306; id., ‘ “Ny de la ponctuation”: sur quelques avatars de la segmentation autographe des Essais’, Nouvelle revue du seizième siècle 17, no. 1 (1999): 147–59.

226

Notes to pp. 139–142

8 ‘Essais’ de Michel de Montaigne, ed. A. Tournon, 3 vols (Paris: Imprimerie Nationale, 1998); Essais de Michel de Montaigne, ed. E. Naya, D. Reguig-Naya and A. Tarrête, Folio classique, 3 vols (Paris: Gallimard, 2009). 9 Montaigne, Les Essais, ed. D. Bjaï, B. Boudou, J. Céard and I. Pantin, La Pochothèque (Paris: Librairie Générale Française, 2001); Montaigne. Les Essais, ed. J. Balsamo, M. Magnien and C. Magnien-Simonin, Bibliothèque de la Pléiade (Paris: Gallimard, 2007). 10 For her work on this front, see Ph. Desan, ‘Marie de Gournay et le travail éditorial des Essais de 1595 à 1635’, Journal of Medieval and Renaissance Studies 25 (1995): 363–80; Marie de Gournay et l’édition de 1595 des ‘Essais’ de Montaigne, ed. J.-C. Arnould (Paris: Champion, 1996); Montaigne et Marie de Gournay, ed. M. Tetel (Paris: Champion, 1997). Claude Blum also offers perspicacious remarks about Gournay’s editions and her role as an editor in Marie de Gournay, Œuvres complètes, ed. J.-C. Arnould and others, 2 vols (Paris: Champion, 2002), I, 27–43. 11 M. Simonin, ‘Aux origines de l’édition de 1595’, in L’Encre et la lumière (Geneva: Droz, 2004), 523–50 (previously in Arnould et al., Montaigne et Marie de Gournay, 7–51); and id., ‘Montaigne, son éditeur et le correcteur devant l’exemplaire de Bordeaux des Essais’, Travaux de littérature 11 (1998): 75–93. The judicious appraisals of the 1595 edition by Sayce and Maskell are to be found in R. A. Sayce, ‘L’Édition des Essais de Montaigne de 1595’, Bibliothèque d’humanisme et Renaissance 36 (1974): 115–41; D. Maskell, ‘Quel est le dernier état authentique des Essais de Montaigne?’, Bibliothèque d’humanisme et Renaissance 40 (1978): 85–103; id., ‘Montaigne correcteur de l’exemplaire de Bordeaux’, Bulletin de la Société des Amis de Montaigne, Ve série, no. 25–6 (January–June 1978): 57–71. 12 See the descriptions of the different states in R. A. Sayce and D. Maskell, A Descriptive Bibliography of Montaigne’s ‘Essais’, 1580–1700 (London: The Bibliographical Society and MHRA , 1983), nos. 7A, 7A† and 7B, 25–35. 13 ‘Exemplar’ is Philippe Desan’s term for the printer’s copy: see his ‘L’Exemplar et l’exemplaire de Bordeaux’, in Montaigne dans tous ses états (Fasano: Schena Editore, 2001), 69–120, and his Reproduction en quadrichrome de l’exemplaire avec notes manuscrites marginales des Essais de Montaigne (Exemplaire de Bordeaux) (FasanoChicago: Schena Editore and Montaigne Studies, 2002), xiii–xix. 14 EB , fols 42v, 47r and 290v. 15 See Alain Legros, ‘Essais de 1588 et exemplaire de Bordeaux’, on the website MONLOE , https://montaigne.univ-tours.fr/essais-1588-exemplaire-bordeaux (accessed 16 July 2017): ‘Montaigne set down the first word and Gournay, under his direction, wrote the rest’ (‘c’est bien Montaigne qui a placé le premier mot et Gournay, sous son contrôle, a écrit la suite’). 16 Simonin, ‘Aux origines’, 531, 543. 17 Maynooth University, Russell Library, RB Case 122. The edition is L’Angelier’s, first issue, on which see Sayce and Maskell, Descriptive Bibliography, no. 7A, 25–9. For this

Notes to pp. 142–149

18 19

20 21

22 23 24

25 26 27 28 29 30

31

227

particular discovery, see J. O’Brien, ‘Gournay’s Gift: A Special Presentation Copy of the 1595 Essais of Montaigne’, The Seventeenth Century 29, no. 4 (2014): 317–36. I am grateful to the Editor of The Seventeenth Century for permission to reproduce some of the material here. The English translation here is my adaptation of Frame. The difficulties with seeing the interpolation as a fabrication of Gournay’s were long ago highlighted by J. Zeitlin, ‘The Relation of the Text of 1595 to that of the Bordeaux Copy’, in The Essays of Montaigne, trans. J. Zeitlin, 3 vols (New York: Knopf, 1934), I, 421–2, 430. The editors of the 2007 Pléiade Essais comment that the stylistic coherence of the passage and what we know of Gournay’s relations with Montaigne belie the view that she added the praise section out of vanity: see Balsamo et al. (eds), Les ‘Essais’, 1654, note 2 to 703. See Legros, ‘Essais de 1588 et exemplaire de Bordeaux’, discussing EB , fol. 290v. See Balsamo, ‘Le Destin éditorial des Essais’ in Balsamo et al. (eds), Les ‘Essais’, xxxii–lv. Among other changes made on the exemplar or during the print run, one can also cite a famous passage in chapter 2.8 in which Montaigne expresses his grief over the loss of his friend, La Boétie, beginning with the interjection, ‘O my friend’ (‘O mon amy’). The passage does not appear on EB (fol. 163v), but does appear in 1595 (253). Les Essais de Michel Seigneur de Montaigne (Paris: Hulpeau, 1625), 587–8. Les Essais de Michel Seigneur de Montaigne (Paris: Rocolet and Bray, 1635), 517. Montaigne, Les Essais (1625), sig. [¶¶¶¶¶¶¶ ijr]; with one change (‘par cette voye’ substituted for ‘si Dieu prolonge mes années’), the same passage is also printed in Montaigne, Les Essais (1635), sig. ¶¶¶ ijr. My translation. As confirmed by Sayce and Maskell, A Descriptive Bibliography, 103, observation 6. M. McKinley, ‘An Editorial Revival: Gournay’s 1617 Preface to the Essais’, Montaigne Studies 8 (1996): 193–201 (200). Montaigne, Les Essais (1625), sig. ũv. McKinley, ‘An Editorial Revival’, 201. On this courtier, see É. and E. Haag, La France protestante, 6 vols (Paris and Geneva: Cherbuliez, 1846–59), II (1847), 195–6; J. Chavannes, ‘Madame des Loges et la famille de Beringhen. Une lettre de Madame des Loges à Henri de Beringhen son neveu (1628?)’, Bulletin de la Société de l’histoire du protestantisme 9 (1860): 84–96. See Gournay, Œuvres complètes, I, 1081, note C. On Marie des Loges, see Haag, La France protestante, III (1852), 47–8. On the salon itself, see L. Timmermans, L’Accès des femmes à la culture (1598–1715): Un débat d’idées de saint François de Sales à la marquise de Lambert (Paris: Champion, 1993), 77–84; J. Pannier, L’Église réformée de Paris sous Louis XIII de 1610 à 1621, 3 vols (Strasbourg: Istra; Paris: Champion, 1922–32), I, 338–43; II , 341–7; III , 51–65.

228

Notes to pp. 150–154

32 On Besold, see B. Zeller-Lorenz and W. Zeller, ‘Christoph Besold, 1577–1638. Polyhistor, gefragter Consiliator und umstrittener Konvertit’, in Lebensbilder zur Geschichte der Tübinger Juristenfakultät, ed. F. Elsner (Tübingen: Mohr, 1977), 9–18; B. Pahlmann, ‘Christoph Besold’, in Deutsche und Europäische Juristen aus neun Jahrhunderten. Eine biographische Einführung in die Geschichte der Rechtswissenschaft, ed. J. Schröder and G. Kleinheyer, 4th edition (Heidelberg: Müller, 1996), 56–9. 33 These quotations are from, respectively: De Consilio politico axiomata aliquámmulta (Tübingen: Wild, 1622), 107; De Verae Philosophiae fundamento (Tübingen: Brunnius, 1630), 24 and 19; De Nuptiis iuridico-politicus discursus (Tübingen: Cellius, 1621), 6. For more on Besold and his use of Montaigne, see J. O’Brien, ‘Montaigne beyond the Rhine: The Essais in the Work of Christoph Besold (1577–1638)’, in Montaigne in Transit: Essays in Honour of Ian Maclean, ed. N. Kenny, R. Scholar and W. Williams (Oxford: Legenda, 2017), 171–86. 34 C. Besold, Principium et finis politicæ doctrinæ (Strasbourg: Heirs of Zetzner, 1625), preface, sig. A 2v. The English translation is mine. 35 Besold must have purchased his copy of Montaigne before 1615, for he referred to his copy by page number in the first edition of his work De Aerario publico in that year. 1608 and 1611 are the only two editions before 1615 which would fit the page numbers he gives. For the bibliographical details, see Sayce and Maskell, A Descriptive Bibliography, 54–9 (1608) and 62–7 (1611). 36 For the details, see O’Brien, ‘Montaigne beyond the Rhine’, 173. 37 See further J. O’Brien, ‘Montaigne, Sir Ralph Bankes and other English Readers of the Essais’, Renaissance Studies 28, no. 3 (2014): 377–91. 38 Les Essais de Michel Seigneur de Montaigne (Paris: Bechet and Billaine, 1657). Kingston Lacy shelfmark: M.1.1. 39 For case studies of the English aristocratic dimension, see Boutcher, The School of Montaigne, II , 219–31.

7 Rescuing Shakespeare: King Lear in Its Textual Contexts 1 Ed. Stanley Wells, on the basis of a text prepared by Gary Taylor (Oxford: Oxford University Press, 2000); ed. Jay L. Halio (Cambridge: Cambridge University Press, 1992; 2nd edition, 2005); ed. R. A. Foakes (Walton-on-Thames: Nelson, 1997 = The Arden Shakespeare, third series). Important modern editions of the consolidated text (F and Q combined) include those of Kenneth Muir (London: Methuen, 1952, revised 1972 = Arden 2) and G. I. Duthie and John Dover Wilson (Cambridge: Cambridge University Press, 1960 = The New Cambridge Shakespeare) and G. K. Hunter (Harmondsworth: Penguin, 1972, revised 1996; with an introduction by Kiernan Ryan, 2005). 2 In relation to King Lear the fundamental issues are drawn together in The Division of the Kingdoms: Shakespeare’s Two Versions of ‘King Lear’, ed. Gary Taylor and Michael

Notes to pp. 154–158

3

4

5

6

229

Warren (Oxford: Clarendon, 1983). For a recent re-statement of the traditional view, which takes account of and rebuts the reviser-theory, see Sir Brian Vickers, The One ‘King Lear’ (Cambridge, MA : Harvard University Press, 2016). A selection of the most significant earlier responses to Division defending the practice of combining Q and F includes: Sidney Thomas, ‘Shakespeare’s Supposed Revisions of King Lear’, Shakespeare Quarterly 35 (1984): 506–11; T. H. Howard-Hill, ‘The Challenge of King Lear’, The Library, 6th series, 7 (1985): 161–79; Richard Knowles, ‘The Case for two Lears’, Shakespeare Quarterly 36 (1985): 115–20; Kenneth Muir, ‘The Texts of King Lear: An Interim Assessment of the Controversy’, in Shakespeare: Contrasts and Controversies (Norman: University of Oklahoma Press, 1985), 51–66; Marion Trousdale, ‘A Trip through the Divided Kingdoms’, Shakespeare Quarterly 37 (1986): 218–23; Walter C. Carroll, ‘New Plays vs. Old Readings: The Division of the Kingdoms and Folio Deletions in King Lear’, Studies in Philology 85 (1988): 225–44; Anne Meyer, ‘Shakespeare’s Art and the Texts of King Lear’, Studies in Bibliography 47 (1994): 128–46; and Robert Clare, ‘ “Who is it who can tell me who I am”: The Theory of Authorial Revision between the Quarto and Folio Texts of King Lear’, The Library, 6th series, 17 (1995): 34–59. ‘King Lear’ from Study to Stage: Essays in Criticism, ed. James Ogden and Arthur H. Scouten (Madison: Fairleigh Dickinson University Press, 1997), contains essays on the issue from different perspectives. Cf. two critical reviews of a significant predecessor to Division, Steven Urkowitz, Shakespeare’s Revision of ‘King Lear’ (Princeton, NJ : Princeton University Press, 1980): Philip Edwards, The Modern Language Review 77 (1982): 694–8; Richard Knowles, Modern Philology 79 (1981): 197–200. For a study of the twenty or so manuscripts of early modern English drama extant from the c. 3,000 plays written in the period and how these bear on the editing of Shakespeare, see James Purkis, Shakespeare and Manuscript Drama: Canon, Collaboration and Text (Cambridge: Cambridge University Press, 2016). W.W. Greg’s categories of ‘foul papers’ and ‘prompt books’ have recently been challenged by Paul Werstine, Early Modern Playhouse Manuscripts and the Editing of Shakespeare (Cambridge: Cambridge University Press, 2012), with potential implications for editing which (in part because of the limitations of the documentary evidence) are difficult to estimate. ‘Hamlet’: The Texts of 1603 and 1623, ed. Ann Thompson and Neil Taylor, The Arden Shakespeare, third series (London: Thomson Learning, 2006), scene 7, ll. 115–25. For a recent attempt to argue that Q1 represents Shakespeare’s first (1589) version of the play, see Terri Bourus, Young Shakespeare’s Young Hamlet, History of Text Technologies (Houndmills: Palgrave Macmillan, 2014). A. W. Pollard, Shakespeare Folios and Quartos: A Study in the Bibliography of Shakespeare’s Plays 1594–1685 (London: Methuen, 1909), chapter III . Greg’s views are drawn together in W. W. Greg, The Editorial Problem in Shakespeare: A Survey of the Foundations of the Text (Oxford: Clarendon, 1942; 3rd edition, 1954).

230

Notes to pp. 159–160

7 R. E. Burkhart, Shakespeare’s Bad Quartos: Deliberate Abridgements Designed for Performance by a Small Cast (The Hague: Mouton, 1975). The issue of the ‘bad’ quartos and theatrical abridgement is also discussed by David Bradley, From Text to Performance in the Elizabethan Theatre: Preparing the Play for the Stage (Cambridge: Cambridge University Press, 1992). 8 On the issue of how published texts may relate to texts as performed see Andrew Gurr, ‘Maximal and Minimal Texts: Shakespeare v. the Globe’, Shakespeare Survey 52 (1999): 68–87; and (arguing for a more extended performance time that could have accommodated even the complete Hamlet) Michael J. Hirrel, ‘Duration of Performances and Lengths of Plays: How Shall We Beguile the Lazy Time?’, Shakespeare Quarterly 61 (2010): 159–82. On Shakespeare as writing for readers as well as for play-goers (and on the ‘bad’ quartos as texts abbreviated for performance), see Lukas Erne, Shakespeare as Literary Dramatist (Cambridge: Cambridge University Press, 2003; 2nd edition, 2013). 9 Laurie E. Maguire, Shakespearean Suspect Texts: The ‘Bad’ Quartos and Their Contexts (Cambridge: Cambridge University Press, 1996), 323. Critique of the Pollard–Greg view of ‘bad’ quartos, initiated in the late 1970s, was also given influential expression in Paul Werstine, ‘Narratives about Printed Shakespeare Texts: “Foul Papers” and “Bad” Quartos’, Shakespeare Quarterly 41 (1990): 65–86. 10 See Henry VI, Part Two, ed. Roger Warren (Oxford: Oxford University Press, 2002), 78–87; and Henry VI, Part Three, ed. Randall Martin (Oxford: Oxford University Press, 2008), 106–13. 11 Ben Jonson, ed. Ian Donaldson, The Oxford Authors (Oxford: Oxford University Press, 1985), 539 (Discoveries) and 455 (‘To the Memory of My Beloved . . . Mr William Shakespeare’). 12 Purkis (Shakespeare and Manuscript Drama, 191–246) encourages a tentative scepticism about the otherwise widely accepted identification of Hand D as Shakespeare’s. 13 References to Shakespeare (including to the consolidated King Lear) are to The Riverside Shakespeare, ed. G. Blakemore Evans (Boston: Houghton Mifflin, 1974). Quotations from the Q1 and F texts of King Lear (in modernized spelling and punctuation) are taken from ‘King Lear’: A Parallel Text Edition, ed. René Weis (London: Longman, 1993; 2nd edition, 2010). For these texts in old spelling see The Complete ‘King Lear’, 1608–1623, ed. Michael Warren (Berkeley : University of California Press, 1989): parallel texts of Q1 and F, and photographic facsimiles of Q1, Q2 and F. 14 The theory that Shakespeare may have revised some of his plays has a history which begins with the editing of his works from variant texts in the early eighteenth century: see Grace Ioppolo, Revising Shakespeare (Cambridge, MA : Harvard University Press, 1991), chapter 1.

Notes to pp. 160–165

231

15 In keeping with its fundamental policy of ‘single-text editing’ – that is, editing the text, not the work – the latest Norton Complete Shakespeare (ed. Stephen Greenblatt et al., 3rd edition, 2016) offers each text separately, printing Richard III based on Q1, and including the F text among its digital additions. Contrary to this, and thereby writing textual debate into its practice, the Norton edition does, however, include what it calls a ‘hybrid’ Hamlet, a text based on Q2 but incorporating (in a different typeface) passages found only in F. 16 Harold Jenkins (Arden 2, 1982) assumed that these represented a single work, and combined Q2 and F; Ann Thompson and Neil Taylor (Arden 3, 2006) print Q2 and F in separate volumes (F with Q1, 1603), while the main edition (of Q2) includes F-only text in an appendix. 17 There are also eight short passages (about 160 lines in all) in the Folio text of 2 Henry IV that do not appear in the quarto, but there is no agreement on the source of these differences. Multiple inconsistencies in Timon of Athens suggest that it may have been left incomplete, in a final stage of revision (perhaps not entirely by Shakespeare); but there is only one text (F). 18 See note 10. 19 The main views are summarized by Stanley Wells, ‘The Once and Future King Lear’, in Division, 1–22. Central to the reviser theory has been the view of Peter W. M. Blayney, The Texts of ‘King Lear’ and Their Origins. Vol. 1: Nicholas Okes and the First Quarto (Cambridge: Cambridge University Press, 1982), who argues that Q1 is based on an authorial manuscript with an authority distinct from that which lies behind F; and that the poor printing (verse printed as prose, prose as verse, mislineation, erratic and misleading punctuation) result from the printer’s inadequate equipment. On this view Q1 has independent authority and its severe peculiarities are not an index of fundamental textual corruption. The view is fundamentally re-examined by Sir Brian Vickers (see note 2), who concludes that the Q1 changes were made by Okes and the F cuts by Shakespeare’s company without his involvement. 20 Michael Warren, ‘The Diminution of Kent’, in Division, 59–73 (70). Cf. Warren, ‘Quarto and Folio King Lear and the Interpretation of Albany and Edgar’, in Shakespeare, Pattern of Excelling Nature, ed. David Bevington and Jay L. Halio (Newark, NJ : University of Delaware Press, 1978), 95–106. 21 See David Fuller, ‘Beauty, Pain, and Violence: Through Lessing and Nietzsche to King Lear’, in The Recovery of Beauty: Arts, Culture, Medicine, ed. Corinne Saunders, Jane Macnaughton and David Fuller (Basingstoke: Palgrave Macmillan, 2015), 87–107. 22 Warren, ‘The Diminution of Kent’, 69. 23 Urkowitz, Shakespeare’s Revision of ‘King Lear’, 93. 24 Q has other wrong speech prefixes at 1.4.219; 2.4.158, 263, 265 and 267; and 5.3.79, 154 and 303. While all these attributions can be tenuously defended, none but 5.3.154 has commanded any general assent.

232

Notes to pp. 165–174

25 The main differences are 4.2.53–9 (France spreading banners in England), 4.3.1–8 (details of the French leadership), and Q 3.1.22–34 / F 3.1.14–21 (Kent on French infiltration). Of these, the last is the most important – before revision theory seen as a place where Q and F need to be combined. Samuel Johnson (notes to King Lear, 1765) saw that the F cut (supposed by revision theorists authorial) destroyed the purpose of the scene (the Gentleman is sent ‘he knows not why, he knows not whither’): see Richard Knowles, ‘Revision Awry in Folio Lear, 3.1’, Shakespeare Quarterly 46 (1995): 32–46, who argues that the (stylistically incoherent) F version, with its incompetent understanding of the plot, cannot be authorial (and proposes a new way, consistent with the plot and less stylistically awkward, of combining Q and F). 26 This supposed authorial cut appears to have been executed by somebody who did not know the play well: ‘Then let them anatomize Regan’ (Q, l. 70; F, l. 33) refers back to Q only ‘Arraign her first. ’Tis Goneril’ (Q, l. 43). 27 Roger Warren, ‘The Folio Omission of the Mock Trial: Motives and Consequences’, in Division, 45–57. 28 Clare, ‘ “Who is it who can tell me who I am” ’, 46. Jonathan Croall, Performing ‘King Lear’: Gielgud to Russell Beale (London: Bloomsbury, 2015), based on interviews with directors and actors, gives no precise details of cuts. 29 Interview in King Lear, ed. Jonathan Bate and Eric Rasmussen, The RSC Shakespeare (London: Macmillan, 2009), 201. 30 Peter Holland, English Shakespeares: Shakespeare on the English Stage in the 1990s (Cambridge: Cambridge University Press, 1997), 39. Despite his pro-Folio-only view, Holland admits that in Hytner’s production the mock trial ‘became a pivotal moment of transition and recapitulation’ (40). 31 Film and television productions directed by Andrew McCollough (1953, with a text prepared by Peter Brook cut to c. 73 minutes), Peter Brook (1969), Grigori Kozintsev (1970), Jonathan Miller (1982), Michael Elliott (1984) and Brian Blessed (1999). 32 A striking example of the need for both texts occurs in 2.2 (Q, 71–4; F, 73–6) where, within four lines, Q and F four times alternately require correction from each other of readings that almost no editor defends. 33 W. W. Greg, The Shakespeare First Folio: Its Bibliographical and Textual History (Oxford: Clarendon, 1955), 379. 34 T. S. Eliot, ‘The Need for Poetic Drama’, The Listener XVI (411, November 1936): 994–5. Cf. id., ‘Poetry and Drama’, in Selected Prose of T. S. Eliot, ed. Frank Kermode (London: Faber, 1975), which is in part an analysis of the quasi-musical structure of the first scene of Hamlet, and the importance of that to the scene’s emotional effects. 35 On Valéry’s views in the context of the performance of Shakespeare’s poetry more generally see David Fuller, The Life in the Sonnets, Shakespeare Now! (London: Continuum, 2011), Part 2, ‘Dwelling in the Words’.

Notes to pp. 176–178

233

8 Textual Evidence and Musical Analysis: Once More on the First Movement of Beethoven’s ‘Tempest’ Sonata, Op. 31, No. 2 1 For a concise introduction to the history and practice of textual musical scholarship, see J. Grier, The Critical Editing of Music: History, Method, and Practice (Cambridge: Cambridge University Press, 1996). 2 For a magisterial appraisal of this issue, see J. Hepokoski, ‘Dahlhaus’s BeethovenRossini Stildualismus: Lingering Legacies of the Text-Event Dichotomy’, in The Invention of Beethoven and Rossini: Historiography, Analysis, Criticism, ed. N. Mathew and B. Walton (Cambridge: Cambridge University Press, 2013), 15–48. A condensed explanation of music’s emergence as a textual culture can be found in R. Taruskin, ‘Text and Act’, in Text and Act: Essays on Music and Performance (Oxford: Oxford University Press, 1995), 353–9. 3 The term ‘classical’ music is highly unsatisfactory as a description of Western art-musical culture, but given its widespread usage, I retain it here. I will need below to describe practices of the ‘Classical’ style, that is, the Viennese style of the late eighteenth century, which I will distinguish from the catch-all term ‘classical’ music by capitalization. 4 N. Cook, Beyond the Score: Music as Performance (New York and Oxford: Oxford University Press, 2013), especially 8–32. 5 C. Small, Musicking: The Meanings of Performing and Listening (Middletown, CN : Wesleyan University Press, 1998), 2. 6 Ibid., 8, italics in original. 7 Ibid., 220. This perception finds resonances in the work of Pierre Bourdieu, for whom the consumption of classical music is primarily a way of using cultural capital to define class difference; see P. Bourdieu, Distinction: A Social Critique of the Judgement of Taste, trans. R. Nice (London: Routledge, 1984). 8 R. Ingarden, The Work of Music and the Problem of Its Identity, ed. J. G. Harrel, trans. A. Czerniawski (Basingstoke: Macmillan, 1986). 9 L. Goehr, The Imaginary Museum of Musical Works (Oxford: Oxford University Press, 1992). Goehr’s argument is apostrophized in the well-known statement that ‘Bach did not intend to compose musical works. Only by adopting a modern perspective – a perspective foreign to Bach – would we say that he had. This implication proves to be correct as we examine with hindsight how the concepts governing musical practice before 1800 preclude the regulative function of the work-concept’ (8). 10 I think primarily of the agenda laid out in M. Foucault, The Archaeology of Knowledge (London: Routledge, 1972). 11 R. Parker, ‘Two Styles in 1830s London: “The Form and Order of a Perspicuous Unity” ’, in The Invention of Beethoven and Rossini, 123–38.

234

Notes to pp. 179–182

12 Ibid., 129–33; and see also W. Weber, The Great Transformation of Musical Taste: Concert Programming from Haydn to Brahms (Cambridge: Cambridge University Press, 2008). 13 Hepokoski reduces these alternatives to four approaches: the construction of alternative histories; the investigation of ‘peripheral’ repertoires; a focus on performance and the event rather than scores as texts; and an insistence on the primacy of ‘immediacy’ or the listening experience. See Hepokoski, ‘Dahlhaus’s Beethoven-Rossini Stildualismus’, 24–37. 14 This dependence is manifest as soon as Small tries to explain the experience of listening to a symphony rather than the sociology of its performer–audience relations or the space in which it is played. His refusal to see Beethoven’s Fifth Symphony as a score-defined object, for instance, leads him to make crass errors. The initial theme is not played ‘in unison by the whole orchestra, with the exception of certain softer-toned instruments’, as he claims, but by the strings and clarinets alone. And the opening thematic presentation is not coloured ‘by the presence of the chord known as the diminished seventh’; in fact, there are no diminished sevenths in the theme (bars 1–21), and no unequivocal diminished sevenths until the climax at bar 52, which is to say, at the climax of the transition. These errors could easily be avoided by acknowledging the epistemological value of the musical object: that is, by looking at a score. See Small, Musicking, 173–4. 15 This family also surely includes Wagner’s Tristan Prelude. 16 As reported in Carl Czerny’s memoir, first published in 1842; see Erinnerungen aus meinem Leben, ed. W. Kolneder (Baden-Baden: Heitz, 1968). On the provenance and meaning of this remark, see W. Kinderman, Beethoven (Oxford: Oxford University Press, 1995), 51–7. 17 A. B. Marx, Anleitung zum Vortrag Beethovenscher Klavierwerke (Berlin: Breitkopf und Härtel, 1863), 122. 18 On Marx’s side we find H. Leichtentritt, Musikalische Formenlehre (Leipzig: Breitkopf und Härtel, 1927), contrasting H. Riemann, L. Beethovens sämtliche KlavierSolosonaten: Ästhetische und form-technische Analyse mit historischen Notizen, 3 vols (Berlin: Max Hesse, 1918–192), L. Misch, ‘Das Problem der D-moll Sonata von Beethoven: Versuch einer neuen Formdeutung’, in Beethoven-Studien (Berlin: Walter de Gruyter, 1950), 42–55 and in English, D. F. Tovey, A Companion to Beethoven’s Pianoforte Sonatas (London: Associated Boards of the Royal Schools of Music, 1935), 128–30. For an overview of the Sonata’s reception, see S. Vande Moortele, ‘The First Movement of Beethoven’s Tempest Sonata and the Tradition of Twentieth-Century Formenlehre’, in Beethoven’s Tempest Sonata: Perspectives of Analysis and Performance, ed. P. Bergé, W. E. Caplin and J. D’hoe (Leuven: Peeters, 2009), 293–312. 19 For a definition, see W. E. Caplin, Analyzing Classical Form: An Approach for the Classroom (New York and Oxford: Oxford University Press, 2013), 33–72. 20 See Caplin, Analyzing Classical Form, 34–7 and 133–7.

Notes to pp. 183–186

235

21 C. Dahlhaus, Nineteenth-Century Music, trans. J. Bradford Robinson (Berkeley and Los Angeles, CA : University of California Press, 1989), 15. 22 Ibid., 10. 23 Ibid., 9. For a penetrating recent critique of Dahlhaus’ argument, see Hepokoski, ‘Dahlhaus’s Beethoven-Rossini Stildualismus’, 5–24. 24 J. Schmalfeldt, ‘Form as the Process of Becoming: The Beethoven-Hegelian Tradition and the Tempest Sonata’, Beethoven Forum 4 (1995): 37–71, and id., In the Process of Becoming: Analytic and Philosophical Perspectives on Form in Early NineteenthCentury Music (New York and Oxford: Oxford University Press, 2011). 25 As Dahlhaus has it, Beethoven asks us ‘to exchange categories’, as a result of which ‘the listener becomes aware of himself [sic] as subject, and of his creative role in the formal process’ (‘Erst die Zumutung, Kategorien auszutauschen, bringt dem Hörer seine Subjektivität als Konstituens des Formprozesses zu Bewußtsein’). See C. Dahlhaus, Ludwig van Beethoven und seine Zeit (Laaber: Laaber-Verlag, 1987), 72 and also Schmalfeldt, In the Process of Becoming, 36. 26 Schmalfeldt, In the Process of Becoming, 10. 27 Ibid., 23. Beethoven and Hegel were both born in 1770; there have been numerous attempts to affiliate Hegel’s philosophy and Beethoven’s music. For a recent formulation of the relationship, see M. Spitzer, Music as Philosophy: Adorno and Beethoven’s Late Style (Bloomington, IN : Indiana University Press, 2006). On the relationship between dialectics and the analysis of music in the Beethovenian tradition, see also J. Horton, ‘Dialectics and Musical Analysis’, in Aesthetics of Music: Musicological Perspectives, ed. S. Downes (New York: Routledge, 2014), 111–43. 28 Principally A. B. Marx, Die Lehre von der musikalischen Komposition, praktischtheoretisch, 4 vols (Leipzig: Breitkopf und Härtel, 1841), and see also Musical Form in the Age of Beethoven: Selected Writings on Theory and Method, trans. and ed. S. Burnham (Cambridge: Cambridge University Press, 1997). 29 Th. W. Adorno, Hegel: Three Studies, trans. S. Weber Nicholson (Cambridge, MA : MIT Press, 1993); id., Beethoven: Philosophy of Music, ed. R. Tiedemann, trans. E. Jephcott (Stanford, CA : Stanford University Press, 1998); and id., Philosophy of New Music, trans. R. Hullot-Kentor (Minneapolis, MN : University of Minnesota Press, 2006). For an overview, see M. Paddison, Adorno’s Aesthetics of Music (Cambridge: Cambridge University Press, 1993). 30 Th. W. Adorno, ‘Late Style in Beethoven’ and ‘Alienated Masterpiece: The Missa Solemnis’ in id., Essays on Music, ed. R. Leppert, trans. S. H. Gillespie (Berkeley, CA : University of California Press, 2002), 564–8 and 569–83. 31 The concept has its classic formulation in the writings of E. T. A. Hoffmann, particularly ‘Review of Beethoven’s Fifth Symphony’, Allgemeine musikalische Zeitung 12 (4 and 10 July, 1810), cols 630–42 and 652–9, and ‘Beethoven’s Instrumental Music’, in Source Readings in Music History: The Romantic Era, trans. and

236

32

33 34

35

36

37

38 39

40

Notes to pp. 186–188

ed. O. Strunk (New York and London: Norton, 1965), 35–41, especially 45: ‘When we speak of music as an independent art, should we not always restrict our meaning to instrumental music, which . . . gives pure expression to music’s specific nature recognizable in this form alone? It is the most romantic of all the arts – one might almost say, the only genuinely romantic one – for its sole subject is the infinite.’ For a recent investigation of the relationship between eighteenth-century theory, pedagogy and composition, see R. O. Gjerdingen, Music in the Galant Style (Oxford: Oxford University Press, 2007). See C. Czerny, School of Practical Composition, trans. J. Bishop (London: Cocks, 1848). For Marx, see note 28 above. Schmalfeldt, In the Process of Becoming, 7 and Goehr, The Imaginary Museum of Musical Works, 218–23, especially 218: ‘as music came to be seen as the product of a free person’s labour, a change was deemed necessary in ownership rights.’ This concept is elaborated in relation to the nineteenth-century symphony in M. E. Bonds, After Beethoven: Imperatives of Originality in the Symphony (repr. Cambridge, MA : Harvard University Press, 2013). Bergé, Caplin and D’hoe (ed.), Beethoven’s Tempest Sonata: Perspectives of Analysis and Performance and also the essays on the Tempest by C. Reddick, W. Caplin, J. Hepokoski and M. Cheng-Yu Lee, Music Theory Online 16, no. 2 (2010), ‘Special Issue: Form as Process’ (www.mtosmt.org/issues/mto.10.16.2/toc.16.2, accessed 14 July 2017). W. E. Caplin, ‘Beethoven’s Tempest Exposition: A Springboard for Form-Functional Considerations’ and J. Hepokoski, ‘Approaching the First Movement of Beethoven’s Tempest Sonata through Sonata Theory’, in Beethoven’s Tempest Sonata, 87–126 and 181–212; W. E. Caplin, ‘Beethoven’s Tempest Exposition: A Response to Janet Schmalfeldt’ and James Hepokoski, ‘Formal Process, Sonata Theory and the First Movement of Beethoven’s Tempest Sonata’, Music Theory Online 16, no. 2 (2010): 16.2.6 and 16.2.7 (www.mtosmt.org/issues/mto.10.16.2/toc.16.2, accessed 14 July 2017). Caplin, ‘Beethoven’s Tempest Exposition’, 91. Hepokoski, ‘Approaching the First Movement of Beethoven’s Tempest Sonata through Sonata Theory’, 187. Hepokoski’s essay applies the principles of a theory developed extensively in J. Hepokoski and W. Darcy, Elements of Sonata Theory: Norms, Types, and Deformations in the Late-Eighteenth-Century Sonata (New York and Oxford: Oxford University Press, 2006), in which terms the Largo is defined as a ‘P0 module’: an element of the main theme (P), which is preparatory to its first stable feature (P1). See Elements of Sonata Theory, 86–91. Hepokoski and Darcy consider this passage in Elements of Sonata Theory, 87–8. In ‘Approaching the First Movement of Beethoven’s Tempest Sonata through Sonata Theory’, Hepokoski cites the first movement of Haydn’s Symphony No. 95 as a precursor; see 189.

Notes to pp. 191–197

237

9 Fragments Shored against Ruin: Reassembling The Waste Land 1 G. Hough, Image and Experience: Studies in a Literary Revolution (London: Duckworth, 1960), 28. 2 Quoted in J. S. Brooker, T. S. Eliot: The Contemporary Reviews (Cambridge: Cambridge University Press, 2004), 115. 3 I. A. Richards, ‘A Background for Contemporary Poetry’, The Criterion 3, no. 2 (1925): 520 note. 4 T. S. Eliot, ‘A Note on Poetry and Belief ’, The Enemy: A Review of Art and Literature (ed. P. W. Lewis), no. 1 (1927): 15–17 (14). 5 Quoted in The Poems of T. S. Eliot, Volume I: Collected and Uncollected Poems, ed. C. Ricks and J. McCue (Baltimore, MD : Johns Hopkins University Press, 2015), 575. 6 Ibid., 72. 7 T. S. Eliot, On Poetry and Poets (London: Faber and Faber, 1957), 109–10. 8 The Letters of T. S. Eliot, Volume I, 1898–1922, ed. Valerie Eliot and H. Haughton (London: Faber and Faber, 2009), 707. 9 Ibid., 729. 10 Ibid., 748. 11 Quoted in J. McCue and C. Ricks, ‘Masterpiece in the Making’, The Times Literary Supplement, 9 October 2015. 12 T. S. Eliot, ‘The Art of Poetry. I’, The Paris Review, no. 21 (Spring/Summer 1959): 52. 13 T. S. Eliot, The Waste Land: A Facsimile and Transcript of the Original Drafts, Including the Annotations of Ezra Pound, ed. Valerie Eliot (London: Faber and Faber, 1971), xxix. 14 The Complete Prose of T. S. Eliot: The Critical Edition: Tradition and Orthodoxy, 1934–1939, ed. I. Javadi, R. Schuchard and J. Stayer (Baltimore, MD : Johns Hopkins University Press, 2017), 604. 15 The Complete Prose of T. S. Eliot: The Critical Edition: The War Years, 1940–1946, ed. D. Chinitz and R. Schuchard (Baltimore, MD : Johns Hopkins University Press, 2017), 761. 16 The Letters of T. S. Eliot, Volume I, 1898–1922, 626. 17 Ibid., 549. 18 Ibid., 557. 19 The Poems of T. S. Eliot, Volume I: Collected and Uncollected Poems, 66. The letter to Schiff is quoted from The Letters of T. S. Eliot, Volume I, 1898–1922, 602. 20 Ibid., 62. 21 The Letters of T. S. Eliot, Volume I, 1898–1922, 626. 22 Ibid., 629. 23 Ibid., 630. 24 The Poems of T. S. Eliot, Volume I: Collected and Uncollected Poems, 558.

238 25 26 27 28 29 30 31 32

33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48

Notes to pp. 197–203

The Letters of T. S. Eliot, Volume I, 1898–1922, 707. Ibid., 752. Ibid., 786–7. D. Gallup, T. S. Eliot: A Bibliography, Including Contributions to Periodicals and Foreign Translations (London: Faber and Faber, 1952), 87. The Letters of T. S. Eliot, Volume I, 1898–1922, 746. The Poems of T. S. Eliot, Volume I: Collected and Uncollected Poems, xii. C. Ricks, T. S. Eliot and Prejudice (London: Faber and Faber, 1988), 191. The Complete Prose of T. S. Eliot: The Critical Edition: English Lion, 1930–1933, ed. J. Harding and R. Schuchard (Baltimore, MD : Johns Hopkins University Press, 2015), 133. The Letters of T. S. Eliot, Volume I, 1898–1922, 557. The Letters of T. S. Eliot, Volume II, 1923–1925, ed. Valerie Eliot and H. Haughton (London: Faber and Faber, 2009), 202–3. The Poems of T. S. Eliot, Volume I: Collected and Uncollected Poems, 77. The Waste Land: A Facsimile and Transcript of the Original Drafts, 126. The Poems of T. S. Eliot, Volume I: Collected and Uncollected Poems, 60. Ibid., 321. The Waste Land: A Facsimile and Transcript of the Original Drafts, vii. The Letters of T. S. Eliot, Volume I, 1898–1922, 748. The Poems of T. S. Eliot, Volume I: Collected and Uncollected Poems, 581–2. Ricks, T. S. Eliot and Prejudice, 38. A. Julius, T. S. Eliot, anti-Semitism, and Literary Form (Cambridge: Cambridge University Press, 1995), 132. McCue and Ricks, ‘Masterpiece in the Making’, 556. The Waste Land: A Facsimile and Transcript of the Original Drafts, 1. The Letters of T. S. Eliot, Volume I, 1898–1922, xix. Ibid., 765. D. F. McKenzie, Bibliography and the Sociology of Texts (Cambridge: Cambridge University Press, 1999), 15.

Index of subjects and themes athetesis 42, 46 audiences 37–41, 70–1 authorial annotated copies 136–47 composition and revision 5–6, 136–48, 159–61, 193–203 control over own work (or lack thereof) 102, 133–48, 156–7, 159–61 draft versions 192–203 identity 14, 46, 62–87, 94–101, 130 manuscripts and typescripts 6, 192 punctuation 138–9 versions 4, 133–48, 192–203 authorship 14, 19–20, 28–37, 46, 62–87, 98, 130 autography 1–2, 93–4, 115–52, 156–9 binding 8–9 book format and composition 7–8, 48–61, 115–19, 129–30 ‘boards’ see wax tablets cerae see wax tablets codex 7–8, 115 and roll 48–61 composite manuscript (It. zibaldone) 119, 129–30 membranae see parchment leaves pandects 59–61, 210 parchment leaves (membranae, tabellae) 49–53, 209 pecia (‘piece’) 12–13 roll/scroll 4, 7, 23, 48–61 tabellae see wax tablets / parchment leaves wax tablets (‘boards’, cerae, tabellae) 52–3 censorship and censoring practices 23, 26, 62, 201–2 citation 24 compilations 63–86 composition 5–7, 23, 28

and musical analysis 179–89 formulaic 37–41 oral 36–40, 44, 46, 71 conjecture 41–2 conservation 26 contamination 63–86, 104 dating of manuscripts 66, 74 editing 4–6, 16–19, 102–12, 118, 136–7, 139–49, 154–75 and bibliographical information 154–75 and competing variant readings 104–9 and diction 31 and external evidence 105–7, 191–203 and geographical distribution of witnesses 14, 62–92, 109–12, 196–202 and grammar 31 and internal consistency 31–4, 103–4 and literary judgment 31–4, 43–6, 107, 154–75 and theatrical practice 154–75 and the consolidated text 162, 167, 169, 171–2, 175, 228, 230 and the Editorial Composite 201 collaborative 136–48, 193–203 divergent versions 62–92, 133–75 facsimiles 193–203, 218–20, 230, 237–8 first editions 13–14, 16 forgery 66, 68, 76–8, 80–1 Glosses 17–18 illumination 8, 10, 101 instability of the published text 14–15, 198, 200 languages Arabic 62–92, 211–17

239

240

Index of subjects and themes

Middle Arabic 66, 81 Syrian Arabic 66 Armenian 56, 209 Coptic 56, 209 English 1–3, 154–75, 191–203 Ethiopic 56, 209 French 62–3, 70, 76–8, 81, 84, 133–53 German 80–1, 84–5 Greek 3, 70, 20, 23–4, 28–61 Hebrew 7, 20, 59, 132, 210 Italian 93–132 Florentine and North Italian 110–12, 121 Romagnole 110–11 Tuscan 121 Latin 8, 14, 16, 19, 20, 34, 48–61, 111–14, 120–1, 123–4, 128, 150 Middle Egyptian 213 Persian 63, 65, 71, 79, 212 Provençal 63 Syriac 56, 69, 209 Turkish 63, 65–6, 82, 86 layout and distribution of text 7, 13, 16, 53–4, 58–61, 131–2, 198–200 chapter standardization 16 cross-referencing 16–17, 56–7 headings 16 paraphs 16 lectio difficilior 1, 107 libraries 26, 34, 58–61 marginalia 21–2, 102–3, 133–53, 220 online texts 14 accessibility 22 inferior copies 14 longevity 22–3 ‘pay-walls’ 14 orality and transmission 3–4, 37–41, 68, 71, 75 its influence on language 130–1 paper 13, 23–4 papyrus 7, 23, 25, 28, 31–2, 49–50, 72 parchment 8, 22–3, 49–50 pen names and pseudonimity 19–20

performance 11–12, 37–41, 69–71, 81, 168, 174, 176–90, 232 and musical textuality 176–9, 189–90 pricing of early printed books 14–16 printing 12–13, 23–4 Arabic type 69 Greek type 24 Latin type 24 private circulation of ancient and medieval manuscripts 21 of digitized texts 21 ‘prompt books’ 156 recension 4, 16, 40, 66, 75–86 and interpretation 104 technique 103 revision of transcripts and proofs 195–8 serialization 7 stemmatic method 109–12 translation 20, 42, 59, 63–87, 119–21, 210, 223 and adaptation 20, 63–87 and self-translation 20 transmission difficulties of 23 errors in manuscripts 8, 12, 14, 26, 154 errors in typesetting 12, 198–9 incomplete copies 63–86 lost witnesses 31, 63, 68, 75, 93, 100, 102, 116, 125, 220 oral 4, 54, 68 scribal 6, 8, 13, 26, 39, 50, 62, 77–8, 101–10 unfinished works 5, 60 and projectuality 128, 132, 133–53 variant readings 34, 42–6, 103–9, 154–75, 192–203, 208 watermarks 195 word and image 10–11, 94–6, 115–32 writings left unpublished 5, 117

Index of manuscripts and annotated printed copies Printed copies are marked by an asterisk. For a list of manuscripts containing The Thousand and One Nights see also pp. 88–92. Abba Garima Abba Garima Monastery Garima 1 and 2 (‘Garima Gospels’) 209 Barcelona Fundació Sant Lluc Evangelista P.Barc. Inv. 1 (= P67) 51 Berkeley, CA University of California, Bancroft Library P.Tebt. 4 (2390) 31–2 Bologna Archivio di stato Registri della Curia del Podestà (‘Memoriali bolognesi’) 102 Bordeaux Bibliothèque municipal *S. 1238 Rés Coffre (‘EB’) 136–47, 225–7 Cambridge University Library Qq 106–9 215 Chicago University of Chicago, Oriental Institute E17618 72–3, 213 Cologny Bibliotheca Bodmeriana P.Bod. XIV-XV (= P75) 51 Dublin Chester Beatty Library P. Chester Beatty I (= P45) 51 Durham University Library *Cosin W.2.11 16 *SB 2537 9

Cathedral Library A.II .3 17 A.III .9 18 Florence Archivio di stato Riformagioni, Capitani di parte, Numeri Rossi, 20 (‘Libro del Chiodo’) 96, 100–1, 218–19 Riformagioni, Capitoli, Registri, 19 A 100–1, 219 Biblioteca Medicea Laurenziana Amiatino 1 (Codex Amiatinus) 7–8, 60 Ashburnham 36 125 Ashburnham 828 102, 110 Plut. 1. 56 (‘Rabbula Gospels’) 209 Plut. 26 sin. 1 103–4, 110, 221 Biblioteca Nazionale Centrale B.R. 228 130 Biblioteca Riccardiana 1005 110 1035 103 Museo degli Uffizi, Gabinetto dei Disegni 8P 119 Gotha Forschungsbibliothek Orient A 2632–5 78, 215 Orient A 2637 215 Orient A 2638 215 Orient A 2639 215 Kayseri Ras.it Efendi Yazma Eser Kütüphanesi 614 73

241

242

Index of manuscripts and annotated printed copies

London British Library Add. 43725 (Codex Sinaiticus) 7, 58 Arundel 263 (Codex Arundel) 117–18, 122, 125, 127, 130 *C.45.e.1820 21–2 DEL AR 1308 79, 215 IO ISL 2699 79, 215 P. Egerton 2 (‘Egerton Gospel’) 51 Royal 1. D. V-VIII (Codex Alexandrinus) 210 Victoria and Albert Museum, National Library of Design Forster I 117, 123, 125 Forster II 117, 124 Forster III 117, 123 Madrid Biblioteca Nacional 8936 (Codex Madrid II ) 116, 122–3 8937 (Codex Madrid I) 122–3 Manchester John Rylands Library Arabic 647 66, 77, 212, 214 Arabic 706 73–4 Maynooth Russell Library *R B Case 122 142–9, 226–8 Milan Archivio Storico Civico e Biblioteca Trivulziana 1080 103, 110, 220 N 2162 (Codex Trivultianus) 117, 120–1 Biblioteca Ambrosiana Codex Atlanticus 117–20, 122–3, 127, 130 Biblioteca Nazionale Braidense AG XII 2 110 *Aldina AP XVI 25 102–3, 110, 220 Montecassino Biblioteca dell’Abbazia 512 (589) 221–2 Munich Bayerische Staatsbibliothek arab. 629 214

New York New York Public Library Berg Collection, MSS Eliot, TS 193–203 Oxford Bodleian Library Or 550–6 77, 80, 214 Magdalen College Library Gr. 17 (= P64) 51 Sackler Library P.Oxy. L. 3554. 25 Trinity College, Danson Library L. Binyon, ‘For the Fallen’ 1–3 Paris Bibliothèque nationale de France arabe 3595–7 78, 215 arabe 3598–601 78, 215 arabe 3602–5 78, 215–16 arabe 3609 66, 77–8, 82, 84, 211–12 arabe 3610 66–7, 77–8, 82, 84, 211–12 arabe 3611 66, 77–8, 82, 84, 211–12 arabe 3613 76, 214 arabe 3614 76, 214 arabe 3616 76, 214 arabe 3617 215 arabe 3637 214 arabe 3645 65, 211 arabe 4678–9 78, 214–15 fr. 15277 63–4, 211 fr. 18277 80–1, 216 lat. 2201 116 Suppl. gr. 1120 ii 3 (= P4) 51 Institut de France 2172 + 2185 (Codex A) 117, 122, 130 2173 + 2184 (Codex B) 117, 119–20 2174 (Codex C) 117, 121 2175 (Codex D) 117, 125 2176 (Codex E) 116–17, 126–7, 132 2177 (Codex F) 117, 126 2178 (Codex G) 117, 126, 132 2179 (Codex H) 117, 123 2180 (Codex I) 117, 122–3 2181 (Codex K) 117, 125 2182 (Codex L) 117, 123 2183 (Codex M) 117, 123

Index of manuscripts and annotated printed copies Piacenza Biblioteca Comunale Passerini-Landi 190 102, 220 Riyad King Faisal Center for Research and Islamic Studies 2415 72, 213 Seattle Collection of Bill and Melinda Gates ‘Codex Leicester’ 117, 125 Strasbourg Bibliothèque nationale et universitaire 4278–81 81–2, 216 Toledo Biblioteca Capitular 47.10 223 Zelada 1046 103 Tübingen Universitätsbibliothek M.a.VI .33 72, 213 Turin Biblioteca Reale Varia 95 (Codex of the Flight of Birds) 117, 125 Venice Biblioteca Nazionale Marciana gr. 822 (Venetus A) 34–5, 205

Vincennes Service historique de la Défense (SHD ) Paris, ‘XL /37j, 77–8, 214 Vatican City Biblioteca Apostolica Vaticana Chigi L VI 213 103 Sbath 254 62, 68–70, 211–12 Urb. Lat. 366 109–11, 222 Urb. Lat. 1270 (‘Libro di pittura’) 117–18, 130–1 Urb. Lat. 1757 130 Vat. Ar. 782 66, 212 Vat. Gr. 1209 (Codex Vaticanus) 58 Vat. Lat. 3197 220 Vat. Lat. 3199 103 Wimborne, Dorset Kingston Lacy House *M.1.1. 151–2, 228 Windsor Eton College Library 129 223 Royal Library 12275–727, 19000–152 117, 127 19059 121 Yerevan Mesrob Mashtots Matenadaran 10450 (‘Zeyt’un Gospels’) 209

243

Index of names Abardo, Rudy 221 Abbott, Nabia 213 Abdel-Halim, Mohamed 211 Accius, Lucius 113 c Adawī, Muh.ammad 82, 86, 217 Adorno, Theodor 185–6, 190, 235 Aesop 122 Agamben, Giorgio 223 Akel, Ibrahim 212–17 Aland, Kurt 209 Albertazzi, Marco 219 Alberti, Leon Battista 20, 128 Albertus Magnus 124 Alexander the Great 31 Alf Layla wa Layla (Thousand and One Nights) 62–92, 211–17 Alighieri, Dante 10, 23, 25, 83–114, 120, 122, 192, 217–23 Alighieri, Jacopo 98, 111, 222 Alighieri, Pietro 108–9, 221 Alloatti-Boller, Sara 222 Andalò, Loderingo degli 102, 220 Anne of Austria, Queen of France 149 Anselm of Canterbury 4 Antonio da Vinci 130 Archibald, Elizabeth xii Aristarchus 33, 40, 204–5 Aristeas, Letter of 206 Aristotle 30, 114, 204, 223 Arnould, Jean-Claude 226 Arp, Hans 11 Arrigo de’ Semintendi 119 Arthurian Cycle 24 Asor Rosa, Alberto 224 Asselin de Cherville, Jean-Louis 78–9 Assemani, Giuseppe Simone (Joseph) 69 (see Simcānī, Yūsuf) Augustine 59–60, 116, 210 Ausonius, Decimus Magnus 4 Bach, Johann Sebastian 233 Balbi, Giovanni 113

244

Baldassarri, Guido 224 Balsamo, Augusto 220 Balsamo, Jean 140, 226–7 Bambach, Carmen C. 224 Bambaglioli, Graziolo de’ 111, 222 Bankes, Sir Ralph 151–3 Barański, Zygmunt G. 223 Barbi, Michele 104 Barbier de Meynard, Charles 212 Barde, Paul 76 Barnes, Djuna 198 Bartoli, Cosimo 20 Bartoli, Elisabetta 218 Bate, Jonathan 232 Battezzato, Luigi 207 Bede (the Venerable) 13, 23, 210 Beethoven, Ludwig van 176–90, 233–6 Bell, Acton (pseudonym of Brontë, Anne) 19 Bell, Currer (pseudonym of Brontë, Charlotte) 19 Bell, Ellis (pseudonym of Brontë, Emily) 19 Bellomo, Saverio 218–19, 222 Bembo, Pietro 103, 220 Benjamin, Walter 222 Bentley, Richard 20 Beowulf 4, 20, 23, 26 Bergé, Pieter 187, 234, 236 Beringhen, Henri de 149 Bernard of Clairvaux 19 Bertoni, Giulio 220 Besold, Christoph 150–3, 228 Beutler, Ernst 205 Bevington, David 231 Bianca, Concetta 221 Bianchi, Thomas-Xavier 216 Bible 7, 15–17, 20, 25, 41, 48–61, 66, 69, 208–11 Biblia pauperum 10 Bigi, Emilio 221 Binyon, Laurence 1–3, 9 Bird, Graeme D. 207 Bishop, John 236

Index of names Bjaï, Denis 226 Blake, William 10, 21–2 Blakemore Evans, G. 230 Blayney, Peter W. M. 231 Blessed, Brian 232 Blum, Claude 225–6 Boase, Alan 225 Boccaccio, Giovanni 98, 100, 103, 109, 218–19 Bodoni, Giovan Battista 103 Bonagiunta da Lucca 107–8 Bonaparte, Napoleon see Napoleon I, Emperor of the French Bonaventura (ps.-) 4 Bonds, Mark Evan 236 Boni & Liveright (publishers) 197–200 Boniface VIII , Pope 96 Borgia, Cesare 124 Boschi Rotiroti, Marisa 222 Boswell, James 155 Botticelli, Sandro 10 Bottigheimer, Ruth B. 211 Boudou, Bénédicte 226 Bouffard, Élodie 211–13 Bourdieu, Pierre 233 Bourus, Terri 229 Boutcher, Warren 225, 228 Bowman, Alan K. 208 Boz (pseudonym of Dickens, Charles) 19 Bracciolini, Poggio 121–2 Brach, Pierre de 139 Bradford Robinson, J. 235 Bradley, David 230 Bramante, Donato 123 Branca, Vittore 217–18 Brilli, Elisa 219 Brontë, Anne 19 (see Bell, Acton) Brontë, Charlotte 19 (see Bell, Currer) Brontë, Emily 19 (see Bell, Ellis) Brook, Peter 167, 232 Brooker, Jewel Spears 237 Brooks, Cleanth 191–2, 200 Brugnoli, Giorgio 223 Brunelleschi, Filippo 95, 129 Bruni, Leonardo 83, 100, 217 Bunyan, John 19 Burckhardt, John Lewis 78 Burkhart, R. E. 230 Burne-Jones, Edward Coley 9

245

Burnham, Scott 235 Burton, Sir Richard 84–5, 217 Calvi, Girolamo 118, 224 Cangrande della Scala 97–8, 113, 218, 223 Canova, Andrea 222 Caplin, William 187–8, 234, 236 Carpi, Umberto 218 Carpianus 209 Carpus of Beroea 49–50 Carroll, Lewis (pseudonym of Dodgson, Charles) 19 Carroll, Walter C. 229 Caruso, Carlo 220 Casadei, Alberto 218, 223 Casella, Mario 104 Cassio, Albio Cesare 206 Cassiodorus Senator, Flavius Magnus Aurelius 58–61, 116, 210 Castiglione, Baldassarre 127 Catalano de’ Malavolti 102, 220 Cato Uticensis, Marcus Porcius (the Younger) 133–6 Caussin de Perceval, Jacques 78, 80 Cavalcanti, Guido 108 Caxton, William 13–14 Cazotte, Jacques 76, 85 Céard, Jean 139, 226 Cecchini, Enzo 218 Cecco d’Ascoli 122–3 Céline, Louis-Ferdinand (pseudonym of Destouches, Louis-Ferdinand) 19 Ceolfrith 210 Cervantes Saavedra, Miguel de 14 Champollion, Jean-François 80 Charteris, Leslie 19 Chaucer, Geoffrey 5, 9, 14–15, 23 Chauvin, Victor 216–17 Chavannes, Jules 227 Chavis, Dom Denis 76–8, 80, 85 (see Shāwīsh, Diyūnīsiyūs) Cheng-Yu Lee, Mike 236 Chettle, Henry 159 Chiamenti, Massimiliano 221–2 Chiavacci Leonardi, Anna Maria 107, 221 Chinitz, David E. 237 Chopin, Frédéric 184 Chraïbi, Aboubakr 211–17 Cicero 4, 53, 151

246

Index of names

Cipollone, Annalisa 222 Clare, Robert 167, 229, 232 Clemens, Samuel 19 (see Twain, Mark; Sieur Louis de Conte) Clement of Rome (ps.-) 55 Cobden-Sanderson, Richard 197 Coke, Thomas, Lord Leicester 116 Colbert, Jean-Baptiste 63 Coleridge, Samuel Taylor 195 Coller, Ian 215 Colli, Giorgio 205 Condell, Henry 158–9 Connery, Sean 12 Conrad, Joseph 196, 202 Conroy, Mary 193 Contini, Gianfranco 219 Continuation des Mille et Une Nuits 76–7, 214 Cook, Nicholas 177, 233 Corbeau, André 224 Coste, Pierre 137, 225 Crane, Ralph 156 Croall, Jonathan 232 Curtius, Ernst Robert 98–9, 101 Cypria 29–30 Czerniawski, Adam 233 Czerny, Carl 186, 234, 236 D’hoe, Jeroen 187, 234, 236 Dachs, Hans 205 Dahlhaus, Carl 183–5, 188, 190, 233–5 Dal Pozzo Toscanelli, Paolo 130 Danek, Georg 206 Darcy, Warren 236 Dardi, Andrea 222 Dati, Gregorio 126 Davidsohn, Robert 218 Davies, Charles T. 218 De Robertis, Teresa 218 De Smet, Daniel 213 Dee, John 21 Dei, Benedetto 122 Della Torre, Marcantonio 126 Demades of Athens 138 Demetrius the Ephesian silversmith 49 Desan, Philippe 226 Destouches, Louis-Ferdinand 19 (see Céline, Louis-Ferdinand) Dīb, But.rus 69 (see Dipy, Pierre)

Dickens, Charles 19, 24, 196 (see Boz) Dickey, Eleanor 205 Dihle, Albrecht 207 Dionisi, Gian Iacopo 103, 220 Dionisotti, Carlo 224 Dionysius Thrax 40, 206 Dipy, Pierre 69 (see Dīb, But.rus) Disney, Walt 62 Diyāb, H.annā 62–4, 68–70, 76, 81, 211–12 Dodge, Bayard 212 Dodgson, Charles 19 (see Carroll, Lewis) Domenico di Michelino 94–5 Donaldson, Ian 230 Donatus, Aelius 24 Doré, Gustave 10 Dostoevsky, Fyodor 24 Dover Wilson, John 173, 228 Downes, Stephen 235 Dué, Casey 205, 207 Dungan, David L. 209 Dupin, Armandine Aurore 19 (see Sand, George) Dürer, Albrecht 5, 10–11 Duthie, George Ian 173, 228 Ebbott, Mary 207 Eco, Umberto 12 Edwards, Philip 229 Egidi, Francesco 219 Elgar, Edward 1 Eliot, George (pseudonym of Evans, Mary Ann) 19 Eliot, Henry 196, 202 Eliot, Marian 196 Eliot, Thomas Stearns 6, 14, 174, 191–203, 232, 237–8 Eliot, Valerie 193, 195, 200–2, 237–8 Eliot, Vivienne (Vivien) 6, 194, 196, 200, 202–3 Elisséeff, Nikita 216 Elliott, Michael 232 Eragny Press 9 Erne, Lukas 230 Estienne, Henri II 16–17 Estienne, Robert I 16 Euclides 124 Eusebius of Caesarea 56–8, 209 Eustathius of Thessaloniki 33–4, 205 Evans, Mary Ann 19 (see Eliot, George)

Index of names Farmer, David Hugh 210 Fenelli, Laura 219 Fernández de Avellaneda, Alonso 20 Ficino, Marsilio 20 Finkelberg, Margalit 207 Fiore di virtù 123 Firuski, Maurice 197 Fleischer, Heinrich Leberecht 86, 215 Fleming, Ian 19 Flower, Barbara 221 Foakes, Reginald A. 169–70, 172–3, 228 Folena, Gianfranco 219–21 Foley, Jonn Miles 206 Ford, Andrew 204 Foucault, Michel 178, 233 Frame, Donald M. 225, 227 Francesco da Barberino 101, 112, 219 Francesco di Ser Nardo da Barberino 103 Frederick Henry, Prince of Orange 149 Fuller, David 202, 231–2 Galland, Antoine 63–9, 72–3, 75–9, 81–5, 87, 211–12, 214, 216–17 Gallup, Donald 193, 198, 238 Gamberini, Andrea 219 Gamble, Harry 54, 209 Gates, Bill 117 Gathercole, Simon J. 209 Gawain and the Green Knight 23 Gébelin, François 136, 225 Genet, Jean-Philippe 219 Geoffrey of Vinsauf 108 Gercke, Alfred 221 Gerhardt, Mia I. 211 Ghiberti, Buonaccorso 129–30 Ghiberti, Lorenzo 129 Giacomo da Lentini 108 Gilgamesh, Epic of 43 Gillespie, Susan H. 235 Ginzburg, Carlo 218 Giovanni di Paolo 10 Gjerdingen, Robert O. 236 Godeau, Antoine 149 Goehr, Lydia 178, 186, 233, 236 Goethe, Johann Wolfgang 36, 205 Goitien, Solomon D. 213 Gombrich, Ernst H. 131, 224 Goodacre, Mark 209 Gordon, John D. 193

247

Gospel of Mary 51 Gospel of Peter 51 Gospel of Thomas 51, 55 Gospels 48–61 Gournay, Marie de 139–53 Grafton, Anthony 205 Grahame, Kenneth 10 Gray, Patrick xii Graziosi, Barbara 43, 204, 206–8 Greek Epic Cycle 43 Greenblatt, Stephen 231 Greenwood, Emily 208 Greg, Walter W. 158, 173, 175, 229–30, 232 Gregory XIII , Pope 69 Gregory XV, Pope 69 Grier, James 233 Grotzfeld, Heinz 211–14, 216 Grotzfeld, Sophia 212–14 Grynaeus, Simon 20 Gualtieri, Piero 219 Guez de Balzac, Jean-Louis 149 Guido Novello da Polenta 97 Guidotti, Angela 224 Guillaume, Jean-Patrick 211 Guinizzelli, Guido 108 Guittone of Arezzo 108 Gumbrecht, Hans Ulrich 206 Gurr, Andrew 230 Gustavus Adolphus, King of Sweden 149 Gutenberg, Johannes, 15 Haag, Émile 227 Haag, Eugène 227 Habicht, Christian Maximilian 79–80, 84, 86–7, 211, 215, 217 Hackstein, Olav 207 Haddawy, Husain 87, 217 Haffenden, John 202 Hagen, Friedrich Heinrich von der 87 Halflants, Bruno 212 Halio, Jay L. 228, 231 Halporn, James W. 210 Harding, Jason 238 Hardy, Thomas 4 Harrel, Jean G. 233 Haslam, Michael 207 Haubold, Johannes 43, 206–7 Haughton, Hugh 237–8 Haydn, Joseph 183, 187–8

248

Index of names

Heaney, Seamus 20 Hegel, Georg Wilhelm Friedrich 185, 188, 190, 235 Heller-Roazen, Daniel 223 Heminges, John 158–9 Henderson, James W. 193 Hepokoski, James 187–8, 233–6 Herodotus 29, 204 Herwagen the Younger, Johann 13 Heyberger, Bernard 212 Heywood, Thomas 158 Hezār Afsāneh (‘One thousand fairytales’) 71, 75 Hilgard, Alfred 206 Hirrel, Michael J. 230 Hitler, Adolf 19 ‘Hitler Diaries’ 19 Hoffmann, E. T. A. 235 Høgel, Christian 218 Holbein, Hans 10 Holland, Peter 167, 232 Homer 3, 20, 23, 25, 28–47, 192, 204–8 Horace 108, 113 Horton, Julian 235 Hough, Graham 191, 237 Howard, Thomas, Lord Arundel 116 Howard-Hill, Trevor H. 229 Huet, Pierre-Daniel 65 Huizinga, Johan 20 Hullot-Kentor, Robert 235 Hume, David 155 Hunter, George K. 173, 228 Hurtado, Larry W. 51, 208–9 Hytner, Nicholas 167, 232 Iacomo della Lana 107, 111, 221 Ibn al-Nadīm, Muh.ammad 212 Ibn Yūh.annā, Rizq Allāh 65 Index librorum prohibitorum 23 Indizio, Giuseppe 219 Ingarden, Roman 178, 233 Inglese, Giorgio 220, 222 Ioppolo, Grace 230 Irwin, Robert 87, 217 Isaiah 60, 208 Isidore of Seville 113 Jaggard, William 162 Jahshiyārī, Abū ‘Abd Allāh 72, 75

James I, King of Great Britain 168 James, Henry 154 James, Montague Rhodes 6 Janko, Richard 207 Javadi, Iman 237 Jean, Duke of Berry 8 Jenkins, Harold 231 Jephcott, Edmund 235 Jerome 14, 59 Johannine Letters 49 John the Evangelist 51, 57, 60 Johnson, Samuel 155, 175, 232 Jonson, Ben 157, 159–60 Jordanus Nemorarius (de Nemore) 124 Jowett, William 215 Joyard, Anne-Alexandra 211–13 Joyce, James 191, 193, 200 Julius, Anthony 201, 238 Juvenal 113 Kafka, Franz 5 Kalīla wa Dimna (Kalila and Dimna) 65, 71 Kallas, Ellie 211 Keaney, John J. 205 Kelly, Henry Ansgar 223 Kelmscott Press 9 Kempe, Margery 6 Kenny, Neil 228 Kermode, Frank 232 Kinderman, William 234 Klein, Francesca 218 Klein, Giovanni Battista 218 Kleinheyer, Gerd 228 Knowles, Richard 229, 232 Kolneder, Walter 234 Kotsidou, Haritini 206 Kozintsev, Grigori 232 Krumpholz, Wenzel 179 Labarbe, Jules 207 La Boétie, Étienne de 227 La Charité, Raymond C. 225 Lamberton, Robert 205 Landino, Cristoforo 103, 120 Lane, Edward W. 70, 81, 84–5, 87, 212, 216–17 L’Angelier, Abel 226 Langland, William 154 Langlès, Louis-Mathieu 79

Index of names La Noue, François de 145 Lanza, Antonio 104 Lapo Gianni 108 Larzul, Sylvette 211, 217 Latini, Brunetto 99–100, 218 Lawrence, D. H. 23 Lazzarini, Lino 221 Legros, Alain 140, 226 Leichtentritt, Hugo 234 Lentin, Jérôme 211, 214–16 Leonardo da Vinci 5, 115–32, 224 Leoni, Pompeo 117 Leppert, Richard 235 Leumann, Manu 207 Lewis, Bernard 217 Lewis, Percy Wyndham 191, 195, 237 Lindon, John 221 Lippi, Caterina di Meo 130 Littmann, Enno 87, 217 Liveright, Horace 197, 199 Livy 7, 53 Locke, John 151 Loges, Marie des 149, 227 Lord, Albert 37–9, 206 Louis XIII , King of France 149 Lucas, Paul 63, 68–70 Luke the Evangelist 49, 51, 54–5, 57, 166, 208–9 Lyons, Malcolm C. 87, 217 Lyons, Ursula 87, 217 Maas, Paul 104, 221 Maber, Richard xii McCollough, Andrew 232 McCue, Jim 195, 200–2, 237–8 MacDonald, Dennis Ronald 208 MacDonald, Duncan B. 80–1, 213–16 McKenzie, Donald Francis 203, 238 McKinley, Mary 148, 227 McLeod, Randall 203 Macnaghten, Sir William Hay 82, 86, 216–17 Macnaughton, Jane 231 Magnien, Michel 226 Magnien-Simonin, Catherine 226 Magritte, René 11 Maguire, Laurie E. 230 Mahābhārata 43 Mahdi, Muhsin 78, 81, 86, 211, 213–17

249

Maillet, Benoît de 75, 78, 80, 83 Malachi 60 Malato, Enrico 222 Malherbe, François de 149 Malory, Thomas 13 Mandelbaum, Allen 99 Mandola, Eleonisia 223 Manfredi, Valerio Massimo 217 Mardrus, Jean-Charles 84–5, 87, 217 Marinoni, Augusto 118, 224 Mark the Evangelist 51, 54–5, 57, 60 Marshall, I. Howard 208 Martial 7, 51, 53–4, 209 Martin, Randall 230 Martini, Francesco di Giorgio 123, 125, 129 Martini, Luca 102–3, 220 Marullo, Michele 126 Marx, Adolph Bernhard 180, 183, 185–6, 234–6 Marx, Karl 4, 25 Marzolph, Ulrich 211–13, 216–17 Maskell, David 226–8 Masuccio Salernitano 121 Mascūdī, Abū ’l-Hasan 71, 212 Mathew, Nicholas 233 Matthew the Evangelist 51, 55, 57, 60, 209 Mazzoni, Luca 220 Mazzucchi, Andrea 222 Mecca, Angelo Eugenio 220 Melzi, Francesco 116–18 Mendelssohn Bartholdy, Felix 184 Mercuri, Roberto 223 Meyer, Anne 229 Mi’at layla wa Layla (‘101 Nights’) 72 Middleton, Thomas 19 Milani, Giuliano 100, 218–19 Millais, John Everett 10 Miller, Jonathan 232 Millet, Olivier 225 Milne, Alan Alexander 10 Milton, John 25 Minio-Paluello, Lorenzo 223 Misch, Ludwig 234 Mitchell, Stephen 42, 206 Moerbeke, William of 114, 223 Monachis, Dom de 79 (see Zākhūr, Rafā’īl Ant.ūn) Montaigne, Michel Eyquem de 116, 133–53, 225–8

250

Index of names

Montanari, Franco 206 Montinari, Mazzino 205 Moore, Edward 103, 220–1 Morgan, Teresa 204 Morris, Ian 207 Morris, William 9 Most, Glenn W. 205 Motolese, Matteo 224 Mozart, Wolfgang Amadeus 183 Muir, Kenneth 173, 228–9 Munday, Anthony 159 Mutschler, Fritz-Einer 207 Mynors, R. B. A. 210 Nadīm, Ibn 71 Naggiar, Mordechai 80 (see Najjār, Murdīkhāy) Nagy, Gregory 42, 44, 205–7 Najjār, Murdīkhāy 80 (see Naggiar, Mordechai) Nanni, Romano 224 Napoleon I, Emperor of the French 75, 77, 117 Nardelli, Jean-Fabrice 207 Nardi, Bruno 105, 221 Naya, Emmanuel 139, 226 Nestle, Eberhard 209 Nestle, Erwin 209 Newman, Daniel 216 Nietzsche, Friedrich 36, 205 Noble, Adrian 167 Norden, Eduard 221 North, Michael 198 Nünlist, René 204–6, 208 Nunn, Trevor 167 O’Brien, John 225, 227–8 O’Donnell, James J. 210 O’Neill, Michael xii Ogden, James 229 Ostasio da Polenta 97 Otho IV, Emperor 107 Ott, Claudia 87, 213, 217 Ovid 53 Pacioli, Luca 123–4 Paddison, Max 235 Paget, Sidney 10 Pahlmann, Bernhard 228

Palmarocchi, Roberto 218 Pannier, Jacques 227 Pantin, Isabelle 226 Panzera, Cristina 219 Papias (lexicographer) 113 Parker, Roger 178–9, 233 Parry, Milman 37–9, 42, 206 Pasquali, Giorgio 222 Paul the Apostle 48–51, 54, 61 Pavet de Courteille, Abel 212 Pavier, Thomas 162 Payne, John 85, 87, 217 Pedretti, Carlo 118, 224 Pellegrini, Paolo 223 Perotti, Niccolò 123–4 Persius 113 Peter Comestor 105 Pétis de la Croix, François 65–6, 68, 76, 85, 211 Petrarca, Francesco 21, 103, 116, 119, 122 Petrocchi, Giorgio 104–5, 109, 218, 220, 222 Petrucci Nardelli, Franca 219 Petrucci, Armando 219 Pfeiffer, Rudolf 204 Pherecides 150 Piattoli, Renato 218 Pieri degli Useppi da San Gimignano, ser 102 Piero della Francesca 123 Pisistratus 40–1 Pissarro, Esther 9 Pissarro, Lucien 9 Plato 17, 30, 177, 204, 207 Plautus 113 Pliny the Elder 7, 120, 123 Plummer, Charles 210 Plumptre, E. H. 96, 218 Pollard, Alfred William 158, 229 Pontani, Filippomaria 205 Pope, Alexander 20, 154, 156, 175, 194 Porphyry 31, 205 Porter, James I. 204 Portinari, Folco 96, 98 Pouillon, François 211 Pound, Ezra 6, 193–7, 200–3 Powell, Barry B. 207 Powis Mathers, Edward 87 Procaccioli, Paolo 224 Pucci, Antonio 122

Index of names Pulci, Luca 119, 122 Pulci, Luigi 121–2 Purkis, James 229–30 Quainton, Malcolm 225 Quinn, John 193–4, 197, 199, 201 Quinn, Julia 193 Quintilian 51–3, 151, 208 Quondam, Amedeo 221 Qurt.ī, Ibn Sa‘d 72 Rainey, Lawrence 195, 197–8, 200 Ramusio, Paolo 119 Rank, Louis Philippe 207 Rasmussen, Eric 232 Rat, Maurice 137 Reddick, Carrisa 236 Regnicoli, Laura 218 Reguig-Naya, Delphine 226 Reid, Ben 193–4 Reinhardt, Carl 82–3, 216 Rengakos, Antonios 207 Reynolds, Sir Joshua 21–2 Reynolds, Suzanne 223 Ricci, Pier Giorgio 218–19, 221 Ricciardelli, Fabrizio 219 Rice, Richard 233 Richards, Ivor Armstrong 191, 237 Richelieu, Armand Duplessis, Cardinal de 149 Ricks, Christopher 198, 200–2, 237–8 Riemann, Hugo 234 Rieu, Emile Victor 20 Risch, Ernst 207 Roberts, Colin H. 50, 208–9 Rocca, Luigi 220 Roddewig, Marcella 220 Rosenstock, Samy 11, 19 (see Tzara, Tristan; Samyro, S.; Ruia, Tristan) Rossi, Luca Carlo 222 Rossi, Luciano 222 Rossi, Massimiliano 224 Rossini, Gioachino 184, 233–4 Rufus, Lucius Varius 5 Ruia, Tristan (pseudonym of Rosenstock, Samy) 19 Russell, Alexander 69, 77, 212, 214–15 Russell, Donald A. 208 Russo, Emilio 220, 224

251

Ryan, Kiernan 228 Ryrie, Alec xii Sabbagh, Michel 77–8, 80 (see S.abbāgh, Mikhā’īl Ibn Niqūlā Ibn Ibrāhīm) S.abbāgh, Mikhā’īl Ibn Niqūlā Ibn Ibrāhīm 77–8, 80 (see Sabbagh, Michel) Saccenti, Mario 220 Sacy, Silvestre de 78–80 S.ahyūnī, Jibrā’īl 69 (see Sionita, Gabriel) Salles, Georges 222 Salt, Henry 82 Salutati, Coluccio 104, 221 Samyro, S. (pseudonym of Rosenstock, Samy) 19 Sand, George (pseudonym of Dupin, Armandine Aurore) 19 Sangallo, Giuliano da 129 Sanguineti, Federico 109–12, 222–3 Sankovitch, Tilde 225 Sappho 23 Sartori, Manuel 212 Saulnier, Verdun-Louis 137–40, 225 Saunders, Corinne 231 Sayce, Richard A. 226–8 Sayers, Dorothy L. 20 Schachter, Marc xii Schall, Karl 87 Schenker, Heinrich 177 Schiavone, Oscar 220 Schiff, Sydney 195–6, 202, 237 Schmalfeldt, Janet 184–8, 190, 235 Scholar, Richard 228 Schrader, Hermann 205 Schröder, Jan 228 Schubert, Franz 184 Schuchard, Ronald 237–8 Schumann, Robert 184 Schwab, Raymond 211 Scott, Jonathan 80 Scouten, Arthur H. 229 Seetzen, Ulrich 78–9, 215 Sforza family 120 Sforza, Francesco I, Duke of Milan 121 Sforza, Ludovico “Il Moro”, Duke of Milan 119, 123 Shakespeare, William 1, 14–16, 25, 154–75, 192, 228–32

252

Index of names

Shāwīsh, Diyūnīsiyūs 76–8, 80 (see Chavis, Dom Denis) Shayyāl, Jamāl al-Dīn 216 Shepard, Ernest Howard 10 Shepherd of Hermas 59–60 Sherman, Thomas 19 Sieur Louis de Conte (pseudonym of Clemens, Samuel) 19 Simcānī, Yūsuf Simcān 69 (see Assemani, Giuseppe Simone) Simonetta, Giovanni 120 Simonin, Michel 140, 226 Sionita, Gabriel 69 (see S.ah.yūnī, Jibrā’īl) Skeat, Theodore Cressy 50, 208–9 Small, Christopher 177–8, 190, 233–4 Smith, Grover 195 Spencer, Theodore 202 Spitzer, Michael 235 Squire, Sir John 191 Stabili, Francesco see Cecco d’Ascoli Stallbaum, Gottfried 205 Stanton, Graham N. 208 Stayer, Jayme 237 Strang, Veronica xiii Strowski, Fortunat 136, 140, 225 Strunk, Oliver 236 Supple, James 225 Symmachus, Quintus Aurelius 4 Taccola, Mariano di Jacopo, called 130 Tanturli, Giuliano 221 Tanūkhī 65 Taplin, Oliver 206 Tārīkh Mis. r (‘History of Egypt’) 72 Tarrête, Alexandre 226 Taruskin, Richard 233 Tauer, Felix 87, 214 Taylor, Gary 228 Taylor, Neil 229, 231 Terence 113 Tetel, Marcel 226 Thales 150 Thayer, Scofield 197 Theagenes of Rhegium 40 Theodoret of Cyrrhus 49–50, 208 Theodoric the Great 58 Thibaudet, Albert 137 Thiel, Helmut van 41–2, 206 Thomann, Johannes 211

Thomas, Sidney 229 Thompson, Ann 229, 231 Thousand and One Nights (Alf Layla wa Layla) 62–92, 211–17 Tiedemann, Rolf 235 Timmermans, Linda 227 Timothy 48–9 Timpanaro, Sebastiano 222 Tolkien, J. R. R. 5 Tonello, Elisabetta 220 Tournon, André 137, 139–40, 149, 225–6 Tovey, Donald Francis 234 Trivulzio family 116 Trollope, Anthony 10 Trousdale, Marion 229 Trovato, Paolo 105, 111, 220–3 Tucca, Plotius 5 Twain, Mark (pseudonym of Clemens, Samuel) 19, 26 Tzara, Tristan (pseudonym of Rosenstock, Samy) 11, 19 Uguccione of Pisa 113 Urkowitz, Steven 229, 231 Valensi, Lucette 211, 215 Valéry, Paul 174, 232 Valla, Giorgio 125, 129 Valturio, Roberto 119–20 Vande Moortele, Steven 234 Vandelli, Giuseppe 104 Varsy, Jean-Georges 80–1 (see Wārisī, Ibn Yūsuf (Yūh.annā)) Vecce, Carlo 224 Vermeulen, Urbain 213 Verrocchio, Andrea del 130 Vessey, Mark 210 Vickers, Sir Brian 229, 231 Vico, Giambattista 36, 205 Villey, Pierre 136–40, 225 Virgil 5, 18, 23, 25, 53, 101, 106–7 Vittoz, Roger 196 Voiture, Vincent 149 Volpi, Mirko 220 Wagner, Richard 198, 234 Walton, Benjamin 233 Wārisī, Ibn Yūsuf (Yūh.annā) 81 (see Varsy, Jean-Georges)

Index of names Warren, Michael 164, 228–31 Warren, Roger 166–7, 230, 232 Watson, Francis 209 Webb, James Francis 210 Webber, Edwin J. 223 Weber Nicholson, Shierry 235 Weber, William 234 Wedding Song of Meho Smailagić, The 39 Weil, Gustav 87 Weis, René 230 Wells, Stanley 228, 231 Werstine, Paul 229–30 West, Martin Litchfield 41–2, 44–5, 204–8 West, Stephanie 207 Weston, Jessie 192 Williams, Wes 228 Winckelmann, Friedhelm 209 Witte, Carl 103, 220 Wohlfs, Gerhard 219 Wolf, Friedrich August 36, 205 Woodward, Daniel H. 201

253

Woolf, Virginia 198–9 Wordsworth, William 154, 195 Wortley-Montague, Edward 77, 80 Yamanī, Ah.mad Ibn Muh.ammad Shīrwanī 79, 86, 215 Yeats, William Butler 10 Zaccarello, Michelangelo 222 Zakharia, Katia 81, 216 Zākhūr, Rafā’īl Ant.ūn 79 (see Monachis, Dom de) Zamponi, Stefano 218 Zeitlin, Jacob 227 Zeller, Wolfgang 228 Zeller-Lorenz, Barbara 228 Zenodotus 33, 205 Zetzel, James E. G. 205 Zola, Émile 24 Zorzi, Andrea 219 Zotenberg, Hermann 79, 82–3, 214, 216 Zwijnenberg, Robert 224

254

255

256

257

258