Papers on Quintilian and Ancient Declamation 0198836058, 9780198836056

Citation preview

PAPERS ON QUINTILIAN AND ANCIENT DECLAMATION

Frontispiece. Professor Michael Winterbottom, Corpus Christi Professor of Latin Emeritus, University of Oxford. Photograph courtesy of the Oxford Geology Group.

Papers on Quintilian and Ancient Declamation M I C HA E L WIN TE RBO TTO M Edited by

ANTONIO STRAMAGLIA with

F R A N C E S C A RO M A N A N O C C H I and

GIUSEPPE RUSSO

1

3

Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © Michael Winterbottom 2019 The moral rights of the author have been asserted First Edition published in 2019 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2018963938 ISBN 978 0 19 883605 6 Printed and bound by CPI Group (UK) Ltd, Croydon, CR0 4YY Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

Editor’s Introduction In the late spring of 1954, a promising Oxford undergraduate was working hard for the Hertford scholarship (the main prize then in Latin studies). He needed some extra tuition for Latin composition, but his tutor at Pembroke College was recovering from a serious illness and directed him to ‘a friend at St John’s’. This latter on one occasion . . . did what I gather his old tutor at Balliol (Roger Mynors) used to do, and put before me, after discussing my compositions, a piece of unfamiliar Latin: something from the Elder Seneca. I noted in my diary for 4 June: ‘In the evening I read about Roman declamation.’¹

The Oxford student—tall and slim, with mild but piercing blue eyes—was obviously M(ichael) W(interbottom). His tutor at St John’s College was, also obviously, Donald Russell; and the spark he lit in his pupil was to revolutionize a whole field of study. MW’s doctoral thesis was a commentary on Quintilian, Book Two: a key part of the Institutio in matters of rhetorical training. The work was successfully defended in 1962,² but remained unpublished for more than forty years: MW was by now devoting his energies to a new OCT edition of Quintilian. He produced it—with astonishing speed and skill, considering the bulk of his task—in 1970, together with a companion volume of Problems in Quintilian, and after a string of preliminary studies. This early set of publications immediately commanded praise and respect. It fully illustrates MW’s ‘holistic’ view of scholarship: research on the manuscript tradition (including its human subjects, e.g. Almeloveen) is tightly intertwined not only with the editing of a text, but also with a constant effort to elucidate and contextualize what that text means. With his characteristic candour, MW has recently summarized his creed: Ever since childhood, I have always been concerned to understand the meaning of what I read, and I am not much troubled when I am told by critics that the Author is dead and his meaning a matter for the Reader to decide. I feel in my naïve way that classical prose writers meant something definite by what they wrote, and that, if I do not understand it, that is either my fault or that of the scribes.³

¹ Winterbottom per litteras (28 August 2017). ² See here p. 218 n. [3]. ³ Winterbottom (2017c), 403 (my italics). For all abbreviations see the general bibliography below, pp. 351 61; articles and chapters reprinted in this book are referred to as A.1, A.2 . . . , book reviews as R.1, R.2 . . . For full details of MW’s publications alluded to throughout these pages, see the complete list below, pp.  .

viii

Editor’s Introduction

Quintilian would never subsequently fade from MW’s horizon; nor would the Latin texts of the Anglo-Saxon Middle Ages, the object—since MW’s early years—not of a side interest but of a whole ‘parallel career’, deserving a survey in its own right.⁴ By the early 1970s, however, MW (then Tutorial Fellow in classics at Worcester College, Oxford [1967–92]) was working intensively in the field that had aroused his interest since his undergraduate days: declamation. The study of Roman declamation, in particular, was not so much dormant as nearly non-existent in those years. The texts themselves were often barely intelligible: some of the relevant editions (any explicit mention may charitably be dispensed with) were among the most defective that classical scholarship has ever produced; and only a few aids were available for the interpretation of the genre (most notably Stanley Bonner’s evergreen Roman Declamation (1949)). A drastic and beneficial change was effected by four great scholars: Lennart Håkanson, Donald Russell, D. R. Shackleton Bailey, and MW; by 1989, thanks to them, first-rate editions of all the main texts in the field of Roman declamation, and a better understanding of the whole genre, were at last at hand. Unlike Håkanson and Shackleton Bailey, MW did not concentrate his work in declamation ‘only’ on ecdotic tasks: his Elder Seneca for Loeb (1974), and the massive commentary accompanying his edition of the ‘Quintilianic’ Minor Declamations (1984), splendidly showcase his gift for deep and clear elucidation—even of the most difficult and corrupt texts. In the same years, in a number of seminal papers he set out to shed fresh light on the relationship between rhetorical precepts (Quintilian’s in primis) on the one hand, declamatory theory and practice on the other; and in Russell’s footsteps he investigated the interaction between Greek and Roman declamation—the book on Sopatros (1988, with Doreen Innes) being the largest, but by no means the only, product of this effort. What is really striking, throughout these researches, is MW’s unprecedented breadth of view. He has written on declamation from Gorgias up to Ennodius, singling out (mostly for the first time) constants and variables over the centuries. This he has done thanks to his admirable learning, but also to his being immune to all the stock assumptions which would have hampered progress. See for instance his words on the ‘early stages’ of declamation in Rome: It was not that declamation somehow became more important in the course of the first century . Our impression that it does is largely a delusion, re sulting from the accidents of our evidence. Declamation will have come to Rome with the Greek teachers who brought rhetoric there in the second century . . . . [It] did not increase in importance in the first century: it merely remained

⁴ This will be given in a volume containing a selection of MW’s medieval papers, to be edited by Roberto Gamberini and published by SISMEL (Florence).

Editor’s Introduction

ix

important, and perhaps became, in schools less austere than Quintilian’s, more extravagant in conception.⁵

Without MW’s work, our whole understanding of ancient declamation would now be much more narrow and superficial. Between the 1970s and the 1990s, the indefatigable scholar—in spite of his ever growing teaching duties, to which we owe a valuable anthology of Roman Declamation (1980)—found time for fundamental research in various other domains. First of all, he consolidated his reputation as a specialist in manuscript traditions and critical editions, with his OCT texts of Tacitus (Opera minora, 1975) and Cicero (De officiis, 1994). This was ‘obviously’ accompanied by a number of papers and book reviews (on which latter see below); but special mention should be made of the many entries MW wrote for the standard work on Texts and Transmission edited by Leighton Reynolds in 1983. For decades now, a student’s first approach to the textual tradition of many a Roman author or work has—beneficially—been, more often than not, one of the admirable surveys contributed by MW to this book. Ancient literary criticism is another recurrent field of study for MW: the anthologies he prepared with Donald Russell have fully deserved to be standard since their publication (1972, 1989). More occasional—but no less serious—interest has been lavished on a number of authors, mostly for textual and exegetical issues: Virgil in primis, but also Lucretius, Ovid, Apuleius, Cyprian, Ambrose . . . And the medieval favourites have always been there of course, with William of Malmesbury in the front row. Meanwhile MW was appointed Corpus Christi Professor of Latin at Oxford (1992), a tribute to his rank among the leading living classicists. In 2000, one year before his retirement, the present writer (very timidly) contacted him, to involve him in the Cassino project of re-editing the 19 Major Declamations ascribed to Quintilian in individual volumes, with translation and commentary (1999–). MW declined taking on one or more declamations himself, but generously accepted to comment on the single volumes. The amount and quality of his contributions to each book in this collection (from 2005 on) call for special emphasis. It is very often ‘Winterbottom per litteras’ who finally heals or gives sense to a passage vexed for centuries, throughout these exceptionally difficult texts. All the same, the Maiores remained for several years only a side interest for MW; but in general, declamation and ancient rhetorical texts (with their teaching procedures) came to the forefront of his scholarship as a classicist⁶ throughout the 2000s—an activity more intensive than ever since his ⁵ Winterbottom (1982), 254 6 (= A.5 below, pp. 78 9); my italics. ⁶ No idle qualification: in the same years, MW’s ‘medieval self ’ (Freud might have talked of a Doppelgänger) publishes no less than five books and seventeen articles and an enormous number of book reviews.

x

Editor’s Introduction

retirement. In 2006 a revised version of his D.Phil. dissertation on Quintilian, Book Two was finally published as a book (in collaboration with Tobias Reinhardt); in 2008 his major contribution to the monumental commentary on Cicero’s De oratore (begun by Leeman and Pinkster) appeared; and his papers over these years display an increasingly wider range of issues and approaches. More will be said on this presently; here I should point out MW’s commitment to the recovery and publication of the Nachlass of his friend, the great Swedish scholar Lennart Håkanson, who tragically died in his prime in 1987.⁷ Nothing has been said so far of another hallmark of MW’s scholarship: his countless book reviews. Over the decades, they all demonstrate an unprejudiced—and sometimes memorable⁸—candour in assessing the reviewed author’s merits and shortcomings: something quite hard to come by, these days. What is more, they always offer some acute new insight to the reader: both on the specific subject of the book under review, and on broader— ecdotic, most often—methodological issues. MW’s ‘collected reviews would serve in themselves as a manual of editing’, Michael Reeve wrote some years ago;⁹ few will disagree. Vis à vis such a broad and varied array of publications, this book offers a selection of papers from two especially representative and intrinsically connected fields of interest. The choice is primarily based, of course, on the scholarly ‘weight’ of the single items; but the place and circumstances of publication have also played a role: many of MW’s most acute contributions originally appeared in conference proceedings, Festschriften, and rare periodicals—often hard to find even in leading libraries—or embedded in reviews that even modern search tools may easily miss. It seemed appropriate to give special consideration to such materials. The author has obviously taken part in the selection process. He decided that any items on Quintilian’s text and transmission prior to his edition (1970) should not be included, for that edition took full account of them; and he excluded any non-specialist pieces.¹⁰ What this book does include, as for Quintilian, are papers on some key aspects of the Institutio oratoria—regarding morals, style, structure, rhetorical technicalities, auctores (A.1, 3, 5, 8, 9, 13, 18); and the more recent contributions on Quintilian’s text (A.15, 19; R.2–4, 7–8). ⁷ On this demanding enterprise, carried out en équipe, see the editors’ prefaces in Håkanson (2014) and (2016). MW has provided a sensible appraisal of Håkanson’s scholarship (in Håkanson (2016),  ), and has himself brought to publication the most complicated piece of his Nachlass (Håkanson Winterbottom (2015)). ⁸ Cf. e.g. the opening words of Winterbottom (1978), 685 (= R.3 below, p. 322): ‘These two volumes . . . follow closely upon the first, and they share the merits and demerits of their predecessor. The text makes no pretension to novelty. The translation is fluent and generally accurate. The notes are informative. The apparatus criticus is a disaster.’ ⁹ Reeve (2000), 204 n. 51. ¹⁰ Such as Winterbottom (1985) and (1997).

Editor’s Introduction

xi

In the field of Greek and Roman declamation, this book is more inclusive. Only a few items have been left out, mostly long papers involving textual (re-)editions.¹¹ All other relevant articles and reviews have been reprinted here, on authors and topics ranging from classical Greece to the Latin Middle Ages (A.2, 4, 6–8, 14, 16, 20–3; R.1, 5–6, 9–12); room has been made also for the introduction to the 1988 book on Sopatros (A.10): this impressively learned and wide-ranging piece makes indispensable reading for anyone working on ancient declamation. Finally, the overall scope of the volume suggested the inclusion of two important short papers on rhetorical terms and concepts (A.11–12); and a brilliant survey of some striking novelties on the ancient rhetorical curriculum (A.17). Those who are familiar with MW’s scholarship may wonder if one of his basic tenets has been given sufficient consideration in this book: I have a few rigid principles in life, but one is never to speak or write on a Latin subject without mentioning prose rhythm.¹²

No specific article or book review on prose rhythm is included here, but each selected item does contain at least a case in point, and some papers feature detailed discussions.¹³ Those crystal-clear pages make one regret that MW has more frequently confined himself to brief mentions: his lucidity would have been particularly welcome in this field, as important as it is difficult (and nowadays neglected). All items are here reprinted according to uniform editorial guidelines, and in the process misprints have been removed, OLD has been referred to throughout according to the second edition (2012), a few formal adjustments (e.g. in cross-references) have been made, and occasional clarifications or references to new standard editions have been entered (in square brackets). The author has also worked in a number of addenda or corrigenda to some papers, mostly at their end, when some crucial point had to be made or a recent bibliographical item stood out for its relevance. In general, however, no attempt has been made at systematic updating: this would have implied rewriting the contributions, uprooting them from the historical and intellectual context in which, and for which, they were conceived. To conclude, something must be said about the last paper in the present collection (A.24). This brilliant new assessment of the manuscript tradition of the Major Declamations has been written expressly for this book, and it results from MW’s current main commitment: a Loeb edition of the Maiores. He is officially in charge of the translation and part of the notes (I am handling the

¹¹ Håkanson Winterbottom (2015); Winterbottom (2017d) and (2018). ¹² Winterbottom (2017c), 410. ¹³ See e.g. Winterbottom (1983a), 59 ff. (= A.7 below, pp. 105 ff.); (2017b), 151 (= A.22 below, pp. 271 2).

xii

Editor’s Introduction

Latin text, Biagio Santorelli the rest of the notes and the introduction(s)); but his ‘holistic’ approach to scholarship has never changed, so his work on the translation has soon given rise to repeated discussions of loci critici, dozens of (always astute, often decisive) conjectures, and an unprejudiced approach to the transmission of the text. Being involved in such a dialogue with such a scholar—and man—is a unique experience, still ongoing. But there is something for which MW has to turn elsewhere: who might revise his translation of these tricky and twisted Latin texts? Some special help is called for again, as in 1954 . . . Well, the old St John’s scholar is still there, as learned and acute as ever, ready to vet his former pupil’s translations—and contribute some formidable conjectures of his own. All this is taking place over sixty years since MW’s (b. 1934) and Donald Russell’s (b. 1920) first session on Roman declamation. Friendship may sometimes defy Nature’s laws. * * * This book would have never been produced without the unselfish and enthusiastic help constantly provided by Francesca Nocchi and Giuseppe Russo: my deep gratitude goes to them both. Special thanks are also due to Stephen Harrison, who facilitated and guided contacts with OUP, and was ever prompt with advice and support; and to OUP itself, for accepting and felicitously bringing to publication an anything but easy book. In recent years I have had the privilege of meeting Donald Russell, enjoying his generosity, and profiting from his advice also in relation to this book: it is a pleasure to thank him most warmly for all this. Auctori amicoque carissimo Michaeli, qui semper mihi praesto fuit in hoc opere absolvendo, postremas reddo easque maximas gratias: sit hic libellus longae nostrae eximiaeque amicitiae pignus. A. S. Bari June 2018

Publications of Michael Winterbottom * = included in this volume. Periodicals are abbreviated according to L’Année philologique, whenever applicable (add JML = Journal of Medieval Latin).

BOOKS 1. M. Fabi Quintiliani Institutionis oratoriae libri duodecim, 2 vols. (Oxford, 1970). 2. Problems in Quintilian (London, 1970). 3. & D. A. Russell, Ancient Literary Criticism (Oxford, 1972). 4. Three Lives of English Saints (Toronto, 1972). 5. The Elder Seneca. Declamations, 2 vols. (Cambridge [Mass.] and London, 1974). 6. & R. M. Ogilvie, Cornelii Taciti Opera minora (Oxford, 1975). 7. Gildas. The Ruin of Britain and Other Works (London and Chichester, 1978).¹ 8. Roman Declamation. Extracts Edited with Commentary (Bristol, 1980). 9. The Minor Declamations Ascribed to Quintilian (Berlin and New York, 1984). 10. William of Malmesbury. On Lamentations (Turnhout, 2013). 11. & R. M. Thomson, William of Malmesbury. The Miracles of the Blessed Virgin Mary (Woodbridge, 2015).² *12. & D. Innes, Sopatros the Rhetor, with an introduction by M. Winterbottom (London, 1988) (*pp. 1–20). 13. & D. A. Russell, Classical Literary Criticism (Oxford, 1989). 14. & M. Brett, C. N. L. Brooke, Hugh the Chanter. The History of the Church of York 1066–1127 (Oxford, 1990). 15. & M. Lapidge, Wulfstan of Winchester. Life of St Æthelwold (Oxford, 1991). 16. M. Tulli Ciceronis De officiis (Oxford, 1994). 17. & R. A. B. Mynors, R. M. Thomson, William of Malmesbury. Gesta regum Anglorum, vol. 1 (Oxford, 1998). 18. & R. M. Thomson, William of Malmesbury. Saints’ Lives (Oxford, 2002). 19. & T. Reinhardt, Quintilian. Institutio oratoria. Book 2 (Oxford, 2006).

¹ Reissued in 2002 with a new bibliographical foreword. ² Reissued in 2017 (paperback).

xiv

Publications of Michael Winterbottom

20. William of Malmesbury. Gesta pontificum Anglorum, vol. 1 (Oxford, 2007).³ 21. & J. Wisse, E. Fantham, M. Tullius Cicero. De oratore libri III, vol. 5 (Heidelberg, 2008). 22. & R. M. Thomson, Willelmi Meldunensis monachi Liber super explanationem Lamentationum Ieremiae prophetae (Turnhout, 2011). 23. & M. Lapidge, The Early Lives of St Dunstan (Oxford, 2012).

A R T I C LE S A N D CH A P T E R S IN MISCELLANEOUS V OLUME S 1. 2. 3. 4. *5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. *18. 19. 20. *21.

‘Almeloveen’s manuscript of Quintilian’, CR  12 (1962), 121–2. ‘The textual tradition of Quintilian 10.1.46 f.’, CQ  12 (1962), 169–75. ‘Quintilian, v. 10. 91’, CR  14 (1964), 14. ‘More about Almeloveen’, CR  14 (1964), 243. ‘Quintilian and the vir bonus’, JRS 54 (1964), 90–7. ‘Some problems in Quintilian Book Two’, Philologus 108 (1964), 119–27. ‘The beginning of Quintilian’s Institutio’, CQ  17 (1967), 123–7. ‘Quintilian, . 1. 3’, CR  17 (1967), 264. ‘Quintilian and Boethius’, BICS 14 (1967), 83. ‘Fifteenth-century manuscripts of Quintilian’, CQ  17 (1967), 339–69. ‘The style of Æthelweard’, MAev 36 (1967), 109–18. ‘On the Hisperica famina’, Celtica 8 (1968), 126–39. Revision of the Dialogus, in Tacitus, vol. 1 (London and Cambridge [Mass.], 1970), 217–347. Various contributions to N. G. L. Hammond and H. H. Scullard (eds.), Oxford Classical Dictionary, 2nd edn. (Oxford, 1970). ‘Six conjectures’, CR  22 (1972), 11–12. ‘The transmission of Tacitus’ Dialogus’, Philologus 116 (1972), 114–28. ‘Three lives of Saint Ethelwold’, MAev 41 (1972), 191–201. ‘Problems in the Elder Seneca’, BICS 21 (1974), 20–42. ‘The preface of Gildas’ De excidio’, Transactions of the Honourable Society of Cymmrodorion. Sessions 1974 and 1975, 277–87. ‘On epitrochasmos’, Glotta 53 (1975), 297–8. ‘Quintilian and rhetoric’, in T. A. Dorey (ed.), Empire and Aftermath. Silver Latin II (London, 1975), 79–97.

³ ‘with the assistance of R. M. Thomson’. Vol. 2 (Introduction and Commentary) was also published in 2007; the title page read: ‘by R. M. Thomson with the assistance of M. Winter bottom’.

Publications of Michael Winterbottom 22. 23. 24. 25. 26. 27. 28. 29. 30. *31. *32. 33. *34.

*35. *36. 37. 38. 39.

40. 41. 42. 43. *44. *45.

46. 47. 48. 49. 50.

xv

‘The manuscript tradition of Tacitus’ Germania’, CPh 70 (1975), 1–7. ‘Columbanus and Gildas’, VChr 30 (1976), 310–17. ‘Fiery particles’, CQ  26 (1976), 317–18. ‘Notes on the text of Gildas’, JThS  27 (1976), 132–40. ‘Variations on a nautical theme’, Hermathena 120 (1976), 55–8. ‘Virgil and the confiscations’, G&R  23 (1976), 55–9. ‘A “Celtic” hyperbaton?’, Bulletin of the Board of Celtic Studies 27 (1977), 207–12. ‘Aldhelm’s prose style and its origins’, ASE 6 (1977), 39–76. ‘The other Virgil’, BICS 25 (1978), 146–56. ‘The text of Sulpicius Victor’, BICS 26 (1979), 62–6. ‘Cicero and the Silver Age’, in W. Ludwig (ed.), Éloquence et rhétorique chez Cicéron (Vandoeuvres and Geneva, 1982), 237–66 (‘Discussion’, 267–74). ‘Literary criticism’, in The Cambridge History of Classical Literature, vol. 2: Latin Literature (Cambridge, 1982), 33–50. ‘Schoolroom and courtroom’, in B. Vickers (ed.), Rhetoric Revalued. Papers from the International Society for the History of Rhetoric (Binghamton [NY], 1982), 59–70. ‘Declamation, Greek and Latin’, in A. Ceresa-Gastaldo (ed.), Ars rhetorica antica e nuova (Genoa, 1983), 57–76. ‘Quintilian and declamation’, in Hommages à Jean Cousin (Paris, 1983), 225–35. Various contributions to L. D. Reynolds (ed.), Texts and Transmission. A Survey of the Latin Classics (Oxford, 1983; corr. repr. 1986). ‘The Roman orator and his education’, Akroterion 30 (1985), 53–7. ‘Mankind and other animals: the Georgics’, in R. A. Cardwell and J. Hamilton (eds.), Virgil in a Cultural Tradition. Essays to Celebrate the Bimillennium (Nottingham, 1986), 1–16. ‘Tot incassum fusos patiere labores?’, CQ  36 (1986), 545–6. ‘Notes on the Life of Edward the Confessor’, MAev 56 (1987), 82–4. ‘Pelagiana’, JThS  38 (1987), 106–29. ‘The Life of Christina of Markyate’, AB 105 (1987), 281–7. ‘Quintiliano (M. Fabius Quintilianus)’, in Enciclopedia Virgiliana, vol. 4 (Rome, 1988), 374–6. ‘Cicero and the Middle Style’, in J. Diggle, J. B. Hall, and H. D. Jocelyn (eds.), Studies in Latin Literature and its Tradition in Honour of C. O. Brink (Cambridge, 1989), 125–31. ‘Speaking of the gods’, G&R  36 (1989), 33–41. ‘New light on the X tradition of Cicero’s De officiis’, MD 24 (1990), 135–41. ‘Roger Aubrey Baskerville Mynors’, PBA 80 (1991), 371–401. ‘Aeneas and the idea of Troy’, PVS 21 (1993), 17–34. ‘The transmission of Cicero’s De officiis’, CQ  43 (1993), 215–42.

xvi

Publications of Michael Winterbottom

51. ‘Conjectures on some insular texts’, in D. Conso, N. Fick, and B. Poulle (eds.), Mélanges François Kerlouégan (Paris, 1994), 667–72. *52. ‘On impulse’, in D. Innes, H. Hine, and C. Pelling (eds.), Ethics and Rhetoric. Classical Essays for Donald Russell on his Seventy-Fifth Birthday (Oxford, 1995), 313–22. 53. ‘The Gesta regum of William of Malmesbury’, JML 5 (1995), 158–73. 54. ‘The O.C.T. De officiis: a postscript’, CQ  45 (1995), 265–6. 55. ‘The text of Ambrose’s De officiis’, JThS  46 (1995), 559–66. 56. & S. J. Harrison, ‘The new passage of Tiberius Claudius Donatus’, CQ  45 (1995), 547–50. 57. ‘The integri of Cicero’s Topica’, CQ  46 (1996), 403–10. 58. ‘Introduzione’, in S. Corsi, Marco Fabio Quintiliano. La formazione dell’oratore, vol. 1 (Milan, 1997), 5–26. 59. ‘De studiis Latinis Oxoniensibus’, VoxLat 33 (1997), 542–9. *60. ‘Quintilian the moralist’, in T. Albaladejo, E. del Río, and J. A. Caballero (eds.), Quintiliano: Historia y Actualidad de la Rétorica, vol. 1 (Logroño, 1998), 317–34. 61. ‘Tacitus, Dialogus 13.4’, CQ  49 (1999), 338. *62. ‘An emendation in Calpurnius Flaccus’, CQ  49 (1999), 338–9. 63. ‘In praise of Raphael Regius’, in S. Döpp (ed.), Antike Rhetorik und ihre Rezeption. Symposion zu Ehren von Professor Dr. Carl Joachim Classen (Stuttgart, 1999), 99–116. 64. ‘Notes on William of Poitiers’, JML 9 (1999), 121–30. 65. & J. J. Murphy, ‘Raffaele Regio’s 1492 Quaestio doubting Cicero’s authorship of the Rhetorica ad Herennium’, Rhetorica 17 (1999), 77–87. 66. ‘Three emendations in Columella’, CQ  49 (1999), 633–4. 67. ‘Lucretius 5.845–854’, Hermes 128 (2000), 505–6. *68. ‘More problems in Quintilian’, BICS 44 (2000), 167–77. 69. ‘The earliest Life of St Dunstan’, SCI 19 (2000), 163–79. 70. ‘A new passage of William of Malmesbury’s Gesta pontificum’, JML 11 (2001), 50–9. 71. ‘Leighton Durham Reynolds’, PBA 111 (2001), 659–76. 72. ‘Returning to Tacitus’ Dialogus’, in C. W. Wooten (ed.), The Orator in Action and Theory in Greece and Rome. Essays in Honor of George A. Kennedy (Leiden, Boston, and Cologne, 2001), 137–55. 73. & S. J. Harrison, ‘The prologue to Apuleius’ Metamorphoses’, in A. Kahane and A. Laird (eds.), A Companion to the Prologue of Apuleius’ Metamorphoses (Oxford, 2001), 9–15. 74. ‘William of Malmesbury versificus’, in S. Echard and G. R. Wieland (eds.), Anglo-Latin and its Heritage. Essays in Honour of A. G. Rigg on his 64th Birthday (Turnhout, 2001), 109–27. 75. ‘Believing the Pro Marcello’, in J. E. Miller, C. Damon, and K. S. Myers (eds.), Vertis in usum. Studies in Honor of Edward Courtney (Munich, 2002), 24–38.

Publications of Michael Winterbottom

xvii

76. & M. Deufert, J. F. Gaertner, ‘Critical notes on the Heroides’, Hermes 130 (2002), 502–6. *77. ‘Ennodius, Dictio 21’, in B.-J. and J.-P. Schröder (eds.), Studium declamatorium. Untersuchungen zu Schulübungen und Prunkreden von der Antike bis zur Neuzeit (Munich and Leipzig, 2003), 275–88. 78. ‘The Language of William of Malmesbury’, in C. J. Mews, C. J. Nederman, and R. M. Thomson (eds.), Rhetoric and Renewal in the Latin West 1100–1540. Essays in Honour of John O. Ward (Turnhout, 2003), 129–47. 79. ‘Grillius on Cicero’s De inventione’, CQ  54 (2004), 592–605. 80. ‘Perorations’, in J. G. F. Powell and J. Paterson (eds.), Cicero the Advocate (Oxford, 2004), 215–30. *81. ‘Something new out of Armenia’, Letras clássicas 8 (2004), 111–28. 82. ‘An edition of Faricius, Vita S. Aldhelmi’, JML 15 (2005), 93–147. *83. ‘Approaching the end: Quintilian 12.11’, AClass 48 (2005), 175–83. 84. ‘Faricius of Arezzo’s Life of St Aldhelm’, in K. O’Brien O’Keeffe and A. Orchard (eds.), Latin Learning and English Lore. Studies in AngloSaxon Literature for Michael Lapidge, vol. 1 (Toronto, Buffalo and London, 2005), 109–31. *85. ‘Declamation and philosophy’, Classica (Brasil) 19 (2006), 74–82. *86. ‘Quintilian 12.11.11–12’, CQ  56 (2006), 324–5. 87. ‘Cyprian’s Ad Donatum’, in S. Swain, S. J. Harrison, and J. Elsner (eds.), Severan Culture (Cambridge, 2007), 190–8. 88. ‘Bede’s castella’, Quaestio Insularis 10 (2009), 1–7. 89. ‘Conversations in Bede’s Historia ecclesiastica’, in E. Dickey and A. Chahoud (eds.), Colloquial and Literary Latin (Cambridge, 2010), 419–30. 90. ‘William of Malmesbury and the Normans’, JML 20 (2010), 70–7. 91. ‘Bede’s homily on Benedict Bishop (Hom. 1.13)’, JML 21 (2011), 35–51. 92. ‘On ancient prose rhythm: the story of the dichoreus’, in D. Obbink and R. Rutherford (eds.), Culture in Pieces. Essays on Ancient Texts in Honour of Peter Parsons (Oxford, 2011), 262–76. 93. ‘De vita patris’, in D. Damschen and A. Heil (eds.), Brill’s Companion to Seneca (Leiden and Boston, 2014), 695. 94. ‘Moving the goal posts: the re-writing of medieval Latin prose texts’, Ars edendi Lecture Series 3 (2014), 29–48. *95. ‘William of Malmesbury’s work on the Declamationes maiores’, S&T 12 (2014), 261–76. 96. ‘The earliest passion of St Alban’, InvLuc 37 (2015), 113–27. 97. & †L. Håkanson, ‘Tribunus Marianus’, in L. Del Corso, F. De Vivo, and A. Stramaglia (eds.), Nel segno del testo. Edizioni, materiali e studi per Oronzo Pecere (Florence, 2015), 61–90.

xviii

Publications of Michael Winterbottom

98. ‘Lennart Håkanson: der Mensch, der Gelehrte’, in L. Håkanson, Unveröffentlichte Schriften, vol. 2: Kritischer Kommentar zu Seneca Maior, Controversiae, Buch I, ed. by F. Citti, B. Santorelli, and A. Stramaglia (Berlin and Boston, 2016), –. 99. ‘The style of Ælnoth’, in M. Münster-Swendsen, T. K. Heebøll-Holm, and S. Olsen Sønnesyn (eds.), Historical and Intellectual Culture in the Long Twelfth Century: The Scandinavian Connection (Durham and Toronto, 2016), 119–30. 100. ‘Karsten Friis-Jensen’s preliminary findings towards a new edition of Sven Aggesen’, ibid., 295–317. 101. ‘Text and transmission of some Bedan texts’, MLatJb 52 (2017), 445–59. *102. ‘The editors of Calpurnius Flaccus’, in M. Dinter, C. Guérin, and M. Martinho (eds.), Reading Roman Declamation. Calpurnius Flaccus (Berlin and Boston, 2017), 141–60. 103. ‘The pleasures of editing’, RHT  12 (2017), 393–413. 104. ‘The Tribunus Marianus and the development of the cursus’, in P. Chiesa, A. M. Fagnoni, and R. E. Guglielmetti (eds.), Ingenio facilis. Per Giovanni Orlandi (1938–2007) (Florence, 2017), 231–47. 105. ‘The vocabulary of William of Malmesbury’, Aevum 91 (2017), 377–409. 106. ‘Words, words, words . . . ’, in R. M. Thomson, E. Dolmans, and E. A. Winkler (eds.), Discovering William of Malmesbury (Woodbridge, 2017), 203–18. 107. ‘Cyrus, On the Differentiation of Issues’, S&T 16 (2018), 209–46. *108. ‘The words of the master’, Maia 70 (2018), 73–83. *109. ‘The manuscript tradition of [Quintilian]’s Major Declamations: a new approach’, in this book. 110. ‘Beginning a history’, JML 29 (2019), 101–21. 111. ‘The manuscripts of Berengar of Poitiers’, MLatJb 54 (2019), 157–61. 112. ‘Pope Urban’s speech at Clermont. William of Malmesbury, Gesta Regum Anglorum 347’, MLatJb 54 (2019), 162–7. 113. ‘Notes on the text of the Major Declamations’, MD 82 (2019) [forthcoming].

SE LE C T ED R E V IEWS 1. A. E. Douglas, M. Tulli Ciceronis Brutus (Oxford, 1966), CR  17 (1967), 301–3.⁴ 2. A. Campbell, Aethelwulf De abbatibus (Oxford, 1967), MAev 38 (1969), 60–4.

⁴ This is the first in a long series of reviews in CR, only a few of which are listed here.

Publications of Michael Winterbottom

xix

3. G. Luck, Untersuchungen zur Textgeschichte Ovids (Heidelberg, 1969), CR  21 (1971), 208–9. 4. J. W. Smit, Studies in the Language and Style of Columba the Younger (Columbanus) (Amsterdam, 1971), MAev 41 (1972), 243–4. 5. H. Weiskopf, P. Corneli Taciti Annalium libri XI–XII (Vienna, Cologne, and Graz, 1973), CPh 76 (1975), 283–4. 6. M. W. Herren, The Hisperica famina. : The A-Text (Toronto, 1974), MAev 45 (1976), 105–9. 7. T. Janson, Prose Rhythm in Medieval Latin from the 9th to the 13th Century (Stockholm, 1975), MAev 45 (1976), 298–300. 8. S. Usher, Dionysius of Halicarnassus. The Critical Essays, vol. 1 (Cambridge [Mass.] and London, 1974), CR  26 (1976), 173–4. *9. L. Håkanson, Textkritische Studien zu den grösseren pseudoquintilianischen Deklamationen (Lund, 1974), CR  26 (1976), 276. *10. J. Cousin, Quintilien. Institution oratoire, Tome  (Paris, 1975) and Recherches sur Quintilien (Paris, 1975), Gnomon 49 (1977), 574–9. 11. J. N. Hillgarth, Sancti Iuliani Toletanae sedis episcopi Opera, Pars  (Turnhout, 1976), JThS  28 (1977), 571–4.⁵ *12. J. Cousin, Quintilien. Institution oratoire, Tomes – (Paris, 1976), Gnomon 50 (1978), 685–7. 13. G. Pompella, Francisci Robortelli Utinensis De arte sive ratione corrigendi antiquorum libros disputatio (Naples, 1975), CR  28 (1978), 197–8. *14. J. Cousin, Quintilien. Institution oratoire, Tome  (Paris, 1977), Gnomon 51 (1979), 388–9. *15. S. F. Bonner, Education in Ancient Rome: From the Elder Cato to the Younger Pliny (London, 1977), CR  29 (1979), 73–4. *16. L. A. Sussman, The Elder Seneca (Leiden, 1978), CR  29 (1979), 231–2. *17. J. Cousin, Quintilien. Institution oratoire, Tomes – (Paris, 1978; 1979), Gnomon 52 (1980), 785–6. *18. J. Cousin, Quintilien. Institution oratoire, Tome  (Paris, 1980), Gnomon 53 (1981), 197–9. 19. M. W. Herren, The Hisperica famina. : Related Poems (Toronto, 1987), Peritia 6–7 (1987–8), 331–2. 20. G. Achard, Rhétorique à Herennius (Paris, 1989), Gnomon 63 (1991), 459–61. *21. L. Håkanson, L. Annaeus Seneca Maior. Oratorum et rhetorum sententiae, divisiones, colores (Leipzig, 1989), CR  41 (1991), 338–40. 22. J. B. Hall, Ioannis Saresberiensis Metalogicon (Turnhout, 1991), JEH 43 (1992), 470–1.⁶

⁵ This is the first in a long series of reviews of patristic texts in this journal, not listed here. ⁶ Other reviews of Christian texts appear in later issues of this journal.

xx

Publications of Michael Winterbottom

23. K. M. Coleman, J. Diggle, J. B. Hall, and H. D. Jocelyn (eds.), F. R. D. Goodyear, Papers on Latin Literature (London, 1992), CR  44 (1994), 196–8. 24. R. Granatelli, Apollodori Pergameni ac Theodori Gadarei testimonia et fragmenta (Rome, 1991), CR  44 (1994), 203–4. 25. M. Irvine, The Making of Textual Culture: Grammatica and Literary Theory, 350–1000 (Cambridge, 1994), JML 5 (1995), 273–6. 26. D. Lassandro, XII Panegyrici Latini (Turin, 1992), Gnomon 67 (1995), 560–1. 27. O. Prinz, Die Kosmographie des Aethicus (Munich, 1993), Peritia 9 (1995), 430–2. *28. L. A. Sussman, The Declamations of Calpurnius Flaccus. Text, Translation, and Commentary (Leiden, New York, and Cologne, 1994), CR  45 (1995), 40–2. 29. J. Stevenson, The Laterculus Malalianus and the School of Archbishop Theodore (Cambridge, 1995), Notes and Queries  43 (1996), 457–9. 30. G. Di Maria, Marci Tulli Ciceronis Topica (Palermo, 1994), Gnomon 69 (1997), 647–8. 31. R. Wardy, The Birth of Rhetoric (London and New York, 1996)— W. J. Dominik (ed.), Roman Eloquence (London and New York, 1997)— C. Lévy and L. Pernot (eds.), Dire l’évidence (Paris and Montreal, 1997), SCI 17 (1998), 238–42. 32. H. M. Hine, L. Annaei Senecae Naturalium quaestionum libri (Stuttgart and Leipzig, 1996) and Studies in the Text of Seneca’s Naturales quaestiones (Stuttgart and Leipzig, 1996), SCI 17 (1998), 242–5. 33. J. Briscoe, Valeri Maximi Facta et dicta memorabilia (Stuttgart and Leipzig, 1998), SCI 18 (1999), 191–4. 34. G. W. Most (ed.), Editing Texts—Texte edieren (Göttingen, 1998), SCI 19 (2000), 328–30. *35. A. Stramaglia, [Quintiliano]. I gemelli malati: un caso di vivisezione (Declamazioni maggiori, 8) (Cassino, 1999), CR  50 (2000), 305–6. 36. S. Gwara, Aldhelmi Malmesbiriensis Prosa de virginitate cum glosa Latina atque Anglosaxonica (Turnhout, 2001), Speculum 79 (2004), 1017–18. 37. B. Löfstedt, Virgilius Maro Grammaticus. Opera omnia (Munich and Leipzig, 2003), Gnomon 77 (2005), 362–4. 38. J. C. Yardley, Justin and Pompeius Trogus: A Study of the Language of Justin’s Epitome of Trogus (Toronto, Buffalo, and London, 2003), IJCT 12 (2005–6), 463–5. *39. M. Weissenberger, Sopatri Quaestionum divisio—Sopatros: Streitfälle. Gliederung und Ausarbeitung kontroverser Reden (Würzburg, 2010), Gnomon 83 (2011), 394–6. 40. F. Wendling, Hugonis de Miromari De hominis miseria, mundi et inferni contemptu (Turnhout, 2010), JML 21 (2011), 333–8.

Acknowledgements We should like to thank the following publishers or institutions for kindly granting permission to reprint Michael Winterbottom’s papers listed below:

BINGHAMTON UNIVERSITY— CENTER FOR MEDIEVAL & RENAISSANCE STUDIES

Articles and chapters ‘Schoolroom and courtroom’, in B. Vickers (ed.), Rhetoric Revalued. Papers from the International Society for the History of Rhetoric (Binghamton [NY]: Center for Medieval & Early Renaissance Studies, 1982), 59–70. (A.6)

CAMBRIDGE P HILOLOGICA L SOCIETY

Articles and chapters ‘Cicero and the middle style’, in J. Diggle, J. B. Hall, and H. D. Jocelyn (eds.), Studies in Latin Literature and its Tradition in Honour of C. O. Brink (Cambridge, 1989) [= Cambridge Philological Society, Supplementary Volume 15], 125–31. (A.11)

CAMBRIDGE UNIVERSITY P RESS

Articles and chapters ‘Quintilian and the vir bonus’, Journal of Roman Studies 54 (1964), 90–7. © Michael Winterbottom, 1964. Exclusive Licence to Publish: The Society for the Promotion of Roman Studies. Reprinted with the permission of Cambridge University Press: http://www.cambridge.org/rights/permissions/authors.htm. (A.1) ‘An emendation in Calpurnius Flaccus’, Classical Quarterly  49 (1999), 338–9. © The Classical Association, 1999. Reprinted with the permission of Cambridge University Press: http://www.cambridge.org/rights/permissions/authors.htm. (A.14)

xxii

Acknowledgements

‘Quintilian 12.11.11–12’, Classical Quarterly  56 (2006), 324–5. © The Classical Association, 2006. Reprinted with the permission of Cambridge University Press: http://www.cambridge.org/rights/permissions/authors.htm. (A.19)

Reviews L. Håkanson, Textkritische Studien zu den grösseren pseudoquintilianischen Deklamationen (Lund, 1974), Classical Review  26 (1976), 276. © The Classical Association, 1976. Reprinted with the permission of Cambridge University Press: http://www.cambridge.org/rights/permissions/authors.htm. (R.1) S. F. Bonner, Education in Ancient Rome: From the Elder Cato to the Younger Pliny (London, 1977), Classical Review  29 (1979), 73–4. © The Classical Association, 1979. Reprinted with the permission of Cambridge University Press: http://www.cambridge.org/rights/permissions/authors.htm. (R.5) L. A. Sussman, The Elder Seneca (Leiden, 1978), Classical Review  29 (1979), 231–2. © The Classical Association, 1979. Reprinted with the permission of Cambridge University Press: http://www.cambridge.org/rights/permissions/authors.htm. (R.6) L. Håkanson, L. Annaeus Seneca Maior. Oratorum et rhetorum sententiae, divisiones, colores (Leipzig, 1989), Classical Review  41 (1991), 338–40. © The Classical Association, 1991. Reprinted with the permission of Cambridge University Press: http://www.cambridge.org/rights/permissions/authors.htm. (R.9) L. A. Sussman, The Declamations of Calpurnius Flaccus. Text, Translation, and Commentary (Leiden, New York, and Cologne, 1994), Classical Review  45 (1995), 40–2. © The Classical Association, 1995. Reprinted with the permission of Cambridge University Press: http://www.cambridge.org/rights/permissions/authors.htm. (R.10) A. Stramaglia, [Quintiliano]. I gemelli malati: un caso di vivisezione (Declamazioni maggiori, 8) (Cassino, 1999), Classical Review  50 (2000), 305–6. © The Classical Association, 2000. Reprinted with the permission of Cambridge University Press: http://www.cambridge.org/rights/permissions/authors.htm. (R.11)

C. H. BECK

Reviews J. Cousin, Quintilien. Institution oratoire, Tome  (Paris, 1975) and Recherches sur Quintilien (Paris, 1975), Gnomon 49 (1977), 574–9. (R.2)

Acknowledgements

xxiii

J. Cousin, Quintilien. Institution oratoire, Tomes – (Paris, 1976), Gnomon 50 (1978), 685–7. (R.3) J. Cousin, Quintilien. Institution oratoire, Tome  (Paris, 1977), Gnomon 51 (1979), 388–9. (R.4) J. Cousin, Quintilien. Institution oratoire, Tomes – (Paris, 1978; 1979), Gnomon 52 (1980), 785–6. (R.7) J. Cousin, Quintilien. Institution oratoire, Tome  (Paris, 1980), Gnomon 53 (1981), 197–9. (R.8) M. Weissenberger, Sopatri Quaestionum divisio—Sopatros: Streitfälle. Gliederung und Ausarbeitung kontroverser Reden (Würzburg, 2010), Gnomon 83 (2011), 394–6. (R.12)

EDITRICE MORC ELLIANA

Articles and chapters ‘The words of the master’, Maia 70 (2018), 73–83. (A.23)

FONDA TION HARDT POUR L ’ ÉTU DE DE L ’ ANTIQUITÉ CLASSIQUE

Articles and chapters ‘Cicero and the Silver Age’, in W. Ludwig (ed.), Éloquence et rhétorique chez Cicéron (Vandoeuvres and Geneva: Fondation Hardt pour l’étude de l’Antiquité classique, Entretiens sur l’Antiquité classique Tome XXVIII, 1982), 237–66 (‘Discussion’, 267–74). (A.5)

I NSTITUTO DE E STU DIOS RI OJ ANOS

Articles and chapters ‘Quintilian the moralist’, in T. Albaladejo, E. del Río, and J. A. Caballero (eds.), Quintiliano: Historia y Actualidad de la Rétorica, vol. 1 (Logroño, 1998), 317–34. (A.13)

xxiv

Acknowledgements

ISTITUTO DELLA ENCICLOPEDIA ITALIANA

Articles and chapters ‘Quintiliano (M. Fabius Quintilianus)’, in Enciclopedia Virgiliana, vol. 4 (Rome, 1988), 374–6. By courtesy of the Istituto della Enciclopedia Italiana Treccani. Any use in addition to or beyond the inclusion in this volume will be subject to request from and approval by the Istituto della Enciclopedia Italiana. (A.9)

JOHN WILEY & SONS

Articles and chapters ‘Problems in the Elder Seneca’, Bulletin of the Institute of Classical Studies 21 (1974), 20–42. © Institute of Classical Studies. School of Advanced Studies, University of London, 1974. Reproduced by permission of John Wiley & Sons. (A.2) ‘The text of Sulpicius Victor’, Bulletin of the Institute of Classical Studies 26 (1979), 62–6. © Institute of Classical Studies. School of Advanced Studies, University of London, 1979. Reproduced by permission of John Wiley & Sons. (A.4) Introduction to D. Innes & M. Winterbottom, Sopatros the Rhetor (London, 1988), 1–20. © Institute of Classical Studies. School of Advanced Studies, University of London, 1988. Reproduced by permission of John Wiley & Sons. (A.10) ‘More problems in Quintilian’, Bulletin of the Institute of Classical Studies 44 (2000), 167–77. © Institute of Classical Studies. School of Advanced Studies, University of London, 2000. Reproduced by permission of John Wiley & Sons. (A.15)

L E S BE L L E S LE T T R E S

Articles and chapters ‘Quintilian and declamation’, in Hommages à Jean Cousin (Paris, 1983), 225–35. © Les Belles Lettres, 1983. (A.8)

Acknowledgements

xxv

OXFORD UNIVERSITY PRESS

Articles and chapters ‘On impulse’, in D. Innes, H. Hine, and C. Pelling (eds.), Ethics and Rhetoric. Classical Essays for Donald Russell on his Seventy-Fifth Birthday (Oxford, 1995), 313–22. Reproduced by permission of Oxford University Press: https://global.oup.com/. (A.12)

SO CIE DAD E BR A SILE IR A DE E STUDO S CL Á SSIC O S

Articles and chapters ‘Declamation and philosophy’, Classica. Revista Brasileira de Estudos Clássicos 19 (2006), 74–82. DOI: http://doi.org/10.24277/classica.v19i1.105. (A.20)

TAYLOR & F RANCIS GROUP

Articles and chapters ‘Quintilian and rhetoric’, in T. A. Dorey (ed.), Empire and Aftermath. Silver Latin II (London, 1975), 79–97. (A.3)

THE CLASSICAL ASSOCIATION OF S OUTH AFRICA

Articles and chapters ‘Approaching the end: Quintilian 12.11’, Acta Classica 48 (2005), 175–83. Reprinted with the permission of the Classical Association of South Africa. (A.18)

xxvi

Acknowledgements

UNIVERSIDADE DE SÃO PA ULO

Articles and chapters ‘Something new out of Armenia’, Letras clássicas 8 (2004), 111–28. (A.17)

UN IVERSITÀ DEGLI STU DI D I CASSINO E DE L L A Z I O ME R I D I O N A L E

Articles and chapters ‘William of Malmesbury’s work on the Declamationes maiores’, Segno e Testo 12 (2014), 261–76. (A.21)

UNIVERSITÀ DEGLI STUD I DI GENOVA—DIPARTIMENTO DI ANTICHITÀ, FILOSOFIA, STORIA

Articles and chapters ‘Declamation, Greek and Latin’, in A. Ceresa-Gastaldo (ed.), Ars rhetorica antica e nuova (Genoa, 1983), 57–76. (A.7)

W A LTE R D E G R U Y T E R GM B H

Articles and chapters ‘Ennodius, Dictio 21’, in B.-J. and J.-P. Schröder (eds.), Studium declamatorium. Untersuchungen zu Schulübungen und Prunkreden von der Antike bis zur Neuzeit (Munich and Leipzig, 2003), 275–88. © 2003 by K. G. Saur Verlag GmbH, Munich and Leipzig (now De Gruyter). (A.16) ‘The editors of Calpurnius Flaccus’, in M. Dinter, C. Guérin, and M. Martinho (eds.), Reading Roman Declamation. Calpurnius Flaccus (Berlin and Boston, 2017), 141–60. © Walter de Gruyter GmbH, Berlin/Boston 2017. (A.22) The publisher and editors apologize for any errors or omissions in the above list. If contacted they will be pleased to rectify these at the earliest opportunity.

1 Quintilian and the vir bonus* Sit ergo nobis orator quem constituimus is qui a M. Catone finitur, vir bonus dicendi peritus, verum, id quod et ille posuit prius et ipsa natura potius ac maius est, utique vir bonus.¹ Why did Quintilian insist so strongly on the moral qualities of the orator? The question has not been persistently enough asked. Austin, for example, thinks that it is only ‘a steadfast sincerity of purpose throughout’ that redeems the first chapter of Book 12 from ‘mere moralizing’.² And it only takes the problem a stage further back to say that this is a matter of Stoic influence.³ Even if Posidonius did formulate in connection with rhetoric a maxim on the lines of Strabo’s οὐχ οἷόν τε ἀγαθὸν γενέσθαι ποιητὴν μὴ πρότερον γενηθέντα ἄνδρα ἀγαθόν,⁴ we must still ask why Quintilian troubled to give this Stoic view such new prominence. After all, oratori . . . nihil est necesse in cuiusquam iurare leges.⁵ And it is clear that Quintilian realized that he was innovating. Cicero, he writes, despite the width of his conception, thought it enough to discuss merely the type of oratory that should be used by the perfect orator: at nostra temeritas etiam mores ei conabitur dare et adsignabit officia.⁶ This is perhaps to schematize the contrast a little overdramatically. Cicero had certainly written in the De oratore: quarum virtutum expertibus si dicendi copiam tradiderimus, non eos quidem oratores effecerimus, sed furentibus quaedam arma dederimus.⁷ But there is no doubt that Cicero was not primarily concerned with the moral aspect. As the leading orator of his day, he may have thought it indelicate or superfluous to stress that the perfect orator must be a good man. Moreover, it was not clear that the troubles of Cicero’s day were the result of morally bad orators: one had to

[Journal of Roman Studies 54 (1964), 90 7] * An earlier draft of this paper has been read to the Oxford Branch of the Classical Asso ciation and to the London Classical Society. ¹ Quint. 12.1.1. ² Austin (1954²),  (Austin’s italics). ³ Ibid.,  . See also the notes on 12.1.1. Austin is rightly cautious. ⁴ Strab. 1.2.5; Morr (1926 7), 47. ⁵ Quint. 12.2.26. ⁶ Quint. 12.pr.4. ⁷ Cic. De orat. 3.55.

4

Papers on Quintilian and Ancient Declamation

look back to Saturninus and Glaucia for examples of the evils caused by unscrupulous use of words.⁸ This paper will suggest that there was a very good reason for Quintilian’s newly moralistic approach: and that this was a matter of historical fact, not of rhetorical theory. Tacitus in the Dialogus set himself to explain why nostra potissimum aetas deserta et laude eloquentiae orbata vix nomen ipsum oratoris retineat.⁹ Aper denies that oratorical glory is, in fact, dead. His main examples are Eprius Marcellus and Vibius Crispus.¹⁰ The choice is significant. For the outstanding fact about first-century oratory is that the only orators to achieve any prominence or influence by means of their oratory are the delatores.¹¹ The rest were decorative but impotent: the Dialogus tells us why—education lacked touch with reality, and political conditions took away all scope. Hence one delator, Publius Suillius Rufus, who had been terribilis ac venalis under Claudius, could contrast himself tellingly with Seneca.¹² Seneca was used to academic inertia and the callowness of youth; he was jealous of those such as Suillius who used in the defence of their fellow-citizens an eloquence that was bright, alive, and untarnished—vividam et incorruptam eloquentiam tuendis civibus exercerent. Tacitus’ archetypal delator, as we read in his Annals,¹³ dedit exemplum quod secuti ex pauperibus divites, ex contemptis metuendi, perniciem aliis ac postremum sibi invenere. The history of delation in the first century shows that this summary remained true; but the delator gradually added to these qualities something approaching an official position, and (in some cases) something approaching a theory of oratory. We may start with an Augustan orator, who was not a delator in the strict sense at all: Cassius Severus.¹⁴ Quintilian’s key-word for him is acerbitas.¹⁵ He was finally banished for the libido with which viros feminasque inlustris procacibus scriptis diffamaverat.¹⁶ He was a professional satirist rather than a professional accuser. All the same, Quintilian reproves him for a remark that betrayed quaedam accusandi voluptas.¹⁷ And we are told by Seneca the Elder that he specialized in accusation.¹⁸ In view of this, and because of interesting parallels between Cassius and some later delatores, he deserves discussion here. The evidence does not lie in the Institutio, where, though Cassius is ⁸ Cic. Brut. 224. The Gracchi are also mentioned. Quintilian takes over these examples (2.16.5). ⁹ Tac. Dial. 1.1. ¹⁰ Tac. Dial. 8.1. ¹¹ The eloquence of the delatores is discussed by Froment (1880), 35. ¹² Tac. Ann. 13.42. If Tacitus invents, his invention is of archetypal significance. See also Syme (1958), 331 2. It will be obvious how much I owe to this book. ¹³ Tac. Ann. 1.74.2; Syme (1958), 326 n. 5. ¹⁴ Sources for him are gathered in Meyer (1842²), 545 61 [= Balbo (2007²), .223 43]; Schanz Hosius (1935⁴), .345 ff. ¹⁵ Quint. 10.1.117; 12.10.11. ¹⁶ Tac. Ann. 1.72.3. ¹⁷ Quint. 11.1.57. ¹⁸ Sen. Con. 3.pr.5.

Quintilian and the vir bonus

5

often mentioned, it is almost always to stress his biting wit and his hatred of pomposity. Turn, however, to the Dialogus: here, in the big speech of Aper, an attack is developed on those who habitually reckon Cicero, Caesar, and other Republican orators superior to the orators of Aper’s own day.¹⁹ Aper quibbles about the exact meaning of antiqui in this context, and then asserts that new circumstances breed new styles of oratory. It was all very well for the ‘admirers of antiquity’ to draw a sharp line and to proclaim that with Cassius came the deluge.²⁰ They might say that Cassius was the one who had first diverged from the straight and narrow path; but in fact Cassius knew very well what he was doing. Times had changed. Under the Republic audiences were still impressed by a smattering of philosophy and by rhetorical subtleties prescribed in the dry-as-dust handbooks of Hermagoras and Apollodorus. Once this excitement wore off, there was need for new methods, a vigorous attempt to stave off boredom and monotony. It was on purpose—and for cogent reasons—that Cassius had taken a new course. What exactly had Cassius put in the place of the old techniques? Messalla’s reply to Aper in the Dialogus is significant here. Messalla doesn’t deny that Cassius was a notable orator, though his speeches had plus bilis . . . quam sanguinis. This is much what Quintilian said in his brief notice of Cassius.²¹ But, Messalla goes on, Cassius was the first to despise organization, and cast aside modesty of language;²² he was so eager to strike that he often fell over in the process: a brawler, no true fighter. We may add the evidence of the Elder Seneca. Everything in Cassius’ oratory had a direct purpose—omnia intenta, aliquid petentia.²³ It was strong stuff, elegant, ingentibus plena sententiis. Cassius relied rather on his native wit than on his education.²⁴ He declaimed occasionally, but he was under no illusions about the value of declamation. In scholastica quid non supervacuum est, cum ipsa supervacua sit?²⁵ It is not impossible that in the Dialogus Tacitus is replying to Quintilian on the topic of Cassius Severus.²⁶ Admittedly, Quintilian did not, in the Institutio, assert that Cassius started the decline of Roman oratory. The orators mentioned in 10.1.113 ff. are treated atomically, analysed for the virtues they may illustrate rather than fitted into trends and patterns. But there is a good chance that in the earlier De causis corruptae eloquentiae Quintilian did take a more historical line. From 2.4.41–2 we know that he discussed there whether or not

¹⁹ Tac. Dial. 16.4. ²⁰ Tac. Dial. 19.1. ²¹ Quint. 10.1.116 17; cf. Tac. Dial. 26.4. ²² Tac. Dial. 26.5; cf. Quint. loc. cit.: Cassius lacked gravitas and consilium. ²³ Sen. Con. 3.pr.2. ²⁴ Sen. Con. 3.pr.4: maioris ingenii quam studii; cf. Quint. 10.1.117: ingenii plurimum. ²⁵ Sen. Con. 3.pr.12. ²⁶ I have no new arguments with which to rejuvenate the hoary topic of the date of the Dialogus. This article proceeds on the assumption that it post dates Quintilian’s De causis. Cf. Syme (1958), 112 ff.

6

Papers on Quintilian and Ancient Declamation

suasoriae and controversiae were invented by Demetrius of Phaleron. Now Cicero²⁷ remarks on the agreeable but academic virtues of Demetrius. He it was who primus inflexit orationem et eam mollem teneramque reddidit. Quintilian himself recalls this:²⁸ Demetrius primus inclinasse eloquentiam dicitur. It is not then impossible that Quintilian’s De causis gave a historical sketch of both Greek and Roman oratory; in both there was a clear point where decay started—the time of Demetrius and the time of Cassius: equally, in both, new educational techniques, centred upon declamation, played a leading part in causing this decay.²⁹ If this is correct, Aper’s speech in the Dialogus takes on a further significance. Admirers of the ancients, such as Quintilian, he implies, are irrevocably stuck in the past: they don’t see that Cassius Severus was not an end but a new and hopeful start. They think you can get by, as in the old days, with a little philosophy and a lot of rhetorical rules. They think that audiences will put up with speakers who spend all day on their feet.³⁰ No more pertinent criticism of Quintilian’s general position could be imagined. In reply, Messalla is allowed to say, as Quintilian would have said, that Cassius fell below his predecessors but excelled his successors.³¹ But even he cannot explain why the old days should be thought relevant, even if they were indisputably better. If Quintilian was one of those who attributed to Cassius Severus the drastic change in the direction of oratory in the first century, it becomes necessary to make it quite clear what Cassius was being blamed for. Messalla tells us at Dial. 26.5: Primus enim contempto ordine rerum, omissa modestia ac pudore verborum . . . non pugnat sed rixatur. And we may recall Aper’s speech,³² where the new style started by Cassius is connected with the abandonment of rhetorical subtleties and pedantic rules, quidquid . . . aridissimis Hermagorae et Apollodori libris praecipitur. Cassius relied on his ingenium rather than on training, and he despised declamation. Quintilian’s De causis, it may be conjectured, defended training, asserted that declamation, properly carried out, was advantageous,³³ and demanded a return to the Ciceronian virtues of respect for education and for rhetorical dogma. The De causis, in this light, paved the way for the Institutio, where the despised longa principiorum praeparatio et narrationis alte repetita series et multarum divisionum ostentatio et mille argumentorum gradus were painstakingly recapitulated. It was a plea too for the wider education that people such as Aper so scorned.

²⁷ Cic. Brut. 37 8. ²⁸ Quint. 10.1.80. ²⁹ This conclusion is approached in Norden (1898), 248; cf. also Reuter (1887), 8. But I vis ualize a sketch of the history of oratory, not merely of declamation. ³⁰ Such as Quintilian’s pupil, Pliny: . . . perstitit . . . horis septem. Nam tam diu dixi (4.16.2 3). ³¹ Tac. Dial. 26.4. ³² Tac. Dial. 19.3. ³³ Cf. Quint. 2.10, and 5.12.17 23 where the De causis is actually referred to.

Quintilian and the vir bonus

7

We may next examine the case of Domitius Afer. The interest here lies in the contrast between the impression given of this orator by Tacitus and by Quintilian. If we relied on the Annals alone, we should be hard put to it to see Afer as more than an earlier and more evil Suillius Rufus. He was a tool in the campaigns against Agrippina in 26 and 27:³⁴ later, almost a victim of Caligula, then consul under him.³⁵ Tacitus describes his motives as unsparingly as those of any other delator. He was in a hurry to be famous, at any price: in defence and prosecution alike he was more eloquent than principled—and in old age not even eloquent. He had never been rich; he misspent the rewards of delation, and was lured on to further crimes. Little of all this appears in the pages of Quintilian; senility alone finds a place.³⁶ Instead, Afer is a summus orator, as good as the old-timers, witty and ripe.³⁷ We hear of many of his cases—but almost always of his defences: the only accusation mentioned is that of a libertus of Claudius,³⁸ and that would be to his credit. Afer had been Quintilian’s boyhood hero³⁹—but by then Afer was an old man, author of books on the examination of witnesses,⁴⁰ respectable as never before. Quintilian, then, was disposed to the most favourable judgement possible. Only once does the Tacitean Afer peep out (by accident) in the Institutio. A clever saying of his is quoted, for its cleverness: Ego accusavi, vos damnastis.⁴¹ One can imagine a context, imagine too the cynical smile that this would have aroused in Tacitus. Suillius used his eloquence to defend citizens. Afer was accuser, not judge. How then could we blame either? I now move on to Flavian delation, and examine the careers of a pair closely linked in the Dialogus, Eprius Marcellus and Vibius Crispus (flourishing, it may be remembered, during Quintilian’s professorship in Rome). Tacitus describes them as potentissimi civitatis under Vespasian,⁴² and it is on their public fame that I shall concentrate. For as long as it pleased them, they had been foremost in the forum: now they were foremost in the friendship of Caesar—and indeed the object of his respectful regard. Others might depend on Vespasian’s goodwill: Marcellus and Crispus brought to their amicitia something that they had not received from Vespasian. By this, Aper means that the oratory of these two was their making; and we soon learn from Maternus what sort of oratory it was, lucrosae huius et sanguinantis eloquentiae usus . . . ex malis moribus natus atque . . . in locum teli repertus.⁴³ This weapon had already under Nero been at the service of the emperor, when Marcellus had been the accuser of Thrasea Paetus, torvus ac minax, voce vultu oculis (ardescens).⁴⁴ The reward, five million sesterces.⁴⁵ For Thrasea, death. Crispus had done nothing so spectacular, but by Nero’s death he was rich, ³⁴ ³⁷ ⁴⁰ ⁴⁴

Tac. Ann. 4.52 and 66. ³⁵ Dio Cass. 59.19. ³⁶ Quint. 12.11.3. Quint. 12.10.11; 10.1.118. ³⁸ Quint. 6.3.81. ³⁹ Quint. 5.7.7. Quint. 5.7.7. ⁴¹ Quint. 5.10.79. ⁴² Tac. Dial. 8.3. ⁴³ Tac. Dial. 12.2. Tac. Ann. 16.29.1. ⁴⁵ Tac. Ann. 16.33.2.

8

Papers on Quintilian and Ancient Declamation

powerful, inter claros magis quam inter bonos:⁴⁶ and men remembered that he too had been a delator. In the fourth book of the Histories⁴⁷ Tacitus gives a brilliant picture of the attack on Marcellus and Crispus in the senate during the early days of Vespasian’s reign. After a well-received speech by the fierce Curtius Montanus, Helvidius Priscus took up the attack, the Senators approving: Quod ubi sensit Marcellus, velut excedens curia, ‘Imus,’ inquit, ‘Prisce, et relinquimus tibi senatum tuum. Regna praesente Caesare.’ Sequebatur Vibius Crispus, ambo infensi, vultu diverso, Marcellus minacibus oculis, Crispus renidens . . . ⁴⁸ Crispus, that agreeable man,⁴⁹ might well smile: perhaps he knew what was going to happen: Mucianus’ long speech in favour of the delators next day, and the sudden melting of senatorial free-speech.⁵⁰ It is clear that the new régime had put its shield over Marcellus and Crispus. But, as Tacitus half-tells us in the Dialogus,⁵¹ the deal was not merely one-sided. Rich, powerful, and eloquent delatores were essential to the running of the new Flavian administration. Vespasian could only trust a limited circle, especially if the senate proved hostile. He could use his relations: appoint Titus to the Jewish command, Domitian praetor,⁵² give military posts to Arrecinus Clemens, Caesennius Paetus, and Petilius Cerealis.⁵³ But this was not enough. The delatores were a ready-made answer. They could be thrown to their enemies in the senate if they caused trouble. Meanwhile, they could work. Marcellus became proconsul of Asia, for three years, then, for a second time, consul, in 74. Vibius Crispus governed Africa and Tarraconensis, besides his cura aquarum. Another proconsul of Africa was Paccius Africanus, accuser of the Scribonii under Nero, who like Marcellus and Crispus had come under fire in the senate in 70.⁵⁴ Silius Italicus passed on from Neronian delation to a Flavian proconsulship of Asia.⁵⁵ The delatores were now not merely powerful, they had long been that. Now they were positively members of the Establishment. It was not so much that they were ‘eager to repair their credit’:⁵⁶ rather that Vespasian both needed them and had a hold over them. This was not merely a passing phase of Vespasian’s reign. Admittedly, Eprius Marcellus came to a sudden end in 79, after involvement, real or apparent, in a conspiracy against the throne (significant, this, of the heights to which delatores could by now aspire: and of the basic insecurity of their position). But Vibius Crispus was still making his elegant jokes under Domitian,⁵⁷ and the emperor’s

⁴⁶ Tac. Hist. 2.10.2. ⁴⁷ Tac. Hist. 4.41 ff. ⁴⁸ Tac. Hist. 4.43.3. ⁴⁹ Quint. 10.1.119: delectationi natus. ⁵⁰ Tac. Hist. 4.44.1 2. ⁵¹ Tac. Dial. 8.3. ⁵² Tac. Hist. 4.3.7. ⁵³ Syme (1958), 594 5, where the references for Marcellus and Crispus, and for other dela tores, also appear. ⁵⁴ Tac. Hist. 4.41.4. ⁵⁵ Plin. Ep. 3.7.3. ⁵⁶ Syme (1958), 594. ⁵⁷ Suet. Dom. 3.1.

Quintilian and the vir bonus

9

pronouncement, princeps qui delatores non castigat, irritat⁵⁸ affected petty accusers rather than the mighty. A new generation of delatores flourished— Fabricius Veiento, notably, and Catullus Messallinus, linked in Juvenal’s concilium satire and again in Pliny, who witnesses that Veiento was favoured even by Nerva—and that Catullus no doubt would have been had he lived.⁵⁹ It is not surprising that Trajan’s ruthless stamping out of delatores was the subject of some sections in Pliny’s Panegyricus.⁶⁰ Yet even under Trajan one of the most important accusers of all lived on, and had influence—Marcus Aquillius Regulus, spanning dynasties and generations, still factious, feared and courted after the death of Domitian.⁶¹ He had ruined noble families under Nero, while still unknown. Attacked in the unruly senate of early 70, he, like Marcellus and Crispus, survived: perhaps for the same reason, perhaps thanks to the efforts of his brother, the Messalla who appears in the Dialogus. ‘[H]is subsequent conduct,’ writes Syme,⁶² ‘though highly objectionable, [did] not involv[e] him in the prosecution of any notable members of the senatorial opposition.’ But he launched savage attacks on the memories of Arulenus Rusticus and Herennius Senecio, two Stoic victims of Domitian’s last years: and Regulus’ hand may have been at work in their actual ruin. Periculum foverat, says Pliny⁶³ in connection with Rusticus; and he remarks that Regulus’ crimes under Domitian were no less heinous than those under Nero, merely better concealed.⁶⁴ We do not have details of Regulus’ official career, though he was consul at some time unknown; no doubt he was more prominent in the forum, less in the palace, than Marcellus and Crispus. And, correspondingly, we hear far more of his oratory. Martial, a diligent admirer, praises his eloquence, remarks on his ingenium, and emphasizes his celebrity as a defender.⁶⁵ One remembers Suillius’ claim to use his oratory tuendis civibus: and one is sceptical. More illuminating are the letters of Pliny, which draw Regulus with bold impressionistic strokes. He emerges as a new version of Cassius Severus. Let us recall Severus, his anger, his enthusiasm for accusation and abuse, the oratory that had omnia intenta, aliquid petentia, but lacked order, brawling instead of fighting: a man who despised rhetorical doctrine, maioris ingenii quam studii. Compare with him Regulus, as we see him through the eyes of Pliny (and Pliny, we remember, was a pupil of Quintilian): imbecillum latus, os confusum, haesitans lingua, tardissima inventio, memoria nulla,⁶⁶ nihil denique praeter ingenium insanum . . . ⁶⁷ Relevant to Quintilian also the next words: et tamen eo impudentia ipsoque illo furore pervenit ut orator habeatur. What would Quintilian have thought of this prostitution of oratoris illud sacrum nomen?⁶⁸ ⁵⁸ Suet. Dom. 9.3. ⁵⁹ Plin. Ep. 4.22; cf. Syme (1958), 4 6. ⁶⁰ Plin. Pan. 34 ff. ⁶¹ Plin. Ep. 1.5.15. ⁶² Syme (1958), 77. ⁶³ Plin. Ep. 1.5.2. ⁶⁴ Plin. Ep. 1.5.1. ⁶⁵ For his eloquence e.g. Mart. 5.28.6; his ingenium 5.63.4; his abilities in defence e.g. 4.16.6. ⁶⁶ Clear references to at least three of the rhetorical partes listed by Quint. 3.3.1: inventione, dispositione, elocutione, memoria, pronuntiatione. ⁶⁷ Plin. Ep. 4.7.4. ⁶⁸ Quint. 12.1.24.

10

Papers on Quintilian and Ancient Declamation

We may add the violence of Regulus’ language: he called Rusticus ‘that Stoic ape’, Vitelliana cicatrice stigmosum.⁶⁹ ‘You recognize the style of Regulus’, commented Pliny wryly. And, what is more, Regulus himself knew what his oratory was like. He enjoyed contrasting himself with Pliny: Tu omnia quae sunt in causa putas exsequenda; ego iugulum statim video, hunc premo.⁷⁰ There, uniquely and memorably, speaks the violent oratory of the delatores. On another occasion Regulus contrasted Pliny with Satrius Rufus, cui non est cum Cicerone aemulatio et qui contentus est eloquentia saeculi nostri.⁷¹ Regulus, in fact, was not just a violent and unscrupulous orator. He could see himself in a wide literary context, and recognize that he, and orators like him for nearly a hundred years, were for good reasons quite unlike Cicero. It was not that Cicero was a bad orator: but merely that times had changed. Pliny was going the wrong way—towards a dead past. All this of course reminds us irresistibly of the Dialogus. ‘To defend the modern style in oratory, the author introduces, as it were, a purified and sympathetic Regulus, namely Marcus Aper.’⁷² But there is no need to narrow the case down so much. Aper stands for all the orators of his type that the century produced. Aper, Tacitus tells us, had attained fame in eloquence ingenio et vi naturae,⁷³ rather than by education (institutio is the word, perhaps significantly). But he was learned enough in his way, he despised literature rather than was ignorant of it.⁷⁴ Now this is almost exactly what Aper later says of Cassius Severus, non infirmitate ingenii nec inscitia litterarum transtulisse se ad aliud dicendi genus . . . sed iudicio et intellectu.⁷⁵ And even Regulus habebat studiis honorem.⁷⁶ All were purposeful and intelligent men: and it is impossible to be sure that Tacitus felt no sympathy for the case they put up.⁷⁷ How did Quintilian react to all this? I have suggested above that the Institutio in general is an answer to the Apers of the day who thought rhetorical doctrine outdated: and it can now be seen that Regulus was one important contemporary representative of this view. Regulus, however, as still living is not mentioned by name in the Institutio. What must now be noticed is that Quintilian equally omits to connect with any of the names he mentions the characteristics common to Regulus and, it might seem, many of his delator predecessors: contempt for rhetorical rules, violence of language, increasing political influence, moral failings of the first order. When it is remembered that the delatores were the most important oratorical phenomenon of the century, it becomes clear that Quintilian is glossing over the extent to which he

⁶⁹ Plin. Ep. 1.5.2. ⁷⁰ Plin. Ep. 1.20.14. ⁷¹ Plin. Ep. 1.5.11. ⁷² Syme (1958), 109 with n. 4. ⁷³ Tac. Dial. 2.1. ⁷⁴ Tac. Dial. 2.2. ⁷⁵ Tac. Dial. 19.1. ⁷⁶ Plin. Ep. 6.2.2. ⁷⁷ Note also how Aper spoke: acrius, ut solebat (Dial. 2.1) the pale reflexion of Regulus’ sav age style?

Quintilian and the vir bonus

11

himself is swimming against the tide in proclaiming a new Ciceronianism. Of the Flavian orators, for instance, Eprius Marcellus does not appear at all, even to be criticized. Perhaps he was better left out, in view of the ambiguities of his end.⁷⁸ His rival Vibius Crispus is given a whitewashed picture, much like that of Domitius Afer. For Quintilian,⁷⁹ Crispus was iocundus: so too for Juvenal.⁸⁰ He was only recently dead, and Domitian was still alive. Quintilian might have said much more, but he could hardly have said less. Crispus, he noted, was better⁸¹ in private than in public cases: we could if we liked see this as the faintest of hints that Vibius Crispus was not exactly the most fitting example of the vir bonus dicendi peritus. Other first-century orators mentioned by Quintilian are not delatores. But we should be wrong to say that Quintilian ignores the tendencies he found flourishing around him, even if he did not name their most notorious exponents, or, naming them, did not connect them with those tendencies. We may first remark on his criticism of ‘naturalists’ (my word), people who relied on their ingenium alone. In the second book, before starting on the Ars as such, Quintilian pauses to observe that quosdam in ipso statim limine obstaturos mihi, qui nihil egere eius modi praeceptis eloquentiam putent, sed natura sua et vulgari modo scholarum exercitatione contenti rideant etiam diligentiam nostram.⁸² Quintilian immediately makes it appear as though these objectors are merely professional declaimers who feel that one could declaim without any detailed technical instruction. They rely on their ingenium alone, and boast that they speak impetu, by inspiration,⁸³ claiming that there is no need for dispositio or proof in cases that in any case are imaginary: what is wanted is rather grandes sententiae.⁸⁴ All this is uncannily reminiscent of what we know of Cassius Severus;⁸⁵ and it could be that Quintilian wants us to recognize more important figures behind those foolish declaimers; at any rate, these declaimers have much in common with the attitudes of Regulus. ⁷⁸ Syme (1958), 109 n. 1. ⁷⁹ Quint. 10.1.119; cf. 12.10.11; 5.13.48. ⁸⁰ Juv. 4.81. ⁸¹ Quint. 10.1.119: melior. ⁸² Quint. 2.11.1. Cf. Winterbottom (1964c), 120 ff. ⁸³ Quintilian is never loath to use insanity to explain or abuse the excesses of contemporary rhetoric, cf. 2.12.9: iactatione gestus, motu capitis furentes, and often elsewhere. So even the sage Chilon, according to Diogenes Laertius 1.70, λέγοντα μὴ κινεῖν τὴν χεῖρα· μανικὸν γάρ. We have seen Pliny, echoing his master, talk of Regulus’ furor. Madness, now as always, was connected closely with inspiration, and if the ‘naturalists’ boasted that they were speaking impetu, they were perhaps taking up the criticism of their opponents and making a virtue of it (for impetus of inspiration cf. e.g. Ov. Pont. 4.2.25: Impetus ille sacer qui vatum pectora nutrit. See too Quint. 10.7.14, where Cicero is quoted as saying that according to old orators a god is present in successful extemporary effusion). [This issue is much expanded in A.12 below.] Behind Quinti lian’s use of ratio to mean method at 2.2.4 (cf. 7) may lurk the implication that the Institutio offered reason in place of the madness that now prevailed (Ov. Met. 14.701: postquam ratione furorem / vincere non potuit). ⁸⁴ Quint. 2.11.2 3. ⁸⁵ Sen. Con. 3.pr.4 and 2: maioris ingenii quam studii . . . ingentibus plena sententiis (sc. oratio); Tac. Dial. 26.5: contempto ordine rerum.

12

Papers on Quintilian and Ancient Declamation

In the next chapter, however, Quintilian moves away from the declamation school. The transition is imperceptible, but the change is clear: there is mention of the litigant,⁸⁶ and of audience reaction in court.⁸⁷ We now have, parallel with the naturalist declaimer, the naturalist orator. He is compared with a gladiator rushing without training in rixam.⁸⁸ We shall remember that Messalla in the Dialogus⁸⁹ makes exactly the same criticism of Cassius Severus. Moreover, the naturalist orator, ineruditus as he is, is overprone to abusiveness:⁹⁰ so too Cassius Severus (plus bilis . . . quam sanguinis), so too Regulus (Agnoscis eloquentiam Reguli). ‘And so’, says Quintilian, ‘let them be called ingeniosi so long as it is understood that this is not a word we could use in praise of anyone who was truly eloquent.’⁹¹ Here then Quintilian comes to grips with his real adversaries: those who thought there was no point in rhetorical rules. ‘Let us congratulate them’, he concludes ironically. ‘They are eloquent without work, without method and without discipline.’ These three qualities were what Quintilian proposed to put into the Institutio, and they constituted a good deal of what he recommended in it. We can now see whom he is criticizing, the spiritual descendants of Cassius Severus: and among these, at least by implication, may be numbered the delatores, and in particular Regulus himself.⁹² But more generally than this, the whole background of delator eloquence that has been sketched above throws light on much that might seem irrelevant or academic in the Institutio. Half of the second book, for example, consists of quaestiones about the status of rhetoric. Is it an art? (c. 17). If so, what kind of art—a good one, or merely a neutral one: is it a virtus? (c. 20). Is it utilis? (c. 16). Does nature or education contribute more to the great speaker? (c. 19). All these questions, of course, had for long been discussed by writers on rhetoric. But Quintilian goes over them again, and at length, because they were topics of the moment. People were actually saying, and had for years been saying, that rhetoric was a mere knack, to be picked up by experience in the courts, a matter of ingenium schooled only by practice. Quintilian had at

⁸⁶ Quint. 2.12.4. ⁸⁷ Quint. 2.12.6. ⁸⁸ No need to search, as Spalding searched, for cases of rixa used of gladiatorial combat: Quintilian is saying that a contest between untrained gladiators is a brawl, not a fight (not dissimilarly, Sen. Dial. 10.12.2 talks of puerorum rixantium: they were wrestling, Seneca was being scornful). ⁸⁹ Tac. Dial. 26.5: non pugnat sed rixatur. ⁹⁰ Quint. 2.12.4. ⁹¹ So at much the same time Mart. 7.9: Cum sexaginta numeret Cascellius annos, / ingeniosus homo est: quando disertus erit? ⁹² For naturalists of a rather different kind see Quint. 12.10.40 ff. More relevantly, 9.4.3: Neque ignoro quosdam esse, qui curam omnem compositionis excludant, atque illum horridum sermonem, ut forte fluxerit, modo magis naturalem modo etiam magis virilem esse contendant. Cf. Suillius on his vividam et incorruptam eloquentiam (Tac. Ann. 13.42.3). See also 11.3.10 11, whose tone can instructively be compared with that of 2.12.12.

Quintilian and the vir bonus

13

least to make a show of disproving these contentions: because otherwise the whole mass of his doctrine fell to the ground. But there is more to it than that. It is my contention that the same background answers my opening question also: why did Quintilian insist that the orator should be a good man? Orators are still famous, says Aper:⁹³ quos saepius vulgus . . . transeuntis nomine vocat et digito demonstrat? His examples are delatores, Marcellus and Crispus, representatives of a class who in this century agerent verterent cuncta odio et terrore.⁹⁴ Thus, as Maternus rejoins, this fame was bought at a price; it, like the eloquence that produced it, was ex malis moribus natus.⁹⁵ In the light of this, I suggest, we do not need to look further for the reason for Quintilian’s emphasis on the moral qualities of the perfect orator. There could be no question, of course, of open condemnation of the malpractices of the day. Domitian was still alive; so, no less importantly, was the dangerous Regulus. What is more, Quintilian was in all probability writing during the final Domitianic reign of terror, when Carus Mettius’ victoriae were increasing in number, and sententia Messallini strepebat beyond the four walls of the Alban villa.⁹⁶ Moreover, there was a matter of propriety. Quintilian was a Flavian office-holder, appointed by Vespasian, and, under Domitian, tutor for a while of royal children. Along with kind words about Domitian’s poetic prowess had to go a discreet silence about the less agreeable aspects of Flavian power. Indeed it is striking that Quintilian says as much as he does. In Book 2 he says that he thinks et fuisse multos et esse nonnullos . . . qui facultatem dicendi ad hominum perniciem converterint.⁹⁷ Later, nam et minari et deferre etiam non orator (even a non-orator) potest.⁹⁸ In Book 12, most openly of all, accusatoriam vitam vivere et ad deferendos reos praemio duci proximum latrocinio est.⁹⁹ Indeed, it is surprising that this should have been written, or at least published, under Domitian at all.¹⁰⁰ We find the same tone of voice under Trajan: Pliny in the Panegyric calls delators latrones.¹⁰¹ I suggest then that Quintilian was, like Plato, led to a moralistic view of the function of rhetoric by what he saw going on around him. He found himself disgusted by the way rhetoric was being misapplied: and we should not forget that this was a matter of emotion to the academic Quintilian. Eloquence for him was honesta ac rerum pulcherrima.¹⁰² Nature non parens sed noverca

⁹³ Tac. Dial. 7.4. ⁹⁴ Tac. Hist. 1.2.3. ⁹⁵ Tac. Dial. 12.2. ⁹⁶ Tac. Ag. 45.1. ⁹⁷ Quint. 2.20.2. ⁹⁸ Quint. 4.1.22. ⁹⁹ Quint. 12.7.3. Also in Book 12, note the emphasis on pecuniariae quaestiones in which veritas had to be defended against calumnia by the good orator. ¹⁰⁰ The Institutio was ‘presumably published before Domitian’s death in 96. At least it seems unlikely that if the murder had taken place before publication the complimentary passages would have been allowed to remain’ (Colson (1924),  n. 5). ¹⁰¹ Plin. Pan. 34.1. So, it is true, did Columella (1.pr.9). ¹⁰² Quint. 1.12.16.

14

Papers on Quintilian and Ancient Declamation

fuerit si facultatem dicendi sociam scelerum, adversam innocentiae, hostem veritatis invenit.¹⁰³ It was on these convictions that Quintilian based his assertions that the orator must be a good man, skilled in speaking. And it will be no coincidence that Herennius Senecio drew the moral when he described a contemporary orator as vir malus dicendi imperitus.¹⁰⁴ That contemporary was Marcus Aquillius Regulus. A postscript may be added. Quintilian thought his age rich in rhetorical talent;¹⁰⁵ there could be great orators once again, if the lessons of the Institutio were thoroughly learnt. In a way he was wrong; ironically, with the crushing of the delatores, Trajan seemed to kill oratory also. When Regulus died Pliny found himself writing with a conscious paradox that he missed the man, despite everything.¹⁰⁶ In the same mood, Tacitus wrote in the Agricola¹⁰⁷ of stagnation under the first years of Trajan—the numbness and inertia of the new peace. There was room only for panegyric now, and the driest of legal advocacy: only occasionally the spice of a trial for misdemeanours in the provinces.¹⁰⁸ There was, basically, nothing to do in the senate: Sunt quidem cuncta sub unius arbitrio, qui pro utilitate communi solus omnium curas laboresque suscepit.¹⁰⁹ And this is exactly the wistful note of Maternus’ last speech in the Dialogus: Quid enim opus est longis in senatu sententiis, cum optimi cito consentiant? Quid multis apud populum contionibus, cum de republica non imperiti et multi deliberent, sed sapientissimus et unus?¹¹⁰ In these circumstances Quintilian’s view of the orator as one whose primary task it was to ‘guide the counsels of the senate and bring the errant people back to better courses’¹¹¹ was absurdly out-of-date. Oratory could no longer have its traditional¹¹² political justification.

¹⁰³ Quint. 12.1.2. ¹⁰⁴ Plin. Ep. 4.7.5. ¹⁰⁵ Quint. 10.1.122. Quintilian is less defensive here than at 2.5.23 4 ( . . . novos, quibus et ipsis multa virtus adest. Neque enim nos tarditatis natura damnavit), with which compare Plin. Ep. 6.21.1: Neque enim quasi lassa et effeta natura nihil iam laudabile parit. The period had no great literary self confidence. ¹⁰⁶ Plin. Ep. 6.2.1. ¹⁰⁷ Tac. Ag. 3.1. ¹⁰⁸ Note especially Pliny on his lack of subjects for letters (Ep. 9.2); contrast the long letter about the Priscus trial (2.11). ¹⁰⁹ Plin. Ep. 3.20.12. ¹¹⁰ Tac. Dial. 41.4. This suggests a clue to the varying tones of the Dialogus. Aper’s bright and brash optimism reflects Tacitus’ youth under Vespasian, when the visitor from, say, the north of Italy would look out for Eprius Marcellus in the streets of Rome. Messalla speaks for the Quintilian view, formulated rather later in the century. Maternus’ final speech, filled with the vague nostalgia of the early Trajanic period, dispels the different optimisms of Aper and Messalla. ¹¹¹ Quint. 12.1.26. ¹¹² Cic. De orat. 2.55: Nemo enim studet eloquentiae nostrorum hominum nisi ut in causis et in foro eluceat.

Quintilian and the vir bonus

15

But perhaps in the long run Quintilian was more successful. Cicero was remembered, and Regulus forgotten.¹¹³ And the view of the orator as vir bonus dicendi peritus echoes, at intervals, down the centuries.¹¹⁴ [This early piece, much influenced in style and content by R. Syme, should be read in conjunction with Goldberg (1999). For Tacitus’ Dialogus de oratoribus, see also Winterbottom (2001).]

¹¹³ Except, apparently, by Martianus Capella RLM p. 453.1 Halm [= 3.432 Willis]. ¹¹⁴ So Fortunatianus RLM p. 81.5 Halm [= p. 65.5 Calboli Montefusco]; Cassiodorus RLM p. 495.5 Halm; Isidore RLM p. 507.16 Halm.

2 Problems in the Elder Seneca The text of the Controversiae and Suasoriae of the Elder Seneca is very corrupt. For thirty years at the end of last century it was a happy hunting ground of critical endeavour. Many wounds were healed; many remain. Others, however, were inflicted by the very scholars who sought to heal. Where declaimers so cunning and elusive as those excerpted by Seneca are at play, and where, so often, one has to make up one’s own context for isolated epigrams, the greatest care has to be taken in emendation. Is one curing—or merely misunderstanding? It is easy to be hypnotized by past emendations once they achieve the sanctity of print. One of the principal tasks of a new editor¹ of the Elder Seneca would be, with a clear head, to sift through the discoveries of the past, and appraise their varying merits. To take a few examples at random. In Con. 1.1 a son has been disinherited by his father and adopted by his uncle; now his uncle too is disinheriting him. The declaimer sings the son’s virtues: Quam multi patres optant similem filium! Bis abdicor.² ‘Any other father would be glad to have me—yet I get disinherited, twice!’ No very distinguished epigram, but a point. Yet Vahlen’s ab his abdicor was accepted by Müller and continued into Bornecque; it is infinitely feebler. In Con. 1.2 a priesthood is

[Bulletin of the Institute of Classical Studies 21 (1974), 20 42] ¹ That, on the basis of my recent Loeb (1974), I can hardly claim to be. This article does not discuss all the conjectures of mine there printed (most of the others are trivial variations on earlier efforts), nor does my Loeb print all the suggestions I make here. An earlier version of this paper was read to the Oxford Philological Society on 30 October 1970. The ensuing discussion produced some helpful comments, some of which are recorded in my footnotes. ² Con. 1.1.9 (20.11 M.). References are by page and line to H. J. Müller’s edition (1887; repr. Hildesheim 1963), on which I rely for all manuscript readings. Conjectures not mentioned by Müller are given references. I have not assembled lists of emendations with which I disagree. [The reference edition of the Elder Seneca’s declamations is now that of Håkanson (1989). An unpublished commentary by Håkanson on Sen. Con. 1 has just been edited by F. Citti, B. Santorelli, and A. Stramaglia: see Håkanson (2016).]

Problems in the Elder Seneca

17

sought by a woman who once was in a brothel and venientes ad se exorabat stipem. The declaimer says in accusation of her: Ego illam dico prostitisse; illa se dicit etiam mendicasse.³ ‘I say she was a prostitute; she says she was a beggar too.’ She asked, that is, for alms; for her it is a defence,⁴ for her accuser it is an aggravation of her guilt—she was on public sale, and she begged into the bargain. A palpable hit. Yet editors have persisted with variations on Lipsius’ vindicasse. At Con. 7.7.18 (337.23 M.) we have Fuscus’ colour for the crucified general who said Cavete proditionem from his cross: dixit alienatum iam suppliciis animum et errantem has voces effudisse sine argumentis, sine reo. Without, that is, naming a specific traitor and accusing him. What need of Thomas’s sine ratione? Many such defences of the paradosis could be mentioned;⁵ the mark of success is that what is transmitted is, once the penny has dropped, seen to be cleverer than what editors substitute. All the same, no one would claim that the paradosis of the Elder Seneca is faultless. This paper will be mainly concerned to tamper with it. I group my suggestions under the heads, ‘The Manuscripts’, ‘Using the Excerpta’, ‘Clausulae’, ‘Deletions’, ‘Additions’, ‘Other Conjectures’, and ‘The Greek’: though a certain amount of overlap between these categories is unavoidable.

T H E MA N U S C R I P T S One complication is the relationship of the three most important manuscripts of the ‘main’, that is the non-excerpted, text of Seneca, A, B, and V. Müller supposed that A and B descended from a hyparchetype x, while V descended from a hyparchetype x¹. And I gather from Dr H. D. L. Vervliet, to whom I am exceedingly grateful for much information on the manuscripts, that his further researches on the tradition do not alter this basic picture of the top of the stemma. Now, while AB give us a defective but uninterpolated text, V ‘ex alio codice (x¹) descriptus est, quem ipsum homo doctus, multis locis haud dubie suo tantum ingenio usus, non raro autem alium, ut opinor, eumque optimae notae librum secutus, felicissime emendaverat, cui tamen idem vir doctus, cum

³ Con. 1.2.10 (36.2 M.). ⁴ Professor Goold, to whom I am indebted for his helpful comments, objects that etiam must imply that the woman admitted the charge of being a prostitute. But she would really have said: ‘Yes, I was in the brothel, but I only took alms.’ The accuser twists this to: ‘I was a prostitute, and I begged as well.’ ⁵ I name honoris causa e.g. Summers’s defence (1911), 19 20 and 20 1, of adoravit at Con. 1.2.20 (41.5 M.) and of deprehensus at 1.4.2 (51.22 M.); Shackleton Bailey’s (1969), 323 of petis at 2.1.4 (108.7 M.); and Otto’s (1888), 132 3 of reducere at 2.1.18 (116.10 M.).

18

Papers on Quintilian and Ancient Declamation

scripturam depravatam facili negotio sanare posse sibi videretur, multas et eas manifestas interpolationes⁶ . . . inculcaverat.’⁷ It would, I think, be wrong to say that all the good readings in V that do not appear in AB are due to conjecture, though I suspect that a very large number are; I doubt if (looking at a sample fifteen pages of text) a medieval emender could have achieved Con. 1.1.5 (17.17 M.) naturam mutare (AB have natura militare), 1.1.7 (19.7 M.) essem scivisti cum (esse miscuisti eum), 1.1.23 (28.4 M.) arbitrum (arbitriudi), though very many of the others are easily within the grasp of anyone who knew Latin. It would be perverse, that is, to say that the coincidence of AB should always be preferred, at least as a starting point, to V. But so often do AB point towards, V away from the truth, that it is always worth, in any doubtful case, pondering the possibility that V is interpolated rather than transmitting inherited truth.⁸ x was, I take it, infinitely more like the archetype than V. And how Müller distinguished cases where V was drawing on a good outside source I do not know. I do not wish to labour this point: though it accounts for my practice in many of the passages I discuss of tacitly ignoring the phenomena of V. I point merely to the matter of the words in V that are omitted by AB, not the common-orgarden interpolations, but words that the editors accept with a will. First, note an instructive case where V is caught out making an addition of its own. At Con. 1.6.2 (64.7 M.) the text goes as follows: Artius nos fortuna alligavit quam ut orba posset divellere. Vidisses tectum pannis corpus, omnia membra vinculis pressa. So the manuscripts, except that (a) after alligavit ABV all give : and (b) when omnia . . . vinculis comes round again only V gives the word membra, while AB omit it. The conclusion is clear: when (in the archetype⁹ or behind it) (pan)nis corpus omnia vinculis intruded into the previous sentence, membra was not present to be carried along with the intruders. V or its ancestor x¹ will be responsible for the addition—and, I should judge, acting suo Marte, though no doubt correctly. Hagendahl gives an instructive list of words that appear in V but are lacking in AB.¹⁰ He starts from a striking case, by him misinterpreted. The thema of Con. 7.6 (318.15 M.) starts thus: Tyrannus permisit servis (servisi AB: servis ut V) dominis interemptis—the E(xcerpta) omit the last two words, but no ⁶ A flagrant instance in the poem of Albinovanus Pedo (Suas. 1.15 (529.19 M.)): . . . audaces ire . . . / †asperum† metas extremaque litora mundi. V gives hesperii (sic), Haupt, finely, ad rerum. ⁷ Müller (1887),  . ⁸ I give a trivial but typical example. The quarrelsome Scaurus litiganti similior quam agenti cupiebat evocare aliquam vocem adversariorum et in altercationem pervenire (Con. 10.pr.2 (447.10 M.)). Does not pervenire jar? It is V’s word; AB give vervenire. Did not the archetype give that too, by corrupt reduplication for venire? ⁹ For those who see point, as I do not, in calculating the length of the line in the archetype this passage will be of service. So will the anticipation of (carce)re vixerunt at Con. 9.2.1 (382.13 M.). ¹⁰ Hagendahl (1936), 312 13.

Problems in the Elder Seneca

19

matter. Then E continues the sentence dominas suas rapere: the right sense (cf. below: Cum omnes servi dominas suas vitiassent and esp. 7.6.13 (325.12 M.): tyrannus permisit dominas rapere, non coegit), and with no linguistic objection (for the infinitive see 7.6.13 (325.12 M.) again). But AB omit the words, V has a stop-gap: dominabus suis nubant. These words are perversely ignored by the editors, according to Hagendahl: ‘Quo rarior est forma illa dominabus . . . eo minus est, quod putemus eam interpolatam esse.’ On the contrary: the rarer the form, the more we require a Senecan parallel. None is forthcoming. And why nubant?¹¹—it is clear from everything in the controversia that the decree allowed rape, while the father who wanted his daughter to marry the ex-slave was going beyond the decree: plus servo dominus permisit quam tyrannus.¹² An interpolation, then, and one that reeks of its later date: -abus was particularly affected in the fourth century and later.¹³ The Thesaurus cites dominabus¹⁴ from Baudonivia’s life of St Radegund (c.600). Hagendahl appends a summary list, again instructive. a, ad, in, de, sub, per, et, si, sed, ut, quae, quam, se, me, non are found in V when AB omit them: a total of thirty instances,¹⁵ all the simplest and most necessary of additions, often with parallelism nearby to show the way. Hagendahl can register only seven more substantial items. They are worth examining individually. Con. 2.1.28 (122.15 M.): Cestius illo colore: quos abdicatione non potuit terrere, putat se castigaturum adoptione. ‘Non ille tuum filium concupiscit: suos corrigit. Dum illos correctos (correptos V: om. AB) putaverit, te satis minatum abdicabit.’ Cum (so C. F. W. Müller) will be right for dum. And if we give me for te (so Bursian) we get the good sense that it is the son who has done the threatening. But it is perfectly on the cards that Seneca wrote: cum illis putaverit me satis minatum, abdicabit. And even if he did not, the fact remains that correctos¹⁶ is the simplest and dullest addition: it picks up corrigit from the sentence before. The interpolator may be at work; he may even be wrong. At Con. 7.3.1 (298.10 M.) I am prepared to believe that V had reus vivet: vivet transmitted to it, despite AB’s omission of one vivet. But the extreme ease with which the word could be coincidentally omitted reduces the value of this instance. The same is true of 9.1.11 (378.20–1 M.), where V gives beneficium twice, correctly, AB only once, and of 9.2.9 (385.22 M.) atqui quid interest

¹¹ Quite apart from the fact that the word is properly used of women (as Mr M. D. Reeve re minded me). ¹² Con. 7.5.2 (319.15 M.). ¹³ Kühner Holzweissig (1912), 419 21. ¹⁴ TLL s.v. domina 1935.22. ¹⁵ But delete Con. 1.1.3 (17.6 M.), where non sit is probably wrong. At 2.6.4 (178.12 M.) coercet vitia qui provocat might stand as a question without non. ¹⁶ The interpolator may even have written correptos, as V gives it. Corrigo and corripio were constantly confused, at least in certain forms, and medievals may not have sharply distinguished between their meanings.

20

Papers on Quintilian and Ancient Declamation

convivium in forum an forum (an forum om. AB) in convivium attrahas, though this might be within the range of the alert emender. The case at 9.5.3 (415.12 M.) is helpful here: filios quos perdidisti non quaeris, quem quaeris non perdidisti. AB omit quem quaeris, understandably, but we know from 9.5.6 (421.6 M.) that Votienus said just this. V gives quaeris quem: the mark of the interpolator, ignorant of the term homoeoteleuton but alert to sense. Two final items. The gods, says one declaimer, have given their verdict in favour of an ex-captive and ex-prostitute who wishes to be a priestess: inter tot pericula non servassent illam dii nisi sibi (Con. 1.2.19 (40.15 M.)). All well. Yet V tacks on servata fuisset; the Excerpta add servaturi fuissent. Both interpolate; and, what convicts both, in different terms. Then at 7.2.3 (291.4 M.) we have Non magis quisquam alius occidere Ciceronem potuit praeter Popillium praeter Ciceronem defendere. The evident gap was filled by interpolation in V with the ungrammatical quam nemo pupillium after Popillium. Thomas was right to stress that V, or a forebear, is merely inventing, and right too, I think, to plug the gap with quam quisquam alius Popillium.¹⁷ The matter does not quite end there, for Hagendahl’s list admits of expansion. I can give a further twenty-two cases (there may be a few more). Significantly the pattern is very similar. The majority are small words, that a corrector could have added with ease. Thus ut (Con. 1.1.17 (25.9 M.); 2.4.1 (161.6 M.); 10.4.11 (486.1 M.): all essential to obvious constructions), an (1.7.11 (77.10); 6.th. (558.7 M.): both ‘formulaic’), in (2.1.34 (126.6 M.)), non (2.6.7 (180.11 M.); 10.1.7 (461.3 M.)), a(b) (2.7.9 (191.5 M.); 7.5.14 (318.4 M.)), de (7.6.2 (331.1 M.)), eo (9.pr.4 (371.20 M.), between usque and ut), cum (9.2.24 (393.2 M.), to support a pluperfect subjunctive), and si (2.4.13 (160.13 M.); 10.3.11 (477.19 M.)). Of the others, 1.6.1 (63.16 M.) spei merely completes an anaphora (perhaps unnecessarily, despite support from the Excerpta); 2.5.19 (174.11 M.) publicis is dictated by the contrast with privatis; while at 9.5.17 (421.18 M.) a sentence starting multa referam quae Montaniana Scaurus vocabat and proceeding uno hoc contentus ero cries out for a negative (ne, C. F. W. Müller), which the interpolator fumbled with non. At Suas. 1.1 (520.6 M.), the completion of post omnia oceanum nihil does call for some critical sense (omnia V), but this could be a case of coincidental error in AB. As to the three remaining cases: At Con. 1.5.7 (61.18 M.) the discussion is of three possible choices. Both girls choose death, both marriage, or one marriage, one death. Corruption has pruned this to aut nuptias optabunt aut altera mortem altera nuptias. Proceeding, AB give¹⁸ optaverunt, non poterit fieri quod utraque volet: uno modo poterit fieri quod utraque volet, si utraque mortem optaverit. It was left to modern scholars to supplement the opening

¹⁷ Thomas (1899), 161 n. 6.

¹⁸ Here as elsewhere I ignore minor variations.

Problems in the Elder Seneca

21

statement: best Gertz, who inserted mortem utraque aut before the first nuptias.¹⁹ In what follows it is clear that optaverint (so V²) is necessary, and that the bulk of a conditional clause giving the other two eventualities apart from si utraque mortem optaverit has fallen out: parallelism demands optaverint, as Gertz gave it. We cannot do without the first aut, and editors should not have been deceived by the fact that V has optaverint into omitting it. V’s addition is an interpolator’s—and the interpolator had forgotten the possibility of the girls making different choices. At Suas. 2.14 (540.15 M.) Cestius, cum descripsisset quos habituri essent si pro patria cecidissent, adiecit: per sepulchra nostra iurabitur, V has honores before quos: in that context not a difficult addition, though maybe not made in the right place (better before habituri?). Finally Abdica, inquit. Hoc pater versus. Quid adoptavit sperare possum? (Con. 2.1.3 (107.8 M.)). So, with much corruption, AB. V improves with quid , surely rightly.²⁰ It is difficult to see an interpolator at work here. Coincidental omission in AB? Or a proof of the independence of V from the common source of AB? If it is a proof, it is, I suggest, the only case among the fifty or so places where AB omits words given by V that is not attributable to an interpolator.²¹ It is a meagre haul; I do not dwell on it further, except to suggest that this is the sort of situation one would expect where three manuscripts, one highly interpolated, descend from an archetype with no lower common links. It remains true that other types of error in AB against V suggest that the traditional stemma is correct. It is odd that the evidence of omissions is so indecisive. Caution about ‘additions’ in V should be matched by caution about additions found in D, an agreed descendant of V.²² I note a cluster in Con. 10.4. At 10.4.8 (484.3 M.) we have A te (A patre Kiessling) fortasse aliquis acceptam stipem ad deos. D adds portat after stipem: it may, or may not, be the right verb in the right position (rhythmic considerations would place it at the end of the clause). Just below Lupa expositis infantibus, oblita feritatis, placida velut fetibus suis ubera praebuisse fertur. Sic lupa venit ad infantes.²³ Expositis infantibus is the addition of Haase, the first lupa appears in D but not in the primary manuscripts. Neither addition is necessary. The context, as almost

¹⁹ Though it is true that below utraque is followed, as is more proper, by a singular verb. ²⁰ The corrector of the Toledo manuscript rightly restored the sentence thus: Abdico, inquit. Hoc pater verus! Quid ab eo qui adoptabit sperare possum? ²¹ Contrast the sort of omissions that mark off A from BV (Suas. 1.14 (529.9 11 M.) quis mihi ponam), B from AV (Con. 1.8.8 (87.6 M.) patrem quia), V from AB (Con. 9.4.9 10 (407.14 16) possit patrem). Those gaps could not be filled by conjecture. ²² I do not enquire about the relationship between D and the corrector of T. That does not affect the present argument. But see Hagendahl (1936), 313 14. ²³ Con. 10.4.9 (484.16 M.).

22

Papers on Quintilian and Ancient Declamation

always, is lost, but the declaimer has clearly been talking of Romulus and Remus. The subject can be suppressed, and the dative be understood, till we come to lupa and to ad infantes in the next sentence. That is certain. Less clear 10.4.12 (486.13 M.): Illi singulos exponunt, tu omnes debilitas: illi spem, tu instrumenta vivendi detrahis. Here tu instrumenta vivendi is the contribution of D; tu is right—for the rest we cannot know.²⁴

USIN G THE EXCERPTA Clearly, the E(xcerpta), for all the adaptation which they have undergone,²⁵ are vital to the constitution of the text of the corresponding Controversiae where those are extant. And editors since Bursian have naturally made use of them for this purpose. But they have not always resisted the temptation to assume that words present in the excerpta must be inserted in the main text if they are lacking there. At Con. 9.2.3 (383.11 M.) we have in the main text the excellent epigram: Facilius est ut qui alia meretrici dederit homicidium neget quam ut qui hoc quoque dederit quicquam. Müller adds negarit from E 9.2 (435.16 M.); Bursian did better to add another neget. But E is making sure the reader understands—and providing a proper clausula (hence the form?). We do not need to follow. Editors are not always careful in comparing the main text with E. In Con. 7.1.24 (285.23 M.) Hispanus’ colour is twice given, thus: Hispanus duro colore usus est: Hoc, inquit, supplicium tamquam gravius elegi . . . et hoc colore per totam declamationem usus est, ut diceret hoc se tamquam gravius elegisse. E, summarizing this, gives: Hispanus, duro colore usus, hoc se tamquam gravius elegisse dixit supplicii genus.²⁶ Müller uses this as a justification for adding supplicii genus in the main text after elegisse:²⁷ speciously. Other cases may be more disputable. Tyrannus suspicatus est nescio quid istum de tyrannicidio cogitare, sive isti aliquid excidit, sive magna consilia non bene voltus (benivolis ABV) exigunt (so AB: exibuit V). So the main text at Con. 2.4.13 (161.16 M.). E gives: . . . sive non bene tegit vultus māgnă cōnsĭlĭă.²⁸ That may certainly stand in E—and Kiessling was wrong to suggest texit. What of the main text? Should not the generalizing present appear there also?²⁹

²⁴ But we may guess: illi spem detrahis. For loss of hope as worse than loss of life see Cic. Catil. 4.8: Eripit . . . spem, quae sola hominem in miseriis consolari solet etc. A verb may have dropped out of the illi clause. ²⁵ Hagendahl (1936), 299 ff. ²⁶ E 7.1 (348.21 2 M.). ²⁷ Con. 7.1.24 (286.7 M.). ²⁸ E 2.5 (198.26 M.). ²⁹ So Castiglioni (1927), 117, suggesting vultus contegit. Cf. Sen. Thy. 330 1 multa sed trepidus solet / detegere vultus.

Problems in the Elder Seneca

23

Perhaps then tegit; or, bearing in mind exigunt, tegunt (cf. Cic. De orat. 2.148 voltus . . . perspiciamus omnis, qui sensus animi plerumque indicant). At Con. 10.4.6 (482.21 M.) Fuscus appeals to the judges to pity a group of cripples in court just as they pitied them individually on the streets: Miseremini horum, iudices, [et] misereri etiam singulorum soletis. E has: Miseremini omnium, iudices, quorum singulorum misereri soletis.³⁰ Müller compromises in the main text with Miseremini horum omnium, iudices, quorum misereri etiam singulorum soletis. We may well do without quorum—two parallel clauses give a good bite. And I think we should do without horum as well; it will be ABV’s corruption of omnium (which is itself, of course, essential to the contrast). At Con. 9.5.14 (420.6) Varius’ epigram for the grandfather who kidnapped his grandson after two previous grandsons had died is thus given by the main text: Quae est ista aut tam (aut tam AB: aucta V) praepostera? Quaerere tuos a tertio incipis. Something is clearly missing, and E confirms: Quae ista est tam sera pietas, tam praepostera? Quaerere tuos a tertio incipis.³¹ The logic of this is right. The father’s desire to get his third son back shows affection that is both late in the day and topsy-turvy, not ‘si tardive ou si intempestive’ (Bornecque). We should therefore delete aut in the main text (dittography between -a and t-), and supplement merely tam , not, keeping the aut, (so editors since Bursian). As to the manuscript tradition of the Excerpta, I can only judge from the information in Müller’s apparatus. Müller’s own judgement is that by far the best manuscript is the Montepessulanus (M), because it alone ‘interpolationibus, quibus reliqui scatent, prorsus liber sit . . . nec tamen reliquos neglexi, cum et complura in M perierint et satis multa in M perperam scripta ex uno vel altero recentiorum codicum medelam accipere viderem.’³² In a sense the supremacy of M should, I think, be put more highly. But the impression that it contains almost all the evidence of value is artificially heightened by the activities of the ‘recent’ corrector called by Müller M³, who saves the manuscript from many errors³³ not found in other manuscripts, and may, in part, have been drawing on those other manuscripts. It would be wrong to suppose that all the other manuscripts descend from M. But so much more reliable is M in general that it is always worth pausing before rejecting its readings. Thus at E 6.6 (262.17 M.), where the main text is not available, an advocate is replying on behalf of a wife accused of poisoning and adultery, partly because she had said of her daughter ‘She will die sooner than be married’: exciderunt illi verba quae non minus quam pater filiam luget. This makes ³⁰ E 10.3 (513.15 M.). ³¹ E 9.5 (442.15 M.). ³² Müller (1887),  . ³³ Often trivial or orthographical. M³ also introduces errors (sometimes available in other manuscripts): e.g. at random E 1.3 (95.18 M.) dubitari (M³P); E 2.7 (203.5 M.) animus (M³P); E 7.7 (356.15 M.) nec (M³P).

24

Papers on Quintilian and Ancient Declamation

sense, though unexciting sense. But pater is supplied by the editors from two minor manuscripts, and M omits it. Without it we have point: she regrets her words, and she regrets her daughter too (both regrets suggesting that she was not the poisoner). In a second case (E 5.3 (246.17 M.)) M stands alone against five other MSS quoted by Müller. A mother accuses her husband of encouraging their two sons to kill each other in the pancration at Olympia: Misera mater non odisse non potest qui filium suum occidit. Sense of a sort: the mother hates the father.³⁴ But why is only one son mentioned when both died? Bornecque actually emends to filios suos. But observe that M omits the first non. Now there is point. The misery of the mother lies, partly, in the fact that she cannot find consolation in hating the killer of either son taken individually, because A killed B—but B killed A, and both are her own kin. Nor should undue trust be placed in M’s corrector. In the lex of Con. 6.5 and again in the thema (260.7, 14 M.) M gives qui (quod) vim iudicio fecerit; in each case a corrector adds in before iudicio. But the force is applied to the judging; and there is a decisive parallel for the omission of in in the parallel theme of ‘Quint.’ Decl. min. 386.

CLAUSULAE Much still remains to be discovered about clausulae in the declaimers.[³⁵] Bornecque claimed that they differed in their practice, and differed consistently—so that, for instance, Latro is ‘beaucoup moins régulier que Fuscus et Pompeius Silon qu’Albucius’.³⁶ What is more certain is that the excerptor imposed on what he found in Seneca a much tighter conformity with the rules of prose rhythm. This was convincingly demonstrated by Hagendahl³⁷ and need not be further discussed here. The textual critic, then, will be unwise to disturb, but equally unwise arbitrarily to impose, a clausula in the text of Seneca himself; while in the Excerpta he should never disturb and may venture to impose. In Seneca, then, we shall not tamper with Suas. 5.2 (555.17 M.): Potest (sc. Xerxes) maior venīrĕ quām vīctŭs ēst³⁸ nor with Con. 10.5.1 (493.16 M.): Nemo, ut naufragum pīngĕrēt, mērsīt̆ (despite the excerptor’s out-of-character addition of hominem at the end of the sentence): in each case the increased snap of the sentence rewards the conservative. In a highly rhythmic passage from the infant prodigy Alfius Flavus we can hardly keep ³⁴ Cf. the preceding epigram: Moriuntur non alter ab altero sed uterque a patre. ³⁵ [But see now Håkanson (2016), 10 16.] ³⁶ Bornecque (1902), 26 8. ³⁷ Hagendahl (1936), 300 7. ³⁸ And as Thomas (1899), 209 observes, if you can say maior venit you can say maior victus est. For similar cases see Löfstedt (1936), 26 8.

Problems in the Elder Seneca

25

̆ (Con. 1.1.23 Audimus . . . īmpĭās ĕpŭlās, detestabili parricidiō fŭtūrūm dĭēm (28.5 M.)), but we should replace the offending word with a metrically equivalent one: I suggest fugatum (Aen. 10.257: noctemque fugarat). The allusion is clearly to Thyestes,³⁹ in horror of whose crime the sun turned back in its course: cf. Prop. 3.22.30: nec tremis Ausonias, Phoebe fugate, dapes; Sen. Thy. 776–8: O Phoebe patiens, fugeris retro licet / medioque ruptum merseris caelo diem, / sero occidisti (cf. 120); and parallels amassed by Owen (1924) on Ov. Tr. 2.392. Similarly, when additions have to be made, they may as well produce rhythm from lack of it: so at Suas. 7.6 (578.18 M.) supplement: Lepidus . . . nōstĕr dŏmĭnūs̆ rather than dominus ; and perhaps at 2.19 (544.10 M.) at nunc cuilibet orationes in Verrem tuto licet pro suis (suo BV: sua A) we should place Kiessling’s dicere at the end of the sentence.⁴⁰ Thus, more confidently, in the excerpts. We may assume that the true reading at E 3.6 (219.4) is quare nullam aliam domūm tyr̆ ānnūs pĕtît (petiit M.): cf. Quint. 7.1.61: quando id praestiterit quod advocatus petît. In E 3.9 (223.14 M.): interiit omnis potestas si vivorum imperia neglexerint, mortuorūm trĭbūnī, where sense demands the addition of servi (so C. F. W. Müller), wordorder and rhythm⁴¹ demand that the word should follow neglexerint. And in E 8.6 (367.14 M.): Lacrimis inter verba manāntĭbūs vĕnĭō: talis et filiae nuptiis fui, we may flirt with nūptĭīs ādfŭī ( filiae nuptiis Castiglioni).⁴²

DELETIONS Glosses undoubtedly disfigure the text of Seneca and his excerptor. For the latter one need point only to Con. 4.pr.2 (225.7 M.): Ille triumphalis senex ἀκροάσεις [tuas id est declamationes] suas numquam populo commisit, in all manuscripts. For Seneca himself there is a clear case at Suas. 6.13 (565.9 M.): Varius Geminus says enemies are often won over, ipsum (i.e. Cicero) exoratum a Vatinio Gaio quoque Verri adfuisse. Müller very properly deleted a, but Gaio quoque Verri should go as well. Declaimers did make mistakes and they did invent, but this would be extraordinary as mistake or invention in the first century; in any case quoque points to a marginal addition. In a number of other passages the text yields more easily to deletion of words than to their manipulation into the context. At Con. 1.6.12 (69.17 M.) Glyconis valde levis et greca sententia est has not been improved either by Müller’s ut Graeca ³⁹ As in Con. 1.1.21 (27.5 M.). ⁴⁰ But this is the only casually rhythmic Seneca himself, not a declaimer; and I suspect that he wrote: at nunc quilibet (so ABV) . . . tuto dicet pro suis. The plural orationes, assailed by Müller, is certainly hyperbolic; it is meant to be. For pro suis cf. Con. 1.pr.10 (6.3 M.). ⁴¹ Rhythm alone would not suffice, for impĕrĭă nēglēxĕrīnt would serve. ⁴² Castiglioni (1927), 121.

26

Papers on Quintilian and Ancient Declamation

(‘considering that it was Greek’?) or by Thomas’s e Graecis (what of the order?); et greca should go—it will have originated in a marginal note drawing attention to the inscrutable hieroglyphs of uncial Greek that follow.⁴³ In Con. 2.1 Fuscus is making his division: I say that I had the right to refuse adoption. I say that even if I did not have the right, recte tamen recusasse: et quod iure fit verum est et quod sine iure, quod quidem rationem habeat, recte fit (Con. 2.1.19 (117.4 M.)). Bornecque translates verum est ‘est juste’, which could be the meaning of the words.⁴⁴ But they are surely a marginal comment by an approving reader, and without them the balance of the sentence is much improved. In any case, Fuscus is talking about the proper use of the word recte, and verum blurs the issue. At Con. 2.6.10 (182.5 M.) the glossator may have been more ambitious: Silo Pompeius patronum adulescenti dedit; quod non putabat in accusatoris persona Latro faciendum, ut aliquis per patronum accusaret patrem. The sentence can hardly stand as it is, and Gertz supplied: numquam certe esse factum before ut. But it may be that we should delete ut—patrem as the addition of someone who felt that quod required further elucidation. At Con. 2.1.8 (110.1 M.) Fuscus has remarked that both Croesus and Crassus came to bad ends despite their riches. Dicta praeterea ill(i)a corruentium non refero quotiens enim inter divitias suas exemplo is tam posueritis domum meliores perdentem divitiis suis liberos. One thing is clear: Fuscus is saying that the father (hence Ribbeck’s posueris) has often pointed out to his son, as an example of disaster amid riches, the house of the rich man who had to disinherit his three sons. The complication is that Fuscus has already said this, in very similar words: Mille corruentium inter divitias suas exempla referebas et inter illa ponebas et divitis domum.⁴⁵ And for once this is a continuous piece of declamation where, though an idea may be repeated, its wording should not be. Wiles was surely right to call for the deletion of corruentium and inter divitias suas exemplo.⁴⁶ But he accepted the emendation Sescenta (Ribbeck) praetereo (C. F. W. Müller) alia (Gertz), and regarded the deleted words as a note explaining that. Rather we should regard illia (illa in V) as the remains of mille in the earlier passage, and dicta praeterea as the glossator’s own words introducing his parallel. That leaves us with non refero quotiens enim istam . . . Enim must go, and we cannot do without Gertz’s inter illa (cf. 108.2)⁴⁷ to give us the link we need with the examples of Croesus and

⁴³ gr(aeca) is sometimes found in certain twelfth century manuscripts of Quintilian to replace words that the scribe despaired of copying. ⁴⁴ Cf. Caes. Gal. 4.8.2: neque verum esse qui suos fines tueri non potuerint alienos occupare; Cic. Quinct. 48; Dom. 21, etc. ⁴⁵ Con. 2.1.4 (108.2 M.). ⁴⁶ Wiles (1922), 69. That is not exactly what he says, but it is, I think, what he means. ⁴⁷ Illa used generally = ‘those cases’. In the earlier passage exempla is understood. But here we can hardly write inter illos because of istam domum. Perhaps we should read inter illa exempla?

Problems in the Elder Seneca

27

Crassus. The words may best be placed after quotiens, where inter divitias suas exemplo has displaced them. There is a doubtful case at Con. 10.4.11 (486.2 M.). A man is charged with ‘harming the state’ because he has crippled children and forced them to beg for him on the streets. It is argued that the state is not harmed if one child is crippled—for someone who kills once is charged with murder, not with harming the state. So with two and even higher numbers of murders. Duo debilitantur: nondum res publica iuvenes qui suadere infantes perdidit et infelices. Potuerunt, inquit, duces fieri. We have clearly by now got on to a new argument. But the old one was never finished. The declaimer has yet to deal with the case of more than two being crippled and to draw his conclusion. Vahlen was therefore right to supply laesa est after res publica; and we can either mark a lacuna after these words, or, with Gertz, insert plures debilitantur after duo debilitantur. What, then, of iuvenes—infelices? Minus iuvenes, Vahlen and Müller insert the words a little lower down: Attamen crudelem rem facit et lanista . . . The idea is ingenious, and methodologically sound. Yet by this stage we hardly need definition of the subject of facit, and the addition of a second facit is enough to heal that context. I incline to suppose that iuvenes—infelices is the remains of some title to the whole piece, and should be deleted. In other cases, dittography is responsible for additions to the text. At Con. 1.5.4 (59.6 M.), Latro argues that not all seducers can have their lives saved at the wish of their victim: Si legatus . . . exire debet, peribit; si militare debet, peribit; si lege ducere debet, peribit . . . If Müller was right to see in the third instance an allusion to judges, we should read not, as he did: si lege ius dicere, but merely si ius dicere. Lege will come from legatus in a similar position after si above, and will have displaced the rightful ius. In the declamation of Fabianus in Con. 2.1 many problems will never be solved. But two minor ones may be due to dittography. At 2.1.11 (111.10 M.) Fabianus is attacking civil strife: An, ut convivia populis instruantur et tecta auro fulgeant, parricidium tanti fuit? Magna enim vero laucia sunt propter quae mensam et lacunaria sua ⁴⁸ potius quam lucem innocentes intueri maluerint. V bravely gives us et laudanda for laucia, and Müller (rather less well) et lauta. But laucia is better out of the way; it will be an anticipation of lacunaria, or a misplaced variant on it. Again, at 2.1.13 (113.5 M.): Vix possum credere quemquam eorum vidisse silvas patentisque eamme campos, V was perhaps right to delete eamme⁴⁹ (virentisque gramine Müller). It may have arisen from (quemqu)am e(orum) above. ⁴⁸ Supplied by Novák (1909), 260: much the best contrast to innocentes. At Con. 10.1.7 (460.16 M.) the same word is supplied by the same scholar to provide the same contrast; there it should perhaps replace rather than precede nos. ⁴⁹ Others have come to this conclusion; but my explanation of the intrusion is, I think, new.

28

Papers on Quintilian and Ancient Declamation

At E 4.7 (239.19 M.) the declaimer says that if anyone kills a tyrant, it should be a patriot, not a rival for the tyrant’s wife: Occidat illum civis iratus, misceat maledicta vulneribus, qualia in adulterum maritus adulterat ab adulterae osculis ad praemium curris: nolo tyrannicida imitetur antequam occidat tyrannum (the addition is Gertz’s). With ab starts a new, though related, epigram. We should read adulter at the end of the previous one (so Gertz, Müller), and regard at as dittography of ab (or conceivably as the result of a desperate attempt to provide a verb in a lacunose sentence). There is no need for tu, let alone at, at the start of the new epigram. At Con. 7.1.4 (275.15 M.) the declaimer tells how he found an old wrecked boat on the shore and planned to set his condemned brother afloat in it: Inveni relictum etiam a naufragis navigium,⁵⁰ fragmentum, infelix etiam navigaturis omen, quod si quis gubernator vidisset, iter suum distulisset. Müller was right to feel that there is an etiam too many; but he chose the wrong one to delete. The first has its point: even the shipwrecked sailor can find no use for such a hulk. The second is not wanted. The boat is a bad omen for those proposing to sail in it: what group would etiam contrast with the navigaturi? We have a parallel at 7.1.8 (277.9 M.): the same boat is an infelix omen navigationis. At Con. 7.7.13 (335.18 M.), Albucius is arguing that at an election aiebant alii imperatorem fieri debere ,⁵¹ qualis Scipio fuisset, alii senem, qualis Maximus fuit. The variation fuisset/fuit would in any case sound grotesque. But fuit appears only in V; A has fent, B fecit. I suggest that B too was trying to make sense of a word it could not read, and that A comes nearest, in its gibberish, to the truth: fent is a dittography of senē.⁵² In the controversia on the prostitute who asked Flamininus to let her watch an execution, Fulvius Sparsus gives her words: Hominem occidi numquam vidi. His reply to her is: Alio quid (quod B¹: quem V) Flaminino praetore omnia alia vidisti.⁵³ There seems little point in Bursian’s alioqui,⁵⁴ which Bornecque translates—as the sense but not the word demands—‘donc’. Surely delete alia as an anticipation of alia and be content with quid? At Con. 10.pr.3 (447.14 M.), where in a preface transmitted both by M and by ABV our tradition is at its best, Seneca, after various compliments to Scaurus, starts his criticisms of that declaimer thus: Sed ex his omnibus sciri potest non quantum oratorem praestaret ignarus carus sed quantum desereret. There is no doubt that Scaurus lurks here; and indeed M transmits prestares caurus, corrected to prestaret scaurus. This correction is right, and editors

⁵⁰ Mr N. G. Wilson suggests the deletion of this word; fragmentum then goes directly with relictum. ⁵¹ Surely this is the place for the addition? And the manuscripts give deberem. ⁵² For the type of error x y see Winterbottom (1970b), 189, 200. ⁵³ Con. 9.2.5 (384.7 M.). ⁵⁴ ‘Otherwise’? But that spoils the punch of omnia alia.

Problems in the Elder Seneca

29

should not have tried to rescue ignarus carus (ignavus Scaurus Bursian): icnarus and carus were competing shots at scaurus. I feel more doubt about Con. 9.2.15 (388.13 M.): Is laedit populi Romani maiestatem qui aliquid publico nomine facit: tamquam legatus falsa mandata adfert, sic audiuntur tamquam illa populus Romanus dederit; imperator foedus percussit, videtur populus Romanus percussisse . . . An elegant writer would not have written the first tamquam, and it may be an anticipation of the second. But Seneca uses the word to introduce an example just below, in 9.2.21 (391.10 M.), and he may be being careless. I append two possible cases of intrusion from a distance.⁵⁵ At Con. 9.2.10 (386.4 M.) an execution is being described: Fit a praecone silentium. Adhibentur deinde legituma verba. Canitur ex altera parte classicum. Agreeable as it is to have a trumpet sounding from the other side in a classical context, one is hard put to it to give a meaning to ex altera parte. Could the words come from 9.2.21 (391.8 M.), where they might have, as a sort of heading, been at least partly in the margin? Less extravagantly, at Suas. 2.17 (542.10 M): Seneca fuit . . . qui cupiebat grandia dicere adeo ut novissime morbo huius rei et teneretur et rideretur, it may be that dicere intrudes from 2.19 (544.11 M.), where editors have quite independently wished to insert that word (see above, p. 25). That the word should go, as Gertz desired, is undoubted; the examples that follow are of Grandio’s taste in outsize slaves and silverware. A number of other passages would be improved by deletions. In Con. 2.5.4 (163.3 M.) Fuscus has given two vignettes of tyrannical torture, each followed by the word tacet—the woman keeps quiet, so as not to betray her tyrannicide husband. Fuscus summarizes: Plus tibi praestare non potuit si de te liberos sustulisset. He proceeds with another vignette, on which he comments: Res publica an sit tibi ista datura liberos nescio: tyrannicidam dedit. Vignette and comment are separated in the manuscripts by et (licet in V). Schultingh emended to tacet, and he could be right. But I feel that Fuscus, after his summary, has gone off on a new tack, and dropped the tacet motif. There is something else. In the two certain tacet cases, the word completes the clausula: mināndŏ⁵⁶ tōrquēt: tăcēt and oculōs mĭnācēs: tăcēt.⁵⁷ This would not be the case in our sentence, whether we add the necessary sanguis after vitalibus (as I should prefer: we thus gain a clausula without tacet) or elsewhere. ⁵⁵ I may mention in this connection the words Ceteri quidem fugerunt at Suas. 2.4 (535.2 M.). They make no point there, and embarrass when the next sentence starts Si me quidem. But at 2.7 (536.10 M.) they fit very snugly before the retort Ego vero quod discesserunt (treceni discesserunt D, interpolating) gaudeo and indeed Gertz there added at ceteri fugere. ⁵⁶ ‘[B]ei den Späteren seit dem Tragiker Seneca ŏ im Abl. des Gerundiums’ (Kühner Holzweissig (1912), 112). The trick of making the clausula cover a break is common; see e.g. [Quint.] Decl. mai. 4.17 (79.12 Håkanson): ‘Perveniet ad iuventae rōbŭr:’ ădŏlēvī, and the following sentences. ⁵⁷ et tacet codd., correctly pruned by Kiessling.

30

Papers on Quintilian and Ancient Declamation

Et should also be deleted at Con. 7.1.21 (284.21 M.): In narratione hoc colore usus est [et]: dixit: Hoc unum mihi praesta beneficium. The editors’ usus est et dixit: destroys the formula,⁵⁸ and leaves unclear the context of the following words. Without et we can take dixit as: ‘My brother said to me.’ Gertz’s supplement: et dixit goes to too great lengths to save the intrusive et. In Con. 1.1.9 (20.15 M.) Hispanus is arguing that it is unreasonable to agree that a man should receive alms qua human being, citizen, etc. but to deny them to him qua father. Homo est, civis est, amicus est, propinquus est. Ergo non erit vitium porrexisse stipem nisi dixero: pater est? Ergo starts the sentence as it does in 1.1.3 (17.2 M.): Ergo fame morientem videbo per cuius cineres iuraturus sum? and in the same tone of voice. In the manuscripts condicione intervenes between est and ergo. Editors have boosted it with ea or ista; but though I cannot explain the intrusion, I should prefer to see condicione firmly removed. At Con. 1.2.23 (43.3 M.) Grandaus is discussing the controversia de illo qui tribadas deprehendit et occidit; cum diceret in eadem controversia: ‘Non ideo occidi adulteros non paterentur’, dixit: Εἰ δὲ φηλάρρενα μοιχὸν ἔλαβον. Schultingh patched the Latin up with an ut after occidi. But surely Grandaus must, before εἰ δέ, have been discussing the hypothesis of one of the tribades not being φηλάρρην. In that case, ‘they would not allow male adulterers to be killed on that pretext’ (i.e. merely being in bed together: cf. up to a point 1.7.3 (72.12 M.)). ‘But if, as was the case, I found them taking more active measures (or if I had found them) . . . .’ This requires the deletion of the second non. At Con. 2.7.7 (190.6 M.), Latro (in an unusual piece of continuous declamation) laments that if he is to add a note to his will explaining why he leaves his property to his wife, he is forced by events to use the exact words used in his will by the rival to her affections: iam moriturus tabellas occupare si voleti cum muneribus meis inponere elogium, ex testamento adulteri petendum est. Here volo et goes without saying, whereas volo et ei (printed by Müller) is motivated by desire for a dative with inponere as much as by palaeographic considerations. But the elogium is attached to will rather than wife, and I suggest the deletion of cum as well as the stray i; both perhaps derive from creditum just above. A variant may be in question at Con. 7.6.8 (322.12 M.). A son is asking that his sister be allowed to marry a respectable citizen, not a manumitted slave: Habeamus generum, si possumus, parem similem: si minus, non erubescendum . . . There is little difference in sense here between parem and similem (note the variatio in 7.6.20 (328.11 M.): feci filiam meam ceteris similem, ⁵⁸ For which see e.g. Con. 7.1.22 (285.3 M.). Contrast 7.1.21 (284.15 M.): Et cum colore dixit, 7.16.18 (327.16 M.): hoc colore usus dixit; and see below, p. 40.

Problems in the Elder Seneca

31

fortunam meam publicae parem),⁵⁹ and rather than link the two we should delete one. Rhythm would suggest that similem should stay, but the passage is not highly clausulated.⁶⁰ At Con. 7.8.2 (340.10 M.) a ruthless victim of rape is not moved by the tears of relatives, and insists that the rapist go to court: Ecquid te horum lacrimae movent? Non, inquit, ad illum magistratum veniat. Madvig wrote magistratus, but there is no question of the magistrate doing the coming. Rather than change to ille and have to alter the word order too (so Bornecque’s text), I prefer to delete illum. In the same controversia (7.8.8 (343.9 M.)) the girl argues that she has not yet had her optio because the forms of the law were not fulfilled. Si praetor defuisset, numquid optionem vocares? Si rapta defuisset? Raptor defuit (defuit Bursian: fuit AB: non fuit V): non est ista optio—sermo est. As the text stands it is defensible, though not altogether elegant: if si rapta defuisset stood after si praetor defuisset (or even before it), one would avoid the gap that seems to loom before raptor defuit. Editors, however, with one mind change to si raptor defuisset, which removes the snap altogether. We shall do best to delete si rapta defuisset (which Thomas did)⁶¹ and keep the rest unaltered (which he did not): ‘If there had been no praetor, you wouldn’t call it a choice, would you? (Worse!) There was no seducer.’ The pointedness is typical of declamation. At Con. 10.1.3 (458.11 M.) Fuscus says: Cum inspoliatum cadaver meum inventum sit, quis fuerit percussor nescio. It was the speaker’s father’s body that was found, not his own. Kiessling may have been right to give cadaver mei. But whose the body was does not need to be laboured, and we should perhaps merely delete meum. I should also like, no less arbitrarily, to delete a word in Con. 10.2.17 (472.15 M.): Timeo ne ob hoc ipsum patri vilior fiam ego: scimus quam gloriosus sit. Novák was right to stress the wrong-headedness of ego (he might have mentioned that vīlĭōr fīām gives rhythm), wrong, I think, to do anything but delete it (it may intrude from line 14 just above).⁶² I shall not argue here about the general way in which the perplexed sentence at Suas. 6.16 (566.20 M.) should run; I merely set down the results of a series of emendations of it (mainly by Gertz): sed fortasse efficiam ut his sententiis lectis solidis et verum habentibus recedatis; et, quia hoc †si tam† recta via consequi non potero, decipere vos cogar . . . By hoc Seneca means the enticement of his sons away from the declaimers towards

⁵⁹ If that is right; but parem is Gertz’s emendation for partis, and the sentence should perhaps end with ām mĕām pūblĭcae. ⁶⁰ Cf. Quint. 10.1.55: cui se parem credidit, where one group of manuscripts gives cui se aequa lem credidit parem, and another, inevitably, cui se aequalem credidit. ⁶¹ Thomas (1899), 251. ⁶² Novák (1915), 273. ⁶³ (1927), 124.

32

Papers on Quintilian and Ancient Declamation

better things (just as at the beginning of the preface to the last book of the Controversiae he had himself felt the need ab istis iuvenilibus studiis ad senectutem meam reverti: 10.pr.1 (446.5 M.)). The sentence runs smoothly without si tam; Bursian’s propositum could be right, but it hardly imposes itself, and Gertz’s si eam is more ingenious than convincing. I propose the deletion of the two words, and suppose that their intrusion is not unconnected with salutarem below.

ADDITIO NS The text of Seneca is lacunose as well as burdened with glosses and accidental additions. Some of the gaps result naturally from the peculiar nature of the text: homoeoteleuton may abound where so many clauses are juxtaposed in jingling parallel. So at Con. 10.2.2 (466.19 M.). A son who has been disinherited by a father who had lost to him in competition for a prize in bravery says: Dicam abdicanti: non luxuriabor, non amabo. Hanc emendationem criminum meorum non possum promittere. Ego vero pugnabo et fortiter et fortissime. We clearly need some explicit statement of the emendatio that he cannot offer, the promise not to fight bravely. It should go not before hanc (so Müller) but after promittere, and preferably in the order non fortiter pugnabo (not, as Gertz, non pugnabo fortiter)—the three words dropped out because the scribe leapt on to another trio of words that ended with pugnabo. Again, rhetorical parallelism will often, at the least, hint that something is amiss. In Con. 2.4 a disinherited son betakes himself to a prostitute and has a son by her. On his death, he commends their son to his father, who adopts the child. The father is accused of madness by a second son. For the father one declaimer says: Omnes aliquid ad vos inbecilli, alter alterius onera, detulimus:⁶⁴ accusatur pater in ultimis annis, nepos in primis abdicatur nullus.⁶⁵ Seneca has just told us that the epigram forms a tricolon;⁶⁶ and the shape is dictated by what is left. The father is accused in his last years, of madness; the grandson— the word order insists—must be accused in his first years (of being a prostitute’s son or a bastard? cf. e.g. 2.4.5 (154.2 M.): Pater istius incertus est: bene cum ipso ageretur si et mater); the third colon can hardly be very different from mātĕr īn mĕdĭīs—the mother is accused of being a prostitute. Abdicatur nullus is quite unintelligible; if it (as Gronovius thought) covers abdicatur filius, that could be no more than an interpolation attempting to complete the sentence— it would do so with no relevance or elegance. ⁶⁴ This sentence I do not understand. ⁶⁵ Con. 2.4.12 (158.18 M.). ⁶⁶ That what I conjecture is a proper tricolon may be seen from the tetracolon pointed out by Seneca at Con. 9.2.27 (394.6 M.).

Problems in the Elder Seneca

33

Rather similarly, it may be that in Con. 9.6.17 (429.13 M.) nominabo⁶⁷ istam quae patri adfuit, istam quae mortuo fratre †a† (ac V) matre non flevit we should follow the call of rhetoric and give fratre matre non flevit rather than take account (as does the traditional ) of the a(c) of the manuscripts. Other lacunae are more casually motivated. Con. 1.5 concerns a man who seduces two girls on a single night. One chooses his death, the other marriage to him. Fuscus (1.5.7 (61.11 M.)) shows that only if the seducer dies can both girls have their desired revenge (see above, p. 20). Before examining all the possibilities of the girls’ choices, he says: Quod vult eligat. What is the subject? Clearly the girls—what the seducer wishes is irrelevant. Why then the singular? Hence Schultingh’s Quod volunt elegant—a bold change. Easier, surely Quod vult, eligat (cf. 1.8.11 (88.13 M.): quos ut quisque volet interpretetur). At Con. 1.6.4 (65.11 M.) there seems a failure of point. In imperial Rome nihil est humili casa (that of Romulus: see 2.1.5 (108.8 M)) nobilius: fastigatis supra tectis auro puro fulgens praelucet Capitolium. The golden Capitol is there to provide a luxurious contrast with the lowly hut; yet the contrast is hardly pressed home, and praelucet even gives the impression (clearly not in the declaimer’s mind) that the Capitol is being said to be more brilliant than the hut. We can translate,⁶⁸ as I do: ‘though above it shines out the Capitol with its sloping roofs, gleaming in pure gold.’ But I should prefer to add cui nec before fastigatis: even the gilded Capitol cannot outshine the dwelling of Romulus. For the dative, cf. Hor. Ep. 1.1.83: Nullus in orbe sinus Bais praelucet amoenis. A woman has avoided betrayal of her husband under torture. Alia pro incolumitate mariti vicaria morte decidit (i.e. Alcestis): creditisne hanc in tormentis oppressam? Mortem amplius pro viro praestitisset, si quid amplius exegisset tyrannus. So the received text at Con. 2.5.8 (166.8 M.). Bornecque translates the last sentence: ‘Elle aurait donné davantage, sa vie, pour son mari . . . ’ This sounds better in French than the corresponding Latin. Is it not clear that the snap of the epigram demands that amplius should answer to amplius uncluttered by mortem? And it is equally clear that the previous pair of sentences would be improved if morte were picked up by an allusion to death in the creditisne . . . sentence. We need to transfer mortem to the previous sentence, and we may do so by inserting a verb of fearing, perhaps horruisse, before it; or better still, as Mr D. A. Russell has suggested to me, by substituting horruisse for oppressam. ‘Alcestis died for her husband; this ⁶⁷ Opitz (1888), 280 rightly pointed out that there is no need to add anything here. The mother (revealing her true motive) speaks; and nominabo = ‘I will name as my accomplice’ (so Con. 9.6.16 (429.2 M.)). ⁶⁸ Not, as Bornecque, ‘au dessus des toits qui dressent leurs faîtes élevés d’or pur, brille . . . le Capitole’.

34

Papers on Quintilian and Ancient Declamation

woman would not have hesitated to die too—she would have done more for her husband (i.e. died rather than merely undergoing torture) if the tyrant had demanded more.’ Luxuria suits the young, not the old: Luxuriosus adulescens peccat, at⁶⁹ senex luxuriosus insanit.⁷⁰ So far so good. There follows: aetas exhaurit vitia lasciviunt. Many attempts have been made to put this to rights; but if we add the one word virtutes after exhaurit (a word that could easily slip out before vitia) we gain an epigram: ‘length of years drains away virtue, while vice goes wantoning on.’ This will be related, though not intimately connected, to the epigram that precedes. In the thema of E 6.1 (254.6 M.) the scene is set thus: Abdicato fratri chirographum dedit dimidiam se partem daturum hereditatis si non respondisset. Bursian changed fratri to frater. Themes certainly do exist where a relationship is thus briefly stated (E 6.4 (258.19 M.): Proscriptum uxor secuta est); but it may well be that quidam has slipped out before abdicato. The word constantly introduces themata: so e.g. E 6.2 (255.14 M.), 6.3 (257.7 M.), 6.6 (261.13 M.). In Con. 6.3 an illegitimate younger brother has to choose either the whole of the estate left to him and to his legitimate brother or his own slave mother. The declaimer exhorts him: Elige ut aut patrimonio careas aut scelere.⁷¹ But these are not the alternatives. The scelus is not to choose the mother; in choosing to do without the estate the son would in fact be choosing his mother. The declaimer is being illogical—or his text is lacunose. We need a second verb with scelere to contrast with careas: perhaps fruaris—‘enjoy your possession of a guilt-ridden property.’ Part of the point would then lie in the fact that fruor is normally used of good things; cf. the contrast in Sen. Ep. 55.7: incommodis illarum caret, voluptatibus fruitur. At Con. 7.pr.9 (272.7 M.) we have as a result of Schultingh’s emendation: Cum dixisset Albucius in illa (sc. controversia: cf. 7.pr.8 (272.4 M.)) fratre qui . . . ABV give illo. This hardly sounds like good Latin. Perhaps in illo fratre qui . . . Cf. 10.5.13 (498.15 M.): in illa controversia . . . de illo qui . . . At Con. 7.7.19 (338.9 M.) Seneca has been explaining the device that Cestius called ‘echo’—starting and finishing one’s declamation with the same words. Hoc sententiae genus Cestius echo vocabat et dicenti discipulo statim exclamabat: ἱμερτὴν (or whatever) ἠχώ. Dicenti will hardly suffice. Novák suggested dicente ita discipulo,⁷² Wachsmuth desinente or finiente. There seems nothing

⁶⁹ Surely to be deleted. It appears (as ad in AB, at in V) after senex in the manuscripts, and may be merely a throwback to adulescens. ⁷⁰ Con. 2.6.4 (178.2 M.). ⁷¹ E 6.3 (257.18 M.). ⁷² Novák (1915), 170.

Problems in the Elder Seneca

35

against the dative, but it is easiest perhaps to write dicenti⁷³ (or sic finienti, though sic finivit precedes rather closely in 7.7.19 (338.8 M.)). In Con. 9.pr.3 (371.11 M.) Seneca is talking of the advantages of the declaimer compared with his colleague in the courts. Adice nunc quod ab illis nullius interventu excutitur. The editors write animus for ab, and the sense is clearly right; but we may also think of mens,⁷⁴ or less dramatically and perhaps more realistically memoria. In Con. 9.4 a son had flogged his father in the citadel because the tyrant told him to. Later his father has to defend him on a charge; and Musa makes him say: Tecum, fili inconsideratae pietatis, queror: validius patrem cecidisti quam iussit tyrannus.⁷⁵ It is clear from other epigrams that Musa’s colour is that the father acquiesced in being beaten by his son, because he knew that the son was plotting to get into the tyrant’s good books and eventually take the opportunity of killing him, as the theme states eventually happened. Cf. above: Aiebam: fili, fortiter feri; tyrannus spectat.⁷⁶ The father then cannot be complaining of the severity of the beating—he should be glad that the son, to give verisimilitude to the scene, had laid it on hard. The complaint should be that the son, being naturally affectionate despite what circumstances demanded in the way of lack of affection (inconsideratae pietatis), had not played his part with the brio demanded by the plot. Hence Haase’s languidius for validius (not printed by Müller). Easier, surely, after queror, validius. Lesbocles and Potamon reacted differently to the deaths of their sons. Lesbocles scholam solvit: nemo umquam amplius animo se gessit Potamon.⁷⁷ Various attempts have been made to fill the obvious lacuna; neatest, I think, would be: nemo umquam amplius⁷⁸ animo se gessit Potamon. Müller should not have trusted V’s correction of amplius to ampliore. At Suas. 2.1 (532.9 M.) Fuscus’ descriptio of Thermopylae is complicated by corruption. As Müller’s text reads we have: et huius quoque (i.e. even a small boat) remigium arcet inquietum omne quod circumfluit mare, fallentia cursus vada altioribus internata, aspera scopulorum et cetera quae navigantium vota decipiunt. This will pass; but in view of the fact that for vada the manuscripts give vadet (vadat V), I suggest that the balance of the description would be much improved by reading vada et. The vada are then described in only one participial phrase (that they are ‘born among’ deeper waters would be hardly more than a tautology), while the second is reserved, much more vividly, for the aspera scopulorum. Thus we have three clauses linked by et, the first of

⁷³ ⁷⁴ ⁷⁶ ⁷⁸

Considered by Novák, loc. cit. For this conjecture see Winterbottom (1974b), ad loc. Cf. Con. 1.7.16 (80.17 M.). ⁷⁵ Con. 9.4.2 (403.13 M.). Con. 9.4.2 (403.10 M). ⁷⁷ Suas. 2.15 (541.9 M.). Walter (1935 6), 516 compares Con. 7.pr.7 (271.11 M.): numquam amplius in foro dixit.

36

Papers on Quintilian and Ancient Declamation

three words, the second of four, the third of five. Rivalling the shallows in their danger are the jagged rocks that suddenly appear in the deeper waters. Just below this, the declaimer says: Pudet, inquam, Lacedaemonios et armatos quaerere quemadmodum tuti sint.⁷⁹ The inquam harks back to 2.1 (532.5 M.): Pudet Lacedaemonios sic adhortari. The shame, there, is the speaker’s; and it may be that we should make it his in the second case as well, by adding apud after inquam.

OTHER CONJECTURES A number of other changes are less readily classified. I take them in order. In Con. 1.2 a girl seeks a priesthood after a series of adventures has seen her in a pirate galley and a brothel. Vultis auctionis exitum audire?, asks Vinicius;⁸⁰ and he answers: vendit pirata, emit leno, excipitur nihil. No point here, surely, and Opitz proposed excipit lupanar, comparing 1.2.8 (34.13 M.).⁸¹ With the same sense, but easier palaeographically, I suggest excipit fornix; for the word cf. 1.2.21 (42.5 M.; with double-entendre), 9.2.28 (394.16 M.). In the same controversia a speaker who wishes to stress the incompatibility of priestess and prostitute remarks: Non sine causa sacerdoti lictor apparet: occurrent (occurrenti V) tibi meretricem summovisset (34.5 M.).⁸² The second sentence ought surely to be generalized, and it may be that Seneca wrote summovet, perhaps with a sed following at the start of the next epigram. For the rest, Seneca may have written merely occurrenti meretricem: at 1.2.1 (30.13 M.), despite the lacuna, it is clear that the priestess it is who occurrit the prostitute and gets her lictor to summovere the unclean object. Cf. also 9.2.2 (382.16 M.): Maiestatem laesam dixissem si exeunti tibi lictor a conspectu meretricem non summovisset. At § 13 (37.13 M.) we are still concerned with, and apparently addressing, the impure priestess. Puta enim virginem quidem esse te, sed contrectatam osculis omnium; etiamsi citra stuprum, cum viris tamen †voluntate†: es casta talis qualis videri potest cui lex nocere vult matrem quoque incestam. It may be that Gertz was right to delete te, and to proceed (after Faber) . . . tamen volutatam: an est casta talis . . . Certainly the generalizing puta jars with the specific te. But that is defended by 2.6.1 (175.16 M.); and it seems best to follow up the resounding clausula osculis omnium with a new statement asserting the previous hypothesis rather than a continuation of the same sentence. Read, then, etiamsi citra stuprum, cum viris tamen volutata , and proceed not es casta talis (which, quite apart from its awkward expression, anticipates, at ⁷⁹ Suas. 2.1 (532.12 M.). ⁸² Con. 1.2.7 (34.5 M.).

⁸⁰ Con. 1.2.3 (31.17 M.).

⁸¹ Opitz (1888), 285.

Problems in the Elder Seneca

37

half-cock, the point of the epigram) but merely es talis, for which casta will be gloss (or even dittography?). We come to the priestess thrown from the rock. At Con. 1.3.4 (45.14 M.) we have: . . . et, quam a saxo nusquam reverti fas est nisi ad saxum, quanto minus quam in templum resiluit. This (which Müller prints as a question) does not seem to trouble the editors. Bornecque’s translation (‘ . . . c’est à peine si le saut qu’elle a fait ne la rend pas au temple’) is neither close nor intelligible. Is not the clause an exclamation, and should not another comparative replace minus (improbius?), quam disappearing?⁸³ Or could minus understand fas? Then, with quod for quam, ‘how much less right that she should have bounced back into the temple.’ Exaudierunt dii immortales publicium voti praeces:⁸⁴ perhaps . . . publicae vocis preces (publica vox in a rather different sense in Cornelius Severus’ poem at Suas. 6.24 (573.10 M.)). In a controversia on the handless hero who could not kill the man he caught in bed with his wife, Hispanus says (according to the manuscripts): O dignum cui aut pudica contingat uxor, aut inpudica dum armatus est.⁸⁵ The epigram does not have to be taken closely with what follows (where the hero speaks), and the clear indication of the manuscripts is that the apostrophe is not in the first person. Cf. 1.4.3 (52.8 M.) Ante patriae quam patri negavit manus.⁸⁶ Nor do we need to emend, with Gertz, to the past tense. The epigram will have been set in the past, before the hero was injured. The only change needed is from est to sit. In the case of the double rapist whose victims made different choices, Latro makes the point that the accused is playing off one of his crimes against the other: stupro accusatur, stuprum defendit.⁸⁷ E gives this as stupro accusatur, stupro defenditur.⁸⁸ We arrive at a similar point more economically in the main text by reading stuprum accusatur, stuprum defendit: ‘a rape is the charge, a rape is his defence.’ At Con. 1.7.3 (72.13 M.) a son tells his father, reluctant to ransom him from the pirates: Sed rogare illos potes et audacter roga: in misericordes piratas incidi. The corrector of D made roga out of eroga. Should he not have made rogare? Many details of Albucius’ mixture of colours in Con. 1.7.17 are unclear. One colour is to deny that the man who killed his brother realized that he was a ⁸³ (Mi)nus quam a dittography of nusquam? ⁸⁴ Con. 1.3.10 (49.2 M.). ⁸⁵ Con. 1.3.1 (51.10 M.). ⁸⁶ Which does not need to be prefixed, as in E, by the explanatory in bella non venit et. Like modern editors, the excerptor is not happy with abrupt epigrams. But sententiae do not even need to have a main verb: see Con. 10.1.3 (458.18 M.), where the declaimer breaks off after mortuo patre meo to comment Timeo enim ne quis sibi iniuriam fieri putet, si dixero occiso (E leaves this alone). The previous, quite separate epigram finishes at times, with far greater snap than the editors have allowed it. ⁸⁷ Con. 1.5.1 (57.10 M.). ⁸⁸ E 1.4 (97.20 M.).

38

Papers on Quintilian and Ancient Declamation

tyrant—hence no credit for the tyrannicide. Exsecutus est omnia: hoc illi inter cetera obiecit, quod occupat fratrem suum ignorasset aut dissimulavit ait (BV: dissimulasvit A¹) tyrannicidio quoque eius commendationem fratre.⁸⁹ This can hardly be more than patched up. But Faber seems to have been right in supposing that what was seized was the citadel, and in adding arcem before fratrem. Occupat must clearly become an infinitive: not, however, occupare, as Schultingh, but occupasse: tyrants are in full possession. At the other end, a combination of old and new ideas (ut . . . faceret Faber; voluit . . . detrahere Kiessling) produces ut . . . detraheret; while the remaining words, wherever they intrude from, must just go: pretending the brother was not a tyrant is something that would hardly affect the credit to be gained from his killing. Latro uses a variant of the ‘over my dead body’ theme: Abdicato quoque non permittam exire, iniciam manus, tenebo, novissime ante limen exeuntis cadaver hoc sternam: ut ad hostem pervenias, patrem calca.⁹⁰ Perhaps exeunti—dative, as they say, of disadvantage, rather than genitive of possession. I am grateful to Mr M. W. Frederiksen for advising me on the (apparently real-life) case described in Con. 2.1.34–6. It is unclear how previous editors and translators have visualized the circumstances: they have been content with the juxtaposition of pro matrona (Gertz: hac re ne ABV) adulteri rea (reo ABV) at 2.1.34 (125.16 M.) with servum adulteri postulatum at 2.1.34 (126.4 M.). It is surely clear that only the slave is being accused; the wife is concerned only in so far as the court proceedings may prejudice her own case in any future trial (mulier in quam petebatur praeiudicium: 2.1.34 (126.5 M.)). Hac re ne then must cover a word meaning ‘slave’ (verna occurred to Mr Frederiksen and myself independently); or, better, it should be deleted, the specification of the defendant coming only later at servum. Reo is the correct gender, but we must then proceed in quem Syriacus Vallius . . . calumniam iuraverat. Non tulit hanc contumeliam Latro, et pro Pythodoro Messalae orationem disertissimam recitavit, que (AB: quam V) compositam quem (AB: quam V) suasoriam Theodoto declamavit per triduum.⁹¹ The connection (atque compositam aeque . . . : Gertz) is most easily supplied by writing compositamque and deleting the earlier que. For this type of error see above, n. 52. Agrippa liked to keep his deplorable nomen Vipsanius quiet. One joker cried: Concurrite! Agrippa malum habebis respondi diis eam arce uterque.⁹² We shall not want to disturb the last twelve letters—Marce uterque will mean Marcus Agrippa and his alter ego, Marcus Vipsanius. Of the rest Ribbeck made responde, si dis placet; simpler, perhaps, responde sis.⁹³ Habebis will be better than habebit (Gronovius). The situation is explained to the crowd, then Agrippa is called upon to reply. The scene is the court-room—fuit qui diceret (160.6 M.) will understand accusator from 160.5 M.: Agrippa (witness, ⁸⁹ Con. 1.7.17 (81.10 M.). ⁹² Con. 2.4.13 (160.7 M.).

⁹⁰ Con. 1.8.15 (90.3 M.). ⁹¹ Con. 2.4.8 (156.6 M.). ⁹³ Cf. Con. 8.5.3 (415.10 M.): Mitte sis.

Problems in the Elder Seneca

39

I suppose, rather than defendant) is in for a worsting, and the accuser calls for popular attention. A woman has failed to bear a child, after being tortured because her husband is suspected of plotting to kill the tyrant. The declaimer says to her, ironically: Ita fit, mulier, non vis parere? Delicata es, cruciatus puerperi times?⁹⁴ The editors follow Haase in writing Ita tu, mulier. Perhaps, if the text cannot stand, Itast, mulier. The Ciceronian usage is itane est (e.g. Quinct. 38), but see Sen. Ep. 17.8: Ita est? . . . voles? In the midst of a locus de saeculo Fabianus says: in deterius luxu fluente muliebris ambitio certamine mutuo usque in publica damna privatis insanit.⁹⁵ The dative privatis seems forced, and though the excerpts support⁹⁶ I propose privatim. Later in the same declamation⁹⁷ we hear that Blandus in ultima parte controversiae, qua de re publica disputatur, quaestionem fecit an quinquennium numerari debeat excepta tyrannide. It is not clear why discussion of the reckoning of the five years should be described as de re publica. The state came in earlier, in connection with motive;⁹⁸ this discussion concerns time. Read, then, de tempore; rep. is a constant abbreviation for re publica. A father takes up debauchery on the example of his son, and is accused of madness. He tells his son: Puta te patrem: dic quid me velis facere. Si †tam† bona fide frugi es, et hoc imitor.⁹⁹ I am not sure how closely the epigrams are linked, but we should surely read tu for tam (I cannot understand Haase’s tum): ‘If you are in fact of good character, well, I am only following your lead in this too’ (i.e. as well as in high living). Pauper says to Dives: Vos possidetis agros, urbium fines, urbesque domibus impletis.¹⁰⁰ There is sense of a sort here: ‘Vous possédez, comme champs, la banlieue des villes’ (Bornecque), or rather ‘fields that are properly the territory of whole cities’. But the word urbium spoils the snap of the contrast of agros and urbes. Read perhaps gentium—the rich man’s estates are province-wide. This would be the same sort of exaggeration as Hor. Carm. 3.16.25–7: splendidior . . . / quam si quidquid arat impiger Apulus / occultare meis dicerer horreis.¹⁰¹ A son accused of parricide is set adrift in a rudderless boat. He appeals to the gods to protect him: Si nihil umquam impie cogitavi, si patrem meum etiam damnatus diligo, di immortales . . . adeste. The declaimer then switches to indirect speech: Si aliter sentiret, infelicia sibi imprecatus est maria.¹⁰² ‘S’il mentait en parlant ainsi’ (Bornecque)—or, more exactly, ‘If his feelings were

⁹⁴ ⁹⁷ ⁹⁹ ¹⁰¹

Con. 2.5.4 (163.7 M.). ⁹⁵ Con. 2.5.7 (165.6 M.). ⁹⁶ E 2.5 (199.17 M.). Con. 2.5.13 (169.10 M.). ⁹⁸ Con. 2.5.12 (168.18 M.). Con. 2.6.1 (175.16 M.). ¹⁰⁰ E 5.5 (249.4 M.). See Nisbet Hubbard (1970) on Hor. Carm. 1.1.9 10. ¹⁰² Con. 7.1.5 (276.2, 4 M.).

40

Papers on Quintilian and Ancient Declamation

different from his words’. Maybe so; but perhaps read sentirent (sc. di): ‘If the gods cast their vote against him’.¹⁰³ In the same controversia we hear how the son is rescued by pirates, and how his father set out abroad, only to fall into the hands of those same pirates. Nec hoc tantum divinitus gestum est quod pervenit (the son, that is) tutus in portum: excipitur classe praedonum. Next, the manuscripts give habeat pater mentem navigandi: capietur iudex, ut illum paeniteat sententiae suae.¹⁰⁴ This is a new epigram, and the second part of it may stand. As Thomas saw,¹⁰⁵ iudex means the father, who had given his sententia by condemning his son in a private tribunal and regretted it when his son rescued him from captivity. What of habeat pater mentem navigandi? This should surely be narrative, a statement of the father’s intention to sail the seas, on which the declaimer comments that repentance will be the outcome of his voyage. Habet, then. Later in the same controversia, Pompeius Silo suggests that the stepmother who had been an evil influence over the father should try her luck in an open boat. Let her pray. Si neminem innocentem accusavit, si privignum inmerentem non oppressit, in eos piratas incidet (Schultingh: incident AB: incidat V) qui nesciant captos dimittere.¹⁰⁶ Bornecque saw that nesciant cannot do. If the stepmother, like the stepson, is innocent she will fall in with unusually kind pirates—her son, the pirate chief, in fact, will let her go, as he let his father go. Bornecque wrote sic sciant; rather, just sciant. Otho pater hoc colore usus est pro patre: dixit enim molestum fuisse imperatori quod illum suffixum legati intuebantur.¹⁰⁷ Hoc must refer forward (the context makes that clear), and enim will not do. There are variations in Seneca’s phraseology in such formulas, e.g. Con. 7.6.18 (327.16 M.): Silo Pompeius hoc colore usus dixit: Exhaustum . . . patrimonium; 7.6.20 (328.6 M.): Accaus Postumius hoc colore usus est: Nihil est, inquit (and see above, n. 58). But in our case we must either emend hoc to a specific adjective (cf. 7.2.13 (296.15 M.) Hispo Romanius vehementi colore usus est et duro: patronum enim dedit . . . ) or delete enim (9.1.15 (381.7 M.): Hispo Romanius hunc colorem secutus est: dixit adulescentem . . . ), or alter enim (interim Müller, to no great point), perhaps to hoc.¹⁰⁸ Necessity is an excellent pretext: haec excusat Saguntinos (quamvis non ceciderint patres, sed occiderint); haec excusat Romanos, quos ad servile dilectum Cannensis ruina compulit; quae quidquid coegit defendit.¹⁰⁹ Quae seems very feeble; surely the declaimer continued his anaphora with haec?

¹⁰³ Cf. Con. 7.1.25 (286.14 M.): Triarius et ipse quasi sententiam (no need for an adjective) de fratre ferri voluisset egit, then another prayer. ¹⁰⁴ Con. 7.1.11 (279.3, 4 M.). ¹⁰⁵ Thomas (1899), 217. ¹⁰⁶ Con. 7.1.15 (281.6 M.). ¹⁰⁷ Con. 7.7.20 (338.17 M.). ¹⁰⁸ For the corruption see Winterbottom (1970b), 111. ¹⁰⁹ Con. 9.4.5 (404.21 M.).

Problems in the Elder Seneca

41

In the same controversia, another declaimer comments on the fact that a father is defending a son who has flogged him. Quoniam usque eo saeculum mutatum est ut parricidae pater adsit, non istius advocationi adsum (adsumus B).¹¹⁰ The son may help his father, but why should I? That seems to be the sense. Nos for non (so Bursian) gives the right emphasis, but deprives us of sense. Perhaps, with nos, read adsimus, and make the whole an indignant question. Or better, again as a question: nos istius advocationi adsumet—‘shall he bring us in as partner in his son’s defence?’ A prepositional phrase would be normal, but for the dative in a related sense cf. Plin. Nat. 35.197: Umbrica non nisi poliendis vestibus adsumitur. Latro¹¹¹ has been showing that there are occasions when a son may strike his father with impunity. Nondum de propria sed de communi causa loquor. Si efficio si quis qui cecidit patrem possit absolvi, pro hoc animosius agam . . . Müller was unwisely tempted by the Renaissance conjecture officiose (which covers the last of the preceding examples but hardly the cases of the lunatic or the infant father-beater). We may rest content with si efficio ut (ut driven out by a dittography of si); I suppose we could proceed si quis,¹¹² but it looks inelegant, and qui is as much as we need. ‘If I succeed in showing that a father-beater might be acquitted, I can go on with more confidence (to the particular case).’ In Con. 9.6.10, we hear of Votienus Montanus’ mocking of an epigram of Cestius, who, in a case involving the complicity of a mother and daughter in the poisoning of a stepson, made the mother say: Da fratri venenum, and the child reply: Mater, quid est venenum? Triarius¹¹³ went one worse: induxerat novercam dicentem: Da fratri venenum; fecit illam respondentem: Mater, et mihi da. Montanus’ comment on this is: Quid enim est tam absurdum quam matrem sic locutam cum puella: Da fratri venenum? It is not clear why this objection is brought against Triarius alone and not Cestius as well. But it is conceivable that we must in fact understand it of both. Whatever the case, there is something strange in Triarius’ epigram. A child says ‘Mummy, give me some too’ when it knows that its brother is being given something it covets, not when it is instructed to give its brother something.¹¹⁴ Schultingh was therefore on the right lines in suggesting Dabo fratri venenum; easier, however, Do. In the comment that follows we could still retain Da and suppose that only Cestius’ epigram is under explicit criticism. But it is more natural to read Do here also; and that surely fits in with sic locutam cum puella better than an imperative.

¹¹⁰ Con. 9.4.6 (405.14 M.). ¹¹¹ Con. 9.4.9 (407.13 M.). ¹¹² Quisquis will do in the different context of Con. 9.4.10 (407.15 M.), not, of course, here. Si quis may intrude from 407.11 M. ¹¹³ Con. 9.6.11 (427.1 M.). ¹¹⁴ In that case it says: ‘Can I take some too?’

42

Papers on Quintilian and Ancient Declamation

Et illud . . . : Proderunt plures mendici quam membra.¹¹⁵ This is clearly a piece of scene painting to arouse miseratio for the crowd of cripples: so too at Con. 10.4.22 (491.12 M.), Producitur miserorum longus ordo, maior pars se sine se trahit. That epigram reminds us that a historic present is the vivid tense to use; and we should prefer prodeunt to Bursian’s prodierunt. ‘I shall not give my vote’, says Seneca to his sons.¹¹⁶ Vestri arbitrii erit utrum explicationes eius (Fuscus, that is) luxuriosas putetis an ut petas (A: ut poetas B: ut poeta V). Various adjectives have been proposed to replace petas (Müller prints Gertz’s vegetas). None comes so close as laetas: which is strongly supported by Quint. 12.10.80: Sic erunt magna non nimia, sublimia non abrupta . . . laeta non luxuriosa . . . Both this and Sen. Con. 2.pr.1 (103.6 M.): magis lasciva quam laeta show that laetus is the complimentary adjective of the pair. Ut must go, however we explain its intrusion; perhaps the archetype had an letas with vel petas above it. In the sixth suasoria¹¹⁷ Latro says that anything Cicero does will offend Antony: Aliquid erit quod Antonium offendat, aut factum tuum aut dictum aut silentium aut vultus; and he added a sententia: aut erit (haud enim V) placiturus es. V’s patching hardly constitutes an epigram. And what Latro said is quite uncertain. Perhaps sic (ita?) placiturus es, that (i.e. by offending Antony) is the way you will please him (i.e. by giving him an excuse to get rid of you). Aut will be a repetition of the aut before vultus, an inadvertent continuation of a long series. Erit may be influenced by erit above. Or a repetition of adiecit may have been corrupted into aut erit. In the same suasoria we have a fragment of the historian Cremutius Cordus; Cicero’s head and hands are exposed on the rostra: Itaque, quo saepius ille ingenti circumfusus turba processerat, quae paulo ante coluerat piis contionibus, quibus multorum capita servaverat, tum per artus sublatus . . . ¹¹⁸ Other things are doubtful here (I give the passage as Müller prints it); but quae . . . coluerat seems unintelligible. Quae . . . praebuerat piis orationibus is Müller’s suggestion. We can only patch, and we can do so more economically with quam. Cicero had by his patriotic speeches that saved so many lives striven to win the affections of the people.¹¹⁹

THE GREEK There are many Greek epigrams in Seneca’s collection, often grouped together at the end of a controversia. The words were copied into our manuscripts by scribes who could not understand them; and gaps often reflect their incomprehension or ¹¹⁵ Con. 10.4.23 (491.18 M.). ¹¹⁶ Suas. 2.10 (537.17 M.). ¹¹⁷ Suas. 6.8 (562.14 M.). ¹¹⁸ Contr. 6.19 (568.18 M.). ¹¹⁹ For parallels see OLD² s.v. colo 7b.

Problems in the Elder Seneca

43

desperation. But non-comprehension has its advantages, and it should be possible to make advances in the restoration of these epigrams. At Contr. 1.3.12 (50.10 M.) Diocles of Carystos says to the erring priestess who has survived her first fall from the Tarpeian rock: hla aτuaa (A: hal aaπιaa BV) καὶ δεύτερον καὶ τρίτον καὶ μέχρι ἂν¹²⁰ πέσῃς ἐϕ’ ὃ βέβλησαι. Bursian’s Καὶ ἅπαξ can hardly be right: the problem had arisen after the first fall. Gertz’s¹²¹ Καταβαλῶ gives a better sense. I suggest Καταπήδα. Of course, the girl would be pushed (cf. the related epigram in 1.3.1 (44.4 M.)) rather than jump, but the verb gives the contemptuous tone of virgo desultrix;¹²² cf. also resiluit at 1.3.4 (45.16 M.). At Con. 1.8.16 (91.8 M.) we have a perplexing epigram or series of epigrams by Dorion. A son has disobeyed his father and gone off to battle; the father defends his own attitude: ‘Why these bloodthirsty aspirations? Φοβοῦμαι μή που παράταξις, μή που ΛЄΤΜΟC, μή που πάθη σ’ ἕλῃ.’ So Gertz, who gave λιμός; but Haase’s λοιμός surely comes nearer to the traces: and plagues have been associated with war since Iliad 1. But the choice is very open: cf. [Dion. Hal.] Ars rhet. p. 372.5 Us.–Rad. on subjects for ἔκϕρασις: τὸ χειμῶνα γράϕειν καὶ λοιμοὺς καὶ λιμοὺς καὶ παρατάξεις καὶ ἀριστείας (and again below at Contr. 9.pr.5 (372.18 M.)). There follows Φοβοῦμαι περὶ τῆς ΟΙΚΟΥ ΜeΝe—why not οἰκουμένης? The father’s fears extend beyond his son to the whole world, which may suffer from the battles and plagues listed. Then: Τί, τέκνον, eΡΥΛ (ΘΡΨa V). Perhaps θρυλῶ: the father’s attoniti adfectus are being expressed, we have been told, and he pulls himself together to ask what his babble means. In Con. 2.3.23 (150.12 M.) Diocles provides a ‘speech in character’ for a father who will not forgive his son for a rape. єΒω ΟΝa (AV: aΤΙΟ B) aπΟΟΛє. Müller writes Ἐγὼ θανατιῶ.—Ἀπόθανε. The son tries to arouse pity, but the father is unrelenting. But θανατιῶ would mean ‘I want to die’, not ‘I am fated to die’ (‘je mourrai’, Bornecque), and ἐγώ is superfluous. Perhaps the son’s words are Ἡβῶ θανάσιμος—‘I am young, yet near to death.’ Cf. Soph. Phil. 819 Ὦ γαῖα, δέξαι θανάσιμόν μ’ ὅπως ἔχω. In Glycon’s remarks at Con. 7.1.26 (288.1 M.), I think that Gertz¹²³ was right to read Ἰδία κριτοῦ ἑνὸς οὐκ ἀρκεῖ καταδίκη, and to end the first sentence there. Equally he was right to take Εὑρίσκει τὸ μηδὲν ἀδικεῖν τύχην as the last sentence (Qui nihil commisit fortunam invenit). But the remaining words eππ aΤΚ e ΝaaΙ e ΤeaeΙ have not yet been elucidated; it may be that we should shorten this (lacunose?) sentence still further by adding ἀεί to the third, generalizing sentence. [I should now be wary of arguing from ‘coincidental’ error in A and B. I suspect that B is descended from A.]

¹²⁰ So Novák (1913), 134. The omicron which appears in the manuscripts after ἂν is presum ably an anticipation of that after ἐφ’. ¹²¹ (1888), 294. ¹²² Con. 1.3.11 (49.16 M.). ¹²³ (1888), 294.

3 Quintilian and Rhetoric* Marcus Fabius Quintilianus, public Professor of Rhetoric, tutor of royal children, rich, academic, dry: this hardly sounds the portrait of a revolutionary. And indeed the revolution that he preached in his twenty years of teaching in Rome and in the Institutio that was the product of his retirement was rather in the nature of a reaction. The message was that oratory had become corrupt, that it could only be cured by a return to the standards and methods of a century before. We might gauge the strangeness of such a plea by imagining, at the present time, a professor of English Literature who constantly proclaimed that the whole course of poetry since Keats was a mistake and that redemption could only come if poets, with a proper seriousness, returned to addressing nightingales and Grecian urns. But this analogy would miss the richness of implication in Quintilian’s standpoint. Oratory—as another chapter in this book explains¹—was the king-pin of the Roman educational system. Further, oratory was the literary art par excellence, and from it poetry and history tended to take their lead. In demanding oratorical reaction, Quintilian was calling for a re-thinking of educational dogma and for a reversal of tendencies that had made the Silver Age of Latin literature. The return to the past, for Quintilian, was a return to Cicero. An anecdote in the Letters of Quintilian’s most distinguished pupil, the Younger Pliny,² illuminates for us what must have been a common attitude to such a return. The orator Regulus on one occasion sought out Pliny because he felt uneasy that he might have given offence by contrasting with Pliny one Satrius Rufus, ‘who does not pretend to rival Cicero but is content with the eloquence of his own century’. The sarcasm is clear. We know from the Dialogus of Tacitus that to the orators of the Silver Age Cicero seemed old-fashioned, crude, and longwinded; times had moved on—and brought a new and arguably better style of [From T. A. Dorey (ed.), Empire and Aftermath. Silver Latin II (London, 1975), 79 97] * I have left this chapter largely as it was written in 1967. In the meantime Kennedy (1969) has provided an excellent general survey. ¹ See Clarke (1975). ² For this and other matters arising in this chapter see Winterbottom (1964b) [= A.1 above].

Quintilian and Rhetoric

45

oratory with them. At the same time, Pliny’s reply to Regulus demonstrates the forces to which Quintilian could appeal against this assertion of progress and the inevitability of history. ‘Yes, indeed, I do set up as a rival to Cicero,’ he said, ‘and I am not content with the eloquence of our century. For I regard it as the height of stupidity not to take the best as one’s model for imitation.’ The concept of imitation, of course, did not rule out the possibility of progress. ‘Nothing grows by imitation alone’, Quintilian himself remarks.³ All the same, ‘no one can doubt that imitation plays a large part in art.’⁴ To have the right models was all-important. The feeling of literary tradition that lies behind such a view was what enabled Quintilian, without patent absurdity, to try to re-direct the course of literary history. The false gods must be replaced by the true; all, then, would be well. ‘Let the student who comes to take pleasure in Cicero know that he has made real progress.’⁵ This was Quintilian’s challenge. Cicero’s style, of course, was a thing of the utmost variety: it could be plain or pompous, learned, witty, pathetic. But it had a recognizable unity in humanity and seriousness of intent that could justly be contrasted with the new style. ‘When he had matured and learned by experience what the best style of speaking was’, a modern orator says in the Dialogus,⁶ ‘Cicero attempted passages of greater brilliance and even struck off a few epigrams.’ If anything, however, it was the young Cicero who had most vividly presaged the Silver Age—ranting, for instance, on the punishment of a parricide in the Pro Roscio Amerino (72) in terms that he himself later admitted to have been altogether too extravagant. And it is significant that the passage is closely imitated in a declamation (Decl. min. 299.3) from the collection attributed (perhaps wrongly) to Quintilian. The mature Cicero controlled himself better; and the prose of the first century  was a reaction against his dignified wordiness, not a natural development of it. Aper, the modern orator of the Dialogus, was pleading a case—and pleading it, in this respect, misleadingly. But Maternus, another speaker, representative, surely, of the view of Tacitus himself, is irrefutable. His thesis is that oratory has changed since the days of Cicero; and changed in a way that leaves no point in mere exhortation to return to Ciceronian techniques. ‘What need of long speeches in the senate? Our great men swiftly reach agreement. What need of constant harangues to the people? The deliberations of state are not left to the ignorant many: they are the duty of one man—the wisest. What need of prosecutions? Crime is rare and trivial. What need of unpopular defences? The clemency of the judge meets the defendants half-way.’⁷ To put it less rhetorically, the rule of the emperors—even when they were beneficent—left no real scope for either forensic or deliberative oratory, at least of the kind in which Cicero had made his mark. Words no longer affected important actions ³ Quint. 10.2.8. ⁶ Tac. Dial. 22.2.

⁴ Quint. 10.2.1. ⁷ Tac. Dial. 41.4.

⁵ Quint. 10.1.112.

46

Papers on Quintilian and Ancient Declamation

or swayed judgements that counted. All that remained was the panegyric; and it is no coincidence that this is the only sort of speech preserved for us other than fragmentarily after the death of Cicero (apart from Apuleius’ strange Apologia). Nor was there room for the influential orator himself. He was a product of the old turbulent days. ‘We are speaking of no inert and passive thing that rejoices in goodness and modesty. Great oratory is the nursling of licence (which fools call liberty); companion of revolution, inciter of the unruly mob; uncontrolled, unprincipled, insolent, rash, arrogant. It does not arise in well-regulated states’ (Dial. 40.2). This analysis was faultless. An independent political orator was, under the conditions of the principate, unthinkable; at most, a speaker could be a tool of the régime. Quintilian did not see, or chose to disregard, these restrictions. His perfect orator, like Cicero’s, was decked out with all talents and all arts; he would defend the innocent and crush crime; but, far more than that, he would shine brightest in greater matters—directing the counsels of the senate and leading the errant people on to the right path.⁸ To this extent, Quintilian was not so much preaching a revolution as crying for the moon. He did see, however, that exhortation was not enough, and that conditions must be changed before states of mind could alter; but the only conditions that he felt at liberty to criticize were educational, and even here he was oddly timid. In the Dialogus the character who most nearly represents the views of Quintilian, Messalla, draws a damning contrast between the education of Cicero’s day, when orators were trained by example and personal instruction to come up to the merits of the great speakers of the time, and the present system by which youths are automatically sent to a rhetor. Here they engage in exercises more concerned with tyrannicide, rape, and plague than with sober legal realities. Messalla’s diatribe is broken off for us by a defect in the manuscripts mid-way through a sentence that starts: ‘But when he comes before real judges . . . ’⁹ And it is certain from parallel sources that what Messalla proceeded to say was that being exposed to reality made young men trained on declamation look merely foolish. It could well be that the lacuna also hides a plea to do away with declamation schools altogether, or at least radically revise their methods. Even this, of course, would have been to treat a symptom of the decline of oratory as though it was the cause; and Maternus’ historical exposé would be intended to point this moral. But Quintilian himself was unadventurous even in his recommendations for reforming education. For basically he regarded declamation as a good thing; it was the teachers who had ruined it. It is to be supposed that Quintilian’s lost book on the causes of the corruption of oratory (published a little before the Institutio) went into greater

⁸ Quint. 12.1.26.

⁹ Tac. Dial. 35.5.

Quintilian and Rhetoric

47

detail on this point. It probably drew a parallel between the decline of oratory in Greece after Demetrius of Phaleron, at about whose period declamation was introduced in Athens,¹⁰ and the decline of Roman oratory after the Augustan period when declamation began to flourish in Rome. But, if so, it does not seem to have drawn the conclusion that it was declamation itself that was at fault. At least in the Institutio Quintilian, while asserting that the licence and ignorance of declaimers was among the primary causes of the corruption of eloquence, put the blame just on the methods of the rhetores. The cure was to make the subjects treated by declaimers as near as possible to reality—to the subjects they would have later to treat in court. ‘We shall look in vain for wizards and plagues and oracles and melodramatic stepmothers among the legal phraseology of securities and interdicts.’¹¹ Declamation was instituted as a mirror of reality, and it should not be deflected from reality. But the trouble came when Quintilian considered the practicability of adjusting the exercise in this way. Declamation, when all was said and done, had something of display about it, and could not wholly shrug off brilliance and ornament. Moreover, it was good for young declaimers to ‘fatten’ themselves from time to time on less dry topics. And how difficult it would have been to interfere with the hallowed rules of the game is shown by Quintilian’s wistful remark: ‘Would that it could be added to normal practice that we should use names, set themes longer and more complex in character, shudder less at ordinary everyday language, and put in jokes.’¹² Why, it may be asked, could not so distinguished a teacher carry out such reforms at least in his own school? The answer seems to have been that the parents, the rich conservatives of Rome, would not have liked it. They may have demanded public declamation on set open-days, when they could come and rejoice in their sons’ success: ‘They think their children are working only if they declaim as often as possible’¹³—and that, no doubt, in accordance with the rules to which they had been accustomed in their day. And they, after all, were paying. Nevertheless, practicably or not, Quintilian made his challenge. It is easy, looking at the massive Institutio and observing the scholarly assembling of sources, to regard it as fairly and squarely in a rhetorical tradition. But if it is in a tradition, it is a Greek tradition. As far as Rome went, and forgetting encyclopaedists like Cornelius Celsus, a long book of rhetorical precept was something that went out with Cicero; and even Cicero, once his youthful De inventione was out of the way, took care to disguise the precept behind a more urbane front than Quintilian assumes. If Quintilian was challenging opinion by harking back to Cicero, he was challenging it by writing a rhetorical handbook at all. Oratory had changed; and audiences had changed accordingly. They liked the short brilliant speeches of today. It was only in the old

¹⁰ Quint. 2.4.41.

¹¹ Quint. 2.10.5.

¹² Quint. 2.10.9.

¹³ Quint. 2.7.1.

48

Papers on Quintilian and Ancient Declamation

days that a people still unused to rhetoric would put up with elaborate daylong speeches. ‘Then prestige attached to a long preparatory introduction, a narration back to the year dot, a display of elaborate headings, a thousand stages of proof, and all the other things dictated by the dry-as-dust handbooks of Hermagoras and Apollodorus.’¹⁴ And Aper, the modern orator of the Dialogus, goes on to say that nowadays all that is out of date; everyone has a smattering of education, and looks for new and exotic tricks in oratory. Now the judges are in a hurry; and they have to be wooed by epigrams and brilliant descriptions. Yet here was Quintilian writing twelve books that constantly allude to Hermagoras and Apollodorus, and inculcate just those technicalities that modern audiences scorned. Aper’s invective goes a long way towards summarizing the content of the Institutio. The first two books deal with grammar, various subjects (such as geometry and music) subordinate to rhetoric, and questions about the purpose and definition of rhetoric. Book 3 discusses the so-called ‘status-lore’—the method, that is, of determining the basis or key issue of a case—and also, rather cursorily, display and deliberative oratory. But Quintilian is most concerned with forensic oratory—both because his tradition always had had this emphasis and because, while epideictic oratory gave little room for extended precept, deliberative oratory was virtually extinct. Accordingly, the rest of the Institutio has the law-court in mind—at least when it is not directed to the schoolroom. Books 4–6 deal with the parts of a speech—proem, narration, propositio, argument, peroration. Book 7 concerns the methods of treating cases of different status. Books 8 and 9 are about style, particularly the decoration given by tropes and figures and the attainment of rhythmical effects. In Book 10 Quintilian recommends reading that will nourish the now fully fledged orator, and discusses imitation and extemporary speaking. Book 11 has to do with propriety (that is, with what is fitting in various circumstances), memory-training and gesture. Book 12 concerns the moral character of the orator and the various genres of oratory. This is the barest of summaries. Quintilian goes into great detail, generously illustrates his points from oratory and poetry, and is not afraid to digress. But even from the summary it may be seen where Quintilian’s book differs from a normal rhetorical handbook. At the start of Book 3 we are told that most writers on rhetoric restricted themselves to discussing how materials for speaking should be discovered and dealt with. Quintilian innovated in looking to the child rather than the technique, and in seeing that the whole development was relevant. That is why we are given detailed instruction for the training of boys from birth to their arrival in the school of the rhetor; and why the youth emerging from the school is shown the way to a literary culture

¹⁴ Tac. Dial. 19.3.

Quintilian and Rhetoric

49

that transcends technicalities. Further, Quintilian was concerned, for reasons that will be seen later, with the morals of his budding orators. And the preface to Book 12, which pictures the author setting sail on a sea uncharted even by Cicero himself, stresses the importance and originality of this concern. Nevertheless, it remains true that the bulk of the Institutio is in the tradition of the ‘dry-as-dust handbooks of Hermagoras and Apollodorus’. Let us see how it modifies that tradition. Rhetoric was a dogma that had been elaborated for centuries—ever since the Sophists of the fifth century . The death of political oratory after the Macedonian conquest if anything stimulated the elaboration; but it was not until the second century  that Greek rhetoricians introduced the status theory that is the hallmark of the mature system, the triumph of academicism. By Quintilian’s day the need was not for innovation but for retrenchment and simplification; and this need went happily with Quintilian’s own tidy mind. He more than once protests against the tendency of rhetoricians to strive for originality—their ‘keenness to pass on something different’.¹⁵ All the same, he took the minutest points seriously. In the same chapter, on status, he apologizes solemnly, and with allusions to similar apologies by Hippocrates and Cicero, for changing his mind on the correct way to sub-divide the subject (64). But, significantly, the change of mind was in the direction of simplicity. Typical of his sane attitude is a remark in Book 5:¹⁶ These are the traditional kinds of proof; it is not enough to pass them on to my reader in mere generalities for from any one of them an infinite variety of arguments may arise; at the same time, it is humanly impossible to go through all the different sub headings one by one. Those who have tried to do this have come up against two difficulties at the same time: they said too much, yet they still did not say all there was to be said. And as a result many pupils, falling inextricably into these snares, lost all their natural flair in their addiction to rigid rules; keeping their eye always on their master, they ceased to follow the lead of nature.

This strikes a note very common in the Institutio, and it is what Quintilian was answering to those who scorned rules. A traditional triad made the orator the product of three components—natural talent, rhetorical precept, and actual practice. No one disputed the value, limited but indispensable, of the last. But the century was much perplexed by the claims of the other two. In Tacitus’ description, the orator Aper, though sufficiently well-educated, gave the impression of despising letters ‘as though he would win greater fame for his industry if his natural gifts were not thought to depend on the support of acquired skills’.¹⁷ We hear something very like this of Cassius Severus, the first of the ‘new’ orators: he relied more on talent than on training, and despised ¹⁵ Quint. 3.6.22.

¹⁶ Quint. 5.10.100 1.

¹⁷ Tac. Dial. 2.2.

50

Papers on Quintilian and Ancient Declamation

even the declamation schools.¹⁸ And, later on, we gain much the same impression of the orator Regulus, whom we saw earlier sneering at Pliny’s Ciceronian pretensions. Quintilian names no names, but he is clearly much concerned to counter these subversive views. ‘I see’, he says, ‘that some people will block my way on the very threshold, contending that eloquence needs no precepts of this kind. They think the ordinary education of the schools sufficient, plus their own talents; and they merely laugh at my painstaking care for detail.’¹⁹ There follows a vivid description of the excesses to which this sort of attitude could lead. Such people ‘boast that they speak on impulse and use strength; they think that, when the theme is fictitious, there is need not for proof or order but merely for splendid epigrams—that is what fills the hall with listeners.’ This is the ‘naturalist’ in the declamation school. But he is let loose in the law-court also; and there his violent abuse and lack of principle has greater dangers. ‘They shrink from having to prove their points. They avoid the chill inevitably caused in decadent courts by the details of proof, and look for nothing but what will soothe the ear with pleasures, even corrupt ones. They are concerned only with epigrams, which certainly show up better when their context is so dim and despicable.’²⁰ On the platform, they shout, gesticulate, throw themselves about; and they dub those who have more respect for letters ‘witless, bare, cold and weak’.²¹ Quintilian’s answer to this insult was that an educated orator would be imbued with an instinctive moderation that would avoid such faults. He might be less showy than the untrained, but he would have true force beneath the polished surface. This was what rhetorical training had to offer; and this might be how Quintilian would have answered the criticism that he was too timid in his attack on the educational system of the day: declamation was dangerous if left to the untrained, but in the hands of a responsible teacher, and allied to a thorough training in rhetorical precept and wider literary culture, it could be a worthy method. An orator trained thus would not want to titillate the audience with a stream of epigrams; he would know that other things were more important. At the same time, rules were not enough. They should not become a superstition. In a sensible chapter²² Quintilian firmly puts himself in the opposite camp to those who allowed no exceptions to rules. Rules were originally formulated to help an orator win his case; and if his case so demands he will logically ignore the rules. ‘I do not want young men to think themselves sufficiently educated once they have learned off one of the current rhetorical summaries—safe, as it were, behind the dogmas of the experts. Oratory consists in many things: much labour, constant application, varied exercises, ¹⁸ Sen. Con. 3.pr.4 and 12. ¹⁹ Quint. 2.11.1. ²¹ Quint. 2.12.11. ²² Quint. 2.13.

²⁰ Quint. 2.12.6 7.

Quintilian and Rhetoric

51

frequent practice, deep forethought, the most alert planning.’²³ Thus Quintilian’s rules may seem complex and rigid; but they are always to be read with this flexibility of principle in mind. And at the start of Book 8 he protests that his rules are not so frightening as they seem: it is just as bad to be deterred from learning by thinking them infinitely complex as to be self-satisfied enough to think them infinitely applicable. All the same, to modern ears, the Institutio does often go beyond acceptable limits of what can be reasonably taught with any hope of success. And it is only right to give, by extended quotation, instances of this failing. Quintilian admits in 6.3.11 that wit and humour are a matter rather of natural gifts and the opportunity of the moment than of precept. But this does not prevent him spending eighteen pages in a minute analysis of types of joke, of which the following gloomy passage is a fair sample:²⁴ There is an even more ingenious application of similarity when we borrow for one thing what is at home in another; you might call it ‘make believe’. For instance, at a triumph of Caesar ebony models representing cities were carried by in the procession; then, a few days later, the models in Fabius Maximus’ triumph were made of wood, and Chrysippus remarked that they were the boxes for Caesar’s cities. Again, Pedo said of a gladiator who was chasing another but not striking him: ‘He wants to catch him alive.’ A resemblance may be joined to an ambiguity. To someone going after a ball with little zest, Gabba said: ‘You compete as energetically as Caesar’s candidate in the elections’ (for the word petere is ambiguous, while the resemblance lies in the indifferent attitude). That is enough on this point.

As indeed it is. The same sort of doubts arise particularly in the chapters on style. It is one thing to know how to classify good qualities in the style of other orators; it is quite another to employ them oneself. For example, Quintilian writes of hyperbole:²⁵ Hyperbole is a fitting exaggeration of the truth; it has equal power in two different directions in increasing and diminishing. It occurs in a number of ways. We may say more than the actual fact (e.g. ‘In vomiting he filled his lap and the whole platform with scraps of food’ and ‘Twin rocks threaten heaven’), or we may exaggerate things by using a resemblance (‘You might believe there floated there the Cyclades uprooted’) or a comparison (‘Swifter than the wings of a thunder bolt’) or certain signs (‘She could fly through the surface of unreaped corn without harming the tender ears by her passing’) or a metaphor (like the word ‘fly’ above). Sometimes a hyperbole can be increased by adding another on top, as Cicero says against Antony: ‘What Charybdis was so voracious? Do I say Charybdis? If Charybdis existed, she was a single animal: the ocean itself might scarcely seem able so suddenly to have sucked under so many things so widely scattered and dispersed.’ ²³ Quint. 2.13.15.

²⁴ Quint. 6.3.61 2.

²⁵ Quint. 7.6.67 70.

52

Papers on Quintilian and Ancient Declamation

All this is very well, but in practice it could hardly have been of much service. And Quintilian certainly realized that far more important than any such precepts was the reading of good authors and the building up of the hexis— the state of mind, the settled habit—of fine expression. This is how Quintilian came to insert the famous passage in Book 10 on the Greek and Roman authors that for many is their first and last acquaintance with the Institutio.²⁶ Taken out of context, Quintilian’s remarks often seem both casual and trivial. It is essential to realize what he was trying to do here. He was not writing a handbook to extant literature, or even attempting, within the limits of space that he set himself, a balanced assessment of the worth of different authors. ‘I propose to go through the types of reading that I think particularly suitable for those setting out to become orators.’²⁷ It was normal in ancient literary criticism for judgements to be framed in terms of rhetorical theory; even the excellent ‘Longinus’ thinks in the rhetorician’s rigid categories, and a good deal of his tract On the Sublime is as technical as anything in the Institutio. But Quintilian, by design, is even narrower than this: he tells us about each author what may help the budding speaker, and no more. Sunt qui Propertium malint²⁸ may seem a brutal dismissal of a fine poet. But the whole genre of Latin elegy is given only four lines: there was nothing for the orator here, unless it was the terseness and elegance that Quintilian felt to reach its height in Tibullus. Indeed there was much positively to be avoided: Ovidius utroque lascivior—it was not that he was more pornographic than Tibullus and Propertius, but that his style was unrestrained. He could not master his genius—too much the admirer of his own talent, as Quintilian had remarked earlier²⁹—any more than, in prose, could Seneca, whom ‘one could wish had written employing his own talent and someone else’s judgement . . . If he had not loved everything that he wrote . . . he might win the approval of the educated as well as the admiration of schoolboys.’³⁰ Ovid and Seneca are the distinguished relations of the orators attacked by Quintilian for relying too much on their natural gifts and too little on training. The long passage on Seneca, unchronologically delayed to the end of the chapter, is the key to the rest of Quintilian’s discussion of literature. He protests that he was no enemy of his compatriot; his concern was purely educational—all the young men, at the time when Quintilian was a practising teacher, read Seneca and no one else. These youths had to be brought to realize that there were better authors available; and Quintilian’s attempt to ‘bring back oratory to more austere standards after it had become corrupt and weakened by every sort of fault’³¹ inevitably brought him into conflict with admirers of the most extreme exponent of the modern style. The survey of literature from Homer to Vibius Crispus and Julius Secundus is designed to ²⁶ Quint. 10.1.46 131. ²⁹ Quint. 10.1.88.

²⁷ Quint. 10.1.45. ³⁰ Quint. 10.1.130.

²⁸ Quint. 10.1.93. ³¹ Quint. 10.1.125.

Quintilian and Rhetoric

53

show where these more austere standards were to be sought, who those better authors were. And it is significant that the survey, though it reaches an emotional climax in Cicero (‘born by some gift of Providence as one in whom eloquence might try out her full powers’),³² does not stop with Cicero. Cassius Severus, pioneer of the anti-Ciceronian reaction, might be placed among the foremost if his speeches had had more colour and weight.³³ Domitius Afer, whom Quintilian had heard as a young man, ‘one would not hesitate to include in the company of the great old orators’.³⁴ ‘Future writers will have great scope for sincere praise of orators now flourishing: the talents that adorn the forum today are of the first quality.’³⁵ Quintilian, that is, did not wish to forget everything that had happened in the past century. Progress was still possible, but only on conditions. ‘I should advise mature students’, Quintilian says in another place, ‘to read both the ancients (for if we can take from them their robust virile force and wipe off from it the grime of their period, then this smartness of ours will shine the more brilliantly) and the new authors, for they too have much to be said for them: nature has not condemned us to dullness—it is our fault that we have changed our style of speaking . . . Older writers surpassed us by their principles rather than by their genius.’³⁶ Quintilian hoped to change the ‘principles’—the propositum— of his readers. His day had an elegance—cultus—that was new and in itself desirable. If it could be harnessed to the seriousness and breadth of Cicero, the perfect oratory might yet be attained. It is not, then, to be regarded as surprising that Quintilian himself does not write like Cicero. Even in technical matters, there is a gulf between the rather prolix urbanity with which the De oratore dealt with details of rhetoric and the concise businesslike methods of the Institutio. That is not to say that Quintilian took no trouble to make his book interesting, even, in a mild way, artistic. In 3.1.4 he quotes Lucretius’ famous dictum on the need to smear the rim of a bitter cup with honey. Elsewhere, he says, he has added a certain amount of ‘brightness’ (aliquid nitoris; there is a similar remark in 2.10.12 on the style of declamation compared with real oratory); and it worries him that in the third book the subject may cause there to be more wormwood than honey to the reader’s taste. Yet even here he contrives to be more interesting than his rivals; and in general he tries to vary the tone of his treatise, and not allow precept to go on too long without relief. It may be observed how the prefaces to individual books, while marking off the firm structure of the work, bring variation by striking a more personal note; at the same time, they are cunningly related to the subject of the succeeding chapters. Thus the preface to Book 4 aims to capture the goodwill and attention of the reader in the same way as the exordia that he proceeds to discuss; and the famous preface to ³² Quint. 10.1.109. ³⁵ Quint. 10.1.122.

³³ Quint. 10.1.116. ³⁶ Quint. 2.5.23 4.

³⁴ Quint. 10.1.118.

54

Papers on Quintilian and Ancient Declamation

Book 6 on the death of Quintilian’s wife and children precedes discussion of epilogues and the appeal to emotion. It remains, however, that Quintilian rarely writes a sentence that could be mistaken for one of Cicero’s. The Silver Age shows even in a passage like the following where it is only a question of words that have fallen out of currency:³⁷ Neque enim tuburchinabundum et lurchinabundum iam in nobis quisquam ferat, licet Catō sĭt aūctōr, nec hos lodices, quamquam id Pōllĭōnī plăcēt, nec gladiola, atqui Messala dixit, nec parricidatum, quod in Caelio vix tolerābĭlĕ vĭdētūr, nec collos mihi Calvus persuaserit: quae nec ipsi iam dicerent. No one in our day would tolerate tuburchinabundus and lurchinabundus, even though they have the authority of Cato, or lodices in the masculine, although Pollio likes that, or gladiola yet Messala said this or parricidatus, which is scarcely to be borne in Caelius; nor will Calvus persuade me to say colli. They themselves would not use these words nowadays.

The variety of ways in which Quintilian gives the authorities for the words he cites, the flashes of cretic rhythm and the final epigram all show him using the rhetorical techniques of his age to expound something not in itself very amenable to rhetoric. Moreover, when he does pull out all the stops, Quintilian is clearly not trying to write like Cicero. The lament for wife and children is moving enough; but its rhythms and mood are those of the first century .³⁸ Tuosne ego, o meae spēs ĭnānēs, labentis oculos, tuum fugientem spīrĭtūm vīdī? Tuum corpus frigidum exsānguĕ cōmplēxŭs, animam recipere auramque com munem haurire āmplĭūs pŏtŭī, dignus his cruciātĭbūs quōs f ĕrō, dignus his cogitationibus. Child of my empty hopes, did I see your fading eyes, your departing breath? Did I endure, once I had embraced your body, cold, bloodless, and received your spirit, to breathe any longer the common air? Worthy am I of these agonies that I bear, worthy of these thoughts.

This is grave, and, in its way, effective; but it is not Cicero. And when Quintilian says that his dead wife was so young that her loss could be regarded as inter vulnera orbitatis (the phrase is hardly translatable, but the meaning is that her death was like the death of one of Quintilian’s children), the declamation school is very close.³⁹ Similarly when a diatribe is launched against the modern method of bringing up children to precocious debauchery:⁴⁰ When a child is not yet bringing out his first words, he already knows the meaning of ‘purple’, already insists on the choicest dye. We train their palates before their tongues. They do their growing in litters; if they touch the ground, ³⁷ Quint. 1.6.42.

³⁸ Quint. 6.pr.12.

³⁹ Quint. 6.pr.5.

⁴⁰ Quint. 1.2.6 7.

Quintilian and Rhetoric

55

they hang on hands that support them on each side. We are gratified if they say something particularly daring; words that we should scarcely tolerate from Alexandrian slaves we greet from our children with a kiss and a laugh.

All this scarcely differs in language and sentiment from Seneca. We know that Quintilian meant what he said; but this was a century in which even sincerity was expressed artificially. A fortiori, in matters of language, we do wrong to impose on Quintilian a desire to be Ciceronian. ‘renuntiare officiis occurs thrice in Sen[eca], in Pliny the Younger and even in Ciceronian Quintilian.’ So, at one point, Summers,⁴¹ one of the best commentators on Silver Age Latin. Quintilian did not proscribe—and could not have proscribed if he had so wished—words that originated after Cicero. He is an author of the Silver Age; and he was in revolt not against anything that had happened to the Latin language, or even against the new smartness in itself: only against the slickness and superficiality that all too easily went with that smartness. The answer was breadth of interest; and Cicero was the key to this, not because he had lived a century before, but because he had been broad in the way that mattered. Professor Clarke elsewhere in this volume⁴² shows how meagre was the influence of Quintilian on the educational methods of the Middle Ages. The question of his influence in rhetoric is hardly separable from that other problem. But it may be remarked here that, striking as is the lack of interest in the Institutio during the Middle Ages, his neglect in the later centuries of antiquity is even more extraordinary.[⁴³] The name of Quintilian attached itself to two collections of declamations, known as the Minor and the Major. The Minor, which may perfectly well have some direct connection with the teaching of the master, are preserved only by the most tenuous thread of manuscripts. The Major were more popular; and they at least can claim to have been quoted by respectable authors of late antiquity. The Institutio,⁴⁴ however, scarcely causes a ripple in the same period. The use made by St Jerome of a single chapter of the first book is quite exceptional. Scholars have hardly been able to dredge a convincing reminiscence out of Augustine; rather later, Boethius seems to have had a phrase from Quintilian’s criticism of Seneca at the back of his mind while he was writing his Consolatio philosophiae.⁴⁵ It may not be coincidental that the Seneca passage comes in a part of the Institutio that we find excerpted for a brief ‘Quintilian on authors’ during the Middle Ages. The Institutio, in fact, was too long a book, and too technical, to win many readers. Only another Spaniard, Isidore, seems to have ⁴¹ (1910), 335. ⁴² Clarke (1975). ⁴³ [Important advances in our knowledge of the use made of Quintilian in late antique texts have been made by Ulrich Schindel, especially in Schindel (2001).] ⁴⁴ See Colson (1924),  ff. ⁴⁵ Winterbottom (1967c). [I now think this parallel no more than a coincidence.]

56

Papers on Quintilian and Ancient Declamation

paid much attention to it; and, ironically enough, though he employs it on grammar and even on music, he looks elsewhere for information on rhetoric. One or two shadowy rhetoricians, especially Julius Victor, made use of Quintilian; extracts from the Institutio attached themselves to certain manuscripts of the great Cassiodorus (who indeed himself seems in his commentary on the Psalms to have gone to Quintilian for a few definitions), and there are more in a wonderful eighth-century manuscript (Paris. lat. 7530) from Monte Cassino. That is virtually all. Manuscripts of the Institutio emerge for us first in the ninth century, one group from central France, the other probably from Italy. There are not many now, and there probably never were very many more. One—indeed the most important—of the French group (the Berne MS 351) was, it would seem, the property of the proto-humanist Servatus Lupus, abbot of Ferrières on the Loire. The details of Lupus’ dealings with Quintilian are revealing. Twice he wrote in search of manuscripts of the author, once to York when he had no text at all, and once to Rome when he had merely a mutilated text (as indeed the Berne manuscript is). But he profited little from such of Quintilian as he managed to come by. Perhaps he did not read all that he had. Lupus’ method with his books was to read them carefully, making corrections where necessary and noting interesting words in the margin. This process in Quintilian proceeds little beyond Book 2; and I have seen no sign in Lupus’ letters of any use made of the Institutio. A monk of St Gall two hundred years later was rather more energetic. He had access to a complete text of Quintilian (now Zurich, Zentralbibliothek, 288), and he corrected it for about three-quarters of the way; then, shortly after writing in the margin a despairing couplet about the difficulty of emendation where one has no exemplar to help one, he too turned to other matters. A still more persistent monk, this time at Bec, produced an elegant and intelligent abbreviation of the Institutio in the middle of the twelfth century (MS Paris. lat. 14146). He, like Lupus, had only a mutilated version to work from; but he spread his interests widely, and did not fight shy of the technical details, though he tended to omit the examples. All the same, it was these details that presumably put people off in the Middle Ages. Quintilian was a great name, a revered ‘author’. But when it came to reading him, it became clear how little there was in common between ancient rhetoric—elaborate dogma for the speaker in Roman court or senate—and the medieval ars dictandi that concerned itself with less august situations. Only a John of Salisbury could get much out of the Institutio; and he, after all, read everybody. The ordinary man was more at home with the brief extracts on morality and etiquette that went the rounds of medieval France as part of the Florilegium Gallicum. In the fourteenth century Petrarch, like Lupus before him, searched for Quintilian; found him; then gloomily deplored his mutilated state. Even the part of the book available to him, however, was more than enough for

Quintilian and Rhetoric

57

Petrarch. His tiny notes and shrewd emendations in the margins of the Paris. lat. 7720 stray little outside the first and twelfth books. And indeed it was only in the fifteenth century that Quintilian really met with any wide appreciation. Gasparino Barzizza was even prepared to provide a completion of the still mutilated Institutio. But this kindly thought was made unnecessary by the discovery of a complete text—made by Poggio in a squalid tower at St Gall, where the very manuscript laboured on so lovingly over three hundred years before at last came to light. This in 1416; and perhaps a hundred manuscripts dated between then and the end of the century testify to the popularity of the book in Renaissance Italy, France, and Germany. The great Lorenzo Valla covered the margins of his manuscript (Paris. lat. 7723) with commentary; and his emendations win for Valla the right to be called the first editor of Quintilian.[⁴⁶] The story after that is of a slow fading, marked by a steady decline in the number of editions produced as the centuries passed. Nowadays the average classics undergraduate never opens the Institutio. If this were a sign that rhetoric is dead, there would perhaps be no need for grief. And indeed, for the moment and in this country, it is dead. But persuasion lives on, and always will live. It is here that Quintilian remains relevant. For it is the great virtue of the Institutio that it faces up to the dangers of persuasion. It makes some show, at least, of replying to the criticisms that Plato had voiced long before. ‘Most people,’ writes Quintilian,⁴⁷ ‘content to read a few clumsy excerpts from Plato’s Gorgias . . . , have fallen into the serious error of believing that Plato held that rhetoric was no art but a mere knack of giving pleasure.’ Despite the arguments that follow, Quintilian himself overestimates Plato’s enthusiasm for rhetoric and confidence in the possibility of a just orator. But Plato’s criticisms are always in the back of his mind. This is one of the factors that led him to reject definitions of rhetoric that made it only a means of persuasion (2.15.2 ff.). If it were that and that only, it would have to be judged simply on its results, on its success in winning cases. But an orator can be good even if he loses his case; and others—prostitutes, flatterers, seducers—are persuaders without being orators. The solution for Quintilian was to import a value judgement into his definition. Rhetoric is the ‘knowledge of how to speak well’;⁴⁸ and the orator—as Cato had said—is the good man who is skilled at speaking.⁴⁹ Quintilian devotes a whole chapter to a quasi-philosophical ‘proof ’ of the virtues of this definition. ‘Well, let us grant, what is in fact impossible, that we have found some bad man who at the same time is supremely eloquent. Nevertheless, I shall deny that he is an orator . . . He who is called in as counsel ⁴⁶ [To this survey of the manuscript tradition of Quintilian add e.g. Winterbottom (1967b) and (1970b).] ⁴⁷ Quint. 2.15.24. ⁴⁸ Quint. 2.15.38. ⁴⁹ Quint. 12.1.1.

58

Papers on Quintilian and Ancient Declamation

for the defence needs to have trustworthiness that avarice cannot corrupt, influence deflect or fear break down. Or are we to give the sacred name of orator to a traitor, a turncoat, a conniver?’⁵⁰ In this passage, Quintilian’s procedure is particularly transparent. An orator must be a good man by definition. Anyone who is not good is ruled out automatically—ruled out, that is, by force of logic; not, however, ruled out of worldly influence. And this was where the rub came. You can refuse to give the name of orator to the corrupt barrister (or the unscrupulous advertising agent); but if he is doing well out of his corruptness, he will not be unduly offended. It was the custom of the ancient literary critic to avoid mention of living writers. This meant that Quintilian was relieved of the necessity of mentioning one of the most notable and notorious speakers of his day, one who looms large in the letters of Pliny—the Regulus, in fact, who thought he might have offended by commenting on Pliny’s Ciceronianism. This man was one of the class of delatores, the informers, who, as was natural, flourished under the principate. They were encouraged by a law that gave successful prosecutors a share in the estates of their victims, and further encouraged by emperors who found them useful agents of their regimes. A number of the most distinguished orators of the century—Domitius Afer and Vibius Crispus among those mentioned by Quintilian—were tarred with this brush. By the end of the century the delatores were important people. Regulus, who had ruined noble families under Nero, came into his own under the Flavians; his crimes under Domitian, remarked the benign Pliny,⁵¹ were no less heinous than those under Nero—merely better concealed. These crimes brought him the consulship, at a date unknown; their instrument was an abusive and violent style of oratory. He called Rusticus ‘that Stoic ape’; and we can measure his quality from another remark to Pliny—‘You think one has to go through all the details in a case; I see the throat at once—and grip that.’⁵² Here was the spirit that felt the prolixity of Cicero to be out of date. But Regulus’ own talents were, at least in Pliny’s view, unimpressive: ‘Weak lungs, indistinct speech, hesitant tongue, snail-like imagination, no memory, nothing but an insane genius. Yet, thanks to this very shamelessness and madness he has come to be regarded as an orator.’⁵³ Not, we imagine, by Quintilian. This Regulus—and there were other speakers like him—was on two counts opposed to every Quintilianic ideal. He was lacking in the technical make-up of the orator; and he was a bad man. Quintilian’s book might have been written as a protest that such a man could go so far. Regulus, someone said of him, was a bad man unskilled in speaking.⁵⁴ Cato’s definition, and Quintilian’s, was overturned.[⁵⁵]

⁵⁰ Quint. 12.1.23 4. ⁵¹ Plin. Ep. 1.5.1. ⁵³ Plin. Ep. 4.7.4. ⁵⁴ Plin. Ep. 4.7.5. ⁵⁵ [For this topic see in greater detail A.1 above.]

⁵² Plin. Ep. 1.20.14.

Quintilian and Rhetoric

59

On the practical level there was nothing that Quintilian could do about Regulus or his like; though there was a revulsion against informers at the beginning of the reign of Trajan not long after the publication of the Institutio, there is no reason to suppose that it was caused by the Institutio. For one thing, writing under Domitian, Quintilian could not be too specific in his plea for moral standards to be applied to oratory. But the plea remains impressive; it is one that is rarely made, but that should never be silenced: I have tried, as far as a man could, to contribute something to the technique of speaking; and I should deserve ill of humanity if I were making ready this weapon for the use of the highwayman, not the soldier. And, quite apart from myself, Nature, in the very gift to man by which she shows herself to have been particularly generous to him and particularly to have separated him from other animals, Nature herself will prove to have been no parent but a stepmother if she designed the power of speech to be the companion of crime, the opponent of innocence and the enemy of truth.⁵⁶

⁵⁶ Quint. 12.1.1 2.

4 The Text of Sulpicius Victor Nec enim volo haec in multorum manus pervenire: thus the modest Sulpicius Victor, dedicating his Institutiones oratoriae to his son-in-law. His wish has been granted. Karl Halm edited the little book among his Rhetores Latini Minores in 1863, but since then scholars have done almost nothing to improve his text. I make no apology for venturing into a stable which, without being Augean, is in some need of spring-cleaning.[¹] There is, so far as I know, no manuscript of the work. We possess it only because it is contained in a precious Basle edition of 1521² ‘ex cod. Spirensi cum omnibus vitiis quibus scatebat diligenter expressa’. Apart from this (B), Halm’s apparatus cites two other editions, that of Pithoeus (1599) and that of Capperonnier (1756). I cite throughout from Halm by his pages and lines. 314.9 ff. Sulpicius distinguishes thesis, res rationalem³ disputationem recipiens sine definitarum personarum circumstantia, and hypothesis, res rationalem disputationem recipiens cum definitarum personarum circumstantia. But, he continues, the definitions may on occasion be reversed: Nam et thesis potest habere definitae personae circumstantiam, si ita ponamus: ‘An Ciceroni post consulatum eundum in provinciam fuerit’; vel . . . : ‘An Socrati uxor ducenda fuerit.’ Neque enim esset thesis, si illo modo poneretur: ‘Deliberat Cicero, an post consulatum eat in provinciam’, aut ‘Deliberat Socrates, an ducat uxorem’: quae sunt causae deliberativae. Sed quoniam nos hoc disputamus, debueritne ille facere, est thesis, quamvis sit definita persona. In B what follows ducenda fuerit is Neque enim est hypothesis, and this suggests that Sulpicius made his point positively: Haec enim esset hypothesis (answered by Sed . . . est thesis, which parallels the singular referring to the two examples preceding). Compare to [Bulletin of the Institute of Classical Studies 26 (1979), 62 6] ¹ [A Teubner edition of Sulpicius Victor by R. Jakobi is forthcoming.] ² ‘Veterum aliquot de arte Rhetorica traditiones . . . in inclyta Basilea an. M.D.XXI.’ With a foreword ‘Ioannes Frobenius lectori s.’ The quotation in the main text is from Halm’s preface (1863),  . I have checked the passages I cite against B, and note one or two variations not observed by Halm. ³ Therefore read λογικήν in Theon RhG .120.13 Spengel [= p. 82 Patillon, who so reads]. Compare also Max. Plan. RG v.252.29 Walz.

The Text of Sulpicius Victor

61

some extent Quint. 2.4.25: Sunt et illae (sc. theses) paene totae ad deliberativum pertinentes genus: ‘Ducendane uxor’, ‘Petendine sint magistratus’; n a m q u e et h a e personis modo adiectis suasoriae erunt.⁴ 315.28–9 Sulpicius is discussing cases of the kind known as ἀσύστατα. One type is ἄδοξος, id est inhonesta, turpis et de rebus personisve deformibus. B has deformis, and a third adjective parallel to inhonesta and turpis would be welcome (though compare 317.16–17: Haec (sc. causa) non de turpi, sed de re parvula et humili est), and deformis may be retained if we delete de as an anticipation of it. Cf. 322.14: in eo genere causarum, quod est rebus atque personis honestum. 316.13–17 Pathetica est causa, cum personae eius quae loquitur repraesentandus adfectus est. Si necesse erit vel indignatione aliqua atque ira vel dolore aliquo vel, ut plerumque accidit, luctu † excitatos excire, non erit otiosum, ut commota sit et excitans omnis oratio, perinde (proinde B) atque ipsa res exiget. B starts the second of these sentences si sese debebit, and one might think si decebit more appropriate to the context (debebit is corrupted to decebit at line 28). In what follows, it is reasonable to suppose that excire has been influenced by excitatos and may be replaced by a verb with regard rather to sense than to palaeography. Imitari, then, or agere: for Sulpicius is speaking of the ‘representation’ of emotion, the orator playing the part of persons subject to various passions. 316.23–4 Post haec intellegendum, quemadmodum diximus, ex quo modo causarum sit controversia. Modi autem causarum sunt quinque. The reference back is to 315.15–17: Cum hypothesin esse intellexerimus, id est controversiam, intellegendum erit an consistat, tum ex qua specie sit, deinde ex quo modo . . . ⁵ This suggests that we should delete the first causarum, which is needed neither with modo (one would if anything expect causae) nor with controversia. 321.29–31 Hic erat ordo re vera, ut de statibus protinus traderemus, si non esset a Zenonis vestigiis recedendum: sed professi sumus usuros nos esse iudicio, videbitur exigere aliquid inserendum esse de meo. The reference back is to 313.3–6: Zenonis praecepta maxime persecutus, ita tamen, ut ex arbitrio meo aliqua praeterirem, pleraque ordine immutato referrem, nonnulla ex aliis quae necessaria videbantur insererem. That supports Halm’s addition of si (though one might suggest , which could easily disappear after iudicio; compare 326.36: sic rerum suarum ratio deposcat), but the verb should be videretur (videtur B), the mood because of the oratio obliqua, the tense because videbantur is being recalled. 322.24–6 Vitia exordiorum et alia quidem, sed quae maxime accident maximeque dicenda sint, haec fere sunt: . . .

⁴ For haec cf. 314.21.

⁵ Cf. also 315.8: intellegere debemus, cuius modi causa sit.

62

Papers on Quintilian and Ancient Declamation

Dicenda is pitifully weak; rather vitanda. Cf. Victorin. RLM 200.26–7 Halm [= 66.12–13 Riesenweber]: Verum haec maxima vitia exordiorum esse dicit atque haec omnino vitanda esse praecipit, as well as 322.35: Omnia haec vitia vitanda sunt. 322.33–4 Contra praecepta (sc. exordium erit), si, cum debemus iudicem docibilem facere, attentum utcumque faciamus. Attentum cannot be right, for contra praecepta est quod nihil eorum efficit quorum causa de exordiis praecepta traduntur,⁶ and attentiveness is one of those things, as Cicero goes on to say. Read then infestum, comparing Victorinus on that very passage of Cicero:⁷ contra praeceptum est . . . quod iudicum animos sibi non conparat, interdum etiam infestos facit. 323.9 ff. Sulpicius remarks that the clarity of a narration is furthered if it opens with a nominative, ut ferme narrationes sunt Tullianae. After brief citations from S. Rosc. 15 and Mil. 24, he concludes: nec facile apud Ciceronem est ulla coepta narratio. It was Pithoeus who added aliter, rightly, he too who changed B’s facere to facile. Facile would go easily enough with a verb like invenitur, but it is out of place here; we should prefer fere, despite the proximity of ferme. The word is often used by Sulpicius (e.g. 336.1–2: Non enim fere accidit . . . ). For the combination with ulla see TLL s.v. fere 496.9. At Quint. 6.pr.10 fere is corrupted to fecere. 323.16–20 A narration will be probabilis, si argumentorum et quaestionum semina quaedam fuerint ubique respersa, modo ne argumentandi genera ponantur: ut tempus, quo factam esse dicimus, adsit, et causa cur facta sit, et persona quae fecit, et facultas quare facere potuisse credatur, et locus ubi facta est. B reads: ut omnibus quae facta esse dicimus adsit et causa cur facta sint, and to accept this we need only change to ubi facta sunt below. There is no reason why tempus should be placed first; below, Sulpicius exemplifies merely causa, which comes first in B’s version. We may do without it altogether (Sulpicius is being rather short with the list: compare for example Victorin. RLM 207.1 ff. Halm [= 75.1 ff. Riesenweber]), or add tempus et before locus below. 324.6–9 In altero stultitiae erit crimen, si videbimur (i.e. in a partitio) dicere id, quod necessarium non fuerit causae: in altero diffidentiae, si confessi fuerimus ipsi nobis esse dicendum, et videbimur tamen per diffidentiam et metum praeterire. Halm suggested praeterisse. More urgent, perhaps, the addition of quid before confessi. 325.28–9 Et ut istorum locorum (those of a coniectura perfecta) vim perspicere possimus, eamus per singulos: sit et controversia, quam supra posuimus, exemplum. For et, rather ea, or haec, or et .

⁶ Cic. Inv. 1.26.

⁷ Victorin. RLM 201.14 15 Halm [= 67.11 12 Riesenweber].

The Text of Sulpicius Victor

63

Sulpicius continues: Probationum expetitio primus locus ex rei parte, ut si hic dicat adulescens, crimen maiestatis ex manifestis probationibus oportere convinci: indignum enim esse, ut ex suspicionibus levibus in tantum periculum devocetur. B’s expetitione points to expetitio est; compare 326.33: Derivatio est locus ex rei parte. And its magis et manifestis points to magnis et manifestis, for we know the nature of the charge from line 23, and it is adfectata tyrannis, not maiestas. Cf. (besides Act. Apost. 2.20: antequam veniat dies Domini magnus et manifestus) Cic. Clu. 50: crimen tantum ac tam manifestum. I have no parallel for magna probatio, but the adjective is common with argumentum (TLL s.v. argumentum 547.13).[⁸] 326.13–17 Sed reus poterit haec eadem (i.e. arguments from motive) quoque ad defensionem suam vertere, si res suggeret: ut si hic dicat adulescens, quod dives esse dicatur, in neminem minus quam in se suspicionem istius criminis cadere, cui optimum sit et optatissimum, copiis suis et statu rerum suarum perfrui tranquillaque quiete rei publicae. Something is clearly wrong here, and B’s enim after neminem is confirmation. We should accept enim, place a colon before in neminem, and posit a lacuna before that. The sort of thing that has fallen out can be seen from parallels like 329.3–5: quod ipse dives et ille pauper, hoc pro illo debeat valere, quod abiectum hominem in simultate magis despicere voluerit quam odisse (also 330.30–1). 329.35–330.2 . . . controversia illa: ‘Pauper et dives inimici. Dives non comparente filio suo rapuit pauperis filium et in domo sua torsit et incertum quid dicentem (quo dicente B) in tormentis necavit. Ita et pauper est reus diviti conscientiae.’ Ita goes against all the conventions of style in declamatory themes. But there is more. Later, we hear that Aliud signum criminis ponitur ex parte contraria, quod taceat (331.1–2), and that poterit pauper haec dicere: si caedes illa a filio per conscientiam ipsius fuisset admissa, non se fuisse taciturum, neque ex silentio suo posse esse suspitionem (331.7–9). The last sentence of the theme should therefore read: Tacuit pauper, et reus est diviti conscientiae, or the like. Compare to some extent [Quint.] Decl. mai. 19.th. (371.9–13 Håkanson): Speciosum filium, infamem tamquam incestum cum matre committeret, in secreta domus parte torsit et occidit in tormentis. Interrogat illum mater quid ex iuvene compererit. Nolentem dicere malae tractationis accusat. 331.10–14 Etiam ille modus est duplicis coniecturae, qui a Graecis συγκατασκευαζόμενος στοχασμός dicitur, ideo scilicet, quia una eademque actione duo pariter efficienda sunt et probanda. Ιd fit, cum vel ipsa res, quae causam criminis dicitur attulisse, in dubium devocatur, alia coniectura probanda est.

⁸ [For magna probatio cf. Aug. Serm. 88.652 (ed. Verbraken (1984), 100); Ep. ad Aug. 6.1 (CSEL 34/1.12).]

64

Papers on Quintilian and Ancient Declamation

It is clear from the succeeding discussion that Halm’s addition of vel misleads us as to the true nature of the two sorts of case intended. At 332.27–8 we have: Hoc modo ex causa duplex coniectura est, sed diximus duplicem ex signis criminis fieri. This needs to be given a reference, and we must therefore add vel signum criminis before vel ipsa res (ipsa thereby gaining in point as well) at 331.12. That the signum too in dubium devocatur is confirmed by 333.12–15. Now both res and signum, naturally, must be proved alia coniectura to make the coniectura ‘double’, and the connective after devocatur must therefore be et. 336.32–3 Sulpicius lists the topics for discussion in causae finitivae: finis, contraria definitio, legislatoris , voluntatis coniectura, qualitas conclusiva. It is not only voluntas that B omits, but a further topic, maius, to be added before voluntatis, as the discussion in 337.16–23 shows. 338.25–7 . . . ut si haec causa sit posita: ‘S     . Patrem suum reum proditionis fecit et optinuit; petiit sacerdotium.’ The consistent practice of themes shows that we must read the present petit. 340.33–5 Nam utrumque horum (sc. poena and accusatio/defensio) nomine actionis significatum potest videri, magis etiam, si Graece controversia explicetur, qui δίκην tam poenam quam actionem vocant. Qui can be defended, but quia is preferable. 343.18–20 Secundus locus est a summo ad imum, per quem singula quae praecesserunt quaeritur, iustumne sit ut petant filios qui necessitate misera coacti sint exponere. A passage below (343.31–2: Cetera fere per ipsum locum, qui est a summo ad imum, considerantur: an ex his quae acciderant iusta sit causa mortis) supports Halm’s per quem, but it also suggests the possibility of reading then singula . . . quaeruntur: iustumne . . . 343.28–31 Ponunt quidam nonnulli etiam generalem quaestionem huiusmodi: an sola sit causa moriendi (i.e. taking one’s own life) debilitas corporis; et rursum, an satis sit ad causam moriendi, cum nemo , nisi causa sit, velit. B has causam morum, cum nemo sine causa sit, velit, from which we should make causam mori , cum nemo, sine causa sit, velit. 349.31–2 Erit igitur divisio in huiusmodi causis, ut sit propositio criminis a summo ad imum, quibus commemoret accusator, quae beneficia praestiterit, quibus iuverit . . . A comma should be placed after criminis to separate the two topics (for the first compare 348.34). In what follows, the first quibus seems to be an anticipation of the second. The reference should be restricted to the locus a summo ad imum, and one expects in quo, as at 326.24. 352.9–12 et est fere (sc. the status legum contrariarum) quasi duplex scripti et voluntatis status, quia utrorumque scriptorum interpretatio est. Ergo etiam loci hi sunt: propositio scripti et interpretatio scripti, propositio alterius scripti et interpretatio scripti. Add alterius before the last scripti.

The Text of Sulpicius Victor

65

I end with a brisk list of some passages where it seems to me that readings in Halm’s apparatus, of B unless otherwise stated, have been unjustly rejected or suspected. 313.13 orator. 317.27 haec, corresponding to his below. 319.10 dicimus: here and in a number of other places Halm emends a present of dico or the like to the future; the author was probably inconsistent, but there is no need to expel the presents everywhere (observe for example 319.19; 326.6, 10, 24, 27–8; 335.9; 336.13–14; 337.18–19: some more difficult than others). These are not so much presents of precept as statements of customary practice. 319.27 Halm suggests the addition of sumenda; but cf. 326.1–2. 327.9–10 dividi Pithoeus, rightly; cf. 327.19 dividuntur posterioribus locis (not to speak of line 8). 328.4 privignus semper infestus. 329.13 imperare (cf. Sen. Con. 7.1.6: difficile esset parricidium facere, etiam quod imperaret pater). 333.33 The omission of quidam thus is common in declamation themes, despite my remarks elsewhere.⁹ Cf., for a start, 338.26 (also 350.9). 339.8 ut (compare lines 9–10). 344.29 hic (in the contio: cf. line 27). 345.3 tota actione (defended by the rhythm, while TLL s.v. dispergo 1409.28 ff. shows that the ablative is as proper as the accusative; it does not seem quite necessary to bring the words into line with 346.16).

⁹ Winterbottom (1974a), 32 [= A.2 above, p. 34].

5 Cicero and the Silver Age Materials for a proper assessment of the influence of Cicero on the literature of the first century  do not exist. The only Latin speech to survive from that century, the Panegyricus of Pliny, is indebted in ideas and wording to the Pro Marcello;¹ but the genre is peripheral, and Pliny, as a devoted emulator of Cicero, may be equally untypical. The only corpus of philosophy to come down to us from this century, the Letters and Dialogues of Seneca, is hardly comparable with Cicero’s philosophical writings:² style, sect, and personality pull in quite other directions. We may suspect that Tacitus’ Dialogus was quite exceptional in its anxiety to follow the tradition of Cicero’s rhetorica.³ One would not expect Cicero’s poetry to have had any resonance so late.⁴ And though there was use made and admiration expressed of his letters,⁵ Pliny⁶ found that the difference of political background dictated—as Seneca’s interests and character had dictated—a different manner and form. In these circumstances, there is room for speculation rather than analysis. But I shall begin with some remarks on a more tangible topic, the reputation of Cicero in the first century as a historical and literary figure, and go on to consider how his writings were employed by grammarian and rhetorician. Only then shall I try to gauge what sort of gulf separates Cicero from the Silver Age.

[From W. Ludwig (ed.), Éloquence et rhétorique chez Cicéron (Vandoeuvres and Geneva, 1982), 237 66 (‘Discussion’, 267 74)] ¹ Mesk (1911), 81 4, who notices a more general debt to the Pro lege Manilia (recommended, it may be noted, by Fronto to Marcus Antoninus: .30 Haines [= 225.3 van den Hout²]). ² The De amicitia aroused interest in Gellius (1.3.11 ff., on Cicero’s superficiality; 17.5, defence against a caviller). Didymus’ Περὶ τῆς Κικέρωνος πολιτείας was presumably political, not philosophical; hence Suetonius’ answer (Schanz Hosius Krüger (1922³), .60). ³ Though Columella cites the Orator in pr.29. ⁴ The criticisms are familiar; see e.g. Gudeman (1914²), 350. ⁵ Nep. Att. 16.3: undecim volumina epistularum . . . : quae qui legat, non multum desideret historiam contextam eorum temporum, and what follows. Suetonius at least made a show of exploiting this source. See also Fronto, .158 Haines [= 104.12 van den Hout²]: Omnes autem Ciceronis epistulas legendas censeo, mea sententia vel magis quam omnes eius orationes. Epistulis Ciceronis nihil est perfectius. He was interested in both wording and content. ⁶ Ep. 9.2.2; cf. Sen. Ep. 118.1 2.

Cicero and the Silver Age

67

In following the course of Cicero’s reputation,⁷ it is unreal to separate the historical from the literary. The persuasiveness that Cicero could command in his spoken speeches extended to their written counterparts, and his own view of the two major crises of his career, the consulship and the struggle against Antony, imposed itself.⁸ Catilinarians and Philippics moulded opinion after his death even more masterfully than when they were delivered. Thus, from Sallust on, the Catiline affair is given a good press.⁹ Catiline is seen as wholly bad, and no awkward questions are raised as to the legality of Cicero’s actions. It was, of course, possible to be less polite. For the Pseudo-Sallust, the consulship caused the conspiracy.¹⁰ This sort of line emerges once in Appian’s account: the plotters plan to kill Cicero and accuse him in the assembly as ‘a cowardly war-monger who was turning the city upside down when there was no danger’.¹¹ The violent declaimer Romanius Hispo called him, in similar vein, turbator oti.¹² Such criticism will have its roots in the propaganda against Cicero after his consulship, which was stoked up again for his own purposes by Antony.¹³ And it will have found literary expression in the history of Asinius Pollio,¹⁴ who even in his laudatory obituary said that Cicero ‘displayed more spirit in picking quarrels than in carrying them through’.¹⁵ But almost invariably we find such criticism in a dialectical context, where it is immediately balanced by a neighbouring defence. Varius Geminus, one of the few declaimers to exhort Cicero to beg Antony’s pardon, and a man accustomed to voice scurrilia,¹⁶ himself pleads the other side a section earlier in the Elder Seneca. The pseudoSallust’s assault is partnered by a comforting invective of a Pseudo-Cicero. Even Calenus’ long attack in Dio Cassius (46.1 ff.) is the answer to an equally

⁷ For which see Richter (1968), 161 97. ⁸ ‘Nusquam laudes minuuntur, quas Cicero ipse sibi tam large attribuerat’: Petzold (1911), 59. Note Dio Cass. 37.42.1: Κατιλίνας . . . ἐπὶ πλεῖόν γε τῆς τῶν πραχθέντων ἀξίας ὄνομα πρὸς τὴν τοῦ Κικέρωνος δόξαν καὶ πρὸς τοὺς λόγους τοὺς κατ’ αὐτοῦ λεχθέντας ἔσχε. ⁹ For use made of this in declamation see Zieliński (1908²), 345 6; Kurfess (1914), 512 ff. Elsewhere, see (after Verg. Aen. 8.668 9) e.g. Juv. 8.231 ff.; Flor. Epit. 2.12; Gell. 5.6.15; Plut. Cic. 22; contrast Dio, for whose views on Cicero see Millar (1964), 46 55, modified by Richter (1968), 192 7. Plin. Nat. 7.116 thought Cicero’s consulship enough to ensure his fame; but he mentions his dealings with Antony too (117) another topic that received a good press (naturally enough): see e.g. Livy, ap. Sen. Suas. 6.17; Vell. 2.64.3; Juv. 10.122 ff.; Plut. Ant. 2.1 2 and 20.2 4, all sympathetic to Cicero. Declamation had its role here too: see below, pp. 76 7. ¹⁰ Ps. Sal. Cic. 3. ¹¹ App. B. civ. 2.3. ¹² Sen. Con. 7.2.13. ¹³ As can be deduced from e.g. Cic. Phil. 2.15 ff. Cf. Dio Cass. 46.2.3: οὗτός ἐστιν . . . ὁ τὸν . . . Κατιλίναν ἐκπολεμώσας ἡμῖν. Cicero was represented as a trouble maker between Pompey and Caesar (Phil. 2.23), a charge answered by Vell. 2.48.3 5 (Petzold (1911), 59 60). Note also how the charge that Cicero was cruel in 63 (Sul. 7 8) reappears in Ps. Sal. Cic. 6. ¹⁴ Gabba (1957), 317 39. Doubted by Richter (1968), 194 n. 107 (see also 178). ¹⁵ Sen. Suas. 6.24. ¹⁶ Sen. Suas. 6.12.

68

Papers on Quintilian and Ancient Declamation

long speech of Cicero’s. If we see the case against Cicero from time to time, that is largely because rhetoric thrives on cut and thrust.¹⁷ Generally, however, Cicero’s career, and even his character,¹⁸ was viewed sympathetically. His talent for self-praise (illum ipsum consulatum . . . non sine causa sed sine fine laudatum)¹⁹ found a defender only in the loyal, and perceptive, Quintilian.²⁰ But it was agreed that he had something to praise: non sine causa. Nor was the verdict unanimous that he lacked constantia.²¹ Livy said he faced none of the disasters that confronted him ut viro dignum erat;²² but in a crisis things were different. Asconius comments on his constantia in agreeing to defend Milo (p. 38.6 Clark), even though the actual speech was deficient in that respect (p. 42.1 Clark); Velleius talks of his bravery in 63.²³ Lucan gives him a brave and indeed aggressive speech (7.68 ff.).²⁴ All in all, Cicero was seen as a man whose life could provide a sympathetic exemplum: not only of the troubles of the human condition²⁵ but of the ingratitude of the state²⁶ and of an undeserved death.²⁷ The anger we feel at Clodius’ and Antony’s treatment of Cicero is not seen as unreasonable by Seneca;²⁸ similarly, Velleius comments on the erumpens animo ac pectore indignatio that informs his account of the murder.²⁹ An exemplum: Cicero has faded into the past, and taken his place among the Roman heroes whose grandeur absolved their posterity from the effort of cool assessment of their actions. Only while memories were very fresh could a critical eye be brought to bear; and then it tended to be jaundiced, as Pollio’s was, by political differences.³⁰ So, too, with Cicero’s literary reputation. In the last years of his life, his style came under attack from the Atticists (whoever they were). If Calvus was one of them, his motives might be supposed to have included rivalry with one against whom he struggled in the 50s to attain the principatus eloquentiae.³¹ After Cicero’s death, Asinius Pollio is in the vanguard of literary as well as historical assault. If he was infestissimus famae Ciceronis,³² it did not help that he had been a rival of his eloquence. He was offended, because he took it personally, when Sextilius Ena recited: Deflendus ¹⁷ It is perhaps in this context that we should regard the speeches purporting to be by Catiline and C. Antonius known to Asconius (p. 94.1 Clark; cf. Schol. Bob. on Sul. 2.2 and Quint. 9.3.94), though he thought they were the work of obtrectatores Ciceronis. Compare the reply by Cestius to the Pro Milone (Quint. 10.5.20; also below, p. 69). ¹⁸ Though note the defensiveness of Quint. 12.1.16 ff. ¹⁹ Sen. Dial. 10.5.1; cf. Plut. Cic. 6.3 5 and 24.1 4. ²⁰ Quint. 11.1.18. ²¹ Sen. Con. 2.4.4. ²² Sen. Suas. 6.22. ²³ Vell. 2.34.3. ²⁴ But notice 67: Addidit invalidae robur facundia causae, the sophist’s gift. We might see the speech as Cicero the πολεμοποιός in action again (see above, p. 67 with n. 13). ²⁵ Sen. Dial. 10.5.1. ²⁶ Sen. Ben. 5.17.2; cf. Vell. 2.45.2. ²⁷ Sen. Dial. 9.16.1. ²⁸ Sen. Dial. 4.2.3. ²⁹ Vell. 2.66.3. ³⁰ Quintilian comments: postea . . . quam triumvirali proscriptione consumptus est, passim qui oderant, qui invidebant, qui aemulabantur, adulatores etiam praesentis potentiae non respon surum invaserunt (12.10.13, on which see Güngerich (1950), 246 7). ³¹ Sen. Con. 7.4.6. But note the remarks of Gruen (1966), 215 26. ³² Sen. Suas. 6.14.

Cicero and the Silver Age

69

Cicero est Latiaeque silentia linguae: ‘I do not propose to listen to someone who thinks I am dumb.’³³ And Quintilian remarks that Asinius and his son³⁴ attacked the faults of Cicero’s oratory like enemies, etiam inimice.³⁵ A stylistic gulf ³⁶ separated the two—the result of their rivalry, or a cause of it. Seneca finds himself contrasting them when it came to compositio.³⁷ Again, there was personal animosity behind the attitude of Cestius. Not only did this rhetor put on his syllabus only those speeches of Cicero to which he had composed replies:³⁸ he was positively infestus to Cicero (little though he admired anybody)—and got whipped for his pains by Cicero’s son.³⁹ These views did not disappear in the first century .⁴⁰ Largius Licinus, whose book was provocatively entitled Ciceromastix, is coupled by Gellius⁴¹ with Asinius Gallus among people who pedantically found fault with details of Cicero’s diction.⁴² The schoolmen could be represented, however maliciously, as regarding themselves superior to Cicero—though not to Gabinianus.⁴³ But all this was nothing compared with the torrent of encomium. Romani maximus auctor / Tullius eloquii. That is Lucan’s summary,⁴⁴ and it is typical:⁴⁵ apud posteros . . . id consecutus ut Cicero iam non hominis nomen sed eloquentiae habeatur.⁴⁶ More interesting the testimony of those who might have been expected to be less than enthusiastic. Gellius⁴⁷ did not think of his favourite archaic writers as superior to Cicero; his comparison of Cato, Gracchus, and Cicero on a similar topic finds the Verrines a clear winner⁴⁸—for

³³ Sen. Suas. 6.27. ³⁴ Who continued the old feud by comparing his father with Cicero (Plin. Ep. 7.4.3; cf. Gell. 17.1.1): answered by an emperor (Suet. Cl. 41.3). ³⁵ Quint. 12.1.22. ³⁶ Quint. 10.1.113. ³⁷ Sen. Ep. 100.7. ³⁸ Sen. Con. 3.pr.15. ³⁹ Sen. Suas. 7.12 13. ⁴⁰ But Zieliński (1908²), 44 seems to overestimate (and overdramatize) anti Ciceronianism in the schools. If their style was different from his, that was at least partly a matter of genre; and Cicero’s own declamations may not have been in his forensic manner. ⁴¹ Gell. 17.1.1. ⁴² Normally regarded as exemplary, and not only by Quintilian (1.5.44: Sed quem potius ego quam M. Tullium sequar?): see Sen. Ep. 58.6; 111.1; Ascon. p. 24.12 Clark: merita viri auctoritate and especially p. 76.21: Inducor magis librariorum hoc loco esse mendam quam ut Ciceronem parum proprio verbo usum esse credam. ⁴³ Tac. Dial. 26.8. Cf. the ironical advice in Lucian. Rh. pr. 17 to read declamation rather than Isocrates, Demosthenes, and Plato. Conversely the devotees of archaic writers rhetorum nos trorum commentarios fastidiunt (Tac. Dial. 23.2). ⁴⁴ Lucan. 7.62 3. ⁴⁵ For Quintilian Cicero is just divine (e.g. 1.6.18: divine ut omnia). Cf. also e.g. Nep. fr. 58 Marshall (Locuples ac divina natura); Vell. 2.66.5; V. Max. 2.2.3; Plin. Nat. 7.117; Fronto, .100 and .142 Haines [= 180.8 and 124.21 van den Hout²] (he uses Tullianus as an adjective of com mendation, e.g. .122 [= 20.6]). ⁴⁶ Quint. 10.1.112. ⁴⁷ Cf. Gell. 17.13.2. For the extreme archaist position see Quint. 8.5.33; SHA Hadr. 16.6 (cf. Tac. Dial. 23.2). ⁴⁸ Gell. 10.3.

70

Papers on Quintilian and Ancient Declamation

interesting reasons, to which I shall return. Seneca, to the dismay of Gellius,⁴⁹ waxed sarcastic about Ennian elements that he detected in Cicero; but the context is uncertain. The tone hardly seems very serious, and the criticism is perhaps rather of Ennius than of Cicero. In any case, Seneca elsewhere shows no hostility to Cicero’s style,⁵⁰ so different from his own. He found Cicero’s ordering of words sine infamia mollis,⁵¹ just as it was devexa et molliter detinens.⁵² Seneca was clearly not one of those who found fault with Cicero’s rhythms, which for some made him exultantem ac paene . . . viro molliorem;⁵³ naturally he was not—another point to which I shall return. Cicero, in fact, remained, as Plutarch interestingly says, consistently reputable, καίπερ οὐ μικρᾶς γεγενημένης περὶ τοὺς λόγους καινοτομίας.⁵⁴ By the end of the century, Quintilian and his pupil Pliny were confessedly carrying the flag for him, and the criticisms of Aper⁵⁵ in the Dialogus have something of the same dialectical purpose that we found in the historical criticism of Cicero: they are there to give weight and point to the praise of Messalla. Plures hodie reperies qui Ciceronis gloriam quam qui Vergili detrectent.⁵⁶ But there would not be many of either, in Aper’s day or in the earlier years of the century. Cicero, here too, was an exemplum: a monument of Roman culture comparable with Virgil.⁵⁷ Hae tibi erunt artes. But not only those. Here was something ‘to set alongside or even above’ insolens Graecia.⁵⁸ If we believe an anecdote in Plutarch, one Greek at least had seen such a rival on the horizon even in Cicero’s student days.⁵⁹ And, like Plutarch himself, Caecilius of Caleacte⁶⁰ and ‘Longinus’⁶¹ unbent far enough to perform συγκρίσεις of Cicero and Demosthenes.⁶² An exemplum, then, in literary as well as historical aspect. But on the literary side Cicero received at times more measured criticism than ever on the historical. Quintilian, even, true to Cicero’s own precepts, was still in search of the perfect orator. ‘Cicero’, he says,⁶³ ‘admittedly stood at the peak of eloquence. I can scarcely find anything lacking, though I could perhaps find something that he might have cut away: for that is the general view of scholars, ⁴⁹ Gell. 12.2.3 ff. ⁵⁰ Seneca joins the encomiasts at Ep. 40.11 (a quo Romana eloquentia exiluit; cf. Vell. 1.17.3: erupit); 107.10; 118.1. He liked Cicero’s philosophical style (Ep. 100.9). ⁵¹ Sen. Ep. 100.7. ⁵² Sen. Ep. 114.16. ⁵³ Quint. 12.10.12; cf. Tac. Dial. 18.4. ⁵⁴ Plut. Cic. 2.5. ⁵⁵ And even he agrees that Cicero Primus . . . excoluit orationem (Tac. Dial. 22.2; cf. Nep. fr. 58 Marshall: perpoliverit). ⁵⁶ Tac. Dial. 12.6. ⁵⁷ Mart. 5.56.5 and 11.48; also Tac. Dial. 12.6, just cited. ⁵⁸ Sen. Con. 1.pr.6. Cf. Suas. 7.10; Vell. 2.34.3 (also Nep. fr. 58 Marshall). With the related topic at Sen. Con. 1.pr.11 (illud . . . ingenium quod solum populus Romanus par imperio suo ha buit) compare Plin. Nat. 7.117 (ingenii/imperii). ⁵⁹ Plut. Cic. 4.6 7. ⁶⁰ Plut. Dem. 3.2. ⁶¹ Subl. 12.4. ⁶² Though ‘Longinus’ (Subl. 12.5) says that a Roman would do the job better. For a Roman who made the comparison, see Quint. 10.1.105 ff. (also 12.1.14 ff.), and cf. Gell. 15.28.6 7. ⁶³ Quint. 12.1.20.

Cicero and the Silver Age

71

that he had many virtues, but some faults.’ Quintilian thought that Cicero could have spoken better if he had lived longer, in a less troubled age.⁶⁴ But the point is that he was capable of criticizing the great man, however modestly⁶⁵ and hesitatingly. And he and others could do this, and do it with authority, because they had the sort of intimate knowledge of Cicero’s speeches that came from teaching them to the young. How was such teaching conducted? A good modern commentary on Cicero would not normally restrict itself to either a purely historical or a purely philological (including rhetorical) approach. But it would seem that ancient commentaries did specialize more rigidly.⁶⁶ Asconius’ excellent work is almost entirely taken up with historical problems, and his prefaces show a concern to set the speech in a historical context. On the other hand, the much later Bobbio scholiast⁶⁷ is predominately interested in the meanings of words and in rhetorical comment, some of it presupposing knowledge of the stasis system associated with Hermogenes. He does, at times, explicate matters of fact: but in much the same tone of voice as Servius uses to deal with historia in Virgil. And though we have Asconius’ commentary on the Pro Milone and much of the Bobbio scholiast on the same speech, their remarks hardly ever overlap. It is much to Asconius’ credit, however, that he is not unaware of the dangers of regarding a speech of Cicero as a historical document raising merely historical problems. In the introduction to the Corneliana, he remarks that ‘there is extant the speech of the accuser Cominius, which is worth picking up not only because of the speeches of Cicero we possess for Cornelius, but for its own sake’ (p. 61.23 Clark). And it may well be that he says this because he takes a point later made by Quintilian, that ‘it is very useful . . . to read whenever possible the speeches given on both sides . . . Even if they are not always equal in merit [to Cicero’s], they are rightly desiderated for anyone wishing to get to know the point at issue in the case’⁶⁸—and, one might add, to see when Cicero is distorting facts. More explicitly, Asconius is ready to distinguish between the mos historicus and the mos oratorius (p. 13.4 Clark): when Cicero says nobody, he does not necessarily mean quite that.⁶⁹ Nor is Asconius unaware that Cicero may be disingenuous. ‘The orator took refuge

⁶⁴ Quint. 12.1.20. ⁶⁵ Quint. 3.3.7: Quod . . . audacius dixerim and the like. See further Stroh (1975), 300 n. 8. ⁶⁶ Sen. Ep. 108.30 ff. is interesting on this topic; he contrasts the approaches of philologus, grammaticus, and philosophus to Cicero’s De republica. ⁶⁷ It would not be surprising if he drew on the work of much earlier commentators. But they are virtually unknown. Hier. Adv. Rufin. 1.16 mentions (cf. Ep. 70.2.1) Vulcatius as commenting on Cicero’s speeches in the same breath as Asper on Virgil and Sallust (late 2nd c. ) but also as praeceptoris mei Donati. Statilius Maximus seems to have been a lexi cographer rather than a commentator (Schanz Hosius Krüger (1922³), .164 5; again late 2nd c. ?). ⁶⁸ Quint. 10.1.22 3. ⁶⁹ Though see Wiseman (1979), 46 n. 26.

72

Papers on Quintilian and Ancient Declamation

in Metellus’ nobility and Curio’s energy in order to hide what they had done with regard rather to their advantage than to morality’ (p. 63.9 Clark). Again, ‘I do not want you to fail to observe that a skilful orator may, if need be, use the same facts from two angles’ (p. 70.13 Clark): and he proceeds to make an interesting distinction between the sort of things Cicero would say in a contio and in the senate.⁷⁰ Most impressively of all, he says that ‘a reading [of the Cornelian speech] will make clear the rhetorical art by which Cicero contrived at once to preserve the dignitas of his distinguished opponents and yet prevent his client from being harmed by their auctoritas: and with what moderation he dealt with a case so difficult in other ways’ (p. 61.7 Clark). It is not clear to me if Asconius’ commentary reflects his own teaching in a school.⁷¹ If it does, I doubt if there were many teachers like him in his enthusiasm for detailed historical research. But his intelligent awareness of the tricks of Cicero’s rhetorical trade is not unique. The Bobbio scholia are quite capable, even at their rather elementary level, of saying not only what Cicero means by something, but also why he says it in that way or in that place. And these scholia do, or may, reproduce the sort of thing a practising rhetor would have told pupils as they read through a speech of Cicero in class. The procedures employed in one class, that of Quintilian, can be recovered in some detail; and to Quintilian I now turn. In the school of the grammaticus, one traditionally read verse authors: Homer and Virgil are approved of by Quintilian,⁷² together with carefully chosen representatives of other genres. Only when pupils began at the rhetor’s school did they start reading prose. Others put minor writers, or more luxuriant styles, on the syllabus.⁷³ For Quintilian, only the best would do, et statim et semper, with a preference, in the early days, for the simpler writers: Livy, then, rather than the generally more distinguished Sallust. But above all Cicero,⁷⁴ and, as Livy once put it, anyone very like Cicero.⁷⁵ And there is a caveat that in effect expands on that last addition. The ancient orators, the Gracchi and their like, were to be avoided, and, no less, recentis huius lasciviae flosculi:⁷⁶ Seneca, without a doubt, and his like. These were to be reserved for a maturer pupil—the pupil, indeed, for whom the long reading list of 10.1.46 ff. is intended.⁷⁷

⁷⁰ Cf. Schol. Bob. on Mil. 9: Ἐναντία huic argumentatio est in illa oratione quae ‘Pro M. Tullio’ inscribitur. Ibi quippe, quoniam aliud praesentis negotii condicio poscebat . . . ⁷¹ He addresses his sons; but that may be a show, as with the Elder Seneca. ⁷² Quint. 1.8.5. ⁷³ Quint. 2.5.18. ⁷⁴ Cicero would not have been as offended as squeamish poets at being read in school: that was part of fame. Cf. Att. 2.1.3: ea quae nos scribimus adulescentulorum studiis excitati and esp. Q. fr. 3.1.11: praesertim cum . . . meam (i.e. orationem in Pisonem) . . . pueri omnes tamquam dictata perdiscant. See the excellent remarks of Stroh (1975), 52. ⁷⁵ Quint. 2.5.19 20. ⁷⁶ Quint. 2.5.21 2. ⁷⁷ So, of Seneca: iam robustis et severiore genere satis firmatis legendus (Quint. 10.1.131).

Cicero and the Silver Age

73

Quintilian’s recommendation of Cicero even for beginners comes as part of a chapter that had started with a general defence of the practice of careful reading of texts at the rhetor’s school: reading that he specifically compares with a grammarian’s enarratio of poetic texts.⁷⁸ It is clear that Quintilian’s practice was unusual, and not without critics: though he significantly remarks that the Greeks had pioneered it.⁷⁹ It is clear, too, that it was not to be an elementary matter if it was to be done at all; not merely a matter of explaining unusual words as they cropped up, but of pointing out ‘virtues and, if they occur, faults’.⁸⁰ The class would take it in turns to read aloud (notice the oral nature of all this), and the master would, after explaining the case, leave no stone unturned, either in inventio or elocutio.⁸¹ In the following sections Quintilian amplifies those two headings. What he says might serve as the blueprint for a rhetorical commentary on a speech of Cicero.⁸² But it also gives us the headings for the Institutio to come. As Quintilian took his pupils through the theoretical course on which his book is based, he would illustrate his teaching from the speeches which they were concurrently reading;⁸³ and in his enarratio of those speeches he would correspondingly draw on the concepts made familiar in his teaching. ‘Examples of everything that I teach’, he says in Book 10, ‘are to be drawn from reading, and they are far more effective even than textbooks . . . for what the teacher recommended is shown in operation by the orator.’⁸⁴ We glean more detail of the process of careful reading from 10.1. At least at the early stage, one should read almost as carefully as one writes, going over a speech twice so that an orator’s subtle preparations can be properly appreciated.⁸⁵ A speech even of Cicero should not be thought beyond criticism: even Homer nods.⁸⁶ And a speech should be seen in context, compared with the speeches of the opposing counsel or with others on the same theme.⁸⁷ Quintilian, then, brought high standards to the treatment of a speech. He was not, of course, a pioneer. Cicero’s orations had been much trampled over, if only by pedants. And these Quintilian sometimes has in mind. He has no answer, or gives none, to those who found fault, on quasi-logical grounds, with the partitio of the Pro Cluentio.⁸⁸ But at another place he is provoked by such criticism to a memorable analysis. Many people thought frigid⁸⁹ the passage in the Milo (28) where the defendant is described as coming home, changing ⁷⁸ Quint. 2.5.1. ⁷⁹ Quint. 2.5.3. ⁸⁰ Quint. 2.5.5. ⁸¹ Quint. 2.5.6 7. ⁸² That Quintilian is thinking of him, and of Demosthenes, is shown by 2.5.16. ⁸³ Note Quint. 8.3.79: Cuius praeclara apud Vergilium multa reperio exempla, sed oratoriis potius utendum est and he proceeds to give one from Cicero (namque ad omnium ornandi virtutum exemplum vel unus sufficit: 8.3.66). Quintilian, so far as the figures were concerned, wished to transfer his material from the grammatical to the rhetorical mode. For the speeches of Cicero he knew best, and so probably taught most, see Stroh (1975), 271 n. 106 (with 301). ⁸⁴ Quint. 10.1.15. ⁸⁵ Quint. 10.1.20 1. Cf. Quint. 4.2.57. ⁸⁶ Quint. 10.1.24. ⁸⁷ Quint. 10.1.22 3. ⁸⁸ Quint. 4.5.11. ⁸⁹ Quint. 4.2.59.

74

Papers on Quintilian and Ancient Declamation

shoes and clothes, and waiting around a little for his wife to get ready: ut fit, says Cicero, winking at the married men in his audience. But for Quintilian— and rightly—the words have their role, in showing, unobtrusively and almost subliminally, how unhurried was Milo’s departure. Both the details of the scene and the everyday language in which it is described contribute to the effect. This is the way in which Cicero should be commented on; and it is sad that so few of the perceptions that Quintilian must have passed on to his pupils in class found a place in his book. Quintilian shows up well whether he is discussing a brief passage like this or a speech as a whole. He has a sharp eye for the consilium, the shape and strategy of an oration. His strength arises from a conviction, bred from personal experience in the courts, that rules are there to be broken, that τὸ πρέπον and the force of circumstance are what must dictate an orator’s line.⁹⁰ Hence a clear understanding of the point of the three quaestiones that precede the narration of the Pro Milone,⁹¹ or of the handling of Scamander in the Pro Cluentio.⁹² And his appreciation goes beyond the intellectual. If Cicero had such an effect on the judges in the Pro Cornelio, when he reduced them to a state of mental blindness, quo essent in loco ignaros,⁹³ then that was not just the result of reason and lucid argument: ‘It was sublimity, surely, and magnificence and brilliance and personal authority that brought on that uproar.’⁹⁴ I expect that we all sometimes find tedious the more technical books that are the heart of the Institutio. They had to be there; one needed to learn the rules and the terminology. But Quintilian is very far from being a mere labeller. He, like ‘Longinus’, sees that what matters is not knowing what a device is called but knowing what effect it has. When Cicero says: Sed earum rerum artificem quem?—Quemnam? Recte admones, Polyclitum esse dicebant,⁹⁵ we are told why Cicero talks like this: ‘he is making sure that when he is accusing Verres of being crazy for statues and pictures, he is not thought to be keen on such things himself.’⁹⁶ He remains thoroughly didactic. He constantly cites Arch. 19, the heightened passage Saxa atque solitudines voci respondent, bestiae saepe inmanes cantu flectuntur atque consistunt, to illustrate different points, quo sint magis familiaria.⁹⁷ And he knows the arteswriter’s trick of rewriting a passage in another form to show up the essence of a device under discussion.⁹⁸ But this didacticism is pointful. Quintilian keeps us aware that Cicero was not a stringer-together of miscellaneous devices, but an orator who had a client to satisfy, an opponent to out-manoeuvre, and an audience to persuade.

⁹⁰ See esp. Quint. 2.13 and 11.1 (also above, p. 72 with n. 70). Also Schol. Bob. on Mil. 31, quite in Quintilian’s spirit and comparable with Quint. 6.5.10. ⁹¹ Quint. 4.2.25. ⁹² Quint. 11.1.74. ⁹³ Quint. 8.3.4. ⁹⁴ Quint. 8.3.3. ⁹⁵ Cic. Verr. 2.4.5. ⁹⁶ Quint. 9.2.62. ⁹⁷ Quint. 9.4.44. ⁹⁸ e.g. Quint. 4.1.66 7.

Cicero and the Silver Age

75

The sort of passages of Quintilian to which I have been drawing attention are in effect literary criticism of Cicero. But such criticism was by no means an end in itself. It was part of the teaching of imitatio as one route, and an important one, to practical skill in oratory. Thus it is no coincidence that the long passage of 2.5 on what a careful reading will look for in a speech is closely paralleled by another in 10.2 on the kind of imitatio that goes beyond wording. Illuc intendenda mens, quantum fuerit illis viris—for it must be remembered that Quintilian has by no means made Cicero his sole exemplar—decoris in rebus atque personis, quod consilium, quae dispositio, quam omnia, etiam quae delectationi videantur data, ad victoriam spectent: quid agatur prohoemio, quae ratio et quam varia narrandi, quae vis probandi ac refellendi, quanta in adfectibus omnis generis movendis scientia, quamque laus ipsa popularis utilitatis gratia adsumpta, quae tum est pulcherrima cum sequitur, non cum arcessitur. Haec si perviderimus, tum vere imitabimur (27). But there was a stage between such analysis of the qualities of a Cicero or a Demosthenes and the production of a speech designed for the courts. This was the declamation. The master composed fair-copy speeches⁹⁹ to illustrate the doctrines he was expounding and to show how the techniques of the great orators could be taken over. The pupil composed his own declamations to practise in the safety of the schoolroom what he would eventually have to do in the court. And both master and pupil, with more or less sophistication, would employ imitatio in declamation. Thus, as in the sphere of reading and precept, the speeches of Cicero had a role to play in declamation. That was true later (and no doubt earlier also) of Demosthenes. The collection of Sopatros, that forms the bulk of the eighth volume of Walz’s Rhetores Graeci, shows the rhetor telling his pupils to take ideas and wording from the great orator. One example suffices: at p. 11.22 Walz [= 1.7.17 Weissenberger] Sopatros remarks that ‘it is possible at once to tack on τὸ Δημοσθενικόν, that we are born not only for our parents but also for the city’, alluding to a passage of the De corona (205) which is regarded as so familiar that it is not cited more fully. This sort of thing is an indication, I take it, of class reading of Demosthenes parallel with the composition of declamations. The two were brought together by the sporadic use of declamation themes that actually involved Demosthenes and his times. In one of Sopatros’ collection, we are asked to suppose that ‘When money was disappearing from the acropolis, Demosthenes was found writing a speech in defence of sacrilege, Aeschines burying money in a solitary place; and they accuse each other’ (p. 19.16 W. [= 4.th.1 We.]). Here a declaimer actually impersonated Demosthenes on the one side, Aeschines on the other. And marks would clearly be given for ingenuity of pastiche.

⁹⁹ Quint. 2.5.16.

76

Papers on Quintilian and Ancient Declamation

Whether this sort of thing goes back to the very beginning of declamation, which started, surely, in Greece and Asia Minor not long after Demosthenes’ death, or whether it is the product of the renewed interest in the great orators evidenced in the Atticist movement, is uncertain. I should guess the former; Romans rarely innovated in this kind of field, and we find a parallel in Latin declamation of the Augustan period. I think that in the collection of the Elder Seneca use of Cicero is virtually restricted to declamations actually concerned with Cicero. There are three of these, all set during the last days of the orator and his struggle with Antony. And in all three, but especially Con. 7.2 and Suas. 6, Ciceronian pastiche is rampant. The character of declamation becomes very clear. It is not desired that such themes should turn a declaimer to careful research into the historical background, for deliberately fictitious circumstances are posited: it was not believed that Antony bargained for the burning of Cicero’s books,¹⁰⁰ or that Cicero was in a position to beg Antony’s pardon, or, yet,¹⁰¹ that Popilius killed Cicero despite having earlier been defended by him on a charge of parricide. What mattered was to strike Ciceronian poses and make coy, or clumsy, allusion to the great man’s words. It is inoffensive that Haterius should say: Proposito in rostris capite Ciceronis, quamvis omnia metu tenerentur, gemitus tamen populi liber fuit.¹⁰² We may, but do not need to, remember that Cicero had used these last words in the Second Philippic, when he was making much of the deplorable behaviour of Antony in putting the property of Pompey up for sale (64). There is a sort of aemulatio here. Haterius¹⁰³ wishes us to take the point that the groans were far more justly uttered, and thus the words more aptly employed, at the death of Cicero. One is less happy with the idea of Latro: ‘Sulla’s thirst for citizen blood has returned to the state; at the triumviral auctions the deaths of Romans are put up for sale like revenues. One single noticeboard surpasses the disaster of Pharsalus, of Munda, of Mutina. The heads of former consuls are weighed out for gold. Tuis verbis, Cicero, utendum est: ‘O tempora, o mores!’¹⁰⁴ The quotation is too studied, the bathos¹⁰⁵ too insistent. No better when Argentarius describes Antony’s debaucheries, with the comment: Iam ad ista non satis est dicere: ‘Hominem nequam!’¹⁰⁶ Not every quotation from Cicero can turn into a good epigram. This sort of theme continued: for Quintilian knows of the two suasoriae on Cicero.¹⁰⁷ And it is not unreasonable to suppose that if the themes were still ¹⁰⁰ Sen. Suas. 7. ¹⁰¹ This became ‘fact’ (see Cic. Orat. dep.  Schoell [= 81 Crawford]) because the declaim ers parroted it, despite Sen. Con. 7.2.8 (declamatoribus placuit parricidi reum fuisse). ¹⁰² Sen. Con. 7.2.5. ¹⁰³ Compare too Pompeius Silo in Suas. 6.4. ¹⁰⁴ Sen. Suas. 6.3; cf. Cic. Catil. 1.2. ¹⁰⁵ Felt by antiquity no less than by me: cf. Mart. 9.70. ¹⁰⁶ Sen. Suas. 6.7; cf. Cic. Phil. 2.77. ¹⁰⁷ Quint. 3.8.46. Mart. 3.66 and 5.69 look to be from the same stable (Zieliński (1908²), 345).

Cicero and the Silver Age

77

used, it was at least partly because people still wanted to match themselves against Cicero. But it is interesting that when we come to the Minor Declamations, the work if not of the Ciceronian Quintilian then at least of his school, only a very restrained use is made of Ciceronian tags. It is true that no declamation is preserved in this corpus that exploits the events of Cicero’s life. But one senses in these restrained and purposeful speeches a desire not to cheapen the Master by extremes of parody. In the very last Declamation (388), the sermo remarks¹⁰⁸ that a parallel for an attack on a mother in court can be found in Cicero’s Pro Cluentio, from which a short extract is quoted.¹⁰⁹ It will be significant that in the same declamation the argument Si doceo non perisse, nimirum raptus est; si raptum ostendo, doceo etiam vivere¹¹⁰ recalls the form of Clu. 64.¹¹¹ And this restrained form of allusion is the norm elsewhere. Some of the resemblances are very close indeed. Thus in Decl. min. 259.12 Intellego, iudices, quam difficili ac velut scopuloso loco versetur oratio mea comes straight from Div. Caec. 36; but the allusion makes no particular point, and is not heavily insisted upon. Elsewhere, aemulatio is at work. The passage on the punishment of parricides in S. Rosc. 72, of which Cicero later became halfashamed,¹¹² is varied, though by no means beyond recognition, in Decl. min. 299.4. But in general it must be repeated that Cicero is not much employed in the Minor Declamations, less, I should judge, than Demosthenes in Sopatros. If these are in any sense the work of Quintilian, they do not suggest that the master encouraged any slavish imitation of Cicero. That is true of style as well as content, and it conforms with what one should deduce from the Institutio about the nature of Quintilian’s Ciceronian stance.¹¹³ For him,¹¹⁴ Cicero was not, as we have seen, a unique exemplar: merely the nearest that a Roman had come to the ideal of the perfect orator. It was the spirit of Cicero’s speeches, and of his rhetorical works, that most mattered. A superficial Ciceronianus might persuade himself that an over-use of esse videatur was the key to success.¹¹⁵ Quintilian knew that style, even regarded less frivolously than that, was less important than a basic seriousness of approach. When Quintilian set himself in opposition to what one might call the ‘naturalists’, who thought that being born was enough to make one an orator,¹¹⁶ he was being truly Ciceronian. For him, as for Cicero, oratory was a difficult art, to be learned slowly and carefully and with reverence. It was not

¹⁰⁸ Decl. min. 388.32. ¹⁰⁹ Cic. Clu. 12: it is natural to connect this with the passage in 11.1.61 ff. where Quintilian discusses Cicero’s tactics in just this section. ¹¹⁰ Decl. min. 388.11. ¹¹¹ Again cited in the Institutio: 5.10.68. Reminiscence of this passage could explain the oth erwise mysterious iudicium at Decl. min. 388.29. ¹¹² Cic. Orat. 107; again mentioned in the Institutio: 12.6.4. ¹¹³ See Winterbottom (1975) [= A.3 above]. ¹¹⁴ Quint. 10.2.25. ¹¹⁵ Quint. 10.2.18. ¹¹⁶ Quint. 11.3.11.

78

Papers on Quintilian and Ancient Declamation

just a narrow technical matter, either, that could be picked up from a handbook. It called for the whole man, devoting himself to a wide range of learned activity, not necessarily or only because learning paid dividends in the court, but because an orator was to be more than a hack. And here, of course, Cicero was model as well as preceptor. I doubt if his philosophy ever helped him to win a case. But it made him an orator worth the study of posterity. It is in that spirit that the compiler of the Minor Declamations finds room for themes involving Cynicism (283) and the relative merits of oratory, philosophy, and medicine (268). Oratory is, or should be, more than rhetoric. Where the emphasis does change, between Cicero and Quintilian, is in their attitude to declamation. It was not that declamation somehow became more important in the course of the first century . Our impression that it does is largely a delusion, resulting from the accidents of our evidence. Declamation will have come to Rome with the Greek teachers who brought rhetoric there in the second century . Cicero certainly trained in it, both in Rome and in the East, and continued to practise it later in life. If we form the impression that it was somehow in abeyance during the late Republic, that is due to sleight of hand in Cicero’s maturer rhetorical works. He takes over the details of the technical Greek rhetoric, and with them, naturally, the stasis lore which is intimately wedded to declamation; but he is concerned at once to widen it and to make it relevant to the practical needs of Roman youths in a way that it originally was not. Hence, on the one hand, his emphasis on the importance of philosophy as a close ally of rhetoric, and on the other the impression he gives that the rhetorical training leads straight to the forum with little delay in the schoolroom. Declamation rears its head only very occasionally.¹¹⁷ Cicero perhaps saw that it was liable to become an end in itself;¹¹⁸ perhaps he felt uncomfortable about it, even defensive. It was a boyish pastime,¹¹⁹ beneath the consideration of the grave debaters of the De oratore. Cicero sweeps declamation under the carpet. Quintilian, a practising teacher, could not afford to do that, nor did he really wish to. He saw uses in declamation, and contrived to give it a place in his scheme without abandoning the Ciceronian emphasis on the practical nature of his training, and on the need for a wider outlook than the ordinary rhetor fostered. The passages (esp. 2.10) where he assesses the value of declamation are perhaps familiar enough. Less obviously, he manages to give precepts for declamation intermingled with precepts for real-life oratory: the two merge into each other in the Institutio.¹²⁰ To give a single example: the reading list ¹¹⁷ e.g. Cic. De orat. 2.100. ¹¹⁸ Cic. De orat. 1.149. ¹¹⁹ Cic. De orat. 1.244: pueri apud magistros. ¹²⁰ Indeed, Quint. 2.10.1 seems to say that after the progymnasmata comes declamation and that that is the subject of the rest of the Institutio: suasorias iudicialesque materias: quarum antequam viam ingredior . . . Where does he do that if not in Books 3 12? Generally, see Win terbottom (1983b) [= A.8 below].

Cicero and the Silver Age

79

of 10.1.46 ff. is meant for the mature student who wishes to acquire the final hexis in practical oratory. It is from this point of view that Quintilian praises Menander, qui vel unus . . . diligenter lectus ad cuncta quae praecipimus effingenda sufficiat:¹²¹ his plays are omnibus oratoris numeris . . . absolutae.¹²² But then, taking a personal line (Ego tamen), he remarks that Menander will contribute even more to declaimers, because they have to impersonate different sorts of character: In quibus omnibus mire custoditur ab hoc poeta decor.¹²³ While holding firmly to his doctrine that declamation makes sense only as an imitation of and preparation for the courts, Quintilian gives it its due place in his book. And it is very likely that the Minor Declamations show us how he made it train his pupils in a sober and well argued eloquence. I have argued that declamation did not increase in importance in the first century: it merely remained important, and perhaps became, in schools less austere than Quintilian’s, more extravagant in conception. But clearly something did happen to oratory after Cicero. We have seen that Plutarch remarked on the persistence of Cicero’s fame despite the stylistic innovations of the intervening period. Tacitus reports a common view that Cassius Severus, under Augustus, somehow marked a turning point: quem primum adfirmant flexisse ab illa vetere atque derecta dicendi via.¹²⁴ We are much at the mercy of other people’s impressions, and cannot check them for ourselves. It would be perverse to claim that, just as Cicero continued to be highly esteemed as an orator throughout the century, so he continued to mould the style of contemporary oratory.¹²⁵ The language had moved on, for one thing. As early as the Augustan period, a declaimer could attract attention by using quaedam antiqua et a Cicerone dicta, a ceteris deinde deserta.¹²⁶ And, so far as style went, to admire

¹²¹ Quint. 10.1.69. ¹²² Quint. 10.1.70. ¹²³ Quint. 10.1.71. ¹²⁴ Tac. Dial. 19.1. The same sort of thing was said of Demetrius of Phaleron (Quint. 10.1.80, from Cic. Brut. 38); one can imagine Quintilian drawing the parallel in his De causis corruptae eloquentiae. Dr D. C. Innes suggests to me that the comparison of Demosthenes and Cicero (above, p. 70 n. 62) is relevant here: the greatest orators of Greece and Rome, at the end of their line. ¹²⁵ Nor should we underestimate the sort of changes in legal procedure and audience expec tation mentioned in Tac. Dial. 19.5 and 39.1 3. ¹²⁶ Sen. Con. 4.pr.9. A fortiori later; thus Sen. Ep. 108.32: ea quae consuetudo saeculi mutavit, tamquam ait Cicero . . . ‘calce’ . . . Hanc quam nunc in circo ‘cretam’ vocamus, ‘calcem’ antiqui dicebant. In the Dialogus Aper thought that Cicero had actually gone out of his way to ‘imitate’ really antique orators like Galba (18.1; cf. Quint. 10.1.40), and remarks on the vitia antiquitatis that marred his earlier orations (22.3); but that was rather a matter of style. It may be observed that Fronto seems a little disappointed with the choiceness of Cicero’s vocabulary; see .4 Haines [= 57.8 van den Hout²]: is mihi videtur a quaerendis scrupulosius verbis procul afuisse; .6 [= 57.16]: paucissima admodum reperias insperata atque inopinata verba, quae non nisi cum studio atque cura atque vigilia atque multa veterum carminum memoria indagantur (with ve terum carminum cf. the fragment of Sen. Ep. in Gell. 12.2.3 ff., esp. 6: Aput ipsum quoque . . . Ciceronem invenies etiam in prosa oratione quaedam ex quibus intellegas illum non perdidisse operam quod Ennium legit; see also above, p. 70); similarly Aper’s protest in Dial. 20.5 against poeticus decor . . . Acci aut Pacuvi veterno inquinatus. That Cicero cited such poets is observed by

80

Papers on Quintilian and Ancient Declamation

and emulate Cicero was one thing, to reproduce him wholesale quite another. Pliny the Younger was an admirer: Marci nostri.¹²⁷ And he was an avowed emulator,¹²⁸ who despaired nevertheless of getting to the great man’s level.¹²⁹ But when it came to producing a speech of combative nature,¹³⁰ Pliny went for his figurae to Demosthenes and Calvus, masters, he modestly adds, of the vis which he was himself incapable of; Cicero’s λήκυθοι were not neglected,¹³¹ but they were clearly not of primary importance.¹³² Pliny, in fact—and we should remember that he was Quintilian’s pupil—was not a real Ciceronianus: a tribe, in any case, that laid itself open to ridicule.¹³³ The very fact that Quintilian draws attention to them suggests that most people were not conscious imitators of Cicero. But it would perhaps be wrong to suppose that an unbridgeable gulf separated Cicero from the orators of the first century . Contemporaries naturally stressed the novelties of those who followed Cassius Severus, and the Dialogus is founded on conscious and generally accepted perception of a disparity in quality and manner between ancients (including Cicero) and moderns. But enough is said by the writers of the Silver Age to suggest the senses in which we could speak of continuity between Cicero and his successors. I have already touched on Cicero’s quarrel with the Atticists. He thoroughly abuses, and doubtless in part misrepresents, these audacious rivals. They were, he asserts, altogether too devoted followers of Lysias, forgetful of the wide range of Attic oratory, which found room for Demosthenes as well as sparer talents. In the Tusculans Cicero is confident that the Attici have been defeated: iam conticuerunt paene ab ipso foro inrisi.¹³⁴ He would have regarded this as the triumph of his own Demosthenic oratory. His despised critics might have called it, rather, the triumph of Asianism. And we may consider three of the aspects of Cicero’s Asianism—I use the term with due consciousness of the pitfalls of this vocabulary¹³⁵—which link him with the age to come. There is, first of all, rhythm. When, later, the Greek rhetores were converted wholesale to Atticism, they seem at first to have laid aside the Hellenistic rhythms which had marked the heyday of the older, corrupt rhetoric. But in the end they relapsed into an accentual rhythm that is at least the heir, and is

Quint. 1.8.11, but that is a different matter. See on these passages Throop (1913), 39 (his article does little more than assemble material); cf. Gell. 12.21.22: cum (sc. Cicero) insolentias verborum a veteribus dictorum plerumque respueret . . . For Fronto excerpting si quid eleganti aut verbo notabili dictum videretur in Cicero’s letters, see .158 Haines [= 104.8 van den Hout²]. ¹²⁷ Plin. Ep. 1.2.4. ¹²⁸ Plin. Ep. 1.5.11 12; cf. Mart. 10.19.14 17. ¹²⁹ Plin. Ep. 4.8.4 5. ¹³⁰ Plin. Ep. 1.2.3: in contentione dicendi. ¹³¹ Plin. Ep. 1.2.4. ¹³² Still, Cicero was also a source for ‘daring’ oratory (Ep. 9.26.8). ¹³³ Quint. 10.2.18; Tac. Dial. 23.1. ¹³⁴ Cic. Tusc. 2.3. ¹³⁵ I speak of things that could be paralleled from the most sure source of Asian oratory, the Greek extracts in the Elder Seneca [see A.10 below, pp. 146 50].

Cicero and the Silver Age

81

perhaps the progeny, of the old metrical system. Thus a Sopatros, for all his enthusiasm for Demosthenes, makes sure that two or four unaccented syllables separate the accented ones at the cadence of his sentences. A similar fascination attended the clausula system in Latin. It doubtless arrived, with the rest of rhetoric, in the second century . Cicero imbibed it with his mother’s milk. And despite his professed reaction against the rhetores and despite his admiration of Demosthenes, he never saw fit to alter his ways. His critics were correct to fasten upon this point. His rhythm did separate him from the Attic orators and align him with Greek declaimers. This is what was meant by the taunt that he was in compositione fractum, exultantem ac paene . . . viro molliorem.¹³⁶ Quintilian himself saw that they were right, though he argued that Cicero was justified: nec vitium duxerim si Cicero a Demosthene paulum in hac parte descivit¹³⁷—for Latin as a language lacks the inherent Greek venustas et gratia and can reasonably be given extraneous ornaments like rhythm that Greek, he implies, did not need.¹³⁸ The elaborate argument of the long last part of the Orator shows Cicero, too, making the best of an awkward position. And of course it all went on.¹³⁹ The orators will have used rhythm in court. Seneca in his philosophy, Pliny in his Letters and his Panegyricus, Quintilian in his rhetorical handbook, all used it, more insistently than Cicero, perhaps, but on recognizably the same system. Even Gellius, for all his archaizing enthusiasms, and though he does not himself employ rhythm, regards Cicero’s numeri as a virtue. A sentence in the Pro Plancio gives a crispum . . . agmen orationis rotundumque ac modulo ipso numerorum venustum.¹⁴⁰ This, it is true, is put into the mouth of the rhetor Antonius Iulianus. But it is Gellius in propria persona who shows detailed appreciation of Cicero’s practice in compositio and speaks of his modulamenta orationis.¹⁴¹ Cicero’s rhythm, in fact, enabled him to sound familiar to readers a century later in a way that the Elder Cato could never have done. Secondly, epigrams. Aper in the Dialogus, giving the credit side, as he saw it, of Cicero’s oratory, remarks that he quasdam sententias invenit, utique in iis orationibus quas senior iam et iuxta finem vitae composuit, id est postquam magis profecerat.¹⁴² There is, doubtless, special pleading here. Güngerich’s new commentary remarks: ‘überzeugt jede Lektüre von Ciceros Reden . . . dass er dieses Kunstmittel noch nicht gesucht hat, wie es ja auch in der Theorie erst in der Kaiserzeit bei Seneca Rhetor und Quintilian behandelt wird.’ And, as Güngerich points out, even Quintilian thinks that Cicero could have managed ¹³⁶ Quint. 12.10.12. ¹³⁷ Quint. 9.4.146. ¹³⁸ Quint. 9.4.145. Another line of defence was that Demosthenes was rhythmical, but on a different system: Cic. Orat. 234 (cf. Quint. 12.10.26). ¹³⁹ And when Quintilian argues against detractors of rhythm (e.g. 9.4.53; 57; 64) he is fighting Cicero’s battle over again, not a contemporary one. I am not sure that 9.4.1 implies otherwise. ¹⁴⁰ Gell. 1.4.4. ¹⁴¹ Gell. 1.7.19. ¹⁴² Tac. Dial. 22.2.

82

Papers on Quintilian and Ancient Declamation

more epigrams.¹⁴³ But Güngerich might have added that Quintilian, while denying sententiae to ‘the ancients and particularly the Greeks’, did find them in Cicero;¹⁴⁴ and that when he wishes to illustrate the use of sententiae in narrative, he finds two examples, one brief, the other more expansive, in speeches of Cicero,¹⁴⁵ neither particularly late. The Sulla is not a late speech, but the Bobbio scholiast finds a sententia in § 31: though it is true that he remarks: quod genus in Tullianis orationibus rarum est.¹⁴⁶ Again, when Cicero himself has to give an example of the unregenerate style of his youth, he produces from the Pro Cluentio (199) uxor generi, noverca filii, filiae paelex,¹⁴⁷ which, whether sententia or no, has the true ring of declamation: compare Sen. Con. 6.6: Generi adultera, filiae paelex, and especially 9.6.1: Nefaria mulier, filiae quoque noverca. I take it that we have here another symptom of Asianic rhetoric.¹⁴⁸ Sententiae appear in the Greek extracts in the Elder Seneca as freely as in the (derivative) Latin. And I very much doubt if Philostratus is right to say that the late first-century  Assyrian sophist Isaeus was the first to sum up every argument ἐς βραχύ.¹⁴⁹ The Silver Age, again, merely expands on something that Cicero had used with restraint; he learned it, perhaps, from his Asiatic preceptors, but brought to it a moderation they did not practise. And sententiae, it may be added, are closely associated with that enthusiasm for figures of speech and punning that people noticed in Cicero.¹⁵⁰ Seneca the Elder discusses the relationship between Publilius and Cicero;¹⁵¹ even Trimalchio seems to have wind of the topic.¹⁵² As for the word-play of figurae verborum,¹⁵³ it will be this that made Plutarch speak of Cicero ‘as striving with the sophists Isocrates and Anaximenes’.¹⁵⁴ Quintilian, too, finds that mention of Isocrates and Gorgianic figures brings Cicero to his mind: Delectatus est his etiam M. Tullius, verum et modum adhibuit non ingratae nisi copia redundet voluptati, et rem alioqui levem sententiarum pondere implevit.¹⁵⁵ That would be a matter of taste. For the Atticists, this was one of the things that made Cicero redundantem et in repetitionibus nimium.¹⁵⁶ It was the very life-blood of Asianism, and of the Silver Age. ¹⁴³ Quint. 12.10.46. ¹⁴⁴ Quint. 12.10.48. ¹⁴⁵ Quint. 4.2.121. ¹⁴⁶ Gudeman (1914²), 237 gives a number of examples, mostly from early speeches. And if Cicero’s Asianism did lessen after he visited Asia, as he argues in the Brutus, that is what we should expect (note the first type of Asian style, sententiis . . . concinnis et venustis: Brut. 325). But the matter is not amenable to statistics. See also Norden (1898), 232 n. 1. ¹⁴⁷ Cic. Orat. 107. ¹⁴⁸ See n. 146 above. ¹⁴⁹ Philostr. VS 514: πᾶσαν ὑπόθεσιν συνελεῖν ἐς βραχὺ Ἰσαίου εὕρημα (if this does in fact allude to epiphonematic epigram). ¹⁵⁰ Cicero remarks on his own enthusiasm for antithesis in the Gorgianic tradition at Orat. 167 (cf. 165). ¹⁵¹ Sen. Con. 7.3.9. ¹⁵² Petr. 55.5. ¹⁵³ It decreased in Cicero with time: Norden (1898), 225 ff.; cf. Davies (1968). ¹⁵⁴ Plut. Comp. Dem. et Cic. 2.2. ¹⁵⁵ Quint. 9.3.74. ¹⁵⁶ Quint. 12.10.12. Demosthenes (ἥκιστα . . . ἐπιδεικτικός: ‘Longin.’ Subl. 34.3) was praised for his figurae sententiarum: Cic. Brut. 141; Orat. 136 (cited in Quint. 9.1.40). In the latter passage,

Cicero and the Silver Age

83

I come to the third aspect of continuity between Cicero and the Silver Age. I start not from a criticism known to have been made of him by the Atticists, but from a quality of his own that Cicero prided himself upon in explicit contrast with the Atticists. In the Orator he stresses that the perfect orator must master all three styles, and most especially the grand. His opponents, who restrict themselves to the plain, are inevitably disqualified from perfection. And their pretensions to being imitators of the Attic orators are lopsided too. Lysias, their hero, may be a model in the plain style; the far more distinguished Demosthenes is a model for all three.¹⁵⁷ Yet, for Cicero, even Demosthenes non semper implet auris meas.¹⁵⁸ What did Demosthenes lack? Cicero only says that his ears saepe aliquid inmensum infinitumque desiderant; and he is no doubt contrasting with Demosthenes’ tautness the sort of qualities in himself that made Quintilian remark that illic nihil detrahi potest, hic nihil adici.¹⁵⁹ But gaps in Demosthenes’ armour were specified by Quintilian: ‘we are superior (i.e. Cicero was superior to Demosthenes) in two things of the greatest importance in the arousing of emotion, wit and commiseratio.’¹⁶⁰ Quintilian suggests here and elsewhere¹⁶¹ that Demosthenes’ failings in the arousal of pity¹⁶² had an external cause: a law that banned the practice in Athenian courts. Quintilian seems to be wrong over the fact; at least only the Areopagus, in trials for murder, appears to have had such a law. But murder trials are after all very important. And the main point is that Demosthenes could be thought wanting in this field. It was not, of course, that Demosthenes was incapable of arousing emotion in general.¹⁶³ When Cicero describes the effect of grand style oratory,¹⁶⁴ he is saying much the same as Dionysius says when he describes the effect of a speech of Demosthenes (Dem. 22; cf. ‘Longin.’ Subl. 34.4). That Demosthenes had vis is stressed.¹⁶⁵ Pliny¹⁶⁶ looked to him as well as to Cicero for sublimitas; and of course he was a prime exemplar for ‘Longinus’. It is specifically in the Cicero agrees they are maiora. Pliny (Ep. 1.2.2) looked for figurae orationis in Demosthenes (and Calvus): meaning by this, as Dr Innes suggests to me, figures generally, not just verbal ones. ¹⁵⁷ Cic. Orat. 75 ff. ¹⁵⁸ Cic. Orat. 104. [But see below, p. 88.] ¹⁵⁹ Quint. 10.1.106. Cf. for Cicero’s fullness 6.3.5; 12.1.20; 12.10.52 (again contrasted with Demosthenes). Contrast the account of Calvus, imitator Atticorum: fecit . . . illi properata mors iniuriam si quid adiecturus sibi, non si quid detracturus fuit (10.1.115). ¹⁶⁰ Quint. 10.1.107. ¹⁶¹ Quint. 12.10.26; cf. 2.16.4; 6.1.7. ¹⁶² Cic. Brut. 290 might be pressed to say that Demosthenes could arouse tears. If anything, his forte was to ridicule opponents who indulged in pathos (Stevens (1944), 14 with n. 53). Note further ‘Longin.’ Subl. 34.2 3: Hyperides was οἰκτίσασθαι προσφυέστατος, while Demosthenes was τῶν προειρημένων (including οἶκτος) κατὰ τὸ πλέον ἄμοιρος. And of course Hyperides was not the only Attic orator who employed this technique (see Dover (1974), 195 201). Thrasy machus and Aristotle, in their different ways, paid theoretical attention to the topic. ¹⁶³ Cic. Orat. 26 and 133; Quint. 6.2.24 (δείνωσις); 12.10.23; Plut. Comp. Dem. et Cic. 1.2: ἐνεργείᾳ . . . καὶ δεινότητι. ¹⁶⁴ Cic. Orat. 97 9; Brut. 290; cf. Quint. 12.10.62. ¹⁶⁵ e.g. Cic. De orat. 3.28. ¹⁶⁶ Plin. Ep. 9.26.8.

84

Papers on Quintilian and Ancient Declamation

arousal of pity that Quintilian found him wanting. Cicero, on the other hand, so excelled here that the last speech in a trial was habitually reserved for him.¹⁶⁷ It seems possible that Cicero, in exploiting the possibilities of pathos, is the heir to Hellenistic oratory. It is natural to think in this context of the extravagances of the historians criticized by Polybius. Phylarchus Σπουδάζων . . . εἰς ἔλεον ἐκκαλεῖσθαι τοὺς ἀναγινώσκοντας καὶ συμπαθεῖς ποιεῖν τοῖς λεγομένοις, εἰσάγει περιπλοκὰς γυναικῶν καὶ κόμας διερριμμένας καὶ μαστῶν ἐκβολάς, πρὸς δὲ τούτοις δάκρυα καὶ θρήνους ἀνδρῶν καὶ γυναικῶν . . . ἀπαγομένων.¹⁶⁸ Polybius links this with the aims and effects of tragedy; but he might as easily have juxtaposed it with emotional oratory.¹⁶⁹ At least in Rome such displays of passion graced the law-courts; and the orator was taught at school ἀεὶ πρὸ ὀφθαλμῶν τιθέναι τὰ δεινά.¹⁷⁰ I do not think we can be sure what the practice of Hellenistic as opposed to classical Greek law-courts was;¹⁷¹ but it seems likely at least that the arousal of pity, amongst other emotions, was taught and practised in the declamation school. It was indeed particularly appropriate there, according to Quintilian: illic litigatores loquimur frequentius quam ut advocati: orbum agimus et naufragum et periclitantem, quorum induere personas quid attinet nisi adfectus adsumimus?¹⁷² And the appropriate emotion for the three characters chosen by Quintilian would surely be pity. In the first century , at least, we can be sure that emotions were of great importance in declamation: hence, amongst other things, the tendency for aequitas, which came towards the end and merged with the epilogue, to gain the upper hand over ius.¹⁷³ Whatever the Hellenistic background, Cicero was not the first Roman to exploit the appeal to pity. One thinks of Antonius’ account of his own successful defence of C. Norbanus, which fell into two parts, one involving commendatio, the other concitatio: the latter enabling the orator to say that he pro meo sodali . . . et pro mea omni fama prope fortunisque decernere . . .

¹⁶⁷ e.g. Cic. Orat. 130. ¹⁶⁸ Plb. 2.56.7. ¹⁶⁹ Quint. 6.1.30: producere ipsos qui periclitentur squalidos atque deformes et liberos eorum ac parentis institutum. Breasts were notoriously bared in a case conducted, significantly (see n. 162 above), by Hyperides: e.g. Quint. 2.15.9; the story is not necessarily true (Kowalski (1947)), but it could reflect later practice as well as the lubricious imagination of scholarly investigators. Generally note Cic. Brut. 43, historians writing rhetorice et tragice. ¹⁷⁰ Plb. 2.56.8. See Quint. 8.3.61 ff. on ἐνάργεια, esp. 62: oculis mentis ostendi (cf. Gell. 10.3.7: Quae totius rei sub oculos subiectio!; for the context see p. 85); 67: Sic et urbium captarum crescit miseratio (also 6.2.32 3); generally, Avenarius (1956), 130 40. Mr R. B. Rutherford, referring me to Nisbet and Hubbard on Hor. Carm. 2.1.17, remarks that Pollio interestingly combines the roles of orator, historian, and tragedian. ¹⁷¹ But the emotional passage of Hegesias translated in Rut. Lup. 1.7 seems to be forensic. ¹⁷² Quint. 6.2.36; cf. 6.1.25 6. ¹⁷³ See e.g. the sermo to ‘Quint.’ Decl. min. 270.5 (text uncertain), where the teacher argues for a careful treatment of the legal points before illa quae sola dicuntur. Emotional appeal was not enough: Nisi . . . etiam iure defenditur, verendum erit ne illum flentem (leg. flentes?) iudices dam nent (Decl. min. 270.1).

Cicero and the Silver Age

85

Petebam a iudicibus ut illud aetati meae, ut honoribus, ut rebus gestis, si iusto, si pio dolore me esse adfectum viderent, concederent.¹⁷⁴ I need not labour the examples from Cicero himself:¹⁷⁵ merely observing that the many heads for arousing pity detailed in the De inventione (1.106–9) and doubtless inherited from Hellenistic tradition¹⁷⁶ (indeed Cicero mentions the rhetor Apollonius: 1.109) can be abundantly illustrated from his own speeches. As for his successors, Quintilian himself was proud of his achievement in court in this area.¹⁷⁷ And Gellius’ comparative treatment of Cato, Gaius Gracchus, and Cicero is significant. Gracchus may be a fortis ac vehemens orator;¹⁷⁸ but in the passage cited from him there is nothing spoken ampliter insigniterque aut lacrimose atque miseranter;¹⁷⁹ whereas in the Verrines¹⁸⁰ Cicero’s emotional appeal is singled out for praise: Q̱uae . . . miseratio! Quae comploratio! Quae totius rei sub oculos subiectio! (Gell. 10.3.7; see p. 84 n. 170, and cf. 10.3.14: Haec M. Tullius atrociter graviter apte copioseque miseratus est). And pity was much in demand generally in the first century . Lucan and Seneca’s tragedies are evidence of the excesses pursuit of it could bring. All these ways in which Cicero foreshadows the Silver Age can be subsumed under one heading: voluptas. He gave audiences what they wanted;¹⁸¹ and they went on wanting it in the century that followed. There is nothing particularly disreputable about this. Cicero argues in the Brutus for the primacy of public approval.¹⁸² Still, critics might rebel against the lengths to which such pandering went. Quintilian thought voluptas was properly aimed at by Cicero:¹⁸³

¹⁷⁴ Cic. De orat. 2.200 1. Add the case of Servius Galba, miseratione sola . . . elapsum (Quint. 2.15.8). Galba was princeps ex Latinis (Cic. Brut. 82) to use miserationes. ¹⁷⁵ Or his contemporaries. Note e.g. Ascon. p. 20.18 Clark: Ipse quoque Scaurus dixit pro se ac magnopere iudices movit et squalore et lacrimis . . . Also the emotional (and Asianic) Hortensius (Div. Caec. 46: cum commiserari, conqueri . . . coeperit). ¹⁷⁶ I take it that the evidence adduced by Solmsen (1938), 394 6 shows that Hellenistic artes did not treat emotion in Aristotelian depth, not that they ignored it. ¹⁷⁷ Quint. 6.2.36. ¹⁷⁸ Gell. 10.3.1. ¹⁷⁹ Gell. 10.3.4. It is interesting that Gellius finds Cato more satisfactory than Gracchus in miseratio (10.3.15 ff.). Cato, he says, iam tum facere voluisse quod Cicero postea perfecit (16) a judgement to be compared with Cicero’s own discussion of Cato in the Brutus (esp. 65: omnes oratoriae virtutes in eis i.e. Cato’s speeches reperientur). ¹⁸⁰ Gellius quotes from Cic. Verr. 2.5.161 3 passages that in all Quintilian remarks on seven times, often in connection with their vividness (ἐνάργεια is illustrated from Verr. 2.5.6: 8.3.64) and power to arouse pity (esp. 4.2.114). Another famous passage (Verr. 2.5.118 19), seven times used by Quintilian, finds echoes not only in Sen. Con. 7.2.1 but even in Man. 5.621 ff. ¹⁸¹ Cic. Orat. 106. ¹⁸² Cic. Brut. 183 ff.; cf. Tusc. 2.3. ¹⁸³ Cf. Quint. 12.10.45: id fecisse M. Tullium video, ut cum omnia utilitati, tum partem quandam delectationi daret, cum et suam se rem agere diceret, agere autem maxime litigatoris: nam hoc ipso proderat, quod placebat. This is why Pliny looked to Cicero for amoenitas (Ep. 1.2.4; cf. Gell. 10.3.15), and why Fronto was so struck by his ornatus (.4 Haines [= 57.6 van den Hout²]). But it was possible to see excessive flores in Cicero (Quint. 12.10.13, correctly taken by Güngerich (1950), 246 7). Hence the feeling that he was more like an epideictic orator than a combative one, an Isocrates rather than a Demosthenes: see above, pp. 80 and 82, adding Plut.

86

Papers on Quintilian and Ancient Declamation

Cicero did well to spice even his argumentation with it;¹⁸⁴ and, faced with an audience of less than perfectly wise men, he was justified in giving pleasure to their ears.¹⁸⁵ But when it came to his own day, Quintilian felt that voluptas tended to be prava, and there was something sinister about tickling the ear with it.¹⁸⁶ It was all a question of degree, no doubt.¹⁸⁷ But Quintilian, like the rest of us, tended to enjoy the middle-aged reflection that things are steadily getting worse. Perhaps he would have been surprised, had he been miraculously transported to a court addressed by Cicero, to find how Silver the great orator really was.

DISCUSSION M. Calboli: Non ho trovato nella relazione molto bella del collega Winterbottom un riferimento a quella parte della pronuntiatio che è rappresentata dal De gestu, mentre il riferimento ad Antonio e alla sua difesa di Norbano mi ha fatto pensare ad un’altra difesa, quella che Antonio fece di Manio Aquilio; in essa infatti Antonio strappò la tunica di Aquilio, mostrando le cicatrici delle ferite ricevute da Aquilio per la patria (gesto o ispirato a quello di Iperide nel processo di Frine, o, come vuole U. W. Scholz nel suo lavoro su Antonio, a M. Servilio Gemino). Ora la mia domanda è: come vede il collega Winterbottom il ‘gesto’ nel ciceronianismo del tempo da lui trattato? Quanto poi è stato osservato sulle declamazioni mi suggerisce una domanda. Lei sostiene che le declamazioni sono entrate in Roma già dal 125 a.C.—e io sono d’accordo, solo vorrei spostare più in su tale data—e che non c’è differenza tra le declamazioni del periodo più antico e quelle del periodo augusteo, se non perché alcune erano in latino, altre in greco. C’è tuttavia una differenza che può essere causata anche da questo motivo, tra le declamazioni di cui troviamo tracce nella Rhetorica ad Herennium e nel De inventione e quelle di cui riporta brani Seneca il padre. I temi delle prime riguardano argomenti mitologici e della storia di Roma, i temi delle seconde argomenti inventati. A questo punto è naturalmente interessante il confronto con l’impiego di Menandro a cui Lei ha fatto riferimento, perché i temi delle declamazioni di Seneca, specificamente delle controversie, sono argomenti vicini alla commedia anche per il tipo di intreccio. Crede Lei che questo quadro si possa accettare?

Comp. Dem. et Cic. 1.4, where the Demosthenic manner is implicitly contrasted with Cicero’s ὡραϊσμὸς καὶ παιδιά (cf. the sophistic παίγνιον)[, and Tac. Dial. 20.6]. ¹⁸⁴ Quint. 5.14.35. ¹⁸⁵ Quint. 12.10.52 3. ¹⁸⁶ Quint. 2.12.6. ¹⁸⁷ Quint. 12.10.47.

Cicero and the Silver Age

87

M. Winterbottom: Antonius baring the scars of Aquilius is part of a long tradition of forensic miseratio (Quint. 2.15.7–9 himself juxtaposes the case of Phryne; cf. also Ov. Met. 13.262–5), which I conjectured to have been important in Hellenistic oratory. I do not know that Cicero ever behaved thus (observe the irony of Verr. 2.5.32), though he pointed to scars on Rabirius’ face (Rab. perd. 36; cf. fr. 35 Malcovati⁴ of Hortensius from the same case, cicatricum mearum, part of a prosopopoeia). For the later period, Quintilian does not seem critical of emotional displays in court (6.1.30–3; scars mentioned at 6.1.21). As to gestus more generally, he counsels restraint (note 11.3.123, more cautious even than Cicero). Practice must have varied greatly.—As to M. Calboli’s second point, I should be surprised if Greek rhetores in the Rome of the first century  were not using invented ‘Menandrian’ themes: they had surely been using them for centuries in Greece and Asia Minor. Cicero, in his De inventione, and the Auctor ad Herennium presumably avoided such themes (details in Bonner (1949), 23–8) because they thought them less suitable for their Roman readership. M. Leeman: At the beginning of your paper you announce that at the end you will try to gauge what sort of gulf separates Cicero from the Silver Age. At the end you conclude that Cicero himself is very ‘Silver’ already—a conclusion which sounds like a (carefully prepared!) ἀπροσδόκητον, if I may express myself paradoxically . . . Your main thesis is that there is much more continuity between Cicero and the Silver Age than is usually supposed. A disputatio in contrariam partem should certainly take into account Quintilian’s remarks about Seneca’s attitude to Cicero (Quint. 10.1.126), who is certainly implied in the classical authors quos ille non destiterat incessere, cum diversi sibi conscius generis placere se in dicendo posse quibus illi placent diffideret. There certainly is a gulf here! M. Winterbottom: M. Leeman is of course right that there is another side to the picture. Naturally Seneca’s philosophical style is different, and consciously different, from Cicero’s. But at 10.1.126 Quintilian seems much to exaggerate Seneca’s hostility to Cicero and to the potiores generally. M. Classen: When one compares the rhetorical handbooks on which Cicero was brought up (which must have been of the kind of the Rhetorica ad Herennium) and Quintilian’s Institutio oratoria, one is struck by the changes and refinements of the precepts as laid down by Quintilian. To what extent are these modifications due to Quintilian, and what evidence is there of earlier rhetorical analyses of Cicero’s speeches (of the type of which there are traces in the Scholia Bobiensia)? Does Quintilian owe any new idea or modification of the traditional theory to the interpretation of the speeches of an author other than Cicero? M. Winterbottom: Theory clearly moved on a good deal between Cicero’s youth and the Institutio. Quintilian 3.1.16–21 gives the names (I expect that the controversy of Apollodoreans and Theodoreans will have been fruitful in

88

Papers on Quintilian and Ancient Declamation

focusing attention on the need for flexibility in applying precepts). Quintilian exploited these theoretical advances, leavening them with his own practical experience and his study of Cicero’s speeches. As I remarked, there must have been earlier work on Cicero, but we have almost no direct evidence of it. So too for other orators. Quintilian himself can deploy details from Calvus, Asinius Pollio, and others, but I doubt if they much influenced his thinking. M. Stroh: Cicero Orat. 104 wird von den Kommentatoren regelmässig so verstanden, als wolle er hier einer gewissen Unzufriedenheit mit dem Stil des Demosthenes Ausdruck geben: . . . non semper implet aures meas (auch einen Mangel an ‘menschlicher Erfülltheit’ des Wortes wollte man schon finden). Aber man vergleiche den Kontext! Cicero nimmt auf den ‘Mangel’ des Demosthenes nur darum Bezug, weil er die Idealität des—real nie voll erreichbaren—orator perfectus darstellen will: ‘selbst’ Demosthenes (ipse!) bleibt bier natürlich noch zurück. Folglich ist Demosthenes hier gar nicht wegen seiner Mangelhaftigkeit genannt, sondern weil er das Äusserste an bisher überhaupt erreichter Vollkommenheit darstellt, weil er Cicero selbst jedenfalls übertrifft (so ausdrücklich § 105). M. Winterbottom: I am very grateful for M. Stroh’s remarks. They explain why what Cicero missed in Demosthenes was aliquid immensum infinitumque. The relevant paragraph in my paper should be modified accordingly. M. Michel: Je félicite M. Winterbottom pour la précision de sa méthode, qui le conduit à étudier essentiellement les jugements des déclamateurs et de Quintilien. Peut-être faudrait-il, à la fin de l’exposé, nuancer le mot voluptas: dans le sens esthétique, Cicéron préférait delectatio. Je voudrais ajouter quelques suggestions qui sortent plus ou moins du domaine auquel M. Winterbottom a choisi de se restreindre. D’abord, les jugements sur Cicéron apparaissent dans des textes où l’histoire interfère avec la rhétorique (Tite-Live, Asinius Pollion, Tacite . . . ). L’une des formes privilégiées de la déclamation a toujours été l’histoire. Salluste avait préféré la tradition de Thucydide à la conception cicéronienne de l’éloquence; Tite-Live y revient partiellement; Tacite faite une nouvelle synthèse des deux tendances, en s’aidant du langage virgilien. D’autre part, on peut aussi évoquer les problèmes politiques: ils se posent presque toujours lorsqu’il s’agit de pratique, et ils sont présents à 1’esprit de Quintilien quand il formule ses jugements relatifs à Cicéron. Domitien avait reproché aux philosophes leur secessus, par lequel ils refusaient de participer aux affaires publiques. Mais l’orateur, au contraire, s’abstenait par définition d’un tel otium. Or la conception cicéronienne faisait de lui un philosophe. Quintilien pouvait donc supprimer la contradiction entre philosophie et action, grâce à l’idée cicéronienne qu’il se faisait de la vraie culture. De là ce retour à Cicéron, qui s’affirme vers le temps de Trajan. J’ajouterai enfin une remarque sur la comparaison entre Démosthène et Cicéron chez le pseudo-Longin. Elle marque bien, chez l’orateur latin, la tendance à l’asianisme dont a parlé M. Winterbottom. J’ajouterai que l’auteur décrit l’éloquence cicéronienne

Cicero and the Silver Age

89

comme la coulée d’un grand fleuve. La notion de flumen orationis est importante chez les rhéteurs latins et se trouve liée à une certaine idée de l’elocutio. La continuité du flot oratoire s’oppose chez Quintilien au style coupé que pratiquent les amateurs de sententiae. Sénèque célèbre aussi une telle ampleur dans la continuité à propos de Cicéron lui-même ou de Papirius Fabianus. M. Winterbottom: Quintilian does use the Ciceronian word delectatio at 12.10.45, but voluptas appears at 5.14.35. M. Calboli: L’acuta osservazione di Alain Michel sul rapporto tra storici ed oratori mi fa pensare alla famosa lettera di Plinio il Giovane a Titinio Capitone (5.8.1) sullo scrivere storia: Suades, ut historiam scribam eqs. Ora sarei curioso di sapere quanto può avere influito il ciceronianismo di Plinio il Giovane nel suo atteggiamento di fronte alla storia. Il ciceronianismo in quel tempo, e in un uomo come Plinio il Giovane, influisce anche sui giudizi e sugli atteggiamenti letterari. M. Leeman: In Ep. 40 Seneca discusses delivery, not style. Fabianus was a fluent speaker (fundere / effundere verba). M. Stroh: Besonders einleuchtend fand ich, was Sie über die Frühgeschichte der Deklamation gesagt haben und über die Gründe, warum Cicero so wenig darüber spricht. Es gibt eine Trivialvorstellung von der römischen Redekunst, wie sie durch manche Handbücher geistert: Am Anfang war sie danach praktisch, forensisch, und ihr Meister hiess Cicero; dann, nach dem Ende der Republik, musste sie sich vom Leben in die Hörsäle zurückziehen und trieb die traurigen Blüten der Deklamationsrhetorik; ein tiefer Geist wie Tacitus sah diese historischen Zusammenhänge, während der Romantiker und Cicerofan Quintilian eine Beredsamkeit erneuern wollte, deren Zeit vorbei war. Herr Winterbottom hat hier mit Recht widersprochen, schon indem er auf das Alter der Deklamation auch in Rom hinweist. Ich frage mich nur, warum die Deklamation im späteren Sinn von declamatio—also in dem der controversiae et suasoriae, nicht im offenbar älteren Sinn der ‘Sprechübung’—erst in der frühen Kaiserzeit ausführlicher bezeugt wird; man hat doch den Eindruck, dass ihre Bedeutung damals zumindest mächtig zunimmt. Ein Problem bleibt mir bezüglich der Deklamationen in der frühen Kaiserzeit. Warum haben diese Rhetoren wohl eine politisch so brisante Sache wie die Ermordung Ciceros zum Lieblingsthema gemacht? Der Mörder Ciceros war zu einem guten Teil doch Octavian-Augustus selber. Bei Untersuchungen zur Textgeschichte der Philippiken ist mir aufgefallen, dass in den beim älteren Seneca exzerpierten Deklamationen immer nur auf die zweite Philippica angespielt wird, die Rede also, in der Octavian noch nicht erscheint. Das dürfte seinen Grund sicherlich in politischer Vorsicht haben, aber die Sache an sich bleibt trotzdem sonderbar. M. Winterbottom: As to M. Stroh’s first point: if I am right, there was a change of terminology (details in Bonner (1949), chs. 1–2) in the first century  rather than much change of practice. Cic. De orat. 2.100 by itself shows that

90

Papers on Quintilian and Ancient Declamation

the invented controversia themes, familiar from later sources, were known at this time (and also that Cicero was aware of their educational point). If declamation seems more prominent in the Augustan and post-Augustan period, that is partly a reflection of the continuing Hellenization of Rome. As to the second point, M. Stroh makes an interesting observation. The Second was the most popular of the Philippics for Quintilian also, and there may be no need to invoke politics. If the rhetores see Cicero’s death as the work only of Antony (so also Livy, ap. Sen. Suas. 6.17: pro certo habens . . . non magis Antonio eripi se quam Caesari Cassium et Brutum posse), that is after all consistent with the story of the struggle among the triumvirs for Cicero’s life (e.g. Plut. Cic. 46.2–6). The theme was too good to waste, and did not need to reflect on Octavian-Augustus. M. Leeman: There seem to be two types of declamation in Cic. De orat. Apart from the simplified type for pueri mentioned in 2.100, there is the important exercise in 1.149, causa aliqua posita consimili causarum earum, quae in forum deferuntur, still practised by Sulpicius and Cotta at the age of 33 and much more sophisticated in περίστασις than the ὑπόθεσις in 2.100, which seems to belong to the προγυμνάσματα. M. Calboli: Il termine declamare (declamatio), che il collega Stroh nega sia nato per tempo, in realtà trova nel noto passo di Cic. Brut. 310 un preciso terminus ante quem. Dice infatti Cicerone, Commentabar declamitans: sic enim nunc loquuntur. Ciò significa che l’uso del termine non doveva essere antico (nunc), ma certo alla data del Brutus era già in uso per questi esercizi retorici (per gli attori, cf. già Cic. De orat. 251). Sulle declamazioni avrei poi un altro elemento da suggerire, in parte seguendo una osservazione di Leeman. La differenza tra ‘tesi’ e ‘ipotesi’ in Ermagora, io non credo che fosse come in Cicerone. In Ermagora la ‘tesi’ era un esercizio generale che, con l’aggiunta dei μόρια περιστάσεως, diveniva una causa precisa, una ipotesi, e poteva variare col variare dei μόρια περιστάσεως. È Cicerone che introduce la filosofia, come ha ben mostrato il Michel, trasformando la ‘tesi’ da semplice schema generale di esercitazione in meditazione sullo stato del mondo, in espressione di principi generali. Vorrei sapere cosa pensa il collega Winterbottom di questa possibilità. M. Winterbottom: In answer to M. Calboli’s question, I should say that the distinction between θέσις and ὑπόθεσις did not in itself change. But Cicero’s philosophical enthusiasm meant that he gave θέσις more importance than others, both as an independent exercise (Att. 9.4) and within a forensic speech. The norm was to employ θέσις as an elementary προγύμνασμα (so still Quint. 2.4.24–5 with a nod to the Ciceronian position). M. Classen: What kind of public are Asconius’ commentaries addressed to? Is it reasonable to assume that people continued to study Cicero’s speeches, but increasingly felt the need for an explanation of the legal and historical background?

Cicero and the Silver Age

91

M. Winterbottom: As I said, I am not sure whether Asconius wrote as a schoolmaster for pupils or as a scholar for an interested general public (perhaps both). In any case, M. Classen must be right that Asconius’ work presupposes students of Cicero’s speeches in the mid first century  (a time when it seems that there was renewed interest in Cicero’s Letters), and ones who needed the background detail desiderated by Quintilian 10.1.22–3. M. Michel: Je voudrais ajouter une remarque qui ne constitue pas une question, puisque je sors de la période envisagée par M. Winterbottom. Il a montré comment on aboutit à la seconde sophistique. Cela coïncide avec un triomphe du cicéronisme, qui se manifeste notamment chez Fronton. Bien sûr, les Latins, comme l’atteste Aulu-Gelle, mettent de plus en plus l’accent sur l’archaïsme. Mais on ne parle plus de Sénèque et l’Asianisme, comme on nous l’a montré, n’est pas défavorable à l’Arpinate. La seconde sophistique fait la part grande à la déclamation et elle insiste surtout, avec Hermogène, sur la division des styles politique (Démosthène) et épidictique (Platon); Isocrate est considéré comme faisant la synthèse. Dès lors, le succès du cicéronisme, chez les écrivains chrétiens d’Afrique (ou de milieu africain), de Minucius Felix à Lactance, n’a rien d’étonnant. Saint Augustin, qui vient après eux, est nourri de culture cicéronienne. Peut-on suggérer qu’il retrouve, au-delà, la tradition de la tension sallustienne et sénéquienne? M. Winterbottom: I should prefer to put it that Cicero and the declaimers saw antiquity out together.

6 Schoolroom and Courtroom At the very end of what we have of Suetonius’ De grammaticis et rhetoribus¹ we are told of the death of Gaius Albucius Silus of Novara. He was suffering from a tumour, and, returning to his home town, he ceased taking food. But before dying he called the people together and gave the reasons for his decision to take his own life. Suetonius tells us that he did this more contionantis; but it would be more correct to say that it was more declamantis. There seems no reason to suppose that giving reasons for one’s suicide was a custom commonly indulged in ancient Italy: though Valerius Maximus² witnesses to something like it among the Greeks of Massilia. Albucius’ death was surely consciously modelled, with an élan worthy of Petronius himself, on the law that he will so often have treated in his declamation-hall: Q       ,  .³ This same Albucius reminds us in a less bizarre way of the intimate connections between declamation and real-life oratory. He was mainly a declaimer; but, at least before an unfortunate incident in the centumviral court, he also pleaded cases.⁴ It was not only that the schools influenced the courts by training future lawyers; Albucius, like other declaimers,⁵ led a double life between lecture hall and courtroom. I want in this paper, in what must be a random fashion, to discuss some of the ways in which this interaction affected both declamation and legal oratory. Chronology might seem to make it difficult for us to see declamatory elements in the preserved speeches of Cicero. They stretch from the 80s to the 40s . Declamation, on the other hand, was said by the Elder Seneca to

[From B. Vickers (ed.), Rhetoric Revalued. Papers from the International Society for the History of Rhetoric (Binghamton (NY), 1982), 59 70] ¹ Suet. Rhet. 30.7. The passage was juxtaposed with the declamation law by Sprenger (1911), 238; but Sprenger did not draw my conclusion. ² V. Max. 2.6.7 8. Cf. Sprenger (1911), 237. ³ e.g. [Quint.] Decl. mai. 4.th. (60.11 Håkanson). ⁴ Suet. Gram. 30.4 5. ⁵ Quintilian is the most distinguished example. It was true even of grammatici (Suet. Gram. 22.1). So too in Greece: cf. Philostr. VS 511; 516, etc.

Schoolroom and Courtroom

93

have been a thing born later than himself; and he was born not earlier than 55.⁶ But this information can hardly be taken literally. Declamation was something with its roots in the Greece of the era between Demetrius of Phaleron and Hermagoras of Temnos. It was a Hellenistic creation, that brought its apparatus of tyrants and piracy and New Comedy sexuality over the Adriatic to Italy. Such influence will have been reaching a peak in the philhellene years of the late second century , and there is nothing to prove that the lectures Cicero took down as the basis for the De inventione were given anywhere else but in Rome. The declamation themes that occur in that treatise may not be as extravagant as those that we find later in the Elder Seneca. But it may well be that the schools were already in the early first century moving towards the sort of melodramatic and even epigrammatic style that was later characteristic of them. In the Orator⁷ Cicero claims to have found the ears of Rome ‘hungry for this varied style that spreads equally into every type’; ‘I was the first to turn them to an astonishing enthusiasm for this manner.’ Now observe the illustrations Cicero gives us. First—received by what clamours!—the famous extravaganza beginning: Quid enim tam commune quam spiritus vivis, terra mortuis, mare fluctuantibus, litus eiectis?⁸ It may not be a coincidence that this passage found so ready acceptance into the declamatory style of the Quintilianic Decl. min. 299; I have found it echoed in the Elder Seneca and Apuleius too, not to speak of Martianus Capella and a Donatist council. It is perfectly on the cards that what went to feed declamation came originally from declamation; the youth who had to defend Roscius of Ameria on a charge of parricide mentions in the De inventione⁹ the case of a father-murderer who suffered the horrors of the culleus. And if we had a fourth book of the De inventione it would show us, as does the fourth book of the Ad Herennium, a schoolroom milieu perfectly well aware of all the stylistic tricks of the declaimers’ trade. Indeed, that emerges from Cicero’s second example from his early, though mature style: uxor generi, noverca filii, filiae paelex. What are these words of the Pro Cluentio¹⁰ if not declamatory? Was not the champagne bubble of the unregenerate Cicero the product of the declamation school? It is the argument of the Brutus¹¹ that his visit to Asia enabled him to calm down his hectic oratory: a paradox meant to

⁶ Sen. Con. 1.pr.12. For Seneca’s date of birth, cf. Griffin (1972), 5. The difficulties of this pas sage, and the development of declamation generally, are discussed e.g. by Bonner (1949), ch. 1. ⁷ Cic. Orat. 106 7. ⁸ Cic. S. Rosc. 72, echoed in ‘Quint.’ Decl. min. 299.4; Sen. Con. 7.2.3; Apul. Met. 4.11.7; Mart. Cap. RLM p. 477.14 ff. Halm [= 5.522 Willis] (citation for metrical purposes). For the Council cf. Norden (1898), 626. ⁹ Cic. Inv. 2.149. ¹⁰ Cic. Clu. 199. Cf. Sen. Con. 6.6: Generi adultera, filiae paelex; 9.6.1: Nefaria mulier, filiae quo que noverca. ¹¹ Especially Brut. 316.

94

Papers on Quintilian and Ancient Declamation

explain away what critics of his Asianism must have used as a trump card. But he had seen Asian orators before he left Rome: they were the rhetores who ran the schools, the forerunners of persons we meet in Seneca: Grandaus, Asianus aeque declamator; Adaeus, rhetor ex Asianis,¹² Glaucippus of Cappadocia, Hybreas of Mylasa . . . Cicero, however, matured, and grew away from the schools, even grew to look down on rhetores who had done so much for him. He became an exemplum in his turn. It is easy for the historian to think that Cicero’s speeches were published as part of a series of propaganda campaigns intended to recommend Ciceronian policies or gloss over Ciceronian errors. And indeed something approaching that reason is given when Cicero sends to Atticus the corpus of consular orations: ‘Since you like my writings as well as my doings, the same compositions will show you both what I did and what I said.’ But we should notice a little earlier the stress placed on another motive for publication: ‘it appears’, says Cicero to Atticus, ‘that you too find pleasure in these performances which the enthusiasm of my young admirers prompts me to put on paper.’¹³ Adulescentulorum studiis excitati. And it is not unreasonable to expect that, in writing up speeches for the public, Cicero kept in mind the interests of the adulescentuli, still in their schools or already launched on the courts with Cicero, perhaps, as their mentor. He was particularly free to do this in the Pro Milone. The spoken speech, if not a disastrous failure, was shorter and less grand than Cicero wished it to become for posterity. I shall say something later about a particular addition he made to the spoken version. But in general I think we should see in the peculiar faithfulness of our speech to the precepts of rhetoric a reflection of Cicero’s care to educate the youth. This is a speech, as he almost says of his consular orations, that he hopes to see read in the schoolroom with the attention accorded to the Philippics of Demosthenes. But as a bonus it was to exemplify in action the rhetorical rules that seem so lifeless when they are merely placed end to end in an ars rhetorica. That, too, is a theme to which I shall return. The Pro Milone is exceptional, and there are good reasons for the exception. But I shall discuss briefly two other examples of my point. The Pro Caelio was, I take it, in another sense for Caelius; it would be sent to Cicero’s old protégé in the hope that this brilliant example of a defence speech might turn to better ways one who was showing such disconcerting signs of a rabies accusatoria.¹⁴ But it also recalled to Cicero’s old pupil the days when he had been his model in oratory; it is almost as a private joke that he says: ‘I can also make other enquiries of the accuser, after my own custom and that of advocates generally: where did Caelius meet Lucceius’ slaves, how did he gain access? . . . I can take ¹² Sen. Con. 1.2.23; 9.1.12. ¹⁴ Cf. e.g. Cic. Cael. 76 7.

¹³ Cic. Att. 2.1.3 (tr. Shackleton Bailey).

Schoolroom and Courtroom

95

my speech through all the thickets where suspicions may lurk: there will be found no motive, no place, no means, no accomplice, no hope of achieving or concealing the crime, no reason, no trace of so terrible a deed.’¹⁵ This all has word for word correspondence with the Ad Herennium; and I think that Cicero included it, together with the following¹⁶ technical distinction between the πίστεις ἔντεχνοι and ἄτεχνοι, not because his original speech indulged in such coy allusion to the schoolroom, but because the adulescentuli could profitably be reminded that their masters’ precepts were not irrelevant to a real case. He is employing a trick we shall see used later in declamation: pedagogy intrudes into the fair-copy oration. I add a further example of a rather more far-reaching kind. At the start of the fourth speech against Catiline, Cicero says: ‘I see, conscript fathers, that the faces and eyes of you all are turned towards me. I see that you are anxious not only about the danger to yourselves and to the republic, but also, if that is repelled, about the danger that threatens me.’ As Professor Nisbet once wrote of another aspect of these speeches: ‘When faced with a concrete and urgent problem did the Roman senate really waste time on this stuff?’¹⁷ The answer, as Nisbet implied, is surely ‘No’. But it is more interesting to speculate why the fourth Catilinarian has come to us in the form it has. It is partly a matter of changed circumstances. The words si id depulsum sit show right from the start that Cicero wrote up his speech with more than half an eye on what was to follow the events of 63, on the wave of feeling which, fostered by Clodius, was to remove him into exile. Hence—and this is an obvious point, often made before—the unnatural emphasis on the peril to Cicero, on the ‘perpetual war’ that he has undertaken with ‘desperate citizens’.¹⁸ Less obvious, perhaps, is the effect this new viewpoint has had on the very genus of the speech. It is true that the speech remains in its general direction deliberative. Its penultimate sentence is a plea to the senators to make their decision carefully and bravely. But it is significant that the last sentence of all redirects their attention to the consul, a consul prepared to stand up for their decision quoad vivet:¹⁹ again the note of personal concern. And it seems to me that the speech has been recast in ways that bring it close to a legal speech, and one written in selfdefence: not in Catilinam so much as pro Cicerone. In this recast speech the deliberative kernel—which, I take it, represents things spoken on the actual occasion²⁰—takes on a new aspect. The arguments for execution become, as well, arguments for Cicero’s correctness in desiring and carrying out that execution. Around this kernel, Cicero has placed proem

¹⁵ Cic. Cael. 53, with Austin (1960³), ad loc. ¹⁶ Cic. Cael. 54. ¹⁷ Nisbet (1965), 62. ¹⁸ Cic. Catil. 4.22; cf. 3.27. ¹⁹ Cic. Catil. 4.24. ²⁰ It would not surprise me if Cicero spoke only at the end of the debate, and summing it up as a chairman does. For the kernel note e.g. Catil. 4.6 (Illa praedicam quae sunt consulis) and the tell tale Nunc antequam ad sententiam redeo, de me pauca dicam (4.20).

96

Papers on Quintilian and Ancient Declamation

and epilogue hardly conceivable in the original senatorial speech. In the proem, he evokes the tears of his brother, carissimi atque amantissimi, and present to hear him, but says he will die with mind untroubled and prepared.²¹ In the epilogue he commends to his hearers his small son.²² These are topics, surely, more proper to judicial oratory, and through them Cicero appeals not so much to the senatorial audience of 63 as to a wider body of Romans in the years that immediately followed. But this was one of the corpus of consular speeches that, as we have seen, was published by Cicero partly in answer to the enthusiasms of young admirers. And I should not be surprised if Cicero wrote up his senatorial speech to some extent in judicial form because judicial speeches were what the schoolroom wanted. The suasoria was not taken too seriously at school: it was reserved for the young.²³ It was the controversia that engaged the attention of the older pupils. I have come to the schools, and shall now turn to the second half of my topic. It has been a cliché not only of modern but of ancient critics that the declamation schools were disastrously adrift from reality. I need only mention the splendid assault on the rhetores at the start of Petronius’ Satyricon—‘pirates with chains in their hands standing on the beach, tyrants writing edicts instructing sons to cut off their fathers’ heads’—along with the devastating parody of the declamatory style: ‘These wounds I received for the freedom of all, this eye I forfeited for you. Give me a guide to lead me to my children: my knees are hamstrung and cannot support my body.’²⁴ And clearly there was something excessive about some first-century practices. We know this because Quintilian, himself a rhetor, goes on the defensive when he discusses the topic. His premiss is that ‘things that are naturally good can be put to good use’, and declamation is for him a naturally good thing. It is only thanks to its teachers that (he says) ‘it has now so degenerated that the licentious ignorance of declaimers has become one of the principal causes of the corruption of eloquence.’²⁵ Now Quintilian’s point in this chapter is that declamation was started as a preparation for the forum. It should therefore ‘follow the pattern of the speeches it was invented to train for’. Quintilian sees that the whole business, especially the feeling of unreal emotions, trembles on the brink of absurdity: the only defence can be that we are, ‘as it were, on manoeuvres to prepare us for the real battle and the serious fighting’. At the same time, declamation does have an element of display about it, and that justifies it being given a certain ²¹ Cic. Catil. 4.3. ²² Cic. Catil. 4.23. ²³ So at least later Tac. Dial. 35.4. ²⁴ Petr. 1.3 and 1.1. Haec vulnera pro libertate publica excepi, hunc oculum pro vobis impendi is very like the repetitiousness so prevalent in the Declamationes minores (e.g. at random 251.7: quidquid asperrimum leges, quidquid crudelissimum habent iura, occupas). Succisi poplites seems to me to strike a note of absurdity, despite (or because of) its echo of Verg. Aen. 9.762 and 10.699 700. ²⁵ Quint. 2.10.3. In what follows I allude especially to 2.10.4, 8, and 12.

Schoolroom and Courtroom

97

brilliance (nitor) that might be improper in the law-courts. Quintilian tries to have it both ways: declamation as pure training combined with the declamation of ἐπίδειξις. Now, whatever the excesses of some of the rhetores, it is exceedingly difficult to believe that declamation was not on the whole regarded as a practical as well as an attractive way of training the public speaker. Consider it. From 100  or even earlier till the times of Ennodius, the rhetores held sway over the secondary education of a people we are accustomed to laud for their practical genius. Did these same Romans who built aqueducts and amphitheatres and established a masterful legal code really consign their children to an education system that patently had no advantages in the formation of a lawyer or politician or, later, civil servant? We should at least give them the benefit of the doubt, and try to see what declamation could offer besides its superficial absurdities.²⁶ It is difficult to use the evidence of the Elder Seneca in this enquiry. He reports with some favour the forceful views of Cassius Severus on the gulf between schoolroom and courtroom.²⁷ But Seneca’s own book is constructed in such a way that we find it hard to be sure what a complete declamation was like. And that is the crucial thing. Quintilian, or the Pseudo-Quintilian, draws a distinction between the bones of a speech and its flesh:²⁸ a distinction not so much between the argumentation and the less essential elements as between the skeleton of the speech, its formal division, and the language which clothed the skeleton. In this terminology Seneca is particularly interested in the flesh: the bulk of his work consists of striking phrases culled from many speeches, grouped together by theme. But it would be quite wrong to imagine that a speech by, say, Albucius Silus consisted entirely of epigrams. On the contrary, it is clear from the sections that Seneca devotes under each topic to divisio that the declaimers were very much concerned with the structure of their argumentation. The tiresome thing is that nowhere²⁹ are we given a complete declamation from which we could see how much this division dictated the overall pattern—how far, for instance, it was obscured or swamped by more meretricious elements. Where, then, can we look? Primarily, I suggest, to the Minor Declamations that are transmitted under the name of Quintilian. Here are 145 extant declamations, some long, some short, some supported by sermones where the rhetor advises how the subject should be treated, others left on their own. They are the fragments, posthumously published as I should judge, from

²⁶ See now for a sympathetic assessment of declamation Bonner (1977), esp. ch. 21. ²⁷ Sen. Con. 3.pr.8 18. Cf. 9.pr. (also 2.3.13). ²⁸ Decl. min. 270.2. ²⁹ In Con. 2.7 Seneca gives a good deal of a declamation of Latro’s, but this is cut short by a lacuna in the manuscripts. It is noticeable, however, as Dr J. A. Fairweather stresses to me, that much of what remains is (fairly) sober argument. [See further A.7 below, p. 111.]

98

Papers on Quintilian and Ancient Declamation

a practising teacher’s workshop. I stress ‘practising’. In the Elder Seneca we can never be quite sure in what circumstances the various declaimers uttered their more or less scintillating sententiae. I should guess that there was a considerable range of possible situations, stretching from the purely teaching declamation, given as a ‘fair copy’ behind closed doors, to the purely epideictic declamation, given to impress friends and attract publicity for the school: the context we should, I think, imagine for the Major Declamations. But the Minor Declamations do not make sense except as the notes of a rhetor involved in the day-to-day business of teaching the young.³⁰ And from them we can see how it is possible to regard declamation as a training for real life. I should perhaps enter a caveat here. The collection shows the most unmistakable signs of knowledge of the Institutio oratoria; and I do not feel able to confute, on linguistic or other grounds, anyone who thinks them the work of Quintilian himself. In that case, and perhaps in any case, we may have to regard these declamations as exceptions: I mean as composed with exceptional attention to reality by someone who felt strongly that declamation must not lose contact with the courts. A prima facie objection at once emerges: are not these declamations, like all others which we know from the ancient world, based on themes that cannot, ut parcissime dicam (to use the rhetor’s own cliché), often have found counterparts in real life? The tyrants writing edicts, the pirates carrying chains are still familiar figures. Worse still, the laws that govern the declamations are usually unreal. Now there is an answer to this objection, and I am surprised that Quintilian himself did not formulate it. If you make an unreal speech to train your powers of argument and expression, it is positively preferable that it should be on an unreal topic and subject to unreal laws. This is clearest from a ‘historical’ declamation like 323. ‘Alexander’, it is stated in the theme, ‘was besieging Athens, and he burned a temple outside the walls. He began to suffer from a plague. An oracle stated that the plague would only finish if he restored the temple . . . ’ and so on. Now Alexander did not lay siege to Athens at all; a fortiori this particular incident can only be fictional.³¹ It was therefore possible for a schoolboy to treat the theme without any distraction: all he needed to know was that Alexander was a young and bold general, and that Athens was famous for religious observance. The rest was up to his ingenuity. If he had been faced with a real historical incident, he would, or should, have felt the need to read his history books, name names, allude to further events. Faced with a fiction, he can fall back on pure argument. Those who regard the ³⁰ A new idea (I think), which I shall develop in the introduction to a forthcoming edition of the work. [See now Winterbottom (1984), .] ³¹ Cf. Sen. Con. 7.2.8 on fictions over Popillius (also Suas. 6.14 15). This is what Cicero means when he makes Atticus say: concessum est rhetoribus ementiri in historiis, ut aliquid dicere possint argutius (Brut. 42). It was of course possible for declaimers to be historically accurate: so Aristides (see Reardon (1971), 105).

Schoolroom and Courtroom

99

learning of facts as more important than learning how to argue will find this deplorable; but rhetores taught rhetoric, not history.³² The same is true of nonhistorical cases. And here we know that my point was seen in antiquity. In court, says Cicero in the De oratore, we must get to know the case at issue very thoroughly. ‘But’, he goes on, ‘this precept is not applicable to the schoolroom. For the cases that boys treat are simple. “The law forbids a stranger to ascend the wall. A stranger has ascended it and repulsed the enemy. He is accused.” It takes no trouble to learn up a case like this.’³³ Exactly so: the boys could get to know how cases are argued without the distraction of unnecessary detail. Quintilian once suggested that declamations might be improved by the use of proper names, longer and more complex themes, an infusion of ordinary words, and—optimist that he was—jokes.³⁴ But he did not suggest that reallife cases, past or present, would provide better material. They would have too many ramifications; and anyway, in the nature of things, neither master nor pupil would have access to the details. So too with declamatory laws. Even in the early days of Greek declamation they can hardly have been more than a very simplified reflection of Attic practice. By the time of the early Roman empire they are a hopelessly confused mixture of Greek and Roman, fact and fiction. To try to disentangle the ravelled skein is not a waste of time;³⁵ but to commit oneself to the general proposition that declamatory law is somehow a ‘source’ for Roman law is amongst other things to ignore the point that these laws were so regularly chosen just because they were largely fictional. You learned your law at a law school or in the practice of the courts. You learned to argue at the declamation school. And the two were kept apart for a very good reason. The use of fictional laws encouraged flexibility and ingenuity of argument. If the laws had been real one would have had to step more carefully. And one might have got used to asserting things about the law that turned out to be false when one appeared in the courts. Cum ad veros iudices ventum . . . When the great lacuna of the Dialogus³⁶ cuts him short, Messalla is about to use the contrast between school and courtroom as a stick with which to beat the declaimers. But it was not quite so simple. When one first arrived in the courtroom one knew that one did not know the law; that was better than thinking one knew it when one in fact did not. To come more specifically to the Minor Declamations: I want to suggest two ways in which they achieve their pedagogic purpose. First, they show in action what in an ars rhetorica would normally be stated as a precept. Thus in the De inventione Cicero gives detailed heads of arguments to be used in cases with different στάσεις. Each of the Minor Declamations clearly has its own στάσις, ³² Quintilian encouraged his pupils to read historical works (e.g. 2.5.19); but that is a different matter. ³³ Cic. De orat. 2.100. ³⁴ Quint. 2.10.9. ³⁵ See especially Bonner (1949), chs. 5 6. ³⁶ Tac. Dial. 35.5. Cf. e.g. Petr. 1.2.

100

Papers on Quintilian and Ancient Declamation

and the arguments appropriate to it are deployed not abstractly, but with reference to the particular theme.³⁷ Now this, of course, does not count as a link between schoolroom and reality if, as may have been the case, the niceties of the στάσις-system were scorned or neglected in the law-courts. Aper—no sympathetic witness, it is true—speaks in the Dialogus of ‘the items prescribed in the dry-as-dust handbooks of Hermagoras and Apollodorus’³⁸ as being a feature of the old unenlightened courts of the Republican past. But the fact is that, even if the Hermagorean system took its classification of arguments too far, the arguments themselves remained indispensable; and someone trained in the system would know where to look for his proofs, have, as it were, a checklist to ensure he had made no important omission.³⁹ Secondly, the fair-copy declamations in this collection are designed quite consciously to show the ossa controversiae, the bones. Thus, to take an example at random, a speech (339) purporting to be made by Demosthenes supporting a bill to prevent those captured by Philip at Chaeronea from taking part in Athenian assemblies. The speaker starts: Priusquam causas rogationis meae persequar, making it clear that what immediately follows is a proem extra causam. After this proem, he says that the other side contends that bills cannot be accepted if they are against the law or if they are directed at individuals. Each of these points is discussed, carefully marked off at the start of each: neque adversus leges esse existimo and Illud . . . aliquanto minus existimari possum, adversus singulos scripsisse legem. A further divisio, between consideration of the interests of the former captives and the interests of the state, is made explicitly, and it dictates, again with the appropriate signposts, the structure of the rest of the piece. It is just the same impulse, to make the shape of the speech intelligible in advance, that causes the frequent use of such phrases as postea videbo, answered by interim . . . ⁴⁰ These techniques could, of course, be used to mark off the parts of a speech laid out on Hermagorean principles; but they could still teach something to a speaker who when he arrived in court might wish to forget everything he had ever learned about Hermagoras. The rhetor by these methods was inculcating something that would always have its uses, the practice of lucid and organized thought. The point was to train the pupil to speak a declamation that had a shape, that advanced from argument to argument in a logical sequence, and to have the shape of the whole in his head at the same time as he spoke a part. As I read my

³⁷ For example, the topics of the conjectural στάσις are run through in Decl. min. 321. See also Bonner (1949), 15. The system and declamation fed each other; I doubt if the system had much to do with real life. ³⁸ Tac. Dial. 19.3. ³⁹ Cf. e.g. Cic. Brut. 263 (where it is noticeable that Cicero regards an orator trained in the details of the Hermagorean system as exceptional). ⁴⁰ Decl. min. 275.2.

Schoolroom and Courtroom

101

students’ essays, I sometimes wish that the art of divisio were imparted along similar lines in the schools of today. Now one constant method of dividing a controversia was to move from treatment of the law to treatment of equity. This fitted in well with the movement through argumentation to the emotional epilogue, but it was a different level of division. Was it a division with no relevance to the courtroom? It is worth stressing here that the point at issue is not that still argued about in the wake of Stroux’s classic Summum ius summa iniuria.⁴¹ Stroux was concerned to argue that rhetoric—and of course in particular the contrast of ius and aequitas—affected the interpretation of the law far earlier than the Byzantine period to which others so confidently relegate any part of the Digest that smacks of rhetoric. Without being in the least qualified to decide this grave point, I should not be in the least surprised if Stroux were right. Orators and lawyers must have gone to the same secondary schools, and spoken much the same language. Indeed it would be very desirable that the Minor Declamations should be studied by someone with a mastery of legal texts and legal terminology.⁴² Furthermore, I have been struck, in such parts of the Digest as I have read consecutively, by the resemblance between the tricks of the two trades of rhetorician and lawyer. Both had pupils to teach, and both stimulated them in a similar way; as Cicero puts it in the Topica: ‘Fictitious examples . . . have their value, but they belong to oratory rather than to jurisprudence, although even you’ (he is addressing the lawyer Trebatius) ‘are wont to use them.’⁴³ Now Gaius constantly mentions the differing views of the legal schools. And Cicero’s De oratore tells the same tale of dissension among the learned. ‘It is not difficult’, says Antonius, ‘for an orator to find a lawyer to back the side he is supporting, whichever it happens to be.’⁴⁴ If all law were straightforward, the lawyer would be out of business: but so too would the orator. And when the law was doubtful it would often happen that aequitas was brought in to make one view rather than another more plausible: ius . . . dubium aequitatis regula examinandum est is what Quintilian expressly says.⁴⁵ I should be astonished if, even before the decadent days of Byzantine sentimentalism, lawyers themselves did not sometimes find themselves arguing in these terms. But my more confident, and different, point is that whatever the lawyers did, practical oratory, at least in the early empire, was as open to considerations of equity as the declamation school. It is true that the rhetor of the Minor Declamations at one point follows up his detailing of the legal topics by

⁴¹ Republished in Stroux (1949). ⁴² I cannot judge the knowledge of law displayed by Lanfranchi (1938), but his knowledge of rhetoric left something to be desired. ⁴³ Cic. Top. 45. ⁴⁴ Cic. De orat. 1.242. For Gaius, see e.g. Inst. 1.7. ⁴⁵ Quint. 12.3.6.

102

Papers on Quintilian and Ancient Declamation

remarking: ‘After that we may say the things that people tend nowadays to restrict themselves to’⁴⁶—by which he must mean that modern declaimers were too keen on aequitas. It is also true that it was only when Cicero came to write up his Pro Milone for publication that he ended it with the argument that if Milo had killed Clodius he would have been abundantly justified in doing so.⁴⁷ But we can use the Institutio itself to show that practical oratory too laid stress on equity. Stroux demonstrated in his famous libellus how such arguments were the stock-in-trade of the late Republic.⁴⁸ One need only read the Pro Caecina, or Cicero’s various accounts of the famous causa Curiana,⁴⁹ to see that those hours in the classroom were not wasted when it came to the forum. But Stroux’s argument can be carried forward to the next century. Quintilian was no stranger to the law-courts: he several times lays stress on his forensic experience.⁵⁰ He can speak with equal authority of declamation and of actual practice. Thus, as to the schoolroom, he tells us, in connection with a declamation on Alexander and the Thebans, that, ‘as the Amphictyones are the judges, the most effective part of the case will be the tractatio aequi’;⁵¹ or again in a declamatory context he remarks that ‘generally we will find a treatment of aequitas at the end of cases, because judges’ (note the slide over into reality) ‘listen to nothing with greater pleasure.’⁵² But elsewhere Quintilian says much the same when he is undoubtedly speaking of the courts. Thus he says: ‘we must get to know the nature of the judge: is he more inclined to ius or to aequum?’⁵³ Or, speaking of ipsae actiones, he stresses the importance of the decision whether to use first arguments from ius or from aequum.⁵⁴ Or again, in that most real of chapters where he discusses the need for morality in the practising orator, he says: ‘Who does not see that a great part of oratory consists in the treatment of aequum and bonum?’⁵⁵ So too had Cicero said, specifically of iudicia privata, that the contest is often not about the act, but about aequitas and ius.⁵⁶ In the empire, as under the Republic, equity was an important part of a case; and declamation, at least in the hands of a Quintilian, was here as elsewhere a more realistic form of training than it is fashionable to think it.

⁴⁶ Decl. min. 270.5. Cf. Quint. 5.8.1. ⁴⁷ See Ascon. p. 41.9 Clark. ⁴⁸ Stroux (1949), 42 ff. ⁴⁹ e.g. Cic. Brut. 143 ff. (with Douglas (1966), ad loc., valuable on what exactly the opposition of ius and aequum means: a topic also covered by Nörr (1974), 36 ff.). ⁵⁰ Contrast Cicero on the ivory tower academic: quod . . . ipsi experti non sunt, id docent ceteros (De orat. 2.76). Correspondingly Quintilian was not content to use examples from declamation alone, but continually cited real orators, unlike the minor rhetoricians such as Fortunatianus and Sulpicius Victor. In Greek, compare the constant examples from Demos thenes as well as from declamation in e.g. (Ps. )Hermogenes’ De inventione. ⁵¹ Quint. 5.10.118. ⁵² Quint. 7.1.63. ⁵³ Quint. 4.3.11. ⁵⁴ Quint. 6.5.5. ⁵⁵ Quint. 12.1.8. ⁵⁶ Cic. De orat. 1.173.

7 Declamation, Greek and Latin The declamatory texts most familiar to us are Latin; and it is on these that scholars have lavished their attention. Seneca the Elder is read, at least in part, even by those not primarily interested in rhetoric; and my own Loeb¹ will eventually, I am glad to say, be supplemented by a Teubner text edited by Professor Lennart Håkanson of Uppsala.[²] I am myself about to publish an edition, with commentary, of the Minor Declamations, perhaps rightly attributed to Quintilian.[³] As for the Major Declamations, deplorably edited long ago by Lehnert, we now have Håkanson’s vastly improved text.⁴ Yet it remains true that declamation was a Greek discovery, and that the Greek practice is abundantly illustrated by extant texts. One of the most important of these, the Διαίρεσις ζητημάτων of Sopatros of Athens, has been little studied. We still have to read it in a text that advances little upon what the Aldine editor transcribed from an inferior manuscript. For two or three years my Oxford colleague Dr Doreen Innes and I have together been discovering how much can be done for this work by conjectural emendation and by use of a primary manuscript.[⁵] On a wider front, we can look forward to a general book on Greek declamation from Mr Donald Russell of Oxford, as well as the continuation of Professor George Kennedy’s authoritative work on ancient rhetoric.⁶ But meanwhile I offer some remarks on the interconnection of Greek and Roman declamation, remarks that I hope will throw some light on the vexed matter of Asianism, on the development of prose rhythm, and on the history and techniques of declamation itself. Let us start with the more familiar. The Elder Seneca was well aware of the importance of the Greek declaimers of his day. He was writing in Latin, and for readers in the west. But he names thirty-six Greek declaimers, often reserving a place at the end of each theme for extracts from their speeches. Further, he is well aware of the interplay between Greek and Latin speakers. [From A. Ceresa Gastaldo (ed.), Ars rhetorica antica e nuova (Genoa, 1983), 57 76] ¹ Winterbottom (1974b). ² [See now Håkanson (1989).] ³ [Winterbottom (1984).] ⁴ Lehnert (1905); Håkanson (1982). ⁵ [See now Innes Winterbottom (1988).] ⁶ See now Russell (1983) and Kennedy (1983).

104

Papers on Quintilian and Ancient Declamation

It is true that it was a matter of some hilarity that Clodius Sabinus should declaim in both languages on the same day; Maecenas quoted Homer (Il. 5.85): Τυδείδην δ’ οὐκ ἂν γνοίης ποτέροισι μετείη.⁷ Seneca sounds more appreciative of the Greek Argentarius, who was ‘loyal to the principle of never declaiming in Greek’.⁸ But that would not mean that Argentarius, or his more celebrated teacher Cestius, who hailed from Smyrna, could divorce themselves from their Greek upbringing. As for the Latin-born declaimers, Albucius, admittedly someone very prone to imitate, was (according to Seneca) ‘breathless with admiration’ for the Greek Hermagoras, and burned to imitate him.⁹ Scaurus thought that current obscene treatment of a certain theme derived from the Greek declaimers.¹⁰ Murredius ‘wished to reproduce’ an epigram of the Greek Nicetes.¹¹ Of course, there was room for coincidence in epigram where so many treated a theme: finding a parallel between Artemo and Latro, Seneca once comments¹² that ‘Latro cannot be suspected of plagiarism, for he both despised the Greeks and was ignorant of them.’ Again, Seneca was not without a sense of the rivalry of Greek and Latin declaimers, nor innocent of prejudice in favour of ‘our’ speakers. He is glad, when he can, to find a saying that ‘outstrips the Greeks’ (Graecos praeminet).¹³ All the more striking, then, his awareness of the general dependence of Latin declamation on Greek, and his feeling that the Greeks set the contemporary standards of quality and practice. The scribes of the Elder Seneca’s medieval manuscripts found great difficulty with the Greek extracts. Sometimes, to our loss, they just left the mysterious hieroglyphics out in despair. But what remains is uniquely valuable. It shows us the scholastic Greek declamatory style that made the Silver Age of Latin what it was: a style that we can see in no contemporary Greek source, least of all in the complacent Dionysius of Halicarnassus, who felt, quite wrongly, that the salutary influence of the Roman empire had somehow banished from all but a few cities in Asia the manner that had prevailed in Greece for so long, ‘vulgar, frigid and tasteless’. Dionysius thought that this style had begun, as it was allegedly ending, in Asia: it was ‘sprung from some Asian sewer the other day—some Mysian or Phrygian or Carian plague’.¹⁴ For sufferers from this plague we must look, if anywhere, to Seneca’s Greeks. Thirty-six, as I have said, are named. We know the homes of about ten. Only one came from mainland Greece: a Spartan, oddly enough—Laconic brevity at last in fashion. Three came from the islands (Euboea and the Asian Mytilene), one from the Greek city of Massilia in Gaul. The other five, significantly, come from Asia Minor: Smyrna, Tralles, Cappadocia, Mylasa, Pergamum. The Smyrniot is called by Seneca merely Graecus; and others he calls ‘Greek’ ⁷ Sen. Con. 9.3.14. ¹⁰ Sen. Con. 1.2.22. ¹³ Sen. Con. 1.4.12.

⁸ Sen. Con. 9.3.13. ⁹ Sen. Con. 7.pr.5. ¹¹ Sen. Con. 1.4.12. ¹² Sen. Con. 10.4.21. ¹⁴ Dion. Hal. Vet. orat. pr.1.

Declamation, Greek and Latin

105

may also be from Asia. If anything can tell us about Asian rhetoric in the early Empire, it must be these texts, corrupt and fragmentary though they are. One thing antiquity agreed on finding in Asianist oratory was rhythm. Cicero¹⁵ remarks upon the double trochee as Asianic. And the climax of Quintilian’s account of what Atticist opponents criticized in Cicero lies in the words in compositione fractum, exultantem ac paene . . . viro molliorem:¹⁶ Cicero’s positioning of words was dance-like and effeminate. It is odd that, so far as I know, scholars have not thought to analyse the clausulae of the Greek extracts in the Elder Seneca. If one does, one finds a striking fact: out of 134 ‘strong’ endings of sentences (marked by full-stop, colon, or question mark), 72 are rhythmical in the sense that they exhibit double trochee, cretic plus trochee, or double cretic, or resolved versions of those combinations. As many as 42 of the 62 others are placed in doubt by ambiguities of scansion. Particularly regular declaimers are Adaeus (6 rhythmical endings out of 7), Damas (4 out of 5), Dorion (4 out of 6) and Hermagoras (7 out of 11). Adaeus was, Seneca¹⁷ tells us, ex Asianis; Damas came from Tralles. I lay especial stress on these facts because Greek texts scannable on this system, familiar to us from Latin, are so rare. We all know that Eduard Norden¹⁸ illustrated Asianic rhythm from the grandiose inscription raised by Antiochus of Commagene in an incomparable site on the summit of the Nimrud Dağ in south-eastern Turkey. Those bombastic words will have been written by some friendly rhetor. Where else can we turn to find these rhythms? Norden¹⁹ saw traces in Herodian and Favorinus, sophists both: one from Syria, the other a Gaul who frequented the Greek east. I can add Iamblichus, declaimer and novelist. He came from Syria, too. But we are short of these rhythms in Greek, just as we are short of Asianist writing in general: that is no coincidence, of course, for Asianist writing was rhythmical. Later, by paradox or because of concessions to the enemy, the Atticizers became rhythmical too. By then, the system had changed; it was based now not on the length of syllables but on their accentuation. But if we want to get an idea of what critics meant by the sing-song effeminacy of the Asians, we could do worse than to turn to the sixth century and the declamations of Choricius of Gaza. An Austrian study²⁰ has recently shown that his prose is marked by virtually one hundred per cent rhythm, at colon end as well as at sentence end. Every four or five words, the beat recurs: the stressed syllables separated by two, or four, unstressed. According to a famous passage of Cicero’s Brutus (325), there were two types of Asiatica dictio. One type is clear enough: sententiosum et argutum, sententiis non tam gravibus et severis quam concinnis et venustis. To this concinnitas I shall return. The other type is volucre atque incitatum, and ¹⁵ Cic. Orat. 212. ¹⁶ Quint. 12.10.12. ¹⁷ Sen. Con. 9.1.12. ¹⁸ Norden (1898), 141 5. ¹⁹ Ibid., 397 n. 4; 423 7. ²⁰ Hörandner (1981), 76 8.

106

Papers on Quintilian and Ancient Declamation

is marked by admirabilis orationis cursus. Norden diagnosed this sort of impetuous excitement in the Nimrud Dağ inscription. Perhaps he was wrong. But at least that inscription illustrates the bombastic side of Asianism, the side that caused people, according to Quintilian,²¹ to regard Cicero as tumidior. We can see this kind of thing at times even in the snippets preserved by the Elder Seneca from the Greek rhetores. One, indeed, he actually dubs tumidum and inflatum.²² Dorion is paraphrasing Homer, and is, I take it, doing so in the suasoria on Alexander the Great that Seneca is discussing at this point. The Cyclops flings a rock into the sea: ὄρους ὄρος ἀποσπᾶται, ‘mountain is torn from mountain’, and then ‘an island is picked up and thrown’, χειρία βάλλεται νῆσος. We may notice, besides the hyperbole, the cretic rhythms. Alongside this, Seneca placed²³ an epigram of Menestratus ‘when he was describing the huge size of the monsters bred in the Ocean’. He thought it a very corrupt epigram, worse even than Dorion’s; but the Greek is denied to us by the vagaries of the transmission. What is of interest is that Seneca joins to it an epigram of the Latin rhetor Musa on Scylla and Charybdis, a greater portentum even than they: Charybdis ipsius maris naufragium . . . Quid ibi potest esse salvi ubi ipsum mare perit? ‘Charybdis, wrecker of the sea itself . . . What can be safe where even the sea perishes?’ A monstrous sentiment indeed, but one not unrelated to a famous passage of Cicero:²⁴ Quae Charybdis tam vorax? Charybdin dico? Quae si fuit, animal unum fuit: Oceanus, me dius fidius, vix videtur tot res . . . tam cito absorbere potuisse. As with rhythm, so with bombast: both Cicero and the Senecan declaimers illustrate Asianism. Then there is concinnity, a beast with many heads. Let us look at only one, word play. Diocles of Carystos²⁵ said: Ἂν ἐπιτύχῃς, μίαν προσθήσεις ἀριστείαν· ἂν ἀποτύχῃς, τρεῖς ἀριστείας ἀπολέσεις. While μίαν contrasts with τρεῖς, and προσθήσεις with ἀπολέσεις, the antithesis is further pointed by the play of ἐπιτύχῃς against ἀποτύχῃς. And the order of the words is adapted to give rhythm. Again there are parallels in the Latin declaimers and in Cicero. Seneca²⁶ mentions an anonymous rhetor who said: Peribit ergo quod Cicero scripsit, manebit quod Antonius proscripsit? Again the contrasts, peribit/manebit and Cicero/Antonius, are capped by word play: scripsit/proscripsit. In the third section of almost his earliest speech (S. Rosc.), Cicero finds himself playing off ignoscendi ratio against cognoscendi consuetudo. His last published words were: si vivi vicissent qui morte vicerunt.²⁷ It was not without reason that Cassius Severus remarked that the habit of punning, indulged by Pomponius, writer of Atellans, and the mime-composer Laberius, spread to Cicero who ‘said innumerable things in that vein in both speeches and conversation’.²⁸

²¹ Quint. 12.10.12. ²⁴ Cic. Phil. 2.67. ²⁷ Cic. Phil. 14.38.

²² Sen. Suas. 1.12. ²⁵ Sen. Con. 1.8.15. ²⁸ Sen. Con. 7.3.9.

²³ Sen. Suas. 1.13. ²⁶ Sen. Suas. 7.11.

Declamation, Greek and Latin

107

Rhythm, bombast, word play: these Asian characteristics in Cicero were learned by him at the schools of the Greek declaimers who taught him in Rome, and reinforced when he went on his tour of Greece and Asia in his youth. But we can take it further back than that by a long way: to the archpriest of the cult. We do not have much in the way of fragments of Hegesias. But what we have is illuminating. A series of seventeen quotations from the Alexander history is preserved by the mediation of Agatharchides and Photius (FGrH 142). No less than three (6, 7, 14) feature word play. ‘Misfortune has made speechless the place that once spoke most loudly’; ‘we won our name by leaving our city’, with a play on κατελάβομεν and καταλιπόντες; in the third there is play between ἄσπορον and Σπαρτούς. The longest fragment (12), like the shorter, is marked by the familiar rhythms. And, too, it well illustrates the bombast of Asia: ‘In destroying Thebes, Alexander, you have acted as Zeus might were he to cast the moon from the regions of heaven: for, as to the sun, I leave that to Athens. These cities were the twin eyes of Greece, and it is about the other that I am anxious now: one, the city of Thebes, has been struck out.’ The image of the two eyes of Greece has classical antecedent; but the sun and moon go well beyond that. Hegesias, rhetor and sophist, wrote the kind of history of which Polybius was to be so contemptuous. Its extravagances will have fostered the controversiae and suasoriae concerned with Alexander, and may well have been coloured by them in their turn. Hegesias takes us back to the third century  (as it would seem) and to Asia Minor. But declamation goes back further still. Quintilian, in his lost De causis corruptae eloquentiae, confessed²⁹ that he was unsure whether exercises on judicial and deliberative themes were invented by Demetrius of Phaleron. But that they went back to the time of Demetrius, that is to c.325–300 , he had no doubt: fere constat. A similar picture is given by Philostratus.³⁰ We are to think, he says, of two Sophistics. The second is itself ἀρχαία, though not so old as the more philosophical Sophistic that had prevailed in the fifth century . The second one dealt in typed characters, poor men, rich men, heroes, tyrants, and had themes based on history. The inventor of this Sophistic was, according to Philostratus, the orator Aeschines: who went off after his fall to teach rhetoric in Caria and in Rhodes, just off the Asian coast. The time is, again, the end of the fourth century. There is, then, nothing to prevent us imagining the widespread practice of declamation in the Greek east during the Hellenistic centuries. These are years when we are very short of surviving prose texts, and pieces of information come to us fitfully and by chance. P.Hib. 15, from the third century , is a witness from Hegesias’ century to interest in Alexander at school level, for it forms part of a speech advocating action against Macedonia.

²⁹ Cf. Quint. 2.4.42.

³⁰ VS 481.

108

Papers on Quintilian and Ancient Declamation

Another papyrus, P.Berol. inv. 9781, of the same century, gives us some of a speech related to the Leptines case. This is the first sign of a trend familiar in the Second Sophistic, when Demosthenes and his problems formed the basis of countless cases, and when at the same time his speeches were used as a quarry for grand ideas and properly Attic expression. Thus in Sopatros we find the remark³¹ that ‘it is possible to tack on at once τὸ Δημοσθενικόν, that we are born not only for our parents but also for the city.’ The allusion is to a passage of the De corona (205) which Sopatros regards as so well known that he does not bother to cite it more fully. I should speculate that in the Hellenistic period, and especially later, there was class reading of Demosthenes parallel with the composition of declamation. And it may be that this papyrus shows us an infant Atticism, born to counter the Asianism that found such congenial material in the adventures of Alexander. This lost Hellenistic declamation is of wider importance than might appear. Those who trained others or were themselves trained in the schools could be active in the law-courts, and might well write history. We have seen Hegesias at work as a historian. Three more of his fragments (27–9), available to us only in the Latin translation of Rutilius Lupus, appear to be forensic. One (28) has a characteristic play on words, laborem echoing laboriosus. The two others give rein to pronounced pathos, especially fr. 27, where three successive sentences, apparently from an epilogue, start with the word miseremini. Rutilius also preserves for us fragments of other Hellenistic orators. Even the early Charisius, an imitator of Lysias, can be found launching a violent attack on an errant son, with indignant anaphora and polyptoton: pater . . . patrem . . . patris . . . ³² Elsewhere³³ he paints a picture of the res publica begging ‘for the shrines and holy places of the immortal gods, for the monuments of your fathers, for yourselves and your salvation’. Then there is Myron, who exclaims to Fortuna over the plight of a once rich woman, now reduced to a single maid.³⁴ All this forensic emotion is the proper background against which we should set the oratory of Cicero. Quintilian was later emphatic that one of the two points in which Cicero surpassed Demosthenes was his command of emotional appeal (10.1.107). So accomplished was he that it was to Cicero that his colleagues preferred to leave the last speech in the defence of an embattled client. This emotionalism Cicero will have learned in the declamation schools of Rome and Greece as much as from his Roman predecessors. When it came to history, Hegesias was no less concerned to wring the withers of his public. One of his fragments (10) goes: ‘A city that falls victim to the madness of a king is more worthy of pity than is tragedy.’ That sort of approach was still popular centuries later. In one of Sopatros’ themes, Demosthenes is sent to present a crown to Alexander, but returns, his mission ³¹ RG .11.22 Walz [= 1.7.17 Weissenberger]. ³³ Rut. Lup. 2.6. ³⁴ Rut. Lup. 2.1.

³² See Rut. Lup. 1.10.

Declamation, Greek and Latin

109

unaccomplished, when he finds that Thebes has been destroyed. Amid much that is reminiscent of Hegesias, Sopatros makes Demosthenes exclaim: ‘Alas for these evils! How appalling were the deeds I saw that had afflicted Thebes (Θηβῶν δράματα), deeds that surpassed the tragedy of Mount Cithairon itself (τὴν Κιθαιρῶνος αὐτοῦ τραγῳδίαν)!’³⁵ But Hegesias was not the only emotive historian. We all remember the attack launched by Polybius³⁶ against the historian Phylarchus, who, ‘in his enthusiasm to arouse pity in his readers and to make them feel for what he writes, introduces women’s embraces, torn hair, exposed breasts, and the tears and laments of men and women as they are dragged away.’ Polybius ranks Phylarchus with the tragedians rather than with the historians, remarking that ‘the purposes of history and drama are not the same, but diametrically opposed to each other.’³⁷ It is not unreasonable to suppose that in the Hellenistic period emotional history and emotional oratory were blood relations, with, perhaps, a common parent in the declamation school. It was in the schools, certainly, that were elaborated the innumerable loci for the arousal of pity and anger that we find first in the De inventione and the Ad Herennium. Later, Quintilian remarks³⁸ that in the schools ‘we speak more frequently in the role of litigant than in that of advocate: we act the part of the orphan, the shipwrecked mariner, the endangered—and where is the point of assuming these roles unless we take on the appropriate emotions?’ And for the roles that Quintilian mentions the appropriate emotion would surely be pity. Later still, Sopatros relishes themes that allow πάθος. Thus, in the piece on Demosthenes and the destruction of Thebes: ‘The speech also partakes of emotion: for the fact that the city of Dionysus and Heracles now lies low gives the theme πάθος.’³⁹ In this respect, as in so many, declamation fed upon the practices of the courts; and the courts, no doubt, fed upon the practices of the schools. It was a closed circle, perhaps a vicious one. When the curtain goes up on Latin rhetoric, declamation is fully established in Rome. The themes that are so freely mentioned in the De inventione and Ad Herennium may be less exotic and fanciful than those we find later, but they are Greek in origin and presuppose the whole apparatus of declamation. Particularly interesting are two. In one⁴⁰ we have Ajax and Ulysses. Ajax comes to his senses after his madness, realizes what he has done, and falls upon his sword. Ulysses appears on the scene, and, seeing the corpse, pulls out the bloodstained weapon. Teucer comes along, and charges Ulysses with murder. I shall return to this theme later, but we should meanwhile notice that the story is not merely mythological: it is fictitious mythology, unknown to the tragedians. A rhetor, and a Greek one, has clothed the bare bones of a familiar conjectural theme, that of the corpse found in a lonely spot, giving the ³⁵ RG .210.21 W. [= 35.3.79 We.]. ³⁶ Plb. 2.56.7. ³⁸ Quint. 6.2.36. ³⁹ RG .205.10 W. [= 35.1.3 We.].

³⁷ Plb. 2.56.11. ⁴⁰ Rhet. Her. 1.18.

110

Papers on Quintilian and Ancient Declamation

characters names that will help the declaimer in his characterization. Another striking theme, from the De inventione,⁴¹ tells the story of Alexander, tyrant of Pherae, killed in bed by his wife Thebe. As her reward for this tyrannicide, Thebe demands the son she had by the tyrant. In reality, the tyrant was killed by his brothers-in-law, and there was no son. A rhetor has here given historical names to the tyrant and tyrannicide of his theme; or, if you like, falsified history to make a more exciting theme; concessum est rhetoribus ementiri in historiis, ut aliquid dicere possint argutius.⁴² Fictitious mythology is matched by fictitious history: another theme that we can follow down the centuries. In the Minor Declamations (323) Alexander, besieging Athens, burned a temple. A plague ensues. Alexander is told that it will not end until he restores the temple. There can be no truth in this story, for Alexander did not lay siege to Athens. No matter. It would not be surprising if Cicero derived the material for his De inventione from a Greek teacher working in Rome. Certainly he himself declaimed in the east: Plutarch⁴³ tells us how he spoke (in Greek) before a teacher in Rhodes, Apollonius, son of Molon. If we are to believe the details of the story, Apollonius was lost in amazement at Cicero’s performance and said: ‘Greece and her fate I am sorry for. The only glories that were left to us were our culture and our eloquence. Now I see that they have been taken over, in your person, by Rome.’ There will have been Greek rhetores in Rome throughout the period from 160, when Suetonius⁴⁴ proves the existence of the art, to the Elder Seneca and, of course, beyond. Latin writers naturally showed especial interest in the phenomenon of Plotius Gallus, first Latin teacher of rhetoric. But the very letter of Cicero⁴⁵ which witnesses to Plotius’ popularity remarks that he himself had been prevented from attending Plotius’ classes by the authority of those who ‘thought that intellects were better fostered by Greek exercises’. And if, as Quintilian⁴⁶ appears to imply, Plotius taught by means of declamation, it is hardly likely that his Greek predecessors and contemporaries in Rome did not do so. The young Cicero who rhapsodizes about the horrors of the culleus in the speech for Roscius of Ameria⁴⁷ will have often spoken about parricide in Rome before, but behind the walls of the declamation school. It is true that these Greeks only come fully into focus in the pages of the Elder Seneca from which I started. Their activities were largely obscured by Roman nationalism, operating in both of the two main sources. The Elder Seneca’s account cannot begin to claim credence. For him⁴⁸ declamation is not only a new word but a new practice: ‘The name has emerged recently, the practice itself having become popular not long ago: the thing was born after me—that is why it is easy for me to have known it from its cradle.’ Seneca was ⁴¹ Cic. Inv. 2.144. ⁴⁵ Suet. Rhet. 26. ⁴⁸ Sen. Con. 1.pr.12.

⁴² Cic. Brut. 42. ⁴⁶ Quint. 2.4.42.

⁴³ Plut. Cic. 4.6 7. ⁴⁷ Cic. S. Rosc. 71 2.

⁴⁴ Suet. Rhet. 25.

Declamation, Greek and Latin

111

born around 55 , and what he says of the rise of declamation cannot be true even of Rome. However he reaches his false conclusion, it is striking that his discussion pays no attention to the Greek background—by ignorance, perhaps, rather than by malice. Suetonius is equally, though perhaps more consciously, one-sided. In discussing rhetoric, he makes no mention of Greek teachers. Once the art has arrived from Greece, he is concerned⁴⁹ merely to list the Roman exponents of note, starting with Plotius Gallus, who flourished as late as 92. There are only five names to take us down to the time of Augustus. The Elder Seneca, we may recall, named thirty-six Greek rhetores, and the assumption is that they were all operating in Rome. * * * Declamation was designed to train pupils to speak, in court or assembly, and speeches consist of more than clever epigrams. The Elder Seneca, no doubt inadvertently, gives the impression that epigram was of major importance in the exercises of his own day. His sons were keen on point, and he pandered to their keenness. But, as his exposition of the division of each case shows, he was perfectly well aware that the organization of the speech, and particularly of its argument, was crucial. It is very unfortunate that the only speech which (as it would seem) Seneca set out to give us in full, Latro’s in Con. 2.7, is stopped in full flood by a lacuna in the manuscripts. It is sad, too, that the division of this speech, if Seneca provided one, is missing too; and that, even if the speech were preserved as a whole, it would not be a very good instance of ordered argument. The case is one of coniectura: a husband accuses his wife of adultery ‘on suspicion’. And Latro indulges in a good deal of not particularly structured abuse, arising especially from examination of the wording of the will which is the prime evidence for the alleged adultery. Yet even here the speaker is concerned to make the main headings of his speech clear. The proem finishes with a resounding epigram (1): the woman has received so much from her lover ‘that even after she has been punished adultery will still have made a profit’. Then we turn to the narration: ‘I know what instructions I gave my wife on my departure; you must apply to rumour for the rest, for the story of how a handsome young man . . . ’ and so on (2). But the story is very short. Latro is soon rounding off this section: ‘I come here for no other purpose than to complain of my fortune: you know my case better than I.’ And so we arrive at the argumentation, starting off with what the Greek rhetoricians would call an ἀντίθεσις: ‘There is no reason for her to say: “I couldn’t help it”’—and Latro proceeds to argue (3) that she could have helped it. The objection is slipped in neatly and unobtrusively, and refuted colourfully. The speech has its shape, but Latro makes his transitions so skilfully that we do not feel it is being forced upon our attention.[⁵⁰] ⁴⁹ Suet. Rhet. 26 30. ⁵⁰ [I discuss Sen. Con. 2.7 in a little more detail in Winterbottom (1980b), 27 and 89 90.]

112

Papers on Quintilian and Ancient Declamation

It was possible to do these things more blatantly. Thus, even in the highly elaborate Major Declamations, we find coy allusion to the technical terms. In Decl. mai. 13 the end of the narration is marked by the words: ut ipse tristem finiam expositionem (6 [271.12 Håkanson]). There is a very obvious partitio: ‘As far as I could see, the accuser divided his case into two quaestiones: an damnum sit, et an iniuria datum’ (7 [272.19 H.]), words that could come straight from a division in the Elder Seneca, as well as recalling such a division as Cicero’s in the Murena (11: Intellego, iudices, tris totius accusationis partis fuisse). Later, the declaimer marks the end of his argumentation with the words (15 [282.7 H.]): ‘I realize that your good sense requires nothing more to be said about the case’; then, making the transition to the epilogue: ‘and that your good faith and sense of values needs no exhortation to give a true judgement.’ Why any epilogue at all, we wonder. The answer is given (15 [282.11 H.]): ‘Perhaps there may be larger scope for emotional appeal (adfectus materia).’ Even in this sophisticated piece, the teacher is present, showing his pupils how a good declamation should be structured. This is something that goes back to the beginnings of rhetoric. By the time of Plato’s Phaedrus the complicated subdivisions of a speech, with their grandiose technical terms, had become matter for jest.⁵¹ Earlier, the sophistic declamation of Gorgias had shown them in action, if in a less developed form. The Palamedes prefigures the declamation of the Second Sophistic. Odysseus accuses Palamedes on a charge of betraying Greece to the barbarians. There is no narration, for Palamedes glides smoothly from the proem to the foundation of his defence (5): a twofold defence, expressed with Gorgianic antimetabole: ‘If I had wished I should not have been able to embark upon such a course; and if I had been able I should not have wished.’ The epilogue is clearly marked by: ‘But for the rest I have to speak to you concerning yourselves—and with that said I shall have finished my defence’ (33). And in what follows there is conscious allusion to the practice of Athenian speakers in introducing appeals to pity and recapitulation of the facts in their epilogues: both ploys which Palamedes here rejects on the grounds that he is speaking before the first of the Greeks, who do not need such things. But let us return to the argumentation section of the Palamedes. The signposts that we have seen marking the parts of the speech as a whole are apparent within this section also. Thus, to take the first part of the doublebarrelled divisio: ‘I should not have been able if I had wished.’ At 6 Gorgias launches into this with the words: ‘I shall first of all approach the λόγος that I am unable to do this.’ At 12 he sums it up: ‘Altogether, then, and in every way it was impossible for me to do all of these things.’ The second argument is similarly marked off, and at 21 the two are summarized together: ‘What I have ⁵¹ Plat. Phaedr. 265e 7d.

Declamation, Greek and Latin

113

said shows that I could not have betrayed Greece and that I would not have wished to.’ This rather laboured use of signposting is parodied by Plato in the Symposium, significantly enough in the speech of the Gorgianist Agathon (e.g. 196b). But parody could not kill something so useful. The technique is still going strong centuries later. In ‘Quint.’ Decl. min. 315 the law set is that a hero must kill a deserter with his own hands. By the ill luck that afflicts many inhabitants of what Mr Russell has excellently dubbed Sophistopolis, our hero has a son who has deserted. His prize, he hopes, will be permission not to have to administer the death blow in person. In a mixture of sermo and declamation proper, we are taken through the various arguments. The law says that a hero should have any prize he chooses. That may include something that is against another law. If there is a conflict, it is preferable to preserve the law that honours a hero. Anyway, there is no real conflict, for many things can happen to prevent a hero killing a deserter. And the lawgiver did not envisage a situation of the kind that has arisen now. ‘The law spoke of a deserter and a hero. Everything has changed: it was a son who deserted.’ On this epigram the argumentation ends. But before turning to other matters, the hero summarizes: ‘Therefore, I have so far said this: any prize at all is my due. It is my due even if a law is thereby broken. My law is more in the interests of the state. In any case, I am not claiming something against the law, for much can happen to stop a hero killing. I am not claiming something against the law, for it cannot be that the law intended that a son should be killed by his own father.’ Such summaries, and various kinds of signposting of a rather more subtle nature, are common in the Minor Declamations. Their purpose is pedagogical, as it was in the Palamedes. In model speeches of this kind, pupils are being shown how to give declamations, and eventually orations, that have a shape, that advance from argument to argument in logical sequence: and to keep the shape of the whole speech in their heads at the same time as they speak a part. Even in the most epideictic declamations, the scholastic purpose is rarely forgotten altogether. All this shows clearly one virtue of the whole practice of declamation. A master could and did make his class read the speeches of a Demosthenes or a Cicero, pointing out in them the qualities and techniques that rhetorical precept recommended. Thus Quintilian says: ‘It is from these speeches that we get models of everything that I have been giving instruction in—and more effective these models are too than the teachings of the textbooks . . . What the teacher recommended the orator makes clear in action.’⁵² But real published speeches did not always illustrate, with appropriate preciseness and conciseness, the point a teacher wanted to get over. On the other hand, his own model declamation could illustrate as much or as little as he wished it to. All this was particularly true of the intricacies of the stasis-system.

⁵² Quint. 10.1.15.

114

Papers on Quintilian and Ancient Declamation

And here we find another way in which the Palamedes prefigured the methods of the Hellenistic and post-Hellenistic declamation. The theme of the Palamedes is chosen with an eye to the particular interests of rhetoricians in the fifth century. There is no knock-down evidence for or against Palamedes. Both prosecution and defence can only argue from what is probable, τὰ εἰκότα. And the word is actually used in 9. We see here what would later on be called a case calling for στοχασμός or coniectura: did Palamedes betray the Greeks? This is why, for instance, Palamedes says (15): ‘As a trustworthy witness I shall bring forward my past life.’ That is the argument ex ante acta vita, later a key point in the developed argumentation from conjecture. A direct line connects these words of Palamedes with, say, these of Cicero speaking for Sulla (79): ‘it is essential that the cases of honest men be weighed not in accordance with the fickleness of witnesses: instead, the life of each individual is his witness.’ The sophists took over the practices of Athenian courts, and in turn affected that practice: an interweaving of precept and forensic reality is a constant theme down the centuries. So too with the Tetralogies attributed to Antiphon: all three cases are designed to allow argument on probabilities. And this argument goes on in, as it were, a charmed circle or a closed room. The characters are fictional, the cases known only from the simple themes assumed to govern them. If a rhetorician had tried to illustrate his topics for argumentation from a real speech, he would have been hampered by too much extraneous material (think of the complications of Antiphon’s speech De caede Herodis, for example). If he or his pupils had tried to argue a real-life case, they would have had to take account of the awkwardnesses of fact. Fictional themes and fictional characters allowed concentration on what mattered. We can see now how snugly the Ad Herennium case of Ajax and Ulysses fits into this tradition: mythology, and fictional mythology at that, used in the service of coniectura. That brings us again to the Hellenistic period. We should not think of declamation as gaining a foothold, and then becoming popular, because, or only because, real oratory declined after the period of Alexander. Rather it blossomed because it was an essential tool in the pedagogue’s struggle to demonstrate the complexities of the system he had erected, and to train his pupils in it. It is no accident that the Hermagorean De inventione of the youthful Cicero so often illustrates its points by reference to declamation themes: that is how Hermagoras used declamation. Again, we should not suppose that the themes became more fictional because of some deep-seated urge to escapism in the large, cold world after Alexander. Why, for instance, are declaimers so much concerned with the figures of the Hero or the Tyrannicide? The answer will have little to do with the military or political history of the period, or with the psyches of Hellenistic Greeks. It is rather that these personages could plausibly be allowed a prize consisting of whatever they wished to ask in the way of reward for their services. And from this

Declamation, Greek and Latin

115

moment of arbitrary power could arise imaginary cases by the score on which the budding advocate could cut his teeth. Similarly, we should be unwise to assume anything from declaimers’ laws to have any particular relationship to reality. The rhetores made up laws to fit in with the cases that they made up too. For this reason, a good deal of S. F. Bonner’s justly praised and highly useful book on Roman Declamation is in a sense beside the point. No declamatory law can be used as evidence of Greek, let alone Roman, legal practice: it is not, of course, harmful, but it is not very helpful either, to trace the relationship between declamation law and real law. We should take this back to the Tetralogies. Scholars have found difficulty in reconciling the homicide laws presupposed there with the code of Draco. Naturally: they cannot be reconciled with it. They are made up to suit the cases that ‘Antiphon’ wished to argue. The intimate connection between stasis and declamation continued. In a sadly unpublished Tübingen dissertation Professor Joachim Dingel[⁵³] has shown how the Minor Declamations fit into this pattern. It is true that the sermones that introduce a number of the pieces are so elementary that they do not go into the technical details of stasis at all. But the speeches themselves illustrate the arguments that the Hermagorean system, as expounded primarily in the lnstitutio oratoria of Quintilian, regarded as appropriate to the different types of case. Thus in Decl. min. 281 we find successively arguments from past life, motive, place, means, and opportunity, the heads commonly assigned to cases of coniectura. One may imagine the teacher pointing out these correspondences when providing his enarratio of his own declamation. Far less tractable the published speeches of the great orators: though the scholiasts made manful attempts to bring the argumentation of Demosthenes under control. Nor, in this area, would rhetorical precept much influence reality. If Tacitus⁵⁴ says that back in Cicero’s day ‘everything prescribed in the dry books of Hermagoras and Apollodorus was reverenced’, that is malicious and unjustified inference from the precepts found in Cicero’s own works. More realistic, Cicero himself: in the Brutus he regards it on two occasions⁵⁵ as worthy of mention that a Roman orator was, quite exceptionally, versed in these mysteries. Cicero himself was of course versed in them. But it is significant that the speech where Cicero’s practice most corresponds to his theory, the Pro Milone, is the speech where he had most temptation to make his written speech diverge from the one he spoke: for, it would seem, the spoken speech in this case was short and undistinguished. Greek declamation, however, provides the crucial example of the linking of stasis and declamation. The Minor Declamations look like a rhetor’s Nachlass, ⁵³ [Professor Dingel’s dissertation was later published as Dingel (1988). For the sermones see A.23 below.] ⁵⁴ Tac. Dial. 19.3. ⁵⁵ Cic. Brut. 263; 271.

116

Papers on Quintilian and Ancient Declamation

disorganized and perhaps not published in his lifetime. The Διαίρεσις ζητημάτων (i.e. Divisio controversiarum) of Sopatros, on the other hand, is a conscious production. The declamations are grouped according to stasis, starting with conjecture and proceeding, more or less, according to the order laid down by Hermogenes, whose system had after the third century  taken over from that of Hermagoras. At the start of each new stasis, Sopatros normally summarizes the doctrine to which he adheres, listing the heads of the appropriate series of arguments. Again, Hermogenes is the main inspiration. Sopatros does not give complete declamations; rather, he illustrates, in snippets of widely varying length, the sort of way in which you could, more or less elegantly, put the arguments that the theory laid down for you. As we have seen, he is interested in themes with emotive content, and he accordingly pays close attention to proem and epilogue. But there is no doubt that argument is his main concern. There are occasions when he evinces awareness that the framework might seem too schematic, the bones—as it were—sticking through the flesh. He remarks, for example, ways in which smooth transition may be made from one heading to another.⁵⁶ But his overriding concern is with a speech sternly formulated according to a predetermined scheme. He demonstrates to his pupils how to make speeches of this kind, and in making them they train themselves to apply the Hermogenean stasis system. Sopatros is perhaps an extreme case. Choricius’ declamations are a different matter. Not merely are the bones invisible beneath the skin: the skin itself is lit up by the bloom of his talent for characterization. Ἦθος is all. But even here Hermogenes is not forgotten. In the twelfth declamation the scholiast marks in the margin not only the main parts of the speech but also the articulation of the argumentation, with the technical terms προβολή, ἀνθορισμός, πηλικότης [see A.10 below, p. 146 n. 61]. I think that further investigation would show that at least in some degree Choricius constructs his speeches to conform to, and to illustrate in action, the precepts of Hermogenes. * * * I shall end with style. We have seen how, in the Elder Seneca, the Greek declaimers set a pattern of point and sometimes of bombast for their Latin counterparts. Less familiarly, there is a striking point of resemblance between the Greek of Sopatros and the Latin of the Minor Declamations. Leo⁵⁷ drew attention in the latter to what he called ‘eine auffallende stilistische Eigenheit’. Pairs of parallel clauses, often with anaphora, repeat or elaborate a thought in more or less varied form. The phenomenon is everywhere. One example will suffice. Sed quis tandem me innocentem, quis dignum conversatione vestra putaret si aliud optarem? (anaphora; the second clause adds a little to the first,

⁵⁶ See e.g. RG .13.1 and 58.4 W. [= 1.8.10 and 9.4.3 We.].

⁵⁷ (1960), .250.

Declamation, Greek and Latin

117

at least emotionally). Est pro me. Non enim et tum ideo tantum defensus sum ut in civitate remanerem, ut essem vobiscum (anaphora, with identical content), sed ut probarem me innocentem, nihil contra patriam meam esse molitum (the second clause provides some emotive addition). Et nunc non est satis si mihi ignovistis, si me civem vestrum publica necessitas fecit (anaphora; the second clause provides a reason for the first, as well as restating it). Ego vero si merito damnatus sum, si quid tale quale contra me pronuntiatum est feci, non utor isto beneficio (anaphora; identical content).⁵⁸ The trick is over-used in the Minor Declamations, with too little regard to the context and a certain insensitivity to monotony of effect. But a defence could be essayed, and one that brings us back to the same pedagogic motive that I traced earlier. Like the so-called Isidorean style of the later period, it showed the pupil how an idea could be expressed in different ways, how flexible language could be. It widened vocabulary; it built up hexis. And, like a crutch cast away when the leg is healed, it could be discarded by the mature orator. At any rate, it is a style that I have found only occasional parallels for in other Latin literature: and always in authors influenced by declamation. But the striking parallel is the Greek of Sopatros. In a typical passage,⁵⁹ a victorious general, accused of having distributed among his soldiers money paid to him to betray his city, says: ‘I said to myself: “ . . . I myself will find the means, I myself will provide the food.” Thinking this, seeing this, I took the money and distributed it among my poverty-stricken soldiers, saying: “The enemy is hiring you against himself, the foe is turning you against himself by money.” Thinking this, saying this, I provided the money. They took it, went to war and conquered. After the victory and the success, instead of crowning me, instead of providing some reward, they arraign me as a traitor, and turn my good deed into material for a charge’ (φήσας πρὸς ἐμαυτόν· ‘ . . . αὐτὸς εὑρήσω πόρον, αὐτὸς πορίσω τροφάς.’ Ταῦτ’ ἐννοήσας, ταῦτα ἰδὼν ἐδεξάμην τὰ χρήματα, διέδωκα πενομένοις στρατιώταις, φήσας· ‘Oἱ πολέμιοι καθ’ ἑαυτῶν ὑμᾶς μισθοδοτοῦσιν, οἱ δυσμενεῖς ὑμᾶς καθ’ ἑαυτῶν προτρέπονται χρήμασι.’ Ταῦτ’ ἐννοήσας, ταῦτ’ εἰπὼν παρέσχον τὰ χρήματα· ἐστρατεύσαντο λαβόντες καὶ νενικήκασι καὶ μετὰ τὴν νίκην καὶ τὸ κατόρθωμα ἀντὶ τοῦ στεφανοῦν, ἀντὶ τοῦ παρασχεῖν ἀμοιβήν, ὡς προδότην εὐθύνουσι καὶ τὴν εὐεργεσίαν εἰς ἔγκλημα ἄγουσιν). The trick is used five times. In each case, there is asyndeton: in four cases there is anaphora, while in the fifth πολέμιοι is varied by δυσμενεῖς. Even outside the strict examples, there is a tendency to double: thus τὴν νίκην is backed up by τὸ κατόρθωμα. A figure has become a style. What are we to make of this parallel between the Minor Declamations and Sopatros? Does it point to a later date for the Latin text than most people would think of assigning it? Does it point away from Quintilianic authorship?

⁵⁸ ‘Quint.’ Decl. min. 266.7 8.

⁵⁹ RG .201.13 W. [= 33.1.7 We.].

118

Papers on Quintilian and Ancient Declamation

We should be cautious in trying to give answers to these questions. The trick is found in one of the Latin declaimers cited by the Elder Seneca,⁶⁰ as well as in the much later Major Declamations.⁶¹ On the Greek side, it surfaces in Himerius and Libanius and Dio Cassius, as well as in Sopatros. As to Quintilian, there is even an instance in one of the declamatory excerpts in the Institutio.⁶² We seem to be in the presence of something timeless, that is taken to extremes in the Minor Declamations and in Sopatros, works which could, for all that, be centuries apart. The fact is, of course, that declamation was a highly traditional matter. What is more, it was constantly entangled with the practice of the courts and interacting with other kinds of literature. In the end the cross-currents may be impossible to distinguish. We all know the agony of Medea:⁶³ Νῦν ποῖ τράπωμαι; Πότερα πρὸς πατρὸς δόμους, / οὓς σοι προδοῦσα καὶ πάτραν ἀφικόμην; If we think that there is something declamatory about this, we are right. Gorgias said much the same in the Palamedes (20): Ποῖ γὰρ τραπέσθαι με χρῆν; Πότερον εἰς τὴν Ἑλλάδα; Δίκην δώσοντα τοῖς ἠδικημένοις; But then so did Antiphon⁶⁴ and so did Demosthenes⁶⁵ in court. Do we find a certain rhetoric in Catullus’ Νam quo me referam? Quali spe perdita nitor? / Idaeosne petam montes?⁶⁶ and the rest? Yes. He will have known not only his Ennius but his Gaius Gracchus:⁶⁷ Quo me miser conferam? Quo vertam? In Capitoliumne? Cicero used the idea in the courts,⁶⁸ but the Pseudo-Sallust’s invective against him (1) brings us back (it would seem) to the schools again. For both Greeks and Romans, declamation was not something apart. It affected, and was affected by, the world of law and literature for which it was a training. [On Greek declaimers in the Elder Seneca, Citti (2007) is now fundamental; on ‘meta-rhetorical’ clues within declamations see Stramaglia (2016). For the relation of this paper to A.10, see the note added at the end of that piece.]

⁶⁰ Sen. Con. 9.3.2. ⁶³ Eur. Med. 502 3. ⁶⁶ Catull. 64.177 8.

⁶¹ e.g. 19.3 (374.9 H.). ⁶² Quint. 7.1.48. ⁶⁴ Antiph. Myst. 148. ⁶⁵ Demosth. 28.18. ⁶⁷ Ap. Cic. De orat. 3.214. ⁶⁸ Cf. e.g. Cic. Mur. 88.

8 Quintilian and Declamation* The first volume of Professor Cousin’s invaluable Études sur Quintilien concludes¹ with an assessment of Quintilian’s ‘orientation rhétorique’. My own essay, which I offer in memory of Cousin, might be thought of as an appendix to that chapter; it is concerned with Quintilian’s ‘orientation déclamatoire’.² Quintilian was a practising teacher of rhetoric, who employed as a major tool the traditional exercise of declamation.³ He was also a practising barrister. His Institutio oratoria witnesses to his sense of courtroom realities, but also to the abiding utility and fascination of the procedures of the schools. These opposites are reconciled in his dictum that declamatio, quoniam est iudiciorum consiliorumque imago, similis esse debet veritati (2.10.12);⁴ and his order of priorities is revealed in the doctrine that declamation was forensium actionum meditatio (4.2.29).⁵ No doubt it had its epideictic side (2.10.12), and even quite unreal themes were properly conceded to the young (2.10.5–6).⁶ [From Hommages à Jean Cousin (Paris, 1983), 225 35] * All references are to the Institutio oratoria unless otherwise stated. ¹ Cousin (1935), 771 93. ² There is a characteristically informative sketch in Bonner (1949), 80 2. ³ He never suggests abandoning it, though he mentions possible reforms of detail in 2.10.9 (for jokes see also 6.3.15). ⁴ Cf. 2.10.4: Sint ergo et ipsae materiae quae fingentur quam simillimae veritati (reminiscent of Hor. Ars 338: Ficta voluptatis causa sint proxima veris); 5.12.17: illa vera imagine orandi; 5.12.22: adulescens . . . componat se ad imitationem veritatis; 10.5.14: declamations are useful si modo sunt ad veritatem accommodatae et orationibus similes; 12.11.15: imaginem veri discriminis. Else where note ‘Quint.’ Decl. min. 338.5: scholastica controversia complectitur quidquid in foro fieri potest; Sen. Con. 9.pr.4: Non est autem utilis exercitatio nisi quae operi simillima est in quod exercet; itaque durior solet esse vero certamine; [Dion. Hal.] Ars rhet. p. 372.1 Us. Rad.: πρὸς ἀλήθειαν ῥητορεύειν. Critics pointed out that this did not work out in practice; thus Petr. 1.3: nihil ex his quae in usu habemus aut audiunt aut vident (sc. adulescentuli in school); Tac. Dial. 35.4: materiae abhorrenti a veritate. Contrast with all this Sen. Con. 10.pr.12: Nihil est indecentius quam ubi scholasticus forum, quod non novit, imitatur. ⁵ Cf. 2.10.4: actiones in quarum exercitationem reperta est (sc. declamatio); 2.10.8, cited below; 5.12.17: declamationes, quibus ad pugnam forensem . . . exerceri solebamus; 5.13.44: exercitatio nibus quae foro praeparant. ⁶ Compare too 2.4.14.

120

Papers on Quintilian and Ancient Declamation

But declamation was to be no end in itself. It might still have something to offer to the grown-up; but Quintilian thinks not of the professional scholasticus but of the consummatus ac iam in foro clarus (10.5.14), finding in declamation some relaxation from forensic cares:⁷ he is merely scornful of anyone who spent longer than he needed in the schools (12.11.15). Declamation, then, is a mere handmaiden of oratory. We shall see later the role Quintilian allows it. Meanwhile, we may pause on his criticism of declamation as practised by others: his polemic, here as elsewhere, is instructive. In the wrong hands, the hands of those whose licentia atque inscitia were among the prime reasons for the corruption of eloquence (2.10.3; cf. Petr. 2.2), declamation became unduly epideictic. This arose directly from neglect of the link with the courts: nam si foro non praeparat, aut scaenicae ostentationi⁸ aut furiosae vociferationi simillimum est (2.10.8). Quintilian compares declamation of this kind with the ματαιοτεχνία⁹ of the chick-pea virtuosos derided by Alexander (2.20.3–4). Much like the old sophists,¹⁰ such speakers might ask the audience to supply the first word for their declamation (10.7.21: frivolum ac scaenicum). So much for ostentatio. Furiosa vociferatio¹¹ is doubly suggestive. These people were noisy¹² and over-demonstrative (3.8.59: exclamandum; 7.1.44: tumultu et clamore); abusive,¹³ too (3.8.69). But that suggested madness (3.8.59: initio furioso).¹⁴ Quintilian does not mean that quite literally.¹⁵ But in a sense these people did lack reason; just as they did not see the ratio for declamation (2.10.7), they did not give oratory any reasoned basis (cf. 2.11.7: nulla est ratio dicendi). They thought it enough to speak impetu . . . et viribus . . . ; neque enim opus esse probatione aut dispositione in rebus fictis (2.11.3).¹⁶ More

⁷ It would have the same soothing effect as poetry; cf. 10.1.27: velut attrita cotidiano actu forensi ingenia . . . reparantur. ⁸ Cf. 4.3.2: ostentatione declamatoria; Petr. 126.6: histrio scaenae ostentatione traductus. For the Elder Seneca’s tendency to contrast ostentatio and exercitatio see Bonner (1949), 72. ⁹ Cf. [Dion. Hal.] Ars rhet. p. 372.8 Us. Rad.: (ἐκφράσεις) ματαία ἐπίδειξις καὶ λόγου ἀνά λωμα; also p. 372.22: μάταιον μῆκος λόγων. ¹⁰ Cf. Philostr. VS 482 (of Gorgias): παρελθὼν . . . ἐς τὸ Ἀθήνῃσι θέατρον ἐθάρρησεν εἰπεῖν ‘Προβάλλετε’ καὶ τὸ κινδύνευμα τοῦτο πρῶτος ἀνεφθέγξατο. For the modern sophists, see Plin. Ep. 2.3.2; Luc. Rh. pr. 18: ἐπειδὰν . . . οἱ παρόντες ὑποβάλωσί τινας ὑποθέσεις καὶ ἀφορμὰς τῶν λόγων[; Lib. Or. 59.4]. ¹¹ The link is made again in 11.3.45: ne dicamus omnia clamose, quod insanum est. Cf. also Petr. 1.1: Num alio genere furiarum declamatores inquietantur, qui clamant . . . (taken up by Agamemnon at 3.2: cum insanientibus furere). ¹² Cf. 2.11.2: exclamaverunt; 2.12.9: clamant ubique. 2.11 12 concern a particular class of ‘naturalist’ declaimers and orators, who rejected ars altogether; but they fit naturally into the wider picture. For noisy orators in court, see 4.2.37: tumultu et vociferatione. ¹³ That extended to the courts too: see 2.12.4. Cf. also [Dion. Hal.] Ars rhet. p. 368.11 Us. Rad. for criticism of λοιδορία in declamation. ¹⁴ Compare the naturalist orators motu capitis furentes (2.12.9). ¹⁵ See Winterbottom (1964b), 95 n. 19 [= A.1 above, p. 11 n. 83]. ¹⁶ Cf. 2.12.6: illud quaestionum et argumentorum apud corrupta iudicia frigus evitant (the ‘naturalists’ in court). Also Sen. Con. 9.pr.1 2, cited in n. 20.

Quintilian and Declamation

121

generally, they did not believe that eloquence required precepts of the kind that Quintilian’s book abounds in; the vulgaris scholarum exercitatio (2.11.1) was enough.¹⁷ Instead of sober argument, such declaimers over-valued elocutio (5.12.18).¹⁸ They had a special liking for sententiae, preferably daring or obscure (2.11.3: sententiis grandibus, quarum optima quaeque a periculo petatur; 7.1.44: sententiae . . . praecipites vel obscurae (nam ea nunc virtus est)).¹⁹ Epigrams were popular with audiences (2.11.3: cuius rei gratia plenum sit auditorium; 5.13.42: populares), and it was pleasure that these declaimers were in business to purvey.²⁰ Depraved pleasure at that, thought Quintilian.²¹ He analyses this tendency in a scathing passage, comparing it with the excesses of sexual libido (5.12.17–21). Declamation must in any case be deficient in blood and strength by comparison with oratory (10.2.12); these declamations, ad solam compositae voluptatem, lacked masculinity (5.12.17: nervis carent; cf. Petr. 2.2: enervaretur). Quintilian’s own principles can be deduced from this polemic. And they naturally cohere with his precepts for oratory itself: the overall impression of order, the stress placed on argumentation (5.pr.5) and the reservations about elocutio (8.pr.32–3, esp. iucundam non deformi voluptate), including sententiae (8.5.25–34), and the recommendation of moderation in delivery. But we can turn too to the Quintilianic Minor Declamations, product of his school if not his own pen, for confirmation. The message of the so-called sermones, the Master’s words of advice, is entirely concordant. They are concerned with the division, ‘colour’, and tone of the pieces, not with their style. They too preach the primacy of argument (270; 343), and regard the case as more

¹⁷ They would be among the many who believed that declamation was enough ad formandam eloquentiam (2.10.2). ¹⁸ See 3.8.58 for effusiorem, ut ipsi vocant, cultum in suasoriae. Quintilian returns to the attack on a wider front in 8.pr.18 ff.; cf. also Petr. 1.3: mellitos verborum globulos et omnia dicta factaque quasi papavere et sesamo sparsa. In the schools this led to incredibili verborum fastidio, another symptom of the divorce from reality (8.3.23; cf. 2.10.9). It had been different in Quintilian’s youth (8.3.22), though much earlier we notice that Albucius’ idiotismi were the result of his desire not to seem scholasticus (Sen. Con. 7.pr.3 4; cf. 4.pr.9). There are traces of such fastidium even in the sensible [Dion. Hal.] Ars rhet. p. 365.3 ff. Us. Rad. ¹⁹ For obscurity in the schools see also 8.2.20 (cf. 8.5.23). The sententia criticized in 8.5.22 has, significantly, a family resemblance to Cic. S. Rosc. 72 (declamatory). ²⁰ 5.12.17: declamationes . . . ad solam compositae voluptatem. Cf. Sen. Con. 9.pr.1 2: Qui declamationem parat, scribit non ut vincat sed ut placeat. Omnia itaque lenocinia [ita] conquirit; argumentationes, quia molestae sunt et minimum habent floris, relinquit; sententiis, explicatio nibus audientis delinire contentus est . . . Sequitur autem hoc usque in forum declamatores vitium, ut necessaria deserant dum speciosa sectantur. ²¹ Prava voluptas, of the new style: 2.5.22. Cf. 2.12.6: the ‘naturalists’ in court nihil aliud quam quod vel pravis voluptatibus aures adsistentium permulceat quaerunt (they did not even trouble to please the judges: adsistentes is contrasted with iudices at 9.2.76; cf. Cassius Severus in Sen. Con. 3.pr.12: adsuevi non auditorem spectare sed iudicem, though the contrast there is of judge with the audience in the schools). One might end by thinking the quest for popularis adsensio the reason for the downfall of oratory (Sen. Ep. 102.16); a gruesome picture in Plin. Ep. 2.14.6 8.

122

Papers on Quintilian and Ancient Declamation

important than the pleasure of the audience (316: no apology for not providing loci communes). The Minor Declamations themselves illustrate these principles and others preached in the Institutio. This is not the place to go into detail.²² But it is worth looking briefly at the declamation extracts, taken from his files or specially composed, which Quintilian from time to time uses as examples in the Institutio.²³ They at least must give us sure indications of his approved practice in declamation.²⁴ In 4.2.68 ff. Quintilian gives a series of ‘extracts’, at first hardly more than bald statements (68–9) of the argument to be put forward in cases that are obviously unfavourable to the speaker. Here there is no stylistic striving, but it may be noticed that Virginem rapuit, non tamen optio patri dabitur is the obverse of an argument familiar in declamation, that the father has a dominant role in such choices.²⁵ The extracts in 70–1 are of greater interest. Familiar lines of defence²⁶ are ingeniously worked into apparent confessions; and in each case the excuse is given in a sentence of some elaboration, only to be destroyed by brisk declamatory cola. Finally, a more complex treatment in 73–4, passing from oratio recta to oratio obliqua. There are familiar elements here too: the harsh father, the corrupting influence of other youths, the inability, when it comes to the point, to kill a father;²⁷ and the style shows an agreeable economy: parricidium obicitur iuvenibus quorum pater vivit— point employed not for its own sake but, typically, to forward the argument. In 5.10.104 we again have an unadorned argument: one that reappears, in different phrasing, in the Minor Declamations (284.4). There follows in 111–18 a long discussion of the treatment of a theme concerning Alexander and the Thebans. There are no extracts here, and the whole is rather to be compared with a sermo.²⁸ The use of statements of what is agreed to lead up to determination of what is in question (111–12) follows a precept of Quintilian (7.1.5 ff.) which is stated in a less sophisticated form in the sermo to Decl. ²² I hope to publish soon a new edition, with commentary. [See now Winterbottom (1984).] ²³ If from his files, it is surprising that there is no overlap with the Minor Declamations, supposing these are Quintilian’s work: though it is true that the first third of the corpus is lost. ²⁴ The master declaimed as a model for his pupils (2.2.8; 2.4.12; 2.5.16; 2.6.5; also 11.2.39 for a scene from Quintilian’s school). The pupils declaimed too (1.2.23 5 etc.), but in Quintilian’s view they should not do it too frequently; they should not declaim all that they wrote (2.7.1; contrast 5: Aliquando . . . permittendum quae ipsi scripserint dicere). Writing, for them, was the norm. ²⁵ See my notes (Winterbottom (1974b)) on Sen. Con. 3.5 and 8.6; in the Decl. min. e.g. 259. th.: Imperavit filiae dives ut nuptias optaret. ²⁶ For the excuses for rape, cf. e.g. ‘Quint.’ Decl. min. 309.4 (many parallels in declamation and New Comedy). ²⁷ Cf. e.g. ‘Quint.’ Decl. min. 260.28 (harsh fathers); Sen. Con. 10.2.2 (other youths); Decl. min. 281.6 (inability to kill). ²⁸ Compare, too, 4.2.95 6 with the sermones of the ‘coloured’ declamations (351 2) given in oratio obliqua.

Quintilian and Declamation

123

min. 320 and often illustrated in practice in the declamations.²⁹ There follows in 113 an epigrammatic divisio which dictates the course of the rest of the discussion; Quintilian is interested only in the legal argumentation, and aequitas, though said to be of the greatest importance before the Amphictyons who are judging the case, is dismissed in a single sentence (118).³⁰ In a second elaborate discussion of a theme (7.1.42–63), we do find oratio recta extracts, but they are of a functional kind;³¹ the arguments against scriptum in 50–l are a common topic.³² But si lex obstat, nulla lis est, inane iudicium est (48) is reminiscent of the repetitious dicola so characteristic of the Minor Declamations. Aequitas is again placed last, and little discussed, though its importance is again stressed (62–3). Finally, the case of causa mortis discussed in 7.3.31–4. The opposed definitions in 32 reappear, in different wording, in the Declamations, whose author is familiar with Quintilian’s method of dealing with cases of finitio.³³ The detailed working out of the argument for the defence (7.3.33) is paralleled in the Declamations (270.7 and 270.15–16), as is the plea that the father contributed to his own death (270.11–12). Quintilian is little concerned with style here, but he ends his extract with an epigram. Such extracts, then, show us Quintilian taking over the stock situations, characters and arguments familiar not only from the Minor Declamations, but from ancient declamation as a whole. He uses them solely to exemplify methods of persuasion. They may be phrased elegantly and display declamatory point. But Quintilian never employs them as illustrations of his stylistic doctrines; there Cicero is the model.³⁴ If Quintilian was concerned to teach a realistic and forensic declamation, it was because he was worried about the effect the normal kind was having on oratory in the courts. The link was clear: paulo plus scholis demus: nam et in his educatur orator, et in eo quo modo declametur positum est etiam quo modo agatur (9.2.81). Too many of the practices of the school were leaving their mark on real oratory.³⁵ The school-trained speaker could not but find the forum strange. At the root of the trouble was the arbitrariness of the ²⁹ Decl. min. 276.1; 309.7 etc. ³⁰ Not that Quintilian is hostile to use of declamation to train pupils in emotional oratory: 2.10.8. Cf. 4.2.128 (declamatory repetita narratio is used vel invidiae gratia vel miserationis); 7.1.15 (adfectus built into a theme); 11.1.55 (plures in schola finguntur adfectus). ³¹ Note 54: omisso speciosiore stili genere ad utilitatem me submitto discentium. ³² Cf. e.g. ‘Quint.’ Decl. min. 317.1: exemplified in 317.4 ff. ³³ For the definitions see Decl. min. 270.3 ff. General precepts in 246.3 and 247.1. Compare Quint. 7.3.21 ff. ³⁴ For Quintilian and Cicero see Winterbottom (1982) [= A.5 above]. ³⁵ Cf. also Sen. Con. 9.pr.2, cited in n. 20. Not only oratory was affected: Id quoque vitandum, in quo magna pars errat, ne in oratione poetas nobis et historicos, in illis operibus oratores aut declamatores imitandos putemus (10.2.21); the result of neglect of this rule was a Lucan, magis oratoribus quam poetis imitandus (10.1.90). Contrast however 3.8.49: prosopopoeia is a useful exercise for budding historians and poets. For influence of poetry and history on declamation note [Dion. Hal.] Ars rhet. p. 372.9 Us. Rad.

124

Papers on Quintilian and Ancient Declamation

declaimers’ world. The themes were brief, and details could be filled in ad libitum by the speaker (ut videtur: 2.10.14, with a forward reference to 7.2.54–6, where we are told of the harm this may do in the courts; there one could not automatically ask, say, in an adultery case: Quis testis? Quis index?).³⁶ Similarly, in the schools pro facto est quidquid voluimus, whereas in court one cannot assume that the accuser will put on a fierce expression just because that is in your script (6.1.43). From the self-contained and self-satisfied school it was a difficult transition to the glare of the forum (10.5.18–19, where the old tirocinium fori is recommended).³⁷ The baffled products of the school stupent novitate cum in iudicia venerunt, et omnia suis exercitationibus similia desiderant (12.6.5). But it was not the same. Neither judge nor adversary would be sympathetic, and there would be a time-limit to curb your eloquence (ibid.).³⁸ You could no longer assume prior knowledge of the case in your proem (4.1.34),³⁹ or automatically place narration immediately after proem (4.2.28). Still, all this need not be more than an embarrassment to the individual product of the schools. What was worse was that the courts were starting to take over corrupt practices of the schools. Thus the laetus ac plausibilis locus that creeps in after the narration to separate the gracilitas of what precedes from the pugnacitas of the argumentation to come; the search for voluptas, and the opinionated braggadacio of advocates, was responsible (4.3.2). From the schools, too, came extravagant comparisons (8.3.76), the vice of ‘singing’ a speech,⁴⁰ and a tremulous and histrionic gesture of the hand blamed on foreign—doubtless Greek—schools (11.3.103). Hence a certain difficulty. Quintilian saw declamation as the imago of the forensic speech, and needed to stress what was common or analogous in the practice of school and forum. Thus he remarks the parallel between abdicatio and exheredatio (7.4.11, with other instances; cf. 7.4.35);⁴¹ cases involving scriptum et voluntas are fabricated in school to give practice for their frequent occurrence in the forum (7.6.1); Quintilian draws on his own experience in court to show that figured controversiae may have their uses in real life ³⁶ And where si quid tibi ipse sumas probandum est (12.6.5). For the arbitrariness of the schools cf. also Sen. Con. 9.pr.2: adversarios quamvis fatuos fingunt: respondent illis et quae volunt et cum volunt (see also 3.pr.13: Adsuerunt enim suo arbitrio diserti esse); [Dion. Hal.] Ars rhet. p. 372.18 ff. Us. Rad. (!). The theme had to be kept to (Bonner (1949), 51); but there was latitude, for the themes were skimpy. ³⁷ A familiar topic: e.g. Sen. Con. 3.pr.13; 9.pr.3 and 5; Petr. 1.2 (also Petr. 4.4: iuvenes ridentur in foro); Tac. Dial. 35.5 (cut short by the lacuna). ³⁸ Cf. Sen. Con. 9.pr.5: Illic (sc. in the forum) iudici blandiuntur, hic (sc. in school) imperant (and the section as a whole); Tac. Dial. 34.3. Tacitus makes rather similar points in contrasting the old days with the present in the law courts (Dial. 19.5 and 38). ³⁹ Cf. [Dion. Hal.] Ars rhet. p. 367.23 Us. Rad.: ὡς εἰδόσι τοῖς δικασταῖς διαλεγόμεθα προ οιμιαζόμενοι (and what follows). ⁴⁰ 11.3.57 (cf. 11.1.56). For Greek parallels Norden (1898), 375 9. ⁴¹ Note that in 7.4.20 he uses the ‘real’ word exheredare in a case that otherwise sounds declamatory, involving as it does a meretrix (cf. his comment at 11.1.82, quoted below).

Quintilian and Declamation

125

(9.2.73–4, especially Quod non inseruissem . . . nisi probare voluissem in foro quoque esse his figuris locum); even a case involving the very declamatory character of the meretrix is called scholastica materia sed non possit accidere (11.1.82). Yet on the other hand Quintilian had to be careful to tell his pupils where court and school differed, so that, as far as possible, they would not in desuescendis⁴² morentur (3.8.70). In court you would probably not have a case where both sides employed the same colour (4.2.97); a repeated narrative is res declamatoria magis quam forensis (4.2.128); in school, though not in the forum, one normally speaks in character rather than as an advocate (3.8.51);⁴³ school and forum differ in the use of contradictiones (5.13.45);⁴⁴ school themes are far less complex than the circumstances of a real case (7.1.4); in court you need to be more careful in your treatment of a definition case (7.3.20). Most significantly, in discussing which party should speak first, Quintilian remarks that this topic has been treated by others multis milibus versuum, but that in the forum the question hardly arises, for the issue is settled for one by the formula or some other device of procedure (7.1.37). We see clearly here how Quintilian seeks to bring to the academic and theoretical handbook that was current the sense of reality that his experience in the courts made natural to him.⁴⁵ Even more explicitly, Quintilian criticizes Verginius Flavus, a writer of his own time (3.1.21), cuius apud me summa est auctoritas (7.4.40).⁴⁶ Flavus ranged under qualitas cases involving abdicatio, dementia, mala tractatio, and nuptiae orbarum, all declamatory topics with no more than analogues in the Roman court (7.4.24). Sed alia quoque multa controversiarum genera in qualitatem cadunt (32). When he has listed these, Quintilian expresses surprise that Flavus tam anguste materiam qualitatis terminasse (40). But he gives the reason: cum artem scholae tantum componeret. In these sections Quintilian can be seen at work expanding an earlier handbook by introducing further ⁴² Cf. Votienus’ real reason for never declaiming (Sen. Con. 9.pr.1): ne male adsuescam. For differences between school and forum see also ‘Quint.’ Decl. min. 338.4: iudicum personae raro incidunt in scholasticas materias, in foro frequenter; Tac. Dial. 35.5 the everyday happenings of the declamation school crop up in court vel raro vel numquam. ⁴³ Cf. 6.2.17 (plerumque) and 36 (frequentius); 11.1.38 (plerumque), 55 (normal). Also 4.1.46 7 (11.1.59): in school one has a free choice, but in the forum it is rarely wise to defend oneself. All this prosopopoeia explains the utility of Menander to the declaimer (10.1.71); and it gives a clear example of the lasting effect of Attic legal practice on declamation. ⁴⁴ See Winterbottom (1970b), 99 100. ⁴⁵ Earlier, besides Cicero, observe the attitude of Pollio (Bonner (1949), 72 3). Contrast the complete blindness of a Sopatros (RG  Walz [= Weissenberger]) to real life law courts. [Dion. Hal.] Ars rhet. is less culpable (p. 370.13 Us. Rad.): (κεφάλαια) οἷς χρῶνται οἱ μελετῶντες, οὐκ ἂν χρησαμένων τῶν ὡς ἀληθῶς λεγόντων (also p. 372.1, cited in n. 4). ⁴⁶ Quintilian points out in 4.1.23 4 a case where Flavus misreported the doctrine of the (flexible) Theodorus. He also reports a joke of Flavus (11.3.126) at the expense of quodam suo antisophiste. The word is significant of Quintilian’s attitude towards Flavus; he nowhere else uses sophistes of another rhetor, let alone the very rare antisophistes.

126

Papers on Quintilian and Ancient Declamation

kinds of case, which he draws from both declamation and reality. He starts with three⁴⁷ of Flavus’ declamatory types, in great detail (27–31); then he comes to others: iniuria (common to court and school), divinatio (only reallife), tutela (common), male gesta procuratio (which he implies to be scholastic by citing the real-life analogue), inscriptum maleficium (finguntur in scholis), male gesta legatio (veris causis frequens), and then a series of declamatory actions (note 39: finguntur et testamenta). Such a pattern is common in the Institutio.⁴⁸ Advice on restraint of approach to a real-life demand for curatio bonorum comes alongside the declamatory figure of the abdicans pater (11.1.58–9). A familiar quotation from the Pro Milone (4.4.2) prefaces fictive examples from declamation. There may even be a blurring of the two levels, Ciceronian reality dressed up in declamatory clothing. The important discussion of the Caelius case is put in hypothetical terms: ut si defendendus sit M. Caelius (4.2.27; contrast the historical facts, in quibus solis omnis Ciceronis versatur oratio . . . quae est ipsius actione defensa). Contentions of the Pro Cluentio (5.10.108) and Pro Milone (7.1.34) are stated like declamation themes.⁴⁹ It is clear why Quintilian behaves like this. Id quo facilius accipiant iuvenes nondum scholam egressi, primo familiaribus magis ei aetati exemplis ostendam (5.10.96): a declamatory instance is more familiar, that is, to the schoolboy than one from Cicero (98–9).⁵⁰ Quintilian does not write only for pupils trained in his own school, where reading of Cicero supplemented and illuminated the rhetorical training.⁵¹

⁴⁷ The fourth, on nuptiae orbarum, is delayed to 39. ⁴⁸ There is particularly complex interweaving of declamation and reality in 3.8.10 ff. Suasoria is here used of real life deliberative oratory (10, 65), and the discussion largely focuses on real life for some time (though observe 19: examples based on history and stated like fictional themes). After a reference to exercitationes in 43, however, Quintilian turns to declamation (44 morality; 46 themes, known from Seneca, concerning Cicero and Antony; 47 C. Caesari suadentes . . . ad firmabimus . . . ), and this provides a smooth transition to prosopopoeia (called an exercitatio in 49), a matter mostly for the school (n. 43 above; the allusions in 50 and 51 to Cicero and Lysias are distanced from contemporary reality, though for use of deliberative oratory even in Quinti lian’s day see 70). Hence we find contrasts with verum consilium (62) and even with the superior realism of speeches in history (67 ff.). ⁴⁹ Quintilian occasionally indulges in a sort of word play drawn from the terminology of declamation. Thus 5.10.9: Asconius’ statement of the subject of a speech of Cicero is velut thema; 6.4.6: speakers who baulk at taking part in altercatio are said to be content to provide their clients merely with illum . . . ambitiosum declamandi sudorem (cf. Plin. Ep. 2.14.2: adulescentuli obscuri ad declamandum huc (sc. to the courts) transierunt); 12.8.6: orators reluctant to depart from their scripts velut themata in scholis posita custodiunt. Rather similarly, 11.1.83: Alia praetereo: neque enim nunc declamamus. Elsewhere, cf. Petr. 3.1: Non est passus . . . me diutius declamare (against declamation). ⁵⁰ Cf. 7.2.17: utendum est enim et hic exemplis quae sunt discentibus magis familiaria; also other passages where he remarks on a turn to declamation: 9.2.81 (cited above, p. 123); 11.1.55: Quod praecipue declamantibus (neque enim me paenitet ad hoc quoque opus meum et curam susceptorum semel adulescentium respicere) custodiendum est . . . ⁵¹ I refer again to Winterbottom (1982) [= A.5 above].

Quintilian and Declamation

127

But even his own adulescentes (meos enim semper adulescentes putabo) would welcome the fictae controversiae exemplum (7.3.30). In particular, the familiar themes of declamation were useful to illustrate the complexities of the status system, and, what was intimately connected with status, of argumentation. Though Quintilian has used the cases of Rabirius Postumus and Milo to demonstrate the point that status arises from the prima conflictio causarum (3.6.11–12), and though he can discuss simple points without more than a simple Hominem occidisti and Non fuit adulter as illustration (3.6.17), he falls back on a fully stated declamation case when the going gets more complicated (18–19).⁵² Similarly when he comes to consideration of the different status in Book 7, each has its illustrations from declamation (thus 7.2.17 ff.: coniectura; 7.3.19 ff., 30 ff.: finitio; 7.6.5 ff.: scriptum et voluntas; 7.7.2 ff.: ἀντινομία; 7.8.2 ff.: syllogismus; 7.9.4 ff.: ἀμφιβολία). So too when Quintilian needs to illustrate the movement from general to particular (7.1.23 ff.) or the eliciting of hidden questions (7.1.42 ff.). So again, though less frequently, in argumentation, where the examples normally come from the field of logic: even here declamation could prove helpful, giving an instance of argumenta ex causis (5.10.36: significantly attached to finitio)⁵³ and ratio factorum dictorumque (5.10.47). Is it sacrilege to steal a piece of private property from a temple? Quintilian calls that a notissimum exemplum of a problem of definition (7.3.21). Familiar it certainly is: often in Quintilian, and going back to Aristotle, forward to Hermogenes and beyond,⁵⁴ it was easy to erect a declamation theme around the problem: Qui privatam pecuniam de templo surripuit, sacrilegii reus est (ibid.).⁵⁵ Here we see in miniature the story of the interconnection of declamation and the status-system. Hermagoras will, in organizing that system, have harnessed it to the pre-existing exercise of declamation. Declamation themes could be used to illustrate types of case and the arguments ranged under those types. The rhetor’s fair copy could show, better than the speech of a Lysias or Demosthenes, those principles in operation.⁵⁶ When all this was transplanted to Rome, Cicero in the De inventione and especially the Auctor ad Herennium tried to reduce the declamatory element in a rhetorical handbook, seeking to replace Greek and fictive examples with Roman and ‘historical’ ones. As we

⁵² Cf. also some of the examples in the discussion of the novem elementa in 3.6.25 7. ⁵³ So in 4.2.68 ff.; the examples from declamation are suggested by distinguamus igitur genera causarum. ⁵⁴ See Adamietz (1966) on 3.6.38, adding e.g. RG .102.11 Walz. ⁵⁵ Cf. RLM pp. 91.17 and 338.5 Halm. For the Greek, Hermog. RhG .37.8 Rabe [= Stat. 2.2.9 Patillon] and 62.3 R. [= Stat. 4.15.5 P.]. ⁵⁶ This is worked out in detail for the Minor Declamations in Professor Joachim Dingel’s regrettably unpublished Tübingen dissertation [this work was eventually published: see Dingel (1988)]. It is true also of the declamations of Choricius, which can be analysed in terms of Hermogenes’ στάσις system.

128

Papers on Quintilian and Ancient Declamation

can see from the De oratore, Cicero’s own legal experience led him to tip the scales even further in the direction of reality. But, as Quintilian’s polemic against Verginius Flavus shows, the writers of artes continued to go their own sweet way. The Institutio, here as elsewhere, evinces a Roman and Ciceronian desire to return towards the real. But, as this paper has tried to show, Quintilian found that declamation could not be thrown overboard in an ars any more than it could in the school. Quamvis enim omne propositum operis a nobis destinati eo spectet ut orator instituatur, tamen, ne quid studiosi requirant, etiamsi quid erit quod ad scholas proprie pertineat in transitu non omittemus (2.10.15).⁵⁷

⁵⁷ So it is inadvertently that Quintilian might seem to imply in 2.10.1 (tempus adpetet ad grediendi suasorias iudicialesque materias: quarum antequam viam ingredior, pauca mihi de ipsa declamandi ratione dicenda sunt) that the rest of the Institutio will be wholly concerned with declamation.

9 Quintiliano e Virgilio Q(uintiliano) fu retore (Calahorra 35/40 d.C. – Roma? 96 d.C. o poco oltre). L’unica opera che ci sia pervenuta, la Institutio oratoria, pubblicata non molto prima del 96 d.C., testimonia come, nel I sec. d.C., V(irgilio) fosse annoverato fra i classici. Q. lo ritiene inferiore solo a Omero, lo considera un’autorità per ciò che riguarda il trattamento della poesia e lo cita spesso per illustrare questioni grammaticali e retoriche. Altro elemento non meno significativo è il suo servirsi di versi virgiliani, per abbellire passi particolarmente importanti del suo trattato. Il punto culminante delle sue lodi dell’imperatore Domiziano (10.1.92) è la citazione di Ecl. 8.13. In 12.1.27 Q., descrivendo la figura del grande oratore, la cui superiorità verrà messa in luce sulla scena politica, cita Aen. 1.151–3, versi che ai suoi occhi sono la descrizione dell’oratore catoniano, il vir bonus dicendi peritus (analisi, questa, seguita da vicino da Serv. ad loc.). In entrambi i casi (come anche per la citazione di Georg. 2.272 in 1.3.13, che pur essendo meno cruciale segna però un momento culminante) l’autore viene menzionato per nome, mentre Q. presume che i lettori riconosceranno comunque Aen. 3.193 quando descrive la sua sensazione di galleggiare su un oceano sconfinato (12.pr.3), come pure l’accenno casuale di 2.13.8 (= Aen. 3.436: la citazione non è necessariamente inesatta, malgrado l’opinione di Cole (1906)). Analogamente, ci viene detto che l’oratore di alto livello ‘non sopporta i ponti’ (12.10.61), una velata allusione a Aen. 8.728 (a cui si fa riferimento, di nuovo senza nominare l’autore, in 8.6.11, come a un tipo di metafora). Q. (8.3.28), citando come virgiliani 4 versi di Cat. 2, loda il componimento per la sua posizione critica nei confronti dell’artificiosità (mirifice) e aggiunge una nota erudita in cui identifica l’autore a cui tale critica è diretta. Naturalmente la fonte della sua grande ammirazione per V. è la sua poesia in esametri. Il poeta occupa il primo posto nella lista degli scrittori romani (10.1.85), come Omero capeggia quella degli autori greci (10.1.46); Q. sottolinea apertamente il parallelo (si veda anche 12.11.26) e, anzi, giunge al punto di menzionarlo come ‘senz’altro il più [Originally published as: ‘Quintiliano (M. Fabius Quintilianus)’. From Enciclopedia Virgiliana, vol. 4 (Rome, 1988), 374 6. Italian translation by Barbara Macleod]

130

Papers on Quintilian and Ancient Declamation

vicino’ a Omero di tutti i poeti greci e latini ‘di quel genere’ (eius generis): giudizio, questo, rafforzato da una frase del mentore di Q., Domizio Afro (10.1.86). Il ‘genere’ a cui si fa riferimento non è soltanto la poesia epica, bensì la poesia in esametri in generale, come viene immediatamente dimostrato dal l’allusione ai poeti didascalici Macro e Lucrezio (10.1.87; la superiorità di V. viene notata in 12.11.27); ma il paragone con Omero implica necessariamente che in questo passo Q. si riferisca in particolare all’Eneide. Sebbene V. debba inchinarsi alla ‘natura celeste e immortale’ di Omero, curae (cf. 10.3.8, dove ci si richiama all’autorità di Vario per confermare come V. scrivesse soltanto pochissimi versi al giorno) et diligentiae vel ideo in hoc plus est, quod ei fuit magis laborandum, et quantum eminentibus vincimur, fortasse aequalitate pensamus. Vale a dire, il livello generale della poesia virgiliana è addirittura più alto di quello di Omero, benché egli non ne eguagli mai le vette più elevate. Ma egli doveva ‘lavorare di più’, in quanto mancava della φύσις omerica. La contrapposizione è analoga a quella fatta da Ovidio riguardo a Ennio ingenio maximus, arte rudis (Tr. 2.424), tranne per il fatto che, non volendo sostenere che Omero è limitato nell’ars, lo rappresenta invece come meno diligente di Virgilio.

L’intento della Institutio è di esporre i principi generali da seguire nell’educazione dell’oratore. Prima di arrivare alla scuola del rhetor, il giovane doveva venire istruito per alcuni anni da un grammaticus, ed era soltanto a questo stadio che gli veniva impartita un’istruzione formale sui poeti. In 1.8.5 Q. specifica che a Roma la prassi abituale era di cominciare a leggere sui testi di Omero e di Virgilio. Quello che segue è significativo: ‘è necessaria una capacità di giudizio più sviluppata per apprezzare le loro doti, ma per quello c’è tempo: non verranno letti soltanto una volta.’ E, in effetti, egli riteneva che i poeti avessero molto da offrire al giovane più in là negli studi e all’oratore già affermato (10.1.27–30). Ma il fatto resta che erano i ragazzi più giovani, di dodici anni o meno, a essere seguiti nello studio più particolareggiato dei poeti: gli studenti più grandi leggevano V. per conto loro, o non lo leggevano affatto. Comunque, dato che la grammatica veniva considerata in buona parte come propedeutica rispetto alla retorica, il modo in cui i grammatici spiegavano la poesia risentiva considerevolmente dell’influsso della retorica, e un commentatore come Servio riteneva che il sottolineare gli elementi retorici del testo poetico fosse parte del suo lavoro. In effetti, il sistema di figure retoriche e tropi era altrettanto sviluppato per la poesia che per la retorica (e di conseguenza, p. es., nelle Etymologiae di Isidoro il sistema viene incluso nel libro 1, De grammatice, invece che nel 2, De rhetorice). In un trattato di retorica come quello di Q., le figure retoriche e i tropi vengono inclusi nella retorica, ma gli esempi, tratti dalla poesia non meno liberamente che dall’oratoria, dimostrano l’interrelazione con la γραμματική. E di nuovo, così come Servio usa raramente la dottrina retorica al di fuori dell’erudita trattazione di figure retoriche e tropi (fatta eccezione per l’analisi dei discorsi contenuti nell’Eneide), Q. usa V. per illustrare il suo insegnamento retorico principalmente nei libri della Institutio che trattano della elocutio, l’8 e il 9.

Quintiliano e Virgilio

131

È raro che Q. vada oltre la semplice citazione di V.: se il nome del poeta viene menzionato, è di solito nella forma Vergilius (P. Vergilius in 8.3.24); il titolo dell’opera da cui la citazione è tratta è di rado menzionato, e in primo Georgicon (8.3.78) è l’unico esempio in cui anche il libro viene citato; spesso l’autore non viene neppure nominato, e talvolta i commenti suggeriscono una scarsa sensibilità per la poesia.¹ Talvolta, però, quello che egli dice è di più ampio interesse. Nelle Bucoliche (l’osservazione di Orazio, Sat. 1.10.44–5 sulle Bucoliche viene discussa in 6.3.20, dove facetum è inteso implicare non umorismo, bensì bellezza ed eleganza colta), Q. (8.6.46–7), al pari di Servio, è convinto che in realtà il passo di 9.7–10 si riferisca a V. stesso e non a Menalca. In 9.3.8, benché i celebri versi 62–3 della Ecl. 4 siano corrotti nei suoi manoscritti come in quelli di V.,² il commento di Q. sulla struttura grammaticale permise a Poliziano di emendare cui in qui (sarà poi Bonnell a completare la correzione emendando parentes in parenti). E ancora, in 10.1.56, Q., come Servio e Probo, ritiene che Chalcidico versu (Ecl. 10.50) sia un’allusione a Euforione. Per ciò che riguarda le Georgiche, Q., sapendo (10.1.56) che esse traevano spunti da Nicandro, non aveva bisogno di un grande acume per vedere che il tibi e il mihi di 2.298 e 3.435 vanno generalizzati (9.3.21). Riguardo all’Eneide, Q. nota come V., al pari di Omero, avvinca l’attenzione del lettore all’inizio del poema con una breve esposizione della summa rei (4.1.34). Egli sottolinea il grande amore di V. per il linguaggio arcaico (9.3.14),³ citando Aen. 11.406[⁴] e 1.19–20.⁵ Analogamente, in 8.3.24–5, egli sostiene che l’uso che V. fa di parole arcaiche è un segno della sua ‘buona capacità di giudizio’: olli,⁶ quianam (Aen. 5.13 e 10.6, e Serv. ad loc.), moerus (cf. Serv. ad Aen. 10.24), pone,⁷ pelligere (Norden (1957⁴), ad 6.34; un’altra lettura possibile, in Q., è pellicere, su cui Haupt (1870)). E ancora, in 1.7.18 Q. menziona pictai vestis (Aen. 9.26) e aquai (7.464; cf. Servio e Serv. Dan. ad loc.). In 8.3.63 egli, commentando il passo che parte da Aen. 5.426, sostiene che la descrizione dei pugili è così vivida che gli spettatori stessi non hanno potuto vederli meglio dei lettori, il che è esattamente lo scopo della ἐνάργεια (cf. anche 6.2.32–3; 8.3.70). Infine, Q. (9.2.64) commenta Aen. 4.550–1 sotto il punto di vista dell’emphasis: le emozioni di Didone esplodono con tanta forza che ella pensa che una vita senza matrimonio sia accettabile solo per gli animali.⁸ A proposito della redditio, in 8.3.79 Q. dice: praeclara apud Vergilium multa reperio exempla, sed oratoriis potius utendum est. In queste parole cogliamo con chiarezza la già menzionata contaminazione di grammatica e retorica, di poesia e

¹ Kennedy (1969), 85 a proposito di Quint. 8.6.60. ² Sulla limitata importanza di Q., che spesso sembra citare a memoria, per la critica testuale virgiliana, cf. Cole (1906); Odgers (1933). ³ Williams (1960), ad Aen. 5.10. ⁴ [Forbiger (1872 5), .473.] ⁵ Austin (1971), ad loc. ⁶ Ibid. ad Aen. 1.254. ⁷ Austin (1964), ad Aen. 2.725. ⁸ Austin (1955), ad loc.

132

Papers on Quintilian and Ancient Declamation

oratoria. Q. prosegue, com’è logico, con esempi tratti da Cicerone. In vari passi egli può citare parallelamente Cicerone e V. senza farsi scrupoli: p. es. 4.2.2; 8.5.11; 8.6.11 12 e 68 9; 9.2.50 e 54; 9.3.28, 34, 44 e 75. In altri casi, però, è costretto a contrapporre l’uso poetico e quello oratorio: così in 8.3.73 4 (simili tudini in cui l’ovvio è ‘chiarito’ dall’oscuro); 8.6.18 (‘in un’arringa non direi . . . che gli uccelli “nuotano” nell’aria, benché Virgilio si sia servito brillantemente del termine riferendosi alle api e a Dedalo’: cf. Georg. 4.59 e Aen. 6.16; Servio nel suo commento a quest’ultimo luogo mette i due passi in relazione), 29 30 (antono masia), 40 (epiteti), 60 (perifrasi), 66 (tmesi).

Anche il termine reperio (8.3.79) è interessante. Viene da domandarsi se Q. non abbia avuto un ruolo di primaria importanza nella individuazione di citazioni virgiliane (nonché ciceroniane) per spiegare questioni di grammatica e di retorica. Non essendoci pervenuti trattati che risalgano al periodo fra V. e Q., non possiamo saperlo con certezza.⁹ Che Q. conoscesse le opere del poeta tanto bene da poter svolgere tale ruolo, è probabile; è però anche probabile che egli si sia limitato ad ampliare elenchi che già si andavano compilando. Non possiamo considerare come prove sicure le concordanze d’opinione fra Q. e i commentatori virgiliani di età successive, benché sembrerebbe che questi fossero da lui indipendenti e che le coincidenze risalgano a una fonte comune (un grammatico?). È certo che Servio talvolta segue una linea indipendente. Commentando Georg. 3.364, egli tenta una spiegazione di umida che tacitamente confuta quella di Q. (8.6.40), il quale invece lo considera un epiteto superfluo; definisce l’aedificant di Aen. 2.16 una metafora e non una catacresi (come fa invece Quint. 8.6.34),¹⁰ mentre in Aen. 1.65 identifica non un’antonomasia (Quint. 8.6.29), bensì una perifrasi. Ma più spesso, com’è naturale, le loro interpretazioni coincidono. Così in Aen. 1.135 l’interruzione del discorso viene attribuita all’ira (in accordo con Quint. 9.2.54); Georg. 2.541–2 vanno interpretati allegorice (cf. Quint. 8.6.45); vocemque his auribus hausi (Aen. 4.359) è un esempio di pleonasmo (e così sostiene Quint. 8.3.54; cf. anche 9.3.46); tantum sperare dolorem (Aen. 4.419) è un’acyrologia (cf. Quint. 8.2.3). E in certi casi l’analogia è sorprendente: sia Q. (8.6.8) che Servio spiegano Georg. 3.135–6 come intesi a evitare l’uso di linguaggio indecente (cf. anche Quint. 9.3.59 e Serv. ad Ecl. 3.8–9), mentre il commento di Servio Dan. ad Georg. 1.181 potrebbe senz’altro essere basato su Quint. 8.3.19–20 (o la sua fonte). Q. ricorre a V. (la cui ortografia sostiene di conoscere sulla base del manoscritto autografo del poeta: 1.7.20; cf. Gell. 2.3.5) anche per illustrare questioni gramma ticali in senso stretto. In questo V. è uno degli auctores, una di quelle grandi autorità il cui iudicium conta come ratio (1.6.2), che possono trasformare a loro piacimento lo scorretto in corretto (1.5.35 e 1.6.2: entrambi illustrati da irregolarità ⁹ Su Remmio Palemone, il grammatico che fu maestro di Q., come ammiratore di V., cf. Schöll (1879). ¹⁰ Condorelli (1968).

Quintiliano e Virgilio

133

del genere, alcune delle quali vengono commentate anche da Servio). Anche qui Q. ci può offrire qualcosa di diverso da Servio: secondo lui, l’infinito storico di Aen. 11.142 è un’ellissi (8.6.21), mentre Servio Dan. lo definisce un infinitum pro pronuntiativo. Ma più spesso le loro interpretazioni coincidono. Entrambi riten gono che conticuere (Aen. 2.1) non possa essere un duale (Quint. 1.5.43), dato che in latino tale forma non esiste. Entrambi pensano che navigat aequor (Aen. 1.67) sia una figura greca (Quint. 9.3.17); e Q. (9.3.6), come Servio, vede un’analogia tra Georg. 1.183 e Ecl. 8.28. Le osservazioni di Q. non sono sempre corrette: egli è evidentemente in errore (9.3.11) riguardo al congiuntivo velit di Aen. 2.104, mentre in 7.9.10 sembra combinare Aen. 1.155 e 5.212 nelle parole caelo decurrit aperto (la nota di Servio sul primo dei due passi, in cui sceglie fra i due possibili si gnificati, dimostra che questo è il passo che Q. aveva in mente; ritroviamo una analoga fusione di passi virgiliani in 9.4.85 e, forse, in 9.2.49).¹¹

Naturalmente, V. costituiva anche un’utile fonte di materiale per illustrare problemi metrici: p. es., per l’allungamento di vocali per posizione (1.5.28 e 9.4.85) e l’elisione (9.4.40 e 11.3.34: in entrambi i casi Q. si serve di Aen. 1.3, laddove anche Servio commenta mult(um) ille); in 1.5.18 menziona l’insolita scansione di Ītaliam (Aen. 1.2) e di unĭus (1.41), cui anche Servio Dan. e Servio fanno riferimento nel commento. In 11.3.36–8 troviamo una discussione particolarmente interessante sul metodo corretto di dividere i cola e di rimarcare la divisione con la voce:¹² 1’esempio che egli sceglie sono i ben noti primi versi dell’Eneide, presupponendo qui, come in altri casi, che la poesia venisse letta ad alta voce. Il modo di pronunciare può influire sul significato (11.3.176, con diversi esempi: Serv. ad Aen. 11.383 ne collega due), e in 11.3.70 egli sembra avere in mente una lettura in pubblico, poiché descrive i gesti che si adatterebbero a Aen. 3.620 e 1.335 (entrambi parte di discorsi). Q. ricorre inoltre a V. come a un testimone di ‘fatti’. Questo auctor eminentissimus dimostra, grazie al carme di Iopa (Aen. 1.742), che ‘la musica è connessa alla conoscenza di questioni divine’ (1.10.10). Si può ricorrere alle opere di V. come a una fonte autorevole sull’uso di certe parole: numeri = rhythmi (9.4.54: cf. Serv. ad Ecl. 9.45) e argumentum (5.10.10: cf. Serv. ad Aen. 7.791). Egli può spiegare al filosofo logico l’uso di segni (5.9.15–16) e le opposizioni (5.11.14), nonché similarità e dissimilarità (5.11.30, ove una citazione di Ecl. 1.22 è preceduta dalla considerazione che la formica e l’elefante, tanto diversi tra loro, pure sono accomunati grazie al genus: una riflessione, questa, forse non indipendente dal fatto—a noi noto grazie al commento di Servio ad Aen. 4.404—che V. applica alle formiche un emistichio che Ennio, Ann. 465 Vahlen² = 494 Skutsch aveva usato a proposito di elefanti). Solo in un caso Q. sembra criticare V.: quando, senza nominare l’autore, trascrive Aen. 1.109 come un esempio di ‘mescolanza di parole’ (8.2.14), un tipo di oscurità particolarmente indesiderabile (così Carisio, p. 363.3 Barwick: ¹¹ Cole (1906).

¹² Cunningham (1953 4).

134

Papers on Quintilian and Ancient Declamation

Synchysis est hyperbaton obscurum; in Servio non troviamo alcuna critica). In Ecl. 10.11 la ripetizione di nam sembra superflua, ma ‘non manca di attrattiva’ (9.3.18). E Q. (8.3.47) scarta l’opinione di Celso sul possibile significato osceno di Georg. 1.357 (che Ausonio non include nel Cento nuptialis). Egli conosce bene V. e nutre per lui una grande ammirazione. Citandolo così spesso senza menzionarlo, egli presuppone dei lettori che lo conoscano altrettanto a fondo.

10 Sopatros’ Διαίρεσις ζητημάτων I. SOPATROS We can say little about the author of the Διαίρεσις ζητημάτων.¹ He was certainly called Sopatros (RG .55.6; 184.14–15 Walz [= 9.1.13; 29.5.2 Weissenberger]). His addressee is an unknown Carponianos, whom he calls υἱέ ² (pp. 78.8–9; 129.17 W. [= 12.1.2–3; 22.3.16–17 We.]). P. 55.6–7 W. [= 9.1.13 We.] seems to say that Sopatros taught at Athens, but we suggest an emendation that would bring the passage into line with words at the end of Sopatros’ Prolegomena to Aristides: Ταῦτ’ ἐγώ σοι Σώπατρος ἐπιδίδωμι, ὅσα γε ἔμαθον (v.l. μαθὼν) παρὰ τῶν διδασκάλων Ἀθήνῃσι . . . ³ Both passages would then prove that Sopatros was educated in Athens. The words ὁ σοφὸς ὁ ἡμέτερος Ἱμέριος (p. 318.29 W. [= 60.2.9 We.]) are problematic,⁴ but they suggest that Sopatros and his readers⁵ were, like Himerios, Athenian, and may [Original title: ‘Introduction’. From D. (C.) Innes and M. Winterbottom, Sopatros the Rhetor. Studies in the Text of the Διαίρεσις ζητημάτων (London, 1988), 1 20. This section is entirely due to MW] ¹ The title as given ‘in den führenden Hss’ according to RE s.v. Sopatros 10 (Glöckner), 1002. Not, however, in C [Oxford, Corpus Christi College, 90; c.1300 1330], U [Città del Vaticano, Biblioteca Apostolica Vaticana, gr. 901; c.1250 1280], or Par. [Paris, Bibliothèque Nationale, gr. 2983A; c.1275 1300], none of which gives us the beginning of the book. The title is confirmed by Ioannes Doxapatres (see Glöckner (1901), 11): though another citation in Glöckner (1901), 22 adds the adjective πολιτικῶν (cf. RG .25.12 13 Walz). For the meaning of the singular διαίρεσις see below, p. 137 n. 13. [Greek text: RG .2 385 Walz, utterly inadequate; a proper critical edition is not yet available, but a readable Greek text much improved, with paragraphs and subsections and a German translation with some notes are now offered by Weissenberger (2010).] ² See Russell (1983), 7 n. 23, raising the possibility that Sopatros merely meant ‘an intellectual or spiritual “son” or pupil’. But a real son would make a good addressee (as Cato and the Elder Seneca found). ³ Dindorf (1829), .757.24 5 [= Lenz (1959), 151.1 2]. Other resemblances of language make it certain that the Sopatros who wrote the Διαίρεσις was also the author of these prolegomena (which are addressed to an Alexander: Dindorf (1829), .745.2 [= Lenz (1959), 127.5]). ⁴ See Innes Winterbottom (1988), 246. ⁵ The whole atmosphere of the declamations is patriotically Athenian (see below, p. 159 n. 139), but that hardly proves that they were written at Athens by an Athenian.

136

Papers on Quintilian and Ancient Declamation

identify Himerios as one of Sopatros’ teachers.⁶ At the least, this passage demonstrates that Sopatros was contemporary with or later than Himerios, who lived c.310–c.390. Other rhetoricians to whom he alludes are earlier than that: Minucianos (p. 18.10 W. [= 3.3.3 We.]: lived under Antoninus Pius), a commentator on him, Porphyrios (p. 35.21 W. [= 8.4.1 We.], c.232–c.305),⁷ and a commentator on Hermogenes (2nd century ) called Metrophanes (p. 65.18–19 W. [= 9.9.12–13 We.]: perhaps 3rd century). The first writer to cite Sopatros’ Διαίρεσις appears to be George of Alexandria, perhaps to be dated in the 5th century.⁸ The Διαίρεσις itself, naturally enough in this unrealistic genre, gives no topical reference. Sopatros may well have lived in the late fourth century, but there is no certainty about it. We also have extracts reported from τῶν Σωπάτρου Μεταποιήσεων, said to have been addressed to one Ausonios.⁹ A few fragments from Sopatros’ Προγυμνάσματα¹⁰ are of little service. There is a lengthy commentary on Hermogenes’ Περὶ των στάσεων¹¹ attributed to a Sopatros; but it is not likely that it is the work of the author of the Διαίρεσις, and it throws no light on the identity and little on the date of its own author.¹²

II . THE Δ Ι Α Ι Ρ Ε Σ Ι Σ AND HERMOGENES The ζητήματα (‘problems’) of Sopatros’ title are the kernel of what the Romans called controversiae, exercises in which students of rhetoric were trained to give speeches in the law-court. The διαίρεσις (divisio) was the advice ⁶ The tone is quite unlike that of Sopatros’ references to rivals. Note also p. 184.13 15 W. [= 29.5.1 2 We.], where Sopatros contrasts himself by name with ὁ διδάσκαλος ἡμῶν. It is presumably a coincidence that there is extant a letter from ‘Sopatros’ to his brother ‘Hemerios’ (Schmid Stählin (1961⁶), 1086 n. 4). ⁷ See Innes Winterbottom (1988), 40. ⁸ See Schilling (1903), 759 (references to two citations) and 692 3 (date, partly argued from an assumed fourth or fifth century date for Sopatros). There is a verbatim citation of Sopatros in RG .247.11 20 Walz, but the author is unknown. ⁹ Glöckner (1910). For fragments known earlier see RG .1294.6 1295.3 Walz (Demos thenes; expanded in Rabe (1908), 141 3) and RhG .112.24 8, 115.2 12, 119.30 120.10 Spengel (Homeric). ¹⁰ See Rabe (1926), 59 70. For the interconnection of progymnasmata and declamation see p. 249.21 2 W. [= 43.6.8 9 We.]. ¹¹ Available in RG  and .1 211 Walz. ‘Ego a compilatore quodam commentarium volu minis  amplificatum et aliorum technicorum doctrina depravatum, commentarium autem voluminis  in brevius contractum esse, neutra igitur recensione genuinum nobis Sopatri opus servatum esse censeo’ (Glöckner (1901), 2). See further Rabe (1909). ¹² The evidence of RG  is difficult to weigh. The rhetores referred to in RG  are much the same as those mentioned in the Διαίρεσις (Aristides, Libanius, Lollianos, Minucianos, Porphy rios); it is not likely that the two works differ much in date. But I see no stylistic or other evidence to suggest that they are the work of the same author.

Sopatros’ Διαίρεσις ζητημάτων

137

given by teachers on the treatment of such declamations. Division hoc proprium habet, ostendere ossa et nervos controversiae (‘Quint.’ Decl. min. 270.2); and it would normally be given orally as an introduction to each theme (ibid. 314.1; Quint. 2.6.1–2). What Sopatros gives us is a series of such introductions¹³ organized for publication.¹⁴ At varying lengths, he provides the headings according to which no less than 82¹⁵ fictional cases (two with the same theme) were to be argued, often illustrating with extracts of oratio recta declamation. The themes are ordered according to their στάσις or ‘issue’. Sopatros’ general practice is to ‘divide’ the first case falling under a particular στάσις at some length,¹⁶ with technical discussion and occasional polemic against rival scholars; other cases under the same στάσις are dealt with more summarily. The order of the στάσεις¹⁷ is: στοχασμός, ὅρος, ἀντίληψις, ἀντιθετικαί, μετάληψις, παραγραφή, πραγματική, ῥητὸν καὶ διάνοια, ἀντινομία, ἀμφιβολία, συλλογισμός. This list starts in the same order as Hermogenes’ in his Περὶ τῶν στάσεων, but after the ἀντιθετικαί Hermogenes continues: πραγματική, μετάληψις, ῥητὸν καὶ διάνοια, ἀντινομία, συλλογισμός, ἀμφιβολία.¹⁸ We are thus at once confronted with the problem of Sopatros’ sources. ‘[L]ater commentators . . . treat [Hermogenes’] words as they would a classical text.’¹⁹ Yet Sopatros never names Hermogenes, and he differs from him on even the basic matter of the order of the στάσεις. He does not even seem concerned to mark his differences from him. Though he often conducts polemic against anonymous τινες,²⁰ their heresies never coincide with known views of Hermogenes. It looks as though Hermogenes bulks larger in the view of modern ¹³ Sopatros might reasonably have made his title (for which see above, p. 135 n. 1) Διαιρέσεις ζητημάτων (cf. Eun. VS 491: τὰ ζητήματα διαιρέσεις ἔφασκεν); note the implication of p. 65.24 5 W. [= 9.10.5 We.]: ἄλλῃ διαιρέσει. But the singular διαίρεσις is presumably more abstract, ‘the method of dividing’ a series of themes. ¹⁴ Publication (not proved by p. 55.7 W. [= 9.1.13 We.]: ἐξεθέμην) is suggested by the sys tematic layout and the elaborate cross referencing. It is likely that the complete book con tained a prefatory letter addressed to Carponianos (see Innes Winterbottom (1988), 21). ¹⁵ Other cases are referred to allusively (pp. 42.25 43.5 W. [= 7.1.12 18 We.], with Innes Winterbottom (1988), 43); yet others, though not analysed, are mentioned in full detail: p. 51.16 20 W. [= 6.1.11 14 We.] ~ p. 52.11 13 W. [= 6.2.16 17 We.]; p. 127.11 12 W. [= 22.1.10 11 We.]; p. 129.4 7 W. [= 22.3.5 8 We.]; p. 181.22 4 W. [= 28.5.5 7 We.]; p. 246.28 W. [= 43.1.29 We.]; p. 268.12 16 W. [= 46.1.11 15 We.]; p. 324.14 15 W. [= 63.1.2 3 We.]; pp. 332.28 333.1 W. [= 67.3.2 3 We.]; p. 348.28 9 W. [= 70.1.7 8 We.]; p. 377.21 2 W. [= 79.1.2 3 We.]. ¹⁶ Observe especially Cases 1, 13, 22, 31, 35, 50. ¹⁷ For a detailed scheme see Innes Winterbottom (1988), . ¹⁸ For his treatment of παραγραφή see Innes Winterbottom (1988), 186 on p. 245.22 W. [= 43.1.1 We.]. ¹⁹ Russell (1983), 6 7. ²⁰ P. 10.9 W. [= 1.6.4 We.]; p. 16.20 W. [= 3.1.9 We.] (against Minucianos?); p. 35.24 6 W. [= 8.4.5 6 We.]; p. 79.2 W. [= 13.1.7 We.]; p. 93.28 W. [= 14.2.5 We.]; p. 128.5 W. [= 22.2.1 We.]; p. 227.26 W. [= 38.1.1 We.]; p. 231.2 W. [= 39.4.1 We.]; p. 360.26 7 W. [= 74.1.1 2 We.]. See also p. 184.13 W. [= 9.5.1 We.]; p. 317.7 W. [= 58.2.7 We.]; p. 321.12 22 W. [= 61.3.1 9 We.]; p. 343.21 8 W. [= 61.6.1 6 We.]; p. 344.9 11 W. [= 70.2.2 4 We.].

138

Papers on Quintilian and Ancient Declamation

scholars, for whom he is the unique complete Greek source for a late στάσις system, than in that of Sopatros. Nor is Sopatros’ treatment of the sub-divisions of the στάσεις particularly faithful to Hermogenes. Hermogenes does not know the term ἀντεγκληματικός (p. 16.5 W. [= 3.tit. We.]),²¹ nor the division of ἁπλοῦς ὅρος into κατὰ κρίσιν, κατ’ ἀξίωσιν, and κατ’ αἴτησιν (p. 78.21 W. [= 13.tit. We.];²² see RG .455.1 ff. Walz for Hermogenes’ neglect of the distinction). Sopatros has a different order and terminology for the sub-divisions of διπλοῦς ὅρος (p. 98.12 W. [= 16.tit. We.]),²³ and he follows Minucianos’, not Hermogenes’, order of the ἀντιθετικαί (p. 188.9 W. [= 31.th. We.]).²⁴ On the other hand, it would be wrong to suppose that he follows Minucianos as a matter of course: he disagreed with him, we know, on στοχασμὸς παράδοξος (p. 78.6 W. [= 12.tit. We.]).²⁵ And he differed from both Hermogenes and Minucianos on μετάστασις (p. 360.27–9 W. [= 74.1.1–2 We.]).²⁶ Finally, though Sopatros’ κεφάλαια for particular types of case always bear a family resemblance to those of Hermogenes, they are not always identical with them. Thus in στοχασμός he uses the Hermogenean headings in their Hermogenean order, and in ὅρος too he is very close to Hermogenes (p. 82.6 W. [= 13.4.1 We.], though with different terminology for προβολή).²⁷ But when Sopatros comes to give the headings for ἀντίστασις they are only ὁρικὸν παραγραφικόν, τὰ ἀπ’ ἀρχῆς ἄχρι τέλους, ἀντίστασις itself, μετάληψις (p. 189.9–13 W. [= 31.2.1–5 We.]), whereas Hermogenes has a far longer, more complex and inconsistent list (RhG .72.3–9 Rabe [= Stat. 6.1.1–7 Patillon]); again, Sopatros sharply prunes the Hermogenean headings for ἀντίληψις (cf. the structure of e.g. Case 22 with Hermog. RhG .65.10–13 Rabe [= Stat. 5.1.1–5 Patillon]). It may well be that Sopatros was concerned to keep down the number of headings to simplify his task as a teacher of the young. And in general he seems to go his own way, nullius addictus iurare in verba magistri. That is unsurprising. It was of the status lore that Quintilian wrote: cum in aliis omnibus inter scriptores summa dissensio est, tum in hoc praecipue videtur mihi studium quoque diversa tradendi fuisse (3.6.22). Sopatros, however, hardly breaks away from the traditional themes for declamation. He normally uses traditional examples (often from Hermogenes), or variants on them.²⁸ Even the ones that lack exact parallel naturally

²¹ See Innes Winterbottom (1988), 28. ²² See Innes Winterbottom (1988), 68. ²³ See Innes Winterbottom (1988), 85. ²⁴ See Innes Winterbottom (1988), 145. Note, however, that the Hermogenean order is fol lowed in the Aristides Prolegomena (cited ibid.). ²⁵ See Innes Winterbottom (1988), 67. Besides, Sopatros explicitly differs from Minucianos at p. 18.10 ff. W. [= 3.3.3 ff. We.]. ²⁶ See Innes Winterbottom (1988), 283. ²⁷ See Innes Winterbottom (1988), 70 1. ²⁸ Cf. the introductory notes to the cases in Innes Winterbottom (1988), passim.

Sopatros’ Διαίρεσις ζητημάτων

139

move in the world of Sophistopolis, so urbanely described by Russell.²⁹ That is a little-changing world, probably formed in the Hellenistic period and elaborated rather than developed over the next centuries.

I I I. TH E ΔΙΑΙΡΕΣΙΣ AND OTHER COLLECTIONS The Διαίρεσις ζητημάτων has great importance in the study of declamation, both Greek and Latin. That is partly because of its very form: it takes us, as no other work does, into the very workshop of the rhetor. Many moderns form their impression of declamation largely from the Elder Seneca. Yet his collection (which unlike Sopatros’ also covers suasoriae, exercises in deliberative oratory) is highly idiosyncratic. Seneca was not himself a rhetor, or even a declaimer; he was an appreciative auditor of declaimers, hoping to pass on to his sons the flavour of the speakers he had heard during a long life span, which started well before and finished well after the principate of Augustus. To this end, he typically gathers under each of the themes sententiae from the speeches of various declaimers, along with a divisio and the colores (or ‘slants’ on the case) adopted. But he sees declamation from the outside. His divisions are untechnical: there is nothing like the relentless series of headings given by Sopatros.³⁰ He is not instructing budding declaimers, but commenting intelligently on what his extraordinary memory had preserved from the past. Seneca never gives us a complete speech,³¹ and he usually restricts himself to mere snippets. But complete speeches are available. In Latin we have the socalled Major Declamations (before the end of the fourth century ) falsely ascribed to Quintilian; in Greek, among much else, the collections of Libanius (314–c.393) and Choricius (early sixth century). These speeches were consciously worked up for publication by practising rhetores, and their ἐπίδειξις is meant to impress a wider audience than that found in the classroom. They may well have been recited on public occasions. But, orally or in writing, they aimed to enhance the reputation of the teacher and to impress prospective parents. They are more, that is, than mere fair copies. But it is true that

²⁹ Russell (1983), ch. 2. ³⁰ See Winterbottom (1974b), . ; Fairweather (1981), 152 65. Contrast even so complex a division as Latro’s in Con. 1.2.13 14 (schema in Fairweather (1981), 161) with Sopatros’ practice. Division was becoming more complicated all the time (Con. 1.1.13); but we should not assume that Sopatros’ book would have seemed unfamiliar to Greek rhetors in and before Seneca’s day. ³¹ On Con. 2.7, cut short by a lacuna in the manuscripts, see below, p. 142.

140

Papers on Quintilian and Ancient Declamation

Choricius’ speeches, and some in the Libanian corpus, are prefaced by short introductions³² where the speaker comments on the tone and treatment demanded by the theme in question. And all these declamations are, in addition if not primarily, examples of how a speech should be given: the goal towards which the pupil strove. More like the Διαίρεσις ζητημάτων is the Latin collection known as the Minor Declamations,³³ the work of Quintilian or his school. Here, speeches or fragments of speeches are often introduced by or interspersed with sermones, ‘chats’ given by the teacher to advise on the treatment of the theme set. There is no attempt at systematic arrangement, and the collection seems to be no more than the Nachlass of a rhetor. His fair copies, complete and incomplete, are jumbled together with notes for his classroom lectures, and were published, it is to be surmised, out of piety after his death. The non-technical and unsystematic nature of the sermones is on a different level of sophistication from Sopatros. In particular, they give divisions of the Senecan rather than the Sopatran kind: heads of argument, not titled κεφάλαια. More advanced Roman teachers will no doubt have used their declamations systematically as a means of teaching the στάσις lore. Even the Minor Declamations illustrate the arguments advised in the handbooks; but they are directed to an audience too elementary to be battered with technicalities.³⁴ It follows that the particular significance of the Διαίρεσις is that it lays bare the intimate connection between declamation and rhetorical precept that forms an enduring theme in the story of the genre. What is elsewhere assumed or left to be inferred is here stated. The teacher gives instruction in precepts by showing his pupils how to apply those precepts in practice speeches. That is true of style too, as we shall see later. But for the moment we may dwell on dispositio and inventio. ³² Θεωρίαι in Choricius, προθεωρίαι in Libanius [and e.g. Himerius, Themistius]. These short passages were perhaps delivered orally (reference to τις ὑμῶν at Chor. 509.10 Förster Richtsteig; cf. [Lib.] .118.18 21 Förster similar to Him. 9.1.10 11 Colonna and esp. [Lib.] .528.15 16 F.: Πεπεισμένος . . . ὑμᾶς εὐγνώμονας ὑπάρχειν ἀκροατάς). Personal in tone (Chor. 435.10 F. R.: Φιλέλλην γάρ τις ἐγώ; defence of wordiness in a speech set in Sparta: Lib. .444.11 14 F.; prayers for success: Chor. 315.21 3 and 511.12 14 F. R.), they comment on the choice of theme (Chor. 509.15 F. R.: τῷ πλάσματι στρατιώτην εἰσήγαγον) and role (Chor. 253.5 8 F. R.: I am neither a miser nor a father of sons, but have taken τὴν μίμησιν ἑκατέρας ποιότητος ἐκ τῆς τέχνης; cf. 510.10 F. R.), and on the tone adopted and the characters involved. There is never a proper division. Chor. 154.8 F. R. and Lib. .551.12 F. mention κεφάλαια, but do not list them. That is the job of οἱ τὰς διαιρετικὰς τέχνας συγγράψαντες ([Lib.] .649.3 4 F.). A Senecan style division is given in the προθεωρία to Him. 9.2.12 23 C. [See now generally Nolè (2013 14).] ³³ Winterbottom (1984); see esp.  . For a characteristic division see Decl. min. 249.1. There is often comment on tone in the Chorician manner (e.g. 259.1 with Winterbottom ad loc.). ³⁴ It should not be forgotten how young the pupils were; cf. [Dion. Hal.] Ars rhet. p. 331.15 17 Us. Rad.: ὡσπερεὶ παιδαρίων ἐν διδασκαλείῳ ἐπιδεικνυμένων καὶ παραμυθίαν τοῦ διδασκάλου ἀπονέμοντος, ἵνα μὴ κλαίῃ τὰ παιδία (perhaps, however, of progymnasmatists?). For pupils as μειράκια with παιδαγωγοί see Philostr. VS 614 and 618.

Sopatros’ Διαίρεσις ζητημάτων

141

I V . DISPOSITIO This aspect of declamation was given attention as early as the First Sophistic.³⁵ Gorgias’ Palamedes³⁶ is, as G. A. Kennedy³⁷ rightly says, ‘not an epideictic speech, but an example of arrangement, argument, and style’. Palamedes is being accused by Odysseus on the charge of having betrayed Greece to ‘the barbarians’ (3).³⁸ His speech is therefore in form a second speech: the facts would have been rehearsed by the prosecutor, and Palamedes realistically does not repeat them. He therefore passes straight from the proem (1–4) to the argumentation (6–32), which is prefaced only by a partitio explaining what points he proposes to make (5). The speech ends with an epilogue, sharply marked off by its first words: Λοιπὸν δὲ περὶ ὑμῶν πρὸς ὑμᾶς ἐστί μοι λόγος, ὃν εἰπὼν παύσομαι τῆς ἀπολογίας (33). This structure will have been the familiar doctrine of fifth-century τέχναι, including Gorgias’ own;³⁹ by the time of Plato further elaborations had taken place, and they are duly mocked in Phaedr. 266d ff. Gorgias’ technique is sophisticated enough to make pedagogic allusion to precepts that he chooses to neglect: those dictating in the epilogue appeals to pity⁴⁰ (33: ἐν ὄχλῳ . . . οὔσης τῆς κρίσεως χρήσιμα) and recapitulation of the facts (37: πρὸς . . . φαύλους⁴¹ δικαστὰς ἔχει λόγον), both inappropriate in this ³⁵ Philostratus distinguishes the two Sophistics (the second of which began, it should be remembered, in the late fourth century ) by their themes (τὰ φιλοσοφούμενα in the first, while the second τοὺς πένητας ὑπετυπώσατο καὶ τοὺς πλουσίους καὶ τοὺς ἀριστέας καὶ τοὺς τυράννους καὶ τὰς ἐς ὄνομα ὑποθέσεις, ἐφ’ ἃς ἡ ἱστορία ἄγει (VS 481), and by the fact that μετεχειρίζοντο τὰς ὑποθέσεις οἱ μὲν (i.e. those following Aeschines) κατὰ τέχνην, οἱ δὲ ἀπὸ Γοργίου κατὰ τὸ δόξαν (ibid.)). Neither distinction is altogether just. The Palamedes is not philosophical, and Gorgias (and other sophists) wrote rhetorical τέχναι. All the same, the Palamedes does not have the (limited) specificity of later declamation. And one important difference between the two Sophistics should be added: whereas the later rhetoricians trained their pupils to declaim themselves, the early sophists seem to have restricted themselves to providing ready made fair copy speeches which pupils had to memorize (Plat. Phaedr. 228a; Aristot. Soph. el. 183b.36 ff., commenting that such training is ταχεῖα μέν, ἄτεχνος δέ). ³⁶ Even the Helen, while operating at a high level of generality, is formally the defence of a particular person on a kind of charge (8: ἀπολογήσασθαι καὶ τὴν αἰτίαν ἀπολύσασθαι). For its careful division and self consciously marked sections see Norden (1898), 386 n. 2, with later parallels. ³⁷ Kennedy (1963), 170. ³⁸ The trial is not, naturally, part of the original story, for which see Frazer (1921) on Apollod. Epit. 3.8. ³⁹ Duly exemplified in the Tetralogies attributed to Antiphon. Thus in 1.α we have 1 3 proem; no narration (even though this is a first speech: perhaps this head did not much interest the first sophists, who placed their emphasis on argumentation and emotional appeal); 4 9 argumenta tion; 9 11 epilogue. ⁴⁰ Socrates too avoided emotional appeal, for different reasons (Plat. Ap. 34c ff.). It is exemplified in the Tetralogies (e.g. 2.β.11). ⁴¹ Compare later Aristot. Rh. 1415b.4 7: Δεῖ δὲ μὴ λανθάνειν ὅτι πάντα ἔξω τοῦ λόγου τὰ τοιαῦτα· πρὸς φαῦλον γὰρ ἀκροατὴν καὶ τὰ ἔξω τοῦ πράγματος ἀκούοντα· ἐπεὶ ἂν μὴ τοιοῦτος ᾖ, οὐθὲν δεῖ προοιμίου (also 1404a.7 8), leading to Quint. 12.10.52 3 (Cum vero iudex detur . . . populus . . . ).

142

Papers on Quintilian and Ancient Declamation

court of ‘the first of the Greeks’ (37). This is the speech of a rhetor who, no less than Sopatros, would impress on his pupils the importance of clearly separated proem, narration (where appropriate), argument, and epilogue. Later declamation shows the same concern to mark off parts of the speech clearly,⁴² and even to allude archly to those parts.⁴³ That is natural enough in the didactic Minor Declamations, where one proem concludes with the words: Cuius si vos auctoritas, iudices, movet, illud unum in prima parte causae dixisse contentus sum, vocem eius ab utraque parte recitari, while the narration immediately starts thus: Quod si vindicari filium mallet, satis eum vel sola expositio causae tueretur (338.10–11). But it is more surprising to find in the elegant Major Declamations the end of a narratio marked off with ut ipse tristem finiam expositionem (13.6 [271.12 Håkanson]) or allusion made in an epilogue to the τέχνη’s demand for emotional effect at this point (Maior forsitan materia videatur adfectus: 13.15 [282.11–12 H.]). Rather differently in the only complete speech that Seneca set himself to give us, though it is rudely cut short by the manuscripts: the proem is wound up, in a way typical of the Silver Age, by a resounding epigram: tantum in istam dives amator effudit, post poenam quoque expediat fuisse adulteram (Con. 2.7.1). And the summary of events that begins the narrative is given a figured form that reduces its bareness (Con. 2.7.2):⁴⁴ Quae praeceperim uxori proficiscens, scio; cetera, quemadmodum adulescens formosus, dives, ignotus in viciniam formosae et in absentia viri nimium liberae mulieris commigraverit, quemadmodum adsidua satietate cotidianae per diem noctemque libidinis exhaustis viribus perierit, interrogate rumorem (with allusion to ex suspicione in the theme, and with an ingeniously damaging explanation for the death of the young man). Sopatros would have admired that: for it is clear that, for all his array of headings, he was concerned to pass as smoothly as possible from one subsection to another, with no abrupt change of gear.⁴⁵ ⁴² So too in real life oratory, starting with Antiphon (e.g. Her. 25: Τὰ μὲν γενόμενα ταῦτ’ ἐστίν· ἐκ δὲ τούτων ἤδη σκοπεῖτε τὰ εἰκότα). Image and reality continue to intertwine throughout the history of ancient eloquence. ⁴³ So in Luc. Abd. 3 the transition to the narration is marked by the word διηγησάμενος. Does Tyr. 14: Καί μοι πρὸς θεῶν ἤδη ἀπ’ ἀρχῆς ἐς τέλος . . . πάντα ἐξετάσατε give us our first whiff of the technical term ἀπ’ ἀρχῆς ἄχρι τέλους? Note also how in [Quint.] Decl. mai. 13.15 (282.7 9 Håkanson) the advent of the epilogue is marked by the words: Intellego neque prudentiam vestram desiderare plura de causa neque vestram fidem ac religionem egere exhortatione vere iudicandi. ⁴⁴ Contrast the studied bareness of ‘Quint.’ Decl. min. 316.5: Iuvenis frugaliter vixi, patrimo nium auxi, uxorem duxi, filium sustuli, hunc amo (with parallels some from Sopatros given by Winterbottom ad loc.). ⁴⁵ Note the transitions recommended at p. 12.14 15 [= 1.8.1 2 We.]; p. 58.1 5 W. [= 9.4.1 4 We.] (going straight into the narration is σφόδρα . . . ἀρρητόρευτον); p. 272.16 17 W. [= 46.5.1 2 We.]: εὐφυῶς καὶ ἀκολούθως ἐκ τῆς διανοίας εἴσβαλλε εἰς τὴν μετάληψιν (with Innes Winterbottom (1988), 209). See further Sen. Con. 1.1.25: Hermagoras transit a prooemio in narrationem eleganter eqs. (also Quint. 4.1.77 9).

Sopatros’ Διαίρεσις ζητημάτων

143

That could hardly be said of Gorgias. The partitio of the Palamedes (5) establishes a double argument: οὔτε γὰρ βουληθεὶς ἐδυνάμην ἂν οὔτε δυνάμενος ἐβουλήθην ἔργοις ἐπιχειρεῖν τοιούτοις. These two sections, when they come, are self-consciously marked off at beginning and end: (a) Ἐπὶ τοῦτον δὲ τὸν λόγον εἶμι πρῶτον, ὡς ἀδύνατός εἰμι τοῦτο πράττειν (6) ~ Πάντως ἄρα καὶ πάντῃ πάντα πράττειν ἀδύνατον ἦν μοι (12); (b) Σκέψασθε κοινῇ καὶ τόδε· τίνος ἕνεκα προσῆκε βουληθῆναι ταῦτα πράττειν, εἰ μάλιστα πάντων ἐδυνάμην; (13) ~ Ὅτι μὲν οὖν οὔτ’ ἂν ἐβουλόμην προδοῦναι τὴν Ἑλλάδα, διὰ τῶν προειρημένων δέδεικται (21, summing up both together). Such care for articulation is parodied by Plato in Agathon’s speech in the Symposium.⁴⁶ It is not so much a feature of Gorgias’ style as a sign of his desire to encourage pupils to speak in an ordered and planned way.⁴⁷ That is true also of similar phenomena in the Minor Declamations. Thus in 341 Hoc ius quale sit postea videbimus (1) announces that a discussion of the fairness of the law will follow arguments as to its applicability in the present case; and that expectation is kept before us by the words Iterum dicam: si quid iniqui habet lex nostra, postea viderimus (3). The passage on the applicability of the law ends at 5 with haec de iure, and the section on aequitas is introduced by the explicit objection Sed ius ipsum iniquum est.⁴⁸ Similarly motivated are the frequent summaries of the position so far reached in the argument. Sometimes the summary sounds like a sermo (e.g. 252.22), but elsewhere (e.g. 315.14: Ergo haec dixi adhuc: debetur mihi quodcumque praemium eqs.) these recapitulations, however naively introduced, are clearly meant to be part of the declamation itself. Sopatros can occasionally be seen managing these things, and with aplomb. At p. 118.7–10 W. [= 20.7.17–20 We.],⁴⁹ before turning to πηλικότης, he makes elegant allusion to the preceding arguments from συλλογισμός and γνώμη νομοθέτου. Again, at p. 120.21–2 W. [= 20.10.3 We.]⁵⁰ he remarks that an ἐπιλογικὸν νόημα should mark the end of the treatment of the first ὅρος (an illustration follows in 23–5 W. [= 4–6 We.]). Nor were such devices restricted to declamation. Thus Cicero in the admittedly scholastic Pro Milone: Video adhuc constare, iudices, omnia, with what follows (52).

⁴⁶ See the edition of Bury (1932²),  with n. 1. ⁴⁷ It will then be part of his Asianism (i.e. his dependence on Greek declamatory methods) that Hortensius was over meticulous in division (e.g. Quint. 4.5.24). Cicero grew out of it (Norden (1898), 386 n. 2; Militerni Della Morte (1977), 36 n. 42, 63 5), and made fun of Hortensius for it as early as the Verres trial (Div. Caec. 45). ⁴⁸ Similarly in Lucian’s Abdicatus. Argument from the law is clearly introduced (8: ἀπὸ τοῦ νόμου ἄρξομαι) and clearly concluded (13: πολλὰ εἰπεῖν ἔχων ὅμως παύσομαι). The section on aequitas then starts Σκέψασθε . . . ὅντινα ὄντα καὶ ἀποκηρύξει με (13). Both sections are summa rized in 19 (ἱκανῶς . . . δέδεικται). ⁴⁹ See Innes Winterbottom (1988), 97. ⁵⁰ See Innes Winterbottom (1988), 98.

144

Papers on Quintilian and Ancient Declamation

V. I N V E N T IO We have seen how intimately Sopatros’ book is bound up with the στάσις system. The headings of a case vary according to its ‘issue’. Students were not made to master the system merely for the sake of acquiring theoretical knowledge. Rather, they were acquiring the key to the whole procedure of invention: they had to identify the στάσις of a case before they could know the correct way to argue it. Though the system was not formulated expressly—it would seem—until Hermagoras in the second century , there is no doubt that it has its roots much earlier. Antiphon’s Tetralogies are intended to illustrate the treatment of different kinds of case: ‘The first is concerned with a question of fact . . . The second is about a question of responsibility . . . The third deals with . . . putting the blame elsewhere. These themes prefigure the later differences of stasis or “issue”; and it was precisely to illustrate such differences that declamation themes were developed and classified.’⁵¹ The outcome of this later development is to be seen not in extant work of the Hellenistic period, of which there is almost none, but in the way in which the Ad Herennium and Cicero’s De inventione use declamation themes to illustrate their discussion of various στάσεις.⁵² As we see from the hypotheses to the speeches of the Attic orators, it was possible to classify real orations according to their στάσις: naturally, for the system was meant to be logically all-inclusive; and one tried to see in a Demosthenes the arguments elaborated by rhetoricians to deal with cases of the appropriate issue.⁵³ But the complexities of real life veiled the eyes of pupils from the clear light of rhetorical truth; a declamation theme, stripped of all unnecessary περίστασις, was a simpler tool for a preceptor to employ, just as his fair-copy declamation would economically illustrate the argumentation he recommended without distracting elements.

⁵¹ Russell (1983), 17. ⁵² See Bonner (1949), 11 26. But it is not enough to say that ‘there existed in the heart of the Greek rhetorical handbooks of the 2nd century , these germs of the fully blown controversia’ (pp. 15 16). More justly, Russell (1983), 20: ‘Fragmentary and vague as our knowledge is, meletē, we may be sure, was familiar in the Hellenistic age.’ ⁵³ See especially the hypotheses to the In Midiam (and the scholia on the speech: cf. Innes Winterbottom (1988), 4 on p. 85.16 W. [= 13.8.10 We.]), and, less detailed, for the speeches of e.g. Isaeus and Lycurgus. Such classification would, of, course, have been unknown to the orators themselves, but it probably began soon after Hermagoras (see Lossau (1964), 111 ff.). For rhetoricians’ treatment of Demosthenes see also Kennedy (1983), 119. Reading of Demosthenes will have gone on parallel with the composition of declamation, and we can see Sopatros pillaging him (see Innes Winterbottom (1988), 308 s.v.; also Winterbottom (1982), 251 [= A.5 above, p. 75]). It is significant that Hermogenes and Minucianos wrote commentaries on Demosthenes (Glöckner (1901), 25). Note also the allusion in RG .191.30 Walz to ὁ τὸν λόγον (sc. the κατὰ Τιμάρχου of Aeschines) ἐξηγούμενος.

Sopatros’ Διαίρεσις ζητημάτων

145

All this is present in embryo in the Palamedes. The argumentation is ἐκ τῶν εἰκότων,⁵⁴ and hardly any attempt is made to use even available mythological detail. Fascinatingly, too, we can see here clear signs of moves that continue to be made down the ages in a case of στοχασμός or ‘conjecture’. Palamedes, as we have already seen, relies, in the absence of knockdown evidence of his innocence, on two arguments from probability: that he was in no position to betray the Greeks, and that he had no motive for doing so.⁵⁵ The details of this double assertion correspond so neatly with later stochastic procedures that it is tempting to suppose that Gorgias’ own τέχνη contained these headings, so helping to lay down the foundation for later doctrine, and that the Palamedes, like later declamation, consciously illustrates its author’s precepts in inventio as well as dispositio. As to τὸ δυνατόν, Palamedes raises questions of place, opportunity, and time that appear regularly in later theoretical discussion.⁵⁶ But particularly striking is his treatment of βούλησις. He suggests, naturally only to dismiss, various motives that might have led him to betray the Greeks: τοῦ τυραννεῖν (ἕνεκα) (13), πλούτου καὶ χρημάτων ἐρασθείς (15), τιμῆς ἕνεκα (16), φίλους ὠφελεῖν βουλόμενος ἢ πολεμίους βλάπτειν (18).⁵⁷ Three of these motives are among those postulated in Sopatros’ Case 1, where it is suggested that Alcibiades might have wished to be a tyrant for four reasons: πρώτη, τὸ ἐπιθυμεῖν δυναστείας· δευτέρα, τὸ μείζονος τιμῆς ὀρέγεσθαι· τρίτη, τὸ βούλεσθαι τοὺς ἐχθροὺς τιμωρήσασθαι· τετάρτη, τὸ τοὺς φίλους ἐθέλειν εὐεργετεῖν (pp. 10.27–11.2 W. [= 1.6.18–21 We.]). A different three recur in the Ad Herennium’s discussion of causa in conjectural cases (2.3: quaeritur num quod commodum maleficio appetierit, num honorem, num pecuniam, num dominationem).⁵⁸ The Palamedes, then, may well consciously illustrate the arguments recommended in Gorgias’ rhetorical handbook. Such illustration of theoretical

⁵⁴ Note οὐκ εἰκός (9). Probabilities naturally remained crucial for all forensic argument, whether in court or schoolroom. Note in Sopatros esp. p. 7.16 ff. W. [= 1.4.1 ff. We.] (where εἰκότα are contrasted with witnesses). ⁵⁵ Cf. Quint. 7.2.27: Nam is ordo est, ut facere voluerit, potuerit, fecerit. ⁵⁶ Note 10: Πότερα δ’ ἐκόμισαν ἡμέρας ἢ νυκτός; (add. Reiske) Ἀλλὰ πολλαὶ καὶ πυκναὶ φυλακαί, δι’ ὧν οὐκ ἔστι λαθεῖν. [Ἀλλ’] (del. Winterbottom) Ἡμέρας; Ἀλλά γε τὸ φῶς πολεμεῖ τοῖς τοιούτοις. Cf. Rhet. Her. 2.7: Tempus ita quaeritur: quid anni, qua hora, noctu an interdiu. Also 11: αὐτὸς ἔπραττον ἢ μεθ’ ἑτέρων; Ἀλλ’ οὐχ ἑνὸς ἡ πρᾶξις. Ἀλλὰ μεθ’ ἑτέρων; Τίνων; Δηλονότι τῶν συνόντων. Πότερον ἐλευθέρων ἢ δούλων; Cf. Rhet. Her. 2.7: Spes celandi quae fuerit quaeritur ex consciis, arbitris, adiutoribus, liberis aut servis aut utrisque (also Cic. S. Rosc. 74: si per alios fecisse dicis, quaero quos, servosne an liberos; Cael. 53); so too Sopat. p. 46.10 W. [= 7.4.3 We.]. The urgent questions (some cited above) in Palam. 6 12 are common in later declamation (e.g. Sopat. p. 36.4 9 W. [= 8.5.1 6 We.]; ‘Quint.’ Decl. min. 267.8), and are not alien to court practice (see Quint. 7.2.54 5). ⁵⁷ Cf. Aristot. Rh. 1399b.35 6: εἰ . . . ὠφέλιμον ἢ αὐτῷ ἢ φίλοις ἢ βλαβερὸν ἐχθροῖς. ⁵⁸ Note too how Palamedes uses the argument ex ante acta vita: μάρτυρα πιστὸν παρέξομαι τὸν παροιχόμενον βίον (15); cf. Rhet. Her. 2.5: Defensor . . . demonstrabit vitam integram, si poterit. For the wording cf. Cic. Sul. 79: hoc maxime interest . . . vitam unius cuiusque esse testem.

146

Papers on Quintilian and Ancient Declamation

precept on argumentation was always an important feature of declamation. A regrettably unpublished Tübingen dissertation of Professor Joachim Dingel (1972) demonstrated how the Minor Declamations presuppose and illustrate the arguments that the στάσις system organized.[⁵⁹] Thus in 281 there is appeal to past life (3), motive (3, 5), place (ibid.), means (ibid.), and opportunity (5):⁶⁰ all to be found in Quint. 7.2.27 ff. as well as other textbooks. Again, in Choricius ancient readers themselves observed the same conformity to the headings of the rhetores. Thus, in the twelfth declamation the scholiast marks in the margin the Hermogenean κεφάλαια: προβολή, ἀνθορισμός, γνώμη νομοθέτου, πηλικότης.[⁶¹]

V I . E L O C U T I O: THE ‘A S I A N’ STYLE In the fields of dispositio and inventio Sopatros fits snugly into the continuing history of declamation. So far as style is concerned, the matter is more complex. I first survey the characteristics of the ‘Asian’ style,⁶² which, though in part engendered by the First Sophistic, has its main roots in the Hellenistic declamation hall. Throughout antiquity declaimers were interested in style, often a distinctive one, vulnerable to parody⁶³ and hostile comment. The excesses of Gorgias, as displayed for the admiration and imitation of his listeners in the Palamedes and Helen, are well enough known; and they duly influenced writers of more realistic vein, Thucydides and Isocrates. Later, the veil hardly lifts on the declamatory prose of the Hellenistic period. When it does, we find in the historical fragments of the Asian⁶⁴ Hegesias, besides characteristics reminiscent ⁵⁹ [This dissertation was eventually published: see Dingel (1988).] ⁶⁰ Coniectura (3) alludes to the στάσις in operation (see above, p. 142), though the sermo, characteristically (see above, p. 140), does not go into the technicalities. ⁶¹ [See Förster Richtsteig (1929), 518, 519, 520, 521.] ⁶² I illustrate from Hegesias of Magnesia, from the Elder Seneca’s Greek declaimers, who often came from Asia, from his Latin declaimers, who were much influenced by the Greek ones, and from Cicero, who was criticized in his lifetime for being ‘Asian’ and who certainly received part of his education in Asia. I do not mean to imply that the concept of an Asian style is at all fixed: I use features of these authors that were in antiquity regarded as showing Asian signs. Nor do I imply that all Seneca’s declaimers employed the same style (see the valuable discussion of Fairweather (1981), 243 303, with excellent words of warning on p. 302), or that it is sensible to put Cicero, who rightly prided himself on his variety, in this category. ⁶³ Petr. 1.1: Haec vulnera pro libertate publĭca ēxcēpī̆ , hunc oculum pro vobis impendi; date mihi [ducem] qui me ducat ad liberos meos, nam succisi pŏplĭtēs mēmbrā nōn sustĭnē̆ nt, with high language, mock pathos, and rhythm. Similarly, Plato had parodied the style of the first sophists (see e.g. Gorg. 448c with Dodds’s note). ⁶⁴ He was from Magnesia on the Meander. Stylistically he ἦρξε μάλιστα τοῦ Ἀσιανοῦ λεγομένου ζήλου (Strab. 14.1.41), though he thought himself Attic (Cic. Brut. 286) and an imitator of

Sopatros’ Διαίρεσις ζητημάτων

147

of Gorgias, new qualities regarded as making up the so-called Asian oratory of the following centuries: oratory principally exemplified for us by the exiguous fragments of Greek declaimers (many of them actually hailing from Asia)⁶⁵ preserved by the Elder Seneca. Concinnity was one of Cicero’s two types of Asiatica dictio—sententiosum et argutum, sententiis non tam gravibus et severis quam concinnis et venustis (Brut. 325),[⁶⁶] and it accordingly was exploited by Hegesias, whose style, for Cicero, was in ipsa, quam tamen consequitur, concinnitate puerile (Brut. 287). But it was not Hegesias’ invention: In huius concinnitatis consectatione Gorgiam fuisse principem accepimus (Orat. 165). Cicero (Orat. 167) particularly connects Gorgias (and Isocrates) with the use of antithesis, and regards himself as the heir to this tradition (Nos etiam in hoc genere frequentes). Examples from Gorgias are hardly necessary; from Hegesias note the brief ἐκεῖνο Λεωκόριον, τοῦ̅ τ̅ ο̆ Θησεῖ̅ ̅ο̆ν (FGrH 142, fr. 24). In Seneca there is, for instance, Nicocrates’ τοῦ μὲν ἰδίου μάρτυρος ἐφείσατο, τοῦ δ’ ἐμοῦ κατεφρόνησεν (Con. 7.5.15). The Latin declaimers, who so often take their tone from their Greek contemporaries,⁶⁷ do not hold back (e.g. Con. 7.5.9: maritum occidit, adulteram strinxit), and Cicero (Orat. 167) illustrates his own predilection for contrast from the fourth Verrine (115).⁶⁸ Closely related to concinnity is word play. One may illustrate from play on verbal prefixes. Hegesias’ ὄνομα κατελάβομεν πόλιν κᾰτᾰλῐπόντε˘ ς (fr. 6)⁶⁹ Lysias (Orat. 226). Many of our fragments are preserved by Agatharchides via Photius (Bibl. 445b ff.), who hesitates between κομψότης and μανία as a description of his style (441a). ⁶⁵ Thirty six are named. We know the homes of about ten: one came from mainland Greece (Sparta, oddly enough), three from the islands (Euboea, Mytilene), one from the still Greek city of Massilia, and the rest from Asia Minor (Smyrna, Tralles, Cappadocia, Pergamum, Mylasa). One of these, described by Seneca as Graecus, we know to have come from Smyrna (Cestius: Con. 7.1.27); so that other Graeci may not have been from the mainland. ⁶⁶ [On Cic. Brut. 325, a capital source on Asian style(s), see now Lucarini (2015).] ⁶⁷ Seneca, for all his nationalistic feelings, was well aware of the influence of Greek declaimers on Latin (see Winterbottom (1983a), 58 [= A.7 above, pp. 103 4]) in his own day. But his account of the origins of declamation (Con. 1.pr.12) is blithely unaware of earlier Greek background, just as Suetonius’ account of rhetoric in Rome (Rhet. 25 30) restricts itself to the Latin teachers. Romans naturally showed a keen interest in Plotius Gallus and the Latini rhetores; but it was the Greek ones who counted, even in Cicero’s youth. When he was keen to attend Plotius’ classes, Continebar . . . doctissimorum hominum auctoritate qui exstimabant Graecis exercitationibus ali melius ingenia posse (ap. Suet. Rhet. 26.1). And that went on: Augustus was taught by Apollodorus of Pergamum (Suet. Aug. 89.1), Tiberius by Theodorus of Gadara (Suet. Tib. 57.1). Antony (accused of Asian volubilitas by Augustus: Suet. Aug. 86.3, cf. Plut. Ant. 2.8) practised with Sex. Clodius, e Sicilia Latinae simul Graecaeque eloquentiae professor (Suet. Rhet. 29.1), as we know from the Philippics (e.g. 2.101). ⁶⁸ Indeed Cicero is perhaps a better illustration for the continuance of Gorgianic isocola than the declaimers, who become more interested in point than in symmetry; but see Norden (1898), 288 90. Cf. also Plut. Comp. Dem. et Cic. 2.2: Cicero rivalled the sophists Isocrates and Anaximenes. ⁶⁹ For other kinds in Hegesias note fr. 7: φωνήσαντα . . . ἄφωνον, and fr. 14: ἄσπορον . . . Σπαρτούς (on which Agatharchides commented that the writer ἐκ τῶν ὀνομάτων τὴν ἐναντίωσιν

148

Papers on Quintilian and Ancient Declamation

looks forward (for example) to Diocles’ Ἂν ἐπιτύχῃς, μίαν προσθήσεῑς ἀ̆ ρῑστείᾱ̆ ν· ἂν ἀποτύχῃς, τρεῖς ἀριστείᾱς ἀ̆ πο̆λέ̆σει˘ ς (Sen. Con. 1.8.15) and a Latin epigram like the anonymous Peribit ergo quod Cicero scripsit, manebit quod Antonius proscripsit? (Suas. 7.11). Cicero started early⁷⁰ in this manner (S. Rosc. 3: non modo ignoscendi ratio verum etiam cognoscendi consuetudo iam de civitate sublata est), and Cassius Severus thought him the heir of the Atellans in punning (Sen. Con. 7.3.9).⁷¹ Hegesias is linked in bombast⁷² (τὸ παρὰ μέλος οἰδεῖν) by ‘Longinus’ (Subl. 3.2) with Gorgias and others: πολλαχοῦ γὰρ ἐνθουσιᾶν ἑαυτοῖς δοκοῦντες οὐ βακχεύουσιν,⁷³ ἀλλὰ παίζουσιν. Hegesias’ fragments give us (FGrH 142, fr. 12) the following inflated passage: Ἀλέξανδρε, Θήβᾱς κᾰτᾱσκάψα˘ ς ὡς ἂν εἰ ὁ Ζεὺς ἐκ τῆς κατ’ οὐρᾰνὸν με̆ρί̆δο˘ ς ἐκβά̆λοι τὴν σε̆λήνη˘ ν· ὑπολείπομαι γὰρ τὸν ἥλῐον ταῖς Ἀ̆ θήναι˘ ς. Δύο γὰρ αὗται πόλεις τῆς Ἑλλάδος ἦσᾰν ὄψει˘ ς. Διὸ καὶ περὶ τῆς ἑτέρας ἀγωνῐῶ νῦ˘ ν. Ὁ μὲν γὰρ εἷς αὐτῶν ὀφθαλμὸς ἡ Θηβαίων ἐκκέ̆κοπταῑ πό̆λι˘ ς.⁷⁴ Even the snippets left us by the Elder Seneca from Greek declaimers provide something; when Dorion paraphrases Homer (presumably in a suasoria on Alexander the Great), he produces an exaggeration that Seneca himself calls tumidum and inflatum (Suas. 1.12): ὄρους ὄ̆ ρο̆ς ἀ̆ ποσπᾶται˘ καὶ εἴληφεν, οὐκ ἐκ τοῦ πράγματος, citing parallels from Hermesianax of Colophon). For play on prefixes see Traina (1987⁴), 85; the book has excellent material on word play generally. ⁷⁰ His last published words (Phil. 14.38: si vivi vicissent qui morte vicerunt) illustrate another kind of word play, where different parts of the same word are set off against each other. Note in Seneca Con. 7.1.26: Διήγησαι νῦν, πάτερ, πῶς ἀπέλυσεν οὕτως ἀπολυθείς (complete with pun); 8.5: Uterque nostrum cum rogatur fastidit, cum relinquitur rogat. See Bonner (1949), 69 70 for paronomasia in Seneca generally. ⁷¹ Cf. Quint. 6.3.2 and Plut. Comp. Dem. et Cic. 1.4 5 on Cicero’s excessive enthusiasm for jokes. ⁷² Concinnity and bombast are two types of κακοζηλία for Diomedes (GL .451.10 Keil: aut nimio cultu (cf. Sen. Suas. 2.23) aut nimio tumore). Similarly, in a longer list, Quint. 12.10.73: aut puerilibus sententiolis lascivit aut inmodico tumore turgescit. Cicero’s distinction of two types of Asiatica dictio (Brut. 325: volucre atque incitatum, with admirabilis orationis cursus, as opposed to concinnity: see p. 147) is perhaps somewhat different. Note also the contrast not explicitly made by Cicero of bombast with declamatory point (cf. Sen. Con. 9.6.1: Nefaria mulier, filiae quoque noverca; also 6.6 [163.14 Håkanson]: Generi adultera, filiae paelex) at Orat. 107: though it may be significant (the observation of an anonymous pupil of mine) that the epigram there quoted from the Pro Cluentio (199: uxor generi, noverca fili, filiae paelex) is preceded in that speech by a period as bombastic as, and rather similar in structure to, the extravaganza from the Pro Roscio (72) previously cited in Orat. 107. ⁷³ I do not press the parallel of that ‘Dithyrambus in Prosa’ (Norden (1898), 143), the inscription of Antiochus of Commagene on the Nimrud Dağ (studied, with its related in scriptions, in Dörner Goell (1963), 36 91: I owe this reference to Mr P. J. Parsons). That is self encomium and religious legislation, not declamation. Its sentence structure is more like Isocrates’ than Hegesias’, and it does not seem as consistently rhythmical as Norden implied (see further Waldis (1920), 3 11, 58 62). ⁷⁴ Exaggerations such as those in fr. 8 are also relevant. Agatharchides comments on the μεταφορὰ σκληρά in fr. 9 (cf. Phld. Rh. .180.15 ff. Sudhaus); cf. frr. 16 17 and the extravagant metaphors attacked by Seneca in Con. 10.pr.9 in the course of his assault on Musa, in whom omnia usque ud ultimum tumorem perducta.

Sopatros’ Διαίρεσις ζητημάτων

149

χεῑρί̆ ᾱ βάλλε̆ται̅ νῆσο˘ ς.⁷⁵ The manner transplants readily into Latin. Musa’s Charybdis ipsius maris naufragium . . . Quid ibi potest esse salvi ubi ipsum mare perit? (Suas. 1.13: the first sentence a portentum itself in Seneca’s eyes) finds its counterpart in Cicero’s Quae Charybdis⁷⁶ tam vorax? Charybdim dico? Quae si fuit, fuit animal unum: Oceanus, me dius fidius, vix videtur tot res . . . absorbere potuisse (Phil. 2.67). And his critics saw him as tumidiorem et Asianum (Quint. 12.10.12). Intimately connected with the bombast of Hegesias is his desire to arouse strong feeling. We have seen how he strove to rouse indignatio against Alexander over the destruction of Thebes (fr. 12),⁷⁷ and other fragments seek to call forth the counterpart of anger, pity. Thus Βασιλικῇ μανίᾳ προσπταίσασα πόλις τραγῳδίας ἐλεεινοτέρα γέγονε (fr. 10), or Αἱ δὲ πόλεις αἱ πλησίον ἔκλαιον τὴν πόλιν, ὁρῶσαι τὴν πρότερον οὖσαν οὐκέ ̅ ̅ ̆ τ’ οὖσα ̅ ̅ ˘ ν (fr. 13). A fragment only preserved to us in Latin translation, and apparently from a forensic speech, has threefold anaphora of miseremini (fr. 27).⁷⁸ Hegesias, it seems, felt in history the orator’s desire to rouse the emotions so long exploited in the lawcourt,⁷⁹ in this akin to the historians criticized by Polybius for confusing

⁷⁵ Maecenas, that expert in good taste, contrasted this passage unfavourably with Verg. Aen. 8.691 2 and 10.128. Cf. also ‘Demetr.’ Eloc. 115: (frigidity) καθάπερ ἐπὶ τοῦ Κύκλωπος λιθοβολοῦντος τὴν ναῦν τοῦ Ὀδυσσέως ἔφη τις· ‘Φερομένου τοῦ λίθου αἶγες ἐνέμοντο ἐν αὐτῷ.’ Among the Senecan Greek passages note also Suas. 1.11 (Artemon): Οὐ ταῖς Ἑλλησποντίαις ᾐό̆σῐν ἐ̆ φεστῶτε˘ ς οὐδ’ ἐπὶ τῷ Παμφυλί̆ῳ πε̆λά̆γει˘ τὴν ἐμπρόθεσμον καραδοκοῦμε̆ν ἄμπωσι˘ ν· οὐδ’ Εὐφράτης τοῦτ’ ἔστῐν οὐδ’ Ἰνδό˘ ς, ἀλλ’ εἴτε̆ γῆς τέρμα, εἴτε φύσεως ὅρος, εἴτε πρεσβύτατον στοιχεῖον, εἴτε γένεσις θεῶν, ἱερώτερόν ἐστιν ἢ κατὰ ναῦς ὕδωρ. Also Con. 7.1.25 (Greek original of an epigram by Triarius); Suas. 1.16 (Glycon). Among the Latin declaimers, Seneca criticizes Murredius for speaking tumidissime at Con. 9.2.27 (he remarks on the way in which Murredius padded out a tetracolon ut impleretur numerus; if this refers to the rhythm, rather than to the number of cola, cf. Cic. Orat. 230: Apud . . . Asiaticos maxime numero servientes inculcata reperies inania quaedam verba quasi complementa numerorum (with Kroll’s note); note also the criticism of Isocrates by Dion. Hal., e.g. Dem. 19). Cf. also Con. 10.1.14; Fairweather (1981), 217 18. For declaimers who liked to ‘talk big’ cf. Suas. 2.17; Sen. Ep. 114.11; Tac. Dial. 35.5: ingentibus verbis. ⁷⁶ Mart. Cap. RLM 473.24 5 = 5.512 Willis: nec longe petita debent esse translata, ut si dicas luxuriosum ‘Charybdim’. For harsh metaphors see n. 74 above. ⁷⁷ Indignatio merges with miseratio, as often. Agatharchides (cf. e.g. ‘Demetr.’ Eloc. 28) comments perceptively that Hegesias is merely ‘jesting’ (σκώπτειν): and seems σκοπεῖν πῶς ἂν τάχιστα συγκόψαιτο τὸν λόγον, οὐ πῶς τὸ πάθος ὑπὸ τὴν ὄψιν ἀγάγοι διὰ τῆς ἐνεργείας (v.l. αργ ). Equally acutely, he says of the word play of fr. 6 (above, pp. 147 8) that Τοῦτο πάθος μὲν οὐδαμῶς ποιεῖ, ἀλλὰ συνάγει πρὸς τὴν ἔμφασιν καὶ ποιεῖ ζητεῖν τί λέγει: δισταγμός destroys τὸ δεινόν; and on fr. 8 that Δεῖ . . . τὸν οἰκτιζόμενον ἀφέντα τοὺς ἀστεϊσμοὺς τὸ πρᾶγμα σημαίνειν ᾧ οἰκείωται τὸ πάθος, εἰ μέλλοι μὴ τῇ λέξει διακοσμεῖν, ἀλλὰ τῷ τῆς συμφορᾶς αἰτίῳ προσεδρεύειν. Compare the comment of Dionysius (Comp. 123) on fr. 5: Τούτων οὐκ ἂν ἔχοι τις εἰπεῖν δεινότερα πάθη οὐδ’ ὄψει φοβερώτερα· πῶς δὴ ταῦτα ἡρμήνευκεν ὁ σοφιστής, ἄξιον ἰδεῖν, πότερα σεμνῶς καὶ ὑψηλῶς ἢ ταπεινῶς καὶ καταγελάστως. ⁷⁸ Cf. the pathetic passages of Charisius cited by Rut. Lup. 1.10, 2.6 and 16. Cicero thought him Attic, an imitator of Lysias and a model for Hegesias (Brut. 286). He looks from his fragments rather Ciceronian. ⁷⁹ See above, p. 141 n. 39.

150

Papers on Quintilian and Ancient Declamation

history and tragedy.⁸⁰ The story of πάθος in declamation is not over. Even in the Elder Seneca’s Greeks we can glimpse touches of miseratio: Glycon (Con. 2.3.23) says something like: Βραδέως ἐλεεῖς με· κηρὸς ῥῦσις οὐκ ἔστι. Φθίνω κρυεροτέραν θανάτου μέριμναν· οὐ περιμενῶ σου τὸν ἔλεον. And there are ‘pathetic’ passages in the Latin declaimers.⁸¹ As for Cicero, his prowess in miseratio is well known. He was clear that his control over his audience’s emotions was what marked him out as a great orator. His sterner Atticist critics will have been less enthusiastic. We come finally to rhythm, a particular mark of ‘Asianism’: Cicero was criticized as in compositione fractum, exultantem ac paene . . . viro molliorem (Quint. 12.10.12), and he gives such space to the matter in the Orator precisely because he was himself vulnerable here. The progymnasmatist Theon speaks of τὴν ἔμμετρον καὶ ἔνρυθμον λέξιν as exemplified in τὰ πολλὰ τῶν Ἡγησίου τοῦ ῥήτορος, καὶ τῶν Ἀσιανῶν καλουμένων ῥητόρων (RhG .71.9 Spengel [= p. 16 Patillon]), and we can see in Hegesias⁸² the cretic-based clausulae of the familiar system brought into Latin from Hellenistic Greek and popularized by Cicero. That the Senecan Greek declaimers are equally conscious of the system is less well known.⁸³ It is difficult to give satisfactory statistics in view of the corruptness of the text and the ambiguity of many clausulae. But of 134 ‘strong’ endings of sentences, marked by full-stop, colon, or question mark, 72 are rhythmical in the sense that they exhibit double trochee (known to the ancients as characteristic of Asian rhythm),⁸⁴ cretic plus trochee, or double cretic, or resolved versions of these combinations. As many as 42 of the 62 others are placed in doubt by ambiguities of scansion. Particularly regular declaimers are Adaeus (rhetor ex Asianis) with 6 rhythmical endings out of 7, Damas of Tralles with 4 out of 5, Dorion with 4 out of 6, and Hermagoras with 7 out of 11.

V I I . ‘A T T I C’ DECLAMATION We may classify the elements that were in antiquity regarded as characterizing the Asian style, and we may find illustrations of these elements in Hegesias ⁸⁰ See Winterbottom (1982), 263 4 [= A.5 above, p. 84]. Not that historians had ever rejected emotional appeal: for history and declamation see Plb. 12.25a.5, b.4, k.2, 8, 11; 26d.6. ⁸¹ See Bonner (1949), 62 3. Note also Sen. Con. 7.pr.3: Albucius adfectus efficaciter movit; 7.4.6: aiebat (sc. Latro) . . . ipsam orationem ad habitum eius quem movere volumus adfectus molliendam. ⁸² In the fragments not so far cited note especially ἐπῑστρᾰφεὶς δ’ οὐκέ̆τ’ εἶ δο˘ ν (8). The ex tended narrative in fr. 5 is not rhythmical. ⁸³ There is a very tentative remark in Wilamowitz Möllendorff (1900), 6 n. 1. I have marked rhythm in the Greek fragments from Seneca, as from Hegesias, which I have cited for other reasons. ⁸⁴ Cic. Orat. 212.

Sopatros’ Διαίρεσις ζητημάτων

151

and in the Greek (and Latin) declaimers excerpted by the Elder Seneca. But we should not deceive ourselves: we have no real idea what a complete forensic declamation given in an Asian hall in the early Roman empire, let alone the Hellenistic period, would have been like. Hegesias was not declaiming when he wrote the history from which we have such exiguous fragments, fragments in any case chosen to illustrate his vices rather than his virtues. Seneca does not normally give us more than isolated epigrams. The Asian declaimers will have been as diverse as the Latin ones whom Seneca describes at length,⁸⁵ and their emphasis on the different elements of the style will have varied. Their critics, it must always be remembered, used Asianus as a term of abuse rather than a carefully defined categoriser. For Dionysius of Halicarnassus, writing under Augustus, Asianism was in any event doomed. An alien Muse, ‘sprung from some Asian sewer the other day—some Mysian or Phrygian or Carian plague’, had driven out the Attic Muse. But now, under the beneficent influence of Roman rule, all had changed: ‘Apart from some few cities in Asia, where ignorance makes good learning slow to penetrate, the liking for the vulgar, frigid, and tasteless in literature has ceased.’⁸⁶ Dionysius seems to identify this change with the rise of the enthusiasm, which his treatises forwarded, for the Attic language and Attic literature of a bygone age. Meanwhile, in Rome, Cicero had rejoiced with equal fervour at the discomfiture of the Attici, qui iam conticuerunt paene ab ipso foro irrisi (Tusc. 2.3). He was not here celebrating the triumph of Asianism: but if the point and brevity of the Senecan declaimers is Asian, then the Roman Silver Age is Asian too; certainly it was not Ciceronian. Yet treacherous as these terms are in Greek literature, they are almost meaningless in Rome. The criticism of Cicero as Asianus had some foundation in reality; but the rise of a school of Roman orators who looked for models to fifth—and fourth—century Greece was a mere aberration that could not but be ephemeral. It was parasitic upon the Greek Atticist movement,⁸⁷ and was largely irrelevant to Roman conditions. In Greece, however, though bizarre and stultifying, Atticism made some sense. Its aim was to purify Greek literature linguistically as well as stylistically, by a return to the vocabulary and standards of the classical period. There was no wish to get rid of the system of rhetorical education that had established itself in the centuries following the death of Demosthenes. But Atticism could alter the style of some declaimers, and, by its continuing dominance, ensure that we arguably have no example of an ‘Asian’ declamation extant from any ⁸⁵ See Fairweather (1981), 243 303. ⁸⁶ Dion. Hal. Vet. orat. 1 and 2 (trans. D. A. Russell). ⁸⁷ Contra e.g. Bowersock (1979), 57 75; Gelzer (1979), 16 (‘der Attizismus ist wohl . . . in Rom und für Römer konzipiert worden’). For a recent statement of the other view see Bowie (1970), 36.

152

Papers on Quintilian and Ancient Declamation

period. What we do have is a large number of forensic declamations that strive to be Attic in vocabulary⁸⁸ and syntax, and avoid the excesses that had been blamed on the Asian rhetores. No doubt such excesses continued, but they have left little mark on the record.⁸⁹ Norden felt able to classify under ‘Asianismus’ extracts from four speeches of the Second Sophistic, all preserved by Philostratus.⁹⁰ But it is notable that the themes are hardly those of forensic declamation. Onomarchus of Andros was speaking an ethopoeia of a man in love with a statue:⁹¹ a progymnasma, then, and certainly not a controversia. It is not surprising that he lets himself go on the subject of love, where tumidity was in order, as Plato’s Symposium shows.⁹² The major fragment of Philagrus comes from an introductory διάλεξις. Dionysius and Apollonius were declaiming on historical themes, but not ones to evoke much in the way of focused argument. Little that these speakers say could not be paralleled in ‘Attic’ declamation; and, as usual, we cannot confidently argue from extracts to the tone of the full declamation. If we must have an Asianist, let him not be the point-loving Polemon, who, according to Procopius of Gaza, ‘purified ancient oratory from the Asian monstrousness’,⁹³ but rather the Latin Apuleius, as seen especially in his Florida; but they are hardly more than προλαλιαί, and give no certain idea what tone their author would have adopted in a serious forensic declamation.⁹⁴ There is no need for speculation about Lucian, Libanius, and Choricius: their declamations are extant. And for all their differences⁹⁵ they give us a ⁸⁸ Cf. Philoxenos’ criticism of the incorrect speech of contemporary rhetores (Dihle (1977), 170); Cic. Orat. 25 (insolens verbum). Wilamowitz Möllendorff (1900), 6 n. 1 draws attention to the ‘vulgar’ βαυνός at Sen. Con. 10.5.21. Cf. also Schmid (1887 96), .44 n. 18. More generally, see [Dion. Hal.] Ars rhet., with Russell (1979), 113 30. ⁸⁹ Rohde (1886) makes the most of what there is; but his list of Asianists (‘Polemo, Philos tratus, Aelian, Libanius, Himerius, Choricius, Heliodor, Longus, Achilles Tatius u.s.w.’, pp. 173 4) lacks nuance. ⁹⁰ Norden (1898), 413 16, citing VS 522 (Dionysius of Miletus), 580 (Philagrus), 601 2 (Apollonius of Athens), and 598 9 (Onomarchus). ⁹¹ Similarly: Τίνας ἂν εἶπε λόγους ζωγράφος γράψας κόρην καὶ ἐρασθεὶς αὐτῆς; (Sev. Alex. RG .546.11 ff. W. [= Ethop. 9 Amato]). See further Rohde (1914³), 621. ⁹² For Sopatros see below, p. 159 with n. 140. ⁹³ Cited by Norden (1898), 368. But this may be a linguistic observation rather than a stylistic one: τερατείας is reminiscent of Aeschines on Demosthenes (Ctes. 167: ῥήματα ἢ θαύματα). For Polemon’s point see below, n. 98. Another, perhaps stronger, candidate is Iamblichus (see n. 99). ⁹⁴ Or perhaps the Apologia does that. ⁹⁵ In particular, Libanius and Choricius pay attention to ἦθος rather than πάθος. See esp. Russell (1983), ch. 5. It should always be remembered that a major feature of declamation was the acting of the character personated; note e.g. Chor. 253.5 6 Förster Richtsteig: τὸν πατε ́ρα μιμη ́σομαι το ̀ν φιλάρ γυρον, 510.8 9: Τὸν . . . στρατιώτην τοιοῦτον ἄν τις μελετήσειε τρόπον. In Sopatros note p. 274.23 W. [= 46.8.1 2 We.]: ὁ ἀποκήρυκτος ἐπιείκειαν ὑποκρινεῖται (cf. p. 286.10 11 W. [= 50.1.2 We.]); cf. Philostr. VS 541 (ὑποκρι νασθαι) and Luc. Salt. 65 ́ (ὑπόκρισις). Thus Cicero’s impersonations (e.g. in the Pro Caelio) would come very naturally; hence too Quintilian’s view of declamation and comedy as close relations: 3.8.51; 10.1.71. For a

Sopatros’ Διαίρεσις ζητημάτων

153

consistent picture of an ‘Attic’ declamation,⁹⁶ conscious in language and sometimes in theme of the fifth—and fourth—century past. They eschew bombast and pathos, except in proem, epilogue, and prosopopoeia. They are not averse to the clever ending,⁹⁷ but they are in no way obsessed by point,⁹⁸ which had taken over the Latin Silver Age, or by word play,⁹⁹ except of an austere kind. Lucian is early enough to reflect the initial impulse to ignore the tempting rhythms of the Asianists.¹⁰⁰ But Choricius follows Himerius down a categorization of types see Hermog. RhG .29.12 ff. Rabe [= Stat. 1.5.1 ff. Patillon]. It may be relevant to all this that writing in the Second Sophistic is so fond of allusion to the drama: complex cases at Luc. Tyr. 20 (also end of 22); Polemo, Cyn. 48 9 (cf. Call. 45), sparked off by Cynegiros being brother of Aeschylus. Less elaborate: δρᾶμα (e.g. Lib. .562.3 Förster; [Lib.] .325.15 F.; see also Rohde (1914³), 376 7 n. 3); σκηνή ([Lib.] .349.16 F.); θέατρον ([Lib.] .328.17 and 332.20 F.; Sopat. Metap. 511.30; cf. θεατρίζειν at [Lib.] .341.20 F. and Innes Winterbottom (1988), 45). For τραγῳδία see below, p. 157 with n. 125. Generally, Schmid (1887 96), .405 6, 408, 416; .255 and 261; .301; .481 and 486. Observe how Choricius (509.10 F. R.) refers us to Menander’s Thrasonides in connection with the character ization of a braggart soldier. ⁹⁶ Note Schmid (1887 96), .216 (on Lucian): ‘μελέται . . . welche nicht die wilde asianische Manier, sondern eher den gedämpften Ton der herodeischen Richtung zeigen’. Contrast Rohde (1886), 177: ‘Nun war es so gut wie unmöglich, nach den Vorschriften attischen Styls diese Gaukelkünste zu betreiben.’ That is not true of forensic declamation. ⁹⁷ See Luc. Abd. 32: Οἴμοι, πάτερ, ταῦτ’ ἦν σου καὶ τῆς πάλαι μανίας τὰ προοίμια (cf. ‘Quint.’ Decl. min. 290.5: Crede, pater, iam aliquas similes rerum imagines video, redit dementiae prior causa); Lib. .316.1 5 Förster. ⁹⁸ The declamatory tone of voice, so familiar from Latin, is rarely sounded in Greek. But notice use of what may be called the ‘pregnant colon’ (type: Sen. Con. 10.4.20 (Artemon): Τὰ μὲν τῶν ἄλλων εὔρωστα· πλεῖ, γεωργεῖ (then Τὰ δ’ ἡμέτερα ἀνάπηρα· τρέφει ἄρα τὸν ὁλόκληρον); 9.2.5: Virum nobilissimum et tantis honoribus functum turpiter meretrix clementem fecisset: crudelem fecit; Cic. Agr. 2.46: Est in imperio terror: patientur. Est in adventu sumptus: ferent. Imperabitur aliquid muneris: non recusabunt). Thus Lib. .629.13 14 Förster: Κεκράτηκας πολεμίων· σπεῖσαι. Παῖδα ἐσωφρόνισας ἐκβαλών· ἀναλάμβανε; Aps. RhG ².325.15 17 Spengel Hammer [= Ars rhet. 10.45.18 19 Patillon]: Λέγει παράνομα· σιωπῶ; Καταλύει τοὺς νόμους· οὐ φθέγξομαι; (also Philostr. Ep. 23: Πρόσταξον ὡς ἔοικέ σοι· πείθομαι. Πλεῖν κέλευσον· ἐμβαίνω. Πληγὰς ὑπομεῖναι· καρτερῶ. Ῥῖψαι τὴν ψυχήν· οὐκ ὀκνῶ. Δραμεῖν διὰ πυρός· οὐκ ἀναίνομαι; for the topic see Winterbottom on ‘Quint.’ Decl. min. 257.4). There is a striking late Latin example in [Quint.] Decl. mai. 4.17 (79.10 14 Håkanson). More generally, Polemon makes much pointed use of the loss of Cynegiros’ hand (Cyn. 1: ἄνδρα τοῖς βαρβάροις καὶ κατὰ μέρος μεμαχημένον κτλ.) in the tradition of Sen. Con. 9.1.2: nec verebar ne Cynaegirus suas pluris aestimaret manus with Winterbottom (1974b), .217 n. 3, adding V. Max. 3.2.22 (cf. also the handless hero of Sen. Con. 1.4, where note 12 (Murredius): Reliqui in acie pugnantes manus the Greek epigrams on this theme are lost; add Sopatros Case 45, with Innes Winterbottom (1988), 200 [and now Citti Pasetti (2015), 133 41]). ⁹⁹ Naturally there is antithesis (see e.g. Schmid (1887 96), .418 on Lucian): it is hardly possible to write Greek without it. But as for the rest the declaimers seem, if anything, to be more restrained than the epideictic sophists (see e.g. Schmid (1887 96), .273 ff. on Aristides). Contrast the abundance of word play in Iamblichus (cf. Hinck (1873) 46 9 [= Iambl. Bab. fr. 35 Habrich: a declamation embedded in a novel], starting as the speaker means to go on with Ἔστω δὲ καὶ τοῦτο̆ τεκμήρῐ ο˘ ν τοῦ τὴν κατηγορίαν εἶναι τὴν γινομένην ἀ̆ ληθῆ˘, τὸ καὶ τὴν ὑπόθεσιν εἶναι τῷ μέλλοντι κατηγορεῖν ἀ̆ ηδῆ˘ (pp. 46.23 47.2 Hinck [= p. 27.14 16 Habrich]). ¹⁰⁰ The old metrical rhythms are to be seen in Iamblichus (see e.g. the extract cited in n. 99); Norden (1898), 397 n. 4 and 424 7 remarks them in Herodian and (?) Favorinus. That keeps us within the second third century. Presumably rhythm was one of the first targets of the new

154

Papers on Quintilian and Ancient Declamation

new rhythmical road, to a point where every colon is marked by a limited range of accentual effects.¹⁰¹ We may indeed with Choricius be thought to have returned to the sing-song that their critics objected to in the Asian orators.¹⁰² We can see in all this, if we will, a moderate use of Asian techniques, applied in a language that strove for the patina of classical antiquity. Three stylistic features demand attention. When pathos¹⁰³ is sought, there is use of what one may call the ‘ὤ (ὦ) style’. There are hints of this in Seneca: Hermagoras says Ὢ τύχης δεινῆς (Con. 2.6.13), Glycon Ὢ κακῶν συμφωνιῶν (Con. 10.4.22). In Lucian we have Ὢ πατρὸς μισοῦντος ἀδίκως· ὢ παιδὸς φιλοῦντος ἀδικώτερον (Abd. 18, in an excited passage of the argumentation). Polemon has twenty-seven ὢ’s in a row (Call. 51–3).¹⁰⁴ Libanius and Choricius seem to hold back in forensic declamation. But Libanius has no inhibitions in μονῳδία (e.g. Or. 17.14–15, 27, 34, 36). And that is the true home for this sort of exclamatory style. When Dionysius represents Demosthenes ‘denouncing himself to the boulē of Chaeronea’ and using the words Ὦ Χαιρώνεια πονηρὸν χωρίον . . . Ὦ αὐτομολήσασα πρὸς τοὺς βαρβάρους Βοιωτία, Philostratus introduces the quotation with the words: ἐς τήνδε τὴν μονῳδίαν . . . ἐτελεύτησεν (VS 522). Menander prescribes apostrophe in monody: Ὦ γένος λαμπρόν . . . (RhG .435.30 Spengel; cf. 419.25–8). In a quite natural way (cf. Quint. 2.10.12; 3.4.16), the procedures of epideictic invade appropriate passages of the controversia. Secondly, there is the ‘list style’. This has its roots in Gorgias, and a strong whiff of Asian flamboyance. At least it is the constant whim of Apuleius¹⁰⁵ (e.g. Apol. 14.7: quod luto fictum vel aere infusum vel lapide incusum vel cera inustum vel pigmento illitum; cf. also Fronto, .116 Haines [= 30.3–5 van den Hout²]: ficus Pompeiana vel holus Aricinum vel rosa Tarentina vel nemus amoenum vel densus lucus vel platanus umbrosa). It conveniently allowed an author to exhibit width of vocabulary and, often, Gorgianic sound-effects and varied knowledge. We find it in one of Philostratus’ declaimers, Pollux of Naucratis (in a διάλεξις: Ὁ Πρωτεὺς . . . ἐς ὕδωρ αἴρεται καὶ ἐς πῦρ ἅπτεται καὶ ἐς λέοντα θυμοῦται καὶ ἐς σῦν ὁρμᾷ καὶ ἐς δράκοντα χωρεῖ καὶ ἐς πάρδαλιν πηδᾷ

Atticism. It is the odder that the alleged Roman Atticist, Calvus, did not avoid it more scru pulously. See (with caution) Fairweather (1981), 99, esp. fr. 28.3 Malcovati⁴: āmbĭtu iudĭcarē̆ nt and fr. 32: turpĕ mĭsĕrerī̆ (from an emotional epilogue). ¹⁰¹ The figures in Hörandner (1981), 160 3. He does not consider Libanius’ forensic decla mations; but they do not appear to be any more consistently rhythmic than the Diegemata. ¹⁰² The evidence in Norden (1898), 294 5, 375 9. ¹⁰³ Often pity, but also strong feeling of the other kinds (joy, anger, etc.). ¹⁰⁴ And many others, collected by Schmid (1887 96), .65. For the practice of Aristides see Schmid (1887 96), .290 1. In Cicero see Verr. 2.3.28: O praeclarum et commemorandum iudicium! O severum edictum! O totum perfugium aratorum!; 5.163: O nomen dulce libertatis! O ius eximium nostrae civitatis! O lex Porcia legesque Semproniae! O graviter desiderata . . . tribunicia potestas! (cited under apostrophe by Quint. 9.2.38); Att. 4.19.1 (joy and relief). Note also Rhet. Her. 4.12 (O feros animos! . . . : grand style). ¹⁰⁵ For Latin see Winterbottom (1977a), 74 5, adding Tert. Adv. Marc. 1.1: Scytha tetrior eqs.

Sopatros’ Διαίρεσις ζητημάτων

155

καί, δένδρον ἢν γέ̆νηταῑ, κο̆μᾷ˘ (VS 593)¹⁰⁶ and in Polemo (Call. 47: οὐκ ἀκινάκῃ Μηδικῷ, οὐ κοπίδι Περσικῇ, οὐκ αἰχμῇ Βαβυλωνίᾳ, οὐ πελέκει Φοινικίῳ), but also in Lucian (e.g. Abd. 3: μῖσος ἄλογον καὶ νόμον ἀπηνῆ καὶ βλασφημίας προχείρους καὶ δικαστήριον σκυθρωπόν) and in Libanius (e.g. .299.8 Förster: δεῖπνα παρασκευάζειν καὶ κρατῆρας ἱστάναι καὶ δαιτυμόνας συγκαλεῖν καὶ ὑμέναιον ᾄδειν).¹⁰⁷ This list style deploys short cola that provide, successively, new pieces of information. In the third feature to which I draw attention, ‘theme and variation’,¹⁰⁸ two or more synonymous phrases follow each other. This manner is frequent in Lucian’s Abdicatus (e.g. 1: ἀποκηρυττόμενος ὑπ’ αὐτοῦ πάλιν καὶ τοῦ γένους ἀλλοτριούμενος δεύτερον and 7: οἴεται γὰρ ἓν εἶναι μανίας εἶδος καὶ μίαν τὴν νόσον καὶ τἀρρώστημα ταὐτόν; also 11: τοῦτο γάρ . . . , and Tyr. 20: παρατείνει . . . ). The technique is only sporadically employed elsewhere. Norden remarks as exceptional a passage from Libanius.¹⁰⁹ But there is a clear relevance to the sophistic enthusiasm for repeating speeches in new wording (Philostr. VS 572 and 619).¹¹⁰ And one particular type will prove important in Sopatros, repetition with asyndetic dicolon:[¹¹¹] παραχωρῶ τοῦ γέρως, ἐξίστα μαι τῆς δωρεᾶς (Luc. Tyr. 10).

VIII. THE STYLE OF S OPATROS ’ DECLAMATION Our special privilege in the Διαίρεσις ζητημάτων is to see behind the scenes. The teacher often remarks upon the style and content of his illustrative ¹⁰⁶ Also in Onomarchus’ ethopoeia (Philostr. VS 599: προσώπου στάσις, χρόας ἄνθος, βλέμμα τος κέντρον, μειδίαμα κεχαρισμένον, παρειῶν ἔρευθος, ἀκοῆς ἴχνος). With Pollux’s list cf. Him. 68.9.68 70 Colonna: Πρωτεὺς ὡς πυρσὸς ἁπ ́ τεται . . . ¹⁰⁷ See also Luc. Tyr. 4: πρὸς τὰς ὀργὰς ἠπιώτερος . . . , 10: ἐλευθερία σαφὴς . . . , 16: τύραννος χαλεπώτερος . . . ; Phal. 2.12: ὁ Πύθιος χρᾷ . . . ; Lib. .353.6 8 and .197.12 13 Förster. Outside declamation note e.g. Plut. Fort. Alex. 326e: πρὸς ἀμάχους δυνάμεις . . . , 336c: στόλους ἐπλήρου . . . , 341e: πλήθεσιν ὅπλων . . . ; Philostr. Ep. 8: ἡ ψυχὴ τοῦ σώματος . . . , 16: οἱ μὲν βάρβαροι ὅπλοις . . . , 19: Τὸ μὲν οἴκημα . . . ; Long. 1.22.2: καὶ φωνῇ πείθεσθαι . . . , 31.3: γάλα κατέσπεισαν . . . ¹⁰⁸ Perhaps more familiar in Latin poetry. See esp. Sen. Con. 9.5.17: Ovidius nescit quod bene cessit relinquere and the interesting context. ¹⁰⁹ .68.1 3 Förster: δακρύων μὲν ἐπὶ τοῖς κειμένοις, στένων δ’ ἐπὶ τοῖς σεσυλημένοις, ἀλγῶν δ’ ἐπὶ τοῖς ὑβρισμένοις (cited in Norden (1898), 403 n. 2). Norden (1898), 396 7 also cites some thing similar at Dio 56.36.2. For this style in Latin see Winterbottom (1977a), esp. 44, 62 3, 74 5 (adding Sen. Ep. 121.6: Hoc edita protinus faciunt; cum hac scientia prodeunt; instituta nascuntur). ¹¹⁰ Cf. the ταυτολογία that the well educated condemned in Philippos of Side (Norden (1898), 370) and that was perhaps part of his Ἀσιανὸς χαρακτήρ. Also Sen. Con. 4.pr.7: quotiens velles eandem rem et quamdiu velles diceret (sc. Haterius), aliis totiens figuris, aliis tractationibus (Seneca is commenting on his Greek style facultas). ¹¹¹ [On this type of repetition see Winterbottom (1983a), 74 6 (= A.7 above, pp. 116 18); in the Minor Declamations, see also Winterbottom (1984), . Much more material from the Major Declamations could be cited.]

156

Papers on Quintilian and Ancient Declamation

extracts. What he says makes clear the crucial importance of the character of the impersonated speaker, the circumstances of the case, and, not least, the part of the speech that is in question. Sopatros does not often allude to characterization.¹¹² In Case 1 he remarks upon Alcibiades’ high and stubborn spirit (p. 2.2 W. [= 1.1.2 We.]), traits that find expression in his proud proems (esp. p. 3.2–5 W. [= 1.1.14–16 We.]) and epilogue (precept pp. 13.29–14.1 W. [= 1.10.1–3 We.]; example p. 14.9–22 W. [= 1.10.9–20 We.]) and in his ‘Caesarism’.¹¹³ Case 57 is called ἠθικόν (p. 315.23 W. [= 57.1.3 We.]), for a miser, favourite material for Libanius and Choricius, is the speaker;¹¹⁴ but not enough illustration is given to show how Sopatros would have put over such a character. In Case 20 however, where the speaker’s piety has to be established (pp. 111.7–9; 112.1–3 W. [= 20.1.11–13; 20.2.8–10 We.]), we can see the precept carried out (e.g. pp. 112.4–30; 119.25–120.2 W. [= 20.3.1–35; 20.9.7–10 We.]). And Sopatros’ remark in this case that Δεῖ πρῶτον ἐπὶ παντὸς ζητήματος ἐπισκοπεῖν τὴν ποιότητα (p. 110.24–5 W. [= 20.1.1 We.]) reminds us that characterization is important even where Sopatros does not expressly allude to it. Detailed analysis of the procedures of ‘Demosthenes’ in Cases 4, 7, and 35 would be illuminating.¹¹⁵ The fictive orator argues vigorously, with much apostrophe and rhetorical question; and he, like Alcibiades, employs Caesarism. But, unlike Libanius and Choricius, Sopatros seems—on the surface at least—more interested in πάθος than in ἦθος. He often remarks on features of a case that make it παθητικόν (e.g. p. 28.8–11 W. [= 5.1.1–4 We.]; p. 78.26–8 W. [= 13.1.1–3 We.]),¹¹⁶ and he contrasts the arousal of emotion with the argumentative aspect of a speech (ἀγών: p. 33.7–8 W. [= 8.1.3 We.]; p. 329.29–30 W. [= 66.2.4–5 We.]; pp. 339.30–340.5 W. [= 69.1.2–6 We.]).¹¹⁷ Strong emotion could be aroused in the narrative (p. 371.14–15 W. [= 77.2.12 We.]; note the abuse at p. 29.19–20 W. [= 5.2.7–8 We.]) and the argumentation (p. 107.29 W. [= 19.3.17 We.] in the ἀνθορισμός). But the natural places for it were the proem and the epilogue.¹¹⁸ It is partly a matter of

¹¹² Note pp. 60.29 61.1 W. [= 9.5.4 5 We.] (making a defence one’s own ὡς ἐπ’ ἀληθείας). ¹¹³ See Innes Winterbottom (1988), 22 on p. 4.24 5 W. [= 1.3.10 11 We.]; Norden (1957⁴) on Verg. Aen. 6.510 11. ¹¹⁴ See Innes Winterbottom (1988), 238 on Case 52. ¹¹⁵ Observe p. 206.8 9 W. [= 35.2.15 16 We.]: πληκτικὸν καὶ δριμὺ καὶ Δημοσθενικῆς μετέχον βαρύτητος. Note, for example, in Case 4 the vigorous counterattack in the κατάστασις (p. 20.12 ff. [= 4.2.1 ff. We.]); in the βούλησις the contrasted questions of p. 22.5 13 and 18 21 W. [= 4.3.15 22 and 26 8 We.] may be meant to be Demosthenic. ¹¹⁶ Cf. p. 58.1 2 W. [= 9.4.26 We.]: πάθος ἐστὶν ἐν ὅλῳ τῷ πλάσματι. ¹¹⁷ For the contrast see Winterbottom on ‘Quint.’ Decl. min. 270.1. ¹¹⁸ Παθητικὰ γάρ ἐστι κατὰ φύσιν ἀμφότερα (p. 56.17 18 [= 9.2.32 We.] with Innes Winterbottom (1988), 53). For the proem note p. 28.13 W. [= 5.1.5 6 We.] and p. 56.29 30 W. [= 9.3.5 6 We.] (a declaimer mentions his own πάθος in a proem at p. 79.25 W. [= 13.2.12 13 We.]); for the epilogue p. 40.21 2 W. [= 8.13.2 3 We.].

Sopatros’ Διαίρεσις ζητημάτων

157

content. One embattled pauper threatens suicide in mid-epilogue (p. 41.14–15 W. [= 8.13.18–19 We.]).¹¹⁹ It is a common ploy to appeal to the sentient dead.¹²⁰ One may call for the witness of a mother,¹²¹ or allude to a threatening δαίμων (p. 93.3–4 W. [= 13.18.1–2 We.]). Tombs, to be built, inscribed, and wept over, figure prominently (e.g. p. 238.1–3 W. [= 40.9.12–14 We.]; p. 374.9–13 W. [= 77.12.2–7 We.]). There is an express θρῆνος at p. 32.10–16 W. [= 5.5.17–22 We.]. But along with subject matter go linguistic techniques. The παθητικά of pp. 107.29–108.4 W. [= 19.3.17–20 We.] are couched in a pair of sorrowful questions, introduced by anaphoric τίς (so too at p. 109.9–11 W. [= 19.5.7–8 We.]). Ethopoeia, remarked upon as ‘giving life’ to a narrative (p. 58.28–9 W. [= 9.4.23 We.]),¹²² is more naturally associated with the emotional parts of the speech (note the implication of p. 333.24–5 W. [= 67.4.9–10 We.]). Here, naturally, is the home of the ‘ὤ (ὦ) style’, along with other ἀνακλητικά (p. 56.29 W. [= 9.3.5 We.]).¹²³ Φεῦ τοῦ δράματος sets off an extract from an epilogue τοῦ πάθους ἔχων τὴν αὔξησιν (p. 31.23 W. [= 5.5.1 We.]), but such exclamations are no less at home in narrative and argument.¹²⁴ Much play is made of τραγῳδία and related words;¹²⁵ and when Sopatros comes to lament Alexander’s destruction of Thebes, he seems, consciously or unconsciously, to look back across the centuries to the Asianic Hegesias.¹²⁶

¹¹⁹ See Winterbottom on ‘Quint.’ Decl. min. 259.24. ¹²⁰ e.g. p. 50.17 20 W. [= 7.10.18 23 We.] (see Winterbottom on ‘Quint.’ Decl. min. 332.14). The topic is said to be ἐπιλογικόν at p. 57.15 16 W. [= 9.3.20 1 We.]; but it invades the proem at p. 33.21 4 W. [= 8.1.14 16 We.] and p. 67.22 5 W. [= 10.1.12 15 We.]. ¹²¹ P. 50.21 2 W. [= 7.10.26 7 We.] (led up to by p. 49.23 6 W. [= 7.9.13 16 We.]). ¹²² There are several in the long κατάστασις of Case 1 (pp. 4.17, 4.27, 6.5, 6.8, 7.3 W. [= 1.3.5, 13, 42, 44, 51 We.]). ¹²³ See Innes Winterbottom (1988), 53; also ibid., 261 ad p. 335.10 W. [= 67.7.4 We.] on ἀναφωνητικῶς. ¹²⁴ Thus φεῦ in κατάστασις at p. 6.8 W. [= 1.3.44 5 We.] (ἠθοποιΐα), and argument at p. 62.27 W. [= 9.6.36 We.]; οἴμοι in κατάστασις at e.g. p. 252.27 W. [= 44.1.18 We.] and p. 310.11 W. [= 54.1.21 We.], in argument at e.g. p. 212.11 W. [= 35.4.21 We.]. Ὤ (ὦ) is commonest (in narrative at e.g. p. 81.28 W. [= 13.3.34 We.]; in argument at e.g. p. 178.9 W. [= 27.3.22 3 We.]). And it is quite often used twice in succession (e.g. p. 32.12 13 W. [= 15.5.9 20 We.]; thrice at p. 41.26 7 W. [= 8.13.29 We.]), but Sopatros does not over use it. ¹²⁵ Τραγῳδία: p. 159.22 W. [= 23.8.19 We.]; p. 251.18 W. [= 43.9.14 We.]; p. 323.18 W. [= 62.2.12 We.]. Σκηνή: p. 28.12 W. [= 5.1.5 We.] (with Innes Winterbottom (1988), 36) and Metap. 513. The two together: p. 70.27 8 W. [= 10.5.17 We.]; p. 98.20 W. [= 16.1.5 We.]; Metap. 507.22 3 (παθητικῶς) and 510.19 20 (μελετηρῶς). Δεύτερος ἀγωνιστὴς τοῦ δράματος: p. 74.6 W. [= 10.11.4 We.]. See esp. p. 232.11 14 W. [= 39.7.4 6 We.], and generally above, pp. 152 3 n. 95 (also p. 20.19 21 W. [= 4.2.6 8 We.] with Innes Winterbottom (1988), 31). ¹²⁶ See p. 210.20 and 22 W. [= 35.3.77 and 79 We.] with Innes Winterbottom (1988), 161 ad locc. (also p. 173.20 W. [= 26.4.13 We.] with Innes Winterbottom (1988), 134); and above, pp. 149 50.

158

Papers on Quintilian and Ancient Declamation

All this may give the impression of a hysterical declaimer. On the contrary: Sopatros relies mainly on reason. Measured antithetical proems¹²⁷ and elaborate narrations¹²⁸ presage carefully organized argumentation sections, which dominate the whole speech and rarely admit emotion or much stylistic excitement.¹²⁹ Indeed, the style is generally restrained. Sopatros is hardly interested in point¹³⁰ or word play.¹³¹ The ‘list style’ only appears once, and in a modified form.¹³² The most noticeable trait is a fondness for full expression, with repetition in variation. Sometimes this is not given any particular structure;¹³³ but often it is decorated with anaphora in tricolon (e.g. pp. 34.30–35.2 W. [= 8.3.9–11 We.]: Ἂν οἰκείας ὀδύρηται συμφοράς, ἂν θρηνήσῃ τὴν τύχην, ἂν οἰκείοις σχολάσῃ τοῖς πάθεσιν)¹³⁴ or, characteristically, dicolon (e.g. p. 228.13–14 W. [= 38.2.10–11 We.]: Ἐπὶ τούτοις ἡ κρίσις, ἐπὶ

¹²⁷ The prevalence of an opening with μὲν οὖν (see Innes Winterbottom (1988), 38 on p. 33.10 W. [= 8.1.5 We.]; note Metap. 511.14: προοιμιακῶς· ‘Ἤδη μὲν οὖν . . . ’) is symptomatic of Sopatros’ desire to give antithetical shape to his proems. Observe how in Case 13 the contrast of πάθος and the need to speak dominates the second proem (pp. 79.24 80.5 W. [= 13.2.11 22 We.]) and extends into the προκατάστασις (p. 80.22 7 W. [= 13.3.1 5 We.]). On the style of proems see p. 56.11 18 W. [= 9.2.27 32 We.] with Innes Winterbottom (1988), 52 3. ¹²⁸ They need not lack πάθος (see above on the ‘ὤ (ὦ) style’, and p. 311.14 15 W. [= 54.2.26 7 We.]), but should not degenerate into the tone of an epilogue (p. 291.13 14 W. [= 50.5.52 We.]). ¹²⁹ Note however Sopatros’ precepts at p. 56.19 21 W. [= 9.2.33 5 We.]. ¹³⁰ There is a certain cleverness in the treatment of the epilogue in pp. 123.18 124.16 W. [= 20.16.1 24 We.] (ending with an explicit παράδοξον), and a pointed end in p. 182.15 16 W. [= 28.6.8 9 We.]. Ἀποκηρύκτου πατὴρ ἀκήκοεν· ὀνομαζέσθω καὶ τοῦ προσαγγείλαντος (p. 339.7 8 W. [= 68.6.14 15 We.]) strikes a declamatory note that is rarely sounded (not even in the illustrations for τὰ μελετηρά and τὰ θεατρικά: see Innes Winterbottom (1988), 45 and 53 on p. 45.8 W. [= 7.2.51 We.] and p. 56.21 W. [= 9.2.35 We.] for a list of the relevant passages). P. 45.7 8 W. [= 7.2.51 We.] shows that applause was sought. ¹³¹ See however p. 34.11 W. [= 8.2.10 We.] (pun on ἀρχή); p. 36.13 W. [= 8.5.10 We.]: δυστυχῶν παίδων δυστυχῆ πατέρα (cf. p. 74.2 3 W. [= 10.11.1 2 We.]: δυστυχοῦς . . . δυστυχεστέρας); p. 133.14 15 W. [= 22.6.41 2 We.]: μαθόντες . . . παθόντες; p. 311.5 6 W. [= 54.2.18 We.]: πόνῳ . . . χρόνῳ; p. 335.17 18 W. [= 67.7.10 11 We.]: ἐπὶ τοῖς μικροῖς με γάλας τὰς ὀργὰς . . . ἐπιδεικνύμενος; p. 342.6 W. [= 69.5.6 We.]: ὑπὲρ ἑνὸς τροπαίου δύο δωρεάς (with what follows); p. 349.21 4 W. [= 72.1.6 10 We.] (pun); p. 358.22 W. [= 73.6.4 5 We.] (Gorgianic echo?); p. 364.30 W. [= 75.3.18 19 We.] (paronomasia?); p. 373.29 30 W. [= 77.10.5 We.] (pun). All this is very mild, like the familiar πολλοὶ πολλάκις (see Innes Winterbottom (1988), 96 on p. 116.6 W. [= 20.5.18 We.]). Note how Sopatros avoids pointed repetition of the same word in p. 293.26 7 W. [= 50.7.20 1 We.]: Μεταλλάξας . . . τὴν τύχην τὴν . . . γνώμην οὐκ ἤμειψεν; p. 296.22 3 W. [= 50.9.15 16 We.]: Ὑπὲρ τοῦ πολέμου τοῦτον ἀνείληφας· ὑπὲρ χρείας τῆς αὐτῆς παραχώρησον. The only type of figure of speech favoured is anaphora, and that to excess. The in jokes perhaps present at p. 27.1 6 W. [= 4.7.2 6 We.] are reminiscent of Choricius’ rhetor in Decl. 12 Förster Richtsteig. See also Innes Winterbottom (1988), 44 on p. 43.30 W. [= 7.2.18 19 We.]. ¹³² P. 32.2 4 W. [= 5.5.10 12 We.]: a Lucian would have made this a far more regular list. Note the same reluctance in p. 240.9 10 W. [= 41.3.4 5 We.] (but the text may be corrupt). ¹³³ Note e.g. p. 48.5 15 W. [= 7.7.5 14 We.]; pp. 66.24 67.2 W. [= 9.11.23 31 We.]; p. 270.23 8 W. [= 46.3.9 13 We.]; p. 301.3 10 W. [= 50.11.18 22 We.]. ¹³⁴ Also e.g. p. 276.11 13 W. [= 46.10.4 6 We.]; p. 314.11 13 W. [= 56.1.13 15 We.]; p. 315.4 5 W. [= 56.2.10 12 We.].

Sopatros’ Διαίρεσις ζητημάτων

159

τούτοις τὸ δικαστήριον).¹³⁵ Even without anaphora, the manner is unmistakable, whether in dicolon (e.g. p. 42.7–8 W. [= 8.13.37 We.]: τὴν ἀπανθρωπίαν ἐννοῶν, τὴν ὠμότητα λογιζόμενος; p. 74.3–4 W. [= 10.11.2–3 We.]: ἐριννύες ἄρα τῆς ὁδοῦ προεπόμπευον, χορὸς ἀλαστόρων τῆς πορείας ἡγεῖτο; p. 198.8–9 W. [= 31.10.10–11 We.]: Πάλιν συλλήψονταί μοι αἱ θεαί, συμμαχήσουσιν οἱ κρείττονες; p. 359.26–6 W. [= 73.7.12 We.]: μὴ μεμετρημένως, μὴ πεφεισμένως) or in tricolon (e.g. p. 266.1–4 W. [= 45.6.6–8 We.]: Ἐγκαλῶ σοι τανῦν ὑπηρετησαμένῳ τῷ δράματι, διακονίαν ἣν ἐποιήσω μέμφομαι, κατὰ σοῦ γραφὴν τὴν τοιαύτην ὑπηρεσίαν ποιῶ: following upon a dicolon in anaphora). This trick, on which Sopatros never comments,¹³⁶ is commonly used to give liveliness to the narrative (e.g. p. 201.15–24 W. [= 33.1.9–16 We.]), but it is at home in any part of the speech, except perhaps the proem. That, together with his consistent use of rhythm,¹³⁷ is Sopatros’ main concession to the sophistic manner. His version of ‘Attic’ declamation is markedly restrained. He keeps emotion in its place and he is extremely sparing of verbal figures¹³⁸ (apart from anaphora). He may have a penchant for patriotic πομπεία,¹³⁹ and there is the odd rhapsody on love.¹⁴⁰ But on the whole he remembers the lesson of the sophists preceding him in the long line that started with Gorgias: style is there to support argument, and argument is what wins cases.

¹³⁵ Also e.g. p. 39.20 2 W. [= 8.11.4 6 We.]; p. 224.24 W. [= 37.2.11 12 We.]; p. 266.7 8 W. [= 65.6.10 11 We.]; Metap. 509.9. ¹³⁶ In Metap. 506.6 18 Sopatros cites a dicolon (with connective and without anaphora) as an example of ῥυθμὸς σεμνός. It is noticeable that he often comments on the need for variety in other fields: see e.g. p. 8.14 15 W. [= 1.5.5 6 We.]; pp. 9.22 10.1 W. [= 1.5.37 40 We.]; p. 12.12 13 W. [= 1.7.28 9 We.]; p. 61.11 14 W. [= 9.5.15 18 We.]; p. 294.9 15 W. [= 50.8.3 7 We.]; p. 365.3 6 W. [= 75.3.21 4 We.]. ¹³⁷ In an elaborate passage like pp. 31.24 32.25 W. [= 5.5.2 30 We.] (epilogue), the rhythm is consistent at colon as well as sentence ends. My impression is that Sopatros observes rhythm in a high degree in his declamation, and to some extent in his precepts. But I have no statistics. ¹³⁸ There are occasional rhyming isocola (p. 31.24 5 W. [= 5.5.2 We.]: νεανίας ἐν πολέμοις ἀήττητος, ἐν ἀριστείαις ἀοίδιμος; p. 32.12 14 W. [= 5.5.19 20 We.]: ὦ καὶ τῆς ἐνεγκαμένης ἐμοὶ προτιμώτερε, ὦ καὶ ἀγχιστέων ἐμοὶ προσηνέστερε). A pointed antithesis like p. 93.5 6 W. [= 13.18.3 4 We.]: ζῶντα μὲν γὰρ ἀπεκήρυξε, τελευτήσαντα δὲ ἐτυμβωρύχησε is rare. At p. 147.23 4 W. [= 23.3.3 4 We.] Sopatros remarks on the antitheses of the following de clamatory passage (with pregnant colons: ἔγραφε νόμους· οὗτος εἱστήκει κατήγορος κτλ.) as avoiding τὸ ὕπτιον. ¹³⁹ Rich in exhortation and promises, with appeal to Athenian or panhellenic sentiment: e.g. p. 14.11 22 W. [= 1.10.11 20 We.]; pp. 27.8 28.2 W. [= 4.7.9 26 We.]; p. 50.4 8 W. [= 7.10.6 10 We.]; p. 102.2 8 W. [= 16.4.3 9 We.]; p. 109.22 4 W. [= 19.6.1 3 We.]; pp. 143.14 145.8 W. [= 22.15.11 54 We.] (for πομπεία in this case see pp. 126.28 127.3 W. [= 22.1.1 3 We.]; p. 131.16 19 W. [= 22.5.1 4 We.]); pp. 174.19 175.10 W. [= 26.6.3 18 We.]; pp. 197.30 198.16 W. [= 31.10.1 17 We.]. ¹⁴⁰ P. 70.13 17 W. [= 10.5.6 9 We.]; p. 95.10 12 W. [= 14.4.15 We.]; p. 373.4 6 W. [= 77.8.9 11 We.] (cf. also p. 365.5 W. [= 75.3.23 We.]; p. 366.18 W. [= 76.2.11 We.]; p. 373.30 2 W. [= 77.10.5 We.]). See above, p. 152.

160

Papers on Quintilian and Ancient Declamation

[This piece is in part an expansion and documentation of A.7; some of the additions pointed to at the end of that paper apply here too. On prose rhythm in Greek see Winterbottom (2011), 262–5; also Fraenkel (1968), 186 (on Hegesias); Stramaglia (2015), 140–6 (on transition to accentual clausulae, to be dated in Greek no earlier than in the first decades of the fourth century ). On certain salient stylistic features of both Greek and Roman declamation see now Citti–Pasetti (2015).]

11 Cicero and the Middle Style* In the introduction to his edition of the Orator,¹ Wilhelm Kroll remarked on the oddity of the fact that Cicero twice excludes from the purview of his treatise epideictic oratory and everything like it. Kroll rightly explained this emphatic exclusion as being the result of personal considerations. Cicero wished to differentiate his oratorical from his philosophical writings, and in particular to mark his oratorical style off from the style in vogue among the Asian speakers to whom his critics so opprobriously compared him. But it may also surprise that, although the central doctrine of the Orator, the need for the perfect speaker to be a master of all three styles, inevitably implies that he will have the Middle Style in his repertoire, yet this style is virtually identified with the epideictic manner which Cicero so firmly rejects. I here explore this paradox, and suggest that the explanation lies here too in considerations personal to Cicero. * * * Cicero’s first rejection of epideictic starts as follows: Sed quoniam plura sunt orationum genera eaque diversa neque in unam formam cadunt omnia, laudationum scriptionem et historiarum et talium suasionum qualem Isocrates fecit Panegyricum multique alii qui sunt nominati sophistae reliquarumque rerum formam quae absunt a forensi contentione eiusque totius generis quod Graece ἐπιδεικτικὸν nominatur, quod quasi ad inspiciendum delec tationis causa comparatum est, non complectar² hoc tempore: non quo neglegenda sit; est enim illa quasi nutrix eius oratoris quem informare volumus . . . ³ [From J. Diggle, J. B. Hall, and H. D. Jocelyn (eds.), Studies in Latin Literature and its Tradition in Honour of C. O. Brink (Cambridge, 1989), 125 31] * This paper represents an abbreviated version of my contribution to the colloquium on ‘Republican and Augustan Latin literature’, Cambridge, Gonville and Caius College, 19 and 20 March 1987. I am most grateful to Dr D. C. Innes for commenting on an earlier draft, and to Professor D. A. Russell for a pertinent question. ¹ Kroll (1913), 10. ² Cf. Cic. De orat. 1.22, where the same limitation of subject matter is made. See also 2.341 (where there is the same emphasis on the written nature of epideictic: Graeci . . . laudationes scriptitaverunt . . . ; nostrae laudationes . . . scribuntur ad funebrem contionem; see n. 4). ³ Cic. Orat. 37.

162

Papers on Quintilian and Ancient Declamation

Cicero is here describing a type of literature that is essentially written, and that is thereby sharply marked off from the combative oratory of the forum. He is, in fact, following the Aristotelian⁴ distinction of λέξις between ἀγωνιστική and γραφική, the former appropriate to delivered speeches, whether forensic or deliberative, the latter to everything else, including epideictic ‘oratory’. He speaks particularly of the suasiones⁵ of Isocrates and ‘many’, apparently ancient, sophists; and when he comes to sketch the history of epideictic in 39–40 he does not take it later than Isocrates. But both in 37 (est . . . quasi nutrix) and later in 42 (educata huius nutrimentis eloquentia est⁶—ipsa se postea colorat et roborat), the present tenses show that Cicero is thinking of an influence of epideictic on ‘real’ oratory that continues to the present day.⁷ He is thinking, that is, of the fact that an orator like himself was trained in declamation schools, at Rome or in the east, before passing to the dust of the forum. His final words in 42: Verum haec ludorum atque pompae; nos autem iam in aciem dimicationemque veniamus primarily contrast ‘games’ and reality. But the games were played in schools; the ambiguity is pointed. It should also be noticed that in 37–8 the educative value of epideictic is seen to lie in the copia verborum, the rhythms and the Gorgianic σχήματα that are proper to it; those σχήματα are employed in the forum et rarius multo . . . et certe occultius (38). Indeed at 42 Cicero goes so far as to say that this genus orationis is spretum et pulsum foro. It has a place in the orator’s education; but having grown up thanks to it, he then grows out of it. Cicero’s second rejection of epideictic (62–8) takes the form of a differentiation of the eloquence of the ‘real’ orator from that of philosophers, sophists, historians, and poets: λέξις ἀγωνιστική, that is, from γραφική. The philosophical style (like the historical and poetic) is easily enough distinguished. Philosophers lack nervos . . . oratorios ac forenses. Their oratio is umbratilis. They soothe men’s minds rather than stir them. They aim at giving at least some pleasure. A little less easily distinguished, but still quite separate, is the style of the sophists. They too soothe rather than move. They please rather than persuade. They seek out concinnas magis sententias . . . quam probabiles, employ

⁴ Aristot. Rh. 1413b.3. For the background see Quadlbauer (1958), 60 ff. The written nature of epideictic is stressed at Cic. Orat. 208 (scripsit) and by Dionysius of Halicarnassus (Isoc. 1: Isocrates gave up public speaking and ἐπὶ τὸ γράφειν ἃ διανοηθείη κατέφυγεν; 13 (view of Hierony mos); Dem. 4: ἀγῶνα μὲν . . . οὐδένα πώποτε τἀνδρὸς ἀγωνισαμένου, γραφὰς δὲ συνταξαμένου). All this supports Havet’s conjecture scriptionem in Orat. 37. ⁵ Cf. Dion. Hal. Isoc. 10: πολλοὺς . . . λόγους πρὸς πόλεις τε καὶ δυνάστας καὶ ἰδιώτας γραφέντας. ⁶ So the MSS, followed by Westman (1980); the reading certainly gets the emphasis where it should be, but the past tense is surprising and the consequent punctuation awkward. Colorat et roborat in any case tie the remarks to the (generalizing?) present. ⁷ Quint. 2.10 is relevant. Declamation aliquid in se habet epidicticon (12), and therefore contains florid elements on which the pupil can fatten himself for the moment in the knowledge that he will have to be ‘slimmer’ in court (6).

Cicero and the Middle Style

163

noticeable metaphors, and favour Gorgianic figures. This picture of the sophistic manner is consistent with and complementary to that given in 37 ff. When Cicero comes to characterize the Middle Style in 91–6, he describes it in terms constantly reminiscent of what he has said of the sophists, with a glance at the philosophers also. He quickly gives Demetrius of Phaleron⁸ as the supreme exponent of the style, commenting on his brilliant metaphors. In the Brutus (37–8) he had laid stress on Demetrius’ background in the philosophical schools. If he came out into the sunshine and the dust, it was from the shades (umbraculis) of Theophrastus. He delighted rather than inflamed his fellow countrymen. He preferred to be suavis rather than gravis, and to leave a memory of his concinnitas behind in men’s minds. To return to the Orator: of the Middle Style in general, Cicero says that it lacks nervi. It is the home for all the ornamenta dicendi, and especially for all figures of speech. It is the proper medium for latae eruditaeque disputationes and loci communes spoken sine contentione (that is, not in the λέξις ἀγωνιστική),⁹ both of which pertain to the realm of philosophy. It is therefore no surprise when Cicero sums up by remarking that e philosophorum scholis tales (apparently speakers in this manner) fere evadunt: and a respectable manner it is, flowery, polished, attractive. Nor is it surprising that Cicero should go on to allude specifically to the sophists. But the words in which he does so take us to the heart of our problem: Hoc totum e sophistarum fontibus defluxit in forum, sed spretum a subtilibus, repulsum a gravibus, in ea de qua loquor mediocritate consedit. The Middle Style, which Cicero has been so careful to identify with an epideictic which he has twice emphatically marked off from the oratory of real life, now suddenly finds its way into the forum. It is seen as a river, taking its rise from the springs of the sophists, that has flowed downhill to the market-place below (below, are we meant to think, the hill into which is set the theatre where the sophists disport themselves?) and there settled, as it were, at a middling depth. The Middle Style in these sections is given a substance it had hitherto seemed to lack. Whatever part it played in the stylistic theories of Theophrastus,¹⁰ it had hitherto been regarded by Cicero himself as no more than a colourless halfway house on the road up from the Plain to the Grand. Even in the Orator itself he has attributed it to an orator interiectus inter hos medius (cf. Opt. gen. 2) et quasi temperatus, and discouragingly dubbed him in neutro excellens, utriusque particeps (cf. Gell. 6.14.3) vel utriusque, si verum quaerimus, potius expers (21). In the same passage he has endowed the Middle Style orator with ornamentis modicis verborum sententiarumque, ⁸ It was a question whether he was to be regarded as a practical speaker as well as a sophist (Cic. Off. 1.3). See the subtle discussion of Heldmann (1982), 98 122. ⁹ For loci communes in forensic oratory see Cic. Orat. 126 7 (nervosius). ¹⁰ See most recently Innes (1985), 260 2.

164

Papers on Quintilian and Ancient Declamation

words only by main force to be reconciled with the sophistic Middle Style, whose ornaments (95: verborum . . . lumina omnia, multa etiam sententiarum; 96: omnes verborum, omnes sententiarum . . . lepores) are employed with so much more freedom and openness than in the forum (38). What is said at Orat. 21 is, on the other hand, entirely consistent with Cicero’s earlier allusions to the Middle Style in the De oratore, where its ‘middleness’, not its floridity, is in question (3.177; 199: particeps utriusque generis, cf. Orat. 21; 212). It is consistent, too, with Ad Herennium 4.11, where the author thinks of a scale from gravia to pervulgatissima verba on which the Middle Style comes halfway.¹¹ Why then does Cicero in Orat. 91–6 both give new epideictic content to the Middle Style¹² and assert that in this guise it has ‘flowed down into the forum’, though we have earlier been told that the epideictic genus was spretum et pulsum foro (42)?¹³ I suggest that the single answer to these two questions is that only by these means could Cicero bring himself—for it is Cicero, naturally, who lurks behind the perfect orator he is describing—into his own picture of oratory. He was as well aware as his contemporaries that his eloquence had been ‘nurtured’ by the declamation schools of Rome and Asia Minor; that his forensic style was unmistakably marked by the declaimers’ enthusiasm for figures of speech and for rhythm; and that he pleased as well as persuaded his audience—indeed persuaded them partly by pleasing them.¹⁴ It was he, primarily, who diverted the waters of the sophists into the forum, for it was he most of all who had brought into Roman oratory the beguiling devices of epideictic. He asserts this himself in the Orator when describing the success of his early speech Pro Roscio Amerino: Ieiunas igitur huius multiplicis et ¹¹ Cf. also Rhet. Her. 4.13: si haec . . . aliquantulum demiserimus neque tamen ad infimum descenderimus. The emphasis is on vocabulary. Figures, on the other hand, need to be used, not too seldom and not too often, in each style (4.16). ¹² Quintilian clearly distinguished the two ways of looking at the Middle Style (12.10.58: tertium alii medium ex duobus, alii floridum . . . addiderunt); Proclus (ap. Phot. Bibl. 318b) shrewdly criticized their identification. Fronto, on the other hand, is strikingly unaware of the second way: Omnia ἐν τῷ ἐπιδεικτικῷ ἁδρῶς dicenda, ubique phaleris utendum; pauca τῷ μέσῳ χαρακτῆρι (misunderstood by Haines in his Loeb I.107 [= 49.16 17 van den Hout²]). Dionysius went a way of his own (see esp. Bonner (1938)): his Middle Style is neither ‘halfway’ nor ‘florid’. ¹³ Another way of measuring Cicero’s change of stance is from the standpoint of the officia oratoris. In the De oratore they are, as ever, three in number, but alongside movere and docere we find conciliare (2.115, 121, 128, 310). Delectare appears at Orat. 69 (where igitur is a quite illegitimate link to the rejection of epideictic that precedes), and here Cicero first makes the connection between the officia and the styles. That commits him to the new view of the Middle Style that duly appears in 91 6. It is interesting that Cicero has already made pleasure giving the third officium in the Brutus (185; 276); yet when he describes Marcus Calidius in terms that make him sound like a Roman Demetrius of Phaleron (Douglas (1966) on Brut. 274), he is not ready to make him a Middle Style orator: the Brutus knows only two styles, and the colourless Middle Style that was all Cicero envisaged at this stage held no attractions for the categorizer of Roman orators. ¹⁴ Cf. Winterbottom (1982), esp. 258 66 [= A.5 above, pp. 80 6].

Cicero and the Middle Style

165

aequabiliter in omnia genera fusae orationis aures civitatis accepimus, easque nos primi . . . ad . . . audiendi incredibilia studia convertimus (106).¹⁵ The Roman public was hungry to have the third style added to the other two, and Cicero gave them what they wanted. His theory therefore needed full and coherent account to be taken of an epideictic element in practical oratory. It is extremely likely that it was Cicero’s innovation to identify the Middle Style with the epideictic manner¹⁶ and find it a place in the forum; one can see the development taking place before one’s eyes between Orat. 21 and 96. And if it was not Cicero’s innovation, it was that of a Greek theoretician who saw in the Hellenistic world some such merging of the epideictic and the forensic traditions as Cicero pioneered in Rome.¹⁷ The apparent inconsistency between the rejection of epideictic and calm acceptance of its influx into the forum is at least partly palliated by the way in which Cicero draws limits for its reception there: a limitation that is at the same time a statement of his own practice as he saw it. The sophists give pleasure apertius . . . quam nos et crebrius (65); the implication is that the forensic orator will be restrained in entertaining his hearers. Similarly, as we have seen, the Gorgianic figures are employed in veritate causarum et rarius . . . et certe occultius (38), though Cicero boldly cites examples from his own speech for Milo for a concinnitas he traces back to Gorgias (165), and even calls himself frequens in the use of antithesis, citing a passage from the Verrines just after remarking how common the figure is in Isocrates and Gorgias (167). So too with rhythm. Cicero, who traces it back to Thrasymachus, Gorgias and especially Isocrates (175–6), is careful to restrict its forensic use, in a passage (210) to which sufficient attention is not always paid.¹⁸ It should be used in forensic speeches only in encomia, or in ἐκφράσεις, or in amplificatio, especially peroratory. The first two of these heads are what one

¹⁵ The succeeding passage (Orat. 107) from the Pro Roscio Amerino is high spirited and even declamatory, with a high degree of concinnitas. It is natural to think of it as in the Middle Style, yet it was clearly intended to move the listener: clear evidence of the difficulties caused by the alignment of the styles and the officia. I owe to an anonymous pupil the attractive suggestion that, when Cicero goes on to cite uxor generi, noverca filii, filiae paelex from Clu. 199, he is by shorthand referring to the whole context, where another sentence marked by elegant corre spondences immediately precedes (cuius ea stultitia est ut eam nemo hominem, ea vis ut nemo feminam, ea crudelitas ut nemo matrem appellare possit). ¹⁶ Kroll (1907), 89 seems to date the identification a good deal earlier than Cicero. Why then was it not employed in the De oratore? ¹⁷ Schmid (1894), 143 approaches the point: ‘Ist etwa zu Ciceros Zeit ein selbständiger blühender Stil erst neu in die Litteratur eingeführt worden und äussert nun bei Cicero seine ersten Ansprüche auf Anerkennung in der rhetorischen Theorie?’ Yes: a style that formed part of Cicero’s own armoury. ¹⁸ See also Orat. 209 (neque totum adsumendum est ad causas forenses neque omnino repudiandum) and 221 (non modo non frequenter, verum etiam raro in veris causis aut forensibus circumscripte numeroseque dicendum est). Cicero is clearly answering Atticist criticism that he over used rhythm.

166

Papers on Quintilian and Ancient Declamation

would expect: places within forensic oratory where the topics of epideictic bring their style with them. The third is a give-away. Perorations are meant to move, and the Grand Style should be hostile to rhythm. Cicero defends his own practice.¹⁹ Cicero, then, who was in effect attacked for being a Roman Isocrates,²⁰ ripostes by showing that he is a Roman Lysias and a Roman Demosthenes as well. What he did not and could not show was that the almost perfect Demosthenes, too, had an epideictic element. When he asserts Demosthenes’ control over all three styles, he gives him a Middle Style that reverts to colourlessness; and he gives no examples to prove his point.²¹ It was not in fact possible to argue that Cicero was altogether like Demosthenes. That left him only one card to play, the composition of the Philippics: a card, surely, that trumped the carpings of the Atticists.

¹⁹ Rhythm is ejected from the Plain Style in Orat. 77, but not placed in the Grand Style in 97 9 (a passage generally short on technical detail). But the mention of epilogues in Orat. 210 may be taken over without due thought from a Greek source: see ‘Longin.’ Subl. 12.5 (pointed out to me by Dr Innes), where Ciceronian χύσις is said to be appropriate to τοπηγορίαις . . . καὶ ἐπιλό γοις . . . καὶ παρεκβάσεσι . . . καὶ ἐπιδεικτικοῖς. ²⁰ It is interesting to observe how closely Dionysius’ criticisms of Isocrates are paralleled by his contemporaries’ criticisms of Cicero. With Quint. 12.10.12 and Tac. Dial. 18.4 (for figures of speech see Winterbottom (1982), 261 [= A.5 above, p. 82]) compare Dion. Hal. Isoc. 3, 12, 13; Dem. 4, 18, 20; Comp. verb. 19. Note Brutus’ distaste for Isocrates (Orat. 40) contrasted with his enthusiasm for Demosthenes (Orat. 105). ²¹ Nothing, that is, to correspond to his citing of his own Pro lege Manilia (significantly a virtually epideictic speech) as an example of the Middle Style in Orat. 102. For Demosthenes he can only say that he is uniquely temperatus as well as uniquely gravis and uniquely callidus (23), and that he moves easily from Grand to Middle (eo potissimum) that being next on the scale down from Grand (Orat. 111): he has given examples for Plain and Grand speeches and also ‘varied’ ones, i.e. partly Plain, partly Grand, but he gives none for the Middle Style. That would have been difficult or vacuous in the ‘halfway’ sense of Middle; and quite impossible in the ‘florid’ sense: Demosthenes was ἥκιστα ἐπιδεικτικός (‘Longin.’ Subl. 34.3), and notable for figures of thought (Orat. 136), not figures of speech. Note how Dionysius praises his μετριωτάτη κατασκευὴ λέξεως (Dem. 14), and shows how he avoided Isocrates’ ψυχρὰ καὶ μειρακιώδη σχήματα (Dem. 21). It is true that he says Demosthenes merged ἀληθινὴ διάλεκτος (Dem. 8), but that is in a context where Demosthenes is said to have been able to combine a whole series of opposites, and Dionysius is not to be held to his words too strictly.

12 On Impulse* Iuba, Scipio, Labienus in legionarios impetum fecerunt (B. Afr. 52.1). But one might suffer impetus from something less rational than a general: from a wild beast, or a disease. One might be carried along by one’s own impetus when making an attack on another: fert impetus ipsum.¹ The impetus assailing one might come from within oneself, a mental impulse, a burst of passion. When the word came to be used in literary criticism, it brought with it more than a hint of lack of control and even violence. All the same, the great orator could not do without impetus: Quid denique Demosthenes? Non cunctos illos tenues et circumspectos vi sublimitate impetu cultu compositione superavit?² A speaker needed drive, the onward movement that would carry the audience with him; and, with it, force.³ Hence the appropriateness of the common comparison with a river; for a river too has its impetus.⁴ On this model, the slender and circumspect Lysias hardly measured up at all: puro tamen fonti quam magno flumini propior.⁵ The middle style at least was a river, lenior . . . amnis et lucidus quidem, sed virentibus [From D. Innes, H. Hine, and C. Pelling (eds.), Ethics and Rhetoric. Classical Essays for Donald Russell on his Seventy Fifth Birthday (Oxford, 1995), 313 22] * Some of the same material was exploited in a short paper given at the conference on ‘Understanding the passions in Roman literature and thought’ at the University of Exeter in July 1992, and in a lecture at the University of Helsinki in September 1993. This final version is offered to Donald Russell in admiration and affection; like anything else one might have written for him, he would have done it better himself. I hope that he will at least find it appropriate, in the sense that its origin lies in my (unpublished) thesis on Quintilian Book Two, of which he was, with R. G. Austin, an examiner in 1962. ¹ Verg. Aen. 12.369. ² Quint. 12.10.23. ³ For the combination see also Quint. 6.2.10; 12.2.11 (impetu quoque ac viribus . . . est opus, ut vis amnium . . . ); 12.10.64 (if my text is right); Cic. Orat. 229. ⁴ TLL s.v. impetus 604.56 7. ⁵ Quint. 10.1.78. The Callimachean echo perhaps not carried over from Dionysius was not noticed by Peterson (1891), ad loc.: ἥτις καθαρή τε καὶ ἀκράαντος ἀνέρπει / πίδακος ἐξ ἱερῆς ὀλίγη λιβὰς ἄκρον ἄωτον (Call. Hymn. 2.111 12). The big river (Ἀσσυρίου ποταμοῖο μέγας ῥόος, 108) comes just before. Other river passages in Quintilian include 5.13.13; 5.14.31 (eloquence, like big rivers rather than ieiuni fontes, should make a way for itself if it finds none available to it); 9.4.7; 9.4.61 (in a long period the ears ductae . . . velut prono decurrentis orationis flumine tum magis iudicant cum ille impetus stetit).

168

Papers on Quintilian and Ancient Declamation

utrimque ripis inumbratus.⁶ But it was the grand style that had the real thrust: ille, qui saxa devolvat et pontem indignetur et ripas sibi faciat multus et torrens,⁷ iudicem vel nitentem contra feret cogetque ire qua rapiet.⁸ There might be, as in the case of Cicero, such a blending of styles as to disguise the force, ut ipsa illa quae extorquet impetrare eum credas, et cum transversum vi sua iudicem ferat, tamen ille non rapi videatur sed sequi.⁹ The judge, in his helpless boat athwart the current, might have some illusion of being in control. In fact, vis was sweeping him away.¹⁰ But if the boat is out of control, how controlled is the river? The orator might make the audience go where he would, but might he not be carried away by his own impetus?¹¹ A safer analogy was that of the horseman. Sometimes, indeed, the orator had to ensure that the impetus of his own speech was not lost.¹² Too much attention to fitting words into pleasant mosaics could lead to a failure of heat¹³ and drive, ut equorum cursum delicati minutis passibus frangunt.¹⁴ On the other hand (though Quintilian is here talking about writing), when we find ourselves going too quickly, we should stop ut provideamus et ferentes¹⁵ equos frenis quibusdam coerceamus, quod non tam moram faciet quam novos impetus dabit.¹⁶ The horse is an irrational beast, but he can be ruled by rational man, made to run in the right direction and at the right speed. This is a comforting picture: the orator in

⁶ Quint. 12.10.60. ⁷ Tac. Dial. 24.1: Quo torrente, quo impetu saeculum nostrum defendit! (with Gudeman (1914²), ad loc.). Also Quint. 3.8.60 (tumultuosius atque turbidius follows) and 10.7.23: Id potius quam se inani verborum torrenti dare quasi tempestatibus quo volent auferendum. ⁸ Quint. 12.10.61. ⁹ Quint. 10.1.110. ¹⁰ No fun: non aliter quam qui adverso vix flumine lembum / remigiis subigit, si bracchia forte remisit, / atque illum in praeceps prono rapit alveus (9.4.7 suggests that Quintilian understood this as the river) amni (Verg. Georg. 1.201 3). Virgil pairs this passage with horses out of control (Georg. 1.512 14): see what follows. ¹¹ Sen. Con. 7.1.20: Quidam principia tantum habuerunt in sua potestate, deinde ablati sunt impetu; Sen. Ben. 1.10.1: Sed longius nos impetus evehit provocante materia. At Quint. 12.pr.3 suo iam impetu fertur means ‘carried along by’, as in Aen. 12.369 cited earlier. ¹² Cf. Quint. 3.8.60; 9.4.35 (worry about hiatus); 10.7.14 (infelix verborum cavillatio); 11.2.46 (weak memory); 11.3.134 (sitting down); 12.9.18 (over preparation). Also Sen. Ben. 7.8.2 (elo quentiae . . . non concinnatae nec in verba sollicitae contrasted with prout impetus tulit). ¹³ Also associated with impetus at Quint. 10.3.6 and 17 (note also 10.7.13: si calor ac spiritus tulit). At 10.3.18 it is used in connection with emotions and contrasts with diligentia. See also n. 16 below, and calescimus in Ov. Fast. 6.5, cited below, p. 174; also Plin. Ep. 2.19.2 (lost in rec itation) and 7.9.6 (recalescere ex integro et resumere impetum fractum). ¹⁴ Quint. 9.4.113. ¹⁵ Si vera lectio. It would, I think, be transitive (‘carrying us away’). I am not sure the analogy works well; perhaps better that with long jumping at Quint. 10.3.6 (et velut repetito spatio sumit impetum). ¹⁶ Quint. 10.3.10. Cf. also the slave employed by C. Gracchus to play the flute in order to regulate his pronuntiatio, quia ipsum calor atque impetus actionis attentum huiusce tempera menti aestimatorem esse non patiebatur (V. Max. 8.10.1). Something similar is related of Hate rius (Sen. Con. 4.pr.8).

On Impulse

169

control of his mettlesome medium. More comforting still if he is, like the charioteer in Plato’s Phaedrus,¹⁷ in full charge of lower elements of the soul. He should certainly govern his own emotions: bonus altercator vitio iracundiae careat; nullus enim rationi (i.e. the speaker’s own rational judgement) magis obstat adfectus, et fert extra causam plerumque, et deformia convicia facere ac mereri cogit (6.4.10). But it was all very difficult. Quintilian followed Cicero in believing that to move others one must be moved oneself, and he explains how such emotion should be induced: Nos illi simus quos gravia indigna tristia passos queremur, nec agamus rem quasi alienam, sed adsumamus parumper illum dolorem.¹⁸ He had been able to do the trick himself: frequenter motus sum ut me non lacrimae solum deprenderent, sed pallor et veri similis dolor.¹⁹ Like the real thing: so perhaps it was only an act,²⁰ and the speaker’s self-control was not prejudiced.²¹ At least it was only for a little. What, meanwhile, of the auditors? They are not even the horses of this analogy. We have, for them, to return to the raging river; and in the face of its fury, the judges, though rational, are under force majeure that makes them more helpless than a bridled horse. And they are thus dominated because the orator can play on their emotions and deprive them for a time of their reason. In the end their pity will weaken: fatigatur lacrimis auditor et requiescit et ab illo quem ceperat impetu ad rationem redit.²² But in the meantime, cum irasci favere odisse misereri coeperunt, agi iam rem suam existimant,²³ et, sicut amantes de forma iudicare non possunt quia sensum oculorum praecipit animus, ita omnem veritatis inquirendae rationem iudex omittit occupatus adfectibus: aestu fertur et velut rapido flumini obsequitur.²⁴ The impetus of the speech is all-powerful because its counterpart is the impetus of the hearer’s emotions. Such a doctrine implies some degree of contempt for the judges thus manipulated. Impetus was characteristic of children, of youth, of animals, of the Roman mob;²⁵ and it might have been tactful to the propertied persons who made up Roman juries to represent them as less easily swayed. The ideas were Greek, and perhaps reflected an aristocratic disdain for the large juries of democratic Athens. It was Greeks again who had raised moralizing objections ¹⁷ Plat. Phaedr. 253e 4b. ¹⁸ Quint. 6.2.34. ¹⁹ Quint. 6.2.36. ²⁰ Cf. the tragedians in Quint. 6.2.35. ²¹ Cf. Cic. Tusc. 4.55: Oratorem vero irasci minime decet, simulare non dedecet (so also Sen. Dial. 4.17.1). But note Orat. 130 2 as well as De orat. 2.189 90. ²² Quint. 6.1.28. ²³ Cf. Quint. 6.2.34 (cited above) on the orator: nec agamus rem quasi alienam. The speaker takes over his client’s feelings and transfers them to the judges. One is reminded of the analogy of the magnet and rings in Plato’s Ion: the rhapsode’s tears (535c.5), aroused by pity for the Homeric characters, are matched by those of the audience (535e.1). ²⁴ Quint. 6.2.6. ²⁵ Quint. 1.3.10 (boys); Sen. Cl. 1.1.3 (youths); 2.20.9 (lions); Cic. Rep. 1.9 (insanos atque indo mitos impetus volgi).

170

Papers on Quintilian and Ancient Declamation

to such methods, as Quintilian well knew: Fuerunt et clari quidem auctores quibus solum videretur oratoris officium docere, . . . primum quia vitium esset omnis animi perturbatio, deinde quia iudicem a veritate depelli misericordia gratia ira similibusque non oporteret.²⁶ But Quintilian could no more accept so austere a view than Cicero: Nihil est enim in dicendo (he makes Antonius say to Catulus) maius quam ut faveat oratori is qui audiet, utique ipse sic moveatur ut impetu quodam animi et perturbatione magis quam iudicio aut consilio regatur: plura enim multo homines iudicant odio aut amore aut cupiditate aut iracundia aut dolore aut laetitia aut spe aut timore aut errore aut aliqua permotione mentis quam veritate . . . aut legibus.²⁷ It was just a fact that such emotions influenced judgements; and in that case it was essential that an orator should appeal to those emotions. Antonius put very frankly what Plato had found so shocking about the art of rhetoric. So far we have seen nothing to surprise us in the use of the word impetus in the discussion of oratory. The speaker and his speech have impetus like the horses and rivers to which they are compared. That onrush corresponds to the onrush of emotions that override the hearer’s reason. In this connection too impetus is commonly used in Latin.²⁸ Impetus and ratio make a natural contrast, for sunt quidam inrationabiles impetus animorum.²⁹ The assault of various strong emotions—grief, love, anger—is regularly called impetus.³⁰ Such onrushes overwhelm the defences of reason. And the most notable case of loss of self-control, madness itself, is described in the same terms.³¹ The image of the horse returns in such a passage as Seneca’s Moderare, alumna, mentis effrenae impetus, / animos coerce.³² Nor is river imagery found inappropriate for πάθη: Potius fugientia ripas / flumina detineas . . . / quam

²⁶ Quint. 5.pr.1. ²⁷ Cic. De orat. 2.178. ²⁸ Our evidence suggests that Cicero was all important in the extension of the use of impetus to mental assaults (from Inv. 2.17: Impulsio est quae sine cogitatione per quandam affectionem animi facere aliquid hortatur, ut amor, iracundia, aegritudo, vinolentia (cf. Dig. 48.19.11.2) et omnino omnia in quibus animus ita videtur affectus fuisse ut rem perspicere cum consilio et cura non potuerit et id quod fecit impetu quodam animi potius quam cogitatione fecerit; cf. 2.19, but not thus in Rhet. Her.). The influence of the Stoic ὁρμή is obvious (Ambr. Off. 1.228: appetitus il le qui quasi quodam prorumpit impetu, unde Graece ὁρμή dicitur quod vi quadam serpente proripiat, building on Cic. Off. 101); as Harry Hine remarks to me, ὁρμή like impetus could get out of hand: πᾶν πάθος ὁρμὴ πλεονάζουσα (SVF .205). The extension to discussion of oratory comes in De oratore. ²⁹ ‘Quint.’ Decl. min. 325.13. For the contrast see TLL s.v. impetus 609.23 4. Note Cic. Inv. 2.164: Temperantia est rationis in libidinem atque in alios non rectos impetus animi firma et moderata dominatio. ³⁰ Quint. 6.pr.14 (grief: contrast alia cogitatio); Sen. Ep. 104.13 (love: see n. 32 below); V. Max. 5.9.1. ³¹ e.g. Cic. Dom. 119: omni impetu furoris in eum civem inruerit for madness can lead to physical attack (cf. Sen. Dial. 5.3.2: ruat). ³² Sen. Phaed. 255 6. Or (again) Sen. Ep. 104.13: cupiditates refrenavit . . . indomitos amoris impetus fregit.

On Impulse

171

miseros lugere vetes.³³ We come now, however, to a more specialized usage, though one that arises naturally enough from the others.³⁴ In 2.11–12 Quintilian, embarking on the preliminary questions traditionally dealt with in the προλεγόμενα to an ars rhetorica, confronts opponents who deny the very need for such an ars.³⁵ These ‘naturalists’ rely on their own talents and on the ordinary school exercises; they think that eloquence has no need of praecepta. They make do, that is, with two of the customary triad, φύσις and ἄσκησις; τέχνη they repudiate. (2.11.3) Igitur impetu dicere se et viribus uti gloriantur: neque enim opus esse probatione aut dispositione in rebus fictis, sed . . . sententiis grandibus . . . (4) Quin etiam in cogitando nulla ratione adhibita aut tectum intuentes magnum aliquid quod ultro se offerat pluribus saepe diebus expectant . . . (6) Qui plurimum videntur habere rationis non in causas tamen laborem suum sed in locos intendunt . . . (7) Unde fit ut dissolute . . . oratio cohaerere non possit . . . Magnas tamen sententias et res bonas (ita enim gloriari solent)³⁶ elidunt; nam et barbari et servi, et, si hoc sat est, nulla est ratio dicendi . . . (2.12.9) Verum hi pronuntiatione quoque famam dicendi fortius quaerunt; nam et clamant ubique et omnia . . . emugiunt, multo discursu anhelitu, iactatione gestus, motu capitis furentes . . . (11) At illi hanc vim appellant quae est potius violentia: cum interim non actores³⁷ modo aliquos invenias sed . . . praeceptores etiam qui, brevem dicendi exercitationem consecuti, omissa ratione ut tulit impetus passim tumultuentur, eosque qui plus honoris litteris tribuerunt ineptos et ieiunos et tepidos et infirmos, ut quodque verbum contumeliosissimum occurrit, appellent. (12) Verum illis quidem gratulemur sine labore, sine ratione,³⁸ sine disciplina disertis. It is clear why Quintilian disliked these people. As teachers they were rivals who might seem to offer pupils an easier and more attractive ride than the ³³ Stat. Silv. 5.5.62 4. ³⁴ For Greek use of ὁρμή (see above, n. 28) in literary criticism, Doreen Innes points out to me e.g. Philostr. VS 568: χολή τε γὰρ ἄπεστι τοῦ λόγου καὶ ὁρμαὶ πρὸς βραχύ (‘impulsive outbreaks on the spur of the moment’ Wright) and, for inspiration, Philostr. VS 533 (Polemo): προοίμιον ποιούμενος τοῦ λόγου τὸ μὴ ἀθεεὶ τὴν περὶ αὐτοῦ ὁρμὴν γενέσθαι. But, as she remarks, φορά is also relevant: ‘Longin.’ Subl. 2.2 (φορᾷ καὶ ἀμαθεῖ τόλμῃ) and 20.2 of πάθος (cf. 21.2); [Luc.] Dem. enc. 7 (of narrative flow?). The two combine of inspiration in θεοφορήτῳ ὁρμῇ at Philostr. VS 509, cited in n. 61 below. ³⁵ Cf. ‘Longin.’ Subl. 2.1. See later Aug. Doctr. Chr. pr.4 (praecepta tractandarum scripturarum not needed, divinum munus enough). We hear of naturalists elsewhere in Quintilian. But whereas in Book 2 they sound like Asian declaimers (epigram, pravae voluptates), they later resemble the neo Atticists; cf. 9.4.3: quosdam . . . qui curam omnem compositionis excludant, atque illum horridum sermonem, ut forte fluxerit, . . . magis naturalem . . . esse contendant; 11.3.10: Sunt . . . qui rudem illam et qualem impetus cuiusque animi tulit actionem iudicent fortiorem . . . sed non alii fere quam qui etiam in dicendo curam et artem et nitorem . . . ut adfectata et parum naturalia solent improbare; 12.10.40: quidam nullam esse naturalem putant eloquentiam nisi quae sit cotidiano sermoni simillima (and what follows). ³⁶ Cf. Sen. Con. 7.pr.9. ³⁷ For naturalists in the courts see Winterbottom (1964b) [= A.1 above]. ³⁸ Cf. Quint. 2.20.21.

172

Papers on Quintilian and Ancient Declamation

austere school of Quintilian. As pleaders, they were a disgrace to the profession, yet might be successful without deserving to be. If they were right, there was no point in Quintilian writing the Institutio at all. His criticisms are accordingly edged with satire and irony. We saw that Quintilian found admirable in Demosthenes his vis and his impetus.³⁹ The naturalists seem to be boasting of the same qualities. As to vis, Quintilian argues that in their case it is rather to be called violentia. He devotes a good deal of the passage to denying that such speakers truly speak fortius than the educated.⁴⁰ One notices, particularly, his disapproval of indiscriminate abuse, which may bring danger to speaker and client alike:⁴¹ a warning that coheres with the already cited advice that the good altercator should avoid anger. The naturalists are represented as lacking self-control as well as education. Indeed, the claim to speak impetu Quintilian turns back against its proponents by exploiting the familiar contrast between impetus and reason. It is of course true that the primary sense of ratio in these sections—nulla ratione adhibita; qui plurimum videntur habere rationis; sine ratione disertis—is ‘method’; these people lack the ὁδός which is an integral part of a τέχνη.⁴² But there is some hint of the deeper connotation of ‘reason’. These shouting and wildly gesticulating speakers are described as furentes.⁴³ Less emotively, Quintilian argues that the uneducated speaker is liable to lack order.⁴⁴ The speech will not hang together,⁴⁵ with the result that the speakers passim tumultuentur;⁴⁶ for (7.pr.3) oratio carens hac virtute (sc. ordo) tumultuetur necesse est et sine rectore fluitet nec cohaereat sibi. ‘Good things’ and ‘big epigrams’ are, in Quintilian’s eyes, no substitute for the basic virtue of organization. So far as the appeal to nature is concerned, Quintilian answers elsewhere that to be an orator it is not enough just to be born.⁴⁷ On the contrary, id est maxime naturale quod fieri natura optime patitur.⁴⁸ Seen in

³⁹ Quint. 12.10.23. ⁴⁰ See Quint. 2.12.1 3. For fortiorem cf. 11.3.10 (cited in n. 35). ‘Force’ went with ‘virility’ (9.4.3; 11.3.10: solam viris dignam); compare the Atticist attacks on Cicero as effeminate (12.10.12: viro molliorem). See Cic. De orat. 1.23.1: fortem et virilem (also Sen. Ep. 114.22). ⁴¹ Quint. 2.12.4. Cf. 12.9.8 13. Even the naturalists’ abuse of people like Quintilian is random: ut quodque verbum contumeliosissimum occurrit (2.12.11, cited above). ⁴² See Quint. 2.17.41 (Cleanthes): Ars est potestas via, id est ordine, efficiens; Dion. Hal. Comp. verb. 206: ὁδοῦ τε καὶ τέχνης. ⁴³ Quint. 3.8.59: cur initio furioso (? ose) sit exclamandum non intellego; 11.3.45: ne dicamus omnia clamose, quod insanum est; Cic. Brut. 233 (cf. De orat. 3.136: clamore et . . . verborum cursu). For shouting cf. 4.2.37: tumultu et vociferatione; 7.1.44: pulchre fuerit cum materia tu multu et clamore transactum (associated with sententiae praecipites); Luc. Rh. pr. 15. ⁴⁴ Cf. Sen. Con. 4.pr.9 (the extemporizer Haterius): Is illi erat ordo quem impetus dederat (on the same man Tac. Ann. 4.61: impetu magis quam cura vigebat). ⁴⁵ Quint. 2.11.7. ⁴⁶ Quint. 2.12.11. ⁴⁷ Quint. 11.3.11; cf. Sen. Ben. 3.30.4. ⁴⁸ Quint. 9.4.5.

On Impulse

173

this light, natura is a guide to be followed in the pursuit of the highest standards, not a licence to ignore the precepts of one’s betters.⁴⁹ To a degree, the naturalists were pressing the claims of extempore oratory. Quintilian describes unsympathetically the disorganized nature of such preparation as they did;⁵⁰ and when it came to the actual speech abrupta quaedam, ut forte ad manum venere, iaculantur.⁵¹ It is illuminating to compare, and contrast, with their procedures what Quintilian will say about extemporization. At the writing stage, gazing at the ceiling and rousing one’s thought processes by humming⁵² will not do; only an orderly consideration of quid res poscat, quid personam deceat, quod sit tempus, qui iudicis animus will enable us to approach composition humano⁵³ quodam modo.⁵⁴ Nor is it any better first decurrere per materiam stilo quam velocissimo . . . sequentes calorem atque impetum: even the revised version of such effusions leaves traces of the original superficiality.⁵⁵ We come to extempore speech in 10.7. It is praised as an indispensable weapon in an orator’s armoury, one which he must be able to use even though he might prefer not to.⁵⁶ But it is noticeable that Quintilian’s account stresses the features that make his ideal of extemporization so different from the practice of the naturalists. In particular, it is not just a gift of nature. An usus inrationalis, analogous to the eye’s ability to read without conscious thought, is conceivable (11), but it must be preceded by ars, ut ipsum illud quod in se rationem non habet in ratione versetur (12). That ars is as learnable as anything else in rhetoric, and it only comes with long practice. And, in a particular case, there is a set method, a via, to be followed consciously from start to finish of a speech (5–6). The declaimers criticized in 10.7.21, who exposita controversia . . . verbum petant quo incipiant⁵⁷ are the extreme case of the naturalists who certa sibi initia priusquam sensum invenerint destinant.⁵⁸ All those who do not speak disposite ornate copiose seem to Quintilian to rant (tumultuari, 10.7.12) just like the naturalists ‘all over the place’ (passim) in 2.12.11. As for impetus, the word is tamed and found a safe home here. A successful extemporization is explained in terms of well-conceived emotions and fresh images that continuo impetu feruntur,⁵⁹ like any good passage of oratory, and Quintilian remarks, as he has done

⁴⁹ Typical are Quint. 4.5.3; 5.10.101; 7.1.40; 8.3.71; 12.10.44: quo quisque plus efficit dicendo, hoc magis secundum naturam eloquentiae dicit. ⁵⁰ Quint. 2.11.4. ⁵¹ Quint. 2.11.6. ⁵² Cf. Quint. 2.11.4. ⁵³ A striking adjective: impetus is less than human (Sen. Dial. 4.16.1: Errat qui ea (sc. animalia) in exemplum hominis adducit quibus pro ratione est impetus: homini pro impetu ratio est). ⁵⁴ Quint. 10.3.15. ⁵⁵ Quint. 10.3.17. ⁵⁶ Quint. 10.7.4. ⁵⁷ This all goes back to the First Sophistic (Philostr. VS 482: Gorgias’ Προβάλλετε). For the Second, see Russell (1983); an extemporizer might prepare for a short time (cf. Quintilian’s ad vice at 10.7.20). ⁵⁸ Quint. 2.11.5. ⁵⁹ Quint. 10.7.14.

174

Papers on Quintilian and Ancient Declamation

before, on the danger of over-anxious search for words. Impetus is given its place here subordinate to or hand in hand with art. The declaimers who sat looking at the ceiling for aliquid quod ultro se offerat⁶⁰ are in our parlance ‘waiting for inspiration’. And there hangs over Quintilian’s description of the naturalists a faint air of the mad poet. In 10.7.14 Quintilian himself points to an old saying that a god was present when an extempore stroke proved successful.⁶¹ But we have seen that he explains such success away without recourse to supernatural aid (ratio (!) manifesta est). A Plato might be imagined to be inspired (10.1.81: ut mihi non hominis ingenio sed quodam Delphico videatur oraculo instinctus), but for the rest of us it is ars and studium that must make up the deficiencies of ingenium. A poet, of course, might see things differently.⁶² Ovid, in Tomi, felt he could not compose as he had before (Pont. 4.2). That could be put in the language of ingenium (15: Nec tamen ingenium nobis respondet ut ante) or of the Muses (27: Vix venit ad partes . . . Musa); but another way of putting it was that Impetus ille sacer qui vatum pectora nutrit, / qui prius in nobis esse solebat, abest (25–6). Or, more obscurely, in the proem of Fast. 6: Est deus in nobis; agitante calescimus illo; / impetus hic sacrae semina mentis habet (5–6). Is the impetus the onrush of the inspiring god, who stirs up the poet to compose, or something internal to the poet that could be thought of as a god? In either case, the connection of impetus with the divine remains clear. Impetus in extemporization and inspiration⁶³ deserved to find some mention in the Oxford Latin Dictionary.[⁶⁴] But the extension of usage is slight. We are still concerned with something to be contrasted with ratio, something we can and perhaps should rein in with the help of reason, something which is ultimately alien and mysterious as well as indispensable. * * * Even the most devoted admirer of Quintilian will concede that for acuteness and eloquence he must bow to ‘Longinus’. But the two critics come together interestingly in the matters I have been discussing. If there is an answer to my ⁶⁰ Quint. 2.11.4; cf. 10.3.15: expectaverimus quid obveniat. ⁶¹ Cf. Philostr. VS 509: Τὸ γὰρ θείως (cf. μὴ ἀθεεί in 533, cited in n. 34) λέγειν . . . ἀπ’ Αἰ σχίνου . . . ἤρξατο θεοφορήτῳ ὁρμῇ αὐτοσχεδιάζοντος, ὥσπερ οἱ τοὺς χρησμοὺς ἀναπνέοντες (for this verb cf. 515), reminiscent of ‘Longin.’ Subl. 13.2, cited below (also Sen. Suas. 3.6 7). For im petus of divination see Fronto, p. 5.7 8 van den Hout²; cf. Cic. Div. 1.111. ⁶² Cic. Fin. 4.10: Quod etsi ingeniis magnis praediti quidam dicendi copiam sine ratione con sequuntur, ars tamen est dux certior quam natura; aliud est enim poetarum more verba fun dere, aliud ea quae dicas ratione et arte distinguere (cf. Arch. 18). ⁶³ Sen. Nat. 3.27.13: tantum impetum ingenii of Ov. Met. 1.292 (picked up in 14 by impetus used of the flood); Petr. 118.6: hic impetus on the civil war (contrasted with ultimam manum); Tac. Dial. 10.6: fortuitae et subitae dictionis impetu; Ann. 14.16.1: impetu et instinctu; Suet. Aug. 85.2: magno impetu opposed to succedente stilo. ⁶⁴ [The new edition (2012) has no changes under this entry.]

On Impulse

175

criticism of Quintilian’s admiration of the great orator sweeping his audience away on a flood of emotion, it might be on the same lines as what Donald Russell has written of ‘Longinus’: ‘[He] has a further expectation: he requires that the “emotion” shall be not only vehement but of a certain moral quality . . . This is why the personality of the writer is important. He cannot be expected either to think grand thoughts or to generate and excite grand emotions if he is ravaged by desire for gain or money, or deaf to the calls of honour and posthumous fame.’⁶⁵ Quintilian’s ideal of the vir bonus, whom the Institutio aims to train no less than the vir peritus dicendi, is reaching in the same direction. As for inspiration, ‘Longinus’, like Quintilian, connected it with the Pythia at Delphi: ‘she is in contact with the tripod near the cleft in the ground which (so they say) exhales a divine vapour, and she is thereupon made pregnant by the supernatural power and prophesies as one inspired. Similarly, the genius of the ancients acts as a kind of oracular cavern, and effluences flow from it into the minds of their imitators.’⁶⁶ But such imitation could only be the product of years of devoted study of the great models. For both ‘Longinus’ and Quintilian genius was an infinite capacity for taking pains. [The use of the word impetus in later and especially Christian authors would be worth investigation. In William of Malmesbury’s twelfth-century commentary on the book of Lamentations I note (CCCM 244 (Turnhout, 2011), p. 180): ‘Considerandus est sanctissimi prophetae [Jeremiah] affectus, quomodo sermonem variet modo huc modo illuc, sicut eum sancti Spiritus agebat impetus’—the Holy Spirit intervening in a matter of style.]

⁶⁵ Russell (1981), 82.

⁶⁶ ‘Longin.’ Subl. 13.2 (trans. Russell).

13 Quintilian the Moralist This paper will be concerned with Quintilian’s use of moral language and his deployment of moral concepts. The picture that will emerge is that of a man who comes to see, as the over-sixties do, that things are by no means so satisfactory now as they were in his youth, let alone in bygone eras; but of a man too who exhibits the acquired Roman attitudes of a native of Calahorra. I start from a sentence in the prologue to the first book of the Institutio (1.pr.11). ‘Although’, says Quintilian, ‘I confess that I shall be using certain things that are contained in the books of the philosophers, yet I can claim, with right and truth to support me, that this material is part of our subject and belongs properly to the art of oratory.’ This sentence is significant in two ways. It affirms the philosophical content of the Institutio, and at the same time it acknowledges the prolonged demarcation dispute that had made such topics seem out of place in an ars rhetorica.¹ The first of these points, the philosophical content, is in effect the subject of this lecture. But I need also to say something about the relations between rhetoric and philosophy generally. Philosophers of Quintilian’s own day are a minor but diverting part of the story. They first appear in the same prologue, concealing their deplorable behaviour behind their austere looks and their unorthodox clothes.² And later (12.3.12) we hear of their schools as a place where lazy but arrogant students sit for a while, assuming a severe expression and sporting a long beard, before launching themselves on the world as claimants to auctoritas on the basis of a melancholy appearance that is merely a cover for vice: in publico tristes, domi dissoluti. One senses in such passages the tone of a rhetor whose school is sometimes judged less grown-up than the rival establishment down the road; and also, perhaps, the tone of a favourite of Domitian who is taking on, or affecting, some of the Flavian suspicion of practising philosophers. But these

[From T. Albaladejo, E. del Río, and J. A. Caballero (eds.), Quintiliano: Historia y Actualidad de la Retórica, vol. I (Logroño, 1998), 317 34] ¹ Cf. the similar wording of Quint. 1.10.11 (music); also 2.1.1 6 (grammar). ² Quint. 1.pr.15; cf. 2.16.5.

Quintilian the Moralist

177

moderns are open to a yet more serious criticism, that links them to their predecessors down the ages. They ‘think everything comes second to their precepts and their writings’ (5.11.39). Yet it goes no further than books. ‘Which philosopher has practised the administration of the res publica (πολιτική) for which so many of them give precepts?’ Philosophers, Quintilian assumes, are theorists; his ideal pupil is to be a Romanus quidam sapiens, who will, unlike the impractical Greeks, prove himself in action a vir vere civilis.³ In what follows this passage in 12.2, Quintilian gives typical expression to his views on what I called the demarcation dispute with the philosophers. He has already recalled the Ciceronian line that the faculty of speaking flows from the inmost founts of philosophy, and that there was once a time—very long ago indeed—when the same teachers served for both morals and speaking (12.2.6). Quintilian’s argument, as he now develops it, is as critical of rhetors as of philosophers. Teachers of rhetoric have long ago given up concern for philosophy. But it remains necessary for orators, and they are forced to turn to the written doctrine of those ‘who give precepts on virtue’ (12.2.8). That allusion to virtue takes up the earlier remark about teachers of morals. But it transpires that Quintilian is thinking on a wider front. And he goes through the three major divisions of philosophy—natural, moral, rational—in turn (12.2.10–23). The rational branch (λογική) is seen rather as providing techniques, patterns of analysis, to the orator. Moral and natural philosophy, on the other hand, are regarded as supplying the ability to talk on abstract themes: justice, bravery, abstinence, and the like in the one case, augury, oracular responses, and religion generally in the other. In this wider context there seems no particular reason why Quintilian should be worried that philosophy should have to be learned from philosophy teachers, by book or in person, rather than from rhetors. Earlier in the Institutio Quintilian had had little to say about the desirability of philosophical—that is, abstract—passages in oratory.⁴ The fact is that the discussion in Book 12 is a throw-back to the teachings of Cicero, teachings that had arisen from the particular interests of the great man and had no necessary validity for lesser mortals. Even in the case of Cicero it is not clear that his occasional abstract passages—like that on glory at the end of the Pro Archia—add much to his argument; and I do not need to recall the shaming sections of the Pro Murena,⁵ where Cato is ridiculed for his philosophical beliefs in words that play to an ignorant audience. Cicero thought a philosophical training important to the perfect orator because he regarded himself

³ Quint. 12.2.7; cf. 1.pr.10; 11.1.35. For sapiens see also 1.10.6. ⁴ But note 1.pr.12 (cf. 16); 2.1.21 3; 2.21.12 13; 3.4.16 (cf. 3.5.3). In 10.1.35 it is the phi losophers who talk of such things. ⁵ Cic. Arch. 28 30; Mur. 61 2 (Fin. 4.71: Apud imperitos tum illa dicta sunt, despite Mur. 61 ad init.).

178

Papers on Quintilian and Ancient Declamation

as approaching perfection—and he had always loved and studied philosophy: quod erat demonstrandum. That made little sense for his less intellectual contemporaries, even though they could go to study philosophy at Athens, as his own son made a show of doing. As for Quintilian, his sketch of the usefulness to the orator of the various schools of philosophers has a faded and out-of-date air about it; and it ends in the affirmation that the orator does not need a particular school to follow at all.⁶ He is—and here we approach the real concerns of Quintilian—to be himself ‘perfect in distinction both of life and of eloquence’; and just as he will take as exemplar the most eloquent speaker he can find, so he will look for ‘the most high-minded precepts and the straightest path to virtue’.⁷ This, I take it, is more important to Quintilian than any philosophical content in oratory: the orator is to be a good man as well as a good speaker,⁸ and the tragedy of the split between philosophy and oratory had been that—as we saw he put it earlier in 12.2—the same person had ceased to teach both morals and the art of speaking. Quintilian confuses the issue, as Cicero had confused it, by his talk of what an orator could say that would exhibit the influence of a philosophical training. His real aim was the production of a morally good speaker; and the counterpart of that was a teacher who regarded himself as inculcating both ἀρετή and ῥητορική. That was to go back to Gorgias and the ideals of the earliest sophists. And if the sought-for orator was ever achieved, he would retrieve from the philosophers what the orators should never have conceded, the whole realm of morality.⁹ That is clear from 1.pr.10: it is the orator who is aware of the ratio rectae honestaeque vitae, and who in conformity with it carries out his task of ‘guiding cities by his counsels, establishing them by his laws, and correcting them by his judgements’ (cf. 12.7.2). One wonders indeed if these echoes of the start of Horace’s Epistle to Augustus do not suggest that Quintilian had in mind, in this picture of the political orator, an idealized emperor figure. He had himself taken on the education of Domitian’s nephews; and he perhaps half-fancied himself as the Isocrates who could train them in political virtue or even as the Aristotle poised to educate a new Alexander (1.1.23). The rhetor, then, is to be a teacher of morals as well as of rhetoric, and his product, the vir bonus dicendi peritus, is to take back ground lost to philosophy so dramatically that he will appropriate the task of correcting vice

⁶ Quint. 12.2.24 6 (cf. 30: exempla matter more). ⁷ Quint. 12.2.27. In 28 Quintilian stresses the benefit an orator gains from practising on abstract philosophical themes (cf. 20). That is different from learning philosophy and using it in real speeches. ⁸ Already in 1.pr.9. ⁹ Conceding: Quint. 10.1.35. For the retrieval: 12.2.9 (vindicet sibi . . . velut rebus repetitis); cf. 1.pr.17: reposcere.

Quintilian the Moralist

179

and forwarding virtue. For his part—the part of a teacher—Quintilian does everything he can to come up to his ideal, and loses no opportunity to contrast his practice with that of others. He is indeed concerned with the behaviour of his pupil while he is still a baby: ‘What will a child who crawls around clothed in purple lust after when he is grown up?’ (1.2.6). It is important, Quintilian thinks, to pay attention to the morals of the nurse as well as to her correctness of speech (1.1.4); and so too with the παιδαγωγός (1.1.9). Again, what a child copies out for writing practice should ‘give some morally good advice’, for what you learn at that age stays with you for ever (1.1.35–6). Morality too is an important consideration in the debate between education at home as opposed to education at school (1.2.2–8). In sum, ‘there is no age so infirm that it cannot learn from the start what is right and what is wrong’ (1.3.12). Quintilian naturally pays special attention to the character of the rhetor himself and the behaviour to be expected of him: not, he says, that the moral excellence of teachers is not always vital, but a child at the age when he goes to the rhetorical school is particularly vulnerable (2.2.2). So the rhetor must ‘speak much of what is right and good: the more he admonishes, the less he will need to punish’ (2.2.5). A clear warning against homosexual practices, whether between younger and older students or between teacher and pupils, concludes the chapter (2.2.14–15). There is little room for moral precept in the technical books that follow 2; but they of course represent the syllabus followed by the morally sound teacher who has been described so carefully. And when we come to the reading list to be presented to the fully trained pupil, morality is not forgotten. Love elegy, unsuitable for the younger student (1.8.6), is now permissible (10.1.93); but Afranius’ homosexual amours are disapprovingly said to have left their mark on his togatae (10.1.100). On the other hand, Alcaeus is said to contribute much to mores (10.1.63), and Seneca, however deplorable his style, is praiseworthy as an ‘excellent flayer of vices’— he is in fact author of much that is to be read morum gratia, for its moral effect (10.1.129). But it is of course in Book 12 that we hear most of morality. The prologue deploys a fanfare of trumpets. Even Cicero is now to be surpassed, as Quintilian sets out across uncharted seas in his temerarious attempt to give the orator mores and assign him officia (12.pr.4). And the contrast with Cicero is more subtly marked by a reworking of a phrase of his. Cicero, as we have seen, and as Quintilian points out later in the book, saw excellence in oratory as arising from the inmost founts of philosophy (12.2.6). For Quintilian the orator under training is to seek ‘greater aids from the very penetralia of wisdom’ (12.pr.3). The role of philosophy is now to help with the moral education of the orator far more importantly than with the content of his speeches. And just as Quintilian thought an orator needed training as well as natural talent, so he regarded goodness as something that requires training as

180

Papers on Quintilian and Ancient Declamation

well as a basis in one’s inborn character (12.2.1–2).¹⁰ That is firmly stated at the start of a chapter in which Quintilian eventually seems to conclude that Roman exempla were at least as important as Greek philosophical doctrines. The aim of all this was of course the production of a vir bonus. If we ask what are the characteristics of such a person, we receive no systematic answer: no doubt Quintilian thought that any Roman would know what the words meant (cf. 5.10.17–18). But we can get an idea of his conception by examining his explicit statements about the good man, together with his own moral precepts. As one might expect, he has no quarrel with conventional morality: it is good to be brave, liberal, sparing, as opposed to suffering from the neighbouring vices, rashness, prodigality, parsimony (3.7.25). That means little in itself; it is the application to individual cases that would be revealing.¹¹ A little more helpful are other passages. You will have to avoid too much histrionic elegance in delivery if you want to retain the authority of a man who is bonus—but also gravis (11.3.184). (The passage is related to many in which Quintilian stresses the practical advantages to an orator of being a good man; for his goodness will make him more credible and persuasive—he will, that is, improve his ἦθος.)¹² Again, since Quintilian demands that, in humour and wit, the orator should speak with due regard to his dignity and modesty (6.3.35), we can get some idea of his standards of prudery from the long and illustrated discussion of the whole subject. Anger is to be avoided by the ‘good’ altercator (6.4.10)—and I take it that means or includes moral goodness. Quintilian, who often disapproves of boastfulness in others,¹³ apologizes for telling a story about his own career because he is afraid he may be thought subject to the vice himself (9.2.74). When we come to the morality of the courtroom, he is, as one would expect, strongly supportive of good faith towards a client (10.7.2); the good man will prefer defence to prosecution (12.7.1); and if one is accusing, one should show no pleasure in the task (11.1.57). The good man’s friends will be the best people, optimus quisque, in a moral sense (12.7.5). And, Quintilian touchingly says almost at the end of his book, the good man’s duty is to teach what he knows (12.11.8). It is striking, though, and consonant with the grand role assigned to the orator, that Quintilian is very aware of the public virtues of the vir bonus. It is not just that he praises the political career of Cicero in moral terms: the consulate nobly conducted, the province honestly administered, and—very remarkably—the choice of the best side in the civil wars, id est, rei publicae (12.1.16). Quintilian goes beyond this in assigning his good man characteristics ¹⁰ The will to virtue is also important: Quint. 12.11.11 (cf. 31). Quintilian goes on in 12.11.11 12 to note the place of precept (cf. also 16) and nature in the process. There are interesting remarks in 1.3.2 on the connection between cleverness and morals. ¹¹ Note the discussion of Cicero’s claim to bravery at 12.1.17. ¹² See Quint. 3.8.13; 4.1.7 and 46; 4.2.125 7; 12.1.11 (cf. 12.2.17). ¹³ See Winterbottom (1997), 14.

Quintilian the Moralist

181

that place him firmly on the political scene. He should defend his friends, but also guide the senate with his counsels, and lead the people and the army into courses he approves (2.16.19). He will hate bad men, be concerned at turns in public events, go off to avenge crimes and injustices—et omnia honesta, all is to be upright (11.1.42). Finally, in the last chapter of all, the orator is explicitly called a bonus civis, whose place is at the centre of affairs: law-suits, councils, the senate, speeches before the people. We are reminded that, earlier in the book, Quintilian has said that legal activity will be the least of the tasks of the summus orator: ‘he will shine brighter in greater roles, guiding the counsels of the senate, leading the errant people into better courses.’ It is here that Quintilian uses the famous simile from the Aeneid (1.148–56), where Neptune calming the seas is compared to a statesman quelling a rebellious mob: tum pietate gravem ac meritis si forte virum quem / conspexere, silent arrectisque auribus adstant (151–2). That, explains Quintilian, is the vir bonus. Then: ille regit dictis animos et pectora mulcet (153)—that adds the element of oratory, dicendi peritus (12.1.26–7). A man equipped with these virtues was in some ways handicapped in the real world. And it is one of the most impressive features of the Institutio that Quintilian does not try to minimize the extent to which ideal and reality might clash. The problems all come up, in one form or another, in 2.16–17. In the first of these chapters the question is tackled ‘Is rhetoric useful, or does it do positive harm?’ Quintilian’s answer, in effect, is that it may be harmful, but that a speaker who employed it for evil ends would not be worthy of the sacred name of orator. The riposte is not as feeble as it sounds. Quintilian’s whole aim was to train a speaker who would not, though he could, misuse the art of speaking. Then in 2.17, under the pedantic-sounding heading ‘Is rhetoric an art?’, we approach the difficulty that rhetoric from time to time deals in falsehoods. Quintilian was no doubt aware that his preliminary answer is a mere sophistry: ‘the orator, when he substitutes falsehood for truth, knows that it is false and that he is employing it instead of what is true; therefore he is not himself deceived, but only deceiving others’ (2.17.20). But later in the chapter (2.17.26–8) he reveals a stronger position that he will develop in later passages of the Institutio. ‘People’, he says, ‘make it a charge against rhetoric that it employs vices, as no art would, because it both speaks falsehoods and arouses emotions. Neither of these things is shameful, so long as the motive is good . . . For even a sapiens is at times permitted to tell a lie, and an orator has no choice but to arouse emotion if a judge cannot be brought to see what is aequum in any other way; for judges are unsophisticated, the sort of people who often need to be deceived to prevent them going astray.’ It would be different if wise men made up juries and deliberative bodies; but they do not. In that earlier passage in 2.17 Quintilian was arguing against philosophers with a philosopher’s weapons, and quite unconvincingly. Now he counters the primarily Stoic view that emotions are not to be roused or falsehood

182

Papers on Quintilian and Ancient Declamation

spoken—the two merge in this passage—with the practical point that the world is not like that; emotion and a measure of deceit are essential if the truth is to be arrived at. Audiences are ignorant and irrational, and cannot be trusted to see the point for themselves. Of course, no hostile critic of rhetoric would think this anything other than a concession of the point he was making. But we must recognize again what underpins an otherwise weak case: the assumption of the rectitude of the hoped-for orator. The vir bonus can be trusted, as a vir malus could not, to judge what is the aequitas in the interests of which the ignorant need to be deceived. In 2.17 it is not only bona ratio to which Quintilian appeals (27), the good motives in the orator that can recognize a good reason for deceit when he sees it, but communis utilitas,¹⁴ the common good. That is a theme, and a Ciceronian one, that recurs several times later. In the De officiis the common good can be a reason for doing something that in other circumstances would be less than right and honest. In Quintilian it is employed as a reason that may justify the orator in misemploying evaluative terms (brave for rash, or rash for brave), or rousing emotions despite the objections of the Stoics.¹⁵ And in 12.1 the Stoics are confronted head on. Even the most rigid of them will admit (says Quintilian) that a vir bonus may sometimes tell a lie if the resulting good is great enough: the welfare of a sick child (and Quintilian had lost two young sons), not to speak of the diversion of a robber from his victim or the deception of an enemy for the sake of the patria. Quintilian deduces that there must therefore be many circumstances in which an orator will be right to take up the sort of case that he would not, lacking the special reason, have accepted (12.1.38–9). The philosophers’ own arguments are being taken over to make the point that an advocate cannot, given the complexity of the moral problems he faces, be as pure white as a philosopher theorizing in his study. It is against this background that we must judge certain other passages in the Institutio. There are some things Quintilian is not prepared to tolerate under any circumstances: the suborning of witnesses (5.7.32), or collusion (9.2.87). But elsewhere there is room for moral manoeuvre. Quintilian faces the problem of someone who needs to give less than honourable advice to a good man when the interests of that good man are at stake. ‘Are you actually advising such a thing? Do you think that right and proper?’ he imagines people objecting. And he can only console himself with words of Cicero to the effect that the role of an adviser is to have regard to the interests of the person he is advising, and that advice does not always leave room for what is right (3.8.41–2). That is the way of the world. He is aware that an orator may need to invent—and he recalls without disapprobation a popular saying that a liar should have a good memory, and so be able to stick convincingly to his

¹⁴ Quint. 2.17.36. Cf. 12.1.37 and 43. In the De officiis, note 1.31.

¹⁵ Quint. 3.7.25; 6.1.7.

Quintilian the Moralist

183

deceptive story (4.2.91). He is perfectly happy that the demands of figures may lead to pretence. Thus in 9.2 he remarks that figures ‘adapted to the increasing of emotions are particularly bound up with pretence: we pretend to be angry, to be glad, to be afraid, to be surprised, to be grief-stricken, to be indignant . . . ’ (9.2.26). That is all part of the process of playing on the judges’ emotions, of making them feel what we pretend to feel ourselves. And this, as we have seen, is part of what the philosophers objected to in rhetoric. Quintilian gives the philosophers their say on several occasions; but his answer to them is always the same. This is the only way to deal with ignorant juries, the only way to ensure that the common good, or at least the good of our client, prevails. So far, we have seen Quintilian facing as honestly as he can the insuperable problems of morality that arise from the trade of the lawyer and the politician: problems raised definitively by Plato, and never answered, for there is no answer to them. To define the orator as the vir bonus, and to deny the title of orator to those less than good, is to express a pious wish, to put forward an aim, but in no way to solve a problem. But at least Quintilian is on the side of the angels. I come now, however, to a different, less familiar, and perhaps less attractive side of Quintilian the Moralist. Part of the problem arises from the fact that Quintilian is a workaholic, and expects the rest of us to be workaholic too. Studendum semper et ubique. He pronounces: ‘There is scarcely any day so busy that some moment cannot be won from it for writing or reading or speaking’ (10.7.27). But, as he says in his last chapter of all, we make our own time short: ‘for how small a portion do we allot to studies? Our hours are taken up by the pointless labour of the salutatio, leisure devoted to fabulae (plays? gossip?), the games, dinner-parties. Add in all kinds of sport, the insane concern for one’s body, travel, country pursuits, worry about one’s accounts, vice, wine, pleasure.’¹⁶ One might have thought this list all-inclusive, but one can add more from other passages: the Campus Martius, dicing, casual conversation, sleep (I suppose he means siestas) (1.12.18), daydreaming (called a vice: suggested topics are—interestingly— travel, seafaring, warfare, addressing the people, spending money we do not have: 6.2.30), and the hunt (12.1.6). Worse than the time-wasters are the devotees of the palaestra, who spend half their lives on oil and half on wine, swamping their minds in their concern for their bodies (1.11.15). And then there are the new customs that have invaded the city—or rather, Quintilian corrects himself, not customs at all. For it is ex consuetudine that we bath and cut our hair and eat dinner—but these latest fads cannot count in that class: plucking body-hair, having permanent waves, carousing in the baths . . . (1.6.44). To combine an extreme compulsion to study, at the expense of all the traditional Roman ways of passing the time, with a dislike of new trends

¹⁶ Quint. 12.11.18; cf. 1.12.7.

184

Papers on Quintilian and Ancient Declamation

enables the moralizing sexagenarian to take a happily jaundiced view of the world around him. He can dwell, as a laudator temporis acti should, on things that distinguish the present from a somewhat better past. It is only a throwaway line—in one of several interesting allusions to the relativity of taste¹⁷— when he remarks that for the old Romans luxuria was a crime (3.7.24). More considered are aspersions on the sullen behaviour of modern youth: in his praesertim moribus—manners nowadays being what they are—‘the more spirited students go so far as to resent admonition and to sulk in silence’ (2.6.3). But—and one does sympathize here—it is modern music that really touches a raw nerve. ‘I must make it quite clear’, says Quintilian, ‘that I am not recommending the music that you hear on the stage nowadays, made effeminate by unseemly rhythms, which has cut out in us any manly strength that remained’ (1.10.31). This association between modern trends and sex is entirely characteristic of our author, and it runs through much of what I shall be discussing in the rest of this paper. Quintilian thoroughly confuses literary and moral criticism. He does this to describe, with persuasive intent, oratorical practices of which he approves or disapproves. But I shall suggest that, whether Quintilian intended this or not, the process goes the other way too. In promoting a particular kind of oratory in this manner, he is also promoting a particular view of the way in which people should behave. He is treading firmly on philosophical ground, and dealing as he promised he would with topics that rhetoric had in his view for too long conceded to the philosophers. I am in fact talking about Quintilian’s use of emotive language in the discussion of oratory and particularly of style. I am not of course suggesting that he is doing something that literary critics in antiquity had never done before. But it is my impression that he does it far more systematically and on a larger scale. It is in itself an entirely natural tendency. At the most general level, it comes easily, for example, to call aspects of style one approves of ‘virtues’ and those one disapproves of ‘vices’.¹⁸ And, building on that, one could extend the use of moral language in striking directions. To take examples from a single chapter, 1.5: Quintilian starts off with the statement that oratio has three virtues, correctness, clarity, ornateness, and as many vices, the contraries of these three. The vices of barbarism and solecism are said to display foulness, foeditas (5). Those who make mistakes of accentuation are said to sin (peccant, 23). A speaker confused between conjunctions like an and aut has committed a ‘delict’, deliquerit (49), or, at the very least, confusions related to this count as error (47). It is very easy to go on beyond this sort of thing into more elaborate and extended metaphorical language. Narration that ¹⁷ See 5.10.24 5 and 40; 11.3.137 8. ¹⁸ For vices of various kinds see 1.1.5; 2.10.13; 2.12.1; 3.8.69; 4.1.13 and 71; 4.2.43; 10.1.18; 11.1.56; 11.3.91.

Quintilian the Moralist

185

goes to the opposite extreme from the dry and jejune may be said to ‘wanton’ (lascivire)¹⁹ with its far-fetched descriptions. Speeches that are ‘faulty’ (vitiosae) are also called ‘corrupt’ and can be analysed to display features that are ‘improper, obscure, tumid, abject, sordid, wanton, effeminate’—in a word, prava, ‘depraved’, the sort of thing that appealed to jurors who were themselves depraved (iudiciorum pravitate).²⁰ I return for a moment to the contrast of past and present: for this is in a way the essence of Quintilian’s moralizing criticism, both of style and of life. Just as people had behaved better in the past, so they had spoken better. Then the old orators hid their eloquence, so very different from the show-offs of today (4.1.9). It is now a ‘virtue’ of epigrams to be obscure and what Quintilian calls ‘headlong’ (7.1.44). It is now the faulty custom to bring too many short syllables together, with a sound like a child’s rattle (9.4.66); and equally now the deplorable custom to sing one’s speech (11.3.57). Most basic of all, perhaps, there was the split between the old style and the new—the split that so exercised Tacitus in the Dialogus. Quintilian describes it in 10.1.43 in terms that suggest that if he had to choose he would choose the older, though he is on the surface merely describing extreme positions: ‘some think that only the veteres²¹ are to be read, and judge that no others have natural eloquence and strength worthy of true men; others are pleased by this modern wantoning and foppery, everything aimed at giving pleasure to an unlettered mob.’ The moral attitudes so often revealed by Quintilian are clearly in evidence here. It is worth dwelling for a little on his ambivalent views on pleasure.²² This was a problem, like that of the emotions, raised by the philosophers. The Stoic ἀπάθεια, in a modified form, appealed to a certain puritanical instinct in the Romans; and the Epicurean goal of pleasure could easily be misrepresented as something more voluptuous than it really was. Accordingly, Quintilian makes one of the marks of his distinction between the old days and modern decadence the increased pursuit of pleasure for the audience. Then judges were more austere, and a speech was composed for its utility rather than for display; now pleasure has burst in even where a man is on trial at risk of fortune or even life (4.2.122). As for the speakers, some of them nowadays (for Quintilian uses the present tense) avoid arguments as being horrida et confragosa, wild and broken terrain, in favour of loci amoeniores; Quintilian characteristically compares them to the sailors of Ulysses who, beguiled by the taste of lotus or the song of the Sirens, preferred pleasure to safety—the safety in this case being that of the poor client, victory in whose interest is of less importance to ¹⁹ Quint. 2.4.3. Cf. for words from this root 2.5.22; 4.2.39; 9.4.6, 28, 108, 142; 11.1.56; 12.10.73. ²⁰ Quint. 2.5.10. For ‘corruption’ cf. 2.12.6; 10.1.125; 12.10.76. For ‘depravity’ cf. e.g. 4.2.39; 9.4.6; 10.1.19. ²¹ On whom see 8.5.33. ²² See 5.pr.1 (cf. 5.8.3).

186

Papers on Quintilian and Ancient Declamation

such speakers than the pursuit of the mirage of praise (5.8.1). On the other hand, Quintilian is unable to deny that the pleasure of the audience has its place in the persuasive process, alongside the emotions whose arousal he describes in such detail. For pleasure leads to belief: voluptate ad fidem ducitur.²³ The way had been clearly shown by Cicero, who for our author is virtually beyond criticism. Quintilian describes eloquently how the great orator’s treatment of the case of Gaius Cornelius caused the Roman people to express its admiration by applause as well as shouts of approval. ‘It was Cicero’s sublimity, splendour, brilliance and authority that drew that thunderous reception . . . I think that those present did not know what they were doing, and did not applaud of their own free will: they were as though bewitched, unaware of their surroundings, when they burst out into that expression of their pleasure’ (8.3.3–4). Quintilian’s epigram for this in another place (12.10.45) is: hoc ipso proderat, quod placebat—it was to the benefit of the client that Cicero gave pleasure. There was need then to distinguish between pleasures. Those produced by a Cicero were all right, others might not be. Quintilian in one place (1.12.18) classes a list of vulgar pleasures—games, dicing, and the like—as ‘unscholarly gratifications’, to be contrasted with, and neglected in favour of, geometry and music. Or the undesirable pleasure aroused by too fluent a prose rhythm could be dismissed as ‘the soft and voluptuous pleasure of melody’ (9.4.31). Style should be ‘agreeable—but, with a pleasure that is not unseemly but is linked with renown and dignity’ (8.pr.33). For Quintilian, as for Epicurus, some pleasures are better than others. An equally fluid concept was that of the natural. Quintilian sometimes polemicizes against opponents who, unimpressed by the need for so elaborate a training as he provided, asserted that all that was required to be an orator was to be born.²⁴ That was going too far. But natural remained a potent word of commendation, just as unnatural was a potent word of disapproval. In 2.5 we are told that language expressed secundum naturam is favoured as being straightforward and unaffected, whereas the admiration we give to things that are deviant and far-fetched is like the way in which some ‘prize bodies that are distorted and in some way monstrous over those that have lost nothing of the good things that we should all be granted by nature; again, those who are attracted by superficial appearances think that plucked and pumiced skin, hair waved and pinned, faces shining with colour that is not their own, are more beautiful than uncorrupted nature can provide, so that bodily beauty seems to arise out of faults in character’ (2.5.11–12). Here the topic of the naturally well-formed body is joined by the topic of beauty

²³ Quint. 4.2.119; cf. 5.14.35; 8.3.5. ²⁴ e.g. 9.4.3 (where they use the language of virility); 11.3.11.

Quintilian the Moralist

187

unaided by cosmetics, familiar to us from such poems as Propertius 1.2; and the connection between corrupt style and corrupt morals is strongly hinted at. There is an even more striking development of these themes at 5.12. Declaimers who have forgotten that their speeches are intended as training for real-life oratory are said to be depriving of their due nervi speeches that are composed for pleasure alone. That nervi is here to be taken in a sexual sense is made quite clear from the simile that follows. Such declaimers are like slavetraders who preserve the beauty of their boy-slaves by castrating them. ‘They think strength and muscle and beard and everything that nature has bestowed on the male is not beautiful enough, and they soften what would be strong if only it were allowed to be so; in the same way we’—Quintilian sardonically classes himself with these deplorable declaimers—‘we cover over the virile appearance of speech, its power to communicate robustly and sparely, with a soft stylistic skin, and forget about strength so long as there is smoothness and shine.’²⁵ But Quintilian is clear that a man is better than a eunuch, and a eunuch no more admirable if made so by a knife than he would be if born thus. Bad morals, Quintilian hopes, will not prevail so far as to make things good just because they are financially profitable. By the time Quintilian returns to eloquence, we are thoroughly prepared to have modern oratory called ‘libidinous’ and the audience that takes pleasure in it slaves of resupina voluptas (they lie back and enjoy it), and to have contrasted with it any indication of ‘a man who is masculine and uncorrupted, let alone gravis et sanctus’ (5.12.17–20). The intermingling of moral and literary judgement could hardly go further. Quintilian is using emotive language to make his literary points. In so doing he displays his own moral attitudes. He seems to be confident that he can appeal to similar attitudes in those for whom he writes; if they do approve, he is reinforcing those attitudes, if they do not, he is hopeful of instilling them. I do not need to dwell too long on other examples of these attitudes. They flow from or are related to those that we saw in the passage from 5.12. The admired figure is male, in the prime of life, heterosexual, and conventional. Deviants from that ideal are decried. It is men who give speeches, and they should in all respects behave like men (and not homosexual men either). Nowhere do such attitudes appear more clearly than in the chapter on rhythm, 9.4. To end a verse with four long syllables is praemolle, ultra-soft (65). When Thucydides (1.8.1) wrote the words ὑπὲρ ἥμισυ Κᾶρες ἐφάνησαν (one notices the Ciceronian clausula) he did not realize that he was employing ‘a very soft type of rhythm’ (78). Rather than this sort of thing, ‘I should prefer’—says Quintilian—‘the placing of words to be harsh and rigid to its being effeminate and without nervi’ (142). So too with the voice: we do not want it to be so restricted as to be

²⁵ Cf. 8.3.6. For gravis et sanctus below, cf. 1.8.2.

188

Papers on Quintilian and Ancient Declamation

as thin as that of ‘eunuchs, women or the sick’ (11.3.19). And a man has to be very careful of the height above the ground of the lower edge of his tunic: rather below the knee at the front, halfway down the hams at the back. For below that level is the mark of a woman, above it the mark of a centurion (11.3.138). It is difficult to know which Quintilian thought worse. It is well enough known that Seneca wrote a letter (114) underlining the interconnection of morals and style. Quintilian’s concurrence with that view has been less noticed. Seneca famously took Maecenas as an example of the point. It was not to be wondered at if a man who, deputizing for Augustus, could give the watchword for the day when not wearing a belt, should write sentences whose word order was so convoluted (6). Quintilian goes some way towards making the same point when he gives his own illustrations from Maecenas’ writings: such hyperbata are ‘vicious’ (the usual sub-moral language), and chosen specifically to gambol about and be wanton (exultent atque lasciviant); one of the examples is particularly discreditable, for it offends against the principle of τὸ πρέπον by ‘playing about’ on a serious subject (in re tristi ludit).²⁶ But the principle is laid down with all clarity in 11.1: ‘What one says generally displays one’s character and opens up the secrets of the heart;²⁷ and it is not without reason that Greeks have stated that every one speaks in the way in which he lives’ (30). And Quintilian thought the principle true of the listeners too: for in 11.3 he remarks that there are those who are allured by the pleasure of listening continually to things that tickle the ear—and that such people act thus ‘in line with the other faults in their lives’ (60). Quintilian sees morals and language going downhill in parallel. Commenting on the way in which words properly used by Sallust—dicta sancte et antique, he says characteristically— are now subjected to ridicule, he remarks that ‘we have lost perfectly good words thanks to our morals, and one has to give way even to vices when they are triumphant’ (8.3.45). Or again, recommending the early reading of the old Latin writers, he says that they will supply sanctitas . . . et ut sic dicam virilitas (words now very familiar to us), ‘for we have slipped into all manner of vicious pleasures in our manner of speaking too’ (1.8.9). It is not surprising that, in the sort of extended metaphorical passages found in 5.12 and the prologue to Book 8, one is hard put to it to distinguish, at times, between what is being said of life and what of style. Nor is it surprising that in such a passage as 2.4.16 pupils are described in terms that waver around between their moral and their stylistic progress: too early indulgence in extemporary effusion gives them, according to Quintilian, ‘a contempt for hard work, total shamelessness, the habit of speaking as badly as possible, practice in bad habits, and . . . an arrogant conceit in their own accomplishments’. The full justification, then, for Quintilian’s insistence on the vir bonus

²⁶ 9.4.28.

²⁷ Cf. 8.pr.20: cultus . . . muliebris . . . detegit . . . mentem.

Quintilian the Moralist

189

dicendi peritus now becomes clear. It is not merely that such a man will put over a more persuasive ἦθος, and be a more effective answer to philosophical critics of rhetoric. He will be, by virtue of his better way of life, ipso facto someone likely to be better in every aspect of style. In praising the vir bonus, and formulating his rhetorical rules around such a paragon, Quintilian is able to give added persuasiveness to his own doctrines; but he can also recommend to a wider public his own ideal of a purer older Rome, where people both spoke and lived better than they did in the present day. * * * The Institutio oratoria was not Quintilian’s first work. He had, not very long before, written a shorter book On the causes of the corruption of eloquence. It was a book that I should guess from what we know of it to have introduced some political dimension into its argument. It seems likely that Quintilian argued that, just as Demetrius of Phaleron had marked the decline of Greek eloquence after Demosthenes, so Cassius Severus, in the age of Augustus, marked the turning point of Latin oratory: and that both watersheds were not unconnected with political changes. In that sense, the argument may not have been very different from that of the Dialogus of Tacitus. But one is on safer ground in assuming that Quintilian, in the De causis, made the same sort of connection between the decline of eloquence and the decline of morals that I have been talking about in this paper. Corrumpere can have the same overtones of sexual, or at any rate moral, fault as other words habitually used by Quintilian of stylistic decadence. And it may be significant that the scanty allusions to the book in the Institutio include the over-wrought passage on the castration of slave-boys that I have already discussed.²⁸ Another concerns cacozelon, stylistic bad taste, and the familiar half-moral words appear: tumida, exultantia, compositione fracta.²⁹ Yet another points out that all hyperbole is a form of lying, though (as Quintilian says) not a type employed in order to deceive.³⁰ It looks as though the book De causis was the moral kernel of the Institutio: I mean that having identified in it the stylistic and moral faults of his age, Quintilian proceeded to a longer work in which he hoped to cure both kinds of fault. This will have been a preoccupation of Quintilian well before he became a crusty old man. He takes us back in 10.1 to an earlier day when Seneca was the only author young men were prepared to read and when Quintilian had won a reputation for being an enemy of his fellow-Spaniard, for ‘I was trying to recall to severer standards a type of speaking that was corrupt and broken down by every kind of vice’ (125). But now, in his last years, he brings his efforts to their ²⁸ Quint. 5.12.17 19 (the De causis is referred to at 23). ²⁹ Quint. 8.3.56 7 (the De causis is alluded to in 58). ³⁰ Quint. 8.6.74 (allusion to the De causis at 76).

190

Papers on Quintilian and Ancient Declamation

culmination. In his last chapter he considers what a retired orator should do to pass his last days. One suggestion is that he should ‘compose an art of rhetoric or give worthy expression to the finest precepts on how to live one’s life’ (12.11.4). In Quintilian’s eyes perhaps there was no real dichotomy here. The Institutio tries to do both, and so to bring back rhetoric and moral philosophy under the same umbrella. I do not pretend that my picture of Quintilian’s attitudes has been an altogether pretty one; certainly they are not what we nowadays call ‘politically correct’. But: Haec erant, Marcelle Vitori, quibus praecepta dicendi pro virili parte adiuvari posse per nos videbantur, quorum cognitio studiosis iuvenibus si non magnam utilitatem adferet, at certe, quod magis petimus, bonam voluntatem (12.11.31).

14 An Emendation in Calpurnius Flaccus The theme of the second declamation of Calpurnius Flaccus is Matrona Aethiopem peperit. Arguitur adulterii. In one of the excerpts,¹ the accuser is arguing that for a white woman with a white husband to produce a black child is certain proof of adultery, for individual races have fixed physical characteristics to distinguish them. I give the text as argued for by Watt:² Sua cuique genti etiam facies manet: rutili sunt Germaniae vultus et flava proce ritas; Hispaniae non eodem omnes colore tinguntur? Ex altera parte, qua convexus et deficiens mundus vicinum inficit orientem, hic effusiora corpora, illic collectiora nascuntur. Diversa sunt mortalium genera, nemo tamen est suo generi dissimilis. incolae suppl. Watt || inficit Schultingh: mittit codd. || hic Schultingh, favente Watt: illic codd.

Each race has its permanent appearance. Germany exhibits red faces and tall bodies capped by blond hair. Are not the peoples of Spain dyed with one and the same colour? In the other direction, where the sky, dipping down and about to come to an end,³ colours the East (i.e. the peoples of the East) which is so near to it, bodies are in some regions more sprawling, in others more compact. The races of men are different from each other; but no one fails to conform to the char acteristics of his own race.

As to inficit, Schultingh’s note reads in part: ‘Oriens vero Asiam Asiaticosve populos notat: hos inter et Indi: quos . . . tingit et inficit sol . . . ’[⁴] That is clearly a desirable sense in a context concerned with colour more crucially than size.⁵

[Classical Quarterly NS 49 (1999), 338 9] ¹ Calp. Decl. 2.6 11 Håkanson. ² Watt (1996), 123. ³ The earth is visualized as flat, and the sky (and in particular the sun) as descending and getting nearer to it as it approaches its eastern edge. Cf. to some extent Tac. Ger. 46.1; Thomson (1948), 329 with n. 2. I am grateful to Professor D. A. Russell for discussing this and other points with me. ⁴ [Schultingh ap. Burman (1720b), 794.] ⁵ In the West, the inhabitants of Germany and Spain are separately characterized partly by colour, partly by size. In the East, where two unspecified areas are given bodies of different sizes, a mention of their colour cannot be foregone. And only thus is point given to vicinum: see the

192

Papers on Quintilian and Ancient Declamation

But the verb should rather be mutat, which is nearer palaeographically and which preserves the rhythm.⁶

passages cited by Nisbet Hubbard (1970) on Hor. Carm. 1.22.21, esp. Servius on Aen. 4.481: Et dicta Aethiopia a colore populorum, quos solis vicinitas torret. ⁶ For this sense of muto see TLL s.v. 1724.71 8 and OLD² s.v. 11, adding (surely) Luc. 1.684 7, where editors adduce Catull. 11.7 8.

15 More Problems in Quintilian In 1970 I published as BICS Supplement 25 a monograph called Problems in Quintilian,¹ to accompany my Oxford Classical Text of the author published in the same year.² Recently, I have worked through Quintilian again in detail, discussing with Professor D. A. Russell his new text and translation, which will appear as a Loeb.³ The following notes supplement, or in some cases revise, what I wrote in 1970. Professor Russell has been kind enough to read and comment on them, but his assent is not to be assumed. I am also grateful to Professor W. S. Watt for allowing me to mention some new suggestions. Citations are from my own text. 4.3.2 Plerisque moris est, prolato rerum ordine, protinus utique in aliquem laetum ac plausibilem locum quam maxime possint favorabiliter excurrere. (2) Quod quidem natum ab ostentatione declamatoria iam in forum venit, postquam agere causas non ad utilitatem litigatorum sed ad patronorum iacta tionem repertum est, ne, si pressae illi qualis saepius desideratur narrationis gracilitati coniuncta argumentorum pugnacitas fuerit, dilatis diutius dicendi vo luptatibus oratio refrigescat.

Diutius is not in itself objectionable (‘put off still longer’ after the rigours of the narration), but one might perhaps expect the adverb to precede the participle (Russell indeed suggests transposition). I have considered dulcius, to be taken with dicendi. For dulciter with dicere cf. 4.2.62, 9.4.14, and 12.10.71. The comparative is found at 12.10.27. 5.7.26 Reliquae interrogandi sunt partes: . . .

Q. proceeds to discuss rules for interrogating witnesses. Corsi,⁴ not fudging the issue as others had done, translates ‘Rimangono da considerare le varie parti dell’interrogatorio’; but this seems to lack point. Read rather artes, soon to be [Bulletin of the Institute of Classical Studies 44 (2000), 167 77] ¹ Winterbottom (1970b). For editions and translations preceding 1970, see the ‘Index edi torum’ in my text, . . ² Winterbottom (1970a). ³ [See now Russell (2001).] ⁴ Calcante Corsi (1997), .789.

194

Papers on Quintilian and Ancient Declamation

picked up by Illae vero pessimae artes in § 32. Cf. also ‘Quint.’ Decl. min. 314.15: Erat, erat illic potentior testis: non quidem apud iudicem dicet nec oratorum interrogabitur artibus, sed loquetur ore vestro. 5.11.30 Scio quosdam inani diligentia per minutissimas ista partis secuisse, et esse aliquid minus simile, ut simia homini et [ut] marmora deformata prima manu, aliquid plus, ut illud ‘non ovum tam simile ovo’ . . .

See my discussion in Problems.⁵ I now think that we should read prima manu deformata , comparing, to some extent, Cic. Top. 69, where perfecta incohatis form a contrasting pair. I have changed the order to preserve the rhythm as well as to make it clear that prima manu is to be taken with deformata. Alternatively (Watt) deformata prima manu . 5.12.14 Quaesitum etiam potentissima argumenta primone ponenda sint loco, ut occupent animos, an summo, ut inde dimittant, an partita primo summoque, ut Homerica dispositione in medio sint infirma †aut animis† crescant.

I discuss this crux in Problems.⁶ I now think that animis is not impossible (despite animos just before, an objection I did not mention), if we assume that aut has been collapsed from ac velut. ‘animi can hardly be attributed to arguments’, I wrote before. But perhaps they can in a simile. And there is no need to keep too closely to the Homeric original after infirma. For ac velut, see e.g. 3.3.7, 12.2.9 (and velut is supported by a related passage at 4.2.102). For animis crescere compare phrases like Liv. 38.54.1: Morte Africani crevere inimicorum animi. 6.1.40 Alius imaginem mariti pro rea proferre magni putavit, at ea risum saepius fecit. Nam et ii quorum officii erat ut traderent eam, ignari qui esset epilogus, quotiens respexisset patronus offerebant palam, . . .

The unfortunate aides kept producing the picture whenever the advocate looked their way. Their problem was not that they ‘had no idea of the nature of a peroration’,⁷ but that they could not time their act properly because they did not know what an epilogue was at all (or the meaning of the word ‘epilogue’). The Latin for that is quid esset epilogus. 6.2.6 Nam cum irasci favere odisse misereri coeperunt, agi iam rem suam existimant, et, sicut amantes de forma iudicare non possunt quia sensum ocu lorum praecipit animus, ita omnem veritatis inquirendae rationem iudex omittit occupatus adfectibus: . . .

I printed praecipit understanding it to mean ‘anticipates’. If that is not acceptable, praecurrit may be considered as an alternative (Q. uses the verb at 1.3.3). The sense is in any case the same. The mind, taken over, like the ⁵ Winterbottom (1970b), 95 6.

⁶ Ibid., 97.

⁷ Butler (1920 2).

More Problems in Quintilian

195

judge, by emotion, ‘outruns’ the eyes by ignoring their evidence (whereas normally the eyes transmit the emotion to the mind; cf. 11.3.14, cited below on 11.2.4). Cf. Cic. Fat. 44: confitentes non fieri adsensiones sine praecursione visorum (also Top. 59). 7.3.18 Non negaverim . . . esse id tum elegans, tum etiam fortissimum, si modo erit illa (sc. finitio) inexpugnabilis.

The PHI disk confirms that Quintilian always writes cum . . . tum etiam . . . in such collocations,⁸ with the exception of this passage and 9.3.28: tum gratiam tum etiam vires accommodat. Both passages are only transmitted by A, and both should be emended. I mention cum as a variant (from an early edition) in the latter place. 7.3.22 Actor ergo ita finiet: ‘Sacrilegium est surripere aliquid sacri’.

The main lines of the supplement (that of P, modified by Halm) are of course correct, but it remains to tidy up a small point of wording. A few lines later, in 24, Q. writes: adiciendum enim ‘aut ex sacro’; and there is no reason why the preposition should be different here. Obrecht in secundis curis (according to Burman) read e sacro, but Q. normally (62 to 5) has ex before s. 7.3.27 (26) Circa propria ac differentia magna subtilitas, ut cum quaeritur an ad dictus, quem lex servire donec solverit iubet, servus sit. Altera pars finit ita: ‘Servus est qui est iure in servitute’, altera: ‘Qui in servitute est eo iure quo servus’, aut, ut antiqui dixerunt, ‘Qui servitutem servit’. Quae finitio, etiam si distat aliquo, nisi tamen propriis et differentibus adiuvatur, inanis est. (27) Dicet enim adversarius servire eum servitutem aut eo iure quo servum. Videamus ergo propria et differentia . . .

The sentence that starts § 27 raises a problem of wording; Spalding remedied it by adding esse after eo, though we can perhaps understand servire. Worse, a problem of sense. Despite enim, the sentence does not explain what precedes. It seems merely to repeat what is said about the definition of the opponent above. It should be deleted as a gloss, perhaps added by someone (needlessly) concerned to explain distat aliquo. Ergo then directly takes up the need for propria et differentia expressed in the sentence ending inanis est. 7.6.5 6 Unum est in quo ipso patet semper id servari non posse: ‘Liberi parentis alant aut vinciantur’: non enim alligabitur infans. Hinc erit ad alia transitus et divisio: ‘Num quisquis non aluerit? Num hic?’ (6) †Propter hoc† quidam tale genus controversiarum in quo nullum argumentum est quod ex lege ipsa peti possit, sed de eo tantum de quo lis est quaerendum est.

A, the only primary manuscript available here, gives propter, changed by its corrector to proponunt. As often, it is hard to be sure if the corrector is ⁸ There are 17 instances (starting at 2.5.6). In no case (except 9.3.28) do I signal a variant.

196

Papers on Quintilian and Ancient Declamation

emending suo Marte or recording a reading from another source. In either case, the correct word for the setting of a theme in declamation is ponere (e.g. 3.8.53), not proponere (5.10.110 and 7.1.41 are different). But all is still not well, especially with hoc. And one perhaps misses some connection with what precedes. Spalding therefore suggested Praeter hoc proponunt. But he is perhaps too dismissive of Obrecht’s⁹ suggestion (followed by Halm) that propter hoc goes with the previous sentence; though we might there precede it with a third num. The following sentence may then start Ponunt quidam . . . ; but the verb could be added later in the clause. 8.3.55 At vitium erit quotiens otiosum fuerit et supererit, non cum * adicietur.

Q. is discussing pleonasm. I previously¹⁰ suggested supplementing with (‘[a]t minimum’) proderit quod. But the tenses vary oddly, and I now wonder about reading: quotiens otiosum fuerit quod supererit, non cum adicietur. If that is the truth, it would be in itself an example of pleonasm, and Q. would be teaching by example. 8.3.85 Supprimitur vox, ut fecit pro Ligario Cicero: ‘Quod si in tanta fortuna bonitas tanta non esset quam tu per te, per te, inquam, optines: intellego quid loquar.’ Tacuit enim illud, quod nihilo minus accipimus, non deesse homines qui ad crudelitatem eum inpellant.

Quintilian cites and explicates Lig. 15. Intellego quid loquar (also in our Cicero manuscripts) seems an odd remark: orators should understand what they mean, without drawing attention to their prowess. Did Cicero write intellegis? 8.4.13 Est tamen quamquam diversarum rerum quaedam vicinia: repetam itaque hic quoque idem quo sum illic usus exemplum, sed non in eundem usum. (13) Nam hoc mihi ostendendum est, augendi gratia . . .

‘For what I have now to demonstrate is . . . ’¹¹ That is certainly the emphasis. Read then hic for hoc, or if it is wished to avoid repetition, hoc . Watt suggests that hoc should be replaced by nunc. 8.5.1 Sententiam veteres quod animo sensissent vocaverunt. Id cum est apud oratores frequentissimum, tum etiam in usu cotidiano quasdam reliquias habet: . . .

Quintilian naturally contrasts orators with poets (e.g. 1.5.52, 8.6.19 and 27), and it may be that then the word may in effect cover other prose writers as well as speakers. Note in particular how at 8.6.20 prorsa picks up oratoribus, and in orando is immediately followed by an example from Livy. This Quintilianic ⁹ Cf. Sen. Con. 1.7, where the same law is set. In the discussion of the division at § 11, Seneca tells us that most declaimers raised the point an . . . omnis pater a filio alendus sit (= num quisquis). Latro, after dealing with that, made the next question: an hic alendus sit (= num hic), with two sub divisions, the first of them an alendus sit quod (qui Boot) . . . (= (num) propter hoc). ¹⁰ Winterbottom (1970b), 140. ¹¹ Butler (1920 2), with my italics.

More Problems in Quintilian

197

usage of both noun and verb should be added to the lexica. There is thus no need to emend to the wider auctores (cf. OLD² s.v. 8–9a) in our passage. 8.6.1 Circa quem (sc. tropum) inexplicabilis et grammaticis inter ipsos et philo sophis pugna est quae sint genera . . .

‘ . . . teachers of literature, who have quarrelled no less violently with the philosophers than among themelves . . . ’¹² That would require the deletion of the first et and the addition of cum after the second. Better, take inter ipsos (for which cf. 10.2.15) with both nouns: grammarians and philosophers alike have had internecine quarrels about this weighty matter. 8.6.37 Nos quis ferat si Verrem ‘suem’ aut Aelium ‘doctum’ nominemus?

This example of transumptio was not understood until the discovery of the Anecdota Ecksteinii, from which Halm supplied Catum. The omission would be better explained, and the point clearer, if we wrote aut Aelium Doctum’ nominemus. 8.6.43 Sunt autem quibus non videatur hic (sc. tropos, i.e. epitheton) omnino tropos quia nihil vertat, nec est semper, sed cum id est adpositum, si a proprio diviseris, per se significat et facit antonomasian. Nam si dicas ‘ille qui Numantiam Carthaginem evertit’, antonomasia est, si adieceris ‘Scipio’ adpositum: †non potest ergo esse iunctum†.

The final sentence has been much emended. Recently, Schindel has advocated retention of the transmitted wording, translating ‘verbunden mit seinem proprium kann das epitheton in der Tat kein tropus sein’.¹³ This asks a lot of a reader, but it may be what was originally meant by a glossator who was trying (unnecessarily) to summarize the point of the passage. 9.3.89 (88) Quaedam verborum figurae paulum figuris sententiarum decli nantur, ut dubitatio. Nam cum est in re, priori parti adsignanda est, cum in verbo, sequenti: . . . (89) Item correctionis eadem ratio est: nam quod illic dubitat, hic emendat.

The last phrase is suspicious. It does not explain why correction, like dubitation, is sometimes a figure of speech, sometimes a figure of thought. And one would like to know what the subject of the verbs is supposed to be. Halm countered the second objection (though not the first) by suggesting dubitatur . . . emendatur. He might have compared 5.11.3: Hoc in oratione fieri non potest, sed quod illic interrogatur, hic fere sumitur. It is interesting that critics have been tempted to delete that sentence. Perhaps this one should go too.

¹² Butler (1920 2).

¹³ Schindel (1987), 135 [= 29] n. 113.

198

Papers on Quintilian and Ancient Declamation

9.4.48 Proinde alia [ad] dimensio versuum: . . .

The exact wording is uncertain, but the general sense is clear: verse, unlike prose rhythm, does not allow substitution of e.g. anapaest for dactyl. The problem is how proinde can link this to the discussion of prose that has gone before (so Spalding: ‘Proinde sane nihili est, neque enim ulla argumentationis consecutio’). Emendation seems called for. I suggest prorsus, ‘quite different’ (as indeed Butler translates; PHI, it is true, does not register this collocation). Alternatively profecto (Watt). Holmes ingeniously suggests re-punctuating: Proinde alia. Ad dimensionem (so A) versuum pro . . . ¹⁴ But as Russell points out, one would expect alii (sc. pedes); and what follows seems to require supplement. 9.4.78 Nam et initium (of Plato’s Timaeus) hexametri statim invenias, et ana creontion protinus colon efficies, et si velis trimetron . . .

There seems no reason for the variation between subjunctive and future. Read efficias (as editors till Radermacher), or invenies. The condition si velis rather suggests the subjunctive: such identifications of metrical feet are optional. 9.4.97 Spondius quoque,¹⁵ quo plurimum est Demosthenes usus, non eodem semper [prae] se habet: optime praecedet cum creticus . . . (98) Potest, etiam si minus bene, praeponi anapaestos . . .

Watt suggested non eundem (sc. pedem) semper prae se habet.¹⁶ This is clearly preferable palaeographically, but the indicative verb seems out of place (it is obvious that spondees are not always preceded by the same foot): this should be precept, not statement (notice praecedet and recte praeponitur in 99; also 101: erit; 102: bene praeponitur). Read (accepting Watt’s view) habebit (or perhaps habeat). 9.4.140 Itaque †tragoediae ubi recesset adfectatus etiam tumor rerum et† spon diis atque iambis maxime continetur: . . .

Discussed in Problems,¹⁷ where I suggested tragoedia, ubi decet vel adfectatus etiam tumor rerum [et] . . . Russell suggests successit, ‘where even swollen material has been successful’, and this is certainly nearer to the transmission than decet vel. Alternatively, with the same sense, bene cessit. See TLL s.v. cedo 732.43 ff.; OLD² s.v. cedo 7a, esp. Sen. Con. 7.1.27: At Vergilio imitationem bene cessisse; Q. 10.2.16: cum iis felicissime cessit imitatio. These datives also suggest the possibility (preferable palaeographically) of tragoedia, cui. For the

¹⁴ Holmes (1997 8), 66 7. ¹⁵ Rather illogical. I take it that Q. means: ‘There is also the spondee (cf. above: Est et doch mius), and (but) it . . . ’ ¹⁶ Watt (1993), 318. ¹⁷ Winterbottom (1970b), 186 7.

More Problems in Quintilian

199

general sense cf. ‘Longin.’ Subl. 3.1: tragedy a genre ‘naturally magniloquent and tolerant of bombast’ (trans. Russell). 9.4.140 At ille comicus aeque senarius, quem trochaicum vocant, pluribus cho reis, qui trochaei ab aliis dicuntur, pyrrhichiisque decurrit, . . .

Quintilian is talking about the comic senarius, which like the tragic line just discussed (aeque) has six feet. This is not known to have been called ‘trochaic’, nor is there any reason why it should be (I do not know why Spalding says the appellation is ‘per se a verisimilitudine non abhorrens’); hence Bonnell’s deletion of that clause. Next, the appearance of chorei, i.e. trochees, perplexes. Hence Spalding’s pluribus trochaeis, qui tribrachi ab aliis dicuntur, . . . , that is, tribrachs, the proper partner for pyrrhics. Spalding appeals to 9.4.82: Tres breves trochaeum, quem tribrachyn dici volunt qui choreo trochaei nomen imponunt (cf. also 80: huic contrarium e longa et brevi choreum, non ut alii trochaeum nominemus). But that very passage informs us that there were those who called ‘chorei’ trochees. That seems to suggest that our passage is correctly worded. But it does not, of course, explain the oddity of trochees appearing here at all. Quintilian may be thoroughly muddled. If not, we should perhaps read pluribus trochaeis, deleting qui trochaei ab aliis dicuntur as the gloss of someone recalling §§ 80 and 82 when confronted with the corrupt choreis. The glossator might be the same person responsible for quem trochaicum vocant just above. 10.2.27 Illuc intendenda mens . . . : quid agatur prohoemio, quae ratio et quam varia narrandi, quae vis probandi ac refellendi, quanta in adfectibus omnis ge neris movendis scientia, quamque laus ipsa popularis utilitatis gratia adsumpta, quae tum est pulcherrima cum sequitur, non cum arcessitur.

We can happily understand sit with ratio, vis, and scientia, but hardly with laus.¹⁸ An explicit subjunctive is needed, such as valeat or prosit (which should follow quamque, rhythm marking a pause at popularis): ‘how useful popular acclaim is, acclaim that is exploited for our own ends . . . ’ It may be that this would necessitate the further change to quantumque.¹⁹ 10.5.8 Ipsa denique utilissima est exercitationi difficultas.

The difficulty is useful to the exercised, not to the exercise (of paraphrase). We should consider exercitationis.

¹⁸ Peterson comments: ‘adsumpta (sit): “how popular applause itself has been worked in”, made useful for winning the case.’ Russell suggests adding sit. ¹⁹ Cf. e.g. 1.12.2: non satis perspiciunt quantum natura humani ingenii valeat. Observe, however, 12.11.10: quantam rem petant quamque nullus sit . . . labor recusandus, where, as here, quamque takes up quant . The Renaissance correction quantaque does not solve the problem, for laus is out of line with the preceding scientia.

200

Papers on Quintilian and Ancient Declamation

11.1.14 Minora vero quaeque sunt ex mediis plerumque sunt talia ut aliis sint concedenda, aliis non sint, aut pro persona tempore loco causa magis ac minus vel excusata debeant videri vel reprehendenda.

The variation between excusata and reprehendenda seems insupportable. Read either excusanda (so old editors, and preferable palaeographically) or reprehensa. 11.1.40 Quam dignae Milonis in peroratione ipsa voces eo viro qui pro re publica seditiosum civem totiens compescuisset quique insidias virtute superasset!

There is no obvious reason to emphasize peroratione by the addition of ipsa. On the other hand, voces, which should refer to the famous prosopopoeiae in Mil. 93–4, 98, needs some support. The solution is less certain than the problem. Perhaps dignae Milonis in peroratione [ipsa] voces, ‘the well-known words of Milo in (Cicero’s) peroration’.²⁰ Cf. 12.6.4: totus ille Ciceronis pro Sexto Roscio locus. Watt suggests reading illa for ipsa. 11.1.71 Haec est profecto ratio et certissimum praeceptorum genus illius viri (sc. Ciceronis) observatio, ut, cum aliquid detrahere salva gratia velis, concedas alia omnia: . . .

Watt put et . . . observatio in parentheses.²¹ Less dependent on modern typographical convention²² would be something like ratio et, (ratio, ut est already Kiderlin)²³ certissimum praeceptorum genus, illius viri observatio: ‘The following is surely the (right)²⁴ method, and—what constitutes the surest kind of precept—the practice of Cicero, namely that . . . ’. Id might be dispensed with (Russell). 11.1.88 (87) In quibus omnibus commune remedium est ut . . . reprehensa [alia] laude compenses: (88) sic cupidos dicas: sed non mirum, quod pericu lorum ac sanguinis maiora sibi deberi praemia putent; eosdem petulantes: sed hoc fieri quod bellis magis quam paci consuerint.

Sic is Spalding’s conjecture for si. Si itself is perhaps supported by si asperum dicas in 90, but we might also consider deletion. In that case, as Russell points out, milites should precede cupidos; cf. the word order of the next sentence: Libertinis detrahenda est auctoritas: licet iis testimonium reddere industriae . . . , where si is not inserted. Also possible is si dicas (cf. 6.1.4).

²⁰ Cf. 6.1.27: preces non dat Miloni . . . , accommodavit tamen ei [verba] convenientis etiam forti viro conquestiones, making exactly the same point. If it were not for this parallel one might suppose that the reference is rather to the remarks attributed to Milo in 104, and that peroratione ipsa means ‘right at the end of the speech’ (though I do not know a parallel for such a use). Even in that case, I should wish to add illae. ²¹ Watt (1988), 157. ²² An objection also to Spalding’s et, certissimum praeceptorum genus, illius . . . ²³ Kiderlin (1889), 83. ²⁴ Kroll added vera before ratio, Watt recta.

More Problems in Quintilian

201

11.2.4 Non arbitror autem mihi in hoc inmorandum, quid sit quod memoriam faciat, quamquam plerique inprimi quaedam vestigia animo, velut in ceris anu lorum signa serventur, existimant. Neque ero tam credulus ut †quam abitu tardiorem firmioremque memoriam fieri et actem quoque ad animum pertire†.

Q. says he will not dwell on the way memory works. He mentions one theory, that traces are left on the mind, like the mark of a seal in wax (cf. Cic. Tusc. 1.61, part of a passage much in Q.’s mind here). What follows is doubtful in the extreme, but, on the assumption that Q. is not introducing a new theory but casting doubt on the one already mentioned, I suggest ut putem attritu tardiorem infirmioremque memoriam fieri, et tactum quoque ad animum pervenire: ‘I am not so credulous as to think that memory becomes slower and weaker because it is being physically worn away, and that touch too (i.e. touch as well as sound and sight) gets through as far as the mind.’ Q.’s objection starts from the empirical fact that memory often becomes impaired (e.g. in old age). How can this be if it is a series of traces in something analogous to wax? That would mean that the traces were being somehow worn away. But how could that be, when the mind cannot be touched (as wax can be)? In another passage (11.3.14) Q. says that the eyes and the ears are the two senses per quos . . . omnis ad animum penetrat adfectus.²⁵ Penetrat there might suggest penetrare here, but the rhythm would be inferior to that given by pertinere (an old conjecture) or the more graphic pervenire (Russell). Infirmus is used of memory in 11.2.49. Attritus is not attested in Q., but cf. 10.1.27: velut attrita cotidiano actu forensi ingenia; the noun is found several times in the Younger Seneca. 11.2.29 Non est inutile iis quae difficilius haereant aliquas adponere notas, qua rum recordatio commoneat et quasi excitet memoriam: (29) nemo enim fere tam infelix ut quod cuique loco signum destinaverit nesciat. At si erit tardus ad hoc, . . .

Kiderlin suggested the addition of erit after fere.²⁶ I now think the supplement correct—est perhaps could be understood, but not the (necessary) future— though the word should perhaps follow infelix (cf. 7.1.21: nemo tam demens fuerit ut . . . ). Alternatively (Russell) erit could replace fere. 11.3.16 Nam opus est omnibus, sicut non oris modo suavitate sed narium quo que, per quas quod superest vocis egeritur, dulcis esse †tamen† non exprobrans sonus.

²⁵ This passage (where Q. is talking specifically of emotions) should probably not be exploited to explain quoque in our passage, for it could easily be argued that touch metaphorically ‘gets through to the mind’ as much as other senses (most easily of all, according to Lucr. 5.102 3; cf. Hollis (1990) [= (2009²), 293 4] on Call. Hecal. fr. 109). But if I am right, Q. is merely trading on the obvious fact that the mind cannot be literally ‘touched’ by anything. ²⁶ Kiderlin (1890), 476.

202

Papers on Quintilian and Ancient Declamation

Murgia suggests: dulcis est et non exprobratus sonus.²⁷ I should prefer something that gives a precept (like Halm’s dulcis sit). And if exprobrans is to be changed, why not rather to exprobrandus? There is no reason to call the sound ‘free from (past) reproach’, but it might be ‘irreproachable’. 11.3.27 Ita, si dicendum in sole aut ventoso umido calido die fuerit, reos desere mus? Nam crudum quidem aut saturum aut ebrium aut eiecto modo vomitu, quae cavenda quidam monent, declamare neminem qui sit mentis compos puto.

The first sentence clearly concerns the courtroom. In the second we abruptly move to the schoolroom. Unless Quintilian is being very careless, he should mark the change in some way: perhaps by writing vel before declamare. It is crazy for those with a stomach problem even to declaim, let alone take part in a real case (where the outcome matters). 11.3.43 Nam prima est observatio recte pronuntiandi aequalitas . . . Secunda va rietas: quod solum est pronuntiatio.

Both Shackleton Bailey (qua salem habet pronuntiatio)²⁸ and Watt (e.g. solum est)²⁹ have felt that the last phrase, strangely phrased and untrue, cannot stand. Neither change seems satisfactory, and rather than rewrite we should perhaps delete the words (the Idiot Glossator³⁰ seems to be at work). 11.3.92 Est autem gestus ille maxime communis, quo medius digitus in pollicem contrahitur explicitis tribus, et principiis utilis cum leni in utramque partem motu modice prolatus, simul capite atque umeris sensim ad id quo manus feratur obsecundantibus, et in narrando certus, sed tum paulo productior, et in expro brando et coarguendo acer atque instans: . . .

This all-purpose gesture is said to be ‘useful’ in the proem, certus in the narration, and urgent in invective. Spalding glosses gestus certus by ‘qui narrantem nihil dubitare ostendat de iis, quae referat’. That seems a lot to get out of the word, and one might consider gratus, ‘attractive’. Cf. 6.3.94: Est gratus iocus qui minus exprobrat quam potest, which supports a contrast in our passage between gratus and in exprobrando . . . acer. Alternatively, Russell suggests aptus (cf. §§ 95 and 96), or, even more attractively, certe aptus. 11.3.125 Procursio oportuna brevis moderata rara conveniet: . . .

The pile up of four adjectives is suspicious. If one is to go it should be rara, not happily chosen after oportuna (if it were opportune, one might do it a lot), and conceivably the result of dittography of -rata.

²⁷ Murgia (1991), 210. ²⁸ Shackleton Bailey (1983), 228 9. ²⁹ Watt (1988), 157 8. ³⁰ Well seen at work when, on Aut enim expugnatur intentio aut adsumptio aut conclusio, nonnumquam omnia (5.14.20), he comments: Sed omnia haec tria sunt.

More Problems in Quintilian

203

11.3.144 Sed haec amictus observatio dum incipimus: procedente vero actu, iam paene ab initio narrationis, sinus ab umero recte velut sponte delabitur, et cum ad argumenta ac locos ventum est reicere a sinistro togam, deicere etiam, si haereat, sinum conveniet. (145) Laeva a faucibus ac summo pectore abducere licet: . . .

It would be normal to write ventum erit before the future conveniet. But the future verb is itself out of line with delabitur. Read therefore convenit?³¹ Rhythm supports (conversely it supports conveniet at 125, discussed above), and the present is found similarly e.g. at 161. 11.3.169 Est his diversa vox et paene extra organum, cui Graeci nomen amari tudinis dederunt, super modum ac paene naturam vocis humanae acerba: . . .

The PHI disk tells us that the locution super modum appears only seven times in Classical Latin: three times in Festus, once in Gellius, once in the Vulgate, and twice in Quintilian—once here, one at 12.10.39, cited below. Old editors were no doubt right to bring the two Quintilian examples into line with the normal supra modum (eight times, without variant, in our author, four times, as in 12.10.39, with tollo or cognates); for supra naturam cf. 9.4.19. 12.1.12 Bonus quidem et dicet saepius vera atque honesta. (12) Sed etiam si quando aliquo ductus officio (quod accidere, ut mox docebimus, potest) falso haec adfirmare conabitur, maiore cum fide necesse est audiatur.

‘[I].e. maintains, but untruly, that something is verum atque honestum’ (Austin on haec adfirmare). If so, the Latin is very loose. Judging by what has preceded, haec should mean vera atque honesta, not vera atque honesta esse. I am not sure how to mend matters. Perhaps falsa ac adfirmare, corresponding to dicet vera atque honesta. Alternatively, as Russell suggests, e.g. falso (falsa?) aliqua adfirmare (cf. 11.3.67), or, on the same lines, si quando deleting haec (Watt). 12.2.17 Sed ille vir bonus, qui haec non vocibus tantum sibi nota atque nominibus aurium tenus in usum linguae perceperit, sed qui virtutes ipsas mente complexus ita sentiet, nec in cogitando laborabit sed quod sciet vere dicet.

Austin ad loc.: ‘The sentence is clumsily constructed (Sed ille . . . sed qui . . . sed quod . . . ) . . . The main clause begins at nec (= ne . . . quidem).’ It seems more natural to take nec as a connective. In that case we should add erit after complexus. The man who has embraced virtues in his mind will feel like that (i.e. virtuous), and will not . . . but . . . Alternatively (Russell) read laborabit et (commended by Spalding), deleting qui before virtutes. 12.10.53 (52) Quod si mihi des consilium iudicum sapientium, perquam multa recidam ex orationibus non Ciceronis modo sed etiam eius qui est strictior multo,

³¹ Kiderlin (1895), 235 suggested convenit, et laeva.

204

Papers on Quintilian and Ancient Declamation

Demosthenis . . . (53) Cum vero iudex detur aut populus aut ex populo laturique sententiam indocti saepius atque interim rustici, omnia quae ad optinendum quod intendimus prodesse credemus adhibenda sunt . . .

Naturally Roman courts did not consist of wise men, but nor did they regularly consist of ‘the people’ or ‘those drawn from the people’. Indeed, Cicero, mentioned before and after (54) this passage, usually addressed panels drawn from the higher ordines. It is therefore very odd that Quintilian should imply that judges were always low-class, and often or sometimes illiterate boors into the bargain.³² Unless Quintilian is exaggerating, or thinking of Athenian conditions (Demosthenes too is named before and after our passage), we should consider emending to datur.³³

³² That they could often be country folk is confirmed by 4.2.45: iudicem rura plerumque in de curias mittant. ³³ I append some marginalia and corrections to Winterbottom (1970b), cited by page: 12 In 8.3.21 A may be right to omit in; see my n. on ‘Quint.’ Decl. min. 335.5. 20 Note the sophisticated and just remarks of Ekkehard IV on the dangers of emending Julian of Toledo ‘ad solitum stilum’ cited by J. N. Hillgarth in CCSL 115. n. 2. 23 The conjecture about Montpellier MS H 126 made at n. 6 is wrong. 57 Housman (1972), 943 argued that at 10.1.98 the correct form of the genitive of (the tribrach) Varius is Varii. That may be so; but Housman was unwise to point to the spellings Spurii (5.9.13) and Marii (5.11.15) in A. B, whose spelling is far more reliable, has the shorter forms at those places. 101 5.14.31: Those who do not like ieiuni may care to consider tenues; cf. Lact. Inst. 3.1.6: tenui fonte; Tert. Spec. 7.4: rivulus tenuis ex suo fonte. 101 2 6.pr.13: Cf. (up to a point) Sen. Tro. 955: vivit in poenas Phrygum. 189 10.1.27: Libertate (despite its repetition below) can perhaps be defended by 10.5.15: dialogorum libertate gestiendum (‘the easy freedom of dialogue’: Peterson). 201 11.2.10: For cogitatio = ‘something thought out in advance’ cf. 12.9.20. 216 12.11.11: Cf. also [Quint.] Decl. mai. 17.6 (337.17 18 Håkanson): Abstulisti quidem mihi partem, ut exclamarem hoc loco . . . (where perhaps read partem or partem).

16 Ennodius, Dictio 21 Ennodius left twenty-eight Dictiones, of which Jacques Sirmond, in his edition of 1611, classified ten as controversiae.¹ Only one, no. 21, is a fully worked out exercise, and it has the further interest of purporting to be a reply to a declamation going under the name of Quintilian. The purpose of the present study is to provide a translation of the Dictio (the first in English), equipped with a short commentary on its text and language, and briefly to discuss its technique and merits.

TRANSLATION

Theme 1. L: C       . A man had two sons, a frugal one and a wastrel. Both fell into the hands of pirates. They wrote to their father asking to be ransomed. He sold up his property and sought out the pirates. They gave him the choice of ransoming whichever of the two he chose, for he had brought too little money (to ransom both). He chose to ransom the wastrel, because he was ill. The son died on his way home. The frugal son escaped from the pirates. His father requests support from him. He speaks against. The case is brought in opposition to Quintilian. For he defends the father, I the son. I ask the reader not to think this arrogant of me. [From B. J. Schröder and J. P. Schröder (eds.), Studium declamatorium. Untersuchungen zu Schulübungen und Prunkreden von der Antike bis zur Neuzeit (Munich Leipzig, 2003), 275 88] ¹ The most recent edition is still that of Vogel (1885); our dictio is printed on pp. 260 5. All references to Ennodius are given by page and line of Vogel’s edition. [For the paragraph marks (§) I have frequently added in my translation see below, p. 217.] The pseudo Quintilian Declamationes maiores (MD) are cited by page and line from the edition of Håkanson (1982) (H.). I should like to thank Bianca Schröder very warmly for helping me to grapple with Ennodius.

206

Papers on Quintilian and Ancient Declamation

Preface 2. Surely it cannot be right to speak against Quintilian—except in defence of the truth. Or is it that a man forgets himself in his eagerness to talk, when he is putting a just cause into words? Cleverness is appropriate to those who aim to deceive; their style, daubing itself with the paintbrush of mendacity and praiseworthy only word-deep, takes on a beauty that is foreign to it. Let the claims of equity be put forward without any help from rhetoric. My preference is for the display that will arise naturally from the defence of uprightness. Against so eloquent a man let there advance an open nature fortified by the support of the gods. That I venture to speak results from my confidence in my side’s cause. 3. It is for you, noble sirs, merely to listen with good will, guarding your ears and your minds to prevent the higher style that confronts us from believing that victory is in its grasp just because of its superior power of speech. What else can help us against its assaults but to escort justice home after it has been abroad for so long?

Proem 4. Most distinguished judges, I should rejoice to have escaped the filth of the prison and the pirates’ chains, were it not that on my return I find my father standing in for the privateers. § He asks sustenance from me after having failed to free me from my enemies whether by gold or by tears. He thinks that he should benefit from the return of someone whom he did not ransom when he had the chance. By your actions you broke the laws of Nature; it is in vain then that you rely on the law to beg for food. 5. The law decrees the provision of a living to fathers. This solemn relationship does not lie merely in a name. Even the pirate may say if you are really a father. § You often make out that I was better than the dead man when you speak; but you never show me to be that when you make a judgement. § Observe, judges, the guilty road followed by my dying brother: the son who was preferred came closer to the character of his father. As for me, the ill omen under which I led my life amid actions of quite a different sort from those was made obvious when my father of his own free will left me to the pirates. 6. § He imagines that he is bringing convincing arguments to bear on his offspring when he proclaims that the son who was blameworthy and near to burial deserved readier favour. Let your verdict, judges, proclaim what I owe (to my father)! Meanwhile, it is quite apparent that I was put second even to the ashes of a vicious man. § He does not blush to use a father’s authority to ask anything at all from a son in whose case he did not wish to act up to what he is called. 7. § If only the pirates’ chains had preserved me in the health that would have enabled me to feed him! By offering sustenance I do not owe, I should make my father a captive—of

Ennodius, Dictio 21

207

shame; the unjust judge would groan at the unfairness of his own sentence. For food that is proffered by someone for whom you have done nothing good fails to sustain: a belly unaware of any service previously conferred is hungry even when it is fed full. 8. I cannot imagine what a man thinks he has left to hope for from his two sons, when he has buried one of them and allowed the other to be held by the lasting tyranny of those who carried him off. If he has the face to do it, let him receive whatever limbs exhausted with tortures have to offer in the way of groans. I do not know what an old man can demand of a cripple. For the present, as I remark, we both need feeding. § Why do you make play of your straitened circumstances? The man who wished to ransom only one of his sons sold up his property in such a way that the price would not be enough to pay for both. 9. § Who could portray my unhappiness, my misery? After my experiences with the Cilician pirates, I am told that even the liberty I regained is subject to restraint. I ask you, judges: if even in my own country I am going to be proclaimed to be a debtor, from what have I escaped? § My revered begetter added the point that I am grieved that my brother was released from his bonds. As if the ties of nature so enslave us that the joy of my brother’s rescue could compensate for the fact that I had to stay behind! It is when we have been released that we owe love to those who share in our flesh and blood. 10. Believe me, for what I say is perfectly obvious: joy for another cannot enter the heart of someone who is afraid for himself. § After the manner of the crowd of whom we disapprove, my father cries that the man from whom something is asked is somehow the richer for it. This is the trick of a specious cleverness in begging, to use an assumed humility as a weapon, in such a way that the pretence of a proud abasement leaves the one who is asked no room for manoeuvre. § He asserts that his property was sold off for the sake of us both, when it is agreed that it was spent for one only; he expects double gratitude for a single service. 11. § We hardly need to discuss what he thought when what he did is obvious to all. Your side’s defence was used up by the pirate, when he gave you a choice between us. § Let the father who ransomed only the son who was doomed to die keep of that son the tomb which was all he wanted to keep. For when he preferred the sick son to the healthy he proclaimed his abhorrence for the one who was destined to live on. I do not know what you can ask of the survivor when you have as your consolation the tears for which you paid.

Narration 12. Judges, I always continued to practise upright behaviour, even at the time of life that looks kindly on vices. I avoided expenditure on what a boy longs for, in order that my mature age might have something to enjoy. By my sober behaviour in this regard, I brought it about that my father had the money left that he could offer for the ransom of his prodigal son. Why do you cavil

208

Papers on Quintilian and Ancient Declamation

about the narrowness of the family means? If I had been like (my brother) in my behaviour, things would have come to such a pass that neither of us could have been freed. 13. Only an unfair judge could have wished what I saved by my principled behaviour to go to help the other. To judge by what my stern father says, what caused my long captivity was my own probity. A sober and diligent person, attentive to father and brother—that is the sort of person who does not deserve to be brought back home; and if we had been prisoners on the same terms, our joint return would have been blackened by suspicion. But we were equal in our plight, yet far apart in the love our father bore us. Happy the young man whom his prodigality could benefit!—earlier, he went through a large part of the family wealth, and then himself became the beneficiary of what remained. 14. Wretched that I am, doomed ever to be the prisoner of my father’s judgement! I played in advance and in peacetime the role I was to suffer later, and while keeping to the straight and narrow path I learned from my own parsimony what life would be like in the hands of an enemy. That is how I managed to support life, content with what food the barbarous pirate provided; however dire the circumstances, I survived thanks not to the kidnapper’s humanity but to the force of habit. I ask you: for what did those frugal practices preserve me, if being upright wins the hostility of fathers?

Objection 15. What of the fact that he claims that the barbarians said: ‘You have brought the price of a single man?’ Objections always receive their due answer. You were not held in anxious suspense about any bargaining over the sale: you offered the gold for the enemy to see before making any plea. While everything was going like this for you, no one could have doubted that you would be satisfied with the freeing of one son. The moment you went into those caves, full of the filth of death, the pirate knew what you were going to choose, even before offering you the choice. The trepidation you showed was not the result of affection; there is no doubt about the intentions of a man who by every sign of his concern proclaims that of two sons he has come for one only. After your departure, I was guarded still more closely. 16. I owe thanks to the gods, if here too, in the place I escaped to, they ensure my liberty. Your choice, meantime, seems to have said aloud to the enemy: ‘Unless the boy I leave behind is carefully guarded, he will slip away.’ But the gods led me, and I made good my escape in full view of the foe.

Digression 17. Further: my father has rich resources of speech, and has lived long enough to have learned how to bring home what he says to his listeners, well able as he

Ennodius, Dictio 21

209

is to use their hearing as a means of exerting irresistible force upon their minds. He says that for all mankind food is in common, and demands from one man what it is wrong for anyone to refuse; his generalization is that whatever the earth produces is owed to everyone. 18. Yet in the days when it was for the common use that corn produced its wheaten offspring and vineyards poured forth their liquor, when the land had not been divided up among cultivators one man to a plot, no one left his son in enemy hands in such a manner that he might be thought to have paid more attention to groans than to slavery. He chose tears; judges, I know not what he expects of the survivor. § But he asks me to fear the law, and wants the provisions of a statute covering fathers and sons to apply to him even though he has shown no signs of being a father. He looks to my will, even while I am being subjected to compulsion, and asks me to grant to piety what he himself has asserted one who refuses cannot deny to fear. 19. Is there any place for pity where power is urgent? We can hardly say that liberality confers what cannot be withheld. Clemency is not properly so called if it is not freely given; you can count as a gift (if you have given it) something that it is in your power not to give. § But the constant repetition of the word ‘nature’, the holy name of begetter, stands in our way: as though the temporary enjoyment of the light of life, the gift of the gods, could be counted as one of the things that men can grant. It is sacrilege to equate the generosity of the gods with the seed of mortals: a cruel father can claim credit only for his lust. 20. § In his request for help for one destitute, he adds to my problems by arguing that if I had not returned the city could have come to his aid. What is the implication of that, if not that the pirate should take me back again? If you find burdensome the sight of a son for which you did not pay, you will indeed possess my tomb without expense to yourself—though you are used to paying even a high price for a tomb. My age promises me life, but the severity of the tortures that were inflicted on me (and the more harshly when you had gone away) promises my demise.

Examples 21. Next, you brought forward examples, and ones appropriate to a father. But they were cheapened when you used them in the way you did, as if hoary antiquity offered models for your whims rather than patterns for upright behaviour. What did great Aeneas’ shoulders feel of the burden of carrying his father? Did his father’s body weigh hard on a son whom he had often carried by his actions? Do you wish to know that he deserved to be snatched from the midst of the enemy? 22. It is crystal clear that he did not wish to be rescued, for he was afraid he might be a burden. He believed that his sepulchre was to be the ruins of falling Ilium; and to avoid his old limbs

210

Papers on Quintilian and Ancient Declamation

holding back his advancing son, who was more worthy to live, he proclaimed that the fall of his city was sufficient burial for himself. He wanted to stay to be the victim of plunderers, while his son escaped; whereas you left your son behind and returned to your home country. Why do you recount the noble deeds of Scipio on his begetter’s behalf? Offering himself for his father with all zeal, he gave everything, but he did not come to the end of all that he owed. 23. § The motive for good intentions is always similar, and impulses to be kind are not differentiated. That father declares what he deserves from his son who judges that he will come to perform the duties of nature only under compulsion. The law threatens imprisonment unless I obey. What difference does it make? The long period of torment that you ensured for me made me used to such suffering. § Even though long endurance of punishments has not made me defiant, I must say I am surprised if my escape from the barbarians was the doing of my father. § There is no reason to avoid pirates if the same lot pursues one back home. 24. To my mind, there is a single misfortune at home and abroad, the bitterness is the same for those who return and those who stay captive. I can find no difference whatever, and I do not try to determine what chain around my neck is preferable: it is all one whether I fall into the toils of unyielding iron or pay toll to luxury. Shall I, myself starving, bring banquets to one who is already sated, and, worse still, does not love me? Think of all the scorn with which he will spurn food I offer! Think of the virulence with which my assiduous attentions will be trampled underfoot by one whom I could not make feel any pity for me even when I was at my wits’ end! 25. He asserts that it was on level terms that my brother and I wrote to be ransomed, that it was with the same amount of tears that we asked to be freed, that it was under the same conditions that we dictated our letters. I do not deny it, or reject his assertion; as to this objection he can claim victory. But I do demand to know why it was that what two asked, what was owed to two, was received by only one. It was for both of us that you entered on the seas, and, according to you, it was with love as your guide that you crossed the perils of the watery element. 26. Yet who would give credence to a man who talks of his love for two sons if he comes back with only one? Pity is only praiseworthy when it has faced danger; when a result is lacking how can anyone have the effrontery to boast of his affection? Are the secrets of the heart always truly revealed when speech turns the key in the lock? Love which, even if it is silent, is not guaranteed by actions is a mere monstrosity. Love stands firm between all men, its roots unshaken, if this boast of yours is valid. § The law says: ‘Feed or be imprisoned.’ My father cries: ‘This food is owed to a father in any case, even if he does not ransom.’ 27. You armed yourself on this front when you neglected the duty to which nature bound you; what was shamefully denied to true purity is of service to your flatulent defence. Believe me, when the law said ‘parent’ it also meant ‘ransomer’. It exhorted, ordered, insisted that all the obligations of a sacred act are not

Ennodius, Dictio 21

211

sufficiently met by the invocation of a mere name. My plight blinded your senses to the understanding of this: the redeeming of a son, which is part of a close relative’s business, could not but be part of the instructions laid upon him. 28. § You cite the Dioscuri, yet accuse me of lack of affection, of jealousy at the freeing of my brother. But you don’t realize that if they had been subject to your ‘equal’ shares it would have caused a mistake. For you would have adjudged one to the eternal beauty of light, the other to possess night for ever. I lie if your judgements on your sons were different from that. What did you think the result was when the pirate let go one son, never to own him again, while the other he kept, never to let him go, as though the father was handing him over? 29. § All that I suffered in that dire land after you saw me I have no doubt had its origin in you. All hope was lost to me, once the ransomer left the barbarians’ shore without me. What ships making for those savage shores was I not first to spy? If my failing sight could not follow something with my eyes, I perceived it in my mind, for the more expectantly one waits the more keenly one sees. 30. How often, before your arrival, if ever the pirate permitted us to go out in the open air, did I glimpse my father in the person of strangers, and saw likenesses that love paints for itself with no disgrace of deceit! Hearts long distressed are struck by images of this kind. Alas, how often, when I refused food the pirate proffered, did the barbarian console me thus: ‘Fear not; do not sigh for this affliction; if your begetter is alive, you are not far removed from freedom.’ 31. Behold, at last what I had longed for came to pass. Even the pirate hurried to announce this man’s arrival: not far from the love generated by unsullied clemency was the man who was happy to welcome a ransomer. Judges, our masters were not inhumane: they were such as might have been softened even by entreaty. If tears had made up for the shortfall in gold, he would have brought back both at one time. He showed that he did not wish to happen what he did not make much effort to beg for. Calculate, I ask you, what he can lawfully ask of me when he did not help me in my extremity even by entreaty! 32. § You also said that you are fortified behind the privileges of nature, and that your offence against me is not subject to punishment. But the man who rests his defence on his power cannot claim the rights of innocence. Let the innocent glow in the splendour of his action; but he who thinks he can found his defence on the plea that he is superior is confessing to his crimes with no attempt at pretence. The guilty are not protected by authority, and where someone has committed a sin the rights arising from a name vanish away.

Epilogue 33. Most holy judges, after my life in the custody of brigands I have made these claims with a final effort of my breast as though I were able to do what is demanded even if I were judged to be in debt. He pays no regard to my bones,

212

Papers on Quintilian and Ancient Declamation

stripped though they are of flesh, filthy with the cruelty of my gaolers, or to the secret parts of nature opened up by the torturers. In my body all that makes up a man was brought into the light of day by my torments; the deepest recesses of my vitals became public knowledge when a barbarian pressed on his tortures while my father did not see fit to help me. 34. Beyond all measure is a severity that does not first of all ask for proof of ability to pay even when dunning a debtor. Yet though you see my youthful vigour weakened under the burden of such sufferings, the strength of my years worn down by a variety of misfortunes, though I have the needs of an old man while being envied as a youth, you do not cease to demand of me what even an enemy would not have asked of a man so enfeebled. Are you expecting to make me your tributary, and yet still assault me with the greatness of a father’s name and with a father’s prestige, disdaining to show even to a son the mercy you should offer to any man whatever? 35. May the powers above enrich your purposes with this reward, that you look a second time upon your child’s longed for funeral, and attain the height of your desires by burying him, seeing that you find his living presence a burden to you! I, Ennodius, Emended My Speech with My God’s Help

COMMENTARY

Notes to Vogel’s Text 1. pater vero: The connective vero is both inappropriate in itself and alien to the asyndetic style of themes. It should be deleted (note that ve(nditis) follows). 2. peniculo fucata mendacii: Cf. 24.26: ficta . . . et peniculo decorata mendacii; an image peculiar to Ennodius (TLL s.v. peniculus 1074.22–5). parturiet: Perhaps read parturiat: ‘the kind of display that arises . . . ’ 3. quid est aliud . . . iuvare: This might seem to be a faulty substitute for quid est aliud quod iuvet. But Bianca Schröder (henceforward B. S.) suggests that the words mean: ‘My helping myself against Quintilian’s assault is the same as my bringing justice home’, i.e. they come to the same thing. peregrinantem: Partly metaphorical (cf. 68.29), partly literal of the returned son. 5. dicat . . . si . . . es: Virtually ‘judge if ’; cf. 11.26: si sum . . . aestimate; further cases in Vogel’s Index s.v. si, adding 176.5; Hofmann–Szantyr (1972²), 544. iudicas: Alluding to the choice the father made (thus, e.g., in 7 he is an arbiter, giving a sententia). viam culparum: Cf. 72.5: via scelerum, and contrast the recti iter (cf. 200.12: pietatis tramitem; 274.8: innocentiae calle) of the good son in 14. The speaker proceeds to contrast his own life (note the emphatic ego). decedentis (decidentis MSS): Dying, that is, at the time of the ransoming.

Ennodius, Dictio 21

213

But (B. S.) the meaning may be ‘dead’; cf. 177.4. paruit: Hartel’s par fuit (for which Vogel compared 12; note also 128.23) is surely right. 6. quod vocatur: i.e. pater; cf. 130.24 and its context. The nomen is not enough (cf. 130.23, 175.19). 7. ad hos usus: i.e. to feed him. reservasset: Rhythm might suggest that Ennodius wrote servasset (and cf. 14). The same issue arises at 237.35. cui . . . mereris: The construction seems impossible, and Hartel’s merueris imposes itself. 8. spei . . . residuum: Cf. 206.36. gemerent: gemuerint (T²) is necessary. The speaker means that, in his present state, he has only groans to offer. 9. post Cilicas: See Nisbet–Hubbard (1970) on Hor. Carm. 1.18.5, and perhaps compare post oculos tuos below, 29. resumpsit: sc. infelicitas? Or read resumpsi? 10. inprobati . . . examinis: Apparently an unfriendly allusion to Christians, who hold that Beatius est magis dare quam accipere (Acts 20.35). Cf. perhaps 298.11: Mentiti sunt homines qui se iurabant accipere beneficia, si dedissent. For this use of examen see TLL s.v. 1163.68–73. technam: tegnam codd., as elsewhere in Ennodius’ MSS. urbanitas: Ennodius often uses urbanitas and urbanus pejoratively: see 2, and Vogel’s Index. abiectionis: Such a beggar glories in his own inferiority: anyone who gives him money will be ‘more blessed’ than he. 11. exhausit . . . : i.e. once you were given that choice, and acted (quid fecisset) as you did, you left yourself no leg to stand on. 12. huius rei: One expects rather huiusmodi (used at 138.12). 13. districtio: A favourite Ennodian word for ‘severity’, unrecognized by TLL; see Vogel’s Index. et nisi . . . regressio: The sense seems to be: If we had both been loved and both ransomed, people would have (on the principle stated in Talis . . . in patriam) suspected the worst of our morals (but, as the son goes on to say, we were not on a level: though our plight was the same, only one of us was loved). That requires not nisi but si. 14. indicio: The parsimony which he had practised gave him an indication of what life in an enemy’s hands would be like. usus ille frugalior: His frugal habits enabled him to stay alive in prison, only to find his father hostile to his virtues. cui: Or (B. S.) ‘for whom’ (i.e. if not for my father). 15. accipiunt: Perhaps read accipiant. descendit: I (hesitantly) take the verb as perfect, not present, with ista referring to the father’s emotions in the pirate camp rather than in court. amore: Love, he seems to mean, for both sons. quasi: Apparently more or less = quod, with adnuntiat. 16. gratias: It. ‘grazie’. Many parallels in Vogel’s Index; TLL s.v. gratia 2226.18–27. quo (quod MSS) evasi: Vogel’s emendation (though worryingly against the rhythm) makes good sense. The gods helped the son to escape the pirates, but they need also to help him in the present trial if he is not to lose his liberty again (by becoming a debtor to his father: cf. esp. 9, and also 33). 17. praesentare: One naturally thinks of rhetorical enargeia (TLL s.v. praesento 863.36–50; Quint. 4.2.123: credibilis rerum imago, quae velut in rem praesentem perducere audientis videtur; cf. Ennod. 74.8), but callens . . . per auditum shows that the son is thinking of a more general persuasiveness. 18. servi: So B. The other

214

Papers on Quintilian and Ancient Declamation

MSS give (the rhythmical) servitiis, and a dative is desirable. If my translation is correct, the point is feeble enough, for both sons were in ‘slavery’. excusantem: = recusantem; see Vogel’s Index; TLL s.v. excuso B (1306.23 ff.). formidini: Fear of the law. 19. inter munus (inuter munos B, ceu inter munus cett.): I do not know how Vogel could regard this as Latin. The sense suggests something like id munus or id ut munus. 20. vel quod . . . : Read, as a parenthesis, quod soles vel magno conparare. 21. quasi . . . institutis: Examples, that is, are intended for moral instruction (as indeed the father had tried to use them), not as excuses for indulging one’s own wishes. For formam, cf. 89.4: Vive tu tamen, vive, exemplum et forma bonorum operum. Aetas cana is the long-ago period when exempla were set: cf. esp. 321.9: canae aetatis exempla recolantur. The son goes on to say that the parallel of Aeneas does not work: it was no problem for him to carry Anchises, and Anchises in any case deserved to be carried. actione: Balancing corpore; but the antithesis is false. vis scire: Familiar declamatory language: see my n. on ‘Quint.’ Decl. min. 266.12. That he deserved is proved by the fact that he did not wish. 22. credens . . . : cf. Aen. 2.638–49. procedentem: Perhaps prae-. relicto: sc. filio. Cf. Aen. 2.657–8: Mene efferre pedem, genitor, te posse relicto / sperasti . . . ? Scipionis: Apparently an allusion to the battle of the Ticinus (218 ), when Scipio Africanus the elder is said to have saved his father’s life (Liv. 21.46.7–8, with Walbank’s n. on Plb. 10.3.3–7). totum dedit . . . quae debuit: Scipio gave his all to save his father; but he was so indebted to him that he could never give enough. This son, on the other hand, owes his father nothing. So, again, the example does not fit. 23. quamvis . . . tamen: The words correlate, and a full stop should therefore follow diuturnitas. miror: A sarcasm; for the constr. (and the unusual subj. intulerit) see Hofmann– Szantyr (1972²), 666. It was all the father’s doing (cf. 29). But perhaps read mentior, comparing 28, and 204.11: Mentior, si umquam dispositis tuis inpedimentum exhibuit ardor; for the subj. cf. (with Mentior nisi) 17.8–9. 24. tamen: Used with little adversative force. rancore: Vogel’s Index provides three further examples of this late use. 25. ex aequo . . . : The father seems to argue (unconvincingly) that the equality of the pleas proves the evenhandedness of his own response to them; the son makes the obvious retort. habet: There is much to be said for the conjecture habeat reported by Vogel. liquentis elementi: 8.5: incerta liquentis elementi (TLL s.v. elementum 347.28 ff.); 118.28: liquidum maris elementum. 26. clave sermonis: The metaphor (apparently almost unique to Ennodius: TLL s.v. clavis 1318.41–9) is better worked out at 75.4–5: Bene secretum pectoris reseratur clave sermonis (cf. 125.19 and other passages in Vogel’s Index). pietas . . . : For the sentiment B. S. compares Theognis 979–82. tenetur: Or (B. S.): ‘can be known by’. ostentum: A ‘prodigy’, not as TLL s.v. ostendo 1136.79–80. subsistit: There seems no reason to reject the transmitted future, perhaps to be classed as that of probability (Hofmann–Szantyr (1972²), 311). If the father’s boast of loving

Ennodius, Dictio 21

215

his son is accepted, we will find that love is universal (a likely story!). iam: i.e. even before circumstances are taken into account. 27. ventoso: ‘empty’ (cf. 65.2 and 84.30). Munimini (cf. also 32: munitus) carries on the metaphor of armasti. quod . . . puritati: What is the allusion? Presumably the same as that in cum . . . cessares, the failure to ransom the frugal (‘pure’) son. But neither riposte seems to have any force against a father who says: ‘Feed me even if I did not ransom.’ sufficienter: To be prefaced with non (as in 313.7). The argument is, as in 6 and 32 (where sacramenta recurs), from the need for a father to live up to that name. But the sentiment is not here lucidly expressed (one expects not ‘a sacred act’ but ‘a sacred relationship’). calamitas . . . obnubit: But why should it have had that effect? praeceptis: the orders of the law, emphasized by the tricolon of the previous sentence. 28. Ledaea pignora: The devoted brothers Castor and Pollux, each allowed (in one story) to live for ever but on alternate days. They are mentioned as exempla at 18.17. relicturus: One would expect dimissurus. Can relinquere mean ‘relinquish’ in such a context? 29. comprehendi: Rhythm suggests that this was here spelt (or at least pronounced) comprendi. 30. aerem: For the ‘open air’, see TLL s.v. aer 1051.77–1052.5. 32. non liceat vindicari: There being, in the world of declamation, no action against a father except for madness (‘Quint.’ Decl. min. 346). defensionem suam locat in apice: ‘= est pro defensione ei potentia’ (Vogel in his Index, p. 368). actionis luce: Cf. 52.16: Quod illi concessum est pro actuum luce; 303.32: de actionis nitore conscientiam. 33. alio: Read, I think, ultimo. crucibus: cruciatibus MSS apart from B. But cf. 54.30. 34. principe loco: Mannered variation for primo loco, frequent in Ennodius, e.g. 109.21; TLL s.v. princeps 1277.28–32. tributariam functionem: Cf. 310.12; TLL (s.v. functio 1547.75–7) adds examples from Cassiodorus. The father is as it were imposing a direct tax on his son, who protests in an indignant question. genio: Cf. 61.18 Apostoli genio.

Quintilian Ennodius takes trouble to draw our attention to his confrontation with Quintilian. First, in his ‘own’ authorial voice, he asks the reader (not a declamation audience) to forgive him for this boldness (1). Then, taking on the role of declaimer, he speaks a praefatio (the Greek prolalia) that contrasts his own reliance on plain-spoken truth with Quintilian’s higher style (2–3). The natural assumption would be that he is ‘replying’ to the extant fifth Major Declamation (MD 5), believed in late antiquity to be the work of Quintilian. But as the piece proceeds that expectation is disappointed. At a number of places Ennodius responds to points made by his opponent;² but none is more ² See 261.34 (addidit; cf. perhaps MD 5.13 (97.14 H.)); 262.25 (perhibet); 262.38 (dicit); 263.16 (dicens; when the topic of the city’s support is raised in MD 5.22 (108.24 ff. H.), it is employed in

216

Papers on Quintilian and Ancient Declamation

than vaguely to be found in MD 5, and most are not to be found there at all. Most decisively, the exempla (21: protulisti exempla) of Aeneas, Scipio (263.30: commemoras) and the Dioscuri (264.20: laudas) are not used in the earlier declamation.³ It is interesting, but perhaps no more than coincidence, that MD 9 employs a Scipio and the brothers Castor and Pollux within a few pages of each other;⁴ but no exemplum at all is used in MD 5. How is this to be explained? Ennodius may have been replying to a lost declamation attributed to Quintilian; but it would be odd if two pieces on the same theme were going around under the great name.⁵ He may have been replying to a lost Minor Declamation; the extant collection (which could well derive from the real Quintilian’s school) only starts with no. 244. In that case, the praefatio might be read ironically. For the Minor Declamations are far more concerned with content than with style, and Ennodius would be pointing out how much more stylishly he is performing than the master. Perhaps most likely, Ennodius knew MD 5, but is not in the least concerned to react to it in detail. In the worst declamatory tradition he assumes that his opponent has said what he finds it most convenient to reply to.⁶

Technique Ennodius makes a show of carefully articulating his declamation. Headings mark off the various parts of the speech, and within the piece there are some self-conscious allusions to technical terminology (15: obiecta; 21: exempla; 25: obiectione).⁷ But that does not mean that the declamation could have served as a model to show a pupil how to argue a case consecutively and with due regard to the stage the speech has reached. What is important to this speaker is mannered cleverness of invention and expression, not the pedagogic aims so brilliantly demonstrated by our honorand[⁸] in the Minor Declamations. Thus, he is not interested in the status lore and the arguments flowing from it. On a

a quite different way); 264.3 (denuntiat); 264.12 (clamat); 264.39 (adiecisti). At 263.12 the de claimer mentions the father’s naturae iteratio; the concept obviously comes up in MD 5 (e.g. 5.7 (91.22 3 H.)), as does exploitation of the sacred name of father (e.g. 5.7 (91.12 13 H.); 5.8 (92.9 10 H.)); but this can hardly prove anything. The idea at 7 that the son could make his father ashamed by offering food he does not deserve is faintly reminiscent of MD 5.21 (107.10 11 H.). ³ Reitzenstein (1908), 108 n. 4, while mentioning this point, exaggerates the similarity of the two pieces. ⁴ MD 9.17 (192.7 H.); 9.22 (197.5 6 H.). ⁵ One of two possibilities canvassed by Reitzenstein (1909), 3 4. The other is that MD 5 and the declamation answered by Ennodius descend from a lost (second century?) ‘archetype’. ⁶ Cf. e.g. Sen. Con. 3.pr.12 (Cassius Severus): adsuevi non mihi respondere sed adversario. ⁷ If my conjecture ultimo at 33 is right, the word would signal the start of the epilogue. ⁸ [Professor Joachim Dingel.]

Ennodius, Dictio 21

217

related theme Porcius Latro⁹ dixit legem hanc pro malis patribus scriptam esse, bonos etiam sine lege ali, and made a division: Si non omnes alendi sunt, an hic alendus sit . . . . an alendus sit qui filium a piratis non redemit. Hoc loco quaesit an non potuisset redimere, an noluisset . . . Such precision of thought, and consequent lucidity of structure, is alien to this declamation. A point may, one feels, come anywhere. The impression is often of a series of extracts from a speech rather than a continuous piece; and I have by frequent paragraph marks (§) in my translation shown where at least some of these ‘extracts’ begin and end.

Virtues Ennodius would have thought my last paragraph quite beside the point. His dictio is designed not to instruct but to entertain. It comes to us straight from what Donald Russell felicitously called Sophistopolis. And it comes to us in some sense timelessly. Only style and language differentiate it from earlier display pieces. As far as content goes, perhaps only the term cognitores alongside iudices points to a date.¹⁰ There seems to be a hit at the Christians in 10; and the future bishop who corrects his speech Deo meo iuvante impersonates a declaimer who relies on the favour of caelestes (2) and who in turn impersonates a son who talks of superi (19).¹¹ On these terms, the piece has something to offer. Ennodius is no fool, and what he makes the son say, for all its playfulness, exercises the mind at some level. It can even move. We may feel that the horrors of the pirate prison are overdrawn (though there is much that is comparable in MD 5), but there is a genuine pathos, and the fruit of genuine observation, in the picture (30) of the son seeing in casually met faces the likeness of his longed for father.[¹²] The long and distinguished tradition of declamation, that goes back ultimately to Gorgias and his contemporaries, is not yet dead. [I regret to say that when I wrote this piece I overlooked the important discussion of the subject by Håkanson (1986), 2285–90; equally important is now the specific study by Bureau (2007).] ⁹ Sen. Con. 1.7.11. Quint. 7.1.55 and 7.6.5 are also relevant for declamatory precept. Natur ally Latro’s points are basic even to our speech (see 18.26 7), but they are not systematically employed. ¹⁰ TLL s.v. cognitor 1487.70 ff. gives instances from Ammianus onwards. The principes viri of 3 seem to be not the distinguished audience attending the declamation, but the judges who are to decide the case (cf. e.g. 221.12); or perhaps they are in some sense both. ¹¹ Kennell (2000), 153 7 unconvincingly finds the whole argument basically Christian. ¹² [Cf. the touching final words of J. G. Farrell’s novel Troubles (1970): ‘But he was still troubled by thoughts of Sarah. His love for her perched inside him, motionless, like a sick bird. For many weeks he continued to think about her painfully. And then one day, without warning, the bird left its perch inside him and flew away into the outer darkness and he was at peace. Yet even many years later he would sometimes think of her. And once or twice he thought he glimpsed her in the street.’]

17 Something New out of Armenia When I retired in 2001, my thoughts, mindful of the demands of ring composition, returned to my earliest topic for research. So my last professorial class concerned the final chapter of Quintilian’s Institutio oratoria, which begins with remarks on the orator’s retirement.[¹] It is no time for idleness. The old man should, Quintilian thinks, write history, or a legal treatise, or a rhetorical handbook, or something philosophical; and he will go on teaching, his house crowded with the young men seeking advice from the oracle. My retirement has not been much like that. But ring composition has continued, to the extent that I find myself working, with a much younger colleague, Tobias Reinhardt, of Frankfurt and Oxford, on revising for publication my long-ago doctoral thesis, a commentary on the second book of the Institutio.[²] Since I gained my doctorate in 1963,[³] naturally, a good deal has been published that needs to be taken into account in 2004—but nothing, it seems to me, of more interest and significance than the 1997 Budé of Aelius Theon’s Progymnasmata. The text of Theon, as it was known hitherto, stopped abruptly well before its proper end, though the topics of what was lost could be deduced from allusions in what was preserved. But an Armenian version, giving a complete text, became known during the last century; and Giancarlo Bolognesi (of the Catholic University of Milan) and Michel Patillon seven years ago produced a Budé edition in which the known Greek is filled out by the new Armenian material, together—fortunately for me—with a French translation. This fine edition points out some, though not all, of the parallels between this new material and Quintilian, whose closeness to Theon had long been recognized. But I think that there is room for further working of this seam, and from (as it were) the Quintilian end. This paper aims to make a start

[Letras clássicas 8 (2004), 111 28] ¹ [See Winterbottom 2005 (= A.18 below).] ² [Eventually published as Reinhardt Winterbottom (2006).] ³ [The viva voce exam having taken place in 1962. I take this opportunity to correct the careless dating of my doctorate to 1970 in the preface to Reinhardt Winterbottom (2006). Mr Reinhardt is now in his turn Corpus Christi Professor of Latin in the University of Oxford.]

Something New out of Armenia

219

on this task. In the process I shall be trying to establish precisely what went on in an ancient school of rhetoric at the pre-declamation stage; I shall hope to throw some new light on Quintilian’s tenth book; and I shall argue that what a boy learned before he came on to declamation ‘proper’ was of great importance to his education, and had a disproportionate effect on anything he might write in later life. I do not need to remind you that Quintilian’s Institutio is not constructed like a typical Ars rhetorica, or handbook of rhetoric. Instead, it takes us through an orator’s life, describing and prescribing an educational process that starts in infancy and ends only with death. I say infancy, because it seems to me that in 2.4.15 Quintilian is alluding to a method of helping very young children to learn to talk. He seems to say that it is useful, after telling them some simple story, to make them repeat it, first in the same order and then in a different order. We have perhaps here a witness to a personal experience. Quintilian, marrying late, had two children, perhaps in his sixties; and we know from an affecting passage⁴ that he was especially close to his younger son, who died at the age of 5. As early as that, a child would not have started his formal education; but nothing is more characteristic of Roman educational practice than the close interest displayed by the father. An older child of Quintilian died at the age of 9, after showing great promise.⁵ In his last days, deficiens iamque non noster, he babbled of scholae and litterae. He had, then, already started at the school of a grammaticus, who taught both formal grammar and the understanding of poetry. The age of entry to this school probably varied a good deal from one pupil to another. And equally it is hard to generalize about the age at which a boy would normally pass (if he did pass at all) from the school of the grammaticus to that of the rhetor. But a child who made this transition would seem to have stayed with the rhetor from the age of 15, at the very most, to roughly 18. This variation of practice caused difficulties for teachers, and dispute: should the grammaticus or the rhetor teach the so-called progymnasmata, the preliminary rhetorical exercises that are my topic today? The evidence concerning this dispute is too complex for me to discuss now. But one thing is obvious. The longer the grammaticus retained a boy, the more tempting he found it to teach him the progymnasmata. For a rhetor like Quintilian this was a deplorable invasion. But he contented himself with a compromise. The grammaticus was to teach the most elementary of the exercises (those named in 1.9); the majority, however, were to be taught by the rhetor—and these Quintilian lists and discusses in 2.4. The corollary of this compromise was that the age of transition should not be too late. Tiresomely, Quintilian specifies no ages. Indeed he asserts that it is not a question of age at all, but of

⁴ Quint. 6.pr.6 8.

⁵ Quint. 6.pr.10 11.

220

Papers on Quintilian and Ancient Declamation

how far the student has progressed: let him go on to the school of the rhetor when he is capable of it (cum poterit: 2.1.7). What Quintilian says in 2.1 of contemporary Greek practice (by which he may mean the practice of Greek teachers in Rome) is difficult to follow. As for the Greek writers on progymnasmata, they do not raise the problem at all. Theon, like the others, appears to assume that all the progymnasmata will be taught by the rhetor. That is not very far from Quintilian’s position; and so far as this paper is concerned, no more need to be said on the matter. I do not need to discuss the exact relationship of Quintilian and Theon. The Italian scholar Italo Lana⁶ was sure that Quintilian had read, and drawn on, the Progymnasmata. If he had not, he had surely read something very like them (I mention in passing that Malcolm Heath⁷ has recently argued that Theon wrote centuries later than Quintilian). Nor do I need to discuss the complications of the ordering of the exercises as laid down in Quintilian and Theon. Both took the sensible view that easier exercises should come before more difficult ones. But they seem to have disagreed on the way in which this principle should be applied. This matters not at all for my purposes. For I shall be concerned not, for the moment, with the individual Exercises (Narration, Theses, Commonplaces, Praise, and the rest), but instead with a group of what I shall call Techniques. Though, as we shall see, they were known to Quintilian, they are discussed systematically only in Theon, whose Budé editors called them ‘exercices d’accompagnement’; for whereas the Exercises were taught, one at a time and in succession, pupils were trained in the Techniques throughout the progymnasmatic stage. It is the Techniques on which the new ‘Armenian’ chapters of Theon throw particular light. They are headed (I translate from the French): Reading (c. 13), Listening (c. 14), Paraphrase (c. 15), Elaboration (c. 16), and Contradiction (c. 17). We had already heard something of these from the Greek version. The Greek words are ἀνάγνωσις, ἀκρόασις, παράφρασις, ἐξεργασία, and ἀντίρρησις. Theon states that ‘we shall use Reading, Listening, and Paraphrase from the start, but Elaboration and especially Contradiction only when we have achieved some ἕξις’.⁸ I shall return to the implications of this remark about ἕξις later. It is not clear from this passage, or any other in Theon, when he thought it had been attained sufficiently to allow the introduction of the Techniques of Elaboration and Contradiction. One assumes, in any case, that the three other Techniques—Reading, Listening, Paraphrase—were taught with increasing sophistication as the course proceeded and the pupil matured. We may observe that, just as the Technique chapters of Theon follow detailed discussion of the Exercises, so in Quintilian 2.4 on the Exercises is ⁶ Lana (1951), 108 51. ⁷ Heath (2002 3), 141 58. ⁸ Theon RhG .65.22 5 Spengel = p. 9 Patillon[ Bolognesi].

Something New out of Armenia

221

followed by five chapters of a more general nature. It is from these parts of Quintilian and Theon especially that we may hope to build up some picture of what went on in a rhetorical school at the progymnasmatic stage. Even the mise-en-scène is problematic. One must apparently visualize the teaching as going on in one, presumably large, room. For when Quintilian⁹ voices his views on the need to keep younger pupils separate from older ones, he says: Pueros adulescentibus permixtos sedere non placet mihi, not (perhaps) a natural way of recommending their being placed in different rooms. This reminds us of the wide age range of the pupils concerned. As we have already been told,¹⁰ pupils arrived in the rhetor’s school adulti fere pueri, and were still there iuvenes etiam facti. Even if pupils attended over an age range of only two or three years, the practical difficulties faced by a single teacher in controlling and organizing them, let alone teaching them different parts of the curriculum, must have been formidable. Quintilian does mention the possibility of employing assistants,¹¹ but not in such terms as to suggest that this was a frequent practice in Roman schools. And with or without them we seem to have to imagine a scene in which, for example, while one age group indulged in tumultuous applause of an effusion of one of their peers (applause much deprecated by Quintilian: 2.2.12), other pupils somehow got on with their work in another part of the room. Nor is it at all clear how any scope was available in such a system for attention to individual pupils. Yet Quintilian sometimes talks in terms that suggest this was possible (the problem does not arise for Theon, for he seems to envisage not classroom instruction but the private teaching of a single student). The teacher’s aim, for Quintilian,¹² was to observe the strengths and weaknesses of each pupil, and to adapt his teaching to them. Equally, the words secundum cuiusque vires¹³ seem to tell us that the teacher, when providing pre-formed material to his pupils, should make sure that he took account of their different abilities; and in the analogy with the feeding of birds that follows in 2.6.7, the parent birds share food out among their nestlings, rather than (the implication should be) leaving it for them to fight for their own portion. The fifth chapter of Book 2 is relevant to this problem of attention to individual pupils, while also forcing us to confront the implications of Theon’s discussion of Reading. Quintilian tells us (1) that it is valuable for the rhetor to equip his students with the reading of history and oratory in exactly the same way as the grammaticus was expected to provide poetarum enarratio, exposition of the poets. It is typical of the lacunae in our knowledge of ancient educational practice that we are ill-informed precisely how this vital part of the grammatical syllabus was conducted. Indeed, when the most helpful English ⁹ Quint. 2.2.14. ¹² Quint. 2.8.4 5.

¹⁰ Quint. 2.2.3. ¹³ Quint. 2.6.5.

¹¹ Quint. 2.5.3.

222

Papers on Quintilian and Ancient Declamation

book on Roman education, that by Bonner,¹⁴ comes to this matter, it has to use as primary evidence our chapter of Quintilian on enarratio of prose by the rhetor. Nor is Quintilian himself very lucid. He remarks¹⁵ that this practice is one that he had himself sometimes adopted. And it was, he says, current in Greek schools (I suppose in Rome), ‘but’ (I quote) ‘rather by means of assistants, for clearly time is not available for teachers themselves legentibus singulis praeire’. What is meant here, apparently, is a procedure by which the teacher gave a model reading of a passage; then each pupil read the passage aloud, either to the teacher or to an assistant. At some point would come exposition of the content, directed to the class as a whole. Quintilian explains (4) that he is not talking, at this progymnasmatic stage, of the sort of ‘reading’ that might have taken place in the school of grammaticus, with attention given to (for instance) the meanings of individual words. Rather, the rhetor would do something more appropriate to his profession, by pointing out the stylistic virtues and also vices of the passage under discussion. No question, he adds (5), of calling each pupil up to the feet of the master, and ‘playing the slave to him’ in the reading of whatever text the boy chose. Here Quintilian seems to be especially dismissive of the idea that the text might not be the rhetor’s own choice.¹⁶ And one can see that to allow each child to come up with a different text would make exposition on a class basis impossible. But presumably that each should come up to do his own reading of the same text was just what did happen in a class where the employment of assistant teachers (as in Greek schools, or in Quintilian’s own occasional practice) provided the time for legentibus singulis praeire to be possible. Quintilian, however, dismisses that idea too for his present purposes; for he is going in 2.5.6 to put forward a compromise solution—one, presumably, that he himself normally employed. In the school of the grammaticus the solution may have been to make the class read the (poetic) passage together. This may be what is implied by Seneca the Elder’s reference¹⁷ to a group of people speaking singuli as opposed to omnes simul (but semel is transmitted) tamquam in choro manum ducente grammatico, ‘individually, not all together like a choir with the grammaticus conducting’. But the passage may well bear a different interpretation. Again, Macrobius¹⁸ speaks of Virgilian verses qualiter eos pueri magistris praelegentibus canebamus, which ties the occasion down to enarratio poetarum without making it absolutely clear that the ‘singing’ was done by a group in unison. In any case, Quintilian does not mention such a solution. Instead, he recommends (2.5.6) choosing one of the pupils—turn and turn about—to act as the reader for the occasion. That would give all the pupils the chance to perform, ¹⁴ Bonner (1977), 225 6. ¹⁵ Quint. 2.5.1 3. ¹⁶ We may compare how Theon in c. 14 (p. 107 [Patillon ]Bolognesi) stresses the need for the master alone to choose the topic for his lecture; but the passage is highly obscure. ¹⁷ Sen. Suas. 2.13. ¹⁸ Macr. Sat. 1.24.5.

Something New out of Armenia

223

and in the process to get practice also in pronuntiatio, which would doubtless cover gesture as well as vocal effects. This combination of reading with practice in delivery is found also in Theon c. 13;¹⁹ the Budé editors suggest that this shows that the boy read from a book supported on a lectern; but it is possible that Theon is here thinking, as a little further on and in a different context,²⁰ of recitation from memory. As for Quintilian, even though in 2.5 he talks in terms of ‘reading’, an earlier passage (1.11.14) makes matters much clearer: ‘when the time comes to read orations, and the pupil is beginning to appreciate their virtues, he must be supervised by a careful and experienced teacher, who will not just train him in reading, but make him learn by heart chosen passages from the speeches, and speak them on his feet, clearly and in the same manner he will eventually have to adopt in court, thus training delivery, voice, and memory at one and the same time.’ In precisely the same way, Theon speaks of the teacher telling the pupil to learn off examples from literature of the exercises being worked.²¹ Another relevant passage is Juv. 7.152–3, where a big class quaecumque sedens modo legerat, haec eadem stans / perferet atque eadem cantabit versibus isdem; but this raises too many complications to be discussed here.[²²] The way the master should conduct his enarratio is laid down in 2.5.7–12. We should notice carefully the terms in which Quintilian makes his first point. The reader once chosen, tum exposita causa in quam scripta legetur oratio (nam sic clarius quae dicentur intellegi poterunt) . . . The future tenses here (especially legetur) seem to show that this exposition of the circumstances of the case from which the passage is extracted is meant to precede the reading itself. Are we then to suppose that the whole enarratio, the identification of the virtues (and the faults) of the passage in question (as described in 7–12), is also to precede the reading? Quintilian does not say so specifically, while the discussion in Theon (c. 13) does not seem to refer to a classroom situation at all, but—as we have seen—to private instruction of a single student. It may perhaps be that Quintilian, who is after all talking about something whose details he and his Roman readers knew very well, is giving us highlights of a more complex procedure that may have gone like this. The teacher dictates the passage to his group. They copy it down on their tablets or papyrus sheets, and take it away to study at home overnight. Next day the teacher appoints the reader for the day, and does his exposition. Finally, the pupil reads, or recites, the passage. This is entirely conjectural. But a student listening to the reading would be able, on this hypothesis, to recognize the features he had been told to look for, and understand better a passage that he might have found puzzling ¹⁹ Theon p. 103 [Patillon ]Bolognesi. ²⁰ Theon p. 105 [Patillon ]Bolognesi. ²¹ Theon RhG .65.29 66.2 Spengel = p. 9 Patillon[ Bolognesi]; cf. an obscure passage at the end of c. 13. ²² [The Juvenal passage has been elucidated by Stramaglia (2017²), ad loc.]

224

Papers on Quintilian and Ancient Declamation

during his work at home the evening before. One might compare the process by which we might read a portion of an unfamiliar classical text, then study a good commentary, and finally reread the text. I have talked of a ‘passage’ being read in this way in a rhetorical school, with the implication that it would not be a particularly long one. We have seen that 1.11.14 talks of parts chosen from speeches. Yet Quintilian does not in 2.5 speak in such terms. He talks (7), as we have seen, of the oratio to be read (he does not return to the alternative of history until much later in the chapter, at 19), and he recommends the teacher to comment on details that would span a whole speech, from proem to epilogue, and on a whole variety of tones. But one might have expected that Reading would be tied in with the particular Exercise currently being studied. Thus, if the Exercise were the Commonplace, and the subject pro amore,²³ it would make sense for the class to read a portion of Cicero’s speech for Caelius. But Theon too, like Quintilian, talks in c. 13 in terms of a whole speech. It is true, however, that in c. 2 he has given parallels to the various Exercises that take the form of extracts from prose texts, Plato, Herodotus, and the like. Practicality alone would suggest that the model of an extracted passage is preferable. It is beyond belief that a class could, at one sitting, listen to even a short speech of, say, Cicero, together with an exposition of its virtues. We might, of course, conjecture that, as the year went by in practice of the various Exercises the separate sessions devoted to Reading dealt with extended texts, a little at a time; rather as in England today an A-level class might read a book of Virgil over a period of some months, in two or three lessons each week. But one feels that, if this system had been adopted in antiquity, it would have unduly narrowed the range of texts introduced to the pupils (as indeed it does in England). Selection, not just of authors but of passages within authors, seems the natural way of proceeding, and that indeed seems to be hinted at later on in Quintilian’s chapter,²⁴ as well as being clearly stated in 1.11.14. If that is right, it looks as though we must understand Quintilian 2.5.7–12 as providing general headings for a commentary from which the teacher would choose what was appropriate on any given occasion. Thus if (for example) the proem of Cicero’s Pro Caelio were being studied, the teacher would describe the circumstances of the case, discuss the tone of the passage, and show how it lays the basis for the argumentative strategy of the whole speech. Theon’s remarks in c. 13 should perhaps, then, be taken in the same way. Quintilian 2.5 finds other parallels in Theon’s chapters on what I have called Techniques. These suggest (as usual) a close connection between the two authors, or at least a convergence of Roman and Greek didactic practice. I note three instances in particular. Quintilian’s remarks (16) on the ‘correction’

²³ See Quint. 2.4.23.

²⁴ Quint. 2.5.24 5.

Something New out of Armenia

225

of a published speech of a distinguished author (so much more agreeable to everyone than the correction of a pupil’s efforts) should be seen in the light of Theon’s chapter (c. 17) on Contradiction, where the pupil is ‘invited’ to contradict the arguments found in a written text; the example given is from Demosthenes. Second, the authors to be studied: just as Quintilian (19) recommends, among the historians, Livy rather than Sallust for the youngest (doubtless with the implication that Sallust could be read by older pupils), so Theon (c. 13) mentions first Herodotus, with his style ‘d’une grande simplicité’,²⁵ then Theopompus and Xenophon, Philistus and Ephorus, and finally Thucydides, whose obscurity Theon goes on to stress. Quintilian’s equation of Livy with Herodotus, and of Sallust with Thucydides, seen also in 10.1.101, is obvious. Finally, just as Theon (c. 13) remarks that imitation of Lysias alone will make pupils ‘indigents, faibles et sans grâce’,²⁶ so Quintilian (21) warns against excessive concentration on reading the old orators like Cato and the Gracchi, for pupils thus fient . . . horridi atque ieiuni. The harm caused by concentration on a single model is also stressed in Quintilian’s chapter on imitation.²⁷ This process, the enarratio of prose authors, is explicitly modelled on the grammaticus’ explanation of poetic texts. We are not surprised therefore to find parallels to Quintilian’s recommendations here in his earlier discussion of the task of the grammaticus,²⁸ who will ‘imprint’ (infigat: I will return to this metaphor) on the minds of his pupils the various virtues and technical devices of the poetry they are studying. It is interesting, too, to find echoes when Quintilian comes, in Book 10, to give advice to students who have reached an advanced level. He there tells us what this older reader should look for, in terms similar to those used in laying down the sort of thing a rhetor should point out to his pupils. Thus²⁹ a reader should be acquainted with the cases in connection with which the speeches he reads were spoken. And in 10.2, the chapter on imitation, Quintilian makes the point (27) that a model is to be followed not just in vocabulary but in the whole conduct of a speech; one needs to pay close attention to the strategies of the speech and to its emotional effect. Thus in childhood the grammaticus and rhetor taught a student to look for certain features of a text he read; when he is older, he is able to look for them himself. I shall come back later to some other aspects of Book 10. Such is Quintilian’s conviction of the value of reading that he compares it very favourably with anything a teacher can provide in the way of a spoken model. He asks: ‘If the teacher declaims to provide his pupils with models, will not reading Cicero and Demosthenes make a bigger contribution?’³⁰ Earlier, true, he has said that ‘Though plenty of examples for imitation accrue from

²⁵ Theon p. 104 [Patillon ]Bolognesi. ²⁶ Theon p. 105 [Patillon ]Bolognesi. ²⁷ Quint. 10.2.23 6. ²⁸ Quint. 1.8.16 17. ²⁹ Quint. 10.1.22. ³⁰ Quint. 2.5.16.

226

Papers on Quintilian and Ancient Declamation

reading, more satisfying nourishment comes from the “living voice” of the teacher.’³¹ Any contradiction is explained by the context of the two passages, and Quintilian surely did not mean to depreciate the value of models provided by the rhetor’s own oral efforts. And we may now turn to the part these played in the schoolroom. Ipse aliquid, immo multa cotidie dicat quae secum auditores referant:³² the teacher should make formal ‘speeches’ every day. Similarly Theon in c. 14³³ speaks of ‘daily Listening’,³⁴ just as earlier he had spoken of daily writing in association with Listening.³⁵ And the pupils, Quintilian says, are to take such speeches back home with them, not, I think, metaphorically, but in the form of notes in which the master’s words, or some of them, were taken down. Quintilian talks elsewhere³⁶ of different methods of correcting a boy’s (I suppose written) work. The teacher may make alterations, sometimes interposing something of his own. Or he may ‘dictate’ a whole exercise, so that a boy can imitate it and ‘love’ it until he can produce something good for himself (the use of the word puer reminds us that Quintilian is here thinking of the progymnasmatic stage). But a more regular practice would have been for the teacher to speak a model exercise to the class. This certainly happened at the later full-declamation stage, as we know from several passages. One is 2.5.16, which I mentioned earlier. Another is 2.6.1, where one method is to supply a whole speech, complete with argument and emotional appeal, as opposed to a mere ‘division’, the bare headings of a speech (Quintilian here seems, rather disconcertingly in the context of these chapters, to be talking in terms not of the progymnasmata, but of the declamations that were taught after them; but I take it that what he says can be taken to apply, mutatis mutandis, to progymnasmatic exercises also). All this, of course, went on in the privacy of the schoolroom, and is to be distinguished from the display speeches that a master might give in public, to advertise his school and show off his talents. This sort of thing goes back to the very roots of declamation: to the speeches given by the Greek sophists when teaching their pupils. The aim, as ever, was to provide a model for imitation. Isocrates, in a striking passage (13.18) adduced by Theon’s Budé editors, says that it is essential that a teacher should present himself as a παράδειγμα, ‘so that those who have received his imprint (τοὺς ἐκτυπωθέντας) and are capable of imitating him may at once become more accomplished orators than the others’. That terminology continues into Theon: thus τυπούμενοι γὰρ τὴν ψυχὴν ἀπὸ καλῶν παραδειγμάτων κάλλιστα καὶ μιμησόμεθα: ‘for if our souls receive the stamp of good examples, we will also imitate in the best possible way’;³⁷ compare ὅπως τυπωθέντες κατὰ τὴν ἐκείνων ἀγωγὴν μιμήσασθαι δυνηθῶσιν, ‘so that pupils may receive the stamp ³¹ Quint. 2.2.8. ³² Quint. 2.2.8. ³³ Theon p. 107 [Patillon ]Bolognesi. ³⁴ Theon RhG .62.9 Spengel = p. 4 Patillon[ Bolognesi]. ³⁵ Note also Cic. Brut. 305. ³⁶ Quint. 2.4.12. ³⁷ Theon RhG .61.30 1 Spengel = p. 4 Patillon[ Bolognesi].

Something New out of Armenia

227

of their teachers’ models and be able to imitate them’,³⁸ and perhaps, from the Armenian (c . 13), ‘Lorsque quelqu’un admire ce qu’il y a de bon chez tous et entreprend d’y conformer sa pensée, du fait qu’il existe en lui une sorte de matrice du discours, que chacun peut modeler d’après sa propre nature . . . ’³⁹ And this is Quintilian’s imagery too. He remarks that, by learning off pieces of oratory or history, pupils ‘will always have within them something to imitate’, et iam non sentientes formam orationis illam quam mente penitus acceperint expriment.⁴⁰ The pupil was ‘stamped’ with the imprint of the models his teacher provided orally, or by what he read, aloud or in private, from authoritative texts. The teacher then gives his model, let us say of a Narration. In, for instance, the Libanian collection, the topic would be taken from myth. For Quintilian that sort of Narration would be the province of the grammaticus; he wishes the rhetor to start with stories taken from history.⁴¹ The master would give a model speech, telling the story. If there was, as I have suggested, some attempt to gear Reading to the Exercise currently being taught, the associated lesson in Reading might be that very passage of Livy, or some other from the same author; that would be appropriate in another way, for Quintilian (we recall) approved of the reading of Livy by the young, that open and accessible writer.⁴² Finally: ‘I want’, says Quintilian, ‘Narrations to be composed in writing, with all possible care.’⁴³ We may imagine then that, after the teacher has spoken his model Narration, the pupil takes home notes of what had been said, and with their help, excogitates a written narrative of his own. This brings me to a mysterious chapter in the ‘Armenian’ part of Theon, that concerning Listening (c. 14). This is how I translate ἀκρόασις, for, like the names of the other Techniques, the word refers primarily to an activity of the student. But ἀκρόασις may also mean what the student listens to, something spoken by the teacher. Now, at the start of the chapter, Theon tells us how, by stages, a pupil may come to be able to recall (for example) every detail of the style, content, and structure of an oratorical composition that he has listened to. The point of this process of recollection, it emerges, is that the student should be able to reproduce what he has heard, not orally, but in writing. Theon seems to be talking of what we think of as a declamation. And, as usual, he talks in terms of a complete speech. But, granted that the Technique of Listening was inculcated throughout the progymnasmatic course, it may be that what Theon says needs to be understood to cover also the sort of case ³⁸ Theon RhG p. 71.1 2 Spengel = p. 15 Patillon[ Bolognesi]. ³⁹ Theon p. 105 [Patillon ]Bolognesi. ⁴⁰ Quint. 2.7.3. ⁴¹ Of course, history might have more than a touch of myth: when Quintilian comes on to the exercise of refuting Narrations, he gives the example of the wolf that suckled Romulus (2.4.19). But that story was to be found in the historian Livy. ⁴² Quint. 2.5.19. ⁴³ Quint. 2.4.15.

228

Papers on Quintilian and Ancient Declamation

to which I referred above: the teacher gives a model Narration, orally. The student, having listened to it, works to internalize it in the way described by Theon, and reproduces it in writing.⁴⁴ The later part of Theon’s chapter on Listening is very obscure, as his editors frankly acknowledge. Clear enough is the notion that, ‘if there is no Listening’, i.e. if the teacher gives no oral presentation, the pupil will instead be asked to write on a topic suggested by, for example, some recent event—a procession, perhaps, or a spectacle. But what are we to make of the paragraph in which Theon talks of certain works that have been wrongly attributed to famous authors? I can only suggest that there is a lacuna before this paragraph, in which has been lost a transition to a different sort of lecture, where the master did not declaim, but talked about literary topics connected with prose literature. Robert Kaster, in his lucid way, remarks in his Suetonius commentary⁴⁵ that ‘as an orator’s ἀκρόασις would be a “declamation” . . . or a poet’s a “recitation” . . . a gramm(aticus)’s would be a “lecture” or some similar display of his expertise’. It may then be that Theon, like Quintilian, thinks of the rhetor extending to prose the sort of treatment that a grammaticus gave to poetry, and that this treatment might sometimes take the form of a lecture on the authenticity of a prose text. For that, of course, was what a grammaticus, exercising his κριτική, might do for a poetic text. But what the following paragraph may mean, I have little idea, and I will not discuss it now, except to repeat the important new information that Listening occurred every day. To return to Quintilian: What happened after the pupil had written his little effort following the model of his teacher? It would be natural to expect that the pupil would read his effort out in class, or even memorize it and recite it. And perhaps indeed he sometimes did, even in Quintilian’s class (Theon does not help on this topic), no doubt with appropriate measures to discourage the other pupils applauding in the unseemly way described at 2.2.12. And the passage I have mentioned about the teacher’s correction of a pupil’s exercise⁴⁶ might after all concern the oral and public criticism of a spoken exercise. Clearer is the evidence of 2.7.1, where we find disapproval of the custom by which the pupil learned by heart everything (sic) he had written in order to speak it on a formal occasion, perhaps in the presence of the fathers who are said to have insisted on such displays. Quintilian’s policy was to restrict this sort of thing, to make sure that only worthy pieces of composition were thus publicized. We may, incidentally, observe the phrasing of this concession: Aliquando tamen permittendum quae ipsi scripserint dicere.⁴⁷ The implication is perhaps that pupils could and should recite what other people had written; ⁴⁴ I earlier made a similar suggestion about the precepts of Theon (and Quintilian too) con cerning the enarratio of prose texts: that, though they are framed in terms of complete texts, they may be thought of as applicable to parts of those texts. ⁴⁵ Kaster (1995), 60. ⁴⁶ Quint. 2.4.12. ⁴⁷ Quint. 2.7.5.

Something New out of Armenia

229

and this seems to take us back to the practice of Reading, where (as we have seen) a pupil read aloud a passage from an established author. And the emphasis in this chapter on the desirability of a pupil learning by heart not his own work but that of classic authors fits in with my earlier assumption that in Reading the pupil learned the passage by heart and recited it, with appropriate gestures. In any case, it is clear that Quintilian was thoroughly in favour of his pupils learning passages of prose, and he gives cogent reasons for this favour:⁴⁸ such learning exercised the memory, and provided the internal forma for imitation which I mentioned earlier. And later⁴⁹ Quintilian remarks that he thinks Demosthenes is to be ‘read or rather learned off by heart’. At the very earliest stage he has recommended the memorizing of ‘chosen passages from the poets’.⁵⁰ Prose was more difficult to learn off than poetry;⁵¹ but that only made the practice the more valuable. It will be recalled that Theon remarked: ‘we shall use Reading, Listening, and Paraphrase from the start, but Elaboration and especially Contradiction only when we have achieved some ἕξις.’ As his editors acknowledge, we cannot tell at what point the two last Techniques began to be applied. For Quintilian, however, ἕξις was something not to be attained until a much later stage than the progymnasmata. He does not speak of such established ability, indeed, until the start of Book 10. He says at the opening of that book⁵² that it is debatable what most contributes towards this quality: writing, reading, or speaking. And the book contains, amongst other things, chapters on reading (1), writing (how to write: 3; what to write: 5), and the closely associated imitation (2). Amidst all this there is much to remind us of Theon. It seems that Quintilian, entirely approving of the Greek’s procedures, felt that they should be extended over a much longer time scale, perhaps even over a lifetime. Ἕξις could not possibly be attained early on: far more experience was required. Naturally, the Techniques are adapted to an older student. Reading is now an entirely private activity, and, as we have seen, the reader must now himself look out for the stylistic features formerly pointed out to him by a teacher. Listening⁵³ is now attendance not at some lecture, but at a real-life oration. Correction, discussed in 10.4, is neither the process described in 2.4.10–12,⁵⁴ where a kindly teacher points out errors in a pupil’s work, nor yet that seen in Theon c. 17, the Contradiction that dares to find fault in a Demosthenes. Rather, it is the process of going over one’s own work and giving it a final polish on the basis of one’s mature judgement. But it is in the two chapters on Writing that the echoes of the second book are most obvious. One must write ⁴⁸ Cf. Quint. 2.7.2 4, supplemented by 1.11.14. ⁴⁹ Quint. 10.1.105. ⁵⁰ Quint. 1.1.36. ⁵¹ Quint. 11.2.39. ⁵² Quint. 10.1.1. ⁵³ Quintilian uses the word auditio at 10.1.10. ⁵⁴ Cf. Quint. 2.2.7.

230

Papers on Quintilian and Ancient Declamation

with all care and as much as possible.⁵⁵ Writing is a remedy against ‘empty loquacity and words that are born only on the lips’.⁵⁶ Writing stores away treasures of words for future use.⁵⁷ This, and more, from 10.3. In 10.5, the discussion of what is to be written by those striving for ἕξις (and note that this is still not yet achieved), Quintilian starts with translation from Greek, but soon comes on to Paraphrase, to which Theon had devoted a separate chapter (15) as one of the Techniques. For Quintilian it is something to be applied to poetry, to Latin speeches in a spirit of emulation, even to one’s own work. He remarks strikingly⁵⁸ that the exercise gives us the incentive to look very carefully at a piece of writing, ‘not passing it over in casual reading’, but weighing up its virtues. This, in fact, is a means of encouraging critical reading of texts. Quintilian had mentioned paraphrase in passing at 1.9.2, as something to be applied to a piece of poetry with the aim of abbreviation or ornamentation. This was appropriate at a very early stage of rhetorical education, so early that even Quintilian was happy to leave it to the grammaticus. And Suetonius too mentions it in this context.⁵⁹ Theon, on the other hand, thought it should be taught from the start of the course in progymnasmata. Quintilian (who has not mentioned it in Book 2) now promotes it to an even higher class. It is notable, too, that he associates with it the (thoroughly sophistic) exercise of ‘expanding what is by nature brief and amplifying the insignificant’;⁶⁰ he may here be working into his system what Theon (c. 16) had called Elaboration, a Technique that became suitable as the progymnasmatic curriculum progressed, though not right from its start. As 10.5 goes on, we realize that we are being given a partial reprise of elements of Book 2. We are recommended to write Theses (with the comment that Cicero exercised himself on them even when he was a great man), and to employ the associated exercise of supporting and refuting them.⁶¹ Then come Commonplaces, again with a comment suggesting their utility not just for the older student but even for the mature: ‘we know that orators wrote them out’.⁶² And finally (10.5.14) declamation itself recurs, recommended specifically as something which, if properly exercised, will be of high use ‘not just while one is growing up, but when one is the complete orator, already famous in court’. Quintilian is quick to warn that one can spend too long in the declamation school;⁶³ but he is clear that declamation, like the simpler Exercises that preceded it, continues to be of service as worth writing (and surely also sometimes speaking) right on into adult life.

⁵⁵ ⁵⁶ ⁵⁸ ⁶¹ ⁶³

Quint. 10.3.2; cf. 2.4.15. Quint. 10.3.2; cf. 2.4.15: ‘the garrulity of extemporization’. ⁵⁷ Quint. 10.3.3; cf. 2.7.4. Quint. 10.5.8. ⁵⁹ Suet. Gram. 4.5. ⁶⁰ Quint. 10.5.11; cf. 1.9.2. Quint. 10.5.11 12. ⁶² Quint. 10.5.12, looking back to 2.1.11. Quint. 10.5.17 18; cf. 12.11.15.

Something New out of Armenia

231

It should not be assumed that what Quintilian recommends in the Institutio took place in every detail in reality, even in his own school. But that students spent much time and trouble on the progymnasmata, that what they learned at this stage became ‘building-blocks’ when they came to compose complete declamations,⁶⁴ that even the grown Cicero thought it worthwhile to continue to practise on these lines, all this is beyond question. And I propose to conclude this lecture with some remarks on the wider importance of these exercises, preliminary as they were, yet of fundamental and continuing value. If we are seeking for traces of the influence of rhetorical education on Latin literature, it is most natural, and most fruitful, to look in speeches: which means, primarily, in the orations of Cicero, but also secondarily, in formal utterances represented in other genres, especially epic and the novel. Particularly important here is rhetorical precept on the articulation of a speech. When we try to go further than that, we come up against the problem of the inclusiveness of precept. To remark, for example, that an argument used in a speech in Ovid’s Metamorphoses can be found catalogued in Quintilian is not to make a great discovery; for Quintilian aims to catalogue as many arguments as possible. It might be more significant if it were not found in Quintilian. Or again: Susanna Braund has shown that certain passages in Juvenal 6 can be assigned to the loci listed by Cicero for the arousal of indignation.⁶⁵ Well, they would correspond, wouldn’t they?—for those loci are intended to cover as many ways as possible of making a listener angry. Nor, again, is it helpful to say that the various tropes, and figures of word and thought, which one can find in a literary text are ‘rhetorical’ in the sense that a rhetor inculcated them. The problem of inclusiveness arises here also. It is not easy, perhaps impossible, to express oneself without using tropes and figures; and any trope or figure one does use will probably find its place in the handbooks. In any case, though Quintilian, like the Ad Herennium and the De oratore, discusses such matters (in Books 8 and 9), they were really the preserve of the grammatici.⁶⁶ It was part of the task of the grammaticus to point out to his pupils tropes and figures in poetry; and, if Quintilian thought the rhetor should do the same in his exposition of prose texts, that was avowedly parasitic on the practice of the grammatici. Fifteen centuries later, in the English ‘grammar school’, schoolboys of the Elizabethan period were taught ‘to know the name, definition and use of a large number of figures of speech’.⁶⁷ They would first learn the figures, then identify them in whatever they read, thirdly use them themselves. That is the training that led to the extravagances of early Shakespeare. It is not specifically rhetorical. To come closer to the rhetor’s school: at least some of what we tend to call ‘rhetorical’ expression or tone in Latin literature might more helpfully be ⁶⁴ Cf. Quint. 2.10.2. ⁶⁷ Vickers (1970), 48.

⁶⁵ Braund (1988), 3 5.

⁶⁶ Quint. 8.5.35.

232

Papers on Quintilian and Ancient Declamation

called declamatory. Inclusiveness is relevant here too. Rhetoric, allied with grammar, aimed to classify all means of expression; it would not be absurd to say that everything we write is in some sense ‘rhetorical’. But there is a certain tone, especially dominant in post-Augustan literature, that can be reasonably called ‘declamatory’, a tone that I once described as ‘hectic, hectoring and melodramatic’.⁶⁸ Such tone expresses itself in paradox, exaggeration, and ingenuity of all kinds; it favours not so much the long period as the short phrase (especially to mark articulation by closure). It is reasonable to suppose that this sort of tone began to impose itself at the progymnasmatic stage, at least where a teacher was so inclined (and, as a few passages in the Institutio show, even Quintilian could write like this when he wished to). This is where the model speech was so important: tone and style no less than content. All the same, it would be difficult for a 13- or 14-year-old to master such a tone quickly. It was surely the content that affected him most profoundly. No age is more impressionable than this. What one learns then stays with one for ever. And, in any case, the content of a boy’s later rhetorical studies was of no great relevance if he chose to be a poet or essayist rather than an orator. But the progymnasmata taught him to discourse upon a very wide variety of topics, using techniques that would be relevant later, not merely in declamation or oratory, but in any form of literature. The wide utility of Narration is an obvious example. Inculcated by instances from history, it could feed back into history again. The exercise in Praise not merely provided frameworks (and checklists) for eulogy in other genres; it also, as Quintilian puts it, equipped a pupil with cognitio rerum, and exempla that were indeed in omni genere causarum potentissima, but might illuminate a philosophical point too.⁶⁹ Quintilian appends to Praise the Comparison of people. The comparison of Milo and Clodius is a Leitmotif of the speech for Milo; but comparison is also vital for a Plutarch writing biography. Then there are Commonplaces, directed, say, against sexual excess. Add a defendant, and this becomes an accusation, says Quintilian;⁷⁰ and so Caelius found. But a philosopher like Seneca might use such material too. As for Thesis, Quintilian gives as an example the merits of life in the country as opposed to life in the town.⁷¹ That topic could find a place in declamation, as Decl. min. 298 shows, or in real-life oratory (Cicero exploits it in his speech for Roscius of Ameria). But it is reflected in literature of many kinds, including satire and didactic. It is hard to think that it does not have some bearing on the tone (and weaknesses) of the closing passage of the first Georgic. Most influential of all, perhaps, was an exercise found in Theon (though not in Quintilian 2.4). This was Prosopopoeia, where the pupil was asked to ⁶⁸ Winterbottom (1980b), 60. ⁶⁹ Quint. 2.4.20. This is perhaps the only point where history in the modern sense was taught. ⁷⁰ Quint. 2.4.22. ⁷¹ Quint. 2.4.24.

Something New out of Armenia

233

imagine ‘what sort of words would be spoken’ by someone in a particular situation. Other progymnasmatists often give examples from mythology: what would Medea say when Jason married another woman?⁷² But Theon’s examples draw on real life or on history: what would a husband say to his wife when he was setting off abroad, or Datis to the Great King after the Persian defeat at Marathon?⁷³ A pupil trained in this exercise could go on successfully to write a declamation in character. When grown up, he might be a Cicero, able to give words to Appius Claudius Caecus inveighing against Clodia. But equally he might be an Ovid writing a speech for a mythological character in the Metamorphoses or for a lover in the amatory works; or a Propertius who found in his fourth book that he could widen his scope by the use of a variety of speakers; or a Tacitus writing a speech for the Scottish chieftain Calgacus on the eve of a battle with invading Romans. In such ways, the progymnasmata left an imprint on a boy’s mind, and on his style, that might well last a lifetime. Here, perhaps, more than in the details of rhetorical precept, may lie a crucial part of what we mean when we talk about the influence of rhetorical education on Latin literature.

THEON

QUINTILIAN

exercises: 3 12

exercises: 2.4

techniques (only in Armenian version): reading: 13 listening: 14 paraphrase: 15 elaboration: 16 contradiction: 17

techniques: 2.5 2.2.8, 2.5.16, 2.6.1 cf. 1.9.2, 10.5.3 8 cf. 1.9.2, 10.5.11 cf. 2.5.16 enarratio, followed by reading (aloud) model speeches, followed by writing (and recitation?)

[This piece, printed as it was given at a lecture in São Paulo (hence the personal remarks at the start), brings together material on Quintilian and Theon which can be found dispersed in Reinhardt–Winterbottom (2006), summarized there at pp. –. On the role of assistants in reading and other school practices add now Nocchi (2017).]

⁷² Quintilian may well have thought such exercises should be taught by the grammaticus. ⁷³ This type Quintilian seems to have taught at the rather later suasoria stage.

18 Approaching the End: Quintilian 12.11* What we mark as the final chapter of Quintilian’s Institutio oratoria was found wanting by a sympathetic editor. R. G. Austin wrote of it: ‘[W]e see his orator retiring with powers still unimpaired and holding the position of an elder statesman to whom all the young men will turn with veneration for help and advice. There is a simple grandeur in this aspect of his conception: is it too much to claim that he has Cicero in mind, with a vision of what might have been? It is a pity, I think, that the book does not end on this note. Instead, a number of pages are devoted to a somewhat dull and unconvincing postscript, intended to show that, after all, the path to perfect eloquence is not so difficult as it might seem.’¹ Accordingly, commenting on section 7 of the chapter, Austin wrote: ‘At this point, one feels that the book ought to have ended.’ Ac nescio an eum tum beatissimum credi oporteat fore, cum iam secretus et consecratus, liber invidia, procul contentionibus, famam in tuto conlocarit, et sentiet vivus eam quae post fata praestari magis solet venerationem, et quid apud posteros futurus sit videbit. This resounding period is indeed markedly closural, with its hints of death and fame after death. In fact, up to conlocarit we might suppose that the retired orator is thought of as dead; for only in death would a great man normally be ‘far from strife’, or his fame quite safe. The securest seclusion from the world (secretus) comes after death (‘The grave’s a fine and private place’), and only then would one expect ‘consecration’. The author of the Octavia said of Augustus that he was post fata consecratus (529). But Quintilian’s orator is, it turns out, still alive to receive nothing less than veneration: he is—as it were—Divus Cicero in his own [Acta Classica 48 (2005), 175 83] * I first started to think about Quintilian 12.11 when I chose it as the topic of my last Oxford seminar before my retirement in 2001. This led to lectures, in different forms, at the Catholic University of Milan and the University of Cassino during 2003. I am most grateful to my hosts there, and especially to Elisabetta Matelli and Antonio Stramaglia. I hope that the piece, now much revised, will be of interest to John Atkinson, who, with Val, did so much to make my two visits to the University of Cape Town happy and rewarding. I hope too that he will enjoy his retirement as much as I am enjoying mine. ¹ Austin (1954²),  .

Approaching the End: Quintilian 12.11

235

lifetime, or the Augustus to whom Horace wrote (Ep. 2.1.15): Praesenti tibi maturos largimur honores. But this is not the end of the book, only of one theme in it, the one which takes us from the birth of a promising son (1.1.1) to the suspended animation of the retired orator; and that theme does indeed give the whole book a purposeful forward movement. But a no less important theme is provided by Quintilian’s concern to arouse his reader’s enthusiasm for the whole project of oratory. Far from being a postscript, the final sections of 12.11 constitute a vital culmination of this aspect of the Institutio. To return to the opening part of 12.11. Quintilian here talks of the end of a career. He was himself in retirement when he wrote his book, as we are told in 1.pr.1² and again in 2.12.12, a passage of particular interest, in which consolemur otium nostrum³ is indicative of the author’s ambivalent attitude towards retirement, while desinere dum desideraremur is taken up in 12.11.3, where Quintilian tells us of a mot directed at the orator Domitius Afer, who had gone on speaking in the courts despite the waning of his powers: Afer, someone remarked, preferred deficere quam desinere. Quintilian, whose mentor Afer had been, had chosen to avoid that. But the early sections of 12.11 are only allusively concerned with Quintilian. Rather, as Austin suggested, Quintilian has Cicero much in mind. We are pointed explicitly to a passage of Cicero’s dialogue De oratore, where Crassus is made to look forward to a retirement whose solitude is alleviated by the throng who will come to consult him on matters of law; Est . . . sine dubio domus iuris consulti, Crassus says, totius oraculum civitatis (1.200). Quintilian picks that up (12.11.5) by drawing a picture of the home of his retired orator thronged by optimi iuvenes seeking from him the ‘path of speaking’ as though from an oracle (velut ex oraculo). To this we should add a passage apparently overlooked by the commentators, in Cicero’s De senectute (28–9), where the old Cato (who addresses his friends as optimi adulescentes, 39) remarks on the pleasures of an old age stipata studiis iuventutis, ready to teach the young. Still, Quintilian’s words do remind us vividly of the boni iuvenes of his first preface (1.1.7): Quintilian’s own pupils, whom in another place (7.3.30) he calls adulescentibus meis (meos enim semper adulescentes putabo). These studious young men will, as we shall see, reappear at the very end of the chapter, and of the book. Two further connections with Cicero suggest themselves. Quintilian’s stress on the physical failings of the old, the increasing weakness of voice, lungs, and ² Where the length of Quintilian’s career is given as twenty years. Kennedy (1998), .153 had the pleasant idea that Quintilian consciously connected this period with the twenty years laid down by Augustus for legionary service; but it would seem that by the time of the Flavians that service had been extended to twenty five years. ³ Compare Mart. 12.pr.: in hac provinciali solitudine, ubi nisi etiam intemperanter studemus, et sine solacio et sine excusatione secessimus.

236

Papers on Quintilian and Ancient Declamation

general health, is taken not just from life but from De oratore, where Crassus foresees (1.199) a time when his powers will begin to fail (deficere), and also from De senectute, where lungs, strength, and voice are mentioned (28). It is noticeable that both Cicero and Quintilian gloss over the possibility of mental decline and even collapse; Quintilian just says that knowledge increases with age (12.11.2); and later he will tell us how Cato learnt Greek for the first time as an old man (12.11.23, from De senectute 3 and 26). Secondly, Quintilian wants his orator to make for port with vessel unharmed. The idea of retirement as a port after the storms of one’s career appears in De oratore (1.255): what Crassus saw as a solitude to be feared, Antonius thinks of as a quiet harbour to look forward to. Quintilian sides with Crassus here. The retired orator must keep busy; his port is like the one hankered after by Cicero himself in the last years of his life (Brut. 8), a portum . . . non inertiae neque desidiae, sed oti moderati atque honesti. There must, then, be fruits of old age (12.11.4).⁴ Quintilian loyally mentions Cicero’s idea of giving legal advice, but his other suggestions are designed to leave something tangible behind: books on history (written, as he says, for posterity), philosophy, and, put in to cover himself, rhetoric. The climax comes with talk of going on ‘shaping’ the young (formabit, 5): Quid . . . est honestius quam docere quod optime scias? (6). And Cicero the preceptor of the young is made an explicit exemplum. The retired orator is motivated amore quodam operis: nemo enim minui velit id in quo maximus fuit. Quintilian’s own voice comes through clearly. The work must go on, and his retirement is not to be a rest. He will teach by writing, and by going on talking to the young. We have seen the strong closure at the end of 7. Still, when Quintilian does go on, it is to a section (8) that links neatly with what precedes. He is confident, he says, that he has, as well as he could, brought the fruits of his knowledge and research to the notice of all who wished to learn: that puts him, in his way, alongside Cicero the teacher of youth. Id viro bono satis est, he now says of himself, docuisse quod scierit. Being a good man is one attribute of an orator; and the echo of 6 is obvious and intended: Quintilian has himself done what he recommended the ideal orator should do. But now a dark note: ‘I fear however’ (9). The link is: I have taught as well as I could; but is my programme practicable? This is the new topic. 18–19, an assertion that it is our own fault if we find we have no time to complete such wide studies, and 21–4, an inspiriting list of men who have in fact mastered vast tracts of knowledge, are straightforward enough. But we need to find some clue to the labyrinth of 9–17. It seems not to have been noticed that a key element of the passage is recapitulation of the topics of the Institutio as a whole, and especially of those ⁴ Cf. Plin. Ep. 3.7.14: quidquid est temporis futilis et caduci, si non datur factis . . . , certe studiis proferamus, et quatenus nobis denegatur diu vivere, relinquamus aliquid quo nos vixisse testemur.

Approaching the End: Quintilian 12.11

237

of the advanced programme of studies prescribed in the early chapters of Book 12. Nor is recapitulation inappropriate to this final chapter. In a speech, so went the precept, the peroration should link recapitulation with emotional appeal. I shall come back to the emotions later. A preliminary statement of the main lines of Quintilian’s programme is given in 12.11.9, in the shape of a distinction between the weight (magna) and the number (multa) of the items on his syllabus. The former covers the requirement eundem virum bonum esse et dicendi peritum; this summarizes the essence of 12.1 (which began with a statement of Cato’s famous definition of the orator). The latter takes in earlier features of the Institutio, the ‘arts’ learned by boys (referring to 1.10, where see 1: Nunc de ceteris artibus quibus instituendos . . . pueros existimo) and ea quae de eloquentia tradebantur (covering Books 2 to 11), but subordinates them (note praeter and adiecerim) to the advanced programme of Book 12, from which Quintilian for the moment picks out praecepta morum (12.2, where in 1 notice mores . . . oratori studiis erunt excolendi) and scientia iuris civilis (12.3; see its opening words). After bracing general remarks in 12.11.10, Quintilian goes through the programme again, in more detail. The virtue of the orator is so important⁵ that Quintilian expands on it in 11–13, before coming in 16–17 to the other items in the syllabus (cetera, 13). These recapitulations are subordinate to arguments designed to allay the fears⁶ of those who find Quintilian’s programme too demanding. Quintilian is concerned first (11–13) to show the practicability of the attainment of virtue. At the start of 12.2 we were told that virtus, etiam si quosdam impetus ex natura sumit, tamen perficienda doctrina est (we may be reminded here of Seneca’s remark in De otio 4.2 that one thing the retired could enquire into is natura an ars bonos viros faciat). Nature and art accordingly intertwine in an unschematic way in our passage. Thus in 12.11.12: natura enim nos ad mentem optimam genuit, adeoque discere meliora volentibus⁷ promptum est ut vere intuenti mirum sit illud magis, malos esse tam multos. Nature and precept, therefore, work together in the same direction. But nature is ultimately the more important, and it is given the climax of the section (13): Nam ut aqua piscibus, ut sicca terrenis, circumfusus nobis spiritus volucribus convenit, ita certe facilius esse oportebat secundum naturam quam contra eam vivere. Again we may note the closural effects: the examples of fish, animals, and birds leading up to humankind, the mild word play of secundum naturam and contra eam. The purport of 9–13 is quite clear: nature has prepared us in

⁵ Quod prius quodque maius est (11) echoes 12.1.1: id quod et ille (sc. Cato) posuit prius et ipsa natura potius ac maius est. ⁶ Perhorrescant, 12.11.9; at § 12 we should perhaps read praemetuuntur for the corrupt prae muntur: cf. Winterbottom (2006) [= A.19 below]. ⁷ Cf. 12.11.11: (becoming good) voluntate maxime constat.

238

Papers on Quintilian and Ancient Declamation

advance, and it is not difficult to learn what needs to be learned. To become good is natural, and moral precepts work with the grain of our natures. The closure at the start of 12 paves the way for defence of the practicability of the rest of the programme, cetera. In 16, Quintilian, rather oddly, again mentions learning moral precepts, together with getting experience of the forum:⁸ this alludes to 12.6, on the age for starting to plead in court. In the lacuna before cognoscere (12.11.16) we surely need a mention of law, as in 12.11.9 above, to recall 12.3. Kiderlin was therefore right to add ius civile.⁹ The next item, rerum cognitio (12.11.17), duly recalls the short chapter 12.4, on historiae, to be understood in the grammarians’ sense, of the background knowledge of texts from which the orator can accumulate exempla. We may observe how in 12.4 Quintilian stresses that one does not need to wait till old age to win authority here: study (that is, the reading of books) will give you an old man’s experience before you grow old. So here in 12.11.17: Rerum cognitio cotidie crescit; but experience is to be supplemented by bookreading. The argument is puzzlingly set out. But Quintilian seems to exclaim: ‘How many books we need to read, even if we set ourselves to read only useful material!’¹⁰ But we are not to despair, for, as is stressed in 18–19, we waste an enormous amount of time that might be devoted to study. It does now begin to look as though Quintilian is starting to move away from the position that the programme can really be completed in adolescence.¹¹ This is suggested especially by the mention of ‘anxious calculation of accounts’, not something adolescents normally much worry about. I shall return to this point. Before Quintilian started this recapitulation of the programme of Book 12, he had already, in 10–11, laid the foundations of the emotional appeal of this peroration. The human mind is capable of so much. Even minor arts can aspire to discover the courses and numbers of the stars, and almost measure the universe itself. The hyperbole is emotive, and contains an argument a fortiori: if a mere geometer can do so much, then, we the students of the grandest of all arts, can do anything! The goal is great, no labour to reach it is to be grudged. And there is emotion too in 18–20, combined with further scorn for the professors of other arts. In a very characteristic passage, Quintilian gives us an impassioned statement of the human tendency to waste time. Most other things a man, especially a young man, might wish to do are ruled out ⁸ In foro nos experiri implies not just ‘attending real cases’ (Austin), but ‘trying oneself out’ in the making of speeches; cf. Plin. Ep. 7.4.3: Latinos elegos . . . feci. Expertus sum me aliquando et heroo. ⁹ Winterbottom (1970a), ad loc. Perhaps more has dropped out. ¹⁰ But 17 remains obscure: et tamen . . . omnia sounds like an imagined objection, that the reading list is too long. Perhaps, then, emend et tamen to at enim (answered by sed, as in ‘Quint.’ Decl. min. 371.5). ¹¹ Note earlier (13): etiam si aetatem nostram non spatio senectutis sed tempore adulescentiae metiamur (with the implication: how much more can we do if we employ old age to the full!).

Approaching the End: Quintilian 12.11

239

with distaste. Some of them are familiar targets of a satirist’s criticism, and are given adjectives to point the moral: morning visits to one’s patron (vanus salutandi labor), undue concern for one’s body (insanam corporis curam). But Quintilian could be thought to be banning almost everything on which a Roman (or anyone else) might spend his leisure: conversation (fabulae), shows, dinner parties, travel, concern for one’s estates, sex (and wine too, unless we delete et vinum).¹² This sort of complaint is not to be dismissed, in the way Austin dismisses it, as just a locus de saeculo. It is that, but it is used to a purpose: we waste time we can ill afford to waste. Quintilian is surely inspired by, even corrects, a passage of Seneca’s De brevitate vitae (1.4): non accipimus (v.l. accepimus) brevem vitam sed facimus (MSS fecimus), echoed in our chapter (12.11.18) by breve nobis tempus nos fecimus.¹³ Seneca’s treatise gives parallel examples of how time is wasted (e.g. 2.1; 7.1; 13.1). His purpose, though, is to encourage us to find time for philosophy, particularly—it may be noted—in old age (cf. especially 7.4; 19.1): whereas Quintilian is thinking of the need to make time for study in youth, and (it begins to appear) throughout one’s life. But the general argument is remarkably parallel. The similarity illuminates Quintilian’s praise for Seneca as an egregius . . . vitiorum insectator in 10.1.129. Conversely, we should look upon this passage of Quintilian, and others like it, as being signs of his desire to ensure that the student of rhetoric is trained in morals also. We may notice how, back in 12.11.4, he said that the retired orator will write a rhetorical handbook or give expression to the precepts of philosophy. Quintilian’s book does both. So here: when he talks about time-wasting, he is not ranting: he is teaching. We saw that, before Quintilian turned to recalling the details of his advanced programme, he gave some general encouragement (12.11.10). And he gave it, we now discover, in terms that look forward to later parts of the chapter. The prize is great—and in 29 we find that the reward is not, or not primarily, a matter of money and praise. The human mind is capable of extraordinary feats—and in 21–4 Quintilian unobtrusively uses his own wide reading to give examples of great men who have mastered many skills. These are exempla to prove his point (again youth is forgotten, as we shall see); and a list that started with Homer ends with calculated bathos (24): if the mediocre Cornelius Celsus can master so vast a field, anyone can! We come to the last paragraph, 25–30. Quintilian uses the phrase exhortatio studiorum, but exhortation to study has been an important element of the

¹² Other passages of the Institutio (1.11.15; 1.12.18; 6.2.30; 12.1.6) cite yet more distractions of which Quintilian disapproves: exercise in the Campus Martius; dicing (approved for old men by Cato in Cic. Sen. 58); the hunt; the palaestra; sleep [siestas, arguably], and, while awake, day dreaming. ¹³ Cf. André (1983), 19 n. 63.

240

Papers on Quintilian and Ancient Declamation

Institutio,¹⁴ not least earlier in our chapter. The pupil is now urged to work hard in the confidence that what he is aiming for is not, as it might seem, unattainable; being an orator is (the argument proceeds) something that is possible as well as eminently desirable. The chapter, and indeed the whole of Book 12, is being tied together. The assertion (25) that nature allows the attainment of such heights looks back to similar words in 12.1.31. Oratory is the opus pulcherrimum (12.11.29): that phrase had been used in the first section of our chapter, and early in 12.1 too (4). And the very fact that Cicero had achieved so much, implied earlier and now stated more openly, should be an incentive to try at least to come near to him. It follows that 21–30 is an integral part of the argument that started at 9. The programme is attainable, and the goal is very well worth striving for. Who is being encouraged in this chapter? In the proem to Book 12 (3) we are told that the student has been a dicendi magistris dimissus; he is now either going along under his own steam, aut maiora sibi auxilia ex ipsis sapientiae penetralibus petit—a grand way of alluding to a philosopher’s school.¹⁵ Either way he is out of Quintilian’s hands, and that is why exhortation is so necessary. These are, as it were, the words the Master might have used at the end of the last semester, when addressing pupils who must now bring their own motivation to private studies. But now, writing a book that not only students would read, Quintilian addresses others too. We find him criticizing other rhetors (15). And one might guess that he also had fathers in mind, fathers glad enough not to be paying any more fees to a rhetor, but in need of some encouragement themselves to support sons lounging round the house, and occasionally picking up a book on moral philosophy. Still, Quintilian is mainly talking to the young in this part of the chapter, even if he sometimes (e.g. 12 and 14) uses the first person plural to soften his exhortations to them. But is there perhaps some element of self-exhortation here too? We have seen how in 2.12.12 Quintilian talked of writing the Institutio as a ‘consolation’ of his leisure. Now the long book is nearly finished, and the dread question ‘What next?’ looms. Though retirement is not mentioned after 12.11.7, there is continued allusion to old age. It is striking how the examples amassed in 21–4 (Gorgias, Aristotle, Cato) dwell on long periods of time and on activity in advanced age. And the mention of Cato takes us back, as we have seen, to the De senectute, and to the idea elaborated there that one should never stop studying. Cicero (26) alludes to famous words of Solon, whom (he says) ‘we see boasting in his poetry that he grows old learning something new every day’: γηράσκω δ’ αἰεὶ πολλὰ διδασκόμενος (fr. 18 West² [= 28 Gentili–Prato²]). Finally, in 19–20, Quintilian, after listing ways we find ¹⁴ Thus the (Ciceronian) ideal of the perfect orator, first broached at 1.pr.9, is designed to set a goal towards which the student may strive. ¹⁵ Cf. Quint. 12.2.6: Cicero . . . testatur dicendi facultatem ex intimis sapientiae fontibus fluere.

Approaching the End: Quintilian 12.11

241

to waste time (a list which, as we have seen, sometimes sounds more appropriate to an older man than to youths), goes on to say that if all that misspent time were spent on study, we should find we had a long enough life and plenty of scope for learning. Nec vero si geometrae et musici et grammatici . . . omnem suam vitam . . . in singulis artibus consumpserunt, sequitur ut pluris quasdam vitas ad plura discenda desideremus. All this is not much to the point if Quintilian is merely encouraging the young to master moral philosophy and law; rather, he has himself, and his retirement, much in mind. He distances himself from those who stop learning their own subject, and yet are content to know only that; he, it is implied, will go on studying rhetoric, but not rhetoric alone. In Book 1 (1.4.5) he had said of grammar that it is necessaria pueris, iucunda senibus, dulcis secretorum comes. He is telling himself that work must go on: that there is life after the Institutio, even if it is only devoted to reading yet more, in rhetoric and in other arts. Austin thought of much of the chapter as a postscript. But the real, and very short, postscript comes in the last section of all (31). The reader most obviously envisaged here is the dedicatee of the book, Marcus Vitorius. But Quintilian’s thoughts are with the young, the studiosi iuvenes whom he hopes will carry out his programme even after they have left his school. No less are his thoughts with his book, which he hopes will prove useful to the young. But if it does not, he knows it will give them ‘good intentions’ (bona voluntas). For Quintilian, being good is what matters. And he may have the last word on a teacher’s retirement (12.11.8): id viro bono satis est, docuisse quod scierit.

19 Quintilian 12.11.11–12 Nam id quod prius quodque maius est, ut boni viri simus, voluntate maxime constat: quam qui vera fide induerit, facile eas idem quae virtutem docent artis accipiet. Neque enim aut tam perplexa aut tam numerosa sunt quae †praemuntur† ut non paucorum admodum annorum intentione discantur.

Quintilian argues¹ that to become good is not so hard as it might seem to those who have read his precepts in 12.2, a chapter that began with the admonition that virtue, even if grounded in nature, tamen perficienda doctrina est: mores ante omnia oratori studiis erunt excolendi atque omnis honesti iustique disciplina pertractanda, and continued with detailed advice on what he now calls ‘the arts that teach virtue’. Critics, seeing no sense in praemuntur, have at times looked for a verb meaning ‘are handed down’ or the like (promuntur, praecipiuntur, traduntur). On a different tack, I suggest praemetuuntur. The studies that Quintilian demands are feared in advance, though they will seem less frightening once they are embarked upon. This takes up a remark in 12.11.9: Vereor tamen ne aut magna nimium videar exigere, qui eundem virum bonum esse et dicendi peritum velim, aut multa, qui tot artibus in pueritia discendis morum quoque praecepta et scientiam iuris civilis praeter ea quae de eloquentia tradebantur adiecerim, quique haec operi nostro necessaria esse crediderint velut moram rei perhorrescant et desperent ante experimentum.

Praemetuere is found only occasionally in classical Latin, but in respectable authors: Lucretius, Caesar, Virgil, and Phaedrus, oddly always in the present participle. Quintilian uses other rare compounds in prae-; note praecogito (once each in Livy and Seneca), praeformo (once each in Scribonius Largus and Silius Italicus), and praemollio (Quintilian’s are the first two instances). Especially pertinent is Quintilian 4.5.5: Nam est nonnumquam dura propositio, quam iudex si providit non aliter praeformidat quam qui ferrum medici prius quam curetur aspexit. Elsewhere that verb is only found once, in Silius. [Classical Quarterly NS 56 (2006), 324 5] ¹ For discussion of the whole context cf. Winterbottom (2005) [= A.18 above].

20 Declamation and Philosophy* Two familiar ancient sources formulated models for the history of the relationship between oratory and philosophy from the fifth century  on. Philostratus, writing in the first half of the third century , distinguished two Sophistics.¹ The ‘Old’ Sophistic, founded by Gorgias, he sums up as ‘rhetoric philosophizing’. Unlike philosophy itself (Philostratus means philosophy as it developed later), it assumed the truth of what it spoke of; and it dealt with philosophical themes ‘diffusely and at length’ (Philostratus here implies another contrast with developed philosophy). The examples given are courage, justice, heroes and gods, and the manner in which the universe received its shape. The rather later ‘Second’ Sophistic, on the other hand, was concerned with type characters, poor men and rich men, war heroes and tyrants, and treated τὰς ἐς ὄνομα ὑποθέσεις, particular cases involving named persons. Hypothesis is the ordinary term for the theme of a declamation. The rhetorician Hermagoras had in the Hellenistic period contrasted it with the general thesis. And though Philostratus does not use that term, he is clearly thinking of the Hermagorean distinction. The fifth-century sophists spoke on abstract ‘philosophical’ topics, theseis. Later sophists, what we tend to call rhetores or declaimers, treated particularized though fictional cases, hypotheseis. The other model was that of Cicero.² He looked back to a period when philosophy and rhetoric were united, taught and practised by the same persons. But then came a split, attributed to Socrates, as a result of which the two had been lastingly separated. Philostratus’ model does not explicitly state that the first sophists concerned themselves with all topics, not just

[Classica (Brasil) 19 (2006), 74 82] * I have done little more than lightly annotate a paper I gave in Ouro Preto, which was in tended to be provocative. Two friends were kind enough to comment on it: Tobias Reinhardt beforehand, Jaap Wisse afterwards. The latter, who expressed strong reservations, hopes to re turn to the subject himself. ¹ Philostr. VS 480 4. ² Cic. De orat. 3.60 1.

244

Papers on Quintilian and Ancient Declamation

philosophical ones. But that is the implication of the story both he³ and Cicero⁴ tell, of how Gorgias would speak on any topic suggested by his audience. In that case, Cicero’s model is in essentials identical to Philostratus’. But whereas Philostratus talks of sophists, Cicero, no friend of rhetores, talks of oratory and philosophy. Both, however, see a crucial break in the fourth century. For Cicero, in blaming it on Socrates, is really thinking of the influence of the anti-rhetorical dialogues of Plato (d. 347), while Philostratus thinks of Aeschines (d. c.322) as founder of the Second Sophistic. Cicero himself had changed his mind about philosophical topics. In his youthful De inventione,⁵ perhaps echoing his philosophy teachers, he was sharply, and no doubt unfairly, critical of Hermagoras for including in the field of the orator such (Old Sophistic) questions as ‘Is anything good except honestas?’ and ‘What is the shape of the world?’ But in his maturity, he was highly sympathetic to some kind of rapprochement between oratory and philosophy. We may distinguish three aspects of his mature view. Two concern oratorical, the third educational practice: (a) Cicero came to see that even the forensic orator could not avoid talk of abstract matters. Thus he makes Crassus say⁶ that passages ‘very often’ come up in which an orator must talk of the gods, or the ius gentium, or the various virtues. He represents the philosophers as crying out that all this is their field. But there was the crucial matter of style. If orators discussed such topics, it was ‘with all pleasantness and gravity’, in contrast to the plainness of the philosophers:⁷ much the same contrast as we saw implied in Philostratus. Later, in the important preface to the Paradoxa Stoicorum, Cicero commends the Stoic Cato for being able to make his ‘grave’ passages from philosophy acceptable even to senate and people: something easier for the less doctrinaire Cicero. In praising Cato, Cicero praises himself. Both brought philosophical themes into oratory. They will not, perhaps, have been the only Roman orators who did so. (b) In particular, Cicero became convinced that it was the mark of a great orator to move freely, where possible, between the general and the particular.⁸ It is not clear how he came to this view; perhaps from reflection on what he had instinctively done in his own speeches. We may think, for example, of the generalized question discussed in Pro Milone 7–11. (c) Finally, so far as education was concerned, Cicero was (at least in the case of his nephew) happy enough that it should be conducted in the usual manner, involving declamation on imaginary but particular themes. Yet, he writes to his brother in 54 ,⁹ you know that my preferred method is ‘a little more scholarly and philosophical (θετικώτερον)’; and when the young man is

³ Philostr. VS 482. ⁶ Cic. De orat. 1.56. ⁹ Cic. Q. fr. 3.3.4.

⁴ Cic. De orat. 1.103. ⁷ Cic. De orat. 1.57.

⁵ Cic. Inv. 1.8. ⁸ See e.g. Cic. Orat. 45.

Declamation and Philosophy

245

with Cicero in the country, he looks forward to initiating him in this deeper kind. Five years later Cicero¹⁰ was declaiming on a number of general questions for practice, choosing topics poignantly relevant to his own position (the one he mentions first is: ‘Should one remain in one’s country under a tyranny?’). He spoke on both sides of these questions. That was the philosophers’ way. But there was practical point: Cicero was trying to make up his mind. This three-pronged attempt to reconcile oratory and philosophy Cicero represents as a novelty of his own. A modern scholar, Jakob Wisse, has used the words ‘idiosyncratic and surprising’.¹¹ But I shall suggest that Cicero was doing little more than bring into the realms of theory what had long been practised under the teachers of rhetoric who had, as he says in the letter to Quintus I have quoted, trained him in declamation. It is in the nature of the case that we know little of declamation before Seneca the Elder wrote in the early Empire. But the rhetores were a conservative lot, and I see no reason why we should not in this matter argue back from Seneca, and from even later evidence, Greek and Latin. In particular, there seems no doubt that progymnasmata, the preliminary exercises in rhetoric, went back well before the Christian era. I touch first on the inclusion of abstract material in a speech. Seneca on several occasions draws our attention to passages of what he calls philosophizing, some spoken by an avowed philosopher, Fabianus, others by an admirer of Fabianus, the older C. Albucius Silus. I give a single example. Seneca introduces¹² with the words Albucius philosophatus est a passage on slavery: no one is free or slave by nature, for these are only names imposed on men by Fortune. It is easy to find parallels for this idea in the Younger Seneca, a practising philosopher.¹³ But one did not have to be a philosopher to declaim on such a theme; there is an eloquent passage on the topic in Decl. mai. 13.8 (273.8–274.7 Håkanson), and the author of the Minor Declamations puts it succinctly: sortimur genus, non eligimus.¹⁴ Many declamation themes involved slaves, and the topic was inevitable. It is telling that¹⁵ Alcidamas, back in the First Sophistic, remarked that ‘the gods made all men free, and Nature has enslaved no one’. To a modern ear all this does not sound exactly like philosophy. But that is just the point. At this level the concept of a reconciliation of rhetoric and philosophy is vacuous, for such moralizing permeated most types of discourse. When Mayor gives parallels for Juvenal’s utque animas servorum et corpora nostra / materia constare putet paribusque elementis,¹⁶ he mentions, among others, Philemon, Petronius, and Cyprian. And it is at this level that Cicero is talking. When he says that orators often ¹⁰ Cic. Att. 9.4. ¹¹ Wisse (2002), 397. ¹³ Cf. Winterbottom (1974b), .138 9 n. 2. ¹⁵ As we know from Aristot. Rh. 1373b.18.

¹² Sen. Con. 7.6.18. ¹⁴ ‘Quint.’ Decl. min. 388.22. ¹⁶ Juv. 14.16 17.

246

Papers on Quintilian and Ancient Declamation

speak, say, of the gods, he does not mean a technical Stoic or Epicurean disquisition; he means an uplifting passage on Iuppiter Optimus Maximus. A second passage where the Elder Seneca catches Albucius philosophizing may serve as a transition to Cicero’s call for free movement between general and particular. Cestius is said¹⁷ to have criticized Albucius for treating as problemata philosophumena what Cestius thought were merely minor offshoots of the quaestio: An haec (a Vestal who survived precipitation from the Tarpeian Rock) deorum adiutorio servata sit; namely An di immortales rerum humanarum curam agant. Etiamsi agunt, an singulorum agant. Si singulorum agunt, an huius egerint. I shall return to the substance of the quaestio here, the role of providence in human life; meanwhile, we should note how Albucius moves from the general (Do gods care?) to two successive stages of particularity (Do they care about individuals? Did they care about this woman?). This movement is highly typical of the procedures of declaimers in Seneca, most of whom were not philosophers at all. They will have been taught to structure their arguments like this at the school of the rhetor; and those who wonder what Hermagoras did with his distinction between thesis and hypothesis may look for their answer here. Nor is Seneca our only source. In the Minor Declamations, we find (299.2) the Master prescribing as a first question An utcumque (‘under all circumstances’) sepultum eruere non liceat; there follow two more of increasing particularity, and yet others that he calls speciales quaestiones. Some philosophical teaching may lie behind this sort of thing. But its roots lie deep in the mentality of classical writers; recall how Pindar moves from maxim to example, or Horace¹⁸ from Miserarum est neque amori dare ludum . . . to Tibi qualum Cythereae puer ales, tibi telas . . . Here again Cicero is not really innovating; declamation already did what he required of the orator, and the classical mind was well attuned to such a course. I come now to theseis. When Cicero practised them himself, he did so, as we have seen, in the proper philosopher’s manner, taking first one side and then the other. But such pro and contra treatment was not essential, as we see when we consider the practice of the rhetores. For they made thesis one of the progymnasmata that, from some (I take it Hellenistic) date, were used to introduce students to rhetoric; and it was surely not expected of a pupil at this tender age that he should speak on both sides of such topics on the same occasion. And the examples given by, for example, Libanius point that same way. More important, the link between progymnasmatic thesis and philosophers was clear to rhetorical writers. Thus Quintilian’s remark¹⁹ that the antiqui spoke theseis is to be glossed by a passage in Book 12,²⁰ where he comments on the way in which the Peripatetics virtually (fere) instituted the speaking of theseis for training in speaking. More explicitly, the Greek rhetor ¹⁷ Sen. Con. 1.3.8. ²⁰ Quint. 12.2.25.

¹⁸ Hor. Carm. 3.12.1 ff., 4 ff.

¹⁹ Quint. 2.1.9.

Declamation and Philosophy

247

Theon, probably writing in the first century , tells us that examples of theseis are easily to be found in Aristotle and Theophrastus, who wrote books with that title.²¹ When Cicero commended his ‘more philosophical’ method for the education of his nephew, he was pushing at an open door. It may be objected that the sort of thing a child of 12 might come up with, even with a lot of prompting, on a general topic would hardly be sophisticated enough to count as philosophical. All the same, the exercise did help a boy to think and talk in general terms. It is in any case crucial to remind ourselves that practice in the thesis did not stop at the progymnasma stage. On the contrary, like all the exercises that made up that early course, it was only introduced there. When a boy came on to declamation proper, he would discover that the thesis was a building block that he could insert into the fabric of a full speech (I shall come back to this matter). Similarly, it might prove useful when he eventually made a speech in court. This is what Quintilian means when he says that all the early exercises ‘crop up in forensic cases’.²² Here again, Theon confirms: he says²³ that we may look for instances of thesis in, for example, two (apocryphal) speeches by Lysias. And Quintilian points to the use made by Cicero in his Pro Murena of the thesis comparing law and soldiering,²⁴ as well as to the way in which the great orator practised oratory in his maturity by speaking on such general topics.²⁵ In this way, the thesis could play a role in exercising for oratory from childhood to old age. It may be added that the old Seneca²⁶ encouraged his son Mela to study eloquence if only as a means to other arts, citing the case of Fabianus, who went on from declamation to dialectic (he was equally obscure, it seems, in both departments). It follows that it was not a matter of choosing between rhetoric and philosophy. It was possible and even inevitable to start in the school of the rhetor before moving to the school of the philosopher: a move, indeed, that might well take a student from Rome to Athens. Cicero, and Quintilian after him, were over-ready to polarize the subjects. We have a good deal of information on the way in which general questions were employed in declamation. Theon²⁷ distinguished two kinds of thesis: (a) practical, e.g. ‘Should one marry?’, and (b) theoretical, e.g. ‘Do the gods take thought for the world?’²⁸ It is obvious, he goes on, that the practical ones are ‘more political’ and are ‘in accordance with the “character” of rhetoric’, while the theoretical ‘are suitable rather to the philosophers’. But, he adds ²¹ Theon RhG .69.1 3 Spengel (= p. 13 Patillon[ Bolognesi]). ²² Quint. 2.1.10. ²³ Theon RhG .69.3 13 Spengel (= p. 13 Patillon[ Bolognesi]). ²⁴ Quint. 2.4.24. ²⁵ Quint. 10.5.11, perhaps with a reference wider than just to the sort of thing we saw in Cic. Att. 9.4. ²⁶ Sen. Con. 2.pr.4 5. ²⁷ Theon RhG .121.7 26 Spengel (= pp. 83 4 Patillon[ Bolognesi]). ²⁸ He gives detailed headings for this at pp. 126.3 128.24 Spengel (= pp. 91 4 Patillon [ Bolognesi]).

248

Papers on Quintilian and Ancient Declamation

significantly, it is possible for rhetoricians to treat them too. It was possible, we may note, because the rhetores made it possible. Albucius was criticized²⁹ for asking in one controversia: ‘Why does a cup break if it is dropped, but a sponge doesn’t?’ (on which Cestius commented: ‘Go and listen to him tomorrow: he’ll tell you why thrushes fly and pumpkins don’t’). The objection to this question was presumably that it was irrelevant to the theme. But themes could be so designed as to make ‘philosophizing’ desirable or inevitable. To take the example of the role played by the gods in the world. Quintilian comments interestingly on it.³⁰ The thesis ‘Is the world ruled by providence?’ is, he says, part of the eternal battle between Stoics and Epicureans. As for the particular question, it would (Quintilian says) vary according to the case and the type of forecasting of the future it involved. Seneca’s third Suasoria may provide an example. The theme is ‘Agamemnon deliberates whether to sacrifice Iphigenia, it being Calchas’ pronouncement that otherwise it is impossible to set out to sea.’ The particular case here concerns augury; note how Pompeius Silo says (4) that even if some methods of divination work, no belief should be placed in augury. As for the thesis, we may turn to the division by Cestius (3): ‘(a) the gods do not make their wishes felt in human affairs; (b) even if they do, men cannot know their will; (c) even if men do know it, the fates are irrevocable.’ (It may be observed how vividly this recalls the argument of Gorgias’ On Nature back in the First Sophistic: nothing exists; even if it does, it is incomprehensible; even if it is comprehensible, it is not communicable. Plus ça change . . . ) Less schematically, Arellius Fuscus argued that the delay at Aulis was due to natural causes, sea and wind; as for the will of the gods, that was not to be known by mortals. Seneca of course gives us no more than short extracts from declamations, or summaries of their arguments. We have to turn to the fourth of the Major Declamations, of uncertain date but much later than Seneca, to see the thesis argued in detail. The theme makes discussion of the point inevitable, and was chosen, surely, with that in mind. A man opposes the suicide of his son, of whom it had been predicted by an astrologer (mathematicus) that he would go on from being a war hero to killing his father. The declaimer, speaking as the son, devotes two pages to the general topic.³¹ We may note two points in particular: (1) The son represents his father as having argued that there is no fate, only chance; but that even if providence rules all things (etiam ut providentia regantur), it is not open to human comprehension. The continuity of declamation from the fourth century  to (perhaps) the fourth [³²] is splendidly obvious. ²⁹ Sen. Con. 7.pr.8. ³⁰ Quint. 5.7.35. ³¹ [Quint.] Decl. mai. 4.13 14 (74.14 77.4 Håkanson).

³² [But see the next footnote.]

Declamation and Philosophy

249

(2) Much of the wording of these paragraphs is taken almost without change from Concerning the Error of Profane Religions, which Firmicus Maternus wrote in the fourth century after his conversion to Christianity.[³³] Declamatory topics were not to be rendered out of date by a mere change of the empire’s religion. It might be thought that more ‘physical’ theseis could not readily be introduced into declamation—such topics as Hermagoras’ ‘What is the shape of the earth?’ Yet look at the first Senecan Suasoria, in which Alexander, having reached India, is advised on whether or not he should proceed to sail across the Ocean. The philosophers have a field day here, and, even if wordy, they are not being irrelevant. Fabianus says that those who have succeeded in calculating the movements of the stars, ‘men to whom no part of the universe is a mystery, are still in doubt as to the Ocean’ (4); and he proceeds to specify competing ideas about the Ocean’s nature. More concisely, Albucius asserts (3) that ‘Even the earth has its end; the very universe sets somewhere.’ But anyone talking on such a theme could hardly avoid taking some explicit stand on the nature of the world. It was probably not a philosopher who said (1): ‘Beyond everything, the Ocean; beyond the Ocean, nothing.’ The theme, attractive in other ways, cries out for some treatment of this ‘physical’ thesis. Quintilian, as appears from two passages,³⁴ thought of all the progymnasmata as being in some sense general, presumably because they did not treat hypotheseis, particular cases. But the most obviously general of the whole lot was the locus communis or commonplace. And of this I should say a little. A precise distinction between this and the thesis is formulated by Theon,³⁵ who makes the commonplace the amplification of something that is agreed, aiming at the punishment of an offender, the thesis the amplification of something that is disputed, aiming at persuasion. This distinction was not universally accepted. While Quintilian thinks³⁶ of the discussion of, for example, witnesses in general as being a locus, Theon³⁷ calls the comparable discussion of rumour and torture thesis. What matters for our purposes is that the progymnasmatic exercise of commonplace, like that of thesis, was something that gave a student practice in generalized, if not abstract, thought; and commonplace, too, was a building block for declamation and found its way into real oratory.³⁸ Thus, the early exercises lay a firm foundation, by no means restricted to theseis, for generalizing oratory. ³³ [The Major Declamations are now convincingly dated before the fourth century , so it was quite certainly Firmicus Maternus who drew on them, not the other way round.] ³⁴ Quint. 2.1.9; 2.4.36. ³⁵ Theon RhG .120.16 25 Spengel (= pp. 82 3 Patillon[ Bolognesi]). ³⁶ Quint. 2.4.27. ³⁷ Theon RhG .69.14 17 Spengel (= p. 13 Patillon[ Bolognesi]). ³⁸ Theon RhG .67.31 68.3 Spengel (= pp. 11 12 Patillon[ Bolognesi]) gives examples from Demosthenes.

250

Papers on Quintilian and Ancient Declamation

After all these abstractions, I turn finally to some real people: just as Philostratus, having set up the model of the two Sophistics that I began with, gave potted biographies of individual sophists. We may note that he says that earlier in the Second Sophistic the name ‘sophist’ was given not just to distinguished orators but also to philosophers who ‘expressed themselves with fluency’;³⁹ and it is with such persons that he begins. Six precede Dio of Prusa. The best known is Carneades, on whose forceful oratory Philostratus comments.⁴⁰ It is well known from other sources that on an embassy to Rome in 155 he spoke for and against justice on consecutive days. We think of him as a philosopher; but this sort of bravura performance reminds us how closely such persons followed in the footsteps of the old sophists, shocking audience with their paradoxical assertions. Similarly, the much earlier Eudoxus of Cnidus is said by Philostratus⁴¹ to have had an ornate style and to have improvised well; Philostratus talks of his links with the Academy, but he might also have mentioned his remarkable versatility in philosophy as geometer and astronomer. We may add to Philostratus’ list Apuleius of Madaura, who was both Platonist and declaimer, not to speak of law-court orator and novelist. We hear of Euphrates of Tyre from Latin sources as well as Greek. Both Philostratus⁴² and Pliny⁴³ regarded him as a philosopher. Pliny describes him as disputing subtiliter graviter ornate, often attaining the sublimity and width of Plato with a style that was copiosus et varius, dulcis in primis: all this despite his formidable appearance, the tall frame, the long hair, the enormous white beard. It would seem that the borders between philosophy and even dialectic on the one hand, and rhetoric on the other, were easily crossed. Seneca⁴⁴ frowned on the effusive style of the philosopher Serapio (‘more came to him than a single voice is capable of uttering’). But, he reflects, such licentia is a Greek fault.⁴⁵ And he praises for a proper Roman eloquence our old friend Fabianus,⁴⁶ whom he commends for ‘disputing’ with facilitas rather than celeritas. The same philosopher is praised in another of Seneca’s letters⁴⁷ for sentences that are not ‘forced into epigram’ but spoken latius. Seneca is contrasting different types of oratorical style as employed by philosophers; he is not (as we saw Crassus doing in the De oratore) associating one style with philosophy, another with oratory. Cicero, in short, seems much to exaggerate the split between rhetoric and philosophy. We have seen evidence from both before and after his day to suggest that the old First Sophistic tradition of eloquent philosophy and philosophical eloquence was by no means forgotten. It may be that Cicero exaggerated in order to show in the best light his own proclaimed achievement ³⁹ Philostr. VS 484. ⁴² Philostr. VS 488. ⁴⁶ Sen. Ep. 40.12.

⁴⁰ Philostr. VS 486. ⁴³ Plin. Ep. 1.10. ⁴⁷ Sen. Ep. 100.5 6.

⁴¹ Philostr. VS 484. ⁴⁴ Sen. Ep. 40.2. ⁴⁵ Sen. Ep. 40.11.

Declamation and Philosophy

251

in reconciling the two. Of course, the sort of symbiosis that I have been describing, in the progymnasmata and declamations of the rhetors, and in the display rhetoric cultivated by Greek philosophers for public performance, is a superficial phenomenon compared with the achievements of a Eudoxus or a Carneades when properly philosophizing. But Cicero’s solution is superficial. His own true achievement was to bring the flexibility and verve of oratory to the exposition of technical matters in his own philosophical works. But even here one has to remember that just as his material is very often Greek in origin, so too is his aspiration to express it rhetorically. In a remarkable passage of the Tusculans⁴⁸ he says: ‘I have always thought the perfect philosophy to be the kind that could discuss the great questions in a full rich style. And to this practice I devoted myself with such enthusiasm that I even ventured to hold scholae in the Greek manner.’ The model he was following in this practice was, he tells us, Aristotle. [Professor Wisse’s views (see headnote) may now be gleaned from Wisse– Winterbottom–Fantham (2008), 42–54, esp. 42–5 and 50 (he contends that rhetoricians did not deal with theseis and, by implication, with philosophical questions). The relationship between declamation and philosophy has been treated, for the Major Declamations, by Pasetti (2008).]

⁴⁸ Cic. Tusc. 1.7

21 William of Malmesbury’s Work on the Declamationes maiores 1. In spring 2014 I gave a paper, as yet unpublished, on the editors of the declamations of Calpurnius Flaccus, from Pierre Pithou to Lennart Håkanson, trying to give the names a human face.[¹] Now I offer a short study of the work of a medieval scholar on the Major Declamations. He is in some sense their first known editor; but less can be said of his life and personality than we should wish. Malmesbury is a little town in the Cotswolds, a picturesque area of rural England, west of Oxford and north-east of Gloucester. The river Avon, one of several of that name, curls round a hill on which from time immemorial a Benedictine monastery lay. In the twelfth century a grand Norman church was built, the remains of which have now been converted to serve the needs of the parish. To this house a boy called William was sent to train as a monk some time around 1100. He had been born and bred in England, but he had both English and Norman blood. The date of his birth is controversial,² but on any calculation he must have been remarkably young when, around 1125, he completed the first drafts of two interrelated books, the Gesta regum Anglorum and the Gesta pontificum Anglorum, which between them covered the history of England, secular and ecclesiastical, from the Anglo-Saxon conquest to William’s own day. This achievement was the more remarkable in that it rested on an extraordinary wealth of reading. It was not merely that William had read all that he could find (including documents) that threw light on the history of England. He was also familiar with all, or almost all, the classical, patristic, and early [Segno e testo 12 (2014), 261 76] * This paper was read in an earlier draft at a colloquium at the Fondation Hardt, Geneva in September 2014. I am grateful to Rod Thomson, Robert Kaster, and Ermanno Malaspina for their helpful comments. ¹ [See now Winterbottom (2017b) (= A.22 below).] ² Thomson (2003²), 199 201. This ground breaking book (1987¹) is the standard work on many aspects of a remarkable man.

William of Malmesbury and the Declamationes maiores

253

medieval Latin texts that were available in England at this period. It will have helped that he (as he tells us) was in his youth an assistant in the library at Malmesbury, and eventually became librarian and precentor. But he must have often asked for research leave, to go round England (and perhaps also France) looking at manuscripts, especially, it would seem, at Canterbury. There is no point in listing the countless authors of whom he shows knowledge.³ But I draw attention to a part of his literary activity that is especially relevant here. He produced, or supervised the production of, collections of material that still survive, in the original or in copies. There is for example a massive fifteenth-century volume including no fewer than twentytwo works of Cicero, now in the Cambridge University Library.⁴ William also collected material concerning canon law and ancient and Carolingian history,⁵ and produced what may be called an edition of the Liber pontificalis.⁶ The list of texts known to have been worked on by William continues to grow. Robert Kaster has recently discovered in a couple of manuscripts of Suetonius, descended from a lost witness, nearly 2,000 readings that diverge from the inherited text. Their quality varies wildly. They can be put on a spectrum that ranges from acute corrections to wilful tampering. Kaster at first called them the work of Impiger, a busy and indeed aggressive corrector. Later, he came to realize that this unknown person was in all probability William of Malmesbury. I shall return to Kaster’s findings later (§ 4). I add that Ermanno Malaspina has found similar alterations in another pair of manuscripts, this time of Cicero’s Lucullus, and again there is every reason to identify the energetic corrector as William.⁷ The manuscript that concerns us here is Oxford, Bodleian Library, Rawlinson G. 139 (s. XII¹; I call it R).⁸ It contains three main texts: Cicero’s Partitiones oratoriae, his De officiis, and the Declamationes maiores. These works were not transcribed by William himself, but he seems to have supervised their transcription. They all contain notes, largely corrections to the text, in his unmistakable hand. It is on this basis that we may think of William as a kind of editor of the Maiores. I shall discuss the nature of his interventions, and the light they throw on the problems of reading this, or any old Latin text, in twelfth-century England. 2. First, the manuscript that William was correcting. One always hopes that an English text of a classical author may be a little unusual. That is certainly true of the De officiis found in the Rawlinson manuscript, which does not ³ See Thomson (2003²), 202 14. ⁴ Cambridge, University Library, Dd. 13.2; see Thomson (2003²), esp. 51 5. ⁵ Ibid., 63 6 and ch. 7. ⁶ Ibid., ch. 6. ⁷ Cf. Kaster (2016); Malaspina (forthcoming). I am grateful to both authors for allowing me to see these articles before publication. ⁸ On which see Thomson (2003²), 85 6.

254

Papers on Quintilian and Ancient Declamation

present the ordinary Vulgate text common on the Continent at this period. That Vulgate is a twelfth-century redaction of what is known as the ζ group, and it is present in a large majority of the 700 or so surviving manuscripts. The other group, ξ, has only a handful of witnesses.⁹ R is one of a select group of English and French books that, while not belonging to the ξ group, show contamination from it. The details have yet to be worked out. But I was able to use the Rawlinson manuscript, together with another from Avranches, as an independent witness to Off. 1.40.¹⁰ What is more, I cited one correction made by William.¹¹ It is worth stressing that there is nothing unusual in finding good corrections in medieval manuscripts. Scribes who wrote manuscripts knew Latin, and would inevitably correct what they thought to be errors as they went along. So too would readers. When I worked on De officiis I chose a number of books, especially ones that could be precisely dated, in which to search for conjectures.¹² It was rarely that I found nothing at all of interest, either in the original text or in the margin. It was a real pleasure to find, occasionally, that an unknown person in the Middle Ages or the Renaissance had anticipated a conjecture made by a scholar in print very much later. What, then, of the original text of the Major Declamations in R? There was a chance that it too might be of interest. The only other roughly contemporary English manuscript of our text that I know of is also in the Bodleian Library, Arch. Selden B. 36, detected by Hugo Dessauer to be of impressive ancestry.¹³ It was, it seems, copied from the French MS Leiden, Universiteitsbibliotheek, Voss. lat. Q. 111 (V), one of the two most important witnesses to the Major Declamations. The date of the Vossianus is disputed; Birger Munk Olsen¹⁴ dates it as late as the second half of the twelfth century, and the Selden manuscript to ‘S. XII/XIII’. Our Rawlinson text does not have as impressive a pedigree as that. It was unknown to Dessauer, and for that matter to Håkanson.¹⁵ But it is quite clear from my examination that it is a close relation of a manuscript known to both of them:¹⁶ Paris, Bibliothèque Nationale, lat. 7802 (D). Munk Olsen¹⁷ dates this to the mid-twelfth century, and judges it to be French. Håkanson¹⁸ notes three places where D alone gives the truth, presumably by conjecture; our Rawlinson manuscript coincides at each place. I add just one error common to R and D that I have noticed (there will be many more). At Decl. mai. 3.2 (43.24 H.)¹⁹ both ⁹ See generally Winterbottom (1993). For the Rawlinson and Avranches manuscripts, see ibid., 229 31. ¹⁰ See Winterbottom (1994), . ¹¹ At Off. 1.135. ¹² See Winterbottom (1994), . ¹³ Dessauer (1898), 20. ¹⁴ Munk Olsen (1982 5), .297; 299. ¹⁵ I refer throughout to Håkanson (1982). ¹⁶ Dessauer (1898), 28 9; Håkanson (1982), . ¹⁷ Munk Olsen (1982 5), .302. ¹⁸ Håkanson (1982), . ¹⁹ Here as elsewhere, in citing the Maiores, I add Håkanson’s page and line number in brackets.

William of Malmesbury and the Declamationes maiores

255

manuscripts give deliberasset for deliberasses; it is²⁰ characteristic that William, or another person, corrected this to deliberasse, the reading of the important manuscripts B²¹ and M.²² There is no need to try to work out the exact relationship between D and R. But R will come close to D in the stemma drawn in Håkanson’s preface:²³ very low down. The stemma divides the manuscripts into two groups: V (mentioned above), B, and M on the one side, and on the other various sub-groups, called by Håkanson δ, β, and γ, all going back to an unnamed hyparchetype independent of BVM. I must say that I am not quite convinced of this independence. I suspect that the text could be edited from B, V, and M with no reference to the other manuscripts except as sources of conjecture (and I might add that I think that unique readings of V may not always have been taken seriously enough).[²⁴] But this does not matter for our purposes. What concerns us here is that R is close to D, well down below γ. This means that we are not to look to it for anything but conjectural truth. And it also means that it is not likely to be a very correct text.²⁵ If one imagines reading this very difficult collection of declamations in a text full of error, with no apparatus criticus, no help from commentaries, translations, indexes, or dictionaries, let alone data bases, one will begin to get an idea what William was up against. The manuscript is quite legible; but the light will not always have been very good, especially in winter; and it is not as warm in Malmesbury as—say—in Italy. 3. To turn to William’s marginalia: they are normally easy to distinguish from corrections made by the first hand; he wrote with a very fine pen, and was able to insert words legibly above the line as well as in the margin. But of course where words are crossed out, or under-dotted to signal deletion, it is not usually possible to know who was responsible; and the original scribe quite often makes corrections of his own, no doubt after consulting his original. I shall give some examples of William’s corrections, starting with something of limited interest except to William’s own editors. He was very concerned with orthography. This we know because we have, astonishingly, the autograph of one of his major works, the Gesta pontificum: a tiny book, written in a tiny hand on closely-packed lines.²⁶ The spelling conforms to a rigorous

²⁰ As we shall see: below, p. 257. ²¹ Bamberg, Staatliche Bibliothek, Class. 44 (M.IV.13); s. X. ²² Montpellier, Bibliothèque Interuniversitaire. Section Médecine, H 226; s. XII². ²³ Håkanson (1982), . ²⁴ [For my latest and quite different view on the whole issue see in detail A.24 below.] ²⁵ All the same, I should judge from my sampling that R is not markedly more difficult to understand than Lehnert’s very conservative text (Lehnert (1905)). But I myself vividly remem ber how frustrating it was to work with that. ²⁶ Oxford, Magdalen College, lat. 172.

256

Papers on Quintilian and Ancient Declamation

system, which extends to the way in which Anglo-Saxon names are turned into Latin equivalents. So far as proper Latin is concerned, William (for example) has a rule, almost never broken, for the use of t instead of c.²⁷ But he makes no attempt to impose such a scheme on our text. Instead, he classicizes where his text seemed to him unduly medieval. He very often reacts to omission of the aspirate, both at the beginning of words (ibernum changed to hibernum) and in mid-word (trait to trahit); and he does not like final d (velud to velut). It is more interesting that he is unaware of some classical norms: he corrects perit to periit because he does not recognize the contracted form of the perfect; similarly he corrects genitive Servili to Servilii. To come to more substantial matters. I start with a few places where R has a false reading not found in our principal manuscripts. We, with our hindsight, know what the answer is; William did not. At Decl. mai. 14.2 (288.18–289.1 H.) modern editions read: Miseremini, iudices, ne vobis venefica sic imponat

R has miserrime. William changes that to miserum me, an exclamation that appears several times in the Declamations, as he perhaps remembered (he had an extraordinary memory for words and phrases). Some supplements: At 3.13 (54.11–12 H.) modern texts read: Libenter (libere Håkanson) te, imperator, interrogo: (a question follows)

R omits interrogo. William supplies the word percunctor. He had no database to tell him that the word percontari is not used in the Declamations (though it is in Quintilian). His addition is not right, but it deserved to be. At 10.18 (217.16–17 H.) the declaimer wrote: Sceleratus ille, ille impius, quicumque defunctum filium vidit

R omits one instance of ille. William was not to know this when he added et before impius. At 10.6 (204.20–3 H.) we now read a sentence concerning the son conjured up from the grave as follows: Nec iam nisi cum luce certa fugatisque sideribus invitus ille vanescebat ex oculis²⁸ multum resistens, saepe respiciens, et qui se promitteret etiam proxima nocte venturum.

R omits sideribus . . . respiciens. William knew that something was missing, and he added tenebris, a perfectly appropriate companion for fugatis. It is odd that ²⁷ Details in Mynors Thomson Winterbottom (1998),  ; see also Winterbottom (2007),  . ²⁸ Where we should add a comma, as the rhythm shows.

William of Malmesbury and the Declamationes maiores

257

he did not try to find a verb for the sentence, but so far as he went he was on the right lines. These three passages show how difficult it is to be sure of making the correct supplement, even where we are able to diagnose a lacuna. Conversely, two deletions by William: At 13.17 (284.19–20 H.) we read: in publicum vivitur, et communes opes congeruntur in medium

R corrupts et to ad, which William then deletes. This could well have been right, but it was probably not. At 15.7 (308.23–4 H.): Bibi medicamentum crudele, saevum

R corrupted saevum to the meaningless servum. William deleted the word. I am all too conscious of the number of words I have myself deleted in classical and medieval texts when I should have been emending or supplementing them (a principle that used to be firmly stated by the late W. S. Watt). At fifteen or more places I have noted William introducing a new reading with the abbreviation for vel, normally used in manuscripts to signal a variant reading found in another book. It would seem unlikely that William was in fact doing that, for almost always the readings so signalled are not found in Håkanson’s apparatus. It would seem, rather, that he uses this abbreviation merely to preface a conjecture of his own. I mention one case in particular, as it is repeated within a few lines, and is of some wider interest. This is in the theme of Decl. mai. 13, where the law, stated (264.11 H.) and then alluded to (264.16 H.), is Damni per iniuriam dati sit actio. William changed dati in both places to illati. If this was his own idea, he was not being stupid: he was perhaps even showing off his learning. The law stated in Sen. Con. 3.6 (excerpt) is Damni inlati actio sit, and illati is also in Fortunatianus (Rhet. 1.15 (p. 88.3 Calboli Montefusco)); dati, on the other hand, is paralleled in ‘Quint.’ Decl. min. 385.²⁹ Quite a number of other changes made by William coincide with readings of known manuscripts, including important ones like M and V. But there is no pattern suggesting that there is more than coincidence at work here, and again it does not seem that William was drawing upon another manuscript (apart, that is, from the exemplar of R itself, from which some of his corrections doubtless come). For that matter, readings of R that William sees fit to correct sometimes coincide with those of other manuscripts, again including primary witnesses like B and M. But all this is not relevant to my present topic, and

²⁹ I owe this information to the database of declamatory themes under construction by Lucia Pasetti and Catherine Schneider.

258

Papers on Quintilian and Ancient Declamation

further investigation would in any case require wider information on the whole tradition than I possess or plan to accumulate. I come now to give some snapshots of William at work correcting passages which are the same, or more or less the same, in texts we now read. Here, at least, he and we work in the same field and face the same textual difficulties. Where, by our standards, he fails, we have no reason whatever to feel patronizing. I start with five places where a cheer is certainly called for. In them, William anticipates modern conjectures. 5.4 (88.3 4 H.) rus, servulos, penates et omnia viliora properantius, festinatione perdentis, addixi properantius William, Dessauer: properantibus codd.

6.14 (125.13 15 H.) si quaedam fortuna parentes nos deprehenderit, quamvis exiguo divisus spatio inpune opem non feret divisus codd., Håkanson: divisos William, Burman

8.12 (162.25 6 H.) Sibi quisque firmatur, sibi quisque componitur firmatur codd., Håkanson: formatur William, Schultingh [and already π, as I have now discovered]

16.3 (322.6 7 H.) ex duobus magis amatur, quem tyrannus paratus est alligare amatur codd., Håkanson: amat William, Winterbottom ap. Håkanson³⁰

Finally: 16.8 (328.4 5 H.) quantum ille praestitit , qui [mihi] [in]vidit poenam meam nec recessit! vidit William, Håkanson: invidit codd.

It is true that William’s manuscript gives invidit as two words, in vidit, so that deleting in was an easy move. Nevertheless . . . In many other passages, however, William strikes out on his own. I shall illustrate the sort of thing he does. I start with an unusual case. At 7.4 (140.18 H.) the manuscripts present a phrase that is deleted by the editors as a gloss on the word talionis: Talio dicitur quasi tali poena id est simili, ut cui oculus aufertur oculum auferat, vel si quid tale. Isidore explains similarly,³¹ but the exact form of our gloss is not to my knowledge found elsewhere. In R William added glosa above the word Talio. I do not know if he meant by this what we would mean, that is an extraneous explanation added by someone other than the author of the work.

³⁰ It is said that the English scholar Porson used to weep tears of joy when he found that he had made a conjecture that Bentley had thought of before him. ³¹ Etym. 5.27.24: Talio est similitudo vindictae, ut taliter quis patiatur ut fecit.

William of Malmesbury and the Declamationes maiores

259

Alternatively, he may just mean that here we are being given an explanation, without implying that it should be expunged. We have seen various places where William produced conjectures that anticipate modern suggestions. But the majority of his changes do no more than correct R back to the ordinary transmitted reading. When this happens, we cannot tell if William is correcting on his own initiative, or just profiting from the consultation of the exemplar of R. But when William makes other changes, he is undoubtedly out on his own, and the likelihood of his making a correct change, or even a plausible one, is not very high. In what follows, I shall not so much be looking for brilliant conjectures as trying to see how William’s mind works, and how good a Latinist he was. 2.19 (37.15 H.) haec est omnium natura rerum, ne . . .

William changed haec to hoc. Conversely, in 5.6 (90.23–4 H.): Nondum hoc caritas est he changed hoc to haec. He clearly felt some uncertainty about this sort of construction.³² 4.8 (69.24 70.1 H.) Felicior mehercules qui moritur antequam debeat, antequam velit

William added quam ille qui after debeat. There is no virtue in this addition. William seems to have misunderstood the pair of parallel antequam clauses, oddly, considering that he often writes like this himself. 5.6 (91.2 H.) ignotis cadaveribus humum gerimus

Håkanson prints ingerimus from the deteriores, also citing E’s congerimus. William (oddly) corrects to egerimus, the opposite of the required sense. 7.11 (148.9 10 H.) Scis profecto, percussor, facinus ordinare, disponere ordinate William

This change suggests that William did not recognize the coupling of two verbs of similar meaning in asyndeton: common enough in certain types of Latin, but (I think) not part of William’s own armoury (though, in the tradition of the declaimers, he very much affected parallel clauses of similar meaning in asyndeton: see above on 4.8). Equally, he did not see that ordinare is ensured by the rhythm. But his change is by no means absurd. William may, or may not, have known a passage of the Ad Herennium (4.69) that reads: acute et cito reperimus, distincte et ordinate disponemus. 8.4 (154.13 14 H.) sive captavit ex hoc vultu quandam artis imitari, quod sanare non potuit

³² Discussed in Stramaglia (2013), 199 n. 340.

260

Papers on Quintilian and Ancient Declamation

This is how R gives the text. Schultingh eventually corrected vultu to velut, and Håkanson later added imaginem as an object before imitari. M gives industriam at this point, presumably by conjecture (the word could easily have fallen out before imitari; and the phrase turns up several times, even in Cicero). William, too, felt the need for an accusative noun, and supplied constantiam, on rather the same lines as M’s addition. I suppose he was thinking of the look of a doctor’s resolution on the man’s face (vultus). We should, by the way, in emending this passage, remember the rhythm: artis imitari is not lightly to be tampered with. 8.16 (167.14 16 H.) cum . . . in latentes meatus haec quibus tegimur medi cina descendat

The addition of per was not known to William. Instead, not very convincingly, he wished to take quibus with meatus, and accordingly changed the order so that haec immediately preceded medicina. 10.6 (205.13 H.) iuvenem videbis et forte etiam †die speres†

William changed to desperes, with no sense or syntax; many other conjectures have been proposed over the centuries.³³ I wonder if we might read: et forte etiam diu speres, ‘and perhaps even as long as you could hope’ (i.e. not just ‘see’, but converse with at length; the woman goes on to say that she has been able to be with her son totis noctibus). 10.19 (218.14 H.) ut ferrum tuum refigas

William, perhaps not recognizing the sense ‘unloose’, changed to the more violent refringas. This verb is not found in the Major Declamations. 12.2 (233.12 13 H.) quasi ego non confitear illum etiam nimium multum attulisse

Stramaglia in his Cassino edition³⁴ translated nimium multum ‘fin troppo’ with no comment, but William took exception to the phrase and deleted multum. He was not being silly, and only a single passage in Plautus³⁵ parallels this exact locution. Nimium + multa (or the like) is fairly frequent. 13.1 (265.22 266.1 H.) si cuius mihi conscius culpae . . . iram tulissem

William corrected to cuiusquam, though he must have been aware of the idiom si quis.³⁶ 13.9 (275.6 7 H.) lascivientem luxuria fugam tinnitu conpescimus

William found luxuriam in R, and it is the reading of most manuscripts. Håkanson prints luxuria from V. William deleted the word. The text is not ³³ Fully reviewed by Schneider (2013), 168 9 n. 179. ³⁴ Stramaglia (2002), 35. ³⁵ Merc. 479. ³⁶ The original reading of R is not quite certain.

William of Malmesbury and the Declamationes maiores

261

perhaps quite secure; Dessauer wrote luxuriam. Fugam is boldly used = ‘the bees in flight’. 13.10 (276.5 H.) plenas vel cremare igni vel aquis inmergere alvos plenos William

Lewis and Short, which was not on William’s shelves, tells us that alvus, though normally feminine, is masculine in some early and/or poetic Latin. The Thesaurus linguae Latinae³⁷ adds that the masculine is found in the verse of Venantius Fortunatus, a poet known to William.³⁸ It is striking that he knows—or appears to know—that there is doubt about the gender of the word (but he does not challenge the feminine at 13.3 (268.16 H.)). Less eruditely, he changes publicas aures to publicos aures at 15.12 (315.17 H.); but there seems no doubt about the gender of this noun. 15.1 (302.2 4 H.) nihilque gravius adficiat conscientiam bonorum quam quoties cumque nulla merita ceciderunt

Gronov changed nulla to irrita, but I take it that Håkanson understood nulla in much the same sense: the deserts came (proleptically) to nothing. Burman suggested multa, and on the same lines William changed to nonnulla; but neither change carries conviction. 15.14 (317.2 4 H.) Toto licet infelicem terrore convenias, non exosculabitur ma nus; mortem, suprema denunties, te non rogabit.

Not seeing that licet also covers denunties, William added si before mortem. 4. I return for a moment to Kaster’s work on Suetonius (above, § 1). He classifies the corrections made by his Impiger, identified now as William of Malmesbury, under various heads. They especially concern changes of order. William, for instance, was anxious that attributive adjectives should follow their noun (libro tertio, not tertio libro; and by extension capite obvoluto, not obvoluto capite); that adverbs should precede the word or phrase to which they are attached (tunc clarissimo, not clarissimo tunc; diu apud se, not apud se diu); and that hyperbaton should be eliminated. My findings for the Rawlinson manuscript of our declamations do not show up this sort of thing at all. Why so? It looks as though William worked in two different ways on the books he read. In Suetonius, it may be, he was producing an ‘edition’ of a classic text that pupils in the monastery would find easier to read than the original. In the Maiores, on the other hand, he was not producing a text at all, but correcting a corrupt copy as he went along: writing in ideas as they came to him. Most of these ideas were to our eyes wrong. But the same could be said of the sort of

³⁷ TLL s.v. alvus 1800.28.

³⁸ Thomson (2003²), 214.

262

Papers on Quintilian and Ancient Declamation

thing we all write in our margins without thinking that anyone else will ever read them. I end with two general points. First, it is strange that the book seems to have left little mark on William’s own works. Among his other preserved productions is a sort of commonplace book, called the Polyhistor,³⁹ into which he copied stories from a variety of classical and patristic sources, especially the Elder Pliny and Jerome. Even if nothing in the Major Declamations was suitable for that compilation, it is very likely that William made collections of other material for use in his historical and hagiographical works. Perhaps only so can we explain his remarkable ability to quote and adapt other authors, often using them in a very sophisticated (we should nowadays call it an intertextual) way. But I know of only one place where the Major Declamations are drawn upon by William. This is at 13.2 (266.14 H.), where we read: Sic quoque me latentem invenit invidia. As Håkanson notes in his apparatus, these words are found in two identical passages of Jerome.⁴⁰ Now in William’s Vita Wulfstani, he writes at one point (1.8.1): ita latere volentem invenit et carpsit invidia.⁴¹ Again, in his commentary on Lamentations (2.369–70) we find: sed hunc quoque latentem Ismaelis invidia invenit et trucidavit (also ibid. 1.383–4: sed ibi quoque florentem carpsit invidia).⁴² We cannot in these circumstances know if William was drawing on the declamation or on Jerome, with whom he was very familiar. But it is worth mentioning that this passage is marked in R by a symbol in the margin. Rod Thomson tells me that this is of a design unique to William. Its main body is a capital N. There is a tiny o at the top of the lefthand ascender. The right-hand ascender turns into a T at the top, while at the foot there is a tiny a. So, very economically, we have all the letters of Nota. Again we cannot be quite sure what is going on. Perhaps William saw this sentence here for the first time, and marked it for later use. Perhaps he had read it in Jerome, gone on to use it in the Life of Wulfstan and the Lamentations commentary, and then, finding it in R, marked it as a striking parallel. A second point arises from what I have just been discussing. What interested William when he read this difficult book? The marginal Nota I have mentioned is found in about thirty-five places. At one place William drew further attention to a passage by adding the picture of a hand. This comes at 5.11 (95.21–2), with the words Pessimus est mortalium, qui amari fratrem suum sine sui caritate putat. That is, ‘Worst of mortals is he who thinks that his brother is being loved (i.e. by his father) with no affection given to himself.’ ³⁹ Ed. Testroet Ouellette (1982). ⁴⁰ Quaest. Hebr. in Gen., pr. (p. 1.20 2 de Lagarde = CCSL 72.1): me vero procul ab urbibus, foro, litibus, turbis remotum, sic quoque (ut Quintilianus ait) latentem invenit invidia; Vita Malchi 6.1 (SC 508.196): Sic quoque me latentem invenit invidia. ⁴¹ Winterbottom Thomson (2002), 36. ⁴² Winterbottom Thomson (2011), 144 and 21.

William of Malmesbury and the Declamationes maiores

263

The words, though obviously not meant by the declaimer in a Christian sense, could have clear resonances for a monk who might think other brothers were loved by God, or by the abbot, more than he was. It seems worthwhile then to glance at some of the other thirty-odd passages where the Nota symbol is used. As to distribution: there are none in Decl. mai. 2, 10, or 12. There are seven in Decl. mai. 11, four in Decl. mai. 9. William the historian is interested by imperium populi Romani ad hanc diem militari disciplina stetit (3.14 (55.1–2 H.)), and, it may be, by Nihil est facilius quam in quemlibet adfectum movere populum (11.7 (226.24–5 H.)). He was also attracted by an allusion to a Terentius and Scipio Africanus (9.20 (195.7–12 H.)). William the amateur in medicine⁴³ points out a feeble remark on types of illness in 8.12 (162.20–1 H.). William the Benedictine monk directs attention to Haec enim condicio superiorum est, ut quicquid faciunt praecipere videantur (3.15 (55.17–19 H.)), and to quo ad altiorem quisque honorum gradum escendit, magis in exemplum spectantibus patet (3.13 (54.14–16 H.)), applied by the declaimer to soldiery rather than the monastic hierarchy. But for the most part William notes, or wants us to note, passages relevant to the life of Everyman, and not least to his own; for his works, and especially the commentary on Lamentations, amply illustrate a concern for his own soul, torn between reason and emotion, plagued by the pleasures of the senses; and bring before us, too, a man in a social situation, with friendships and affections that may not always be easy to deal with. Thus on innocence: Innocentia per gradus certos ab homine discedit (1.6 (6.11–12 H.)); on anger: haec animi suspicienda moderatio, vincere iram et inter simultates quoque meminisse hominis, followed by more snatches of Roman history (9.17 (192.3–8 H.)); on love: ego perpetuam quandam mihi gloriam reor, quod ille caelestis animus me potissimum quem amaret elegerit (9.14 (188.20–2 H.)); on friends: si vitae praestes omnia secunda, amicus otiosa res est (a sentiment William would have deplored: 16.7 (326.13–14 H.)); on pietas: habet pietas impetum suum, nec ullum dominum novit adfectus (6.20 (131.11–12 H.)); on the illicit: diliguntur inmodice sola quae non licent (14.8 (296.6–7 H.)); and finally on life: tota vita hominis unus est dies (4.4 (70.23–4 H.); with the preceding passage about all the pleasures of the world). I have tried to give this early worker in our field a human face. Here is a final touch. He read the Major Declamations through from the start, but not quite to the finish. He laid down his fine-nibbed pen on fol. 138r, at what we know as 17.9. He left no comments on the speeches Infamis in matrem. Enough, it seems, was enough.

⁴³ See e.g. Gesta regum Anglorum 2.pr.1: phisicam, quae medetur valitudini corporum, ali quanto pressius concepi.

22 The Editors of Calpurnius Flaccus* Calpurnius’ work, as we have it, does not make things easy for the reader. And I have been prompted to look at the successive editors of our author, from 1580 to the present day, and ask the question: what, if anything, did they contribute to his understanding? The first exhibit is Pierre Pithou (‘Pithoeus’: 1539–1596).¹ He alone of our editors was not based in a university. He was a French grandee, procureur-général in the Parlement of Paris, who, as we shall see, moved in exalted circles. He had a remarkable track record as a purchaser of important manuscripts. All three were written in the ninth century: two in France, one in Germany. The most famous is the Montpellier Juvenal,² by far the best witness to the exiguous purer strain of this tradition. Pithoeus edited the author in 1585. Another P is what is recognized to be quite the most important witness to Phaedrus,³ whom Pithoeus edited in 1596. A third of his purchases is another Montpellier manuscript,⁴ which contains a considerable batch of the Minor Declamations ascribed to Quintilian (the only medieval witness to this text), together with the Elder Seneca—in the excerpted version—and our Calpurnius. Pithoeus edited the Minor Declamations in 1580. It was by no means the editio princeps (it had been in print since 1494), but Pithoeus’ manuscript put the text on a new and far sounder basis.

[From M. Dinter, C. Guérin, and M. Martinho (eds.), Reading Roman Declamation. Calpurnius Flaccus (Berlin and Boston, 2017), 141 60] * I am very grateful to Robin Briggs for help with the French seventeenth century, and to Antonio Stramaglia (and through him Massimo Pinto) and the late Martin West for help in tracing the career of Georg Lehnert. Professor Stramaglia was also kind enough to read a draft of this chapter and make valuable suggestions. My principal debt, however, is to Donald Russell, who discussed with me the text and translation of Calpurnius 13. ¹ I do not supply references for easily accessible details of the careers of the earlier editors. See generally the still very useful survey in Sandys (1903 8), vol. ; also Eckstein (1871). For other transmissions mentioned en passant, see Reynolds (1983). ² Montpellier, Bibliothèque Interuniversitaire. Section Médecine, H 125 (?Lorsch); see e.g. Munk Olsen (1982 5), .575. ³ New York, Pierpont Morgan Library, M.A. 906 (?Reims); Munk Olsen (1982 5), .227 8. ⁴ Montpellier, Bibliothèque Interuniversitaire. Section Médecine, H 126 (Reims); Munk Olsen (1982 5), .298 (Decl. min.); .419 (Seneca Rhetor, excerpta); .53 (Calpurnius Flaccus).

The Editors of Calpurnius Flaccus

265

Unfortunately the section of his manuscript containing Calpurnius was sadly mutilated, and Pithoeus had to fall back on a much younger book. He had acquired his old manuscript ‘Claudii Falceti . . . beneficio’.⁵ This was Claude Fauchet (1530–1601), historian and antiquary. He was premier président in the Cour des monnaies, a Paris sovereign court established in 1552. It may be added that Pithoeus dedicated his edition to Christophe de Thou (1508–1582), premier président of the Parlement of Paris: Pithoeus addresses him as ‘Equiti, Regni Curiae Praesidi primario, et sacri Consistorii Senatori’.⁶ He was father of the great historian and book collector Jacques-Auguste de Thou (1553–1617), whom Pithoeus must have known. It will have been on some similar network that he got wind of a substitute for the defective old book. He does no more than refer to this as an Italian exemplar, ‘non adeo vetusto’.⁷ But Håkanson showed that it was what we call N (Bern, Burgerbibliothek, lat. 149; s. XVI).⁸ This book was once owned by Jacques Bongars (1554–1612), diplomat, bibliophile, and scholar, whose library, partly inherited from Pierre Daniel, eventually ended up in Bern. I am going to digress on the subject of N—I hope it will be thought with good reason. Håkanson was well aware of the large number of places where readings of N coincide with corrections in B (known as B²), and was convinced that in such cases the corrections were taken from N (and not from N’s twin in Munich, M).⁹ Some of these corrections he thought might be due to conjecture; others he judged (I think rightly) to have been brought in from a source outside our tradition.¹⁰ In his apparatus to the not very long Decl. 13, the piece which I shall be using throughout to illustrate this paper, Håkanson (1978) cites no fewer than seven instances where N agrees with B² against BC. In three cases he prints N’s reading; and he says in his preface (Håkanson (1978), ) that two of these are not likely to be the result of conjecture. Several of the other readings are matters of dispute. In any case, it is remarkable that, despite the new witnesses that have emerged since Pithoeus, the text he printed in 1580 (see the Appendix) is basically almost identical to that printed by Håkanson (1978) four centuries later, if we ignore later conjectures made or accepted by Håkanson. There are three new readings from BC, all in the last three sentences: nativum for notum (highly dubious: Håkanson (1978) expresses no doubts ad loc., but puts a question mark against it in his list on p. ), moris for mortis (which Pithoeus had already conjectured), and me for nec before venenum right at the end (also, in fact, in N); I shall return to these ⁵ See Pithoeus’ dedicatory letter to Christophe de Thou and his final note on the Minor Declamations, reprinted in Burman (1720b), 419 and 790 respectively. ⁶ Burman (1720b), 413. ⁷ See Pithoeus’ note referred to in n. 5 above. ⁸ Håkanson (1978), . ⁹ For the manuscripts of Calpurnius see Håkanson’s (1978) praefatio and list of sigla. There is no need to go into detail here. ¹⁰ Håkanson (1978),  .

266

Papers on Quintilian and Ancient Declamation

passages. This does not suggest any great superiority of BC to N: rather the contrary. It looks then as though Pithoeus’ luck was holding out. The manuscript to which I assume he had access through Bongars was by no means a dud, though of course the much damaged Montepessulanus would have been more valuable if it had survived intact. However that may be, what Pithoeus printed remained the base text until the early twentieth century. It was what his successors strove to emend. He had done just a little himself to improve what he found in his manuscript. He suggested changing consequitur to the future consequetur, and he may have been right to do so. And, as we have seen, he proposed emending mortis to moris, which proved later to have manuscript support: a rather odd change, in my opinion, for mos can hardly have a genus. Pithoeus, then, transcribed his manuscript, and made only minimal suggestions as to how its text might be improved. And he did not explain anything, either his conjectures or any difficulty in the Latin. This is in sharp contrast to his work on the Minor Declamations, where, besides exploiting his own old manuscript, he (to quote my own assessment) produced ‘acute solutions to problems that Aerodius had failed to touch’.¹¹ This difference of approach is natural considering the different importance of the two texts, and the excitement of using an old primary manuscript for the first time. Pithoeus did not make much of Calpurnius Flaccus: ‘Quis autem hic Calpurnius fuerit, alii fortasse dicturi sunt.’¹² But others, I fear, have not been able to tell us. Pithoeus in this way provided a text of Calpurnius, and not at all a bad one, for others to try to understand. When Calpurnius was next edited, in 1665, we have moved from the grand political and bibliophilic world of Pithoeus and his friends to the world of Dutch universities. That too was grand enough in its own way. Johann Friedrich Gronov (‘Gronovius’: 1611–1671) was born in Hamburg, but studied at Leiden and Groningen. His predecessor in the chair at Leiden was Daniel Heinsius, and Gronovius was a friend of the brilliant Niklaas Heinsius, Daniel’s son, the consummate textual critic. Gronovius was no mean critic himself, with an interest in prose (including the Elder Seneca) as well as poetry. E. J. Kenney cites a fine letter of his to the young Heinsius where he talks of the art of textual criticism in terms that would be taken up by Bentley and Housman later.¹³ You need to be expert in languages, ancient customs, history, philosophy. ‘When all this is brought together,’ he goes on, ‘and reading and diligent labour is attended by talent and judgement, the result is the ability to judge each ancient writer in his own terms.’ The climax

¹¹ Winterbottom (1984), . ¹² See again Pithoeus’ note referred to in n. 5 above. ¹³ Kenney (1974), 58 (Latin in n. 1): ‘Haec cum conjuncta sunt, et lectio diligensque labor ingenio & judicio non destituitur, existit inde praeclarum illud egregium, posse de unoquoque Scriptorum veterum . . . judicium ferre, genium uniuscujusque penetrare, . . . locos corruptos restituere.’

The Editors of Calpurnius Flaccus

267

of Gronovius’ list of the outcome of such wide erudition is ‘to restore corrupt passages’. This he himself did with panache. In the Minor Declamations I speak of ‘a galaxy of fine emendations’ on his part.¹⁴ His ideas shine in the Major Declamations too, and in the Elder Seneca. No one could have been better qualified to understand and, where need be, correct Calpurnius Flaccus. Explanation, as with Pithoeus, is in short supply. In Decl. 13 he reports Pithoeus’ two conjectures, but does not print either in his text. Moris, indeed, he improves to timoris, superficially a good deal more convincing, and to be accepted by Håkanson: the context does after all concern fear. But I am not very happy with the resulting phrase. I have wondered about keeping Pithoeus’ text, including mortis. The city doctor has been given an objection: ‘I had a better opportunity’ (that is, than you, the citadel doctor) ‘to administer poison: the tyrant himself had provided it’ (that is, by summoning him to his bedside). To this the citadel doctor ripostes: ‘(But poisoning) is a common cause of death, so that previous experience results in a cautious attitude to everything.’ Accordingly, the argument goes, the city doctor did not have a better opportunity. But, as we shall see, there is even more to say about this crux. I have spoken as though Gronovius edited our work. In fact, the 1665 edition of the declamations ‘ex officina Hackiana’ (Leiden and Rotterdam) is a Variorum text: ‘cum variorum notis’, as it proclaims. The companion volume, also of 1665, which was devoted to Quintilian’s Institutio oratoria, mentioned the notes of Turnebus, Camerarius, Gronovius, ‘and others’; but only Gronovius had, it seemed, worked on Calpurnius. Nor was the whole operation coordinated by him. Perhaps it was just a question of notes from his own private margin being taken over into the common stock. Further on in the 1665 Hackiana we find a letter, dated in the same year, from another scholar who made his mark on declamatory texts, Johann Schultingh.¹⁵ It is addressed ‘Henrico Bentingio’, that is, Hendrik Bentinck, scion of an important family.¹⁶ The Bentincks came from the Netherlands, where their family estate has for centuries been Schoonheten House in Overijssel; but they were to make a wider fortune when Hans Willem accompanied the Prince of Orange to England in the Glorious Revolution of 1688. In my own country they are still represented by the 12th Earl of Portland, who plays David Archer in the BBC Radio 4 programme The Archers (fame at last!). It appears that Hendrik and Johann had both been pupils ‘summi Gronovii’. Schultingh expresses the very proper conviction that he regarded it as part of his job to remove any injury inflicted on ancient texts by the ignorance or audacity of scribes or others, and to restore corrupted passages. He has begun ¹⁴ As above, n. 11. ¹⁵ At pp. 759 60. The notes to Calpurnius follow on pp. 761 84. ¹⁶ He is addressed as ‘nobilissime’. Bentinck is styled ‘Domino in Werckeren’. Werckere was a manor house in the hamlet of Mastenbroek in the north west of the municipality of Zwolle. Thanks are due to Jasper de Mooij for his help with Dutch geography.

268

Papers on Quintilian and Ancient Declamation

this task by turning his attention to ‘the Quintilians’, but time has been short, and he can only offer his friend, in memory of their condiscipulatus, a specimen of his work. There follow around fifteen pages of ‘emendationes et notae’ on Calpurnius’ declamations. The letter is dated from Nijmegen, where Schultingh held a chair, and the year is given as 1665. Poignancy is added by the fact that Johann died in the following year, only seven years after the birth of his son Anton, who was to become professor in Leiden in the middle age of a life that was presumably a good deal longer than his father’s. We learn more specifically from Burman (of whom more later) that Johann had hoped to write a commentary on the Major Declamations and on Calpurnius, but had not got far into the task when he died, leaving somewhat chaotic notes behind.¹⁷ These Burman was able to draw on with the permission of Anton Schultingh. We therefore have news of Schultingh’s work both directly from the Hackiana and indirectly via Burman: though I must say that I have not found anything in Burman’s reports of his ideas on Calpurnius that goes much beyond what had been published in the Hackiana. Before saying something of Schultingh’s work on Decl. 13, I must stress that, whatever its quality, it marks a distinct advance on Pithoeus and Gronovius in a significant sense. The idea of producing notes as well as emendations was new for Calpurnius, and it reflects Schultingh’s realization that this is a difficult text that requires explication as well as correction. The idea of writing a commentary was a sound one, even if it was not completed. Burman took the idea further, and at last in our own day we have a full commentary by Lewis A. Sussman (1994). Sussman also provides a translation, and that too was a happy thought. The more recent French version published under the name of Paul Aizpurua (2005) is not to be ignored, as we shall see. Declamation, apart from the Elder Seneca, has little attracted translators until quite recently. I name honoris causa the quaint and acute English translation of the Major Declamations by (as the title page splendidly puts it) ‘a learned and ingenious hand’, that of John Warr, which came out in 1686. I know of hardly any successors until the remarkable Cassino series began in 1999. What then of Schultingh, as we see him grappling with Decl. 13? Before discussing his emendations, I should stress his ability to provide declamatory parallels for particular usages, as here of perdere, and to explain (rather than emend) difficult phrases. Thus, on Poenas meas hinc cogitate, in quibus nec ira nec natura cessavit, he comments that ira refers to the torturers and the cruel tyrant, while natura ‘seems’ to refer to the flames employed by them. He compares Decl. 7, where the general is iratus and ignes are deployed. I doubt if this is correct, but seeing problems is one step towards solving them and may spark off ideas in others. Burman, after citing Schultingh’s note, said he

¹⁷ Burman (1720a), praefatio, **** 3.

The Editors of Calpurnius Flaccus

269

preferred to refer natura to the innate cruelty of the tyrant, ira to the temporary access of rage caused in him by the situation. I myself wonder if we should not, while taking ira of the tyrant, refer natura to the patientia of the doctor under torture. In any case, one would like some help with the word hinc: is it in effect explained by the relative clause that follows? The harshness of the torture can be judged both from the cruelty of the tyrant and the ability of the doctor to withstand the pain. If one now turns to some of Schultingh’s emendations, one is struck by his desire to give point where point seems to be lacking. In adding huic before virus serpebat and non before longa, he comments on the ‘languida oratio’ of the received text. But in this instance, where the previous sentence is so doubtful and we are not sure if a new sententia starts at virus, the addition seems over-bold. Again, in emending the last sentence of the piece to In arce enecavi: non deseruit me venenum, he comments that this would be ‘declamatorium’ as well as restoring the sense. We now know what Calpurnius really wrote here; but this feeling for declamatory Latin is one of Schultingh’s great strengths in conjecture. Finally, we come to his treatment of the major crux, or series of cruces, in this piece. He felt, as Håkanson did after him, that Ultio . . . vindicavit should be transposed to follow emergit. For, as Håkanson put it, ‘these words belong to the description of the torture, whereas praemium . . . polliceri is probably [!] logically connected with o quam facile etc.’.¹⁸ Håkanson later dropped this idea, no doubt rightly: one cannot be sure about connections of this sort when all we have is a series of sententiae. As in the Elder Seneca, a great difficulty for the critic is not being sure where the breaks between quotations should be set. As to what follows here in Calpurnius, Schultingh saw what for me is the key point: that illecebra refers to the temptation of the prize for tyrannicide, which replaces in the mind of the city doctor his previous desire to win a reward by saving the tyrant’s life. He accordingly wrote inlecebraque in contrarium transferuntur (the passive was later found in BC). This seems to me to be on the right lines, though what should be made of the preceding gerunt is very doubtful. Schultingh thought of cedunt, where persuasiones is nominative, and vertunt, where it is accusative. I prefer it to be accusative, and suggest something like suggerunt, ‘supply’. The subject of both this and transferuntur is the city doctor together with his advocate, as in confingunt earlier. It should always be borne in mind that Schultingh’s notes on Calpurnius are only a kind of draft; he did not live to perfect them. But he consistently displays an acute mind at work, concerned to help the reader to understand a difficult text, and to encourage him to think further for himself. What he did not do, any more than Gronovius, was to look for new manuscript evidence.

¹⁸ Håkanson (1974a), 57.

270

Papers on Quintilian and Ancient Declamation

He did a little in this direction for the Institutio oratoria, but not for the declamations. When Ulrich Obrecht came to edit Calpurnius in Strasbourg in 1698, he added little, and indeed nothing at all in Decl. 13, where he in effect reproduces Gronovius’ text. But that his flair was unabated is clear from a passage in Decl. 12, where from O noverca! Est alta crudelitas he produced O novercae stulta crudelitas! Lehnert rejected this palmary emendation, Håkanson accepted it: no further comment on the comparative judgement of these two editors is necessary. Similar strokes aroused my admiration when I was editing the Minor Declamations. This was a clever man, who did not perhaps give his full attention to Calpurnius. He lived from 1646 to 1701 (he died not long indeed after the publication of his work on Quintilian and Calpurnius). He was a Strasbourg man, and presumably related to another Obrecht from that city (Elias, born in 1653 and professor in Uppsala). Ulrich’s teacher, and his father-in-law, was Johann Heinrich Boekler (d. 1672), who also taught at Uppsala; there was a link between the two universities, fostered by Queen Christina of Sweden. Boekler worked on prose texts too, though historical ones. Ulrich followed him to the extent of editing the pseudo-historical Scriptores Historiae Augustae and Dictys Cretensis, as well as Quintilian. His career merits further investigation. With Pieter Burman the Elder (1668–1741) we return to the Netherlands and to Variorum editions. Indeed he is the natural heir of Gronovius, whose son Jakob had taught him at Leiden. He himself ended up as professor at that university after a spell at Utrecht. No one contributed more freely to the genre of Variorum quarto editions, which so flourished in his time. In forty years he produced editions of around eleven major Latin authors. His declamation collection is more elaborately annotated than Gronovius’ Variorum. For Calpurnius, as we have seen, he adduces Schultingh’s notes, as well as those of Pithoeus and Gronovius. And he adds a good deal of his own from a wellstocked mind: rather surprisingly, considering that he did not value the declamations very highly. He only adds this second volume, he admitted, ‘so that nothing should be found to be missing in this edition rather than because I think they deserve the expenditure of much time and trouble’.¹⁹ He did adduce some new manuscript readings for the Major Declamations, but none for the Minor, let alone for Calpurnius. Burman, no doubt, worked too fast and spread himself too widely. But he was nobody’s fool. In commenting on Decl. 13, he reacts to the notes of the scholars whose work he draws upon rather than branching out on his own. But he has helpful things to say in the process. We have seen him countering ¹⁹ Burman (1720a), praefatio, *** 3: ‘ . . . de Declamationibus, quas magis ne quid desiderari in hac editione posset, quam ut [sic] eas admodum dignas censeam, in quas multum temporis et laboris impendatur, subjiciendas putavi . . . ’.

The Editors of Calpurnius Flaccus

271

Schultingh’s view of ira and natura. Further on, though, he approves of Schultingh’s longa poenarum dilatione, comparing the haste seen in instantem and festinans nearby. I myself feel that the addition is not necessary. The punishment of the citadel doctor went on a long time, during which the condition of the tyrant got worse, giving him the opportunity to identify the poison (or, as I once suggested, the poisoner: see the apparatus to my text of Decl. 13 below, p. 280). What is perhaps more important is to explain dilatione, which does not mean ‘continuation’. When Burman does suggest changes of his own, he does not much convince, at least in Decl. 13. Near the end, he suggests Notum hoc genus morbi; perhaps not very persuasively, but he knows parallels for the confusion of this word with mortis (though he does not cite more than one). That sort of argument is not often found in the material I have examined, and has a modern ring. He would also like clamabant in the last sentence but one, and Lehnert accepted this. But there is perhaps no real problem in tota arce: the tyrant started to range all over the citadel, shouting ‘Doctor!’. Burman thought he was asking for the citadel doctor; but the city doctor must be meant, as in the phrase medicum flagitabat above, which defends both the singular clamabat and its tense. What is the point, though, of quasi ego de tyrannicidio non negassem? Aizpurua, whose translation I mentioned earlier, noted:²⁰ ‘Le tyran savait donc parfaitement, malgré les dénégations du médecin de la citadelle, que celui-ci l’avait empoisonné, et il en réclamait un autre, ne voulant plus des “soins” du premier.’ This gives point: the citadel doctor is using the tyrant’s call for a second opinion to prove that he had administered poison and so merited the prize. We may compare for this quam et medicus confirmaverit et tyrannus earlier. The tyrant (it is argued) acted as though he did not believe the citadel doctor’s denials. Here as elsewhere the speaker is reacting to the difficulty that the citadel doctor had, despite torture, claimed that he had not administered poison. I do not understand why in an earlier sentence Burman proposed to read Unde venenum tam celeriter parasti rather than praeparasti. But here I hope to contribute something of use to critics of our author, while at the same time keeping to my principle of never writing a paper without mentioning prose rhythm. When Calpurnius ends a sentence with a word forming a dichoreus or double trochee, his normal practice is to precede it with a proparoxytone word. Thus in a passage recently mentioned, the (paroxytone) dichoreus flagitábat is preceded by the proparoxytone médicum. The rule is observed even when the dichoreus is divided into monosyllable + trisyllable: thus at the end of the piece non venénum is preceded by the proparoxytone deséruit.

²⁰ Aizpurua (2005), 226 n. 44.

272

Papers on Quintilian and Ancient Declamation

I call it a rule. It holds good in about thirty-six cases out of the forty-five (80 per cent) which I count in the whole of Calpurnius.²¹ I shall not enlarge on the fact that this clausula is what is in the cursus system called the velox. I have written in one of my more impenetrable articles about the way in which dichorei over the centuries turn from elements in the metrical system of rhythm into elements of the cursus.²² But Calpurnius is very relevant to this story, and it may be relevant to the question of his date (later rather than earlier, I should personally judge). In view of this rhythmical practice, it would perhaps be unwise to abandon praeparasti in favour of parasti in the passage from which we started. We may also look at an early sentence in the piece, printed by Håkanson as Confingunt nocendi voluntatem, postquam sanandi rationem perdiderunt. Rationem is a conjecture introduced by Håkanson, the manuscripts being divided between rem (BC) and remedium (B²MN). The latter had never been questioned until B and C were discovered, for Pithoeus bequeathed N’s remedium to his successors. Rem, and rationem too, break my rule, so far as it is a rule.²³ But remedium conforms to it. I do not say that this decides the matter (for the locution is strange), but rhythm is a factor to be weighed, here and in other places in Calpurnius. To return to my story: I should not omit mention of Pieter de Fransz, Latinized as Francius, whose ideas are occasionally reported in Burman’s Variorum. He has not had a good press, but he was certainly not without gifts. He was part of the Dutch network. Younger than Schultingh but older than Burman, he lived from 1645 to 1704, serving first as professor of History and Eloquence at Amsterdam, then of Greek in the same university. He was not an editor. Rather, his métier of eloquence led him to compose poetry and many orations, two on the Ratio declamandi.²⁴ In his preface Burman speaks of having access to a codex which had once belonged to Francius.²⁵ It contained the Institutio oratoria and the declamations (what could it have been?—it would be nice to track it down). Burman got it from ‘Johannes Bouersius’ (sic), professor at Deventer, whom I cannot trace. It is apparently from this manuscript that Burman cites some conjectures by Francius. For instance, he thought to read tyrannicidio in the first sentence of Decl. 13, and the dative is certainly normal after the markedly Quintilianic locution

²¹ Details in Håkanson (2014), 124; Håkanson, counting on a different system, reckoned about 65%. He did not apply these findings to the cruces I discuss. See also Santorelli (2017). ²² Winterbottom (2011), 262 6. ²³ Rationem is classified by Håkanson (2014), 125 under a much less common type of rhythm. ²⁴ I have looked at these; they are hardly relevant to what we think of as declamation. ²⁵ Burman (1720a), praefatio, page following **** 2. For more on Francius’ library see Win terbottom (1962a) and (1964a).

The Editors of Calpurnius Flaccus

273

detrahere fidem.²⁶ But Schultingh (Burman attests) had made this emendation before him. Ever since Pithoeus first published Calpurnius, progress had been made only in conjectural emendation. No search was made for new manuscripts; editors merely felt it their task to improve and sometimes explain the text as they found it in printed editions. The nineteenth century saw new conditions that made tracking down manuscripts easier than before. Books passed from monasteries or private collections into city libraries. Catalogues began to be made. Lachmann, theorist as well as practitioner, laid the foundations for a scientific approach to textual criticism. Such developments brought distinguished results in some authors. Karl Halm in the 1860s recognized for Quintilian’s Institutio oratoria three primary manuscripts: an Ambrosianus that had been known but little exploited before him, and books from Bern and Bamberg that had not been known at all. These three remain the basis for any modern text. But, equally important, Halm brought to his task an admirable acuteness in conjecture. His new witnesses were the foundation of his masterly text, but they did not dictate it. So too with the Elder Seneca. H. J. Müller in 1887 combined identification of the primary witnesses with excellent critical acumen. The Minor Declamations did not need such basic work on the manuscripts, for, as we have seen, Pithoeus had edited this text from the only old witness. And the editio princeps of 1494, for all its faults, did give access to a different, though much less good, branch of the tradition. It was left to Constantin Ritter to build on these foundations. But the value of his Teubner of 1884 lies especially in his own excellence in conjecture, and that of his mentor Erwin Rohde. The fate of the Major Declamations and of Calpurnius Flaccus during this period was less happy. Both were edited by Georg Lehnert: the declamations in 1905, Calpurnius a little earlier, in 1903. For the declamations, Hugo Dessauer had recently produced a thorough analysis of the manuscripts, which Håkanson later did little to modify.²⁷ Lehnert himself worked out the relationships of the far fewer witnesses to Calpurnius. On this basis he might have been able to produce satisfactory texts for both Calpurnius and the Major Declamations. That he did not was the result of his own deficiencies. He did not lack intelligence or diligence, but he was in the grip of a textual conservatism that might have been appropriate to some classical texts, but made no sense at all in the books he chose to edit. Few ancient books in Latin are more difficult to understand than these; and, to make matters worse, they are not well transmitted. It is just not good enough for an editor to print what he judges to be the archetype of his authors with insufficient citation—let alone

²⁶ Quint. 2.17.15 (and nine other instances); also [Quint.] Decl. mai. 10.13 (211.18 19 H.). ²⁷ Dessauer (1898).

274

Papers on Quintilian and Ancient Declamation

acceptance—of earlier conjectures, and with little attempt to explain any difficulties.²⁸ Lehnert’s texts have minimal value apart from their apparatuses, and even these Håkanson found could not be trusted in detail. The texts themselves represent an abnegation of the task of an editor. That task, I take it, is at a minimum to print a text founded on the manuscripts and to emend it where the editor judges it necessary. If the printed text is hard to understand, the apparatus should suggest possible remedies or refer us to printed discussions by earlier scholars. An obelus should be used in moments of despair. These principles were unknown to Lehnert, or rejected by him. In Lehnert’s defence I can only cite his own words in the preface to the Major Declamations. After remarking on the proliferation of emendations on display in Burman’s edition, of which he promises to cite a selection in his apparatus, he goes on: ‘I myself have purposely made few innovations. The wording of the declamations is better transmitted than you might think at first sight. If one more carefully examines the connection of sentences, and takes account not only of the (so to say) corrupt eloquence of the declamations but also of the manner of expression, which needs to be observed as carefully as possible and to be illuminated from texts that indulge in the vulgar language, it will become evident that many of the emendations that looked easy (lenes) in fact lack weight (leves). Accordingly, my only aim in this edition was to put on display the readings of the manuscripts as carefully as possible, and with their help, so far as possible, to try to recover the pristine form of the text after the removal of blots and errors.’²⁹ ‘Care’ is appealed to three times in this apologia, but that is not the only virtue required of an editor. And it is not clear how blots and errors in such a text are to be removed if not by conjectural criticism. Lehnert’s next sentence (with a further appeal to carefulness) looks forward to a felix eventus of work by others on these declamations. He certainly did not achieve such a happy result himself. I know nothing in Lehnert’s philological background that accounts for his editorial practice. He seems to have trained at Leipzig when Otto Ribbeck (1827–1898) was still in post; but Ribbeck was by no means a conservative critic. The slightly younger Alfred Klotz (1874–1956), whom he mentions as a friend in both volumes, was not (I think) conservative either. Another friend was the

²⁸ This is not true of the Major Declamations, on which he published some adversaria in Lehnert (1903b) (not listed in Håkanson (1982),  ). ²⁹ Lehnert (1905), : ‘Ipse pauca novavi neque id sine consilio. Melius enim quam primo aspectu credideris, declamationum verba tradita sunt. Accuratius enim sententiarum nexu perspecto et ratione habita non modo corruptae, ut ita dicam, declamationum eloquentiae sed etiam loquendi usus qui quam accuratissime observandus atque ex scriptis quae indulgent sermoni vulgari, illustrandus est, apparebit multas earum emendationum quae videbantur esse lenes re vera leves esse. Quam ob rem meum tantum erat hac in editione codicum lectiones quam accuratissime proferre atque earum auxilio quantum fieri potuit, maculis detersis mendisque deletis conari pristinam textus formam recuperare.’

The Editors of Calpurnius Flaccus

275

excellent Dessauer, mentioned earlier for his work on the manuscripts of the Major Declamations (Dessauer 1898). He died very young, and did not live to be an editor. His friend’s edition is dedicated to his manes. Lehnert himself was born in 1871. It is presumably he who wrote a Leipzig dissertation De scholiis ad Homerum rhetoricis (1896). In 1901 he signed himself as being at Munich when he briefly introduced an article by the now deceased Dessauer.³⁰ But in January 1903 he was back in Leipzig, where he dated the preface of his Calpurnius (dedicated PATRI OPTIMO). There is no similar indication in the Major Declamations edition of 1905. But he worked in the Giessen University Library from 1903 to 1913, and was later occupied in library and archival activities in and around Giessen.³¹ Two brief pieces on declamatory topics came out in the early 1930s³² and, for the Hepding Festschrift in 1938, a piece on ‘Zauber und Astrologie in den erhaltenen römischen Deklamationen’.³³ Lehnert also contributed to Bursians Jahresberichte, the final item being a report on Greek and Roman Rhetoric in the years 1915 to 1925.³⁴ This was published (for some reason) only in the volume dated 1944–55; its author died in 1944. We can see Lehnert’s principles, and his judgement, at work in miniature in Calpurnius 13. He cites several conjectures, but accepts only two—one of them Burman’s perhaps unnecessary clamabant. He accepts from BC the new reading nativum, without mentioning that B², like Pithoeus, has the more plausible notum. He does register that all the manuscripts give not deseruit me but deseruit nec in the final sentence. Håkanson was able to base on this a brilliant and elegant conjecture, but Lehnert does not seem aware that his own new version of the sentence is as meaningless as the earlier. In sum: Lehnert gives us materials for a text, but not a text, or at least not one superior to that of Pithoeus. I am aware of having spoken harshly of this scholar, and shall make some amends by quoting from a letter written me by the humane Antonio Stramaglia while we were jointly investigating Lehnert’s hitherto obscure life: ‘È bello avere un’immagine un po’ più concreta di uno studioso che non è certo stato in prima linea fra i “più bravi”, ma che comunque si è dato da fare, pur dovendosi contentare di un lavoro di secondo piano.’ I come now, and finally, to Lennart Håkanson, about whom such reservations do not need to be expressed. Lennart, whom I was privileged to know as a friend, died at the age of 47. He had done a great deal that had won him the highest admiration of textual scholars working on Latin prose and Latin verse. But he was cut off in his prime. He should be alive today, and should have contributed to this volume. ³⁰ Dessauer (1901), 416. ³¹ Cf. Gundel (1957), 206. ³² [Lehnert (1930) and (1932).] ³³ [Lehnert (1938).] ³⁴ Among his many surveys, Lehnert (1920), (1935), and (1944 55) are of particular relevance.

276

Papers on Quintilian and Ancient Declamation

I shall, however, try to avoid hagiography, tempting though that is.³⁵ First, as a background to the dry details of the career, some words from Alf Önnerfors’s obituary on the man:³⁶ ‘[T]he thoughts he exchanged with his friends over the chessboard or during fishing expeditions on Lake Bosarp’ (where the family owned a summer house) ‘were about other (sc. than academic) matters, the essential ones. Lennart was an exceptionally wellintegrated and harmonious human being. Those of us who knew him cannot forget his open, unspoiled boyish spirit, his genuine modesty, and the spontaneous pleasure he took in the good things of life; but most of all his warm friendship.’ But he did, as people often said, live for his work. When I went to stay with him in Lund, he retired to work in the evening. Meanwhile, his wife Monica (with whom I am still in touch) turned over the pages of family photograph albums with me in the sitting room. His colleagues in the university told me that he was well known for forgetting to turn off the lights of his Volvo, and had regularly to summon assistance when his battery failed. I also met him on academic occasions, first when he lectured in Oxford on the text of Lucan. I think we felt in advance that this young man could not improve on A. E. Housman. By the time the lecture was over, we were all aware that a new star had swum into our view: at least as acute as Housman, and quite without Housman’s absurdly polemical tone. Later, I heard him give a paper at Cambridge for an occasion in honour of C. O. Brink; his topic, the historical fragments in the Suasoriae of the Elder Seneca, was a precursor of his posthumously published work, the Teubner of that difficult author. But for me the climax of our acquaintance was our correspondence while I was editing the Minor Declamations. It was typical of him that, when he heard of the progress I had made on my edition, he not only abandoned his own project of editing this text with the German scholar J. Dingel, but sent, over a long period, more and more emendations and interpretations for me to make free use of. When I say that I was also at the same time receiving similar help from D. R. Shackleton Bailey and W. S. Watt, you will be able to judge how exciting that period was for me. Håkanson was born in Karlsborg in 1939, and at the University of Lund was taught by another Latinist of the highest distinction in Latin textual criticism, Bertil Axelson, himself the pupil of Einar Löfstedt: an apostolic succession indeed. Axelson found fault with his thesis on Statius’ Silvae: he wrote of ‘numerous faulty judgements’, of an ‘irritating lack of rigour and pregnancy in presentation’. Later, when Håkanson was a lecturer at Lund, the tone ³⁵ In what follows, I draw (with the permission of the editors) on my account of Håkanson’s career, which has appeared, in German translation, in the second volume of his Unveröffentlichte Schriften (Håkanson (2016),  ); this book contains a complete list of his writings (ibid.,  ). ³⁶ Önnerfors (1988); kindly supplied to me (with other material) by Gerd Haverling, and kindly translated for me by Francis Lamport. See ibid., 160 for Axelson’s words quoted below.

The Editors of Calpurnius Flaccus

277

changed: ‘H. has matured remarkably quickly: latterly he has tackled the major pseudo-Quintilian declamations and in our seminars has produced a surprisingly long series of really excellent emendations.’ Axelson noted correctly, though in apparent depreciation of Håkanson’s efforts, that the Major Declamations had been seriously neglected by scholars. That observation cuts both ways. Håkanson was breaking new ground, and it needed an original talent to do justice to a whole genre that—in both Greek and Latin—called for more sympathetic and scholarly treatment than it had ever received before. That of course was true also of the edition of Calpurnius (1978), which came out four years before that of the Major Declamations (1982). It had been preceded by three articles in Eranos discussing individual passages (published in 1972, 1974, and 1976). In a letter to me dated 1975 enclosing offprints of two of these, he says: ‘I must admit, that I have changed my mind on more than one or two of the problems discussed, partly because I have now collated all MSS existant [sic] to this author and prepared an edition, which will be published as soon as I can find anyone willing to print it (for the moment I am trying to tempt the venerable Teubner Verlag).’ Teubner was indeed tempted, and the book came out in 1978. Let us look, for almost the last time, at Decl. 13, and at the concluding sentence: In arce enecavimus: deseruit nec venenum. This is what Lehnert printed, noting that BC read arcem. Håkanson had already seen in 1972 that in this unintelligible string of words we must read: In arce me nec animus deseruit nec venenum. Clarity and point result from the change of a u/v to an n. This is the sort of emendation that looks obvious once someone else has made it; it was not obvious to Lehnert, and it might not have been obvious to any of us either. For all Håkanson’s work on manuscripts, it was of course textual criticism that was his real forte, as can be seen from the long list of his works. He published, beside the editions I have mentioned, major collections of conjectures and interpretations on Statius and Silius Italicus, and articles on the text of many other authors, some of them late. It may well be that, had he lived, he would have done yet more for late and medieval Latin, in the great tradition of Swedish scholarship. But the authors he most favoured were those who raised the problems in which he excelled, by their often tortured cleverness not only of expression but of argument that constantly keeps readers on their toes. It is by no means always a matter of emending the wording. A solution may often lie in the elucidation of a bizarre train of thought, or the re-positioning of a comma misplaced in previous editions. In this kind of writing Håkanson moved as easily as if he had himself sat at the feet of a Roman rhetor. In working on the texts he did, he showed a sure sense of what he did best. Håkanson buttressed his critical work by some publications on wider topics: on homeoteleuton and adverbs in Latin dactylic poetry, and on prose rhythm. He also published a translation into Swedish of Apuleius’ Metamorphoses,

278

Papers on Quintilian and Ancient Declamation

judged by an obituarist to be ‘surely one of the most successful in any modern language’.³⁷ What he did not pronounce upon, at least in public, was what we distinguish as ‘literary’ criticism. In an age of theory he may well have felt that silence was the best course of action, and that the authors of antiquity are best served by trying to understand exactly what they were saying. At any rate, he told Alf Önnerfors that for the moment he was much too occupied with projects in textual criticism to turn in that other direction. By 1980 Håkanson had been elected to the chair of Latin at Uppsala, only to die in the sea off Crete on 19 June 1987. He left behind, as eventually transpired, a corpus of unpublished work. Two volumes drawn from it have recently been published.³⁸ If, then, we look back over the story I have been telling, we can discern a pattern familiar (mutatis mutandis) from the story of many, though by no means all, classical texts. In the early period, down to 1800 or so, scholars, freelance (Pithoeus) or, later, holding academic appointments (Obrecht, Burman), worked to correct texts without taking much interest in tracking down further manuscripts. In the nineteenth century, it became possible, and normal, to search for better manuscripts than those used hitherto. Lehnert did this service for Calpurnius Flaccus. In the twentieth century, it became easier and easier, and it was seen more and more to be desirable, to try to look at as many witnesses as possible. But at the same time the law of diminishing returns was asserting itself. R. A. B. Mynors, for all the thoroughness of his work on the manuscripts, produced a text of Pliny’s Letters that hardly differed from H. Keil’s. The same is true of L. D. Reynolds’s work on the Younger Seneca, and my own on Cicero’s De officiis. If modern texts do improve on their predecessors, it is normally because they draw on the work of exceptional textual critics. Håkanson’s Calpurnius, and his editions of the Major Declamations and Seneca Rhetor also, fall into this category. He by no means neglected the manuscripts. But the basic work on them had been done before him. What he brought to them was something only he could bring: a sure judgement and a conjectural flair that marks him off from the other editors I have discussed, good as some of them were. D. R. Shackleton Bailey did not praise lightly; but of Lennart Håkanson’s critical powers he judged that they ‘would have been remarkable at any period in the history of philology’.³⁹ * * * * * * I should like to finish with a word or two about someone who has done a great deal for declamation without ever editing a declamatory text. Lennart ³⁷ Staffan Fogelmark (Sydsvenska dagbladet snallposten, 27 June 1987). ³⁸ Håkanson (2014) and (2016). [Add Håkanson Winterbottom (2015).] ³⁹ Shackleton Bailey (1976), 73.

The Editors of Calpurnius Flaccus

279

Håkanson was cut off in his prime. His elder, Donald Russell, is still very much alive at the age of 97. I had the pleasure of discussing Calpurnius 13 with him some weeks before the Paris conference,[⁴⁰] and very much benefited from our talk; he was as sharp and engaged as he had been when he first introduced me to the Elder Seneca in 1954. We disputed over some of the cruces I have been talking about. On one, near the end of the piece, he suggested that we should accept moris, Pithoeus’ conjecture confirmed by BC, but disjoin it from genus: ‘it is customary that . . . ’. If that is right, we may perhaps take the short phrase Notum hoc genus to mean ‘This sort of thing is well known’. It was sad that we could not proceed to email Lennart and ask him what he thought.

APPENDIX

a. Calpurnius Decl. 13, ed. Pithoeus XIII. Medicus tyrannicida.  . Tyrannus suspicatus sibi venenum datum ab eo medico quem in arce habebat, torsit eum. ille pernegauit. Misit ad medicum ciuitatis. dixit datum illi ab illo venenum, sed se remedium daturum, dedit poculum, quo exhausto statim perijt tyrannus, contendunt de praemio.   . Asit, sanctissimi iudices, vt hanc vos fidem tyrannicidij detrahatis, quam & medicus confirmauerit & tyrannus. Confingunt nocendi voluntatem, postquam sanandi remedium perdiderunt. Poenas meas hinc cogitate, in quibus nec ira, nec natura cessauit. Tolerabilis vis est vbi ad consuetudinem mali, caussa necessitatis emergit. Præmium consequitur qui ausus est & confirmare meum venenum, & suum remedium polliceri. Vltio quidem illa non quæstio. Tyrannus venenum quæsiuit? sed vindicauit. O quam facile gerunt persuasiones, inlecebramque in contrarium transferunt. Vt virus serpebat interius, & artus omnes longa pœnarum dilatione languebant, veneficium iam tyrannus agnouerat, quia instantem interitum sentiebat. festinans medicum flagitabat. vnde venenum tam celeriter præparasti? dicis forte, Maior mihi dandi veneni fiebat occasio, quæ ex ipsius voluntate veniebat. Notum hoc genus mortis est, vt ex sensu priore, ad cuncta cautior sollicitudo procedat. nonne iam apud tyrannum cuncta suspecta præsens formido faciebat? Medicum tota arce clamabat. quasi ego de tyrannicidio non negassem. In arce enecauimus. deseruit me venenum. ⁴⁰ [To whose proceedings the present paper belongs.]

280

Papers on Quintilian and Ancient Declamation

b. Proposed text⁴¹ 13. Medicus tyrannicida  . Tyrannus, suspicatus sibi venenum datum ab eo medico quem in arce habebat, torsit eum. Ille pernegavit. Misit ad medicum civitatis. Dixit datum illi ab illo venenum, sed se remedium daturum. Dedit poculum; quo exhausto statim periit tyrannus. Contendunt de praemio. 1. Absit, sanctissimi iudices, ut hanc vos fidem tyrannicidii detrahatis, quam et medicus confirmaverit et tyrannus. 2. Confingunt nocendi voluntatem, postquam sanandi remedium perdiderunt. 3. Poenas meas hinc cogitate, in quibus nec ira nec natura cessavit. 4. Tolerabilis vis est ubi ad consuetudinem mali causa necessitatis emergit. 5. Praemium consequitur qui ausus est et confirmare meum venenum et suum remedium polliceri? 6. Ultio quidem illa, non quaestio. 7. Tyrannus venenum quaesivit? Se vindicavit. 8. O quam facile suggerunt persuasiones, illecebraque in contrarium transferuntur! 9. Virus serpebat interius, et artus omnes longa poenarum dilatione languebant: veneficium iam tyrannus agnoverat. 10. Quia instantem interitum sentiebat, festinans medicum flagitabat. Unde venenum tam celeriter praeparasti? 11. Dicis forte: ‘Maior mihi dandi veneni fiebat occasio, quae ex ipsius voluntate veniebat.’ Notum hoc genus: moris est ut ex sensu priore ad cuncta cautior sollicitudo procedat. Nonne iam apud tyrannum cuncta suspecta praesens formido faciebat? 12. Medicum tota arce clamabat, quasi ego de tyrannicidio non negassem. 13. In arce me nec animus deseruit nec venenum. Thema venenum datum B²N: verecundatus BC 1 tyrannicidio Schultingh, Francius 2 remedium B²N: rem BC: rationem Schaub ap. Håkanson 5 consequetur Pithoeus 6 7 ultio . . . vindicavit post emergit (4) transpos. Schultingh 8 suggerunt tempt. Winterbottom: gerunt codd.: cedunt vel vertunt Schultingh | inlecebraque Schultingh: inlecebramque codd. | transferuntur BC: transferunt ut B²N 9 serpebat . . . longa Schultingh | serpebat B²N: scribebat BC | veneficum Winterbottom (1995), 42 [= R.10 infra, p. 343] 10 parasti Burman 11 veneni B²N: bene BC | notum B²N: nativum BC | post genus distinxit Russell | moris BC, Pithoeus (e coniectura): mortis B²N: timoris Gronovius: morbi Burman 12 clamabant Burman ⁴¹ The apparatus given here is not what I should print in an edition of Calpurnius. It includes all the emendations mentioned in the course of this paper, and excludes manuscript variants except those mentioned there.

The Editors of Calpurnius Flaccus

281

13 in arce me nec animus Håkanson (1972), 65: in arce (arcem BC) enecavimus codd.: in arce enecavi: non Schultingh | nec² codd.: me Pithoeus, errore legendi, et sic edd. usque ad Lehnert

c. Translation 13. The doctor who killed the tyrant  -     . A tyrant, suspecting he had been given poison by the doctor he kept in the citadel, tortured him. He absolutely denied it. The tyrant sent word to the city doctor,⁴² who said that he (the tyrant) had been given poison by him (the citadel doctor), but that he would supply a cure. He gave him a draught, on draining which the tyrant fell dead on the spot. They dispute the prize. 1. Far be it from you, most reverend judges, to refuse to believe in a tyrannicide that has been confirmed both by a doctor and a tyrant.⁴³ 2. They (the doctor and his advocate) are fabricating an intention to harm , after failing to cure him. 3. Picture my punishment from this: neither anger (the tyrant’s) nor nature (the doctor’s endurance) failed. 4. Force can be borne when necessity results in⁴⁴ habituation to evil. 5. Is the prize going to one who had the face both to confirm my poison and to promise a remedy of his own?⁴⁵ 6. That was revenge, not an investigation.⁴⁶ 7. The tyrant did not enquire into poison, but revenged it. 8. How easily they supply convictions,⁴⁷ and are put into reverse by the lure (of the prize)! 9. The poison crept further in, and all the limbs grew languid as the punishment was prolonged: by now the tyrant had recognized he had been poisoned/who had poisoned him. 10. Because he felt death was upon him, he asked for a doctor in a hurry. How did you manage to prepare a poison so quickly?

⁴² Rather than ‘sent for’? ⁴³ The (city) doctor by saying the citadel doctor gave poison to the tyrant (cf. § 5), the tyrant by suspecting he had. ⁴⁴ Though emergit is odd. ⁴⁵ Perhaps not a question. ⁴⁶ i.e. the tyrant had decided in advance what the answer was. ⁴⁷ Persuasio in the plural is found in Quintilian, e.g. 3.1.6 with mutare.

282

Papers on Quintilian and Ancient Declamation

11. Perhaps you say: ‘I had a better opportunity to administer poison because it arose from his own wish.’ This is a familiar situation: it is customary that as a result of previous experience one is more cautious towards everything. Did not his present terror now make everything suspect?⁴⁸ 12. He called for a doctor all over the citadel, as though I had not denied I had tried to kill him. 13. In the citadel I was failed neither by my spirit nor by my poison.

⁴⁸ Moris est is common in Quintilian. The city doctor claims that he had intended to give poison, and that he had had a better opportunity to give it than the citadel doctor because the tyrant had sent for him. The reply is: no, you didn’t, because the tyrant would have been more cautious this time round.

23 The Words of the Master 1. The text that provides the nearest extended parallel to the ‘Quintilianic’ Minor Declamations is Sopatros’ The Division of Controversiae. Here the rhetor presents eighty-two cases, grouped according to the issue (status) underlying each. Each is ‘divided’, that is, given a structure based on a system not unlike that of the second-century AD Hermogenes. Technical discussion is accompanied by passages of declamation, varying in length, to illustrate the points being made. Nothing could be more orderly, or, in its obsessive way, more sophisticated.¹ The Minor Declamations² too combine advice from a rhetor with declamatory illustration. But they are far less systematic, and far less technical. The status system is hardly mentioned,³ and though cases are often given a ‘division’ it is one that typically just lists ‘questions’ to be dealt with.⁴ There is no obvious intent to order the material, and rarely any sense of books or life

[Maia 70 (2018), 73 83] ¹ Stramaglia (2010) (a contribution which very helpfully brings together the limited evidence for procedures in ancient rhetorical schools) puts alongside each other the sermo of 247 and a passage of Sopatros, contrasting the level of sophistication they display. Russo (2013) shows the riches that are emerging in this area from the evidence of papyri. Giovanna Longo promises a survey of this exciting field. ² I cite them by number and where necessary section from my own edition (Winterbottom (1984)). By ‘(D. R.) Shackleton Bailey’ I mean that scholar’s Loeb (Shackleton Bailey (2006)), not his earlier Teubner (Shackleton Bailey (1989b)). ³ The word status appears in its technical sense only at 320.1 2, where the Master (not, he says, for the first time) gives the ‘easiest’ method of establishing the issue in question (see also Winterbottom (1984) [henceforth usually: ‘my n.’] on 276.1; Dingel (1988), 66 7). In 247.1 6 the Master ‘shows the way’ (cf. Quint. 2.6.2 with Reinhardt Winterbottom (2006), 144) to deal with a case involving definition (see also 270.3 4; 292.1; 349.2). At 299.2 qualitas does not allude to the status of that name (the word is picked up by the topic quale huius factum, but that only as an element of the epilogue). Here and elsewhere we cannot be sure what the Master may have said in earlier lost sermones (hence, no doubt, references back that do not work for us, such as 261.1: quoque; 316.2: paulo ante (!); 320.1: saepe vobis dixi; 331.19: dicebam). We should naturally not infer that the Master was unaware of the details of the status lore; thus for the topics of coniectura see my n. on 281. ⁴ In this respect resembling the ‘division’ sections in the collection of the Elder Seneca.

284

Papers on Quintilian and Ancient Declamation

beyond the classroom.⁵ In these circumstances, it seems likely ‘that our collection represents the Nachlass of a rhetor, inefficiently edited and published after his death. We have before us, on this view, not notes taken down at another’s lecture, but notes made by a lecturer for his own purposes. The lack of system will reflect the editor’s inefficiency or lack of understanding in the ordering of a confused archive . . . The sermones will be the notes, to be expanded in front of the class, on the treatment of each topic. The declamationes will be fair copies for dictation or performance.’⁶ We have before us, then, with all its puzzles and defects, the combined work of a dead Master and a floundering Editor. I propose here to consider the words of the Master—what the manuscripts call his sermones—, his informal addresses to his class,⁷ to see if anything can be usefully made of them.⁸ 2. Out of the original 388, 145 items have come down to us. In fifty-eight of these (40 per cent), there is declamatio but no sermo. In thirteen (9 per cent) there is sermo but no declamatio.⁹ Where a single sermo is present, it most often precedes declamatio (thirty-six cases, 25 per cent). In seven cases (nearly 5 per cent) a single sermo follows declamatio. These ‘normal’ cases total 114 (nearly 79 per cent). The remaining items show a variety of combinations, including a single sermo surrounded by declamatio (eight, rather over 5 per cent) and two sermones separated by declamatio (three cases, 2 per cent). In the cases—a quarter of the whole—where sermo precedes and introduces declamatio, the topic is often divisio (I shall return to 270, where the concept is discussed: below, pp. 293–4). 249.1 is typical. Without introduction or overt grammatical construction, we have a list of questions, culminating in ‘What comes last in all controversiae’, what is aequius.¹⁰ Each of these topics, except the last, is taken up in the declamatio, and in the same order. Or to put it the other way round, we could deduce the quaestiones from the speech, where the declaimer often echoes the words of the sermo; thus 3: non utique semper exigi ut prius agatur cum adultero (varied at 14); 8: egi cum adultero (cf. 14: et a me actum); 13: a more extended plea, but reprised at 14 with words closely ⁵ There is a not very impressive use of Cicero’s Pro Cluentio at 388.32. ⁶ I cite Winterbottom (1984), . I still hold this view. ⁷ Contrast the much more formal (and ‘published’) introductions found in later Greek collections; the field is helpfully surveyed by Nolè (2013 14). ⁸ Earlier, Dingel (1988), 11 13; recently, Oppliger (2016). I did not profit from Gunderson (2003), chapters 2 4. I have naturally drawn on my commentary (Winterbottom (1984); see above, n. 3), but without citing it much in extenso; and I hope to have added some novelties. ⁹ For this group see below, p. 287. ¹⁰ A sermo sometimes identifies, in mid declamation, the transition to aequitas: most overtly in 245.4; 266.6; 309.11; 336.12; 340.10; 374.7. At 336.12 (summae partes) and 374.7 (ultimae partes) the Master makes it sound as though the epilogue too concerns equity; for such phrases of the peroration see my n. on 259.23. Note also 271.10: Post haec dicemus domui etiam honestam esse hanc contentionem; 273.13 (where circumscriptione implies emotive treatment of the cheat ing, illustrated in 14).

The Words of the Master

285

recalling the sermo. The summary at 14 marks the turn from the legal questions to the intention of the state in giving this praemium (16: voluntatem . . . rei publicae . . . tempore illo quo praemium dabat, with elaboration of the quaestio summarized in the sermo: Utrum . . . an . . . dederit). The structure of the speech is made so clear that the formal sermo was hardly necessary. Sermo and declamatio work hand in hand.¹¹ By no means all of the sermones are taken up like this in the model speeches. I register in a footnote some others that are (more or less),¹² but I shall not dwell on the exceptions; they witness, like so much else, to the disorganized state of the collection. At the extreme, we have cases like 366, where the Master says that the killer will defend himself on a point of law, whereas the declamatio is put into the mouth of his accuser. In 249, and some other cases,¹³ the sermo is no more than a list of quaestiones. But of course such lists might accompany other material, as at 254.2; 280.2; 281.1 (all worked out in the succeeding declamationes). Equally, the bare list of headings may be phrased more expansively, as in 266.1–4 (with illustrations in reported speech, as well as the snatch of declamatio in 5)¹⁴ and 359, where no declamatio follows. Most didactic of all is 271, where, after the initial sermo has given the first question, short sermones point out the structure,¹⁵ with brief illustrations until we come to a longer passage on aequitas. Here the manuscripts distinguish SERMO and DECLAMATIO,¹⁶ but the whole is hardly different from what we see in 266, except that here the illustrations are given in direct speech. Equally bitty is 315.1–7, where the snatches of declamatio hardly merit the title.¹⁷ There is an unusual

¹¹ To give one more example, in 307 the sermo: Duo nobis efficienda sunt (for this formulation cf. 345.1, where it is picked up by (dives) non potest aliter efficere quam si . . . ), ut conscium habuerit, ut hunc habuerit is taken up in the declamatio by nego fieri potuisse ut proditor conscios non haberet (3) and Videamus quos (sc. conscios) habuerit (5). ¹² See my nn. on 252; 255; 280.2; 284; 317; 333; 343. Other declamations, though they lack sermones, are given clearly marked structures instead: 253 (quasi suasoria); 268; 296 (repetition of obicio quod); 322; 330; 332. ¹³ See my n. on 249.1, where I cite the sermones of six other cases (284; 286 7; 299; 303 (no declamatio); 317) as parallels. But only in 284 and 317 do we find the speech taking up the sermo so closely. ¹⁴ Discussion in Dingel (1988), 11 12: ‘ein so ausführlicher und gut gegliederter sermo die declamatio überflüssig macht’ (p. 12). ¹⁵ Discussed by Oppliger (2016), 107 11. Sermones often signal transitions; see above, n. 10 (to aequitas). So from verba legis to its voluntas (308.16), from one side of the case to the other (331.14, in addition to the heading PARS ALTERA; at 274.10 the other side is given a new sermo of its own). There is a lively one at 388.19: Summovimus peregrinum cadaver; restat ut inducamus filium (vigorously rendered by Shackleton Bailey: ‘We have got rid of the foreign corpse; remains to bring in the son’; but perhaps ‘bring on’: cf. OLD² s.v. induco 3a). ¹⁶ Shackleton Bailey quarrels with them in §§ 5 6. ¹⁷ Compare Oppliger’s analysis ((2016), 113 15) of 293 as a ‘forme hybride entre la déclama tion et le sermo’ (p. 116).

286

Papers on Quintilian and Ancient Declamation

‘narrative illustration’ in 315.9 (Hic clamor ille—‘the familiar’: meta-rhetorical— utriusque exercitus . . . ). It does not surprise that when such illustrations are in oratio recta, they are normally introduced explicitly: thus 245.4 (illa dicere solet);¹⁸ 247.1–6.¹⁹ Where they are in reported speech, introduction is sometimes provided (266.1: illud dicere solemus and 2: illud dicetur; 349.2: ut dicamus;²⁰ 359.2: dicit enim; 385.8: dicemus), sometimes not (374.9; 376.6;²¹ 380.2, introduced by Hic, sc. dicemus). This observation supports Rohde’s addition of a title at 342.16 (see also my n. on 15). We have seen that in these circumstances there is room for blurring between sermo and declamatio,²² that may affect the way we present the text. In 261 (see my n. on 1) editors need to decide the whole nature of the piece (see my nn.).²³ Is Hoc ad meritum pertinet, illud ad pericula (343.14) sermo (cf. e.g. 309.11) or declamatio? How should we present the similar Haec ad ius, illa ad aequitatem at 250.8, where my n. cites parallel problems (add 342.13)? I suggest there that the practice of the Editor may not have been consistent. The Master may appeal to common²⁴ practice in suggesting quaestiones. Thus, almost all cases concerning madness pass from quid dementia sit to an haec dementia sit (316.1; note also 349.2);²⁵ so too with cases of harming the state (326.3)²⁶ and ones involving optiones (266.1: Initia communia, with my n.) and magistrates (314.2).²⁷ But such precedents did not always make the

¹⁸ At 246.3, an introduction may have been lost in the lacuna. ¹⁹ Dicit several times; possumus dicere (4); si ita finiretur (4) though note also two fictiones (5 and 6: fingamus) not introduced at all. At 343.2 the illustration is introduced only by a phrase suggesting the elimination of the persona of the defendant (Shackleton Bailey’s wording). It may be that we should not use inverted commas here. ²⁰ In 349.3 illustration is introduced by aliquid . . . adplicabimus, quod . . . (then a further sen tence in direct speech). ²¹ Here the oratio obliqua illogically spills over into the advice to thank those who gather up exposed babies (Referendam esse gratiam). ²² In 343.4 sequentia illa sunt, ut probem me et ab hoc circumscriptum esse et indignissime circumscriptum repeat in the declamatio the words of the sermo (1): Non tantum nobis in hoc laborandum est, ut circumscriptus sit adulescens, quantum in illo, ut circumscriptus sit a patre. Ritter wanted to extend the sermo at 315.9. See further Dingel (1988), 13 with n. 28. ²³ 261.1: per summas digeri potest seems to allude to the unusual nature of the declamatio here. Discussion in Dingel (1988), 12: an example of how the declamatio can ‘den sermo ersetzen’ (also p. 119, of 287: here, as often, we must regard ‘den sermo als Ergänzung, nicht als Rezept der declamatio’). ²⁴ ‘Common’ to all cases of a kind, but also commonly employed. ²⁵ Such a movement from general to particular is of course familiar: cf. 270.5; 281.1 (similarly 307.1). The principle is stated and exemplified (of loci) at 244.5; for the fluidity of the concept, see my n. on 299.1. The prima quaestio was accordingly often general (as often in the Elder Seneca; note e.g. Con. 1.1.14); observe 337.2 (in fact the only quaestio suggested). ²⁶ Similar is 271.1 (abdicatio). ²⁷ Cf. 348.13: In summa parte controversiarum talium quaeri solet ecquo animo fecerit hoc imperator; 308.21: Illa iam communia (‘the usual commonplace’ Shackleton Bailey) pro omnibus testamentis.

The Words of the Master

287

path clear. Pleraeque controversiae sub hac lege (sc. inscripti maleficii) duas quaestiones habent, but the Master has often had to say that both do not always need to be raised (252.1), and in this particular case a third has to be added. He acknowledges doubts about the division of 254 (§ 1), and about the ordering of quaestiones in 280 (§ 3). 3. The Master implies in 337.1²⁸ that he regards his job in dividing a case as (at times) including the provision of an appropriate colour. That is seen in action at 285.1, where, after a single quaestio, we are given our orders: Colorate (with a suggestion, or precept, in the direct speech that follows—there is no declamatio). The manuscripts point towards rubrication of the verb here and at 384.1; the more reason not to specify DECLAMATIO in either place. In the string of cases following 352²⁹ where no declamatio is provided, a colour is often given at some length, always in oratio obliqua, though (usually) with careful attribution of the side in question; thus e.g. 353.1: Petitor necesse est . . . dicat and 4: Infitiator dicat oportet; 358.2 (after comment from the Master): Dicat (sc. adulescens).³⁰ But 351.1 and 352.1 have only dicemus. 359 gives only quaestiones. In 360 and 361, though there is some declamatio, the pattern continues: 360.1: nurus dicet and 3: nurus queretur . . . Contra illa dicet; 361.2: Contra ille . . . dicet. When unaccompanied sermones return for a while, there is some variety: 362.1–2 oratio obliqua, with no attribution; 363.1: Maritus illo colore defendendus est, ut dicatur . . . ; 364.1: A parte divitis color ille introduci potest, ut dicatur . . . ³¹ Later, a colour for the other side is given at 386.2 in direct speech (with doubled inquit). Elsewhere, the word color is used in its technical sense remarkably rarely. It appears in declamatio at 333.7 (meta-rhetorical: perhaps to be connected with earlier sermo in 3, where see my n.). The meaning at 280.2: modus et color declamationis and 316.3: contra colorem talis animi is quite different (see my nn.). The grouping together of ‘colour’ cases in 352 ff., a unique sign of ordering, seems to suggest some attempt on the part of the Editor to separate out divisio from colores, as the Elder Seneca systematically did. What the Master thought of sententiae, so prized by Seneca,³² is made clear by his treatment of them in the collection as we have it. He does not use sententia

²⁸ Antonio Stramaglia, whose general help and encouragement I gratefully acknowledge, refers to his note on [Quint.] Decl. mai. 4.6 (67.14 17 Håkanson; see Stramaglia (2013), 120 1 n. 103), which puts into context the disapproval of figured controversiae displayed here. ²⁹ Just before this, we have sermones that supply a colour: 349.1, then 351.1 7, which especially foreshadows the following cases, though here there is a snatch of declamatio. ³⁰ In 357, the sermo begins (1): Haec mulier dicet; at the end the Master sums up (5): Hoc colore optime defendetur mulier. ³¹ This last colour is interrupted by two interesting comments by the Master in §§ 2 and 4. For the meaning of constitutio in 4 see my n. on 315.17. ³² For the three components of Seneca’s work, see e.g. Bonner (1949), 54 7.

288

Papers on Quintilian and Ancient Declamation

in the sense of ‘epigram’ at all, though that does not mean he altogether eschewed their use in declamatio.³³ His attitude is much like that of Quintilian himself; the locus classicus is 8.5, where 25: Duae sunt diversae opiniones, aliorum sententias solas paene sectantium, aliorum omnino damnantium, quorum mihi neutrum admodum placet is echt quintilianisch. 4. In the lengthy sermo prefacing 338, the Master ‘gives a simplified and abbreviated version of Q[uintilian]’s teaching (4.1 and 6.1) on proem and epilogue, tailored to the needs of declamation’.³⁴ The first paragraph of the discussion (1–3) begins with an emphatic statement of the difference between the proem and the other parts of a speech, enforced by an anaphora hardly paralleled in these sermones:³⁵ Prohoemium propriam formam, propriam legem, proprium modum habet. Almost at once, however, we find that the proem, though free of narration and loci,³⁶ is in fact like the epilogue in two important respects. But then come features that make them different again, summarized in a pointed antithesis: ita prima parte efficimus ut omnia audiat iudex, summa parte efficimus ut meminerit eorum quae audierit. Another contrast: the epilogue should make the judges feel well-disposed to the speaker, just like the proem, in fact, but with more emotional appeal and more freely. The proem does use pleading, but shouldn’t tire the judge out. But then the epilogue shouldn’t either. A familiar saying rounds off the paragraph: it is all too true that tears dry up quickly. Quite apart from the infelicitous repetitions in 1 (res (del. Shackleton Bailey) . . . rem . . . re . . . res (del. Shackleton Bailey)),³⁷ the passage hardly impresses. The Master keeps qualifying his basic contrast of proem and epilogue in a way that does not conduce to lucidity, despite the stylistic pretension of the opening emphases and the closural devices.³⁸ This extract from a longer passage raises a question: what is the purpose of this sermo? Though by no means irrelevant, it tells us much more about proem and epilogue than pupils needed to know when composing a speech on this ³³ I remark on one below, p. 289. ³⁴ Winterbottom (1984), 527. My notes supply the parallels with the Institutio; cf. Dingel (1988), 35 7, 158 60. ³⁵ But note below § 5: si iam . . . , iam . . . ; 325.2: non est causa quae recipiat testem, non est causa quae recipiat consignationem; 349.2: qui non . . . qui non . . . (typical of the declamationes: ‘Nur Anaphern sind häufig’ Ritter (1881), 221). Stylistic flourishes are so rare in sermones that I can relegate a further instance to this footnote: 346.2 agere mitem, indulgentem (for ‘asyndeton in paired words’ in the declamatio sections see my Index in Winterbottom (1984), 604). ³⁶ It is made much clearer in the ‘source’, Quint. 4.1.60, that the point is that the proem should not be similar in style to these other elements of the speech. One wonders why the Master does not add mention of the argumenta, as Quintilian did; they are more important than the loci, and may well have dropped out by accident. ³⁷ As for interim solet (2: see my n.), the phrase appears not only at 274.1, but several times in the Institutio (e.g. 8.4.17) and once in the Declamationes maiores, at 3.7 (48.19 H.): hardly anywhere else. See also Dingel (1988), 159 n. 374. ³⁸ For which cf. the final sentence in § 6 (quod dicturus fuerit ~ quod dixerit).

The Words of the Master

289

particular theme, or when admiring the Master’s model. It is § 7 alone that would help a pupil to orient himself on this particular occasion (huic mulieri hodie). Perhaps, then, what we have here is a ‘handout’ to be preserved for future reference, as a substitute for consultation of multiple places in the enormous Institutio. It might have been more tightly written. But it does not read like mere notes made in preparation for a lecture. It is possibly the companion piece for one or more lost sermones dealing with other partes orationis, or at least with narratio. That is not a topic much mentioned in the sermones left to us, though see 281.1 (Narratione praeparandum est ut . . . ; for this role of narration cf. e.g. Quintilian 5.pr.4 neque prohoemii neque narrationis est alius usus quam ut huic (sc. argumentation) iudicem praeparent).³⁹ But the Editor (or even the Master) added the heading NARRATIO at 332.3, 333.6 and 388.3. This last is buttressed by direct statements in the ensuing narration (5): Haec est nostra narratio: ista narret suos piratas (and then Habetis narrationem pueri, habetis et matris: audite liberti). And indeed the Master is quite capable of making the structure of a declamation clear without using headings.⁴⁰ Thus in the declamatio of 338 itself,⁴¹ the proem is clearly marked: it starts as so often, in oratory and declamation, with a concessive clause (Etiamsi . . . ), while the initial address to the iudices recurs at the end in 10, where in prima parte causae dixisse contentus sum confirms that this section of the speech is over.⁴² What will follow is at once made clear by vel sola expositio causae;⁴³ and the end of the narration is more subtly signalled by the recurrence of the theme of the ambiguity of the nurse’s evidence (16). Filium matri vindico contra eum qui adserere coepit nuper (17) is a manifest propositio, leading into the argumentation, marked by a new address to the judges and by the word probationis. The same theme of the nurse’s evidence seems to conclude the argumentation at the end of 27; the freedom and pathos of the epilogue is marked at its outset by the doubled question (Cui non . . . ? Quis non . . . ?) and by the use of the word miserebor. Anaphoras dominate 28–30; and, after a pause marked by the sententia: Nulla potest esse tam longa simulatio, we end with the brief sentences, exclamations, and apostrophes of the last sections

³⁹ In the declamatio at 283.2: Ceteros . . . quos abdicant patres sine narratione culpae abdicare non possunt is meta rhetorical. For 338.1: narrationis forma see above, p. 288. ⁴⁰ In Sopatros, division distinguished not only ‘constituent arguments’ but also ‘some of the “parts” of a speech’ (Russell (1983), 138). ⁴¹ It will be observed that the Master does not comment on, let alone explain, the difficulties of the piece (to which Shackleton Bailey devotes many notes). ⁴² Closure is also supplied by the brief statement of a crucial assertion, that the nurse’s ev idence works both ways. ⁴³ Cf. [Quint.] Decl. mai. 3.7 (48.9 10 H.): Hoc expositionis loco, imperator, malo accusatori bus credas. Illi narrarunt . . . See Stramaglia (2016) for meta rhetoric in that corpus (p. 38 on the passage just quoted).

290

Papers on Quintilian and Ancient Declamation

(31–2). There is a final address to the iudices in 32, and Shackleton Bailey is perhaps too dismissive of his own voletisne in the last sentence of all. 5. It is now time to look at a few other sermones that offer something beyond normal division, especially as they often involve difficulties of interpretation. We saw (above, p. 286) that in 314.2–3 the Master remarks on the fact that this case shares features with others on the same kind of theme. He prefaces this by explaining (1) why he had to repeat material in different sermones. One reason is propter interventum novorum, the changing make-up of the class he addresses. That is clear enough: newcomers needed to be told things that older hands were familiar with. The other reason is obscure: propter condicionem divisionum, glossed, though not much clarified, by the words: ad praesentis materiae controversias nonnihil interest quomodo ego diviserim. The point is perhaps this: It is in the nature of division that a rhetor made it clear how he personally dealt with a particular kind of material;⁴⁴ so (in the present instance) it matters, when cases concerning magistrates come up, that the class as a whole is aware how I (ego is emphatic) have in the past divided them. He feels the need to repeat his own line on the matter. praesentis materiae controversias means not ‘controversies based on a (sic) current theme’⁴⁵ but ‘controversiae concerning the kind of material we are considering today’,⁴⁶ or, as he puts it in 2, controversiis huius materiae. In 272.1, the Master distinguishes lis and controversia. Shackleton Bailey kindly passed over my misunderstanding of this contrast, noting: ‘A controversia may consist entirely of a lis, the point(s) in dispute, but the lis may be accompanied by other elements which are not disputed.’⁴⁷ I am not sure that this necessitates his emendation of potior to potentior,⁴⁸ and think my view of the former as meaning ‘preferable’, i.e. ‘more amenable to treatment’ might still stand. I now offer for comparison 254.23: existimo commodiorem esse partem diversam (‘easier’ Shackleton Bailey). For 292.1, where Shackleton Bailey prints the same emendation, see below, pp. 291–2. There remains the contrast of lis and controversia. Elsewhere (hence my original misconception) this distinguishes between a real-life case and a fiction of the schoolroom.⁴⁹ But this is not true of a parallel in our corpus, 270.3: Tota enim lis et omne discrimen controversiae in hoc positum est; here the second phrase varies the first: both refer to the ‘nub’ of a school exercise. Similarly, in ⁴⁴ With the implication that others might treat it differently. ⁴⁵ So Shackleton Bailey. But what does that mean? ⁴⁶ Cf. 252.19 (praesentis . . . iudicii); 267.6 (defensionem); 325.14 (liti); 334.2 (cognitionem). ⁴⁷ Shackleton Bailey (1989a), 378. ⁴⁸ For which cf. 357.5: ex contraria parte infinitum potens est illud . . . (I register the informal infinitum [adv.] in Winterbottom (1984), XV). ⁴⁹ Of course the Master, like Quintilian, is very conscious of the gulf between them: see 325.5 and 338.5 (with my n.).

The Words of the Master

291

the Institutio, we find passages like 9.2.87, where, in a discussion of school practice, Quintilian says: neque litem intellego in qua pars utraque idem velit (i.e. I can see no point of dispute where both sides agree in their aims).⁵⁰ So in 272.1: Lis . . . circa unum factum versatur the Master must mean that ‘[T]he point at issue is about one act only’ (Shackleton Bailey). There is then a contrast between this nub (for which the Master lists quaestiones) and the controversia to which it is central. The Master’s phrase (ibid.): Potior in hac causa controversia quam lis est seems to be shorthand for ‘The controversia as a whole is more amenable to treatment (or make a more powerful impression) than the legal point at issue.’ Speaking from this note, he would have said something more lucid.⁵¹ Indeed the whole section is ill-phrased, most notably § 2. Shackleton Bailey tolerates, as I did not, Cetera . . . controversiae, as meaning ‘The other elements in the controversy’;⁵² I am now inclined to delete controversiae. I commented on the odd use of themata to mean ‘elements in the case’.⁵³ But maiorem cumulum habent now strikes me as more strange than it did when I glossed it ‘add to the effect’;⁵⁴ Shackleton Bailey’s ‘are more cumulative’ (than the nub?) shows up the oddity. Again, we are in the presence of shorthand that the Master would not have been happy to see ‘published’. The sermo of 292 opens (1): Duplex quaestio est, iuris et facti. Nam etiamsi vis inlata est, quaeritur an causa mortis sit. Sequens potior, an propter vim perierit. Circa ius illud est, ut finitione tractetur. The Master is again being not quite lucid. The first quaestio, concerning fact (coniectura), is: Was force brought to bear? Even granted that it was, a second⁵⁵ quaestio remains, and one ‘preferable’⁵⁶ to the other: Did the boy die because of that act of force? That is a matter involving definition.⁵⁷ And, as the Master proceeds to say, ‘He who brought death to bear is properly the cause of that death.’ In the normal case, of murder of A by B, that is clear enough: B killed A and is the cause of his death. But in the present case, A killed himself because (it seems) of force brought to bear by B. If B is to be found guilty, he must be shown not merely to have made A want to kill himself, but to have made him think he had no

⁵⁰ Other relevant passages are 4.1.26; 5.10.40; 7.2.7; 7.3.7; 7.6.6. ⁵¹ At 345.1 the Master uses the Greek phrase κατὰ συναθροισμόν (see my n.); I assume that he would have explained it in class. Greek words are glossed at 299.1 (pathetice, pragmatice: the forms show that they are being treated as Latin; but patheticos (adv.) follows in 2). The Master ventures on probativae (quaestiones; the Greek will be ἀποδεικτικαί, cf. Quint. 5.10.7: Ἀπόδειξις est evidens probatio) in the same sermo; the word is not found again till the Middle Ages. ⁵² Meaning ‘in this exercise’. ⁵³ Winterbottom (1984), 373. ⁵⁴ Ibid. ⁵⁵ Sequens takes up primum in Quint. 7.9.9. ⁵⁶ For potior here, see above, p. 290. ⁵⁷ Compare closely 270.3 4 and 289.3. The wording in our passage is very inelegant, and savours of the Master’s shorthand; but I take it that circa ius means ‘as to ius’ (bringing us back to the first quaestio; for the phrase cf. e.g. 245.4), while illud est introduces ut finiatur, as we might say: ‘what is needed is definition’. A word may have dropped out. Similar but more easily in telligible is 270.3: Finiamus ergo necesse est quid sit causa mortis. In short, I agree with Pasetti (2018), 132 as against Shackleton Bailey.

292

Papers on Quintilian and Ancient Declamation

alternative but to do so. Especially relevant here⁵⁸ is the formulation in Quint. 7.3.32: Hic finitio est accusatoris: ‘Per quem factum est ut quis periret, causa mortis est’, rei: ‘Qui fecit quid sciens per quod perire homini necesse esset’. The declamatio following this sermo discusses the first quaestio, despite its not being judged potior (or potentior for that matter).⁵⁹ I draw attention finally to the long and unusual sermo prefacing 325 (1–5). § 1 points out the difficulties presented by the case, and does so with vigour. First, a call to attention: Quid dicimus? (more arresting than Quid dicamus?);⁶⁰ then three interrogatives—Quomodo, Quomodo, and Numquid—which structure not only the section but also the declamatio to come (see my n.). As often, sermo can be seen either as giving prescriptions for a speech to be made by pupils or as explaining to them how the model declamatio works (or, of course, both). The same vigour marks the climax of the Master’s interesting remarks on the problems posed by such a case: Quomodo tamen probamus? . . . Ergo debet restituere alicui; si non huic, statim interrogo: cui? (4). Shackleton Bailey finds what follows unintelligible. But I remain happy with the explanation of Habent . . . descendunt, proposed by Håkanson and offered in my commentary: in the nature of school declamation there is no answer to the question just posed (the cut and thrust of the sermo continues); and analogies can be found in the courtroom too. The logic of et ideo difficultatibus quoque utendum est is indeed far from clear; but it may be that the Master means: ‘and so one should take advantage even of the difficulties presented by declamation to train oneself for the forum’. On this view (and perhaps on any view), the sermo at 9 (not signalled in the manuscripts) is making a different point, though I do not know what that point is. 6. From the sermones we can build up a picture of the Master’s views on various topics. ‘Granting’ an advocate—an important matter, cf. Quint. 4.1.46: hoc primum intuemur: I refer to my full note on 250.1. The employment of an advocate to avoid arrogant self-praise (260.3) finds a parallel in the same passage of Quintilian (4.1.45). Restraint: I again refer to a full note (on 259.1). Add 245.1 with my n.⁶¹ The reservations about showing anger (at least in this case) expressed at 346.1–2 ⁵⁸ See Winterbottom (1984), 410; Dingel (1988), 85. ⁵⁹ In 2: Quis queritur, quae proclamatio? I wonder if we should read not proclamatio but comploratio. ‘Lamentation’ might seem to go better with queritur than ‘outcry’ (Shackleton Bailey). For parallels we have to look to the Major Declamations, where proclamatio(nes) and gemitus are linked in 7.7 (144.15 16 H.), 8.21 (173.4 H.), and 18.4 (357.2 3 H.); but then so are complorationes and gemitus in 5.16 (102.8 H.). Note also 16.1 (319.11 12): Excedit omnem querelae meae complorationem. ⁶⁰ PHI turns up three parallels from the Digesta; this is not a frequent locution. ⁶¹ Shackleton Bailey translates persona nobis proponitur amici ‘we are presenting the persona of a friend’. Perhaps rather ‘we have the persona of a friend to deal with’; the same scholar translates 259.16: quatenus forma iudicii proponitur ‘since we have the form of a trial to deal

The Words of the Master

293

sound typical of the cautious Master (and a good Roman): non semper sequendi sunt adfectus, et plerumque consilium his praeponendum est. He does not of course ban emotional appeal, at least in the epilogue (338.3; cf. 294.7: In summa parte paene irate dicit (dicet Rohde) hic accusator); but it will never be enough by itself (270.1; 328.17; 343.2). The topic of restraint is closely bound up with the persona of those involved in a case, attention to which is stressed in 316.2. Relevant is 289.1 (Custodienda est amici persona). Examination of the use of the word in the declamatio sections might be profitable. A diverting passage opens 291. What sort of father am I to be, gravem et severum an facilem et ignoscentem (1).⁶² Intellego me, iudices, fictae huic personae sufficere non posse (2) sounds metarhetorical. Loci: At 316.7 the Master sounds defensive but defiant on the topic of commonplaces. He starts: Nolo quisquam me reprehendat tamquam vobis locos non dem. But from what he goes on to say, it would seem that pupils who chose to insert commonplaces in their speeches did so at the risk of their teacher’s frown. I am not sure what dem implies here: perhaps recommendation of commonplaces in sermones, something he does not do in the corpus as we have it. In 324.3 we find: Hic subici poterit et locus in quo laudemus hanc legem; that is not a generalized topic, and in the progymnasma such praise was of individual laws (Quint. 2.4.33).⁶³ We need not then think that poterit is meant to be grudging and discouraging, and locus will mean no more than ‘passage’ (pace Shackleton Bailey). All the same, the declamatio passages do not eschew commonplace: see my Index s.v. locus communis.⁶⁴ The Master’s words at 316.7, characteristically imprecise, seem to reflect the ambiguous attitude of the Institutio (see my n. there). The Master does not discuss descriptio (Gk. ἔκφρασις), though he recommends one at 366.8 for an epilogue and exemplifies it in some declamations (see my Index s.v. ἔκφρασις).⁶⁵ I refer to my note on 315.9, where the word describere appears (see also 329.17 with my n.). Figuratio is used, apparently uniquely, in the same sense at 380.2 and 385.7 (in each case again ultima). The Master knows about ἐνάργεια (281.3 and 329.17 with my nn.). 7. A brief mention of parts of the long sermo introducing 270 may properly close a discussion that has necessarily had much to say of the Master as

with’. There is perhaps even a touch of ‘is given us by the thema’ (note 316.3, where pupils are advised not to contradict the thema in portraying a father). ⁶² One thinks of Cic. Cael. 33 6 (note 35: ita gravem personam induxi, with Austin (1960³) on 33 4), and of the two fathers in Terence’s Adelphoe. ⁶³ See Reinhardt Winterbottom (2006), 112 13. ⁶⁴ Winterbottom (1984), 613. ⁶⁵ Winterbottom (1984), 608. For Quintilian’s attitude, see Reinhardt Winterbottom (2006), 82 on 2.4.3.

294

Papers on Quintilian and Ancient Declamation

formulator of divisiones.⁶⁶ I am now more struck than before by the concession to declaimers in fortasse recte, though secundum meum quidem iudicium remains emphatic enough (2). And it now strikes me that the closing words, Post haec licebit nobis dicere illa quae sola dicuntur (5), imply rather oddly that the epilogue was the only place where the modern declaimers spread themselves so deplorably. Echoing Quintilian’s language (see my n. on 2) and Quintilian’s attitudes, the Master is emphatic that the headings laid down for a speech should inform that speech in practice (2): vestienda sunt haec, ut ex illis interiores vires habeat. Shackleton Bailey filled this out elegantly: ut ex illis interioribus vires habeat.⁶⁷ The supplement avoids the problem of ‘the reference of haec and illa to the same ossa’ (see my n.). As an alternative, I now suggest vestienda est haec (sc. caro), ut ex illis (= his de quibus locutus sum) . . . But if anything has emerged from the present enquiry, it is that these sometimes cryptic and over-brief sermones may often be not so much corrupt as what the Master wrote, without bringing the ultima manus to them.

⁶⁶ I refer to my n. on 2 for elucidation of the general argument of the passage. ⁶⁷ Thus Shackleton Bailey (1989b). Scholars have also sought to fill out quae sola dicuntur in 5.

24 The Manuscript Tradition of [Quintilian]’s Major Declamations: A New Approach* I. THE ANCESTRY OF M AND Π In the late nineteenth century Hugo Dessauer¹ first drew attention to M, the Montpellier manuscript (Bibliothèque Interuniversitaire. Section Médecine, H 226: s. XII², ?France) of the Major Declamations.² After Dessauer’s premature death, his friend Georg Lehnert exploited it in his pioneering edition (Teubner, 1905); and Lennart Håkanson followed suit when he produced his magisterial Teubner text in 1982. All these scholars recognized M’s lack of sincerity when compared with two other members of what became known as the α class, B (Bamberg, Staatliche Bibliothek, Class. 44 [M.IV.13]: s. X, N. Italy) and the much later V (Leiden, Universiteitsbibliotheek, Voss. lat. Q. 111: s. XII², France). But Håkanson’s stemma³ makes it one of four primary manuscripts in its class—the fourth being π (Paris, Bibliothèque Nationale, lat. 1618: s. XII¹, France).⁴ I shall begin by challenging the primacy of M and π.

* I am very grateful indeed to Oronzo Pecere and Antonio Stramaglia for their help and en couragement. ¹ Dessauer (1898), 14 16. We are told a little about this talented and short lived man by Lehnert (1905),  n. 1 (and by himself, in the Lebenslauf that concludes his dissertation). ² Details of these and other manuscripts are readily available in Håkanson (1982); but for date and origin I follow where possible Munk Olsen (1982 5), .292 305. I rely for the most part on Håkanson’s reports of their readings; but I have drawn on Lehnert for further information about corrections in M (occasionally consulting a scan), and have checked many readings of π in a scan very kindly supplied by Francesco Citti. All references to the Major Declamations are by page and line of Håkanson’s edition. ³ Håkanson (1982), ; redrawn below, p. 310 fig. 1. ⁴ Håkanson was undoubtedly right to make V independent of B. But we should distinguish between the good readings it inherited (those shared with the other branch of the tradition families βγδ , which I shall be calling Φ) and those it owes to scribal conjecture. Neither B nor V shows overt signs of contamination.

296

Papers on Quintilian and Ancient Declamation

Lehnert⁵ remarked of B: ‘Initia declamationum XIV, XVI, XVII, XVIII, XIX, I⁶ desunt, cum rubricator munere suo non functus sit.’ These omissions, or phenomena resulting from them, are found in M and π. I draw the obvious conclusion that M and π descend from B.⁷ Others seem not to have drawn the conclusion, and it of course conflicts with Håkanson’s stemma. Dessauer explicitly rejects it (for π), objecting especially that not all the omissions seen in B reappear in π and that good readings appear in π that could not have been obtained by conjecture (he gives no details).⁸ But the contamination to be seen in both M and π nullifies these objections. To that contamination I now turn. First, M and the contemporary M². M² is often caught introducing readings from Φ manuscripts, and an ancestor of M was tampered with too.⁹ In the following list I give (a) cases of agreement in truth, against BVπ, between M and Φ (or some but not all manuscripts of that group, signalled by ‘–’); (b) other cases, where M is right, Φ wrong. In (a) contamination can have taken place, in (b) it cannot, and scribal conjecture must be responsible.¹⁰ (a) MΦ/BVπ: 1.17 probaverit; 2.3 erant (–); 2.9 acta essent; 10.16 cur (–); 18.16 gravius; 21.2 affertur (M²; –); 26.2 caecitas (–); 26.20 contingit (–); 30.21 cur ante (–); 40.23 mater; 60.8 opera (M²; –); 65.14 eventu; 79.8 ecquando (–); 81.8 excedit; 82.2 futurum; 87.9 in inaequalitate (M²; –); 93.19 praestant; 96.25 dicitis (M²; –); 97.7 inpatientia (–); 98.19 differris (–); and many more (over fifty), very few of them involving M². Note especially 258.17 supervacuis (superba quis BVπ); 330.8 videns (om. BVπ). (b) M/BVπΦ: 12.6 luce; 30.7 putet; 73.1 accedere; 79.9 ullius; 147.20 plurimam; 148.20 dicerem; 183.8 suppl. me; 193.20 oratio; 197.7 quae; 239.1 morbidi nam; 260.17 redemptum; 272.11 coarguendi; 275.16 forte potest; 332.11 negat. Next, π and its contemporary corrector, π². I subdivide the list as for M. (a) πΦ/BVM: 13.4 telum (–); 16.13 vertuntur (+ C²D²); 32.16 habet; 49.1 gratia (–); 95.23 artiore (–); 117.3 assideo (–); 172.16 aegro (–); 173.6 inpotentissima (–); 259.14 respondet; 269.5 insidant (–); 308.1 venefica (–); 314.3 nunc (–); 328.23 illud; 333.23 vestri (–); 335.19 hausurus. ⁵ Lehnert (1905), . ⁶ The order of pieces in B is 2 19, 1. The (clearly related) order in both M and π is 2, 1, 3 19 (though M adds 3a, the Miles Marianus). See Håkanson (1982),  and (2014), 39 43. ⁷ For the details see Appendix 1. It is not impossible, of course, though perhaps not likely, that B precisely reproduces the lacunose state of an ancestor from which M and π descend. ⁸ Dessauer (1898), 12. ⁹ See below, p. 297. For M see Håkanson (1982),  ; for the corrector, ibid.,   (he agrees with Dessauer (1898), 35 that the variants derive from γ class manuscripts). ¹⁰ That must be true on any plausible view of the relationship of these manuscripts (including Håkanson’s stemma).

The Manuscript Tradition of the Major Declamations

297

(b) π/BVMΦ: 10.18 subvertetur; 15.12 ad quem; 19.17 quaerebat; 67.21 infirmitatis (π²); 71.21 districti; 120.16 obsta; 121.12 quod; 129.15 pone tergum; 155.27 admissa; 196.4 inclusus; 204.3 diductis; 243.5 (transposition); 257.4 quod; 262.16 cadit; 268.10 vel ne (vel BV: ne MΦ); 276.19 sceleribus; 277.11 omni; 285.13 has (eas Obrecht); 299.5 haec (hunc/ nunc); 313.19 cui; 347.3 tam; 362.14 hoc; 368.2 continuas; 375.1 suspenso.¹¹ So too, occasionally, Mπ together give the truth against the rest: e.g. Mπ/BVΦ: 32.10 respondet (but + Sγδ); 60.5 plus (M²; + E); 291.1 qui; 381.2 vis once (+ SE). We have seen much evidence for new truth shared with some or all of the Φ manuscripts. I now list agreements in error pointing in the same direction.¹² First, errors of M, M² and π individually against BV but with the rest of the manuscripts (or most of them, signalled by ‘–’).¹³ M(M²)Φ/BV: 4.17 ingressus (–); 15.8 complorare (–); 15.12 ad te; 15.16 ictum (–); 50.4 illuserit; 98.19 redimitur (so too π, as a variant);¹⁴ 99.13 in altero; 129.15 post tergum (M²); 146.11 numquid non; 170.11 quo; 193.20 trepidat et ratio (M²); 223.15 inducat. πΦ/BV: 58.17 contemplari; 248.13 in praesenti adicere. Finally I list cases of M and π agreeing in error with (most of) Φ against BV. MπΦ/BV: 46.7 minor (+ βγH); 46.7 pugna (+ Φ); 50.11 vicem (+ Φ); 56.14 ignominiae (+ Φ (not AS)); 98.19 redimitur¹⁵ (+ Φ); 171.12 descendens (+ Φ); 223.15 merori (+ Φ);¹⁶ 280.11 culpam (+ Φ).¹⁷ It is clear from this evidence that contamination from Φ—bearing both good and bad readings—affected (at least) the ancestor common to M and π (ε on my Fig. 2), as well as M itself via its corrector (who employs the formula ‘in aliis sic’: see Appendix 1). There is then no obstacle to believing that the ancestor common to M and π displayed omissions and other errors seen in its own forebear, B, and that these were eliminated by correction from a Φ manuscript (or more than one). Such correction could have been accompanied by scribal emendation. ¹¹ Cf. Lehnert (1905), : ‘cum ea quibus a Bambergensi dissentit maximam partem meris debe antur coniecturis’. ¹² Håkanson (1982),   lists some of these passages to demonstrate the superiority of BV over M. In my view they show that M and π share a common ancestor below B; many individual errors separate them. Here as elsewhere I do not speculate about further lost inter mediate manuscripts. ¹³ Håkanson (1982),  draws attention to some agreements of M with βγ. ¹⁴ At 128.2 Håkanson accepts the omission of tenebrae in MΦ. ¹⁵ Added as a variant in π. ¹⁶ Merori corr. from morori π. Note also 260.7 deprehenderis V: condeprehenderis B: compre henderis MπΦ. ¹⁷ Mπ by no means always agree with Φ manuscripts; not, e.g. at 5.3 incassum; 43.24 nunc; 44.11 quem stantae; 107.21 carcere miserit; 116.1 num; and so on: perhaps fifty items in all.

298

Papers on Quintilian and Ancient Declamation

Dessauer’s objection to the idea that π descends from B therefore falls; and so would the same objection if brought against M. M and π are therefore not primary manuscripts. They can and should be used as carriers of scribal conjecture. But otherwise their evidence can be disregarded, and on two counts: they descend from a surviving manuscript, and they are subject to contamination from a source that can more readily be reconstructed without their help.

I I. Φ AS HYPARCHETYPE? We come now to that source, which I have called Φ. It has not hitherto been discussed in the detail it deserves; indeed Håkanson does not set out his apparatus in a way that makes it easy to assess the value of its contribution. For him it was a hyparchetype independent of α and equipollent to it.¹⁸ But serious difficulties stand in the way of this view. That the non-α manuscripts go back to a common source is not to be doubted.¹⁹ In Håkanson’s apparatus their shared errors show up most obviously as agreements of βγδ, and I list, pretty inclusively, places where they do not agree with M and/or π.²⁰ Errors shared by βγδ:²¹ *18.19 om. at; *36.23 om. en (as also at *229.4); 65.10 exilii; *68.20 om. non; 88.4 utiliora; *89.4 om. tenuit; 118.18 sed; *123.20 poterit (‘fort. recte’ Håkanson); 127.16 tametsi; 128.9 quippiam; 161.11 datur; 206.16 at; 209.18 orbitati; 222.6 captavit; 225.20 illo; 241.7 citra;²² 255.14 om. in; 274.1 demus; 277.17 iam; 284.9 tutas; 313.4 om. et (as also at 338.10); *319.12 om. me; *330.19 apponit; 356.16 illius; 377.6 miseratur; 386.3 patrem (compare the reading of π). This list complements those on p. 297 where I illustrated agreements in error of Φ with M and π, separately and together. But this is only the tip of an iceberg, for a complete list would include the unquantifiable number of Φ errors that survived into sub-groups δ and β (see just below), only to be eliminated in β and γ respectively. ¹⁸ Much remains to be discovered about the descendants of Φ. But, so far as I can check it, I see no reason to question this part of Håkanson’s stemma. ¹⁹ The subscription at the end of Decl. 10 is not a marker of the whole of Φ: it is found only in sub group β. The second subscription follows Decl. 18, and is carried by αβA. See Håkanson (1982),  ; and in general now, on these subscriptions and their implications, Stramaglia (2017), 195 6; 207 8 (with further bibliography). ²⁰ Merging and slightly revising Håkanson’s lists ((1982), ). ²¹ Some of the items appear to be not so much errors as attempts to emend: I have asterisked some possible examples. ²² Approved by Stramaglia (2002), 50 and 122 n. 92.

The Manuscript Tradition of the Major Declamations

299

I come now to the errors of α.²³ I hope to have demonstrated above that M and π descend from B, so that ‘α’ means now not the agreement of BMVπ²⁴ against the others, but the agreement of B and V alone. The details of these errors are crucial to my case, though they are so many that I have relegated them to Appendix 2. What follows now draws attention to the most important points that emerge from them. The first list in the Appendix registers places—about fifty-five in number— where errors of BV are also found in a sub-group (especially δ), or sub-groups, of Φ. These errors were corrected further down the stemma, and were still present in Φ itself. They can be ignored here, though they testify to the shrewdness of medieval correctors. The second list, of places where BV are wrong and Φ right, is much longer, nearly 250 in all. My contentions are two in number, and complementary to each other (see p. 310 fig. 2): (1) that the list of α errors registers not the variants of a hyparchetype,²⁵ but errors that stood in α itself, before an intelligent reader emended them away by free conjecture; and (2) that (similarly) the list of Φ errors registers not the variants of a hyparchetype, but such errors as we can know of that ensued when the corrected α was copied.²⁶ I propose this picture in order to explain: (a) The shortness of the list of both α and Φ errors, when compared with the length of the text (389 Teubner pages). There are fewer than one a page: and very much fewer that one would expect to find in any copy of so difficult a text. (b) The remarkable fact that the lists throw up hardly a single variation in word order between α and Φ, even though this is one of the commonest errors made by a scribe copying out any text.

²³ Those who have discussed these matters seem shy of listing the common errors that link the α manuscripts; but see Lehnert (1905),  . ²⁴ This is Håkanson’s definition, though his inclusion of π is fitful (note his p. : ‘B V M et saepius π’). I have re examined π where it seemed desirable, and with profit. ²⁵ In my stemma α is the ‘archetype’. Its (ultimate) ancestor was the (late antique) manuscript whose correction is recorded in the subscriptions (see p. 298 n. 19). ²⁶ If this hypothesis is true, the question arises as to where and when all this took place. As to the place, one would like to think that France was the scene, and that B travelled to Bamberg just like the older part of the Bambergensis of the Institutio. But if B was really written in northern Italy that is not a possible scenario. As to the date, one thinks of the twelfth century as the age of correction par excellence, and that is when π and (rather later) M were written. Their common ancestor need not have been very much earlier.

300

Papers on Quintilian and Ancient Declamation

(c) The equally remarkable fact that they throw up no omissions, except of the most trivial kind. Again, this is an extremely common error when any text is copied. (d) The errors in α are by and large such as could be easily corrected by an alert reader. I comment on these points in turn. (a) For purposes of comparison, I note that there are 250 variants between the primary manuscripts A and B in the first eighteen pages of Quintilian Book 9 (Oxford Classical Texts). They include over thirty omissions, and four variations of order. (b) There are, on my view, (almost) no variations in word order discernible because the corrector I envisage did not change the word order unless he felt forced to. The one case I know is at 106.9, where BV have strictisque circa vitalibus dolorem suum membra riguerunt (actually printed by Lehnert); the accusative had to follow the preposition governing it (circa dolorem vitalibus Φ). (c) Similarly, there are—I hold—so few and so trivial supplements because it is notoriously difficult to fill lacunae. I note the following: 123.16 (where an is easily supplied to link dubitat and conveniat); 196.24 (where ad had fallen out in the phrase se ad exhortandam amicitiae fidem magna quaedam composuisse carminibus putent); 207.1 (where et is readily supplied to link the parallel participles in audientem forsitan and exire cupientem); 285.18 quod (readily supplied); 337.12 (see my comment in p. 308 n. 38). For the special cases at 318.9 and 319.3 see Appendix 1 below. That is all. I add that there are nine omissions in my list of βγδ errors (above, p. 298), though it is true (and could be significant) that I marked six of those omissions as possible cases of conjectured improvement. (d) This is the trickiest part of the argument, for what could and what could not have been emended by a medieval scribe is a matter of judgement, here very liable to be swayed by parti pris considerations. We can all agree that a scribe could have elicited the verb grassari from crassari (so twice, at 102.1 and 293.19), or reduced amortalitatis to mortalitatis at 299.13. But beyond that sort of simplicity, opinions will differ. I shall try to play fair with my readers by listing (though also commenting on) the passages that most make against my thesis. 13.10 alter tantum pro se cogitat, alter pro parte utraque] Leaving aside the knotty question as to what the words mean, a scribe here is asked to have emended cogitata (cogitat V) iter to cogitat alter (Φ). Probably he could. V, which displays some good corrections of its own (below, p. 303 n. 28), got as far as cogitat (a finite verb is after all clearly wanted). And the preceding alter, and the shape of the whole sentence, points the way to a second alter.

The Manuscript Tradition of the Major Declamations

301

23.27 infamatura velut deprehensi trepidationem] The exact text here is argued about to this day, but the emender is only required to get from infamatur ac to infamatura (Φ). The stepmother is being talked of, and there is no place here for a passive finite verb. 31.24 et, si bene nomen diceres, hauriet, bibet] What is to be drunk is of course the poison, and a negative is obviously needed. So venenum non (Φ) is a conceivable conjecture to replace the jarring bene nomen. But it is to be observed that it is bene that jars (not nomen, sc. veneni), and that Reitzenstein did not feel happy with venenum (see Krapinger–Stramaglia (2015), 188 n. 275). Perhaps we might think of merely deleting bene, though non is naturally still needed; dices did not emerge until βγ, and Schultingh wrote dixeris. 83.19 Per ego, si fas est, quicquid feci, per hanc ipsam mei caritatem] Anyone who arrived at per ego (Φ) from perago (BV) had a sure feeling for Latin idiom; but per, with another just to come, is obvious enough. There is a parallel our scholar may have known, though not in the same declamation: 109.17 Per ego te, iuvenis, . . . per . . . (where β did not recognize the idiom and deleted the first per). 86.14 Dii (better di?) immortales] BV have dum mortales, but the minims in α may have been interpreted better by our corrector. For the apostrophe, see just below at 88.14 (where V again has dum mortales). 95.15 impudentissime generis humani, tu non feres ut frater tuus vel magis ametur?] B²V have feris; but I am not sure that fers (B) is wrong. 99.9 non fratrem tibi praetulisset quod in te fratri praetulissem] Any reader might have balked at praetulisset, and an alert one would have seen that a first person verb was needed. A clever one could, perhaps, have made the move to praetuli sed (Φ). 108.10 Intellegit, iudices, et ipse iuvenis non esse se calamitatium nostrarum iustitiae parem] BV give essent for esse se (Φ); the correction is perhaps manageable. 109.5 quicquid faciebam mendicitas est, ex quo reversus es] As Stramaglia (2018), 69 n. 372 explains, the father had previously been begging in order to pay his son’s ransom. Now, the son having returned, he feels that, if he continues to beg, people will refuse alms, feeling his son should be supporting him and that the begging is just begging. I am not very ready to believe that a scribe could have taken this point sufficiently to emend sed (BV) to ex (Φ). This is one of the weakest links in my argument. But it is true that ex quo has been used temporally in this piece at 86.12 (of the son’s return) and 102.1. 125.9 primum illud, ut etiam quia deerit, possit inpune nihil praestare] quia deerit (BV) looks wrong, but it was not perhaps quite simple to change to the correct qui aderit (Φ). 126.8 tribunos didicimus, candidatos ferimus] It is easy enough to see that didicimus (BV) is wrong, harder to emend it to deducimus (Φ). It is possible that if this is an emendation, it was not meant as we understand it; and ferimus too might have been taken differently.

302

Papers on Quintilian and Ancient Declamation

171.4 abacta est aperitur ut prima mater] So BV. The change to a perituro prima (Φ) would be meritorious but perhaps possible. 196.6 ut me nectum recipere non posses, victorem nolles] The contrast with victorem might have enabled an emender to see the truth here (victum Φ). 209.1 Gravis testisque pelle suae orbitas est] If our hypothetical emender was able to change this to testis querelae (Φ) he would have been justifiably pleased with himself. suae would have set him looking for a noun ending -ae. 214.4 nunc omne terrenum (codd. pervium) numen et religio, quam istum rogo, propius adeste] Again, the change to isti inrogo (Φ) would be meritorious. But a corrector perhaps again counted the minims better (cf. my remarks on 86.14). 241.13 Ut maior urgere necessitas putaremus coepit] putaremus is a strange (‘mirum’ Håkanson) intrusion, which on my hypothesis the emender would have deleted (om. Φ). 245.20 ad sepulturam residua conferuntur, et tantundem cadavera igni permittimus] ‘At last’ (tandem Φ) we are able to burn corpses, rather than being forced to eat them. A possible conjecture. 263.16 Adeo ne apud inferos quidem ulla poena est fame maior] non audeo (BV) would only turn to adeo (Φ) at the hand of a bold emender. This is perhaps the best example to confute my view. 275.20 Nisi fallor, esset aestimatio et apum. An (apuma in BV) tandem quas subripere non liceret, liceat occidere?] The bees would come easily enough in such a piece. α might have had in corrected to an. 280.7 Non enim ideo qui aliud patrocinio tuo conferat haec denuntiatio] Our emender is asked to make ideo qui into video quid, perhaps a move within his powers. 286.1 ipsis enim se invicem angulis haerent] Φ has ipsi . . . sibi anguli, but I am not very happy with the hiatus at the colon end, or with the overall sense, and it may be that the truth is still to seek here. 294.17 confectoque quod petebatur, reliqua quoque viribus vicinae tabis expirant] BV divide the words vicina et ab is (his B²V), but our man could have read α’s perhaps unarticulated²⁷ text correctly. 316.7 illi quos in voluntate superfluentium facultatium mittit secura felicitas] mittit in asks for an accusative, and the rest follows (in voluptates Φ). 325.13 Ut turpis et quod recedo, in eo tantum est, ad quem revertor] ut asks for a verb in the subjunctive (turpe sit Φ). 328.4 Dii deaeque, quantum ille praestitit qui me invidit poenam meam nec recessit] This difficult sentence was only healed (if then) by Håkanson; and if it was a corrector who made mihi (Φ) out of me, it was because he thought (hopefully) that the sense might be ‘envied me my punishment’.

²⁷ See Håkanson (1982),   (of the archetype).

The Manuscript Tradition of the Major Declamations

303

333.1 nemo umquam volet innocentem filium videri, quem voluerit occidere] nec novum becomes nemo in Φ, a bold move. 335.8 sed si dicunt solent qui mori volunt . . . ab omnibus . . . secessi] sicut solent (Φ), a possible conjecture if one has divined the sense. 336.15 Who would believe the following story? Venenum filii, quod invenerat paternum, ille potius effundit] So BV. The change to the past tense effudit (Φ) imposes itself (stetit follows). It was the father who had discovered the poison (pater Φ), but he did not pour it away (non Φ). A clear-headed reader could have made this change. But perhaps pater, num (π) ille could be right? 350.15 Nihil est profecto inprovisorum dolor et horrentius] It would take a master stroke to arrive at dolore torrentius (Φ). The latter word is by no means an obvious one: see Pasetti (2011), 206 n. 442. 368.14 Praepostera rara res est filium occidere, deinde erubescere] Φ omits rara, and a bold emender could have removed it. 373.15 Maiorem et inpatientia, maior urit adfectus] An emender might have been led to maior me inpatientia (Φ) by the parallel maior and by the need for this object for the verb. 380.12 Quid agimus, anime, quemadmodum effugimus, evadimus?] BV have quaenam modum. quemadmodum (Φ) is no doubt right (cf. 97.21, with effugere), though π’s quonam modo (cf. Quint. 1.3.6) has its attractions. Many readers will think that this array sufficiently disproves my hypothesis. But I would ask them to compare the examples of scribal correction visible in M and particularly π (above, pp. 296–7: esp. 129.15 and 243.5), in the subgroups of Φ (below, p. 304 n. 30), and in B and V themselves.²⁸ The questions raised in (a)–(c) (pp. 299–300) remain to be answered by those who regard Φ as a normal hyparchetype. I conclude, with much hesitation, that B, V, and Φ descend independently from α. Each (but especially Φ) has been subject to conjectural emendation. It should be emphasized that Φ was an excellent text even before it was corrected, for it avoids the multitudinous individual errors of B and V.

III. A SLIM APPARATUS CRITICUS My speculations about the origin of Φ are not of great practical importance. Whatever picture we draw, there remain three witnesses to the text: the extant ²⁸ The examples of correction by B² (for which see Håkanson (1982), ) which turn up in my list of BV errors (Appendix 2) are not particularly impressive. But elsewhere note 323.3 amari, 367.9 sivit. As for B itself, see especially 170.10 vix. In V observe e.g. 42.9 tuis; 100.18 nostra; 160.17 dis placet; 189.9 ne si; 208.11 es; 209.3 inique ei; 217.6 a te; 260.7 deprehenderis; 347.20 mollis; 376.20 crede.

304

Papers on Quintilian and Ancient Declamation

B and V, and the reconstructed Φ. Either B and V stand together against Φ, as Håkanson wished, or (on my hypothesis) the three are independent of each other. In the first case, the transmitted reading would, in a Maasian world, be either that given by BV or that given by Φ. In the second, again in a Maasian world, it would be given by either BΦ or VΦ. Ours is not, of course, the world of Maas. But my elimination of M and π does at least reduce the blurring element of contamination, which seems (especially) to operate from Φ towards those two manuscripts. A rule of thumb, whichever picture we follow, is that two witnesses outvote one (unless, in the first case, that one witness is Φ). When this rule breaks down, contamination (or scribal correction) has taken place. I print in Appendix 3 an experiment in the formulation of a slimmer apparatus criticus for this text. We need to distinguish between two kinds of apparatus. One is the kind we see in Håkanson, where he ‘shows his working’: that is, he aims to tell us everything we need to know in order to see how the manuscripts relate to each other. The other kind is called for, and sometimes seen, in the series of Oxford Classical Texts (‘recognovit brevique adnotatione critica instruxit’). The two approaches could be combined in the present case if fuller information were reserved for an appendix.²⁹ The basis of the apparatus proper would be the readings of B and V. These would be confronted by the evidence of Φ (sometimes to be reconstructed only with a degree of hazard). But the three sub-groups,³⁰ and their constituent manuscripts,³¹ would, like M and π, only be adduced as sources of conjecture.³² A rough and sometimes arbitrary order of priority would need to be established, perhaps somewhat as ²⁹ I think of Ermanno Malaspina’s first edition of Seneca’s De clementia (Malaspina (2001)). My review of his subsequent Teubner text (Malaspina (2016)) in Winterbottom (2017a) makes an I hope reasoned case for preferring his earlier approach. But clearly both approaches are entirely valid. ³⁰ I list here corrections accepted by Håkanson and attributed by him to whole groups. γ: 313.18 diduxerit. δ: 57.20 impudicum (+ E; a splendid correction); 68.14 om. nemo (+ E); 103.1 egerat (+ C²); 178.9 quodam; 293.8 unumquodque (+ AE); 335.6 inponerem (+ E); 376.19 suffeci (+ E). β: 38.24 cum clusus (+ AC); 40.20 quantae (+ ACD); 87.9 in inaequalitate (+ AEH); 89.7 morbis (+ C); 129.13 digne; 212.5 videras; 295.2 potes; 330.1 osculis; 360.23 deprehendetur; 364.15 laminas. γδ: 126.19 exorabit. γβ: 70.13 obitu; 145.12 liventia; 332.4 quidquid. δβ: 234.21 exponamus. The list can be supplemented by cases where errors of BVδ were corrected lower down the stemma (see Appendix 2 below). ³¹ A fair number of individual manuscripts are known to carry good conjectures. I pick out, honoris causa, the twelfth century corrector of C (see Håkanson (1982),  and n. 9), together with the rather later E (ibid., ). C and E are in Håkanson’s group ψ (a sub group of γ), but their relationship with δ (see the second list in the previous note) would bear investigation. A and O also score well. ³² This is the way in which, in the footsteps of R. A. B. Mynors’s Catullus, I treated the recentiores of Quintilian’s Institutio and Cicero’s De officiis in my editions for Oxford Classical Texts.

The Manuscript Tradition of the Major Declamations

305

follows: XII¹: π C; XIIm: D H; XII²: M; XII–XIII: E; XIII: J O; XIV: A N P; XV: S. The earliest in each of the three groups gives a rough order for them (different from what one might expect from Håkanson’s stemma): γ: C; δ: H; β: P. I am under no illusions: the Major Declamations, after the forthcoming Loeb,³³ may never be edited again. Meanwhile, we, like so many of our medieval predecessors, are still engaged in trying to make sense of this difficult but fascinating text.

APPENDIX 1 : MISSING RUBRICS On this topic see above, p. 296. In the present appendix, I (a) assume the correctness of my view that M and π derive ultimately from B and (b) foreshadow my further view that Φ is no more than a corrected version of α. At 318.9 BV omit the first words of the theme of Decl. 16, as would have been obvious to any reader of those books, for appliciti sunt now has no subject. That subject was supplied differently in B’s progeny: M gives uterini duo fratres and π duo iuvenes, both by conjecture. M’s corrector notes: ‘in aliis sic incipit thema duo amici ex quibus uni mater erat’, a clear (and typically worded) case of contamination from Φ. These are words that we have to read twice before realizing that they are meant to convey that the boy’s mother, though not his father, was still alive. It is obviously a possibility that an emender of α could have contrived this not quite happy formulation on the basis of the speech that follows. But something on those lines is clearly what is needed. A few lines later, at 319.3, there is an omission that raises a question that will have to be considered in a wider context. The start of the speech itself (Etsi . . . olim) is missing in B (before much later correction), π (naturally), and V (which also omits omnes). M has a concoction that cannot be right, but its second hand, in the familiar formula, informs us: ‘in aliis libris sic incipit haec causa: etsi, sanctissimi viri, olim.’ This is the reading of Φ, and editors have always printed it. If this is a transmitted reading, it is fatal to my case. It might, however, be the result of conjecture. The opening concessive, paralleled in other declamations, is necessitated by the answering tamen (319.6). The address to sanctissimi viri could have been imported from just below (319.18; again, pretty soon, in 320.14).³⁴ But the constant practice of these pieces is to ³³ In preparation (Latin text by A. Stramaglia, English translation by M. Winterbottom, intro duction(s) and notes by B. Santorelli and M. Winterbottom). ³⁴ Also in the peroration of Decl. 17 at 352.1.

306

Papers on Quintilian and Ancient Declamation

start with an immediate address to the iudices;³⁵ and in 289.5, where sanctissimi viri are apostrophized well into the piece, the iudices had been addressed in its second word (287.12).³⁶ I submit that one might expect our piece to start Etsi, iudices (cf. 302.1; 353.7), with address to the sanctissimi viri reserved till later. What we read in Φ, then, may be no less a concoction than what we read in M. These omissions early in Decl. 16, affecting both B and V, presumably go back to lost rubrics in their parent, α. The slackness of B’s own rubricator affected several other pieces. I discuss them in the order in which these pieces are placed in B. Decl. 14: The argument (287.10–11) is given in V, M², and Φ. The corrector of M knew it by contamination. It was transmitted from α. Decl. 15: The argument (301.12–13) is given in V and M mg. (with the usual formula ‘in aliis sic tema incipit’), but not in BMπ. It is widely omitted in the Φ manuscripts, no doubt because it is the same as that of the preceding Decl. 14. Decl. 17: The argument (331.2–6) lacks its first seven words (Filium . . . invenit) in B (and therefore π); M carries a concocted replacement. V has the argument in full, as does Φ. The first words of the speech itself are altogether lacking, thanks, presumably, to a lost rubric in α (cf. above on Decl. 16). VΦ transmit one word lost in BMπ, lassatusque (V) or lassatus ((β)γδ); it was therefore, on any view, present in α. Decl. 18: Here much of the argument (353.3–6) is missing in V, but the scribe refers the reader forward to the identical argument in Decl. 19 (371.9–13). Then the start of the speech itself, Etsi . . . ordinavit, is missing in B (and so π); as in the previous piece, M carries a concoction of its own. VΦ (= α) transmit the lost words. Decl. 19: In this case, the argument is present in the α manuscripts, but the first words of the speech (Debebatur . . . misero) are omitted in B and so π (M again offers a concocted substitute). As in Decl. 18, VΦ (= α) transmit the lost words, though in V they have become detached from their rightful place (see Håkanson’s apparatus at the foot of p. 371). Finally, in Decl. 1, the argument was lost altogether (Lorenzo Valla supplied one much later). The opening words of the speech, Si iuvenis innocentissimus, iudices, uti vellet, are omitted in B and therefore π; they are transmitted to us by VΦ (= α).

³⁵ Always, except in Decl. 3 (where there is no panel of judges; Marius is not named until as late as 43.17, but he is alluded to in the third line of the piece (42.9), tuis . . . sanctissimis (!) auribus, and in 43.7 he is addressed as summe imperator); 4 (P.C. addressed at once (61.4)); 11 (but the judges are addressed in 1.2 (220.6)); 17 (for the missing start see below); 19 (where iudices appears in the ninth line (372.9)). ³⁶ Similarly sanctissimi viri in 336.5 had been preceded by an address to the judges at 332.9; the start of the piece is lost (see below).

The Manuscript Tradition of the Major Declamations

307

APPENDIX 2: E RRORS OF BV For discussion see above, pp. 299–303. I list first cases where BV are joined in error by one or two of the sub-groups of Φ (γδβ).³⁷ BV = sub-group(s) of Φ: 4.20 et (corr. B²; + β); 9.15 obiecit (+ δ); 10.16 videre (+ δ); 17.16 illam (+ β); 19.3 rogabat (+ δ); 31.25 diceres (+ δ) (but see p. 301); 36.25 agitante se (+ γβ); 57.15 allegatas (+ β); 57.21 negavit (+ β); 75.11 siderumque (+ δ); 76.7 sacro . . . ingenio (+ δ); 79.3 praedicabar (+ δ); 121.1 haec (+ δ); 123.15 quis (+ δ); 125.21 praeditum (+ δ); 127.22 comparares (+ δ); 143.8 si (+ δ); 156.13 dirum (+ δ); 159.20 om. quem (+ γδ); 159.28 repeated words (+ γδ); 161.5 plangore (+ δ); 183.16 omnis (+ β); 232.14 civibus (+ δ); 235.26 interficientes (+ δ); 238.19 a real muddle (+ δ); 251.4 ac (+ δ); 253.9 superat (+ δ); 255.3 faciet (+ γ); 262.12 abrumpimus (+ γ); 270.21 possint (corr. B²; + δ); 273.9 sed in (sed et in δ); 279.9 inmorabantur (+ γδ); 283.21 ipsa . . . stipata (+ δ); 284.21 nec fas] nefas (+ δ); 285.9 referuntur (+ δ); 288.13 id (+ δ); 290.6 fuissem per (+ β); 296.1 mea defectu (still influencing δ); 302.4 efficient (+ δ); 313.19 quod (+ δ); 313.19 add. aut ad (+ δ); 313.20 tumultuosa (+ δ); 314.4 inprodigiosque (cf. δ); 314.8 quem bis (corr. M²; + δ); 314.8 patentibus (+ δ); 314.9 quo (+ δ); 314.20 remedia (corr. M²; + δ); 316.22 consurgere (+ δ); 321.11 inter explicabilis (+ δ); 321.18 illas (+ δ); 323.4 inimicis (+ δ); 331.3 tenentem (+ β); 336.4 tenerem (+ β); 336.4 quod cui id (quod cui δ); 336.17 trepidus (+ δ); 356.11 possit (+ δ); 357.11 iudicium (+ γβ). These items (around 55 in number) are to be ignored (see above, p. 299); but they are valuable as a demonstration of what medieval correctors were capable of (see also above, p. 304 nn. 30–1). I come now to the cases where BV agree in errors where Φ gives the truth. BV alone against Φ: 7.25 scit (corr. B²V²); 7.28 ipso; 8.7 noverce; 8.15 qui (corr. M²); 11.1 intrasse quid; 11.17 suum; 12.5 esse; 12.14 mensura; 13.11 iter; 14.15 suspensas (corr. M²); 15.13 addictus; 17.5 illa; 18.3 quis; 18.3 novercam; 19.13 te gladius habere (corr. M²); 19.14 miseri; 23.27 infamatur ac; 23.28 trepidatione; 25.20 qui; 31.25 bene nomen; 35.3 praemittunt; 36.10 tumultatur; 38.7 concurrent; 43.11 ictum; 47.1 re; 50.7 faciat (corr. B²); 54.14 concessa (corr. B²); 55.11 luxuriae; 65.10 densissima; 65.22 circumstantes; 70.16 quandam (corr. B²); 70.17 interrogatis; 71.13 succendentium; 77.6 confundit (corr. B²); 82.3 quantos; 82.16 sentit (corr. M²); 83.19 perago; 86.2 querellam; 86.14 dum mortales; 87.23 dissimiles; 90.17 qui; 95.16 feris (fers B before corr.);

³⁷ In both this list and the next one, I have included information about corrections in B and V (to show what may have been possible independently) and M (to show contamination and/or free conjecture at work). I exclude some cases where BV are joined by one or more individual Φ manuscripts, all cases where the truth does not appear in Φ but was seen by a medieval or modern scholar, and a number of cases where the text is in doubt.

308

Papers on Quintilian and Ancient Declamation

99.10 praetulisset; 101.19 mersum (corr. π²); 102.1 crassari; 102.7 exanguis; 102.15 corpori; 106.9 circa vitalibus; 108.10 essent; 109.6 sed; 113.16 potest; 119.13 morsit; 121.3 cumularunt; 122.23 deis; 123.16 om. an; 124.4 autem; 124.4 permittit et; 124.11 dubitationem; 125.9 quia deerit; 125.20 natos; 126.8 didicimus; 126.12 carcere; 126.19 exoravit; 130.5 poneretur; 130.12 lectulo; 130.22 litigato; 132.12 contingat; 132.9 ne (‘fort. recte’ Håkanson); 133.15 patre; 133.27 possis; 135.6 non; 135.10 creditis; 135.20 letam; 148.21 intremescis; 154.12 languorem; 165.12 desperaverint; 171.4 aperitur ut; 176.3 qui in; 177.17 humilitatis; 177.23 finem quem; 181.15 trepidantem; 184.21 locos (corr. M²); 187.9 aut; 188.17 quantum; 192.14 tum (‘cum vel dum’ M²); 192.16 tum; 194.14 constet; 196.6 nectum; 196.24 om. ad; 197.14 vitam; 197.23 illis; 198.8 rogavimus; 203.17 planguorem; 206.3 sepulchrum; 207.1 om. et; 209.1 testisque pelle; 214.5 quam istum rogo; 229.10 debeat; 234.13 omnis; 234.14 fame; 239.18 morientis; 239.19 me me ipsum; 241.14 ; 244.12 abiecero; 245.20 tantundem; 249.23 sinum; 251.3 contraria; 253.7 retulerit; 256.15 apibus; 263.16 non audeo; 269.23 pellere; 270.2 proxima; 270.16 corrumpit; 271.19 privata; 273.11 defendes; 274.9 custodi (corr. B²); 275.21 apuma in; 277.11 societatem; 277.12 magistratos (corr. B²); 277.19 dominorum (B²V); 278.7 versosque; 280.7 ideo qui; 280.14 procedat; 282.7 si neque; 282.8 neque de; 283.18 de eorum; 284.22 qui; 285.1 congerent; 285.12 admirant; 285.18 om. quod; 285.18 ne; 286.1 ipsis se angulis; 287.1 pertinent (corr. B²); 288.9 meretrice; 288.9 possit; 288.10 praevaricationem; 289.1 imponant; 290.14 potui (corr. M²); 292.1 demissum; 292.18 flagrantesque; 293.18 quis; 293.19 crassantur; 294.10 medicinam; 294.18 vicina et ab (h)is; 297.12 respiciant; 297.13 possideris; 297.19 conpesci; 297.22 adiuvavit; 298.14 misertus; 299.5 tenent; 299.13 amortalitatis; 299.19 mandareturque; 300.18 illas; 300.15 adsistor; 303.11 invidia (corr. B²); 303.22 movet; 303.24 hanc patientiam; 304.9 amovit; 304.17 moverentur; 304.21 grave; 306.4 vagedum; 307.12 fleri (corr. V²); 309.14 lauda; 311.12 censum; 311.12 laboris; 312.6 sanandum; 314.1 fueriis; 315.6 gratia; 316.5 fures; 316.7 voluntate; 316.18 quid; 316.20 dedisse (corr. B²); 317.4 rogavit; 317.7 ultionis; 318.1 quis; 318.9 om. duo . . . erat (see above, p. 305); 319.3 om. etsi . . . olim (see above, pp. 305–6); 319.19 tamen credibilis; 320.16 nullus; 320.18 hominis; 321.7 accidisse; 324.12 hominis; 325.13 turpis et; 327.3 fide si; 328.4 me; 328.14 tractabit; 328.14 exigit; 329.18 carcere; 330.4 irem (corr. V²); 332.14 credulitatis; 333.1 nec novum umquam; 333.2 nocentem; 333.10 vivere; 333.14 accusatur; 334.11 nos; 334.20 cedentis; 335.8 si dicunt; 336.16 paternum; 336.16 effundit; 337.12 om. ut;³⁸ 339.20 iratus; 339.24 ut; 340.11 favorem (corr. M or M²); 342.22 confugiret (corr. B²); 343.8 posse; 343.10 posset; 344.4 accipere; 344.23 habes; 346.4 ipso ³⁸ Håkanson’s apparatus is not very clear here. The fact is that α read et, the others ut; Reitzenstein (1909), 55 made out of this et, ut. Here ut might well be a conjecture of someone concerned to give circumdaret a construction.

The Manuscript Tradition of the Major Declamations

309

(corr. B²); 346.20 propia (corr. B²); 346.24 tunc; 347.9 parricida bis; 349.1 bibes; 349.2 palpitatemque; 350.15 dolor et horrentius; 351.8 si (corr. M²); 353.16 ipse; 354.1 comperit; 354.12 calidissimam (corr. B²); 356.24 incerto; 357.20 tractionis (corr. B²); 359.18 materia mirari; 360.21 captavit; 362.8 a matre (rightly?); 363.10 nudus; 364.10 insistere; 368.14 praepostera ; 369.3 potest (corr. B²); 369.6 solas; 369.25 pars; 370.26 praestitisti; 371.3 damnatus; 371.6 nisi (corr. M²); 371.12 compereret (corr. B²); 372.4 damus (corr. B²); 373.3 contempta; 373.15 maiorem et; 373.19 condam; 376.1 relictam eorum; 380.13 quaenam modum.

APPENDIX 3: S AMPLE OF AN ED ITIO MINOR O F TH E MAJOR DECLAMATIONS I here illustrate the possibility of drastically reducing Håkanson’s apparatus, without much—if any—loss of important information. I give a positive apparatus, to show the state of affairs clearly (more clearly at any rate than Håkanson did). I give individual variants only of BVΦ; other manuscripts (including M and π) are used as carriers of good readings not found in these three, only the earliest carrier being adduced (for a possible order of priority see above, pp. 304–5). Φ is reconstructed where possible; otherwise a split among the sub-groups is shown. Håkanson’s commentary matter, some rejected conjectures, and various trivia have been removed, but his text is assumed. Some comments of my own are added in footnotes. Asterisks signal items that might be omitted. p. 15.8 colorare BV: complorare Φ | isti VΦ: ista B | 9 hae Schultingh: ac BVΦ | 12 ad quem π: autem ad quem BV: ad te Φ | 13 praeparare VΦ: praepara B | illuc B: illud VΦ | ad ictus Φ: addictus BV | potes VΦ: potest B | 15 visceris βγ: visceri BVδ: viscerum π | 16 tibi Håkanson: ubi BVΦ | de spiritu sanguinis del. Winterbottom | hoc cum Håkanson: hoccum BV: ictum Φ | explorare VΦ: exploraret B | 17 potes V: potest BV²Φ p. 81.6 ista VΦ: istae B | 8 excedit Φ: expedit BV | 11 cogitationem BΦ: cogitionem V | 16 *porrexerim BΦ: prorexerim V | 22 citra Håkanson: contra BVΦ p. 113.3 *rixamur BΦ: rixamus V | 4 lege VΦ: legi B | praescribitur BΦ: perscribitur V | 5 tutum Helm: totum BVΦ | 16 potes Φ: potest BV | 20 parentium Bγ: parentum Vβδ³⁹ | 22 diduceret P: deduceret BVΦ p. 181.2 *veniret B²VΦ: inveniret B | 3 *putabam B²VΦ: putabat B | 4 *cruentum . . . confectorem VΦ: cruentem . . . confictorem B | 7 in VΦ: inter B | 8 offulsit V: effulsit BΦ | *obstupui VΦ: obstipui B | 11 rediit B: redit VΦ, ³⁹ I think this sort of variant should be hived off to a praefatio for bulk treatment.

310

Papers on Quintilian and Ancient Declamation

fort. recte | iter VΦ: usque huc iter B | 15 trepidante Φ: trepidantem BV | 23 fecerat ϛ:⁴⁰ fecerant BVΦ p. 235.3 aliqui BVβγ: aliquis δ, fort. recte | 13 alias Lucas: aliis BVΦ | 16 reliquit BΦ: relinquit V | 19 sitientis Vβ: sitienti Bγδ | soli VΦ: sole B | 21 maturitatem VΦ: maturitate B | 22 herbae solum vicerant BΦ: solum vicerant herbae V | 23 inanis BV: inanes Φ⁴¹ | 25 et rudes VΦ: crudes B | 26 inter scientes βγ: interficientes BVδ p. 347.1 terendum VΦ: tenendum B | 3 tam π: quam BVΦ | 6 occidere VΦ: occide B (occidi B²) | 11 *cur ergo B(corr. ex crego)Vβγ: cur ego δ | 14 cum dico VΦ: cum iudico B | 17 accepere ϛ: accipere BVΦ | 18 desperatione B²Φ: despirationem B: desperationem V | 19 colla BΦ: collo V | 20 mollis V: molis BΦ *

* *

α

V

M

[Φ]

π

B

δ

β

γ

Fig. 1 Håkanson (1982),  redrawn Ω α B

Φ

ε V

π

M + M2

γδβ

Fig. 2

⁴⁰ I should prefer to replace stigma (passim) with a positive attribution to a particular manuscript or printed edition. ⁴¹ Again, I think this sort of variant should be hived off to a praefatio for bulk treatment.

1 L. Håkanson, Textkritische Studien zu den grösseren pseudoquintilianischen Deklamationen, Gleerup (Lund, 1974) Few standard classical editions are as unhelpful as G. Lehnert’s of the Major Declamations.¹ The bland apparatus, innocent, so far as the editor could manage it, of the conjectures of scholars earlier than Hugo Dessauer, rarely assists a reader wrestling with the gibberish that so regularly masquerades as a text. This is a monument to conservatism run mad, to a kind of criticism that (in Reitzenstein’s trenchant words) ‘mit der christlichen Liebe um den Ruhm streitet, alles zu ertragen, alles zu dulden, alles zu glauben und alles zu hoffen’.[²] Reitzenstein himself produced in 1909 a series of Studien where obvious truth rubs shoulder with extravagant error. Dr Håkanson has now given us a work of comparable distinction and superior consistency. Håkanson has worked on the manuscript tradition, without markedly modifying earlier views. But his main contribution is in emendation, of which he is a delicate and skilful practitioner. He excels in the slight touch, the shifted comma, that illuminates a whole sentence. He is naturally assisted by modern awareness of prose rhythm: hence convincing repunctuation at e.g. 2.6 (25.19 L. [25.22 H.]); 4.15 (82.12 L. [77.25 H.]); 9.16 (182.1 L. [190.23 H.]) and, though he does not make the point, at 1.4 (7.4 L. [5.11 H.]) as well. He has excellent ideas of his own: e.g. 3.9 (49.3 L. [50.8 H.]) opponet; 9.12 (177.27 L. [186.6 H.]) deletion of est; 12.17 (233.18 L. [250.2 H.]) repimus, and he can recognize excellent ideas in others (thus for Burman—to be spelt thus—e.g. 5.19 (106.3 L. [104.21 H.]) squalentes; 6.19 (129.2 L. [131.7–8 H.]) fleres . . . rescinderes; 19.1 (337.1 L. [372.11 H.]) intra pectoris; and for Schultingh—to be spelt I think thus—e.g. 8.4 (148.23 L. [154.13 H.]) velut)). He rarely writes an unnecessary note—that on 6.9 (119.11 L. [120.3–4 H.])

[Classical Review NS 26 (1976), 276] ¹ Lehnert (1905). ² [Reitzenstein (1909), 85.]

314

Papers on Quintilian and Ancient Declamation

might qualify. And he equally rarely emends a passage that is sound—but at 4.9 (75.31 L. [70.10 H.]) aliter surely implies ‘in the absence of any sanction’. Where so many solutions are propounded, not all can persuade equally. I cannot believe in those at e.g. 1.14 (16.12–16 L. [15.13–17 H.]); 5.3 (90.23–7 L. [87.8–11 H.]); 13.13 (259.7–13 L. [280.6–11 H.]). And one feels the itch to tinker further with good ideas at e.g. 1.9 (10.28 L. [9.16 H.]) (quaeret is excellent: but then delete et?); 1.10 (12.10 L. [11.4 H.]) (perhaps timore rather than tremore); 1.13 (15.14 L. [14.13 H.]) (‘ab secreta’ would better explain the corruption). Håkanson promises us a new edition. It will be enormously superior to Lehnert’s. But could one ask him to include in his apparatus frequent hints for the struggling reader? These declamations will never be easy reading; they need a kindly as well as a skilful editor.

2 (1) J. Cousin (ed., tr., comm.), Quintilien. Institution oratoire, Tome  (Livre ), Les Belles Lettres (Paris, 1975) (2) J. Cousin, Recherches sur Quintilien, Les Belles Lettres (Paris, 1975) Cousin’s Études sur Quintilien, a vast and valuable survey of the sources, appeared as long ago as 1936. Now, in 1975, he begins the crowning of a lifetime’s preoccupation with this author by giving us a study of the manuscripts and the first volume of a new Budé. If I seem to dwell in this review, as reviewers will, on what I see as the defects of these two books, it is not because I do not appreciate C.’s past and present services to Quintilian, but rather because of our common concern that books in manus hominum quam emendatissimi veniant. 1. The lengthy introduction to the Budé deals with five topics: ‘Biographie’, ‘L’avocat et le professeur’, ‘L’homme et la doctrine’, ‘Dédicace’, and ‘Établissement du texte’. The last I reserve for later treatment. The first two cover much the same ground as an early article of C.’s own,¹ and much the same conclusions are come to; the discussion is thorough, though little hangs on attaining certainty, say, as to the date of the Institutio: if we could date the Dialogus, the topic would be much more important. As to the doctrine, C. argues, as I should say correctly,² that Q.’s interest in the morality of the orator is to be linked with contemporary activities of delator-orators. For the rest, C. treats Q. from a more philosophical standpoint than I personally find helpful; a few pages on the overall structure of the Institutio would perhaps have been more valuable. As to the ‘Dédicace’, I entirely agree with C.’s hesitant suggestion that the preface to Book 1 is ‘postérieure à l’ensemble du traité’ (p. ). [Gnomon 49 (1977), 574 9] ¹ Cousin (1931). ² Cf. Winterbottom (1964a) [= A.1 above].

316

Papers on Quintilian and Ancient Declamation

Some details: P. : Whatever the merits of the identification of Q.’s Iunius Bassus with the Elder Seneca’s Iulius Bassus, it should not be imagined that the manuscript evidence gives any support; at all the three places where Bassus is mentioned in Q., the one primary MS always gives Iunius. There is no variation at 6.3.57 (C. has read Radermacher’s apparatus too hastily). Pp.  : The argument about the twenty years of Q.’s teaching career is thoroughly confusing (esp. p.  n. 1). Pp.  : I see nothing suggestive about the fact that Q. refers to his imperial tutorship merely as officii delegati mihi at 6.pr.1; he had specified the post at 4.pr.2, and had other preoccupations in the later preface. And if the allusions were so embarrassing, why did he not delete them when the work was published as a whole? P.  n. 1: ‘Je me demande toutefois si Cicéron ne songe pas à Théophraste plus qu’à Aristote lui même, quand il vante le style des péripatéticiens’ (e.g. Orat. 46). But see Nat. deor. 2.95: Praeclare, with the suc ceeding quotation.

Book 1 is provided with a long preliminary ‘Notice’ of its own; illustrative material is awkwardly divided between—and sometimes repeated in—this, the notes to the translation, and the full ‘Notes complémentaires’. Indeed, one wishes that C., a master of the sources and of the bibliography, had given us full commentaries on some of the untouched books of Q., rather than embarking on his present venture. Discussion of individual passages naturally causes overlap between these parts of the book, the translation, which is accurate and flowing, and the text itself: though the last I discuss further below. 1.1.9 n.: I cannot believe that Hier. Ep. 107.4 consulted Diogenes of Babylon on Leonides. He employs Q. 1.1 passim in this letter, and I doubt if his addition of et in incessu proves very much. 1.4.3 (see p. 34): I do not agree that emendatio (or rather emendate lectio, which is what Q. says) is ‘un choix, qui permet d’écarter toute érudition inutile ou toute fantaisie puérile’; Colson’s note repeated, indeed, by C. in his own footnote is clearly right, that Q. is speaking of correct reading. 1.4.7: C. translates At grammatici . . . while reading Aut.³ 1.4.27: C. reads teruntur, and to some extent follows my own view;⁴ but his ‘la flexion n’est pas banalisée’ does not make it clear how he takes declinationibus. 1.4.29: ‘Dictu et factu sont, en effet, à vrai dire, des formes participiales’: see however Winterbottom (1970b), 63. 1.5.7 and 9: C. reads barbarismum and barbarismi, then complains (pp. 13 14) that Q.’s ‘notion de barbarisme ne s’accorde pas avec la nôtre’. That is precisely why we must read barbarum, barbari with B: then only the technical sense is called barbarismus (§ 10). 1.5.52: haec sequentia = ‘the last two’. I am not sure what C. means by ‘ces séquences’, but hardly that. 1.5.61: The note on p. 168 is inconsistent with that for 1.6.42 on p. 171. Surely Caelius in each case is the orator? 1.6.38: C.’s translation ignores hoc.⁵ 1.8.14: C. may well be right to delete et schematismus with Faber; but he should not have at the same time (p. 176) said that ‘c’est le groupe et σχήματα qui doit être supprimé’ for then ut dixi

³ For a defence of the latter, see Winterbottom (1970b), 62. ⁵ See ibid., 66.

⁴ See ibid., 63.

Reviews

317

(see 1.5.52) has no reference. 1.10.2 ff.: Undoubtedly an answer to critics, as C. argues on p. 38, but surely to ones who thought that rhetoric could be taught without the ἐγκύκλιος παιδεία. Q. may have read Sen. Ep. 88 (so C., p. 41), but he had other adversaries in mind: § 3, for instance, is hardly a report of Seneca’s views. 1.10.28: I find it hard to believe that crassiore . . . Musa does not attach itself to imperitiores.⁶ 1.10.49: C. reads augendi, and defends it pp. 182 3; but he translates augenda.

It would be ungenerous of me to find great fault with the text, which follows my own⁷ (not listed on p. ) very closely. C. inevitably punctuates differently, but he spells almost exactly as I do: I have noticed only half a dozen orthographical variations (for ἐπισυναλιφήν at 1.5.18 see LSJ s.v. συναλιφή, and Fuhr (1906); so too at 9.4.36). More substantial disagreements are rare, about 25 in all; and I have noticed only three places where C. prints a reading not to be found at least in my apparatus. 1.1.5: If we do add vasa, rhythm recommends its placing after nova. 1.4.13: C. returns to the old conjecture ac lases et asa fuerunt. But he seems to do so by perpetuating (and compounding) an old misconception (e.g. Halm, Raderma cher) about the MS readings. For him, B reads ac lases as fuerunt et, A fuerunt aetatis atque. In fact, B has ac lases fuerunt as et, A aetatis fuerunt atque. This reduces the attraction of Sarpe’s (unrhythmical) change. 1.4.29: C. changes mutantur to mutuantur without defence or need. 1.7.5: It may be observed that the conjecture proposed in the apparatus is Halm’s. 1.12.1: C. follows F in transposing tamen to follow eodem. Cf. however 9.4.111: ac praecedente spondio tamen . . . There are few misprints in the text, but observe 1.1.37: volant; 1.5.19: h omitted; 1.8.2: virilis est; 1.9.1: instituent.

I cannot, again, grumble about the MSS employed by C.; his list of sigla (pp. –) includes all the MSS used by me except the Monacensis (no great loss), and adds none except Parisinus lat. 7724, the fruits of which are not relevant to Book 1. But he would be a rash man who trusted any statement in C.’s apparatus. I give some examples (pauca e multis) of his errors of omission or commission: C.’s statement first, the truth in brackets. ep. Tryph. 1 infiniti A: stituti H (institi H; C. omits to say that for operis just before H gives instituti operis, corrected, I think, from institi operis). 2 novum H (novumque H: the words have been corrected, but the first hand did not omit que). 1.pr.4 in eloquentiae H: eloquentiae T (T has in). 26 sint A: sunt HT (sint AH: sunt T). 27 ingeni. quique adiuvante HT (ingenita cuique adiuvante H: inge**** quique adiuvante T: quidem quae adiuvant t). sicut AH: sicut et Bg (sicut A: sicut et Bg H). 1.1.4 posset Bg H (posset aBg H). 5 quae deteriora is printed without mention that this is a conjecture. 19 pro gatum Bg (prorogatum Bg). 22 putat et Bg (putet et Bg). 28 fere AH: ferre Bg (fere a in ras., b H corr.: ferre ⁶ For an advocacy of et for etiam cf. Winterbottom (1970b), 68. ⁷ Winterbottom (1970a).

318

Papers on Quintilian and Ancient Declamation

Bg H). 30 nominibus Bg H: in omnibus A (Bg and H both have in). 32 lectione A ( oni). 35 id est quas (quam Bg) bH: quas A (id est quas Ab: id est quam Bg: id est qua s H). 37 durantur AH (durantur AbH). 1.3.tit. A reads qua tractanda. 1.4.8 C. has merged Philander’s addition of in with a variation on ut just above. 1.5.62 C. does not tell us the author of the deletion he prints (so too at 1.6.5). 1.7.8 illud n enim] illud n ABH: illud enim nm Colson (illud enim K: illud n ABH). 1.7.16 quod quidem t: quid quod BH: quo Ab (quod quidem t H corr.: quid quod quidem B: quo quidem Ab and apparently H before correction). 1.7.20 caussae b in marg. (Bn in marg.).

I can only pass on to C., in all due humility, the fruits of my own experience: one cannot trust one’s collations, and each statement in an apparatus has to be checked against the actual manuscripts. Worse still, in a way, (a) C.’s use of the older MSS; and (b) his failure to explain his (or rather my) use of the later ones. (a) C. systematically employs Bg and H as well as A and Bn (his B = the consensus of Bn and Bg). As to Bn and Bg, nothing could be more certain than that the latter is a copy of the former; C. assents to this (p. ), and offers no defence of his use of Bg nor could any be offered. There is more reason to employ, even where A is present, the corrections (b) and supplements (G) of the Bambergensis; and C. disagrees with my attempt to show that b and G have no stemmatic value whatever where A is available.⁸ But we are not told why C. uses H, though he seems to acknowledge that it is a copy of Bg as corrected and supplemented. I do not know what C. means by saying (p. ) that ‘il a bénéficié des corrections de Bg et le scribe de H s’est lui même à plusieurs reprises corrigé, de sorte que le codex est un bon garant à placer sur une ligne parallèle de la descendance de Bn et de A à travers Bg’ (my italics). To my mind, almost all the readings that C. gives from H after 1.pr.6 and from Bg after 1.2.5 are worthless. (b) C. says of his (my) use of the fifteenth century MSS that they occasionally have ‘des leçons qui paraissent sûres et dispensent de recourir à l’ingéniosité des modernes’ (pp.  ). That the same is (or should be) true of his use of the mutili descended from Bn and all other eliminable MSS, C. does not state. Nor does he make the point that the main use of Vaticanus lat. 1766 and Harl. 4995 is to date the corrections in P.⁹ As to Vat. lat. 1761, the reader may well wonder why use should be made though only up to 1.pr.4 of a MS which is said on one and the same page to have a ‘texte utile’ and scribes who ‘ne comprennent pas ce qu’ils écrivent’ (p. ). The answer is that, to the end of Book 3, this MS is a direct descendant of A, and that it can therefore be employed in making up for the loss of some of the first folio of A itself. The same is true, mutatis mutandis, of my use of descendants of Bn up to the point where Bn today starts. All this is explained in Winterbottom (1967a), to which C. refers earlier, but of which he does not seem to make use.

⁸ Winterbottom (1970b), 5 18; see Cousin (1975), 6 n. 2, without detailed argument. ⁹ See Winterbottom (1970a),  .

Reviews

319

C., after 1.2.6, almost never (perhaps seven times in all: six available from Halm, Niedermann, or Radermacher) gives readings from recentiores that he could not find in my edition. I do not know why he suddenly calls in J as support for B at 1.4.14. At 1.8.11 we are told ‘videmus R: -imus ABH’; in fact R has vidimus; but in any case A has videmus. We have already begun to see some of the defects of the chapter on the establishment of the text; I add some further comments. P. : C. tells us that Bn lost its first two folios in the sixteenth century, with the result that it now starts at 1.2.5, although just below he says that ‘manquent en effet la lettre à Tryphon, le prooemium du livre I et I, 1 6 jusqu’à verim nec de patribus difficile à lire’ (C. seems here to be thinking of the Bambergensis); while on p.  he says that ‘[c]’est peut être à cet endroit [sc. 1.1.6] que commençait la partie ancienne [which is that?] du Bernensis, au moment de la copie par le copiste de Bg’ (in the tenth century). C. also seems to regard the other lacunae of Bn as the result of physical damage to this MS; that would indeed be a proof of the descent of all/other such mutili from it but the matter is not so: the lacunae do not come at the end of versos, except at 5.14.12, where the scribe has adjusted the lineation to make this possible. C., in any case, speaks ‘de filiation, ou, si l’on préfère, de descendance’ of the other mutili from Bn, though he only sporadically mentions, without acknowledgement, my attempts¹⁰ to prove their dependence: thus for Ambros. F 111 sup. and J, though not for N or D or R. P. : J’s hand is ‘une seule main’, but it ‘change au f o 57’. I think C. may mean that the hand is the same throughout, but adopts a new manner at that point. This may be true; but he should mention why the change is textually significant.¹¹ P. : It is exceedingly misleading to say that D ends at 12.10.15; but these pages show the same misconceptions about the Bec sub group and the excerpta from Books 10 and 12 as appear in the Recherches¹² (see below). And C.’s descriptions generally are often inaccurate or unclear. P. : The verso on ff. 232v 233r of T is in the hand of the eleventh century corrector, and marks the end of his corrections. It has nothing whatever to do with Poggio, as C. suggests. P. : C., discussing the spelling of pre(he)ndo words, should mention the matter of rhythm. Nor is it true (p. 92 n. 3) that ‘[n]os mss. écrivent toujours comprendere’ at 2.15.20, the first passage I checked, A has comprehendendo.¹³

2. The section ‘Établissement du texte’ is culled from the other work under review, where C. follows the tradition of the Institutio from late antiquity to the present day, with detailed descriptions of almost all the MSS and a useful catalogue of printed editions. C. gives details of the decoration and ownership of the codices (I am glad to be corrected on the Lassbergianus: p. 109) that will be useful to future investigators; it is a pity that he shares with myself a lack of the expertise in humanistic palaeography that would have made our parallel ¹⁰ Winterbottom (1970b), 22 30. ¹¹ See ibid., 28. ¹³ See further Winterbottom (1970b), 58.

¹² Cousin (1975).

320

Papers on Quintilian and Ancient Declamation

researches more fruitful. C. has devoted great labour to this book, and I take no pleasure in recording my judgement that he does almost nothing to advance, and much to obscure, the study of Q.’s transmission. C. is to be congratulated on having located at least three complete fifteenthcentury Quintilians not previously signalled,¹⁴ including Bononienses 1624 and 2299 (not 2229), which I am particularly ashamed to have overlooked. It is a pity that he tells us nothing that would enable us to fit these new manuscripts into the classifications which I proposed in Winterbottom (1967b), and which C. himself accepts. Against these successes, he omits mention¹⁵ of the Anecdota Parisina; two thirteenth-century books, Clare College, Cambridge 26 James and Laur. Med. 50.27; and the fifteenth-century Marcianus 1965, all mentioned in my own work. Nor does he know of Marcianus 4087, which but for a typing error I should have added to the list of descendants of P at Winterbottom (1967b), 362 (where, further, Vat. lat. 1764 is a misprint for 1765). The chapters on the medieval period and on Petrarch display no knowledge of Winterbottom (1970b), 22–30 or (1962b).¹⁶ C. therefore is unclear about the relationship of the Beccensis group with the Joannensis group, and of both with Bn; quite apart from innumerable misstatements as to the content of these mutili. Nor does he realize that the excerpts 10.1.46–131 and 12.10.10–15 formed a separate tradition from at least the eleventh century into the fifteenth. On the other hand, he has read my remarks on the descent of certain late mutili from Ambros. F 111 sup. in Winterbottom (1967b), 344–5, for he cites them later (p. 84: but I am sorry that, as appears from pp. 88 and 104, he should think that by (K) I meant an intermediary between Bn and the Ambrosianus); but he does not exploit them in his discussion of Petrarch’s MS on p. 40 and of Paris. lat. 7720 on pp. 43–4. By a similar process, the chapter on Poggio moves in an era prior to discussion in Winterbottom (1967b), which C. is later to follow: hence misunderstanding over (amongst many other things) Vindobonensis lat. 3135. It may be worth adding that in the letter to Poggio cited on p. 58 Bruni meant by syncopis illis grandioribus the lacunae of the pre-Poggio mutili; he is making the correct point that the text of Q. is less sound where the mutilus tradition fails us.¹⁷ When we come to the post-Poggio period, C. largely follows the thread of exposition in Winterbottom (1967b); but he blurs, misunderstands, or misrepresents much of what I say there. The classes he accepts from me are rarely

¹⁴ But Monacensis 4023 (p. 151 n. 1) is only ‘Scholia in X librum’, not as C. might seem to imply a full text. ¹⁵ To judge by the Index. But its list of Manuscripts is incomplete; and it rarely gives refer ences that fit the text. ¹⁶ Cf. Winterbottom (1970b), 31. ¹⁷ See Winterbottom (1967b), 345 with n. 1.

Reviews

321

defined properly: ‘C’est en marge de ce groupe’ is the way he introduces my class λ (p. 102), without any mention of its relationship to its most important member, Laur. Med. 46.9, discussed on p. 69. His listing of MSS under my headings is, indeed, regularly interrupted by startling intrusions. Thus on p. 76 ‘Je rangerais volontiers dans la même groupe’ (that is, of primitive descendants of T) Bodley Rawlinson C 48: which is a fragment from the quite separate excerpt tradition;¹⁸ at p. 121 Ottob. lat. 1159, a fourteenth-century mutilus, suddenly appears among a group of fifteenth-century ‘vulgate’ texts; at p. 140 Voss. lat. fol. 80, a late mutilus, intrudes, apparently by association of ideas with a quite unrelated Vossianus. Some assorted details may give an idea of the quality of this book. P. 12: Allusion to the non passage ‘II, 10, 90’,¹⁹ matched by ‘II, 4, 91’ on p. 48. P. 25: Under AG add ‘VIII, 3, 64’ after ‘V, 14, 12’. P. 26: The Harley MS 2664 (not ‘lat. 2664’) is dated at least fifty years earlier than the later part of the MS C. agrees it is copied from (p. 26 n. 1). P. 121 n. 1: ‘[O]n ne connaît pas de traité de sy nonymes de Cicéron’. On the contrary: for this, and for the letter to ‘Valturius’ (elsewhere Veterius or Veturius), see e.g. Brugnoli (1961), 285. Pp. 130 1: C. quite misunderstands the contention of Radermacher as to P and Julius Victor. P. 132: What Peterson is here said to have remarked of Paris. lat. 7725 he in fact said of Harl. 4995 (the error is caused by careless reading of my discussion in Winter bottom (1967b), 361). P. 140 n. 1: The Bessarion Quintilian, sought by C., is Marcianus 1965 (Zanetti 435). Pp. 149 50: The description given of Paris. lat. 7729 seems to have been merged with the description of a second, quite different manuscript. P. 164: As deserta . . . parte is the correction of t at 1.pr.17, we can hardly use the reading to diagnose relationship to the lost S, which was only one of countless texts descending from T at this point. Three personal points. I have never said anywhere of Bernhard Spluges what is imputed to me at p. 62 n. 3; I am misquoted on the Burneianus at p. 117 n. 2; and at p. 134 C. understands gemellus in a way that I do not.

¹⁸ See above, p. 320; for this MS, Winterbottom (1962b), 173. ¹⁹ So also in the Budé, p. .

3 J. Cousin (ed., tr., comm.), Quintilien. Institution oratoire, Tome  (Livres  et ) —Tome  (Livres  et ), Les Belles Lettres (Paris, 1976) These two volumes of the Budé Quintilian follow closely upon the first,¹ and they share the merits and demerits of their predecessor. The text makes no pretension to novelty. The translation is fluent and generally accurate. The notes are informative. The apparatus criticus is a disaster. The t e x t takes my own as a basis, as is elegantly proved by the common error iudubitata at 5.10.19. C. rarely quarrels with my spelling, perhaps twenty times over these four books. I am surprised that he favours dirigere at 2.6.1 when he imports derectione at 3.6.30, while Plisthenem at 3.7.19 (though A there has -en, despite C.) is less incorrect than Ulixen at 2.13.13.² I take it that nunquam at 5.10.48 is a misprint. I should not expect to agree with a Continental editor on many details of punctuation; but there are occasions when C.’s produces unhappy effects (e.g. 2.1.2 destruction of correlation of et illi / et hi; 2.1.4–5 destruction of correlation of et grammatice / et rhetorice, which leads to incorrect translation; 2.17.24 et gubernator separated from 25 et medicus; and especially 2.18.1–2 where the structure is Cum sint . . . aliae . . . aliae . . . aliae . . . , fere iudicandum est). As to differences of wording from my own text, they are few, perhaps sixty over the four books, and usually trivial. C. is no lover of the obelus, even printing at 5.12.5 a reading he says in a note ‘ne donne pas un sens acceptable’, and he is a little readier than I to plump for a conjecture that avoids the use of the dagger. There are a few places where I should judge C. clearly wrong. At 2.4.37 I cannot believe in the construction quaeritur an . . . an . . . (unless the

[Gnomon 50 (1978), 685 7] ¹ Reviewed by Winterbottom (1977b), 574 8 [= R.2 above, pp. 315 19]. ² See Winterbottom (1970b), 51.

Reviews

323

two questions are quite separate as in the next sentence). At 3.5.6 it seems perverse to make Q. use primum of the first of two where prius is a variant (cf. 3.6.36, again contrasted with sequens). At 4.1.40 the indicative videtur suggests that sunt should be omitted. At 5.14.27 I cannot believe that Gertz’s partes is wrong (let alone worthy of neglect even in the apparatus). In general I think C. is more enamoured of the somewhat interpolated A stream than I am (thus at 5.11.31, where C. himself cites a good parallel for frigida without aqua). But in general it is cheering that we disagree so little. I am only sorry that I cannot follow C. in his two new conjectures. At 5.12.14 ut animis crescant is inelegant after ut . . . sint infirma and open to objections.³ At 5.13.50 I cannot see that et narrationibus et contradictionibus occurremus improves on my own et narrabimus et . . . ⁴ There are places where C. does not conform to his own conventions, for example in the use of italics to mark divergences from the manuscripts, especially with iis (as at 4.2.68). As to misprints, I noticed: 2.4.36: ea (eae); 2.12.7: circa (circa illas); 2.13.8: iterumque (iterumque iterumque); 2.16.5: section number omitted; 2.21.4: da (ad); 3.4.5: comprenderit (-int); 3.5.5: omnen (-em); 3.5.16: seni (an seni); 3.6.14: proprius (-pius); 3.6.96: patrus (-ris); 3.7.4: dubia (-iae); 3.8.16: se (si); 3.11.12: qui (qui sui: see app.); 3.11.16: illus (illud); 4.1.26: locabantur (-buntur); 4.5.3: adeo (adeo ut); 4.5.6: sognoscenti (cog-); 5.7.29: testis (testis testi); 5.7.37: hoc (hoc genus); 5.10.63: si (sit); 5.10.115: causa (causae); 5.13.57: as (ad). My soundings in the t r a n s l a t i o n suggest a certain patchiness: some chapters impeccable, others surprisingly prone to error. Thus: 2.1.4: tantum = ‘so far’, not ‘seulement’; 2.1.9: hoc . . . genus is defined by the ut clause; 2.1.12: ne . . . quidem goes with statuam, with the sense ‘a statue too is not begun’; traditus is translated as though the reading were tradendus (read by A); 2.2.8: ut dicitur comments on the cliché viva vox; 2.2.11: the sense of the last sentence is: ‘thus facility will come about by writing, judgement by listening’; 2.3.7: gradum . . . minuat means ‘shorten his step’; 2.3.10: the last sentence means: ‘this may be taken by some in a way that makes my present opinion contrary to that I held before’; 2.19.2: utrisque is ignored; 3.7.10: Ante hominem means not so much ‘Avant de parler de l’homme’ as ‘As to the time before his day’ (cf. [tempus] quod ante eos fuit above); 3.7.23: Interesse tamen Aristoteles putat is translated: ‘Il est intéressant cependant, selon Aristote’; 3.7.28: ut . . . sic ignored. Bornecque was right at most of these places, and his Garnier may well remain the most reliable translation in English or French. C.’s n o t e s, though full, are almost exclusively factual. They give the parallels in technical treatises so familiar to the author of Études sur Quintilien, but they are little concerned with matters of interpretation. One of the very longest is a very

³ See ibid., 97.

⁴ Cf. ibid., 99 100; Quint. 4.2.29; 7.1.38.

324

Papers on Quintilian and Ancient Declamation

Gallic note on wine, with only a tangential relevance to 2.4.9. On the other hand, there is no indication at 4.2.27 how crucial is the variation between A and B over the word quam, and no reference to Heinze’s discussion of the consequences for the Pro Caelio.⁵ I add a few comments: 2.5.3: C. proposes to read adiutor for auditor at Suet. Gram. 10, a change perhaps made less rather than more plausible by the appearance of adiutorium earlier in the sentence. 2.10.5: The change in pestilentiam responsa here suggested was attributed by Spalding to Fareus. 2.17.21: On Cicero’s deception of the jury in the Cluentius case, C. might have added the witness of Plut. Cic. 25. 2.21.1: C. might have mentioned that the Greek πραγματική in this sense is as odd as Q.’s negotialis. 3.4.2: I cannot believe C.’s suggestion that Cicero is the maximus temporum nostrorum auctor. 3.6.77: I cannot understand C.’s defence of the transmitted Non licet abdicare syllogismo. 4.1.54: I do not see why C. says that ‘[l]’actio diversae partis est ici la formule délivrée par le magistrat pour saisir le juge’; is it not the other side’s speech? In general, there is some carelessness abroad in the notes: in volume  there is overlap on Celsus between pp. 155 and 257–8, and a note on 3.9.9 appears both on p. 216 and on p. 276; in volume  a note appears on both p. 75 and p. 76. I come finally to the a p p a r a t u s. It has what seem to me imperfections in principle. It makes room for far too trivial orthographical variants (so, at random, 3.1.14: coepit / cepit; 3.8.23: alioqui / alioquin); and conversely it contains far too few rejected though plausible conjectures. To come to the manuscripts cited: I spoke in my earlier review⁶ of C.’s use of H. This manuscript is descended from the extant Bg, as corrected (b) and supplemented (G), and it rarely offers even a correct conjecture; still less is there need for the parading of its countless errors. It is true that C.’s misreporting of it makes it appear to have virtues it can lay no claim to: it does not have sicut at 4.4.2 (except in the hand of a late corrector), efficitur at 5.10.11 (ditto), insequentis at 5.10.46, citra at 5.10.53, de at 5.11.12, tu habes at 5.11.28, conduxerit at 5.14.9 or sit at 5.14.17. It is also true that by suppressing the witness of b C. can make H seem surprisingly virtuous. Thus at 5.11.19 ‘quoque fabellae AH: om. B’ hides the fact that b adds the words. It is only by similar errors that C. is occasionally able to make Bg itself seem other than what it is, a copy of Bn. Thus at 5.8.6 ‘credibiles alia [by which is meant ‘aliae’] ABgH: om. Bn’ should read ‘credibiles ali(a)e AbH: om. B [C.’s siglum for the agreement of Bn and Bg]’. The fact is that C.’s use of H and of the first hand of the Bambergensis is pointless. There is, on C.’s view (though not mine), point in reporting readings of b, and it is a pity that C. does not do that more consistently and more accurately; b (G) undoubtedly has its interest in giving us a clue as to the readings of A before correction, at least in certain parts of the text,⁷ and I now ⁵ Heinze (1925), 201 n. 2. ⁶ Winterbottom (1977b), 577 [= R.2 above, p. 318]. ⁷ See Winterbottom (1970b), 6 7.

Reviews

325

regret that I did not use it more for this (one) purpose (so at 5.14.7, where b shows that the first hand of A had aliquis). All these are, doubtless, matters on which opinions can differ. Less disputable is the flagrancy of the carelessness exhibited on every page. We all make mistakes in collating and in reporting our collations, but C. goes beyond normal limits. Some of his entries are wrong even on inspection. The apparatus conflicts with the text e.g. at 2.5.19 et hic, 2.15.22 quibus sunt non, 3.8.3 quae, 4.2.101 obruentur, 5.2.3 inveniundum, 5.13.13 ut ea. Manuscripts are forgotten (2.1.1: ‘cotidie / BgH: cott- A’) or mentioned on both sides of the equation (4.2.25: ‘indicium etiam ABH: om. B’), or both (4.5.3: ‘partitionem H: partionem BH’). At 4.5.13 and 5.13.60 a reading is given from A in a passage that C. has already (correctly) stated to be omitted in that manuscript. To come to matters of substance: C. at one point reproduces my own misprint (5.10.78 hod- for rhod-), a compliment of a kind, perhaps, but not one that encourages belief that he has checked his entries against the manuscripts. On the other hand, C. is able to tell us the reading of A before correction in many places where my own naked eye despaired—even at 5.11.34, where he tells us himself that the corrector’s reading is in rasura. It is C.’s treatment of A, indeed, that is particularly misleading. He often neglects to tell us that A has been corrected (thus at 4.1.43 it is entirely relevant that the a of quidam is a correction); and this is especially disconcerting when we are told merely that A omits a word or words without being told that a supplies them: thus at 2.13.9 et iuncti pedes is added supra lineam, I think by the first hand. The fact is that no entry in this apparatus can be confidently taken as complete or true. I give some typical examples: 2.16.14: ‘eiusque A: eiusdem BH’: eiusque is read by a (in ras.), B; 2.17.25: ‘contingit BnH: -tigit ABg’ should read ‘contingit A, Bn corr., BgH: -tigit Bn’; 3.8.8: ‘minor A: maiorne B: maior nec H’ (read ‘maiorve A’); 4.5.17: ‘credat ABnH; -dere Bg’ should read ‘credat AbH: cre B’; 5.10.33: ‘quae D’Orv. 13: quaeque a: aeque A: namque B: neque H’ should read ‘quae Regius: quaeque a (quid A¹, incertum): namque B: neque H’. C. in fact is capable of getting anything wrong. At 5.12.22 my abbreviation ‘Kiderlin 1889–2’ is expanded to the meaningless ‘Kiderlin 1889–1892’. At 2.16.1 Gensler’s indignissime is cited without the fit that would make it intelligible. It is particularly dispiriting that at 5.11.31 C. repeats the old error that B has disparata despite my apparatus and despite my explicit warning,⁸ and that at 5.10.56 he returns to the view that a is Badius’ addition without troubling to check my implicit statement that A has the word. Does progress towards the truth in our knowledge of Q.’s manuscripts need so disconcertingly to resemble the climbing of a greasy pole?

⁸ In Winterbottom (1970b), 96.

4 J. Cousin (ed., tr., comm.), Quintilien. Institution oratoire, Tome  (Livres  et ), Les Belles Lettres (Paris, 1977) The fourth volume of Cousin’s Budé Quintilian is no better, and no worse, than its predecessors. With a general reference to my reviews in this journal,¹ I may pass at once to detail. In the text, the punctuation is sometimes wayward. At 6.pr.4 C. actually starts a new paragraph in the middle of a sentence, the second half of which is naturally left without a main verb, while at 7.10.17, though unwilling to delete ut, he still places a full-stop at complectantur, thus crippling the whole sentence. To turn to readings: 6.3.26: I assume that dicenti is meant to be obelized rather than deleted; C. seems to translate Gesner’s sicut, though he does not cite the conjecture. 6.3.55: The manuscripts do not omit quam, here printed in pointed brackets, nor do they omit et at 6.3.60. 6.3.62: Buecheler intended to read Gabba, not L. Gabba. 6.3.92: C. is quite at liberty to reject my autem, but I cannot believe that haec will do: perhaps it would have been better if I had not observed the letter in A in the first place. 6.3.100: C. prints and translates my legatario, but disputes it on p. 209; I still do not know how the legate comes into it (Gaius’ reference to peregrini at 3.134 does not tie down his remarks to the provinces). 7.7.4: There cannot possibly be no lacuna here: the words Magistratus ab arce ne discedat do not constitute a controversia (in my apparatus ad loc., where I suggest what may have fallen out, opat should of course be optat). 7.9.10: By persisting in A’s Lacheten C. produces a metrical howler. 7.9.13: I do not know why C. places casu nominativo in a separate sentence from what precedes, or why he thinks that Regius had access to Vaticanus lat. 1766.

[Gnomon 51 (1979), 388 9] ¹ Winterbottom (1977b), 574 8 [= R.2 above, pp. 315 19]; (1978) [= R.3 above].

Reviews

327

C. makes a number of conjectures: 6.3.78 (p. 206): Suillio (the identification was already canvassed by Spalding); 6.3.100: aspere (the order seems odd if we have an adverb); 7.2.22: etenim in superiore (I can understand Halm’s etenim in posteriore, but not this); 7.2.26 (p. 219): delete marginos; 7.2.33: enceniis (as unthinkable for me as Radermacher’s ἐνκαινίοις; Gertz’s innocentia now seems to me certain); 7.3.2: male tractare (hardly necessary; observe incidentally that Spalding thought uxorem should replace rem publicam); 7.4.2 (p. 226): delete nam tum ignorabatur; 7.10.1: quo spectatur (there seems no place for any kind of relative here, as C.’s own translation indeed shows). At 7.10.10 C. attributes to himself coloris (before modos), which e.g. Halm read and which the early manuscripts themselves give (colores in A only after correction).² The apparatus pullulates with error. I reserve my protests for one main matter: that by constant misreporting C. makes it seem that G and H are far more often right against A than is in fact the case, with the result that G is made to appear independent of A, and H a good deal more than a faithful duplicate of G. As to the supposed merits of G, it may be worth pointing out that, despite C.’s reports, G does not read 6.1.10: animum; 6.1.18: sollitudinem; 7.4.19: ponendum est. On the other hand, A or a reads 6.pr.12: tuosne; 6.1.3: responsurus; 6.1.5: respondeant; 6.1.38: et quidem; 6.2.9: spectanti; 6.2.36: convenit (I think); 6.3.38: diceret; 6.3.43: dictorum; 6.3.47: amphibolia; 6.3.53: frigida; 6.3.87: cascellio; 7.3.6: de vi; 7.7.5: poni . . . casu; 7.7.10: ut; 7.10.2: amphiboliam. Further, 7.2.49: tyrannicidam, 7.4.3: agitur, 7.5.2: contineatur, and 7.9.12: fanni et eius are the result of correction, and we cannot know that A¹ did not read what G reads. At 7.10.2 A¹ had *ocis, not ocis, and at 7.10.11 it certainly finished the word oeconomia correctly. At 7.2.35 A cannot have read ad quidquid, but that can hardly be the truth in any case. The entries at 6.pr.1 victori and 7.2.25 huic are muddles that I shall not pause to unravel. As to the even slimmer merits of H, this manuscript does not have any of the following readings, attributed to it by C.: 6.pr.4: despicere; 6.1.1: a; 6.1.14: peroratione; 6.3.8: et quae; 6.3.38: iam; 6.3.52 dabantur; 6.5.9: quo. One block of strikingly good readings is (correctly) reported from H in the later part of 6.1 and in 6.2 (e.g. 6.1.51: sint reservanda), but C. does not mention here³ that these all come on two folios in a fifteenth-century hand that provides us with 6.1.41: cognomen acciderit . . . 6.2.25: legi (vel). In his introduction to Book 6, C. himself indulges in pathos when discussing the proem. Q.’s words are certainly affecting, but in a cool moment C. ought to remark that they are placed to introduce a book on adfectus as cunningly as the invocation of ‘all the gods’ is placed in 4.pr. at the start of a book that will discuss proems. ² Cf. Winterbottom (1970b), 131. ³ That these folios ‘ont été refaits’ is mentioned by Cousin (1975),  n. 1.

5 S. F. Bonner, Education in Ancient Rome: From the Elder Cato to the Younger Pliny, Methuen & Co. (London, 1977) In 1949 Mr Bonner gave us what shows every sign of remaining the standard work on Roman declamation. Now he has written a bigger and equally authoritative book on Roman education generally from the time of the Elder Cato to that of the Younger Pliny. This is to pull off a fine double, and one’s only regret is that nearly thirty years should have separated the two books. Bonner writes lucidly and with a certain reticent humour; he has collected intriguing photographs of abacuses, ink-pots, and wax tablets; and he thoroughly deserves to attract the general reader he has half an eye to. But this remains a scholar’s book that scholars will value. The notes are exhaustive, and directed almost entirely to the primary texts; the bibliography is wide ranging. The whole is enormously more sophisticated than Gwynn (1926), and a great deal more detailed than the corresponding sections of Marrou (1975⁷). The book has three parts. The first is historical, tracing the development of education from a private, family matter to such elaboration of public organization as it reached under Trajan. The second, in some ways the most interesting and original, discusses the conditions of teaching: the difficulties of teachers trapped between parents reluctant to pay and pupils reluctant to learn, working often in rented shops or in the open air, looking gloomily forward to a pensionless old age, are sympathetically portrayed. Finally, in Part Three, we are told in great detail what the Roman schoolboy learned, and how he learned it. So elaborate is this section that chapters 20 and 21 (for instance) form excellent introductions to the topics of rhetorical precept and declamation. And the information, for example, about the use of the abacus (pp. 183 ff.) or punctuation (pp. 220 ff.) makes this a useful reference book on a wide variety of topics. All this is largely, it is true, an account of the known. But Bonner has thought through his material, and without any striving for [Classical Review  29 (1979), 73 4]

Reviews

329

originality he comes unemphatically to his own conclusions. Not that one can always assent to them. He suggests, for instance, that Crassus’ edict of 92  was directed against teachers who encouraged a noisy and brawling eloquence that (Crassus feared) might prove dangerous in the courts (pp. 71–4). This certainly makes good sense of ludus impudentiae, but it is hard to believe that it is the whole story (and why Latin schools only?). From this wide field I select one topic for comment. Both here and in his earlier book, Bonner stresses more than others have done the parallels between Roman and declamatory laws, and infers from this the practical utility of declamation. There is nothing illegitimate about this, but it is perhaps worth underlining just what it is that Bonner is doing. To take the case of declamatory abdicatio, discussed on pp. 312–13: it is true that this has much in common with the Roman practice of relegatio, and it was only natural that it takes on some Roman colouring from that source. But it remains true also that the kernel of abdicatio is the Greek practice of ἀποκήρυξις, or at least that practice as it is represented in the Greek declamation on which the Roman rhetores undoubtedly drew; and they took over with the Greek concept some items, like the public legal process involved, that had no place in Roman custom. It is true that Quintilian, at 7.4.11, points to a real-life parallel to that public process. But he is not in any way asserting that what goes on in declamation is itself Roman. Bonner’s juxtaposition of declamation with Roman laws follows in the wake of Quintilian, who was concerned to defend declamation by showing its at least partial contact with reality. But we should not imagine that this kind of special pleading tells us anything about the way the themes and laws of declamation originated, or how indeed the majority of declaimers at Rome probably employed them.¹ It is after all perfectly arguable that an imaginary case subject to an imaginary law is better for the training of a schoolboy than involving him in the complexities of a real case, which he could only know imperfectly, and of Roman law, which even as an advocate he might well get away with knowing superficially. Bonner very rarely says anything that is palpably false: though in the theme of Decl. min. 341 Quidam rem furtivam transtulit per publicanos does not imply ‘a situation . . . in which someone steals an article as he passes through’ (p. 317). And there are no gaps that I have noticed—except that I should like to have seen at least speculation on the fascinating problem as to how a Roman schoolboy was taught to orate rhythmically.

¹ Even the Minor Declamations, which have a good claim to be considered ‘school of Quin tilian’, imply a iudicium in cases of abdicatio (see e.g. 378.1).

6 L. A. Sussman, The Elder Seneca, Brill (Leiden, 1978) S. F. Bonner’s standard work looks at Roman declamation largely through the pages of the Elder Seneca. Professor Sussman’s new book looks more directly at the old man himself, his life, the artistic design of his collection of Controversiae and Suasoriae, his literary criticism, and his lost Histories. It is not a faultless book, but it is scrupulous in giving the evidence from which it argues, and up to the minute with the bibliography; nothing previously published covers this ground so fully and so conveniently. A serious defect of the book (one it shares with my own Loeb of the Elder Seneca) is its neglect of the Greek background. This comes out in small ways, for example in the attempt to explain the technical use of color without attention to the fact that τὴν μετάθεσιν τῆς αἰτίας οἱ Ἑρμαγόρειοι χρῶμα καλοῦσι (RG .308 n. Walz). More spectacular is the remark (p. 167) that ‘Lucian displays (i.e. in his Abdicatus) an uncanny number of parallels with Con. 4.5. . . . Seneca’s works appear to have circulated in the Greek speaking world.’ What these parallels show is that Roman declamation was based, not only in broad outline but in tiny details of argument and style, on a preexisting Greek tradition, followed by the Romans as slavishly as by Lucian and the rhetoricians of the Empire. Firm grasp of this principle would have encouraged Sussman to throw overboard the discussions of early Roman declamation in Sen. Con. 1.pr.12 and Suet. Rhet. 1 [= 25], which, so far as they are not muddled and mistaken, pay inordinate attention to the backwater of the rhetores Latini, and to draw a clear picture. The Greeks brought the controversia and the suasoria (not just the progymnasmata: p. 4) to Rome, doubtless in the second half of the second century . And whatever the deviations of the Latin teachers may have been—and Cicero too resists the wholesale Graecizing of Roman education—Greek teachers had by the days of Augustus taken a stranglehold: Latro the Spaniard and Hybreas of Mylasa play the same game. [Classical Review  29 (1979), 231 2]

Reviews

331

It is Sussman’s view that Seneca’s ‘stated reliance on memory is a conventional literary fiction’ (p. 77). It is true that there was such a convention, but it is less clear that Seneca, who makes much of the present decay of his wonderful memory, and uses it to explain the randomness of his book (Con. 1.pr.2–4), fits very snugly into it. Why did he need to be random if, as Sussman seems to assert, the bulk of his book rests on published sources? Sussman cites, but does not try to explain away, Seneca’s awkward remark that he is doing a service to the declaimers: Fere enim aut nulli commentarii maximorum declamatorum extant aut, quod peius est, falsi. Itaque ne . . . ignoti sint . . . suum cuique reddam (Con. 1.pr.11). I do not see how Sussman can go on to say that ‘Seneca must have had access to the commentarii of many declaimers’ (p. 81); nulli extant does not even imply that such works were unpublished: it means they did not exist. We should, then, regard the sources that Seneca does refer to (like Otho’s book of colours) as exceptional; and we should be more cautious than Sussman is of adding to the exceptions: in particular, the pseudepigrapha assigned to Latro (Con. 10.pr.12) may well be what Seneca primarily had in mind in adding aut . . . falsi at Con. 1.pr.11, and they do not make publication of genuine Latro declamations likely (p. 80 n. 149). I add some notes on details. P. 3 A commonplace was not a ‘narrative’. And, despite Augustine (RLM 139.28–30), hypothesis and thesis are best regarded not as species and genus but as species of the same genus (so Sulpicius Victor in RLM 314.6–7). P. 36 Con. 1.1.1–3 does not seem to me to be ‘in connected form’. P. 38 Much of Seneca’s collection cannot be used as evidence that ‘the period . . . gave way to a more choppy arrangement of numerous sententiae strung loosely together’. And the exceptional Con. 2.7 does not suggest that the generalization is true. P. 40 Con. 1.1 does not state or imply a law ‘that a son can be disinherited for disobedience’, nor does any other theme that I know. Where the abdicatio law is stated (as e.g. in Decl. min. 371), it is merely abdicare liceat. This is not, then, a case of conflict of laws at all. P. 45 Clarke’s citation from Hudibras does not deserve Sussman’s ‘sic’, nor does Ben Jonson on p. 170. P. 48 Seneca’s son Mela was not thinking of becoming a rhetor: a philosopher, rather. P. 101 None of the words listed here seems ‘poetic’, least of all nasutus. P. 111 The division is not ‘a part of the speech which regularly occurred after the narration’.¹ Pp. 156–7 I cannot believe Grisart’s view that Gaius’ harenam . . . sine calce alludes to the Elder Seneca. The Elder did not ‘compose’ the extracts from the declaimers that he lists, and it would not be a very telling insult to call declamations commissiones. Pp. 163–6 Still less do I believe that ‘Quintilian harbored an incredibly deep antipathy towards the Annaei’. Sussman’s view rests, at least in part, on a misinterpretation (Butler’s) of Quint. 10.1.90: et (translated by Butler, characteristically, ‘but’) . . . magis ¹ Cf. Winterbottom (1974b), . .

332

Papers on Quintilian and Ancient Declamation

oratoribus quam poetis imitandus is in context almost a compliment, not ‘a clever piece of spite’. P. 168 I should be glad to know the evidence for ‘the continued importance of declamation . . . after the fall of Rome and until well into the Renaissance’. P. 169 n. 79 Not Decl. mai. 363, but 5. Sussman, to my mind, over-values his author. There is nothing especially subtle about the unity that (as Sussman well argues) Arellius Fuscus gives to the Suasoriae, and the unity Sussman finds in the first preface of the Controversiae is tenuous. Sussman is unable to demonstrate any distinction in the thought processes of one who, while holding that things have declined ever since Cicero, gives extracts from post-Ciceronian declaimers to provide models for his sons, and (if Sussman is right: p. 91) to prove that Roman declamation surpassed Greek. Nor is ‘profound’ (p. 105) a word I should have applied to any of the judgements penned by the pleasant old gentleman about whom Sussman has given us this pleasant book.

7 J. Cousin (ed., tr., comm.), Quintilien. Institution oratoire, Tome  (Livres  et )—Tome  (Livres  et ), Les Belles Lettres (Paris, 1978; 1979) These two volumes bring the Budé Quintilian near to completion. The previous four were reviewed in this journal¹ and the judgements passed there continue to impose themselves. Before I proceed to further unfavourable criticism, I should like to stress that the translation, for which most people will turn to this book, is a model of accuracy and elegance. The text differs little from my own, except in punctuation. C., with his translation to think of, naturally tends to print conjectures where I was freer to obelize. I am sorry that he has rejected Gertz’s elegant repunctuation of 11.3.30 and 126. At 9.4.58 he reads et figuris mutare casus; the translation suggests that he means [et] casus, but the matter is left in doubt by the apparatus, which tells us both that et was added by the ed. Ven. and (implicitly) that it was deleted by Spalding. At 9.4.82 C. does not succeed in printing Gertz’s conjecture, though he apparently intended to do so. I have noticed few misprints, though causarum tum has dropped out in 8.pr.10 and vero in 8.4.10. The apparatus is as unsatisfactory as ever. It is inconsistent with the text e.g. at 9.3.97, where C. prints my deletion of et but notes that he accepts Halm’s de, and at 9.4.95, where he prints absit tam, translates Christ’s absit enim, and implies in the apparatus his choice of absit tamen. It contains palpable absurdities at e.g. 8.2.20 (vitam on both sides of the equation), 8.4.5 (a reading of A in a passage correctly stated to be omitted from A), 9.2.38 and 9.4.106 (A said to read both si and etsi). I see no point in repeating my strictures on the constant use of the codex descriptus H, but in these volumes C. exhibits a

[Gnomon 52 (1980), 785 6] ¹ Winterbottom (1977b), 574 8 [= R.2 above, pp. 315 19]; (1978) [= R.3 above]; (1979) [= R.4 above].

334

Papers on Quintilian and Ancient Declamation

new—though fitful—enthusiasm for a copy of that, T, whose only virtue lies in its corrector, the excellent Ekkehart IV. Not that T is reported any more accurately than H (or any other manuscript): in Book 8, readings of t are given to T at 2.15 (ystorici); 3.41; 6.48, 49, 50, 52, 61, 62; neither T nor t has caecilius at 8.3.35, nisu sed at 8.4.9 or oculos at 8.6.19; at 8.3.11 t (not T) has vero (not vera); at 8.3.47 C. has got into a muddle between putant and putat (other muddles e.g. at 8.pr.30; 8.3.59; 8.4.29; 9.2.26, 49 and 60; 9.3.22; 9.4.63; 10.1.101). C. attributes to A many readings that are in fact its corrector’s, so that a quite false impression is given of the manuscript’s relation to G. The manuscripts thus maltreated are otherwise abused. Argentoratensis and Monacensis (the latter, I am glad to say, still with us) are confused at .175. t is cited in the sigla for Book 11, though his corrections cease during Book 10. Paris. lat. 7530 (which is not employed everywhere it might be) is called P in one passage where it does appear (8.6.37 ff.), though we were promised P 7530 in the list of sigla. P is also regularly, and conventionally, used as the sign for Paris. lat. 7723, though the sigla lists vary between P and Par. lat. 7723 and Par. 7723. At 10.1.38 it is called P 7732, and at 11.1.44 its corrector appears as p 7723. Finally, at 9.2.54 C. records from H quam gaudeo et iam si quid ab eo abstulisti et abs ce lius, with the comment ‘quae verba nusquam leguntur’. They come from two sections before; and they prove, as Peterson showed,² the interrelationship of Bg, H, T and F. I note thirteen conjectures which C. attributes, implicitly or explicitly, to himself: 8.2.13 (p. 279): radix corporis (compare Unger’s radius corporis); 8.3.26: ‘prolem’ dicendum (anticipated by Osann) in versu et (anticipated by Doerry) ‘prosapiam’ insulsum (hardly convincing); 8.3.39: et iam respondi talibus (anticipated by Halm, and perhaps left over from his apparatus); 8.3.59 (p. 286): Lydica for Iadica (put forward very tentatively in the course of a confused discussion); 9.3.72: venustatem quamdam (why not quandam?) ex verbo non dissono accipiat (in essence Christ’s reading, as refined in my own apparatus; I think that non eadem is best deleted, and aliquam retained); 9.4.68: spiritum (anticipated by Spalding); 9.4.143: perdit utique et; 11.2.7: extulisset; 11.3.13: [nisi] (already deleted by b) libera vitiis (which deprives the sentence of meaning); 11.3.21: fluctus (anticipated in early editions). I often disagree with C. about the attribution of conjectures. He disagrees with himself on Serranum at 10.1.89 (Lange on the left-hand page, Sarpe on the right-hand) and prava at 10.1.130: did Sarpe think of this in 1815 (lefthand page) or 1819 (right-hand page)? I should like to have been responsible for the deletion of usu at 10.1.83, but I was not. Others will find more to their taste than I do C.’s prefaces to the four books, with their emphasis on sources. His notes are strong in Realien (not always

² Peterson (1891),  .

Reviews

335

very relevant: 8.6.49: currents of the Euripus; 11.3.16: the internal arrangements of the nose) and bibliography (but let us have Castorina (1946) at 10.1.115), weaker on interpretation. Some details: .3–14: An obscure attempt to relate Book 8 to the stylistic background, not helped by characterization of the new rhetoric as marked by ‘redondances, . . . traits recherchés, . . . périodes compliquées’ (5), and by C.’s apparently regarding Q. Haterius as a contemporary of Q. (10 n. 1). 8: The real joke about the suicide of Albucius Silus was that he was, consciously, bringing to life a ‘law’ of the declamation schools (e.g. [Quint.] Decl. mai. 4). 13 n. 2: I cannot believe that ‘[l]a plupart des allitérations en latin se rencontrent dans des contextes religieux ou techniques formulaires’. 317 (on Q. 9.3.17): I presume that Q. is thinking of the Greek φιλεῖ γίγνεσθαι, and that Sal. Jug. 34.1 is not the passage he had in mind (unless he was badly misremembering it). .314: C. assumes without argument that Demetrius of Phaleron was author of the Περὶ ἑρμηνείας. 330: C. seems to assume that the treatment of Eprius Marcellus and Vibius Crispus in the Dialogus is uniformly laudatory; but Aper’s praises are implicitly refuted by Maternus (13.4).

8 J. Cousin (ed., tr., comm.), Quintilien. Institution oratoire, Tome  (Livre ), Les Belles Lettres (Paris, 1980) This seventh volume concludes the Budé Quintilian. It is a set that will be found useful for its reliable translation of a reliable text, and for at any rate some parts of its notes. Cousin deserves our congratulations; he doubtless recorded the final words of the Ambrosian manuscript, FINIT AMEN, with the same mixture of satisfaction and melancholy the reviewer felt when he did the same ten years before. After my earlier strictures on Cousin’s apparatus,¹ it is a pleasure to report that matters here are much improved. The constant irritation of the unnecessary presence of H remains; and conversely C. does not employ the evidence of the excerpta in 12.10.10–15, though he used it in 10.1.46–131. There are still blots (B on both sides of the equation at 8.7—this and similar later references are to Book 12—, the variant que for quae recorded at 10.28) and muddles (as at 1.7). C. habitually reports Petrarch’s corrections in Parisinus lat. 7720 as the readings of the first hand (pr.1 etc.); and where he cannot deduce from my apparatus the provenance of a reading he puts it down to ‘edd.’ (so e.g. 3.11 formularii) without making further investigations. But a good deal of what he records here is both true and rationally presented. The text is very little different from my own. At 10.15 C. prints Stoer’s optendunt, whose aim was (one supposes) to bring the mood into line with the following delitescunt;² but he then punctuates in such a way as to dissociate the two verbs. At 6.3 he presumably meant to print veniae spes est (not veniae et spes est); this is not Becher’s conjecture (veniae est spes et) but Davies’s.³ (At 10.50, incidentally, voluptates is the reading of G, not a conjecture of Gertz.)

[Gnomon 53 (1981), 197 9] ¹ Winterbottom (1977b), 574 8 [= R.2 above, pp. 315 19]; (1978) [= R.3 above]; (1979) [= R.4 above]; (1980a) [= R.7 above]. ² See Buttmann’s n. in Spalding (1798 1834), .616. ³ Spalding (1798 1834), .557.

Reviews

337

C. prints two conjectures of his own: 9.16 scalpta (perhaps right) and 11.12 (praesumuntur). He argues for the latter on p. 238, mysteriously: he rejects Radermacher’s praemuniuntur while not finding my argument against it ‘très convaincant’; regards Buttmann’s praecipiuntur as a ‘conjecture séduisante’, though my argument against Radermacher tended in its favour; and then proposes a word praesumuntur that involves Q. saying that ‘ce qui est inné, antérieur à toute expérience’ is simple enough to be learned in a few years. C.’s translation ‘les notions que l’on présume’ is something else again, and open to my objections to Radermacher.[⁴] The preliminary ‘Notice’ and succeeding ‘Notes complémentaires’ have, for my taste, a great deal too much information on the ‘artistic’ sections 10.1–9: superfluous pages made longer by apologies for their superfluity (‘ce qui n’a aucun intérêt pour notre texte’, ‘qui n’est pas en cause ici’, ‘pour aller à l’essentiel’), and by our being told twice on successive pages (pp. 212–13) which room of the Uffizi houses the same picture of Botticelli. More reasonably, there is a good deal on the Atticist/Asianist quarrel, not always convincing: P. 165 It is for Q. (10.16), not for Santra, that the dispute is antiqua. P. 188 ‘Ceux qui critiquent la manière de Démosthène, ce sont les Attiques, tels Calidius et Calvus’: not Calvus, anyway, who imitated him. P. 229 I can see no irony, voluntary or involuntary, in Q.’s calling Scipio, Laelius and Cato velut Attici Romanorum (10.39): he is merely classing them as the old Roman classic orators. P. 242 Cicero by no means says that Cato ‘a été l’un des modèles pour les Attiques’; indeed he says: Catonem vero quis nostrorum oratorum . . . legit? Aut quis novit omnino?⁵ Some miscellanea: P. 189 Is it any longer believed that Q. was Tacitus’ teacher? P. 191 For further information about b’s reading at 1.42 see Winterbottom (1970b), 209. P. 192 aliquamdiu at 2.6 has nothing to do with the rhetores Latini. Ibid. C.’s list of philosophers who were ‘dans l’ombre du pouvoir’ does not invalidate the text from which he starts (2.7). P. 203 C. exaggerates the technicality of negotiis in 7.10; Hor. Carm. 3.5.53 is more relevant than Q. 3.5.17. P. 203 The tituli of Q.’s chapters are common to AB, and are surely late antique in origin, not medieval. P. 204 On the spelling of plaudere and cognates see further Winterbottom (1970b), 55. P. 210 C. seems to believe ‘l’histoire horrible’ of the Olynthian tortured by Parrhasius in Sen. Con. 10.5; it is of course declamatory fiction. P. 227 Was Isaeus really ‘célèbre’ for δεινότης? Dion. Hal. Isae. 20 merely says that he provided τὰ σπέρματα καὶ τὰς ἀρχάς of Demosthenes’. Pp. 231–2 Nothing that C. says in defence of his retention of the manuscript text at 10.61 touches the real difficulty, that Q. is talking of the grand-style orator (Hic orator, mistranslated by C.) in general. How can he then proceed to say that apud hunc (‘c’est-à-dire Cicéron’: quite

⁴ [For a new conjecture see Winterbottom (2006) = A.19 above.]

⁵ Cic. Brut. 65.

338

Papers on Quintilian and Ancient Declamation

wrong) ‘la patrie . . . parfois . . . parlera à Cicéron [my italics] dans le discours prononcé par lui au sénat contre Catilina’? P. 236 C. argues against Stroux’s Iulio at 11.3 that ‘il est impensable que le mot Africano ait disparu’ (why so?) and unthinkable that this man should have spoken thus of Domitius Afer ‘étant données les relations entre les deux hommes’ (what do we know of them?). That some name lurks here seems certain; Campanus’ aliis, ‘tentant’ to C., is very feeble. P. 242 C. tells us twice that Cato wrote both de agricultura and de (re) rustica; indeed, he goes so far as to prophesy a Budé of both. P. 243 On the asyndeton laudem praesentem futuram (11.29) see Winterbottom (1970b), 70–1; 218. C.’s volume ends with an ‘Index vocum graecarum’, an ‘Index hominum et locorum’ (I should here apologize for omitting Iulius Africanus from my own; this orator is still hard done by in C., who omits the reference in 12.10.11), a useful ‘Index rerum’ and a ‘Conspectus operis’ listing chapter titles: something I regret not having included myself.

9 L. Håkanson (ed.), L. Annaeus Seneca Maior. Oratorum et rhetorum sententiae, divisiones, colores, Teubner (Leipzig, 1989) Appended to the preface of this book are noble words [p. ]: ‘Paucis diebus post quam huic editioni summam manum imposuit librumque imprimendum Lipsiam misit Lennart Håkanson de Seneca rhetore optime meritus diem supremum obiit. Senecae suo satis vixit, litteris et nobis non item.’ All who knew Håkanson and admired the man and his work will echo this sentiment. His tragically early death robs scholarship of one whose name deserves to stand alongside those of the most illustrious critics of Latin prose and poetic texts. This posthumous book will for a very long time be the standard edition of the Elder Seneca, and it shows H.’s virtues exercised at their fullest stretch, grappling with a deeply corrupt text: an acute assessment of a transmission, a clear eye that sees infelicities where we had earlier passed unwarily by, and the most delicate touch in emendation. The preface argues for a stemma in which the three ninth-century manuscripts (so dated by B. Bischoff, as we learn from a citation [p.  n. 1] of Dr Vervliet’s regrettably unpublished dissertation), A, B, and (the parent of) V, descend separately from a hyparchetype α (the other hyparchetype, β, gave rise to the excerpted tradition, and is often not available to us): whereas it has been generally thought that V, though cruelly interpolated, is at root independent of an ancestor of AB. H.’s arguments fall well short of proof. The figures of omissions given on p.  do not seem to prove that V is descended from an apograph of α. I agree (and have myself argued) that omissions common to AB against V do not prove the independence of V (pp. –). But I am not sure that the list of readings for the splits AV/B and BV/A prove [Classical Review  41 (1991), 338 40]

340

Papers on Quintilian and Ancient Declamation

that A, B, and V’s parent descend separately from a common ancestor (p. ): coincidental error and conjecture are in any case needed to explain these phenomena. And when (pp. –) H. has to admit the possibility that V contains readings from outside the traditional altogether, he is bowing to facts that have led others to see V as independent from, though even more corrupt than, the common parent of A and B. H. does, however, seem to be right (p. ) that V’s corrector, at least at Con. 1.5.2, gives us something not found in the archetype of the manuscripts we possess and not easily won by conjecture (in the Excerpta, where he—I think rightly—regards M as without rival, he remarks that at Con. 7.pr.8 M’s corrector supplies words which are omitted in αM but which must be genuine (p.  n. 2); but the omission could easily be coincidental, and M² may well be correcting from M’s exemplar). Still, as H. himself says, little hangs on the resolution of these problems. The editor must in any case pay far more heed to AB than to V. H. goes on to make interesting observations on the character of α (pp. –).¹ H.’s text indicates by small raised square brackets the portions also witnessed to by the excerpted tradition, and marks the start of new epigrams by capitalization (occasionally this results in ambiguity: do ‘Antonius’ and ‘Lepidus’ start new epigrams at Suas. 7.6 or not?). The apparatus is extremely rich. It reports the errors of A, B, and V, as it would seem, in full; other MSS of the complete tradition appear only as bearers of conjecture (is it not time that someone identified ‘cod. Vat. Schotti’ at Suas. 1.16?). At times H. gives welcome help in interpretation, but one often feels the lack of the companion commentary, which remains in an incomplete state in the library of Uppsala University.[²] There is a very useful bibliography of editions and critical work on the text. And there are four separate indices of names.³ The beautifully produced volume is a credit to the House of Teubner at Leipzig (whither now?). It would be impossible in a short review to discuss all H.’s adjustments to the received text, which run into hundreds, if not thousands. I illustrate from one Suasoria (7) the difficulties he faced and his success in overcoming them. In five and a half pages I count eighty-five places where the transmitted text has (always with good reason) been abandoned. Thirteen of these are H.’s own changes. At §§ 1: cupientis, 10: esse quam, 11: insignita (the order insania M. would explain the omission more neatly), and 14: more, ut quam primum tantum tumeant quantum potest, a iure iurando et dixisset multa, (where the transposition leaves a awkwardly far from coepisset) he is building upon earlier suggestions. Elsewhere he provides new solutions to old problems: 2: interim for in te, more ¹ For errors of the type xy and xy cf. Winterbottom (1970b), 189 and 200. ² [See pp.  and . The commentary on the first book of the Controversiae is now published: cf. Håkanson (2016).] ³ Syme (1978) might have been exploited: see e.g. p. 143 with n. 4 on Fabium in Con. 10.pr.13.

Reviews

341

allusive and nearer to the transmission than Gertz’s ante te; ibid.: eloquentiam, Cicero, nec perituram, rogo hardly certain (rogo seems so long delayed as to become superfluous), but an advance on the frigid received text; 8: pateris perire (avoiding emendation of the verbs; but Ciceronis seems inelegant before Cicero). Most notably, he strikes out into new territory: 1: Difficilis est (sc. Antonius), looking forward to non feres (perhaps a colon should follow est?); ibid.: Pendent nefariae proscriptionis tabula tot praetorii . . . ;⁴ 8: Qualis est pactio? Aufertur Ciceroni ingenium; sine vitam (sc. auferri), immeasurably more pointed. The stroke is characteristic of H. Characteristic of the troublesome text is that difficulties remain: 1: Can luerentur—uneasily explained by Edward—really stand (more may be corrupt here)? 2: Read Antonium? Note how, to avoid another ambiguity, Cestius says in 3: M. Antonium illum indignum hoc successore generis. 3: Should not At videris start a new extract, continuing to nocentiorem? But persuasisse seems out of place, and corruption may lurk. 4: Edward’s objection to sui still stands; H. does not mention C. F. W. Müller’s pusilli, but something on those lines seems called for (parvi?). 8: I much prefer Vive ut to Videlicet. But one hesitates to differ from a scholar as acute as H. How sad it is that friendly argument on such points is now cut short.

⁴ Cf. OLD² s.v. pendeo 2d, where the first example suggests rather nefariae here.

10 L. A. Sussman (ed., tr., comm.), The Declamations of Calpurnius Flaccus, Brill (Leiden, New York, and Cologne, 1994) Lewis Sussman has served declamation well, with a book on the Elder Seneca (1978) and a pioneering translation of the Major Declamations that go under Quintilian’s name (1987). He now gives us a text (without apparatus), facing translation, and commentary for Calpurnius Flaccus. The lavish volume is rounded off by a full bibliography and index. In the most interesting section of the introduction S. inclines towards a late first-century or second-century date. Other indications confirm, or point later: 1.7 (all references by page and line of Håkanson, as in S.): ante . . . dum (cf. TLL s.v. dum 2217.77: Ammianus); 24.19: ea propter (TLL s.v. is 486.6 ff.); 27.21: fatŏ in a clausula; 29.6: ut . . . itidem (ante- and post-classical); 32.12: in tempore suo (see n. on p. 206); and 37.12: perpetes (n. on p. 227). S. has set himself to revise a text edited in 1978 by Lennart Håkanson. It is no harsh criticism to say that he is unable to advance beyond the work of one of the finest critics ever to touch a Latin text. But it is much to his credit that he worries away at the remaining problems. He could be right to punctuate 4.8–11 as questions, and his objection to Håkanson’s praestiterint at 37.11 is cogent. But he misunderstands Håkanson’s intentions at 2.9 mittit, at 16.1–2 (where his view is the same as Håkanson’s) and at 20.4, where Håkanson certainly did not think uti an infinitive. And there are quite a number of places where Håkanson’s text is superior, not least in punctuation: especially at 1.5 and at 30.2 (where the translation is wrong); I should myself suggest some adjustments to Håkanson’s punctuation, e.g. at 5.2 (iudices?) and 9.7 (sis:). There is a certain laxity of editorial convention (a conjecture in angle-brackets at 2.13, another in angle-brackets and obelized at 39.2). S. very occasionally makes conjectures of his own: at 20.11–12 (where the order is certainly very odd), 39.2: afflictos (after Håkanson’s afflictum), and 39.23: deam (but, apart [Classical Review  45 (1995), 40 2]

Reviews

343

from anything else, ream goes better with the remarks on human responsibility that follow). S. discusses the problems of translating Calpurnius on pp. 22–3, remarking that ‘I have felt free to use colloquial touches which would be normal under such circumstances’ (would they?). That seems to conflict with the observation that ‘[t]he style used is slightly formal and occasionally old fashioned’. And S.’s version is sometimes unhappy (3.20: aliquid impudici habet ‘is a closet homosexual’; 34.14: puella ‘girlie’; 36.10: fescennina ‘the off-color marriage songs’). Indeed it is not infrequently just wrong, as at 4.22, 5.2, 5.13–14 (patere is imperative), 7.6 (rather ‘I must reveal the way my host felt: he thought this treatment a kind of tyranny’), 11.8 (petit is perfect; so too perit at 35.17?), 17.2 (the passives are impersonal), 23.1 (rather ‘who had been offended’, i.e. enough to want his son dead), and 34.5–6 (but see p. 213). S. is right to emphasize that these fragmented epigrams are not easy to interpret beyond cavil. He has been so fortunate as to have the advice of D. A. Russell, and might, I think, well have taken it even more often than he has (thus at 3.4–5, 3.11, 10.11–12, 26.17, 34.17–18). S. contributes constructively e.g. at 27.3–4, 27.6, 30.18–19, 31.9, and 36.6–7 (in several of these places he might be firmer in deciding between his own alternatives). The following notes are offered towards a continuing debate: 2.12: Why cannot amavit stand? It gives rhythm (something S. hardly ever mentions); the subject of inquit will be the wife’s advocate. 3.20: Perhaps qui would be less ambiguous. 5.1: Håkanson’s conjecture destroys the rhythm. 5.18: Surely (p. 108) the fovea was covered over before the youths arrived? 9.8–9: It is perhaps time that I explained what I think the transmitted text means: ‘Your character is such, young man, that you wanted to blind your father—or else he wants to blind you.’ That is, either the son did blind the father in secreto, or (though he did not) the father is trying to get him punished for it. Either way, it is (not very convincingly) argued, the son can hardly be a good boy. One might think of ut . 10.18–20: These are reported words of the father, as he faces the problem of whether to tell his wife of the dream. I think that arcesso in line 20 means simply ‘send for’: the father is about to tell his wife. Then (10.21–2) not as Håkanson, but ‘at the sudden appearance of the dream’. In 10.25–11.1 all the father means is that he again saw a dead son. 13.17: Perhaps veneficum: the tyrant had recognized the castle doctor as the poisoner—and so was demanding a second opinion. 15.5–6: explicare is here used in a financial sense (TLL s.v. 1731.17 ff.). 16.18: Perhaps nihil : questa est. 19.1: The son alludes in mortem . . . necessariam to the plague. 19.17–18: Is not the reference merely to prostitution? 19.23: proelium seems oddly used of the μονομαχία of the brothers (pugna in ‘Quint.’ Decl. min. 271.9). 23.16: One expects pro alia. 24.23–25.1: The proposed ellipse (p. 176) does not seem easy to me. Perhaps quis . . . filii (or rather fili). 25.12: The rhythm enforces feriat (subj. orbitas). And then rapuisti is supported by the rhythm (cf. also 32.4–5).

344

Papers on Quintilian and Ancient Declamation

Similarly in 25.19 honore praestatis, hic sanguine will be what Calpurnius wrote (what construction could the participle have?). 26.4–5 (n. on p. 180): Not ‘[b]ut I waited too long’, but ‘I waited for a long time’. 35.6: Delete pater, who should come as a surprise in what follows. 35.17 (n. on p. 220): Perhaps then moreretur, with rhythm. 37.12 (n. on p. 227): The nights were ‘free’ not because of the absence of sexual demands from the wife, but because he had the nights with his wife, while his brother had to snatch furtive meetings with his lover. 38.1: Delete imperator. 38.20: Why not quis (cf. Meister’s change at 7.20)? 39.9 (n. on p. 239): forte is preferred for its rhythm. 40.4: reddit (N) is right (cf. esp. 31.21; the same corruption at ‘Quint.’ Decl. min. 335.th.). The commentary wastes space in raking over the past (Lehnert’s edition should be allowed merciful oblivion), and is over-hospitable to emotive alliteration (‘hissing s’s’, ‘explosive alliteration of p’, ‘mournful m’s’; see Winterbottom (1984) on Decl. min. 306.th.). I noticed a dozen or so misprints. In the text delete carcerem after cum¹ at 5.3, and read necessariam at 19.1, stuprandam at 33.12.

11 A. Stramaglia (ed., tr., comm.), [Quintiliano]. I gemelli malati: un caso di vivisezione (Declamazioni maggiori, 8), Edizioni dell’Università degli Studi di Cassino (Cassino, 1999) The important Italian contribution to the study of Roman declamation continues with a new project, ‘un ampio progetto di ricerca’ on the pseudoQuintilian Major Declamations, at the University of Cassino. When complete, this will provide both a complete Italian translation of these difficult pieces (Lewis A. Sussman has already blazed the trail in English) and studies of individual declamations, with introductory matter, translation, and notes. The volume under review is the first of these individual volumes. The introduction proper, by Lorenzo Greco, interestingly summarizes ancient views on the two key features of Decl. 8, vivisection and twins, and usefully lists the points of resemblance between its wording and that of the closely related Decl. 5. Stramaglia himself provides an important ‘Nota al testo’, a learned and acute analysis of the problems raised by the transmission of the corpus as a whole. Admirable though it is, this analysis looks rather out of place in a volume whose Latin text is printed without any apparatus criticus (a lack, by the way, that makes the angle brackets round recipe quem mihi credidisti in § 5 (155.20–1 Håkanson) very misleading). But despite that, the editor has thought through the many problems (though I should welcome a note on perituro . . . fato at 1 (152.11 H.); ‘una malattia senza speranze’ is the right sense, but how do we get to it?). S. discusses them intelligently in his full notes, and his translation is always worth consulting. The bibliography is invaluable (though it lacks an old article by S. Vassis,¹ praised in n. 120).

[Classical Review NS 50 (2000), 305 6] ¹ Vassis (1906 7).

346

Papers on Quintilian and Ancient Declamation

S.’s independence of mind is well illustrated by his willingness to take issue with the great and much regretted critic Lennart Håkanson, who edited these declamations for Teubner in 1982: rightly, I think, at nn. 80 (where (§ 15 = 165.25 H.) one might think of deleting the first duobus), 111, and 116; wrongly, in my judgement, at nn. 33, 118, and (perhaps) 44 (on § 8 (158.20–1 H.), though I am not sure that the truth has yet been discovered there; see below). He makes conjectures of his own at 3 (154.5–6 H.): artis suae , and 17 (169.3 H.): ex secto (but the theme employs exseco), and reports emendations of W. Hübner at 13 (163.20–1 H.) that I am not astronomer enough to judge. Many passages in such a difficult text naturally remain uncertain: 4 (155.9–10 H.): What does morte qua medicus parabatur mean? 7 (157.20–1 H.): magnorum parentum can hardly be right (the run of the sentence does not favour S.’s view that magnorum goes with what precedes). 14 (165.9–10 H.): Can numquam ex hoc rationem reddas stand? I make some minor suggestions of my own: 4 (155.4–5 H.): quod etiam de similibus. Read merely quod de tam similibus. Here etiam is without point; it could have arisen from a tam added above the line and entering the text in the wrong place. 6 (156.16–19 H.): The point of the vexed sentence starting Facinus est . . . (seen by Burman) is that it would be a scandal if the husband were to be let off the hook just because he is being accused under a restricted law that does not (the defence may successfully argue) apply to something so heinous as killing his son. Then read: si illa de minore dolore quereretur, ‘if the mother were complaining of something less serious’. The next sentence should perhaps start: Itane (so Schultingh) matrona (transposed) impudenter facit quod pro detracto cultu . . . 8 (158.20–1 H.): Perhaps read: Frustra captas videri simulatione (MSS ultione[m]) magnae caritatis ab omni curae ratione sepositus. The husband, says the accuser, is trying to represent himself as having been excluded from the decision about his sons by the affection that led him to leave it to the doctor. In fact he felt no such affection, and it is mothers who get ‘excluded’ (6 (157.12 H.): seposita). Cf. 4 (154.23–4 H.): Vultis intellegere, iudices, nihil inpatientia caritatis fecisse patrem? Non retulit ad matrem. 9 (159.28–160.3 H.): Sepono paulisper immanitatem patris qui credidit; [et] . . . queri totius generis humani nomine volo. For the asyndeton cf. 16 (166.25–7 H.): Differo paulisper quod . . . ; publico potius mortalitatis contendo nomine . . . 17 (168.5 H.): deprenderint (for the rhythm) (perhaps also reprendere at 9 (159.19 H.)). In the notes, I miss a reference on 12 (162.22–3 H.): Nihil . . . voluit esse rerum natura tam simile quod non aliqua proprietate secerneret to Quint. 10.2.10. I have doubts about the translation e.g. at 2 (153.5–6 H.): et morborum . . . concessit (in effect ‘made illness part of the art of medicine’; cf. 16 (167.3–4 H.): ratio sanitatis intercidit si consumit medicina tantundem). I found almost no misprints, though in n. 120 read ‘non interpungeva dopo ingratae’.

12 M. Weissenberger (ed., tr., ann.), Sopatri Quaestionum divisio—Sopatros: Streitfälle. Gliederung und Ausarbeitung kontroverser Reden, Königshausen & Neumann (Würzburg, 2010) In 1988 the reviewer, with Dr D. C. Innes, published an apparatus criticus without a text;¹ twenty years on, Michael Weissenberger has published a text without an apparatus criticus. The text is that of an important and lengthy collection of treatments of controversia themes by an apparently fourthcentury rhetorician, a corpus which has largely been ignored even by scholars interested in this somewhat remote area. That neglect was for a long time the result, in large part, of the appalling state of the text, which in Walz’s RG  (1835) was hardly more intelligible than it had been in the Aldine of 1508. In producing our own volume, we were well aware that we were only going part of the way towards remedying this sad state of affairs. One of our reviewers justly remarked that what was needed was ‘a completely new text, . . . arranged in such a way that the reader can easily grasp the subdivisions within each διαίρεσις’.² Meanwhile the ‘philologi Britannici’, however ‘optime meriti’, had—Weissenberger says—left readers sunk in deep water, for the result of their labours has been ‘ut Sopatri lectori duo nunc semper in manibus libros esse necesse sit oculique huc illuc assidue distracti citissime defatigentur’ (p. 7). This ‘calamitas’ (as Weissenberger dubs it, p. 8) has been remedied to the extent that we now have, thanks to his labours, a proper text³ (numbered declamations

[Gnomon 83 (2011), 394 6] ¹ Innes Winterbottom (1988). Dr Innes kindly commented on a draft of this review. ² Rutherford (1990), 22. ³ Two kinds of type (not quite as different as one might like) distinguish Sopatros’ own re marks from the ‘fair copy’ extracts.

348

Papers on Quintilian and Ancient Declamation

equipped with numbered sections),⁴ followed into the bargain by a German translation. It is only fair to say, however, that anyone who wishes to work in any detail on this intriguing text needs now to have not just two but three books to hand: Walz (Wa), Innes–Winterbottom (IW), and Weissenberger (We). As the reviewer has found, the absence in We’s text of frequent marginal indication of Wa’s page and line numbers makes this juggling act the more frustrating; and it does not help that the book refuses to open flat. Further, all that appears below We’s text is a register of the places where he has added his own conjectures to a text conflated from Wa and IW.⁵ Only by constantly consulting Wa and IW can the reader tell what elements in the conflation are due to Walz, what to manuscripts unknown to Walz,⁶ and what to conjecture. The ordinary reader (if such a one ever strays this way) will be enormously grateful for what Weissenberger has done; but he (or she) must be warned that the text is a construction whose underpinning has to be (painfully) looked for elsewhere. ‘Restat, ut contendam affirmemque hanc editionem nulla codicum perscrutatione confectam ideoque criticam, quam dicunt, neque esse neque haberi debere’ (p. 8). Such a critical edition would not be an insuperable task, and it should be encouraged by the Bibliotheca Teubneriana, which at one time was undertaking the revision of Walz. It might rest virtually on MSS U and C alone, though U is hard to deal with.⁷ Meanwhile, one wonders if We’s text could be displayed on TLG, with Walz’s page numbers added. Readers need all the help they can get with this text, and We gives it in various forms. First of course by the translation, the first complete one in any language,⁸ and to be turned to constantly and rewardingly. But we are also given: an ‘Einleitung’ (251–67) explaining with exemplary lucidity the stasis system employed by Sopatros and the ways in which it modifies that of Hermogenes, and summarizing what can be said about the author and the circumstances of composition; notes on the translation, quite full on the ‘historical’ declamations (to which most readers may indeed be attracted) but scanty elsewhere;⁹ a very helpful ‘Glossar’ (531–42); an ‘Index nominum’

⁴ The numbering is by paragraphs, and they are sometimes very long (e.g. in no. 22). It would have much aided reference to the translation (would that it were en face!) if quite short sub sections had been introduced within the longer paragraphs. ⁵ In those places, and those alone, he registers the readings recommended by IW and/or, though without detail (under the siglum ‘IW*’), those reported by them from the manuscripts they employed. ⁶ The manuscript we called C is not a Bodleian MS, as We states (p. 7); it belongs to, and is currently kept in, the library of Corpus Christi College, Oxford. ⁷ Cf. Innes Winterbottom (1988),  n. 1. ⁸ We mentions one or two versions of individual declamations. He does not know Russell (2006), from a master hand. ⁹ An engaging feature of the notes is We’s willingness to answer Sopatros back, criticizing his cogency and suggesting counter arguments (thus nn. 105, 126, 132, 135, 158, 171, 185, 300, 304). That is in the spirit of the genre; Sopatros would have been delighted.

Reviews

349

(247–50); and at the end a ‘Gesamtschau’ of the declamations, grouped, as Sopatros arranged them, by the staseis they illustrate (543–6). The whole adds up to a very considerable achievement, for which students of ancient rhetoric will be warmly grateful. The text is naturally what cost We most effort. He has made numerous emendations. Some are obviously right (and IW should have made some of them already), others likely or possible. Many others are violent changes, but it is quite understandable that We wished to print a text he could translate (he almost never employs the obelus). I make some comments (selected from many). Sopatros employs (accentual) rhythm, especially in the declamation extracts (cf. Innes Winterbottom (1988), 12 with n. 25 [= A.10 above, p. 159 and n. 137]), and one should not lightly propose changes that do not take this into account. One instance out of many: We deletes the article in 4.6.8 (26.18 19 Wa): ἐπ’ ἐρημίας τὰ χρήματα. We is occasionally led astray by misunderstanding the reports of IW as to the crucial manuscript C.¹⁰ Thus: 15.3.18 (97.4 5 Wa): C does not have ἀπήλλαγμαι or any equivalent verb; 13.6.2 (83.13 Wa): it does not have διαιρεῖ; 19.6.8 (109.29 Wa): it reads ἔντιμος; 29.1.5 (182.27 Wa): it has καλούμεναι, not καλούμενοι. At 46.1.16 (268.17 Wa) δὲ for γὰρ should be credited to C; so too the addition of ὅτι at 72.9.12 (355.21 Wa).¹¹ Some points of Greek usage: 1.3.46 (6.10 Wa): ἠβούλησε is not a correct form. 1.3.55 (7.7 Wa): For μᾶλλον + comparative see Schmid (1897), 158. 3.1.7 (16.17 Wa): For πλεονεκτεῖν + accusative see LSJ s.v. .2. 9.1.2 (54.18 Wa): Is not διαιρήσειας (IW) a possible late form? 9.1.13 (55.6 Wa): Is a genitive after παραλαμβάνειν without a preposition tolerable? See too the parallel passage cited by Innes Winterbottom (1988), 1 [= A.10 above, p. 135]. 14.4.1 2 (94 22 3 Wa): Cannot μάθῃ stand imperativally? Cf. 1.10.12 (14.12 Wa) ἴδωσιν (also emended). 20.4.36 (115.2 Wa): Comparative for superlative, as e.g. at 20.3.12 (112.17 Wa); see Schmid (1897), 24. 22.1.14 (127.15 Wa): We’s (unnecessary) change seems to involve misuse of the particle μέν. 22.7.37 (136.27 Wa): Can an imperatival sentence be a question? Professor D. A. Russell, who has kindly read a draft of this review, suggests that no more is needed in this passage than the addition of μὴ after εἰ. 26.3.19 (172.30 Wa): We objects to perfect δεδώκασιν with ἄν. It certainly looks odd here alongside an aorist, but compare 22.7.39 (136.30 Wa), also emended. 36.6.1 (222.2 3 Wa): Here and elsewhere We emends a form of πίπτω, used in connection with ἀντίθεσις, to the normal ἐμπίπτω. It seems bad method to normalize so frequently; at least reference might have been made to IW on 52.14 Wa. 48.1.15 16 (281.15 16 Wa): We emends πότερον . . . καὶ to πότερον . . . ἢ. But is this tolerable Greek (rather than εἴτε . . . εἴτε)? 66.5.7 (330.28 Wa): Is ἀπάγω (as opposed to ἐπάγω) used like this? ¹⁰ At 46.4.14 (272.7 Wa) the reading of MS U should have been mentioned. ¹¹ Emendations should have been credited to IW at 46.1.9 (268.9 10 Wa) and 70.3.4 (344.20 Wa).

350

Papers on Quintilian and Ancient Declamation

I contribute some new emendations sparked off by suggestions of We. 1.5.36 7 (9.21 2 Wa): Rather than change to oratio obliqua read ἢ ?¹² 5.4.13 (31.8 Wa): Rather πάντων μᾶλλον, as at 7.5.6 7 (46.28 Wa). 22.8.9 (137.17 Wa): Perhaps rather εἶτα should be replaced by εἰ. 22.9.8 (138.2 Wa): Perhaps better delete αὐτὸ. 23.9.29 30 (161.12 13 Wa): Perhaps punctuate as a question. 24.3.19 (164.16 Wa): Perhaps μαχοίμεθα? 24.5.15 (165.26 Wa): The point of We’s μονήρη is unclear. Dr Innes suggests that the general sense requires something like ὅλως. 25.5.8 (169.22 Wa): Deleting τὸν destroys the rhythm. Perhaps then read τοὐργαστήριον just above. 43.8.16 (250.27 Wa): Perhaps rather τῷ νόμῳ. 76.5.9 10 (368.5 Wa): One might think rather of deleting ὁ ἀπολουτρωσάμενος. Dr Innes remarks that the simple verb is regularly used in this declamation, and compares especially 76.7.18 (370.3 Wa), where our passage is picked up for rebuttal. She also points to another likely gloss just below, at 76.6.4 5 (368.18 Wa), where either ὁ συγγενής or ὁ τῆς κόρης πατήρ should probably be deleted. I noticed very few misprints. But the omission of τοῦτον at 23.4.13 (151.17 Wa) has led to an unnecessary change just before.

We remarks (p. 267), of the footnotes to his translation: ‘Immer sind IW zu vergleichen, die in vieler Hinsicht reicheres Material bieten, aber beim Leser mehr Kenntnisse voraussetzen und deshalb geringeren Wert auf Erklärung legen.’ Readers of We’s own monumental book will need to be pretty knowledgeable themselves; but they will be richly rewarded.

¹² For another approach cf. Schenkeveld (1991), 215.

(p.351) References Michael Winterbottom , Antonio Stramaglia, Francesca Romana Nocchi, Giuseppe Russo Bibliography references: Anastatic reprints have not been recorded. Periodicals are customarily abbreviated according to L’Année philologique (or its predecessors, for earlier journals). Adamietz, J. (1966), M. F. Quintiliani Institutionis oratoriae liber III (Munich). Aizpurua, P. (2005), Calpurnius Flaccus. Les plaidoyers imaginaires (Extraits des déclamations) (Paris). André, J. M. (1983), ‘L’otium chez Quintilien: pédagogie et magistère moral’, in Hommages à Jean Cousin (Paris), 193–206. Austin, R. G. (19542), Quintiliani Institutionis oratoriae liber XII (Oxford; 19481). Austin, R. G. (1955), P. Vergili Maronis Aeneidos liber quartus (Oxford). Austin, R. G. (19603), Cicero. Pro Caelio (Oxford; 19331). Austin, R. G. (1964), P. Vergili Maronis Aeneidos liber secundus (Oxford). Austin, R. G. (1971), P. Vergili Maronis Aeneidos liber primus (Oxford). Avenarius, G. (1956), Lukians Schrift zur Geschichtsschreibung (Meisenheim am Glan). Balbo, A. (20072), I frammenti degli oratori romani dell’età augustea e tiberiana, 2 vols. (Alessandria; 20041). Bonnell, E. (1866–9), M. Fabii Quintiliani Institutionis oratoriae libri duodecim, 2 vols. (Leipzig). Bonner, S. F. (1938), ‘Dionysius of Halicarnassus and the Peripatetic mean of style’, CPh 33: 257–66. Bonner, S. F. (1949), Roman Declamation in the Late Republic and Early Empire (Liverpool). Bonner, S. F. (1977), Education in Ancient Rome (London). Bornecque, H. (1902), Les déclamations et les déclamateurs d’après Sénèque le père (Lille). Bornecque, H. (19322), Sénèque le Rhéteur. Controverses et Suasoires, 2 vols. (Paris; 19021). Bowersock, G. W. (1979), ‘Historical problems in late Republican and Augustan classicism’, in Flashar (1979), 57– 78. Bowie, E. L. (1970), ‘Greeks and their past in the Second Sophistic’, P&P 46: 3–41. Braund, S. H. (1988), Beyond Anger: A Study of Juvenal’s Third Book of Satires, Cambridge et al. Brugnoli, G. (1961), ‘I Synonyma Ciceronis’, in Atti del I Congresso internazionale di Studi ciceroniani, vol. 1 (Rome), 283–99. Bureau, B. (2007), ‘Ennode de Pavie adversaire de “Quintilien”. Éthique et éloquence autour de la controverse liberi parentes alant aut vinciantur (Ennod. Dict. 21, Ps. Quint. Decl. Maior. 5)’, in M. Ledentu (ed.), Parole, Media, Pouvoir dans l’Occident Romain. Hommages offerts au Professeur Guy Achard (Lyons), 147–72. Burman, P. (1720a), M. Fabii Quinctiliani de Institutione oratoria libri duodecim (Leiden). Burman, P. (1720b), M. Fabii Quinctiliani ut ferunt Declamationes XIX maiores…et Calpurnii Flacci Declamationes (Leiden). Bury, R. G. (19322), Symposium (Cambridge; 19091). (p.352) Butler, H. E. (1920–2), The Institutio oratoria of Quintilian, 4 vols. (London and Cambridge [Mass.]). Calboli Montefusco, L. (1979), Consulti Fortunatiani Ars rhetorica (Bologna). Calcante, C. and Corsi, S. (1997), Marco Fabio Quintiliano. La formazione dell’oratore, 3 vols. (Milan). Castiglioni, L. (1927), ‘In Pomponium Melam animadversiones criticae’, in Raccolta di scritti in onore di Felice Ramorino (Milan), 101–29.

Castorina, E. (1946), Licinio Calvo (Catania). Citti, F. (2007), ‘La declamazione greca in Seneca il Vecchio’, in L. Calboli Montefusco (ed.), Papers on Rhetoric, vol. 8 (Rome), 57–102. Citti, F. and Pasetti, L. (2015), ‘Declamazione e stilistica’, in M. Lentano (ed.), La declamazione latina. Prospettive a confronto sulla retorica di scuola a Roma antica (Naples), 115–48. Clarke, M. L. (1975), ‘Quintilian on education’, in Dorey (1975), 98–118. Cole, C. N. (1906), ‘Quintilian’s quotations from the Latin poets’, CR 20: 47–51. Colson, F. H. (1924), Marcus Fabius Quintilianus. Institutionis oratoriae liber I (Cambridge). Condorelli, S. (1968), ‘Ecum aedificant. A proposito di una citazione virgiliana in Quintiliano’, Helikon 8: 380–2. Cousin, J. (1931), ‘Problèmes biographiques et littéraires relatifs à Quintilien’, REL 9: 62–76. Cousin, J. (1935), Études sur Quintilien, vol. 1: Contribution à la recherche des sources de l’Institution oratoire (Paris). Cousin, J. (1936), Études sur Quintilien, vol. 2: Vocabulaire grec de la terminologie rhétorique dans l’Institution oratoire (Paris). Cousin, J. (1975), Recherches sur Quintilien (Paris). Cousin, J. (1975–80), Quintilien. Institution oratoire, 7 vols. (Paris). Cunningham, M. P. (1953–4), ‘Some principles of Latin phrasing’, CW 47: 17–22. Davies, J. C. (1968), ‘Molon’s influence on Cicero’, CQ NS 18: 303–14. Dessauer, H. (1898), Die handschriftliche Grundlage der neunzehn grösseren pseudo-quintilianischen Declamationen (Leipzig). Dessauer, H. (1901), ‘De codice rescripto Parisino 7900 A’, RhM 56: 416–22. Dihle, A. (1977), ‘Der Beginn des Attizismus’, A&A 23: 162–77. Dindorf, G. (1829), Aristides, 3 vols. (Leipzig). Dingel, J. (1988), Scholastica materia. Untersuchungen zu den Declamationes minores und der Institutio oratoria Quintilians (Berlin and New York). Dinter, M., Guérin, C., and Martinho, M. (eds.) (2017), Reading Roman Declamation. Calpurnius Flaccus (Berlin and Boston). Dorey, T. A. (ed.) (1965), Cicero (London). Dorey, T. A. (ed.) (1975), Empire and Aftermath. Silver Latin II (London). Dörner, F. K. and Goell, T. (1963), Arsameia am Nymphaios (Berlin). Douglas, A. E. (1966), M. Tulli Ciceronis Brutus (Oxford). Dover, K. J. (1974), Greek Popular Morality in the Time of Plato and Aristotle (Oxford). Eckstein, F. A. (1871), Nomenclator philologorum (Leipzig). Fairweather, J. (1981), Seneca the Elder (Cambridge). Flashar, H. (ed.) (1979), Le classicisme à Rome aux Iers siècles avant et après J.-C. (Vandoeuvres and Geneva). (p.353) Forbiger, A. (1872–5), P. Vergili Maronis Opera, 4 vols. (Leipzig). Förster, R. and Richtsteig, E. (1929), Choricii Gazaei Opera (Leipzig). Fraenkel, E. (1968), Leseproben aus Reden Ciceros und Catos (Rome). Frazer, J. G. (1921), Apollodorus. The Library, 2 vols. (London and Cambridge [Mass.]). Froment, T. (1880), ‘L’éloquence des délateurs’, Annales de la Faculté des Lettres de Bordeaux 2: 35–57.

Fuhr, K. (1906), ‘

’, BPhW 23: 733.

Gabba, E. (1957), ‘Note sulla polemica anticiceroniana di Asinio Pollione’, RSI 69: 317–39. Gelzer, T. (1979), ‘Klassizismus, Attizismus und Asianismus’, in Flashar (1979), 1–41. Gertz, M. C. (1888), ‘Zu Seneca Rhetor’, JKPh 137: 293–8. Glöckner, S. (1901), Quaestiones rhetoricae (Breslau). Glöckner, S. (1910), ‘Aus Sopatros Μεταποιήσεις’, RhM 65: 504–14. Goldberg, S. M. (1999), ‘Appreciating Aper: the defence of modernity in Tacitus’ Dialogus de oratoribus’, CQ NS 49: 224–37. Griffin, M. (1972), ‘The Elder Seneca and Spain’, JRS 62: 1–19. Gronov, J. F. (1665), M. Fab. Quintiliani Declamationes undeviginti. M. Fabii Avi et Calpurnii Flacci Declamationes. Auctoris incerti dialogus De causis corruptae eloquentiae (Leiden and Rotterdam). Gruen, E. S. (1966), ‘Cicero and Licinius Calvus’, HSPh 71: 215–33. Gudeman, A. (19142), P. Cornelii Taciti Dialogus de oratoribus (Leipzig and Berlin; Boston, 18941). Gundel, H. G. (1957), ‘Die klassische Philologie an der Universität Giessen im 20. Jahrhundert’, in LudwigsUniversität, Justus-Liebig-Hochschule 1607–1957. Festschrift zur 350-Jahrfeier (Giessen), 192–221. Gunderson, E. (2003), Declamation, Paternity, and Roman Identity. Authority and the Rhetorical Self (Cambridge). Güngerich, R. (1950), Rev. of Austin (19481), Gnomon 22: 243–7. Gwynn, A. (1926), Roman Education from Cicero to Quintilian (Oxford). Hagendahl, H. (1936), ‘Rhetorica I: In controversias Senecae patris quaestiones’, in Apophoreta Gothoburgensia Vilelmo Lundström oblata (Gothenburg), 283–338. Håkanson, L. (1972), ‘Some critical remarks on Calpurnius Flaccus’, Eranos 70: 59–71. Håkanson, L. (1974a), ‘Some more critical remarks on Calpurnius Flaccus’, Eranos 72: 53–64. Håkanson, L. (1974b), Textkritische Studien zu den grösseren pseudoquintilianischen Deklamationen (Lund). Håkanson, L. (1976), ‘35 Emendationsvorschläge zu den grösseren pseudoquintilianischen Deklamationen’, Eranos 74: 122–38. Håkanson, L. (1978), Calpurnii Flacci Declamationum excerpta (Stuttgart). Håkanson, L. (1982), Declamationes XIX maiores Quintiliano falso ascriptae (Stuttgart). Håkanson, L. (1985), Apuleius. Den gyllene åsnan (Höganäs). Håkanson, L. (1986), ‘Die quintilianischen Deklamationen in der neueren Forschung’, ANRW II.32.4: 2272–306. Håkanson, L. (1989), L. Annaeus Seneca Maior. Oratorum et rhetorum sententiae, divisiones, colores (Leipzig). Håkanson, L. (2014), Unveröffentlichte Schriften, vol. 1: Studien zu den pseudoquintilianischen Declamationes maiores, ed. by B. Santorelli (Berlin and Boston). (p.354) Håkanson, L. (2016), Unveröffentlichte Schriften, vol. 2: Kritischer Kommentar zu Seneca Maior, Controversiae, Buch I, ed. by F. Citti, B. Santorelli, and A. Stramaglia (Berlin and Boston). Håkanson, L. and Winterbottom, M. (2015), ‘Tribunus Marianus’, in L. Del Corso, F. De Vivo, and A. Stramaglia (eds.), Nel segno del testo. Edizioni, materiali e studi per Oronzo Pecere (Florence), 61–90. Halm, K. (1863), Rhetores Latini Minores (Leipzig). Haupt, M. (1870), ‘Varia’, Hermes 4: 335–6. Heath, M. (2002–3), ‘Theon and the history of the progymnasmata’, GRBS 43: 129–60. Heinze, R. (1925), ‘Ciceros Rede Pro Caelio’, Hermes 60: 192–258.

Heldmann, K. (1982), Antike Theorien über Entwicklung und Verfall der Redekunst (Munich). Hinck, H. (1873), Polemonis Declamationes quae exstant duae (Leipzig). Hofmann, J. B. and Szantyr, A. (19722), Lateinische Syntax und Stilistik (Munich; corr. repr. of 19651). Hollis, A. S. (20092), Callimachus. Hecale (Oxford; 19901). Holmes, N. (1997–8), ‘Perinde and proinde’, Glotta 74: 59–75. Hörandner, W. (1981), Der Prosarhythmus in der rhetorischen Literatur der Byzantiner (Vienna). Housman, A. E. (1972), ‘The Thyestes of Varius’, in id., The Classical Papers, ed. by J. Diggle and F. R. D. Goodyear, 3 vols. (Cambridge), III.941–9 (= CQ 11 (1917): 42–8). Innes, D. C. (1985), ‘Theophrastus and the theory of style’, in W. W. Fortenbaugh, P. M. Huby, and A. A. Long (eds.), Theophrastus of Eresus. On His Life and Work (New Brunswick and Oxford), 251–67. Innes, D. (C.) and Winterbottom, M. (1988), Sopatros the Rhetor. Studies in the Text of the Διαίρεσις ζητημάτων (London). Kaster, R. A. (1995), C. Suetonius Tranquillus. De grammaticis et rhetoribus (Oxford). Kaster, R. A. (2016), ‘Making sense of Suetonius in the twelfth century’, in A. Grafton and G. Most (eds.), Canonical Texts and Scholarly Practices. A Global Comparative Approach (Cambridge), 110–35. Keil, H., Mayor, J. E. B., and Rendall, G. H. (1880), Pliny’s Letters. Book 3 (London). Kennedy, G. A. (1963), The Art of Persuasion in Greece (Princeton). Kennedy, G. A. (1969), Quintilian (New York). Kennedy, G. A. (1983), Greek Rhetoric under Christian Emperors (Princeton). Kennedy, G. A. (1998), ‘Quintilian on retirement’, in T. Albaladejo, E. del Río, and J. A. Caballero (eds.), Quintiliano: Historia y Actualidad de la retórica, 3 vols. (Logroño), I.151–8. Kennell, S. A. H. (2000), Magnus Felix Ennodius. A Gentleman of the Church (Ann Arbor). Kenney, E. J. (1974), The Classical Text (Berkeley, Los Angeles, and London). Kiderlin, M. (1889), ‘Zum XI. Buche des Quintilianus’, Philologus 48: 76–87. Kiderlin, M. (1890), ‘Zum XI. Buche des Quintilianus’, Philologus 49: 469–78. Kiderlin, M. (1895), ‘Kritische Bemerkungen zu Quintilians Lehre von dem Gestus und zu C. Sittls Edition derselben’, BBG 31: 226–39. Kowalski, G. (1947), ‘De Phrynes pectore nudato’, Eos 42: 50–62. (p.355) Krapinger, G. and Stramaglia, A. (2015), [Quintilian]. Der Blinde auf der Türschwelle (Größere Deklamationen, 2) (Cassino). Kroll, W. (1907), ‘Randbemerkungen’, RhM 62: 86–101. Kroll, W. (1913), M. Tullii Ciceronis Orator (Berlin). Kühner, R. and Holzweissig, F. (1912), Ausführliche Grammatik der lateinischen Sprache, vol. 1: Elementar-, Formen- und Wortlehre (Hanover). Kurfess, A. (1914), ‘Invektiven der Kaiserzeit. Ein Beitrag zur Geschichte der Invektive’, ZGW 68: 512–24. Lana, I. (1951), Quintiliano, il Sublime e gli Esercizi preparatori di Elio Teone. Ricerca sulle fonti greche di Quintiliano e sull’autore Del Sublime (Turin). Lanfranchi, F. (1938), Il diritto nei retori romani (Milan). Lehnert, G. (1903a), Calpurnii Flacci Declamationes (Leipzig). Lehnert, G. (1903b), ‘Zum Texte der Pseudo-Quintilianischen declamationes maiores’, Philologus 62: 419–44.

Lehnert, G. (1905), Quintiliani quae feruntur Declamationes XIX maiores (Leipzig). Lehnert, G. (1920), ‘Bericht über die Literatur zu den lateinischen Deklamationen bis 1914’, JAW 183: 204–56. Lehnert, G. (1930), ‘Zu Seneca, Suas. 3, 3 (S. 548, 10 Müller)’, PhW 50: 766. Lehnert, G. (1932), ‘Der Rehdigeranus der sogenannten grösseren quintilianischen Deklamationen’, PhW 52: 1051– 4. Lehnert, G. (1935), ‘Bericht über die rhetorische Literatur (1907–1914). III’, JAW 248: 1–147. Lehnert, G. (1938), ‘Zauber und Astrologie in den erhaltenen römischen Deklamationen’, in Volkskundliche Ernte Hugo Hepding dargebracht am 7. September 1938 von seinen Freunden (Giessen), 131–9. Lehnert, G. (1944–55), ‘Griechisch-römische Rhetorik: 1915–1925’, JAW 285: 5–198. Lenz, F. W. (1959), The Aristeides Prolegomena (Leiden). Leo, F. (1960), Ausgewählte kleine Schriften, 2 vols. (Rome). Löfstedt, E. (1936), Vermischte Studien zur lateinischen Sprachkunde und Syntax (Lund). Lossau, M. J. (1964), Untersuchungen zur antiken Demosthenesexegese (Berlin and Zurich). Lucarini, C. M. (2015), ‘I due stili asiani (Cic. Br. 325; P. Artemid.) e l’origine dell’Atticismo letterario’, ZPE 193: 11–24. Malaspina, E. (2001), L. Annaei Senecae De clementia libri duo (Alessandria). Malaspina, E. (2016), L. Annaeus Seneca. De clementia libri duo (Berlin and Boston). Malaspina, E. (forthcoming), ‘A tradição manuscrita do Lucullus de Cícero: do Corpus Leidense a William de Malmesbury e à fortuna no período humanístico’, in M. Martinho and I. Tardin Cardoso (eds.), Ciceronianíssimos, vol. 2 (São Paulo). Marrou, H. I. (19757), Histoire de l’éducation dans l’Antiquité (Paris; 19381). Mesk, J. (1911), ‘Zur Quellenanalyse des plinianischen Panegyricus’, WS 33: 71–100. Meyer, H. (18422), Oratorum Romanorum Fragmenta (Zurich; 18321). Militerni Della Morte, P. (1977), Studi su Cicerone oratore. Struttura della Pro Quinctio e della Pro Sexto Roscio Amerino (Naples). Millar, F. (1964), A Study of Cassius Dio (Oxford). Morr, J. (1926–7), ‘Poseidonios von Rhodos über Dichtung und Redekunst’, WS 45: 47–63. (p.356) Müller, H. J. (1887), L. Annaei Senecae Patris Scripta quae manserunt (Vienna). Munk Olsen, B. (1982–5), L’étude des auteurs classiques latins aux XIe et XIIe siècles, 2 vols. (Paris). Murgia, C. E. (1991), ‘Notes on Quintilian’, CQ NS 41: 183–212. Mynors, R. A. B. (1963), C. Plini Caecili Secundi Epistularum libri decem (Oxford). Mynors, R. A. B., Thomson, R. M., and Winterbottom, M. (1998), William of Malmesbury. Gesta regum Anglorum, vol. 1 (Oxford). Nisbet, R. G. M. (1965), ‘The speeches’, in Dorey (1965), 47–79. Nisbet, R. G. M. and Hubbard, M. (1970), A Commentary on Horace: Odes. Book I (Oxford). Nocchi, F. R. (2017), ‘Assistant professor: ruoli e pratiche didattiche fra antico e moderno’, in L. Mecella and L. Russo (eds.), Scuole e maestri dall’età antica al medioevo (Rome), 45–57. Nolè, M. (2013–14), ‘La

: riflessioni su un genere’, InvLuc 35–6: 227–54.

Norden, E. (1898), Die antike Kunstprosa. Vom VI. Jahrhundert v. Chr. bis in die Zeit der Renaissance, 2 vols. (Leipzig).

Norden, E. (19574), Aeneis Buch VI (Stuttgart; Leipzig, 19031). Nörr, D. (1974), Rechtskritik in der römischen Antike (Munich). Novák, R. (1909), ‘Textkritische Studien zu Seneca Rhetor’, WS 30: 104–35; 243–68. Novák, R. (1913), ‘Kritische Studien zu Seneca Rhetor’, WS 35: 131–42. Novák, R. (1915), ‘Kritische Studien zu Seneca Rhetor’, WS 37: 166–76; 272–88. Obrecht, U. (1698), M. Fabii Quinctiliani Declamationes innumeris locis emendatae (Strasbourg). Odgers, M. M. (1933), ‘Quintilian’s use of earlier literature’, CPh 28: 182–8. Önnerfors, A. (1988), ‘Lennart Håkanson’, Vetenskapssocietetens i Lund årsbok: 159–62. Opitz, R. (1888), Rev. of Müller (1887), JKPh 137: 273–93. Oppliger, C. (2016), ‘Quelques réflexions sur la méthode (ou les méthodes?) du Maître des Petites déclamations’, in A. Casamento, D. van Mal-Maeder, and L. Pasetti (eds.), Le Declamazioni minori dello Pseudo-Quintiliano. Discorsi immaginari tra letteratura e diritto (Berlin and Boston), 103–16. Otto, A. (1888), Rev. of Müller (1887), ZGW 42: 127–33. Owen, S. G. (1924), P. Ovidi Nasonis Tristium liber secundus (Oxford). Pasetti, L. (2008), ‘Filosofia e retorica di scuola nelle Declamazioni Maggiori pseudoquintilianee’, in F. Gasti and E. Romano (eds.), Retorica ed educazione delle élites nell’antica Roma (Pavia), 113–47. Pasetti, L. (2011), [Quintiliano]. Il veleno versato (Declamazioni maggiori, 17) (Cassino). Pasetti, L. (2018), ‘Un tema storico nelle Minores. Per una lettura della decl. 292’, Maia 70: 129–39. Patillon, M. and Bolognesi, G. (1997), Aelius Théon. Progymnasmata (Paris). Peterson, W. (1891), M. Fabi Quintiliani Institutionis oratoriae liber decimus (Oxford). Petzold, P. (1911), De Ciceronis obtrectatoribus et laudatoribus Romanis (Leipzig). Pithou, P. (1580), M. Fab. Quintiliani declamationes, quae ex CCCLXXXVIII supersunt, CXLV: ex vetere exemplari restitutae. Calpurnii Flacci Excerptae X. Rhetorum minorum LI, nunc primum editae, Dialogus de oratoribus, sive De causis corruptae Eloquentiae (Paris). (p.357) Quadlbauer, F. (1958), ‘Die genera dicendi bis Plinius d. J.’, WS 71: 55–111. Rabe, H. (1908), ‘Aus Rhetoren-Handschriften. 5. Des Diakonen und Logotheten Johannes Kommentar zu Hermogenes ’, RhM 63: 127–51. Rabe, H. (1909), ‘Aus Rhetoren-Handschriften. 11. Der Dreimännerkommentar W IV’, RhM 64: 578–89. Rabe, H. (1926), Aphthonii Progymnasmata. Accedunt anonymi Aegyptiaci, Sopatri, aliorum fragmenta. (Leipzig). Reardon, B. P. (1971), Courants littéraires grecs des IIe et IIIe siècles après J.-C. (Paris). Reeve, M. D. (2000), ‘Cuius in usum? Recent and future editing’, JRS 90: 196–206. Reinhardt, T. and Winterbottom, M. (2006), Quintilian. Institutio oratoria. Book 2 (Oxford). Reitzenstein, R. (1908), ‘Zu Quintilians grossen Declamationen’, Hermes 43: 104–19. Reitzenstein, R. (1909), Studien zu Quintilians grösseren Deklamationen (Strasbourg). Reuter, A. (1887), De Quintiliani libro qui fuit de causis corruptae eloquentiae (Breslau). Reynolds, L. D. (1965), The Medieval Tradition of Seneca’s Letters (Oxford). Reynolds, L. D. (ed.) (1983), Texts and Transmission. A Survey of the Latin Classics (Oxford; corr. repr. 1986). Richter, W. (1968), ‘Das Cicerobild der römischen Kaiserzeit’, in G. Radke (ed.), Cicero. Ein Mensch seiner Zeit (Berlin), 161–97.

Ritter, C. (1881), Die quintilianischen Declamationen. Untersuchung über Art und Herkunft derselben (Freiburg and Tübingen). Rohde, E. (1886), ‘Die asianische Rhetorik und die zweite Sophistik’, RhM 41: 170–90; repr. in id., Kleine Schriften, vol. 2 (Tübingen and Leipzig, 1901), 75–97. Rohde, E. (19143), Der griechische Roman und seine Vorläufer (Leipzig; Wiesbaden, 18761). Russell, D. A. (1979), ‘Classicizing rhetoric and criticism: the pseudo-Dionysian Exetasis and Mistakes in Declamation’, in Flashar (1979), 113–34. Russell, D. A. (1981), Criticism in Antiquity (London). Russell, D. A. (1983), Greek Declamation (Cambridge). Russell, D. A. (2001), Quintilian. The Orator’s Education, 5 vols. (Cambridge [Mass.] and London). Russell, D. A. (2006), ‘The initiate and the dream: a lesson in declamation from Sopatros’, Topoi (Lyon) 7: 41–52. Russo, G. (2013), ‘L’uomo che rubò a se stesso. Una declamazione con notazioni didattiche (P.Lond. Lit. 138, coll. III 5–IV 18)’, APF 59: 301–25. Rutherford, I. (1990), Rev. of Innes–Winterbottom (1988), CR NS 40: 20–2. Sandys, J. E. (1903–8), A History of Classical Scholarship, 3 vols. (Cambridge). Santorelli, B. (2017), ‘Metrical and accentual clausulae as evidence for the date and origin of Calpurnius Flaccus’, in Dinter–Guérin–Martinho (2017), 129–40. Schanz, M. and Hosius, C. (19354), Geschichte der römischen Literatur bis zum Gesetzgebungswerk des Kaisers Justinian, vol. 2: Die römische Literatur in der Zeit der Monarchie bis auf Hadrian (Munich; 18901). Schanz, M., Hosius, C., and Krüger, G. (19223), Geschichte der römischen Literatur bis zum Gesetzgebungswerk des Kaisers Justinian, vol. 3: Die Zeit von Hadrian 117 bis auf Constantin 324 (Munich; 18961). (p.358) Schenkeveld, D. M. (1991), Rev. of Innes–Winterbottom (1988), Mnemosyne s. IV 44: 213–16. Schilling, L. (1903), Quaestiones rhetoricae selectae (Leipzig). Schindel, U. (1987), Anonymus Ecksteinii. Scemata Dianoeas quae ad rhetores pertinent (Göttingen). Schindel, U. (2001), Die Rezeption der hellenistischen Theorie der rhetorischen Figuren bei den Römern (Göttingen). Schmid, W. (1887–96), Der Atticismus in seinen Hauptvertretern, 4 vols. (Stuttgart). Schmid, W. (1894), ‘Zur antiken Stillehre aus Anlass von Proklos’ Chrestomathie’, RhM 49: 133–61. Schmid, W. (1897), Der Atticismus in seinen Hauptvertretern. Registerband (Stuttgart). Schmid, W. and Stählin, O. (19616), Geschichte der griechischen Literatur, vol. 2.2: Die nachklassische Periode der griechischen Literatur. Von 100 bis 530 nach Christus (Munich; 19241). Schneider, C. (2013), [Quintilien]. Le tombeau ensorcelé (Grandes déclamations, 10) (Cassino). Schöll, F. (1879), ‘Q. Remmius Palaemon’, RhM 34: 630–2. Schultingh, J., ap. Burman (1720a–b). Shackleton Bailey, D. R. (1969), ‘Emendations of Seneca Rhetor’, CQ NS 19: 320–9. Shackleton Bailey, D. R. (1976), Rev. of Håkanson (1974b), AJPh 97: 73–9. Shackleton Bailey, D. R. (1983), ‘Notes on Quintilian’, HSPh 87: 217–40. Shackleton Bailey, D. R. (1989a), ‘More on Quintilian’s (?) Shorter Declamations’, HSPh 92: 367–404. Shackleton Bailey, D. R. (1989b), M. Fabii Quintiliani Declamationes minores (Stuttgart). Shackleton Bailey, D. R. (2006), Quintilian. The Lesser Declamations, 2 vols. (Cambridge [Mass.] and London).

Solmsen, F. (1938), ‘Aristotle and Cicero on the orator’s playing on the feelings’, CPh 33: 390–404; repr. in id., Kleine Schriften, vol. 2 (Hildesheim, 1968), 216–30. Spalding, G. L. (1798–1834), M. Fabii Quintiliani De institutione oratoria libri duodecim, 6 vols. (Leipzig). Sprenger, J. (1911), Quaestiones in rhetorum Romanorum declamationes iuridicae (Halle). Stevens, E. B. (1944), ‘Some Attic commonplaces on pity’, AJPh 65: 1–25. Stramaglia, A. (2002), [Quintiliano]. La città che si cibò dei suoi cadaveri (Declamazioni maggiori, 12) (Cassino). Stramaglia, A. (2006), ‘Le Declamationes maiores pseudo-quintilianee: genesi di una raccolta declamatoria e fisionomia della sua trasmissione testuale’, in E. Amato (ed.), Approches de la Troisième Sophistique. Hommages à Jacques Schamp (Brussels), 555–84. Stramaglia, A. (2010), ‘Come si insegnava a declamare? Riflessioni sulle routines scolastiche nell’insegnamento retorico antico’, in L. Del Corso and O. Pecere (eds.), Libri di scuola e pratiche didattiche. Dall’Antichità al Rinascimento. Atti del Convegno Internazionale di Studi (Cassino, 7–10 maggio 2008), 2 vols. (Cassino), I.111–51 (pl. 1–2). Stramaglia, A. (2013), [Quintiliano]. L’astrologo (Declamazioni maggiori, 4) (Cassino). (p.359) Stramaglia, A. (2015), ‘Quint. 9, 2, 85–86 e P.Hamb. II 134, fra tecnica declamatoria e Satzrhythmus’, GLB 20: 133–48. Stramaglia, A. (2016), ‘The Hidden Teacher. “Metarhetoric” in Ps.-Quintilian’s Major Declamations’, in M. Dinter, C. Guérin, and M. Martinho (eds.), Reading Roman Declamation. The Declamations Ascribed to Quintilian (Berlin and Boston), 25–48. Stramaglia, A. (2017), I frammenti delle Declamazioni maggiori pseudo-quintilianee, SIFC S. IV 15: 195–214. Stramaglia, A. (20172), Giovenale, Satire 1, 7, 12, 16. Storia di un poeta (Bologna; corr. repr. of 20081). Stramaglia, A. (2018), ‘Pseudo-Quintilianus, Declamationes maiores, 5: Aeger redemptus,’ PhilolAnt 11: 25–76. Stroh, W. (1975), Taxis und Taktik. Die advokatische Dispositionskunst in Ciceros Gerichtsreden (Stuttgart). Stroux, J. (1949), ‘Summum ius summa iniuria. Ein Kapitel aus der Geschichte der interpretatio iuris’, in id., Römische Rechtswissenschaft und Rhetorik (Potsdam), 7–66 (= Festschrift Paul Speiser-Sarasin zum 80. Geburtstag am 16. Oktober 1926 (Leipzig and Berlin, 1926), 115–56). Summers, W. C. (1910), Select Letters of Seneca (London). Summers, W. C. (1911), ‘Notes on the controversiae of the Elder Seneca’, CQ 5: 17–23. Sussman, L. A. (1978), The Elder Seneca (Leiden). Sussman, L. A. (1987), The Major Declamations Ascribed to Quintilian. A Translation (Frankfurt am Main, Bern, and New York). Sussman, L. A. (1994), The Declamations of Calpurnius Flaccus (Leiden, New York, and Cologne). Syme, R. (1958), Tacitus, 2 vols. (Oxford). Syme, R. (1978), History in Ovid (Oxford). Testroet Ouellette, H. (1982), William of Malmesbury. Polyhistor (Binghamton [NY]). Thomas, E. (1899), Schedae criticae novae in Senecam Rhetorem (Leipzig). Thomson, J. O. (1948), History of Ancient Geography (Cambridge). Thomson, R. M. (20032), William of Malmesbury (Woodbridge; 19871). Throop, G. R. (1913), ‘Ancient literary detractors of Cicero’, WUS 1.2.1: 19–41. Traina, A. (19874), Lo stile ‘drammatico’ del filosofo Seneca (Bologna; 19741). Vassis, S. (1906–7), ‘De locis quibusdam Quintiliani declamationum quae feruntur’, EEAth 3: 113–16.

Verbraken, P.-P. (1984), ‘Le sermon LXXXVIII de saint Augustin sur la guérison des deux aveugles de Jéricho’, RBen 94: 71–101. Vickers, B. (1970), Classical Rhetoric in English Poetry (London). Vogel, F. (1885), Magni Felicis Ennodi Opera (Berlin). Waldis, J. (1920), Sprache und Stil der großen griechischen Inschrift vom Nemrud-Dagh in Kommagene (Heidelberg). Walter, F. (1935–6), ‘Zu beiden Seneca: Seneca Rhetor’, Philologus 90: 515–16. Walz, C. (1832–6), Rhetores Graeci, 9 vols. (Stuttgart et al.). Watt, W. S. (1988), ‘Notes on Quintilian’, GB 15: 139–60. Watt, W. S. (1993), ‘Notes on Quintilian’, in H. D. Jocelyn and H. Hurt (eds.), Tria lustra. Essays and Notes Presented to John Pinsent (Liverpool), 315–20. Watt, W. S. (1996), ‘Ten notes on Calpurnius Flaccus, Declamationum excerpta’, Eranos 94: 123–7. (p.360) Weissenberger, M. (2010), Sopatri Quaestionum divisio—Sopatros: Streitfälle. Gliederung und Ausarbeitung kontroverser Reden (Würzburg). Westman, R. (1980), M. Tulli Ciceronis Orator (Leipzig). Wilamowitz-Möllendorff, U. von (1900), ‘Asianismus und Atticismus’, Hermes 35: 1–52; repr. in id., Kleine Schriften, vol. 3 (Berlin, 1969), 223–73. Wiles, J. J. (1922), ‘Emendations of Quintilian and the Elder Seneca’, CR 36: 69. Williams, R. D. (1960), P. Vergili Maronis Aeneidos liber quintus (Oxford). Winterbottom, M. (1962a), ‘Almeloveen’s manuscript of Quintilian’, CR NS 12 (1962a), 121–2. Winterbottom, M. (1962b), ‘The textual tradition of Quintilian 10.1.46 f.’, CQ NS 12 (1962), 169–75. Winterbottom, M. (1964a), ‘More about Almeloveen’, CR NS 14: 243. Winterbottom, M. (1964b), ‘Quintilian and the vir bonus’, JRS 54: 90–7. Winterbottom, M. (1964c), ‘Some problems in Quintilian Book Two’, Philologus 108: 119–27. Winterbottom, M. (1967a), ‘The beginning of Quintilian’s Institutio’, CQ NS 17: 123–7. Winterbottom, M. (1967b), ‘Fifteenth-century manuscripts of Quintilian’, CQ NS 17: 339–69. Winterbottom, M. (1967c), ‘Quintilian and Boethius’, BICS 14: 83. Winterbottom, M. (1970a), M. Fabi Quintiliani Institutionis oratoriae libri duodecim (Oxford). Winterbottom, M. (1970b), Problems in Quintilian (London). Winterbottom, M. (1974a), ‘Problems in the Elder Seneca’, BICS 21: 20–42. Winterbottom, M. (1974b), The Elder Seneca. Declamations, 2 vols. (Cambridge [Mass.] and London). Winterbottom, M. (1975), ‘Quintilian and rhetoric’, in Dorey (1975), 79–97. Winterbottom, M. (1977a), ‘Aldhelm’s prose style and its origins’, ASE 6: 39–76. Winterbottom, M. (1977b), Rev. of Cousin (1975–80).I and Cousin (1975), Gnomon 49: 574–9. Winterbottom, M. (1978), Rev. of Cousin (1975–80).II–III, Gnomon 50: 685–7. Winterbottom, M. (1979), Rev. of Cousin (1975–80).IV, Gnomon 51: 388–9. Winterbottom, M. (1980a), Rev. of Cousin (1975–80).V–VI, Gnomon 52: 785–6. Winterbottom, M. (1980b), Roman Declamation. Extracts Edited with Commentary (Bristol). Winterbottom, M. (1981), Rev. of Cousin (1975–80).VII, Gnomon 53: 197–9.

Winterbottom, M. (1982), ‘Cicero and the Silver Age’, in W. Ludwig (ed.), Éloquence et rhétorique chez Cicéron (Vandoeuvres and Geneva), 237–66 (‘Discussion’, 267–74). Winterbottom, M. (1983a), ‘Declamation, Greek and Latin’, in A. Ceresa-Gastaldo (ed.), Ars rhetorica antica e nuova (Genoa), 57–76. Winterbottom, M. (1983b), ‘Quintilian and declamation’, in Hommages à Jean Cousin (Paris), 225–35. Winterbottom, M. (1984), The Minor Declamations Ascribed to Quintilian (Berlin and New York). Winterbottom, M. (1985), ‘The Roman orator and his education’, Akroterion 30: 53–7. Winterbottom, M. (1993), ‘The transmission of Cicero’s De officiis’, CQ NS 43: 215–42. Winterbottom, M. (1994), M. Tulli Ciceronis De officiis (Oxford). Winterbottom, M. (1995), Rev. of Sussman (1994), CR NS 45: 40–2. (p.361) Winterbottom, M. (1997), ‘Introduzione’, in Calcante–Corsi (1997), I.5–26. Winterbottom, M. (2001), ‘Returning to Tacitus’ Dialogus’, in C. W. Wooten (ed.), The Orator in Action and Theory in Greece and Rome. Essays in Honor of George A. Kennedy (Leiden, Boston, and Cologne), 137–55. Winterbottom, M. (2005), ‘Approaching the end: Quintilian 12.11’, AClass 48: 175–83. Winterbottom, M. (2006), ‘Quintilian 12.11.11–12’, CQ NS 56: 324–5. Winterbottom, M. (2007), William of Malmesbury. Gesta pontificum Anglorum, vol. 1 (Oxford). Winterbottom, M. (2011), ‘On ancient prose rhythm: the story of the dichoreus’, in D. Obbink and R. Rutherford (eds.), Culture in Pieces. Essays on Ancient Texts in Honour of Peter Parsons (Oxford), 262–76. Winterbottom, M. (2017a), Rev. of Malaspina (2016), Eikasmos 28: 479–82. Winterbottom, M. (2017b), ‘The editors of Calpurnius Flaccus’, in Dinter–Guérin–Martinho (2017), 141–60. Winterbottom, M. (2017c), ‘The pleasures of editing’, RHT NS 12: 393–413. Winterbottom, M. (2017d), ‘The Tribunus Marianus and the development of the cursus’, in P. Chiesa, A. M. Fagnoni, and R. E. Guglielmetti (eds.), Ingenio facilis. Per Giovanni Orlandi (1938–2007) (Florence), 231–47. Winterbottom, M. (2018), ‘Cyrus, On the Differentiation of Issues’, S&T 16: 209–46. Winterbottom, M. and Thomson, R. M. (2002), William of Malmesbury. Saints’ Lives (Oxford). Winterbottom, M. and Thomson, R. M. (2011), Willelmi Meldunensis monachi Liber super explanationem Lamentationum Ieremiae prophetae (CCCM 244) (Turnhout). Wiseman, T. P. (1979), Clio’s Cosmetics. Three Studies in Greco-Roman Literature (Leicester). Wisse, J. (2002), ‘The intellectual background of Cicero’s rhetorical works’, in J. M. May (ed.), Brill’s Companion to Cicero. Oratory and Rhetoric (Leiden, Boston, and Cologne), 331–74. Wisse, J., Winterbottom, M., and Fantham, E. (2008), Marcus Tullius Cicero. De oratore libri III, vol. 5: Book III, 96–230 (Heidelberg). Zieli ski, T. (19082), Cicero im Wandel der Jahrhunderte (Leipzig and Berlin; 18971). (p.362)

(p.363) Index nominum et rerum acting (impersonation) of characters by declaimers 43, 61, 75, 84, 109, 125, 152, 156; see also characterization, ethopoeia advocates: declaimers speaking as 109, 125 ‘granting’ of 292 aemulatio, emulation 10, 76–7, 230 Aeschines: as character in declamation 75 founder of Second Sophistic 141, 244 as rhetor 107 Albucius Silus, C.: admiration for Greek declaimer 104 his colloquialisms 121 philosophizes 245–6, 248 his suicide 92, 335 Alcidamas 245 anaphora: in declamation 116–18, 158, 289 in Hellenistic oratory 108, 149 in a sermo 288 see also asyndetic dicola Antiphon: as orator 114, 118 his Tetralogies 114–15, 144 see also themes Aper, M., in Tacitus’ Dialogus: on Cicero 81 defends modern oratory 4–6, 10, 13, 48, 70 and delators 7, 10, 335 on handbooks 5, 48, 100, 115 seemed to despise education 49 Apollodorus, handbook of 5, 48, 100, 115 apparatus criticus, types of 304 Apuleius: Apologia 46, 154 Florida 152 imitates Cicero 93 Metamorphoses 277 as philosopher 250 Aquillius Regulus, M. 9–14, 44, 50, 58 argumentation: in Decl. min. 113, 115, 121–2 in Gorgias 112–13, 141–2, 145 in Greek declamation 116, 154, 156–7 in Latro 111 as part of a speech 48, 97, 142, 289 pleasure introduced into 86 in Quintilian 121–3 Aristotle 83, 127, 178, 240, 247, 251 Asconius 68, 71–2, 90–1, 126 Asian style, seen in Cicero and elsewhere 80–5, 93–4, 105–7, 146–50 assistants, teaching 221–2, 233 asyndetic dicola 96, 116–18, 123, 155, 158–9; see also variation asyndeton: adjectives in 288, 338

verbs in 259 Atticism 76, 80 Atticists: Greek 105, 108, 151, 153–4 Roman 68, 80, 82, 105, 150–1, 154, 166, 171, 337 Bobbio Scholia on Cicero 71–2, 74, 82 bombast 153, 199 as sign of Asianism 105–6, 116, 148–9 Callimachus 167 Calpurnius Flaccus: edited by Sussman 342–4 editors of 264–82 emended 191–2 Cassius Severus: on Cicero 106, 148 marks change in oratory 4–6, 79, 189 as orator 4, 9–10, 12, 49 Quintilian on 53 in Seneca the Elder 97 Cato the Elder: archaic words used by 54 in Cicero’s De senectute 235, 240 his definition of orator 3, 57–8, 129, 237 Gellius on 69, 85 imitation of 225 in Quintilian 240, 337 characterization, prosopopoeia: and emotion 84 of a father 43, 293 in Greek declamation 43, 116, 152, 156 and Menander 79 in oratory 200 in Theon 232–3 useful for poets, historians 123 see also acting; Choricius; ethopoeia (p.364) Choricius of Gaza 127, 139, 152–4 characterization in 116, 156 introductions to declamations 140 rhythm in 105, 153–4 scholiast on 116, 146 Christians 175, 213, 217 Cicero: rhetorical works: Brut. 85, 93–4, 105–6, 115, 163 De orat. 53, 78, 90, 99, 101, 128, 164, 231, 235, 250 Inv. 47, 85, 87, 93, 99, 109–10, 114, 127, 144, 244 Orator 83, 93, 150, 161–6 speeches: Catil. iv 95–6 pro Caelio 94–5 pro Milone 73–4, 94, 102, 115, 126, 143, 165, 232, 244 pro Murena 112, 115, 177, 247 pro Sulla 114 anthologized by William of Malmesbury 253 commentators on 71 declaims 69, 78, 110, 230, 246 declamatory elements in 81–2, 92–4

on Demetrius of Phaleron 5 disguises his force 168 on emotion 170 epigrams in 45, 81–2 held scholae 251 on impetus 170 quoted by Quintilian alongside Virgil 132 read in school 126 as represented in declamation 67, 76 reputation of 67–70 on rhetoric and philosophy 177, 243–6 and the Silver Age 66–91 on status system 114 on types of Asian oratory 147 used in Quintilian 12.11 235–6, 240 clausulae, see prose rhythm closure 232, 234, 236–8, 288–9; see also sententiae colon, ‘pregnant’ 153, 159 colores, colours 121–2, 139, 287, 331 common good 182 commonplaces, loci communes 331 and Decl. min. 122, 293 in Middle Style 163 and Quintilian 224, 230, 232, 249 in Theon 220, 249 comparison 51, 124, 232 concinnity 106, 147; see also figures; word-play coniectura, conjecture: cases of 111, 114–15, 145 headings of 115 illustrated in declamation 127 in Sopatros 116 conjecture, medieval 18, 252–61, 295–305 contradiction, in Theon 220, 225, 229, 233 controversiae: of Ennodius 205 figured 124–5 lis and controversia 290–1 origin of 5 see also Seneca the Elder ‘correction’, in school 225, 228 corrupt(ion): of advocate 58 of eloquence (oratory) 44, 46, 52, 96, 120, 185, 189 epigram 106 pleasures 50 rhetoric 80 schools 124 uncorrupted: men 187 nature 186 oratory 4 see also Quintilian, De causis corruptae eloquentiae deceit, falsehood 181–2, 189 declamation: Asian 103–9, 146–55 ‘Attic’ 150–5, 159 coloured 122

declamatory language, tone of voice, style 82, 93, 96, 104, 116–18, 122–3, 146, 152–9, 214, 231–2, 269 and drama 157 Francius on 272 Greek and Latin 103–18 ‘historical’ 98 history of 92–3, 107–11, 330 in later life 230 and legal oratory 92–102 public 47, 98, 139, 226, 228 school, procedures of 218–33 as training for forum 75, 79, 96, 111, 120, 124–6, 162, 292 value of disputed 5–6, 46–7, 49–50, 78, 98 delators, delatores 4, 7–14, 58, 315 delivery, oratorical 89, 121, 180, 223 Demetrius of Phaleron: and declamation 5–6, 93, 107 and decline of oratory 47, 189 and Middle Style 163 and the 335 (p.365) Demosthenes: appears in declamation 75, 100, 108, 154 and Cicero 70, 81, 83, 88, 94, 108, 166, 204 and Pliny 83 and Quintilian 73, 81, 83, 108, 167, 204, 229 and rhythm 81 and Roman Atticists 80 his scholiasts 115, 144 in Sopatros 75, 77, 80–1, 108–9, 156 studied in school 75, 108, 113, 225 in Theon 225, 229 and three styles 83 and a topos 118 description, 43, 48, 185, 293 Dio Cassius 67, 118 Dionysius of Halicarnassus 83, 104, 151, 166 divisio, division: in Choricius 118 in Decl. min. 121, 140, 283, 287, 293–4 in Gorgias 112 in handbooks 6 in Seneca the Elder 97, 111, 139 in Sopatros 116, 136–7 two types of 283 Domitian: delators under 8–9, 13, 58 and Quintilian 58, 176, 178 Domitius Afer: in Quintilian 7, 53, 58, 130, 235, 338 in Tacitus 7 editors, principles to be followed by 274 education, Roman: Bonner on 222, 328–9 role of fathers in 219 effeminacy, see virility elaboration, in Theon 220, 229–30, 233 emotion:

aroused by Cicero 83–5, 108, 150 aroused in other Roman oratory 84–5, 87 Cicero and Quintilian on 169 in declamation 43, 96, 109, 112, 123, 150, 226 and Demosthenes 83, 108 felt by Quintilian 13, 54 and figures 183 in history 109, 149–50 and impetus 170, 173 and philosophy 183, 185 played on by orator 169, 175, 182–3 representation of by orators 61 Sopatros on 156–9 and William of Malmesbury 263 see also epilogue; pathos enargeia 84–5, 131, 213, 293; see also description enarratio 73, 115, 221–3, 225, 233 Ennius 70, 118, 130, 133 Ennodius 97, 205–17 Epicureanism 185–6, 246, 248 epideictic, display: in Cicero’s Orator 161–6 and declamation 47, 98, 113, 119, 154 and Gorgias 141 precepts lacking for 48 epigrams, see sententiae epilogue, peroration: in a consular speech 96 in Decl. min. 288–9, 293–4 and emotion 53, 101, 108, 112, 142, 153, 156, 166, 237–8, 293 and equity 84 as part of a speech 48, 141 in Sopatros 116, 156–7 see also recapitulation Eprius Marcellus 4, 7–8, 11, 335 equity, aequitas: and declamation 84, 101–2 in Decl. min. 284 in Ennodius 206 and ius 101, 143 and Quintilian 123, 182 ethopoeia 152, 155, 157; see also acting extemporization 11, 48, 173–4, 188, 230 Fabianus 27, 39, 89, 245, 247, 249–50 figures, schemata 48, 73, 197 in Cicero 82, 165 in Elizabethan schools 231 figured controversia 124–5, 287 figured narratio 142 Gorgianic 82, 162, 165 impossibility of avoiding 231 in Middle Style 163–4 and poetry 130 Quintilian on 183 in Sopatros 159 Fronto: on Cicero 66, 69, 79–80, 85 employs ‘list’ style 154

on impetus 174 on Middle Style 164 and ‘un triomphe du cicéronisme’ 91 Gellius, Aulus 66, 69–70, 81, 85, 203 general and particular 127, 244–6, 248–9, 286; see also hypothesis; thesis gesture 48, 124, 202, 223, 229 good man, vir bonus 3–15, 182 means what? 180–1 (p.366) orator needs to be 3, 13–14, 57, 129, 175, 178, 180–2, 188–9, 236–7 see also orator Gorgias: and declamation 217, 243 his Palamedes 112–14, 118, 141–3, 145 and philosophy 178, 248 in Quintilian 240 his style 147–8, 154, 248 see also figures; Plato grammar 48, 56, 241 grammaticus, grammarian: and historiae 238 school and scope of 72–3, 130, 219, 221–2, 225, 227–8, 230–1 handbooks, rhetorical (artes) 5, 48–9, 74, 87, 100, 125, 127, 140, 144–5, 218, 239 Hegesias: echoed in Sopatros 157 style of 107–9, 146–50, 160 translated into Latin 108 Herennius, see Rhetorica ad Herennium Hermagoras of Temnos 93, 115–16 handbook of 5, 48–9, 100, 116 and hypothesis/thesis 243, 246, 249 and status system 115, 127, 144 Hermogenes: commentaries on 136 and Sopatros 136–9, 348 status system of 71, 116, 127, 283 hexis, 52, 79, 117, 220, 229–30 Himerios 118, 135–6, 153 history, historians: in declamation school 98, 109–10, 127, 152, 221, 227, 232 and epideictic 161–2 fictitious 110 Hellenistic 107, 109, 146–7, 149–51 to be written in old age 218, 236 see also themes Homer 72, 194, 239 compared with Virgil 129–31 in declamation 104, 106, 148–9 homosexuality 179, 187, 343 Horace 131, 178, 235, 246 horme, 170 horses 168–70 hyperbole 51, 106, 189, 238 hypothesis 60–1, 90, 243, 249, 331; see also general; thesis imitation: declamation imitates courts 79 ‘Longinus’ on 175 Quintilian on 45, 48, 75, 225–6, 229 Theon on 225–7 impersonation, see acting

impetus, impulse 11, 50, 120, 168–75 ingenium 6, 9–12, 130, 174 inspiration 11, 174 Isocrates: and Cicero 82, 166 and Gorgias 146, 165 on teaching 226 as writer of epideictic 161–2 Juvenal 9, 11, 231, 245, 264 Latro, M. Porcius: on Cicero 76 despised Greeks 104 extended citation of 111, 142, 331 prose rhythm of 24 pseudepigrapha of 331 laws: declamatory 98–9, 115, 329 praise of 293 Libanius: as declaimer 139–40, 246 style of 118, 152–5 listening, in Theon 220, 226–8, 233 literary criticism: ancient 52, 58, 75 eschewed by Håkanson 278 use of moral terms in 167, 184 Livy 68, 72, 90, 225, 227 ‘Longinus’: on bombast 148 compares Cicero and Demosthenes 70 on Demosthenes 83 on effect of figures 74 and Quintilian 174–5 as rhetorician 52 Lucan 68–9, 85, 331–2 Lucian 69, 143, 152–5, 330 Lysias: apocrypha of 247 and Atticists 80, 83 bad for imitators 225 and Cicero 166 an imitator of 108 madness of poet or orator 11, 58, 120, 170 Maecenas 104, 149, 188 Martial 9 Maternus, in Tacitus’ Dialogus 7, 13–14, 45–6, 335 Menander, playwright 79, 125, 153 Menander, rhetorician 154 (p.367) Messalla, in Tacitus’ Dialogus 5–6, 9, 12, 14, 46, 99 meta-rhetoric 95, 112, 118, 142, 216, 286–7, 289, 293 Minucianos 136–8 moralizing, in Quintilian 176–90; see also good man music, modern 184 myth: examples from 235 in Gorgias 145

in Libanius 227 mythology, ‘fictitious’ 109–10, 114 narration, narrative: in Decl. min. 289 dicola in 159 elaborate 6, 48, 158 emotion in 156 epigrams in 82 exclamations in 157 as exercise 227, 232 figured 142 lacking 112, 141 as part of a speech 48, 142 position of 124 repeated 123 signalled 112, 142 ‘wantoning’ 185 natural and unnatural 186 ‘naturalist’ orators and declaimers 11–12, 50, 77, 120, 171–3 nature and art 237 joined in triad with practice 49, 171 orator: as good man skilled at speaking 3, 13–14, 57, 178 officia oratoris 164 perfect, as portrayed by Cicero 3–4, 46, 83, 161, 177–8 as portrayed by Quintilian 3, 13, 46, 70, 77 in retirement 190, 218, 234–41 sacred name 58, 181 oratory: deliberative 45, 48, 95, 126, 139, 162 forensic (judicial) 45, 48, 95, 124, 162, 165–6, 244 modern 3–15, 46, 52–3, 80 as opus pulcherrimum 240 Republican 5, 72, 84–5, 225 Ovid 52, 130, 174, 231 panegyric 14, 46; see also Pliny the Younger paraphrase: of Homer 106, 148 in Theon 220, 229–30, 233 partition, partitio 73, 112, 141, 143 pathos, pity, miseratio: in Cicero 84, 150 in declamation 23, 42, 84, 108, 150, 153–4, 289 in Demosthenes 83 in epilogues 112, 141 in Hellenistic period 84, 87, 108–9, 149 in Quintilian 327 tears do not last 288 Peripatetics 246 peroration, see epilogue Petrarch 56, 320, 336 Petronius 92, 96, 245 philosophy: and Cicero 78, 90, 163, 177 and declamation 78, 243–51 in old age 236, 239

and Quintilian 176–9, 190 and rhetoric 78, 176, 178–9, 183, 189–90, 243 suspect under Domitian 176 see also themes Philostratus: cites sophists 152, 154, 250 on Isaeus 82 on two Sophistics 107, 243–4, 250 Plato: Gorgias 57 Ion 169 parodies sophists 143 Phaedrus 112, 141, 169 Quintilian on 174 on rhetoric 13, 183, 244 Symposium 113, 143, 152 as teaching material 224 pleasure, voluptas: given by Cicero 94, 164, 186 (corrupt) oratory 57, 85, 124, 185–8 declaimers 50, 121 philosophers 162 nuances of word voluptas 88–9 of old age 235 taken by judges in aequitas 102 vulgar 186 Pliny the Younger: as Ciceronian 45, 50, 66, 70, 80, 85 on delators 9–10, 13, 58 on Demosthenes 83 rhythm in 81 Plotius Gallus 110–11, 147 poet, inspired, mad 174 (p.368) poetry: and grammarians 73, 130, 219, 222, 225, 228, 231 and Quintilian 48, 129–34, 196–7, 229–30 Polemon 152 Polybius 84, 107, 109, 149 prepon, 74, 188 probability 114, 145 proem: in a consular speech 95 in declamation 111–12, 116, 141, 153 in Decl. min. 142, 288–9 extra causam 100 as part of a speech 48, 142 in Sopatros 116, 156–9 in Quintilian 124 progymnasmata: as building blocks for declamation 136, 231, 247, 249 ethopoeia among 152 and general questions 247, 249, 293 influence on Latin literature 232–3 Theon and Quintilian on 218–31 when invented 245 see also Theon; thesis propositio 48, 289 prose rhythm, clausulae:

always mentioned by MW XI, 271 and Asianism 80–1, 105–6, 150, 153 and Cicero 70, 81, 107, 165 clausula spans sentence-break 29 and epideictic 162 Greek 159–60 as guide in textual criticism 21, 24–5, 29, 31, 36, 65, 192–3, 199, 201, 203, 213, 215, 256, 259–60, 271– 2, 317, 319, 343–4, 346, 349–50 and Quintilian 48, 54, 186–7 and Sopatros 159, 349 prosopopoeia, see characterization Quintilian: as advocate 74, 88, 92, 102, 119, 124–5, 169 on character of rhetor 178–9 and Cicero 3, 44–5, 47, 53–5, 68, 70–5, 108, 177–80, 186 De causis corruptae eloquentiae 5–6, 46–7, 107, 189 on declamation 46, 96–9, 119–28 declamation extracts in 122–3 on descriptio 185, 293 as father 182, 219 on impetus 11, 120, 141, 167–8, 171–3 Institutio oratoria: Bk. 10 229–30 early editors of 267–70, 273 edited by Cousin 315–27, 333–8 emended 193–204, 242 in the Middle Ages and later 55–7 on Middle Style 164 on proems 288 on retirement 234–41 and Theon 218–33 on theseis 247 and Vespasian 13 and the vir bonus 3–15 workaholic 183 [Quintilian], Major Declamations (Decl. mai.) 55, 98, 139 apparatus slimmed down 309–10 Decl. 5 and Ennodius 215–16 Decl. 8 edited by Stramaglia 345–6 Håkanson on 313–14 Jerome uses 262 manuscript tradition of 295–310 philosophizing in 245, 248, 251 repetition with variation 117 rival stemmata for 310 William of Malmesbury’s work on 253–63 ‘Quintilian’, Minor Declamations (Decl. min.) 97–8, 121–2, 140, 216, 283–4 and Cicero 45, 77 sermo(nes) 84, 97, 115, 121–2, 140, 283–94 signposting in 113 and status system 115 reading: and declamation 75, 108, 113 of poetry aloud 133 recommended by Quintilian 51, 71, 73, 78–9, 179, 188, 225 in retirement 238, 241

in Theon and Quintilian 220–7, 229–30, 233 reason, ratio 74, 120 and emotion 169–70, 172, 263 recapitulation 112, 141, 143, 237–8 repetition, see variation rhetor(es): age of beginning pupils 219–20 attacked by Messalla 46 ‘corrupting eloquence’ 47, 96, 120 and enarratio 222 falsifying history 98 Greek 80, 85, 94, 106, 109–11 and progymnasmata 219, 227 rhetores Latini 330, 337 (see also Plotius Gallus) satirized by Petronius 96 as teacher of morals 178 and theseis 246 (p.369) rhetoric: Posidonius on 3 and Quintilian 44–59 status and value of 12, 48, 57 Rhetorica ad Herennium 86, 259 on style 93, 164, 231 traces of declamation in 86, 109, 114, 127, 144 on various loci 95, 109, 145 river-imagery 163, 167–70 Sallust 67, 72, 88, 188, 225 ps.-Sallust 67, 118 scribal error, some types of 26, 28, 38, 40, 340 Seneca the Elder: on Cassius Severus 4–5, 49, 97 and Cicero 67, 76 division in 139 edited by Håkanson 339–41 excerpta of 22–4 Greek clausulae in 105, 150 Greek extracts in 42–3, 103–5, 148 on history of declamation 92–3, 110–11 nature of his book 97, 111, 139, 287 on philosophizing 245–6 Sussman on 330–2 textual problems in 16–43 see also divisio; Latro Seneca the Philosopher: on anger 68 on Cicero 69–70, 87 on Greek licence 250 ‘jealous of orators’ 4 Quintilian and 52, 55, 72, 87, 179, 188, 237, 239 on retirement 239 his tragedies 85 sententiae, epigrams: in Cicero 45, 81 favoured by ‘naturalists’ 11, 50, 121, 171 giving closure 113, 123, 142 Quintilian on 121, 288 sermones, see ‘Quintilian’, Minor Declamations

Servius, commentator on Virgil 71, 130–3, 192 ‘signposting’ 100, 112–13 Silver Age: and Cicero 44–5, 66–91 literature of 104, 142, 151, 153 and Quintilian 44, 54–5 Sopatros 135–60 edited by Weissenberger 347–50 nature of his book 103, 116, 283 style of 116–17, 155–9 see also Demosthenes; themes Sophistics: two distinguished by Philostratus 107, 243–4, 250 First (old/‘philosophical’) 107, 243, 245, 248 Asianism prefigured in 146 importance of dispositio in 141 Second 107, 112, 152, 243, 250 see also Aeschines sophists: discussed by Cicero 161–5 and extempore oratory 173 forerunners of declamation 226 some individual sophists 82, 105, 107, 163 as philosophers 250 sophistes, antisophistes 125 style of 163–4 see also Antiphon; Gorgias stasis, status, issues 48–9, 144 in Bobbio scholia 71 and declamation 78, 114, 127, 146, 216 in Sopatros 116, 137–8, 283, 348 Stoics 248 on emotions 181–2, 185 view of oratory 3 see also Cato the Elder style: discussed by Quintilian in moral terms 184–9 Isidorean 117 list style 154–5, 158 of Quintilian 53–5 of Sopatros 155–9 three styles 83, 161, 166 Grand 83, 166 Middle 161–6 Plain 83, 166 ( ) style 154, 157 see also declamation suasoriae: on Alexander the Great 106–7, 148, 249 on augury 248 origin of 5 in real life 126 reserved for the young 96 not in Sopatros 139 see also Seneca the Elder Suetonius 92, 110, 230, 253, 261 Suillius Rufus, P. 4, 7, 12 Sulpicius Victor 60–6, 331 Tacitus: and Cicero 66 on first-century oratory 4, 7–8, 10, 14, 44–5, 49, 79, 115, 185 and Quintilian 189, 337

on stagnation under Trajan 14 (p.370) textual criticism, Gronovius on 266–7 theme(s), declamation: in Antiphon 144 in Cicero 93, 244 involving historical persons 75–6, 98, 107, 110, 122, 152 involving philosophy 78, 248–9 involving slavery 245 odd use of thema 291 phrasing of 24, 34, 63, 65, 212 Quintilian on 47, 123, 127 in Sopatros 138–9 unreal 98, 119, 329 word-play involving 126 Theon, Aelius: on Asian rhetors 150 and Quintilian on progymnasmata 218–31 his ‘Techniques’ 220, 233 Theophrastus 163, 247 thesis, theseis: composed by Cicero 230, 244–6 as progymnasma 90 Quintilian on 232 Sulpicius Victor on 60–1, 331 Theon on 220, 230, 246–7 see also hypothesis time-wasting 183, 238–9, 240–1 tragedy 157, 199 and history 84, 108–9, 149–50 Trajan 9, 14, 58, 88, 328 transitions, smooth 111, 116, 142 translation 230 tropes 48, 231 variation, variatio 30, 53–4, 158–9 Verginius Flavus 125–6, 128 Vespasian 7–8, 13 Vibius Crispus 4, 7–9, 11, 13, 52, 58, 335 Virgil: in Quintilian 72, 129–34 in school 222 virility (masculinity) and effeminacy, of style 53, 105, 121, 172, 184–7 virtues and vices, of style 73, 151, 184, 222 William of Malmesbury: on impulse 175 marks passages with ‘nota’ 262–3 orthography of 255–6 varying treatment of classical texts 261–2 works on Decl. mai. 252–63 word-play, punning: in Asianists 106, 147–8 in ‘Attic’ declamation 153 in Cicero 82, 106, 148 lacking in Sopatros 158 in Quintilian 126, 237 see also themes writing, in school 122, 226–30, 233 written literature 162

(p.371) Index of manuscripts discussed Antwerp, Stadsbibliotheek, 411 17–22, 43 Bamberg, Staatliche Bibliothek, Class. 44 (M.IV.13) 255, 295–7 Bern, Burgerbibliothek, lat. 149 265 Brussels, Bibliothèque Royale, 9581–9595 17–22, 43 Leiden, Universiteitsbibliotheek, Voss. lat. Q. 111 254, 295–7 Montpellier, Bibliothèque Interuniversitaire. Section Médecine, H 126 23, 204, 264 Montpellier, Bibliothèque Interuniversitaire. Section Médecine, H 226 255, 295–7 Oxford, Bodleian Library, Arch. Selden B.36 254 Oxford, Bodleian Library, Rawlinson G.139 253–63 Paris, Bibliothèque Nationale, lat. 1618 295–7 Paris, Bibliothèque Nationale, lat. 7802 254 Vatican, Biblioteca Apostolica Vaticana, lat. 3872 17–22 Zurich, Zentralbibliothek, 288 56