Ekphrasis-imagination-persuasion.pdf

Ekphrasis, Imagination and Persuasion in Ancient Rhetorical Theory and Practice

This page has been left blank intentionally

Ekphrasis, Imagination and Persuasion in Ancient Rhetorical Theory and Practice

Ruth Webb

© Ruth Webb 2009 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, electronic, mechanical, photocopying, recording or otherwise without the prior permission of the publisher. Ruth Webb has asserted her moral right under the Copyright, Designs and Patents Act, 1988, to be identified as the author of this work. Published by Ashgate Publishing Limited Ashgate Publishing Company Wey Court East Suite 420 Union Road 101 Cherry Street Farnham Burlington Surrey, GU9 7PT VT 05401-4405 England USA www.ashgate.com British Library Cataloguing in Publication Data Webb, Ruth, 1963– Ekphrasis, imagination and persuasion in ancient rhetorical theory and practice 1. Ekphrasis 2. Rhetoric, Ancient I. Title 809.9’3357 Library of Congress Cataloging-in-Publication Data Webb, Ruth, 1963– Ekphrasis, imagination and persuasion in ancient rhetorical theory and practice / Ruth Webb. p. cm. Includes bibliographical references. ISBN 978-0-7546-6125-2 (alk. paper) 1. Ekphrasis. 2. Rhetoric, Ancient. I. Title. PN56.E45W43 2009 808’.0481–dc22 09ANSHT 2008035799

ISBN 978-0-7546-6125-2 EISBN 978-0-7546-9330-7

Contents

List of Tables Abbreviations Acknowledgements Preface Introduction

vii ix xi xiii 1

1. The Contexts of Ekphrasis

13

2. Learning Ekphrasis: The Progymnasmata

39

3. The Subjects of Ekphrasis

61

4. Enargeia: Making Absent Things Present

87

5. Phantasia: Memory, Imagination and the Gallery of the Mind

107

6. Ekphrasis and the Art of Persuasion

131

7. The Poetics of Ekphrasis: Fiction, Illusion and Meta-ekphrasis

167

Conclusion

193

Appendix A: Translations Appendix B: Subjects for Ekphrasis Bibliography Index

197 213 215 233


List of Tables

Table 2.1

The subjects of ekphrasis

56

Table 3.1

Comparison of the subjects for ekphrasis with the parts of narration and the subjects for enkōmion in the surviving Progymnasmata

64


Abbreviations

AJP American Journal of Philology BAGB Bulletin de l’Association Guillaume Budé BASP Bulletin of the American Society of Papyrologists BICS Bulletin of the Institute of Classical Studies BMGS Byzantine and Modern Greek Studies CP Classical Philology CQ Classical Quarterly DOP Dumbarton Oaks Papers GRBS Greek, Roman and Byzantine Studies HSCP Harvard Studies in Classical Philology JRS Journal of Roman Studies JWCI Journal of the Warburg and Courtauld Institutes MD Materiali e discussioni per l’analisi dei testi classici Or. Oration PCPS Proceedings of the Cambridge Philological Society REG Revue des études grecques RhMus Rheinisches Museum für Philologie TAPA Transactions of the American Philological Association Walz, Rhetores graeci Rhetores graeci, ed. Christian Walz (9 vols, Stuttgart: Sumptibus J.G. Cottae, 1832–36) ZPE Zeitschrift für Papyrologie und Epigraphik


Acknowledgements

The length of time over which this project has been evolving means that I owe a huge debt to a great many people, from my original supervisors at the Warburg Institute, Jill Kraye and Liz McGrath, and Jean-Michel Massing who first mentioned Philostratos to me, to colleagues at King’s College London, Princeton and the Université Paris X – Nanterre. The following people have made particular contributions through invitations to contribute to conferences or to joint publications, by reading various drafts or simply by being willing to discuss various aspects of the subject. They are, in alphabetical order, Michael Baxandall, Susanna Braund, Averil Cameron, Alejandro Coroleu, Sandrine Dubel, Jaś Elsner, Christopher Gill, Françoise Graziani, Liz James, Bob Kaster, Mario Klarer, Margaret Mullett, Laurent Pernot, Stéphane Rolet, Charlotte Roueché, Agnès Rouveret, Suzanne Saïd, John Smedley, Oliver Taplin, Philip Weller, Barbara Zeitler and Froma Zeitlin.


Preface

My work on ekphrasis in the Greek rhetorical tradition began with the research for my Ph.D. thesis, ‘The Transmission of the Eikones of Philostratos and the Development of Ekphrasis from Late Antiquity to the Renaissance’ (Warburg Institute, London, 1992). The thesis itself contained a brief overview of the ancient theories of ekphrasis and enargeia as found in the Progymnasmata and other rhetorical treatises and has provided the core of the present book. The material on the subject contained in the thesis has been greatly expanded: the treatment is in greater depth and more sources are discussed. The plan of this book has undergone several permutations over the years. In the end, I have opted for a smaller chronological range (first to fifth centuries CE) and to restrict the study mostly to the rhetorical handbooks. Regrettably, this has reduced the space available for the discussion of Christian texts and of the later, Byzantine, material. There is also less analysis of examples of ekphrasis than I had at first planned. However, the elucidation of the main sources for the theory of ekphrasis and enargeia, many of which are neither well known nor easily approachable, seemed to me to be the priority. I hope to be able to fill some of the lacunae I have identified at some point in the future. A word about the transliteration of the Greek: I have tried to reproduce as far as possible the Greek spellings (‘-os’ rather than ‘-us’, ‘ai’ rather than ‘ae’ or ‘e’, ‘k’ rather than ‘c’) but have kept the more familiar forms of well-known names, such as Thucydides and Achilles. The result, as usual, is not perfect. Ruth Webb


Introduction

‘A speech that brings the subject matter vividly before the eyes’

This is the definition of ekphrasis taught to students in the Greek schools of the Roman Empire as they began their studies of rhetoric. It is a very different definition from the one which has become familiar in modern literary criticism for, however ekphrasis is defined in modern critical discourse, it is usually seen as a text or textual fragment that engages with the visual arts. Over the last few decades, ekphrasis has been defined as ‘the poetic description of a pictorial or sculptural work of art’, ‘the verbal representation of visual representation’ or ‘words about an image’. For all their variety, these definitions place a central importance on a certain type of referent: the visual arts (a category which sometimes includes and sometimes excludes buildings and monuments). But this was not its ancient sense. This book is an exploration of the range of meaning of the term as it was used in antiquity. It is a positive attempt to understand and explain a set of ideas about language and its impact on the listener that are expressed in the rhetorical handbooks of the first centuries CE, when Greek rhetoric reached a height of sophistication under the Roman Empire. Ideally it ought to be possible to start straight away with this analysis, but the popularity of the modern definition of the term makes it necessary to include some preliminary remarks about what is and what is not included in this book and why. What This Book Is Not About This is not a book about word and image, in the sense of ‘words about pictures’. There was indisputably a strong tradition of describing real or imaginary works of art in oratory, historiography, epigram, epic and other poetry. But there is no evidence that these were considered to form a �� Leo Spitzer, ‘The “Ode �� on a Grecian Urn”, or content vs. metagrammar’, in Essays on English and American Literature, ed. Anna Hatcher (Princeton, 1962); J.A.W. Heffernan, Museum of Words (Chicago, 1993), p. 3; Shadi Bartsch and Jaś Elsner, ‘Introduction: eight ways of looking at an ekphrasis’, CP, 102 (2007): i. �� See, in particular, Jaś Elsner, ‘The genres of Ekphrasis’, in Elsner (ed.), The Verbal and the Visual: Cultures of Ekphrasis in Antiquity = Ramus, 31 (2002): 1–18.

Ekphrasis, Imagination and Persuasion

single genre, or that that genre had a name, still less that that name would have been ‘ekphrasis’. Painting, sculpture and architecture certainly were among the subjects of ekphrasis as it was conceived and defined in antiquity: the Progymnasmata – the elementary exercises in rhetoric which contain the first definitions of ekphrasis – mention the Shield of Achilles in Iliad, 18 as an example and contain advice on describing sculptures, paintings and buildings. Outside these elementary exercises, the Younger Philostratos refers to his grandfather’s Eikones as ‘ekphraseis of works of graphic art’ and many descriptions of such subjects seem clearly to fit the ancient definition by describing their subjects in such vivid detail that the reader does seem to ‘see’ them. Such subjects certainly could be evoked in ekphrasis, but they were not its defining feature. I have argued elsewhere that the existence of this intermediate category of ekphraseis (in the ancient sense) of works of art and architecture (like Philostratos’ Eikones or Paul the Silentiary’s verse ekphrasis of Hagia Sophia) provided part of the impetus towards the modern definition as scholars in the late nineteenth and early twentieth centuries focused attention on this particular group of texts which then came to stand for the whole category of ekphrasis. But at no point in antiquity (or Byzantium) was ekphrasis confined to a single category of subject matter, nor can every text about images be claimed as ekphrasis in the ancient sense. There are, for example, many epigrams about sculptures which do not seek to ‘bring the subject matter before the eyes’. The examples recently analyzed by Simon Goldhill, for example, consider the act of viewing and meditate on naturalism, but their function as comments on the act of viewing is different from the central function of ekphrasis: making the listener ‘see’ the subject in their mind’s eye. An ekphrasis may itself constitute a commentary on the act of viewing, but this common feature is not central to the definition of ekphrasis that interests me here. So, the epigrams, like certain passages of Pliny or Pausanias, while very relevant to understanding constructions of viewing in antiquity, are at most only tangentially relevant to the rhetorical practices that are the subject of this book. For that reason they are not included. �� As Graham Zanker, ‘New light on the �� “ekphrastic epigram”’, ZPE, 143 (2003): 59–62 points out, the term ‘ekphrastic epigram’ is a modern coinage. �� Philostratos the Younger, Eikones, Proem, 2. �� See Ruth Webb, ‘Ekphrasis ancient and modern: the invention of a genre’, Word and Image, 15: 7–18. �� Simon Goldhill, ‘The naïve and knowing eye: ecphrasis and the culture of viewing in the Hellenistic world’, in Simon Goldhill and Robin Osborne (eds), Art and Text in Ancient Greek Culture (Cambridge: Cambridge University Press, 1994), pp. 197–223; ‘What is ekphrasis for?’, CP, 102 (2007): 16–19.

introduction

This study focuses on the rhetorical theory and practice of ekphrasis for the simple reason that it is in the rhetoricians’ schools that ekphrasis was defined, taught and practised and it is therefore in the domain of rhetoric that we can find a substantial explanation of what ekphrasis was, how it functioned and what its purpose was. This rhetorical technique of ekphrasis can also be found in poetry, of course, but to study poetic examples in detail would require a separate book. Many of the poetic descriptions of works of art – the Shield of Achilles, the Hesiodic Shield of Herakles, Moschos’ Europa, Catullus 64, Virgil’s Shield of Aeneas – do fulfil the basic requirement of ‘placing before the eyes’ and seem to rival the visual arts, as ekphrasis should. It would therefore have been possible to append a chapter to this book examining these usual suspects from the perspective of rhetorical ekphrasis. But there are many reasons why I decided against this. Firstly, these passages have already been abundantly and fruitfully analyzed by others and I would have very little to add to the existing studies. Secondly, the inclusion of a chapter devoted to one type of subject matter would have created a serious imbalance in the book as a whole. Since my analysis of the rhetorical sense of ekphrasis stresses the inclusiveness of the term, such a final chapter ought to analyze examples of poetic ekphrasis of all types of subject matter: battles, people, animals, landscapes. Needless to say, this is a gargantuan task that is far better undertaken in a series of separate studies of authors, genres or categories of subject. What This Book Is About This book is an exploration of a particular phenomenon in ancient rhetorical theory and practice: the use of language to try to make an audience imagine a scene. It mines the rhetorical handbooks of the first centuries CE in order to clarify as far as possible how a particular phenomenon was understood and taught and how it fitted into the wider system of rhetorical theory and practice. This means first of all analyzing the (often frustrating) definitions of ekphrasis in the Progymnasmata and supplementing their laconic comments by reference to more forthcoming sources such as the first-century Roman rhetorician Quintilian, who

�� See, to name but three examples, Alessandro Perutelli, ‘L’inversione speculare: per una retorica dell’ekphrasis’, MD, 1 (1978): 87–98; Andrew Laird, ‘Sounding out ekphrasis: art and text in Catullus 64’, JRS, 83 (1993): 18–30; Froma Zeitlin, ‘The artful eye: vision, ekphrasis and spectacle in Euripidean theatre’, in Simon Goldhill and Robin Osborne (eds), Art and Text in Ancient Greek Culture (Cambridge, 1994), pp. 138–96. �� See, for example, Jean-Pierre Aygon, Pictor in fabula: l’ecphrasis-descriptio dans les tragédies de Sénèque (Brussels, 2004) and Janice Hewlett Koelb, The Poetics of Description: Imagined Places in European Literature (New York, 2006).


supplements rhetorical precepts with accounts of his first-hand experience in the courts. From this study there emerges a clear set of ideas about the word’s ability to summon up images in the listener’s mind and about the ways in which an orator might need to make use of this technique in very specific ways to bolster his case. The Progymnasmata, with the help of Quintilian’s insights, help explain the uses of ekphrasis and similar techniques discussed in the more advanced Greek rhetorical manuals on epideictic (occasional speeches) and declamation (judicial and deliberative speeches treating fictive cases). The use of ekphrasis in epideictic is familiar, and Menander Rhetor’s manuals on the topic are relatively accessible in their presentation of the topic. But the Greek technical handbooks on declamation are a very different matter. Fortunately, recent studies and translations by Donald Russell, Malcolm Heath, Michel Patillon, George Kennedy and others have opened up this highly technical and specialized field, making it possible to trace the uses of ekphrasis in declamation. Looking at ekphrasis in the context of the rhetorical handbooks is revealing. First of all, this approach underlines the close interconnections between Quintilian and the Greek sources on rhetoric. It also shows that the use of ekphrasis was certainly not confined to rhetoric, all our sources make clear that it was used in history and poetry too. But, in the period we are dealing with, rhetoric enjoyed great prestige; it was an active practice studied by a large proportion of male members of the elite and was the focus of much of their intellectual energy. In the case of rhetoric, we have evidence for the methods of training in the handbooks while the surviving examples of declamation and epideictic speeches show the direct results of this training whose indirect results are also evident in other sorts of composition. It is also important to specify the type of relationship between rhetorical theory and rhetorical practice that is assumed in this book. The interest of the handbooks does not lie simply in the provision of a schema against which finished compositions can be measured, identified and judged.10 Instead, the rhetorical manuals reveal to the modern reader the rich network of ideas and assumptions that underlay the composition and

�� See, in particular, Donald R. Russell, Greek Declamation (Cambridge, 1983); Lucia Calboli Montefusco, La dottrina degli status nella retorica greca e romana (Hildesheim, 1986); Michel Patillon, La Théorie du discours chez Hermogène le Rhéteur: essai sur les structures linguistiques de la rhétorique ancienne (Paris, 1988) and his translation and commentary of Hermogenes: Hermogène: l’art rhétorique (Paris, 1997); Malcolm Heath, Hermogenes on Issues: Strategies of Argument in Later Greek Rhetoric (Oxford, 1995) and George A. Kennedy, Hermogenes, Invention and Method: Two Rhetorical Treatises from the Hermogenic Corpus Translated with Introduction and Notes (Atlanta, 2005). 10 �� For example, Francis Cairns, Generic Composition in Greek and Roman Poetry (Edinburgh, 1972) represents a vitally important development in the study of ancient

introduction

reception of ancient texts. These assumptions are often very different from our own, as might be expected when we take into consideration the fact that the culture of the Imperial period was still very much an oral culture, despite the importance of the written word. The phenomenon of ekphrasis that emerges from this study belongs to a conception of the word as a force acting on the listener, a conception that is familiar from Gorgias’ Enkomion of Helen but which clearly continued to be active throughout antiquity and beyond, into Byzantium. An investigation of ekphrasis in this sense also reveals some of the energies that dwell within the texts that, to us, are black words lying still on the white page but which, to the ancient reader, were alive with rich visual and emotional effects. The nature of ekphrasis, its defining quality of enargeia (or ‘vividness’), and the role of the imagination in both mean that this is almost as much a study of ancient psychology as of rhetoric. The study of the ancient definition of ekphrasis is therefore far from being the restrictive move that it is sometimes claimed (whether explicitly or implicitly) to be. It is, I hope, a positive contribution which opens up new perspectives on the rhetorical culture of the Imperial period and on the attitudes to language and verbal representation that were current at that time. The aim of focusing on the ancient definition is not to close down discussion of the phenomenon of ‘words about images’ either in ancient or modern literature nor to brand certain usages of the word as ‘incorrect’. Rather it is to create a space for the ancient definition and to underline quite how different it and its underlying concepts are to our own ideas about texts and literature. The Modernity of the Modern Definition In considering the difference between ancient and modern ekphrasis it is important to bear in mind exactly how recent the modern definition is. One searches in vain for any unambiguous use of the term to mean ‘description of a work of art’ in any source before the late nineteenth century.11 It did not become current in critical discourse until the second half of the twentieth century and only then was it applied regularly to literature and its reception but tends to be over-prescriptive in its use of rhetorical theory. See the general comments of Goldhill, ‘What is ekphrasis for?’, p. 7. 11 �� Ekphrasis is not used, to my knowledge, in Gotthold Ephraim Lessing’s seminal work on painting and poetry, Laokoon. Paul Friedländer, Johannes von Gaza und Paulus Silentiarius: Kunstbeschreibungen Justinianischer Zeit (Leipzig and Berlin, 1912) uses the term ‘Bildbeschreibung’ [description of a picture] and reserves ekphrasis only for texts that fell into the ancient category. Othmar Schissel von Fleschenberg, ‘Die Technik des Bildeinsatzes’, Philologus, 72 (1913): 83–114 and Jean Seznec, ‘Flaubert and the graphic arts’, JWCI, 8 (1945): 175–90 also discuss the subject without recourse to the term ‘ekphrasis’.


modern literature as well as to Greek and Latin texts. In earlier periods, on the rare occasions when the term was employed in the modern European languages, it retained its ancient range of meaning.12 The modern usage is the result of a series of transformations which took place gradually and unevenly, with the result that it is possible to find scholars using the word in different ways at the same period. Broadly, it is possible to distinguish four types of usage. The first is a simple transliteration from the Greek: ekphrasis (or its equivalents in other European languages) is used in the same contexts and of the same texts as the ancient Greek term. In the second, the term is used of those ancient examples of ekphrasis that happen to describe works of art (and/ or architecture). One example of this second stage is J.D. Denniston’s definition of ekphrasis as ‘the rhetorical description of a work of art, one of the types of progymnasma’, citing Philostratos’ Eikones, the lost Eikones of Nikostratos, Lucian’s de Domo and Kallistratos’ ekphraseis of statues as examples in the first edition of the Oxford Classical Dictionary, published in 1949.13 The third usage includes any ancient texts about art and is exemplified by Glanville Downey’s entry on ‘Ekphrasis’ published ten years later in the Reallexikon für Antike und Christentum, which starts from the Shield of Achilles and includes descriptions of works of art from epic and tragedy. The fourth sense in which the term is used is the far broader ‘words about art’, conceived as encompassing texts of any genre from any culture or period of history (though very often appealing to the Homeric Shield of Achilles as the inescapable ur-ekphrasis). These meanings are genetically related and it is possible to trace the gradual and uneven process of evolution. I have argued elsewhere that interest in texts like Philostratos’ Eikones (ekphraseis that have works of 12 �� The first usage of ‘ecfrasi’ in Italian is a case in point. According to Battaglia’s Grande dizionario della lingua italiana, the term occurs in Gregorio Comanini’s Il Figino of 1591 (ed. Barocchi, p. 310). The passage in question is taken directly from the Latin translation of the Acts of the Council of Nicaea in Concilia omnia, ed. F.L. Sarius (Colonia Agrippina, 1567) vol. 3, p. 94 (Synodi Nicenae Secundae action quarta), which transliterates the term directly from the Greek original. Although the ekphrasis in question is of paintings, it is not necessary to assume that Comanini understood the term as referring to a genre specialized in this type of description or of ‘art criticism’. The first usage in English, in the Edinburgh Review of 1815, is similarly ambiguous. The anonymous author refers only to ‘an ecphrasis of Libanius’ and, as the fourth-century orator and teacher Libanios composed ekphraseis of the whole range of subjects, it is far from sure that he was referring only to his ekphraseis of works of art, as is assumed by Grant F. Scott, ‘The rhetoric of dilation: ekphrasis and ideology’, Word and Image, 7 (1991): 301–10 and The Sculpted Word: Keats, Ekphrasis, and the Visual Arts (Hanover, NH, 1994) in what is otherwise an extremely perceptive analysis. 13 �� Koelb, The Poetics of Description, pp. 2–3, underlines the importance of Denniston’s new definition, which, as she points out, is contradicted by the entry on the Progymnasmata in the same volume.

introduction

art as their subject) led to those particular ekphraseis coming to be seen as representative of ekphrasis tout court.14 I shall retrace some aspects of this process at the end of Chapter 1; here it is sufficient to cite one example that reveals one way in which the process of restriction took place. The origin of Denniston’s definition of ekphrasis as a rhetorical description of a work of art lies in the German Reallexikon des klassischen Altertums published by Friedrich Lübker in 1914 which served as a model for the first Oxford Classical Dictionary.15 Here ekphrasis is defined as ‘rhetorical description, mostly of a painting, one of the Progymnasmata’ (‘rhetorische Beschreibung, zumeist eines Bildes, die man zu den Progymnasmata zählte’) and illustrated by the same examples as in the OCD entry. Lübker may overstate the importance of paintings among the ancient subjects of ekphrasis but he makes clear that these ekphraseis are only part of a larger group. Denniston’s entry, however, leaves out the crucial adverb zumeist, a simple change that turns ekphrasis into an ancient genre specializing in the descriptions of paintings. A seemingly tiny detail of translation in a work considered as authoritative may thus have contributed to the transformation of ekphrasis.16 Ekphrasis and Description If the only difference between the ancient and modern definitions were the presence or absence of a certain category of subject matter, the question would deserve a few lines of discussion at most. And, if the absence of ‘works of art’ meant that ekphrasis was fully assimilable to ‘description’ as commonly understood nowadays, there would also be little more to say. But this is not the case. The most important difference between ekphrasis and the category of ‘descriptions of works of art’ lies not in the categories of subject matter envisaged for each but in the criteria by which the two groups are defined. It is the common type of referent – ‘the work of art’ – that makes it possible to classify an epigram on Myron’s cow alongside the Shield of Achilles and a passage from Pausanias (not to mention Keats’ Ode). Yet in the ancient definition the referent is only of secondary importance; what matters, as we have seen, is the impact on the listener. The ancient and modern categories of ekphrasis are thus formed on

14

�� Webb, ‘Ekphrasis ancient and modern’. �� Friedrich Lübker, Reallexikon des Klassischen Altertums (Leipzig and Berlin, 1914). 16 �� Koelb, The Poetics of Description, p. 2 suggests that Denniston’s entry may have inspired Leo Spitzer’s redefinition of ekphrasis in ‘The “Ode on a Grecian Urn”, or content vs. metagrammar’, Comparative Literature, 7 (1955): 203–25. For further discussion of Spitzer’s article, see Chapter 1. 15


entirely different grounds, and are ultimately incommensurate, belonging as they do to radically different systems. If it seems natural to us now to group descriptions according to subject matter it is probably because description itself is popularly conceived as treating a particular class of referent: static objects or persons assimilated to static objects. (�� The neoclassical critical tradition had, moreover, encouraged the idea that descriptions could and should be classified by their subject matter.)17 �� As summed up by Don Fowler, ‘narrative is about people, description deals with things.’18 This very basic distinction is famously identified by Gérard Genette as one of the main characteristics of the modern conception of literature, and has naturally been refined and questioned by many critics who have pointed out, among other things, the practical difficulties involved in separating narration from description once one begins to analyze a text.19 These �� ideas – the strict division between narration and description and the association of description with static, non-human or dehumanized referents – are absent from the ancient accounts. Instead we find a marked continuity between ekphrasis and narration and explicit statements that the two modes share the same group of referents, as set out in Chapter 3 below. By contrast, the emphasis given in the ancient definitions of ekphrasis to effect, over and above any formal or referential characteristics, is striking: an ekphrasis can be of any length, of any subject matter, composed in verse or prose, using any verbal techniques, as long as it ‘brings its subject before the eyes’ or, as one of the ancient authors says, ‘makes listeners into spectators’. Mere words are credited with the ability to make absent things seem present to the spellbound listeners, to control the contents of the most intimate of faculties, the imagination. So, while the visual arts may be literally absent from this definition of ekphrasis, and from most of the discussions by ancient rhetoricians, the idea of the visual underpins this 17

See Jean-Michel Adam, La Description (Paris, 1993), pp. 32–9. Terms such as chronographia and topographia do occur in ancient rhetorical treatises but, as we shall see, were not central to the treatment of ekphrasis. 18 Don Fowler, ‘Narrate and describe: the problem of ekphrasis’, JRS, 81 (1991), p. 26 (with a useful survey). See also Philippe Hamon, ‘Qu’est-ce qu’une description?’, Poétique, 112 (1972), p. 465; Mieke Bal, Narratology: Introduction to the Theory of Narrative (Toronto, 1985), p. 130: ‘we will … define a description as a textual fragment in which features are attributed to objects’; Georg Lukács, ‘Narrate or describe?’, in Writer and Critic, ed. A. Kahn (London, 1978), p. 127 points to the modernity of this conception of description when he associates it with the objectification of humanity by capitalism. 19 Gérard Genette, ‘Frontières du récit’, in L’Analyse structurale du récit (Paris, 1981), p. 162: ‘L’opposition entre narration et description, d’ailleurs accentuée par la tradition scolaire, est un des traits majeurs de notre conscience littéraire’ [The opposition between narration and description, which, I should add, has been accentuated by traditional methods of teaching, is one of the most important characteristics of our understanding of literature].

introduction

mode of speech which rivals the effects of painting or sculpture, creating virtual images in the listener’s mind. It is these basic underpinnings of ancient ekphrasis that make it of interest as well as distinguishing it from the modern usages of the term. The aim of an investigation of ancient ekphrasis is, far from simply pointing out how different the modern definition is, to set the ancient theory and practice in its own context, one in which the concepts of ‘literature’ and ‘art’ did not have their modern contours and in which language played a different role. In this context, it is difficult to find exact equivalences between modern literary terminology which was developed mainly for the analysis of written texts (predominantly prose fiction and poetry) and ancient categories of thought. ‘Description’ is a case in point. Although it is the nearest equivalent to ancient ekphrasis (‘descriptio’ in Latin) and will be used frequently in this book as a translation, �� its connotations are very different, as is only to be expected of a term that has been defined and discussed with reference to the written word rather than live, oral performance. Outline Chapter 1 sets the scene and, after a brief overview of the cultural context to which the ancient definition and use of ekphrasis belong, explores the importance of the visual imagination in the ancient reception of texts of all kinds. The chapter finishes with a more detailed analysis of the formation of the modern meaning of the term which stresses the very different cultural contexts within which this took place and the different interests that motivated it. My exploration of the ancient conception of ekphrasis begins in Chapter 2 with the definition of the term in the Progymnasmata and continues in Chapter 3 which focuses on the range of subject matter prescribed for ekphrasis. The interest of the subjects lies not only in their variety (from battles to crocodiles, from figures to battlements) but in the way in which the precise categories used serve to tie ekphrasis into the broader rhetorical context. These handbooks represent a stage in the students’ rhetorical training (a point made by Paul Friedländer) and give us a glimpse of a pedagogical practice designed to instil certain habits in students.20 They present the advantage of revealing the types of assumptions about language that were assimilated by the student at a very early stage but, for the same reason, there is a limit to what they can tell us in isolation. As textbooks they lack the background of oral explanation, practice and example which would 20 See Ruth Webb, ‘The Progymnasmata as practice’, in Yun Lee Too (ed.), Education in Greek and Roman Antiquity (Leiden, 2001), pp. 289–316.


10

have complemented them in antiquity. As elementary exercises they do not begin to answer questions such as how and why ekphrasis should be used, or why it should be included in a rhetorical training at all. If, however, one places the Progymnasmata in the wider contexts of rhetorical training in general and of ancient assumptions about language and psychology which were so widespread as to hardly need articulating, a fuller picture emerges. Ekphrasis was a technique used to make the audience feel involved in the subject matter, to make them feel as if they were at the scene of a crime, or that they themselves witnessed the achievements for which an emperor is being praised. Chapters 4 and 5 turn to the first-century Roman rhetorician Quintilian for an explanation of the rhetorical uses of ‘placing before the eyes’ and the psychological background to enargeia and ekphrasis. In particular, ancient conceptions of two closely linked faculties – memory and imagination – are essential to the power of ekphrasis to appeal to the emotions, and also help to explain why rhetoricians could be so confident about the effect vivid language would have on an audience. Chapter 6 analyses some examples of the use of ekphrasis in both declamation and epideictic as prescribed by the more advanced Greek handbooks of the Imperial period that represent the next stage in rhetorical training after the Progymnasmata. This analysis shows first that ekphrasis was indeed conceived as a means of achieving persuasion, of altering the listeners’ perception of the subject in a way that helped the orator to win their assent. Secondly, we see the close interconnections between ekphrasis in the Greek tradition and Quintilian’s comments on enargeia and, finally, ekphrasis in epideictic contexts is shown to have a persuasive function. Rhetoricians stress the ability of the word to create illusions, for obvious reasons. But their discussions are littered with phrases which betray the ‘as if’-ness of both ekphrasis and enargeia and the ambiguous status of the mental images they produce. The audience both sees (metaphorically) and fails to see (literally) the subject matter. Ekphrasis is therefore a powerful type of fiction and is central to the explorations of the language’s ability to create a universe of likeness that Barbara Cassin has identified as typical of the Second Sophistic.21 Chapter 7 therefore explores the fictional nature of ekphrasis by looking at instances outside purely rhetorical texts, in the novel, for example, where authors seem to draw attention to the fictional nature of their ekphraseis. It is also possible to apply the reading of ekphrasis as fiction to properly rhetorical usages. Declamations are themselves fictional speeches, a fact which makes the ekphraseis they contain a fiction within a fiction, engaging the audience in a complex combination of acquiescence to and awareness of the illusion. 21

Barbara Cassin, L’Effet sophistique (Paris, 1995).

introduction

11

Even the uses of ekphrasis in epideictic speeches, referring as they do to the world known to the audience, can be shown to rely to a certain extent on the ultimate failure of the word to fulfil the task of ‘placing before the eyes’. Loss and separation were recurring themes in the epideictic speaker’s repertoire as his speeches marked the departures and arrivals, births and deaths that punctuated the lives of the elite and of their cities. The ekphraseis in epideictic in fact often rely for their effect on the acknowledgement that the word cannot replace the actual presence of a person or of a place. Conclusion It should be clear by now to anyone expecting to read about descriptions of works of art (if they are still reading) that they have open�� e�� d the wrong book. However, I hope that they will continue reading. Descriptions of paintings, sculptures and buildings (even the Shield of Achilles) are discussed as the context requires and there is much else that is relevant to the study of the interaction of text and image. It is also important to stress that this book is not based on the assumption that only ancient categories can be used to analyze texts, simply that ancient categories are worthy of interest in their own right, particularly if we wish to understand more fully the contexts in which the texts we have were composed and some of the ways in which they might have been received and understood by contemporaries. It is ironic that the modern meaning (which after all has been current for only about half a century at the time of writing) should have almost totally eclipsed the ancient meaning.22 The late-twentieth-century fascination with the phenomenon of ‘descriptions of works of art’ has proved extremely fruitful, in particular in encouraging interdisciplinary exchanges between classical scholars and specialists in other periods of literature and, to some extent, between literary scholars and historians of art and archaeologists. However, this achievement has been at the expense of the ancient meaning, which is often acknowledged only to be ignored. The present book aims to fill this lacuna.

22 Page DuBois, ‘Reading the writing on the wall’, CP, 102 (2007): 45 recalls a publisher’s unwillingness to allow the term ‘ekphrasis’ to be included in the title of a book because of its unfamiliarity. Now it can be difficult to publish a book with the word ‘ekphrasis’ in the title that does not focus on descriptions of works of art.


1. The Contexts of Ekphrasis

Interest in ancient art and aesthetics was a vital impetus to the creation of the modern definition of ekphrasis. Another, altogether less positive, factor was the general lack of curiosity in the first part of the twentieth century about the rhetorical culture of the Roman period (particularly the Greek rhetorical culture which could only be seen as a disastrous falling off from the sublime heights of the classical period). It is this disdain that may well have allowed so meticulous a linguist as Denniston to disregard the ancient definition. The results of these combined phenomena can be seen in the vision of both ekphrasis and the rhetoric of the Imperial period in Roland Barthes’ overview of ancient rhetoric, published in 1970. Here Barthes cites ekphrasis as the typical product of an age when, he claims, rhetoric had given up any claim to persuasion and was purely for show. Ekphrasis, defined as a self-contained, detachable fragment, was typical of the type of discourse that resulted – that is to say a loosely connected patchwork of passages. Barthes’ picture derives from a once pervasive view of the Greek rhetorical practice of the Roman period as the decadent pastime of the disenfranchised who, without a proper forum in which to flex their rhetorical muscles, engaged in sterile semblances of debate. The picture offered by Barthes is a significantly updated version that rightly stresses the role of improvisation in the rhetorical performance of the time and the interaction between rhetoric and ‘literature’ in the case of the novel. However, it is difficult, if not impossible, to �� accept the characterization of declamation as a disconnected series of passages after reading theoretical works on the subject such as those by Hermogenes which reveal a highly structured approach in which persuasion was still the main goal.

�� Roland Barthes, �� ‘L’ancienne rhétorique: aide-mémoire’, Communications, 16 (1970):� 183: ‘�� Le discours étant sans but persuasif mais purement ostentatoire, se déstructure, s’atomise en une suite lâche de morceaux brillants, juxtaposés selon un modèle rhapsodique. Le principal de ces morceaux (il bénéficiait d’une très grosse cote) était la descriptio ou ekphrasis. L’ekphrasis est un fragment anthologique, transférable d’un discours à un autre …’�� [Since speeches had no persuasive purpose but were purely a matter of display, they lost all structure and broke down into a loosely connected series of brilliant passages, strung together like a rhapsode’s song. The most important of these passages – it was highly prized – was descriptio or ekphrasis. Ekphrasis is a select fragment, which can be transferred from one speech to another …]. On the idiosyncrasies of Barthes’ overview of ancient rhetoric, see David Cohen, ‘Classical rhetoric and modern theories of discourse’, in Ian Worthington (ed.), Persuasion: Greek Rhetoric in Action (London, 1994), pp. 76–7.

14


The Context of Ancient Ekphrasis The Progymnasmata which offer the definitions of ekphrasis as ‘a speech that brings the subject matter vividly before the eyes’ belong to the first centuries CE. The version by Ailios Theon is usually accepted as the earliest and dated to the first century, while those by a certain Nikolaos are dated to the fifth century. Between lie the third-century version wrongly attributed in antiquity to the famous rhetorician Hermogenes of Tarsos and those of Aphthonios from the fourth century. To the information offered by the Progymnasmata can be added Quintilian’s discussion of enargeia and the advice on the use of ekphrasis in the context of larger speeches to be found in the more advanced rhetorical treatises by Hermogenes (second century), Menander Rhetor (later third century), Sopatros Rhetor (fourth century) and Syrianos (fifth century). All these authors are witnesses to the rich rhetorical culture that flourished in the Greek-speaking areas of the Roman Empire and survived in the Byzantine Middle Ages (to a far greater extent than in the medieval West). Throughout this period the study of rhetoric dominated the education of the elite and mastery both of the Attic dialect and of rhetorical forms of exposition was a prerequisite for many careers, even for acceptance as a male member of the elite, and a central element in certain conceptions of Greekness. For more humble families who could nevertheless afford to educate their sons, a training in rhetoric offered a chance for the talented to improve their social position. This is the picture drawn by the ‘autobiography’ of the second-century Syrian Lucian, who depicts his young self torn between his own desire to study rhetoric (paideia) and his family’s demands that he earn a living as a sculptor. Paideia personified offers fame, fortune and travel to the young Lucian in contrast to a life of toil in the workshop. The type of fame and fortune to which Lucian refers is exemplified in Philostratos’ Lives of the Sophists, a collective portrait of the most famous Greek exponents of the art of rhetoric in the second and early third centuries (among whom Lucian is not counted). The accounts of charismatic star teachers and speakers described by Philostratos, who coined the term ‘Second Sophistic’ to describe the phenomenon, give a vivid impression �� See, for example, Theon, Progymnasmata, 118, l. 7: ἔκφρασίς ἔστι λόγoς περιηγηματικὸς ἐvαργῶς ὕπ’ ὄψιv ἄγωv τὸ δηλoύμεvov. �� Malcolm Heath, �� ‘Theon and the history of the Progymnasmata’, GRBS, 43 (2002/3): 129–60 argues �� for a much later date for Theon, identifying him with the fifth-century rhetorician of the same name. I prefer to retain the earlier date because of the parallels with Quintilian and the unusual use of Hellenistic historians while acknowledging that these are by no means decisive criteria. �� Lucian, The Dream or His Life, 1–13.

The contexts of ekphrasis

15

of the glamour and popularity of rhetorical display at the period: speakers drew large audiences who adulated them but who could also be skilled listeners able to criticize the performances they listened to. Philostratos’ Sophists performed declamations (meletai), fictional speeches that also formed part of the rhetorical training delivered in schools. These meletai were speeches on imaginary cases in which the speaker took on the persona of a character in a situation specially formulated to pose a particular rhetorical problem. Many of the cases were set in the classical Greek past (none post-date the death of Alexander in 323 BCE) and involved characters such as Perikles or Demosthenes in situations more or less loosely based on history. Others were imaginary but involved a stock cast of characters drawn from the world of the classical polis: the young hero, the rich man, the general, the tyrant, the orator. Declamation demanded a certain dramatic talent from its exponents who had to speak in persona (Philostratos mentions Polemo’s habit of leaping up from his chair at the climax of his argument and of stamping on the ground, while Herodes Attikos is said at one point to have had tears in his eyes as he declaimed on a particularly emotive subject). But, above all, it required precise skills of analysis and argumentation and a mastery of presentation and style (all in irreproachable atticizing Greek). It was the structures provided by this training (rather than the lack of them as Barthes claims) that allowed the best declaimers to improvise lengthy and complex speeches. The other principle public activity of Philostratos’ sophists was epideictic oratory: occasional speeches marking significant moments in citizens’ lives or in the life of the city. By the Roman period, the range of occasions for such speeches was vast: they marked the arrivals and departures of dignitaries or even pupils within a school, invitations to governors, weddings, deaths and funerals and festivals. Nor was there a complete absence of occasions for more obviously practical uses of rhetoric: Philostratos mentions several cases where these rhetorical performers and teachers had to use their art in their own defence in court, and city councils – boulai – still provided a forum for debate among the wealthy elite. In the fourth century, when power was concentrated more directly in the person of the emperor, Libanios used his rhetorical skills to try to persuade Theodosios of various changes that should be made in

�� Philostratos, Lives of the Sophists, 537 and 574.

�� See Laurent �� Pernot, La Rhétorique de l’éloge dans le monde gréco-romain (2 vols, Paris�� , 1993) and La Rhétorique dans l’Antiquité (Paris, 2000), pp. 104–7 (on the survival of political rhetoric after the battle of Chaeronea); John Ma, ‘Public �� speech and community in the Euboicus’, in Simon Swain (ed.), Dio Chrysostom: Politics, Letters, and Philosophy (Oxford, 2002)�.


16

the administration of Antioch. Just as importantly, Malcolm Heath has shown how the skills taught in the rhetorical schools could be put to use in actual court cases. The case of Augustine in fourth-century North Africa illustrates the continued importance of a rhetorical education: though not wealthy, Augustine’s family were determined to allow him to develop his talent by sending him to Madauros and then to Carthage in the hope that his studies would lead to a distinguished career as an advocate. His trajectory is very similar to that depicted in Lucian’s Dream and illustrates the uses to which a rhetorical training could be put in an increasingly Christian context. Augustine was certainly not alone. In the Greek East the fourth century saw the continuing importance of rhetorical training as offered by men like Libanios and his rival teachers, many of whose pupils were Christians. The talent of these Greek Christian rhetors of the fourth century – Gregory of Nyssa, his brother Basil of Caesarea and Basil’s friend Gregory Nazianzen, who studied rhetoric with him at Athens – has led to them being identified as part of a ‘Third Sophistic’, a title that emphasizes the continued value and relevance of rhetoric beyond the third century.10 Recent studies of the Second Sophistic have rightly emphasized the social, political and cultural functions of rhetorical performance as�� a means of communicating power and negotiating identity.11 The predominance of classical themes made declamation a means of asserting and exploring Greek identity.12 So, while orators may no longer have been at the forefront of politics, as in classical Athens or Republican Rome, rhetorical performance provided an important forum for the Greek citizens of the Empire to assert their identity, to achieve social status among their peers

�� For one example, see Bernard Schouler, ‘Un enseignant face aux prisons de son temps’, Pallas, 72 (2006): 279–96. �� Malcolm �� Heath, ‘Practical advocacy in Roman Egypt’, in Michael J. Edwards and Christopher Reid (eds), Oratory in Action (Manchester, 2004), pp. 62–82. �� Augustine, Confessions, II, iii (5) and III, iii (6) – iv (7). 10 �� See, for example, Averil Cameron, Christianity and the Rhetoric of Empire: The Development of Christian Discourse (Berkeley, 1991);�� Peter Brown, Power and Persuasion in Late Antiquity: Towards a Christian Empire (Madison, 1992); �� Eugenio Amato (ed.), Approches de la Troisième Sophistique: hommages à Jacques Schamp (Brussels, 2006). 11 See, for example, Maud Gleason, Making Men: Sophists and Self-presentation in Ancient Rome (Princeton, 1995)�� ; Thomas Schmitz, �� Bildung und Macht: Zur sozialen und politischen Funktion der zweiten Sophistik in der griechischen Welt der Kaiserzeit (Munich,�� 1997); Tim Whitmarsh, Greek Literature and the Roman Empire (Oxford, 2001). 12 �� See the seminal work of Ewen L. Bowie, ‘Greeks and their past in the Second Sophistic’, in Moses I. Finley (ed.), Studies in Ancient Society (London, 1974), pp. 116–209 and Paolo Desideri, ‘Filostrato: la comtemporaneità del passato greco’, in Fernando Gascó and Emma Falque (eds), Pasado renacido: Uso y abuso de la tradición clásica (Seville, 1992), pp. 55–70.


17

and their contemporaries and was one of the principal media in which relationships with Rome and the representatives of the Empire were constructed.13 It is equally important not to lose sight of the more technical aspects of the art of rhetoric and its continued utility as an intellectual training with many applications. Malcolm Heath, for example, has recently stressed the value of the rhetorical education offered in the schools of the Imperial period and the very practical considerations that ensured its survival.14 He has also shown the continuing vitality of the rhetorical tradition beyond the second- and early-third-century period portrayed by Philostratos. The Progymnasmata textbooks belong to this long history of rhetoric, spanning as they do the first five centuries CE, and showing the continued processes of adaptation and reflection that took place. Rhetoric: Theory and Practice The principle sources for the rhetorical conception of ekphrasis, the Progymnasmata, consist primarily of a set of definitions and instructions for the various exercises, of which ekphrasis was one. The value of these exercises for us lies precisely in their elementary status. As the gateway through which every rhetorically educated person passed (and the final stage in the education of those who could not find the time or the money to achieve a full rhetorical training), they reveal assumptions about language and ways of reading exemplary classical authors which were inculcated at an early age.15 In particular, the Progymnasmata represented a process of transition from reading to speaking, the moment when the schoolboy, whether in Egypt, Syria or Asia Minor, now primed with examples and mastery of the classical Attic idiom still used in high-level discourse, first began to put together his own compositions and to learn to be heard as well as to listen. The most important thing that students learned by working through the Progymnasmata was not rules as such but

13 �� See on this point Laurent �� Pernot, ‘La rhétorique�� de l’empire ou comment la �� rhétorique� grecque a inventé l’empire romain’, Rhetorica, 16 (1998): 131–48. 14 Malcolm Heath, Menander: A Rhetor in Context (Oxford, 2004), esp. pp. 277–331. 15 �� On ancient education and its social implications, see Robert A. Kaster, Guardians of Language: The Grammarian and Society in Late Antiquity (Berkeley, 1988); Teresa �� Morgan, Literate Education in the Hellenistic and Roman Worlds (Cambridge, 1998)�� ; Yun �� Lee Too (ed.), Education in Greek and Roman Antiquity (Leiden�� , 2001). �� Raffaella Cribiore, The School of Libanius in Late Antique Antioch (Princeton and Oxford, 2007), p. 146, suggests that for many students the Progymnasmata would have represented the bulk of the rhetorical training they received.


18

a set of practices and skills that could be put to use in (or transferred to) the composition of full-scale speeches or other types of composition.16 The Progymnasmata were therefore neither abstract nor isolated from the rest of the cultural context. Their purpose was to prepare students for a life of speaking in which the failure to use the socially sanctioned forms at the macro level of speeches or the micro level of grammar and vocabulary could lead to serious embarrassment.17 They were also part of a preparation for a life of critical and agonistic listening. The mention in the definition of ekphrasis of ‘placing the subject before the eyes’ (hup’opsin) is therefore far from theoretical. This was an effect that students were taught to expect to feel for themselves when they read Homer or Thucydides, the most frequently cited sources. But it did not end there. The point of this reading was ultimately to enable students to work the same effect on others as they themselves became active users of rhetoric, first of all in their elementary ekphraseis and later in the full-scale speeches they would compose and perform for their peers in the rhetorical schools and in the wider world. The discussions of ekphrasis in the Progymnasmata and in other rhetorical treatises show how future citizens were taught to participate in its power both as readers or listeners and as speakers. They thus learned to situate themselves as part of a continuous tradition stretching from Homer to the Roman present and to see themselves as involved in a reciprocal process, reproducing the effect that the classical models had on them on their own audiences. Above all, the rhetorical texts that form the basis of this study were part of the living culture of their epoch. The definitions and classifications that they contain were not the result of abstract theorizing in an antique ivory tower but reflected and shaped actual practices. The Progymnasmata in particular, poised as they are between the stage of reading and speaking, also tell us about habits of reading that were deeply ingrained. One particular habit derived from the schools, and also encouraged by the surrounding culture, was a deep identification with texts of the past, their authors and the events they relate, something that can be seen clearly in the ways in which the Homeric poems are appropriated throughout Greek and Roman culture, particularly in the way in which Homer himself is cast as a teacher for the present.18 The rhetoricians’ discussions of ekphrasis, the 16

�� See, for example, Jean Bouffartigue, L’Empereur Julien et la culture de son temps (Paris, 1992), pp. 523–33. 17 �� See, for example, Lucian’s self-defence against an accusation that he misused an Attic term in The Mistaken Critic (Pseudologista). See also Simon Swain, Hellenism and Empire: Language, Classicism and Power in the Greek World, AD 50–250 (Oxford, 1996), pp. �� 43–64. 18 �� See especially ps.-Plutarch, On the Life and Poetry of Homer; Robert Lamberton, Homer the Theologian: Neoplatonist Allegorical Reading and the Growth of the Epic Tradition (Berkeley,


19

type of writing that ‘places before the eyes’, tell us about the imaginative engagement that was expected.�� Young readers were encouraged not to approach texts as distanced artefacts with a purely critical eye, but to engage with them imaginatively, to think themselves into the scenes and to feel as if they were present at the death of Patroklos, the making of the Shield of Achilles, or the Athenian disaster in Sicily during the Peloponnesian War. This openness of the past to those with the educational attainments to read the texts is clear from the uses of classical history in the declamations that were set in the classical past. Whether performed by students in schools or by professional sophists these speeches show a creative attitude towards classical history. For all the reverence paid to the past, history was not a fixed, inalterable object; events could be freely manipulated to serve the needs of the present.19 This openness of history, its availability as matter for manipulation, also underlies the irreverent re-imaginings of the classical tradition by an author such as Lucian or the creative re-presentations of moments from myth and tragedy in the Philostratean Heroikos, as well as the better-known Eikones or Imagines. One particular manifestation of this attitude towards the past is the habit of reading for the sensation of being plunged into the scene or transported back into the moment, which emerges clearly from the rhetoricians’ discussions of ekphrasis and is evident in other sources as well. This habit of responding imaginatively to the written or spoken word forms a vital part of the background to the teaching and use of ekphrasis in rhetorical contexts and deserves to be explored briefly here. Seeing Words Poets and prose writers, orators and historians were all credited with the ability to ‘place a subject before the audience’s eyes’. The many reports of the visual impact of reading texts from classical antiquity make it clear that intense imaginative involvement with the scenes described was a common type of response to texts. As mentioned above, Homer and Thucydides were the examples most often cited in the Progymnasmata and their impact on the ancient reader is confirmed in other sources. These 1986) and Froma Zeitlin, ‘Visions �� and revisions of Homer in the Second Sophistic’, in Simon Goldhill, ed., Being Greek under Rome: Cultural Identity, the Second Sophistic and the Development of Empire (Cambridge, 2001)�� , �� pp. 195–266. 19 �� On the manipulation of the past in the declamations, see Thomas �� Schmitz, ‘Performing history in the Second Sophistic’, in M. Zimmermann (ed.), Geschichtsschreibung und politischer Wandel im 3. Jh. N. Chr. (Stuttgart, 1999);�� Ruth Webb, ‘Fiction, �� mimesis and the performance of the Greek past in the Second Sophistic’, in David Konstan and Suzanne Saïd (eds), Greeks on Greekness: Viewing the Greek Past under the Roman Empire (Cambridge, 2006)�� , pp. 27–46. The most famous ancient statement of this freedom is in Plutarch’s Life of Solon, 27.1.


20

ancient writers often use language that is close to the terminology we find in the technical definitions of ekphrasis which is credited with the ability to place ‘before the eyes’ (hup’opsin) or to make listeners into ‘spectators’ (theatai). Plutarch, for example, writing in the late first or early second century CE echoes the Progymnasmata in his judgement of Thucydides’ ability to make his readers feel as if they were present at the events he describes: Thucydides is always striving for this vividness (enargeia) in his writing, as he eagerly desires to make the listener a spectator, as it were, and to produce in the minds of his readers the feelings of astonishment and consternation which were experienced by those who witnessed the events. ὁ γοῦν Θουκυδίδης ἀεὶ τῷ λόγῳ πρὸς ταύτην ἁµιλλᾶται τὴν ἐνάργειαν, οἷον θεατὴν ποιῆσαι τ�ν ἀκροατὴν καὶ τὰ γιν�μενα περὶ τοὺς ὁρῶντας ἐκπληκτικὰ καὶ ταρακτικὰ πάθη τοῖς ἀναγινώσκουσιν ἐνεργάσασθαι λιχνευ�μενος.

As this suggests, the visual impact is not an end in itself but has the further effect of producing an emotional impact, involving the listener in the events. The same enthusiasm for the visual impact of words is shown by ps.-Longinos in his discussion of the sublime. Citing Herodotos’ account of the journey from Elephantine to Meroe (26.2, cf. 9.6), he exclaims: ‘do you see, my friend, how he takes your soul and leads it through these places, turning hearing into sight (tēn akoēn opsin poiōn)?’ In these contexts, the difference between Thucydides the dispassionate reporter and Herodotos the teller of tall tales is nowhere to be seen. Instead, both are sources of visual experience which transports the reader back to the events described, involving him both imaginatively and emotionally. Xenophon, too, was renowned for his ability to make his readers feel that they were participating in the events of his history. Plutarch attributes the same power to him as to Thucydides, claiming that the long account of the battle of Cunaxa in which the younger Cyrus was killed (Anabasis, 1.8) all but showed the events to the reader, making him feel that it was taking place not in the distant past but before his very eyes, and that the reader (akroatēs) was filled with emotion and shared in the danger.20 Lucian has a fictional speaker in his Eikones attribute the same power to Xenophon’s account of the nobility and fidelity of Pantheia, wife of Abradates (Cyropaideia, 6.4.2–8), exclaiming that he feels as if he could actually see and 20

�� Plutarch, Artaxerxes, 8.1. �� Plutarch uses the same formula here as he did of Thucydides, saying that both authors represent events not as having happened, using the perfect tense (gegen�� ē�� mena), but as happening, using the present (gignomena).


21

all but (mononouchi) hear her arming her husband and sending him out to battle and to his death.21 Here the exoticism of the character – a non-Greek woman acting in a manly fashion – intensifies the effect of the passage and the visual pleasure derived from it, as well as serving as a reminder of the relentlessly masculine point of view of educated response. The underlying erotic interest in the figure of Pantheia emerges clearly from Lucian’s wider context – a debate on the representability of the emperor’s mistress – and from Philostratos’ representation of Pantheia’s suicide over Abradates’ body (Eikones, 2.9). Philostratos’ treatment brings out a further habit of ancient readers – that of imaginatively elaborating upon the scenes presented in texts. Philostratos takes Xenophon as his starting point, citing his source in the opening lines of his description. He points out that Xenophon himself did not describe the appearance of his heroine, but merely her character (ēthos) (2.9.1) and claims that the painter of the picture he is describing filled the gaps, painting Pantheia ‘as he deduced her to be from her soul’. The painting that Philostratos goes on to describe therefore corresponds to a way of reading in which a verbal account of a scene provokes a more detailed visualization, a sensual response. In this case the beauty of Pantheia remains tantalizingly elusive; only her posture as she lies over her husband’s body after her suicide is described in any specific detail. Otherwise her appearance is described in only the most general of terms, implying that for the full experience we must turn to the ever-invisible painting and, by implication, to our imaginations. The best-known and most explicit account of such imaginative supplementation of a text is to be found in the compendious guide to the whole rhetorical curriculum by the first-century CE Roman rhetorician Quintilian, the Institutio oratoria. Citing a passage from the Verrine Orations in which Cicero gave a brief tableau of Verres with his mistress, Quintilian freely admits that the image that arises in his mind when he reads those lines contains details that are not in the text.22 What is more, he presents this response to Cicero’s exemplary enargeia not just as normal but as normative, introducing it with the question ‘is there anyone so incapable of (tam procul abest) forming images of things that he does not seem to see …?’ The passage and its implications for our understanding of ekphrasis will be discussed below, for the moment I would just like to highlight the way in which Quintilian presents his response as the norm: anyone 21 �� Lucian, Eikones, 10. On this passage, see Simon Goldhill, �� ‘The erotic eye: visual stimulation and cultural conflict’, in Goldhill (ed.), Being Greek under Rome: Cultural Identity, the Second Sophistic and the Development of Empire (Cambridge, 2001)�� , p. 189. 22 �� Quintilian, Institutio oratoria, 8.3.64–5. For further discussion of this passage, see Chapter 5.


22

who fails to respond as he does falls short of his readerly ideal. The same confidence that imaginative involvement is the educated norm is shown by the Augustan writer Dionysios of Halikarnassos in his discussion of the enargeia of the fourth-century BCE Attic orator, Lysias. ‘No one’, he claims, ‘can be so clumsy, difficult to please, or slow-witted (skaios, dusarestos kai bradus ton noun) that he will not feel that he can see what is being shown (ta dēloumena) actually happening and that he is conversing with the characters introduced by the orator as if they were present’.23 Like Quintilian, he has only pejorative terms to describe those who fail to respond as he does. In his discussion, acutely analysed by Graham Zanker,24 Dionysios claims that Lysias’ enargeia made the reader feel as if he was in the presence of the characters themselves, even able to converse with them (homilein). This enargeia ‘is a certain power to lead the things shown before the senses’ (dunamis tis hupo tas aisthēseis agousa ta dēloumena).25 This definition of enargeia is very close to the language used to define ekphrasis, which can be literally translated as a speech (logos) which leads the thing shown vividly before the eyes (hup’opsin agon ta dēloumena). The difference lies essentially in the mention of ‘the senses’, where the Progymnasmata mention only sight, the supreme sense. But, as we shall see, even the Progymnasmata definition assumes that senses other than sight can be excited by the workings of ekphrasis. Dionysios casts himself not as a distanced spectator but, like those avid readers of battle narratives, as a participant who could almost enter into the scene himself and converse with the characters. This is partly the result of Lysias’ famed skill at conveying the character of the litigants for whom he wrote through the language he gave them to speak, but Dionysios’ language of showing and of vision makes clear that the impact was felt as above all a visual one. Readers of tragedy, too, felt drawn into the absent spectacle just by reading the words. Several ancient commentators note the vividness of tragic passages.26 Dio Chrysostom in the first century prefaces his discussion of the three versions of Philoktetes by the three great tragedians that were still extant in his day by saying that he ‘was magnificently entertained by the spectacle (thea)’ as he read (Or. 52.3). And ps.-Longinos describes the sheer emotional force of merely reading certain passages from tragedy. Vividness (enargeia) in poetry, he explains, has a shattering 23

�� The Augustan writer, Dionysios of Halikarnassos, Lysias 7. �� Graham �� Zanker, ‘Enargeia in the ancient criticism of poetry’, RhMus, 124 (1981): 297–311. 25 �� Dionysios of Halikarnassos, Lysias 7. 26 �� See �� Roos Meijering, Literary and Rhetorical Theories in Greek Scholia (Groningen, 1987), esp. pp. 49–52. 24


23

impact (ekplēxis). In the mouths of Greek writers of the Roman period these sentiments attest to the constant presence of the classical past. Dio imagines himself as a classical producer, staging the three Philoktetes in his head. But, as Froma Zeitlin has shown, it was a response not confined to Greek readers in the case of Homer.27 Nor did the prestige of the past necessarily imply the inadequacy of the present; Dio notes that he was able to compare the tragedians’ plays in a way that would not have been possible for a fifth-century Athenian. As has already been noted, what is striking about the use of the classics throughout the Roman period is the freedom with which the past could be remodelled and reworked. The canonical classical texts, for all the respect paid to them, were not seen as untouchable monuments but as sources of material, as spurs to emulation. Even at the humble level of the rhetorical schoolroom, boys were taught to undercut the epic heroes, finding fault with Achilles, or to argue with the canonical stories.28 The past was exemplary but was part of a common cultural property for all those who were educated, and thus open to endless reworking as readers became speakers and writers in their turn.29 In her review of reader-response through the ages Jane Tompkins notes this imaginative and emotional engagement of ancient readers. She argues that ancient understandings of the relation between text and reader were very different from the more analytical approaches of modern reader-response criticism.30 In particular, she identifies this involvement as a key difference between ancient criticism and modern reader-oriented theories: rather than being concerned primarily with deciphering meaning, ancient critics reveal a ‘concept of language as a force acting on the world’.31 What Tompkins’ analysis underlines is the contrast between ancient and modern attitudes to literature and language. Where the modern professional reader, the critic, tends to treat his or her subject as an object of analysis, the ancient critic stresses the impact of the text. This does not, of course, mean that modern readers do not respond in the ways described by ancient critics, but that these types of responses are 27

�� Zeitlin�� , ‘Visions and revisions of Homer in the Second Sophistic’. �� See Webb, �� ‘The Progymnasmata as practice’,�� pp. 301–2. 29 �� See especially Whitmarsh, Greek Literature and the Roman Empire; Joy Connolly, ‘Problems of the past in Imperial Greek education’, in Yun Lee Too (ed.), Greek and Roman Education (Leiden, 2001), pp. 339–72; Schmitz, Bildung und Macht, in an otherwise brilliant analysis of the social function of sophistic practices, places too much emphasis on the past as an overwhelming burden that crushed the elite. 30 Jane P. Tompkins, ‘The reader in history: the changing shape of literary response’, in Jane P. Tompkins (ed.), Reader-response Criticism: From Formalism to Post-structuralism (Baltimore, 1980)�. 31 �� Tompkins, ‘The reader in history’, pp. 202–3. See also �� Walter Ong, Orality and Literacy: The Technologizing of the Word (New York, 1988)�� , Chapter 3. 28


24

not valued, theorized and articulated in the same way and do not have the same social and cultural significance as, say, an Imperial Greek reader’s response to his classical reading. Discussions of visual response to words, as Tompkins points out, are one area in which these differences emerge particularly acutely. As Ellen Esrock has noted, ‘readerly visuality’ has been neglected as a valid response by modern criticism for a variety of reasons.32 Ancient critics, by contrast, speak as if such imaginative responses to words were the norm. In the case of Quintilian and Dionysios, failure to respond in this way is even seen as a sign of a much greater moral deficiency. Those who do not respond as they do are branded as slow, incapable, difficult to please – the language bristles with terms of distance and negation (‘abest’, dusarestos). Such confidence may seem surprising to us. It goes against our own culture’s tendency to assume that visualization in response to reading is personal and variable in intensity and in content. An average group of twentieth- or twenty-first-century readers will probably contain individuals who admit to similar experiences when reading, and others who claim never to ‘see’ what they read. Many people assume that their experience of reading is universal and seem genuinely surprised to find that others have such different experiences of reading.33 This discrepancy between modern experience and the claims of ancient critics raises the question of whether we should discount the claims of ancient critics, or whether the ancient experience of reading was very different from our own. When Quintilian and Dionysios both ask ‘who could fail to’ respond in the ways they prescribe they open up the possibility that some individuals may not have responded in this way but, at the same time, they make clear what they consider the norm to be. The question of whether this represents a real difference in response between ancient and modern audiences has been raised by Ann Vasaly and I would agree with her suggestion that things were different in the ancient world and that ancient audiences were more consciously attuned to visual effects and did ‘see’ the subject of poems and speeches in their mind’s eye.34 The most striking difference does not perhaps reside in the elusive domain of personal response but in the discussions of that response. It 32

�� Ellen J. Esrock��, The Reader’s Eye: Visual Imaging as Reader Response (Baltimore, 1994). �� These conclusions are the result of several discussions with small seminar groups composed of graduate students and faculty at Princeton University. For a more scientific approach, see �� Jocelyn Penny�� Small��, Wax Tablets of the Mind: Cognitive Studies of Memory and Literacy in Classical Antiquity (London, 1997), pp. 130–31 �� on the work of Michel Denis. 33

34 Ann Vasaly, Representations: Images of the World in Ciceronian Oratory (Berkeley, 1993), p.�� 99. On the question of how rhetors could predict the imaginative response of their audience, which Vasaly raises here, see Chapters 4 and 5 below. I am leaving aside the question of what the mental experience expressed by the claims to ‘see’ actually might have been.


25

is clear that visualization was valued and encouraged within ancient education as part of a larger attitude to literature as a force able to penetrate and shape the individual. The ability of words to affect the imagination is allied to the idea that they can enter and dwell in the soul, leading one authority, according to Plutarch, to suggest giving earguards to the young to protect them from the wrong sort of language.35 Ancient education clearly placed a value on visualization as a response, and created a vocabulary for the identification and expression of such experiences. It is perhaps not surprising that in a culture such as our own, where the attitude to visualization is sceptical and the activity considered as an intensely individual matter, reports of experience should vary so enormously. In antiquity, by contrast, such visualization could be a very public and shared matter, as we shall see. In particular, it is clear that educators expected and encouraged visualization. One area where we have clear evidence of training in visualization in ancient schools is that of memory techniques which relied on the conscious creation, manipulation and storage of often bizarre images.36 And, as Small points out, in a culture in which the technical difficulties involved in reading and writing made memorization a vital skill it is conceivable that people were more prone to use mental images in speaking and listening than we are.37 In all the examples cited above, the readers who felt that they were in the presence of the subject matter were responding to a text from an earlier period. They reveal a concept of classical texts as privileged points of access to the experience of the past, which make not just the subjects seem present but the authors as well. Again, this was clearly a practice encouraged by their education. A discussion of reading aloud by the firstcentury rhetorician Theon of Alexandria recommends that the student reading the text of a classical orator should think himself into the skin of the original speaker – Demosthenes or Aeschines, for example – at the original moment of performance. The point of this method acting is to involve him totally in the text, emotionally as well as intellectually.38 Such deep identification with the authors of the past continued in schools with the exercise of declamation, in which students argued imaginary cases 35

�� Plutarch, On Listening to Lectures (Moralia 37F–38B). Rhetorica ad Herennium, 3.16.28–24.30. �� See also Cicero, De oratore, 2.354–60�� ; Quintilian, Institutio oratoria, 11.2; the seminal work of Frances Yates, The Art of Memory (London, 1966);� and Small, Wax Tablets of the Mind. 37 �� See also �� Agnès Rouveret�� , Histoire et imaginaire de la peinture ancienne (Ve s. av. J.-C.–Ier s. ap. J.-C.) (Rome, 1989), p. �� 312. 38 �� Theon, Progymnasmata, section 13 (p. 103). This discussion of reading aloud comes from the end of the work, which does not survive in the Greek manuscripts and is preserved only in the Armenian translation. 36


26

in the persona of a character from the classical past, often an orator like Demosthenes.39 Theon’s recommendations also help to explain the way in which ancient readers of all types of text cast themselves as listeners. The term translated ‘reader’ in several of the examples above is akroatēs – ‘listener’ – and what is read is often referred to as a logos, with all its implications of live speech. Despite the importance of the written word, and its culturally crucial role in preserving the words of past eras, the reception of texts remained an essentially aural experience. Active listening was considered as an important activity in itself.40 As Theon shows, reading in the school situation meant reading aloud to others, and even solitary readers are known to have pronounced the words out loud, casting themselves simultaneously as speaker and audience.41 This effacement of the written medium brings the author, whether poet, historian or orator, into direct proximity, casting the reader as a live audience member, like Dio at his private ‘performances’ of tragedy. All readers, even of the deadest of poets, are thus assimilated to the audiences of a live performance. The live audiences of spoken orations were also assumed to respond in the same way to the effective use of vivid language. We have fewer testimonies of individual response, but the whole treatment of ekphrasis and enargeia by ancient rhetoricians is based on the assumption that audiences can be made to respond imaginatively to a speech, placing themselves in the situation described. Aristotle, in the Rhetoric (1411 b24– 5), refers to the ability of certain metaphors to place the image ‘before the eyes’ – pro ommatōn – and makes a distinction between metaphors that evoke an image of motion (energeia) which have this effect, and others, whose image is static, which do not. Quintilian (8.3.62) also makes clear that the audience of a judicial speech should have the subject matter ‘displayed to the eyes of the mind’, and that he regards success in this as a vital ingredient of the persuasive force of the speech. The speech will not have the power that it should (‘debet’) have if the judge believes he is hearing a simple narration, rather than having the facts displayed to his mind’s eye, ‘si narrari credit, non exprimi et oculis mentis ostendi’. Quintilian’s purpose in describing his own response to Cicero, or picking out the vivid passages in the Aeneid is therefore to explain to his 39

See Brian Reardon, Courants littéraires grecs des IIe et IIIe siècles après J.-C. (Paris, 1971);�� Donald �� Russell, Greek Declamation (Cambridge, 1983)�� ; Schmitz, Bildung und Macht and ‘Performing History�� ’; and Webb, ‘Fiction, mimesis and the performance of the Greek past’. 40 �� See Theon, Progymnasmata, section 14 and Plutarch, On Listening to Lectures (Moralia 37C–48D). 41 �� See William �� Harris��, Ancient Literacy (Cambridge, MA, 1989), pp.�� 35–6 with further bibliography.


27

own readers how, and why, to work the same effect in their own speeches. This practical orientation is shared by most ancient critics – they aim to tell us how to do it, not simply to analyse the qualities of a particular writer or passage – making ancient ‘criticism’ a very different phenomenon from modern literary criticism. Ancient ‘critics’ were mostly practitioners of rhetoric and their primary goal was to show others how to be active practitioners and how to harness the power of language for their own ends. Just as the ancient reader was simultaneously speaker and listener as he read aloud, educated readers in general expected to become writers and speakers in their turn.42 The elementary exercise of ekphrasis was one means by which students were taught to appreciate the ability of words to spark an image in the mind and to master this power for themselves. The discussions of ekphrasis can therefore help us to understand how texts were read and what impact the spoken word was thought to have upon an audience. They also reveal the strength of the conception of language as a power acting upon the world that was current throughout antiquity. As a special use of language to bring the subject matter ‘before the eyes’ of the listener, penetrating the mind and acting on the most intimate of faculties, ekphrasis and enargeia also lie at the intersection of word and image. Any examination of either has to take account of ancient theories of psychology in which mental images (phantasmata or phantasiai) played a vital part from the classical period onwards. The plain, paradoxical statement we find in the Progymnasmata and elsewhere that language places a subject ‘before the eyes’ depended on a body of assumptions about language and its impact on the human mind. In turn, these ideas can point to the effects that words actually had on their audiences, as their minds became the locus of interaction between word and image. The ancient discussions of ekphrasis define it as a type of speech that creates immaterial images in the mind. The speaker of a successful ekphrasis is therefore a metaphorical painter, the result of his words is a metaphorical painting and this analogy emerges at certain points in the discussions. The Byzantine scholar John Sardianos, for example, commenting on the ancient rhetorical texts that were still in use in the Greek Middle Ages, points out that ekphrasis works by ‘imitating the painters’ art’.43 How this paradoxical feat could be accomplished, and why it was considered useful for students of rhetoric, will be the subject of the 42 �� Harris, Ancient Literacy, pp. 222–3 notes ‘it may seem that only the exceptionally inarticulate members of the Roman upper class refrained from literary composition’. The same can be said of the Greek elite in the Roman period, assuming that ‘literary composition’ includes rhetorical compositions. 43 �� Sardianos, Commentarium, p. 217, ll. 3–5.


28

next chapters. Before addressing these questions, I would like to return to the modern definition to trace the processes by which the definition took on its familiar modern contours. I believe it is particularly important to try to identify the various ideas and interests that gave rise to the modern definition since the intellectual climate in which it developed was radically different from the world of the ancient rhetors. Ancient and Modern Ekphrasis As outlined above, the basic modern definition grew out of the ancient one by a double process of restriction and expansion, resulting in a concept that is related to but radically different from the ancient meaning that is the subject of this book. First, attention was focused on those ekphraseis that described paintings, sculptures and monuments, leading to an association of the term with that restricted range of subject matter; then the term’s frame of reference was expanded to include other texts, both ancient and modern, referring to the arts. The end result of this process was a genre of writing about art that is often considered to be best exemplified by poetic examples.44 In antiquity, by contrast, ekphrasis was part of a rhetorical training. The authors of the Progymnasmata adduced examples from poetry, like the Shield of Achilles or descriptions of characters from epic, but they were part of a system in which the reading of poetry was largely subordinated to the study of rhetoric. It is ironic that an ancient rhetorical technique should have metamorphosed into an essentially poetic phenomenon, but this contrast reveals one of the key differences between the culture of Roman period and modern aesthetic interests. This dual process of restriction and expansion that resulted in the modern definition was motivated by diverse intellectual currents: interest in the aesthetic problems of describing the arts in words and archaeological curiosity and controversy about the realities lying behind Philostratos’ Eikones and other ekphraseis. Although this process culminated in the mid twentieth century, before which time the term ekphrasis was mainly restricted to studies of Imperial literature and often relegated to obscure footnotes, it was the result of centuries of interest in verbal accounts of the arts. Two French studies of Lucian’s works show the general state of affairs in the earlier part of the twentieth century. It was their interest in aesthetic 44

�� Murray Krieger, Ekphrasis: The Illusion of the Natural Sign (Baltimore, 1992) and �� James A.W. Heffernan, Museum of Words: The Poetics of Ekphrasis from Homer to Ashbery (Chicago, 1993) �� concentrate on examples drawn from poetry; John Hollander, The Gazer’s Spirit: Poems Speaking to Silent Works of Art (Chicago and London, 1995) defines the subject of his work as ‘actual ecphrasis’ – that is, poems written in response to real works of art (p. 4).


29

questions that led Henri Piot and Jacques Bompaire to focus on the special category of ekphraseis of works of art while acknowledging that the sense of the term was far wider in antiquity. Discussing �� the ancient definition, Piot points out that any ekphrasis (of whatever type of subject) in some sense rivals the graphic arts and is thus an example of the art of literature attempting to appropriate to itself the function of another art.45 Bompaire, �� writing in the 1950s, also acknowledged that ekphrasis in Lucian’s day was not conceived as a genre dedicated to the work of art but explained that he chose to focus on this type of ekphrasis simply because it was the most intrinsically interesting, showing as it did the interaction between literature, graphic arts and architecture.46 The precise language used by these two scholars suggests that their interest in ekphraseis of works of art was stimulated as much by contemporary aesthetic concerns as by the popularity of works of art and architecture as subjects for ekphrasis among Lucian and his contemporaries. This is particularly clear in Piot’s use of the phrase ‘transpositions d’art’ which recurs frequently in discussions of ekphrasis in its restricted sense from the late nineteenth and early twentieth centuries. The phrase was coined by Huysmans in the early 1880s to express the way in which Odilon Redon seemed to capture in his paintings the atmosphere of Poe and Baudelaire, but it also summed up the enterprise of French decadent poets such as Théophile Gautier �� who attempted to translate the sensual beauty of art into words. For Gautier, doyen of the doctrine of ‘l’art pour l’art’, the verbal and graphic artist shared the task of representing reality through the workings of art.47

45 Henri Piot, Les Procédés littéraires de la IIe Sophistique chez Lucien: l’ecphrasis (Rennes, 1914), p.�� 22:�� ‘L’objet de l’ecphrasis, dit Hermogène “est �� de mettre sous les yeux la chose qu’on veut montrer”. Elle a des rapports très étroits avec les transpositions d’art en honneur chez les Alexandrins’� [The �� aim of ecphrasis, according to Hermogenes, ‘is to place the subject which one wishes to display before the eyes’. It is closely related to the ‘transpositions d’art’ beloved of the Alexandrians]. 46 �� Jacques Bompaire, Lucien écrivain: imitation et création (Paris, 1958), p.�� 707: ‘le �� sens le plus intéressant est celui qui fait de l’ecphrasis d’oeuvre d’art, sc. tableau, édifice, l’ecphrasis par excellence … la littérature se nourrit d’art�� ’ [The most interesting meaning is the one which treats ecphrasis of works of art (paintings, buildings) as ecphrasis par excellence … literature is nourished by art]. Cf. �� Louis Méridier, L’Influence de la Seconde Sophistique sur l’œuvre de Gregoire de Nysse (Paris, 1906)�� , p. 150 (on Gregory of Nyssa). 47 Gautier’s own writing has been described as a ‘transposition écrite du tableau’. See Georges Matoré, �� Le Vocabulaire de la prose littéraire de 1835 à 1845: Théophile Gautier et ses premières oeuvres en prose (Geneva and Lille, 1951), p. 142. In his poem ‘L’Art’, Gautier presents verbal and sculptural artistry as equivalent: ‘Oui, l’oeuvre sort plus belle / D’une forme au travail / Rebelle / Vers, marbre, onyx, émail’ [Yes, more beautiful pieces emerge from forms that resist being worked: verse, marble, onyx, enamel].

30


The part played by such aesthetic doctrines in the shaping of the modern conception of ekphrasis is shown clearly in the studies of Philostratos by two earlier French scholars, the art historian Edouard Bertrand and the philologist Auguste Bougot, both published in 1881. Both attempted to move the study of the Eikones away from the question of whether the text was archaeologically accurate or not, which had dominated the scholarship up until then, particularly in Germany. Bertrand and Bougot attempted instead to understand the cultural background to Philostratos – though Bertrand’s depiction of the Second Sophistic seems to owe a great deal to late-nineteenth-century Parisian culture – and identified Philostratos as an ‘art critic’ whose descriptions showed how art was perceived, an approach that anticipated twentieth-century readings of the text.48 Bertrand and Bougot also took the step of placing the text within a tradition of poetic descriptions of works of art, a type of text which resonated with contemporary developments in French literature and arts. Both, moreover, borrowed the term ‘ekphrasis’ as a label for this tradition, although the ways in which they introduced it into their discussions suggest that they were far from confident about its usage. �� Bertrand enigmatically refers to the descriptions of works of art in authors of the Roman period such as Catullus, Virgil, Statius, Martial, Apuleius and Lucian as belonging to ‘a fashionable genre which had its own name’.49 The name turns out to be ‘ekphrasis’. But Bertrand is curiously coy about saying so. He hides the word itself in a footnote and leaves it in Greek letters. Most importantly, however, he implies that Philostratos consciously saw himself as writing within this tradition. In a passage bordering on historical fiction, Bertrand imagines Philostratos as an ambitious writer, desperate to surpass these predecessors, who suddenly one day had the idea of creating the new genre of art criticism.50 Bougot, whose translation and substantial introduction appeared just before Bertrand’s study, also brought the term ekphrasis into his discussion rather ambiguously, entitling one section ‘L’ecphrasis ou description des oeuvres d’art’, leaving open the question of whether 48 �� See, for example, �� Jaś�� Elsner,� Art and the Roman Viewer: The Transformation of Art from the Pagan World to Christianity (Cambridge, 1995)�. 49 �� Bertrand, Un critique d’art dans l’antiquité, p. 49: ‘c’est un genre en faveur qui a un nom particulier’. 50 �� Bertrand, Un critique d’art dans l’antiquité, pp. 53–4: ‘Tourmenté de la même ambition, Philostrate s’était aussi créé un style singulier composé d’archaïsmes et de néologismes. Mais il eut un jour une pensée neuve, et ce jour là il créa un genre qui lui survécut et suscita des imitateurs: il créa la critique d’art’ [Tormented by the same ambition, Philostratos also developed a unique style combining archaisms and neologisms. But one day he had a new idea and on that day he created a genre which outlived him and inspired imitators: he created art criticism]. The use of unusual vocabulary which Bertrand underlines here was also a characteristic of Parnassian poetics.


31

ekphrasis is to be understood as equivalent to description generally, or to descriptions of works of art in particular. He took the constitution of the genre of ekphrasis one step further, however, by extending its frame of reference backwards in time to include descriptions of works of art in Homer and tragedy.51 All these studies illustrate the process by which ekphraseis of works from the Second Sophistic were gradually detached from their wider rhetorical background and from the broader definition of ekphrasis current in antiquity and placed within a different context: the description of works of art generally. �� Just as literary scholars’ interest in the enterprise of describing the work of art in words led them to detach ekphraseis of works of art from ekphrasis in general and place it in its new context, archaeologists followed precisely the same process in their study of texts like the Eikones and the Late Antique and Byzantine ekphraseis of works of art and architecture. They isolated these texts from their rhetorical background, treating them as precious sources of information about lost monuments.52 The major step in this direction was, of course, the great survey of descriptions of art and architecture in classical literature that forms the introduction to Paul Friedländer’s 1912 edition of two Late Antique ekphraseis of monuments: Paul the Silentiary’s ekphrasis of Hagia Sophia and John of Gaza’s ekphrasis of a bath building. Friedländer’s survey was to become the ultimate definition of the genre of ekphrasis (in the sense of descriptions of works of art and architecture) in ancient literature and includes all�� the standard examples: the Shield of Achilles, Lucian’s Calumny of Apelles, Euripides’ Ion, Greek and Latin epigrams. But while Friedländer’s introduction was the most thorough review of ancient literary descriptions of works of art, he did not take the further step of defining these texts as members of a single genre, still less of reifying that genre by giving it a name. Instead, he was extremely careful to use the term ‘ekphrasis’ only when discussing texts like the Philostratean Eikones or Byzantine ekphraseis of works of art which fell into the intersection of the ancient and modern definitions or which were discussed as examples of ekphrasis in antiquity.53 51

�� Bougot, Philostrate l’Ancien, pp. 1 and 171. Antonio Muñoz, ‘Le ἐκφράσεις nella letteratura bizantina e i loro rapporti con l’arte figurata’, in Recueil d’études dédiées à la mémoire de N.P. Kondakov (Prague, 1926) focused attention onto Byzantine ekphraseis of monuments and icons as a source of information on lost works. As Henry Maguire, Art and Eloquence in Byzantium (Princeton, 1981), pp. 22–3 has noted, the concentration on ekphraseis of monuments and works of art for their archaeological information has had the effect of distracting attention from other types of ekphrasis.� 53 Friedländer,� Johannes von Gaza und Paulus Silentiarius, pp. 84–5. In his �� 1939 edition of Prokopios of Gaza’s ekphrasis of a painting, Fried�� länder�� still used the term 52

32


The reputation that Friedländer’s survey has acquired as the first definitive survey of ekphrasis is particularly ironic in the light of his own deep concern about the correct use of terminology. In particular, he criticizes those anonymous scholars who have seized on the term ‘ekphrasis’ to explain texts like those of the Philostratoi. The rhetorical exercise, Friedländer explains, cannot in itself explain the existence of such texts; instead they should be seen against the wider background of the tradition of describing works of art and architecture in ancient literature.54 His survey makes an important point, that clearly there was such a tradition and that texts discussing or describing the arts are to be found in many classical genres from epic to epigram, from historiography to the magpie-like miscellany of a Pliny. But he does not present these passages as a genre and sees them as distinct from the ancient practice of ekphrasis. In collecting together these passages Friedländer, like Bertrand and Bougot, was following the lead of earlier scholarship on Philostratos: in 1709 Gottfried Olearius had included in his edition of the Philostratoi and Kallistratos a discussion of the works’ origin and speculated that all three of the later writers had found a common inspiration in the Homeric Shield of Achilles.55 Another German editor of Philostratos and Kallistratos, Jacobs (who published his edition jointly with Welcker in 1825), had taken this exploration of the origin of the Eikones one step further, noting the poetic precedents and citing not just Homer but also Hesiod, Theocritus, Apollonius Rhodius, Moschus, Quintus Smyrnaeus, Catullus, Virgil, Ovid and Statius. However, he did not take the further step of naming these descriptive passages ‘ekphraseis’ (the term is used, in Greek, in the general sense of ‘description’) or of claiming that they constituted a genre.56 But the nucleus of the modern genre is here, as it would later be ‘Bildbeschreibung’. 54 �� Friedländer, Johannes von Gaza und Paulus Silentiarius, p. 83: ‘Man pflegt heut literargeschichtliche Fragen, wie sie uns beschäftigen, mit dem Schlagwort “rhetorische Ekphrasis” mehr scheinbar als wirklich zu beantworten und glaubt, dass hier die ‘Rhetorenschule’ schöpferisch gewesen sei.’ See also his comments on descriptions of works of art in the novel: ibid., pp. 47 and 54–5. 55 �� Philostratos, Opera, ed. Olearius (Leipzig, 1709), p. �� 760�� . The decision to group the Elder and the Younger Philostratos’ Eikones together with Kallistratos’ ekphraseis of statues in a single volume is in itself a significant step that divorces the Eikones from the Philostratean corpus to place it in a context of descriptions works of art, rather than a wider rhetorical context. 56 �� Interestingly, Jacobs notes in his introduction that he became interested in the Eikones while working on Greek epigrams about art works some 30 years earlier. He therefore approaches the text from an archaeological standpoint reflected in his use of the term ekphrasis as the equivalent of description as he understood it, i.e. an accurate account of the appearance of an object. This leads him to the ironic claim (pp. xvi–xvii) that Philostratos


33

in a footnote by Erwin Rohde who, in his study of the Greek novels, first published in 1876, suggested that the roots of ‘rhetorisch-sophistischen ekphraseis’ of paintings and statues from the Second Sophistic lay in the earlier poetic tradition.57 This note, too, mentions many of the standard examples of what would now be called ‘ekphrasis’, from the Shield of Achilles, through the Cloak of Jason in the Argonautica and Moschus’ Europa, to Virgil, Catullus and Nonnus. Like Friedländer’s survey all these footnotes and comments served to draw attention to the phenomenon of describing works of art in antiquity but without applying the label ekphrasis to them. The term floats in the vicinity of these discussions, but, unlike Bertrand and Bougot, the authors do not apply it as a unifying label for writing on art nor do they suggest that they are discussing or defining a genre. This explains why Schissel von Fleschenberg, in his study of the use of descriptions of works of art in the novel published in 1913, does not use it but invents instead the term ‘Bildeinsatz’ (‘inset painting’) to designate what would now almost automatically be termed ekphrasis.58 What Schissel von Fleschenberg’s article does show is the growing interest in this technique as a literary phenomenon. In the same way, Bertrand and Bougot were inspired by contemporary cultural debates in their refreshing attempt to wrest the Eikones from the stale debates about the accuracy and reliability of the descriptions and to show that Philostratos’ enterprise was something different: a reflection of the place of art in society and, in Bertrand’s account, the birth of art criticism. As we have already seen, a parallel interest in Imperial ekphraseis of paintings, sculpture and buildings as sources of information on lost works also contributed to the process. However, almost all of the classical and medieval studies acknowledge the ancient sense of the term and often take their starting points from ekphraseis of works of art. The revolutionary step of defining ekphrasis as an essentially poetic genre, totally divorced from the rhetorical form of ekphrasis, was taken by Leo Spitzer in his famous essay on Keats’ ‘Ode on a Grecian Urn’, first published in 1955. Here, Spitzer identifies the ode as part of a long tradition called ekphrasis: failed to produce proper ekphraseis and succumbed instead to the temptations of rhetorical embellishment. 57 �� Erwin Rohde, Der griechische Roman und seine Vorläufer (Leipzig, 1900), p. 360n. 58 �� In a letter to Friedländler written in 1913 and now in the Paul Friedländer Collection, UCLA Special Collections (Box 1). In the letter Schissel uses the term ‘Bildbeschreibung’ but not ekphrasis. Schissel von Fleschenberg criticizes his historicist approach as old-fashioned (as well as taking great pleasure in pointing out various omissions of primary and secondary sources). In his article Schissel justified his deliberate omission of any enquiry into the origins of the Bildeinsatz by distinguishing his type of literary studies (Literaturwissenschaft) from the (implicitly inferior) enterprise of literary history.


34

it [the ode] belongs to the genre, known to Occidental literature from Homer and Theocritus to the Parnassians and Rilke, of the ekphrasis, the poetic description of a pictorial or sculptural work of art, which description implies, in the words of Théophile Gautier, ‘une transposition d’art,’ the reproduction, through the medium of words, of sensuously perceptible objets d’art (‘ut pictura poesis’).59

Spitzer does not give any source for his definition, perhaps not surprisingly, and his calm assurance masks the innovation he was making. His mention of the Parnassians and Gautier, to whom he erroneously attributes the phrase ‘transposition d’art’, do, however, reveal the intellectual currents which shaped his interest in and conception of ekphrasis.60 A further parallel with Bertrand’s and Bougot’s studies of Philostratos in particular emerges later on in Spitzer’s discussion of Keats’ ode when he insists on the poem’s function as a representation of the poet’s response to the sight of the urn. Interestingly, he presents this as a particular development of ekphrasis, which he appears to conceive of as simply a form of objective description: ‘The ekphrasis, the description of an objet d’art by the medium of the word, has here developed into an account of an exemplary experience felt by the poet confronted with an ancient work of art …’61 Spitzer’s achievement was to create a concept of a poetic genre that triumphantly transcended both time and place. Homer, Theokritos, the Parnassian poets and Rilke all partook of the same essence. The way in which he introduces the first of the quotations gives the unsuspecting reader no clue that he has just invented a genre, albeit one that had long been waiting to happen. He presents it as ‘an obvious generic statement’, a straightforward starting point for any study of Keats’ ode. His method is to separate what might count as straightforward description (‘What exactly ... Keats [has] seen (or chosen to show us) depicted on urn he is describing’ (pp. 72–3)) from ‘the symbolic or metaphysical inferences drawn by the poet from the visual elements he has apperceived’ (p. 73). The statement that Keats’ poem is about an art object and that other such poems have 59

Leo Spitzer, �� ‘The �� “Ode on a Grecian Urn,” or content vs. metagrammar’,�� in Essays on English and American Literature, ed. Anna Hatcher (Princeton, 1962)�� , p. 72. 60 �� Spitzer had previously published articles on descriptions of works of art in the poetry of Mörike, ‘Wiederum Mörikes Gedicht “Auf eine Lampe”’, Trivium, 9 (1951): 203–25 and Garcilaso de la Vega�� , ‘Garcilaso, Third Eclogue, Lines 265–271’, Hispanic Review, 20 (1952): 243–8.�� The term ekphrasis appears once, buried in a footnote, in the latter of these. Given Spitzer’s background in French and classics and his interest in the description of works of art, it does seem likely that he knew the work of Bertrand, Bougot and Piot, not to mention the Oxford Classical Dictionary. I am grateful to Alejandro Coroleu for drawing my attention to the article on Garcilaso. 61 �� Spitzer, ‘The “Ode �� on a Grecian Urn”’, p. 89.


35

existed from antiquity onwards is, one would think, incontrovertible, and gives Spitzer the basis for a brilliant analysis. But such a straightforward statement hides the revolutionary nature of his claim for ekphrasis and its implications of unity of poetic purpose across time and space. What is a useful concept and a useful set of comparanda for the critic embarking upon a reading of the ‘Ode on a Grecian Urn’ is presented as literaryhistorical fact. One can speculate on the attraction of such continuity in Western culture for a refugee scholar like Spitzer.62 But the effect of his definition, taken in isolation, was to obliterate cultural distinctions and to remove rhetoric, particularly the rhetorical culture of the Roman period, entirely from the picture. The term ekphrasis is restricted not only to ‘descriptions of works of art’ but to poetic descriptions of works of art and simultaneously expanded to include all periods of ‘Western culture’. The rest is history. Spitzer’s constitution of the new genre of ekphrasis catapulted the word out of the specialized domain of classical and archaeology into the world of English and Comparative Literature, sparking essays, books, colloquia, redefinitions and counter-definitions.63 The popularity and influence of Spitzer’s definition show more clearly than ever that his new genre satisfied an intellectual need. Descriptions of works of art as a group had attracted interest since the Renaissance.64 This interest emerges in the grouping of Philostratos’ Eikones with Kallistratos’ Ekphraseis by Olearius, in the archaeological debates about whether such descriptions were 62 �� Spitzer himself makes an autobiographical comment in the article, describing himself as ‘a European born scholar, nurtured in a centuries-old tradition of both scholarly and aesthetic interpretation, especially in the classical and the French fields’, as he laments the divide between ‘scholars’ and ‘critics’ in US English departments: ‘The “Ode on a Grecian Urn”’, p. 68. 63 �� See, for example, James A.W. Heffernan, ‘Ekphrasis �� and representation’, New Literary History, 22 (1991): 297–316�� and Museum of Words. Jean Hagstrum, The Sister Arts: The Tradition of Literary Pictorialism and English Poetry from Dryden to Gray (Chicago�� , 1958), p. 18, n. 34 proposed his own definition: ‘I use the noun “ecphrasis” and the adjective “ecphrastic” in a more limited sense to refer that special quality of giving voice and language to the otherwise mute art object.’ Hagstrum �� is keen to provide ancient justification for his idiosyncratic usage, claiming that ‘my usage is etymologically sound since the Greek noun and adjective come from ekphrazein, which means “to speak out”, “to tell in full”.’ �� On the function of the ‘ek’ in ekphrasis, see below, Chapter 3 (p. 74). 64 �� One of the first to conceive of literary descriptions of works of art as a coherent group was Erasmus. His rhetorical treatise, De Copia (II, 202),�� includes a list of passages describing paintings and sculptures: ‘Statuarum item: qualis est in epistolis Plinianis signi senilis; tabularum et imaginum: qualis est apud Lucianum Hercules Gallicus, apud Philostratum varia picturarum argumenta’; to this category belong also Ovid’s description of Arachne’s tapestry in Metamorphoses, the Homeric Shield of Achilles and its Virgilian descendent, the Shield of Aeneas, ending ‘ad haec navis, vestis, παvoπλίας, machinae, currus, Colossi, pyramidis, aut si quid est aliud rerum consimilium, quarum descriptio delectet.’


36

records or ‘mere rhetoric’, in the fascination with the ways in which one art represents another and the autonomy of the arts that this suggests. Spitzer’s ekphrasis was a genre, or mode, in search of a label. The vogue for New Criticism meant that Spitzer’s definition fell on particularly fertile ground, as witnessed by Murray Krieger’s study of Keats’ ode and of his conception of ekphrasis.65 Since the mid twentieth century, ekphrasis, generally understood as the description of works of art, usually in poetry, has become a familiar term, though the basic definition has given rise to many different approaches, which it is impossible to survey here. One such development that has been fruitful in the study of classical literature is the reading of descriptions of works of art as metapoetic commentaries on the literary work within which they are found.66 Such developments reflect the ‘linguistic turn’ in twentieth-century approaches to literature, a development that freed the study of description in particular from the expectation that language should depict reality and underlined the problems involved in verbal representation.67 This same development made possible a fresh appreciation of rhetorical texts of all kinds, in particular the rhetoric of the Second Sophistic, including its ekphraseis of works of art. However, the resulting renaissance of interest in the Second Sophistic and later rhetoric has tended to focus on those texts that most resemble modern literature, such as dialogues and the novel, and, in the case of ekphrasis, a certain subgroup of ekphraseis of works of art (such as Philostratos’ Eikones) has come to stand as emblematic of the whole of ekphrasis. The fact that this move misrepresents the category of ekphrasis as it was understood in antiquity does not in any way undermine the interpretative value of individual studies of individual literary texts. But, if we are interested in the wider intellectual and cultural contexts in which these texts were read, heard and composed, it is misleading to assume that ancient categories and assumptions about language were identical to our own. Of course, those ancient categories are irrecoverable in their entirety but there is still a large amount of information that is still to be exploited in sources such as scholia and rhetorical handbooks.68 This study aims to elucidate the 65

�� Murray Krieger, Ekphrasis: The Illusion of the Natural Sign (Baltimore, 1992). �� See, for example, �� Perrine Galand-Hallyn, Le Reflet des fleurs: description et métalangage poétique d’Homère à Erasme (Geneva, 1994)�� and Andrew Becker, The Shield of Achilles and the Poetics of Ekphrasis (Lanham, 1995)�. 67 �� See, in particular, the essays collected in Roland Barthes et al., Littérature et réalité (Paris, 1982) and in Yale French Studies, 61 (1981). 68 �� Two notable examples of such an approach are �� Roos Meijering, Literary and Rhetorical Theories in Greek Scholia (Groningen, 1987) and Michel Patillon, La Théorie du discours chez 66


37

ancient category of ekphrasis and the assumptions that underlay it by appealing to a wider range of rhetorical texts and discovering what could be called the poetics of ancient rhetorical theory. It is important to note, too, the radical differences between the interests and assumptions that drove the modern structuralist and linguistic approaches to literature and the ancient rhetorical theories that are the subject of this book. The ancient rhetoricians assume a live audience upon whom the speaker can exert an impact and their orientation is first and foremost practical (despite the appearance of obsessive categorization and subcategorization that often strikes the modern reader on first encountering their work). It is not, therefore, surprising if their ways of approaching discourses of all kinds are different from those of a modern critic whose task is to analyze a written text. We have already seen the emphasis placed by ancient readers on their imaginative and emotional responses to certain passages, precisely the type of response that is largely ignored by structuralist and post-structuralist approaches. It is remarkable that even reception theory privileges the hermeneutic mode of reading which casts the reader as interpreter engaged in a patient act of decoding. Though such approaches to texts – and images – certainly did exist in antiquity, they were not at the heart of the ancient phenomenon of ekphrasis. Conclusion The purpose of this excursus into the modern redefinition of the term ekphrasis is to show the intellectual interests that shaped it. Though Spitzer, for example, presents his category of ekphrasis as a transcendent phenomenon that has existed throughout literary history, it is very much contingent on a particular time and place and on a particular narrative of ‘Occidental Literature’ which leaps blithely over those aspects of ancient culture that are less readily assimilable to modern ideas of literature than are Homeric epic or Theocritean pastoral poetry. The ancient conception of ekphrasis is just as much a product of its time and its culture. Clarity is vital precisely because of the ultimate closeness of Spitzer’s definition of ekphrasis to the ancient one. We have seen how the isolation of ancient ekphraseis of works of art from ekphraseis in general was accompanied by the gathering together of poetic precedents for the description of works of art from various genres, leading to this collection of illustrative examples being understood as a genre in itself. There is therefore a genealogical connection between the ancient and modern definitions, a connection reflected in the primacy of the visual in both. But Hermogène le Rhéteur: essai sur les structures linguistiques de la rhétorique ancienne (Paris, 1988).�

38


the different role of the visual is key to the profound differences between the conceptions underlying the two definitions. For the modern definition the visual is a quality of the referent, which in some definitions is already a representation of reality. For the ancient rhetoricians the impact of ekphrasis is visual; it is a translation of the perceptible which mimics the effect of perception, making the listener seem to see. The impact on the audience is powerful and immediate – as Plutarch claims for Xenophon and Thucydides; it is a psychological effect, and, as I shall argue below, what is imitated in ekphrasis and enargeia is not reality, but the perception of reality. The word does not seek to represent, but to have an effect in the audience’s mind that mimics the act of seeing. As moderns we cannot hope to understand something like ekphrasis entirely from an ancient perspective: the sources at our disposal represent a fraction of the definitions and paradigms available to the ancient rhetorician.69 I am fully aware that what I am proposing in the following chapters is another modern interpretation of ancient ekphrasis. However, it is one that is based on a more comprehensive study of the ancient theoretical sources than has been undertaken before and therein, I hope, lies its main interest.

69 �� See the remarks of Don Fowler in his essay �� ‘Narrate and describe: the problem of ekphrasis’, JRS, 81 (1991): 25–35.

2. Learning Ekphrasis: The Progymnasmata

Ancient Usages and Contexts As we have seen in Chapter 1, the restriction of the term ‘ekphrasis’ to texts on art is a relatively modern development, inspired by modern disciplines and fields of interest. If we turn to the ancient usages of the word we find ourselves in a very different intellectual world from that inhabited by the literary critics, art historians and archaeologists who contributed in their various ways to the invention of the modern genre of ekphrasis. Although the ancient usages of the term which are most familiar to us are from texts like the Younger Philostratos’ Eikones, to ancient readers it would most probably have brought back memories of their schooldays. Ekphrasis and the related verb ekphrazō were used overwhelmingly in technical, rhetorical contexts: in handbooks for students and teachers of rhetoric, like the Progymnasmata, and in learned commentaries on classical texts. The first recorded usage is preserved in the writings of the Augustan critic Dionysios of Halikarnassos. He criticizes the historian Philistos for failing to use an appropriate style in his ekphraseis of places, battles and the like. But since the passage in question comes from a later epitome it is impossible to ascribe its usage with any certainty to Dionysios himself. It was however certainly well established in the vocabulary of school texts by the first century CE when Theon wrote his Progymnasmata, which may therefore contain the earliest extant usage. As a post-classical term, it was generally avoided in finished speeches written in Attic Greek. When authors needed to refer to the act of describing in their works they preferred other terms, such as hermēneuo (‘to express in words’) or simply diēgoumai (‘to relate’, to ‘go through’), as Dionysios of Halikarnassos does in the parallel passage to the section of the epitome in which the term ekphrasis occurs. ‘Ekphrasis’, in contrast, was mostly restricted to the vocabulary of the classroom, the tools of the teacher’s trade, used to teach composition and to dissect the works of

Dionysios of Halikarnassos, On Imitation, 3.8. See Wilhelm Geissler, Ad descriptionum historiam symbola (Leipzig, 1916), p. 12. The usages of the verb ekphrazō in Euripides and Sophocles which are cited in the older editions of Liddell and Scott’s Greek–English Lexicon have now been rejected as later variants. Dionysios of Halikarnassos, Letter to Pompeios Geminos, 5.6.

40


the classical canon. Thus we find the term in Greek manuals of rhetoric from the Roman period, like Hermogenes’ treatise On Types of Style (Peri Ideōn Logou) and in the commentaries on his other work on declamation, or the guide to epideictic speeches by ‘Menander Rhetor’. It is also used in the commentaries on the classical texts which were read in schools. The scholia to Homer’s Iliad, for example, identify a range of passages as ekphrasis. Some of these overlap with the modern usage, like the description of shields, while others are simply moments in the action in which the commentator found particular appeals to the imagination. All these texts – scholia, Progymnasmata, and more advanced rhetorical handbooks – reflect the ˇteaching dispensed in schoolrooms throughout the Greek-speaking areas of the Roman Empire. They are often anonymous, or wrongly attributed to some famous name (like Hermogenes or Dionysios of Halikarnassos), and are difficult to date with any certainty. Taken together, however, they do reflect a set of coherent ideas and doctrines which make it possible to build up a picture of what ancient readers and writers understood by a term like ‘ekphrasis’. The general ideas underlying the treatments of ekphrasis in these technical sources were not new. Effective poetry, and later prose, had always appealed to the imagination. But what is particularly interesting about the Roman period is that we can see a range of interconnecting attempts to identify and teach this type of writing. We see which passages from the canonical classical texts became widely accepted as models; we see attempts to explain why and how these models should be absorbed and imitated. Finally, we can often see the results of the teaching methods in finished compositions, such as speeches or even in non-rhetorical works, so pervasive were the effects of the rhetorical education. As we have already seen, these technical, pedagogical sources, including the commentaries on classical texts, had a very practical aim: teaching students to express themselves effectively and in forms sanctioned by the prestige of tradition (the two were not necessarily contradictory or distinct). They are fragments of what was once a living and complex process of training in which a great deal must have depended on oral traditions and live interaction between teachers and students. As a result, they give us a privileged glimpse behind the literary scenes. However, for the very same reasons, they can be opaque.

Scholia graeca in Homeri Iliadem, ed. Hartmut Erbse (7 vols, Berlin, 1969–88) to 11.61–6 (shields); 15.237–8 (the descent of Apollo, likened to a hawk); 18.610 (the armour made by Hephaistos to go with the Shield of Achilles) and 23.232 (Achilles sinking into sleep). See the remarks of Raffaella Cribiore, Gymnastics of the Mind: Greek Education in Hellenistic and Roman Egypt (Princeton, 2001),�� p. 143.

Learning Ekphrasis: The Progymnasmata

41

The Progymnasmata textbooks for example were probably never intended to be a source of definitive statements of theory, but were keys to an educational process. As even C.S. Baldwin grudgingly admitted, they can give us some idea of the compositional categories in which students were taught to conceptualize their writings, as well as of habits of thought and analysis inculcated in the young. Theon certainly speaks as if this were the effect of the training he recommends. By offering examples of his own composition, the teacher will imprint them (tupoō) on his students’ minds, so that they in turn can imitate (mimeisthai) them. The two verbs reveal strikingly how Theon wished to think of the educational process he was representing: through his compositions, he shapes, even brands, the students’ minds.10 The student, in turn, learns by imitation, by aping the actions of his master until he is able to produce an analogous work of his own. In this way, he learns to imitate the processes, not merely the forms, of composition. As described by Theon, it is an intellectual formation which takes root at the level of habit. So the Progymnasmata reflect a stage in the shaping of the ‘literary consciousness’ (to borrow Genette’s phrase) of the elite students who studied rhetoric, rather than being critical tools for the analysis of texts. They also derived their full utility and significance from their wider cultural and intellectual context. The Progymnasmata are most likely to prove useful if read, as far as possible, against that context, as I will try to do in the next chapters, by setting them in their larger educational context and by investigating the ideas about vividness in language and about the relationship of mind, language, memory and emotion which are presupposed by the authors of the Progymnasmata. Though the Progymnasmata certainly do not suffice to explain works like the ancient

All the surviving versions of the Progymnasmata have been translated in George A. Kennedy, Progymnasmata: Greek Textbooks of Prose Composition and Rhetoric (Atlanta, 2003). I have supplied my own translations of the chapters on ekphrasis in Appendix A. Charles S. Baldwin, Medieval Rhetoric and Poetic (New York, 1928), p. 38: ‘Arid, impersonal as arithmetic, pedantically over-classified, sometimes inconsistent, these rules are nevertheless illuminating. They expose sophistic oratory. The patterns set forth for boys are recognizably the patterns of the public oratory of men.’ On the formation of habit in rhetorical education, see James J. Murphy, ‘The key role of habit in Roman rhetoric and education as described by Quintilian’, in Tomàs Albaladejo et al. (eds), Quintiliano: Historia y actualidad de la retórica (Logroño, 1998), pp. 141–50. Theon, Progymnasmata, ed. �� and trans. �� Michel Patillon and G. Bolognesi (Paris: Les Belles Lettres, 1997), 70, �� l. 30 – 71, l. 2. Examples of such models are included in the corpus of Libanios, in Opera, vol. 8. 10 Libanios, Letter 337 similarly refers to a grammarian who is able to make the poets ‘dwell in students’ souls’. On ancient images of education as moulding and shaping, see Teresa Morgan, Literate Education in the Hellenistic and Roman Worlds (Cambridge, 1998), pp. 259–60.

42


novels, Philostratos’ Eikones or even a declamation, they can provide one set of clues to the cultural and intellectual background of a work and to the basic assumptions about language and representation which inform the text. It is not surprising that the general influence of the Progymnasmata is detectable – sometimes quite patently, sometimes in a more subtle form – in poetry, historiography and other types of writing. Ekphrasis and the Progymnasmata As exercises, the Progymnasmata were a keystone of the education process of the elite, marking the transition from grammar to rhetoric and presenting the student with a set of literary, linguistic and ethical concepts at a formative stage in his career.11 With these exercises, students had their first taste of composition and their first opportunity to engage creatively with the literary traditions they had absorbed from the grammarians. The rhetorical skills they taught were the badge of the adult educated male.12 While some exercises, like the chreia (a discussion of a saying attributed to some famous figure such as Diogenes the Cynic) might include grammatical exercises (demanding that the students rewrite a simple sentence in different grammatical forms), others were primarily aimed at developing skills of presentation and argumentation.13 In the form in which they have come down to us, these exercises range from a fable (muthos), using animal characters to illustrate a moral, and a simple narrative (diēgēma), usually taken from mythology,14 through exercises in praise (enkōmion) and blame (psogos), to the systematic discussion of a general question such as whether one should marry (thesis) and the introduction of a law (nomou eisphora). Other exercises include the chreia; 11 See Theon, Progymnasmata, 70, ll. 29–32. On the curriculum, see �� Robert A. Kaster, Guardians of Language: The Grammarian and Society in Late Antiquity (Berkeley, 1988)�� ; Malcolm Heath, Menander: A Rhetor in Context (Oxford, 2004) and, on Progymnasmata in particular, see Morgan, Literate Education, pp. 191–2; Cribiore, Gymnastics of the Mind, pp. 220–44 and The School of Libanius in Late Antique Antioch (Princeton, 2007), pp. 143–7; and Ruth Webb, ‘The Progymnasmata as practice’, in Yun Lee Too (ed.), Education in Greek and Roman Antiquity (Leiden, 2001), pp. 289–316. 12 Public speaking was understood as a masculine activity, and in consequence I will assume that students and teachers were male. On female education, see Ewen Bowie, �� ‘The readership of Greek novels in the ancient world’, in James Tatum (ed.), The Search for the Ancient Novel (Baltimore, 1994), pp. 435–59 and Cribiore, Gymnastics of the Mind, pp. 74–101. 13 On the chreia, see Ronald Hock and Edward O’Neil, The Chreia in Ancient Rhetoric: Volume I. The Progymnasmata (Atlanta, 1986) and Volume 2: Classroom Exercises (Atlanta, 2002). 14 These narrative exercises, particularly diēgēma, prepared for the narrative section of a speech (diēgēsis).


43

kataskeuē and anaskeuē, the confirmation or refutation of a story on the basis of certain criteria like probability; koinos topos (common place), a rehearsal of the commonly held opinions about a certain type of person (such as a murderer or a temple-robber); synkrisis, or comparison, in which one thing is proved better or worse than another by systematic comparison of the qualities of both; prosōpopoiia or ēthopoiia, a speech in the words of a character in a certain situation (such as Ajax when he had lost the contest for the arms of Achilles, or Niobe’s lament for her children)15 and the composition of an ekphrasis. Quintilian (Institutio oratoria, 1.9) mentions some of these exercises and notes that in first-century CE Rome this part of the curriculum was being taken over by the grammarians. The discussion of these exercises by Suetonius suggests that by the late first or early second century they were no longer taught by the rhetors at all, and were presumably taught entirely in the grammarians’ schools.16 However, Quintilian warns us against assuming that this was a universal phenomenon. He attributes some of the blame for this situation to the Roman rhetors, many of whom considered such elementary teaching to be beneath them, and he contrasts their behaviour to that of their Greek counterparts who, he claims, recognized the value of the Progymnasmata and did not disdain to teach them. Theon’s Progymnasmata may therefore represent an attempt to ensure that these exercises continued to be taught as a coherent whole within the Greek schools of rhetoric, in contrast to the haphazard approach of the Latin-speaking teachers. It seems that in Greek schools up to the end of antiquity and beyond the Progymnasmata continued to be taught as part of rhetorical studies and were not generally taken over by grammarians (though practice must have varied considerably). The fact that, in the fourth century, Libanios composed fair copies of these exercises suggests that he still saw them all as falling within his duties as a rhetor, as Raffaella Cribiore has recently emphasized.17 The Sources for the Progymnasmata Theon claims to be the first to provide definitions (horoi) of the exercises (so the definition of ekphrasis found in all the versions of the Progymnasmata is 15

Theon calls this exercise prosōpopoiia, a term that is used elsewhere specifically of a passage where words are attributed to an inanimate object, such as a city. 16 Suetonius, De grammaticis et rhetoribus, 25.4. See also the introduction to Theon’s Progymnasmata by Michel Patillon (pp. xiii–xiv). Robert Kaster in his edition of Suetonius (Oxford, 1995) suggests on pp. 279–80 that all the exercises had been relegated to the grammatici by Suetonius’ time. 17

Cribiore, The School of Libanius, pp. 143–7.


44

perhaps his own), though he makes no claim to have invented the exercises themselves.18 Addressing himself to the teachers of rhetoric, he is the most informative of all the authors concerning the ways in which the exercises were to be taught, particularly in the discussion of pedagogical methods which precedes the chapters devoted to the individual exercises. A much shorter version of the Progymnasmata is transmitted with the works of Hermogenes, a rhetorical theorist of the second century CE whose writings on style and on argumentation came to form the core of the Byzantine rhetorical curriculum.19 The ps.-Hermogenean text may date to the third century.20 Its brevity led to complaints of obscurity from Byzantine readers.21 Neither of these two treatises came close to the popularity of the fourth-century version by Aphthonios, identified as a pupil of Libanios.22 Aphthonios secured lasting popularity by appending his own examples of each exercise, rather than simply referring to passages in the ancient authors as his predecessors had done,23 but his instructions are minimal. This deficiency in Aphthonios’ book was clearly felt by its Byzantine users as several commentaries were composed, drawing on Theon among other sources, to explain his laconic text phrase by phrase. One of these is by John Sardianos, probably writing in the ninth century, another by the eleventh-century scholar Doxapatres.24 One further ancient version of the Progymnasmata is by a certain Nikolaos, who is placed by the Souda in the fifth century.25 Like Theon, Nikolaos goes into each exercise in some detail and is a good deal more 18

Theon, Progymnasmata, 59, ll. 19–20. �� Ps.-Hermogenes, Progymnasmata, in Opera, ed. Hugo Rabe (Leipzig, 1913), pp. 1–27. The new edition of this text by Michel Patillon in Corpus Rhetoricum (Paris: Les Belles Lettres, 2008), pp. 180–206 appeared too late for me to take account of it in this book. 20 Pierre Laurens, preface to Hermogène: l’art rhétorique, trans. Michel Patillon (Paris, 1997), p. 41. 21 A scholiast quoted in RE 2, 1 col.2797 [s.v. Aphthonios] commented on the difficulty of understanding the Progymnasmata attributed to Hermogenes, describing them as ‘rather unclear and difficult to comprehend’ (ἀσαφῆ πως καὶ δύσληπτα); cf. Doxapatres, Homiliae, p. 131. 22 �� Aphthonios, Progymnasmata, ed. Hugo Rabe (Leipzig: Teubner, 1926). Corpus Rhetoricum, ed. Michel �� Patillon (Paris, 2008) contains a new edition of this text on pp. 112–62. 23 On the post-Antique popularity of Aphthonios, see Herbert Hunger, Die hochsprachliche profane Literatur der Byzantiner (2 vols, Munich, 1978), vol. 1�� , pp. 92–120�� and Jean-Claude Margolin, ‘La rhétorique d’Aphthonius et son influence au XVIe siècle’, in Raymond Chevallier (ed.), Colloque sur la rhétorique (Paris, 1979), pp. 239–69. 24 Ioannes Sardianos, Commentarium in Aphthonium, ed. Hugo Rabe (Leipzig, 1928); Doxapatres, Homiliae in Aphthonii Progymnasmata, in Walz, Rhetores graeci, 2. On Sardianos and Doxapatres, see Alexander Kazhdan (ed.), The Oxford Dictionary of Byzantium (New York, 1991), pp. 1,067 and 660 respectively. 25 Nikolaos, Progymnasmata, ed. Joseph Felten (Leipzig, 1913). Souda, vol. 3, 469. 19


45

informative than either Aphthonios or ps.-Hermogenes. Nikolaos claims that his aim was to compile rather than to compose a new treatise, and one senses the weight of tradition upon his shoulders when he remarks that the quantity of work left behind by the dead makes it impossible to invent anything new – all he can do is gather information.26 In fact, his compilation contains some valuable and unique passages which contribute a great deal to the overall picture of the Progymnasmata and their place in rhetorical education. Our ignorance of so many of his sources makes it impossible to say whether Nikolaos’ modesty was false or not, but in relation to the surviving sources he shows originality, in some cases making explicit an idea which was only implicit in the shorter treatises, in others reflecting new developments. In particular he is the only one to introduce works of art into the theoretical literature as a subject matter for ekphrasis (see below, pp. 81–3). Finally, there is an anonymous Byzantine treatise on the Progymnasmata, preserved in several manuscripts and printed in Walz’s Rhetores graeci, which presents some interesting developments.27 Exactly how these exercises were taught is unclear: some papyrus fragments of written compositions survive (though not, to my knowledge, any examples of a school ekphrasis), but Theon’s introduction shows that the compositions were also recited in a first step towards rhetorical performance.28 Even when written, the exercises may later have been committed to memory. So these were not just exercises in prose composition the way we would understand it. They were part of a cumulative training in the techniques and processes of rhetorical performance. In Theon’s system, in particular, texts, whether composed by the student or the teacher, were not closed literary forms, but open to endless manipulation, elaboration and discussion: students were taught to subject the content of stories, sayings, even ekphraseis, to a process of confirmation and refutation, arguing for or against their likelihood.29 The Progymnasmata were eminently practical texts, shaped to fulfil a particular role within the ancient curriculum. The surviving examples do represent a remarkably unified group, especially bearing in mind the centuries which elapsed between Theon and Nikolaos. But they were far from being a static or abstract system. The conformity which strikes the modern reader is the result of centuries of selection in which many other versions were lost along the way, such as those by several authors mentioned 26

Nikolaos, Progymnasmata, p. 1. Walz, Rhetores graeci, 3, pp. 595–8. 28 On the evidence for the use of books in education, see Cribiore, Gymnastics of the Mind, pp. 143–56. 29 See Michel Patillon, introduction to Theon, Progymnasmata, pp. xciii–xcvii; Webb, ‘The Progymnasmata as practice’. 27


46

in the Souda.30 We get glimpses of such alternative systems in some of the remarks about what ‘others’ say or do, particularly in ps.-Hermogenes and Nikolaos. And at some point the order in which the exercises were taught underwent a change. The manuscripts of all four versions preserve the same order, beginning with muthos, but Theon’s introduction shows that he originally prescribed a different order, beginning with chreia, and that the text was later rearranged to conform to accepted practice.31 But even the surviving Progymnasmata were more responsive to change than they are sometimes said to be.32 Nikolaos’ acknowledgement (albeit belated) of statues and paintings as a category of subject matter for ekphrasis is one case in point. There is also a clear shift in emphasis from Theon to Aphthonios which reflects developments in rhetorical fashion and practice over the intervening centuries, particularly the rise in importance of epideictic rhetoric. Aphthonios’ remarks about ekphrasis and other exercises are much more oriented towards the composition of epideictic speeches, while Theon, like his contemporary Quintilian, seems more interested in the type of presentation and argumentation needed in the law courts.33 Change was slow and is often difficult to perceive, but it is detectable in the details, and this is true in the case of ekphrasis (see below, pp. 78 – 80). In addition to these sets of instructions, some examples of model Progymnasmata have survived to complement the examples which Aphthonios added to his own instructions. Theon had advised teachers to compose their own models in certain cases and to use passages from classical texts,34 but the surviving examples date from a much later period and are ascribed in the manuscripts to Libanios and Nikolaos.35 Foerster identified seven of the ekphraseis as the work of Libanios, while other ekphraseis of sculptures in the collection conform quite closely to Nikolaos’ precepts, but there may be further anonymous and undatable 30

See the fragments collected in Rabe’s edition of Aphthonios, Progymnasmata, pp. 53–70. 31 See Italo Lana, I Progimnasmi di Elio Teone I: La storia del testo (Turin, 1959); Patillon, Hermogène: l’art rhétorique. Theon’s text also originally contained further exercises. 32 Jacques Bompaire, Lucien écrivain: imitation et création (Paris, 1958), p. 100 defines such scholastic material as being ‘d’aucun temps ni d’aucun lieu’ [of no particular time or place]. 33 See Italo Lana, Quintiliano, Il ‘Sublime’ e gli ‘Esercizi preparatori’ di Elio Teone: ricerca sulle fonti grece di Quintiliano e sull’autor ‘Del Sublime’ (Turin, 1951). 34 Theon, Progymnasmata, p. 15, ll. 32–6 and p. 9, ll. 30–32. 35 Libanios, Opera, vol. 8, pp. 460–546. Some Byzantine model Progymnasmata are printed in Walz, Rhetores graeci, 1. Those by Nikephoros Basilakes (which do not include any ekphraseis) have been edited separately by Adriana Pignani (Naples, 1983). See also Hunger, Die hochsprachliche profane Literatur der Byzantiner, vol. 1, pp. 92–120.


47

schoolmasters’ work in this group.36 These ‘fair copies’ have many interesting features, but as a group they reflect a very late stage in the ancient history of ekphrasis. The Progymnasmata in Relation to Other Treatises It is important to consider the Progymnasmata both as a system of interrelated exercises and as one stage in a far larger educational process. In particular, the authors of the more advanced treatises on epideictic and declamation assumed that they could appeal to their users’ knowledge of the Progymnasmata.37 The way in which they refer to this preparatory training is revealing. They do not seem to consider the exercises as a product, as a set of ready-made patches to be sewn together, but seem rather to think of them as part of a formative process which has provided students with flexible skills and with a stock of commonly accepted things to say and ways to say them.38 Sopatros the Rhetor in his work On the Division of Questions (Diareseis Zētēmatōn) refers to the ‘places’ (topoi) which the student will have learned from the exercise of koinos topos against an adulterer and which he can now adapt (enarmosai) to the speech in question (the defence of a woman who has killed her adulterous husband in a reversal of classical cultural norms).39 In his discussion of the wedding speech, the author of a treatise on epideictic attributed to Dionysios of Halikarnassos refers his addressee, a former pupil of his, to his earlier experience of writing theseis on the topic ‘should one marry?’ This exercise has made available (prokecheiristai) a stock of things to say and forms of argument that can be put to use in praise of marriage. This training is referred to as a ‘gift’ (dōron), which has the effect of ensuring that the speaker is ‘not inexperienced in the things that it is customary to say’, underlining the role of such education as the provider of a vital form

36

Bernhard D. Hebert, Spätantike Beschreibung von Kunstwerken: Archäologischer Kommentar zu den Ekphraseis des Libanios und Nikolaos (Graz, 1983), pp. 8–9. 37 It may well be the case that some individuals followed a different route before embarking on declamation, but it is hardly conceivable that they started this complex training without some preliminary training in composition. 38 �� Baldwin,� Medieval Rhetoric and Poetic, p.�� 18 conceives the relation of progymnasma to finished speech as follows: ‘Apparently a boy could carry this peacock [composed as an example of the exercise of ekphrasis] from school to the platform and continue to use it with merely verbal variations’.�� Compare Roland Barthes’ characterization of Second Sophistic rhetoric as ‘une suite lâche de morceaux brillants’ [a loose sequence of brilliant parts], ‘L’ancienne rhétorique: aide-mémoire’, Communications, 16 (1970), p. 183. 39 Sopatros, On the Division of Questions, in Walz, Rhetores graeci, 8, p. 249, ll. 20–21.


48

of cultural literacy.40 He also tells the recipient that he may keep this ‘gift’ for himself or pass it on to another as a favour for which he will, in turn, receive gratitude (charis). These remarks underline the role of the Progymnasmata in providing a flexible set of skills that prepare the student for the more demanding task of composing and performing epideictic speeches and for the immeasurably more challenging task of mastering the art of declamation. As Sopatros’ comment suggests, one of the most valuable aspects of the training offered by the Progymnasmata was the mastery of topoi, ‘places’ or kephalaia, ‘heads’ of argument which provided a basic framework for argumentation and for praise and blame. Students learned how to construct a basic speech of praise following the ‘encomiastic topoi’ such as birth, education and achievement.41 In the twin exercises of confirmation (kataskeuē) and refutation (anaskeuē) – the most important according to Aphthonios – the student learned to support or demolish a story by applying a set of questions: is it plausible (pithanon)? is it impossible (adunaton)? is it internally contradictory (machomenon)?42 This process of confirmation and refutation was one to which all the exercises could be subject, according to Theon. One could even use the same technique to demolish an ekphrasis, as well as a narration, a reminder that the exercises were an interlinked system and that each had, potentially, several levels of difficulty. So the Progymnasmata formed a system of training that was directed towards the inculcation of a set of habits and practices in the individual student and only secondarily towards the production of compositions. However, the authors of the Progymnasmata take most of this for granted. In the case of ekphrasis, with the exception of a few lines in Nikolaos, they pay no attention to the purpose of learning to ‘place a subject before the eyes’. One of the most valuable aspects of Quintilian’s treatise is that it provides the key to answering this question. His treatment of enargeia, in so far as it coincides with the treatments of ekphrasis elsewhere, shows how this type of composition had a fully rhetorical role, as an aid to persuasion. And the more advanced Greek rhetorical treatises on epideictic and declamation show how students were taught to put ekphrasis, and the other preliminary exercises, to use in the context of a full-scale speech. These clues to the rhetorical (in the technical sense) purpose of ekphrasis help to explain its inclusion among the Progymnasmata. These elementary exercises were, after all, directed towards the study of rhetoric. They aimed 40

Ps.-Dionysios of Halikarnassos, On Epideictic, 261 (translation in Menander Rhetor, On Epideictic, pp. 365–6). 41 See, for example, Aphthonios, Progymnasmata, p. 21, l. 20 – p. 22, l. 9. 42 See Theon, Progymnasmata, 93, l. 5 – 96, l. 14 with Patillon’s introductory comments on pp. xciii–xcvii.


49

to produce students who could speak effectively, or in a manner which was accepted as effective, whether in real cases, in the mock cases of the declamations, or simply in their dealings with others. Though the extant Progymnasmata textbooks frequently use examples drawn from history and poetry in their treatment of all the exercises, these are submitted to rhetorical categorization and treatment. A literary interest in description, as important as this may have been, would not have been sufficient in itself to prompt the inclusion of ekphrasis among the Progymnasmata. Nor was the composition of history or poetry the main aim of the training they offered, though it might be a by-product, an added advantage, as suggested by Theon’s remark that ekphrasis is a useful skill for the historian to master.43 One cannot imagine that the promise that their sons would learn to describe nicely as an end itself would have cut much ice with the ambitious fathers seeking to enrol them in Theon’s school. The skills they wanted their sons to acquire were rhetorical. I stress this point because the Progymnasmata are often assumed to have provided a type of generalized literary training. Theon does indeed state that his exercises are useful for writing history, poetry, or other types of literature. But this statement needs to be read against the overwhelmingly rhetorical focus of the education of which the Progymnasmata were part. The ability to speak eloquently, persuasively and appropriately, according to grammatical and rhetorical norms, was the aim of elite education. Relatively few of the students who passed through the rhetorical schools would have gone on to write history or poetry. But everyone who was able to continue their studies past the elementary stages would have gone on to learn how to argue court cases, through the exercise of declamation, and how to compose epideictic speeches. By mentioning the applicability of the training he offers to other types of composition Theon is surely trying to find as many ways as he can to make it attractive to potential customers. Ekphrasis in the Progymnasmata The Progymnasmata were therefore a system of practical exercises which were interrelated. The order in which they were taught was the subject of much discussion in the commentaries on Aphthonios. Ekphrasis is one of the last of the Progymnasmata in the order preserved in the manuscripts. It is followed only by the more discursive thesis (discussion of a question) and eisphora nomou (introduction of a law). One Byzantine commentator explains that this is because it was considered one of the more difficult 43

Theon, Progymnasmata, 60, ll. 19–22.

50


and elaborate exercises.44 But in the original sequence of Theon’s exercises, ekphrasis was the fifth exercise, coming immediately after koinos topos (common place).45 The perceived difficulty and status of ekphrasis in relation to the other exercises therefore seems to have changed over time. Exactly when this happened is unclear, but ps.-Hermogenes and Nikolaos reveal that the arrangement of the Progymnasmata, and the place of ekphrasis in particular, continued to be debated in the centuries immediately following Theon. Nikolaos simply points out that some prefer to teach ekphrasis immediately after synkrisis (comparison) because both use the aneimenos (relaxed or simple) style.46 The question raised by ps.-Hermogenes is far more substantial. In his treatment of both synkrisis and ekphrasis he notes that certain (unnamed) authors prefer not to treat these two as separate exercises on the grounds that they can be subsumed within other Progymnasmata. Synkrisis could be included within koinos topos, enkōmion (praise) and psogos (blame), while ekphrasis had a still wider application, being contained within muthos, diēgēma and koinos topos as well as enkōmion. Such comments remind us both how interconnected the various exercises were and how flexible ekphrasis was: the techniques it taught could be put to use in a variety of contexts. In particular, the connections between ekphrasis and the four exercises mentioned by ps.-Hermogenes are worth exploring as they can reveal a great deal about how ekphrasis and its role were conceived. This question will be addressed in the next chapter; for the moment we will consider the presentation of the exercise of ekphrasis and some of the many questions it raises. The Chapters on Ekphrasis The chapters on ekphrasis follow a standard pattern: the one-line definition is followed by a list of categories of subject matter and then, in accordance with Theon’s advice to the teacher to select examples from literature, the authors go on to illustrate each category of ekphrasis by citing passages from classical texts. They then add some general remarks about linguistic style and how to link ekphrasis into a larger context. Aphthonios ends 44 Anonymous scholia to Aphthonios, Progymnasmata, in Walz, Rhetores graeci, 2, p. 55 (text and translation in Appendix A). Aphthonios’ order is: muthos, diēgēma, chreia, gnōmē, anaskeuē, kataskeuē, koinos topos, enkōmion, psogos, synkrisis, ēthopoiia, ekphrasis, thesis, nomou eisphora. He differs from ps.-Hermogenes and Nikolaos in presenting the pairs anaskeuē and kataskeuē and enkōmion and psogos as separate exercises. 45 Theon simply calls this exercise topos but I will use the more usual term throughout for the sake of clarity. 46 Nikolaos, Progymnasmata, p. 67, l. 16 – p. 68, l. 5. The short treatise on this style, Romanos, Peri aneimenou, ed. Walter Camphausen (Leipzig, 1922), p. 2, ll. 19–21, also associates it with ekphrasis.


51

with his model ekphrasis of the Alexandrian acropolis; Nikolaos concludes with advice about the use of ekphrasis within the three genres of oratory: judicial, deliberative and epideictic. The Definition The definition varies little between the four sets of instructions. As we saw in the introduction, Theon’s version reads as follows: ἔκφρασίς ἐστι λόγoς περιηγηματικὸς ἐvαργῶς ὑπ’ ὄψιv ἄγωv τὸ δηλoύμεvov [Ekphrasis is a descriptive (periēgēmatikos) speech which brings (literally ‘leads’) the thing shown vividly (enargōs) before the eyes].47 Ps.-Hermogenes and Aphthonios repeat this definition almost word for word, with some minor variations, ps.-Hermogenes adding a diffident ‘as they say’, as he does elsewhere in his Progymnasmata.48 Only Nikolaos changes the vocabulary slightly, using the adjective aphēgēmatikos instead of periēgēmatikos. This phrase, like the other definitions, was probably learned by rote and recited in response to the master’s questions,49 but what did it mean? One of the most striking features of this definition, as was noted above, is the role of the audience. Where the other Progymnasmata are defined in far more formal terms (diēgēma, for example, is a ‘speech which sets out events which have happened or are of the sort which might happen’) ekphrasis is defined primarily in terms of its effect on the listener. This exercise therefore drew the students’ attention more explicitly than did the others to the communicative function of rhetorical discourse and to the live interaction between speaker and audience which it supposed. As we have seen, the idea of ‘placing before the eyes’ goes back in rhetorical theory to Aristotle who discusses the power of metaphor to place its subject pro ommatōn,50 and the definition makes ekphrasis synonymous with other rhetorical terms for vivid description, such as enargeia, diatupōsis, diagraphē (‘descriptio’, or ‘explicatio’ in Latin, or simply ‘sub oculos subiectio’, ‘placing before the eyes’). In some contexts, rhetoricians draw fine distinctions between the meanings of the different terms for vivid language.51 Elsewhere, however, they may be used interchangeably 47 For the moment I have translated periēgēmatikos as ‘descriptive’; its literal meaning, however, is ‘leading around’. For further discussion, see below. 48 Hermogenes, Progymnasmata, p. 22, l. 10; cf. p. 4, l. 6 (diēgēma). 49 �� An example of this question and answer format is preserved on a papyrus fragment. See Aphthonios, Progymnasmata, pp. 52–5.� 50 Aristotle, Rhetoric, 1411b 24–5. 51 The account of the destruction at Phokis in Demosthenes, 19.65, for example, is defined as a diatupōsis, which is then contrasted to ekphrasis, in the Scholia Demosthenica, vol. 2, p. 28 (157c), but cited as an example of ekphrasis by Nikolaos (see below, p. 76). On the fluidity of ancient rhetorical terminology, which is not surprising given the ad hoc

52


in the same text, as in the case of Menander Rhetor (see Chapter 6) or given identical definitions.52 Ekphrasis, though, is always the term used in Greek sources for the elementary exercise of ‘placing before the eyes’, a distinction which may confer a slightly more scholastic flavour on the term in some contexts. The rest of the definition offers little concrete explanation; even the type of communicative situation and the nature of the audience are left vague. Instead – and this is another of the striking features – the authors resort to purely metaphorical language to define their subject: ekphrasis is a speech which ‘leads [the audience] around’; it is a form of language which achieves the linguistically impossible, appealing to the sense of sight, and bringing the referent into the presence of the audience. In the body of their chapters the authors make similar claims: the audience should ‘almost see’;53 one should try to make listeners into spectators.54 The language used to define ekphrasis clearly puzzled later readers of Aphthonios’ version of the Progymnasmata. Sardianos explains patiently that periēgēmatikos is to be taken metaphorically to mean relating (aphēgoumenos) and displaying (deiknus) everything about a subject.55 His choice of the verb aphēgeomai to gloss periēgēmatikos may well have been influenced by Nikolaos’ alternative adjective, which expresses the same idea of ‘leading’ and ‘guiding’. With refreshing lucidity, Sardianos also points out that the visual effect of ekphrasis is in the mind: it makes people conceive of the subject in their mind (noein) and produces a tupos and phantasia, or mental impression, because for language literally to ‘bring anything before the eyes’ is clearly impossible. As he goes on to explain: ‘even if the speech were ten thousand times vivid (muriakis enargēs) this “bringing the thing shown, i.e. described, before the eyes” would be an impossibility’.56 Sardianos’ comment also identifies enargeia, ‘vividness’, as the key to the definition of ekphrasis. For Theon, too, it is and personal nature of ancient education, see Jean Cousin, Etudes sur Quintilien (2 vols, Amsterdam, 1967), vol. 2, p. 2. 52 Tiberius, De figuris Demosthenicis, 43 defines diatupōsis as ‘leading the subject before the eyes’ (epi tēn thean) in a close echo of the language used to define ekphrasis in the Progymnasmata. 53 Theon, Progymnasmata, 119, ll. 31–2: ἀρεταὶ δὲ ἐκφράσεως αἵδε․ σαφήvεια μὲv μάλιστα καὶ ἐvάργεια τoῦ σχεδòv ὁρᾶσθαι τὰ ἀπαγγελλόμεvα. 54 Nikolaos, Progymnasmata, p. 68, ll. 11–12: [ἔκφρασις] πειρᾶται θεατàς τoὺς ἀκoύovτας ἐργάζεσθαι. 55 Sardianos, Commentarium, p. 216, ll. 13–16: μεταφoρικῶς oὖv καὶ ὁ λόγoς, ὁ πάvτα ἑξῆς καὶ τὰ τoῦ πράγματoς καὶ τὰ τoῦ πρoσώπoυ ἀφηγoύμεvoς <καὶ> μετὰ ἀκριβείας δεικvὺς περιηγηματικὸς ὀvoμάζεται. See also Appendix A. 56 Ibid., p. 216, 22–4: κἂv γὰρ μυριάκις ἐvαργὴς εἴη ὁ λόγoς, ἀδύvατov αὐτὸ κατ' ὄψιv ἀγαγεῖv τὸ δηλoύμεvov ἤτoι ἐκφραζόμεvov. See also Appendix A.


53

enargeia which makes one ‘almost see’ the subject, and for Nikolaos this vividness is the distinguishing characteristic of ekphrasis.57 The author of a Greek handbook of the Roman period, nicknamed ‘Anonymous Seguerianus’ by modern editors, in fact defined enargeia as a type of speech using exactly the same formula as the Progymnasmata textbooks used to define ekphrasis: λόγoς ὑπ' ὂψιv ἄγωv τὸ δηλoύμεvov (‘a speech which brings the subject matter [or, more precisely, ‘the thing shown’] before the eyes’).58 Ekphrasis is therefore the exercise which taught students how to use vivid evocation and imagery in their speeches. Enargeia itself raises further questions which will be dealt with in Chapters 4 and 5 (as is evident already, authorities differ in whether they define it as a type of speech, or a quality of speech, as in Quintilian). The adjective enargēs originally meant ‘clearly visible’, so its later rhetorical use to designate speech which appeals to the mind’s eye is itself metaphorical.59 This constant resort to metaphor in the discussions of ekphrasis suggests that the ideas evoked exceeded in complexity the technical language available to express them. But it may also indicate something about ekphrasis itself: that it is an effect which transcends categories and normal expectations of language. An awareness of the inadequacy of these metaphors of sight and vision (an indication that they are not, as Sardianos explains, to be taken literally) is evident in the frequent use of disclaimers such as ‘almost’ to introduce them. In Theon, the listeners ‘almost’ (schedon) see the subject;60 in Nikolaos they ‘all but’ (mononou) become spectators (theatai).61 The language of illusion, approximation and semblance is deeply embedded in the discussions of ekphrasis. If the use of metaphor is significant in itself, so are the associations conjured up by the various images which the authors of the Progymnasmata draw on. The analogy which springs to mind most readily is that of the visual arts: language which ‘places before the eyes’ is comparable to painting, as Sardianos notes.62 The connection between ekphrasis and the idea of visual representation thus runs deep and is part of its very essence. 57

Nikolaos, Progymnasmata, p. 68, ll. 9–10 and p. 70, ll. 3–6. Anonymous Seguerianus, Art of Rhetoric, 96, p. 26. 59 Pierre Chantraine, Dictionnaire étymologique de la langue grecque: histoire des mots (Paris, 1968), s.v. ἐvάργης; Barbara Cassin, ‘Procédures sophistiques pour construire l’évidence’, in Carlos Lévy and Laurent Pernot (eds), Dire l’évidence: philosophie et rhétorique antiques (Paris, 1997), pp. 15–29; Alessandra Manieri, L’immagine poetica nella teoria degli antichi: phantasia ed enargeia (Pisa, 1998), pp. 113–22. 60 Theon, Progymnasmata, 119, ll. 31–2. 61 Nikolaos, Progymnasmata, 70, ll. 5–6: μovovoὺ θεατὰς εἶvαι [παρασκευάζει]. 62 Sardianos, Commentarium in Aphthonii Progymnasmata, p. 217, ll. 3–5. 58


54

But this is just one of the analogies used by the rhetoricians in their attempts to express the effects of ekphrasis. The words hup’opsin agōn (‘leading before the eyes’) suggest the dramatist who literally produces characters and actions on stage, placing them before the eyes of the audience. And the term used by Nikolaos to express the transformed role of the audience of ekphrasis, theatai, more obviously refers to the spectators in a theatre than to viewers of a work of art.63 Theatrical imagery is frequently used elsewhere of vivid language, as in the scholia to the Shield episode in Iliad, 18 where Homer is said to ‘roll out (ekkukleō) the maker [Hephaistos] as if onto a stage and show us his workshop in the open’.64 The adjective periēgēmatikos, by contrast, casts the speaker as a guide showing the listener around the sight to be described, as Pausanias leads the reader around Greece in his Periegesis (which is not mentioned by any of the rhetorical sources). The analogy between a speech and a journey in which the speaker leads the audience through space is frequent in the Greek vocabulary of discourse, in terms such as diēgēsis (‘telling, narrating, setting out’) – literally, a ‘leading through’ the subject – or periodos for the circular journey of the periodic sentence.65 Periēgēsis would therefore suggest a more elaborate form of telling, a winding path instead of the direct through-route of diēgēsis. Sardianos, before offering the interpretation quoted above, explores the implications, explaining that it is like (hoionei) taking a visitor around the city of Athens, showing and commenting on the points of interest. While his ultimate interpretation of periēgēmatikos makes it merely another way of saying ‘showing in words’, the tenor of the metaphor adds to the composite picture of ekphrasis: the guide not only ‘shows’, but directs his or her audience’s attention, adding order and meaning to the undifferentiated mass of sights which is presented to the visitor. Ekphrasis, in some cases, therefore does not only make ‘visible’ the appearance of a subject, but makes something about its nature intelligible, an idea which is encompassed by the verb dēloō which can mean to explain, to reveal to the intellect, as well as to show. Drawn as they are from different domains, these metaphors all suggest slightly different relationships between speaker, addressee and referent: the subject matter may be ‘brought’ into the presence of the audience 63

The formula may derive from Isocrates, who claims at To Nikokles, 49 that drama, in contrast to Homeric epic, makes the audience not only listeners but spectators (theatai). If so, the later rhetorical tradition uses the contrast in a very different way. 64 Scholia graeca in Homeri Iliadem to 18.476–7: δαιμovίως τὸv πλάστηv αὐτὸς διέπλασεv, ὥσπερ ἐπὶ σκηvῆς ἐκκυκλήσας καὶ δείξας ἡμῖv ἐv φαvερῷ τὸ ἐργαστήριov. 65 See, for example, Quintilian, Institutio oratoria, 10.7.23: speech is like a journey out from a harbour.


55

(speaker as theatrical producer), or the audience may be ‘led around’ the subject (speaker as tour guide). The individual metaphors are hardly definitive statements, but represent attempts to express various aspects of ekphrasis to an audience of students. One thing they share in common is the emphasis on vision. They also insist on the agency of the speaker as performer or producer, whether he is leading his audience on a virtual tour or conjuring up the subject matter on a virtual stage. The models of communication which these metaphors evoke will come in and out of focus throughout this study of ekphrasis in the rhetorical tradition, and may on occasion become more or less concrete. The metaphor of periēgēsis, for example, becomes an organizing principle in many of the city descriptions in epideictic speeches, as in Aphthonios’ own model ekphrasis of the Alexandrian acropolis which is structured as a tour around the building. Nikolaos’ choice of the alternative adjective, aphēgēmatikos, is intriguing. As noted above, it retains the idea of the speaker as guide. Nikolaos uses the noun aphēgēsis in his chapter on ekphrasis in a sense that is close to diēgēsis, narration, when he distinguishes ekphrasis from a ‘plain narration’ (psilē aphēgēsis).66 Elsewhere in his Progymnasmata Nikolaos uses the adjective once to distinguish a type of narration (diēgēsis) that is told in the narrator’s persona (his example is Pindar) as opposed to a dramatic narration told by a character, as in comedy and tragedy.67 The sense is thus close to Plato’s definition of diēgēsis in the Republic 392c–394d where it is contrasted to the quotation of direct speech (mimēsis). In the context of the rhetorical definition of ekphrasis, what is of interest here is the apparent emphasis on the speaker’s role as origin of the account. Composing Ekphraseis The authors of the Progymnasmata then move on to the subjects for ekphrasis (see Table 2.1 and Appendix B). These are a varied list including human figures, seasons, events such as battles, places and animals. As mentioned above, works of art (paintings and statues) are mentioned by Nikolaos, but are not central to the definition. The significance of the subjects and, in particular, the links that they indicate between ekphrasis and the other exercise will be explored in detail in the next chapter. For the moment, 66 Nikolaos, Progymnasmata, 70, l. 3. The distinction between ekphrasis and narration will be explored in the next chapter. 67 Nikolaos, Progymnasmata, p. 12, ll. 9–11. Perhaps under the influence of Nikolaos’ use of the term in his chapter on ekphrasis, aphēgēsis is defined in a comment on Aphthonios as a diatupōsis – i.e., a type of vivid writing in contrast to diēgēsis which states the cause. See Walz, Rhetores graeci, 2, p. 578, ll. 2–3: διήγησις μέv ἐστι μετὰ τῶv αἰτιῶv, ἀφήγησις δὲ ἡ τoῦ πράγματoς διατύπωσις. See further below.

56


it is sufficient to note the sheer variety in the subjects proposed before we move on to the rest of the discussions. However, in contrast to the richness of the tantalizing metaphors contained within the definitions of ekphrasis, these sections are far from illuminating, particularly with regard to how and why language should ‘place before the eyes’. The practical advice can be resumed in a few sentences. In terms of organization, the authors recommend starting from first things and working through to the last, so that in the case of a figure one should move methodically from head to toe or, in the case of an event, start from the events which led up to an action and follow up with the consequences.68 Libanios’ model ekphrasis of a battle follows this pattern to the letter, beginning with the diplomatic prelude to hostilities (in which each city’s rhetors naturally play a vital role) and ending with the celebrations in the victorious city and retribution in the defeated one.69 Aphthonios is the only one to give advice about the description of places. Despite the fact that his model ekphrasis is of a place, the Alexandrian acropolis, his instructions are brief to the point of obscurity. In describing places, as well as periods of time, one should include the surroundings (ta periechonta) and contents (ta en autois huparchonta). Some indication as to what this means can perhaps be obtained from Libanios again: his ekphrasis of Spring lists the activities and sensations which occur in that season.70 Further, Theon and ps.Hermogenes note that one should take care to link the ekphrasis into the surrounding context. Nikolaos’ comments on ekphraseis of statues and paintings are among the more interesting passages and will be discussed below. Table 2.1

The subjects of ekphrasis

Author

Subjects for ekphrasis

Theon

Events, persons, places, times, the manner in which something is done (tropos)

ps.-Hermogenes

Persons, events, places, states of affairs (kairoi), times

Aphthonios

Persons, events, seasons, places, mute animals and plants

Nikolaos

Places, seasons, persons, festivals, events, paintings and statues

68 Aphthonios, Progymnasmata, p. 37; cf. Nikolaos, Progymnasmata, p. 69, ll. 12–17 (on figures in art). 69 Libanios, Progymnasmata, pp. 460–64. 70 Ibid., pp. 479–82.


57

The Language of Ekphrasis With respect to language, the medium of ekphrasis, we are told that the style should be relaxed or simple (aneimenos). Aphthonios adds the advice to use a variety of figures and to ‘imitate (apomimeisthai) the thing described completely’. In the light of ps.-Hermogenes’ statement that the style of the language should be similar to (sunexomoiousthai) the subject matter,71 it seems most likely that Aphthonios means that the qualities of the language should somehow reflect those of the subject. Sardianos interpreted Aphthonios’ phrase as a reference to style and, like ps.Hermogenes, he goes on to explain that the language should share the characteristics of the subject matter, so that if one is describing a meadow the language should be flowery, and so on.72 Dionysios of Halikarnassos’ criticism of Philistos for using language which was inappropriate to the subject matter shows how important the fit between language and subject matter was felt to be. In other sources the use of onomatopoiia is said to be a cause of vividness.73 But it is less than clear that our sources have any objectively observable stylistic criteria in mind. In fact, it seems rather that such judgements were often based on the referent, so that the presence of terms referring to flowers and related things would suffice to make a ‘flowery’ style. A passage in the real Hermogenes’ treatise On Types of Style where ‘sweetness’ (glukutēs) in language is said to result from the quality of the subject (p. 331, l. 5ff) rather suggests that the latter may have been the case, particularly as several of the examples he cites are ekphraseis of things that are pleasant to the senses. These examples are Sappho, fr. 4 (‘Around the apple branches flows cold water’), the opening of Plato’s Phaedrus and the account of the miraculous growth of soft grass beneath Hera and Zeus in Iliad, 14.347–8.74

71 Ps.-Hermogenes, Progymnasmata, p. 23, ll. 11–12: ἔτι μέvτoι συvεξoμoιoῦσθαι τὰ τῆς φράσεως ὀφείλει τoῖς πράγμασιv. See Michel Patillon’s comment on ps.-Hermogenes in Hermogène: l’art rhétorique (Paris, 1997), p. 148, n. 6: ‘Notre auteur se contente de métaphores et ne nous apporte pas d’indications techniques sur les procédés du style …’ [Our author is content to use metaphors and does not provide any technical information on the use of style …]. 72

Sardianos, Commentarium, p. 225. An unusually ironic scholiast in Walz, Rhetores graeci, 2, p. 652 notes ‘I suppose one ought to grunt (grulizein) when talking about piglets’. The author of the Byzantine Progymnasmata in Walz, Rhetores graeci, 3, p. 598, ll. 1–6 recommends using harsh, onomatopoeic vocabulary in an ekphrasis of a battle; his example includes terms with a predominance of kappa, gamma and chi such as smaragē, klangē, chremetismos, kraugē. Cf. Anon. Peri tōn tessarōn merōn tou teleiou logou in Walz, Rhetores graeci, 3, p. 579, l. 29 – p. 580, l. 10. 74 Hermogenes, On Types of Style, 331–2. 73

58


Ps.-Hermogenes puts a particular stress on the impact of the language of ekphrasis when he claims that the hermēneia must almost bring about (schedon mēchanasthai) sight through sound. The sense of hermēneia here is that of style, as the expression of thought. Language translates (hermēneuei) thought into words, and, in the case of ekphrasis, it translates the visible into words which somehow communicate a visual experience to the audience. As the further connotations of hermēneia suggest, this translation involves a certain degree of interpretation, so that ps.-Hermogenes (who might have phrased his statement differently if he had been concerned with all these implications) does not imply a transparent, unmediated communication of a sight. As we shall see, and as is already hinted at by Nikolaos’ use of aphēgēmatikos, the speaker’s role in the description is key, and when ekphraseis describe subject matter which exists outside the text (ekphraseis of buildings, for example) the selection of details and the speaker’s explicit comments do add a great deal of interpretation and demand some interpretative effort from the listener in turn. These laconic statements on the language of ekphrasis leave many questions to be answered, but they do point to the intimate relation between words and their referents, and to the role of the speaker, all of which will be explored in more detail below. Within the context of the Progymnasmata, however, the remarks on language seem less than illuminating. In the ancient and Byzantine schoolrooms, these written instructions must have been only a small aspect of the whole teaching of ekphrasis, or the other Progymnasmata. The commentaries give an idea of the type of verbal supplements provided by teachers. But the most important method of teaching must have been from examples. Theon is very clear about the value of examples, whether drawn from classical literature or composed by the teacher. In both cases the students are expected to internalize them so that they can then produce their own imitations (mimeisthai). However, even the examples of ekphrasis chosen from classical literature present no formal unity and vary as greatly in treatment as they do in subject matter. Some, such as the descriptions of Thersites (Iliad, 2, 217 and 219) and Eurybates (Odyssey, 19, 246) used to illustrate ekphraseis of people, take up only one or two lines of epic verse, while others are of considerable length (like the Shield of Achilles or the night battle in Thucydides, 7). Some, such as the night battle, have obvious dramatic appeal, while others, such as the description of the harbour of Cheimerion in Thucydides, 2 or Ekbatana in the first book of Herodotos’ histories, appear to the modern reader to be dry accounts of layout (schēma).75 75 On the night battle, see Andrew D. Walker, ‘Enargeia and the spectator in Greek historiography’, TAPA, 123 (1993): 353–77.


59

Conclusion Read in isolation as disembodied fragments of doctrine, the Progymnasmata are less than illuminating. Their lack of practical recommendations is startling: it is difficult to see how one would arrive at anything approaching Aphthonios’ model exercises on the basis of his precepts alone. It is not surprising, then, to read Philippe Hamon’s statement that classical rhetoric is little help in defining description, or to find critics such as Palm rejecting their definition of ekphrasis altogether as inadequate.76 But these elementary treatments of ekphrasis were never intended as isolated definitions. They were an integral part of a far wider network of ideas and practices. Though the Progymnasmata may not offer anything approaching a clearly articulated theory of description, they represent the elementary introduction to a notion of representation in language, which becomes clearer from the cumulative evidence of other rhetorical sources, particularly in their treatment of enargeia. But, before considering the exercise of ekphrasis as part of a wider rhetorical system, it is also worth examining its place within the smaller system of the elementary exercises themselves. One way to explore the relation of ekphrasis to the other Progymnasmata and to the system of ancient rhetoric in general is through the range of subject matter proposed. There may be no single defining type of subject matter for ancient ekphrasis but, taken as a group, the categories named are significant in themselves.

76

Philippe Hamon, ‘Qu’est-ce qu’une description?’, Poétique, 112 (1972): 465; Jonas Palm, ‘Bemerkungen zur Ekphrase in der griechischen Literatur’, Kungliga Humanististiska Vetenskapssamfundet i Uppsala, Årsbok (1965–66): 116.


3. The Subjects of Ekphrasis

Introduction The most striking aspect of the subjects for ekphrasis for the modern reader is the variety: we find battles, crocodiles, cities, buildings, people and festivals. However, the lists of subjects offered by the authors are by no means as random as they may appear to be at first sight and many of their constituent elements are familiar from other exercises and from elsewhere in the rhetorical system. They thus provide important clues to the relationship between ekphrasis and other exercises. To begin with the earliest version of the Progymnasmata, Theon identifies as potential subjects persons (prosōpa), events (pragmata) such as a battle, a plague or an earthquake, places (topoi) such as a harbour or a city, and times (chronoi) such as the seasons, and also the manner (tropos) [in which something is done]. The later authors have some variations on Theon’s list, as is clear from the List of Subjects (Appendix B) and Table 2.1, the greatest difference being the word used for ‘times’ – which is sometimes chronoi and sometimes kairoi. There is, as usual, no consistency in usage, but a general distinction is made between the regular rhythms of nature (the seasons) and culture (festivals) on the one hand and man-made circumstances (war and peace) or temporary and unpredictable states of affairs (famine) on the other. Elsewhere there is also a degree of fluidity in the categorization: subjects which are subsumed under one category in one author (such as animals or festivals) appear as independent categories in another. Nikolaos feels free to add a discussion of types of subject (paintings and sculptures) which do not even appear in his main list. That classification by subject is not of crucial importance is also suggested by the way in which Theon simply mentions examples of ekphrasis in the prologue to his work which do not correspond neatly to his list of subjects, without making any reference to category. The model ekphraseis by Aphthonios, Libanios and Nikolaos only increase the sense of variety in subject matter, though, unlike the examples from classical texts, they are all relatively substantial in length. Aphthonios’ example is a description of the Alexandrian acropolis (corresponding to the category of ‘place’). Some of the examples in Libanios’ corpus show a closer correspondence to the various instructions. They include ekphraseis of battles on land and sea, a harbour, a season (the ekphrasis of spring) and a festival

Theon, Progymnasmata, 118, ll. 9–11: γίvεται δὲ ἔκφρασις πρoσώπωv τε καὶ πραγμάτωv καὶ τόπωv καὶ χρόvωv … αἱ δὲ καὶ τρόπωv εἰσὶv ἐκφράσεις. Theon, Progymnasmata, 68, ll. 7–22.


62

(the Kalends) as well as a hunt, a drunken man and three paintings showing landscapes with figures and buildings and a mythological scene. There is a temple of Tyche and an intriguing ‘ekphrasis of beauty (kallos)’, in which the narrator describes the effect on him of the sight of a girl at her window. The rest of the models, attributable to Nikolaos or another, are mainly of statues representing mythological persons (goddesses, Herakles, a Trojan woman) or beasts (such as the Chimaira). (These statue ekphraseis are appropriately attributed to Nikolaos and follow his advice on the description of statues; they also provide a means of describing the type of poetic, mythological themes which are typical of the other model Progymnasmata and a sign of their place in the curriculum between poetry and rhetoric.) In addition to the principal categories, the authors recognize a further category of ‘mixed ekphrasis’, such as the night battle in Thucydides (7.43– 4) which is both an ekphrasis of an action (the battle) and of a time (night). This ‘mixed ekphrasis’ may look like another example of a rhetorician’s mania for classification, but it does reflect a recognition that classification by subject is not the most important feature of ekphrasis and that there will always be examples which cannot be reduced to fit a neat schema. Hermogenes hints at this openness when he finishes his list of subjects by adding ‘and many other things’, as does Nikolaos when he simply begins to tell his readers how to write ekphraseis of paintings and statues. There was clearly space for ekphraseis of all types of subject matter, including works of art; the categories are neither exhaustive nor exclusive, and are very different from the rigid classifications of types of description set out in neoclassical textbooks from the renaissance to the nineteenth century. Instead, they point to the place of ekphrasis in the wider system of ideas about rhetorical composition reflected in the Progymnasmata, and in particular represent the interface between ekphrasis and the exercise of narration. Ekphrasis and Narration: The Significance of the Subjects Amid this variety and the apparently constant reorganization and substitution, four categories of subject matter emerge as stable elements: persons (prosōpa), places (topoi), times (kairoi or chronoi) and events

Foerster in Libanios, Progymnasmata (Opera, vol. 8), pp. 438–9 accepts only the ekphraseis of the battle, the three paintings, the Kalends, the drunken man and spring as by Libanios. Aphthonios, Progymnasmata, p. 37, ll. 17–20 (the night battle in Sicily); Hermogenes, Progymnasmata, p. 22, ll. 15–18 (identified by Rabe as Thucydides, 3.22 – the attempt to break out of the besieged city of Plataia by night). The anonymous scholia to Aphthonios in Walz, Rhetores graeci, 2, p. 57, ll. 8–11 note that the Byzantine scholar Geometres also classified ekphraseis of sea battles as mixed, since the sea is a place.

On these, see Jean-Michel Adam, La Description (Paris, 1993), pp. 34–7.

The Subjects of Ekphrasis

63

(pragmata). The rhetor’s students would easily have recognized this group as four of what Theon calls ‘elements of narration’ (stoicheia tēs diēgēseōs): the ‘who, what, when and where’ of the ancient rhetorical schools (see Table 3.1). When they learned how to compose a simple narration, or to analyse a poetic narrative, students were taught to break down the story into the action (pragma) performed by the person (prosōpon), the place (topos) in which it occurred and the time (chronos) when it occurred as well as the manner (tropos) in which the action was carried out and its cause (aitia). One function of these ‘elements’ was to help students to organize their own narrations. The immediate application was to the composition of the narrative section of a speech in which the speaker had to set out clearly his own version of events. However, there were clearly applications to other genres – many of Theon’s examples of diēgēsis are drawn from historiography (Thucydides, Philistos, Herodotos) and Homeric epic, as well as from orators such as Demosthenes. More importantly, these ‘elements of narration’, also known as peristaseis or peristatika (Latin, circumstantia), represented the conceptual framework within which the speaker could organize the complexities of the events in question (Patillon aptly terms them the ‘micro-universe’ of the rhetoricians.) As one Latin rhetorician explains, they were the means by which students of rhetoric were taught to impose intellectual order onto the mass of material which faced them in a legal case. This statement brings out clearly the role of the peristaseis as a conceptual grid, a pattern for organizing experience and verbal accounts of that experience. The reader of the Progymnasmata who came to ekphrasis well versed in the doctrine of the peristaseis would therefore immediately recognize ‘persons, places, times and events’ as rhetorician-speak for ‘practically everything’ (regardless of how easily individual examples fitted in). One conclusion Theon does however use the Odyssey as an example in his chapter on diēgēsis, showing that a continuity was seen between the technical narration of judicial rhetoric and narration in the broader literary sense. Patillon in Theon, Progymnasmata, p. xlv. See Heinrich Lausberg, Handbook of Literary Rhetoric: A Foundation for Literary Study, trans. Matthew T. Bliss et al. (Leiden, 1998), p. 139. Lausberg quotes Fortunatianus, Ars Rhetorica, 2.1, in Rhetores latini minores, p. 102, l. 21 – p. 103, l.2, who identifies the circumstantia (person, deed, cause, time, place, manner, material/equipment) as the divisions which the student should use to analyse the case once he has considered it in general: ‘prius universam causam confuse considerare debemus, tunc omnia, quae reperta sunt, capitulatim quaestionibus ordinare’ [First we should consider the overall case in no particular order, then we should organize everything we have found under headings, using these questions]. The peristaseis refer both to elements of discourse and to their referents: the pragma can be both an action and a verbal account of that action. For an example of the use of the peristaseis in the criticism of literary texts, see ps.-Plutarch, On the Life and Poetry of Homer, 76.

64


we can draw from this is that the subjects of ekphrasis are unlimited; the other is that the subjects named by the manuals, far from constituting a random selection that can be altered at will, are part of the wider system of rhetoric. Table 3.1

Comparison of the subjects for ekphrasis with the parts of narration and the subjects for enkōmion in the surviving Progymnasmata

Author Theon

ps.-Hermogenes

Aphthonios

Nikolaos

Subjects for ekphrasis Events, persons, places, times, tropoi Persons, events, places, states of affairs (kairoi), times Persons, events, seasons, places, mute animals and plants Places, seasons, persons, festivals, events, paintings and statues

Parts of narration

Subjects for enkōmion

Person, event, place, time, manner (tropos), cause

[Persons and inanimate objects/ abstracts]

–––––––––––––––

Persons, abstract entity (pragmata), mute animals, plants, mountains and rivers

Person, event, place, time, manner (tropos), cause

Persons, events, seasons, places, mute animals and plants

Person, event, place, time, manner (tropos), cause.

Persons, abstract or concrete entities pragmata

Theon mentions two further elements of narration in his discussion of the exercise diēgēsis: the manner (tropos) in which the deed was done (how) and the cause (aitia) (why).10 This explains the presence of Theon’s unique category of ekphrasis of the tropos.11 The way in which he introduces 10 Other discussions of the circumstantia or peristaseis mention further materia/hulē (e.g. Anonymous Scholion in Walz, Rhetores graeci, 7, 2, p. 921, l. 4) and facultas (= quibus adminiculis; see, e.g., Augustine, De rhetorica, 5.7, in Rhetores latini minores; Marius Victorinus, Explicationum in Rhetoricam M. Tullii Ciceronis libri II, in ibid., p. 207, ll. 1–2) which do not feature directly in the discussions of ekphrasis. 11 George Kennedy, Progymnasmata: Greek Textbooks of Prose Composition and Rhetoric (Atlanta, 2003), p. 46 translates tropōn … ekphraseis as ‘ecphrases of objects’. The importance


65

it (‘and there are also ekphraseis of the tropos’) suggests that it was an answer to an anticipated question in the mind of the reader who, hearing the first four peristaseis mentioned, might well have been wondering about the remaining two. None of the Progymnasmata texts makes any reference to the cause (aitia) in the discussion of ekphrasis (with the important exception of Nikolaos’ discussion of ekphraseis of statues and paintings, discussed below). But it is mentioned by an anonymous Byzantine commentator on the Progymnasmata who notes ‘all the peristaseis can be the subject of ekphrasis, except the cause (aitia)’.12 The reasons for its absence are explained by Sardianos who points out that ‘a cause is never apparent by itself but [becomes apparent] from the action’.13 Indeed, an abstract notion like cause hardly lends itself to being ‘placed before the eyes’. Hermogenes, in his discussion of a famous ekphrasis of a storm by the second-century star sophist Aristeides, points to a further motivation when he notes that the omission of the cause in ekphraseis of natural phenomena lends solemnity (semnotēs) to a discourse.14 An ekphrasis was concerned, by definition, with perceptible phenomena and their effects. In the case of the storm, the phenomena are all the more awe-inspiring if their cause is unexplained, or left open to interpretation. In fact, as we will see, many ekphraseis in oratory serve to point towards the cause of the state of affairs described, demanding that the audience fill in the missing peristasis for themselves, an example of how the Progymnasmata reflect only an elementary level of a far more complex practice. The standard set of subjects recommended for ekphrasis are therefore not random, but belong to a system for analysing events and their verbal representation, and provide the point of contact with the practice of narration. This close connection between the subjects of ekphrasis and the elements of narration explains why the authorities mentioned by ps.Hermogenes recommended teaching ekphrasis as a part of the exercise of narration, rather than as a separate exercise. The individual elements of a narration could be expanded by means of ekphrasis for any element or combination of elements of a story could be narrated ekphrastically, that is to say with the vividness necessary to appeal to the audience’s imagination. of the manner emerges in his translation of the rest of the sentence but the choice of ‘objects’ to render this first use of tropos shifts the focus away from the action as a subject of ekphrasis. 12 Anonymous scholion to Aphthonios, Progymnasmata, in Walz, Rhetores graeci 2, p. 55: ἰστέov δὲ ὅτι πάvτα τὰ περιστατικὰ ἐκφράζεται, πλὴv τῆς αἰτίας. 13 Sardianos, Commentarium in Aphthonii Progymnasmata, p. 217, ll. 22–3: oὐδέπoτε γὰρ αἰτία καθ’ ἑαυτὴv ἀλλ’ ἐκ τoῦ πράγματoς φαίvεται. Some examples of the way in which the cause could be implicit in the event or action described will be discussed in Chapter 6. 14

Hermogenes, On Types of Style, 244–5. See also the definition of aphēgēsis cited above (Chapter 2, n. 67) in which diatupōsis is contrasted to diēgēsis which is specifically described as being meta tōn aitiōn (with, or including, causes).

66


The commentators on Aphthonios discuss this, pointing out that ekphrasis provides practice in a special form of narration which is more elaborate.15 They adopt a tripartite classification of types of narration which is set out in the handbook On Invention, attributed to Hermogenes in which the plain (haplē) narration is distinguished from the elaborate (endiaskeuos) and the confirmatory (enkataskeuos).16 The second of these, the elaborate, is associated with ekphrasis. One anonymous Byzantine treatise on rhetoric describes endiaskeuos diēgēsis as ‘ekphrastic’ and ‘all but placing the actions before the eyes’. The treatise is particularly interesting in that its author provides examples of each type, retelling the story of Ajax’s slaughter of the Greek cattle in each mode. The ‘ekphrastic’ endiaskeuos version is told with attention to the appearance of persons (the look in Ajax’s eyes, the quality of his movements) and things (his sword glinting in the darkness), and to the details of the actions (Ajax’s movements, his violent attack and its aftermath).17 All the examples of ekphrasis cited in the Progymnasmata are taken from larger narrative contexts and represent passages where the author elaborates on the appearance of a setting, a character, or the time at which some event occurred. Keeping such ekphraseis within their narrative settings was a concern to Theon and ps.-Hermogenes. They both advise their readers to ensure that ekphrasis is connected to the surrounding context.18 For Theon, this is particularly important in the case of ekphraseis of places, times, ‘the manner in which’ and persons – precisely the types of ekphrasis which recall the common modern conception of description as a separate block inserted into the flow of a narrative. (He notes the way Homer ties the arms of Achilles into the narrative context by describing their effect on the internal audience.) The supporting role which seems to be attributed to these ekphraseis in setting the scene corresponds to the conception of description as the faithful ‘ancilla narrationis’ identified by Genette as one of the characteristics of the modern conception of description. And, as has often been noted, ancient critics warned against the danger of these types of ekphrasis becoming over-lengthy and irrelevant: 15 Doxapatres, Homiliae in Aphthonii Progymnasmata, in Walz, Rhetores graeci, 2, p. 509; Anonymous scholia to Aphthonios, in Walz, Rhetores graeci, 2, p. 55. See Appendix A for texts and translations. 16 Ps.-Hermogenes, On Invention (Peri eurēseōs), 2.7, p. 122. See also below, pp. 71–2, on ps.-Hermogenes’ treatment of diaskeuē. 17 Anon., Peri tōn tessarōn merōn tou teleiou logou, Walz, Rhetores graeci, 3, p. 576, l. 21 – p. 578, l. 27. Text and translation in Appendix A. A reference to Michael Psellos as a model of style provides a terminus post quem of the late eleventh century for the treatise as a whole; however, it is quite possible that the author compiled material from sources of different dates. 18 Ps.-Hermogenes, Progymnasmata, p. 23, ll. 6–8; Theon, Progymnasmata, 119, ll. 25–30.


67

Lucian, On Writing History, 20, complained about historians who revel in pointless ekphraseis of places, and Horace, Ars poetica, famously warned against ‘purple patches’.19 However, if some types of ekphrasis correspond to the descriptionas-opposed-to-narration which Genette identified as characteristic of modern conceptions of literature, others blur the boundaries to such an extent that it is clear that these two types of discourse and the relation between them were conceived quite differently in ancient rhetorical thought. When the rhetoricians do articulate the difference between ekphrasis and diēgēsis they do so in entirely different terms. This is most evident in the treatment of two types of subject matter: pragmata (‘events, actions’, found in all the versions of the Progymnasmata) and tropos (‘the manner in which’ something is done or made, found in Theon alone). As Genette notes, the distinction between description and narration is relatively recent and the two are more difficult to separate in practice than in theory.20 For the ancient rhetoricians, ekphrasis could be applied not only to the background to action (time, place, manner, perpetrator) but to the action itself. An ekphrasis was distinguished from a diēgēsis not by the nature of the subject matter, but by the degree of reference to visible phenomena and the effect it had on the audience. These phenomena can be things, but also the actions themselves. Ekphraseis of Actions (Pragmata) The category of actions or events (pragmata) is central to the subjects of ekphrasis in all the versions of the Progymnasmata. It is clear from the examples cited, as well as from the context, that of all the possible meanings of the Greek term, that of ‘actions’ or ‘events’ is foremost, and the authors advise their readers to follow a temporal sequence, starting from what led up to the main event, in this case the battle, and what followed. The category of pragmata seems to have perplexed some modern readers of the Progymnasmata who come to it with a strong conception of description as distinct from narration, a type of text whose referent is characteristically an object. Some readers, whether consciously or 19

See Andrew Laird, ‘Ut figura poiesis: writing art and the art of writing in Augustan poetry’, in Jaś Elsner (ed.), Art and Text in Roman Culture (Cambridge, 1996), p. 93. Theon, Progymnasmata, 119, l. 33 warns against spending time on useless details (achrēsta). 20 Gérard Genette, ‘Frontières du récit’, in L’Analyse structurale du récit (Paris: Seuil, 1981), pp. 162–4; cf. Jean Molino, ‘Logiques de la description’, Poétique, 91 (1992): 363–82 and Perrine Galand-Hallyn, Le Reflet des fleurs: description et métalangage poétiques d’Homère à Erasme (Geneva, 1994), pp. 9–10. On the category of ‘description of action’, see Adam, La Description, pp. 76–89.

68


unconsciously, have written ‘actions’ out of their accounts of ancient ekphrasis, thus making it conform more closely to ‘description’. The entry on ekphrasis in the third edition of the Oxford Classical Dictionary, for example, translates pragmata as ‘objects’, which is within the range of meaning of the Greek but difficult to reconcile with the example of a battle, not to mention the absence of any example that would obviously correspond to our notion of ‘object’.21 In what seems to be a more deliberate act of philological sleight of hand, Roland Barthes spirited this category away altogether and replaced it with ‘works of art’.22 By this simple move he made ekphrasis conform to accepted modern definitions of description and remedied the rhetoricians’ omission of works of art as a prime type of subject matter (rather than an afterthought in one late author). This small and seemingly insignificant adjustment changes the whole balance of ekphrasis and its relation to the other Progymnasmata. The recommendations to begin an ekphrasis of a pragma like a battle with the events leading up to it and to go on to include the consequences make clear that such ekphraseis could involve temporal progression. Indeed, they could be conceived as a type of narration distinguished only by the inclusion of details, as in the example of the ‘ekphrastic’ elaborate version of the story of Ajax (see above and Appendix A). So while some ekphraseis might, like descriptions as broadly defined in modern terms, constitute a narrative pause, or a separable passage, even when woven into their contexts, others, such as Thucydides’ night battle, or Libanios’ ekphraseis of the Kalends or the hunt, constitute narratives (in the sense of accounts of actions unfolding in time) in themselves. Libanios’ ekphrasis of drunkenness is a particularly interesting example: rather than describing a drunken man as an object of fascination and disgust, he depicts his state through his actions at the drinking party and on his way home, depicting character through actions which betray moral choice (proairesis), as Aristotle recommends for judicial narrations.23 Ekphrasis is therefore not by definition separable from narration, nor does it by definition constitute a digression. Some examples may do so, whether their subjects are narrative or not (the making of the Shield of Achilles being a prime example of a narrative ekphrasis that interrupts a larger framing narrative), but this is not a necessary characteristic of ekphrasis as a whole as it was conceived in ancient rhetoric. 21 Simon Hornblower and Antony Spawforth (eds), Oxford Classical Dictionary (Oxford, 1996), p. 515. A similar idea of description may well underlie Kennedy’s choice of ‘object’ as a translation for tropos in Theon mentioned above, n. 11. 22 �� Roland Barthes, ‘L’effet de réel’, in Barthes et al. (eds), Littérature et réalité (Paris, 1982), p. 84�. 23 Libanios, Progymnasmata, pp. 477–9; cf. Aristotle, Rhetoric, 1417a16–24.


69

Ekphraseis of the Tropos Another class of subject matter which abolishes any strict boundary between ‘objects’ and ‘actions’ is the tropos. This category of ekphraseis of ‘the manner in which’ something is made or done is found only in the earliest version of the Progymnasmata by Theon. Though ‘customs’ or ‘manners’ is within the range of the Greek term, as one of the peristaseis it has a precise meaning that is illustrated by Theon’s examples. These are accounts of how equipment and arms were made: Thucydides’ accounts of the fortification of Plataia and of the construction of a siege machine (Book 4.100) and the account in the ninth book of Ktesias’ Persika of a strategy involving the use of effigies by Cyrus during his siege of Sardis to make the defending forces believe that Persian troops had already stormed the battlements.24 It is interesting to see how the influence of the modern association between description and ekphrasis has affected the interpretation of this passage of Theon’s text. In their translations, Patillon and Kennedy both identify Thucydides’ account of the fortification of Plataia as 3.21, a brief description of the finished construction. This passage does seem to be the one indicated in Theon’s introduction where he mentions ‘the plague in the second book and the fortification of Plataia in the third book’ as examples of ekphrasis in Thucydides’ History.25 However, the term Theon uses, periteichismos, would seem to refer more naturally to the process of construction rather than to a constructed object so that 2.75–8, a vivid account of the construction of the fortifications, fits Theon’s argument much better: the passage describes the offensive and defensive measures taken by the Spartans and the besieged Plataeans in narrative form (Thucydides’ use of the verb periteichizō at 2.78.1 adds further support to this identification). Although none of Theon’s immediate followers mentioned the tropos in their chapters on ekphrasis, the idea was clearly not lost. The ekphrastic version of the Ajax story expands principally on the manner in which the deed was done, and the technique of describing a monument through a narrative of its construction is a frequent technique in epideictic rhetoric.26 Doxapatres also discusses the ekphrasis of the 24

Ktesias, Persika, F9b, ed. Dominique Lenfant (Paris, 2004), p. 114. Theon, Progymnasmata, 68, ll. 7–10. 26 See Ruth Webb, ‘The aesthetics of sacred space: narrative, metaphor and motion in ekphraseis of church buildings’, DOP, 53 (1999): 59–74 and �� ‘Ekphrasis, amplification and persuasion in Procopius’ Buildings’, Antiquité tardive, 8 (2000)�� : 67–71 and below. It is interesting to note that Walter Ong, Orality and Literacy: The Technologizing of the Word (New York, 1988), pp. 42–3 identifies the representation of objects ‘in a total context of human action’ as one of the characteristics of oral discourse. His examples are drawn from Homeric epic, but the examples of ekphrasis of the tropos show that this tendency survived throughout antiquity. 25


70

tropos in his commentary on Aphthonios, distinguishing ekphraseis of the finished appearance (schēma) from ekphraseis of the process, the same distinction as the famous one between Homer’s approach to the description of the shield of Achilles as an account of the manufacture and Virgil’s presentation of the shield of Aeneas as finished object.27 Theon’s discussion of the ekphrasis of the tropos is also notable as the only place in the Progymnasmata where the Shield of Achilles, the ur-ekphrasis in the modern construction of the term, gets a mention. Theon could hardly be further from treating it as a description of an ‘objet d’art’, or even as a work of poetry. He simply lists the passage alongside other accounts of how military equipment and machines were made – a categorization which was suggested by the ancient term for the Shield, Hoplopoiia (‘Making of the Arms’), with its stress on the narrative organization as an account of manufacture. Neither ‘art’ nor ‘object’ figure as ideas in his analysis. Strange as this categorization may seem from a literary perspective, it does recognize a key feature of the Shield: its narrative form. But there is nothing definitive about Theon’s classification; it does not recur and seems rather like an ad hoc attempt to include a familiar passage which was clearly recognizable as an example of ekphrasis within the framework of rhetorical categories of subject matter which Theon was using. But his ability to place what is for us the seminal example of a description of a work of art in such company does show how different his preoccupations and organizing schemes were from those of a twentieth-century critic like Spitzer. The Distinction between Ekphrasis and Diēgēsis So, not only was ekphrasis not understood in antiquity as a term for ‘description of a work of art’, it was not even understood in the same terms as our ‘description’. The subject matter is not a factor in the definition; instead an ekphrasis is distinguished by qualities of the language and, most importantly, its effect on the listener. Ancient students of rhetoric were taught to recognize a distinction between ekphrasis and diēgēsis, but that distinction was expressed in quite different terms from those we might expect. The key term enargeia is central to the distinction. Explaining why the adjective enargōs is included in the definition of ekphrasis, Nikolaos says: 27

Doxapatres, Homiliae in Aphthonii Progymnasmata, in Walz, Rhetores graeci, 2, p. 513, l. 33 – p. 514, l. 5. Text and translation in Appendix A. See also Jean-Pierre Aygon, ‘L’ecphrasis et la notion de description dans la rhétorique antique’, Pallas, 41 (1994): 49. Modern theorists have recognized this liminal category with the term ‘Homeric description’: see Philippe Hamon, Du descriptif (Paris, 1993), pp. 189–98; Adam, La Description, pp. 77–81.


71

‘Vividly’ is added because it is in this respect particularly that ekphrasis differs from diēgēsis (narration). The latter sets out the events plainly, while the former tries to make the listeners into spectators. πρόσκειται δὲ ἐvαργῶς, ὅτι κατὰ τoῦτo μάλιστα τῆς διηγήσεως διαφέρει. ἡ μὲv γὰρ ψιλὴv ἔχει ἔκθεσιv πραγμάτωv, ἡ δὲ πειρᾶται θεατὰς τoὺς ἀκoύovτας ἐργάζεσθαι.28

An ekphrasis is distinguished from narration (diēgēsis) by the quality of enargeia, ‘vividness’. The distinction between ekphrasis and diēgēsis is therefore not a question of the type of referent but resides in the effect on the listener. However, this definition raises a new question of distinction: when is a narrative ‘vivid’ enough to be an ekphrasis? Nikolaos illustrates what he understood by ‘vividness’ by a very basic example clearly inspired by Thucydides: It is [characteristic] of diegesis to say the Athenians and the Peloponnesians went to war, but of ekphrasis [to say] that each side made such and such preparations and equipped itself in this manner (tropos). διηγήσεως μέv ἐστιv εἰπεῖv ἐπoλέμησαv Ἀθηvαῖoι καὶ Πελoπovvήσιoι. ἐκφράσεως δὲ, ὅτι τoιᾷδε καὶ τoιᾷδε ἑκάτερoι παρασκευῇ ἐχρήσαvτo καὶ τῷδε τῷ τρόπῳ τῆς ὁπλίσεως.

A narration therefore simply gives the information that ‘the Athenians fought the Peloponnesians’, while an ekphrasis tells how, with what preparations (so that much of Thucydides’ history would be, according to this definition, ekphrasis).29 The definition of ekphrasis therefore depends on the amount of perceptible detail conveyed by the verbal account, the exact quantity remaining to be determined by subjective judgement, or by convention.30 Degrees of Narration The distinction between plain and visually elaborate (i.e. ekphrastic) accounts is found in several sources, with different vocabularies. It appears in the contrast between simple (haplē) and elaborate (endiaskeuos) narration mentioned above, in which it is clear that the elaboration resides in the description of actions as well as of persons and objects. The author 28

Nikolaos, Progymnasmata, 68, ll. 9–10. Nikolaos’ reference to the tropos seems to be a direct reminiscence of Theon. 30 As noted by Aygon, ‘L’ecphrasis et la notion de description’, p. 48 and Sophie Rabau, ‘Narration et description: l’exigence de détails’, Lalies, 15 (1995): 273–90. 29

72


of the ps.-Hermogenean treatise On Invention mentions the three types of narration, of which these are two, in his discussion of narration at 2.7 and then returns in Book 3 to the elaborate account, which he terms diaskeuē.31 This diaskeuē has similarities to ekphrasis, as is pointed out by a scholiast who defines it as being ‘as if depicting (diazōgraphein) and describing (ekphrazein) the subject and as if bringing it before the listener’s eyes.’32 Ps.-Hermogenes himself defines diaskeuē as a form of diatupōsis and compares it in two places to poetry.33 The action (pragma) is central to ps.Hermogenes’ conception of diaskeuē; moreover, he notes that the cause is absent, as is the case with ekphrasis. Of particular interest is his emphasis on the tropos, the manner in which the act took place, which is central to the elaboration required of the diaskeuē and this is something that is well illustrated by the Byzantine example of the elaborate diēgēsis of Ajax (see Appendix A). I would therefore suggest that ekphrasis may also be considered, in some contexts, to be a process of elaboration that is applied to a basic statement of facts, adding details in order to ‘place the subject before the eyes’. This explains the close affinity between ekphrasis and the exercise of diēgēma noted by the commentators on Aphthonios. In terms of judicial oratory, both contribute to the narrative section of a speech, as Nikolaos himself points out in his Progymnasmata.34 Enargeia in Quintilian A similar relationship between the basic statement of facts and the vivid version that could be termed ekphrastic is to be found in Quintilian. In his discussion of enargeia in Book 8 of the Institutio oratoria Quintilian draws a distinction between a detailed account of an event related with enargeia, which ensures that the audience ‘seem to see’ it unfolding before their eyes, and the ‘brief statement’, which simply conveys the information, as 31 Ps.-Hermogenes, On Invention (Peri Heureseōs), 3.15, p. 166, l. 20 – p. 170, l. 18. Michel Patillon, Hermogène: l’art rhétorique (Paris, 1997), p. 274 identifies this chapter as originally belonging to a section discussing the epilogue. Diaskeuē certainly can be used in epilogues (see Chapter 6 below for examples in Sopatros the Rhetor) but there is, as Kennedy points out, no mention of the epilogue in this chapter and the discussion points rather to a link with narration and the endiaskeuos diēgēsis of 2.7. See George A. Kennedy, Invention and Method: Two Rhetorical Treatises from the Hermogenic Corpus Translated with Introduction and Notes (Atlanta, 2005), p. 127. 32 Walz, Rhetores graeci, 7, p. 717, l. 17 – p. 718, l. 3: χρὴ δὲ γιvώσκειv ὅτι κατασκευὴ μέv ἐστι τὸ ταῖς αἰτίαις ἀληθῆ δεικvύvαι τὴv πρότασιv: διασκευὴ δὲ τὸ διὰ τoῦ τρόπoυ αὔξειv καὶ πoλλαπλασιάζειv τὸ ὁμoλoγoύμεvov καὶ oἱovεὶ διαζωγράφειv αὐτὸ καὶ ἐκφράζειv, καὶ ὥσπερ ὑπ’ ὄψιv ἄγειv τoῖς ἀκρoαταῖς.�� As noted above, the endiaskeuos form of narration may also be termed ‘ekphrastic’. 33 Ps.-Hermogenes, On Invention, 3.15, p. 167, l. 1 and p. 167, l. 21 – p. 168, l. 2. 34 Nikolaos, Progymnasmata, p. 70, ll. 2–3.


73

in Nikolaos’ definition of diēgēsis. Quintilian’s example is different: he uses the sack of a city, one of the stock subjects of rhetoric and historiography alike, which he illustrates with his own circumstantial account of the event, describing the destruction of buildings and the distress of the inhabitants.35 Such a tableau is created, he explains, by ‘opening up’ the brief statement of facts and exploring the implications: Doubtless, if one says that the city has been taken one implies everything which such a fate entails, but, like a brief announcement, this penetrates the emotions less. If instead you open up the things which were included within the single phrase, there will appear flames pouring through houses and temples, the crash of roofs falling and the sound made up of the cries of many individuals, some hesitating as they flee, others clinging to their loved ones in a last embrace, the wailing of women and children and old men lamenting that fate has preserved them for such a terrible destiny. Then there will be the plunder of sacred and profane property and people running to and fro carrying booty and prisoners each driven in front of his captor, and a mother trying to keep hold of her child, and fighting among the victors wherever the booty is greatest. Sine dubio enim qui dicit expugnatam esse civitatem complectitur omnia quaecumque talis fortuna recipit, sed in adfectus minus penetrat brevis hic velut nuntius. At si aperias haec quae verbo uno inclusa erant, apparebunt effusae per domus ac templa flammae et ruentium tectorum fragor et ex diversis clamoribus unus quidam sonus, aliorum fuga incerta, alii extremo complexu suorum cohaerentes et infantium feminarumque ploratus et male usque in illum diem servati fato senes: tum illa profanorum sacrorumque direptio, efferentium praedas repetentiumque discursus, et acti ante suum quisque praedonem catenati, et conata retinere infantem suum mater, et sicubi maius lucrum est pugna inter victores. (8.3.67–9)

Although he does not use the Greek term, Quintilian’s conception of the vivid account is very close to Nikolaos’ conception of ekphrasis: both conceive of the vivid, detailed version as the result of the expansion of a basic statement of fact whether ‘the city has fallen’ or ‘the Athenians and Spartans went to war’.

35 See George M. Paul, ‘Urbs capta: sketch of an ancient literary motif’, Phoenix, 36 (1982): 144–55.

74


Quintilian’s Enargeia and the Greek Tradition of Ekphrasis Several parallels are labelled as ekphrasis in Greek manuals. The most frequently cited evocation of enemy destruction in classical Greek oratory is Demosthenes’ description of Phokis (19.65), cited by Nikolaos as an example of ekphrasis in judicial oratory and by the author of the ps.-Hermogenean On Invention as an example of endiaskeuos diēgēsis.36 Moreover, in a Greek instruction manual on declamation by Sopatros the Rhetor, an account of a sacked city which is comparable to Quintilian’s example is referred to as ekphrasis.37 This passage, which is a reworking of the Demosthenic model (the speaker is Demosthenes himself), and its implications for the function of ekphrasis, will be discussed in Chapter 6. The apparent absence of ekphrasis from Quintilian’s treatment of the preliminary exercise has been noted, but his treatment of enargeia shows that he was familiar with the doctrines and procedures of ekphrasis as we find them in the Greek manuals. There is therefore no need to look for ekphrasis in Quintilian’s treatment of digressio or excursus, as has been suggested, again under the assumption that ekphrasis is characteristically digressive; instead it is his discussion of enargeia that comes closest to the Greek discussions of ekphrasis.38 Ekphrasis: Telling in Full Nikolaos’ example also shows us the force of the preposition ‘ek’ in ekphrasis. It does not refer to the separable nature of the descriptive passage, to the way in which it stands out ‘ek’ from the narrative background (an idea grounded in the common conceptions about description identified by Hamon).39 Instead, the preposition has an intensive force, meaning ‘in full’, ‘utterly’. So to compose an ek-phrasis, is ‘to tell in full’, to give all the details.40 The closest parallel in Latin to ekphrasis is therefore not 36 Nikolaos, Progymnasmata, p. 71 ll. 3–5; ps.-Hermogenes, On Invention, 2.7, p. 124, ll. 9–14. This passage is also cited by Alexandros, Peri Schēmatōn, in Walz, Rhetores graeci, 7, p. 456 and in the Scholia Demosthenica, 157c as a diatupōsis and in other scholia to Demosthenes (157a and b) as diaskeuē and endiaskeuos diēgēsis, while in Apsines, Art of Rhetoric, 3.27 (p. 134, ll. 21–3) it is mentioned as an example of hupographē. The corresponding description of Thebes in Aeschines, Against Ktesiphon (3.157) is termed ekphrasis by Herodian, Peri schematōn, in Walz, Rhetores graeci, 8, p. 603, l. 12 (see Aygon, ‘L’ecphrasis et la notion de description’, p. 47). Compare also the Rhetorica ad Herennium, �� 4.39.51�� on descriptio. 37 Sopatros the Rhetor, On the Division of Questions, p. 210, l. 20. See Chapter 6 for further discussion. 38 Ian H. Henderson, ‘Quintilian and the Progymnasmata’, Antike und Abendland, 37 (1991): 90. 39 Philippe Hamon,‘Qu’est-ce qu’une description?’, Poétique, 112 (1972), p. 465. 40 As Perrine Galand-Hallyn, Le Reflet des fleurs, p. 14 has noted, this conception of ekphrasis as a development of a basic idea is close to Philippe Hamon’s definition of description as the exploration of the lexical field and likewise does away with the need to


75

descriptio but explicatio, ‘unfolding’.41 Similarly, the preposition peri, as in periēgēmatikos, may also serve primarily to convey the idea of elaboration.42 To ‘lead someone around’ (periēgeisthai) a subject is to relate it more fully, in more detail than simply to ‘lead someone through’ it (diēgeisthai). Periēgēsis and its related terms do not therefore necessarily imply that the referent is an object around which a person or an eye can wander, any more than the subject of a diēgēsis has to be spatial. It is a mode of verbal presentation which evokes perceptible details of the subject matter, the aim of which, the Progymnasmata tell us, is to ‘bring the subject before the eyes’, to make the listener see the subject described in his or her mind’s eye. It is this detail and the visual impact which should flow from it, not the type of subject matter, which make an ancient ekphrasis an ekphrasis. Ekphrasis, Muthos and Diēgēma The subjects for ekphrasis, particularly the category of pragmata, help to clarify what ps.-Hermogenes meant when he said that some people taught ekphrasis within the two narrative exercises: muthos and diēgēma. Rather than being an optional extra which could be inserted into a narrative, ekphrasis was a process which could be applied to it, in which the basic idea was expanded by reference to its perceptible characteristics. The object of the exercise was to have an imaginative impact on the viewer which, in a rhetorical context, meant contributing to the persuasive effect of a speech.43 The close link between ekphrasis and the exercise of diēgēma points to the rhetorical function of ekphrasis. It could be used in the narrative section (diēgēsis or narratio) of a judicial speech, to lend extra force and credibility to the speaker’s version of events.44 The inevitably partial nature of rhetorical diēgēsis, in which plausibility was far more important than truth, means that ekphrasis is part of a fictional creation which is far from neutral or innocent.

define description on the basis of the nature of the subject matter. See Hamon, ‘Qu’est-ce qu’une description?’ and Du descriptif. 41 Theon, Progymnasmata, ed. Patillon, p. 149 n. 323; Wilhelm Geissler, Ad descriptionum historiam symbola (Leipzig, 1916), p. 26. On explicatio in the Elder Seneca, see Janet Fairweather, Seneca the Elder (Cambridge, 1981), p. 211. Fairweather, ibid., p. 210, also notes that Senecan descriptio is not invariably an extended set piece but is sometimes ‘merely an extra-vivid piece of narration’. 42 Liddell and Scott, Greek–English Lexicon, peri F IV; cf. Demetrios, On Style, 19 on periagōgē. Quintilian, Institutio oratoria, 2.4.3 uses the image of the sinuous path to convey the effect of excessive description. 43 Anonymous Seguerianus, 96, p. 26 introduces his definition of enargeia by stating that it ‘contributes to persuasion’ (συνεργεῖ δὲ πρὸς πεɩθὼ καὶ ἡ ἐνάργεια). 44 Nikolaos, Progymnasmata, p. 70, ll. 2–3 states clearly that ekphrasis is an exercise which prepares for the composition of judicial diēgēseis.

76


The speaker’s hermēneia, or interpretation of events, is all important, an idea that is perhaps implicit in Nikolaos’ use of the adjective aphēgēmatikos. Nikolaos on the Function of Ekphrasis Of all the authors of Progymnasmata, Nikolaos is the only one to include any explicit discussion of the function of ekphrasis in the context of a speech. As well as noting that ekphrasis is an exercise that prepares for the narrative section of a speech, he goes through each of the standard three rhetorical genres in turn and, as he does so, brings back into focus the response of the audience which was so central to all the authors’ definition of ekphrasis. In deliberative (political) oratory, the ekphrasis of the subject helps the speaker persuade his audience (to either accept or reject the subject); in judicial speeches, ekphrasis is a means of amplification (auxēsis: the technique of emphasising the importance of something, such as the heinousness of the crime); while in epideictic the ekphrasis need only create a sense of pleasure (hēdonē) in the audience. And a little further on, he cites the example of Demosthenes’ evocation of the destruction at Phokis to illustrate the use of ekphrasis as a means of achieving auxēsis, the rhetorical underlining of a particular subject and deinōsis (the arousal of indignation). The effect of ekphrasis was therefore varied and depended on context and occasion. Nikolaos’ account of it is further nuanced by the treatments of the various types of oratory in more advanced treatises as will be discussed in Chapter 6. Ekphrasis and Common Place Nikolaos’ reference to the use of ekphrasis in amplification, making the audience feel the necessary emotion about some class of person or act, also helps to explain ps.-Hermogenes’ claim that ekphrasis could be studied as part of the exercise of koinos topos (common place). This was the exercise which contained a rehearsal of commonly accepted opinions about certain categories of person whether heroes or villains – such as tyrants, adulterers, the desecrators of temples and murderers.45 The exercise provided the students with a wealth of ready-made assertions about each type of person to be used as necessary to expand upon the 45 Theon, Progymnasmata, 106, ll. 6–7. τόπoς ἐστὶ λόγoς αὐξητικὸς ὁμoλoγoυμέvoυ πράγματoς ἤτoι ἁμαρτήματoς ἢ ἀvδραγαθήματoς [Topos is a speech which amplifies an action, whether a crime or a good deed, that is not in dispute]. Theon calls this exercise simply topos. See Laurent Pernot, ‘Lieu et lieu commun dans la rhétorique antique’, BAGB (1986): 253–84; Malcolm Heath, ‘Invention’, in S. Porter, Handbook of Classical Rhetoric in the Hellenistic Period (330 B.C.–A.D. 400) (Leiden, 1997), p. 95.


77

vices or virtues of a category of person or action. Thus, a speech accusing someone of adultery could include a section dwelling on the heinous nature of the crime and the wickedness of those who commit it. One technique which students were taught to use within the koinos topos was the vivid description of the crime or good deed. Theon gives the example of a murder described as an action unfolding, which is to be used in the koinos topos against a murderer: For example we shall describe (diagraphō) the perpetrator in the act of committing (ergazomenos) the murder, how cruelly and mercilessly he became, though a human being himself, the murderer of another human being, and drawing the sword and inflicting the wound and, to ensure that the wound was fatal, striking again and again and being spattered with the murdered man’s blood, and [we will describe] the victim’s cries, as he at one time begged the murderer for mercy and at another time called out for help from men and from the gods and all kinds of things of this sort.46 διαγράψoμεv γὰρ oἷoς μὲv ἦv ὁ ἐργαζόμεvoς τὸv φόvov, ὡς ὠμῶς καὶ ἀvηλεῶς αὐτόχειρ γεvόμεvoς ἀvθρώπoυ ἄvθρωπoς ὤv, καὶ τὸ ξίφoς σπώμεvoς, καὶ τὴv πληγὴv καταφέρωv, καὶ μὴ καιρίας εἰ τύχoι τῆς πληγῆς γεvoμέvης, ἄλλας ἐπ’ ἄλλαις πoιoύμεvoς, καὶ μιαιvόμεvoς αἵματι τῷ τoῦ φovευoμέvoυ, oἵας δὲ κἀκεῖvoς φωvὰς ἠφίει, τoῦτo μὲv τoῦ φovέως δεόμεvoς, τoῦτo δὲ βoηθoὺς ἐπικαλoύμεvoς, vῦv μὲv ἀvθρώπoυς, vῦv δὲ τoὺς θεoύς, καὶ ὅσα ἄλλα τoιαῦτα.

Although Theon does not use the term ekphrasis in this context, his example shows why the teachers alluded to by ps.-Hermogenes thought it appropriate to teach ekphrasis as part of the exercise koinos topos. Nikolaos, in his discussion of the same exercise, does use the term ekphrasis interchangeably with both hupotupōsis and diatupōsis when he comes to discuss the inclusion of such vivid evocations of actions.47 However, what ps.-Hermogenes’ comment masks is the fact that the relation between ekphrasis and koinos topos is different from that between ekphrasis and the narrative exercises, diēgēsis and muthos. Ekphrasis (or hupotupōsis, or diatupōsis) stands to koinos topos as part to whole; it is a passage that can be inserted to increase the dramatic effect of the amplification. But with respect to diēgēsis, ekphrasis can be considered to be an expansion of a basic narrative, a process applied to it, as in Quintilian’s image of ‘opening up’ the idea of the sack of a city, or the 46 47

Theon, Progymnasmata, 109, ll. 3–11. Nikolaos, Progymnasmata, p. 45, ll. 9–22.

78


distinction between the simple (haplē) and elaborate (endiaskeuos) types of narration. If we move from the Progymnasmata to the aspects of the larger speech for which they prepare, there is a comparable variety in the uses of ekphrasis. Used as part of a narration of events in dispute, ekphrasis may serve to establish the speaker’s version (see below on the rhetorical uses of enargeia); used as part of a koinos topos, its purpose is first and foremost to arouse the appropriate emotions as part of a larger amplification. Ekphrasis and Enkōmion in the Progymnasmata Despite the apparent uniformity of the surviving Progymnasmata, there are differences in detail which reflect major differences in orientation and suggest a gradual shift over time in perceptions of ekphrasis. This is apparent in the role of confirmation and refutation in Theon, who treats these most argumentative of exercises not as one exercise among many, but as processes which can be applied to all the other exercises. Thus one writes a simple narration or ekphrasis at one stage, and later learns to argue that the details of the story or the ekphrasis are or are not plausible.48 Of all the authors, Theon seems to have his eye most firmly fixed on the courtroom as a venue for the skills he is teaching, a concern he shares with his contemporary Quintilian. This judicial orientation is also evident in the subjects he singles out for ekphrasis, which correspond closely to the parts of narration, even including the tropos, the ‘manner in which’ something is done. Narration was a vital element of judicial speeches, which revolve around past actions. This close relation between ekphrasis and diēgēma in Theon, alongside the references to ekphrasis in Greek handbooks on declamation, such as that of Sopatros, points to a function for ekphrasis in judicial rhetoric. The later versions of the Progymnasmata, however, reflect a shift towards epideictic, which is visible in the treatment of ekphrasis and its relation to enkōmion, the exercise which prepared directly for epideictic oratory. The demands of epideictic are evident in some small but significant changes in the list of subject matter proposed for ekphrasis. Nikolaos introduces festivals (panēgureis) into his list, which otherwise reproduces the standard schema of person, place, time and event.49 Libanios provides an example of this kind of subject in his model ekphrasis of the festival

48 Patillon, introduction to Theon, Progymnasmata, pp. xxviii–xxxi and Ruth Webb, ‘The Progymnasmata as practice’, in Yun Lee Too (ed.), Education in Greek and Roman Antiquity (Leiden, 2001), pp. 289–316. 49 Festival time (heortē) had figured in Theon’s chapter on ekphrasis (118, l. 21) as a sub-type of times (chronoi).


79

of the Kalends.50 But a far more telling example is Aphthonios’ puzzling addition of ‘plants and dumb animals’ to his list (see Table 3.1).51 Plants and animals are a rather incongruous addition to the peristaseis and no examples of these interlopers are given in the discussion which follows; nor does Aphthonios identify any models from classical texts (although Theon had included descriptions of exotic beasts and birds under the heading of prosōpa). However, the motivation for this alteration is clear from Aphthonios’ precepts for other Progymnasmata. It makes his list of subjects for ekphrasis identical with that for enkōmion, as the table shows. To find examples of ‘plants and dumb animals’ we have to go to the earlier chapter where the horse, the ox, the olive and the vine are suggested as objects of praise.52 The exact subjects suggested by each author are thus a further indication of the connections between ekphrasis and the other exercises. As a comparison between the subjects identified for ekphrasis shows (see Table 3.1), between Theon and Aphthonios there was a perceptible change from seeing ekphrasis as closely allied to narration towards an emphasis on its links with enkōmion. Enkōmion, like diēgēma and koinos topos, was one of the other exercises which, according to ps.-Hermogenes, could encompass the practice of ekphrasis. The model enkōmia by Libanios show how the two exercises were interconnected, and how the appearance of a subject, where appropriate, could be evoked as an integral part of the praise. The model enkōmion of the ox alludes to the beast’s beauty and the appearance of the date palm is described.53 Lucian’s ludic Enkōmion of the Fly also lingers over the insect’s shape, texture and the way its carapace shines in the sun, like a peacock. Conversely, the models of ekphrasis by Aphthonios and Libanios have unmistakably encomiastic aims. Their approach to the subject is in itself encomiastic. Aphthonios’ description (which takes the form of a detailed verbal tour around the building) is prefaced by a discussion of the utility of the acropolis (p. 38) and ends with what comes to be the almost obligatory statement that its beauty is ‘more than can be expressed in words’ (p. 41). The encomiastic features of Libanios’ ekphrasis of spring are clear from his

50

Libanios, Progymnasmata, pp. 472–7. Aphthonios, Progymnasmata, p. 37: πρόσωπά τε καὶ πράγματα, καιρoύς τε καὶ τόπoυς, ἄλoγα ζῷα καὶ πρός τoύτoις φυτά. 52 Aphthonios, Progymnasmata, p. 21: ἐγκωμιαστέov δὲ πρόσωπά τε καὶ πράγματα, καιρoύς τε καὶ τόπoυς, ἄλoγα ζῷα καὶ πρὸς τoύτoις φυτά. Hermogenes also mentions plants and animals as subjects of an enkomion, alongside abstractions such as justice (classed as pragmata) and places. Libanios’ model enkōmia include the ox (pp. 267–73), the date palm and the apple tree (pp. 273–7). 53 Libanios, Progymnasmata, p. 479, ll. 11–19. 51


80

introductory phrase stating ‘I love the spring more than the other seasons (hōrai) and wish to tell (diēgēomai) what it is like’.54 Sardianos’ commentary on Aphthonios reflects the increased importance of epideictic. He is very clear about the uses of ekphrasis in enkōmia (i.e. epideictic speeches) citing several types of subject typical of epideictic speeches: ‘in enkōmia you will describe places made by men, harbours, colonnades and the like’ (p. 215, ll. 18–19). Significantly, he is either less confident or less interested in its use in the other genres; discussing judicial speeches he can only suggest that we might need to describe the place of the crime, while in deliberative oratory, he says vaguely, following Nikolaos, ‘we often need to describe the subject of our speech in order to be more persuasive’. Epideictic, and epideictic needs, have become the norm and, although Byzantine homilies often display the use of ekphrasis of events, they did not feature as a genre in didactic handbooks. The effects of the rise of epideictic on the surviving corpus of Progymnasmata are hardly perceptible on the surface, hence the reputation earned by these exercises as being unchanging. But at the level of details like the subjects for ekphrasis there were subtle changes within the prescriptions for ekphrasis which recognized, however belatedly, the changing demands of rhetorical practice and which altered the relation of the exercise of ekphrasis to the other Progymnasmata. It is, however, important to note that within the teaching of rhetoric the rise of epideictic was relative, not absolute, and it did not eclipse declamation, at least up until the sixth century.55 A certain Athanasios went so far as to subordinate all the other exercises to enkōmion in his version of the Progymnasmata.56 But the experiment cannot have been attractive enough to ensure Athanasios’ immortality among the rhetoricians since his work does not survive and is cited as a curiosity. Among the surviving progymnasmatists, even Aphthonios did not share this view of the primacy of enkōmion. He emphasizes the importance of the exercises in confirmation and refutation, essential skills for the law court, claiming that they ‘contain the whole skill of this art’ and the Progymnasmata were still a preparation for declamation. At the same time, enkōmion clearly had a place in Theon’s system, and the type of use of ekphrasis in the service of enkōmion that we see in Aphthonios and Libanios’ model ekphraseis is

54

�� Ibid., p. 479, ll. 17–18. Chorikios’ declamations show that the art of judicial and deliberative speaking was still being taught by this method in sixth-century Gaza, and the Hermogenean Corpus, which is almost entirely devoted to declamation, was the basis of the Byzantine rhetorical curriculum. Malcolm Heath, Menander: A Rhetor in Context (Oxford, 2004) mounts a vigorous challenge to the thesis that epideictic eclipsed declamation after the second century. 56 See Aphthonios, Progymnasmata, p. 53 and Pernot, La Rhétorique de l’éloge, pp. 59–60. 55


81

well represented in Statius’ occasional poetry and probably had a long history in Hellenistic prose and verse.57 Ekphraseis of Works of Art One type of subject matter for ekphrasis that was particularly useful in epideictic contexts was cities and buildings, which could be included under the traditional heading of ‘places’, and it is surely no accident that Aphthonios’ model ekphrasis is of such a place: the Alexandrian acropolis. If architecture is easily subsumed within the standard treatments of ekphrasis, the same is not true, at first sight at least, of the central subjects of modern ekphrasis: painted and sculpted representations. In a way, the absence of these from the rhetorical manuals is neither surprising nor particularly significant. It is clear that the subjects proposed for ekphrasis by the various authors of the Progymnasmata, far from being randomly chosen, formed a coherent group that reveals the interrelationship of ekphrasis with other preliminary exercises and points to its place within the larger system of rhetoric. But, though the list of subjects has a precise meaning, this never appears to have been exclusive: the list of the peristaseis and the various additions – combined with the emphasis on enargeia in the definition – left open the possibility that vivid accounts of any subject, including statues and paintings, could be counted as ekphraseis. Such subjects may not have been central to the ancient definition of ekphrasis, but nor were they ever excluded, as Theon’s reference to the Shield of Achilles suggests. The Byzantine commentators on Aphthonios remark on the absence of paintings and sculptures from his discussion. Sardianos (who cites Philostratos’ Eikones among the principal models for ekphrasis to be followed by students) explains that ekphraseis of images should be classified according to their subject matter: images of persons are to be considered as persons, images of events as events.58 This solution looks like a typical commentator’s attempt to avoid apparent discrepancies in his source, but it does correspond to an observable quality of many ekphraseis of paintings, in particular, which describe the subject matter rather than confining themselves to what could be shown in a single image. The marginality of these subjects from the rhetoricians’ point of view helps to explain the extraordinarily casual way in which they are 57

See Alex Hardie, Statius and the Silvae: Poets, Patrons and Epideixis in the Graeco-Roman World (Liverpool, 1983). The lack of prose sources for the period makes it impossible to judge how close Statius’ poems were to contemporary prose enkōmia. 58 Sardianos, Commentarium in Aphthonii Progymnasmata, p. 219, ll. 12–25. Doxapatres, Homiliae in Aphthonii Progymnasmata, p. 512, ll. 12–20 makes the same point about representations of persons.


82

introduced into the discussion of ekphrasis by Nikolaos, the only author to make explicit reference to works of art as a category. Nikolaos makes no reference to these subjects in the list at the beginning of his chapter but, after his explanation of the difference between ekphrasis and diēgēsis, makes some general comments about achieving enargeia, which he identifies as being particularly relevant to ekphraseis of paintings and sculptures: We must, particularly when we describe statues for example or paintings or things of this sort, try to add reasons (logismoi) why the painter or sculptor depicted things in certain ways, such as, for example, that he depicted the character as angry for such and such a cause (aitia) or happy, or we will mention some other emotion (pathos) resulting from the story about the person being described. Reasons contribute greatly to enargeia in other types of ekphrasis as well.59 δεῖ δέ, ἡvίκα ἂv ἐκφράζωμεv καὶ μάλιστα ἀγάλματα τυχὸv ἢ εἰκόvας ἢ εἲ τι ἄλλo τoιoῦτov, πεῖρασθαι λoγισμoὺς πρoστιθέvαι τoῦ τoιoῦδε ἢ τoιoῦδε παρὰ τoῦ γραφέως ἢ πλάστoυ σχήματoς, oἷov τυχὸv ἢ ὅτι ὀργιζόμεvov ἔγραψε διὰ τήvδε τὴv αἰτίαv ἢ ἡδόμεvov, ἢ ἄλλo τι πάθoς ἐρoῦμεv συμβαῖvov τῇ περὶ τoῦ ἐκφραζoμέvoυ ἱστoρίᾳ. καὶ ἐπὶ τῶv ἄλλωv δὲ ὁμoίως πλεῖστα oἱ λoγισμoὶ συvτελoῦσιv εἰς ἐvάργειαv.

Nikolaos seems here to consider only images of human figures, such as the examples of statue ekphraseis by Kallistratos or many of Philostratos’ Eikones. He emphasizes the importance of the narrative context and its concomitant emotions to the impact of such ekphraseis: it is reference to these, rather than the description of the physical features of the image alone, that brings the necessary enargeia. It is striking that this is the only place in the Progymnasmata chapters on ekphrasis where the notion of the cause (aitia) is invoked explicitly. What Nikolaos means is perhaps best illustrated by the examples of statue ekphraseis which are transmitted under his name. The author of these model ekphraseis describes the image piece by piece, reading each detail of facial expression, gesture and dress as a sign of emotion or character (ēthos). Both the figure of Medea and that of the captive Trojan woman turn their heads away, the first so as not to look at her children, the second because she cannot bear to see that her city has fallen; the bronze Ajax’s eyes dart wildly here and there; each detail of Prometheus’ face reveals his pain and the manner (tropos) in which he suffers.60 In the case of the goddesses Hera 59

Nikolaos, Progymnasmata, p. 69, ll. 4–11. Nikolaos (?), model Progymnasmata in Libanios, Opera, 8, p. 506, ll. 4–6; p. 525, ll. 9–11; p. 512, l. 16 – p. 513, l. 3. 60


83

and Athena, the same method is applied to the depiction of their natures and spheres of influence through their attributes. This is known as a mode of reading statues as is shown in Porphyry’s work on statues preserved in Eusebios’ Praeparatio evangelica.61 It may also be related to a type of school exercise mentioned by Quintilian (�� 2.4.26)�� in which students of rhetoric were asked to explain the details of the iconography of Venus or Cupid. The three ekphraseis of paintings which Foerster accepts as genuine works of Libanios are similar in that they describe details of gesture (two are of scenes in the countryside, peopled with figures in action; the third is a scene from epic, the foot-race from Patroklos’ funeral games). The accounts of these details are often accompanied by the author’s guess as to the meaning, introduced by a phrase like ‘as [seems] likely’ (hōs eikos) or a verb of meaning (mēnuō, semainō).62 One motivation for including ekphraseis of paintings and statues among the examples of Progymnasmata may well have been the opportunity such subjects provided for describing figures and scenes from mythology which are prominent in the examples of the other exercises. In this sense, describing a statue of Hera or Ajax is a means of describing the character. However, as is clear from Nikolaos’ discussion, the idea of representation introduces new elements into these ekphraseis: the artist’s intention and the task of interpretation attributed to the viewer/speaker.63 As I will suggest below, there is a sense in which these particular ekphraseis are programmatic, making explicit the active role of the audience of any ekphrasis who are prompted to supply further details from their own imaginations. There is, of course, a further and much deeper affinity between visual representation and the task of rhetorical ekphrasis, as Sardianos makes clear in his statement that ‘enargeia imitates [the actions and effects of] the art of painting’.64 In this sense, any ekphrasis rivals the visual arts in that it seeks to imitate their visual impact. The connection lies thus at the level of effect rather than residing in the subject matter and means that any ekphrasis is haunted by the idea of the work of art and, even more

61 See Aline Rousselle, ‘Images as education in the Roman Empire’, in Yun Lee Too (ed.), Education in Greek and Roman Antiquity (Leiden, 2001), pp. 373 – 403. 62 For a translation of one of these ekphraseis, see Michael Baxandall, Patterns of Intention: On the Historical Explanation of Pictures (New Haven, 1985), p. 2. The Elder Philostratos uses such terms frequently in his Eikones. 63 Andrew S. Becker, The Shield of Achilles and the Poetics of Ekphrasis (Lanham, 1995), pp. 42–4 proposes four levels within ekphraseis of works of art: the referent (the subject matter); the physical object; the artist and the interpreter. 64 Sardianos, Commentarium, p. 217, ll. 3–6.

84


perhaps, by the idea of theatrical representation which is also present in the reference to the listener as spectator (theatēs). Theory and Practice: Developments in Ekphrasis Nikolaos’ inclusion of statues and paintings, like Aphthonios’ linkage of ekphrasis and enkōmion, shows that the tradition of ekphrasis as outlined in the surviving Progymnasmata instructions is far from unified or simple, though the loss of other texts prevents us from seeing precisely when changes may first have been introduced, or how long it took for practice to affect precept. It is clear, however, that the authors of the handbooks were receptive �� t�� o change and accommodated it as far as possible within the traditional frameworks of the didactic genres. Such changes are hardly perceptible to the naked eye and only become apparent when studied against the wider context of the relation of each progymnasma to the other exercises in the series and to the more advanced treatises. Despite the occasional comments by ancient critics (particularly Romans in the early Empire) on the distance between the schools and the cut and thrust of real rhetorical practice, the educational process was shaped in many ways by the demands of adult life.65 If we look at all the uses of ekphrasis that the student encountered in his more advanced study of rhetoric, the sense of diversity already encountered in the Progymnasmata is only reinforced. Some ekphraseis are of works of art, most are not; some are laudatory (most clearly the model ekphraseis of Libanios and Aphthonios), many are not. Some correspond to our notion of description, simply defined, in that they describe static, spatial entities, many do not. Any single statement that attempts to define ekphrasis in terms of subject or function (or lack of function) will apply to some cases, but will ultimately fall far short of a complete picture. Each case needs to be considered individually in its context. Ekphrasis as Rhetorical Technique Given the rhetorical uses of ekphrasis, it is curious that there are not more examples from oratory cited in the Progymnasmata. The predominance of examples drawn from history and poetry could be taken as a sign that ekphrasis was originally non-rhetorical and was only later adopted by orators. A complaint by ps.-Dionysios of Halikarnassos that declaimers waste words on ekphraseis of plagues, battles and the like under the 65 See E. Patrick Parks, The Roman Rhetorical Schools as a Preparation for the Courts under the Early Empire (Baltimore, 1945). John A. Crook, Legal Advocacy in the Roman World (London, 1995) and Malcolm Heath, ‘Practical advocacy in Roman Egypt’, in Michael J. Edwards and Christopher Reid (eds), Oratory in Action (Manchester, 2004), pp. 62– 82.


85

influence of poetry and history has been cited in support of this point of view.66 But, as I will argue in more detail in Chapter 6, ps.-Dionysios’ complaint needs to be read in the context of the use of ekphrasis in declamation in general: he seems to be criticizing the relevance and the extent of these passages (just as he criticizes over-lengthy narrations elsewhere in his treatise) as well as the special status of ekphrasis in declamation, where there is no agreed basis in fact and therefore no limit to the description. Quintilian (2.4.3) makes a similar remark about students’ tendencies to indulge in elaborate descriptions. For Quintilian, this is a phase students go through, their youthful exuberance needs to be tempered as they learn; in fact, he finds this excess more promising than its opposite since it is easier to restrain existing practices than to create abilities that have never been developed. Quintilian’s use of the verb ‘lascivire’ paints this danger in a moral light, like Theon’s warning against useless detail (achrēstos, the term he uses, being highly morally charged). But it is a question of degree. Ekphrasis in oratory should be discreet and not easy to identify (Lysias’ narrations are often full of vivid detail, but it would be difficult to cut out any passage to show students as an example). It is possible that poetic and historiographical examples were favoured for this very reason. They had the advantage not only of being familiar to the students at this stage of their studies, but of being easy to identify and show to beginning students as models. The Progymnasmata therefore give us an elementary overview of ekphrasis as it was taught in the first stages of rhetorical education, to students who had just finished, or were still completing, their study of grammar. Reading the instructions for ekphrasis alongside the other exercises, and in comparison with other treatises, reveals a conception of ekphrasis as a process to be carried out on a basic statement of fact (diēgēsis), making the subject, whether an action, a person, a season or a place, visible to the mind’s eye of the audience. Ekphrasis is therefore part of an intimate communication between speaker and addressee which has an impact on the recipient which is always imaginative, and often emotional. So it is not surprising that the terms in which it is defined are very different from the terms in which modern description is defined; above all, it does not only have objects existing in space as its referent but has a temporal dimension. In fact, it is interesting to compare Aristotle’s claim in his discussion of metaphor in Rhetoric (1411b 24–5) that subjects in action (energeia) are more vivid (pro ommatōn) – it seems that there was an association between movement, and its rendering of space through 66

See, for example, Graham Zanker, Realism in Alexandrian Poetry: A Literature and Its Audience (London,1987), p. 40 and Alessandra Manieri, L’immagine poetica nella teoria degli antichi: phantasia ed enargeia (Pisa, 1998), p. 153.

86


time, and vividness; between energeia and enargeia.67 There is a certain associative logic in the way in which ekphrasis is both a force acting upon a listener and a means of depicting actions in words, while description, whose domain is (theoretically) confined to the object, is itself treated as object, to be dissected and analysed by the critic. What this contrast points to is the very different assumptions about language which underlie each definition. Ekphrasis, as is particularly clear from the discussions of enargeia which will be the subject of Chapters 4 and 5, belongs to a concept of language as a force, with the power to affect the listener.

67 See Lucia Calboli Montefusco, ‘Ἐvάργεια et ἐvέργεια: l’evidence d’une démonstration qui signifie les choses en acte’, in Mireille Armisen Marchetti (ed.), Demonstrare: voir et faire voir: forme de la démonstration à Rome (Toulouse, 2005), pp. 43–58; Manieri, L’immagine poetica, pp. 101–4; Guido Morpurgo-Tagliabue, Linguistica e stilistica di Aristotele (Rome, 1967), pp. 256–66.

4. Enargeia: Making Absent Things Present

Introduction: Ekphrasis and Enargeia The surviving Greek sources on ekphrasis show how it was taught at the elementary stage of rhetorical training and, as we will see in Chapter 6, help us to understand how ekphrasis could be put to use in the context of a larger speech. The basic information they provide about the use of ekphrasis in rhetorical contexts is invaluable, but they largely leave aside the reasons why an orator might need to use this technique. To find explicit discussion of these questions we need to turn to a Latin source, the Institutio oratoria of Quintilian, Theon’s contemporary. Quintilian covers the whole range of the rhetorical curriculum, reflecting the traditional teaching of Roman schools, theoretical sources and his own extensive experience of forensic pleading. He was clearly familiar with Greek rhetorical theory and many details of his teaching can be directly compared with Greek examples. What Quintilian adds in particular is the practical, personal perspective of the seasoned speaker, telling us how vivid language could actually be used, and what he knew its effect on an audience could be. The connections between Quintilian’s discussion of early rhetorical education and the Greek Progymnasmata have been explored, particularly by Italo Lana. Several of the elementary exercises he mentions have direct equivalents in the Progymnasmata, though nothing corresponds directly to the exercise of ekphrasis. (An earlier Latin treatise, once ascribed to Cicero, the Rhetorica ad Herennium, devotes a section to ‘descriptio’, which does correspond in many ways to the Greek treatments of ekphrasis, but this is presented as a technique to be used in a finished speech, not as an exercise.) In Quintilian’s handbook, something corresponding to the Greek ekphrasis is to be found but not in his discussion of elementary exercises; nor is ekphrasis identifiable with any single figure or trope. Instead, it is to be found in his various treatments of enargeia (Latin ‘evidentia’), the

�� On Quintilian’s sources, see Jean �� Cousin, Etudes sur Quintilien (2 vols, Amsterdam, 1967).�� For a direct comparison with Theon, see Italo �� Lana, Quintiliano, Il ‘Sublime’ e gli ‘Esercizi preparatori’ di Elio Teone: ricerca sulle fonti grece di Quintiliano e sull’autor ‘Del Sublime’ (Turin, 1951)�� . On the uses of enargeia in Cicero, see �� Ann Vasaly, Representations: Images of the World in Ciceronian Oratory (Berkeley, 1993)�� and �� Beth Innocenti, ‘Towards a theory of vivid description as practised in Cicero’s Verrine Orations’, Rhetorica, 12 (1994): 355–81. �� Lana, Quintiliano, Il ‘Sublime’ e gli ‘Esercizi preparatori’ di Elio Teone and Ian �� Henderson, ‘Quintilian and the Progymnasmata’, Antike und Abendland, 37 (1991): 82–99. Rhetorica ad Herennium, 4.39.51. For further discussion, see below.


88

quality of language that appeals to the audience’s imagination, in precisely the way that ekphrasis is said to do, and which is central to the definition of ekphrasis. Enargeia in Quintilian We saw in the previous chapter how Quintilian’s distinction between a plain statement of facts (‘narratio’) and a narration with enargeia corresponds to Nikolaos’ distinction between diēgēsis and ekphrasis. In Book 4 of the Institutio oratoria, Quintilian includes some remarks on this quality of vividness in his treatment of the statement of facts (‘narratio’). This placement suggests that, rather like the authors of the lost Greek Progymnasmata mentioned by ps.-Hermogenes, he considered enargeia as a quality to be added to various types of discourse. �� In addition to this brief mention of enargeia under the heading of narration, Quintilian discusses it at two further main points in his treatise: as a means of arousing the emotions (Book 6) and as a figure of speech which has the particularly vital role of ensuring that the listener is swayed by the speaker’s case (Book 8.3.67–9, cited above). These passages are of vital importance as Quintilian provides much of the information missing from the extant Greek treatises, explaining how enargeia contributes to the overall persuasive effect of a speech. He shows how important the ability to ‘place before the eyes’ was for the orator who wished to make his audience feel involved in the events of the case. More than that, he explains how the orator should achieve enargeia through the use of mental images (phantasiai). The way in which he describes this harnessing of the power of the imagination provides the missing link between language and sight and explains a great deal about how enargeia was thought to work. In particular, Quintilian’s explanation of the role of phantasia in both the creation and the reception of vivid language helps to explain why it was reasonable for him and other ancient theoreticians to speak of language’s ability to ‘make visible’ its subject matter, a claim that has been subjected to ridicule by at least one modern critic. The nature of phantasia, and its close association in ancient thought with the sense perceptions stored in memory, also help to explain how rhetoricians could be so seemingly sure of the precise impact of their

�� See Georges �� Molinié, Dictionnaire de rhétorique (Paris, 1992)�� , p. 145 (s.v. évidence): ‘Par-delà de la naïveté linguistique de la présentation traditionnelle (cette fameuse et ridicule suppression de l’écran du discours, avec l’idée que l’auditeur est transformé en spectateur), on sera particulièrement sensible au caractère social de cette qualité’ [Beyond the linguistic naivety of the traditional formula (the notorious and ridiculous suppression of the screen that of discourse, with the idea that the listener becomes a spectator), it is important to note the social and cultural aspects of this quality] (my italics).

Enargeia: Making Absent Things Present

89

appeals to the imagination, which is to us the most untrammelled and individual of faculties. This chapter and the next will therefore explore the theories of language, memory and the imagination which underlie the use of enargeia – and thus of ekphrasis – in rhetoric. Though the focus of this study is on the rhetorical contexts for ekphrasis, much of this theoretical background is relevant to the use of ekphrasis in other genres such as history or poetry – for the type of imaginative response which Quintilian assumes in an audience was considered typical of readers of different types of text. Enargeia and Persuasion As we have seen, Quintilian makes the same distinction as does Nikolaos between a plain statement of facts (di�� ē�� g�� ē�� sis/narratio) and a developed account which makes the audience feel present at the events described and emotionally involved in them. In forensic oratory this effect has two main functions. One is related to the orator’s duty to inform his audience of the facts of his case (naturally, the facts as he would like them to be seen, rather than as they necessarily happened). This can be achieved through a plain statement of facts, but a vivid account has the advantage of capturing the audience’s attention, of being more immediately memorable and comprehensible, like the paintings which Quintilian (6.1.32) says were sometimes displayed in courtrooms. To mention just some examples from classical Greek oratory, Lysias, identified as a master of enargeia by Dionysios of Halikarnassos, includes fast-moving yet vivid accounts of the domestic arrangements in the house of Euphiletos, who stands accused of murdering his wife’s lover, in his first oration (9–14), and of the murder of his brother in the twelfth oration, Against Eratosthenes (6–16). Demosthenes’ often cited evocation of the devastation wreaked on Phokis has already been mentioned. Another example is his carefully constructed narrative of the response in Athens to the news of the fall of Elatea, which similarly involves the reader through the inclusion of details such as the time of day (‘It was the evening …’) and the preparations for the emergency meeting of the Assembly. Kathy Eden has argued persuasively that theories of enargeia were in fact developed originally in classical Greece for such forensic contexts in which ‘the narrator set out to reproduce the vividness of ocular proof through language’ in the absence of physical evidence. Only later was this theory applied to discussions of poetry and

�� See Dionysios of Halikarnassos, Lysias, 7 and Graham Zanker, �‘Enargeia in the ancient criticism of poetry’, RhMus, 124 (1981): 297–311�. �� Demosthenes, De Corona, 169–171 (which is cited as an example of diatupōsis in Tiberios, De figuris Demosthenicis, 43).

90


other forms of literature. Enargeia in historiography or poetry can create similar effects, as was outlined in Chapter 1, but in forensic oratory it has the additional function of making the audience into virtual witnesses by making them ‘seem to see’ the events described by the speaker. �� In this, enargeia attempts to achieve the unattainable ideal of a judge who has witnessed the crime and thus can reach a decision that is certain and not subject to the vagaries of interpretation and probability�. Inseparable from this representational and informative function of enargeia is its ability to move the audience and to make them feel the emotions appropriate to the events described. This it shares with the pieces of physical evidence which were sometimes brought into Roman courtrooms: the swords, blood-spattered cloaks, fragments of bone and wounds which served both to confirm the speaker’s version of events and to arouse the emotions of the audience (6.1.30). In fact, Quintilian’s treatment of enargeia in Book 6 of the Institutio oratoria is part of a broader discussion of emotional appeals, and this combination of a feeling of presence and the arousal of the appropriate emotions is the key to the rhetorical function of ekphrasis: the audience should feel not just that they understood the facts and arguments intellectually, but that they were with the speaker, whether Demosthenes learning the fate of Elatea or Euphiletos, the wronged husband of Lysias 1, finding his wife in bed with her lover. Some examples of the use of ekphrasis in judicial and deliberative oratory – as reflected in the Greek treatises on declamation – as well as epideictic will be examined in greater detail in Chapter 6. Here, I want to explore some of the questions that are raised by the rhetoricians’ references to ‘placing before the eyes’, in particular to see whether Quintilian can provide some of the information, missing from the Progymnasmata, about how this effect can be achieved. Achieving Enargeia Quintilian’s activities as critic are subordinate to his role as a teacher and practitioner of rhetoric: his concern is not just to analyse texts, but to show his readers how to achieve those same effects. In the case of enargeia his advice varies between the enigmatic and the illuminating. One �� means of Kathy Eden, Poetic and Legal Fiction in the Aristotelian Tradition (Princeton, 1986)�� , pp. 72–3 and 91–2. This does not mean, of course, that the practice of enargeia did not exist in poetry prior to any theoretical developments. On vividness as a kind of proof in Ciceronian oratory, see also Vasaly, Representations, p. 25 and pp. 254–5 and Innocenti, ‘Towards �� a theory of vivid description’, p.�� 374. �� See Françoise Desbordes, ‘L’argumentation dans la rhétorique antique: une introduction’, Lalies, 8 (1990), p. 87: ‘Pour le juge, l’idéal est le flagrant délit …: il voit que X est coupable …’ [For the judge, the ideal is for the criminal to be caught in the act …: he can see that X is guilty.]


91

achieving enargeia which he and others mention is through the inclusion of details, and therefore corresponds to the definition of ekphrasis as an expanded and developed narration offered by Nikolaos. Quintilian provides the example of the sacked city cited in Chapter 3 to illustrate the enargeia that results from details (‘ex pluribus’, 8.3.66). The statement ‘the city has fallen’ may communicate the basic facts perfectly well, as he says (8.3.67), but in order to have an emotional effect on the listener, to ‘penetrate the emotions’, the speaker needs to ‘open up’ this basic statement and to describe the constituent actions: the flames, the collapsing roofs, the flight and lamentation of the inhabitants. Quintilian’s other example of enargeia ‘ex pluribus’ is a passage from a lost speech by Cicero describing the aftermath of a drunken party, in which the event is described through a selection of its visible consequences (Institutio oratoria, 8.3.66): Videbar uidere alios intrantis, alios autem exeuntis, quosdam ex uino uacillantis, quosdam hesterna ex potatione oscitantis. Humus erat inmunda, lutulenta uino, coronis languidulis et spinis cooperta piscium. I seemed to see some coming, others going, some staggering with the effect of the wine, some yawning from the previous day’s drinking. The floor was filthy, smeared with wine, covered with wilting wreaths and fish bones.

Here, the main event, the party, is conveyed through the description of different elements: the actions (pragmata) of the people who had been involved and the appearance of the place (topos) in which it had occurred, just as the account of the sacked city mentions the ruined and burning buildings, the distraught inhabitants and the victors. In the case of the party, the event is described not as it unfolds but through the signs which the speaker claims to have witnessed, and which he aims to make the audience witness as he describes them, a special type of effect which will be discussed in more detail in Chapter 6. In both cases, however, the peristaseis, the parts of narration which play such an important role in the Greek exercise of ekphrasis, would have been a useful aid in teaching students how to achieve a similar effect in their own compositions. Other authors also mention the inclusion of details as productive of enargeia. Dionysios of Halikarnassos attributes Lysias’ vividness to his �� On the use of the peristaseis in this passage, see �� Piet Schrijvers, ‘Invention, imagination et théories des émotions chez Cicéron et Quintilien’, in Actus: Festschrift H.L.W. Nelson (Utrecht, 1982),�� p. 403. See also Jean-Pierre Aygon�� , ‘L’ecphrasis et la notion de description dans la rhétorique antique’, Pallas, 41 (1994), p.�� 48.

92


inclusion of attendant circumstances (ta parakolouthounta) (Lysias, 7), by which he may mean the incidents which play no direct part in the case, such as the wife’s teasing of her husband in Lysias 1, but which play a vital role in establishing character, making us feel that we know the speaker. For the first-century BCE critic Demetrios, enargeia results from reference to detail (akribologia).10 For him, enargeia also derives from the attendant circumstances (ta sumbainonta), illustrated by the description by an unknown author of the sound of a farmer’s footsteps as he approaches, which conveys through a single detail the gait and demeanour of the character: ‘it was, for instance, once said of a countryman’s walk that “the noise of his feet had been heard from afar as he approached”, the suggestion being that he was not walking at all but stamping the ground, so to say’.11 The inclusion of telling detail is also important to the other type of enargeia which Quintilian mentions in Book 8. This is a brief sketch ‘in which an image of the whole is somehow depicted’.12 The examples are the boxing match from the Aeneid (5.426) and the brief description of Verres with his mistress from Cicero’s Verrine Orations, which Quintilian says affected him particularly as a reader. The difference between this and the previous type (‘ex pluribus’) may be that this second type expands the depiction of one element of narration (an action and persons, respectively), while the party and the sacked city both include references to the visual appearance of persons, places and actions. But, in the end, as Innocenti notes, these practical suggestions on how to achieve enargeia in the rhetorical handbooks are less than satisfying.13 Demetrios’ claim that enargeia is achieved by ‘leaving nothing out and cutting nothing short’ (to paralipein mēden mēd’ ektemnein, On Style, 209) appears to ignore the drastic selection of details which is entailed by any verbal account of any subject, his own examples being�� no �� exception. Quintilian does the same when he asks of Cicero’s description of the 10 �� Demetrios, On Style, 209–10, trans. W. Rhys Roberts. Demetrios’ other examples, a Homeric simile of a man diverting water to irrigate a garden (Iliad, 21.257) and a passage from the funeral games in Iliad, 23.379–81 (Diomedes’ pursuit of Eumelos in the chariot race) correspond to the type of scene-painting we are talking about. Some of the examples he goes on to discuss suggest a different sense of enargeia, as a type of verbal emphasis. See Manieri, L’immagine poetica, pp. 133–7. 11 �� Demetrios, On Style, 217. See Roos Meijering, Literary and Rhetorical Theories in Greek Scholia (Groningen, 1987), p. 41. As this example shows, sight is not the only sense involved in enargeia. 12 �� Quintilian, Institutio oratoria, 8.3.63: ‘est igitur unum genus, quo tota rerum imago quodammodo verbis depingitur’. On this type of image, see �� Alessandra Manierii, L’immagine poetica, pp. 141–4. 13 Innocenti, ‘Towards a theory of vivid description�� ’, p. 360.


93

party: ‘what more would anyone who had actually entered the room have seen?’14 Like Demetrios, he leaves aside the question of selection, though it is clear that any impact derives largely from the nature of the precise details Cicero includes. At the same time, Quintilian’s remark completely elides the distinction between the words and their imaginative effect, and between that effect and the perception of reality.15 These evasions reveal the extent to which the conception of words as provokers of images could overshadow attention to the more formal aspects of a text. To account for the effect of these passages we might prefer to focus on the use of figures, on the choice of vocabulary and the arrangement of words. But this is not the case for the ancient critics. Despite the fact that Quintilian’s accounts of the party and the sacked city occur in his discussion of style in oratory, he pays no attention to the language.16 Instead, he is astonishingly vague, saying simply of the latter, ‘If one opens up [the idea of the sack] there will appear …’, as if through an act of verbal conjuring. No reference is made to the verbal medium except as a portrait of what ‘appears’. At the same time, he seems to assume that the orator’s imagination (the scene that appears to him as he ‘opens up’ the brief statement), its verbal expression and the image which ‘appears’ in the audience’s mind as a result of these words are both simultaneous and identical, and that this image can be equivalent to the direct perception of a thing. Enargeia and Phantasia The vagueness about the linguistic aspects of enargeia and the confidence in its powers displayed by rhetoricians are significant. They point to the complexities of a phenomenon which goes beyond the normal functions of language and which can often only be expressed, as in the case of ekphrasis itself, by recourse to metaphor and simile.17 They are also a consequence of 14

�� Quintilian, Institutio oratoria, 8.3.67: ‘quid plus videret qui intrasset?’ �� See the comments of �� Eleanor Leach, The Rhetoric of Space: Literary and Artistic Representations of Landscape in Republican and Augustan Rome (Princeton, 1988), p. �� 17, who notes that ‘The spectator sees far more, since the orator’s selection and balance has conferred a neat and stylized order upon the chaotic picture of degeneracy he wants to evoke.’ 16 �� The figures of speech used in this example (as Aphthonios advised for ekphrasis) are all passed over in silence, despite the fact that Quintilian will go on to enumerate several of them in the course of Books 8 and 9: use of accumulation (sunathroismos – a technique of amplification discussed by Quintilian, Institutio oratoria, 8.4.27) underlined by polysundeton (breathless repetition of ‘et’ 9.3.51) and homoioptōton (9.3.78). It is also worth noting that, according to Tiberios, De figuris Demosthenicis, 43, some authorities differed on whether diatupōsis was a figure of speech or a figure of thought. 17 �� Manieri, L’immagine poetica, p. 148 notes that ‘ἐvάργεια è una nozione difficilmente racchiudibile negli schemi della precettistica retorica’ [enargeia is a notion that was difficult to encompass within the schemes of rhetorical theory]. 15

94


the way in which language and image were thought to interact in the mind of both speaker and listener. Quintilian’s understanding of the process emerges more clearly earlier in the Institutio oratoria when, in Book 6, he discusses the production of enargeia from the orator’s point of view and discloses his understanding of the psychological processes involved. In his discussion of how to arouse emotion in a murder case he suddenly breaks into a display of enargeia which is close in subject matter and in some of its details to Theon’s example of the vivid evocation of a murder to be used as part of a common place (109.3 –11, cited in Chapter 3, above), a subject that seems to have been a schoolmaster’s favourite: When I am lamenting a murdered man will I not have before my eyes all the things which might believably have happened in the case under consideration? Will the assailant not suddenly spring out, will the victim not be terrified when he finds himself surrounded and cry out or plead or run away? Will I not see the blow and the victim falling to the ground? Will his blood, his pallor, his dying groans not be impressed on my mind? This gives rise to ἐvάργεια, which Cicero calls illustratio and evidentia, by which we seem to show what happened rather than to tell it; and this gives rise to the same emotions as if we were present at the event itself. Hominem occisum queror: non omnia quae in re praesenti accidisse credibile est in oculis habebo? non percussor ille subitus erumpet? non expauescet circumuentus, exclamabit uel rogabit uel fugiet? non ferientem, non concidentem uidebo? non animo sanguis et pallor et gemitus, extremus denique expirantis hiatus insidet? Insequetur ἐvάργεια, quae a Cicerone inlustratio et euidentia nominatur, quae non tam dicere uidetur quam ostendere, et adfectus non aliter quam si rebus ipsis intersimus sequentur. (6.2.31–2)

As in Theon’s version, the ‘showing’ is not subjected to any form of linguistic analysis. But it is clear from this passage why this is the case: enargeia is conceived of as the result of an internal, psychological process. For Quintilian, the first step in producing enargeia in a speech was for the orator to imagine the subject for himself. To borrow Quintilian’s own image from Book 8, the idea of murder is ‘opened up’ in the orator’s own mind and some key phases of the action are organized into a temporal sequence, from the aggressor’s first appearance on the scene to the victim’s dying breath. His account of the scene is presented as a description of the mental process in the form of a series of rhetorical questions describing the visual and aural details he would summon up for himself. Quintilian makes no explicit comment on his use of language, preferring instead


95

to enact the production of enargeia. And, in the very act of verbalizing his exercise in visualization, he demonstrates enargeia, like the magician pulling the rabbit out of his hat. His seamless progression from precept to practice, combined with his reticence about language, suggests that words are important to this process only as the means by which an internal, mental image is conveyed from speaker to listener.18 Immediately before this display of enargeia in action Quintilian had given more details about the speaker’s mental processes. The speaker is said to summon up his own vision of the scene by means of phantasiai, a Greek term glossed in Latin as ‘visiones’. It is a process which he compares to the spontaneous habit of daydreaming (he mentions travelling, fighting and possessing wealth as common fantasies) and which, he claims, the speaker can cultivate in himself as part of his training as an orator: What the Greeks call phantasiai (we shall call them ‘visiones’, if you will,) are the means by which images of absent things are represented to the mind in such a way that we seem to see them with our eyes and to be in their presence. Whoever has mastery of them will have a powerful effect on the emotions. Some people say that this type of man who can imagine in himself things, words and deeds well and in accordance with truth is ‘good at imagining’ (euphantasiōtos). Quas φαvτασίας Graeci uocant (nos sane uisiones appellemus), per quas imagines rerum absentium ita repraesentantur animo, ut eas cernere oculis ac praesentes habere uideamur, has quisquis bene ceperit, is erit in adfectibus potentissimus. Quidam dicunt εὐφαvτασίωτov, qui sibi res uoces actus secundum uerum optime finget. (6.2.29–30)

For Quintilian, therefore, effective enargeia in oratory – and in poetry, as he goes on show – is the result of a controlled and conscious process of visualization. He seems to believe that, although this ability, like the others required by the speaker, varied from individual to individual, it could be developed by training. So it does seem, as suggested in Chapter 1, that the expectation of visual response was far more widespread and developed among ancient audiences than among modern readers. Not only was it talked about and theorized, but students were encouraged to imagine and 18

�� See Mireille �� Armisen, ‘La notion d’imagination chez les anciens II: la rhétorique’, : ‘Tout en passant par le logos, écrit ou parlé, la phantasia rhétorique Pallas, 26 (1980), p. 14�� annule cet intermèdiaire et se résout en une nouvelle représentation mentale qui est l’écho dans l’esprit de l’auditeur de l’image initiale conçue par l’orateur’ [Even as it is transmitted by the word (whether written or spoken) rhetorical phantasia cancels out this intermediary and takes the form of a new mental image which is the echo, in the listener’s mind, of the initial image conceived by the orator].


96

to assume that the texts they read were the products of others’ powers of visualization. All this must have encouraged an awareness of visualization as an important element of both reading or listening to others’ work and creating one’s own. Communicating Images: Enargeia and Phantasia In his own examples of enargeia (the murder and the sacked city of Book 8) Quintilian assumes a live performance situation in which the transmission of mental images and their concomitant emotions between a speaker and his audience is a vital part of rhetorical interaction in the forum or school. Ps.-Longinos describes the process in a similar way, with similar vocabulary, explaining phantasia as resulting ‘when, under the effects of inspiration and passion, you seem to see what you are speaking about and bring it before the eyes of your listeners’ [ὅταv ἃ λέγεις ὑπ' ἐvθoυσιασμoῦ καὶ πάθoυς βλέπειv δoκῇς καὶ ὑπ' ὄψιv τιθῇς τoῖς ἀκoύoυσιv].19 This formulation rolls the distinct moments of Quintilian’s account into one single process, termed phantasia. Both make more specific the vague, uncontextualized reference in the Progymnasmata to the ‘eyes’ (opsis) to which the subject of an ekphrasis (to dēloumenon) is displayed. These are the mind’s eyes of the audience to whom the speech is addressed and who form the equivalent of a painting, or rather a set of moving impressions, in their imagination. Ps.-Longinos’ use of the single term phantasia to encompass the author’s imagination, the words he utters and the resulting impression in the listener’s mind reveals the intimate connections between mental images and the words that both result from and create them. Words and mental images are not the only phenomena to be fused in this way, for both Quintilian and ps.-Longinos assume that what the audience will feel that they can ‘see’ is the same as what the orator ‘sees’ and that the listener or reader will share the vision that he has created in his mind. In this way, the speaker’s visual image is assumed to be transmitted to the audience through the medium of words and to give rise to a comparable image in their minds. Accessing the Author’s Imagination Both Quintilian and ps.-Longinos write as if vivid language can actually give access to the mind which gave rise to it. That is, if enargeia arises from mental images, it must be possible to work back up the chain and to reconstruct the creative process, or rather the original mental image which gave rise to the words that prompt the reader’s own mental image. 19

�� Ps.-Longinos, On the Sublime, 15.1�.


97

Both authors make comments to this effect with reference to examples of enargeia in poetry. Ps.-Longinos (15.2) remarks that Euripides, when he wrote Orestes, 255–7 (‘Mother, I beg you, do not set upon me the bloodstained, snakelike virgins! Here they come, pressing upon me!’), must have seen the Furies who Orestes imagined were tormenting him. ‘Here,’ he claims, ‘the poet saw the Furies himself and compelled the audience to almost see what he imagined (ephantasthē)’.20 In his discussion of enargeia in Virgil, Quintilian likewise asks, ‘had the poet not thoroughly (‘penitus’) conceived an image (‘imago’) of death to be able to say [of the Argive Antores] ‘and, dying, he remembered sweet Argos?’21 These claims rely on the assumption that the poet himself worked by creating mental images that he then expressed in words (as Aristotle recommended to the tragic poet for other reasons in the Poetics, 1455a) and that these same words can therefore allow the reader to access and share this image. So it is with the orator: both the speaker and the listener are involved in the creation of mental images and both are metaphorical painters of these immaterial images. In both these examples of poetic enargeia the evocation of a character’s perceptions – Orestes’ hallucination and Antores’ final memories – plays a vital role. In the second example, as Jean-Pierre Aygon has noted, Quintilian has in mind a passage several lines long.22 But the single line quoted does seem to bear a special significance: Antores’ death is conveyed by the briefest of evocations of the final thoughts and images that passed through his mind, while Orestes’ words contain direct allusions to the appearance of the Furies. In both cases, the poet’s mental image, to which the reader has access through the text, is itself a representation of the character’s mental image.23 There is thus a chain of images which ultimately allows the reader or listener both to share in the experience of the character and to admire the skill of the poet.

ἐvταῦθ' ὁ πoιητὴς αὐτὸς εἶδεv ’Εριvύας. ὃ δ' ἐφαvτάσθη, μικρoῦ δεῖv θεάσασθαι καὶ τoὺς ἀκoύovτας ἠvάγκασεv. As Armisen, ‘La �� notion d’imagination chez les anciens II�� ’, p. 15 notes, Euripides’ passage works less by the description of details than by its emotional intensity, as revealed by the viewer’s response. On this passage, see also Jean-Louis Labarrière, ‘Faut voir à voir! Considérations pseudo-Longiniennes’, Métis, n.s. 4 (2006): 71–93. 21 �� Quintilian, Institutio oratoria, 6.2.33: ‘non idem poeta penitus ultimi fati concepit imaginem, ut diceret: “Et dulces moriens reminiscitur Argos?”’ (Aeneid, 10, 782). 20

22 �� Jean-Pierre Aygon, ‘“Imagination” et description chez les rhéteurs’, Latomus, 63 (2004), pp. 118–19. 23 �� Although it is not stated by Virgil, �� it is likely that Quintilian assumed that Antores’ memory of his homeland itself took the form of a visual image, given the close association between sense perceptions and memory in ancient thought, on which, see below.


98

The Impact on the Listener Quintilian’s discussion of enargeia reveals a conception of language as an active force.24 Both he and ps.-Longinos use powerful, physical imagery to express the impact of vivid speech on the listener, a reminder that theirs is still a predominantly oral conception of language.25 For Quintilian, vivid language ‘penetrates the emotions’, while for ps.-Longinos rhetorical phantasia, the end result of which is enargeia, ‘not only persuades but enslaves the listener’ [oὐ πείθει τὸv ἀκρoατὴv μόvov, ἀλλὰ καὶ δoυλoῦται]�.26 When, in Book 8 of the Institutio oratoria, Quintilian contrasts the plain statement of facts (narratio) with the vivid version of the same events he distinguishes the effect of the two types of discourse in terms of their physical impact on the listener. The plain statement reaches only the ears while the vivid version, the equivalent of ekphrasis, ‘displays the subject to the eyes of the mind’: It is a great skill to be able to speak of our subject matter clearly and so that it seems visible. For the speech is not sufficiently effective, nor does it have as complete a control as it should if it merely reaches the ears so that the judge thinks that the facts of the case are being told to him and not expressed in full and displayed to his mind’s eye. Magna uirtus res de quibus loquimur clare atque ut cerni uideantur enuntiare. Non enim satis efficit neque, ut debet, plene dominatur oratio, si usque ad aures ualet, atque ea sibi iudex de quibus cognoscit narrari credit, non exprimi et oculis mentis ostendi. (8.3.62)

The implication is that vivid language reaches different parts of the mind, penetrating more deeply into the listener to reach the ‘mind’s eye’ (the Latin distinguishes between inner and outer senses of sight, where our Greek sources do not). The distinction between words which stay on the surface of the body, by which Quintilian presumably means plain statements of fact and arguments, and those which penetrate inside to appeal to the ‘eyes of the mind’, reveals a conception of the human body as permeable and of words as a quasi-physical force, both of which are familiar from 24

�� As noted by Molinié, Dictionnaire de rhétorique, p. 145: ‘[Le discours] est appréhendé ainsi comme acte, come dynamisme qui peut même modifier la réalité de la situation existentielle, concrète, des partenaires dans la relation oratoire’ [[Speech] is thus perceived as an action, as a dynamic force which can even change the reality of the actual situation in which the partners in the rhetorical exchange find themselves]. 25 �� On an understanding of language as power and action as typical of oral cultures, see Walter �� Ong, Orality and Literacy: The Technologizing of the Word (London and New York, 1988), pp.�� 31–2. 26 �� Ps.-Longinos, On the Sublime, 15.9.


99

other ancient sources.27 The difference is more than a technical distinction between ‘showing’ and ‘telling’. Instead, it lies in the way each mode of discourse is received by the listener: enargeia derives from the innermost recesses of the speaker’s mind and works its way inside the listener to produce its intense effect. So the judge must be made to feel not just that he is hearing the facts of the case, but that he can actually see the events playing out before his eyes.28 Such a display is by no means sufficient by itself. It is clear from Quintilian’s discussion that rhetorical enargeia is just one weapon in the orator’s arsenal and this is echoed by ps.-Longinos when he introduces an important qualification to the dominating power of enargeia in oratory: it� is only when used in combination with factual arguments that phantasia works this effect. When an orator does manage to combine the two he is, by implication, invincible.29 Both supply a term that was missing from most of discussions of ekphrasis in the Progymnasmata: the emotion which accompanies the mental image ‘placed before the eyes’.�� For ps.-Longinos, phantasia, whether in poetry or oratory, aims for emotive effect (to pathētikon) and to stir the listener (to sunkekinēmon).30 For Quintilian, a brief statement (‘brevis nuntius’, diēgēsis) in a speech conveys the basic information adequately but ‘penetrates the emotions’ less than the expanded version (narration with enargeia, ekphrasis).31 Indeed, �� the mental image itself is almost superfluous. Its function is to arouse the desired feeling in the listener, whether the vague feeling of pleasure (hēdonē) that Nikolaos ascribed to ekphrasis in epideictic contexts or the feeling of indignation (deinōsis) or its concomitant, pity, associated with accounts of the fall of

27

�� Even at the elementary level of the Progymnasmata, Theon (72.1–2) recommends that students should aspire to make their words ‘dwell in the mind’ of their listeners. Plutarch, Moralia, 37F–38B recommends ear-guards to protect the young. The eyes were considered particularly vulnerable openings – see below on the effect of the sight of the beloved upon the viewer. On words as physical entities in Stoic thought, see Catherine �� Atherton, The Stoics on Ambiguity (Cambridge, 1993), �� p. 44 on Diogenes Laertios, 7.55–6. See also �� Gorgias’ corporeal logos (Encomium of Helen, 8)�. 28 �� Leach, The Rhetoric of Space, p. 15 interprets Quintilian’s words as implying the ‘production of a static image’, but Quintilian’s wording allows for events to be ‘displayed’ as developing scenarios, as the parallel with narratio suggests. See also above on the association between enargeia and movement in the scene described. 29 �� Ps.-Longinos, On the Sublime, 15.9. The further comments about the potentially overwhelming impact of phantasia in rhetoric at 15.11 need to be read in this context. 30 �� Ps.-Longinos, On the Sublime, 15.2. 31 �� Quintilian, Institutio oratoria, 8.3.67: ‘in adfectus minus penetrat brevis hic velut nuntius’. In Theon, Progymnasmata, 71, ll. 31–2, too, it is words that are vivid (enargēs) that will inhabit the listeners’ minds.

100


cities.32 Even when the effect is not what we might strictly speaking call ‘emotional’, the physical understanding of the impact of enargeia means that a reader/listener who conceives an image of any kind in his or her mind is still undergoing a pathos of some kind as he or she experiences the words’ effect.33 Enargeia and the Feeling of Presence Quintilian’s discussion of enargeia in Book 6 grows out of a broader discussion of emotion and is introduced as a means to an end. The use of visualization is first introduced as a means of ensuring that the speaker is appropriately involved in the version of events he is presenting. For, Quintilian claims, in order to move an audience, the speaker must himself be moved (6.2.26; cf. Horace, Ars Poetica, 102). He draws an analogy with actors whom, he says, he has seen leaving the stage in tears because of their intense involvement in the plot (6.2.27–8 �� and 35)�� . In terms of the effect on the audience of judicial oratory or declamation, the analogy with the actor is particularly apt for, like the audience in the theatre, they are made to be witnesses of events from the past. Quintilian �� emphasizes this time-warping action of enargeia in a further discussion of the art of ‘placing before the eyes’ (‘sub oculos subiectio’) in Institutio oratoria, 9.2.40–1 (here he treats vivid presentation as a figure called variously ‘evidentia’, hupotupōsis and diatupōsis). Because of its capacity to make the audience feel present at past and future events, this effect is also called ‘translatio temporum’ or, in Greek, metastasis or metathesis (a ‘transference [of time]’) (9.2.41).34 The listener may be transported either to the past, or to a hypothetical future, as is shown in the passage Quintilian cites from Cicero’s Pro Milone (33) which paints a picture of the future consequences

32 �� Demosthenes, 19.65 is cited by Nikolaos as an example of the use of ekphrasis to achieve deinōsis (see above, Chapter 3, p. 76). See also the Rhetorica ad Herennium, 4.39.51 and Cicero, De inventione, 1.54.104 and 107. 33 Phantasia itself was understood as a type of pathos. See Frédérique Ildefonse, ‘Evidence sensible et discours dans le stoïcisme’, in Carlos Lévy and Laurent Pernot (eds), Dire l’évidence (Paris, 1997), p. 116. 34 �� See Meijering, Literary and Rhetorical Theories, p. 42. The Greek term metastasis appears in modern editions of Quintilian, but Isidore of Seville, De rhetorica, 34 uses metathesis in this sense: ‘Metathesis est [figura] quae mittit animos iudicum in res praeteritas aut futuras’ [Metathesis is [the figure] which sends the minds of judges to past or future events] (the examples are an appeal to remember the sack of a city and the evocation of future troubles from Pro Milone, 33, an example also used by Quintilian). I will follow Isidore in using metathesis in the sense of ‘transference of time’ to avoid confusion with the technical sense of metastasis in declamation to mean ‘transference of blame’.


101

of an action.35 Metathesis is more than a simple switch in the time referred to in the speech. Instead, the speaker is said to transport his audience in imagination back or forward to the events in questions, making use of the ability of enargeia to make them feel ‘as if’ they were present. Enargeia, as discussed by Quintilian, shares this power with poetry (as is evident from his frequent use of poetic examples and from ps.-Longinos)36 and with historiography. The subtlest discussion of the similarities and differences between enargeia in oratory and poetry is to be found in ps.Longinos, who gives serious consideration to the nature and function of visualization in poetry, unlike the rhetoricians who, on the whole, use poetic examples for their own ends. As we have seen, he states that both aim to move and stir the reader or listener, but this statement is prefaced by an attempt to explain the differences between what he terms ‘phantasia’ in poetry and oratory: It will not escape your attention that rhetorical phantasia has one aim while, when used by poets, it has another nor that the goal of phantasia in poetry is astonishment (ekplēxis) while that of rhetoric is enargeia, though both seek to create an emotional effect and excitement.37 ὡς δ’ ἕτερόv τι ἡ ῥητoρικὴ φαvτασία βoύλεται καὶ ἕτερov ἡ παρὰ πoιηταῖς oὐκ ἂv λάθoι σε, oὐδ’ ὅτι τῆς μὲv ἐv πoιήσει τέλoς ἐστὶv ἔκπληξις, τῆς δ’ ἐv λόγoις ἐvάργεια, ἀμφότεραι δ’ ὅμως τό τε <παθητικòv> ἐπιζητoῦσι καὶ τό συγκεκιvημέvov.

Longinos’ attribution of enargeia to rhetoric alone is unusual in that the term is frequently used elsewhere of vividness in poetry. He makes his distinction clearer at the end of his discussion of phantasia when he returns to its use in rhetorical contexts after an extended discussion of passages from tragedy: Nevertheless, in the poets, phantasia involves a type of exaggeration that is more suitable to myth and, as I have said, goes beyond the bounds of credibility, while the best form of rhetorical phantasia respects the possible (emprakton) and the true (enalēthes) …

35 �� Ps.-Longinos, On the Sublime, 15.9 cites a similar attempt to play ‘let’s suppose’ from Demosthenes, 24.208 (also from the end of a speech). These examples correspond to the third type of diatupōsis identified by Tiberios, De figuris Demosthenicis, 43 (the evocation of hypothetical events). 36 �� As noted in Chapter 3 above, ps.-Hermogenes also compares diaskeuē to poetry. 37 �� Ps.-Longinos, On the Sublime, 15.2.


102

oὐ μὴv ἀλλὰ τὰ μὲv παρὰ τoῖς πoιηταῖς μυθικωτέραv ἔχει τὴv ὑπερέκπτωσιv, ὡς ἔφηv, καὶ πάvτη τò πιστòv ὑπεραίρoυσαv, τῆς δὲ ῥητoρικῆς φαvτασίας κάλλιστov ἀει τò ἔμπρακτov καὶ ἐvάληθες ...38

As Roos Meijering points out, poets have a greater freedom than orators in that they are able to ‘place before the eyes’ scenes that are fantastic and impossible in the real world (such as Furies or Phaethon’s heavenly ride in the chariot of the Sun) as well as scenes that are lifelike.39 For orators to pretend to see Furies (as ps.-Longinos claims contemporary speakers do) is an absurdity that he condemns as inappropriate. As the first contrast between poetry and rhetoric suggests, a further difference lies in the effect on the listener of such visualization. While poets need only create a stunning impact (ekplēxis) on the listener, orators need to shape and control both the subject matter and its presentation, subordinating it to the specific requirements of the speech. Firstly, the subjects of rhetorical enargeia must be like truth (enalēthēs). In this, rhetorical visualization does have the connotations of direct perception that are present in the root sense of the terms enargeia and enargēs: they are at least close to the world that is the object of that perception.40 But this closeness is only relative, in comparison to the fantastic subjects of poetic phantasia. In underlining the need for the images used by orators to be in accordance with truth, ps.-Longinos echoes other sources on enargeia and related terms.�� Several times Quintilian repeats the key advice that, when summoning up an image to be conveyed in words, the speaker should be careful to confine himself to what is ‘like truth’ or ‘what usually happens’ or what is ‘credible’.41 His attribution of power to the orator who is euphantasiōtos (6.2.30) contains the important qualification that the images must be ‘secundum uerum’, ‘in accordance with truth’.�� This requirement is not surprising. Probability and verisimilitude (distinct ideas but with a considerable amount of overlap in practice) were the hallmarks of the

38

�� Ps.-Longinos, On the Sublime, 15.8. �� Meijering, Literary and Rhetorical Theories, pp. 71–2. 40 �� See Barbara Cassin, ‘Procédures sophistiques pour construire l’évidence’, in �� Carlos Lévy and Laurent Pernot (eds), Dire l’évidence: philosophie et rhétorique antiques (Paris, 1997), pp. 15–29; On �� the meanings of the term, see Manieri, L’immagine poetica, pp. 105–12�� and Juliette Dross�� , �� ‘De la philosophie à la rhétorique: la relation entre phantasia et enargeia dans le traité Du sublime et l’Institution oratoire’, Philosophie antique, 4 (2004): 61–93. 41 �� Quintilian, Institutio oratoria, 6.2.30, the speaker should visualize ‘secundum verum’ [according to truth]; 6.2.31 ‘omnia quae in re praesenti accidisse credibile est’ [everything that could credibly have happened in the present case]. In his discussion of diaskeuē in On Invention, 3.15 (p. 167, ll. 11–17), ps.-Hermogenes similarly emphasizes the need for such depictions to be credible (pithanos) and probable (eikos). 39


103

rhetorical presentation of facts.42 The reasons seem obvious: a narrative needs to conform to an audience’s expectations of what is likely or probable and to their experience of the world if it is to be believed. As Quintilian notes (4.2.34), the question of truth is a different matter: many true things are not credible, whereas untrue things can be like truth (‘uerisimilia’). It is this discussion that leads to Quintilian’s first mention of enargeia (4.2.63– 5), where he insists, in opposition to some other nameless authorities, that a false story needs to be as vivid (‘euidens’, enarg�� ē�s) as possible if it is to be convincing. He thus makes clear that likeness to truth is a quality that results from the orator’s presentation of his story, rather than from the underlying facts themselves.43 The concern for verisimilitude that we see in the discussions of enargeia is one indication of its rhetorical function for the quality of ‘likeness to truth’ was vital in narrations just as the ideas of probability and credibility played a vitally important role in rhetorical argumentation. The multiple interconnections of enargeia and ekphrasis with these notions will be explored further in the next chapter. But before moving on to this, it is important to note another, rather different aspect of enargeia. Enargeia and Illusion Ultimately, whether in poetry, rhetoric or historiography, whether it represents credible or incredible things, verbally produced enargeia is always a matter of illusion. It thus exists in a constant tension between presence and absence. The fact that enargeia produces an illusion is present in the terms for likeness which are used everywhere, as in the terms of approximation (‘all but’) we find in the discussions of ekphrasis. For Plutarch, the reader of Thucydides is merely ‘like’ a spectator; Quintilian’s ideal audience are made to feel ‘as if’ (‘quam si’) they were present at the murder. Moreover, even as Plutarch stresses the overwhelming sense of involvement in the events of Athenian history created by Thucydides’ narrative (Moralia, 346F; see Chapter 1, above), he underlines the illusion involved by referring to Simonides’ comparison of poetry to painting, the art of visual representation. The comparison of enargeia to the visual arts, made explicitly by Plutarch and implicitly elsewhere, also draws attention 42

�� Quintilian, Institutio oratoria, 4.2.31 treats the terms ‘uerisimilis’, ‘probabilis’ and ‘credibilis’ as interchangeable. On these terms, see, for example, Claude Moussy, ‘Signum et les noms latins de la preuve: l’héritage de divers termes grecs’, in Jacqueline Dangel (ed.), Grammaire et rhétorique: notions de Romanité (Strasbourg, 1994). 43 �� I would therefore not agree with Cockcroft, ‘Fine-tuning Quintilian’s doctrine of rhetorical emotion’, p. 503, who suggests that ‘Quintilian would be likely to favour a literal and unambiguous factual representation of that res originally taken into his mind and now again visualised.’

104


to the illusion involved in enargeia and to the way in which verbal depiction creates a feeling like that of presence which is not presence. A further analogy, which similarly conveys both the power of enargeia and the unreality of its creations, is with the stage. When Nikolaos compares the audience of ekphrasis to spectators, the term he uses, theatai, implies the theatre just as much as the visual arts. Quintilian’s suggests an even closer link to the stage with his comparison of the orator who uses phantasia to arouse his own emotions with the actor who is so involved in his role that he cries real tears (6.2.27–8 �� and 35)�� . In this respect the orator and the actor both share the paradoxical state of feeling imaginary things as if they were real, a state which they are able to induce in their audiences as well.44 Quintilian attributes some of the force of the speaker’s display of feeling to the fact that he is speaking of real events in which he is to some extent involved (indirectly as advocate, if not directly), but he makes clear that fictional events can affect the listener equally. He even recommends that schoolboy declaimers should think themselves into the roles they assume in their exercises, advice which assumes that emotional involvement in a fictional situation is both possible and desirable (6.2.36). Some sources reveal a certain anxiety about this identification: the Elder Seneca, for example, records anecdotes about orators who lost the ability to distinguish between their own mental images, their phantasiae, and reality and fell into a state of madness akin to that of Orestes.45 It may be to distinguish the type of phantasia involved in rhetorical enargeia from the uncontrolled hallucinations of a madman that Quintilian chose to compare the orator’s imaginings with daydreams, which remain under the dreamer’s conscious control.46 However, it is noticeable that the loss of ability to distinguish between visualization and reality is not attributed to the audience, nor is it a common idea in the Greek sources. The difference may reflect the different contexts of Greek and Roman rhetorical practice for, while the Greek theorists wrote for speakers who, in theory at least, discussed and evoked events that had occurred to them, the Roman practice of using an advocate meant that speakers in judicial cases were commonly required to become emotionally and imaginatively involved in events that concerned their clients rather than themselves. 44

See Schrijvers, ‘Invention, imagination et théories des émotions chez Cicéron et Quintilien’, pp. 395–408. 45 �� See Juliette Dross, ‘De l’imagination à l’illusion: quelques aspects de la phantasia chez Quintilien et dans la rhétorique impériale’, Incontri triestini di filologia classica, 4 (2004– 2005): 273–90. 46 �� Quintilian, Institutio oratoria, 6.2.30. Quintilian describes these reveries as a ‘vice of the soul’ (animi vitium).


105

Different texts play in different ways on the paradoxes of enargeia, on the simultaneous presence of the absent and absence of presence that are produced. Some examples of the ways in which authors of epideictic oratory and creators of the fictional worlds of declamation and the Greek novel exploit the ambiguities of the illusions they create will be discussed in further detail in Chapter 7. For the moment, however, what concerns us is the rhetorical use of ekphrasis and enargeia as a means of persuasion. While never losing sight of the illusion involved, rhetoricians tend to place emphasis on the ability of words to create presence, rather than the problematic nature of that presence. This emphasis is hardly surprising, and particularly appropriate to the communicative situation of oratory, where the audience is caught up in the immediacy of the speech and its performance, with little time to reflect critically on the nature of their experience. Conclusions Enargeia is therefore far more than a figure of speech, or a purely linguistic phenomenon.47 It is a quality of language that derives from something beyond words: the capacity to visualize a scene. And its effect also goes beyond words in that it sparks a corresponding image, with corresponding emotional associations, in the mind of the listener. This process lies behind the uses of enargeia and ekphrasis in oratory and in other types of text. The discussions of enargeia in rhetorical sources are particularly interesting because of the very careful attention paid to audience response and the question of how an audience’s verbally induced imaginings can be predictable, as they must be if the speech is to be successful. On the surface, the ancient orators’ model of ‘opening up’ and developing a basic statement of fact into a detailed and vivid account bears some similarities with the theory of description proposed by Philippe Hamon in his article of 1972, further developed in his book of 1981.48 Hamon proposes a definition of description as the ‘exploration of the lexical field’ in which an introductory theme or ‘pantonyme’, such as ‘garden’ or ‘house’ gives rise to the enumeration of a series of ‘sub-

47 Armisen, ‘La notion d’imagination chez les anciens II’, pp. 13–14 makes the same point about phantasia in rhetoric. 48 �� Philippe �� Hamon, ‘Qu’est-ce qu’une description?’, Poétique, 112 (1972): 465–85 and Introduction à l’analyse du descriptif (Paris, 1981), subsequently reissued under the title Du descriptif (Paris, 1993). I cite the later edition here and elsewhere.�� For a critique of Hamon’s approach to description, see �� Jean Molino, ‘Logiques de la description’, Poétique, 91 (1992): 363–82.

106


themes’ (rose, chimney) and predicates (snow, smoking).49 This strictly linguistic approach has the advantage of breaking with the long-standing tradition of defining description by its subject matter and the further advantage of removing the representation of reality from the equation (since description is seen as a purely lexical phenomenon).50 On �� a formal level, Hamon’s definition of description is comparable to Quintilian’s idea of ‘unfolding’. Just as striking, however, are �� the radical differences in orientation and in the basic assumptions about language which underpin the two approaches. Quintilian is almost perverse in his refusal to analyse the linguistic aspects of enargeia, but his insistence on the imaginative engagement of both speaker and listener draws attention to a factor which is left out of Hamon’s account of description. In this, Hamon typifies the modern critical neglect of imaginative response. What Hamon does share with the ancient rhetorical discussions of enargeia is an interest in the role of the reader/audience and the contribution demanded of them. But, for Hamon, description engages the lexical competence of the reader.51 The two differ so radically in their approaches that it is clear they are speaking from very different worlds. Where Hamon sees only words, Quintilian sees words as the communicable aspect of mental images which are crucial to the formation and reception of those words. Quintilian and Hamon are working with radically different conceptions of language, and of the relation of words to external reality on the one hand, and to their speakers and audiences on the other.52 In order to understand the ancient rhetorician’s perspective a little better, it is necessary to move on to the theories of imagination and memory on which enargeia and ekphrasis are predicated. These explain some of the questions raised by the use of enargeia in rhetoric while shining a spotlight onto some of the paradoxes involved in making absent things present.

49

�� Hamon, ‘Qu’est-ce qu’une description?’, pp. 474–6; Du descriptif, p. 128. �� See particularly Hamon, Du descriptif, p. 91. 51 �� Ibid., pp. 41–2. 52 �� See the remarks of Claude Calame, ‘Quand dire c’est faire voir: l’évidence dans la rhétorique antique’, Etudes de lettres, 4 (1991): 13–14 on the differences between enargeia and description. 50

5. Phantasia: Memory, Imagination and the Gallery of the Mind

Enargeia in Oratory Enargeia, the quality that makes an ekphrasis an ekphrasis and distinguishes it from a plain report of the facts, is thus a paradoxical phenomenon. It is able to arouse emotions through immaterial semblances of scenes that are not present to the listener and may never have taken place. It uses the medium of language to create an impact on the world, the power of which is expressed in physical terms – most strikingly, as we have seen, by ps.-Longinos’ metaphor of enslavement. A question that remains to be answered is how a speaker could be so confident of his effect on the most intimate recesses of his listeners’ minds. Quintilian cites several examples of emotional appeals involving props and extras (particularly children) falling disastrously flat, but he never betrays any doubts about the predictable power of enargeia. This is despite the fact that, as Ann Vasaly has noted, enargeia would seem to have depended on one of the least predictable factors in the equation: the individual’s unique and personal visualization. The listener’s emotional response to enargeia was, moreover, crucial to its function within a speech, which raises the further question of how orators could hope to predict and control such a seemingly individual and subjective process. As Vasaly points out, orators clearly expected their listeners to make their own, personal contributions to the word pictures they heard and to flesh out the details mentioned with further details supplied from their own imagination. Quintilian’s account of his own response to Cicero’s portrait of Verres does show that he assumed listeners would contribute actively to the images provoked by verbal enargeia and that these contained a personal input. Describing the effect this passage worked on his imagination, Quintilian is happy to admit that his own mental image

Quintilian, Institutio oratoria, 6.1.37–42. But see Libanios, Autobiography (Or. 1), 41, discussed below, for an example of the failure of an ekphrasis in an epideictic speech. Ann Vasaly, Representations: Images of the World in Ciceronian Oratory (Berkeley, 1993), pp. 98–9. Interestingly, the modern theoretician of rhetoric Chaim Perelman, in L’Empire rhétorique: rhétorique et argumentation (Paris, 1977), p. 49, expresses reservations about his ancient predecessors’ advice to use physical props (such as Quintilian’s swords), which may distract the audience, but seems to share their confidence in the verbally created sense of presence.

108


contains details not even mentioned by Cicero. What is more, he presents such elaboration as the norm: Is there anyone so incapable of forming images of things that, when he reads the passage in the Verrines, ‘The praetor of the Roman people stood on the shore dressed in slippers, wearing a purple cloak and long tunic, leaning on this worthless woman’, he does not only seem to see them, the place, their appearance, but even imagines for himself some of those things which are not mentioned. I for my part certainly seem to see his face, his eyes, the unseemly caresses of both and the silent loathing and frightened shame of those who were present. An quisquam tam procul a concipiendis imaginibus rerum abest ut non, cum illa in Verrem legit: ‘stetit soleatus praetor populi Romani cum pallio purpureo tunicaque talari muliercula nixus in litore,’ non solum ipsos intueri uideatur et locum et habitum, sed quaedam etiam ex iis quae dicta non sunt sibi ipse adstruat? Ego certe mihi cernere uideor et uultum et oculos et deformes utriusque blanditias et eorum qui aderant tacitam auersationem ac timidam uerecundiam.

Where Cicero’s text provides only the barest of outlines, giving details of posture, clothing and situation, Quintilian supplies for himself further details of appearance, a response which has been termed ‘slow realisation by way of synecdoche’. The mental image which he describes to us includes the detail of Verres’ eyes, and the figures are set in motion, caressing each other in an external display of desire. Quintilian the reader also attributes feelings to the internal audience that presumably mirror his own response to the scene playing in his mind. This passage sheds light on the extravagant claim that Quintilian made for another Ciceronian passage, the description of the aftermath of the party discussed above (8.3.66; see p. 91–3). When he asks ‘what more would anyone who had actually entered the room have seen?’ it may well be that he is not referring to the sparse details provided by Cicero’s text but to the quality of his own, far fuller visualization of the scene sparked by the words. In the case of the passage from the Verrines, where the larger context survives, it is Quintilian, Insitutio oratoria, 8.3. 64–5. My italics. This part of the Verrines was not delivered in court, a detail which Quintilian ignores. Beth Innocenti, ‘Towards a theory of vivid description as practised in Cicero’s Verrine Orations’, Rhetorica, 12 (1994): 355–81 argues that the text that we have of the actio secunda does represent Cicero’s oratorical practice. Robert Cockcroft, ‘Fine-tuning Quintilian’s doctrine of rhetorical emotion: seven types of enargeia’, in Tomàs Albaladejo et al. (eds), Quintiliano: Historia y actualidad de la retórica (Logroño, 1998), p. 504.

Phantasia

109

clear that Quintilian is taking cues from the attack on Verres in the rest of the work. Moreover, as we have already seen in the discussion of ekphrasis, the passages that were thought to have this immediate effect of ‘placing before the eyes’ did not necessarily evoke static scenes. Quintilian’s accounts of the murder and of the sacked city are full of details of human action; Cicero’s evocation of the room after a party begins with the comings and goings of the drunken guests, and even his tableau of Verres leaning on his mistress contains within it the aftermath of action (‘nixus’) which is sufficient for Quintilian to supply further movements and emotions to his own imaginative rendering of the scene. The question still remains of how an orator could control such creative visualization and interpretation of words on the part of his audience. Quintilian interacts with Cicero’s text in exactly the way that Iser envisages the reader interacting with a work of literature, filling in the gaps left by the text. But where Iser places emphasis on the variable and individual response of each reader, Quintilian’s account assumes a predictable response, or at least a restricted range of responses. For him, this is far from a merely theoretical issue: he is purporting to be able to teach his readers how to guide their audiences’ responses in practical situations, and implying that he himself has wielded this power in the courtroom. One �� indication as to how he could be so confident of a predictable response is provided by his insistence that the orator had to remain within certain bounds if he was to create the desired effect. A particularly significant passage in this respect is to be found in the conclusion to his discussion of enargeia in Book 8. Here, he advises the orator to ‘follow nature’ (‘naturam … sequeamur) (8.3.71). But the remark that immediately follows makes it clear that the idea of ‘nature’ is largely conventional, for it is vital that the audience should be able to recognize what they hear: ‘[people’s] minds’ he says ‘accept (‘recipiunt’) most easily things which they recognize’. This suggests that the easiest visualisations to communicate will therefore not be new creations (whatever that might mean), or representations of

See Lucia Calboli Montefusco, ‘Ἐvάργεια et ἐvέργεια: l’evidence d’une démonstration qui signifie les choses en acte’, in Mireille Armisen Marchetti (ed.), Demonstrare: voir et faire voir: forme de la démonstration à Rome (Toulouse, 2005), pp. 43–58. Wolfgang Iser, ‘The reading process: a phenomenological approach’, in Jane P. Tompkins (ed.), Reader-response Criticism: From Formalism to Post-structuralism (Baltimore, 1980), pp. 50–69.

Ibid., p. 55: ‘each individual reader will fill in the gaps in his own way, thereby excluding the various other possibilities; as he reads, he will make his own decisions as to how the gap is to be filled.’ Quintilian, Institutio oratoria, 8.3.71: ‘facillime enim recipiunt animi quod agnoscunt’.

110


idiosyncrasies, but will correspond to the audience’s expectations and prior experience. Quintilian’s account of his response to the portrait of Verres also provides vital clues to how an orator might be able predict his audience’s response. The values and emotions which Quintilian attributes to the scene (and which he seems to have shared himself) are carefully controlled by the use of significant details. Cicero contrasts Verres’ official status (‘praetor populi Romani’, praetor of the Roman people) with the details of his shoes, the purple cloak (‘pallium’) which, as a Greek style of dress coloured with oriental dye, would have had strong connotations of unRoman luxury, not to mention his unmanly stance, as he leans (‘nixus’) on his mistress. Such details would no doubt have been so loaded, so telling, for Roman readers like Quintilian that they might not themselves have been fully aware of the amount of decoding involved in their response. This example reveals the extent to which what must have been immediate, practically unconscious, associations for the original audience were in fact culturally specific and demanded a degree of what one might call ‘cultural competence’ from both speaker and audience if they were to be fully successful. Imagination, Memory and Knowledge Quintilian’s remark about familiar things being most apt to lodge themselves in the mind points to the close relationship between memory and imagination in ancient thought. Moreover, much knowledge was thought of as being stored in visual form so that the type of imagining called for in oratory was closely related to memory: the orator uses his own visual resources to call up images which already exist in the audience’s mind. The close connection between visualization and memory in ancient thought is further underlined by the importance of visual images in ancient theories and techniques of memorization. The artificial memory techniques discussed by the author of the Rhetorica ad Herennium, by Cicero and (with reservations) by Quintilian relied heavily on mental images.10 The images which the Auctor ad Herennium suggests the orator use in order to remind himself of the order and content of the

See Webb, ‘Imagination and the arousal of the emotions in Greco-Roman rhetoric’, in S. Braund and C. Gill (eds), The Passions in Roman Thought and Literature (Cambridge, 1997), pp. 122–3. 10 Rhetorica ad Herennium, 3.16.28–24.30; Cicero, De oratore, 2.354–60; Quintilian, Institutio oratoria, 11.2. Aristotle also makes a brief mention of the use of mental images in memory techniques in On the Soul, 427b 18–20. See, in general, Frances Yates, The Art of Memory (London, 1966) and Jocelyn Penny Small, Wax Tablets of the Mind: Cognitive Studies of Memory and Literacy in Classical Antiquity (London, 1997).

Phantasia

111

parts of his speech were intentionally bizarre so as to be as memorable as possible. These might function by suggesting words to the orator (the image of a testicle, ‘testiculus’, ‘testis’, could serve to remind the speaker to mention a witness, ‘testis’) as in a rebus puzzle.11 But other images bore a more direct, visual resemblance to the subject the orator was supposed to remember: the testicle-witnesses in this example were involved in a case of disputed inheritance in which a wealthy man is supposed to have been poisoned. He should also be pictured lying ill in bed in a direct illustration of the narrative. Interestingly, the orator’s own memory seems to play an important part here: if he did not know the man in question personally, he is advised not to invent an image of some unknown person, but to choose someone known to him so that the image will spring quickly to mind.12 The memory is therefore the readiest source of images for use in such mnemonic systems, and of the architectural backgrounds to place the images in. Memory and the Gallery of the Mind As this suggests, images derived from sense perception were also thought to be the basis of natural memory, and this model had a long and enduring history in Greek and Roman thought. These memory images resulted from impressions received on the mind or soul through the senses. Gorgias refers to the process in his Encomium of Helen (17) when he states that ‘sight engraves upon the mind images (eikonas) of actions (pragmata) which have been seen’. It was Aristotle who developed this idea in ways which were influential on later centuries. He explains that sense impressions (aisthēmata) gathered in daily life somehow impressed themselves on the soul to create memory-images (phantasmata).13 This basic model endured into the Imperial period and underlies much discussion of sight and memory, as well as the theory of enargeia and phantasia. Because it was so familiar it usually needed no explanation for ancient readers. There is one context, however, where the Aristotelian model becomes clear. This is the special case of the heightened perception of lovers: novels and other discussions of love from the Imperial period tell us time and again how 11

Rhetorica ad Herennium, 3.20.33. Ibid. The reader is also advised not to choose a person of low social standing (‘de minimo loco’) to figure in their tableau. The implication is that such a person would not be sufficiently individualized – quite literally ‘distinguished’ – for the orator to call him to mind quickly and easily. 12

13 Aristotle, On Memory and Recollection, 450a 30–32; see Richard Sorabji, Aristotle on Memory (London, 1972), p. 11. Aristotle’s writings on memory and phantasia raise immensely complex issues. In what follows I am concerned only to provide a bare outline of the general features that are relevant to rhetorical phantasia and enargeia.

112


the image of the beloved enters through the eyes and then remains in the lover’s soul.14 This idea runs throughout the ‘Ekphrasis of Beauty’ attributed to Libanios whose author claims that, once he had glimpsed the beautiful girl at her window, his soul became a painter so that the resulting ekphrasis is the verbal expression of that impact.15 Aristotle and others frequently appeal to analogies with the visual arts to express the nature of the impressions made by sense perception upon the soul and their lingering form as memories. Famously, Aristotle compares the impact of sense perception on the soul to the impression (tupos) made by a signet ring on wax.16 This idea of imprinting is evident in the Greek terms for visual language, diatupōsis and hupotupōsis, implying that such language has an effect analogous to that of direct sensation. In the same passage, memory is compared to a painting (hoion zōgraphēma), an analogy which Aristotle also uses in On the Soul (427b 21–4) to explain the activity of contemplating an internal image (phantasma).17 Like the image of the impression in wax, this idea of the mind as a gallery of paintings left behind by sensation has a long afterlife. Ps.-Plutarch extends the metaphor further when he compares the fleeting images produced by ordinary sensation with the enduring soul-painting impressed on the soul by the sight of the beloved, which is like a painting burned on with encaustic (Moralia, 759C). Aristotle himself uses the graphic analogy to illustrate the way in which we can think of these internal images either as what they represent or as images of what they represent (On Memory, 450b 20 – 451a 2). In this they are just like the images created in the mind by enargeia, which create an impression like that of sensation and can be contemplated either as equivalent to what they represent, or as likenesses. And, like the writers on enargeia, Aristotle also emphasizes the physicality 14

See, for example, Chariton, Chaireas and Kallirhoe, 6.5–7; ps. Plutarch, Erōtikos, 759C with further discussion below. Achilles Tatios, Leukippe and Kleitophon, 1.9 offers the further development that the flowing of the image of the beloved into the lover’s soul leads to a form of copulation. See also Aristotle, Rhetoric, 1370b19–22 on the special importance of memory to the lover and Apollonios of Rhodes, Argonautika, 3.453–6. Cicero, De oratore, 2.357 also refers to this theory in his discussion of memory. 15 Libanios, Progymnasmata, pp. 541–6. 16 Aristotle, On Memory and Recollection, 450a 25–32. See Sorabji, Aristotle on Memory, 3; Gerard Watson, Phantasia in Classical Thought (Galway, 1988), p. 25. Quintilian, 11.2.4 refers to the analogy between memory and the imprint of the ring on wax as something that ‘many people think’, a sign of the pervasiveness of the Aristotelian image. 17 Malcolm Schofield, ‘Aristotle on the imagination’, in Jonathan Barnes et al. (eds), Articles on Aristotle 4: Psychology and Aesthetics (London, 1979), p. 119 cautions, however, that Aristotelian phantasmata are not always to be understood as mental images outside On Memory and Recollection. Nevertheless, in On the Soul, 429a 3–5 Aristotle derives both terms from phaos [light], reflecting the role of sight as the primary sense.

Phantasia

113

of these images which are impressed on the body itself.18 There are thus many parallels between the effects of enargeia and the effects of direct perception which further help to explain the ease with which Quintilian can blur the distinction between the two. For what lies behind vivid speech is the gallery of mental images impressed by sensation in the speaker’s mind. The souls of both speaker and listener are stocked with internal images of absent things, and these provide the raw material with which each party can ‘paint’ the images that ekphrasis puts into words. This idea surfaces in the rhetorical handbooks in the shape of warnings about describing shameful or inappropriate subject matter. In his chapter on the koinos topos, for example, Nikolaos warns against including too much detail in ekphraseis of subjects such as adulterers or a seducer of young boys, ‘for in describing such things we will slander ourselves more than him [i.e. the adulterer]’.19 More striking still is the caution urged by Menander Rhetor when he discusses the ekphrasis of the bride and groom in his discussion of the wedding speech (epithalamion).20 Unless it is socially acceptable for the speaker to have seen her (if he is close relative, for example), a verbal description will lay him open to suspicions of harbouring an illicitly acquired memory image of her in his mind. So verbal representation is clearly thought not just to betray knowledge of words (of the lexicography of adultery or of beauty, as in Hamon’s definition of description), but to derive from an internal image of what is being described.21 This image resides deep within the speaker’s mind and may itself derive from perception. As in the case of the passages of Euripides and Virgil analysed by ps.-Longinos and Quintilian respectively, the verbal evocation allows a glimpse into the mind of the describer, which can be fraught with risk for an orator who is bound to his audience through social ties. Thinking and Speaking with Pictures Words, in Quintilian’s discussion of enargeia, serve simply to communicate the orator’s internal image, his phantasia, of a scene to the audience. His insistence on the visual source and effect of vivid language, and his relative 18

Aristotle, On Memory and Recollection, 453a 14–16. See Watson, Phantasia in Classical Thought, p. 31. 19 Nikolaos, Progymnasmata, 45.16: ἐκφράζovτες γὰρ πλέov ἡμᾶς αὐτoὺς ἢ ἐκεῖvov διαβαλoῦμεv. 20 Menander Rhetor, On Epideictic, Treatise II, 404. 11–14. 21 Philippe Hamon, Du descriptif (Paris, 1993), p. 42 stresses the way in which description appeals to the reader’s memory and knowledge, but the relevant competence is above all lexical (p. 43).

114


neglect of the linguistic details, reflect ideas about thought, memory and language in general which are to be found in Aristotle’s writings, but are also pervasive in the texts of later periods, in scholia, in novels and essays, as well as in the rhetorical theory of phantasia. For Aristotle, thought itself (noein) is inseparable from the mental images, the phantasmata, which stock our minds, and cannot take place without them.22 The resulting connection between phantasia and language was explored in far greater detail by the Stoics. It is worth exploring Stoic theory in a little more detail here because it helps to illustrate the profound connections between mental images and language in ancient thought. A source cited by Diogenes Laertios (7.49) shows that phantasia was thought to be at the root of language through the functioning of thought (dianoia): ‘for the impression (phantasia) arises first, and then thought (dianoia), which has the power of talking, expresses in language what it experiences by the agency of the impression’.23 Phantasia is the basis of language and, as in the rhetorical theory of enargeia, language serves as the medium by which phantasiai are communicated from the speaker’s mind to that of the listener.24 The rhetoricians’ enargeia is therefore far from being an anomalous form of language, rather it is a heightened example of the way that all verbal communication could be thought to work, transmitting an internal impression from one mind to the other. Phantasia and Paraphrasis in Theon That various different forms of words could serve to transmit what was essentially the same phantasia is suggested by a passage in Theon’s Progymnasmata where the author discusses the practice of paraphrasis. Rather than being a purely stylistic exercise, paraphrase in Theon’s conception is the reworking of the same thought in different ways. The same phantasia hitting the same mind, or different minds, will give rise to different utterances. The examples he cites include Demosthenes’ account of the destruction of Phokis (19.65), mentioned by Nikolaos in his discussion of ekphrasis, and Aeschines’ similar account of the destruction of Thebes (Against Ktesiphon, 157), which are both identified as paraphrases of the account of the sack of a city embedded within Phoenix’s speech

22

Aristotle, On Memory and Recollection, 449b 30 and On the Soul, 427b 14 –16 and 432a 7. See Sorabji, Aristotle on Memory, p. 6; Schofield, ‘Aristotle on the imagination’, p. 128; Mireille Armisen, ‘La notion d’imagination chez les anciens I: les philosophes’, Pallas 27 (1979): 29. 23 Translation from Anthony A. Long and David N. Sedley, The Hellenistic Philosophers (Cambridge, 1987), Chapter 33 D. See also Anthony A. Long, Hellenistic Philosophy: Stoics, Epicureans, Sceptics (Berkeley, 1986), p. 125. 24 See Catherine Atherton, The Stoics on Ambiguity (Cambridge, 1993), p. 43.

Phantasia

115

in Iliad, 9.593–4.25 All these passages, despite their apparent differences (Demosthenes describes the desolate condition of a specific city after the attack whereas Kleopatra, the speaker in Phoenix’s speech, describes details of the action, presented as a general warning of what happens on such occasions), are therefore thought of as verbalizations of the same underlying phantasia. Phantasia: Philosophy and Rhetoric The various theories of phantasia that were current in the philosophical schools of the Hellenistic and Roman periods have affinities with the basic concept of the mental image, transmitted by words, that we find in Quintilian and ps.-Longinos. In both the philosophical and rhetorical accounts of phantasia, a particular state of mind gives rise to words. The philosophers’ emphasis on the relationship between the phantasia in the speaker’s mind and his perceptions of the outside world helps to clarify the type of imagination that is in play in rhetorical visualization: it is an imagination based on previously acquired sense perceptions. There are, however, some important differences between the rhetoricians’ accounts of enargeia and the role of phantasia in Stoic linguistic theory, as is only to be expected given the different contexts in which they were developed. First of all, when Quintilian and ps.-Longinos speak of a phantasia, they clearly mean an image of something perceptible, and the corresponding phantasia that results in the audience’s mind is similarly akin to sensation. The Stoics’ phantasia, by contrast, could be a more abstract phenomenon than either Aristotle’s phantasmata or the rhetoricians’ phantasia in that it could be, as Anthony Long points out, a representation of sensory or non-sensory objects.26 As such, Stoic phantasia is often translated ‘impression’ or ‘presentation’.27 Theon’s conception of phantasia in his discussion of paraphrasis was similarly broad. While the example of the sacked city constitutes a tableau that is familiar from other rhetorical sources, the other examples he offers are of abstract ideas which had been expressed in different ways by different authors (or, in the case of the prolific Demosthenes, by the same author.) Thus he treats comments

25

Theon, Progymnasmata, 62, l. 32 – 63, l. 13. See Michel Patillon, La Théorie du discours chez Hermogène le Rhéteur: essai sur les structures linguistiques de la rhétorique ancienne (Paris, 1988), pp. 309–11 and his notes to this passage of Theon. 26 Anthony Long, Stoic Studies (Cambridge, 1996), p. 271. 27 See Jean-Baptiste Gourinat, Les Stoïciens et l’âme (Paris, 1996), pp. 36–62.

116


by Thucydides, Theopompos and Demosthenes on the nature of jealousy (phthonos) as different expressions of the same thought (phantasthen).28 The Stoics’ theory of phantasia, moreover, was central to their reflection on epistemology, which distinguished phantasiai deriving from the perception of some real object or from reasoning (termed katalēptikai) from ‘figments’ such as dreams or the illusions of madness, which can be termed phantasmata.29 The ideal of kataleptic phantasia reflected the Stoic belief that true knowledge of the world was possible and that the wise man was able to distinguish between accurate representations, which formed a correct basis for action, and those that did not derive from reality.30 Classic illustrations of the latter were dreams or the illusions of the deranged, such as Orestes’ hallucination of the Furies.31 This is interestingly handled in ps.-Longinos who, as we have seen, uses the common Stoic example of Orestes’ hallucinations as an example of phantasia in poetry, which is able to allow itself, as he explains a little further on, to include fantastic and impossible scenes.32 There is thus an interesting correlation between his poetic phantasia and the Stoics’ ‘empty attraction’, just as there is between his account of the ineluctable force of rhetorical enargeia, when it is mixed with arguments based on facts, and the impact of kataleptic phantasia when there is no impediment to its acceptance (as there was, for example, for Admetos who, on seeing Alcestis really returned from the dead, refused to accept that his phantasia represented reality).33 Unimpeded kataleptic phantasia, according to Sextus Empiricus, was said by later Stoics to ‘all but seize us by the hair and drag us to assent’.34 Significantly, this ‘unimpeded kataleptic phantasia’ is qualified by Sextus as enargēs, in the root sense of the 28

Theon, Progymnasmata, 63, ll. 13–22 citing Thucydides, 2.45; Theopompos, FgrH II, 115, 395 and Demosthenes, 18.135. 29 See, for example, the account of Diogenes Laertios, 7.49 (= Long and Sedley, The Hellenistic Philosophers, 39 A). The Stoic terminology therefore differs from the usage of Aristotle for whom phantasmata are any kind of internal image, with no implications about their truth status. On Stoic thought on phantasia and truth, see further Long and Sedley, The Hellenistic Philosophers, Chapter 40; Gourinat, Les Stoïciens et l’âme, pp. 40–42. 30 Long and Sedley, The Hellenistic Philosophers, vol. 1, p. 249. 31 Some authorities distinguished between two different types of illusion in that Orestes-like hallucinations were based on the mis-recognition of an object (he saw Elektra but thought he saw a Fury) while others were based on no perceptible object. See Sextus Empiricus, Against the Professors, 7.247–52 (= Long and Sedley, The Hellenistic Philosophers, 40 E). 32 Ps.-Longinos, On the Sublime, 15.8: τὰ μὲν παρὰ τοῖς ποιηταῖς μυθικωτέραν ἔχει τὴν ὑπερέκπτωσιν, ὡς ἔφην, καὶ πάντη τὸ πιστὸν ὑπεραίρουσαν. 33 Sextus Empiricus, Against the Professors, 7.253–60 (= Long and Sedley, The Hellenistic Philosophers, 40 K). 34 Sextus Empiricus, Against the Professors, 7.257 (= Long and Sedley, The Hellenistic Philosophers, 40 K). Juliette Dross�� , ‘De �� la philosophie à la rhétorique: la relation entre phantasia

Phantasia

117

term. The difference is that Longinos’ rhetorical phantasia is said to need the support of factual arguments to achieve its effect, while the power of kataleptic phantasia is such that it need only be free of impediments. In general, however, the rhetoricians use the term in a looser way than do the Stoics to mean any mental image (or, in the case of ps.-Longinos, the transmission of mental images through words) and imply no automatic judgement as to the truth of that image or its relation to reality.35 What matters primarily to the rhetoricians is that the speech or poem should have the desired impact on the listener; for this to be achieved what is important is conformity to culturally accepted ‘truth’ (i.e. probability, verisimilitude in the domain of rhetoric). There are therefore differences between rhetorical and Stoic phantasia, and ps.-Longinos says as much when he distinguishes the ‘usual’ meaning of ‘any thought productive of language’ (a phrase which is ‘straight out of the textbook’, as Simon Goldhill has pointed out) from the now fashionable sense in which he chooses to use it:36 The term phantasia is used generally (koinōs) for anything which in any way suggests a thought productive of speech; but the word has come into fashion for the situation in which inspired by enthusiasm and emotion you seem to see what you are talking about and place it before the eyes of the audience. καλεῖται μὲv γὰρ κoιvῶς φαvτασία πᾶv τὸ ὁπωσoῦv ἐvvόημα γεvvητικὸv λόγoυ παριστάμεvov. ἤδη δ'ἐπὶ τoύτωv κεκράτηκε τoὔvoμα ὅταv ἃ λέγεις ὑπ’ ἐvθoυσιασμoῦ καὶ πάθoυς βλέπειv δoκῇς καὶ ὑπ’ ὄψιv τιθῇς τoῖς ἀκoύoυσιv.37

et enargeia dans le traité Du sublime et l’Institution oratoire’, Philosophie antique, 4 (2004), p. 77 has made the same connection between the two passages. 35 The failure to distinguish rhetorical phantasia from the specialized Stoic sense of the word leads Elsner, Art and the Roman Viewer: The Transformation of Art from the Pagan World to Christianity (Cambridge, 1995), p. 27, for example, to claim that ‘ekphrasis … tells the truth’. As we will see, ekphrasis and enargeia are concerned with verisimilitude, not truth. 36 Simon Goldhill, ‘What is ekphrasis for?’, p. 6. See also Dross, ‘Phantasia et enargeia’, p. 68. 37 Ps.-Longinos, On the Sublime, 15.1. Lucian was also able to refer to the notion of phantasia as part of his satire of Stoics in Philosophies for Sale, 21, indicating that it was generally familiar to an educated audience. Dross, ‘Phantasia et enargeia’, p. 76 suggests that ps.-Longinos’ use of the term koinōs indicates that the Stoic idea may have become so widespread that many people were familiar with it without necessarily knowing its origins. Goldhill, ‘What is ekphrasis for?’, p. 7 interprets ps.-Longinos’ distinction as a sign of ‘an argument brewing over the fashionable sense of phantasia and the Stoic philosophical sense’. I would prefer to his remark as a sign of a desire for clarity in the face of the variety of usages current in his day.

118


Ps.-Longinos appears to be applying patterns of thought that appear in Stoicism to the criticism of poetry and rhetoric with the vital difference that there are no cognitive or moral implications.38 The reader who allows him or herself to ‘all but see’ the Furies as a result of the poet’s phantasia is in no way morally or cognitively deficient but is experiencing the impact of the sublime. In particular, though Longinos’ first example of phantasia recalls the Stoics’ notions of phantasma, it is used for different ends: ps.-Longinos is interested in the effect on the listener, rather than the ontological status of the subject of the vision.39 The contemporary declaimers who, ps.-Longinos claims, speak as if they could see the Furies are guilty of overstepping the boundaries that separate poetic from rhetorical phantasia and unwittingly imitating a madman.40 Their mistake is one of generic appropriateness rather than one of cognition. Most importantly, rhetorical phantasiai need only be ‘like truth’: their subjects may no more have existed than did the Furies seen by Orestes, but they derive their power from their verisimilitude and from the larger context of the speeches in which they are found. This is because the rhetoricians’ interests are ultimately practical: for both poets and orators, it is sufficient to create the desired impact upon the listener.41 Ps.-Longinos thus illustrates both the similarities and the differences between rhetorical and Stoic conceptions of phantasia.42 It might be best to consider the rhetorical and Stoic conceptions of phantasia as different, but related, specializations of the same basic model.43 What the two definitions cited by ps.-Longinos certainly do share in common is the idea that a mental impression of some kind is intimately bound up with the production of language. The role of phantasiai as the internal equivalents of words is expressed most clearly by a fourthcentury North African steeped in traditional rhetorical and philosophical 38

See Matthew Leigh, ‘Quintilian on the emotions’, JRS, 94 (2004): 122–40 on the gulf between Stoic ideals and Quintilian’s theory of the emotions. 39 Orestes’ visions of Furies, treated by ps.-Longinos as an example of poetic phantasia, are one of the stock examples of empty phantasmata cited by the Stoics. See, for example, Sextus Empiricus, Against the Mathematicians, 7.249 (= Long and Sedley, The Hellenistic Philosophers, 40 E). 40 Ps.-Longinos, On the Sublime, 15.8. 41 See below, Chapter 7, however, for examples of writers in other genres playing on the gap between the impressions created by enargeia and truth. 42 I do not therefore recognize the claim, attributed to me by Simon Goldhill, ‘What is ekphrasis for?’, p. 7 that ‘the fashionable sense of phantasia and the Stoic philosophical sense … are discrete usages’, if by ‘discrete’ we are to understand ‘unrelated’, ‘entirely different’. I do think that they are ‘distinct but related’, and that the exact nature of the relationship needs to be decided in each case. It is unfortunate that Goldhill does not provide a footnote to support his assertion. 43 See the remarks of Dross, ‘Phantasia et enargeia’, pp. 76–8.

Phantasia

119

doctrine, St Augustine. He explains that when he wishes to speak the name ‘Carthage’ he first finds within him an image (‘phantasia’) created by the senses.44 The pervasiveness of this basic idea, which we find from Aristotle to Augustine, helps to explain why Quintilian is able to slide from speaking of mental images to speaking of the language used to express them and back again, apparently without feeling any incongruity. Painting in the Mind Stoic kataleptic phantasiai, like Aristotle’s phantasmata, derive from sense perception and are thus a kind of memory.45 It is this close connection to memory that helps to explain both the nature of the mental images that the speaker draws upon and the predictability of the audience’s visual response. If we draw the analogy with rhetorical phantasia, the speaker’s visualization of the scene he wants to place before his audience’s eyes draws on elements already residing in his memory and, unless it is a scene he has witnessed himself, is a composite of existing images. The fact that memory images do not remain inert but are subject to manipulation means phantasiai or phantasmata are not to be understood as limited to the quasi-photographic reproduction of things seen.46 By various processes, images that derive from experience can form the raw material of new composites. Stoic linguistic theory certainly allowed for abstract thought to be derived from perception by a number of procedures including resemblance, analogy, synthesis and transposition.47 If we apply this idea to rhetorical phantasia we find that it is possible to visualize things that one has never seen by applying the same procedures to existing mental images. Mythical and fantastic beasts can be imagined through a process of synthesis, putting together man and horse, or in Lucian’s ironically named True History (1.13–6) the fantastic composite warriors made out of a 44

Augustine, De Trinitate, 8.6.9.68–70: ‘et Carthaginem quidem cum eloqui uolo apud me ipsum quaero ut eloquar, et apud me ipsum inuenio phantasiam Carthaginis.’ See Gerard O’Daly, Augustine’s Philosophy of Mind (London, 1987), p. 113. 45 See Long and Sedley, The Hellenistic Philosophers, 39 D–F and Watson, Phantasia in Classical Thought, p. 45. 46 Daniel Babut, ‘Sur la notion d’“imitation” dans les doctrines esthétiques de la Grèce classique’, REG, 98 (1985): 72–92 makes an analogous argument that the Platonic and Aristotelian notion of mimesis is less reproductive and more creative than is often supposed. He argues that Philostratos’ definition (Life of Apollonius, 6.19) of phantasia as able to create what it has never seen (in contrast to mimesis which is simply reproductive) is therefore not as revolutionary an idea as is often claimed. See also Manieri, L’immagine poetica, pp. 50–51 who traces the gradual evolution of the term. 47 Diogenes Laertios, 7.52–3. See Watson, Phantasia in Classical Thought, p. 47; Manieri, L’immagine poetica, pp. 47–9.


120

combination of animal and vegetable elements. The whole system of visual mnemonics discussed by the Auctor ad Herennium and Quintilian relies on such mental synthesis, placing familiar objects in striking combinations, or against equally familiar architectural backgrounds. Thus, Quintilian’s phantasia of the murder need not itself derive from direct perception but can be synthesized from a variety of knowledge derived from the senses as well as from cultural expectations. Comparison with Theon’s murder scene in his discussion of common place (see above) in fact shows how conventional Quintilian’s imagination was. For those with access to them, the visual arts were also a possible source of mental images. Whatever the source, the raw visual materials were already there, residing as phantasiai in the speaker’s mind. Again, the figure of the lover is of particular interest as it embodies a heightened version of perception in which the intimate connection between sense perception, memory, imagination and speech is clearly visible. Despite the analogy with painting, the lovers’ memory-images in the ps.-Plutarchan Erōtikos (759C) do not remain as inert reproductions of sensation but are starting points for imaginative elaboration. They move and speak in the lover’s mind as he imagines himself conversing with and embracing the person. The active role of the rememberer is also clear in the account of the love-struck king in Chariton’s Chaireas and Kallirhoe. His mental image of the heroine is more vivid than his actual surroundings: ‘He saw only Kallirhoe, though she was not there and heard her, though she did not speak’ (6.5). Prompted by Eros, he likens her to Homer’s Nausikaa, (Odyssey, 6.102–4) and ‘painting (anazōgraphōn) and modelling (anaplattōn) this sight, he burned with passion’ (6.7). The image of modelling is particularly apt in that it conveys the way in which a given body of material can be reshaped, like clay, creating a new image out of existing material, just as a plausible fiction (plasma) of the type that an orator was often called upon to present is worked out of existing material.48 If we take these examples as heightened examples of all imagining, they show how imagination is thought to work by a process of recombination, rather than creation ex nihilo. The imagination involved in the production of ekphrasis and enargeia is therefore conceived as neither entirely free and creative, nor as simply reproductive of sensation, but lies between these two poles. A further stage in the process – the verbalization of these combined images – is added by a scholiast’s reading of Phaedra’s delirious expression of her desire to go outside to the mountains and the hunting grounds of her beloved in Euripides’ Hippolytos. The scholiast interprets her cry ‘take me to the mountain’ and her evocation of what she would see and do there as revealing her imagination (literally, ‘remembering’, 48

See Gioia Rispoli, Lo spazio del verisimile: il racconto, la storia e il mito (Naples, 1988).

Phantasia

121

hupomnēsis) of what she desires. Her words are a verbalization of her erotically inspired visualizations and reveal the contents of her mind. By attributing to Phaedra a detailed ekphrasis of her desires, the scholiast says, Euripides has ‘superbly imitated the character of the lover. For by remembering the things they desire and all but painting (zōgraphountes) these things in their words lovers stir their desire to greater intensity’, just as Chariton’s king does.49 The scholiast understands Phaedra’s ekphrasis of what she imagines herself doing on the mountains to betray her most intimate phantasies.50 This casual reference reveals the role of ekphrasis as the verbalization of an internal image created from visual memories, which itself has an emotive power over its creator. It also reveals the potentially dangerous role of ekphrasis as speech that may give access to the visual contents of the speaker’s mind. Reception and Imagination This detour through the heightened memory-images of ancient lovers reveals the connection between perception, memory and language which underlies the rhetorical theories of enargeia. Language derives ultimately from mental images. The speaker, as we saw in Quintilian’s account of the murder in Book 6 of the Institutio oratoria, makes use of knowledge (visually stored) of what ‘usually happens’ derived from experience of analogous events and from shared cultural knowledge. This visualization is expressed in words. At the receiving end the audience goes through a similar process. They, too, have souls stocked with images, derived from sense perception, or from shared cultural knowledge. The speaker’s words act as triggers for the retrieval of stored images which are recombined as necessary, with the addition of related features, as Quintilian’s account of his response to the description of Verres shows. So far, I have pieced this process together from a wide variety of sources. But Augustine gives a uniquely clear account of how the listener calls upon existing visual traces in response to verbal descriptions.�� He explains that when he hears a description of Alexandria, a city he has never seen 49

Scholia in Euripidem, vol. 2, 32.15 cited by Roos Meijering, Literary and Rhetorical Theories in Greek Scholia (Groningen, 1987), p. 51: ἄκρως δὲ ἐρωτικὸv ἦθoς ἀπεμάξατo τῇ λεπτoμερεῖ τῆς ἐκφράσεως περιεργίᾳ. εἰς ὑπόμvησιv γὰρ ἐρχόμεvoι (oἱ ἐρῶvτες) τῶv ἐπιθυμoυμέvωv καὶ μovovoυχὶ ζωγραφoῦvτες αὐτὰ τoῖς λόγoις ἔτι μᾶλλov τὴv ἐπιθυμίαv ἐξεγείρoυσιv. The scholiast elides the transition from image to its verbal expression here. It is surely the mental images, not their verbal expression, which stir up Phaedra’s desire, but it is only through the ‘word-painting’ that we have access to the imaginings. 50

In the same way, the ps.-Libanian ‘Ekphrasis of Beauty’ presents itself as a verbal expression of the image painted by sight on the speaker’s soul. See Libanios, Progymnasmata, pp. 541–6, esp. p. 545, l. 22 – p. 546, l. 5.

122


for himself, he imagines it as best he can by drawing on his knowledge of the closest sight within his experience, the city of Carthage.51 He adopts from the Stoics the distinction between the image which results from direct experience (and therefore corresponds to reality), which he terms phantasia, and the phantasma, which does not. He knows that his phantasma of Alexandria, cobbled together as it is from features of another city, does not correspond to actuality and claims that he would be astonished if someone were to tell him it was indeed accurate. But it represents the best that the listener can do to create a mental image of something he has not seen, as orators’ audiences were so often required to do. While, from the philosophers’ point of view, such phantasmata were to be distinguished from truth, they were sufficient for the orator’s purpose of creating an immediate impact. Augustine’s remarks are vitally important because they confirm the close association between memory and imagination in the audience’s response to description. This suggests that there is more than just a logical concern for credibility behind Quintilian’s concern that the speaker’s visualizations be ‘like truth’. For the more a scene corresponds to the empirical, or culturally acquired, knowledge stored in the audience’s minds, the easier it will be for them to supply the images suggested by the orator’s words, and the easier it will be for the orator to predict the audience’s response. In order to be effective then, enargeia, and thus ekphrasis itself, must therefore be a re-presentation of familiar and accepted material – it is this very familiarity which gives the speech its evocative and emotive power. If we believe the rhetoricians, the impact of enargeia is immediate, leaving any intellectual judgement of the credibility of the images to a later moment.52 The orator does not have the time to wait for a reader’s more leisurely process of deciphering. To be effective his scenes must be instantly recognizable, common place.53 As Tompkins pointed out, the type of reader response assumed by ancient critics is an immediate reaction to the impact of words at the level of the imagination.54 Although it would be wrong to assume that no process of decoding, of interpreting 51

Augustine, De Trinitate, 8.6.68–89. The analogy would be the difference between feeling an emotion and assenting to it in Stoic theory of the emotions. 53 In a very different discipline, Synesios, On Dreams, 18–19 noted what a hard task it was to describe dreams, with their strange juxtapositions, to another person. He therefore recommended it as an exercise demanding the highest skill in rhetoric. 54 Jane P. Tompkins, ‘The reader in history: the changing shape of literary response’, in Jane P. Tompkins (ed.), Reader-response Criticism: From Formalism to Post-structuralism (Baltimore, 1980). 52

Phantasia

123

the signs constituted by the ekphrasis, is involved in this process, this type of response is very different from the careful, intellectual decoding demanded by, for example, descriptions of allegorical images.55 Enargeia and Argument The philosophical discussions of phantasia and the vital role it plays in cognition and the storage of knowledge point to a further aspect of rhetorical visualization. Kathy Eden has pointed out that the imagination is ‘a faculty which shares certain properties with both perception and thinking’ with the result that the image ‘represents the conjunction of sensation and intellection’.56 The type of knowledge appealed to by imagery is not therefore wholly intellectual (a store of statements of ‘what usually happens’ against which a particular image can be judged). But neither are the images themselves devoid of propositional content. Cicero’s portrait of Verres implies a series of statements about the character and actions of the accused, and these statements in turn form part of an implicit argument about Verres, based on common assumptions about character and probability: a man who acts in this manner is morally deficient. Verres acted in this manner (according to the image seared into the listener’s mind by Cicero’s enargeia), therefore Verres is morally deficient and, furthermore, likely to be guilty. There is therefore an affinity between the image that is ‘like truth’ and the arguments based on probability that the orator may use elsewhere in his speech. A particularly interesting example of the way in which vivid illustration can complement argumentation is to be found in Lysias 1, On the Murder of Eratosthenes, analyzed by Anne-Marie Chanet in a brief but highly illuminating article. Here, the explicit arguments from probability put forward by the speaker, Euphiletos, in support of his claim that he killed his wife’s lover on the spur of the moment without premeditation are backed up by the vivid and probable narration that precedes them. In his compelling evocation of the events that preceded his discovery of the adultery, Euphiletos emerges as a slightly simple and naive, but fundamentally honest, man who has often in the past been the dupe of his cunning wife. Here plausible, detailed narration and argumentation complement each other. What is more, Euphiletos’ self-presentation itself 55

Cf. Roger P. Hinks, Myth and Allegory in Ancient Art (London, 1939), p. 12. Some ancient allegories do make use of enargeia in the presentation of an image. But allegory is not a necessary characteristic of ekphrasis and this overview of the workings of enargeia that I have sketched here suggests that allegorical decoding demands a very different type of reception than that assumed for ekphrasis and enargeia by our rhetorical sources. 56 Eden, Poetic and Legal Fiction, p. 92.

124


contains an implicit (and therefore all the more effective) argument: ‘If I acted in this way I must be a simple, honest man; I cannot therefore be guilty of organizing a plot to kill Eratosthenes.’57 It is tempting to assume that it is just this sort of strategy that Dionysios of Halikarnassos had in mind when he noted the power of Lysias’ narrations to ‘help induce assent without us being aware of it’.58 Verisimilitude and Cultural Competence These mental representations, furthermore, were far from neutral reflections of reality. Among other things they would have been attached to a web of associations for the audience. This is true of physical images such as photographs or paintings, which have a connotative function beyond their representational function, as Barthes showed in his analysis of an apparently simple image used to advertise food in the early 1960s.59 It is still more true in the case of immaterial mental images, which are so closely associated with words and ideas. Some of these associations appear natural or universal, such as the pity or indignation that are assumed in ancient rhetorical sources as standard responses to the description of a sacked city.60 Again, Philippe Hamon’s analysis of description provides a close parallel that differs in one important respect from the ancient system. For, although Hamon emphasizes the importance of the reader’s competence and cultural knowledge to the reception of description, this knowledge is largely derived from texts and encoded in language.61 The contrast with the vital role of the visual in ancient theories of knowledge, memory and verbal representation is important in that it underlines once more the different orientation between the modern theoretician and the ancient rhetorician. In practice, however, modern describers may well make as much use of knowledge that is stored in visual form (whether in personal mental images or in the imagery of popular culture) as Quintilian’s readers. One modern example of a false story that initially proved convincing is the case of Susan Smith, the South Carolina woman whose false account of how her two children were abducted in a carjacking sparked a huge manhunt until 57 Anne-Marie Chanet, ‘Lysias, Discours I (Sur le meurtre d’Eratosthène)’, Lalies, 8 (1990): 100–104. 58 Dionysios of Halikarnassos, Lysias 18: αἱ διηγήσεις … τὴv πίστιv … λεληθότως συvεπιφέρoυσιv. 59 Roland Barthes, ‘Rhétorique de l’image’, Communications, 4 (1964): 40–51. 60 Charity appeals for aid after earthquakes, for example, still rely upon such responses to photographic images of destruction and misery. 61 Hamon, Du descriptif, pp. 48–51.

Phantasia

125

she confessed to their murder. Her description of the alleged attacker, a young black man, was convincing precisely because it corresponded in its details to her audience’s knowledge (residents claimed that she modelled her description on a known individual) and because it appealed to deeply rooted stereotypes.62 Her truth-like story was remarkably effective in inducing belief and sympathy in her audience, until the evidence finally told against it. She seems to have known instinctively what Quintilian tells his readers: that an audience will accept most easily stories and images that fit their preconceptions. The example of Susan Smith’s convincing lie also shows what sort of truth is resembled in effective enargeia. She needed to produce a picture of an attacker that not only corresponded in its details with familiar individuals or types, but one which sparked a range of useful associations for its audience, in this case the association of young black males with violent crime. The production of enargeia involved a competence which was more than simply lexical; rather it was a cultural competence, a familiarity with the key values of a culture and the images attached to them. At the time, one journalist, in a perceptive analysis, noted that Susan Smith ‘didn’t have to use much imagination. She just had to reach for the available nightmares.’ If her story convinced her audience at first, it was precisely because she appealed to stereotypes shared by the dominant culture.63 The audience’s own cultural competence was, and still is, a crucial factor in the reception of enargeia and means that we, as modern readers with our own array of potent images, will not always find ancient examples as vivid and compelling as the original audience might have done, possessing as we do a different visual vocabulary with �� different associations. Failed Ekphrasis and Cultural Dissonance By contrast, the failed ekphrasis to which Libanios refers in his Autobiography (Oratio, 1.41) is an example from antiquity of what can occur when speaker and audience do not share the same values and assumptions. Libanios recounts an incident when a rival of his, a certain Bemarchios, gave a repeat performance of a speech he had composed some time previously for the emperor Constantius for which he had been richly rewarded. Despite the fact that Bemarchios, like Libanios, was pagan, the speech celebrated a new church in Antioch. It contained, naturally enough for this type of speech, an ekphrasis of the building which Libanios claims 62

See the contemporary news reports at www.teleplex.net/shj/Smith/ninedays/ ninedays.html. This case also inspired Richard Price’s novel Freedomland, published in 1998. 63 �� The quotation is from Richard Lacayo writing in Time magazine, 14 November 1994 (available online at www.time.com/time/magazine/article/0,9171,981806,00.html).

126


not to have understood at all.�� According to his account, the members of the audience were left glancing at each other in utter confusion as Bemarchios ‘rambled on about pillars, trellised courts, and intercrossing paths which came out heaven knows where’.64 It is surely significant that this incident occurred at a time of intense contestation of images and values, as Christian visual and architectural forms replaced those of the pagan past to which Libanios remained faithful. While Libanios may be trying on the surface to question his rival’s rhetorical skill, the failure of this ekphrasis reveals what can happen when speaker and audience do not share the same images and values. For the most part, however, authors of rhetorical treatises could assume that an audience, whether in the Roman courtroom or in the Greek theatres in which the sophists performed their declamations, or the elite beneficiaries of epideictic discourses, would share a certain, predictable repertoire. In fact, one of the most important functions of rhetorical education itself was to ensure that this was so. Ekphrasis and Common Place Quintilian’s discussion of enargeia also helps to shed light on a curious feature of his contemporary Theon’s discussion of ekphrasis. Quintilian made clear that enargeia worked most effectively if the details of the account conformed to the commonly accepted views of the audience.�� In this respect, r�� h�� etorical enargeia has much in common with koinos topos (‘common place’) which occupies the same gap between truth and fiction as does enargeia in Quintilian’s account: the statements may not be true in every detail, but they represent a set of values and associations that would conform to dominant ideology and would thus be familiar to the audience. One way in which ekphrasis and koinos topos were connected was, as was mentioned above (Chapter 2), through the use of vivid evocations – variously referred to as ekphraseis, hupotupōseis or diatupōseis – within the koinos topos. But this relation of part to whole was not the only connection between the two. Theon refers to the similarities between ekphrasis and koinos topos in a puzzling passage of his chapter on ekphrasis, puzzling that is, until one takes into account the similarities in approach outlined above. He states: ‘this exercise has a certain affinity with the previous one [koinos topos]; they are alike in that they are both concerned with subjects which

64

Libanios, Autobiography, ed. and trans. A.F. Norman (Oxford, 1965), 41: διεξιόvτoς αὐτoῦ κίovας δή τιvας καὶ κιγκλίδας ὁδoύς τε ὑπ’ ἀλλήλωv τεμvoμέvας ἐμπιπτoύσας oὐκ oἶδ’ ὅπoι …

Phantasia

127

are not defined but are common and general’.65 As Patillon suggests, this points to the general nature of ekphrasis, providing a further hint as to the nature of ekphrasis: that it is not a description of specific scenes in all their individual particularities, but more a general characterization.66 And ekphrasis did not necessarily refer to a specific reality, but could draw on generally accepted ideas. A remark by the Byzantine commentator Doxapatres shows that some authorities interpreted Aphthonios’ advice to ‘completely imitate (mimeisthai) the events described’ to mean that one should not confine oneself to known facts, but should supplement them by the addition of details that are acceptable or possible (endechomena), and which therefore accord with audience expectations whether of reality or of a particular genre.67 Enargeia and Mimesis The interweaving of memory and mental images that lies behind enargeia points to a further characteristic of the latter. For what enargeia, and thus ekphrasis, seek to imitate is not so much an object, or scene, or person in itself, but the effect of seeing that thing, as Elaine Scarry says of the modern reader: ‘imagining is an act of perceptual mimesis’.68 By activating the images already stored in the listener’s mind, the speaker creates a

65

Theon, Progymnasmata, 68, 6–8: συγγέvειαv δέ τιvα ἔχει τὸ γύμvασμα τoῦτo τῷ πρoειρημεvῳ. ᾗ μὲv γὰρ περὶ oὐδεvὸς ὡρισμέvoυ ἐστὶv ἀμφότερα, ἀλλὰ κoιvὰ καὶ καθόλoυ, ταύτῃ ὅμoια. 66 See Patillon’s introduction to Theon, Progymnasmata, pp. xlii–xliv. 67 Doxapatres, Homiliai, p. 524, l. 30 – 525, l. 8: τoῦτo oἱ μὲv oὕτως ἡρμήvευσαv, δέov ἐv ταῖς ἐκφράσεσι μὴ μόvov τὰ γεγovότα, ἀλλὰ καὶ τὰ μὴ γεvόμεvα μὲv, ἐvδεχόμεvα δὲ γεvέσθαι λέγειv ὅτι γεγόvασιv. oἷov εἰ πόλεμov ἐκφράζoμεv, μὴ μόvov λέγειv ὅπως συvέβαλov αἱ φάλαγγες, ὑπερεῖχε δὲ τὸ μέρoς ἐκεῖvo καὶ τὰ τoιαῦτα, ἀλλ’ ὅτι καὶ πρὸ τῆς συμβoλῆς συvvεφία καὶ σκότoς καὶ βρovτῶv ἦχoι βίαιoι, καὶ ὅσα τoιαῦτα· ταῦτα γὰρ, φησὶv, εἰ καὶ μὴ ἐγέvovτo, ὅμως ἐπεὶ ἐvδεχόμεvov ἦv γεvέσθαι, λέγειv ἔξεστιv ὅτι ἐγέvovτo [Some people have interpreted this to mean that is necessary in ekphraseis to relate not only the things which happened but also things which did not happen but are accepted as happening, so that in describing a battle we do not just say how the troops met and how one side overcame the other and so on, but that before the engagement there were clouds and darkness fell and loud peals of thunder rang out and things of this sort, for they say that even if these things did not happen it is still permissible to say they happened because they are accepted as happening]. 68 Elaine Scarry, Dreaming by the Book (New York, 1999), p. 6: ‘We habitually say of images in novels that they “represent” or “are mimetic of” the real world. But the mimesis is perhaps less in them than in our seeing of them. In imagining [Emily Brontë’s] Catherine’s face, we perform a mimesis of actually seeing a face; in imagining the sweep of the wind across the moors, we form a mimesis of actually hearing the wind. Imagining is an act of perceptual mimesis.’

128


feeling like that of direct perception, a simulacrum of perception itself.69 It is the act of seeing that is imitated, not the object itself, by the creation of a phantasia that is like the result of direct perception (cf. Aristotle, On the Soul, 432a 9–10). In communicating through words his own mental image of a murder or a sacked city, Quintilian is not primarily attempting to convey information about a specific reality, but rather to prompt his audience to re-enact internally the act of seeing such a sight, and therefore to achieve an approximation of what an actual witness might have felt. The ancient theory of enargeia thus sidesteps the problem of how to represent the visual through the non-visual medium of language because of the connection that is assumed between words and mental images. Words do not directly represent their subjects, but are attached to a mental representation of that subject. Consequently, the theory of enargeia also provides one way of avoiding the problems involved in representing an external reality through the medium of language: it is not reality itself, but the impact of the perception of reality that is represented. And, although language is in itself a non-iconic medium, its production and reception both activate images stored in the speaker or listener’s mind. What is translated into words is not an object, residing in the material world, but a mental representation of that object, of the type that was thought in antiquity to lie behind all speech. Conclusions: Enargeia and Ekphrasis Enargeia, the quality which makes an ekphrasis an ekphrasis, therefore belongs to a conception of language as a quasi-physical force which penetrates into the mind of the listener, stirring up the images that are stored there. The�� pronouncements �� of the ancient theorists make full sense only when considered against the background of contemporary ideas about the soul and the importance of images to both thought and language. In this context it becomes clear that, rather than attempting the paradoxical task of representing the visual through the non-iconic medium of language, enargeia acted upon the mind to recreate the effect of perception, activating traces of perception which were stored there. So the theory of enargeia which underlies the treatment of ekphrasis in the 69 Watson, Phantasia in Classical Thought, p. 25 notes that phantasia is like perception. Schofield, ‘Aristotle on the imagination’, p. 106 notes that Aristotle’s use of the term phantasia encompasses ‘experiences so diverse as dreams and the interpreting of indistinct or puzzling sense data, which may be held to resemble the paradigm of successful sense perception in one way or another, yet patently lack one or more of its central features, and so give rise to the sceptical, cautious or non-committal phainetai’ (my italics). See also Ruth Webb, ‘Mémoire et imagination: les limites de l’enargeia dans la théorie rhétorique grecque’, in Carlos Lévy and Laurent Pernot (eds), Dire l’evidence: philosophie et rhétorique antiques (Paris, 1997).

Phantasia

129

Progymnasmata bypasses the dilemmas posed by ‘realistic’ description, by insisting on the mind and its images as the point of mediation between material reality and language. This enargeia had its place in all kinds of writing, and most probably in everyday speech�� , an area upon which our rhetorical sources rarely touch. But in rhetoric it had the special functions of making the listeners into witnesses of the speaker’s story, of engaging them emotionally in his account and effecting a change in their mental disposition. To be effective, such appeals to the visual and to the emotions had to correspond to what an audience was prepared to accept as probable, or ‘like truth’, and to be backed up by corresponding arguments. When these conditions are met, the effect of enargeia is described by ps.-Longinos as contributing to the ‘enslavement’ of the listener – one more reminder of the conception of language as force that underlies the theory. All this helps to explain why an exercise like ekphrasis was included by some, if not all, teachers in a set of exercises designed to prepare students for the study of rhetoric, and why ekphrasis was an integral part of the training they received in judicial and deliberative oratory (through the practice of declamation) and in epideictic oratory. The �� discussions of both ekphrasis and enargeia reveal a conception of language as a dynamic force that effects a change in the listener or reader’s mental state and this is a phenomenon that is fundamentally rhetorical, that is if one understands the goal of rhetoric as to bring about such a mental change.70 To emphasize the rhetorical nature of ekphrasis is also to draw attention to the vestigial orality of the phenomenon, the way in which the discussions of both ekphrasis and enargeia assume a live interaction between speaker and audience, with language passing like an electrical charge between them. This does not mean, in the ancient context, that ekphrasis and enargeia are not equally at home in other types of discourse. As was emphasized in Chapter 1, ancient readers saw enargeia as a positive quality in history writing and poetry, as well as rhetoric. These genres had much that was ‘rhetorical’ about them in the broadest sense that they were often intended for live performance and made use of many of the same resources as rhetoric.71 Plutarch’s description of Thucydides’ writing as transporting the reader/listener to the events of the Peloponnesian War is one example (see Chapter 1 above) and the many comments on the sense of presence 70

9–10.

See Chaïm Perelman, L’Empire rhétorique: rhétorique et argumentation (Paris, 1977), pp.

71 See, for example, Ruth Webb, ‘Rhetoric and poetry’, in Stanley Porter (ed.), Handbook of Classical Rhetoric in the Hellenistic Period (330 B.C.–A.D. 400) (Leiden, 1997); William H. Race, ‘Rhetoric and lyric poetry’, in Ian Worthington (ed.), A Companion to Greek Rhetoric (Oxford, 2007), pp. 509–25.

130


created by Homer’s poems are another. Only a very narrow definition of rhetoric as pure argumentation (which never existed in practice in antiquity) would distinguish it entirely from poetry. Instead, any speaker always needed to depict character, situations, events in such a way that they seemed present to their audiences, who would be drawn to assent. Conversely, ancient poetry was as much directed towards the audience and their response as was oratory. So, although I am focusing here and in the following chapter on ekphrasis and enargeia as strictly rhetorical techniques, exploring their use and definition in the rhetorical handbooks that are our best source of information on the reception of texts of all types, the observations about the relationship between text and audience may well also apply to other types of text, a question that I do not intend to address here. It is equally the case that the examples of rhetorical composition that are analysed in the next chapters are not strictly rhetorical, if one understands the term in the strictest sense of a live speech whose aim is to persuade an audience of something and to provoke them to make a decision with direct impact on the real world. The declamations I will be discussing are, of course, simulacra of such speeches, and all the speeches we are now able to analyse had a textual afterlife after their original performance, meaning that, amongst other things, we can never be sure to what extent the written version reflects that performance. More importantly, the chemistry of the interaction between the words, the occasion and the audience would have been different on subsequent readings (even on subsequent performances by the author as suggested by Libanios’ anecdote about the unfortunate Bemarchios). The next chapter will therefore read the sources as relatively unproblematic evidence for rhetorical technique while Chapter 7 will explore some of the paradoxes, particularly the depiction of time and space in the speeches in relation to their various audiences.

6. Ekphrasis and the Art of Persuasion

Quintilian’s discussions of enargeia and phantasia explain the functions of vivid language in a rhetorical context as well as providing insights into the psychological processes involved in both the production and reception of such language. The rhetorical functions of enargeia – to make the audience feel involved in the events in question – point to the reasons why the art of ‘placing before the eyes’ was considered a useful part of an orator’s preliminary training. Quintilian is a valuable source because he is infinitely more forthcoming than the Greek sources. Of these, the Progymnasmata are somewhat laconic, as we have seen, while the more advanced treatises tend to be highly technical and do not provide the type of explanation or personal insight we find in Quintilian. As we have seen, there are significant parallels between Quintilian’s treatment of enargeia and the discussions of ekphrasis in the Progymnasmata, and the same is true of the treatment of ekphrasis and its synonyms in the treatises that deal with the more advanced stages of the curriculum. In this chapter I will explore the uses of ekphrasis as a rhetorical technique as it appears in the more advanced Greek treatises on rhetoric that represent the next stage in rhetorical training after the introductory exercises of the Progymnasmata. On declamation, the art of composing fictitious judicial and deliberative speeches, the most useful sources for our purposes are Sopatros the Rhetor’s treatise On the Division of Questions and the commentaries on Hermogenes’ work On Issues (Peri Staseōn). Both of these contain detailed recommendations on how to construct a speech and tell us how, where, when and why ekphrasis should be used. In the domain of epideictic, the second treatise attributed to Menander Rhetor contains several examples of the use of ekphrasis and its synonyms. All of these sources provide an insight into the rich possibilities of ekphrasis in full-scale speeches. These functions were hinted at by Nikolaos in his Progymnasmata when he mentioned the use of ekphrasis in deliberative speeches to bolster the speaker’s argument for or against a course of action and its contribution to amplification (auxēsis) and the arousal of indignation (deinōsis) in judicial speeches, a way of increasing the impact of the subject described. In epideictic, he claims, it could produce a sense of pleasure, hēdonē, in the listeners. These uses of ekphrasis all correspond to Quintilian’s discussion of enargeia, but an exploration of the Greek sources on declamation and epideictic reveals a more complex set of practices. In this chapter I will

Nikolaos, Progymnasmata, p. 70, ll. 7–15.

132


focus on the uses of ekphrasis to help achieve persuasion, the passages in question illustrate ps.-Longinos’ remarks about the power of enargeia when it is combined with arguments. Declamation Studied in depth, declamation, the art of constructing fictional speeches based on fictional or vaguely historical cases, taught a wide range of skills, as is clear from the Greek handbooks on the art. Through the use of stasis (or ‘issue’) theory it taught analytical skills, as the student was required to decide exactly what was at issue in a particular case. This system for analysing questions was developed by Hermagoras of Temnos in the second century BCE and was further developed by Hermogenes in his work On Issues (Peri staseōn), which became the standard text in later antiquity and Byzantium. Stasis-theory provided a system for classifying the type of issue involved in a particular case. As Hermogenes explains, the first question to ask is whether the act is clear or unclear, as in the example of a man found burying a murder victim, for though it is clear that a murder has taken place, it is unclear who committed the crime. In this case the issue is one of conjecture (stochasmos). If, however, the facts are not in dispute the student has further decisions to make: the issue may be one of definition (horos) (is a man who has stolen private property from a temple guilty of robbery or the far more serious offence of temple robbery?) or quality (poiotēs), in which the justice, legality or advantageousness of the act is at question. Hermogenes then sets out a choice of strategies for making the case and establishes a set of headings under which the argument could be developed. Producing a full-length declamation also required a thorough training in argumentation and in exposition, subjects that are treated by the ps.-Hermogenean treatise On Invention and in Hermogenes’ On Types of Style. It was thus a multifaceted training that combined analysis and argumentation with skills that we would consider to be ‘literary’ such as the mastery of style, characterization, narration and, of course, the vivid See Malcolm Heath, Hermogenes on Issues: Strategies of Argument in Later Greek Rhetoric (Oxford, 1995); George A. Kennedy, Greek Rhetoric under Christian Emperors (Princeton, 1983), pp. 73–86; Donald Russell, Greek Declamation (Cambridge, 1983), Chapter 3. A reference to staseis has been noted in Lysias 12.34 which would put the origins of the system as early as the late fifth century. See Michael J. Edwards and Stephen Usher, Greek Orators I: Antiphon and Lysias (Warminster, 1985), ad loc.

Hermogenes, On Issues, 36, 10ff. Hermogenes, On Issues, 37. Both Kennedy, Greek Rhetoric under Christian Emperors, p. 83 and Bernard Schouler, ‘Personnes, faits et états de la cause dans le système d’Hermogénès’, Lalies, 8 (1990): 111–27 set out Hermogenes’ scheme in the form of flow charts.

Ekphrasis and the Art of Persuasion

133

use of language. The rigour of this training and the energy expended on refining it are both evident from Hermogenes’ handbook and the commentaries by Syrianos (fifth century) and others, and from Sopatros’ manual Diaireseis zētēmatōn (On the Division of Questions) (fourth century). These Greek texts repay closer examination as they reveal the mechanics of declamation, and sometimes the role of ekphrasis in persuasion, in a way which the Elder Seneca’s more familiar catalogue of highlights from declamations does not. Seneca, by concentrating on fragments, omits the analytical and argumentative backbone of the art, simply assuming knowledge of it among his readers. As well as developing into a performance art in its own right, as Philostratos’ Lives of the Sophists shows, declamation remained a preparation for judicial oratory and for speaking in the city councils. As noted above (Chapter 1), the professional declaimers described by Philostratos are often said to have spoken in court cases, and other recipients of this training did go on to speak in court. While declamation clearly cannot have been a mirror to practical judicial oratory, the analytical processes and techniques of presentation which it taught must have been put to use to varying degrees. Epideictic Oratory The third-century handbooks attributed to Menander Rhetor and Dionysios of Halikarnassos show the range of epideictic speeches in

A composite commentary by Syrianos, Sopatros and Markellinos is published in volume 4 of Walz, Rhetores graeci. Syrianos’ comments have been published separately by H. Rabe (Leipzig, 1892–93). Sopatros the Rhetor’s On the Division of Questions is published in volume 8 of Walz’s Rhetores graeci. On the text, see Doreen Innes and Michael Winterbottom, Sopatros the Rhetor: Studies in the Text of the Diairesis Zētēmatōn (London, 1988). Michael Winterbottom, ‘Schoolroom and courtroom’, in Brian Vickers (ed.), Rhetoric Revalued (Binghampton, 1982). See also the comments of Robert Kaster, ‘Controlling reason: declamation in rhetorical education at Rome’, in Yun Lee Too (ed.), Education in Greek and Roman Antiquity (Leiden, 2001), p. 321 on what we miss when we read Seneca. Malcolm Heath, Menander: A Rhetor in Context (Oxford, 2004) stresses the continuing practical relevance of this rhetorical training. On examples of sophists speaking in court, see Laurent Pernot, La Rhétorique de l’éloge dans le monde gréco-romain (Paris, 1993), p. 74; John A. Crook, Legal Advocacy in the Roman World (London, 1995) gives a thorough analysis of the role of advocates in Roman legal practice and concludes that the rhetorical training described by Quintilian would indeed have been of practical advantage. Dominic Berry and Malcolm Heath, ‘Oratory and declamation’, in Stanley E. Porter (ed.), Handbook of Classical Rhetoric in the Hellenistic Period 330 B.C.–A.D. 400 (Leiden, 1997), pp. 393–420 explore the interface between theory and practice.

134


the later Empire. Under this heading came speeches celebrating the events which punctuated civic life and the family life of the elite: visits by governors, departures of friends and officials, appeals for aid to the emperor after an earthquake, athletic victories, weddings, birthdays and deaths. The various genres described by Menander Rhetor and others reflect the variety of social events marked by such speeches.10 If anything unifies the types of speech treated as epideictic, it is their engagement with the visible and invisible fabric of their occasions. Aristotle’s characterization of epideictic as the rhetoric of ‘praise and blame’ and the minor place which he assigns to it do not therefore do justice to the role of the genre in the second century CE and later.11 The term epideixis, ‘display’, places the spotlight on one aspect of this type of rhetoric, which was not absent from declamation or even courtroom and judicial oratory. Part of the speaker’s aim was to display his own skills, something that Lucian certainly recognizes in the implicit comparison between himself and the peacock.12 The other ancient term for this type of rhetoric, panēgurikos logos, reflects its embeddedness in occasion.13 Though ‘panegyric’ strictly refers only to speeches composed for festivals (panēgureis), its meaning was extended in antiquity to cover all types of celebratory speeches and draws attention to the fact that these speeches were composed to mark particular occasions (not just festivals). They were therefore grounded in a specific time and place, even if they were then written down and circulated in a textual afterlife.14 These ideas of ‘display-piece’ and ‘occasional speech’ reflect aspects of epideictic, but the individual speeches were often more complex, and the epideictic genre as On the date of the ps.-Dionysian handbook, see Menander Rhetor, On Epideictic, ed. Donald A. Russell and Nigel G. Wilson (Oxford, 1981), p. 362. Donald A. Russell, ‘Rhetors at the wedding’, PCPS, 205 (1979): 104–17. On departures as poignant moments in the life of the elite, see Libanios, Autobiography, passim. 10 On the interconnection between genre and social setting, see Richard Martin, The Language of Heroes: Speech and Performance in the Iliad (Ithaca, 1989), pp. 43–4. 11 Pernot, La Rhétorique de l’éloge has undertaken an extensive re-evaluation of epideictic rhetoric, particularly in the Second Sophistic, questioning the validity of Aristotle’s dismissive comments for the appreciation of the role of epideictic in later periods of antiquity. In particular, he has pointed out the role of argumentation and persuasion in these speeches. 12 Lucian, The Hall, 11. Cf. Dio Chrysostom, 12.2–3. 13 On the terminology, see Pernot, La Rhétorique de l’éloge, pp. 36–40 and Donald A. Russell, ‘The panegyrists and their teachers’, in Mary Whitby (ed.), The Propaganda of Power: The Role of Panegyric in Late Antiquity (Leiden, 1998), pp. 20–21. 14 On the recording and circulation of epideictic speeches, see Pernot, La Rhétorique de l’éloge, pp. 465–75. As ever, we can never know to what extent the texts we have reflect the speech as performed, but I will take the surviving texts as representative of epideictic practice.


135

whole was far more varied in its compass than either of these labels would suggest. The Rhetorical Treatises on Epideictic The elementary exercise of enkōmion taught the basic starting points for such speeches, the ‘encomiastic topoi’ of birth, education, achievement which provided a basic structure for an epideictic speech (though these topoi could also be used in speeches of other types which called for the praise or blame of a person). The more advanced handbooks give detailed recommendations for various types of speech, suggesting an array of alternative strategies to suit the particular circumstances of each speech, or to circumvent difficulties (such as an honorand’s less than illustrious birth or inadequate education). It is somewhat misleading to refer to their contents as ‘rules’. Their tone is often prescriptive (the use of the second person future indicative to tell the reader what he will do in his speech is frequent), but essentially they offer flexible models for their readers to adapt to the particular circumstances in which they find themselves.15 These handbooks appear to be aimed at adult users who are called upon to speak at various events, though the individual to whom the ps.-Dionysian handbook is addressed is clearly a novice, about to give a speech at a wedding.16 Students in the rhetorical schools seem to have composed practice speeches, but an equally important source of instruction would have been their personal experience of listening to speeches both in the schools, where rhetors often composed speeches for the deaths and departures of students, and outside, at special occasions, or during the competitions in epideictic that were part of the programme in many local festivals in the Greek world.17 These speeches of praise (or less usually blame) referred directly to their time and place, praising the addressee(s), or the space in which they were pronounced, so that, unlike in declamation, the speaker’s ‘here’ and ‘now’ were shared by the audience. As the genre most closely bound up with the present moment, epideictic was the genre in which Greek responses to the present conditions of Roman rule, entirely absent from declamation,

15

Compare the remarks by Heath, Menander: A Rhetor in Ccontext, p. 17 on the function of the teaching of declamation. 16 See Chapter 1 above on the reference to the Progymnasmata in this work. 17 At the festival at Oinoanda in Asia Minor, for example, a competition in ‘encomiastic speeches’ took place on the second day of contests. See Stephen Mitchell, ‘Festivals, games and civic life in Roman Asia Minor’, JRS, 80 (1990): 183–93. See also Pernot, La Rhétorique de l’éloge, pp. 84–92.

136


could be expressed.18 The audience of epideictic were involved in far more complex ways than Aristotle’s characterization of them as ‘judges’ of the speaker’s skill would suggest (except, perhaps, in the competitions, about which we know very little). They were gratified, consoled, persuaded, moved by speakers who played on shared knowledge of past and present and the interaction between the two. This array of knowledge, the unspoken background to the speeches, is largely lost to us, but some hint of it is given by Menander’s comment that the Basilikos Logos, the speech in praise of an emperor, will contain only ‘the good things pertaining to the emperor’. The audience would presumably have been well aware of what was being left out. Shared knowledge of the conventions also added to the understanding of a speech, and the omission of a standard topos of praise would not go unnoticed. Dio Chrysostom relies on the predictability of audience expectations to turn the tables on his Alexandrian audience, breaking off a conventional praise of the city’s location to shock them out of their complacency by pointing out that the city is being praised and not its unworthy inhabitants.19 Similarly, the omission of a speech of praise from an occasion where it would have been the norm could be used to send a powerful message. A speaker arriving in a new city would deliver speech of praise on arrival to win the goodwill of the inhabitants. But Philostratos recounts how Polemo, the most self-confident of sophists, deliberately failed to pronounce an enkōmion of Athens on his first visit there on the grounds that the Athenians’ opinion of themselves was high enough as it was.20 One effect of the rise in importance of epideictic rhetoric in the Roman period is the reorientation of the exercise of ekphrasis in Aphthonios’ Progymnasmata towards its use in epideictic contexts. However, as was noted in the discussion of the Progymnasmata, the rise of epideictic in late antiquity and Byzantium was relative. In late antiquity it did not eclipse declamation and, indeed, as Malcolm Heath has argued our view of the relative importance of epideictic and declamation in later rhetorical theory and practice has been influenced by the available sources which often reflect later preoccupations.21 To this should be added the fact the epideictic itself was not devoid of argumentation; the encomiast’s claims needed to be credible and needed to be proved. As I shall suggest below, 18 See Laurent Pernot, ‘La �� rhétorique�� de l’empire ou comment la rhétorique�� grecque a inventé l’empire romain’, Rhetorica, 16 (1998): 131–48. 19 Dio Chrysostom, Oration, 32.35–7. See Michael B. Trapp, ‘Sense �� of place in the orations of Dio Chrysostom’, in D. Innes et al. (eds), Ethics and Rhetoric: Classical Essays for Donald Russell on His Seventy-fifth Birthday (Oxford, 1995). 20 Philostratos, Lives, 535. 21 Heath, Menander: A Rhetor in Context.


137

ekphrasis could be one of the means used by orators to offer proof of the assertions they made in praise of their subjects. The Progymnasmata and the More Advanced Treatises As we saw in Chapter 1, both Sopatros and ps.-Dionysios seem to assume that the student will have acquired from the elementary exercises techniques and material to rework in his speech, rather than prefabricated set pieces to be inserted at will. Poor or inexperienced orators may well have resorted to prepared passages of all types, but there is no evidence that this was the rule.22 The result, as far as ekphrasis is concerned, is that within a speech or other composition the application of ekphrasis can be difficult to spot, especially since it is neither distinguished by a special type of subject matter, nor does it necessarily constitute a ‘narrative pause’ that stops the reader in his or her tracks or a digression that calls for interpretation. One example is provided by the comparison between Libanios’ model ekphrasis of the New Year festival of the Kalends and the enkōmion of the same festival (Or. 9), delivered in front of his students.23 The ekphrasis is entirely made up of detailed accounts of the festival activities and the experiences of the participants, from the feeling of impatience on the day before, to the preparations, the exchanges of gifts, the revelry, role reversal and public entertainments, ending with the sense of calm at the end of the festival. Many of the same elements are evoked in the enkōmion but here they are interspersed with general reflections on the festival. The ekphrasis does not stand in a simple relation of part to whole, nor is it easy to isolate the use of ekphrasis within the enkōmion. Students who followed the lost versions of the Progymnasmata in which ekphrasis was subsumed within other exercises (muthos, diēgēsis, enkōmion and koinos topos) would have understood it from the very first as a flexible technique, a means of expanding material in a vivid, engaging manner which could be exploited in various ways and to various ends.24 Even the authors of the surviving Progymnasmata are keen to stress that ekphrasis should grow out of its context (see above, pp. 66 –7) and that it is essentially a part of a larger speech whose function depends on the rhetorical context, as Nikolaos makes clear. 22

�� Quintilian, for example, advises against the insertion of ready-made common places at Institutio oratoria, 2.4.28–9. 23 Jean Martin, in volume 2 of the Budé edition of Libanios’ speeches (Paris, 1988), prints a French translation of the ekphrasis alongside Or. 9. 24 Discussing the endiaskeuos type of narration (which is related to ekphrasis), ps.Hermogenes, On Invention, 2.7, pp. 124–5 notes that in actual forensic speeches it tends to be used in combination with the simple and confirmatory types.

138


Ekphrasis: The Effect of the Progymnasmata on Practice However, the elevation of ekphrasis to the status of a separate exercise, as we see in the surviving Progymnasmata, may well have encouraged its expansion and development. It is certainly difficult to imagine a student required to compose an ekphrasis of a person or a sacked city as an exercise being able to get away with a mere two lines of physical detail, as in the Homeric and Demosthenic examples cited in the handbooks. Libanios’ model exercises presumably represent the type of end result that the student was supposed to aspire to and these are all fairly substantial pieces. So the training provided by the surviving versions of the Progymnasmata may well have encouraged a conception of ekphrasis as an extended set piece within a larger composition. The growing use of ekphrasis as an independent form of composition, noted by Nikolaos, may be the end result of this tendency and may have further encouraged the expansion of the exercise.25 More significantly perhaps, the treatments of the exercise of ekphrasis in the surviving Progymnasmata tend to place all the elements of narration on the same plane. This means that ekphraseis of places, times and persons were given the same attention and importance as the ekphrasis of the action. In the conception of narration from which these elements or peristaseis derive, however, the person (prosōpon) and the action (pragma) took pride of place while the others designated the attendant circumstances of the action. This is clear in Theon’s chapter on diēgēsis in which he identifies the elements of narration as ‘the person … and the action carried out by the person, and the place in which the act took place, and the time at which the act took place, and the manner of the act and sixth the reason for these.’26 By treating all these elements equally, the exercise of ekphrasis destabilized the hierarchy inherent in the system of the peristaseis, giving equal weight to accounts of places and times – that is, to the types of ekphrasis that correspond most easily to the modern conception of description and that evoke the same kinds of criticisms, as we see in the cases of Lucian and ps.-Dionysios discussed below. The training offered by the Progymnasmata may also have affected the conception of ekphrasis and its relation to the elusive quality of enargeia. Judging from some of the models that have survived (and which therefore must have been admired sufficiently to be copied) it seems that a methodical breaking down of the subject into parts was often

25

Nikolaos, Progymnasmata, p. 70, ll. 16–19. Theon, Progymnasmata, 78, ll. 18–21: τό τε πρόσωπov … καὶ τὸ πρᾶγμα τὸ πραχθὲv ὑπὸ τoῦ πρoσώπoυ, καὶ ὁ τόπoς ἐv ᾧ ἡ πρᾶξις, καὶ ὁ χρόvoς καθ’ ὃv ἡ πρᾶξις, καὶ ὁ τρόπoς τῆς πράξεως, καὶ ἕκτov ἡ τoύτωv αἰτία. 26


139

considered sufficient.27 This is the method followed by the author of the collection of statue ekphraseis passed down under the names of Nikolaos and Libanios, in which the details of posture, gesture and clothing are laboriously catalogued. A comparable example is Libanios’ own model ekphrasis of a battle, which the speaker claims to have seen with his own eyes. As prescribed by Aphthonios and others, the ekphrasis begins with the opening of hostilities and preparations before launching into the battle itself. Here the actions are broken down into a series of antitheses, marked by the particles men … de, creating a sense of proliferation and representing in sequential form the synchronous actions of the fighting: ‘one man’s hand was cut off, another’s eye was knocked out. … the brave were struck in the chest, the cowardly in the back … one man died after killing many, another after killing only a few’.28 The description culminates in the final antithesis of victors and defeated. Such a conception of ekphrasis as a catalogue of parts is particularly evident in one Byzantine version of the Progymnasmata. Here, the term enargēs (‘vividly’) is omitted from the definition and replaced by the words ‘in detail’ (kata lepton),29 emphasizing the procedure instead of the effect. Ekphrasis in Ps.-Dionysios, On Mistakes in Declamation In this way, education may have had an effect on practice, tending to replace the ideal of enargeia with a practice of enumeration, supplemented by references to the sense of sight which signal to the reader ‘this is an ekphrasis’. This unintended consequence of the training offered by the Progymnasmata may help to explain the apparent rejection of ekphrasis by the author of another ps.-Dionysian treatise, On Mistakes in Declamatio, briefly discussed above (p. 27). The author, a particularly bad-tempered teacher of declamation, complains among other things about the overuse of ekphraseis of storms, plagues, famines, battles and deeds of valour.30 He singles out the use of ‘so-called ekphraseis’, which he defines as ‘painting (graphein) storms, plagues, famines, battles and deeds of valour everywhere’. This practice is inappropriate, he claims, because it does not address the issue of the case. He goes on to attribute this fault (hamartēma) 27

This phenomenon has been identified by Michael Roberts, The Jeweled Style: Poetry and Poetics in Late Antiquity (Ithaca, 1989) as typical of Late Antique aesthetics. In rhetoric, however, the practice of emphasizing details instead of the whole is less striking than in the poetic examples studied by Roberts. 28 Libanios, Progymnasmata, pp. 460–4. 29 Anon., Peri tōn oktōn merōn tou rhētorikou logou in Walz, Rhetores Graeci, 3, p. 595, l. 18. 30 Ps.-Dionysios of Halikarnassos, On Mistakes in Declamation, 17, p. 372, ll. 5–6.

140


to the desire to rival (zēlos) history and poetry, which both in their own way ‘bring on the appearances (opseis) of whatever they need to for the listeners’. This historical and poetic practice of evoking appearances is contrasted with the use of the same technique in judicial or mock-judicial debates (agōnes), which must be measured to fit the needs of the case (pros tēn chreian).31 It is important to note, however, that he does not deny the appropriateness of vivid language to declamation and states clearly that ‘in the essential points of the debates (agōnes) themselves there is sufficient movement of the imagination (phantasia)’.32 The problem, as he sees it, is that many declaimers do not realize this and so think that they need to ‘wheel on images (phantasias) with words from the outside’.33 Ps.Dionysios is referring to the practice of debate, when in the moment, or from concentration on the case itself, the speaker manages to appeal to the imagination, without the use of the over-elaborate prepared passages which he associates with the school exercise of ekphrasis. It is possible that his remarks were directed at the consequences of the teaching of ekphrasis in the schools and, more specifically, at the type of practices encouraged by the inclusion of ekphrasis as an independent exercise. It is certainly noticeable that the inappropriate subjects he singles out are, for the most part, subjects mentioned by the Progymnasmata. Ps.-Dionysios’ comments are thus comparable to Lucian’s observations in On How to Write History. While at one point he criticizes an anonymous historian’s inept use of ekphraseis of caves and landscapes, elsewhere he nevertheless recommends enargeia, stating that the listener, like Quintilian’s judge, should think he is seeing the events rather than hearing about them (51). Further on he advises the writer to show selfcontrol (sōphroneō) in his descriptions (hermēneiai) of landscapes (57). As Franco Montanari has pointed out, Lucian thus attacks what he presents as an abuse of ekphrasis, rather than the principle of using enargeia in 31 The Greek is extremely elliptical here. The use of the particles men and de suggests that he is referring to opseis in history and poetry on the one hand (which are themselves distinct) and in oratory on the other. Ps.-Hermogenes, On Invention, in his discussion of diaskeuē at 3.15, p. 168, ll. 1–3, notes that this type of discourse ‘rivals poetry’; he does not, however, seem to consider that this disqualifies it from being used in oratory. 32 The term I have translated, following Russell, as ‘essential points’ (ta epikaira) may also mean ‘the moment’, implying the (improvised) practice of declamation. 33 �� Ps.-Dionysios of Halikarnassos, On Mistakes in Declamation, 17, p. 372, ll. 3–10: ἐvίoις κἀκεῖvo ἁμάρτημα, αἱ καλoύμεvαι ἐκφράσεις, πoλλαχoῦ τὸ χειμῶvα γράφειv καὶ λoιμoὺς καὶ λιμoὺς καὶ παρατάξεις καὶ ἀριστείας. oὐ γὰρ ἐv τoύτῳ ἐστὶv ἡ κρίσις τῆς δίκης, ἐv τῷ διαγράψαι τὸv χειμῶvα ἀλλὰ καὶ ταῦτα ματαία ἐπίδειξις καὶ λόγoυ ἀvάλωμα. εἰσερρύε δὲ τoῦτo τὸ ἁμάρτημα ἐv ταῖς μελέταις κατὰ ζῆλov τῆς ἱστoρίας καὶ τῶv πoιημάτωv. ἀγvooῦμεv γὰρ�� ὡς ἔoικεv, ὅτι ἱστoρία μὲv πεζὴ καὶ πoίησις γραφικὰς τὰς ὄψεις τῶv ἀvαγκαίωv τoῖς ἀκoύoυσιv παράγoυσιv, ἀγὼv δὲ δικαvικὸς μεμέτρηται πρὸς τὴv χρείαv.


141

historical narratives. Interestingly, it is descriptions of static entities, such as landscapes or equipment, which attract his criticisms. So both Lucian and ps.-Dionysios agree in arguing for descriptions which are integral to their context, not ‘wheeled onto the stage from the outside’ to quote ps.Dionysios’ theatrical metaphor. However, the apparent contradiction between ps.-Dionysios’ claim that ekphrasis properly belongs in history and Lucian’s seeming exclusion of it from history is a further warning against placing too much emphasis on isolated statements by individual critics. In particular, we need to be cautious about accepting such polemical statements as representations of literary historical fact. It is important to read ps.-Dionysios’ complaint in the broader context of theory on declamation in the Imperial period for, as we shall see, the frequent references to the use of ekphrasis within these speeches show that it was more usually considered an integral part of declamation. In this context ps.-Dionysios’ statement emerges as a reaction to a widespread practice, comparable to his other complaints – for example the inclusion of overlong narrations (Chapter 14) and on the need to adapt proofs to the case in hand (Chapter 6). What the author attributes to the influence of history and poetry is not the practice of ‘placing before the eyes’ as such, but what he claims is the indiscriminate use made of it in school declamations under the direction of other teachers. His association of ekphrasis with poetry and history may well reflect the high proportion of examples drawn from these two genres in the Progymnasmata, in which case he reflects a perception that some teachers fail to explain the transition between the elementary level of training and what is required of a declaimer. The Uses of Ekphrasis in Declamation To return to the positive evidence for the use of ekphrasis, one of the most informative Greek sources on the uses of ekphrasis in declamation is Sopatros the Rhetor’s Diaireseis Zētēmatōn (On the Division of Questions), which shows how ekphrasis could be used to achieve the rhetorical aim of the speech and how selectively and carefully this was done. Sopatros himself was probably active in the fourth century, but the instructions he preserves must represent a tradition that was by then centuries old. His general procedure is to give a brief account of the case and then to identify the type of issue involved and to proceed to a detailed outline of the speech. Rather than composing a complete model declamation, he intersperses instructions with illustrative examples. The work often shows how the student was supposed to put into practice the skills acquired

142


from his elementary rhetorical studies.34 Sopatros’ uses of ekphrasis can be paralleled in earlier handbooks and model declamations by both Greek and Latin authors, and his comments are often illuminated by the commentaries on Hermogenes’ On Issues. Hermogenes himself only refers to ekphrasis once, in his work on style, Peri Ideōn Logou. Its absence from his work on issues is not surprising, given that here he focuses on the argumentative substructure of declamation with less emphasis on its presentation, a gap that is filled by his commentators. Together, all these works reveal a lesser-known facet of the use of ekphrasis in rhetoric and, by showing how ekphrasis could function within the context of a larger speech, help explain ps.-Longinos’ reference to the impact of rhetorical enargeia in combination with a wider argument. Types of Ekphrasis in Sopatros The ekphraseis Sopatros mentions often coincide with the type of subjects prescribed for ekphrasis in the Progymnasmata. As mock-judicial or deliberative speeches, examining whether a certain person committed a certain act, or should pursue a certain course of action, declamation themes naturally focused on persons and actions (pragmata), so it is not surprising that most of the ekphraseis mentioned by Sopatros evoke pragmata.35 Historical themes demanded ekphraseis of battles: a hero (aristeus) describes battles, with the use of diaskeuē, ekphrasis and ēthopoiia.36 A loosely historical speech in which Perikles is required to defend his military record in front of the Athenians (after he has unnecessarily burnt their crops) required ekphraseis of his achievements: the campaign and the cities sacked on the way.37 In another of Sopatros’ historical themes, Demosthenes describes the scene at Thebes after its sack by Alexander in a passage which has parallels with Quintilian’s description of the sacked city (8.3.67–9) and, above all, with the real Demosthenes’ evocation of the destruction of Phokis (19.65). In a further speech based on a historical incident, the student has to take on the persona of an Athenian general who evokes the storm at sea which forced him to throw the bodies of the dead overboard.38 An ekphrasis of objects is required in one case: a miser makes use of diaskeuē, and ekphrasis this time in combination with diatupōsis, 34

Innes and Winterbottom, Sopatros the Rhetor, pp. 2–3. See Schouler, ‘Personnes, faits et états de la cause’. 36 Sopatros, On the Division of Questions, p. 353. 37 Sopatros, On the Division of Questions, p. 198. See Richard Kohl, De Scholasticarum declamationum argumentis ex historia petitis (Paderborn, 1915), p. 28 on this and related themes. 38 On the popularity of this theme, see Susan A. Stephens, ‘The “Arginusae” theme in Greek rhetorical Theory and practice’, BASP, 20 (1983): 171–80. Hermogenes, On Types of 35


143

to describe the treasure and other trappings of wealth which eluded his grasp when he sold a plot of land, unaware that there was treasure buried on it. This last example is a ‘figured speech’ (eschēmatismenos logos) where the miser ostensibly seeks his own death through self-denunciation (prosangelia) but hopes instead that his extreme action will convince the judge that the treasure was rightfully his.39 Sopatros seems to have been aware that inappropriate use of ekphrasis was a danger, and that students might unthinkingly seize on every opportunity to develop a familiar subject into an ekphrasis. In one theme, a variation on the myth of Menoikeus set in a generic classical past, a plague-stricken city sends a general to consult an oracle. He is told to sacrifice his son but on his return reveals only that he was told to make a sacrifice. On learning the truth, the son commits suicide and the plague abates. The general is then put on trial and the speech is his defence. As plagues were one of the stock subjects for ekphrasis mentioned in the Progymnasmata, the student’s first impulse on hearing the outline of this case may well have been to look for an opportunity to work such a passage into the speech. But Sopatros has other advice (8.233.16–21): we should not describe the plague, since reminding the audience of its severity would harm the general’s case, given his unwillingness to do what was required to bring it to an end.40 Sopatros’ treatment of this last example shows that the use of ekphrasis had to be carefully judged and that it was thought to have an impact on an audience. The Rhetorical Functions of Ekphrasis In the cases where Sopatros’ readers are advised to use ekphrasis it is often for emotive effect, particularly to elicit sympathy for the speaker. Accordingly, the epilogue, as the place for emotive recapitulations, is a favourite position. The miser’s description (termed ‘diaskeuē, ekphrasis and diatupōsis’) of the money, wealth, attendants and t�� h�� e like which the lost treasure could have bought him belongs in the epilogue.41 In the case of the aristeus and the Olympic Games, the hero had previously put his name Style, 244–5 admired Aelius Aristeides’ storm ekphrasis in a lost speech on the Arginoussai theme. 39 See Sopatros, On the Division of Questions, p. 308, ll. 14–15. A similar case is treated in Libanios, Declamation 31, translated in Donald A. Russell, Libanius: Imaginary Speeches (London, 1996), pp. 135–45. 40 Sopatros, On the Division of Questions, p. 233, ll. 16–21. A similar case, attributed to Quintilian, in which an uncharacteristically generous tyrant kills himself to rid his city of the plague does include a brief description, justified by the different outcome. Quintilian, Minor Declamations, 329.17 with Winterbottom’s note ad loc. 41 Sopatros, On the Division of Questions, p. 309, ll. 7–9: ὁ ἐπίλoγoς ἔχει διασκευὴv καὶ ἔκφρασιv καὶ διατύπωσιv ὑπηρεσίας, χρημάτωv, πλoύτoυ, δoρυφόρωv καὶ τ�v τoιoύτωv.

144


down to compete at the games but, on hearing that his city was at war, returned home before the start of the games and distinguished himself fighting on behalf of his city. The aristeus now wishes to compete in the games. However, there is a law forbidding anyone who has put their name down but does not compete from entering the games again. The declaimer in this exercise has to argue the hero-athlete’s case. Here again the student is advised to include in the epilogue ‘an elaborate account (diaskeuē) of his acts of bravery and ekphraseis of the battle, the burial of the dead, and the victory, as well as ēthopoiiai’, all to convince the audience of his worthiness to compete, despite the letter of the law.42 These uses of ekphrasis are therefore equivalent to the other methods of arousing emotion at the end of a speech which often involved props or the introduction of family members into the court to arouse pity.43 A particularly dramatic example from an actual declamation is the end of Lucian’s Tyrannicide, a declamation in which the speaker claims the reward reserved for the killers of tyrants despite the fact that he only provoked the tyrant’s suicide by killing his son. In the dramatic epilogue Lucian imagines the speaker brandishing a sword, even addressing it as his accomplice. Using both ekphrasis and ēthopoiia he then evokes the last moments of the tyrant’s life, imagining his lamentations over his son’s body and the way in which he withdrew the speaker’s sword from the fatal wound in order to kill himself with the same weapon, making its owner his indirect assassin.44 Like Sopatros, the ancient commentators on Hermogenes’ On Issues frequently associate the use of ekphrasis with the arousal of indignation and other emotions (to deinon).45 However, this arousal of emotion is used for very precise rhetorical ends as is clear from one example of the use of ekphrasis in a deliberative type of speech discussed by Syrianos. As Nikolaos says, deliberative orators often need to use ekphrasis to persuade or dissuade. Syrianos agrees and associates the use of ekphrasis with one in particular of the standard heads of deliberative rhetoric (the telika kephalaia): advantage (to sumpheron). The precise position of the ekphrasis, he explains, �� will depend on the nature of the case. One example is a deliberative speech in which�� the orator seeks to persuade the 42

Ibid., p. 353, ll. 5–7: εἰσὶ δὲ oἱ ἐπίλoγoι πoμπικoί, ἔχovτες διασκευὴv τῆς ἀριστείας, καὶ ἐκφράσεις τῆς μάχης, τῶv ἀvαιρoυμέvωv, τῆς vίκης ἐv oἷς καὶ ἠθoπoιίαι. 43 See, for example, Aeschines, On the Embassy, 179 and the comments of Edith Hall, ‘Lawcourt dramas: the power of performance in Greek forensic oratory’, BICS, 40 (1995): 39–58. 44 Lucian, Tyrannicide, 19–21. Translation and analysis in Heath, Hermogenes, On Issues, pp. 178 and 188–9. 45 Walz, Rhetores graeci, 4, p. 694.


145

people of Sicily to abandon their island and move elsewhere because of the frequent earthquakes. If the speaker’s proposal is difficult to accept (as in this case), the ekphrasis should be placed first, as its evocation of the terrible consequences of the earthquakes will make the speaker’s other arguments in favour (kataskeuai) easier to accept: ‘if [the listeners] have been reminded of the horrors [of earthquakes] by the use of ekphrasis they will become more liable (euagōgoteroi) to be persuaded.’46 Ekphrasis here is therefore used in close connection with argumentation, as suggested by ps.-Longinos for rhetorical enargeia. Syrianos’ assessment of the impact is milder – the listener is easier to persuade, rather than enslaved – but ekphrasis still has an active power to influence, or to help to ‘lead’ (agō) the audience towards the acceptance of the speaker’s thesis. Ekphrasis and Interpretation If we look more closely at the types of speech in which ekphrasis is used, a distinct pattern emerges suggesting that the arousal of emotion and the creation of a sense of involvement in the audience through ekphrasis are used in very precise ways. The majority of the cases in which the use of ekphrasis is advised involve situations where the interpretation of a deed is at issue. The Tyrannicide theme and the case of thwarted Olympic athlete are both examples. The former is a well-known example, with many variations, of a question of definition (horos): does the person indirectly responsible for the tyrant’s death deserve to be considered a tyrannicide and to receive the reward prescribed by law? In his commentary on a version of the Tyrannicide theme in which an orator persuades the tyrant to kill himself, Syrianos recommends the use of ekphrasis of the current peace and prosperity in the city to remind the audience of the benefits brought by the death of the tyrant and for which the speaker claims ultimate responsibility.47 In Lucian’s Tyrannicide, the histrionic epilogue has the very precise effect of making the audience into witnesses of a version of the tyrant’s death in which the speaker’s responsibility is emphasized as much as possible. The hero-athlete’s case is an example of the conflict between letter and intent (rhēton kai dianoia): by withdrawing from the games the first time the speaker has technically disqualified himself so he must argue that the law in question was not aimed at persons like himself. The commentators on Hermogenes similarly recommend ekphrasis in such cases.48 Its function is not only to arouse emotion but to ‘place before the eyes’ of the audience a particular version of events and thus to cast a 46 Syrianos, Commentaria, vol. 2, p. 180 (commenting Hermogenes, On Issues, 77): ὑπoμvησθέvτες δὲ τῶv φoβερῶv διὰ τῆς ἐκφράσεως εὐαγωγότερoι πρὸς τὸ πεισθῆvαι γεvήσovται. 47 Syrianos, Commentaria, vol. 2, p. 108. 48 811, ll. 20–25 (Markellinos commenting Hermogenes, On Issues, 82–3).


146

particular light (or chrōma, ‘colour’ or ‘gloss’) on the supposed breach of the law.49 Ekphrasis and the Transference of Blame Usually, Sopatros simply instructs his readers to ‘include an ekphrasis’ at such and such a point, but there are two examples which he demonstrates in more detail. Both are subjects with a long tradition in the Greek and Latin rhetorical schools: the sack of the city of Thebes and the storm at sea. The latter ekphrasis is worked into the defence of a general who has been hauled before the courts for throwing the bodies of the war dead overboard during a storm at sea. This imaginary case is based on the story of the Athenian generals’ failure to pick up survivors after the battle of Arginoussai in 406 BCE. At their trial, according to Xenophon, they blamed their failure on the sudden storm which blew up, a type of strategy which Sopatros calls metastasis, or transfer of blame to some unavoidable circumstance.50 As in this historical precedent, Sopatros’ general had to convince his audience that the circumstances were compelling and that, placed in the same situation, they would have no option but to make the same choice. The speaker must acknowledge the act, but place emphasis on the circumstances. So Sopatros advises the student orator to include a stirring ekphrasis of the event: ‘Then describe the storm, not in a flat manner (huptiōs), but in a manner suited to a debate (agōnistikōs).’51 Like Quintilian in his account of the murder, Sopatros slips effortlessly from precept to practice, from the persona of the teacher to that of the character as he provides his example: On one side the wave was towering over me, on the other the rest of the sea was rivalling the mountains and drowning my voice with its noise, resounding even more loudly than the heavens themselves, and the sea was churning with all its expanse of waves. Seeing this, witnessing this, I yielded to the stronger and complied in all things with the demands of the moment.

49

See Syrianos, Commentaria, vol. 2, p. 135 for an example of ekphrasis as one means (among others) of creating a chrōma. 50 Russell, Greek Declamation, p. 58; Heath, Hermogenes, On Issues, pp. 21, 50–51. In Hermogenes’ classification this argument would strictly be sungnomē, as there was no third party who could be blamed. Ibid., p. 76. 51 Sopatros, On the Division of Questions, p. 224, ll. 19–20: εἶτα ἔκφρασov τὸv χειμῶvα, μὴ ὑπτίως ἀλλ’ ἀγωvιστικῶς. Innes and Winterbottom, Sopatros the Rhetor, p. 171 compare the anonymous scholia to Hermogenes in Walz, Rhetores graeci, 7, p. 365, ll. 27–9 where Isocrates is characterized as huptios in contrast to Demosthenes.


147

πῆ μὲv τoῦ κύματoς ὑπερέχovτoς, πῆ δὲ τῆς ἄλλης θαλάσσης ἀvτιφιλoτιμoυμέvης τoῖς ὄρεσι καὶ δειvόv μoι ἀvτηχoύσης ὡς ἀληθῶς, καὶ τoῖς ἐξ oὐραvoῦ παραπλησίως, στρεφoμέvης δὲ τῆς θαλάσσης αὐτoῖς ὅλoις πελάγεσι, ταῦτα ὁρῶv, ταῦτα θεώμεvoς, τῷ κρατoῦvτι ὑπήκoυσα, καὶ πρὸς τὸv καιρὸv ἡρμoσάμηv τoῖς ὅλoις.52

The example, if it is meant to be a full model of the ekphrasis the student should produce, is brief and to the point, if somewhat hyperbolic in its comparisons.53 Ailios Aristeides’ storm ekphrasis, part of a lost speech based on the historical�� case �� of the ten Athenian generals, was a particularly accomplished example mentioned by Hermogenes.54 Aristeides presented arguments against a proposal that the generals should not be buried after their execution and in it he returned to the storm that prevented them from picking up the Athenian casualties. The ekphrasis is not quoted in its entirety but what little is preserved suggests that it was restrained: after a short introduction emphasizing the role of the storm as a mitigating factor, Aristeides begins: ‘Almost as soon as the battle had begun, the sea began to swell and a strong wind blowing from the Hellespont swooped down on them.’55 Hermogenes commends this passage particularly for its concentration on evoking the grandeur of natural phenomena, which, he says, lend solemnity (semnotēs) to a speech. In this case he notes that the lack of discussion of the causes makes this example particularly suitable for oratory (politikos).56 Sopatros’ general, like the speaker in Aristeides’ defence, needs only to evoke the idea of the storm, a familiar enough event from both life and literature, in a few details. Both add the crucial detail of the effect of the event on those who were present, a detail which both adds to the impact of the description and ties it in to the central argument of the speech – that the generals acted as they did under compulsion. In both Sopatros’ example and Aristeides’ lost speech, the impact of the phenomenon described upon a human perceiver was all important: the audience were not just to ‘see’ the event, but were supposed to ‘feel’ as the 52 Sopatros, On the Division of Questions, p. 224, ll. 20–26. See Innes and Winterbottom, Sopatros the Rhetor, p. 171 on the language (though my translation differs slightly from theirs). I have attempted to translate the text as printed by Walz but it may well be corrupt. 53 Compare Janet Fairweather, Seneca the Elder (Cambridge, 1981), p. 211 on the brevity of some of Seneca’s explicationes. 54 Hermogenes, Peri Ideōn, I, p. 244, l. 17 – p. 245, l. 3. The speech is also mentioned in Philostratos,Lives of the Sophists, 584. 55 Ibid. Translation from Wooten, p. 20. It is difficult to believe that Hermogenes’ short quotation represents the whole of the ekphrasis, though Aristeides in general does not tend to be expansive. 56 On Hermogenes’ use of politikos to mean oratory, see Ian Rutherford, Canons of Style in the Antonine Age: Idea-Theory in Its Literary Context (Oxford, 1998), p. 37.


148

speaker did. The focalization of description through a viewer, emphasized by the repeated verbs of seeing in the ekphrasis (horōn, theōmenos), was therefore important to this type of rhetorical ekphrasis. Moreover, these ekphraseis are a means of making the audience enter into the state of mind of the viewer at the moment of viewing, transporting the listener into the scene for a very precise purpose: to make them see the defendants’ actions in a new light and to accept their argument that they had no alternative but to do what they did. Sopatros’ other developed example of ekphrasis works in a similar way. Like the storm, the ekphrasis of the sack of Thebes is also used in an example of metastasis, or transference of blame, again loosely based on Athenian history. The speaker is none other than Demosthenes himself who, in this historical fantasy, had been sent to present an honorific crown to Alexander on behalf of the Athenians. Upon discovering the ruins of Thebes still smouldering from the destruction visited upon them by Alexander’s troops he refused to accomplish his mission.57 On his return to Athens, he is called to account and has to explain his failure. The aim of the speech is again to convince the audience that the speaker took the best possible course of action in the circumstances: Alexander’s action was so terrible that no representative of the city of Athens could have honoured him. From the outset Sopatros states that the aim of the speech has to be to arouse intense emotion (pathos) for Thebes. More precisely, pity for the inhabitants of the city and the corresponding sense of indignation against Alexander are essential to the argument, and an ekphrasis of the sacked city was the standard way to achieve this effect, as Cicero and the Auctor ad Herennium, as well as Quintilian, point out.58 Sopatros’ ekphrasis is woven into a narration as ‘Demosthenes’ tells how he approached the area full of expectation, turning over in his mind how he would address Alexander, even running through a practice speech.59 This attention to the speaker’s state of mind is carried over into the next section of the narrative as he describes the first signs of destruction: the smoke rising from the city, the sounds of lamentation in the mountains, and his meeting with survivors who tell him the news (p. 210, ll. 7–20). And, as Quintilian advised, the statement of the event is not enough; the catastrophe must be placed before the eyes so Sopatros continues: ‘Then describe (ekphrason) the misfortune to a reasonable extent (metriōs).’ He then breaks into a demonstration with the following words: 57

5–6.

The sack of Thebes is narrated briefly in Arrian, Anabasis, 1.8 and Plutarch, Alexander,

58 Cicero, De inventione, 10.7; Rhetorica ad Herennium, 4.39.51; Quintilian, Institutio oratoria, 8.3.67–9. 59 Sopatros, On the Division of Questions, p. 209, l. 10 – p. 210, l. 6.


149

Alas, what misfortunes! What dramas have I seen Thebes suffer! They surpass even the tragedy of Mount Kithairon itself.60 One man was mourning his wife on Kithairon, another was weeping for his family, yet another mourned a friend. A woman was fleeing from the fighting and, desperate to rescue her loved ones, was turning back towards the enemy, overwhelmed by the event …? But what words could suffice to describe (ekphrasai) the misfortunes of the Thebans? The suffering (pathē) on Kithairon is little compared to the disaster that has now happened there, no one could bear the extremity of the present misfortunes. … εἶτα ἔκφρασov μετρίως τὸ κακόv. ἀλλ’ oἴμoι τῶv κακῶv. oἷα Θηβῶv τεθέαμαι δράματα, καὶ τὴv Κιθαιρῶvoς αὐτoῦ τραγῳδίαv vικήσαvτα; ἐπέvθει τις ἐπὶ Κιθαιρῶvoς τὴv σύvoικov. ἄλλoς τὸ γέvoς ἐδάκρυεv, ἕτερoς τὸv φίλov ὠδύρετo. ἔφυγεv ἀπὸ τoῦ πoλέμoυ γυvή, καὶ λαβεῖv πρoθυμoυμέvη τὰ φίλτατα ἐπὶ τoὺς πoλεμίoυς ὑπέστρεφεv, κατὰ τὴv πρᾶξιv ἀδυvατoῦσα πρὸς τὴv συμφoρὰv ἐμερίζετo. ἀλλὰ τὶς ἂv ἀρκέσει μoι λόγoς κατὰ Θηβαίωv ἐκφράσαι τὰ δυστυχήματα; μικρὰ τὰ Κιθαιρῶvoς πάθη πρὸς τὰς ἐκεῖ vῦv παραβαλλόμεvα συμφoράς· oὐκ ἂv ἐvέγκoι τις τὴv ὑπερβoλὴv τῶv vῦv ὄvτωv κακῶv.61

This ekphrasis is part of a strategy to ‘amplify the suffering of the Thebans as much as possible’.62 As in the previous example, the speaker as witness serves as an internal narrator whose reactions to the sights that meet him guide his audience’s response (just as in both Arrian and Plutarch the enormity of the event is conveyed through the account of the Greeks’ responses, rather than through details of the sack itself). In a speech like this one, the narrator’s reactions to the sight are more than a narrative device; they are an integral part of the case since in this hypothetical situation the speaker is himself on trial, not Alexander, the perpetrator of the crime. Arousing pity for the Thebans is a means of ensuring sympathy for decision of ‘Demosthenes’ not to honour Alexander. Again the ekphrasis aims to induce the audience to share the speaker’s state of mind by placing them imaginatively in his position. The combination of circumstances in this particular case mean that the ekphrasis serves a more complex function than Quintilian’s example of the sacked city, where enargeia aroused pity for the victims. Here, it is presented as a means to explain the viewer’s state of mind as much as his actions. Another striking example comes from the Latin declamations 60

Mount Kithairon was the mythical setting for the exposure of the baby Oidipous and the deaths of Aktaion and Pentheus, among other tragedies. 61 Sopatros, On the Division of Questions, p. 210, l. 20 – p. 211, l.1. 62 Sopatros, On the Division of Questions, p. 210, l. 10: εἶτα αὔξησov ὡς δυvατόv τὸ πάθoς τῶv Θηβαίωv.

150


attributed to Quintilian. Here, a father who has to justify leaving behind one son when he rescued the other from the pirates’ lair in which both were being held captive paints a vivid picture of the scene (he is now being sued by the abandoned son who later escaped): No mortal fear, no ingenuity of human conception can fully or adequately imagine what I saw there. Beneath a huge rock cliff lay a hole excavated in the ground by the pirates’ cunning, with a night far deeper than any kind of natural darkness. Circling it was the entire vastness of the sea, and on all sides the violent surf dashed and beat violently against the overhanging rocks, causing fear that the entire mass would topple down. It bristled everywhere with torture racks. … non possunt humani metus humanarum cogitationum ingenia satis habundeque concipere quae vidi. Iacet sub immensae rupis abrupto tristis et ultra naturalem profundae caliginis nocte[m] mersus piraticis artibus specus, quem tota circumfusi vastitas maris et undique minantibus scopulis illis a tempestas terrore ruiturae molis everberat. Horrent cuncta crucibus, …63

This dramatic evocation of a scene, worthy of a novel, might seem to be the kind of description that might attract accusations of needless elaboration. But it has a precise function within the economy of the speech: once again to put the audience into the speaker’s position at a precise time and place and to make them understand why he chose to save one son (who was ill) rather than the other. Ekphrasis may thus serve to depict the speaker’s state of mind at a particular point in the past by making the audience share their perceptions. In the cases analysed above, the �� audience are encouraged to ‘see’ in their mind’s eye the scene that the speaker claims to have witnessed directly. A striking variation on this occurs in Chorikios’ speech of the general who dressed as a woman (Declamation 11). This speech is given by a victorious general who has saved his city by dressing as a woman to fool the enemy troops. It is the imaginary tradition in this imaginary polis to commemorate victories in an honorific painting, recording for posterity the manner in which the victory was won. Our general’s rival (whose previous failure to defeat the enemy by traditional military means had forced the speaker to take his drastic action) has proposed this embarrassing ‘reward’ and the general now has to argue against it. In one passage the speaker elaborates on

63 Quintilian, Major Declamations, 5.16; translation from Lewis A. Sussman, The Major Declamations Ascribed to Quintilian: A Translation (Frankfurt, 1987), p. 63.


151

the state of mind that made him cross this particularly sensitive boundary.64 He does so by giving a brief and allusive ekphrasis of the mental images that motivated his decision: For I saw that, as our troops’ strength was waning and that of the enemy increasing, the situation required me to come up with a clever stratagem and, picturing (anaplasas) in my mind the capture of the city, I thought of all the terrible things that capture usually (eiōthe) brings with it, and, most bitter of all, the outrages that enemies usually (sunēthē) commit when they take a city, defiling bridal chambers, raping unmarried girls, not sparing young boys. εἶδov oὖv ὅτι ῥώμης συvεσταλμέvης τoῖς ἡμετέρoις, ηὐξημέvης δὲ τoῖς ἐvαvτίoις μηχαvῆς μoι δεῖται τὰ πράγματα, καὶ τὴv πόλιv ἁλoῦσαv ἀvαπλάσας τῷ λoγισμῷ τά τε ἄλλα διεvooύμηv ὅσα πoιεῖv ἅλωσις εἴωθε δυσχερῆ, καὶ τὸ πάvτωv πικρότατov, τὴv συvήθη τῶv ἐv πόλει κρατoύvτωv ἐχθρῶv ἀκρασίαv, παστάδα λυμαιvoμέvωv, παρθέvoυς βιαζoμέvωv, παίδωv ὥρας oὐ φειδoμέvωv.

Although only the detailing of the enemies’ likely actions at the very end of the passage counts strictly as ekphrasis, the passage is of interest because of the way it exploits the mechanics of ekphrasis and enargeia and plays on the audience’s familiarity with the traditional accounts of sacks of cities: it is almost enough for the speaker to refer to ‘what usually happens’ for the audience to share the mental image to which he refers. And, once they have understood his state of mind they will understand equally his decision to adopt his unlikely disguise and to conceal his true nature (phusis) to fulfil his duty of protecting the women and young people of the city.�� In this particular example, then, ekphrasis communicates the speaker’s state of mind, the sight he feared and acted to avoid. In all these examples, ekphrasis plays an important role – alongside argumentation – in making the audience share the speaker’s perspective. It also serves to alter their perception of certain events and their relation to the present: the general’s action was the only course open to him; the failure of ‘Demosthenes’ to crown Alexander was an honourable and patriotic gesture rather than dereliction of duty; the father had no choice but to save the weaker son, given the horror of his captivity, and the general was driven by his sense of manly duty to protect the virtue of the young to adopt his feminine disguise. As a means of creating such a gloss (chrōma) on events, ekphrasis had a full part to play in the argumentative 64

Chorikios, Declamation 11, 33 (p. 486, l. 19 – p. 487, l. 2).


152

strategy of the speech and could be a vital means of ensuring that the audience assented to the speaker’s thesis. Demosthenes, Phokis and the Rhetoric of Signs In this context, it is worth analysing in more detail the only passage from an Attic orator to be cited in the treatments of ekphrasis in the Progymnasmata: Demosthenes’ account of the scene of devastation he found at Phokis, cited by Nikolaos as an example of how ekphrasis can be used in judicial oratory to amplify a given topic: The manner (tropos) in which the wretched Phokians have been destroyed can be seen not just from the decrees [of the Amphictyonic Council] but also from the deeds which have been done (ha pepraktai), a terrible and pitiful spectacle, Athenians. For just now when we were travelling to Delphi we could not help but see all of it: ruined houses, defensive walls razed to the ground, the land bereft of young men, just women, a few little children and some pitiful old men. No one could express in words the terrible state of affairs there now. ὃv μὲv τoίvυv τρόπov oἱ ταλαίπωρoι Φωκεῖς ἀπoλώλασιv, oὐ μόvov ἐκ τῶv δoγμάτωv τoύτωv ἔστιv ἰδεῖv, ἀλλὰ καὶ ἐκ τῶv ἔργωv ἃ πέπρακται, θέαμα δειvόv, ὦ ἄvδρες ’Αθηvαῖoι, καὶ ἐλειvόv. ὅτε γὰρ vῦv ἐπoρευόμεθα εἰς Δελφoύς, ἐξ ἀvάγκης ὁρᾶv ἦv ἡμῖv ταῦτα πάvτα, oἰκίας κατεσκαμμέvας, τείχη περιῃρημέvα, χώραv ἔρημov τῶv ἐv ἡλικίᾳ, γύvαια δὲ καὶ παιδάρια ὀλιγα, καὶ πρεσβύτας ἀvθρώπoυς oἰκτρoύς. oὐδ’ ἂv εἷς δύvαιτ’ ἀφικέσθαι τῷ λόγῳ τῶv ἐκεῖ κακῶv vῦv ὄvτωv.65

The passage illustrates two characteristics of the use of ekphrasis and enargeia in oratory. Firstly, it is used here as a means of amplification: the destruction of Phokis is not a matter for debate in this speech but a given whose elaboration adds emotive effect to Demosthenes’ strategy of attacking Philip and his allies. Secondly, and above all, it adds to Demosthenes’ portrayal of his own arch-enemy, Aeschines, as complicit in the betrayal of the Phocians.66 Demosthenes uses the ekphrasis, in combination with argumentation and the use of inartistic proofs such as documents, in order to create a particular version of events. It is also interesting to consider how this ekphrasis functions in its context. For, instead of narrating Philip’s actions, Demosthenes simply 65

Demosthenes, 19.64–5. On the background, see Timothy T.B. Ryder, ‘Demosthenes and Philip II’, in Ian Worthington (ed.), Demosthenes: Statesman and Orator (London, 2000), pp. 63–70 66


153

sketches a few details which stand for the whole event and bring to mind a vivid scene of devastation. The passage serves to arouse pity for the widows and orphans whose suffering is mentioned, accompanied by a sense of indignation against Philip. This in itself was not as straightforward as it might seem because the Phocians stood accused of committing sacrilege. But what is most noticeable is that Philip and his actions are not described directly. Instead of evoking the sack itself (as Quintilian does in his example), the violence and destruction are indicated by the visible traces they have left behind. So this particular examples passage relies for its ultimate effect on what is not directly stated: the actions which led to the devastation witnessed and described by Demosthenes. In cases like this, the audience has a further task: to imagine the events that led to the state of the affairs that is directly evoked in the ekphrasis. Quintilian’s first example of the achievement of enargeia through detail (ex pluribus) at 8.3.66, discussed in Chapter 4, works in exactly this way. In his evocation of a wild party, Cicero avoids direct description of the events themselves As with the descriptions of the sacks of cities, the audience have to infer the events from their aftermath, supplying the details from their own imaginative resources and cultural knowledge. The workings of this type of ekphrasis are thus far more complex intellectually and temporally than the brief precepts of the Progymnasmata could ever suggest. From a few well-chosen and unambiguous signs the audience is expected to fill in the rest.67 In the case of Cicero’s description of Verres, ‘the rest’ is the further details, while in the case of these descriptions of the aftermath of an event, it is the past events which have left these traces that are brought to mind and, ultimately, the perpetrator and his character. The audience are thus required to bring to mind two slightly different moments in time: the moment when the signs were seen by the witness, the internal viewer, and the preceding events. It is the latter which are really in question, and the audience makes these events present through their own effort of interpretation. The examples of the use of ekphrasis from Sopatros show how ekphrasis could be put to use to aid the semblance of persuasion: description, rather than being a useless form of dilation, could be used to involve the audience, to make them feel sympathy for the participants in some event, or to present the visible traces of some past action. In this last case, verbal evocation takes the�� place of physical evidence. In so doing, the speaker is prompting the audience to imagine�� , to supply images and judgements from their cultural knowledge and their knowledge of probable outcomes. 67

Seneca, Controversiae, 3, pr. 7 commends the explicationes of Cassius Severus which were neither lengthy nor inane (‘non lentas nec vacuas’) but instead hinted at more than was actually expressed in words (‘plus sensuum quam verborum habentes’).


154

The effect of the words which prompt these imaginings can be compared to that of a painting, the difference being that the images are supplied from the audience’s own memory. Declamation and Fiction All these examples show how ekphrasis within declamation could be used to add to the persuasive effect of a speech, in appeals for sympathy or, above all, in attempts to make the audience understand the state of affairs in which the speaker found himself, or his state of mind. It is an integral part of declamation and, in the precise cases in which its use is recommended, has an important role to play in inducing the audience to assent to the speaker’s argument. There are therefore strong reasons not to take too literally ps.-Dionysios’ claims that ekphrasis has no place in declamation (or that it represents a perversion). Ps.-Dionysios goes on to make the further point that ekphrasis in declamation is particularly problematic because the speakers are not thinking of some actual referent, but ‘invent (anaplattousi) for themselves the appearances of plagues and famines and storms and wars, which have not happened in the way they describe’ (or, indeed, at all). Here, ironically, he contrasts declamation with history and poetry which both relate events that have occurred (ta sumbebēkota) thus placing the content of poetry on a par with historical facts.68 It is a rhetorical move designed to underline what is special about the case of declamation in which arguments were based on fictional events. In the context of competitive declamation, where speakers argued both sides of the case, disputes could arise about the nature of the non-existent event. It is this that he defines as a ‘useless waste of words’. Some of the speeches analyzed above help to explain his concerns. It would be possible, for example, for an opponent to produce a counterekphrasis of the storm in Sopatros’ speech claiming that it was by no means serious enough to excuse the general’s action. A debate that turned on the magnitude of an imaginary storm would indeed be ‘a waste of words’. In this particular case, however, the Arginoussai story, in which the magnitude of the storm was not questioned, provided something approaching a common set of facts that were not in dispute, a reminder of the utility of historical themes in which the characteristics of the basic events and characters were well known to all parties. An example from an entirely fictional case shows the potential problems of ekphrasis if the physical qualities of a fictional object are given too much importance in the speech. With regard to a case mentioned by Hermogenes in which a 68

Ps.-Dionysios of Halikarnassos, On Mistakes in Declamation, 17, p. 372, ll. 10–16.


155

hero is suspected of collaborating with the enemy after they set up a statue in his honour, the composite commentary contains the following advice.69 In order to emphasize the importance of this alleged sign of the hero’s treachery, the speaker must use amplification (auxēsis) by elaborating (diaskeuazō) the peristaseis, dwelling, for example, on the place (topos) in which the statue was set up, on the significance of the act (pragma) of setting it up and so on. The commentators then add that it is possible to elaborate on the material (hulē) of the statue, if it is made of gold, for example – again something that could be contradicted in a useless waste of words by the declaimer’s opponent. To a great extent, then, ps.-Dionysios’ unease about ekphrasis in declamation is caused by the fictional nature of the themes themselves. Declamation and the Uses of Ekphrasis The technical sources on declamation show that ekphrasis and related techniques had a distinct role to play in the art of persuasion. They could aid persuasion by arousing the appropriate emotions in the listener but, far more importantly, they could serve to ‘colour’ the audience’s perceptions of the events in question and were thus particularly useful when the case revolved around the interpretation of facts. The examples from declamation help to clarify ps.-Longinos’ statement about the power of phantasia when used in combination with arguments: alone, the ‘general’s’ ekphrasis of the storm or ‘Demosthenes’ ekphrasis of the sack of Thebes would merely create an impression in the mind’s eye, but as part of a larger argument they have the potential to create assent. The same examples also reveal the cognitive complexity of ekphrasis and the nature of the demands made on the audience. More surprisingly, these features of ekphrasis are also visible in epideictic contexts and it is to these that we turn now. Ekphrasis and Epideictic in Menander Rhetor The range of types of ekphrasis mentioned in the treatise on epideictic attributed to Menander Rhetor is closer to the modern understanding of ekphrasis: monuments, places and even a statue figure prominently. The city and its standard attributes, its public buildings, porticoes, its geographical situation, are ever present in these treatises. The first of the two treatises attributed to Menander offers little information on the use of ekphrasis, although topics familiar from the Progymnasmata – harbours, bays and seasons – are mentioned as typical subjects of epideictic speeches. The first Menandrian treatise also reminds readers that a single part of a city might be the subject of a speech, or that they might need to write a 69

Walz, Rhetores graeci, 4, p. 365, ll. 11–24.

156


speech celebrating the building of a bath or a harbour, or the restoration of an area of the city.70 Examples exist both in poetry and prose and are familiar from traditional surveys of ekphrasis in its modern sense (Paul the Silentiary’s verse ekphrasis celebrating the repair of Hagia Sophia is one example; Prokopios of Caesarea’s ekphrasis of the same church in Book 1 of his Buildings is another). Aphthonios’ model ekphrasis of the Alexandrian acropolis represents the type of training given for this type of speech at the level of the Progymnasmata, as do the model ekphraseis of a harbour and a temple of Tyche in the Libanian corpus (8.529–31). The second Menandrian treatise gives more direct information about the role of ekphrasis on a wider range of occasions. In his detailed instructions on how to compose each type of speech, from birthday greetings to festival orations, the author often mentions the need to describe. The term ekphrasis and the verb ekphrazō appear several times, as do synonyms such as the verb diagraphō and the noun diatupōsis. The fluidity of the technical terminology is such that these terms are often used interchangeably. The treatment of the arrival speech (Epibatērios logos) addressed to a dignitary upon his arrival in the city illustrates some of the uses of ekphrasis, and the fluidity of the vocabulary used. In a speech of this type one should describe (ekphrazō and chōrographeō) the arriving dignitary’s country of origin; one may also include an account (diatupōsis) of the misfortunes of the people under the previous governor (while carefully avoiding attributing any blame to him).71 If the speech is to be made by the person arriving (the type of display which Polemo refused to give at Athens), it should include an ekphrasis of the country in which the city in question is situated.72 He should describe (ekphrazō) the beauty of the plains, rivers, harbours and mountains. The epilogue should contain more description (the verb used here is diagraphō) and a fuller account of the city itself including its architectural glories, the visible signs of its status as a city – stoas, temples and baths.73 Similarly, when inviting a dignitary to visit one’s city in an invitation speech (Klētikos logos) one 70

Menander Rhetor, On Epideictic: Treatise I (How to Praise Cities),365, ll. 188–23: χρὴ δέ σε μηδ' ἐκεῖvo ἀγvoεῖv, ὅτι καὶ ἐπὶ μέρει τoύτωv ὅλαι ὑπoθέσεις γίvovται. καὶ γὰρ ἐπὶ λoυτρoῦ μόvoυ κατασκευῇ καὶ ἐπὶ μέρει τιvὶ τῆς πόλεως ἀvoικoδoμηθέvτι ἔστι πρoφωvεῖv. These recommendations are echoed in the ps.-Dionysian treatise on epideictic in the instructions for composing a panegyric (I.3): again one should praise temples, offerings, public and private buildings. 71 Menander Rhetor, On Epideictic: Treatise II, 379, ll. 5–8; 378, ll. 17–21. Diatupōsis does appear to be used of the more emotive types of description, specifically designed to evoke pity as here and in the Ambassador’s Speech. 72 Menander Rhetor, Treatise II, 383, ll. 13–15. 73 Ibid., 386, ll. 21–29 (p. 110). Pausanias, 10.4.1 expresses surprise that a place without any of the usual public buildings could call itself a city. For a critique of this material definition of a city’s worth, see Dio Chrysostom, Or. 32.37.


157

should include ekphraseis of the city to be honoured, as well as the sights that the addressee would see on his journey there.74 Among the types of speech discussed in the second Menandrian treatise, the ‘Sminthiac Oration’ (for the festival of Apollo Smintheus, apparently in Alexandria) contains the most detailed prescriptions for ekphrasis of monuments.75 As well as describing the festival itself, one should also describe the temple of the god and then his statue (ekphrasis and ekphrazō are used throughout). Menander provides detailed advice on how to procede: one should compare the temple to an acropolis and praise its gleaming appearance; and one should compare the statue to that of Zeus at Olympia or that of Athena on the Athenian acropolis with suitably hyperbolic remarks such as ascribing it to the hand of a Daidalos or supposing that it must have fallen from the heavens.76 (One should then go on the describe the surroundings.) The formulaic suggestions give a good idea of how much autopsy and criticism might be involved in producing an average speech of this type. Another type of speech in which ekphrasis figures prominently belongs more to the private domain of the elite household, though the civic world is by no means absent. This is the wedding speech of which Menander and ps.-Dionysios distinguish two types, marking different stages in the ceremony: the public celebrations and the more intimate bedroom speech pronounced outside the bridal chamber. In the public speech (which Menander calls Epithalamios and other rhetoricians, such as ps.-Dionysios, call Gamēlios) one should describe (diagraphō and ekphrazō) the appearance of both bride and groom.77 As noted in the previous chapter, Menander warns his reader to be particularly careful with the ekphrasis of the bride who would have been a carefully protected member of a wealthy family. Unless the speaker is a close relative he should be careful not to appear to know too much about her appearance; it is better to attribute any knowledge to hearsay – anything else might lead to malicious gossip.78 74

Menander Rhetor, Treatise II, 427, ll. 10–13 (p. 188): εἶτα διαγράψεις τῷ λόγῳ τὸv ὅλov τόπov |ὡς ἐπιλoγικόv|, ὃv δὴ διιὼv ὄψεται, καὶ ὅληv παραπέμπεις αὐτὸv τῷ λόγῳ τὴv ὁδὸv ἐκφράζωv ἠπείρoυς, ὄρη, πελάγη. 75 On the location of the speech envisaged by Menander, see Donald Russell and Nigel Wilson, Menander Rhetor, p. 351. 76 Menander Rhetor, Treatise II, 445, ll. 1–4. 77 Ibid., 404, ll. 8 and 12. 78 Again the verbs ekphrazō and diagraphō are used interchangeably: διαγράψεις δὲ καὶ τὸv vεαvίαv oἷoς ἰδεῖv, oἷoς ὀφθῆvαι, ὡς χαρίεις καὶ εὐπρόσωπoς, ὡς ἰoύλoις κατάκoμoς, ὡς ἄρτι ἡβάσκωv. τῆς παρθέvoυ δὲ φυλάξῃ διὰ τὰς ἀvτιπιπτoύσας διαβoλὰς κάλλoς ἐκφράζειv, πλὴv εἰ μὴ συγγεvὴς εἴης καὶ ὡς εἰδὼς ἀvαγκαίως <λέγoις, ἢ> λύoις τὸ ἀvτιπῖπτov τῷ λέγειv ‘ἀκηκόαμεv ταῦτα’. Menander’s reservations are not shared by Himerios who, in the protheoria to his Epithalamium to Severus, mentions the ekphrasis of the bride as an important feature of the

158


(Menander does not tell us whether this warning reflects bitter experience or merely excessive caution.) As in the case of Nikolaos’ advice about describing the act of adultery in the exercise of koinos topos, the contents of the description of the bride are assumed to betray the impressions stamped in the speaker’s mind. The mere possibility that the mental portrait of the bride, the phantasia in the speaker’s mind, might derive from illicit sense perception was enough to concern Menander. As a special display, an ambitious (philotimoumenos) speaker could also include mythological tableaux of divinities or personifications associated with marriage like Gamos (marriage) or Eros.79 This type of ekphrasis belonged, according to Menander, in the thesis section of the speech, where arguments in favour of marriage were presented, so that ekphrasis is once again used as an auxiliary to argumentation. Ps.-Dionysios’ advice on wedding speeches likewise suggests a function for vivid language which surpasses the decorative digression and which takes us back to declamation. He advises a prophetic description (diatupōsis) of the couple’s future life with their children combined with reminders of their own happy childhoods.80 Here verbal evocation of the future is presented as a parallel to personal memories of the past. Vivid language serves to ‘show’ the listener a future eventuality (one which is unambiguously positive within the value system reflected in the epithalamion). The equivalent method for evoking past time is the appeal to remember so that memory images correspond to those that are verbally evoked (and presumably contribute to them). Such use of language to prompt the interaction of memory and anticipation of the future is entirely appropriate to mark the transitional nature of the wedding. Places and persons are not the only types of subject for ekphrasis which Menander discusses. Ekphraseis of battles and military actions had a role in epideictic, as well as in the historico-fantasy world of declamation, as the discussion of the Basilikos logos, or Imperial Oration, shows. As part of his account of the actions (praxeis) accomplished by the emperor, the speaker speech. Menander expresses similar reservations about describing the physical appearance of a young pupil in the Propemptikos logos (398.14–23). 79 Ibid., 404, l. 29 – 405, l. 4: ἐξέσται δέ σoί πoτε καὶ φιλoτιμoυμέvῳ τὸv θεὸv τῶv γάμωv ἐκφράσαι, oἷός ἐστι, κατ' ἀρχὰς τoῦ λόγoῦ ἐv τῇ θέσει, ὅτι vέoς ἐστὶv ἀειθαλὴς ὁ Γάμoς, λαμπάδα φέρωv ἐv ταῖv χερoῖv, ῥαδιvός, ἐρυθήματι τὸ πρόσωπov καταλαμπόμεvoς, ἵμερov ἀπoστάζωv ἑκ τῶv ὀμμάτωv καὶ τῶv ὀφρύωv. ἐξέσται δέ σoί πoτε καὶ ἀvτὶ τoῦ Γάμoυ τὸv Ἔρωτα ἐκφράσαι ἢ ἐv ἀρχῇ τoῦ λόγoυ ἢ πρὸς τῷ τέλει. 80 �� Ps.-Dionysios, Ars rhetorica, 2.6 (p. 264): εἶτα διατυπωτέov λόγῳ oἷov πρoαvαφωvoῦvτα, ὁπoῖoς ἂv ὁ βίoς ὁ μετὰ τῶv παίδωv γέvoιτo, ὅτι τερπvότατoς χoρὸς παίδωv oἱ γέvoιτo γέρovτι, �� καὶ� ὅτι τρόπov τιvὰ ἐξ ἀρχῆς ἀvαvεάζεσθαι αὐτὸv καὶ�� ἀvηβάσκειv σὺv τoῖς αὐτoῦ παισίv. ἐv τoύτῳ δὲ ἀvάγκη �� καὶ�� ἀvαμιμvῄσκειv, ὧv πoτε �� καὶ�� αὐτὸς ἐv vεότητι ἐπoίησεv. ἡδεῖα δὲ �� καὶ�� ἡ μvήμη τῶv ἐv παισίv ἡμῖv πεπραγμέvωv.


159

should describe (diagraphō) the physical appearance of the places in which the emperor performed military feats, mentioning rivers and harbours and mountains and plains.81 But he should also describe (ekphrazō) the military actions themselves: ‘You will describe the ambushes and traps laid for the enemy by the emperor, and by the enemy for him. You will also describe the battles’.82 Examples of this are to be found in the first panegyric of Constantius by his cousin, the future emperor Julian, which contains substantial ekphraseis of the battles won in the West and in the East as well as the extraordinary strategy used by the Persians to take the city of Nisibis by diverting a river to surround it.83 Their role in the commemoration of military victories makes these ekphraseis equivalent to the visual monuments – trophies, inscriptions, reliefs and paintings – which made the same events visible to the citizens. We therefore find the whole range of subjects for ekphrasis, and not just persons and places, in Menander Rhetor and in epideictic speeches. Some of the surviving monodies for earthquake-stricken cities, for example, contain detailed narratives of the actual event. Aristeides’ Rhodian Oration recounts, in order, the silence that preceded the quake (19), the tidal wave that followed (20), the ensuing panic among the citizens (which borrows motifs from the sack of the city) and the different ways in which the dead met their end (22–23).84 The ekphrasis of the event itself culminates in a spectacular and gruesome sequence of paradoxes expressing a scene of devastation in which every type of order, religious, social and civic has been overthrown: ‘Thrown together were dead bodies, altars, ceilings, rubble, blood, utensils, roofs, foundations, slaves, masters, limbs torn from bodies, images, sacrificial victims, tombs, feasts.’85 Libanios’ monody for the temple of Apollo at Daphne, after it was either struck by lightning or set on fire by a Christian fanatic in 362, describes both his memory of the appearance of the statue it had contained and how the priestess came running into the city and the citizens gathered round helplessly as the temple collapsed.86

81

Menander Rhetor, Treatise II, 373, ll. 17–20. Ibid., 373, ll. 20–22: ἐκφράσεις δὲ καὶ λόχoυς καὶ ἐvέδρας καὶ τoῦ βασιλέως κατὰ τῶv πoλεμίωv καὶ τῶv ἐvαvτίωv κατὰ τoῦ βασιλέως. … καὶ μὴv καὶ πεζoμαχίας ἐκφράσεις. 83 Julian, Panegyric of Constantius, I, 16–19; 21–22; 25; 29–30; cf. ibid., II, 11–13. 84 For the arguments supporting Aristeides’ authorship, see Christopher P. Jones, ‘The Rhodian Oration ascribed to Aelius Aristides’, CQ, 40 (1990): 514–22. 85 Ailios Aristeides, Rhodian Oration, 25: συvεvήvεκτo τε εἰς ταυτὸv vεκρoὶ, βωμoὶ, στέγαι, κόvις, αἷμα, ἔπιπλα, ὄρoφoι, θεμέλια, oἰκέται, δεσπόται, μέλη σωμάτωv, εἰκόvες, σφαγαὶ, τάφoι, δεῖπvα. 86 Libanios, Or. 60.12, in Opera, 4, pp. 319–20. 82

160


Presence and Persuasion The uses of ekphrasis and its synonyms in the monody in particular demonstrate the emotional effect that such imagery could have on the listener. In other c�� o�� ntexts, sensual pleasure, the hēdonē identified by Nikolaos as the typical result of ekphrasis in epideictic, was no doubt one of the effects of listening to a successful evocation of a building, a person or the interweaving of descriptions of divinities, idealized portraits and quotations from Sappho that Menander suggests for the wedding speech.87 But the examples discussed by Menander reveal a far more complex set of responses than Nikolaos implies. Some types of speech discussed by Menander did involve an element of persuasion, or at least demanded that the speaker go through the motions of persuasion, and used ekphrasis to achieve their end. Once we are aware of this use of ekphrasis in epideictic to persuade, which echoes its uses in judicial speeches as reflected in declamation, we may wish to look again at some other types of ekphrasis in epideictic where the persuasive element is present, even if it is not made so explicit. The element of persuasion through the appeal to the listener’s imagination is particularly evident in Menander Rhetor’s treatment of two types of speech: the speech of invitation (Klētikos logos) and in the Ambassador’s Speech (Presbeutikos logos). Both play on the theme of the city description; with the second, we return to the theme of the city destroyed, not by imaginary or long-dead invaders, but by the very real threat of natural disaster which plagued the cities of the East. Both of these speeches use ekphrasis and its impact on the audience to affect the addressee’s mental disposition and thus to bring about (in theory at least) a change in the real outcome of events. Imagination, Memory and Persuasion in the Klētikos Logos In Menander’s discussion of the Invitation Speech (Klētikos logos) ekphrasis is clearly thought to play a persuasive role in the speech. As we have seen, Menander suggests that the speaker faced with the task of inviting a Roman governor to visit his city should give an ekphrasis of the lands he will see on his journey and of the city itself with its attractions – the rhetorical equivalent of the photographs of tropical beaches on display in northern European cities in the winter months. Quintilian’s comments about the need for vivid language to conform to expectations, combined with Augustine’s account of how he imagined the unknown city of Alexandria on the basis of his real knowledge, acquired through 87 Syrianos, Commentaria, vol. 1, p. 14, 23–32 discussing Hermogenes, On Types of Style, 219 (glukutēs) cites Sappho’s poetry as an example of sensuous pleasure through ekphrasis.


161

the bodily eyes, of Carthage, suggest how the speaker might have had to approach this task.88 For the audience who did not know the city, the ekphrasis must have been general enough to allow the invitee to create a mental picture on the basis of prior knowledge (just as each tropical beach is sufficiently standardized to allow us to project our ready-made fantasies onto it, while still representing a specific location). Menander goes on to make an important distinction: the approach will depend on the previous experience of the addressee. If the addressee has actually visited the city before, he explains, there is no need to evoke its advantages in a full-blown ekphrasis. Instead, it will be sufficient simply to remind him of what he has already seen with his own eyes.89 Menander’s practical advice relies on the assumption that the experience of the city has left a mental image firmly lodged in the visitor’s memory (just as Augustine was convinced that his long exposure to Carthage had left that city engraved in his mind). The effect of ekphrasis on the person who has never seen the city is therefore equivalent to the effect of a memory image in the mind of the person who has seen it. In this case, ekphrasis is described as having an emotive and persuasive function, spurring the listener to wish to see the sight described with the eyes of the body. Menander is quite specific about the emotion which this appeal to memory should arouse: the speaker is advised to assimilate the former visitor’s desire to revisit the city to the erotic longing (pothos) of a lover (erastēs) who is far from his beloved.90 Menander’s discussion of the Klētikos logos shows how, within epideictic rhetoric too, ekphrasis can perform a specific function and cannot be reduced to gratuitous decoration or to the display of the orator’s talents. Ekphrasis here is still thought of as working on the mind of the listener and even as producing concrete results. Of course, a Roman dignitary’s decision whether or not to visit a certain city was governed by rather more practical considerations. But it is significant that a form of rhetorical persuasion played a part in the elaborate choreography of relations between Greek cities and their Roman governors and that ekphrasis contributed to this persuasion. Ekphrasis and the Emperor’s Tears Evidence for actual audience response to epideictic is rare (Libanios’ anecdote about his rival’s disastrous performance being one exception), but Philostratos does record the dramatic effect of a letter written by Aelios Aristeides on the emperor Marcus Aurelius. The letter was a 88

Augustine, De Trinitate, 8.6.68–89. See Chapter 5, above. Menander Rhetor, Treatise II, 428, ll. 7–9. 90 Ibid., ll. 12–15. 89

162


plea, addressed to Marcus Aurelius and Commodus, for help after the earthquake which devastated the city of Smyrna in 177 or 178.91 Marcus Aurelius in particular, says Philostratos, was so moved by the letter that he shed tears on the page. Aid for the stricken city was immediately forthcoming, and Aristeides (rather than the emperors) gained a reputation as the founder of the city.92According to Philostratos, the words of the master sophist were enough, even at a distance, to cause real cities to rise again. The letter survives and corresponds to the type of speech which Menander calls Presbeutikos Logos, or Ambassador’s Speech.93 According to Menander, the Ambassador’s Speech should contain a diatupōsis and an elaborate account (diaskeuē) of the destruction suffered by the city. He advises the speaker to pay particular attention to the impact on monuments such as baths and aqueducts that were most likely to interest the emperor (and to attract Imperial funding for repairs).94 Aristeides handles this in an interesting way which takes us back to Menander’s advice on the invitation speech. For he does not describe the former splendour of the city (as he had done in his Seventeenth Oration on the city) but instead asks his audience to remember their own experience of it. Remember what you said when you looked upon it as you arrived, remember what you said when you entered, how you were disposed, what dispositions you made … was there a view which did not make you more cheerful? Which sight did you behold in silence and not praise as befits you? These are things which even after your departure you did not forget. The people were celebrating the Theoxenia, and you rested as if in the most civilized of your possessions. What did your gaze touch upon that did not make you happier? What, of all that you saw, did you pass by in silence and not with the words of praise that befit you? All this you remembered even after your departure. Now it all lies in the dust. ἀvαμvήσθητε ὧv ἐπὶ τῆς πoρείας ἐφθέγξασθε ὁρῶvτες εἰς αὐτὴv. ἀvαμvήσθητε ὧv εἴσω παρελθόvτες, ὡς διετέθητε, ὡς διεθήκατε. oἱ μὲv Θεoξεvία ἦγov, ὑμεῖς δὲ ὡς ἐv τoῖς ἡμερωτάτoις ὧv κέκτησθε ἀvεπαύεσθε. πoία πρoσβoλὴ τῆς ὄψεως oὐχ ἡδίoυς ὑμᾶς ἐπoίησεv; τί τῶv πάvτωv ἐθεάσασθε σιγῇ καὶ oὐ μετὰ τῆς πρεπoύσης ὑμῖv εὐφημίας; ὧv oὐδὲ ἀπελθόvτες ἠμvημovήσατε. ἃ vῦv πάvτα ἐv κόvει.95 91 See Ailios Aristeides, The Complete Works, trans. Charles A. Behr (Leiden, 1981), vol. 2, p. 358, n. 1. 92 Philostratos, Lives of the Sophists, 582. 93 On epideictic in letter form, see Pernot, La Rhétorique de l’éloge, pp. 435–7. 94 �� Menander Rhetor, Treatise II, 423, ll. 19–25. 95 Ailios Aristeides, Letter to the Emperors (Or. 19), 2.


163

The memory image again can stand in for the image which, in other circumstances, Aristeides would have had to convey in words. In the lines which follow, Aristeides proceeds to enumerate some of the sights following each immediately with the statement of its loss: The harbor, like an eye, is closed, the beauties of the market place are gone, the adornments of the streets have disappeared, the gymnasia together with the men and boys have been destroyed, some of the temples have fallen, others have sunk beneath the ground. The most beautiful of cities to behold, renowned for its beauty among all men, is the ugliest of spectacles: a hill of ruins and corpses. The west winds blow through a waste land�. δὲ�� ἀγoρᾶς κάλλη, κόσμoι δὲ�� ὁδῶv ἀφαvεῖς, μέμυκε μὲv ἐκεῖvoς λιμὴv, oἴχεται �� γυ�� μ�� vάσια δὲ�� αὐτoῖς ἀvδράσι καὶ παισὶ διέφθαρται, vαoὶ δὲ�� oἱ μὲv�� κεῖvται, �� oἱ� δὲ�� κατέδυσαv. ἡ δὲ�� πρὸς θέαv ὡραιoτάτη πόλεωv καὶ τoῦ κάλλoυς ἐπώvυμoς ἀvθρώπoις ἅπασιv ἀωρότατov θεαμάτωv ἀπoπέφαvται, κoλωvὸς ἐρειπίωv καὶ vεκρῶv, ζέφυρoι δὲ ἐρήμωv ἐπιπvέoυσι.96

It is the culmination of this passage which is supposed to have reduced Marcus Aurelius to tears: the image of the wind whistling through the ruins. Aristeides’ potent and compressed image is reminiscent of a line by the fourth-century orator Deinarchos, who described the sacked city of Thebes in the following terms: ‘swallows fly around the city of the wretched Thebans’.97 It gains its effect from the preceding passages, which moved from the invitation to recall the beauties and pleasures to the stark juxtaposition of the memory of its monuments with their present state of devastation. It seems that Aristeides was well aware of the potential impact of his culminating image for he follows it immediately with his appeal to the emperors for the help that he hopes will be motivated by their sense of pity. In certain cases, therefore, the uses of ekphrasis, combined with appeals to the personal memory images of the audience, could have a persuasive effect inducing a state of mind conducive to the action sought by the speaker. In one final type of epideictic speech, ekphrasis contributes to persuasion in a way that is closer to its function in some types of judicial speeches: by providing a type of proof of the speaker’s claims about his subjects. 96

Ibid., 3. Deinarchos, fr. F 3: περιπέτovται δὲ τὸ τῶv ταλαιπώρωv Θηβαίωv ἄστυ χελιδόvες. Apsines, Art of Rhetoric 31, ed. and trans. Mervin R. Dilts and George A. Kennedy (Leiden, 1997), p. 220 cites Deinarchos’ line as a means of ‘making present’ (paristēmi) the desolation of the city. 97

164


Ekphrasis, Proof and Persuasion in the Imperial Oration The Imperial Oration (Basilikos logos), one of the most long-lived of the genres discussed by Menander Rhetor, does not, at first sight, seem to require any persuasion. It is, as Menander makes clear, ‘a generally agreed amplification of the good things pertaining to the emperor and allows no ambivalent or disputed features.’98 As such it might seem to be the last place that we might expect to find any sort of argumentation or attempt at persuasion. However, even here, the orator could not simply state the achievements of the emperor but needed to offer some proof of his claims. This is the function of the ekphraseis of battles and other military actions whose inclusion Menander advises.99 The audience thus become virtual witnesses of these deeds, through the orator’s evocation of the events. These are to be treated once more as signs, in this case of the emperor’s military prowess and wisdom, but this is not to be left to the audience to interpret. Menander instructs his reader to link up a passage on the emperor’s wisdom, ‘saying that he was himself the planner, the commander, the discoverer of the moment for battle, a marvellous counsellor, champion, general, orator …’100 Menander’s discussion shows that epideictic speeches, rather than being a catalogue of platitudes, had a persuasive function as well and that ekphrasis, along with appeals to more tangible forms of proof such as victory monuments and processions, played a role in this.101 This aspect of epideictic oratory also helps to underline the relevance to epideictic of the training in techniques of argumentation provided by the study of declamation and suggests that the growth in importance of epideictic in late antiquity did not necessarily mean a decline in declamation or in skills of argumentation.

98 Menander Rhetor, Treatise II, 368, ll. 4–7 (Basilikos Logos): αὔξησιv ὁμoλoγoυμέvηv περιέξει τῶv πρoσόvτωv ἀγαθῶv βασιλεῖ, oὐδὲv δὲ ἀμφίβoλov καὶ ἀμφισβητoύμεvov ἐπιδέχεται διὰ τὸ ἄγαv ἔvδoξov τὸ πρόσωπov εἶvαι … 99 Ibid., 373, ll. 16–28. 100 Ibid., ἐvταῦθα δὲ καιρὸv ἕξεις καὶ ἐπισυvάψαι περὶ φρovήσεως, ὅτι αὐτὸς ἦv ὁ διαταττόμεvoς, αὐτὸς ὁ στρατηγῶv, αὐτὸς ὁ τὸv καιρὸv τῆς συμβoλῆς εὑρίσκωv, σύμβoυλoς θαυμαστός, ἀριστεύς, στρατηγός, δημηγόρoς. Julian, Panegyric of Constantius, I, 34–36 explains the qualities illustrated by the preceding account of Constantius’ campaigns. 101 See Pernot, La Rhétorique de l’éloge, pp. 659–724 and Ruth Webb, ‘�� Praise and persuasion: argumentation and audience response in epideictic oratory’, in Rhetoric in Byzantium, ed. Elizabeth Jeffreys (Aldershot, 2003), pp. 127–35�.


165

Conclusion: Ekphrasis as a Rhetorical Technique I have largely limited this analysis of the use of ekphrasis as a persuasive technique to the theoretical sources which are of particular interest in that they reveal some of the functions of ekphrasis in the context of larger speeches. In declamation it served frequently to present facts in a certain light (chrōma) and was one means of inducing the audience to assent to the orator’s version of events. In particular, we have seen how it could be used to make the audience share the experience and the emotions of the speaker and thus to accept the wider argument that, in the circumstances, his actions were reasonable. The examples analysed also help to explain ps.-Longinos’ comments about the power of rhetorical phantasia. This can ‘enslave’ the listener on condition that it is used in combination with proofs and arguments. In epideictic oratory, the wide variety of functions served by ekphrasis reflects the range of topics evoked – from gods to battles, from persons to cities – and the infinite variety of different relationships between the orator, his addressee and the subject matter of the ekphrasis. Whereas judicial oratory needs above all to evoke moments from the past, epideictic ranges over past, present and future, combining verbal evocation with the personal memory images of the addressees, sometimes to achieve an intense emotional impact. Far from being ready-made passages to be transferred from one speech to the next as displays of the orator’s skills, successful ekphraseis needed to be carefully adapted to their context, to their subject and to their listeners. The comments on ekphrasis to be found in the treatises on declamation and epideictic show how it fitted into the wider rhetorical system and what an elementary training in ekphrasis prepared the student for. They also reveal how ekphrasis belongs to a particular way of understanding and using language. The uses of ekphrasis as a means of persuasion that we see in the handbooks on epideictic and declamation correspond to Quintilian’s conception of the use of enargeia and to ps.-Longinos’ remarks on the role of phantasia in rhetoric. But these are practical manuals, not works of literary criticism. The only work that could, perhaps, deserve this title is On the Sublime and even there the analyses of ancient and contemporary texts are made with a view to producing new compositions that will have the elusive quality of sublimity. There is therefore a great deal that they leave aside. In particular, in their practical emphasis on the ability of the word to create images, and on the ways in which those images can be exploited to help create assent, they largely ignore the paradoxes of enargeia outlined in Chapters 4 and 5. The next chapter will therefore explore some of these paradoxes as they are found in rhetoric and in other prose texts of the period.


7. The Poetics of Ekphrasis: Fiction, Illusion and Meta-ekphrasis

Introduction The rhetorical handbooks’ practical orientation makes them valuable sources. However, this same practical focus also places limitations on what their authors are willing to tell us. With the exception of the passing allusion to the philosophical sense of phantasia in ps.-Longinos’ On the Sublime (which is any case an exception that cannot easily be classified as a ‘rhetorical handbook’), their authors are rarely interested in exploring the philosophical implications of enargeia and phantasia. Nor, on the whole, do they pay attention to the paradoxes inherent in enargeia that emerge clearly when one compares the rhetorical usage of the term with the earlier meaning of ‘directly perceptible to the senses’. As teachers, the rhetoricians emphasize the practical over the speculative and, as vendors of a particular method, they are interested first and foremost in stressing the power that the word is able to wield through enargeia and ekphrasis. This chapter therefore looks beyond the explicit discussions of ekphrasis in the handbooks to consider just some of the questions raised by the practice of ekphrasis. As it is not possible in the space available to explore all the implications of ekphrasis, nor to do justice to the complex questions raised, I have chosen to focus on two main areas. First, I will discuss some of the questions raised by the rhetoricians’ own claims for ekphrasis in both epideictic and declamation. In the case of epideictic, the particular relationship of the speech to the time and place of its delivery raises some important questions about the interaction of the orator’s visual evocations with the audience’s personal knowledge and memory and with their perceptions of their surroundings. In particular, the requirements of epideictic oratory often demanded that the speaker describe sights that were literally before the audience’s eyes as they listened to the speech. The fictive nature of the events debated and described in declamation, by contrast, raises problems of a different nature, in that, as the author of On Mistakes in Declamation points out, the events described are imaginary. Moreover, the fact that the audience of a declamation, like the readers of a novel, were fully aware of the fictive nature of the events that they Laurent Pernot, La Rhétorique de l’éloge dans le monde gréco-romain (Paris, 1993), pp. 441–3.


168

were being asked to imagine adds an extra layer of complexity to their response. Second, I will discuss some examples of uses of ekphrasis in texts that, while not primarily rhetorical, play on the rhetorical theory and, in their various ways, offer an indirect comment upon the rhetoricians’ claims. This group contains two main types of source: the Greek novels, particularly Achilles Tatios’ Kleitophon and Leukippe and Heliodoros’ Aithiopka, and certain ekphraseis of works of art, notably Philostratos’ Eikones. I suggest that these texts contain references to the rhetorical theory discussed in the previous chapters and invitations to reflect upon the ambiguities and paradoxes of enargeia, the art of making absent things seem present that is central to the project of fiction. Ekphrasis as Fiction Though the rhetoricians make confident claims for the power of ekphrasis and enargeia to recreate experience, they themselves clearly recognize the limitations of enargeia. Time and time again, as we have seen, their claims are qualified by the vocabulary of approximation and likeness: ekphrasis makes the audience ‘almost’ (schedon) see or ‘tries to turn listeners into spectators’. The effects of ekphrasis and enargeia thus belong to the domain of likeness, of semblance, just as much as the scenes they attempt to convey, which are, as we have seen, characterized by their likeness to truth (‘secundum verum’, enalēthes). This is entirely appropriate to the rhetorical context given that rhetoric dealt in likeness to truth and in verisimilitude and plausibility (to eikos, to pithanon). As we have seen, Quintilian’s first reference to enargeia comes in his discussion of this type of narrative and its need for the vividness that will make it come alive. In its fundamental like-ness, enargeia is intimately related to fiction; it evokes sights, sounds and sensations of absent things that, moreover, have the power to make us feel ‘as if’ we can perceive them and share the associated emotions. Enargeia thus invokes the fundamental duality of fiction, which demands of its audience a state of mind that has been described by Jean-Marie Schaeffer as ‘split’ (‘scindé’). The audience (whether readers, listeners, viewers or spectators) combine a state of imaginative and emotional involvement in the worlds represented with

Theon, Progymnasmata, 119, l. 32; Nikolaos, Progymnasmata, p. 70, ll. 5–6. Quintilian, Institutio oratoria, 6.2.30; ps.-Longinos, On the Sublime, 15.8. On enargeia and fiction, see, in particular, Barbara Cassin, ‘Procédures sophistiques pour produire l’évidence’, in Carlos Lévy and Laurent Pernot (eds), Dire l’évidence (Paris, 1997), pp. 15–29.

Jean-Marie Schaeffer, Pourquoi la fiction? (Paris, 1999), p. 190.

The Poetics of Ekphrasis

169

an awareness that these worlds are not real. Schaeffer’s analysis, which takes into account all types of fiction and imaginative activity, from daydreaming to the novel, cinema and video games, has the advantage of revealing the complexity of our responses to fiction, in which immersion in the fictive situation does not involve the loss of awareness of reality. The closest ancient category to our notion of fiction that is to be found in the surviving sources is the rhetoricians’ plasmata. This category of narratives that had not occurred but which were like truth (unlike the fantastic mythological tales of the Furies, centaurs and winged chariots) was particularly associated with rhetoric. A plasma is modelled (plattō; cf. Latin ‘fingere’) from the pre-existing material provided by reality and thus occupies an intermediate position between truth and lies. Nikolaos, for example, in his chapter on the exercise of narration, defines plasmata as being similar to the fabulous tales of myth in that both were invented but different in that plasmata could have occurred. There is therefore an intimate connection between the conception of enargeia, which is the product of a mental image that is modelled by the orator from elements of experience, and the project of fiction itself. And, like fiction, the products of enargeia and ekphrasis are themselves both present in the imagination and absent from the world perceived by the senses. Ekphrasis and Absence I will start with some examples from Menander Rhetor himself which rely for their rhetorical effect on the ultimate inability of the word and the mental images it provokes adequately to replace the sense of sight – for some of the uses of vivid evocation in epideictic play on absence and emphasize the differences between past, present and future. The difference between memory images and direct perception is alluded to by Menander Rhetor in his comments on the Speech of Arrival (epibatērios logos) to be made by the visitor: Immediately after the prooemia, which are based on joy, you will develop a head (kephalaion) containing an amplification of the opposite

See John R. Morgan, ‘Make-believe and make believe’, in Christopher Gill and T.P. Wiseman (eds), Lies and Fiction in the Ancient World (Exeter, 1993) and Barbara Cassin, L’Effet sophistique (Paris, 1995), pp. 449–60. Cassin stresses the close relationship between declamation and fiction. Nikolaos, Progymnasmata, p. 13, ll. 9–13: ἔτι κoιvωvεῖ τὰ πλασματικὰ διηγήματα τoῖς μύθoις τῷ ἀμφότερα πεπλᾶσθαι, διαφέρει δὲ καὶ ταῦτα ἀλλήλωv, ὅτι τὰ μὲv πλασματικὰ διηγήματα, εἰ καὶ μὴ ἐγέvετo, ἀλλ’ ἔχει φύσιv γεvέσθαι, oἱ δὲ μῦθoι oὔτε ἐγέvovτo oὔτε φύσιv ἔχoυσι γεvέσθαι. This analogy does not mean that fabulous things, like the Furies, cannot be presented with enargeia.


170

emotion as follows: ‘I was distressed, as is only natural, in the past and mourned because I was deprived of the sight (atheamōn) of such beauties and of the city which is the most beautiful city upon which the sun looks down. But when I saw her, I put aside my grief, I shook off distress; I see the spectacle (thea) of everything I longed for, not as images in dreams or like the shadowy reflections in a glass, but the very shrines themselves, the acropolis itself, the temples, harbours and colonnades.’ oὐκoῦv μετὰ τὰ πρooίμια ὄvτα ἐκ περιχαρείας κεφάλαιov ἐργάσῃ ἔχov ἐvαvτίoυ αὔξησιv oὕτως. ὅτι ἐδυσχέραιvov δὲ ὡς ἔoικεv τὸv παρελθόvτα χρόvov καὶ ἠvιώμηv ἀθεάμωv ὑπάρχωv κάλλεωv τoσoύτωv καὶ πόλεως, ἣv μόvηv καλλίστηv πόλεωv ὁ ἥλιoς ἐφoρᾷ. ἐπειδὴ δὲ εἶδov, ἐπαυσάμηv τῆς λύπης, ἀπεσεισάμηv δὲ τὴv ἀvίαv, ὁρῶ δὲ ἅπαvτα ὧv ἐπόθoυv τὴv θέαv, oὐκ ὀvειράτωv εἰκόvας oὐδὲ ὥσπερ ἐv κατόπτρῳ σκιάς, ἀλλ’ αὐτὰ τὰ τεμέvη, αὐτὴv τὴv ἀκρόπoλιv, αὐτoὺς τoὺς vεὼς καὶ λιμέvας καὶ στoάς.

Here, memory images are described as falling far short of direct perception, as mere shadows compared to the experience of seeing the city itself. By implication, the effect of ekphrasis, which is comparable to these shadow images, therefore also fails to substitute for reality. This passage sheds a different light on the account of the Invitation Speech (Klētikos logos) analysed in the previous chapter. In Menander’s discussion of this speech, the ekphrasis, or personal memory image, of the city to be visited is intended as a spur to action, a lure to arouse the listener’s desire to enjoy the spectacle for himself through direct sensory perception. The persuasive impact of the speech thus resides ultimately in the lack of identity between the image and the result of direct perception and the inability of the verbal evocation to substitute for reality. One particular class of epideictic discourse places special emphasis on the distinction between the mental images produced by ekphrasis and the perception of reality. These are the monodies for the dead, usually those who have died young, or for cities destroyed by earthquake or other disasters. As we saw above in the case of Aristeides’ Letter to the Emperors, the careful use of visual evocation could have an intense emotional impact on the listener, but that impact was achieved precisely because of the gulf between the imagined or remembered beauties of the person or city and the present, visible state of ruin. In such speeches the use of ekphrasis, or of the appeals to remember that may substitute for ekphrasis, takes place within a specific temporal economy which constantly moves between past, present and future as the monody juxtaposes the life which existed in the past, the present state of loss and destruction and the future that, in

Menander Rhetor, Treatise II, 382, l. 31 – 383, l. 9


171

the case of a person, will never be.10 Within this, ekphrasis has a complex role to play. Menander advises the reader about to deliver a monody for a young person as follows: ‘Then you will describe (diatupoō) his physical appearance, what he was like, what beauty he has lost, the blush of his cheeks …’ We might suppose that one effect of such diatupōsis could be consolatory, to substitute in some way for the loss of the person evoked. But the advice that Menander goes on to give suggests rather that these evocations of beauty were part of an enactment of that loss. Each element of the intensely sensual description is in fact used as part of an extended contrast between past and present: … what beauty he has lost, the blush of his cheeks, what a tongue has been stilled, the soft bloom on his cheeks is clearly wilted, the locks of his hair will no longer be admired, the radiance of his eyes and the pupils now sleeping, his curved lashes curved no more, all destroyed. εἶτα διατυπώσεις τὸ εἶδoς τoῦ σώματoς. oἷoς ἦv, oἷov ἀπoβέβληκε τὸ κάλλoς, τὸ τῶv παρειῶv ἐρύθημα, oἷα γλῶττα συvέσταλται, oἷoς ἴoυλoς φαίvεται μαραvθείς, oἷoι βόστρυχoι κόμης oὐκέτι�� λoιπὸv περίβλεπτoι, ὀφθαλμῶv δὲ βoλαὶ καὶ γλῆvαι κατακoιμηθεῖσαι, βλεφάρωv δὲ ἕλικες �� oὐκέτι�� ἕλικες, ἀλλὰ συμπεμπτωκότα.11

Each sign of beauty is no sooner named than declared lost, as the diatupōsis oscillates between past and present, exploiting the poetic capacity of language to evoke absent things which are contrasted systematically with the present. The effect is, in miniature, the same as that achieved by Aristeides in his Letter to the Emperors in which the image of Smyrna, carefully constructed by verbal evocation and appeals to memory images, is replaced by the image of wind-blown desolation of the present. But, whereas in other contexts, ‘making absent things present’ is the unproblematic achievement of enargeia, in the type of monody envisaged by Menander, the accent is on the gulf between the verbal image and reality: between the sensuous detail with which the beauty of the youth is enumerated and the corpse. It is highly likely that in the ritual context of the performance of the monody, which is conceived as a brief, spontaneous lament, the body of the deceased was visible to the speaker and his audience.12 These passages therefore involved a juxtaposition of two levels of perception for the mourners: their direct perception of the present fact 10

Menander Rhetor, Treatise II, 435, ll. 16–29. Menander Rhetor, Treatise II, 436, ll.15–21. 12 This is the implication of Lucian, On Mourning, 13, which alludes directly to the custom of uttering a monody over the body. 11


172

of death and their memory of the deceased as he was in his lifetime. The effect is comparable to that of affixing a portrait of the dead as they were in life to their mummy, superimposing one moment on another. Ekphrasis and the Danger of Superfluity The fact that the referents of epideictic ekphrasis were often known to the audience, or even visible as the speaker pronounced his discourse, led to a different type of problem. Ekphrasis in epideictic thus ran the risk of being superfluous when the audience actually had before their physical eyes the sight which the orator was supposed to bring before the eyes of the mind. Speakers were themselves aware of this problem (and of the fact that their speeches were likely to be circulated later to different audiences), and none more acutely than Lucian whose elaborate prolalia (introductory talk) The Hall is almost entirely devoted to a discussion of the problems faced by a speaker performing in sumptuous surroundings. The speech begins with praise for the beauty of the hall, presented as inspiring to the speaker: Seeing a hall immense in its proportions, surpassing all others in its beauty, gleaming with light, sparkling with gold, beautifully adorned with paintings, would one not desire to make speeches within it, if this were one’s occupation, and to enhance one’s reputation and distinction and fill it with sound and, as far as possible, to become part of its beauty oneself …? oἶκov δέ τις ἰδωv μεγέθει μέγιστov καὶ κάλλει κάλλιστov καὶ φωτὶ φαιδρότατov καὶ χρυσῷ στιλπvότατov καὶ γραφαῖς ἀvθηρότατov oὐκ ἂv ἐπιθυμήσειε λόγoυς ἐv αὐτῷ διαθέσθαι, εἰ τύχoι περὶ τoύτoυς διατρίβειv, καὶ ἐvευδoκιμῆσαι καὶ ἐλλαμπρύvασθαι καὶ βoῆς ἐμπλῆσαι καὶ ὡς ἔvι μάλιστα καὶ αὐτὸς μέρoς τoῦ κάλλoυς αὐτoῦ γεvέσθαι.13

This model of inspiration and of the assimilation of the speaker with his venue is immediately contrasted with the behaviour of the uncultivated viewer who stares in mute wonder.14 In language that refers directly to the theories of vision and memory underlying ekphrasis and enargeia, Lucian 13

Lucian, The Hall, 1. See the analyses of Simon Goldhill, ‘The erotic eye: visual stimulation and cultural conflict’, in S. Goldhill (ed.), Being Greek under Rome: Cultural Identity, the Second Sophistic and the Development of Empire (Cambridge: Cambridge University Press, 2001) and Zahra Newby, ‘Testing the boundaries of ekphrasis’, in Jaś Elsner (ed.), The Verbal and the Visual: Cultures of Ekphrasis in Antiquity = Ramus, 31, 1–2 (2002): 126–35�. 14


173

then goes on to propose a direct relationship of cause and effect between seeing and speaking, claiming that beauty flows into the soul through the eyes and then sends out words in its image.15 But the speaker of this passage is soon interrupted by an imaginary objector, a personified logos, who puts forward an alternative position (14): the hall distracts the audience and the speaker himself from the speech. The opposing speaker subverts the language of ekphrasis when he claims that the audience ‘become spectators instead of listeners’ (18), for here the seeing is literal rather than the metaphorical, mental image brought about through ekphrasis. Its speaker implies that the impact of physical sight is far stronger than the mental imagery created through words, so in place of the union between vision and speech on the one hand and speaker and audience on the other proposed by the first speaker, we are presented with a model of antagonism in which direct perception through sight prevails over the word.16 This second speaker goes on to illustrate this point through his description of the paintings inside the hall. Introduced by the disclaimer ‘painting in words is a bald (psilos) thing’ (21), these ekphraseis are indeed ‘bald’; the pictures are treated for the most part in a cursory manner, as if to prove the objector’s point. Anyone familiar with the rich possibilities of ekphrasis of paintings as demonstrated by Philostratos, for example, would have been well aware that this speaker was not making full use of the power of the word and that his introductory statement marks a selffulfilling prophecy. Taken as a whole, The Hall can be read as an elaborate captatio benevolentiae and an elegant introduction to the main performance which was to follow in which Lucian manages to draw attention to his own skill and to the beauty of the setting, which presumably was to the credit of the community, his audience. Rather than present a straightforward enkōmion, or enkomiastic ekphrasis (in the manner of his Hippias, an ekphrasis of a bath building presented as a periegesis of the building), he discusses the issues raised by the idea of achieving sight through words in general and in particular by the epideictic practice of describing and praising sights which were present to the audience. The dialogue, in typical Lucianic fashion, serves to raise questions, to signal awareness of the problems rather than proposing any single solution or point of view.17 The result is that 15

Lucian, The Hall, 4. Ibid., 20, citing Herodotos, 1.8.3. 17 On Lucian’s personae, see Suzanne Saïd, ‘Le “je” de Lucien’, in Marie-Françoise Baslez et al. (eds), L’Invention de l’autobiographie: d’Hésiode à Saint Augustin (Paris, 1993), pp. 253–70; Sandrine Dubel, ‘Dialogue et autoportrait: les masques de Lucien’, in Alain Billault (ed.), Lucien de Samosate (Lyon, 1994), pp. 19–26. 16

174


both arguments merge to modify each other (the first speaker claims that the opposing logos has been interrupting him throughout). But significant tensions remain. Indeed, the first speech is itself acutely ambiguous in its enthusiasm for the sight. It opens by comparing the speaker’s situation with that of Alexander who is said to have been so seduced by the sight of the river Kydnos that he would have still plunged into it even if he had known in advance that it would lead to his death, and goes on to compare the speaker to two animals, the horse (10) and the peacock (11), and the Hall itself to the notoriously deceptive and treacherous sea (12). If, in the end, the first speaker prevails it is not without doubts being raised concerning his confident claims of an unproblematic union between speaker, audience and context. Monuments and Meanings The Hall thus raises an important question. Ekphraseis of places in epideictic are frequently prefaced with disclaimers announcing the speaker’s inadequacy to the task.18 The ritual nature of this topos makes it no less pertinent as an observation of the impossibility of rendering a full account of any sight in words. However, if a full verbal account of any sight is an impossible task, an ekphrasis can still suggest imaginary supplements that may interact with the audience’s perception or memory of the subject. Examples are to be found in Aristeides’ Smyrnaean Oration (Or. 17) where the relatively brief periegesis of the city is interspersed with similes: the city is compared to an embroidered robe (chitōn) and then to a necklace made up of separate elements each of which attracts the viewer’s gaze. These supplements to the ekphrasis serve to introduce other values and associations into the description of the city: both have connotations of craftsmanship, wealth and an ordered opulence which enrich the ekphrasis and both appeal to the analogy of the city with the human body that we have already encountered in the discussion of the monody. In other cases, ekphrasis can also be used to make explicit the spiritual and political meanings that may have been implicit in their subjects, as I have suggested elsewhere.19 In particular, several ekphraseis of buildings (whether originally pronounced within the building, for example Paul the Silentiary’s ekphrasis of Hagia Sophia, or not, for example Prokopios of Caesarea’s account of the same church in his Buildings) effectively organize the description around the manner (tropos) in which the monument was 18

See, for example, Ailios Aristeides, On Rome, 6–13. Ruth Webb, ‘The aesthetics of sacred space: narrative, metaphor and motion in ekphraseis of church buildings’, DOP, 53 (1999): 59–74 and ‘Ekphrasis, �� amplification and persuasion in Procopius’ Buildings’, Antiquité tardive, 8 (2000): 67–71�. 19


175

constructed. Thus, Prokopios�� credits Justinian with the discovery of the solution to the architectural problem posed by the dome of Hagia Sophia, attributing the emperor’s inspiration to a divine source at 1.1.71 with the result that the monument becomes a sign of the emperor’s intelligence and of divine favour. This type of ekphrasis serves the same purpose as do the ekphraseis of military victories that are to be included as proof of the emperor’s qualities in the Basilikos logos.20 In this way, the monuments themselves can be presented as a type of proof of the qualities ascribed to the builder which was clearly visible to all. The anthropologist Alfred Gell has underlined the ways in which artefacts of all types can serve as ‘objective embodiments’ of power. They thus work an effect on the viewer who is prompted to speculate on the origins of the object and on the processes that produced it and to see it as an embodiment of its creator’s agency.21 Those ekphraseis of buildings that describe the construction process would seem to reflect this type of response, making explicit the sources of agency that are to be seen as embodied within the building. However, the orator who presented a monument in this way was not simply using the audience’s existing perception of the building to affect their perception of the emperor, but might also be attempting to alter their understanding of the building itself by attaching narratives to it and thereby suggesting new associations that might influence the listener’s future responses to it. So speakers like Prokopios reflect a habit of reading artefacts and also, I would suggest, attempt to exploit this habit by attempting to associate one particular origin narrative with the monument so that the latter comes to be understood by the listener as a sign of the patron’s action. Declamation If the problems posed by epideictic ekphrasis derive from the close proximity of the speech as performance and its subject matter, declamation raised the opposite problem. As ps.-Dionysios made clear, the subjects of declamatory ekphrasis were largely imaginary events that were supposed to have taken place in the distant past. Like the actor, the declaimer took on a role and created an imaginary world around him through words. If the boundary between orator and actor was always in need of restatement and reinforcement, so much the more so the boundary between acting 20 See Webb, ‘Ekphrasis, amplification and persuasion’, pp. 67–71 and Jaś Elsner, ‘The rhetoric of buildings in the De Aedificiis of Procopius’, in Liz James (ed.), Art and Text in Byzantine Culture (Cambridge, 2007), pp. 33–57. 21 �� Alfred Gell, Art and Agency: An Anthropological Theory (Oxford, 1998)�� , pp. 20–21 and 68–72.

176


and declamation which Quintilian had likened to ‘stage display or the ravings of a madman’ if it was practised for its own sake.22 The analogy with the actor is also relevant to the relationship of the declaimer to his adopted persona. Like the actor, he occupied an intermediate state between being an orator engaged in declamation and ‘being’ Demosthenes, Perikles or Themistokles, roles that he adopted temporarily for the purposes of his performance.23 Like the audience of a play (or the readers of a novel), the listeners were simultaneously invited to enter into the world created by the declaimer and able to retain an awareness that they were indeed listening to an orator pretending.24 Declamation need not, in this interpretation, be seen as a sign of nostalgia for the past nor need we assume that the speaker’s identity was entirely supplanted by that of his adopted persona, any more than the actor is consumed by his or her role. However, the role of the audience in the make-believe of declamation was different from that of the theatre audience. Michel Patillon’s analysis of the complex enunciative situation created by the exercise of ēthopoiia opens up the dynamics of impersonation in declamation: the speaker’s dual identity, he points out, created a dual audience, the reader/audience and the notional audience to whom the character addressed him or herself.25 In the same way, when a declaimer spoke in the persona of a character from the Greek past he cast the audience of his speech in a dual role. They were addressed as if they were classical Athenians listening to their imaginary Demosthenes or Themistokles, a role into which they may have actively entered in imagination, yet they clearly retained an acute awareness that they were listening to a contemporary sophist whose performance they were called upon to judge. In such a context, the subjects of ekphrasis were doubly fictive. The student or performer who described the storm at sea was speaking as if he were a general in classical Athens who was, in turn, attempting through 22 Quintilian, 2.10.8: ‘nam si foro non praeparat, aut scaenicae ostentationi aut furiosae vociferationi simillimum est.’ On orators and actors, see Elaine R. Fantham, ‘Orator and / et actor’, in Patricia E. Easterling and Edith Hall (eds), Greek and Roman Actors: Aspects of an Ancient Profession (Cambridge, 2002), pp. 362–76 and Ismene Lada-Richards, Silent Eloquence: Lucian and Pantomime Dancing (London, 2007), pp. 135–51. 23 See, for example, Richard Schechner, Performance Theory (New York, 1988), p. 179. Aelius Aristeides, Or. 28 (‘On a remark in passing’) uses the verb hupokrinō of his impersonation of figures from the past. 24 For further discussion, see Ruth Webb, ‘Fiction, mimesis and the performance of the Greek past in the Second Sophistic’, in David Konstan and Suzanne Saïd (eds), Greeks on Greekness: Viewing the Greek Past under the Roman Empire (Cambridge, 2006), pp. 27–46. 25 Michel Patillon, La Théorie du discours chez Hermogène le Rhéteur: essai sur les structures linguistiques de la rhétorique ancienne (Paris, 1988), pp. 302–3.


177

his ekphrasis to make his audience feel as if they shared his experience. By responding to the ekphrasis the listeners were in a sense placing themselves imaginatively in the situation of the purported classical addressees of the speech. In this case they were not just imagining that they were caught up in the storm at sea with the classical general but imagining that they were classical Athenian citizens imagining the scene. Moreover, their involvement in the scene could coexist with the critical awareness that it was a fiction and, specifically, that it was a fiction based on particular historical sources. A striking example from one of the few surviving declamations from the Second Sophistic is to be found in the sophist Polemo’s speech in the persona of the father of Kunaigeiros, the hero of Marathon who held onto a Persian ship even as his hands were being cut off by the enemy. In support of his argument that his son’s death was more glorious than that of another hero, Kallimachos (whose body was shot with so many arrows that it remained standing after his death), the father gives an ekphrasis full of bloody detail, describing how first one hand, then the other, was cut off and remained holding onto the ship as he died on land. Polemo fully exploits the potential for paradox and startling imagery, comparing the soldier’s mutilated body to a victory monument and stating how, as Kunaigeiros remained on land, his hand continued to grasp the ship so that in the end ‘he lay dead, a single man who served on both elements with his limbs, divided between land and sea’.26 This paradox cannot fail to draw attention to the role of the Roman sophist and is thus in tension with the invitation to visualize the hero’s death and to enter into the classical moment evoked. Such a duality of response is typical of declamation, and of all fiction. In the present case, awareness of the long-standing historical tradition surrounding Marathon would have added a further layer to the listener’s response; indeed, this tradition took the place, in the context of declamation, of the common cultural references to which the real orator in a real case could appeal. Ekphrasis in declamation exploited a set of shared images and associations that were familiar from the schools, from texts and also from images (as in the case of Marathon which was depicted in art). It is thus a phenomenon in which intertextuality plays a significant role, but intertextuality of a particular type. As we saw in the first chapter, the type of historical sources which provided the basic material for declamation scenarios were not considered simply as words on a page but as provokers of images in themselves.

26

Polemo, Declamation 1, 10–11 in The Severed Hand and the Upright Corpse: The Declamations of Marcus Antonius Polemo, ed. William W. Reader (Atlanta, 1996), pp. 104–5: καὶ vεκρὸς εἶς ἄμφω τὰ στoιχεῖα πληρώσας ἑαυτoῦ τoῖς μέλεσιv ἔκειτo, γῇ καὶ θαλάσσῃ μεμερισμέvoς.

178


Ekphrasis and the Greek Novel This consideration of declamation has brought us to the subject of fiction, and some of the same questions apply to the use of ekphrasis in the prose fiction of antiquity. The use of ekphrasis in the novel has been well explored, particularly when it comes to those ekphraseis that describe works of art. Shadi Bartsch has offered sophisticated reinterpretations of these passages in Heliodoros and Achilles Tatios, pointing out the interpretative work and play involved in the ekphraseis of works of art and of other subjects in these novels.27 More recently, Helen Morales has pointed out the shifting interactions between the fictional world of Achilles Tatios’ novel and the various types of visual representation in images and in dreams, in the course of a stimulating exploration of the theme of vision.28 My purpose here is to underline some passages that I believe make deliberate reference to the theory of ekphrasis as we find it in the rhetorical handbooks and to discuss some passages that are not usually considered in surveys of ekphrasis in the novel, but which the handbooks encourage us to include in this category. That ekphrasis is central to the novel is clear from its prominent role at the beginning of the works by Achilles Tatios, Longos and Heliodoros, whether in the form of ekphraseis of works of art, as in the case of the first two, or of an enigmatic scene in the case of Heliodoros. Both Achilles Tatios’ Kleitophon and Leukippe and Longos’ Daphnis and Chloe begin with a first-person account of a painting in its setting which serves to set the narrative in motion. Achilles Tatios’ lushly detailed ekphrasis of the painting of the rape of Europa serves, among other things, to prompt the narration by the hero, Kleitophon, that forms the body of the novel, while Longos’ minimalist evocation of a painting and the grove in which it stood provides the pretext for his novel, which is presented as a verbal response to the visual artefact.29 As the first of a series of descriptions of art works, Achilles’ opening ekphrasis serves many other functions, 27 Shadi Bartsch, Decoding the Ancient Novel: The Reader and the Role of Description in Heliodorus and Achilles Tatius (Princeton, 1989). 28 Helen Morales, Vision and Narrative in Achilles Tatius’ Leucippe and Clitophon (Cambridge, 2005). On ekphraseis of persons (or their absence) in the novels, see also Sandrine Dubel, ‘La beauté romanesque ou le refus du portrait dans le roman grec d’époque impériale’, in Bernard Pouderon (ed.), Les Personnages du roman grec (Lyon, 2001), pp. 29–58; Tim Whitmarsh, ‘Written on the body: ekphrasis, perception and deception in Heliodorus’ Aethiopica’, �� in Jaś Elsner (ed.), The Verbal and the Visual: Cultures of Ekphrasis in Antiquity = Ramus, 31, 1–2 (2002): 111–25 �� and Koen de Temmerman, ‘Blushing beauty: characterizing blushes in Chariton’s Callirhoe’, Mnemosyne, 60 (2007): 235–52. 29 It is debatable whether Longos’ brief description of the painting counts as an ekphrasis in itself.


179

tempting the reader to speculate on the painting’s relation to the events of the plot that follows and constantly requiring the reader to revise his or her interpretation as the events unfold. Most importantly, as Danielle Maeder has argued, these opening ekphraseis serve the twin purpose of drawing the reader imaginatively into the universe of the novel while simultaneously proclaiming the artificiality of the work.30 They thus explore the duality that is at the heart of fiction and underline the tensions involved in ekphrasis itself as a discourse that embodies likeness. In this way, the project of ekphrasis and enargeia – to make the reader/listener feel as if he or she is in the presence of persons, places, events and moments that are absent and may never have occurred at all – is central to the novel as genre. I would also add a further tension in these descriptions of works of art, between the temptation to immerse oneself imaginatively in the world evoked by the painting and the awareness of the interpretative need to find a link between the painting and the novel. Though by no means incompatible, these are two different attitudes to the text. In the former, the role of ekphrasis in creating an illusion of perception is predominant and the reader or listener is immersed in the fictional world, while, in the latter, the reader is engaged as critic, deciphering details, attempting to predict how the painting will relate to the narrative that follows. Far from seeing this latter response as typical of ekphrasis, I see it as a special problem posed by particular examples of ekphraseis of works of art in particular contexts. The type of interpretation required by Achilles Tatios’ paintings, or other descriptions of allegorical images, is very different from the type of instantaneous imaginative supplementation that we saw at work in the case of Quintilian’s reading of Cicero. The treatments of ekphrasis and enargeia in the rhetorical sources, as well as the responses to vivid language discussed at the very beginning of this study, encourage us to assume that imaginative engagement was valued by the ancient authors and readers of the novels. Achilles Tatios’ novel is rich in ekphraseis of all types: paintings, statues, animals (4.2, 4, 19), a garden (1.15), the city of Alexandria (5.1) and events such as the hyperbolic ekphrasis of the storm at sea (3.1–5). One example that has not attracted attention, most probably because it does not conform to modern expectations of ‘description’, is the ekphrasis of the death of Charikles, a minor character who dies in a riding accident at the beginning 30

Danielle Maeder, ‘Au seuil des Romans grecs: effets de réel et effets de création’, in Groningen Colloquia on the Novel 4 (Groningen, 1991), pp. 1–33. See also Jean-Philippe Guez, ‘Achille Tatius ou le paysage-monde’, in B. Pouderon and D. Crismani (eds), Lieux, décors et paysages de l’ancien roman des origines à Byzance (Lyon, 2005), pp. 299–307 who suggests that the painting contains in nuce the main themes of the novel.

180


of the novel (1.12). The event is presented in a type of messenger speech, which suggests a generic allusion to tragedy, and Françoise Létoublon has connected this incident with the (false) account of the death of Orestes in a chariot race in Sophokles’ Elektra.31 A further tragic intertext is the death of Hippolytos in Euripides’ play (ll. 1213–48), which in fact presents more similarities with Achilles Tatios’ episode: in both the accident results from the horse being startled, in both the setting is the wilds rather than the organized time and space of a race. The novelist’s extended metaphor of the bucking horse as a rough sea also helps to associate the scene with the death of Hippolytos on the shore. The thematic connection is also clear: all that the reader knows about Charikles is that he prefers homoerotic love and this rejection of women is comparable to Hippolytos’ rejection of all erotic relationships. It is particularly significant that both passages are ekphrastic, as is usually the case with messenger speeches, for this means that the visual imagination plays a part in the recognition of the allusion: the image evoked by Achilles Tatios corresponds to the image evoked in the past and stored in the memory by the text of Euripides’ play. There is thus a form of imaginative ‘intervisuality’ in play in which the secondary images of water (evoked by Achilles Tatios’ metaphor) play almost as important a role as the images of the principal event. Other examples of ekphrasis in Kleitophon and Leukippe conform more closely to modern assumptions about description – that is, they evoke static subjects and constitute a pause in the development of the action. Of this group, the ekphraseis of Egyptian animals described in Book 4 are particularly interesting in that they are ‘textbook’ examples of ekphrasis, corresponding to Theon’s use of Herodotos’ animal descriptions as examples. I do not believe that this is an accident: Achilles Tatios is calling attention to the artifice of his verbal account by peopling his ‘Egypt’ with the stock attributes of classical Greek visions of the exotic land. It is interesting to compare Ailios Aristeides’ remarks in his Egyptian Oration, which is concerned with truth and lies about Egypt and the shortcomings of the accounts by Herodotos and Homer. Criticizing one classical travel writer in particular, Aristeides singles out crocodiles and hippopotami as typical elements that a writer might add to his account to make his fiction (plasma) seem like truth.32 In this case, therefore, Achilles Tatios seems to refer explicitly to the textbook accounts of ekphrasis in order to draw attention to the artifice of his verbal creation, to the fact that this is a plasma and that his Egypt is a world of words.

31 Françoise Létoublon, Les Lieux communs du roman: stéréotypes grecs d’aventure et d’amour (Leiden, 1993), p. 100. 32 Ailios Aristeides, Egyptian Discourse, 96.


181

The uses of ekphrasis in Heliodoros’ Aithiopika are often harder to isolate, though this novel, too, is full of vivid evocations. The difference in approach is well illustrated by the contrasting descriptions of crocodiles for, whereas Achilles Tatios gives us a breakdown of the appearance of the beast in motion and at rest (4.19), Heliodoros describes in a few words the impression of its movement as it crosses the path of Knemon and Kalasiris. This latter passage is no less an ekphrasis, as the reader shares the characters’s partial view of the creature and understands Knemon’s alarm. Between them the two descriptions illustrate the contrast between the scholastic breakdown (kata lepton) and the vivid evocation.33 The dramatic tableau with which the Aithiopika opens is also an accomplished example of ekphrasis that slowly reveals to the reader the perplexing scene of devastation on the Egyptian shore through the eyes of a group of uncomprehending bandits. Jack Winkler has pointed out the aporetic nature of this opening which leaves the reader in ignorance of the events that led to this scene of devastation until the middle of the novel.34 Until then, we share the bandits’ ignorance. The opening scene of the Aithiopika, with its bodies, its overturned tables and signs of a feast violently interrupted, is directly comparable to the ekphraseis of aftermaths that we have encountered in the study of rhetorical ekphrasis. Seen from this perspective, the paradoxical nature of Heliodoros’ approach is all the more evident for he choses a scene where (in rhetorical terms) the act (pragma) is clear but the perpetrator (prosōpon) and his motivation (aitia) are not. The answer to what resembles a declaimer’s conundrum, when it is revealed several books later, turns out to be a surprise – a fact that says a great deal about the limitations of conjecture and of the very special nature of the ‘probable’ in this novel. It is also worth noting that the account of the events that led up to the mysterious scene of devastation with which the novel opens is just as much an ekphrasis as the opening scene, but this time of the event itself, rather than its aftermath. Analysed in rhetorical terms, this section finally allows the viewer to see the action (pragma), the manner in which the action took place (tropos) and the reason (aitia). Significantly, too, it is told from a standpoint of full knowledge by a witness who had a clear

33

�� See, on this contrast, John J. Winkler, ‘The mendacity of Kalasiris and the narrative strategy of Heliodorus’ Aithiopika’, Yale Classical Studies, 27 (1982)�� : 101–2, though Winkler does not recognize Heliodoros’ passage as an ekphrasis. 34 �� See, on this scene, Winkler, ‘The mendacity of Kalasiris’ and John �� R. Morgan, ‘Reader and audiences in the Aithiopika of Heliodoros’, Groningen Colloquia on the Novel 4 (Groningen, 1991), pp. 85–104.

182


view, rather than through the partial and uncomprehending eyes of the bandits through whom the opening scene is focalized.35 A further example of the type of connections between passages that become apparent when we apply the ancient criteria of ekphrasis involves the account of the siege of Syene in Heliodoros’ Aithiopika, 9.3. The method used by the Ethiopian king, Hydaspes, to surround Syene (Aswan) with water is very similar to Julian’s account of the historical siege of Nisibis by the Persians in 350 CE and this similarity provides the main evidence for a fourth-century date for the novel.36 The parallel with epideictic also underlines the function of the account as a portrait of the Ethiopian king (who at this point is unaware that he is the father of the heroine, Charikleia). The tropos of the�� works �� reveals the character and manners (tropoi) of their originator. Heliodoros stresses the measure and order involved in Hydaspes’ works as he divides up (katanemei) the circumference of the city walls and allots (apoklērōsas) each section of ten to a group of ten men. These qualities emerge more clearly still in comparison with Julian’s account of the Persian forces as a confused mass containing women, old men, children and servants. Julian’s far vaguer references to the body of water created around Nisibis also contrast with Heliodoros’ meticulous account of the shape of the channel dug around Syene, which further contributes to the portrayal of Hydaspes.37 Moreover, the extended ekphrasis of the siege works forms a pendant to the ekphrasis with which the novel begins, both of these scenes, dominated as they are by water, thus underline the river’s function as an organizing principle for the action of the novel and the linear movement of its hero and heroine from the Egyptian shore to the land of Ethiopia which is their final destination.38 The Novel and the Limits of Ekphrasis The way in which the opening scene of the Aithiopika exploits the conventions of rhetorical ekphrasis, requiring the reader to search for supplements which he or she is unable to find, invites reflection on the act of reading an ekphrasis of this type and on the normally automatic processes of decoding involved. The rhetorically trained reader would 35 Similarly, the ekphrasis of Leukippe’s fake death at Kleitophon and Leukippe, 3.15 is later supplemented by the ekphrasis of the tropos when Satyros explains how it was done at 3.20–21. 36 �� See Pierre Chuvin, Chronique des derniers païens (Paris, 2004), pp. 321–5 and Glen Bowersock, Fiction as History: Nero to Julian (Berkeley, 1994), pp. 149–60. 37 Julian, Panegyric of Constantius, I, 22, ll. 1–5; 11–12 and 25. 38 On the structuring role of the Nile in the novel, see Tim Whitmarsh, ‘The writes of passage: cultural initiation in Heliodorus’ Aethiopica’, in Richard Miles (ed.), Constructing Identities in Late Antiquity (London, 1999).


183

thus have been acutely aware of the aporetic nature of this scene, identified by Jack Winkler. This is not the only example where the reader seems to be invited not just to imagine the scene evoked by the novelist but also to reflect upon the practice and conventions of ekphrasis and enargeia. It is possible to identify other uses of ekphrasis by both Achilles Tatios and Heliodoros that play on the conventions of ekphrasis and, sometimes, call into question the rhetoricians’ claims for ekphrasis in different ways. Some of the most interesting instances of ekphrasis in the novel are cases in which the ekphrasis is addressed by one character to another, allowing the impact on the internal audience to be depicted. The account of the death of Charikles mentioned above is one example. In this case, the response is by the book: the listeners react emotionally to tragic news, presented in suitably tragic fashion. In other cases, however, the reactions are subject to more developed treatment that ultimately raises questions about the nature of the mental images provoked by the word and their relationship to the reality that is the fictional world of the novel. In Achilles Tatios’ novel, two separate male characters fall in love with the heroine, Leukippe, solely on the basis of reports of her beauty, without having seen her for themselves. The young Kallisthenes at the beginning of the novel and the older Thersandros, the main rival of the hero Kleitophon at the end, fall in love after merely hearing about Leukippe’s beauty. The idea is a topos of the novel but, in Achilles Tatios, it is developed in ways that ask interesting questions about the representational power of the word. Kallisthenes is described as ‘in love at first hearing’ (ex akoēs erastēs), a state which the narrator condemns as the sign of an unstable and uncontrollable nature (2.13.1). He is said to torment himself by picturing (anaplattōn) Leukippe’s beauty and imagining (phantazomenos) what he has never seen. His state of mind is thus a particularly acute case of imaginative supplementation, an exaggerated depiction of the state of any addressee of an ekphrasis who is prompted by words to create a mental image of the subject. Thersandros is similarly portrayed as susceptible to imaginary beauties. His passion for Leukippe is kindled when his servant reports her appearance, ‘singing the praises (katatragōdountos) of her beauty’, with the result that Thersander is ‘filled with an apparition (phantasma) as if of beauty’ (6.4.4). As we have seen, the term phantasma could be used to distinguish mental images that did not derive from sense perception from those that did: it is the term used by Augustine to qualify the image of Alexandria that he built up from his knowledge of Carthage and by the Stoics to characterize the mental image based on illusion (like Orestes’ vision of the Furies). So, here, the choice of the term to designate a word-induced mental image does seem to be a deliberate reference to Stoic theories.

184


Here, Stoic cognitive theory does seem to be acutely relevant as the reader is invited to reflect on his or her own imaginative response to the novel and to consider this, too, as an example of phantasma, in that it, like the phantasms of Leukippe, is evoked purely by words rather than resulting from direct perception. However, these two characters are very different from the reader, however immersed in the fictional situation, in that they not only allow themselves to respond emotionally as if to a reality but even act upon their desires. As such, they are fictional paradigms of the opposite to the Stoic sage: characters who act on an illusion. In Kallisthenes’ case, the folly of this decision is demonstrated in comical fashion when he kidnaps the wrong woman, illustrating the failure of his mental image to correspond to reality. Later on, Thersandros’ failed attempts to rape Leukippe reveal his moral and cognitive failings. Both are thus contrasted with Kleitophon who, though far from a perfect hero, at least falls in love as a result of seeing Leukippe, a process which is emphasized throughout the novel in the many discussions of vision and the eye. Heliodoros’ novel contains a similar invitation to the reader to reflect upon the nature of his or her imaginative involvement in the story. In a passage that has been much discussed, Heliodoros’ wily internal narrator, the Egyptian sage Kalasiris, who is responsible for much of the narration of first half of the novel, provides his listener, Knemon, with an ekphrasis of the occasion on which the hero and heroine, Theagenes and Charikleia, first met and fell in love in Delphi. The festival and its procession are described at length, at the request of Knemon who demands to be made ‘a spectator’ (theatēs) of the event by Kalasiris’ speech (logos) (3.1). This is not the only echo of the technical vocabulary of ekphrasis for, a little further on, as Kalasiris gives a detailed account of the costume worn by Charikleia, Knemon interrupts exclaiming, ‘It’s them! It’s Charikleia and Theagenes!’ A few lines further on, he attributes his response explicitly to the enargeia of Kalasiris’ account (diēgēsis): ‘I thought I could see them even though they are not here, so vividly and exactly as I have seen them myself did your account show them’.39 This naive exclamation contributes to the characterization of Knemon as a listener avid for spectacle; he represents a type of sensuous involvement in the narrative that would probably not have seemed as incongruous to the ancient reader as it did to Winkler and which is not necessarily incompatible with the type of hermeneutic reading he brilliantly describes.40 In fact, the wily Kalasiris is himself caught in the net of his own ekphrastic skills as he responds to Knemon’s first exclamation by begging to be told where the couple 39 Heliodoros, Aithiopika, 3.4.7: θεωρεῖv αὐτoὺς καὶ ἀπόvτας ᾠήθηv oὕτως ἐvαργῶς τε καὶ oὓς oἶδα ἰδὼv ἡ παρὰ σoῦ διήγησις ὑπέδειξεv. 40 Winkler, ‘The mendacity of Kalasiris’, p. 143.


185

are: ‘Where on earth are they? Show me, by the gods’.41 In the context of the story, Knemon’s response, and Kalasiris’ reaction, are at least partly explained by the fact that Knemon knows the couple, and can thus judge his imaginings against his own memory image of them, and that Kalasiris knows that Knemon expects them to arrive at any moment. But Hardie aptly notes the ‘doubt as to the plenitude of vision’ that creeps into the exchange as Kalasiris tells Knemon that he has probably not seen them as they were on the day he was describing.42 The exchange between Kalasiris and Knemon thus plays with the conventions of ekphrasis and enargeia, pointing out the limitations of the power of the word to ‘make present’ and to ‘place before the eyes’. Knemon is comparable to the phantasy�� lovers�� of Kleitophon and Leukippe in his real response to description. But a crucial difference between the episodes lies in the respective role of the reader. Whereas Achilles Tatios’ reader does not hear the evocations of Leukippe that have such a dramatic effect on their audiences, Heliodoros’ reader shares every detail of the sumptuous ekphrasis of events at Delphi with Knemon, the internal audience. At the moment of the highest imaginative involvement, the reader is brought up short by a reminder of the limitations of evocation, even for the listener with direct personal knowledge of the subject. The irony is that this reminder is issued by the reference, in the exchange between Kalasiris and Knemon, to the literal sense of ‘placing before the eyes’, from which the verbal evocation falls far short. The passage thus gives the readers cause for reflection on their own sensuous involvement and its limitations, not only through the distanced depiction of characters’ responses, but through the sharing of those responses.43 Descriptions of Works of Art as Meta-ekphrasis Ekphraseis of all types of subjects, and not only those that present works of art, may therefore have a meta-fictional function in the novel, causing the reader not only to reflect upon the nature of his or her experience of fiction but also, through the dialectic of engagement and distance set in place in the episodes analysed above, making him or her experience in various ways the disjunction between the fictional world and reality. In a similar 41

Heliodoros, Aithiopika, 3.4.7: καὶ πoῦ γῆς oὗτoι; δείκvυε πρὸς θεῶv. Philip Hardie, ‘A reading of Heliodorus, Aithiopika, 3.4.1–5.2’, in Richard Hunter (ed.), Studies in Heliodorus (Cambridge, 1998), p. 31. 43 Shadi Bartsch, ‘“Wait a moment, phantasia”: ekphrastic interference in Seneca and Epictetus’, CP, 102 (2007): 83–5 analyses some similar instances where Stoic philosophers make use of enargeia to invite the reader to participate imaginatively in a scene only to disrupt that involvement in order to provoke reflection. 42

186


way, I would suggest that certain examples of ekphrasis of paintings and sculptures perform their own commentary on the nature of ekphrasis in the broad sense of the word. If all ekphrasis, of whatever subject, is like a painting or sculpture in its aim to ‘place before the eyes’, an ekphrasis of visual representation is doubly ekphrastic. This is not the place for a full investigation of this phenomenon which has been much discussed, but I would simply like to stress here that I see these ekphraseis of works of art as a special category of ekphrasis, and to discuss briefly two examples that, in very different ways, can be read as commentaries on the practice of rhetorical ekphrasis that is at the centre of this book. The first of these examples is the collection of ekphraseis of statues attributed to Nikolaos, the second is the Eikones of the Elder Philostratos. They are both works that are firmly rooted in the rhetorical culture of the Imperial period but they show very different approaches to the project of describing. Statues and Signs As noted above (Chapters 3 and 6), Nikolaos’ method of describing statues is ‘by the book’. After a brief introduction, which often notes the place of the sculpture in a chain of artistic transmission that has led from the event to its literary or dramatic representation and then to its representation in bronze, the author proceeds to break down the figure into its parts and to propose a meaning for each expression, posture or ges�� t�� ure. The technique seems to show what Nikolaos meant by the reasons (logismoi) why the artist represented the figure as he did, and, indeed, the artist’s intention figures prominently in these ekphraseis. In the ekphrasis of the Trojan woman after the fall of Troy, for example, we are told that the fact that she is shown sitting on a chair shows (sēmainō) that she is unable to stand when her city had fallen, that her bare head indicates her loss of shame but the fact that her hair is held back in a band shows that this loss is not total.44 The laborious cataloguing of parts illustrates the technique of ekphrasis kata lepton, which I suggested above was a result of the school tradition of teaching ekphrasis. The result is hard to define as ‘vivid’, but it does seem that these curious ekphraseis serve to lay bare and to demonstrate to readers one of the important aspects of ekphrasis in a rhetorical context. We saw in the previous chapter how the significance of a scene such as the destruction at Phokis or the aftermath of a Roman party could lie in the implicit narrative of the events that preceded. Nikolaos’ statue ekphraseis demonstrate exactly that process, identifying each visible detail of the sculpture as the result of the prior action so that the text not only makes 44

[Nikolaos], in Libanios, Progymnasmata, p. 505. For further discussion, see Ruth Webb, ‘The model ekphraseis of Nikolaos the Sophist as memory images’, in Michael Grünbart (ed.), Theatron: Rhetorische Kultur in Spätantike un Mittelalter (Berlin, 2007), pp. 463–75.


187

explicit a certain way of reading images but also provides a commentary on rhetorical ekphrasis in general. Moreover, in their attention to details of appearance, these ekphraseis also pay attention to the way in which the description of external characteristics (as in Cicero’s portrait of Verres) could serve to portray inner feelings.45 Just as in that example, the knowledge appealed to is of a generic, cultural kind: Nikolaos’ statues are all of figures or events familiar from literature, so that the role ascribed to the artist is that of preserving and transmitting this tradition. Philostratos’ Eikones The Eikones of the Elder Philostratos are immeasurably more sophisticated than these statue ekphraseis (as are the statue ekphraseis of Kallistratos). They are also unmistakeably vivid, by any definition of the term, as every description is packed with concrete nouns, adjectives and verbs of movement, all denoting perceptible features. The description of Amphion and his lyre (1.10), to take just one example, provides details of the lyre’s construction: the dark horns with their jagged edges, the smoothness of the boxwood, the pattern of the tortoiseshell. As so often in the Eikones, the description evokes the particulars of the young man’s beauty, the way his hair with its golden highlights falls on his forehead and onto the down on his cheek, and the iridescent shimmering of his multicoloured cloak. But it is not just visual information that is evoked, as Alessandra Manieri has emphasized.46 Sounds, such as Amphion’s song, are mentioned, as is the fragrance of a garden (1.6.1) or of the smoke rising from an altar (2.27.3). Within the Sophist’s discourse, the ability to perceive the sounds and scents of the painting is an important aspect of the perceptual world of the gallery. In addition, there are many appeals to the senses of touch and of taste. The pictures are thus ‘placed before our eyes’, along with the synaesthetic response that they elicit from the viewer. As was first noted by Bougot, a large part of the interest of the Eikones lies in their depiction of intense imaginative response to the visual arts. Philostratos’ text is often seen as a prime example, if not the example, of the ekphrasis of works of art. However, its sophistication makes it a special use of ekphrasis that should be ranked alongside the novels for its conscious play with fiction – for Philostratos’ descriptions are presented as the direct quotation of discourses made in front of the paintings and addressed to an internal audience of young men and one boy. There are 45 See, in this connection, the discussion of diathesis, which can mean both external and internal ‘disposition’, in Roos Meijering, Literary and Rhetorical Theories in Greek Scholia (Groningen, 1987), p. 41, pp. 31–3. 46 Alessandra Manieri, ‘Colori, suoni e profumi nelle Imagines: principi dell’estetica filostratea’, Quaderni urbinati di cultura classica, n.s. 63/3 (1999): 111–21.

188


thus several levels of time involved: the time of the visit to the gallery; the later moment at which the whole is remembered and the speeches are repeated; the mythical time represented in the paintings and, occasionally, the moment at which the artist made his painted representation. The speaker is no more to be identified with the author, Philostratos, than is Kalasiris with his creator, Heliodoros. The audience is also doubled, rather as in ēthopoiia, or declamation, because the original descriptions are said to have been addressed to the boys in front of the paintings while the text we read is presented as a later report of that event for an external audience of readers or listeners who are not in front of the paintings. The ekphraseis are thus constantly dual: within the fictional setting of the gallery they interpret a painting, which is real and visible to the internal audience, part of their present surroundings, but for the external audience of readers the words alone create the paintings, the speaker and his audience. The paintings, as described by the Sophist, are so dazzling that we forget to question the Sophist’s own existence, or that of his internal audience, or to see that both the painting and his interpretation are creations of his words, as he in turn is a creation of the historical Philostratos. Rather as our sense of imaginative immersion in the procession at Delphi as described by Kalasiris in the Aithiopika is abruptly broken by Knemon’s literal response (or rather, Kalasiris’ literal interpretation of Knemon’s response to the enargeia of the performance), Philostratos too invites his reader to distance him or herself from the enthusiastic submission to illusion expressed by our Sophist-guide. At the moments of most heightened involvement on the part of the speaker, the reader is�� reminded of his or her own physical distance from the gallery and the paintings it contains, a distance that is otherwise collapsed by the use of direct discourse to quote the Sophist’s ‘lectures’. Most �� dramatically, in a passage that has been much commented on, the Sophist invites the boy to step forward and to catch the dripping blood of the dying Menoikeus in the fold of his garment (1.4.4).�� Here the reader is confronted with his or her distance from the scene and reminded that the scenes he or she ‘sees’ are in imagination. It is with the image of Narcissus, with its concentration on the themes of representation and illusion, that the impact is greatest. The whole Eikon can be read as a meditation on the theme of illusion in the visual arts, as the Sophist comments on a bee – which may be painted or real – that is deceived (exapatētheisa) by the painted flower on which it rests. Moving on to the figure of Narcissus himself, enrapt by his own image, he starts to address the young man directly, berating him for his absorption in the image in the pool. By now we are used to this enthusiasm, this desire to enter into the world of the painting; the Sophist has already invited the boy to stretch out the fold of his garment to catch the blood of the dying


189

Menoikeus. But the mise en abîme set up at the opening of Narcissus alerts us to the irony of the situation: a man speaking to a painted youth, telling him he is wasting his time on a mere image. In a very similar passage in the Hunters (1.28) the Sophist interrupts himself, observing that he has been deceived by the painting, but here he simply observes that Narcissus cannot hear him because he is so absorbed in the pool (and not because he is a fiction). The Sophist may be an unselfconscious viewer, but Philostratos, who has staged the whole scene within a scene, is drawing our attention to the seduction of illusion. With those words, our own involvement in the scene as readers is disrupted – if, that is, we note the irony of a grown man speaking out loud to address a painted boy and to tell him that his painted reflection cannot hear him. As the Sophist continues his address to his internal audience with a catalogue of details of Narcissus’ posture (and of the characteristics his hair does not have) we cannot help but be aware of the fictitiousness of the whole scene. It is surely responses like that of the Sophist in front of Menoikeus or Narcissus that the Younger Philostratos has in mind when he tells us in the proem to his own Eikones that there is no shame in the deception (apatē) inherent in art, and that there is no harm in ‘being in front of things which do not exist as if they existed and being led by�� them �� to believe that they exist’. But these remarks apply equally to the reader’s own involvement in the whole scene in the gallery, in the dynamic of communication between non-existent Sophist and non-existent boy. Philostratos has written not just ekphraseis of paintings, as his grandson said, but ekphraseis of the tropos, the ‘manner in which’ paintings are viewed. The definition of ekphrasis as an evocation of an event, which draws us in and makes us feel ‘as if we were there’ helps to open up this aspect of a deceptively complex text.47 Philostratos’ work occupies a special place in that it is an example both of ekphrasis in the ancient sense and in the modern sense: the foundation text, in a way, for the modern definition of ekphrasis, as we saw in the Introduction. It also has many points of contact with the pedagogical tradition of the rhetorical schools: many of the texts ‘illustrated’ in the paintings are precisely the ones that would have been studied in the earlier stages and Philostratos’ style in this text (his hermēneia) is recommended by Menander Rhetor as a model for certain types of speech.48 The 47

On the Eikones as fiction, see further: Duncan McCombie, ‘Philostratus, Histoi, Imagines, 2.28: ekphrasis and the web of illusion’, in �� Jaś Elsner (ed.), The Verbal and the Visual: Cultures of Ekphrasis in Antiquity =� Ramus, 31 (2002): 146–57; Ruth Webb, ‘The Imagines as fictional text: ekphrasis, apatê and illusion’, in M. Costantini et al. (eds), Le Défi de l’art: Philostrate, Callistrate et l’image sophistique (Rennes, 2006), pp. 113–36. 48 Menander Rhetor, Treatise II, 390, ll. 2–4.

190


conception of ekphrasis sketched out in the Progymnasmata can also alert us to Philostratos’ achievement in evoking not the static contents of a gallery that may or may not have existed, but in placing before our eyes a depiction of a process of literal viewing, doubled by our own metaphorical viewing of the scenes he ‘places before our eyes’. Philostratos’ Eikones transcend in their sophistication the world of the rhetorical handbooks and hint at the inadequacies of the discussions of ekphrasis there. By placing an image of an image before our eyes, and simultaneously disrupting that presence, Philostratos calls into question the power that is so unproblematically ascribed to the word in their definition of ekphrasis. In this, his work is comparable to the passages in the Greek novels that comment on the paradoxes involved in the ekphrastic project of ‘placing before they eyes’ and explore the nature of the mental images provoked and their relation to reality. It is not surprising if these questions are largely ignored by the elementary Progymnasmata, but they are implicit, as we saw at the beginning of this chapter, in some of the recommendations of Menander Rhetor and in the practice of ekphrasis in many different genres and contexts. Ekphrasis and the Invisible A further function of ekphrasis that is implicit in its rhetorical uses but not explicitly discussed in the theoretical works is a development of the idea of ‘making absent things present’. As we have seen above, ekphrasis can be used to make visible to the mind’s eye features of a subject that are invisible to the physical sense of sight such as a building’s history or its spiritual significance. Not surprisingly, such uses of ekphrasis are particularly frequent in Christian contexts, as Isabella Gualandri has pointed out in relation to Latin poetry, and often serve to bring out the spiritual qualities of a monument or work of art.49 There is no space in the present volume to explore this phenomenon in the detail it deserves, but one particular example deserves mention because of its close connection to the progymnasmatic tradition with which we began. This is John Chrysostom’s use of ekphrasis in his autobiographical work On the Priesthood in which he paints a picture (eikōn) of a battle, as observed by an innocent peasant boy, in order to convey to his reader his own state of mind.50 Here, the visible is used as a means of expressing the invisible. The 49 Isabella Gualandri, ‘Aspetti dell’ekphrasis in età tardo-antica’, in Testo e immagine nell’alto medioevo (Spoleto, 1994), pp. 328–41. See also Liz James, ‘Art and lies: text, image and imagination in the medieval world’, in Antony Eastmond and Liz James (eds), Icon and Word: The Power of Images in Byzantium (Aldershot, 2003), pp. 59–72. 50 John Chrysostom, On the Priesthood, 6.12.


191

technique is close to that used by Chorikios in his eleventh declamation to depict the cross-dressing general’s state of mind and also takes us back to the close link between imagination and emotion in Quintilian. In Chrysostom, however, the priorities are different: rather than the image of the event being used to provoke an emotive response to that event, here the battle is merely a means of conveying the imperceptible and almost ineffable: the speaker’s state of mind at a precise moment in the past.


Conclusion

The study of the ancient treatments of ekphrasis shows that it was understood to be a type of speech that worked an immediate impact on the mind of the listener, sparking mental images of the subjects it ‘placed before the eyes’. As an elementary exercise, ekphrasis gave practice in this type of speech, showing students how to expand the various elements of a narration to make sure their listeners not only knew what happened but felt as if they were witnesses themselves. The key to the nature and function of ekphrasis is its defining quality of enargeia, the vividness that makes absent things seems present by its appeal to the imagination. Though enargeia is used frequently in genres such as historiography and poetry, its effects can also be thoroughly rhetorical, helping the orator to involve his audience (and himself) emotionally and imaginatively in the subject of the speech and thus to promote their acceptance of the ideas he is putting forward. When integrated into a full-scale speech, ekphrasis served to involve the listener imaginatively and emotionally in the events at issue, making them share the speaker’s indignation at a crime or, in the more complex examples, altering their perception of a fact by placing them in the situation of an eyewitness and making them share that viewer’s experience. In this way, ekphrasis and enargeia underline the emotive and communicative aspects of rhetorical discourse and the way in which it involves the action of one mind upon another. In particular, these rhetorical uses of ekphrasis demanded an active engagement from the listener who was prompted by the speaker’s words to supply details, such as the actions and events that preceded the scenes of devastation that are a recurrent subject. In epideictic contexts, the effects of ekphrasis could be particularly subtle, involving an interaction between the audience’s knowledge of an actual sight and the verbally induced mental image. Works of rhetorical theory, such as the Progymnasmata or Quintilian’s Institutio oratoria, can help reveal the assumptions that underpinned the composition and reception of many ancient texts, assumptions that were often too deeply rooted to need full articulation. In many ways, these works are far from ‘theoretical’ as they are strongly oriented towards practice, so that the orator’s experience or received wisdom takes the place of explicit analysis and exposition of the underlying ideas. In the case of ekphrasis, one consequence of this lack of explicit theorization seems to have been that See, in particular, Chaïm Perelman, L’Empire rhétorique: rhétorique et argumentation (Paris, 1977), pp. 23–5.

194


the ideal end result of imaginative and emotional involvement could be lost from sight. The manuals seem to have encouraged a certain formalistic conception of ekphrasis as a detailed presentation that was not necessarily vivid. At the same time, the habit of proposing all the concrete elements of narration as subjects for ekphrasis encouraged the development of ekphraseis of subjects such as places and persons as independent entities rather than as elements of a narrative. The category of ekphrasis thus overlaps in some cases both the modern category of ‘description’ and the category of ‘descriptions of works of art’ without, however, being coextensive with either group. Some ekphraseis are of works of art; some are of objects presented as static interruptions to narration (in Achilles Tatios’ novel in particular and at the end of Heliodoros’ Aithiopika where our desire for closure is repeatedly frustrated by ekphrasis). Many of these have fruitfully been analysed as such using modern insights born of the study of the modern novel. But these are not the defining examples of ekphrasis; rather, I would suggest that they are often special cases that are used to reflect upon the conventions of ekphrasis and to point out, as in the case of Philostratos, the questions that lurk beneath the confident assertions of the handbooks. Ekphrasis and Art The category of ‘ekphraseis of works of art’ may not have been central to the definition of ekphrasis as it was conceived and discussed by the rhetoricians, but the analogy with the visual arts, not to mention the theatre, underlies the idea of ekphrasis and enargeia. As a type of language which makes the listener ‘see’, an ekphrasis of any subject aims theoretically to create an effect similar to that of a painting or sculpture. The relationship between word and image in ekphrasis is thus mimetic not in the sense of ‘producing a copy of’ but in the sense that ekphrasis ‘acts like’ a painting. The analogy goes far further in that the audience of an ekphrasis, like the viewer of a painting, can be required to supply information from his or her knowledge of the narrative background. The analogy with the visual arts also points towards the fictional nature of the products of ekphrasis and enargeia: like painting, they may have the power to create an illusion of presence, making the listeners feel as if they were in the presence of their subjects, but this feeling of presence is combined with an awareness of absence. An author like Philostratos makes use of the idea of painted representation to dramatize and play upon this combination of involvement and distance that is at the core of the reception of both ekphrasis and painting. Ekphraseis of sculptures raise similar questions, except that, in their case, the parallel tensions between the medium and the subject represented, between stillness and

Conclusion

195

movement, the frozen moment and its wider context, are much more emphatically present. Buildings form another prominent class of subject matter for ekphrasis that is shared by both the ancient and some, at least, of the modern definitions. Here, too, the visible details of the building can be used to point towards the invisible attributes of the building, whether the actions of the patron which the speaker wishes to associate with the building or its unseen, spiritual qualities. The Uses of Ekphrasis The ancient definition may seem frustratingly vague and elusive, but its interest lies precisely in this lack of formal precision. The identification of any passage as an ekphrasis is only a first step in its analysis. Each case needs to be studied in its context to reveal its particular qualities and its function, and what is true of one ekphrasis will not necessarily be true of another. Where the unity is to be found is in the underlying ideas and the wider conceptions of reading, or the individual’s relation to the word and of the interaction between language, memory and imagination. The study of ekphrasis and enargeia provides important information about ancient habits of reading and deeply rooted attitudes towards texts, which are seen as inviting imaginative and emotional involvement. These ancient modes of reading can be surprisingly different to our own: in the case of Thucydides’ history, ancient readers saw not a dispassionate and objective account of events but a window onto the violent and turbulent events of the past. In these rhetorically oriented readings, the text opens up to the reader’s imagination: the words on the page dissolve into images as they impact upon the mind. This does not mean that ekphraseis, or any other type of ancient text, can only be read according to these ancient frameworks, nor that modern insights into, say, description cannot be fruitfully applied to individual examples – far from it. But it does mean that we cannot assume that ancient readers and writers shared our modes of reading and our ways of categorizing texts. Sources like the humble Progymnasmata provide precious insights into the practices that shaped ancient and medieval writing and reading habits.

Simon Goldhill, ‘What is ekphrasis for?’, CP, 102 (2007): 5–6.


Appendix A Translations

I. Translations of the Progymnasmata Chapters on Ekphrasis Theon, Progymnasmata 118.6 – 120.11, ed. M. Patillon (Paris, 1997), pp. 66– 69 118.6 Ekphrasis is a descriptive (periēgēmatikos) speech which vividly (enargōs) brings the subject shown before the eyes. An ekphrasis may be of persons, and events and places and times. Of persons, as in the Homeric passage: ‘He was round-shouldered, dark-skinned, with curly hair’ and the passage on Thersites: ‘He had a pointed head, and was lame in one leg’ and so on, and in Herodotus the appearance of the ibis, of hippopotami and crocodiles in Egypt. Of events, such as the ekphraseis of war, peace, a storm, famine, plague, an earthquake. Of places, such as a meadow, seashores, cities, islands, deserted places and the like. Of times, such as spring, summer, a festival and things of this sort. There are also ekphraseis of the manner (tropos), such as those describing the manner in which pieces of equipment or weapons or siege engines were prepared, like the making of the arms [of Achilles] in Homer, or the blockade (periteichismos) of Plataia and the construction of siege engines in Thucydides: [beginning] ‘they cut a long beam in two and hollowed it out completely …’ And in Book 9 of Ktesias: ‘The Lydians, upon seeing at dawn from far away the forms (eidōla) of the Persians set up on long pieces of wood against the ramparts were put to flight, since they believed that the acropolis was full of Persians and had already been taken …’ All extracts from Theon, Progymnasmata, ed. M. Patillon (Paris, 1997) reprinted with permission © Les Belles Lettres. Homer, Odyssey, 19.246. Homer, Iliad, 2.219 and 217 (a composite). Herodotos, 2.76; 71; 68.

Homer, Iliad, 18.468–617. Thucydides, 2.75–8. Ibid., 4.100. Ktesias, Persika, F9b, ed. Dominique Lenfant (Paris, 2004), p. 114.

198


An ekphrasis may also be of a mixed type, such as the night battles in Thucydides and Philistos, for night is a time (kairos) while the battle is an event (praxis). This exercise has a certain affinity with the previous one [i.e. (koinos) topos]: they are alike in that they are both concerned with subjects which are not defined, but are common and general. They differ, however, firstly in that the topos is concerned with the results of deliberate intention (proairesis), while ekphrasis is mostly concerned with inanimate things which are not the result of deliberate intention; secondly in that in the common place when we relate (apangellō) the subject matter (pragmata) we also set forth our own opinion, declaring it to be either good or bad, while in ekphrasis the relating of the subject (pragmata) is simple.10 In ekphraseis of events (pragmata) we will deal with (epicheireō) what preceded, what happened during the event and the consequences; for example, in the case of a war we will first go through what occurred before the war, the enlisting of troops, the expenditure, people’s fears, the ravaged countryside, the sieges, then the wounds, deaths and mourning, and after all this the capture and enslavement of one side and the victory and trophies of the other. If we are composing ekphraseis of places or times or manners (tropoi) or persons, we will find starting points for our discourse in the narration (diēgēsis) about them, and from the beautiful, from the useful, and from the pleasant, just as Homer does with the arms of Achilles, when he says that they were fine and strong and that the sight of them amazed his allies and terrified the enemy.11 The virtues of ekphrasis are the following: above all clarity (saphēneia) and the vividness (enargeia) which makes one almost see what is being spoken about (ta apangellomena); then one should avoid speaking at great length about useless things (achrēsta); in general one should fit the language (apangelia) to the subject (ta hupokeimena), so that if the subject shown (to dēloumenon) is flowery, the style (phrasis) should also be flowery, but if it is harsh or frightening or anything else, the qualities of the language (hermēneia) should not be inappropriate to the nature of the subject. Some people think that we should practise ekphrasis by refuting and confirming the ekphraseis composed (literally ‘spoken’) by certain authors, saying, for example, that Herodotos was mistaken about the appearance of the ibis when he said they have white feathers except on their heads and neck and the end of their tail, because the tail is completely white. But I do not think that they are saying anything different from what I have Thucydides, 2.2–5 and 7.43–4. Theon varies his vocabulary here, using kairos instead of chronos and praxis for the related pragma. 10 ἐv δὲ τῇ ἐκφράσει ψιλὴ τῶv πραγμάτωv ἐστιv ἡ ἀπαγγελία. 11 See, for example, Homer, Iliad, 19.12–18; 22.131–35 and 312–21.

APPENDIX A

199

already said, as I think that this type of exercise falls into the category of refutations and confirmations of narratives (diēgēseis). Ἔκφρασίς ἐστι λόγος περιηγηματικὸς ἐναργῶς ὑπ’ ὄψιν ἄγων τὸ δηλούμενον. Γίνεται δὲ ἔκφρασις προσώπων τε καὶ πραγμάτων καὶ τόπων καὶ χρόνων. προσώπων μὲν οὖν, οἷον τὸ Ὁμηρικόν, γύρος ἔην ὤμοις, μελανόχροος, οὐλοκάρηνος. καὶ τὰ περὶ τοῦ Θερσίτου, φοξὸς ἔην κεφαλήν, χωλὸς δ’ ἕτερον πόδα, καὶ τὰ ἑξῆς. καὶ παρ’ Ἡροδότῳ τὸ εἶδος τῆς ἴβιδος καὶ τῶν ἵππων τῶν ποταμίων καὶ τῶν κροκοδείλων τῶν Αἰγυπτίων. πραγμάτων δὲ οἷον ἔκφρασις πολέμου, εἰρήνης, χειμῶνος, λιμοῦ, λοιμοῦ, σεισμοῦ. τόπων δὲ οἷον λειμώνων, αἰγιαλῶν, πόλεων, νήσων, ἐρημίας, καὶ τῶν ὁμοίων. χρόνων δε οἷον ἔαρος, θέρους, ἑορτῆς, καὶ τῶν τοιούτων. Αἱ δὲ καὶ τρόπων εἰσὶν ἐκφράσεις ὁποῖαι τῶν σκευῶν καὶ τῶν ὅπλων καὶ τῶν μηχανημάτων, ὃν τρόπον ἕκαστον παρεσκευάσθη, ὡς παρὰ μὲν Ὁμήρῳ ἡ ὁπλοποιία, παρὰ Θουκυδίδῃ δὲ ὁ περιτειχισμὸς τῶν Πλαταιέων καὶ ἡ τῶν μηχανημάτων κατασκευή, κεραίαν μεγάλην δίχα πρίσαντες ἐκοίλαναν ἅπασαν. ἐν δὲ τῇ ἐνάτῃ Κτησίου, οἷον. τὰ εἴδωλα τῶν Περσῶν ἐπὶ τοῖς μακροῖς ξύλοις ὁρῶντες ὑπὸ τὸν ὄρθρον πρὸς τὰς ἀκροπόλεις πόρρωθεν οἱ Λυδοὶ εἰς φυγὴν ἐτράποντο, νομίσαντες [p. 119] τὴν ἀκρόπολιν πλήρη εἶναι Περσῶν καὶ ἤδη ἑαλωκέναι. Γένοιτο δ’ ἄν τις καὶ μικτὴ ἔκφρασις, ὡς παρὰ Θουκυδίδῃ καὶ Φιλίστῳ νυκτομαχία· ἡ μὲν γὰρ νὺξ καιρός τις, ἡ δὲ μάχη πρᾶξις. Συγγένειαν δέ τινα ἔχει τὸ γύμνασμα τοῦτο τῷ προειρημένῳ· ᾗ μὲν γὰρ περὶ οὐδενὸς ὡρισμένου ἐστὶν ἀμφότερα, ἀλλὰ κοινὰ καὶ καθόλου, ταύτῃ ὅμοια. διαφέρει δὲ ἀλλήλων πρῶτον μέν, ὅτι ὁ μὲν τόπος περὶ τῶν ἐκ προαιρέσεώς ἐστιν, ἡ δὲ ἔκφρασις τὰ πολλὰ περὶ τῶν ἀψύχων καὶ ἀπροαιρέτων γίνεται, δεύτερον δὲ ὅτι ἐν μὲν τῷ τόπῳ τὰ πράγματα ἀπαγγέλλοντες προστίθεμεν καὶ τὴν ἡμετέραν γνώμην ἢ χρηστὰ ἢ φαῦλα λέγοντες εἶναι, ἐν δὲ τῇ ἐκφράσει ψιλὴ τῶν πραγμάτων ἐστὶν ἡ ἀπαγγελία. Ἐπιχειρήσομεν δὲ τὰ μὲν πράγματα ἐκφράζοντες ἔκ τε τῶν προγεγονότων, καὶ τῶν ἐν αὐτοῖς γινομένων, καὶ ἐκ τῶν συμβαινόντων τούτοις, οἷον ἐπὶ πολέμου διεξελευσόμεθα πρῶτον μὲν τὰ πρὸ τοῦ πολέμου, τὰς στρατολογίας, τὰ ἀναλώματα, τοὺς φόβους, τὴν χώραν δῃουμένην, τὰς πολιορκίας, ἔπειτα δὲ τὰ τραύματα καὶ τοὺς θανάτους καὶ τὰ πένθη, ἐφ’ ἅπασι δὲ τῶν μὲν τὴν ἅλωσιν καὶ τὴν δουλείαν, τῶν δὲ τὴν νίκην καὶ τὰ τρόπαια. ἐὰν δὲ τόπους ἢ χρόνους ἢ τρόπους ἢ πρόσωπα ἐκφράζωμεν, μετὰ τῆς περὶ αὐτῶν διηγήσεως ἀφορμὰς ἕξομεν λόγων καὶ ἐκ τοῦ καλοῦ καὶ ἐκ τοῦ χρησίμου καὶ ἐκ τοῦ ἡδέος, οἷον Ὅμηρος ἐπὶ τῶν Ἀχιλλέως ὅπλων ἐποίησεν, εἰπὼν ὅτι καὶ καλὰ ἦν καὶ ἰσχυρὰ καὶ ἰδεῖν τοῖς μὲν συμμάχοις ἐκπληκτικά, τοῖς δὲ πολεμίοις φοβερά. Ἀρεταὶ δὲ ἐκφράσεως αἵδε· σαφήνεια μὲν μάλιστα καὶ ἐνάργεια τοῦ σχεδὸν ὁρᾶσθαι τὰ ἀπαγγελλόμενα, ἔπειτα τὸ μὴ τελέως ἀπομηκύνειν περὶ τὰ ἄχρηστα· τὸ δὲ ὅλον συνεξομοιοῦσθαι χρὴ τοῖς ὑποκειμένοις τὴν ἀπαγγελίαν, ὥστε εἰ μὲν εὐανθές τι εἴη το δηλούμενον, εὐανθῆ καὶ τὴν φράσιν [p. 120] εἶναι, εἰ


200

δὲ αὐχμηρὸν ἢ φοβερὸν ἢ ὁποῖον δή ποτε, μηδὲ τὰ τῆς ἑρμηνείας ἀπᾷδειν τῆς φύσεως αὐτῶν. Ἔνιοι μέντοι γε ἀξιοῦσι γυμνάζεσθαι καὶ τὴν ἔκφρασιν ἀνασκευάζοντας καὶ κατασκευάζοντας τὰς ὑπό τινων εἰρημένας ἐκφράσεις, οἷον ὅτι Ἡρόδοτος διέψευσται περὶ τοῦ εἴδους τῆς ἴβεως λέγων ὅτι λευκόπτεροί εἰσι πλὴν κεφαλῆς τε καὶ αὐχένος καὶ τοῦ πυγαίου ἄκρου. τὸ γὰρ πυγαῖον ὅλον λευκόν ἐστιν. ἡμῖν δὲ οὐδὲν καινὸν δοκοῦσι λέγειν παρὰ τὰ προειρημένα, διὰ τὸ τοιοῦτον εἶδος ὑποπίπτειν νομίζειν ἐν ταῖς τῶν διηγημάτων ἀνασκευαῖς τε καὶ κατασκευαῖς. Ps.-Hermogenes, on ekphrasis from Progymnasmata, in Opera, ed. Hugo Rabe (Leipzig: Teubner, 1913), pp. 22–3 22 Ekphrasis is a descriptive speech, as they say, which is vivid and brings the subject shown before the eyes. There are ekphraseis of persons, events, states of affairs (kairoi), places, times (chronoi) and many other things; of persons, as in Homer: ‘he was bandy-legged and lame in one leg …’;12 of events such as an ekphrasis of a land battle or sea battle; of states of affairs such as peace, war; of places such as harbours, seashores, cities; of times such as spring, summer, festival time. There may also be mixed ekphraseis, like the night battle in Thucydides; for night is a time (kairos) while the battle is an event (praxis). In ekphraseis of events we will deal with what preceded them and what happened in them and the consequences; [23] for example, if we are pronouncing an ekphrasis of a war we will first tell what happened before the war, the enlisting of troops, the expenditure, the people’s fears, then the military engagements, the slaughter, the deaths, then the trophy, then the victory songs of the victors, and the tears and the enslavement of the other side. If we are describing places or times (chronoi) or persons, we will find some justification (logos) both from the narrative and from the beautiful or the useful or the unexpected. The virtues of ekphrasis are above all clarity and vividness (enargeia); for the expression (hermēneia) should almost bring about sight through the sense of hearing. One should also make the style (phrasis) like the subject matter (pragma): if the subject is flowery, the language should be the same; if the subject is harsh the language should be likewise. One should note that some of the more rigorous authorities (akribesteroi) did not make ekphrasis into an exercise on the grounds that it is already included in fable, narration, common place and enkōmion; for, they say, there too we describe places and rivers13 and events and persons. 12

Iliad, 2.217. An odd subject to single out here. Rabe suggests kairoi or chronoi.

13

APPENDIX A

201

Nevertheless, since some good authorities (ou phauloi) counted this as one of the exercises, I too have followed their example, thus avoiding the accusation of idleness. Ἔκφρασίς ἐστι λόγος περιηγηματικός, ὥς φασιν, ἐναργὴς καί ὑπ’ ὄψιν ἄγων τὸ δηλούμενον. Γίνονται δὲ ἐκφράσεις προσώπων τε καὶ πραγμάτων καὶ καιρῶν καὶ τόπων καὶ χρόνων καὶ πολλῶν ἑτέρων. προσώπων μέν, ὡς παρ’ Ὁμήρῳ φολκός ἔην, χωλός δ’ ἕτερον πόδα. πραγμάτων δὲ οἷον ἔκφρασις πεζομαχίας καὶ ναυμαχίας. καιρών δε οἷον εἰρήνης, πολέμου. τόπων δὲ οἷον λιμένων, αἰγιαλῶν, πόλεων. χρόνων δὲ οἷον ἔαρος, θέρους, ἑορτής. γένοιτο δ’ ἄν τις καὶ μικτὴ ἔκφρασις, ὡς παρὰ τῷ Θουκυδίδῃ ἡ νυκτομαχία. ἡ μὲν γὰρ νὺξ καιρός τις, ἡ δὲ μάχη πρᾶξις. Ἐπιχειρήσομεν δὲ τὰ μὲν πράγματα ἐκφράζοντες ἀπὸ τῶν προγεγονότων καὶ ἐν αὐτοῖς γινομένων καὶ ἐπισυμβαινόντων· [p. 23] οἷον εἰ πολέμου λέγοιμεν ἔκφρασιν, πρῶτον μὲν τὰ πρὸ τοῦ πολέμου ἐροῦμεν, τὰς στρατολογίας, τὰ ἀναλώματα, τοὺς φόβους, εἶτα τὰς συμβολάς, τὰς σφαγάς, τοὺς θανάτους, εἶτα τὸ τρόπαιον, εἶτα τοὺς παιᾶνας τῶν νενικηκότων, τῶν δὲ τὰ δάκρυα, τὴν δουλείαν. ἐὰν δὲ τόπους ἐκφράζωμεν ἢ χρόνους ἢ πρόσωπα, ἕξομέν τινα καὶ ἐκ τῆς διηγήσεως καὶ ἐκ τοῦ καλοῦ ἢ χρησίμου ἢ παραδόξου λόγον. Ἀρεταὶ δὲ ἐκφράσεως μάλιστα μὲν σαφήνεια καὶ ἐνάργεια· δεῖ γὰρ τὴν ἑρμηνείαν διὰ τῆς ἀκοῆς σχεδὸν τὴν ὄψιν μηχανᾶσθαι. ἔτι μέντοι συνεξομοιοῦσθαι τὰ τῆς φράσεως ὀφείλει τοῖς πράγμασιν· ἂν ἀνθηρὸν τὸ πρᾶγμα, ἔστω καὶ ἡ λέξις τοιαύτη, ἂν αὐχμηρὸν τὸ πρᾶγμα, ἔστω καὶ ἡ λέξις παραπλησία. Ἰστέον δέ, ὡς τῶν ἀκριβεστέρων τινὲς οὐκ ἔθηκαν τὴν ἔκφρασιν εἰς γύμνασμα ὡς προειλημμένην καὶ ἐν μύθῳ καὶ ἐν διηγήματι καὶ ἐν τόπῳ κοινῷ καί ἐν ἐγκωμίῳ· καὶ γὰρ ἐκεῖ, φασίν, ἐκφράζομεν καὶ τόπους καὶ ποταμοὺς καὶ πράγματα καὶ πρόσωπα, ἀλλ’ ὅμως, ἐπειδή τινες οὐ φαῦλοι καὶ ταύτην ἐγκατηρίθμησαν τοῖς γυμνάσμασιν, οἷς ἠκολουθήσαμεν καὶ ἡμεῖς ῥαθυμίας ἔγκλημα φεύγοντες. Aphthonios, Progymnasmata, ed. H. Rabe (Leipzig, 1926) 36 Ekphrasis is a descriptive speech bringing the subject vividly before the eyes. [37] Subjects of ekphrasis are persons and events, times (kairoi) and places, mute animals and, in addition to these, plants: persons as in Homer ‘round-shouldered, dark-skinned, with curly hair’;14 events such as sea battles and infantry battles, as in the Historian [Thucydides]; times such as spring and summer, saying how many flowers appear during them; places as when Thucydides himself tells of Cheimerion, the harbour of the 14

�� Homer, Odyssey, 19.246.


202

Thesprotians, according to its layout (schēma).15 In describing persons we should go from the first things to the last, that is, from head to toe; events from the things that happened before them, during them and what tends to result from them; times and places from the surroundings and what occurs in them. Some ekphraseis are simple, others mixed; simple such as those relating (diexerchomai) infantry battles or sea battles, mixed such as those combining events and times, as Thucydides describes (ekphrazei) the night battle in Sicily;16 for along with his account of how the battle was fought he also defined what it was like at night. When describing we must produce the relaxed (aneimenos) style [38] and vary it with different figures and completely imitate (apomimeisthai) the things described. Ἔκφρασίς ἐστι λόγος περιηγηματικὸς ὑπ’ ὄψιν ἄγων ἐναργῶς τὸ δηλούμενον. [p. 37] Ἐκφραστέον δὲ πρόσωπά τε καὶ πράγματα, καιρούς τε καὶ τόπους, ἄλογα ζῷα καὶ πρὸς τούτοις φυτά· πρόσωπα μεν ὥσπερ Ὅμηρος γύρος ἐν ὤμοισι, μελανόχροος, ούλοκάρηνος· πράγματα δὲ ὡς ναυμαχίας καὶ πεζομαχίας, ὥσπερ ὁ συγγραφεύς· καιροὺς δὲ ὡς ἔαρ καί θέρος, φράζων ὁπόσα παρ’ αὐτά προέρχεται τῶν ἀνθέων· τόπους δὲ ὡς αὐτὸς ὁ Θουκυδίδης τὸν λιμένα τῶν Θεσπρωτῶν εἶπε Χειμέριον, καθάπερ ἔχει σχήματος. ἐκφράζοντας δὲ δεῖ πρόσωπα μὲν ἀπὸ τῶν πρώτων ἐπὶ τὰ τελευταία ἰέναι, τουτέστιν ἀπὸ κεφαλῆς ἐπὶ πόδας, πράγματα δὲ ἀπὸ τῶν πρὸ αὐτῶν τε καὶ ἐν αὐτοῖς καὶ ὅσα ἐκ τούτων ἐκβαίνειν φιλεῖ, καιροὺς δὲ καὶ τόπους ἐκ τῶν περιεχόντων καὶ ἐν αὐτοῖς ὑπαρχόντων. Τῶν δὲ ἐκφράσεων αἱ μὲν εἰσιν ἁπλαῖ, αἱ δὲ συνεζευγμέναι· καὶ ἁπλαῖ μὲν ὡς αἱ πεζομαχίας ἢ ναυμαχίας διεξερχόμεναι, συνεζευγμέναι δὲ ὡς αἱ πράγματα καὶ καιροὺς ἅμα συνάπτουσαι, ὥσπερ ὁ Θουκυδίδης τὴν ἐν Σικελίᾳ νυκτομαχίαν ἐκφράζει· μετὰ γὰρ τῆς μάχης ὅπως ἐπράττετο καὶ νυκτός ὅπως εἶχεν ὡρίσατο. Εκφράζοντας δὲ δεῖ τόν τε χαρακτῆρα ἀνειμένον ἐκφέρειν [p. 38] καὶ διαφόροις ποικίλλειν τοῖς σχήμασι και ὅλως ἀπομιμεῖσθαι τὰ ἐκφραζόμενα πράγματα. Nikolaos, Progymnasmata, ed. J. Felten (Leipzig, 1913), pp. 67–71 67 Some people place ekphrasis immediately after synkrisis, explaining that the order of the progymnasmata is indifferent (adiaphoros) as different people arrange them in different ways, and there is nothing to prevent ekphrasis being practised as an exercise immediately [68] after synkrisis, for we have already pointed out that it is permissible to use a relaxed 15

Thucydides, 1.46. Thucydides, 7.42–5.

16

APPENDIX A

203

(aneimenos) style in synkrisis and, as we are taught (paradedotai) that the same type of style (apangelia) is used more in this exercise [i.e. ekphrasis], it is reasonable that ekphrasis should follow synkrisis. This is what others write, but, in accordance with the usual practice, I have placed ēthopoiia after synkrisis and ekphrasis after this and say: ekphrasis is an expository [aphēgēmatikos] speech which vividly brings the subject before the eyes. ‘Vividly’ (enargōs) is added because it is in this respect that ekphrasis differs most from diēgēsis. The latter sets out the events plainly, while the former tries to make the listeners into spectators. We compose ekphraseis of places, seasons (chronoi), persons, festivals, events: places such as meadows, harbours, marshes and the like; seasons such as spring, summer; persons such as priests, Thersites and the like; festivals such as the Panathenaia, the Dionysia and what is done during them; and we will use this exercise generally for many things (pros polla). It [ekphrasis] differs from diēgēsis in this respect: the latter sets out the subject generally (ta katholou) while the former goes into details (kata meros). So, for example, it is characteristic of diēgēsis [69] to say ‘the Athenians and the Peloponnesians went to war’, but it is characteristic of an ekphrasis to say that each side made this or that type of preparation and used this or that manner (tropos) of equipment. We must, particularly when we describe statues for example, or paintings or things of this sort, try to add reasons (logismoi) why the painter or sculptor depicted things in certain ways, such as, for example, that he depicted the character as angry from such and such a cause (aitia) or happy, or we will mention some other emotion resulting from the story about the person being described. Reasons contribute greatly to enargeia in other types of ekphrasis too. We will begin from first things and then come to the last, so that if we have a bronze man or painted man or whatever is the subject of the ekphrasis we will start from the head and go through the details in order (epi ta kata meros). For thus the speech becomes lively (empsuchos) throughout. There being five sections of a speech, as I have frequently [70] said – prooemion, narration, antithesis, lusis, epilogue – ekphrasis will prepare us for the narrative section except in that it does not give a plain exposition (psilē aphēgēsis) but makes use of those elements that create enargeia and bring the subjects of the speech before the eyes and almost make the audience into spectators.17 There being three genres [of oratory], I mean the judicial, epideictic and deliberative, the use of this exercise will be found in all of them. For when 17 The contrast is again between the ‘plain’ style of narration, for which the exercise of diēgēma provided a preparation and the vivid style in which ekphrasis gave practice.


204

we are advising we often need to describe whatever it is we are speaking about in order to be more persuasive, and when speaking in accusation or defence we need the amplification which ekphrasis provides; moreover, in epideictic speeches, it is sufficient for the ekphrasis to create a feeling of pleasure (hēdonē) in the people sitting in the audience (theatra). For the most part, this exercise is one of those which are used as parts [of a speech]. But nothing would prevent it from sometimes being made sufficient in itself for a complete subject, although for the most part it is one of the parts [of a speech]. We need a variety of styles for it, for it is necessary to fit the type of style (apangelia) to the subject matter, when we are speaking sweetly or lamenting disasters [71] or generally presenting some other emotion (pathos). On occasion we only want to inspire a feeling of happiness (euthumia); at other times we want to arouse indignation (deinōsai) and to amplify the subject, as when Demosthenes in On the False Embassy tries to bring the suffering of Phokis before the eyes through his speech.18 Τινὲς μετὰ τὴν σύγκρισιν εὐθὺς τὴν ἔκφρασιν τάξαντες οὕτως ἔγραψαν· ἔστι μὲν ἡ τῶν ἐφεξῆς προγυμνασμάτων τάξις ἀδιάφορος ἄλλων ἄλλως ταττόντων, οὐδὲν δὲ κωλύει τὴν ἔκφρασιν ἐν μελέτῃ ποιεῖσθαι εὐθὺς [p. 68] μετὰ τὴν σύγκρισιν· ἐπειδὴ γὰρ καὶ ἐπὶ τῆς συγκρίσεως ἔφαμεν εἶναι ἄδειαν τῆς ἀνειμένης φράσεως, καὶ ἐπὶ ταύτης δὲ μᾶλλον ἐκείνῳ τῷ εἴδει τῆς ἀπαγγελίας χρῆσθαι παραδέδοται, εἰκότως ἀκολουθεῖν δεῖ τῇ συγκρίσει τὴν ἔκφρασιν. ἀλλ’ ἐκείνοι μὲν οὕτως· ἡμεῖς δὲ τῷ κεκρατηκότι ἔθει ἑπόμενοι μετὰ τὴν σύγκρισιν μὲν τὴν ἠθοποιίαν ἐτάξαμεν, μετὰ ταύτην δὲ τὴν ἔκφρασιν καί φαμεν· ἔκφρασίς ἐστι λόγος ἀφηγηματικός, ὑπ’ ὄψιν ἄγων ἐναργῶς τὸ δηλούμενον. πρόσκειται δὲ ἐναργῶς, ὅτι κατὰ τοῦτο μάλιστα τῆς διηγήσεως διαφέρει· ἡ μὲν γὰρ ψιλὴν ἔχει ἔκθεσιν πραγμάτων, ἡ δὲ πειρᾶται θεατὰς τοὺς ἀκούοντας ἐργάζεσθαι. ἐκφράζομεν δὲ τόπους, χρόνους, πρόσωπα, πανηγύρεις, πράγματα. τόπους μέν, οἷον λειμῶνας, λιμένας, λίμνας καὶ ὅσα τοιαῦτα· χρόνους δέ, οἷον ἔαρ, θέρος· πρόσωπα δέ, οἷον ἱερέας, Θερσίτας καὶ τὰ τοιαῦτα· πανηγύρεις δέ, ὡς Παναθήναια, Διονύσια καὶ τὰ ἐν αὐτοῖς δρώμενα· καὶ ὅλως πρὸς πολλὰ τῷ προγυμνάσματι τούτῳ χρησόμεθα. διαφέρει δὲ καὶ κατ’ ἐκεῖνο τῆς διηγήσεως, ὅτι ἡ μὲν τὰ καθόλου, ἡ δὲ τὰ κατὰ μέρος ἐξετάζει· οἷον διηγήσεως [p. 69] μέν ἐστι τὸ εἰπεῖν ἐπολέμησαν Ἀθηναῖοι καὶ Πελοποννήσιοι· ἐκφράσεως δέ, ὅτι τοιᾷδε καὶ τοιᾷδε ἑκάτεροι παρασκευῇ ἐχρήσαντο και τῷδε τῷ τρόπῳ τῆς ὁπλίσεως. Δεῖ δέ, ἡνίκα ἂν ἐκφράζωμεν καὶ μάλιστα ἀγάλματα τυχὸν ἢ εἰκόνας ἢ εἴ τι ἄλλο τοιοῦτον, πειρᾶσθαι λογισμοὺς προστιθέναι τοῦ τοιοῦδε ἢ τοιοῦδε παρὰ τοῦ γραφέως ἢ πλάστου σχήματος, οἷον τυχὸν ἢ ὅτι ὀργιζόμενον ἔγραψε διὰ τήνδε τὴν αἰτίαν ἢ ἡδόμενον, ἢ ἄλλο τι πάθος ἐροῦμεν συμβαῖνον τῇ περὶ 18

Demosthenes, 19.65.

APPENDIX A

205

τοῦ ἐκφραζομένου ἱστορίᾳ· καὶ ἐπὶ τῶν ἄλλων δὲ ὁμοίως πλεῖστα οἱ λογισμοὶ συντελοῦσιν εἰς ἐνάργειαν. Ἀρξόμεθα δὲ ἀπὸ τῶν πρώτων, και οὕτως ἐπὶ τὰ τελευταῖα ἥξομεν οἷον εἰ ἄνθρωπον χαλκοῦν ἢ ἐν γραφαῖς ἢ ὁπωσοῦν ἔχομεν ἐν τῇ ἐκφράσει ὑποκείμενον, ἀπὸ κεφαλῆς τὴν ἀρχὴν ποιησάμενοι βαδιοῦμεν ἐπὶ τὰ κατὰ μέρος· οὕτω γὰρ πανταχόθεν ἔμψυχος ὁ λόγος γίνεται. Πέντε δ’ ὄντων τῶν τοῦ λόγου μερῶν, ὡς πολλάκις [p. 70] εἴρηται, προοιμίου, διηγήσεως, ἀντιθέσεως, λύσεως, ἐπιλόγου, γυμνάσει ἡμᾶς ἡ ἔκφρασις πρὸς τὸ διηγηματικὸν μέρος, πλὴν ὅσον οὐ ψιλὴν ἀφήγησιν ποιουμένη, ἀλλὰ παραλαμβάνουσα τὰ ἐργαζόμενα τὴν ἐνάργειαν καὶ ὑπ’ ὄψιν ἡμῖν ἄγοντα ταῦτα, περὶ ὧν εἰσιν οἱ λόγοι, καὶ μονονοὺ θεατὰς εἶναι παρασκευάζοντα. Τριῶν δὲ ὄντων εἰδῶν, τοῦ τε δικανικοῦ λέγω καὶ πανηγυρικοῦ καὶ συμβουλευτικοῦ, ἐν πᾶσιν ἡ χρεία τοῦ προγυμνάσματος τούτου εὐρεθήσεται· καὶ γὰρ συμβουλεύοντες πολλάκις ἀνάγκην ἔχομεν ἐκφράσαι τοῦτο, περὶ οὗ ποιούμεθα τοὺς λόγους, ἵνα μᾶλλον πείσωμεν, καί κατηγοροῦντες ἢ ἀπολογούμενοι δεόμεθα τῆς ἐκ τοῦ ἐκφράζειν αὐξήσεως, καὶ μέντοι καὶ ἐν πανηγυρικαῖς ὑποθέσεσιν ἱκανὸν τὸ τῆς ἐκφράσεως ἡδονὴν ἐμποιῆσαι τοῖς ἐν θεάτροις καθημένοις. Ἔστι δὲ ὡς ἐπὶ τὸ πολὺ τοῦτο τὸ προγύμνασμα τῶν ὡς μερῶν παραλαμβανομένων· οὐδὲν δὲ ἴσως ἂν κωλύοι καὶ ὡς ἀρκοῦσάν ποτε αὐτὴν πρὸς ὅλην ὑπόθεσιν ἐργάσασθαι, ὡς ἐπὶ τὸ πλεῖστον μέντοι τῶν μερῶν ἐστι. Φράσεως δὲ ποικίλης ἐν αὐτῇ δεόμεθα· πρὸς γὰρ τὴν ὑποκειμένην ὑπόθεσιν ἁρμόζειν δεῖ καὶ τὸ τῆς ἀπαγγελίας εἶδος, ἢ γλυκαίνοντας ἢ ἐκτραγῳδοῦντας τὰς συμφορὰς [p. 71] ἢ ὅλως ἄλλο τι παριστάντας πάθος· ἔστι γὰρ ὅτε εὐθυμίαν μόνην ἐμποιῆσαι θέλομεν, ἔστι δὲ ὅτε δεινῶσαί τε καὶ αὐξῆσαι, ὡς ὁ Δημοσθένης ἐν τῷ Περί τῆς παραπρεσβείας τὸ κατὰ τὴν Φωκίδα πάθος ὑπ’ ὄψιν ἄγειν πειρᾶται διὰ τοῦ λόγου. II. Extracts from the Byzantine Commentaries on Aphthonios Sardianos, Commentarium in Aphthonii Progymnasmata On the definition (p. 216, l. 8 – p. 217, l. 5): Periēgēmatikos: Instead of ‘as in a painting’ (graphikos), ‘surveying’ (periodeutikos), ‘giving a detailed account’ (diexodikos); as if seeming to go around in the speech and as if showing; just as if someone took a recent arrival in Athens and guided him around the city, showing him the gymnasia, the Peiraeus and each of the rest [of the sights]’; metaphorically, therefore, the speech which relates (aphēgeomai) everything in order, relating to both the action and the person and showing [it] in detail is called periēgēmatikos.


206

Hup’ opsin agōn: Instead of making manifest (phaneron), bringing the subject vividly before the eyes by detailed presentation (kata meros); for the clarity (saphēneia) of the speech makes the listeners imagine (noein) and see (blepein) the subject. Alternatively, ‘hup’ opsin’ as if less distinctly; for one provides a pure vision (katharan thean), while the other [provides] a plain impression (tupos) and mental representation (phantasia) of the thing (pragma);19 for even if the speech were ten thousand times vivid, it would be impossible to bring before the eyes ‘the thing shown’ or described itself. He said ‘vividly’ on account of narration, because narration is composed in a condensed manner (pachumerōs) while ekphrasis is composed in a detailed manner (leptomerōs). So a vivid speech is one that is clear and pure and as if alive (empnous); for, by the word alone, it all but makes one see what one has never seen, imitating the painters’ art. Ἔκφρασίς ἐστι λόγος περιηγηματικός. Ἀντὶ τοῦ γραφικός, περιοδευτικός, διεξοδικός, οἱονεὶ δοκῶν τῷ λόγῳ συμπεριιέναι καὶ οἱονεὶ δεικνύναι· ὥσπερ ἂν εἴ τις ἐπιδημήσαντά τινα Ἀθήναζε παραλαβὼν ὑφηγεῖτο τὴν πόλιν, δεικνὺς τὰ γυμνάσια, τὸν Πειραιᾶ καὶ τῶν ἄλλων ἕκαστον· μεταφορικῶς οὖν καὶ ὁ λόγος ὁ πάντα ἑξῆς καὶ τὰ τοῦ πράγματος καὶ τὰ τοῦ προσώπου ἀφηγούμενος <καὶ> μετὰ ἀκριβείας δεικνὺς περιηγηματικὸς ὀνομάζεται. Ὑπ’ ὄψιν ἄγων ἐναργῶς τὸ δηλούμενον. Ἀντὶ τοῦ φανερὸν ποιῶν, ἐκ τῶν κατὰ μέρος ἐναργῶς εἰς ὄψιν ἄγων τὸ ὑποκείμενον· ἡ γὰρ τοῦ λόγου σαφήνεια νοεῖν καὶ βλέπειν ποιεῖ τὰ λεγόμενα τοὺς ἀκούοντας. ἢ τὸ ὑπ’ ὄψιν οἱονεὶ ἀμυδρότερον· τὸ μὲν γὰρ καθαρὰν20 θέαν δίδωσι, τὸ δὲ τύπον ψιλὸν καὶ φαντασίαν τοῦ πράγματος· κἂν γὰρ μυριάκις ἐναργὴς εἴη ὁ λόγος, ἀδύνατον αὐτὸ κατ’ ὄψιν ἀγαγεῖν τὸ δηλούμενον ἤτοι ἐκφραζόμενον. ἐναργῶς δὲ εἶπε διὰ τὴν διήγησιν, διότι, ἡ μὲν διήγησις παχυμερῶς λέγεται, ἡ δὲ ἔκφρασις λεπτομερῶς· ἐναργὴς οὖν λόγος ὁ σαφὴς καὶ καθαρὸς καὶ οἷον ἔμπνους· ἃ γὰρ μή τις ἑώρακε, ταῦτα μονονοὺ βλέπειν ποιεῖ ῥήματι ψιλῷ τὴν τῶν ζωγράφων τέχνην μιμούμενος. At the end of his discussion of Aphthonios’ remarks on style, Sardianos adds (p. 224, l. 21 – p. 225, l. 4): Theon says that the virtues of ekphrasis are clarity and vividness that makes one almost see the subject; for that which is very plain to the 19

This extremely intriguing sentence is unfortunately unclear and may be corrupt. The distinction appears to be drawn between direct perception on the one hand and the tupos kai phantasia on the other, but it is unclear how this distinction is thought to relate to the definition of ekphrasis. 20 Rabe prints καθαρόν in his text. I have adopted the tentative suggestion made in the apparatus criticus.

APPENDIX A

207

senses (lian phaneron) and lies before the eyes is vivid; if then the speech is clear and vivid, it almost transfers its subject from the sense of hearing into the eyes; for the speech, in contemplating (theōrōn) the things shown, traces (hupographei) the impression (tupon) of it for (or with) the eyes and paints the truth for (or with) the imagination (phantasia).21 Θέων δὲ ἀρετὰς λέγει ἐκφράσεως σαφήνειαν καὶ ἐνάργειαν τοῦ σχεδὸν ὁρᾶσθαι τὰ ἀπαγγελλόμενα· ἐναργὲς γὰρ τὸ λίαν φανερὸν καὶ τοῖς ὀφθαλμοῖς ὑποπῖπτον· εἰ γὰρ σαφὴς καὶ ἐναργὴς εἴη ὁ λόγος, ἀπὸ τῆς ἀκοῆς εἰς τοὺς ὀφθαλμοὺς σχεδὸν τὰ λεγόμενα μεθίστησιν· ὁ γὰρ λόγος τὰ δηλούμενα θεωρῶν τούτων τοῖς ὀφθαλμοῖς ὑπογράφει τὸν τύπον καὶ τῇ φαντασίᾳ ζωγραφεῖ τὴν ἀλήθειαν.

Anonymous scholia to Aphthonios, in Walz, Rhetores graeci, 2 On the place of ekphrasis in the sequence and its relation to diēgēma (p. 55, ll. 8–16): There being three types of diēgēsis, simple (haplous), confirmatory (enkataskeuos) and elaborate (endiaskeuos), [the exercise of] diēgēma gives us practice in the first two, while ekphrasis gives us practice in the elaborate type. It should, since it is related to diēgēma, be placed immediately after that exercise, but because it is one of the more advanced and more complex exercises and concerns the same subjects as enkōmion, psogos, synkrisis and ēthopoiia – that is to say, persons and events, times (kairoi) and places and the like, it is reasonable to place it alongside them. Τριῶν ὄντων εἰδῶν τῆς διηγήσεως, ἁπλοῦ, ἐγκατασκεύου καὶ ἐνδιασκεύου, τοῖς μὲν δυσὶ γυμνάζει ἡμᾶς τὸ διήγημα, τῷ δὲ ἐνδιασκεύῳ ἡ ἔκφρασις· ἔδει οὖν ὡς οὖσαν συγγενῆ τῷ διηγήματι ταύτην μετ’ ἐκεῖνο εὐθὺς τάττεσθαι· ἀλλ’ ἐπεὶ τῶν τελεωτέρων ἐστὶ τοῦτο καὶ ποικιλωτέρον, καταγίνεται δὲ καὶ περὶ ἃ τὸ ἐγκώμιον καὶ ὁ ψόγος καὶ ἡ σύγκρισις καὶ ἠθοποιία, περὶ πρόσωπα δηλαδὴ καὶ πράγματα, καιρούς τε καὶ τόπους καὶ τὰ τοιαῦτα, εἰκότως αὐτοῖς συνετάχθη. Doxapatres, Homiliae in Aphthonii Progymnasmata, in Walz, Rhetores graeci, 2, Introduction (on the place of ekphrasis in the sequence and its relation to narration) (p. 509, l. 5 – p. 510, l. 9): Some people think ekphrasis should be placed immediately after diēgēma because it, too, is an element (topos) of the narrative mode (to diēgēmatikon). For they also say that because there are three types of diēgēsis, simple (haplous), confirmatory (enkataskeuos) and elaborate (endiaskeuos), [the exercise of] diēgēma gives us practice in the first two, while ekphrasis 21

Another tantalizing and obscure passage.

208


gives us practice in the final one, the elaborate type. Others place it after common place;22 however, Aphthonios and the more accurate authorities among his predecessors place it after ēthopoiia. For if ekphrasis is to be placed alongside diēgēma because it is of the narrative type, then thesis and the introduction of a law should be placed before common place, if, at any rate, they belong to the argumentative section while common place belongs to the epilogue. At all events, even if both of these, that is diēgēma and ekphrasis, contribute to the narrative sections of speeches (diēgēseis), it is nevertheless the case that since diēgēma contributes to the plainer and simpler kinds (tropoi) of narration – these would be the ones that relate the events (pragmata) in a condensed form (pachumerōs) – it has been placed earlier since it is easier. Ekphrasis, by contrast, has been placed much later [in the sequence] since it is more complex and requires greater mastery (hexis) in as much as it gives us practice in the more complex kind of narration, the elaborate one (endiaskeuos). Hermogenes in his book On Invention deals with narrations and the kinds of narration and explains how to compose the confirmatory (enkataskeuos) and the simple (haplous) kind but does not explain the expansion (platusmos) of the elaborate narration at the same time, but rather at the end of the third book.23 In just the same way, Aphthonios places many other exercises after the exercise of diēgēma that prepares us partly for the simple kind of narration, because the events are told in condensed form, and partly for the confirmatory kind, because of the use of both the cause and other elements of narration, before finally coming to ekphrasis, which belongs to the elaborate type. Τινὲς μετὰ τὸ διήγημα αὖθις ἀξιοῦσι τὴν ἔκφρασιν τάττεσθαι διὰ τὸ καὶ αὐτὴν τοῦ διηγηματικοῦ εἶναι τόπον. καὶ γάρ φασιν, ὅτι τριῶν ὄντων τῆς διηγήσεως εἰδῶν, ἁπλοῦ, ἐγκατασκεύου, ἐνδιασκεύου, τοῖς μὲν δυσὶ τοῖς προτέροις γυμνάζει ἡμᾶς τὸ διήγημα, τῷ δὲ τελευταίῳ τῷ ἐνδιασκεύῳ ἡ ἔκφρασις. ἕτεροι δὲ καὶ μετὰ τὸν κοινὸν τόπον24 αὐτὴν τάττουσιν, ὁ μέντοι Ἀφθόνιος καὶ οἱ ἀκριβέστεροι τῶν πρὸ αὐτοῦ μετὰ τὴν ἠθoποιίαν. εἰ γὰρ διότι τοῦ διηγηματικοῦ ἐστι τύπου ἡ ἔκφρασις, διὰ τοῦτο συντακτέα ἔσται τῷ διηγήματι, ὀφείλει καὶ ἡ θέσις καὶ ἡ τοῦ νόμου εἰσφορὰ τεθῆναι πρὸ τοῦ κοινοῦ τόπου. εἴγε ταῦτα μὲν τοῦ ἀγωνιστικοῦ εἰσι μέρους, ὁ δὲ κοινὸς τόπος τῶν ἐπιλόγων. ἄλλως τε δὲ εἰ καὶ ἀμφότερα ταῦτα, τό τε διήγημα καὶ ἡ ἔκφρασις πρὸς τὰς διηγήσεις συμβάλλονται, ὅμως τὸ μὲν διήγημα πρὸς τοὺς ἀποικιλωτέρους καὶ ἁπλουστέρους τρόπους τῆς διηγήσεως συμβαλλόμενον, οὕτοι δ’ ἂν εἶεν οἱ παχυμερῶς τὰ πράγματα λέγοντες, προετάγη ὡς εὐκολώτερον. ἡ δὲ ἔκφρασις πολλῷ τούτου ὕστερον τέθειται, ὡς ποικιλωτέρα Walz prints κoιvὸv τρόπov. Ps.-Hermogenes, On Invention, 2 and 3.15. 24 Walz prints τρόπον. I have emended this to τόπον. 22 23

APPENDIX A

209

οὖσα καὶ μείζονος δεομένη τῆς ἕξεως. παρόσον καὶ τῷ ποικιλωτέρῳ τῶν διηγημένων τρόπων ἡμᾶς ἐγγυμνάζει, τῷ ἐνδιασκεύῳ. καὶ ὥσπερ ὁ Ἑρμογένης ἐν τῷ περὶ εὑρέσεως βιβλίῳ περὶ διηγήσεων διαλαβὼν καὶ τούς τε τρόπους ταύτης καὶ τὰ δι’ ὧν ὅ τε ἐγκατάσκευος τρόπος καὶ ὁ ἁπλοῦς γίνεται τὸν τῆς ἐνδιασκεύου διηγήσεως [510] πλατυσμὸν οὐχ ὁμοῦ τούτῳ, ἀλλ’ ἐν τῷ τέλει τοῦ τρίτου τόμου ἐδίδαξεν, οὕτω καὶ ὁ Ἀφθόνιος μετὰ τὸ διήγημα, τὸ πῇ μὲν πρὸς τὸν ἁπλοῦν τρόπον τῆς διηγήσεως ἡμᾶς προγυμνάζον, διὰ τὸ παχυμερῶς ἐν αὐτῷ τὰ πράγματα λέγεσθαι, πῇ δὲ καὶ πρὸς τὸν ἐγκατάσκευον διά τε τὴν αἰτίαν καὶ τὴν τῶν ἄλλων περιστατικῶν ἐν τούτῳ γινομένην συμφόρησιν, πολλὰ θεὶς μεταξὺ ἄλλα γυμνάσματα ἐν ὑστέρῳ τάττει τὴν ἔκφρασιν, ἥτις τοῦ ἐνδιασκεύου ἐστὶ τύπου. Examples of simple and elaborate (endiaskeuos) narrations: Anonymous, Peri tōn tessarōn merōn tou teleiou logou, Walz, Rhetores graeci, 3, p. 576, l. 21 – p. 578, l. 5 There are three ways to write narratives: simple, elaborate (endiaskeuos), and confirmatory (enkataskeuos). The simple narrative is the one that just sets out the facts in a plain manner and obviously without adornment. The elaborate narrative is the ekphrastic one that recounts everything in detail and all but brings the actions before the eyes while the confirmatory is the one that gives the causes for the actions and states the reason for each thing that occurred. The writer (zōgraphos) will use each and every one of these at the right moment: the simple one when it is the time for an unaffected (apheleia) and truth-like style; the elaborate in debates and when amplifying the subjects. You can even combine the confirmatory with the elaborate type whenever you wish to bring the subject before the eyes and to amplify it. However, the plain narrative is not mixed with the others as it is not possible for the same thing to be both plain and complex (poikilos). In the case of the preliminary exercise of narration, for the most part the rhetoricians recommend that it should be simple; however, it is not unreasonable to introduce more complexity on occasion and expand it to provide a more challenging exercise in rhetoric. So then, as an example I will give you a narrative presented in three ways; let this be the story of the madness of Ajax. The plain narrative: Ajax, having been defeated in the contest for the arms of Achilles, was angry with the victor, Odysseus, and plotted his revenge against the sons of Atreus who had made the choice. During the night he advanced on them with hostile intent, his sword at the ready, but Athena confused both his mind and his sight, and the hero fell upon the flocks and stabbed some

210


as if they were men and drove the others to his tent and, after whipping them severely, killed them. Finally, having recovered from his madness, he sentenced himself to death by his own hand. The elaborate narrative: Ajax, having been defeated in the contest for the arms of Achilles, carried a burning rage in his heart against those who had awarded the prize to the man from Ithaka, the sons of Atreus. He indicated (mēnuei) his inner feelings by his wild appearance, by the ruthless and hot-blooded look in his eyes as well as by his fast and deep breathing. During the night he set out against them, his sword in his hand, a sword that was sharp, a sword that was shining, glinting in the darkness of the night. He moved at one moment with stealthy control and at the next he moved quickly in his rage – long was the stride of the gigantic hero – but Athena diverted both his mind and his eyes, darkness fell on his inner and his outer vision. The night stalker fell upon the flocks’ fold and cut down the dumb animals as if they were men; one he slashed in two, another he pierced through; the head of another he separated from the body, into the belly of another he plunged his sword; here were streams of blood, there spilled guts, everywhere piles of corpses; here, those that were still breathing tried to leap away but the fold stopped them, the wolf was inside, the guard dogs slept soundly, they were soon avenged, the hero, recovering his senses, fell, becoming the victim of his own sword.25 Ἡ τοῦ διηγήματος γραφὴ τριπλῆ ἐστιν, ἁπλῆ, ἐνδιάσκευος καὶ ἐγκατάσκευος· καὶ ἁπλῆ μέν ἐστιν ἡ αὐτὸ μόνον τὸ πρᾶγμα γυμνὸν ἀπαγγέλλουσα, καὶ ὡς φανερὸν ξηρόν· ἐνδιάσκευος δὲ ἡ ἐκφραστικὴ καὶ λεπτολογοῦσα τὸ καθέκαστον, καὶ εἰς ὄψιν μονονοὺ παράγουσα τὰ πραχθέντα· ἐγκατάσκευος δὲ ἡ τὰς αἰτίας τῶν πραχθέντων ἀποδιδοῦσα, καὶ τῶν συμπεσόντων ἕκαστον αἰτιολογοῦσα· καὶ ὁ ζωγράφος ἑκάστῃ τούτων χρήσεται πάντως κατὰ καιρόν, τῇ μὲν ἁπλῇ ἐν καιρῷ ἀφελείας καὶ ἀληθείας, τῇ δὲ ἐνδιασκεύῳ ἐν ἀγῶσί τε καὶ αὐξήσεσι τῶν προτεθειμένων· δύνασαι δὲ καὶ μιγνύειν μετὰ τῆς ἐνδιασκεύου τὴν ἐγκατάσκευον, ὁπηνίκα καὶ ὑπ’ ὄψιν θέλεις ἄγειν τὸ πρᾶγμα καὶ αὔξειν· ἡ δὲ ἁπλῆ τούτοις [p. 577] ἀσύγκρατος, οὐδὲ ἐξὸν εἶναι τὸ αὐτὸ καὶ ἁπλοῦν ὁμοῦ καὶ ποικίλον· τὸ μέντοι προγυμναστικὸν διήγημα ἁπλοῦν τὰ πολλὰ γράφεσθαι οἱ τεχνικοὶ διορίζονται, ὅμως οὐκ ἀπεικὸς καὶ ποικίλλεσθαί ποτε τοῦτο καὶ πλατύνεσθαι πρὸς ἄσκησιν φιλοτιμοτέραν ῥητορικῆς· φέρε δὲ ὡς ἐν παραδείγματι παραθήσομέν σοι διήγημα κατὰ τὸ ῥηθὲν τριπλοῦν εἶδος, ἔστω δὲ τοῦτο τὸ κατὰ τὴν μανίαν τοῦ Αἴαντος.

25 There follows an example of a confirmatory, enkataskeuos, narration in which reasons are given for each part of the action.

APPENDIX A

211

Ἁπλοῦν· Αἴας ἐπὶ τοῖς ὅπλοις παρευδοκιμηθεὶς κατὰ τοῦ λαβόντος ὀργίζεται Ὀδυσσέως, κατὰ τῶν ψηφισαμένων τῶν Ἀτρειδῶν ἄμυναν σκέπτεται· νύκτα ὡς ἐπ’ ἐκείνους ξιφηφόρος χωρεῖ, ἀλλ’ Ἀθηνᾶ καὶ νοῦν καὶ ὄψιν ἐσκότωσεν, καὶ ὁ ἥρως εἰσπίπτει τοῖς κτήνεσι, καὶ τὰ μὲν ὡς ἄνδρας κεντᾷ, τὰ δ’ ἐλαύνει πρὸς τὴν σκηνήν, καὶ μετὰ πολλὴν τὴν μάστιγα κτείνει· τέλος τὴν μανίαν ἀνενεγκὼν αὐτοχειρίαν ἑαυτοῦ κατεδίκασεν. Ἐνδιάσκευον· Αἴας ἐπὶ τοῖς ὅπλοις παρευδοκιμηθεὶς κατὰ τῶν δικασάντων τὴν κρίσιν τοῦ Ἰθακησίου, τῶν Ἀτρειδῶν πίμπραται τὴν καρδίαν θυμῷ, τῷ ἀγριωπῷ μηνύει τὴν ἔνδον διάθεσιν, καὶ τῇ ἀναιδήσει καὶ τῷ ὑφαίμῳ τοῦ ὀφθαλμοῦ, τῷ τε τοῦ ἄσθματος πυκνῷ καὶ βαρεῖ νυκτὸς ὡς ἐπ’ ἐκείνους στρατεύει, τὸ ξίφος ἐν ταῖν χεροῖν, εὔθηκτον τὸ ξίφος, στίλβον τὸ ξίφος, ἀνταυγάζον πρὸς τὸ σκότιον τῆς νυκτός· τὸ κίνημα πῆ μὲν ἠρεμαῖον δολίως, πῆ δὲ καὶ ἅλλεται θυμικῶς, μακρὸν τὸ βῆμα τοῦ γιγαντιαίου τοῦ ἥρωος, ἀλλὰ παράγει καὶ νοῦν καὶ ὀφθαλμοὺς Ἀθηνᾶς, σκότος ἅμα τοῖν βλεφάροιν, τῷ ἐνδομύχῳ τῷ ἐξωτίκῳ· εἰσπίπτει τῇ μάνδρᾳ τῆς λείας ὁ νυκτίλοχος, καὶ ὡς ἄνδρας κατατέμνει τὰ ἄλογα, τὸ μὲν κατερράχισε, τὸ δὲ διεκέντησε, τοῦ δὲ τὴν κεφαλὴν τοῦ λοιποῦ σώματος ἐξετίναξεν ἑτέρῳ κατὰ γαστρὸς τὸ ξίφος ἐμβάπτεται· ἐνταῦθα κρουνὸς αἵματος, ἐκεῖσε σπλάγχνα [p. 578] ἐκκεχυμένα, ἐπάλληλα τὰ πτώματα πανταχοῦ, ὧδε τῶν ἔτι πνεόντων ἅλματα πρὸς φυγήν, ἀλλὰ τὸν δεσμὸν ἐπέσχεν ἡ ἔπαυλις, ὁ λύκος ἐντός, οἱ φύλακες κύνες ὑπνοῦσι τὸν χάλκεον, ἐξεπράχθησαν μικρὸν καὶ ὁ ἥρως ἀνενεγκῶν τοῦ οἰκείου ξίφους ἔργον καὶ αὐτὸς πίπτει.


Appendix B Subjects for Ekphrasis

Asterisk indicates that the category is unique to the author. Theon: Pragmata: Prosōpa:

Topoi: Chronoi: *Tropoi:

a battle, peace, a storm, a famine, plague, earthquake. Eurybates, Homer, Odyssey, 19.246; Thersites, Homer, Iliad, 2.219 and 217; animals from Herodotos’ account of Egypt (2.76; 71 and 68). meadows, shores, cities, a wilderness. periods marked off by nature (spring, summer) or culture (festivals). The Making of the Shield of Achilles (Homer, Iliad, 18); Thucydides’ account of the fortification of Plataia and the construction of a (siege) machine (2.75–8 and 4.100); and Ktesias 9 (a strategy involving effigies).

Ps.-Hermogenes: Prosōpa: Pragmata: Kairoi: Topoi: Chronoi:

description of Thersites, Homer, Iliad, 2.217. land battle and sea battle. peacetime, war. harbours, seashores, cities. spring, summer, festival time.

Aphthonios: Prosōpa: Pragmata:

description of Eurybates, Homer, Odyssey, 19.246. sea battle, infantry battle (‘as in the Historian’, i.e. Thucydides). Kairoi: spring and summer. Topoi: Thucycides’ description of the harbour of Cheimerion, 1.46. *Aloga zōa, phuta: no examples.

214


Nikolaos: Topoi: Chronoi: Prosōpa: *Panēgureis:

meadows, harbours, marshes ‘and the like’. spring, summer. priests, Thersites and the like. the Panathenaia, the Dionysia ‘and what happens during them’. Pragmata: no examples. *Statues and paintings: no examples.

Other Examples of Ekphrasis from Handbooks Related to the Teaching of the Progymnasmata Theon:

Nikolaos: Doxapatres:

passages mentioned in Theon’s preliminary discussion of how to teach ekphrasis in his introduction: plague in Thucydides, 2.47–57; (an event) the fortification of Plataia, 3.21; (places) Sais in Plato, Timaeus, 21e– 25d, Ekbatana in Herodotos, 1.98 (Theon mistakenly places this passage in Book 2 [Theon, p. 128, n. 78]), Theopompos on valley of Tempe; Philistos, Dionysios’ war preparations against the Carthaginians (Book 8) and funeral procession (Book 11). destruction of Phokis in Demosthenes, 19.65. cites the text of Thucydides, 2.79 as an example of a land battle and 2.83 as an example of a sea battle.

Bibliography

Primary Sources Achilles Tatios, Leukippe and Kleitophon, ed. Ebbe Vilborg (2 vols, Stockholm: Almqvist and Wiksell, 1955–62). Ailios Aristeides, Opera, I, ed. Friedrich W. Lenz and Charles A. Behr (Leiden: Brill, 1976–80); II, ed. Bruno Keil (Berlin; Weidmann, 1898). English translation by Charles A. Behr (2 vols, Leiden: Brill, 1981–86). Anonymous, Peri tōn tessarōn merōn tou teleiou logou (On the Four Parts of a Complete Speech), in Rhetores graeci, ed. Christian Walz, vol. 3 (Stuttgart: Sumptibus J.G. Cottae, 1834). Anonymous Seguerianus, Art of Rhetoric, in Two Greek Rhetorical Treatises from the Roman Empire, ed. and trans. Mervin R. Dilts and George A. Kennedy (Leiden: Brill, 1997), pp. 1–73. Aphthonios, Progymnasmata, ed. Hugo Rabe (Leipzig: Teubner, 1926). Apsines, Art of Rhetoric, in Two Greek Rhetorical Treatises from the Roman Empire, ed. and trans. Mervin R. Dilts and George A. Kennedy (Leiden: Brill, 1997), pp. 75–239. Basilakes, Nikephoros, Progimnasmi e Monodie, ed. and trans. Adriana Pignani (Naples: Bibliopolis, 1983). Comanini, Gregorio, Il Figino ovvero del fine della pittura, ed. �� Paola Barocchi� (Bari: Laterza, 1962). Concilia omnia, ed. F.L. Sarius (Colonia Agrippina: Apud Geruuinum Galenium et Haeredes Iohannis Quentelii, 1567). Deinarchos, Orationes cum fragmentibus, ed. Nicos Conomis (Leipzig: Teubner, 1975). Demetrios, On Style, ed. William Rhys Roberts (Hildesheim: G. Olms, 1969). Dionysios of Halikarnassos, Lysias, in Opuscula I, ed. Hermann Usener and Ludwig Radermacher (Leipzig: Teubner, 1899), pp. 8–53. English translation in Critical Essays I, trans. S. Usher (Cambridge, MA: Harvard University Press, 1974). Dionysios of Halikarnassos, On Imitation, in Opuscula II, ed. Hermann Usener and Ludwig Radermacher (Leipzig: Teubner, 1929), pp. 197– 217. ps.-Dionysios of Halikarnassos, On Epideictic, in Dionysios of Halikarnassos, Opuscula II, ed. Hermann Usener and Ludwig Radermacher (Leipzig: Teubner, 1929), pp. 255–92. English translation in Menander Rhetor,

216


On Epideictic, ed. and trans. Donald A. Russell and Nigel G. Wilson (Oxford: Clarendon Press, 1981), pp. 362–81. ps.-Dionysios of Halikarnassos, On Mistakes in Declamation, in Dionysios of Halikarnassos, Opuscula IV, ed. Hermann Usener and Ludwig Radermacher (Leipzig: Teubner, 1933), pp. 359–74. Doxapatres, Homiliae in Aphthonii Progymnasmata, in Rhetores graeci, ed. Christian Walz, vol. 2 (Stuttgart: Sumptibus J.G. Cottae, 1835), pp. 69– 564. Fortunatianus, Ars rhetorica, in Rhetores latini minores, ed. Karl Halm (Leipzig: Teubner, 1863). Heliodoros, Aithiopika, ed. Robert Mantle Rattenbury and Thomas Wallace Lumb (3 vols, Paris: Les Belles Lettres, 1935–43). Hermogenes, On Issues (Peri Staseōn), in Opera, ed. Hugo Rabe (Leipzig: Teubner, 1913). English translation in Malcolm Heath, Hermogenes, On Issues: Strategies of Argument in Later Greek Rhetoric (Oxford: Clarendon Press, 1995). ps.-Hermogenes, On Invention (Peri Eurēseōs), in Opera, ed. Hugo Rabe (Leipzig: Teubner, 1913). English translation in George A. Kennedy, Invention and Method: Two Rhetorical Treatises from the Hermogenic Corpus Translated with Introduction and Notes (Atlanta: Society of Biblical Literature, 2005). Julian, Panegyrics of Constantius, in Oeuvres complètes, ed. Joseph Bidez (Paris: Les Belles Lettres, 1932). Ktesias, Persika, ed. Dominique Lenfant (Paris: Les Belles Lettres, 2004). Libanios, Autobiography (Oration 1), ed. and trans. Albert F. Norman (New York and London: University of Hull, 1965). Libanios, Discours II–X, ed. Jean Martin (Paris: Les Belles Lettres, 1988). Libanios, Opera, ed. Richard Foerster (12 vols, Leipzig: Teubner, 1903– 27). Libanios, Progymnasmata, in Opera, ed. Richard Foerster, vol. 8 (Leipzig: Teubner, 1915). ps.-Longinos, On the Sublime, ed. Donald A. Russell (Oxford: Clarendon Press, 1964). Lucian, Opera, ed. Malcolm D. Macleod (4 vols, Oxford: Clarendon Press, 1972–87). Menander Rhetor, On Epideictic, ed. and trans. Donald A. Russell and Nigel G. Wilson (Oxford: Clarendon Press, 1981). Nikolaos, Progymnasmata, ed. Jospeh Felten (Leipzig: Teubner, 1913). Philostratos, Eikones and Lives of the Sophists, in Opera, ed. C.L. Kayser (vol. 2, Leipzig: Teubner, 1871). Quintilian, Institutio oratoria, ed. Michael Winterbottom (Oxford: Clarendon Press, 1970).

Bibliography

217

[Quintilian], Major Declamations = Declamationes XIX Maiores Quintiliano falso ascriptae, ed. L. �� Håkanson�� (Stuttgart, 1982). Rhetores latini minores, ed. Karl Halm (Leipzig: Teubner, 1863). Romanos, Peri aneimenou, ed. Walter Camphausen (Leipzig: Teubner, 1922). Sardianos, Ioannes, Commentarium in Aphthonium, ed. Hugo Rabe (Leipzig: Teubner, 1928). Scholia Demosthenica, ed. Mervin R. Dilts (2 vols, Leipzig: Teubner, 1983– 86). Scholia graeca in Homeri Iliadem, ed. Hartmut Erbse (7 vols, Berlin: De Gryter, 1969–88). Scholia in Euripidem, ed. Eduard Schwartz (2 vols, Berlin: G. Reimer, 1891). Sopatros, Diairesis Zētēmatōn (On the Division of Questions), in Rhetores graeci, ed. Christian Walz, vol. 8 (Stuttgart: Sumptibus J.G. Cottae, 1835). Souda = Suidae Lexicon, ed. Ada Adler (5 vols, Stuttgart: Teubner, 1928– 38). Suetonius, De grammaticis et rhetoribus, ed., trans. and comm. Robert A. Kaster (Oxford: Clarendon Press, 1995). Syrianos, Commentaria in Hermogenem, ed. Hugo Rabe (2 vols, Leipzig: Teubner, 1892–93). Theon, Progymnasmata, ed. and trans. Michel Patillon and G. Bolognesi (Paris: Les Belles Lettres, 1997). Tiberios, De figuris Demosthenicis, ed. Giulio Ballaira (Rome: Ateneo, 1968). Secondary Sources Adam, Jean-Michel, La Description (Paris: Presses Universitaires de France, 1993). Amato, Eugenio (ed.), Approches de la Troisième Sophistique: hommages à Jacques Schamp (Brussels: Editions Latomus, 2006). Armisen, Mireille, ‘La notion d’imagination chez les anciens I: les philosophes’, Pallas, 27 (1979): 3–37. Armisen, Mireille, ‘La notion d’imagination chez les anciens II: la rhétorique’, Pallas, 26 (1980): 11–51. Atherton, Catherine, The Stoics on Ambiguity (Cambridge: Cambridge University Press, 1993). Aygon, Jean-Pierre, ‘L’ecphrasis et la notion de description dans la rhétorique antique’, Pallas, 41 (1994): 41–56. Aygon, Jean-Pierre, ‘“Imagination” et description chez les rhéteurs’, Latomus, 63 (2004): 108–23.

218


Aygon, Jean-Pierre, Pictor in fabula: l’ecphrasis-descriptio dans les tragédies de Sénèque (Brussels: Editions Latomus, 2004). Babut, Daniel, ‘Sur la notion d’“imitation” dans les doctrines esthétiques de la Grèce classique’, REG, 98 (1985): 72–92. Bal, Mieke, Narratology: Introduction to the Theory of Narrative (Toronto: University of Toronto Press, 1985). Baldwin, Charles S., Medieval Rhetoric and Poetic (New York: Macmillan, 1928). Barnes, Jonathan et al. (eds), Articles on Aristotle 4: Psychology and Aesthetics (London: Duckworth, 1979). Barthes, Roland, ‘Rhétorique de l’image’, Communications, 4 (1964): 40– 51. Barthes, Roland, ‘L’ancienne rhétorique: aide-mémoire’, Communications, 16 (1970): 172–223. Barthes, Roland, ‘L’effet de réel’, in Barthes et al. (eds), Littérature et réalité (Paris: Seuil, 1982), originally published in Communications, 8 (1968). Barthes, Roland et al. (eds), Littérature et réalité (Paris: Seuil, 1982). Bartsch, Shadi, Decoding the Ancient Novel: The Reader and the Role of Description in Heliodorus and Achilles Tatius (Princeton: Princeton University Press, 1989). Bartsch, Shadi, ‘“Wait a Moment, Phantasia”: ekphrastic interference in Seneca and Epictetus’, CP, 102 (2007) (= Shadi Bartsch and Jaś Elsner (eds), Ekphrasis): 83–95. Barwick, Karl, ‘Die Gliederung der Narratio in der rhetorischen Theorie und ihre Bedeutung für die Geschichte des antiken Romans’, Hermes, 63 (1928): 261–87. Baxandall, Michael, Patterns of Intention: On the Historical Explanation of Pictures (New Haven: Yale University Press, 1985). Beall, Stephen M., ‘Word-painting in the “Imagines” of the Elder Philostraus’, Hermes, 121 (1993): 350–63. Beaujour, Michel, ‘Some paradoxes of the descriptive’, Yale French Studies, 61 (1981): 27–59. Becker, Andrew S., ‘The Shield of Achilles and the poetics of Homeric description’, AJP, 111 (1990): 139–53. Becker, Andrew S., ‘Reading poetry through a distant lens: ecphrasis, ancient Greek rhetoricians and the pseudo-Hesiodic “Shield of Herakles”’, AJP, 113 (1992): 5–24. Becker, Andrew S., The Shield of Achilles and the Poetics of Ekphrasis (Lanham: Rowman and Littlefield, 1995). Belting, Hans, Likeness and Presence: A History of the Image before the Era of Art (Chicago: University of Chicago Press, 1994). Bertrand, Edouard, Un critique d’art dans l’antiquité: Philostrate et son école (Paris: E. Thorin, 1881).

Bibliography

219

Berry, Dominic and Malcolm Heath, ‘Oratory and declamation’, in Stanley E. Porter (ed.), Handbook of Classical Rhetoric in the Hellenistic Period (330 B.C.–A.D.400) (Leiden: Brill, 1997), pp. 393–420. Bompaire, Jacques, Lucien écrivain: imitation et création (Paris: De Boccard, 1958). Bompaire, Jacques, ‘Questions de rhétorique, I: Image, métaphore, imagination dans la théorie littéraire grecque’, BAGB (1977): 355–59. Bouffartigue, Jean, L’Empereur Julien et la culture de son temps (Paris: Institut des Etudes Augustiniennes, 1992). Bougot, Auguste, Philostrate l’Ancien: une galérie antique (Paris: Renouard, 1881). Bowersock, Glen, Fiction as History: From Nero to Julian (Berkeley: University of California Press, 1994). Bowie, Ewen L., ‘Greeks and their past in the Second Sophistic’, in Moses I. Finley (ed.), Studies in Ancient Society (London: Routledge and Kegan Paul, 1974), pp. 116–209. Bowie, Ewen L., ‘The readership of Greek novels in the ancient world’, in James Tatum (ed.), The Search for the Ancient Novel (Baltimore: Johns Hopkins University Press, 1994), pp. 435–59. Boyd, Hugh Stuart, Select Passages from the Writings of St Chrysostom, St Gregory Nazianzen, and St Basil (London: John Hatchard, 1813). Brown, Peter, Power and Persuasion in Late Antiquity: Towards a Christian Empire (Madison: University of Wisconsin Press, 1992)�. Cairns, Francis, Generic Composition in Greek and Roman Poetry (Edinburgh: Edinburgh University Press, 1972). Calame, Claude, ‘Quand dire c’est faire voir: l’évidence dans la rhétorique antique’, Etudes de lettres, 4 (1991): 3–22. Calboli Montefusco, Lucia, La dottrina degli status nella retorica greca e romana (Hildesheim: Olms-Weidmann, 1986). Calboli Montefusco, Lucia, ‘Ἐvάργεια et ἐvέργεια: l’evidence d’une démonstration qui signifie les choses en acte’, in Mireille Armisen Marchetti (ed.), Demonstrare: voir et faire voir: forme de la démonstration à Rome (Toulouse: Presses universitaires du Mirail, 2005), pp. 43–58. Cameron, Averil, Christianity and the Rhetoric of Empire: The Development of Christian Discourse (Berkeley: University of California Press, 1991). Carey, Max (ed.), The Oxford Classical Dictionary (Oxford: Clarendon Press, 1949). Cassin, Barbara, L’Effet sophistique (Paris: Gallimard, 1995). Cassin, Barbara, ‘Procédures sophistiques pour construire l’évidence’, in Carlos Lévy and Laurent Pernot (eds), Dire l’évidence: philosophie et rhétorique antiques (Paris: L’Harmattan, 1997). Chanet, Anne-Marie, ‘Lysias, Discours I (Sur le meurtre d’Eratosthène)’, Lalies, 8 (1990): 100–104.

220


Chantraine, Pierre, Dictionnaire étymologique de la langue grecque: histoire des mots (Paris: Klincksieck, 1968). Chuvin, Pierre��, Chronique des derniers païens (Paris: Les Belles Lettres, 2004)�. Cockcroft, Robert, ‘Fine-tuning Quintilian’s doctrine of rhetorical emotion: seven types of enargeia’, in Tomàs Albaladejo et al. (eds), Quintiliano: Historia y actualidad de la retórica (Logroño: Gobierno de la Rioja, 1998). Cohen, David, ‘Classical rhetoric and modern theories of discourse’, in Ian Worthington (ed.), Persuasion: Greek Rhetoric in Action (London: Routledge, 1994). Connolly, Joy, ‘Problems of the past in Imperial Greek education’ in Yun Lee Too (ed.), Greek and Roman Education (Leiden: Brill, 2001), pp. 339– 72�. Cousin, Jean, Etudes sur Quintilien (2 vols, Amsterdam: P. Schippers, 1967). Cribiore, Raffaella, Gymnastics of the Mind: Greek Education in Hellenistic and Roman Egypt (Princeton: Princeton University Press, 2001). Cribiore, Raffaella, The School of Libanius in Late Antique Antioch (Princeton and Oxford: Princeton University Press, 2007). Crook, John A., Legal Advocacy in the Roman World (London: Duckworth, 1995). Curtius, Ernst, European Literature and the Latin Middle Ages (Princeton: Princeton University Press, 1953). Dangel, Jacqueline, Grammaire et rhétorique: notions de Romanité (Strasbourg: Université des sciences humaines de Strasbourg, 1994). Delarue, Fernand, ‘Suétone et l’hypotypose’, Lalies, 15 (1995): 291–300. Desbordes, Françoise, ‘L’argumentation dans la rhétorique antique: une introduction’, Lalies, 8 (1990): 81–110. Desideri, Paolo, ‘Filostrato: la comtemporaneità del passato greco’, in Fernando Gascó and Emma Falque (eds), Pasado renacido: uso y abuso de la tradición clásica (Seville: Universidad de Sevilla, 1992). Downey, Glanville, ‘Ekphrasis’, Reallexikon für Antike und Christentum (vol. 4, Stuttgart: Hiersemann, 1959). Dross, Juliette, ‘De la philosophie à la rhétorique: la relation entre phantasia et enargeia dans le traité Du sublime et l’Institution oratoire, Philosophie antique, 4 (2004): 61–93. Dross, Juliette, ‘De l’imagination à l’illusion: quelques aspects de la phantasia chez Quintilien et dans la rhétorique impériale’, Incontri triestini di filologia classica, 4 (2004–2005): 273–90. Dubel, Sandrine, ‘La beauté romanesque ou le refus du portrait dans le roman grec d’époque impériale’, in Bernard Pouderon (ed.), Les

Bibliography

221

Personnages du roman grec (Lyon: Maison de l’Orient méditerranéen, 2001), pp. 29–58. Dubel, Sandrine, ‘Dialogue et autoportrait: les masques de Lucien’, in Alain Billault (ed.), Lucien de Samosate (Lyon: Université de Lyon, 1994), pp. 19–26. Dubel, Sandrine, ‘Ekphrasis et enargeia: la description antique comme parcours’, in Carlos Lévy and Laurent Pernot (eds), Dire l’évidence: philosophie et rhétorique antiques (Paris: L’Harmattan, 1997). DuBois, Page, History, Rhetorical Description and the Epic from Homer to Spenser (Cambridge: Boydell and Brewer, 1982). DuBois, Page, ‘Reading the writing on the wall’, CP, 102 (2007): 45–56. Eden, Kathy, Poetic and Legal Fiction in the Aristotelian Tradition (Princeton: Princeton University Press, 1986). Edwards, Michael J. and Stephen Usher, Greek Orators I: Antiphon and Lysias (Warminster: Aris and Phillips, 1985). Elsner, Jaś, Art and the Roman Viewer: The Transformation of Art from the Pagan World to Christianity, (Cambridge: Cambridge University Press, 1995). Elsner, Jaś, ‘The genres of ekphrasis’, in Elsner (ed.), The Verbal and the Visual: Cultures of Ekphrasis in Antiquity = Ramus, 31 (2002): 1–8. Elsner, Jaś, ‘The rhetoric of buildings in the De aedificiis of Procopius’, in Liz James (ed.), Art and Text in Byzantine Culture (Cambridge: Cambridge University Press, 2007). Esrock, Ellen J., The Reader’s Eye: Visual Imaging as Reader Response (Baltimore: Johns Hopkins University Press, 1994). Fairweather, Janet, Seneca the Elder (Cambridge: Cambridge University Press, 1981). Fantham, Elaine R. ‘Orator and / et actor’, in Patricia E. Easterling and Edith Hall (eds), Greek and Roman Actors: Aspects of an Ancient Profession (Cambridge: Cambridge University Press, 2002). Feldherr, Andrew, Spectacle and Society in Livy’s History (Berkeley: California University Press, 1998). Fowler, Don, ‘Narrate and describe: the problem of ekphrasis’, JRS, 81 (1991): 25–35. Frede, Michael, ‘An empiricist view of knowledge: memorism’, in Stephen Everson (ed.), Companions to Ancient Thought 1: Epistemology (Cambridge: Cambridge University Press, 1990), pp. 225–50. Friedländer, Paul��, Johannes von Gaza und Paulus Silentiarius: Kunstbeschreibungen Justinianischer Zeit (Leipzig and Berlin: Teubner, 1912)�. Galand-Hallyn, Perrine, Le Reflet des fleurs: description et métalangage poétiques d’Homère à Erasme (Geneva: Droz, 1994).

222


Geissler, Wilhelm (Guilelmus), Ad descriptionum historiam symbola (Leipzig: Wiedae Thuringorum, 1916). Gell, Alfred, Art and Agency: An Anthropological Theory (Oxford: Oxford University Press, 1998). Genette, Gérard, ‘Frontières du récit’, in L’Analyse structurale du récit (Paris: Seuil, 1981) = Communications, 8 (1966). Gill, Christopher and Timothy P. Wiseman (eds), Lies and Fiction in the Ancient World (Exeter: University of Exeter, 1993). Gleason, Maud, Making Men: Sophists and Self-presentation in Ancient Rome (Princeton: Princeton University Press, 1995). Goldhill, Simon, ‘The naïve and knowing eye: ecphrasis and the culture of viewing in the Hellenistic world’, in Simon Goldhill and Robin Osborne (eds), Art and Text in Ancient Greek Culture (Cambridge: Cambridge University Press, 1994), pp. 197–223. Goldhill, Simon, ‘The erotic eye: visual stimulation and cultural conflict’, in Goldhill (ed.), Being Greek under Rome: Cultural Identity, the Second Sophistic and the Development of Empire (Cambridge: Cambridge University Press, 2001). Goldhill, Simon, ‘What is ekphrasis for?’, CP, 102 (2007): 1–19. Gourinat, Jean-Baptiste, Les Stoïciens et l’âme (Paris: Presses universitaires de France, 1996). Graf, Fritz, ‘Ekphrasis: Die Enstellung der Gattung in der Antike’, in Beschreibungskunst-Kunstbrescheibung, ed. Gottfried Boehm and Helmut Pfotenhauer (Munich: Wilhelm Fink Verlag, 1995), pp. 143–55�. Gualandri, Isabella, ‘Aspetti dell’ekphrasis in età tardo-antica’, in Testo e immagine nell’alto medioevo (Spoleto: Centro italiano di studi sull’Alto Medioevo, 1994). Guez, Jean-Philippe, ‘Achille Tatius ou le paysage-monde’, in B. Pouderon and D. Crismani (eds), Lieux, décors et paysages de l’ancien roman des origines à Byzance (Lyon: Maison de l’Orient et de la Méditerranée, 2005). Hagstrum, Jean H., The Sister Arts: The Tradition of Literary Pictorialism and English Poetry from Dryden to Gray (Chicago: Chicago University Press, 1958). Hall, Edith, ‘Lawcourt dramas: the power of performance in Greek forensic oratory’, BICS, 40 (1995): 39–58. Hamon, Philippe, ‘Qu’est-ce qu’une description?’, Poétique, 112 (1972): 465–85. Hamon, Philippe, La Description littéraire de l’antiquité à Roland Barthes: une anthologie (Paris, 1991). Hamon, Philippe, Du descriptif (Paris: Hachette, 1993). Hardie, Alex, Statius and the Silvae: Poets, Patrons and Epideixis in the GraecoRoman World (Liverpool: F. Cairns, 1983).

Bibliography

223

Hardie, Philip, ‘A reading of Heliodorus, Aithiopika 3.4.1–5.2’, in Richard Hunter (ed.), Studies in Heliodorus (Cambridge: Cambridge Philological Society, 1998). Harris, William, Ancient Literacy (Cambridge. MA: Harvard University Press, 1989). Heath, Malcolm, Hermogenes on Issues: Strategies of Argument in Later Greek Rhetoric (Oxford: Clarendon Press, 1995). Heath, Malcolm, ‘Invention’, in Stanley E. Porter (ed.), Handbook of Classical Rhetoric in the Hellenistic Period (330 B.C.–A.D. 400) (Leiden: Brill, 1997). Heath, Malcolm, ‘Theon and the history of the Progymnasmata’, GRBS, 43 (2002/3): 129–60. Heath, Malcolm, Menander: A Rhetor in Context (Oxford: Oxford University Press, 2004). Heath, Malcolm, ‘Practical advocacy in Roman Egypt’, in Michael J. Edwards and Christopher Reid (eds), Oratory in Action (Manchester: Manchester University Press, 2004), pp. 62–82. Hebert, Bernhard D., Spätantike Beschreibung von Kunstwerken: Archäologischer Kommentar zu den Ekphraseis des Libanios und Nikolaos (Graz: Dbv Verlag, 1983). Heffernan, James A.W., ‘Ekphrasis and representation’, New Literary History, 22 (1991): 297–316. Heffernan, James A.W., Museum of Words: The Poetics of Ekphrasis from Homer to Ashbery (Chicago: Chicago University Press, 1993). Henderson, Ian H., ‘Quintilian and the Progymnasmata’, Antike und Abendland, 37 (1991): 82–99. Hinks, Roger P., Myth and Allegory in Ancient Art (London: Warburg Institute, 1939). Hock, Ronald. F. and Edward N. O’Neil, The Chreia in Ancient Rhetoric: Volume I. The Progymnasmata (Atlanta: Scholars Press, 1986). Hock, Ronald. F. and Edward N. O’Neil, The Chreia in Ancient Rhetoric: Classroom Exercises (Atlanta: Society of Biblical Literature, 2002). Hollander, John, The Gazer’s Spirit: Poems Speaking to Silent Works of Art (Chicago and London: University of Chicago Press, 1995). Hornblower, Simon and Antony Spawforth (eds), The Oxford Classical Dictionary, 3rd edn (Oxford: Oxford University Press, 1996). Hunger, Herbert, Die hochsprachliche profane Literatur der Byzantiner (2 vols, Munich: Beck, 1978). Ildefonse, Frédérique, ‘Evidence sensible et discours dans le stoïcisme’, in Carlos Lévy and Laurent Pernot (eds), Dire l’évidence (Paris: L’Harmattan, 1997).

224


Innes, Doreen and Michael Winterbottom, Sopatros the Rhetor: Studies in the Text of the Diairesis Zētēmatōn (London: University of London, Institute of Classical Studies, 1988). Innes, Doreen et al. (eds), Ethics and Rhetoric: Classical Essays for Donald Russell’s Seventy-fifth Birthday (Oxford: Clarendon Press, 1995)�. Innocenti, Beth, ‘Towards a theory of vivid description as practised in Cicero’s Verrine Orations’, Rhetorica, 12 (1994): 355–81. Iser, Wolfgang, ‘The reading process: a phenomenological approach’, in J.P. Tompkins (ed.), Reader-response Criticism: From Formalism to Poststructuralism (Baltimore: Johns Hopkins University Press, 1980), pp. 50–69. James, Elizabeth, ‘Art and lies: text, image and imagination in the medieval world’, in Antony Eastmond and Liz James (eds), Icon and Word: The Power of Images in Byzantium (Aldershot: Ashgate, 2003), pp. 59–72. James, Elizabeth and Ruth Webb, ‘To understand ultimate things and enter secret places: ekphrasis and art in Byzantium’, Art History, 14 (1991): 1–17. Jones, Christopher P., ‘The Rhodian Oration ascribed to Aelius Aristides’, CQ, 40 (1990): 514–22. Kaster, Robert A., Guardians of Language: The Grammarian and Society in Late Antiquity (Berkeley: University of California Press, 1988). Kaster, Robert A., ‘Controlling reason: declamation in rhetorical education at Rome’, in Yun Lee Too (ed.), Education in Greek and Roman Antiquity (Leiden: Brill, 2001). Kazhdan, Alexander (ed.), The Oxford Dictionary of Byzantium (New York: Oxford University Press, 1991). Kennedy, George A., Greek Rhetoric under Christian Emperors (Princeton: Princeton University Press, 1983). Kennedy, George A., Progymnasmata: Greek Textbooks of Prose Composition and Rhetoric (Atlanta: Society of Biblical Literature, 2003). Kennedy, George A., Invention and Method: Two Rhetorical Treatises from the Hermogenic Corpus Translated with Introduction and Notes (Atlanta: Society of Biblical Literature, 2005). Koelb, Janice Hewlett, The Poetics of Description: Imagined Places in European Literature (New York: Palgrave Macmillan, 2006). Kohl, Richard, De Scholasticarum declamationum argumentis ex historia petitis (Paderborn: Ferdinand Schoeningh, 1915) (= Rhetorische Studien, 4). Krieger, Murray, Ekphrasis: The Illusion of the Natural Sign (Baltimore: Johns Hopkins University Press, 1992). Labarrière, Jean-Louis, ‘Faut voir à voir! Considérations pseudoLonginiennes’, Métis, n.s. 4 (2006): 71–93. Lada-Richards, Ismene, Silent Eloquence: Lucian and Pantomime Dancing (London: Duckworth, 2007).

Bibliography

225

Laird, Andrew, ‘Sounding out ekphrasis: art and text in Catullus 64’, JRS, 83 (1993): 18–30. Laird, Andrew, ‘Ut figura poiesis: writing art and the art of writing in Augustan poetry’, in Jas Elsner (ed.), Art and Text in Roman Culture (Cambridge: Cambridge University Press, 1996). Lamberton, Robert, Homer the Theologian: Neoplatonist Allegorical Reading and the Growth of the Epic Tradition (Berkeley: University of California Press, 1986). Lana, Italo, Quintiliano, Il ‘Sublime’ e gli ‘Esercizi preparatori’ di Elio Teone: ricerca sulle fonti grece di Quintiliano e sull’autor ‘Del Sublime’ (Turin: Università di Torino, 1951). Lana, Italo, I Progimnasmi di Elio Teone I: La storia del testo (Turin: Università di Torino, 1959). Laurens, Pierre, Introduction to Hermogenes, Art of Rhetoric, trans. and comm. Michel Patillon (Paris: L’Age d’homme, 1997). Lausberg, Heinrich, Handbook of Literary Rhetoric: A Foundation for Literary Study, trans. Matthew T. Bliss et al. (Leiden: Brill, 1998). Leach, Eleanor W., The Rhetoric of Space: Literary and Artistic Representations of Landscape in Republican and Augustan Rome (Princeton: Princeton University Press, 1988). Leigh, Matthew, ‘Quintilian on the emotions’, JRS, 94 (2004): 122–40. Létoublon, Françoise, Les Lieux communs du roman: stéréotypes grecs d’aventure et d’amour (Leiden: Brill, 1993). Long, Anthony A., Hellenistic Philosophy: Stoics, Epicureans, Sceptics (Berkeley: University of California Press, 1986). Long, Anthony A., Stoic Studies (Cambridge: Cambridge University Press, 1996). Long, Anthony A. and David N. Sedley, The Hellenistic Philosophers (2 vols, Cambridge: Cambridge University Press, 1987). Lübker, Friedrich, Reallexikon des Klassischen Altertums (Leipzig and Berlin: Teubner, 1914). Lukács, Georg, ‘Narrate or describe?’, in Writer and Critic, ed. A. Kahn (London: Merlin Press, 1978). Ma, John, ‘Public speech and community in the Euboicus’, in Simon Swain (ed.), Dio Chrysostom: Politics, Letters, and Philosophy (Oxford: Oxford University Press, 2002). Macdonald, R.A., Review essay of M. Krieger, Ekphrasis, Word and Image, 9 (1993): 81–6. Macrides, Ruth and Paul Magdalino, ‘The architecture of ekphrasis: construction and context of Paul the Silentiary’s poem on Hagia Sophia’, BMGS, 12 (1988): 47–82. Maeder, Danielle, ‘Au seuil des romans Grecs: effets de réel et effets de création’, in Groningen Colloquia on the Novel 4 (Groningen: Egbert

226


Forsten, 1991), pp. 1–33. Maguire, Henry, ‘Truth and convention in Byzantine descriptions of works of art’, DOP, 28 (1974): 113–40. Maguire, Henry, Art and Eloquence in Byzantium (Princeton: Princeton University Press, 1981). Manieri, Alessandra, L’immagine poetica nella teoria degli antichi: phantasia ed enargeia (Pisa: Istituti editoriali e poligrafici internazionali, 1998). Manieri, Alessandra, ‘Colori, suoni e profumi nelle Imagines: principi dell’estetica filostratea’, Quaderni urbinati di cultura classica, n.s. 63/3 (1999): 111–21. Margolin, Jean-Claude, ‘La rhétorique d’Aphthonius et son influence au XVIe siècle’, in Raymond Chevallier (ed.), Colloque sur la rhétorique (Paris: Les Belles Lettres, 1979) (=Caesarodunum, 14bis), pp. 239–69. Marrou, Henri-Irénée, A History of Education in Antiquity (Madison: University of Wisconsin Press, 1982). Martin, Richard, The Language of Heroes: Speech and Performance in the Iliad (Ithaca: Cornell University Press, 1989). Matoré, Georges, Le Vocabulaire de la prose littéraire de 1835 à 1845: Théophile Gautier et ses premières oeuvres en prose (Geneva and Lille: Droz-Giard, 1951). McCombie, Duncan, ‘Philostratus, Histoi, Imagines, 2.28: ekphrasis and the web of illusion’, in Jaś �� Elsner (ed.), The Verbal and the Visual: Cultures of Ekphrasis in Antiquity = Ramus, 31 (2002): 146–57. Meijering, Roos, Literary and Rhetorical Theories in Greek Scholia (Groningen: Egbert Forsten, 1987). Méridier, Louis, L’Influence de la Seconde Sophistique sur l’œuvre de Gregoire de Nysse (Paris: Hachette, 1906). Mitchell, Stephen, ‘Festivals, games and civic life in Roman Asia Minor’, JRS, 80 (1990): 183–93. Molinié, Georges, Dictionnaire de rhétorique (Paris: Livre de Poche, 1992). Molino, Jean, ‘Logiques de la description’, Poétique, 91 (1992): 363–82. Montanari, Franco, ‘Ekphrasis e verità storica in Luciano’, in Filologia e critica letteraria della Grecità (= Ricerche di filologia classica, 2) (Pisa: Giardini editori, 1984). Morales, Helen, Vision and Narrative in Achilles Tatius’ Leucippe and Clitophon (Cambridge: Cambridge University Press, 2005). Morgan, John R., ‘History, romance and realism in the Aithiopika of Heliodorus’, Classical Antiquity, 1 (1981): 221–65. Morgan, John R., ‘Reader and audiences in the Aithiopika of Heliodoros’, Groningen Colloquia on the Novel 4 (Groningen: Egbert Forsten, 1991), pp. 85–104. Morgan, John R., ‘Make-believe and make believe’, in Christopher Gill and T.P. Wiseman (eds), Lies and Fiction in the Ancient World (Exeter:

Bibliography

227

Exeter University Press, 1993). Morgan, Teresa, Literate Education in the Hellenistic and Roman Worlds (Cambridge: Cambridge University Press, 1998). Morpurgo-Tagliabue, Guido, Linguistica e stilistica di Aristotele (Rome: Edizioni dell’Ateneo, 1967). Moussy, Claude, ‘Signum et les noms latins de la preuve: l’héritage de divers termes grecs’, in Jacqueline Dangel (ed.), Grammaire et rhétorique: notions de romanité (Strasbourg: Université des sciences humaines de Strasbourg, 1994). Muñoz, Antonio, ‘Le ἐκφράσεις nella letteratura bizantina e i loro rapporti con l'arte figurata’, in Recueil d’études dédiées à la mémoire de N.P. Kondakov (Prague: Seminarium Kondakovianum, 1926). Murphy, James J., ‘The key role of habit in Roman rhetoric and education as described by Quintilian’, in Tomàs Albaladejo et al. (eds), Quintiliano: historia y actualidad de la retórica (Logroño: Gobierno de la Rioja, 1998). Newby, Zahra, �� ‘Testing the boundaries of ekphrasis’, in Jaś Elsner (ed.), The Verbal and the Visual: Cultures of Ekphrasis in Antiquity = Ramus, 31, 1–2 (2002): 126–35�. O’Daly, Gerard, Augustine’s Philosophy of Mind (London: Duckworth, 1987). Ong, Walter J., Orality and Literacy: The Technologizing of the Word (London and New York: Routledge, 1988). Palm, Jonas, ‘Bemerkungen zur Ekphrase in der griechischen Literatur’, Kungliga Humanististiska Vetenskapssamfundet i Uppsala, Årsbok (1965– 66): 108–211. Parks, E. Patrick, The Roman Rhetorical Schools as a Preparation for the Courts under the Early Empire (Baltimore: Johns Hopkins Press, 1945). Patillon, Michel, La Théorie du discours chez Hermogène le Rhéteur: essai sur les structures linguistiques de la rhétorique ancienne (Paris: Les Belles Lettres, 1988). Patillon, Michel (trans. and comm.), Hermogène: l’art rhétorique (Paris: L’Age d’homme, 1997). Paul, George M., ‘Urbs capta: sketch of an ancient literary motif’, Phoenix, 36 (1982): 144–55. Perelman, Chaïm, L’Empire rhétorique: rhétorique et argumentation (Paris��: Vrin, 1977). Pernot, Laurent, Les Discours siciliens d’Aelius Aristide (Or 5–6): étude littéraire et paléographique, édition et traduction (New York: Arno Press, 1981). Pernot, Laurent, ‘Lieu et lieu commun dans la rhétorique antique’, BAGB, (1986): 253–84. Pernot, Laurent, La Rhétorique de l’éloge dans le monde gréco-romain (2 vols, Paris: Institut d’Etudes augustiniennes, 1993).

228


Pernot, Laurent, ‘La rhétorique�� de l’empire ou comment la rhétorique� �� grecque a inventé l’empire romain’, Rhetorica, 16 (1998): 131–48. Pernot, Laurent, La Rhétorique dans l’antiquité (Paris: Librairie générale française, 2000). Perutelli, Alessandro, ‘L’inversione speculare: per una retorica dell’ekphrasis’, MD, 1 (1978): 87–98. Petit, Paul, Les Etudiants de Libanius (Paris: Nouvelles Editions Latines, 1957). Piot, Henri, Les Procédés littéraires de la IIe Sophistique chez Lucien: l’ecphrasis (Rennes, 1914). Porter, Stanley E. (ed.), Handbook of Classical Rhetoric in the Hellenistic Period (330B.C.–A.D.400) (Leiden: Brill, 1997). Rabau, Sophie, ‘Narration et description: l’exigence de détails’, Lalies, 15 (1995): 273–90. Rabe, Hugo, ‘Aus Rhetoren-Handschriften’, RhMus, 64 (1909): 538–90. Race, William H., ‘Rhetoric and lyric poetry’, in Ian Worthington (ed.), A Companion to Greek Rhetoric (Oxford: Blackwell, 2007). Reardon, Brian, Courants littéraires grecs des IIe et IIIe siècles après J.-C. (Paris: Les Belles Lettres, 1971). Reichel, Georg, Quaestiones Progymnasmaticae (Leipzig: Typis Roberti Noske Bornensis, 1909). Rispoli, Gioia, Lo spazio del verisimile: il racconto, la storia e il mito (Naples: M. D’Auria, 1988). Roberts, Michael, The Jeweled Style: Poetry and Poetics in Late Antiquity (Ithaca: Cornell Unversity Press, 1989). Rohde, Erwin, Der griechische Roman und seine Vorläufer (Leipzig: Bretitkopf und Härtel, 1900). Roueché, Charlotte M., ‘The functions of the Roman governor in late antiquity’, Antiquité tardive, 6 (1998): 83–9. Rousselle, Aline, ‘Images as education in the Roman Empire’, in Yun Lee Too (ed.), Education in Greek and Roman Antiquity (Leiden, 2001), pp. 373–403. Rouveret, Agnès, Histoire et imaginaire de la peinture ancienne (Ve s. av. J.-C.Ier s. ap. J.-C.) (Rome: Ecole française de Rome, 1989). Russell, Donald A., ‘Rhetors at the wedding’, PCPS, 205 (1979): 104–17. Russell, Donald A., Greek Declamation (Cambridge: Cambridge University Press, 1983). Russell, Donald A., Libanius: Imaginary Speeches (London: Duckworth, 1996) Russell, Donald A., ‘The panegyrists and their teachers’, in Mary Whitby (ed.), The Propaganda of Power: The Role of Panegyric in Late Antiquity (Leiden: Brill, 1998).

Bibliography

229

Rutherford, Ian, ‘Inverting the canon: Hermogenes on literature’, HSCP, 94 (1992): 355–78. Rutherford, Ian, Canons of Style in the Antonine Age: Idea-Theory in Its Literary Context (Oxford: Clarendon Press, 1998). Ryder, Timothy T.B., ‘Demosthenes and Philip II’, in Ian Worthington (ed.), Demosthenes: Statesman and Orator (London: Routledge, 2000). Saïd, Suzanne, ‘Le “je” de Lucien’, in Marie-Françoise Baslez et al. (eds), L’Invention de l’autobiographie: d’Hésiode à Saint Augustin (Paris: Presses de l’Ecole normale supérieure, 1993), pp. 253–70. Scarry, Elaine, Dreaming by the Book (New York: Farrar, Straus and Giroux, 1999). Schaeffer, Jean-Marie, Pourquoi la fiction? (Paris: Seuil, 1999). Schechner, Richard, Performance Theory (New York: Routledge, 1988). Schissel von Fleschenberg, Othmar, ‘Die Technik des Bildeinsatzes’, Philologus, 72 (1913): 83–114. Schmitz, Thomas, Bildung und Macht: Zur sozialen und politischen Funktion der zweiten Sophistik in der griechischen Welt der Kaiserzeit (Munich: C.H. Beck, 1997). Schmitz, Thomas, ‘Performing history in the Second Sophistic’, in M. Zimmermann (ed.), Geschichtsschreibung und politischer Wandel im 3. Jh. N. Chr. (Stuttgart: F. Steiner, 1999). Schofield, Malcolm, ‘Aristotle on the imagination’, in Jonathan Barnes et al. (ed.), Articles on Aristotle 4: Psychology and Aesthetics (London: Duckworth, 1979), pp. 103–32. Schouler, Bernard, ‘Personnes, faits et états de la cause dans le système d’Hermogénès’, Lalies, 8 (1990): 111–27. Schouler, Bernard, ‘Un enseignant face aux prisons de son temps’, Pallas, 72 (2006): 279–96. Schrijvers, Piet H. (1982), ‘Invention, imagination et théories des émotions chez Cicéron et Quintilien’, in Actus: Festschrift H.L.W. Nelson (Utrecht: Instituut voor Klassieke Talen, 1982), pp. 395–408 Scott, Grant F., ‘The rhetoric of dilation: ekphrasis and ideology’, Word and Image, 7 (1991): 301–10. Scott, Grant F., The Sculpted Word: Keats, Ekphrasis, and the Visual Arts (Hanover: University Press of New England, 1994). Seznec, Jean, ‘Flaubert and the graphic arts’, JWCI, 8 (1945): 175–90. Small, Jocelyn Penny, Wax Tablets of the Mind: Cognitive Studies of Memory and Literacy in Classical Antiquity (London: Routledge, 1997). Sorabji, Richard, Aristotle on Memory (London: Duckworth, 1972). Spitzer, Leo, ‘Wiederum Mörikes Gedicht “Auf eine Lampe”’, Trivium, 9 (1951): 203–25. Spitzer, Leo, ‘Garcilaso, Third Eclogue, Lines 265–271’, Hispanic Review, 20 (1952): 243–8.

230


Spitzer, Leo, ‘The “Ode on a Grecian Urn,” or content vs. metagrammar’, Comparative Literature, 7 (1955): 203–25 (repr. in Spitzer, Essays on English and American Literature, ed. Anna Hatcher (Princeton: Princeton University Press, 1962), pp. 67–97 (= edition cited). Stephens, Susan A., ‘The “Arginusae” theme in Greek rhetorical theory and practice’, BASP, 20 (1983): 171–80. Sussman, Lewis A., The Major Declamations Ascribed to Quintilian: A Translation (Frankfurt: P. Lang, 1987). Swain, Simon, Hellenism and Empire: Language, Classicism and Power in the Greek World, AD 50–250 (Oxford: Clarendon Press, 1996). Temmerman, Koen de, ‘Blushing beauty: characterizing blushes in Chariton’s Callirhoe’, Mnemosyne, 60 (2007): 235–52. Tompkins, Jane P., ‘The reader in history: the changing shape of literary response’, in Tompkins (ed.), Reader-response Criticism: From Formalism to Post-structuralism (Baltimore: Johns Hopkins University Press, 1980). Too, Yun Lee, The Idea of Ancient Literary Criticism (Oxford: Clarendon Press, 1998). Too, Yun Lee (ed.), Education in Greek and Roman Antiquity (Leiden: Brill, 2001). Trapp, Michael B., ‘Sense �� of place in the orations of Dio Chrysostom’, in D. Innes et al. (eds), Ethics and Rhetoric: Classical Essays for Donald Russell on His Seventy-fifth birthday (Oxford: Clarendon Press, 1995).� Vasaly, Ann, Representations: Images of the World in Ciceronian Oratory (Berkeley: University of California Press, 1993). Walker, Andrew D., ‘Enargeia and the spectator in Greek historiography’, TAPA, 123 (1993): 353–77. Watson, Gerard, Phantasia in Classical Thought (Galway: Galway University Press, 1988). Webb, Ruth,�� ‘Imagination and the arousal of the emotions in GrecoRoman rhetoric’, in S. Braund and C. Gill (eds), The Passions in Roman Thought and Literature (Cambridge: Cambridge University Press, 1997), pp. 112–27. Webb, Ruth, ‘Mémoire et imagination: les limites de l’enargeia dans la théorie rhétorique grecque’, in Carlos Lévy and Laurent Pernot (eds), Dire l’évidence: philosophie et rhétorique antiques (Paris: L’Harmattan, 1997), pp. 229–48. Webb, Ruth, ‘Rhetoric and poetry’, in Stanley E. Porter (ed.), Handbook of Classical Rhetoric in the Hellenistic Period (330 B.C.–A.D. 400) (Leiden: Brill, 1997), pp. 339–69. Webb, Ruth, ‘Ekphrasis ancient and modern: the invention of a genre’, Word and Image, 15 (1999): 7–18.

Bibliography

231

Webb, Ruth, ‘The aesthetics of sacred space: narrative, metaphor and motion in ekphraseis of church buildings’, DOP, 53 (1999): 59–74. Webb, Ruth,�� ‘Ekphrasis, amplification and persuasion in Procopius’ Buildings’, Antiquité tardive, 8 (2000): 67–71. Webb, Ruth, ‘The Progymnasmata as practice’, in Yun Lee Too (ed.), Education in Greek and Roman Antiquity (Leiden: Brill, 2001), pp. 289–316�. Webb, Ruth, ‘Praise and persuasion: argumentation and audience response in epideictic oratory’, in Elizabeth Jeffreys (ed.), Rhetoric in Byzantium (Aldershot: Ashgate, 2003), pp. 127–35. Webb, Ruth, ‘Fiction, mimesis and the performance of the Greek past in the Second Sophistic’, in David Konstan and Suzanne Saïd (eds), Greeks on Greekness: Viewing the Greek Past under the Roman Empire (Cambridge: Cambridge Philological Society, 2006), pp. 27–46. Webb, Ruth, ‘The Imagines as fictional text: ekphrasis, apatê and illusion’, in M. Costantini et al. (eds), Le Défi de l’art: Philostrate, Callistrate et l’image sophistique (Rennes: Presses universitaires de Rennes, 2006), pp. 113–36. Webb, Ruth, ‘The model ekphraseis of Nikolaos the Sophist as memory images’, in Michael Grünbart (ed.), Theatron: Rhetorische Kultur in Spätantike und Mittelalter (Berlin: de Gruyter, 2007), pp. 463–75. Whitmarsh, Tim, ‘The writes of passage: cultural initiation in Heliodorus’ Aethiopica’, in Richard Miles (ed.), Constructing Identities in Late Antiquity (London: Routledge, 1999). Whitmarsh, Tim, Greek Literature and the Roman Empire (Oxford: Oxford University Press, 2001). Whitmarsh, Tim, ‘Written on the body: ekphrasis, perception and deception in Heliodorus’ Aethiopica’, in �� Jaś Elsner (ed.), The Verbal and the Visual: Cultures of Ekphrasis in Antiquity = Ramus, 31, 1–2 (2002): �� 111–25. Winkler, John J., ‘The mendacity of Kalasiris and the narrative strategy of Heliodorus’ Aithiopika’, Yale Classical Studies, 27 (1982): 93–158. Winterbottom, Michael, ‘Schoolroom and courtroom’, in Brian Vickers (ed.), Rhetoric Revalued (Binghampton: Center for Medieval and Early Renaissance Studies, 1982). Yates, Frances, The Art of Memory (London: Routledge and Kegan Paul, 1966). Zanker, Graham, ‘Enargeia in the ancient criticism of poetry’, RhMus, 124 (1981): 297–311. Zanker, Graham, Realism in Alexandrian Poetry: A Literature and Its Audience (London: Croom Helm, 1987). Zanker, Graham, ‘New light on the “ekphrastic epigram”’, ZPE, 143 (2003): 59–62. Zeitlin, Froma I., ‘The artful eye: vision, ekphrasis and spectacle in Euripidean theatre’, in Simon Goldhill and Robin Osborne (eds), Art

232


and Text in Ancient Greek Culture (Cambridge: Cambridge University Press, 1994), pp. 138–96. Zeitlin, Froma I., ‘Visions and revisions of Homer in the Second Sophistic’, in Simon Goldhill (ed.), Being Greek under Rome: Cultural Identity, the Second Sophistic and the Development of Empire (Cambridge: Cambridge University Press, 2001).

Index

Achilles 23, 40 n.4, 43; see also Shield of Achilles Achilles Tatios, Leukippe and Kleitophon 168, 178–180, 181, 183–185, 194 actor, orator compared to 100, 104, 175–6 Ailios Aristeides 65, 147 Egyptian Oration 180 Letter to the Emperors 161–3, 171 Rhodian Oration 159 Smyrnaean Oration 174 On Rome 174 n.18 Aeschines 25, 153 3.157: 74 n.36, 114 aitia (‘cause’) 63, 64–5, 82, 181, 203 Ajax 43, 66, 68, 69, 72, 82–3, 209–11 Alexander 15, 142, 148–9, 151, 174 amplification 76, 77–8, 93 n.16, 131, 152, 155, 164, 169–70, 204 anaskeuē see refutation aphēgēmatikos 51, 55, 58, 76, 203 Aphthonios 14, 44, 46, 48, 50–52, 55–6, 57–9, 61–2, 64, 66, 79–81, 127, 136, 156, 201–2 Apsines 74 n.36, 163 n.97 Aristotle 134, 136 On Memory and Recollection 449b–450a 111–12, 114 Poetics 1455a 97 Rhetoric 1411b 26, 51, 85; 1417a 68 On the Soul 427b 110 n.10, 112, 114; 432a 128 Attic dialect 14, 15, 17–18, 39 Augustine 16 De Rhetorica 64 n.10 De Trinitate 119, 121–2, 161, 183 auxēsis 76, 131, 155; see also amplification Barthes, Roland 13, 68, 124

battles inepideictic 164 as subject of ekphrasis inthe Progymnasmata 56–58, 62, 139 Bemarchios 125–6, 130 buildings see ekphrasis Callistratus see Kallistratos character, depiction of 15, 22, 43, 68, 82–3, 92, 121, 123, 130, 132, 153, 176, 182, 184, 203 Cicero 21, 91, 92–3, 94, 100–101, 107–110, 123, 148, 153, 179, 187 circumstantia see peristaseis city, description of 54–5, 61, 73–4, 91–3, 96, 109, 114–15, 121–2, 124, 128, 142, 145, 148–9, 155–7, 159, 160–63, 170, 174, 179, 205 common place see koinos topos competence, cultural 110, 124–5 confirmation (elementary exercise) 43, 45, 48, 78, 80, 199 Constantius 125, 159, 164 n.100, 182 n.37 crocodiles 9, 61, 180, 181, 197 declamation 4, 10, 13, 15, 16–17, 19, 25–6, 80, 100, 129–30, 131–3 ekphrasis in74, 78, 85, 139–152 and fiction 154–5, 175–7 relation to Progymnasmata 47–9 deinōsis 76, 99–100, 131 deliberative oratory 4, 51, 76, 80, 129, 131, 142, 144–5, 203 Demetrios, On Style 75 n.42, 92–3 Demosthenes 25, 63, 115–16 indeclamation 15, 26, 142, 148–9, 151, 155, 176 Or. 18.169–171: 89, 90

234


Or. 19.65: 51 n.51, 74, 76, 89, 100 n.32, 114–15, 152–3, 204 Or. 24.208: 101 n.35 descriptio 9, 51, 75, 87 description 7–9, 62, 66–9, 70, 74, 84–6, 105–6, 124, 129, 138, 194 diagraphē 51 diagraphō 77, 156, 157, 159 diaskeuē 72, 74 n.36, 102 n.41, 140 n.31, 142, 143, 144, 162; see also endiaskeuos diatupōsis 51–2, 55 n. 67, 65 n.14, 72, 74 n.36, 77, 89 n.6, 93 n.16, 100, 101 n.35, 112, 126, 142, 143, 156, 158, 162, 171 diēgēsis (‘narration’) 42, 54, 55, 63–67, 70–78, 82, 85, 138, 184, 198, 203, 207; see also peristaseis Dio Chrysostom 22, 134 n.12, 136, 156 n.73 Dionysios of Halikarnassos On Imitation 39, 57 Letter to Pompeios Geminos 39 Lysias 22, 24, 91–2, 124 Dionysios of Halikarnassos, ps.– On Epideictic 47–8, 133, 137, 157–8 On Mistakes inDeclamation 84–5, 139–41, 154–5 Doxapatres 44, 69–70, 81 n.58, 127, 207–9 dreams 95, 104, 116, 122 n.53, 128 n.69, 170, 178 education 14–17, 25, 40–42, 45, 47–9, 84–5, 126, 139; see also declamation; grammar; Progymnasmata eikos 83, 102 n.41, 168 ekphrasis of buildings 1, 2, 11, 33, 58, 62, 81, 155–7, 174–5, 195 and diēgēsis 62–72, 75–8, 85 and enkōmion 78–81, 137 of festivals 56, 61, 64, 78–9, 137, 157, 184, 197, 200, 203, 213 and interpretation 58, 75–6, 83, 145–54, 174–5

and koinos topos, 50, 76–78, 126–7 modern definition of 1–3, 5–9, 11, 28–38 inthe novel 178–85 of persons 61, 62, 66, 76, 81–2, 92, 127, 138, 157–8, 170–71, 178 n.28 and persuasion 143–55, 160–164 as rhetorical technique 84–6, 165 of storms 65, 139, 142, 146–8, 154–5, 176–7, 179, 197, 213 theory and practice 138–9 of the tropos 64–5, 67, 69–72, 82, 174, 181–2, 189, 197–8 of works of art 46, 55, 62, 81–4, 172–3,185–90,194–5 see also battles; declamation; enargeia; epideictic; fiction; kairos; painting; poetry enargeia 5, 10, 14, 21–2, 26, 27, 38, 48, 51, 52–3, 59, 70–4, 81–3, 87–130, 138–40, 142, 145, 149, 151–3 and argumentation 123–4 and emotion 88, 90, 94–101 and fiction 179, 184–5 inhistoriography 20, 140–41 as illusion 168–9 inpoetry 22–3 endiaskeuos (‘elaborate’) 66, 71–2, 74, 78, 137 n.24, 207–210; see also diaskeuē energeia 26, 85–6 enkataskeuos (‘confirmatory’) 66, 207–210; see also kataskeuē epibatērios logos (Speech of Arrival) 156, 169 epideictic 4, 10–11, 15, 40, 46, 47– 9,133–7, 167 ekphrasis in55, 69, 76, 78–81, 99, 155–64, 169–75, 182, 203–4 epithalamios 157; see also wedding speech Erinyes see Furies ēthopoiia 43, 142, 144, 176, 188, 203, 207, 208 ēthos 21, 82; see also character Euripides 31

Index Hippolytos 120–1, 180 Orestes 97, 113 explicatio 51, 75, 147 n.53, 153 n.67 fiction 104, 120 declamation and 15, 154–5, 175–7 ekphrasis and 10, 75–6, 168–9, 178–85, 187–9, 194 see also plasma Friedländer, Paul 11, 31–3 funeral oration see monody Furies 97, 102, 116, 118, 169, 183 gamēlios logos 157; see also wedding speech Gautier, Théophile 29, 34 Gorgias 5, 111 grammar, teaching of 42–3, 85 Hamon, Philippe 74, 105–6, 124; see also description Heliodoros, Aithiopika 168, 178, 181–83, 184–85, 188 Hermogenes 4 n.9, 13, 14, 40 On Issues 131, 132, 154–5 commentaries on 133, 142, 144, 145, 155 On Types of Style 57, 65, 132, 142, 147 Hermogenes, ps. On Invention 66, 72, 132 Progymnasmata 44, 46, 50, 51, 56, 57–8, 62, 64, 65, 66, 75, 76–7, 79, 200–201 Herodes Attikos 15 Herodotos 20, 58, 63, 180, 198 Hippolytos 180; see also Euripides historiography 1, 32, 42, 63, 73, 90, 101, 103, 129, 140–141, 154; see also Herodotos, Ktesias, Philistos, Theopompos, Thucydides, Xenophon Homer 18, 23, 34, 37, 40, 54, 63, 120, 130, 180, 197, 200, 201; see also Shield of Achilles Horace, Ars Poetica 67, 100

235

hupotupōsis 77, 100, 112, 126; see also diatupōsis image, mental see phantasia; visualization John Chrysostom 190–191 John Sardianos, 27, 44, 52, 53, 54, 57, 65, 80, 81, 83, 204–7 Julian 159, 164 n.100, 182 kairos 56, 61, 62, 64, 198, 200, 201, 207, 213 Kallistratos 6, 32, 35, 82, 187 kataskeuē see confirmation kephalaion 48, 169; see also telika kephalaia klētikos logos (invitation speech) 156–7, 160–1, 170 koinos topos 43, 47, 50, 76–8, 113, 126–7, 158 Ktesias 69, 197, 213 language as force 23, 98–100, 107, 128–9 and phantasia 88, 93–7, 113–15, 117–121 see also Attic; grammar Libanios 15–16, 41 n.10, 43, 44, Autobiography (Or. 1) 125–6 Enkōmion of the Kalends (Or. 9) 137 model ekphraseis 6 n.12, 46, 56, 61–2, 68, 78–81, 83, 84, 112, 121 n.50, 137, 138, 139 model enkōmia 79 Monody for the Temple of Apollo (Or. 60) 159 Longinos (ps–), On the Sublime 20, 22–3, 96–8, 101–2, 107, 115–118, 129, 132, 142, 145, 155, 165, 167, 168 Longos, Daphnis and Chloe 178 Lucian 14, 18 n.17, 19, 28–9, 30–31, 117 n.37 The Dream 14 n.4, 16 Eikones 20–21 Enkōmion of the Fly 79

236


The Hall (De Domo) 6, 134, 172–4 On Mourning 171 n.12 True History 119–20 Tyrannicide 144, 145 On Writing History 67, 140–41 Lysias 22, 85, 89, 90, 91–2,123–4; see also Dionysios of Halikarnassos madness 104, 116, 118, 176, 209–10 Marcus Aurelius 161–3 meletē see declamation memory 10, 88–9, 110–113, 119–123, 154, 158, 159, 161, 163, 165, 167, 169–72, 174, 180, 185 memory techniques 25, 110–11 Menander Rhetor 4, 14, 40, 52, 113, 131, 133–6, 155–164, 169–172, 189, 190 metaphor Aristotle on 85 indefinition of ekphrasis 55–6, 93 metastasis (transference of blame) 100, 146, 148 metathesis (change intime) 100–101 mimesis 55, 119 n.46, 127–8 monody (brief speech of lamentation) 159, 160, 170–2 Nikolaos, Progymnasmata 14, 44–5, 50, 51–2, 53–4, 55, 56, 58, 61, 62, 74–5, 77, 78, 104, 113, 114, 138,152, 158, 202–5 on ekphraseis of works of art 46, 55, 65, 82–4 on ekphrasis and diēgēsis (narration) 70–73, 88 on the function of ekphrasis 48, 76, 99, 131, 144 model ekphraseis 46, 61–2, 82–3, 139, 186–7 orality 5, 9, 40, 69 n.26, 98, 129 painting see ekphrasis of works of art

as analogy for ekphrasis 27–8, 83–4, 96, 103–4, 139, 154, 178–9, 186, 194, 205, 206 as analogy for memory 112, 119–121 incourtroom 89 indeclamation 150 pathos 82, 100, 148, 204 Paul the Silentiary 2, 31, 156, 174 Pausanias 2, 7, 54 peacock 47 n.38, 79, 134, 174 periēgēsis 54–5, 75, 174 peristaseis (‘elements of narration’) 63–5, 69, 79, 81, 91, 138, 155; see also pragma persuasion, 10, 13, 48, 75 n.43, 105, 130–165 phantasia 27, 52, 88, 93–97, 99, 101–2, 104, 107–130 passim, 140, 155, 158, 165, 206, 207 inStoic philosophy 114–118, 119, 167 phantasma 27, 122, 183 inAristotelian philosophy 111–12, 114, 115, 118 inStoic philosophy 116, 118, 183–4 Philistos 39, 57, 63, 198, 214 Philostratos Eikones, 2, 6–7, 21, 28, 30–35, 36, 81, 173, 187–90, 194 Life of Apollonios 119 n.46 Lives of the Sophists 14–15, 133, 136, 161–2 pithanon 48, 168 plasma 120, 169, 180; see also fiction Plato 55, 57, 119 n.46, 214 Plutarch 19 n.19, 20, 25, 99 n.27, 103, 129, 149 Plutarch, ps– 18 n.18, 112, 120 poetry 72, 89, 130, 140–41, 154 descriptions of works of art in1–3, 36 ekphrasis in4, 190 enargeia in22–3, 40, 89–90, 95, 97–9, 101–2, 129, 193 inthe Progymnasmata 28, 49, 84–5, 141

Index Polemo 136, 156, 177 pragma 61, 63, 64, 67–8, 72, 75, 79 n.52, 91, 111, 138, 142, 155, 181, 198, 200, 206, 208, 213 presbeutikos logos (Ambassadors’ Speech) 160, 162 Progymnasmata nature of 17–19, 41–2, 45–6, 48–9 order of 49–50 relation to full–scale speeches 47–9, 137 inrhetorical curriculum 42–3, 45–9, 62 see also Aphthonios, confirmation, ekphrasis, ēthopoiia, ps.– Hermogenes, koinos topos, Libanios, Nikolaos, refutation, Theon Prokopios of Caesarea 156, 174–5 proof (inoratory) 89–90, 137, 141, 152, 163, 164, 165, 175 Quintilian, 3–4, 10, 14, 26–7, 46, 48, 78, 87–130 passim, 131 Institutio oratoria 1.9: 43 2.4.3: 85 2.4.26: 83 2.4.28–9: 137 n.22 2.10.8: 176 4.2.34: 103 4.2.63–5: 103, 168 6.1.30: 90 6.1.31–32: 89, 94 6.1.37–42: 107 6.2.26–8: 100, 104 6.2.29–30: 95, 102, 104, 121, 168 6.2.33: 97 6.2.35: 100, 104 6.2.36: 104 8.3.62: 26, 98 8.3.63: 92 8.3.64–5: 21–2, 24, 107–110 8.3.66: 91, 108–9, 153 8.3.67–9: 72–4, 88, 93, 99, 142, 148 8.3.71: 109

237

9.2.40–1: 100 11.2: 25 n.36, 110 Quintilian [?] Major Declamations 150 Minor Declamations 143 n 40 reading 17–26, 28, 96, 182–3, 195 refutation (elementary exercise) 43, 45, 48, 78, 80, 199 Rhetorica ad Herennium 25 n.36, 74 n.36, 87, 100 n.32, 110–11, 120, 148 Sardianos see John Sardianos Second Sophistic 10, 14–17, 30, 33, 36, 177; see also declamation; epideictic Seneca, the Elder 75 n.41, 104, 133, 147 n.53 Shield of Achilles 2–3, 6, 7, 11, 19, 28, 31, 32, 33, 40 n.4, 54, 58, 66, 68, 70 Sminthiac Oration 157 Smith, Susan 124–5 Smyrna 162, 171, 174 Sopatros the Rhetor 14, 47–8, 74, 131, 133, 137, 141–9, 153, 154 Sophokles 180 Souda 44, 46 Spitzer, Leo 33–35, 37 stasis theory 132; see also declamation; Hermagoras; Hermogenes; Sopatros Statius 30, 32, 81 Statue see ekphrasis of works of art, Kallistratos; Nikolaos, model ekphraseis indeclamation 155 inepideictic 157 storm see ekphrasis sub oculos subiectio 51, 100 Syrianos 14, 133 n.5, 144–5 telika kephalaia (‘heads of purpose’) 144 theatre, as analogy 54, 100, 104, 176, 194

238


Theon 14, 25–6, 39–81 passim, 85, 87, 94, 99 nn.27 and 31, 114–15, 120, 126–7, 138, 180, 197–200, 206 Theopompos 116, 214 Thersites 58, 197, 203 Thucydides 18, 19–20, 38, 58, 62, 63, 68, 69, 71, 103, 116, 129, 195, 197, 198, 200, 201, 202 Tiberios 89 n.6, 93 n.16, 101 n.35, topos (‘place’) see also koinos topos tupos (impression) 52, 112, 206 verisimilitude 102–3, 117–118, 124–5, 168; see also eikos, pithanon

Virgil, Aeneid 3, 30, 32, 33, 70, 97, 113 visualization developed inancient education 24–5, 95–6 inoratory, 95–6, 100, 102, 104, 107–110, 115, 119, 212, 123 inpoetry 101–2 as response to reading inAntiquity 19–25 wedding speech 47, 113, 157–8, 160 Xenophon 20–21, 38, 146

Ekphrasis-imagination-persuasion.pdf

Recommend Documents