[leech.link]__0198738730.pdf 4x5s1i

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

O X F O R D S T U D I E S IN AN C I E N T C U L T U R E A ND REPRESENTAT ION General Editors The late Simon Price R. R. R. Smith Peter Thonemann Tim Whitmarsh

Oliver Taplin

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

O X F O R D S T U D I E S IN AN C I E N T C U L T U R E A ND REPRESENTAT ION Oxford Studies in Ancient Culture and Representation publishes significant interdisciplinary research into the visual, social, political, and religious cultures of the ancient Mediterranean world. The series includes work that combines different kinds of representations that are usually treated separately. The overarching programme is to integrate images, monuments, texts, performances, and rituals with the places, participants, and broader historical environment that gave them meaning.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

Mimetic Contagion Art and Artifice in Terence’s Eunuch

ROBERT

1

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

3

Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a ed trade mark of Oxford University Press in the UK and in certain other countries © Robert 2016 The moral rights of the author have been asserted First Edition published in 2016 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2016935701 ISBN 978–0–19–873873–2 Printed in Great Britain by Clays Ltd, St Ives plc

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

ACKNOWLEDGEMENTS

Taïde è, la puttana che rispuose al drudo suo quando disse “Ho io grazie grandi apo te?”: “Anzi maravigliose!” Inferno 18.133–5

Gratitude is often easy to express for things of short genesis, but this book has been very long in coming to fruition. Its first glimmer of life was in September 2000, my second week of graduate school, as a delirious hallway conversation between Shadi Bartsch, Phil Horky, and myself. It fashioned itself almost immediately into a dissertation project, but was slower in arrival than its first flash presaged, due to my own lumbering pace in connecting the grad school dots. I owe a fine, high debt of thanks to the faculty of the Department of Classics at the University of Chicago for the training and encouragement they gave me and the patience they showed me throughout. But while gratitude is due to all, special thanks must go to my thesis advisors, Shadi Bartsch, David Wray, and Jas Elsner, who were unstinting in their and their guidance with this project. The last couple years of my tenure as a grad student were spent teaching at Trinity University, where once again I found myself the recipient of material and moral well beyond my desert. The faculty and the students at Trinity were truly remarkable in their willingness to think about mimesis and comedy with me as I wrote the better part of this book, and the fact that many of them are still dear friends at some years’ remove is hard to extricate from their love and intelligent conversation as I worked through this material. Among many names that ought to be mentioned, I can hardly fail to single out Erwin Cook, who first taught me to read Homer when I was an undergraduate at the University of Texas and once again at Trinity was an irreplaceable interlocutor and a model of what a classicist might be. An animal is born once, a man perhaps twice, and such creatures are thus limited in their reasonable claim of terroir, but a plant successfully moved from one pot or earth to another gains a new native soil. This book has survived and thrived in transplantation to Haverford College, and the thanks I owe to this academic environment are immense. My colleagues at Bryn Mawr College and other classicists throughout the greater Philadelphia area have been interested and ive and have taught me more than they know or than I can readily com here in words. But I must make special mention of my colleagues at Haverford, Deborah Roberts and Bret Mulligan, whose friendship, , and unfailing encouragement have made my life these past seven years the envy of all who know me.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

vi

ACKNOWLEDGEMENTS

I could not have asked for a more conducive environment for intellectual growth, productive work, or happiness. Without the generosity of Haverford College in funding my junior leave (2011–12) and intermittent course releases throughout, I could never have completed this book, nor would it have the shape it has without the criticism and encouragement I received when presenting pieces of it as papers, twice at Chicago under the auspices of the Rhetoric and Poetics Workshop, in 2005 at CACW/CAPN, in 2011 at the University of Pennsylvania, and in 2013 at the University of Manchester. Terry Snyder and Margaret Schaus and Haverford Libraries have been a great help with locating and purchasing image permissions. At OUP Hilary O’Shea and Annie Rose have been a delight to work with, and the anonymous readers have astounded me with the care they have taken and the intelligence and sympathy of their response. I wish I knew their identities so I could thank them by name for some of the most thoughtful criticism I have ever received. I can only hope that they will read this and know that they have this stranger’s gratitude and iration. In case such a wide dispersion of thanks be felt to mitigate my own responsibility for the mistakes and lapses in judgment in this book, I should be clear that I claim those for my own, attributable to my own oversight or obstinance and not to the hermeneutic and editorial help I have received along the way. I sigh to think that after so many years in gestation such errors must persist, but as Robert Burton said of his own slim tome, “I do easily grant, if a rigid censurer should criticize on this which I have writ, he should not find three sole faults, as Scaliger in Terence, but three hundred.” It is a debt of a different kind that I owe to my mother, whose formative love, thirst for life, and boundless inquisitiveness were always the beginning of everything. Without the of my closest friends and confidants these fifteen years would have been too arid for this book’s growth, but they have also taught me things that show up throughout the book and suffuse its perspective. I name, in order of the date of our first love, Phil Horky, Dawn LaValle, Fr. Pat Reardon, and Alex Petkas. In this same category, though no doubt with an intensity not to be matched, are my four children: Grace, Ada, Elias, and Jack. From them much of this book has been borrowed in the form of time and attention, but they have also fed me all along the way. Above all else my wife, Dianna, without whom nothing. It is to her I dedicate this book in affection, iration, and thanks.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

F R O N T I S P I EC E . Rembrandt’s Danaë (1636) Photo courtesy of the State Hermitage Museum, St. Petersburg

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

CONTENTS List of Illustrations

Introduction

xi 1

1. Judging Chaerea: The Role of the Painting

28

2. Quickening Images: Mimetic Contagion in Cultic and Erotic Art

49

3. Lifelike Likeness: Mimetic Contagion in the Philosophical Tradition

72

4. Mimetic Contagion in Terence’s Rome

95

5. Mimic Rape: Genre Switching and Role Confusion

120

6. The Poetics of Contamination

157

Epilogue

178

Bibliography

183

Index

197

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

LIST OF ILLUSTRATIONS Frontispiece. Rembrandt’s Danaë (1636) Photo courtesy of the State Hermitage Museum, St. Petersburg 5.1. Santia vase, B.M. F233 Photo courtesy of the British Museum

viii 144

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

Introduction

At the physical centre of every eunuch is a cut, an act of violence and irreversible change, and Terence’s Eunuch is no exception. One line after the exact centre of the play, Chaerea emerges from the house he had entered 56 lines before. Over the next 57 lines he will narrate the events that transpired inside, creating a symmetrical diptych of backstage action and staged narration directly around a break at the stichic centre of the comedy.1 Chaerea’s rape of Pamphila, at the very fulcrum of the text, alters the lives of all the main characters and turns out to be central to the plot. Its centrality, however, is not limited to its effects; we immediately understand that the events of the first half of the play were engineered to get Pamphila and Chaerea alone together backstage without the supervision of their older siblings, one of whom is at a party and the other in the country. A brief summary of the first half of the play will reveal this orchestration. Act 1 began with Phaedria, Chaerea’s older brother and Terentian double, complaining to his slave Parmeno of his unrequited ion for his next-door neighbour, the apparently mercenary meretrix, Thais, who has played him for gifts and affections and now locks him out with haughty indifference. Now, after so many capricious reversals, he is ready to forswear love and Thais forever. Parmeno agrees with his master’s assessment of love, but is cynical about his ability to resist it, even knowing its faults. Thais then s them onstage and explains her recent cold shoulder as part of a ruse to trick her old flame, Thraso, into giving her the beautiful young lute girl, Pamphila. He had originally acquired the girl as a gift for Thais, but since his arrival in Athens, he has learned of her affair with Phaedria and is now hesitating to hand her over. Thais explains that, unbeknownst to Thraso, Pamphila is actually her own adopted sister, an Athenian girl of uncertain

Thais takes Chaerea backstage immediately after line 493. He comes back onstage at line 549. His narration lasts until line 606, at which point Antipho abruptly changes the subject and the boys depart. I do not, of course, suggest that this symmetry would have been perceptible to the play’s original theatre audience, but such careful structuring may indeed be further evidence that Terence wrote with the expectation that his plays would enjoy private reading, whether sympotic or solitary. Peter Kruschwitz (2001) has uncovered a similar effect in several of Terence’s plays, where individual scenes are paired by consisting of the same number of lines. These scenes are sometimes far apart from each other, but upon closer examination they reveal important thematic and dramaturgical similarities. 1

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

2

MIMETIC CONTAGION

parentage, who was given to her mother as a small child. Though raised in a brothel, Pamphila is still a virgin, and Thais is eager to get her out of Thraso’s clutches and into her own safekeeping, partly out of sisterly affection and partly because she hopes to prove Pamphila’s Athenian citizen parentage. Reuniting an Athenian family with their long-lost daughter, Thais would make her adopted sister a marriageable Athenian girl and would procure patronage for herself in the city where she has taken up residence but is as yet without protection. To complete the deception she requires Phaedria’s absence for two days. This will give her the chance to wait on Thraso unreservedly and convince him of her total availability and devotion to him. Evidently unimpressed by this story and regarding Pamphila as simply another expensive bauble that has caught Thais’ eye, Phaedria complains that he too has made such gifts: she wanted an Ethiopian slave girl, she wanted a eunuch—he got her one of each. Still, he grudgingly agrees to go, provided that two days not turn into twenty, and begs her to be faithful in spirit, if not in body (cum milite isto praesens absens ut sies, 192). He and Parmeno go into his house to prepare for his journey, and Thais goes into her own to wait for Pamphila’s long-lost Athenian brother. In Act 2 Phaedria departs for the country, leaving instructions with Parmeno to deliver the eunuch and the Ethiopian girl to Thais. Left alone, Parmeno barely has time to bemoan his master’s love-smitten folly, when Thraso’s toady, Gnatho, enters the stage bearing Pamphila to her new mistress. In the previous Act, Parmeno’s sarcastic interjections had served to comment on Thais’ speech, and here, too, Gnatho’s self-satisfied monologue is punctuated by Parmeno’s disgusted asides. His contempt for the parasite sets Parmeno up for a painful reversal, when Gnatho, who knows where things stand with Phaedria, notices Parmeno and cannot resist the opportunity to poke fun at him for being locked out of Thais’ house. Gnatho goes inside, deposits Pamphila, and departs, and Parmeno only has one line to fume before spotting Chaerea, who stumbles onstage breathless with frustration (occidi!), having no sooner seen than lost the girl of his dreams. He describes her to Parmeno and relates how he was fortuitously hindered from getting near her or seeing where she went, except that it would surely have been somewhere onstage (huc advorterat | in hanc nostram plateam, 343–4). Parmeno realizes he means Pamphila, and tells him that she is Thraso’s gift to Thais, considerably outbidding poor Phaedria’s gift of a broken-down old eunuch. Even with the eunuch’s unspeakable deficits, Chaerea still finds him enviable because of his impending proximity to Pamphila. Parmeno suggests that Chaerea impersonate the eunuch and be delivered into the house in his stead, and they depart to make the necessary wardrobe changes. Act 3 opens with one of the funniest and most ethopoetic scenes in Terence. Here the focus is not on getting into Thais’ house, but getting her out of it and keeping control of her. Thraso flounders from one embarrassingly idiotic story to

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

INTRODUCTION

3

another, while Gnatho leads him on with shameless flattery, and Parmeno’s revulsion grows by the moment along with his eagerness to trick all of them. Thraso is unsure of Thais’ affections, and worries how he can keep her attention once he has her at the party. They have evidently noted how keen Thais was to receive Pamphila and get her away from Thraso, and Gnatho explains that she is jealous and therefore vulnerable. If, during their revelry, Thais wants to invite Phaedria, all Thraso must do is return fire by sending for Pamphila, if she praises Phaedria’s handsomeness, Thraso should praise Pamphila’s beauty, and so on, pitting par against pari until she learns how it feels to be a rival (440–5). We know that Thais is unlikely to mention Phaedria at all, since she has made such a point of insuring his absence, but Gnatho’s hypothetical scenario, by casting Phaedria and Pamphila as functional pares, suggests the real similarity between them: Phaedria’s absence in the country and the emphatic seclusion Terence is setting up around Pamphila. Thais’ plans will keep Phaedria off stage right, in the direction of the country, and Thraso’s party off stage left, in the direction of town, leaving her house rather isolated. She comes out when she hears Thraso’s voice and is ready to leave with him immediately, but Parmeno chooses this moment to make trouble for everyone by showing up with Phaedria’s gifts, the Ethiopian slave girl and the ‘eunuch’, played by Chaerea. This untimely presentation is successful in putting Thais in an awkward position with Thraso. She has been waiting for Pamphila’s brother, Chremes, but now Thraso is eager that they be on their way, so Thais tucks the new slaves into the house as quickly as possible and gives Pythias instructions that if Chremes should show up, he must if possible be kept in the house until her return. Only if he will not wait should he be escorted to the party. As soon as everyone leaves, of course, Chremes arrives from the country, grumbling his suspicions about Thais, who has summoned him now for a second time. He is eager to assure anyone who wants to know (roget quis, 511) that he is innocent of Thais and her wiles, whatever they may be. From her line of questioning at their previous interview, he supposes she is trying to seduce him or snag his family’s estate at Sunium or maybe even pretend that she is his long-lost sister, which he says is preposterous given her age. He learns that Thais is out and he is ready to head home immediately, so Pythias tries to entice him into the house. When it becomes clear that this will not work, she offers to have him escorted directly to Thais at Thraso’s party, and he agrees. In every one of these scenes from the first half of the play, there is some tension around the issue of access to Thais’ house. Both Phaedria and Thais would like to be in her house, whether for love or to wait for Chremes, but circumstances force them off the stage in different directions. Parmeno was made to feel like a fool for being temporarily locked out of her house, and Chremes refused to cooperate with the plan and stay there. This convergence of tensions between people who wish to be or are supposed to be there but are emphatically not there creates a

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

4

MIMETIC CONTAGION

dramatic vacuum about the house. To be sure, there are slaves at home, but Chaerea and Pamphila are left unsupervised by older siblings and by anyone else who has been introduced as a major character so far. Moreover it is clear that everything that has happened up to this point in the play has conspired to produce this configuration. Does this transparence of dramaturgical intentionality provide some mitigation of Chaerea’s crime? If the rape was not a random act but the telos towards which the first half of the play was driving, then the scheming is clearly, even spectacularly, on Terence’s side, and Chaerea comes to seem more like a hapless puppet. In case we miss this orchestration in the flurry of scenes, the characters themselves point it out in hindsight. In Act 5, when Chaerea and Thais are reaching an agreement, he looks back at the day’s events with wonder at how everything has mysteriously played advantageously. ‘What if some god has ordained all this?’ he asks (quid si hoc quispiam voluit deus? 875). Thais agrees: ‘That’s exactly the way I’d like to interpret it!’ (equidem pol in eam partem accipioque et volo). Fortune or the controlling god in New Comedy may always be understood metatheatrically, as the invisible hand of the comic poet ex machina. The vagueness of Chaerea’s quispiam deus may encourage such a reading, but we should not miss his two-fold invocation of Jupiter, once when he first emerges from the house after the rape (550) and then again at the end of the play, in a prayer that Jupiter will continue to maintain the day’s blessings (1048).2 This possible doubling of Jupiter as presiding god and figure for the playwright is reminiscent of the prologue of the Amphitruo, where Mercury repeatedly explains that he has been sent on stage by his ‘father’ with explicit orders to deliver this speech pleading for the audience’s attention. He is sure the audience will comply out of fear of the king of gods, but he goes on to say that Jupiter, who was paradoxically born to a human mother and father (28), has instructed him to ask nicely, since they are all dependent on the success of the comedy. This play with paternity is metatheatrical in two distinct senses: Mercury may be understood as a self-aware character speaking about his ‘creator’ or he may be understood as an actor referring to his ‘patron’. Of course, these two senses may well overlap with each other, and both ontological and performative dependence become aspects of Mercury’s relationship to Jupiter throughout the Amphitruo. The Eunuch’s metatheatrical play with the authority of Jupiter is, as we might expect, much more restrained, but when Chaerea modifies his diffuse claim of providence at work to a prayer for Jupiter’s continued beneficence, it is hard to miss the pattern. Of course, we might be tempted to dismiss his earlier Pro

2 Note also Phaedria’s exclamation the moment he learns that his brother was the eunuch/rapist: Iuppiter magne, o scelestum atque audacem hominem! (709).

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

INTRODUCTION

5

Iuppiter (550) as a standard oath in Roman comedy, but as he is about to explain he has just experienced an epiphany of Jupiter in art (583–91): dum adparatur, virgo in conclavi sedet suspectans tabulam quandam pictam: ibi inerat pictura haec, Iovem quo pacto Danaae misisse aiunt quondam in gremium imbrem aureum. egomet quoque id spectare coepi, et quia consimilem lat iam olim ille ludum, inpendio magis animu’ gaudebat mihi, deum sese in hominem convortisse atque in alienas tegulas venisse clanculum per inpluvium fucum factum mulieri. at quem deum! ‘qui templa caeli summa sonitu concutit.’ ego homuncio hoc non facerem? ego illud vero ita feci – ac lubens. While [the bath] is prepared the girl is sitting in a chamber Looking at a certain painted : on it was this picture, How they say Jupiter once sent a golden shower into the lap of Danaë. I too started looking and because he played a very similar game Once upon a time, all the more my heart rejoiced, That a god should turn himself into a man and creep Onto someone else’s roof tiles to trick a woman through a skylight. And what a god! ‘Who shakes the quadrants of the sky with his clap.’ And I, mere mortal, shouldn’t do it? You bet I did it, and I loved it!

Rape is a frequent enough feature in the backstories of Greek New and Roman comedy, but this act of violence in the Eunuch is unique in the genre. It occurs not in the backstory, but backstage during the action of the play, and many have felt its literally obscene (unstageable) brutality to be a provocation to condemn the character and perhaps the poet as a monster. Chaerea’s narration of the rape seems, however, to displace significant responsibility on to the painting, and Chapter 1 attempts to discern how seriously we are meant to take this excuse. At one time many readers apparently adopted a relatively positive view of Chaerea, but more recently disapproval of his crime and dislike of his character have become so intense that scholars have generally either ignored his description of the painting or dismissed it as a disgusting prevarication. Rather than simply pointing out the lack of positive evidence that Chaerea was planning to rape Pamphila before he saw the painting, I argue that the comedy is carefully structured to reinforce the unexpectedness of the rape. Terence plays against the Roman taxonomic discourse of erotic success (the quinque lineae amoris) and the formulaic code for the context of rape in New Comedy (wine and religious festival), in ways that have been either underappreciated or missed entirely, in order to emphasize the centrality of the painting’s intervention in the plot. One of the reasons critics have been likely to miss the mitigating force of the painting, I argue, is that they are generally blinded by revulsion at the attack. Rape was certainly a crime in antiquity, arguably more severely punished then than now, and yet we

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

6

MIMETIC CONTAGION

evidently respond to this literary scenario with greater consternation than the Greeks and Romans seem to have done. One of the things I attempt in this chapter is a cursory archaeology of reading rape in the twentieth century, as a tentative explanation for the lens with which we all approach the Eunuch. This survey of modern responses to this scene and the painting is balanced against two ancient responses: Donatus’ and Augustine’s, both of which, in very different ways, a more expansive reading of the role of the painting. Where Chapter 1’s principal function is to investigate what the painting does, Chapters 2 and 3 attempt to explain how a painting could be thought to have such an imposing effect. This investigation takes us some distance away from the text of the Eunuch, as this expanded perspective is necessary to demonstrate that such coercive influence was broadly attributed to works of art in antiquity. Indeed, this book is not in the first instance intended as a study of Terence’s Eunuch, but rather of the phenomenon I call mimetic contagion as an organizing motif of that play. The conceit that life may be drawn into imitation of art is less interesting in universalized abstraction than in application within a specific cultural moment, as an historically inflected mode of thinking about artifice and response, but before historical and textual specificity is intelligible, the wider context of mimetic contagion is necessary.3 I begin with cultic practices involving artworks designed to implicate the viewer in imitation. Next, I consider the effect of erotic art on the viewer as a closely related form of persuasive analogy. This dynamic is apparent in straightforward ‘pornography’, but it is even more significant within the discourse of didactic erotic art, a tradition I explore at some length here because of its relevance to the painting in the Eunuch. Along the way I am taken more deeply into several such examples, especially underinterpreted ages in Ovid and Propertius, a pair of Pompeian paintings, and a Flavian mirror. Finally, Xenophon of Ephesus’ novel, the Ephesiaca, will provide a useful comparison for the Eunuch, not only due to the inspirational role of an erotic image hanging above a bed, but because this image, once imitated superficially, will prove itself to have an emblematic quality, recurring through the rest of the story in less obvious ways. The focal viewers of this image in the Ephesiaca, as in the Eunuch, are young people, and perhaps they should be assumed to stand in need of the kind of erotic instruction discussed in Chapter 2. In any case, as young people, they would almost certainly be understood as peculiarly vulnerable to mimetic suggestion more generally, whether for good or for ill, and Chapter 3 explores the role of

3 Of course, not all contagion is mimetic, nor is all imitation based in visual art. For the most part such extensions lie beyond the narrow scope of this book. One of the most suggestive of such stories not told here is the development (and eventual suppression) of medical theories of contagion (metadosis) as distinct from miasma. On this topic see Leven 1993 and Nutton 2000. On the likely confusion between the concepts of contagion and miasma, see Holmes 2010, 133.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

INTRODUCTION

7

mimetic contagion in the philosophizing discourses of pedagogy. Any representation, however crude, might in the right circumstances be thought to inspire imitation, and the cultic and erotic images surveyed in Chapter 2 may well rely more on context and use for their effect than on any artistic quality intrinsic to the object. But when the artwork seemed to excel at verisimilitude, the mimetically contagious quality of the image could appear a necessary corollary of the artist’s success at capturing life. Several ancient testimonies report or postulate artistic images wrought with such exacting verisimilitude that they inspire direct imitation in the viewer, and Xenophon’s Socrates regards this persuasive lifelikeness as essential to an artwork’s ability to become an ethical model, whether for good or ill. This philosophical interest in morally infectious images is taken up by Aristotle, and though his focus is not on the role of lifelikeness, he does express concern for the contagious quality of deleterious artworks, particularly for the young. From Aristotle’s tightly circumscribed treatment of this topic, I turn to the more expansive issue of mimesis in the philosophical tradition, specifically focusing on the aspects of Platonic mimesis that relate directly to mimetic contagion. As has been amply demonstrated elsewhere, the word mime ̄sis is used by Plato and Aristotle to cover a variety of interrelated but still distinguishable phenomena, so I try to be as careful as possible in delimiting the range of where this conceptual field intersects with mimetic contagion. In the Republic, famously, Plato takes a suspicious view of mimesis, but some scholars have doubted that this suspicion has anything to do with what I call mimetic contagion. I argue that in Republic 3 Plato does, in fact, regard the morally contagious effect of visual art, particularly on the young, as a paradigm for thinking about one of the dangers of mimesis more generally. Some two decades later, in the Timaeus, Plato will return to the ethical implications of mimesis in a different light, and here too the aspect of mimesis he has in view will be closely connected with mimetic contagion. My reading of mimetic contagion in the Timaeus shows how the ongoing propagation of images from artefact to viewer contributes to a partial recuperation of the concept of mimesis in Plato’s thought, and it also reveals an important connection between Timaeus’ description of the structure of the universe and Critias’ of the generational propagation of the Atlantis story. Chapter 2 examines cultic practices and popular uses of art across Greco-Roman antiquity, and Chapter 3 considers the reflex of these phenomena in philosophical texts influential on Hellenistic constructions of the pedagogical power of images. Though the problem of the Eunuch explored in Chapter 1 is never totally absent from view in this survey, it is in Chapter 4 that the historical focus for mimetic contagion is narrowed to Terence’s cultural context and to the Eunuch itself. I consider the role of images in mid-second-century BCE Roman funerary practice as a local analogue of some of the broader ancient Mediterranean ritual patterns and pedagogical discourses of the previous two chapters. More importantly,

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

8

MIMETIC CONTAGION

I approach the later testimony on the sack of Syracuse in 211 BCE and the effects of Marcellus’ ovatio on the character of the Roman people not in order to speculate on what really happened in 211, but rather to determine how these events were understood and the effects of Greek art formulated, at least in some circles, half a century later when the Eunuch was produced. Having moved from mimetic contagion in ancient Mediterranean culture at large to Terence’s Roman context more narrowly, I turn next to the Eunuch’s most immediate generic neighbourhood. An investigation of the role of plastic art in Plautus’ comedies reveals attitudes that are importantly related to themes developed in Chapters 2 and 3 more generally, and in a few cases Plautus deploys works of art in ways that prefigure the direction Terence will take in the Eunuch. Paintings are introduced, if only in the imagination, as a comparison for a character in the comedy, but the analogy somehow escapes its container and informs the character more broadly. After extensive exploration of the background of this moment in the Eunuch, both cultural and generic, we are prepared, at last, to return to the Eunuch itself for a fuller reading of the role of the painting. In the Plautine examples there is always a clearly metatheatrical dimension to this play with art, and this feature is also present in the Eunuch. I examine the overtly theatricalized elements of the eunuch intrigue, which is clearly handled as a play-within-the-play. This work of dramatic art is derailed by with the painting and then seems to adopt the painting’s aggressive quality, as the role of the eunuch threatens to step outside its theatrical frame and become reality. Chaerea is faced with emasculation, first figurative (when he casts himself on Thais’ mercy and begs her to become his patrona) and then fictional (when Pythias announces that he is about to be castrated). Chaerea’s castration is, of course, only symbolic or narrated as part of a trick—he is never in any real danger of getting literally stuck in the role of the eunuch. But Thais is not so lucky. She proposes in Act 1 to play the greedy meretrix, a caricature of her stock character grotesquely out of step with her ‘real’ personality. Her only fear is that Thraso, the miles, will not be the only one to fall for her act and that her true love, Phaedria, will come to think ill of her. This is precisely what happens. Phaedria, the arch-lover of Act 1 and model for the future elegiac tradition, has by the end of the play been disabused of his ideals of love and agrees to pimp Thais out to Thraso for money. This disgusting resolution has confused and disappointed readers for the last century, but I argue here that when Thais’ assumed role transcends its intended boundary and becomes indistinguishable from her real life, it is only following the pattern introduced by the painting at the centre of the play. Chapter 5 examines the rape scene from a slightly different perspective and explores another aspect of mimetic contagion, namely its ready connection with the theatrical genre of mime. I see no reason to assume, as is often done, that preliterary Italian folk theatre was all essentially alike. The evidence indicates a range

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

INTRODUCTION

9

of theatrical styles and types of entertainment (Fescennine, satura, Atellan, Phlyax, etc.) as different from each other as the diversity of their provenance would suggest. However, the Romans themselves seem to insist on dichotomizing theatre into highbrow fabula and lowbrow mime, a category capacious enough eventually to subsume all other types of entertainment and function as a relatively stable generic Other. From Cicero on we find explicit testimony of this binary dissection of Roman theatrical entertainment into scripted, cultivated fabula and crude, improvisational ‘mime’, and though such external testimony is lacking for the mid-second century, the prologues of Terence’s plays furnish important evidence that this dichotomy is already alive. The prologues repeatedly return to the theme of the fragility of the production and the destructive invasion of some alien type of entertainment into the performance of Terence’s quiet comedy. I survey the prologues from this perspective, including the much studied prologue(s) of the Hecyra, not to determine how much historical truth they tell, but rather to elucidate the self-consciousness of the comedies’ formulation of their own genre, as permanently vulnerable to subversion by some rougher theatrical Other. After thus delimiting the generic space that Terence constructs for his comedy and examining one of the ways in which he thematizes the irruption of lowbrow shenanigans, I explore the connection between mime and the rape scene in the Eunuch. Ancient definitions of mime emphasize the centrality of imitation in mimic acting. Mimes would sometimes copy their audience, but this copying could also prove contagious, and the audience would find themselves imitating the mime. To elucidate this hall-of-mirrors effect in mimic imitation I look more closely at the Latin Anthology’s Vitalis epitaph (Riese 487a) and the end of Xenophon’s Symposium; the significance of this dynamic for the Eunuch will be obvious to anyone who has read Chapter 4. Alternatively the mimic actor might copy another character in the mime, as in the well-attested mimus secundus or stupidus role, or the imitation might even be drawn from a work of art, as on the Santia vase. Another feature of mime that connects it to the rape scene in the Eunuch is its sexual ‘realism’. Female parts were played by mimae, who might be presented naked onstage (e.g. at the Floralia, only a few weeks after the Eunuch), and perhaps in this period they even went so far as to have or feign intercourse onstage. If the skene were inverted and backstage became the stage, the audience would recognize this scene as a transplant from mime, out of place in New Comedy. But the audience is not alone in this regard: Antipho, Chaerea’s friend to whom he is narrating the backstage events, makes only one real interruption in the story, and I believe we can best understand his comment as a recognition that for the rape scene Chaerea must have sported the costuming typical of mime, rather than that of a character in fabula palliata. I close the chapter by looking at Chaerea’s description of the painting itself, where there seems to be a confused

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

10

MIMETIC CONTAGION

slippage of roles, with Jupiter as prurient viewer and paradoxically in the same frame as metamorphic sexual actor. This is the very same confusion Chaerea experiences in his own role under the influence of the painting, but it also suggests the labile plasticity of roles typical of mime and may therefore be the trigger for the mimic element in this scene. Chapters 4 and 5 focus on themes pertaining to the infiltration of a play by an alien work of art, whether plastic or dramatic, and Chapter 6 extends the inquiry of mimetic contagion to the related question of contaminatio in Terence. Kontaminationsforschung was one of the priorities of nineteenth- and early twentiethcentury scholarship on Roman comedy, but these issues have long been dismissed, at least in the English-speaking world, as both insoluble and irrelevant to our understanding of Plautus and Terence as creative playwrights. There is much wisdom and sobriety in this resolve to read Roman comedy in its own right, especially for Plautus where the practice of combining two Greek plays into one has never been securely proven. However, at least for the three Terence plays where the practice is explicitly itted (Andria, Eunuchus, and Adelphoe), our suppression of the earlier scholarly tradition has been unfortunately influenced by twentieth-century political history and predicated on an ever narrower interpretation of the word contaminare in the Terentian prologues. I argue that this trend has left us with a definition inadequate to explain the full evidence of the word’s use, both in the prologues and elsewhere. By returning to a less extreme version of the early twentieth-century technical use of the term as equivalent to ‘corrupt by mixing’, we can better satisfy the ancient testimony and begin to explain Terence’s evident interest in the poetics of contamination. For instance, in the scene where Chaerea narrates the rape to Antipho, Terence makes an explicit wink (Eun. 552) at his earlier contamination of the Andria with a sentence from Menander’s Eunouchos (cf. Don. ad An. 959). Perhaps because recent scholarship has been blinkered by such a restricted definition of contaminare, we have failed to catch Terence’s sophisticated play with his own comedies’ vaunted hybridity. Most significantly for this study, the two ‘contaminating’ characters in the Eunuch may be seen to operate as a contagion in the play, infiltrating its dramatic structure and making a permanent home for themselves in this play’s world. More particularly, Gnatho, the parasite, who describes his flattery as a form of imitative response, becomes, in turn, a model for the other characters’ imitation and draws them into his own ethical paradigm. The structure of this study, as I have outlined it here, is one of spiralling recurrence. We begin with a single scene of a single play and return to it repeatedly from different angles, following the promptings of our culturally and historically wider and narrower perspectives on mimetic contagion. To the extent that various theories of mimesis and representation, ancient and modern, become relevant along the way, they are introduced and discussed as needed, but as mimetic

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

INTRODUCTION

11

contagion itself has not, at least to my knowledge, been the subject of sustained theorization, there is little in the way of inherited jargon to be explained or defended in this introduction. There are, however, two concepts with some measure of fashionable currency in classical studies that should be addressed at the outset, namely metatheatricality and ekphrasis. Both these concepts have been widely applied and, arguably, abused in our discipline, so it will be worth clarifying how they intersect with our topic and mode of reading the Eunuch.

TERENTIAN METATHEATRICALITY

The term ‘metatheatricality’ had no currency in the study of Roman comedy until about thirty years ago, when there was a significant shift in the distribution of scholarly interest between Plautus and Terence. Since the Renaissance there have been oscillations in each poet’s ‘stock’ relative to the other, and most periods have been content to treat them as opposite points on an aesthetic com, so that the favour of one implies distaste for the other.4 Some degree of balance was always retained in this zero-sum game by asserting that the two playwrights were good in different ways: Plautus was the poet of exuberant bad taste, outrageously inventive language, and the kind of comedy that pitches us into belly laughs, while Terence was a playwright of sophisticated manners, charmingly sedate poetry, and a subtle, ironic form of humour more apt to arouse a smile than a guffaw. The event that threw this aesthetic economy off-centre, at least for the English-speaking world, was the publication of Niall Slater’s Plautus in Performance: The Theater of the Mind in 1985.5 The originality of Slater’s book was not so much a ‘discovery’ that Plautus’ characters often seem to know that they are make-believe people on a stage. Indeed, anyone who has read the plays will have noticed the references to theatrical trappings and the jokes about dramatic conventions.6 His decisive contribution lay rather in a general reappraisal of the artistic value of these moments as more than lowbrow silliness. Informed by the theory and the cultural clout of post-World War II Absurdist theatre, Slater effectively inverted the standard of urbanity and disrupted the older distribution of prestige. Instead of calling Plautus the funnier and Terence the more sophisticated playwright, critics following Slater have generally come to give Plautus the double crown, even attributing Terence’s alleged lack of popular success to his excessive reliance on

For a concise of the history of these oscillations, see Duckworth 1952, 396–433. Given the unique history of Italian theatre and its relative circumvention of naturalism (or at least the persistence of non-naturalistic traditions), it is not surprising that Slater’s work was anticipated in some respects by Barchiesi 1969, Petrone 1977 and 1983, and Chiarini 1979. 6 For proof that this observation requires no elaborated theory of metatheatricality, see Knapp 1919. 4 5

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

12

MIMETIC CONTAGION

boring Greek models and inflexible theatrical naturalism.7 This book attempts, among other things, to redress this critical imbalance by showing that Terence did not completely reject the metatheatrical aesthetics of earlier palliata, but rather subsumed it in a modified form into his own grammar of illusion and performance.8 One could distinguish between forms of self-referentiality that cause a complete if only momentary rupture in theatrical verisimilitude, such as explicitly mentioning the audience or the length of the play, and subtler forms that suggest theatrical self-consciousness through double entendre, without necessarily excluding the possibility of a ‘straight’ reading. Such a distinction is artificial, of course, and impossible to maintain in any rigorous sense, but Plautine examples of both types are easy to find, while Terence never unambiguously breaks through the ‘fourth wall’ and fully jettisons theatrical illusion.9 Indeed, the uniformly greater subtlety of these moments in Terence, combined with a general absence of overt lexical reference to stage trappings, is probably responsible for their near invisibility to scholars used to Plautine metatheatricality. For example, in Act 1 of the Eunuch, when Chaerea’s brother, Phaedria, is planning to leave for the country for two days, his slave Parmeno doubts he can carry it through. ‘I don’t think so’, he says. ‘Either you’ll come right back or insomnia will drive you back tonight’ (non hercle arbitror; | nam aut iam revortere aut mox noctu te adiget horsum insomnia, 218–19). Critics have generally read this as a prudent evaluation of his master’s weakness in the face of love, and of course it is that too.10 But the double entendre rests in another kind of prudence. Parmeno is mindful of the conventional Unity of Time: the total length of time that es in the make-believe world of the play ‘Gentle Terence wrote beautiful Latin verse that captured the spirit of New Comedy—and failed repeatedly the crucial dramatic test of a theatre and an audience’ (Slater 1985, 6). Apart from the prologues and the Suetonian Vita Terenti, all the ancient evidence we have points to Terence’s solid and lasting popularity, and there is no suggestion that he was ever anything but a stage success in his own time and in repeated revivals. In this context, as in several others, the Vita has been shown to be almost completely reliant on inference from the prologues themselves, so it cannot be considered independent evidence (cf. Gruen 1992, 210–15; Parker 1996). In other words, there are many reasons to think Terence was very popular in antiquity, and the only self-standing reason to think otherwise is what Terence himself tells us in the prologues. Scholars have only recently come to appreciate the degree to which the prologues are to be read as playfully ironic dramatic poetry, rather than as historically accurate records (cf. Gilula 1989). 8 The exception to the rule of Terence’s exclusion from scholarship on metatheatricality in Roman comedy is the work of Stavros Frangoulidis. In studies of the Eunuch (Frangoulidis 1993, 1994a, 1994b) and more recently the Phormio (1997), he has detected play with theatrical language and situations. See also now, Knorr 2007. 9 Except for the prologues, of course, which are not only innovative, but arguably a form of metatheatricality more radical than anything else in extant ancient comedy. The audience is also addressed at the end of the plays but only very briefly, to be solicited for applause and dismissed. The Andria’s uniquely expansive two line clausula ends with the standard plaudite!, which by itself (Hec., Ad.) or preceded by valete et (Hau., Eu., Ph.) is sufficient closure for all the other plays. 10 See, for example, Büchner 1974, 245. 7

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

INTRODUCTION

13

cannot be more than one day.11 If Phaedria does stay in the country for two days he will have fallen off the edge of the world, as far as the play is concerned. He is not as clever as his slave nor as close to understanding the real constraints on his action, so he is offended by what he takes to be Parmeno’s lack of faith. ‘Why, I could manage just fine out there for three days’, he says. ‘Three whole days!’ exclaims Parmeno, ‘vide quid agas’ (224). This is usually translated as ‘mind what you are doing’ or ‘be careful’, but that would seem to leave a slight gap in sense.12 After all, why would Phaedria be in any danger? The comment only makes full sense if we that three days in the country is a virtual death sentence for a character in a play. Banished from the stage for the duration, he will act no more in this comedy. We note that the verb agere also means ‘to act in a play’, so that vide quid agas could have the distinctly ambiguous feel of ‘consider your role in all this’. This is a subtle form of self-consciousness, one not marked by unmistakably theatrical language, but its more important difference from Plautine metatheatricality is its ethopoetic and structural significance. It is not merely an entertaining moment, quickly forgotten in a series of funny scenes, but a revelation of Parmeno’s character and a preparation for his later bid for the position of playwright, when he authors the eunuch intrigue.13 Parmeno’s awareness of the generically constituted boundaries of the play world fits him for this coup, but the scene in Act 1 is neatly balanced by a scene in Act 5 where Parmeno has met his match and then some. He comes on strutting and cooing his own praises, saying his handiwork deserves the Palme d’Or (palmarium, 930), and Thais’ slave, Pythias, sees her chance to get even. She feigns hysteria over the rough treatment of Chaerea in the house, expressing sympathy for ‘the poor boy’ (adulescentulum, 943, 949) and cursing Parmeno for setting him up for such a fall. Parmeno thinks he is eavesdropping rather than witnessing a performance, but eventually he can hang back no longer and he asks Pythias what is going on. She tells him about the rape and Pamphila’s citizen status, and she alludes to the horrific punishments Chaerea is now suffering. She also makes it clear that they are blaming Parmeno as the source of all this mischief, and when Chaerea’s father comes on, she encourages Parmeno to make a full confession. Pythias’ reference to Chaerea’s punishment may sound elliptical to us, but as I argue later (pp. 111–16), she must mean castration. The obvious irony here is that Chaerea has played the eunuch and is now allegedly about to become one

On the Unity of Time in Roman comedy see Duckworth 1952, 130–1. More recently, and anticipating some of my own arguments for self-conscious engagement with this convention, see Dunsch 2005. 12 Ruhnkenius 1825 ad loc.: hac formula aliquem monemus ne rem difficilem et periculosam suscipiat. 13 See pp. 35; 108. 11

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

14

MIMETIC CONTAGION

for real. What is perhaps not so obvious in this situation is the close functional homology between the intrigue and the painting that Chaerea encountered in the house. Parmeno has crafted the role of the eunuch as a work of dramatic art, and he expects his master will be able to step into it and out of it at will, but, as Pythias tells it, this role has turned out to have the same aggressive quality as the painting, threatening to break outside its frame and become reality. Chaerea has taken up the part Parmeno sketched for him, donned the mask of the eunuch, as it were, and now the mask seems likely to stick. We must not forget that this is all a lie. If Thais had not been so generous or so eager to find a win-win solution, Chaerea might really have faced such dire consequences, but in fact he is never in any genuine danger of castration; this is just a trick Pythias is playing on Parmeno. In doing so, she is giving him a taste of his own medicine, and like some of her namesake’s most infamous oracles, what Pythias says may in one sense be untrue, but it is the kind of pretence that has the power to create reality. She makes Parmeno believe that he has betrayed Chaerea, so when the father arrives at last Parmeno is in the excruciating position of having to confess everything he and Chaerea have done, and also everything Phaedria has done, and scaring the old man out of his wits. In effect, he does betray both boys by telling on them to their otherwise unsuspecting father. However, there is another sense in which Pythias’ announcement of the impending castration is more than just a lie. By suggesting the nightmare of Chaerea’s pretence becoming real, she points back to the painting at the centre of the play and tells a disturbing truth about the effects artifice can have. The fact that the castration itself is her pretence adds another layer of irony to this revelation. We recall that Parmeno is the arch-trickster, a self-aware character who knows about the Unity of Time and thinks he can co-opt the authority and the kudos of the comic poet. By beating him at his own game, Pythias shows herself supreme in this world of deception. The father rushes into Thais’ house sputtering in fear and finds that he is too late—his son’s eunuch outfit speaks for itself! Nobody but Pythias knows why the old man is so scared, but Chaerea is just as mortified to be found as his father is to find him, and the whole mess is Parmeno’s fault. When Pythias comes back out onstage she cannot contain her laughter: in five lines (1004–8) there are three occurrences of some form or derivative of ridere. Parmeno is still puzzled so she explains (1009–11): numquam pol hominem stultiorem vidi nec videbo. ah non possum sati’ narrare quos ludos praebueris intus at etiam primo callidum et disertum credidi hominem. I’ve never seen and never will see a stupider person. Oh, I cannot relay what a show you have given us inside. And at first I thought you such a clever and capable person.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

INTRODUCTION

15

She never will see a stupider person because, as she seems to realize, the play is almost over and her brief life at an end. For today, Parmeno has taken the cake for stupidity. Her reference to the backstage show (ludos) repeats the theatrical language used earlier for the eunuch intrigue and explains her laughter: she has been witness to a comedy and she knows it. Her use of the word callidum is also pointed, as the stock epithet of the servus callidus role that Parmeno has played so well up to this point. She thought he was that character primo, at the beginning of the comedy, and so did we, but now she can gloat over the revelation of his ineptitude. She explains how he really did betray the poor boy to his father, and Parmeno can scarcely believe how he has been duped (1017–20): PA. ehem quid dixti, pessuma? an mentita es? etiam rides? itan lepidum tibi visumst, scelus, nos inridere? PY. nimium. PA. siquidem istuc inpune habueris . . . ! PY. verum? PA. reddam hercle. PY. credo: sed in diem istuc, Parmeno, est fortasse quod minare. PA. What do you mean, you bitch? You lied? Are you still laughing? Do you really think it’s smart to make fun of me, you snake? PY. Totally. PA. Even if you get away with this . . . PY. Yes? PA. I’ll get even. PY. Sure— But perhaps, Parmeno, what you’re threatening will be another day.

She understands what he evidently knew in Act 1 but seems to have forgotten: they are all creatures of a day, and any consequences that matter will have to matter in the next 74 lines. We have moved from the beginning of the play to the end very quickly here in an attempt to elucidate the dynamics of this single aspect of metatheatricality in the Eunuch, so our reading of these few scenes is one-sided and simple almost to the point of caricature, but even at this limited scale some of the consequences of the novelty of our approach are evident. We noted the use of theatrically resonant vocabulary (ludus, agere, callidus, etc.), but the heart of our reading, concerning the self-conscious play with the Unity of Time, was not keyed by a lexical tick, but rather by observation of a feature implicitly connecting these scenes to one of the empirically evident ‘rules’ of fabula palliata. The Eunuch’s play with and against the code system of Roman comedy is more subtle than the kinds of metatheatricality normally detected in Plautus, but it is also richer, in that this motif is not deployed merely as an isolated drollery, but as a framing device for the whole comedy and a barometer of Parmeno’s discomfiture. In the chapters that follow, several other modes of metatheatricality will emerge in our reading of the Eunuch, forms of engagement with theatrical selfreferentiality almost as different from each other as they are from anything in current Terence scholarship. What unifies them is that they all somehow turn on

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

16

MIMETIC CONTAGION

the rape scene at the centre of the play and, more specifically, on the dynamic of the painting in that scene and its broader implications for the Eunuch. The phenomenon of mimetic contagion need not have anything to do with metatheatricality, for it rests on a very broad base of Greek and Roman experience with art, extending far beyond theatre. Conversely, as our brief look at Terence’s play with the Unity of Time in the Eunuch should illustrate, not all theatrical selfreferentiality, even the more sophisticated sort that interests me here, need have anything directly to do with artworks that project themselves into reality. However, when mimetic contagion is thematized within a play, especially when it is given the centrality it has in the Eunuch, its metatheatrical deployment is probably inevitable. By focusing on mimetic contagion, first by itself and then in its instrumentalization as a metatheatrical device, I believe my project escapes one of the more treacherous shortcomings common to studies of ancient metatheatricality. The term ‘metatheatre’ is plagued by a number of difficulties, beginning with its inapposite prefix and the unfortunate circumstances of its birth in a now justly neglected work.14 By explicit clarification we may escape most of these associations, and the term does have valuable conceptual utility and undeniable currency, but there is often, perhaps unavoidably, a tendency to essentialize metatheatricality as a universally self-identical dramatic phenomenon. Aristophanes and Anouilh may both stage characters who seem to be aware of their own theatrical fictiveness, but we must be careful of the assumption that such awareness accomplishes the same work in both plays or that it is even, necessarily, the same phenomenon at all. Rather than beginning from the premise that all metatheatricality can be read with the same formalist lens, I propose to begin with the painting, first by investigating the painting’s role in the Eunuch and by elucidating a broad cultural context for artworks that inspire imitation. Only when this forensic and anthropological work has been set forth and only with the warrant that it gives me do I plan to back into a metatheatrical reading of the play, showing

14

The word was coined by Lionel Abel (1963) in his bantam testament on all theatre. His argument is that tragedy and comedy came to an end in the Renaissance with a new birth of self-consciousness. From Shakespeare and Calderon down to the Absurdists, modern theatre is all of a type unknown in antiquity, specifically a new genre he calls ‘metatheatre’. Beginning with Barchiesi 1969, classicists have found it easy enough simply to appropriate Abel’s term for theatrical self-consciousness, ignoring his naïve claims about ancient theatre and the whole historical dimension of his argument. Rosenmeyer 2002 offers a vigorous critique of this appropriation, because even without the historical aspect, the term makes a special category out of a feature common to almost all theatre. The meta-, then, is redundant. This is a sound point and a good reason for preferring, as many scholars now do, the adjective ‘metatheatrical’ and the abstract noun ‘metatheatricality’ to denote these aspects of any theatre. I agree wholeheartedly with every one of Rosenmeyer’s criticisms of Abel, but he may be confusing origin and essence, underestimating his colleagues’ ability to ignore Abel’s ghost.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

INTRODUCTION

17

how various types of ‘role-playing’ cluster around the painting and share its basic dynamic of mimetic contagion.

MIMETIC CONTAGION AND EKPHRASIS

Another term of art from the critical vocabulary of modern scholarship that may be relevant to the study of mimetic contagion is ekphrasis. Unlike metatheatricality, ekphrasis is a designation with an ancient pedigree; though it languished in obscurity from late antiquity until a few decades ago, it has enjoyed an enormous ascendancy of late.15 This inheritance presents problems of its own, however, principally because of an important difference in ancient and modern usage of the term. The ancient sources that talk directly about ekphrasis are consistent in applying the term to descriptions of any kind of visible object or event (artworks, crocodiles, women, seasons, harbours, etc.). They do not explicitly limit ekphrasis to contexts of public speaking, but they clearly see it, in the first instance, as a mode of rhetorical discourse, and they emphasize its usefulness in rendering absent objects present for the audience because of its clarity and vividness. In modern critical parlance, however, ekphrasis refers more specifically to the description of works of art. Art may, of course, be described in almost any genre, but the notional generic ‘home’ of ekphrasis in modern scholarship is poetry, not public speaking, and there is usually no presumption that vividness is definitive of the ekphrastic mode, still less that this vividness naturally results in an audience swept away with sympathetic imagination. Thus modern ekphrasis is topically narrower and contextually broader than ancient ekphrasis. Our discussion of mimetic contagion and ekphrasis will intersect with both the modern and the ancient definitions in different places; with the moderns I share a specific interest in the description of art, and with the ancients an interest in the vividness thought to be peculiar to ekphrasis and its role in directing the audience’s experience. From the standpoint of the modern definition, the archetypal ekphrasis that set the standard for the subsequent development of the trope was Homer’s description of the shield of Achilles (Il. 18.478–608), and throughout antiquity the description of works of art was a touchstone of the poetic craft. But our oldest extant theoretical discussions of ekphrasis are in the Progymnasmata, schematic handbooks for elementary education in the Roman Empire. It is difficult to

15

The explosion of modern critical interest in ekphrasis may be traced to Leo Spitzer’s famous essay on Keats’ ‘Ode on a Grecian Urn’ (1955). The flood of scholarship on ekphrasis is now so vast that citing a few examples would be pointlessly arbitrary. For a recent bibliographical survey, see Denham 2010, 213–87. Notable recent collections in classics include the special issues of Ramus 31 (2002) and 102.1 (2007).

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

18

MIMETIC CONTAGION

determine how representative the Progymnasmata are of earlier educational practices, but in some cases the uniformity of their language may indicate that they are operating within received conventions that go back to Hellenistic educational and rhetorical discourses.16 The nearly identical wording of the definition of ekphrasis across the Progymnasmata is a perfect example of this kind of collective witness of a clearly elaborated long-standing tradition. ‘Ekphrasis is a descriptive (περιηγηματικός) speech, which vividly (ἐναργῶς) brings the subject shown (δηλούμενον) before the eyes (ὑπ᾽ ὄψιν).’17 The first feature of this definition to note is that it rests on a paradox: how can speech bring anything before the eyes? As in mimetic contagion, where the artwork surprises by transcending its presumed frame as a mere fiction and propagating itself in the world of the viewer, ekphrasis refuses to be bound by the normal constraints on language to remain a solely auditory phenomenon and becomes an occasion for seeing. As strange as such a conception of ekphrasis may sound to modern readers, it fits within a recognizable ancient pattern for describing speech acts in drawn from visual experience. The most famous literary instance is surely the Horatian dictum ut pictura poesis (AP 361), but there are many other examples of the Greco-Roman instinct for synaesthetic visualization of language.18 In fact, the stock definition of ekphrasis as a logos that puts something ‘under the eyes’ (ὑπ᾽ ὄψιν) echoes Aristotle’s very similar formulation of a certain type of metaphor—the type that ‘signifies activity’ (energeia)—as placing the subject ‘before the eyes’ (πρὸ ὀμμάτων).19 If we ask what it is about energeia that enables it to render metaphor a functional equivalent of ekphrasis, we must turn to its near homonym, another element we noted above in the stock definition of ekphrasis in the Progymnasmata, namely enargeia (‘vividness’).20 Nicolaus says that enargeia is what separates ekphrasis from narration (die ̄ge ̄sis) and that where narration is just a bald exposition of facts, ekphrasis ‘tries to make hearers (ἀκούοντας) into spectators (θεατάς)’.21 Similarly Theon says that enargeia is one of the features (aretai) of ekphrasis and that its function is to make the audience ‘almost see the subject’.22 Another Imperial Greek rhetorical treatise, attributed to ‘Anonymous Seguerianus’, defines enargeia in the same used by the Progymnasmata for ekphrasis:

Webb 2009, 40. Theon Prog. 118.6, ed. Patillon; but see also, with only the slightest variation of diction and syntax, Hermog. Prog. 22; Aphth. Prog. 36; Nicol. Prog. 68. Hermogenes even inserts a ὥς φασιν (‘as they say’), thus indicating the formulaic force of this definition with the tag of an ‘Alexandrian footnote’. 18 19 See p. 107. Rhet. 1411b24–5. 20 On the richly interesting conceptual bleed between energeia and enargeia, see Calboli Montefusco 2005. 21 22 Nicol. Prog. 68. Theon Prog. 119. 16 17

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

INTRODUCTION

19

‘a speech that brings the subject shown (δηλούμενον) before the eyes (ὑπ᾽ ὄψιν)’.23 The circularity of these definitions is frustrating but not surprising, given the strangeness of the phenomenon described. Invisible language is wrought visible or, to be more precise, it becomes a visual experience for the hearer—this is the paradox that defines ekphrastic discourse. But it works this miracle because of its enargeia, which can in turn only be defined or described by its capacity to render the unseen visible. Perhaps we can come closest to unravelling this puzzle by regarding enargeia as the quality (rather than a type of speech, pace Anonymous Seguerianus) that most essentially defines ekphrasis by enabling it to accomplish the seemingly impossible task of turning hearers into spectators. The visibility generated by enargeia survives in its Latin calques. In his Institutio oratoria (roughly contemporary with the oldest of the extant Progymnasmata) Quintilian uses the Greek word freely, but he also glosses it with the Latin word evidentia, the term used, he says, by Cicero and Celsus.24 He mentions that Cicero also calls it illustratio and sub oculos subiectio.25 All these expressions are based in vision, and this last especially echoes the definition of ekphrasis and enargeia in the Progymnasmata. No less than the writers of the Progymnasmata, Quintilian wonders at the power (magna virtus) of enargeia to make the objects of its description ‘appear to be seen’ (ut cerni videantur, 8.3.62). Enargeia does not stop at the ears, but makes a presentation to ‘the eyes of the mind’ (oculis mentis ostendi, 8.3.62). ‘Somehow the whole scene is painted with words’ (tota rerum imago quodam modo verbis depingitur), so that the very witnesses of the event would not have a better view (non clarior . . . spectantibus, 8.3.63). This talk of visibility in speech may ultimately be no more than a metaphor, but it is worth considering how literally we ought to take it, especially since it is so frequently encountered in our sources. Some scholars have postulated that ancient audiences actually experienced oratory radically differently than we do, conditioned as they were by their oral culture to a more actively participatory visual imagination.26 Not that the Greeks and Romans routinely saw hallucinations when they heard vivid speech, of course, but it may indeed be true that predominantly non-literate cultures inculcate a more biddable mind’s eye for memory training and that this mental/visual faculty so honed would be peculiarly susceptible to vivid descriptions. This would certainly explain a great deal about why our sources speak about enargeia the way they do and with such perfect consistency. But even if we reject this thesis as unverifiable speculation and assume that the

Anonymous Seguerianus, Art of Rhetoric 96. For Cicero, see Quint. Inst. 6.2.32; for Celsus, see 9.2.40. 25 illustratio (6.2.32); sub oculos subiectio (9.2.40). He says ‘others’ prefer repraesentatio (8.3.61) and ὑποτύπωσις (9.2.41), but he does not identify them. 26 See Vasaly 1993, 99. 23 24

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

20

MIMETIC CONTAGION

oft-repeated notion of ekphrasis as speech making vision through vividness is only a façon de parler in rhetorical writers, it is still a striking formulation, not only because it is so different from how modern critics use the term ekphrasis, but also because it seems to rest on a conception of language as a physical force capable of acting on the hearer. Even if the audience’s ‘vision’ is only figurative, this notion of speech as a thing objectively working in the world and compelling human response could hardly be more typical of ancient ideas about language, both popular and philosophical.27 In this sense, then, ekphrasis does what all language does, only at a heightened level. The force of the prefix ek- is not, as some modern critics have thought, that the ekphrasis stands ‘out’ from the surrounding text, but rather intensive.28 Ekphrasis is just a full (ek-) telling (phrasis) of what could also be told less fully. We recall Nicolaus’ assertion that ekphrasis does the same thing that narration does, only with enargeia. Similarly, Quintilian classes enargeia as an ornamentum, an additive dress-up for a narration that would be clear and acceptable without it (8.3.61). If one knows all the horrors that attend the sack of a city, the phrase ‘the city was taken’ ought to convey everything, but in fact such a bare announcement ‘insinuates less deeply into the emotions’ (in adfectus minus penetrat, 8.3.67) than a fuller description with all the lurid detail of enargeia. Indeed the mere word eversio says it all, ‘but saying it all is less than saying everything’ (minus est tamen totum dicere quam omnia, 8.3.70). Quintilian’s suggestion that enargeia ‘penetrates’ to the inner recesses of the hearer reveals something important about the mechanics of its effect. The conception of language as a physical force seems to be in play here; enargeia gives speech a heightened power to pierce our defensive boundaries. Where a plain narration of facts makes it only as far as the ears (usque ad aures valet, 8.3.62), speech with enargeia gets all the way to the mind’s eye. But what is the faculty that enables us to generate enargeia in the first place, and what exactly happens once this charged speech reaches the soul of the hearer? The answer to these questions requires recourse to another Greek technical term: phantasia. We may be tempted to translate phantasia as ‘mental picture’ (Quintilian proposes the Latin visio), but Quintilian’s use of the Greek word suggests that phantasia is a rhetorical term of art. Ps.-Longinus explains that phantasia can popularly (κοινῶς) denote ‘any idea (ἐννόημα), wherever it comes from, that generates speech’, but the word has taken on a more technical sense for situations ‘where you seem to see what you describe and place it before the eyes (ὑπ᾽ ὄψιν) of those listening’.29 This obviously echoes the definitions of ekphrasis and enargeia we have already seen, but here the visual

Ong 1988 is perhaps still the best place to begin exploring the consequences of orality for cultural studies at large. For the ‘psychodynamics’ of the experience of language in pre-literate cultures, see esp. pp. 31–2. For the physical force of ancient speech and its relevance to enargeia, see Webb 2009, 98–100. 28 29 See Webb 2009, 74. Longin. Rh. 15.1. 27

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

INTRODUCTION

21

experience is not only for the audience, but also for the speaker. Ps.-Longinus likes the term phantasia, but others, he says, call it eidolōpoiia (‘image making’, 15.1). Indeed, there is an objective production of images, visible not only to the ive recipients of the speech, but also to the agent of its production. The practical purpose of phantasia in poetry is ekple ̄xis (‘astonishment’), and in speeches enargeia (15.2), however phantasia is not in itself a feature of language, but an imaginative faculty, a physical process, experienced by the speaker first and then ed on to those listening as if by contagion. Quoting two Euripidean ages portraying the madness of Orestes in vivid , ps.-Longinus remarks that the poet himself saw the Furies, and that what he had thus conceived with phantasia he compelled (ἠνάγκασεν, 15.2) the audience virtually to see as well. So one important difference between phantasia and the comparatively inert modern notion of a ‘mental picture’ is its reliance on compulsion. Indeed, as the text goes on to say, in prose speeches no less than in poetry phantasia does not just persuade the listener, it ‘enslaves’ him (δουλοῦται, 15.9). This treatment of phantasia in On the Sublime is extremely similar to Quintilian’s fuller explanation of the process behind enargeia. ‘When it comes to arousing emotions, the key is being aroused ourselves’ (circa movendos adfectus in hoc posita est, ut moveamur ipsi, 6.2.26). The reason imitation of grief, anger, or indignation is sometimes ridiculous is that ‘we only adapt our words and face without adapting our interior state’ (verba vultumque tantum, non etiam animum accommodarimus, 6.2.26). quare, in iis quae esse veri similia volemus, simus ipsi similes eorum qui vere patiuntur adfectibus, et a tali animo proficiscatur oratio qualem facere iudici volet. an ille dolebit qui audiet me, qui in hoc dicam, non dolentem? irascetur, si nihil ipse qui in iram concitat ei quod exigit simile patietur? siccis agentis oculis lacrimas dabit? fieri non potest: nec incendit nisi ignis nec madescimus nisi umore nec res ulla dat alteri colorem quem non ipsa habet. primum est igitur ut apud nos valeant ea quae valere apud iudicem volumus, adficiamurque antequam adficere conemur. (6.2.27–8) Therefore in situations where we wish to be lifelike, we should assimilate ourselves to the emotions of those who are truly in the grip of ion. Our speech should proceed from the kind of spirit it wishes to produce in the judge. Or will anyone be brought to pain listening to me, if I am in no pain myself when this is my object in speaking? Will he get angry, if the one summoning him to anger feels nothing like the emotion he is trying to stir up? Will he weep, if the speaker’s eyes are dry? Impossible. Nothing starts a fire but flame, we do not get wet unless there is moisture, there is no thing that es on to another a colour it does not itself have. So those feelings should first be strong in us that we wish to be strong in the judge, and we should be affected before we try to create affect.

But how should we go about generating such feelings in ourselves? This is where phantasia enters his , the faculty by which ‘images of absent things are so presented to the mind that we seem to have them present and see them with our

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

22

MIMETIC CONTAGION

eyes’ (imagines rerum absentium ita repraesentantur animo ut eas cernere oculis ac praesentes habere videamur, 6.2.29). With frequent practice and attention this faculty grows strong, and the adept imaginer, the euphantasiot̄ os, as he calls him, ‘will be most powerful in affects’ (is erit in adfectibus potentissimus, 6.2.30). The mental activity usually squandered in daydreaming can be harnessed to the production of phantasiai, and Quintilian gives a vivid example himself of an imagined murder scene complete with dramatic and gory details and then asks, ‘Will not the blood, the pallor, the groans, the very last gasp of the dying victim be imprinted on the mind? The result is enargeia’ (non animo . . . insidet? insequetur enargeia, 6.2.31–2). Whose animus does Quintilian mean here? English translators (Watson 1856, Butler 1920, Russell 2001) seem to assume he means the speaker, but the Latin will allow either the speaker or the listener, and of course Quintilian means both, in that order. Enargeia can more accurately be said to show (ostendere) than tell (dicere), and it generates the same emotions in the audience as if they were present at the events live (6.2.32). He quotes several phantasia-inspired ages from the Aeneid and notes that the poet had deeply conceived an image of death (poeta penitus ultimi fati cepit imaginem, 6.2.33). Thus the audience’s visual experience is immediately caused by and in turn proves the same visual experience deep in the poet’s soul. Lest we imagine this vision sharing is peculiar to poetry, Quintilian transitions directly from poetic illustrations to precepts for rhetorical practice, just as ps.-Longinus did, stressing that these techniques are as necessary for an advocate or other public speaker as for a tragic or comic actor (6.2.34–6). Phantasia is at least as different from ‘mental picture’ as enargeia is from ‘vividness’, and for roughly the same reasons.30 Both ancient concepts are rooted in a strongly physicalist notion of language and of the mind’s eye, and both concepts presume that ekphrastic speech can exert a surprising degree of control over the audience’s thoughts and feelings. Ps.-Longinus, as we have seen, goes so far as to say that phantasia ‘enslaves’ the hearer (δουλοῦται, 15.9), and Quintilian implies that a speech with enargeia ‘completely dominates’ him (plene dominatur, 8.3.61). Even if these expressions of compulsion are taken to be hyperbolic, they indicate something important about how ekphrasis limits diversity of audience response by imposing a unified pattern of thought, feelings, and visual experience on anyone listening. But what is the relationship between this shared vision and the subject of the description itself? In the examples mentioned here, both poetic and rhetorical, ps.-Longinus and Quintilian seem to assume that the characters of the speaker’s narration form a third point on the affective triangle established by For an of phantasia distinguishing it sharply from ‘creative imagination’ and asg it ‘more properly to the object which engenders an impression than to the mind which is impressed’, see Tanner 2006, 259. 30

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

INTRODUCTION

23

ekphrasis.31 The speaker thinks himself into the emotional and visual world of the characters and then foists that imagined reality on to the audience, so that both he and they are pulled into conformity with the pattern of thought and vision of the narrated characters. But in these cases the speaker has access to this pattern only through his own imagination, since the characters are either literary fictions or parties to a legal dispute whose hearts and minds are no more visible than the heroes of myth and thus open for fictionalization in the same ways. Is the situation not somewhat different when the object of description is not an imaginary scene but a visible work of art? In some sense no, for hardly any ancient ekphrasis of a work of art confines itself entirely to what can be seen, and if there is some mythical or other narrative context the speaker routinely uses the image as a springboard into the larger story.32 Furthermore, the speakers in the examples cited by ps.-Longinus and Quintilian are not entirely free to imagine whatever they want, since they are constrained by the datum of the mythological tradition or the indisputable facts of the case. But however much this givenness may limit their imaginative freedom, it is not visual in nature, so it operates on different field from the phantasia itself. By contrast, when the object of description is a visible work of art, it supplies the fundamental visual datum, which the speaker may indeed supplement with collateral vividness, but which can hardly fail to be the basis for the visual pattern experienced by speaker and audience. Because this work of art is itself the product of creative visualization, it too relies on the mechanics of phantasia for its very existence. The image which came to be in the mind of the artist, whether from witnessing live events or other works of art or from the unprompted resources of his own imagination, is given objective form and transmitted to the mind of the viewer, who in this case is the composer of an ekphrasis mediating the same phantasia to his audience or readers. If the phantasia experienced by all these creators, viewers, and readers is really the same, then phantasia creates a hall of mirrors effect down through untold layers of artistic and literary creation, so that the reader of an ekphrastic description of a work of art can peer into the mind of the ekphrast, back to the artwork itself, into the mind of the artist, back to his sources, etc. Where we moderns might be inclined to emphasize the loss, distortion, amplification, or originality that occurs at each of these stages, the ancients were much more interested in the continuity of imaginative visual experience, and their formulation of phantasia provided the conceptual framework to explain how such continuity is possible. In this sense, too, phantasia provides reliable access to reality outside oneself. Rhetorical theorists were not the only ones to use phantasia as a technical term to describe a peculiarly objective form of imagining. The Stoics relied on phantasia

31

Aygon 2004, 118–19.

32

Nicol. Prog. 69.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

24

MIMETIC CONTAGION

to denote the perception of sensory objects or the mental presentation of rational but non-sensory cogitations, in contrast to phantasmata, by which they mean mere figments of the imagination, illusions, dreams, and the like.33 The dichotomy for the Stoics is not between visible and invisible content or vivid and sparsely detailed, but between true and false. Obviously there are important differences between rhetorical and Stoic phantasia. For example, a rational conclusion, however correct it may be, would presumably not fit the rhetoricians’ understanding of phantasia if it did not have a visual component, and conversely a highly visual daydream would not satisfy the Stoic definition. But these two conceptions of phantasia do overlap in their remarkable assumptions about the uniformity of mental picturing; the rhetoricians and the philosophers would agree that all who correctly and clearly perceive a visible object have the same phantasia. As we have seen, the ekphrastic description of a work of art opens up a special case of rhetorical phantasia in that it implies a vista of successively identical visions, and our sources emphasize that such visions must somehow originate in an experience of immediate co-presence with the described object.34 Rhetorical phantasia comes very close to philosophical phantasia here, for if the content of the transmitted image is not just universally identical among participants, but also originates with a clear and direct perception of the object, then the phantasia is not only objective, but also, in some sense, true.35 I would not wish to push this possible connection between rhetorical and Stoic phantasia too far. Rhetoric is not philosophy, and though the implicit role of truth may be drawn out as a hidden and probably unrecognized consequence of their formulation of phantasia, the rhetoricians claim to be interested only in likeness to the truth (veri similia, 6.2.27). However, as different as these two versions of phantasia are from each other, it is interesting that they come closest to agreement 33 A good selection of Stoic texts on phantasia is collected in Long and Sedley 1987, vol. 1, }} 39–40 (pp. 236–53). The editors focus particularly on ages relevant to kataleptic phantasia, the controversial notion of a ‘cognitive impression’, an infallible of sensory data, to which all normal human beings have immediate access, but which only the sage is able to discern perfectly. 34 After a particularly vivid age from Euripides, ps.-Longinus apostrophizes the reader: ἆρ᾽ οὐκ ἂν εἴποις, ὅτι ἡ ψυχὴ τοῦ γράφοντος συνεπιβαίνει τοῦ ἅρματος καὶ συγκινδυνεύουσα τοῖς ἵπποις συνεπτέρωται; οὐ γὰρ ἄν, εἰ μὴ τοῖς οὐρανίοις ἐκείνοις ἔργοις ἰσοδραμοῦσα ἐφέρετο, τοιαῦτ᾽ ἄν ποτε ἐφαντάσθη (‘Would you not say that the writer’s soul is mounting the chariot together with [Phaethon] and has risked it all flying off with the horses? If he were not running alongside those celestial bodies, his phantasia would never have conceived such things’, 15.4). 35 Webb (2009, 117 n.35) criticizes Jaś Elsner for failing to distinguish between philosophical and rhetorical phantasia and thus being led to the erroneous conclusion that ‘ekphrasis . . . tells the truth’, (1995, 27). She is right in principle to insist on the distinction between these two versions of phantasia, but it may be precisely in this sense that the two versions blur together: if we take seriously the claim that rhetorical phantasia establishes a link of shared visual experience between speaker and hearer and if the speaker is describing a visible object, then the validity of the hearer’s apprehension of the object will be in direct proportion to the successful enargeia of the ekphrasis.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

INTRODUCTION

25

when the object of perception and description is a visible work of art. It is also interesting to note that Stoic phantasia, most clearly exemplified in the immediate perception of the truth, is not a fundamentally different faculty than the one that enables us all to see the objects around us.36 Infallible insight is only a developed form of this universal ability. Similarly, as we have seen, Quintilian regarded phantasia as a more disciplined application of the same faculty used in daydreaming, and the Progymnasmata saw enargeia as a supercharged form of narration, doing the same thing as all descriptive language does but with more force. Ekphrasis is defined by its paradoxical ability to make an absent object visible with words, and it accomplishes this by relying on phantasia and enargeia, which at first seem just as miraculous, but turn out only to be a heightened or clarified form of very usual activities. In this study I argue that any act of mimesis, as an imitation of some model, may suggest the possibility of further copying of itself. This suggestion is in most cases entirely implicit and it may even be suppressed because of its unruly potential, but when it is actualized the result is mimetic contagion. Ekphrasis is neatly analogous to mimetic contagion in that it first seems to be an anomalous form of language, but on closer inspection it turns out to be only an unusually focused form of speech, thus revealing the potential force of all language. More specifically, though, the ekphrastic description of a work of art intersects with the topic of mimetic contagion because it is a recapitulative response to an artwork. Not all ekphrasis is mimetic contagion, any more than all mimetic contagion is ekphrasis, but the description of an artwork entails the creation of a new work, in a different medium but in imitation of the original. Among the Progymnasmata writers, Hermogenes says that the style of ekphrasis should ‘match (συνεξομοιοῦσθαι) the subject matter’, using flowery language to describe a meadow, for instance, and Aphthonios says it should ‘completely imitate (ἀπομιμεῖσθαι) the thing being described’.37 More generally enargeia creates a visual experience, so it may naturally be described as a visual medium. John of Sardis, the Byzantine commentator on Aphthonios, says that enargeia works by ‘imitating (μιμούμενος) the painters’ art’.38 It is the contagious quality of phantasia, however, that establishes the relationship between ekphrasis and mimetic contagion at the deepest level. The ekphrast trains himself into imaginative conformity with the character and emotional life of the object of his description and thus impresses its image deep in his soul. By speaking from his own visual experience in this inspired state he impresses the same image on the soul of his audience and draws them with him into parallel conformity. 36 37 38

See Long and Sedley 1987, vol. 1, 249. Hermog. Prog. 23; Aphth. Prog. 38; see Webb 2009, 57. Sardianos Commentarium 217.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

26

MIMETIC CONTAGION

I shall demonstrate in Chapters 2–4 that mimetic contagion is a recurrent topic of interest and a familiar way of thinking about image and reality across all periods of Greco-Roman antiquity and in a great variety of genres and social practices. By contrast, our sources for this elaborated discourse of ekphrasis, enargeia, and phantasia are all rhetorical theorists from the first and second century CE and their Byzantine commentators. Because of the tight uniformity of definitions across these sources and the echoes of both vocabulary and concepts in Aristotle and Hellenistic thought more broadly, it seems likely that these Imperial testimonies reflect older traditions of talking about ekphrasis, at least in rhetorical and educational discourses. Their reliance on a strongly physicalist conception of language certainly makes their theory of ekphrasis seem like a natural extension of a very old and widespread view in antiquity. If nothing else, Quintilian’s repeated references to Cicero’s word choices for translating these Greek would seem to imply that, as in other areas, there was an established Hellenistic technical tradition for which Cicero was an influential figure in shaping the Latin technical vocabulary, but for which he was generally far from the first Roman enthusiast. However, all this must remain speculation. The relevance of this discourse of ekphrasis for our investigation of mimetic contagion is that it represents a tantalizingly similar development in ancient thinking about art and response. Unfortunately, without evidence from before the first century CE, we cannot know whether this elaborated theory of ekphrasis may be too historically and generically limited to be trusted as a hermeneutic key for the rest of ancient literature. So it would probably be hazardous to apply this understanding of ekphrasis directly to Terence’s Eunuch or to rely on it as a straightforward explanation for the contagious force of the painting in the play. What we do, however, indisputably see in the Eunuch is a continuation of a very old tradition of ekphrastic description of a work of art, which had by the second century BCE developed into a self-consciously elaborate art form in its own right. How much of the later theory of ekphrasis was already in place we can only guess, but in the Hellenistic period ekphrasis was obviously taken very seriously as an object of focused literary attention. It will be worth ing in the study that follows that the painting in the Eunuch is backstage, i.e. totally fictional. Whatever role the painting ultimately plays in the comedy, it only exists in ekphrasis.

INTRODUCTORY EPILOGUE

The frontispiece of this book reproduces Rembrandt’s Danaë of 1636, a renowned icon of sexual vulnerability, now hanging in the Hermitage in St. Petersburg. Her legs are simultaneously covered and uncovered, pressed together and exposed; her right hand is raised in a gesture of fear or awestruck expectation, and her body is

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

INTRODUCTION

27

either turning towards or away from the strange golden light that suffuses her bed. Jupiter is not portrayed, except as a radiance focused on Danaë’s hesitant but unprotected body, and in this regard he seems to be a figure for the viewer’s gaze.39 If the viewer’s perspective is therefore written into the painting in sympathy with the role of the god, at least one viewer has been drawn into an excessively literal and horrific recapitulation of Jupiter’s act. In the summer of 1985 a man, variously described as a deviant, a lunatic, or a political dissident, entered the Hermitage and vandalized the painting with a knife and sulphuric acid.40 In an understandable attempt to thwart ill-gotten fame, the man’s identity and purpose have been kept largely secret, but some sense of his motivation may be gleaned from the details of the assault: ‘the knife attack began at the subject’s genitals. Then the attacker put down the blade and turned to a pressurized bottle filled with acid, with which he made four ejaculatory squirts in quite separate areas: her genitals, belly, breasts, and finally her face.’41 The sexual character of the attack, so often latent in vandalism, is here explicit, but more interesting is the man’s assimilation to the role of Jupiter in the painting. The Hermitage is refashioned into another Acrisius’ tower for the rape of the painted Danaë, just as Thais’ cubiculum becomes a ‘stage’ for the rape of the mimic Danaë, with both rapes in imitative response to the same painted image. This observation does nothing to absolve the attacker of guilt for his crime, still less does it imply blame for Rembrandt’s painting, as if it somehow invited the assault. However, when we understand the logic of response as embedded in the work of art itself, violent imitation must not be confused with random brutality. As we have said, the phenomenon of mimetic contagion may be traced across the ancient Mediterranean world and well beyond. A more striking illustration of this continuity could hardly be found than this uncanny parallelism of response between Chaerea and the unnamed vandal of Rembrandt’s Danaë.

The classic feminist treatise on the scripting of eroticized (male) scopic desire in art is Kappeler 1986. She problematizes the distinction not only between pornography and the female nude, but in principle that between pornography and representation tout court, an objection to mimesis (though Kappeler never uses this term) that makes her critique relevant to the philosophical scruples explored in Chapter 3. But where Plato is concerned about the potential effect of the image on the viewing subject, Kappeler draws attention to the violence of representation as experienced by the represented (female) object and the complicit role of the viewer as consumer. See especially her ‘Problem 6: Why Look at Women?’ (1986, 63–81). 40 41 Russell 1997. McKim-Smith 2002, 29. 39

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

ONE

Judging Chaerea The Role of the Painting

It must be confessed openly that here for once Terence presents us with pictures which are or may be deleterious. Gilbert Norwood (1923, 61)

The genesis of crime, whether it be in the human heart itself or in the promptings of some external inducement, is one of the great mysteries of our common experience. The only safe presumption may be that no two cases are exactly alike and that no two cultures approach this question with exactly the same framework of ethical and psychological premises. Within a work of literature words have an especially great significance in shaping our conception of what happens and why, but the slave Pythias, who later comes on and describes Pamphila’s physical and emotional state soon after the attack, relates that Pamphila is in no condition to talk about what happened (659), and she herself was not an eyewitness to the outrage. Chaerea’s narration of the rape is thus our only window on to the backstage event at the centre of the play, and though we may regard his perspective as hopelessly partial, it also occupies a uniquely focal place in any interpretation of this pivotal juncture. It is worth quoting again the text of Chaerea’s ekphrastic confession:1 dum adparatur, virgo in conclavi sedet suspectans tabulam quandam pictam: ibi inerat pictura haec, Iovem quo pacto Danaae misisse aiunt quondam in gremium imbrem aureum. egomet quoque id spectare coepi, et quia consimilem lat iam olim ille ludum, inpendio magis animu’ gaudebat mihi, deum sese in hominem convortisse atque in alienas tegulas venisse clanculum per inpluvium fucum factum mulieri. at quem deum! ‘qui templa caeli summa sonitu concutit.’ ego homuncio hoc non facerem? ego illud vero ita feci – ac lubens.

1

For a fuller recapitulation of the events of the first half of the play leading up to this scene, see pp. 1–4.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

THE ROLE OF THE PAINTING

29

While [the bath] is prepared the girl is sitting in a chamber Looking at a certain painted : on it was this picture, How they say Jupiter once sent a golden shower into the lap of Danaë. I too started looking and because he played a very similar game Once upon a time, all the more my heart rejoiced, That a god should turn himself into a man and creep Onto someone else’s roof tiles to trick a woman through a skylight. And what a god! ‘Who shakes the quadrants of the sky with his clap.’ And I, mere mortal, shouldn’t do it? You bet I did it, and I loved it!

The progression in Chaerea’s narration from painting to mythological subject to consummated crime is quick and clear, leaving little room for doubt as to the painting’s relevance, and his affective response highlights the revelatory quality of the painting’s suggestion. It sounds, from this description, as if the painting marks a turning point in the events backstage, and Donatus provides strong confirmation of such a reading: RESPECTENS TABVLAM QVANDAM PICTAM: bene accedit repente pictura ad hortamenta aggrediendae virginis, ideo quia non ad hoc venerat Chaerea, ut continuo vitiaret puellam, sed ut ‘videret, audiret essetque una’, cum nihil amplius cogitare ausus fuerit, usque dum picturam cerneret. LOOKING AT A CERTAIN PAINTED : it is fitting that the painting comes up suddenly to encourage Chaerea to assault the maiden, since he had not come for the purpose of immediately raping the girl, but to ‘see her, hear her, and be together with her’, for he would not have dared to think of more until the moment when he saw the painting. Several modern critics have quoted this age in Donatus with tentative approval or have made the related point that there is no clear mention of rape before this moment in the play.2 But these scholars have made no attempt to answer the arguments of the majority of critics, who ignore Donatus’ testimony or seek to refute it. Perhaps it is not so surprising that many twentieth-century critics have refused to take Chaerea’s narrative of the painting’s intervention at face value, or that they have been generally ill-disposed to Chaerea and inclined to 2 Barsby (1999a, 198) says Donatus’ interpretation is ‘probably . . . correct’, but he maintains that ‘it would be in character for Chaerea to seize upon a plausible excuse for his behaviour’. Saylor 1975 emphasizes the theme of planlessness throughout, so he makes use of Donatus here (305). Philippides 1995 also gives unusual weight to Donatus’ comment on the painting (281). Hers is the most intense engagement with Donatus’ interpretation of this scene. Donatus sees the bath and laying-out of Pamphila in 593 as a reference to the ritual preparation of the bride by her pronubae. Philippides takes up this argument and sees it as a clear mitigation of Chaerea’s crime. Donatus’ observation is quite shrewd and I am inclined to take his word on such cultural associations, but I am not sure I agree with Philippides that by itself the suggestion of wedding ritual does much to extenuate the crime.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

30

MIMETIC CONTAGION

sympathy with Pamphila. She appears on stage only briefly as a mute character and is quickly ushered into Thais’ house where she will be assaulted and kept hidden for the rest of the play. As the Eunuch’s ‘madwoman in the attic’, Pamphila’s total silence and near invisibility have made her presumed point of view all the more persuasive for many of us, and the concomitant reaction against her assailant has been severe. It is hard to imagine how earlier ages could have been so friendly to this character.3 When sco de’ Nobili, one of the bright lights of early sixteenth-century Venetian public life, took the name Cherea, presumably he did not mean to characterize himself as a brutal rapist.4 Chaerea’s influence shows itself clearly in the warm adaptations of his character in Renaissance and Restoration comedy.5 In the nineteenth century the great Terence editor Philippe Fabia summed up his age’s response to Chaerea, calling him ‘une des figures les plus originales, les plus vivantes et les plus attrayantes de la comédie antique’ and comparing him to the ionate, impulsive Chérubin in Beaumarchais’ Le Mariage de Figaro.6 Chaerea still had fans in the early twentieth century. Casper Kraemer calls him ‘a young, likable boy, with an impetuous, ionate, rather self-centered disposition’.7 But his focus is less on the lovability of the character than on his verisimilitude. ‘Chaerea might well find his present-day counterpart on any college campus or, with some little additional vulgarity, on any street corner in a modern city. He certainly is not unnatural in his enthusiastic attention to female beauty. His own claim to connoissieurship [sic] in matters of this sort is amply ed by the “raffish succulence of phrase,” which may be reprehensible, but is still a familiar characteristic of boys at his age.’8 Edward Kennard Rand finds him ‘impulsive, ionate, tender, resourceful, manly, pious, true, a Catullus in action, scandalously indecorous, irresistibly lovable’.9 Given that

3

And indeed they were not equally friendly to him in all contexts, especially those involving young actors and audiences. For a very helpful survey of the bowdlerization of Roman comedy, particularly of this scene of the Eunuch, when it was regularly performed by boys as the Westminster Latin Play from the sixteenth century to the twentieth, see Brown 2008. 4 On Cherea de’ Nobili’s remarkable career, not just as an actor and a figure in early Commedia dell’ arte, but also as a sometime diplomat, chancellor, and spy, see Szakolczai 2013, 181–6. 5 One thinks most readily of Viola in Twelfth Night (‘Thou shall present me as an eunuch to him’, 1.2.56). The Eunuch has been called Shakespeare’s favourite play, perhaps because of the quotation in The Taming of the Shrew (1.1.158): ‘If love have touched you, nought remains but so; | redime te captum quam queas minimo’. If nothing else, Viola’s reference to her singing ability should make the allusion to Chaerea clear, but of course she is a eunuch of a very different sort. See Elam 1996. Closer English avatars are Erostrato in Gascoigne’s The Supposes (1566), Lionel in Sir Charles Sedley’s Bellamira, or The Mistress (1687), and Courtly, Jr in Thomas Cooke’s The Eunuch, or the Darby Captain (1736). Sedley’s Lionel is arguably fairly dastardly, but certainly no more than any other lover in Restoration comedy. 6 Fabia 1895, 17 n.1; ‘one of the most original, lively, and attractive figures of ancient comedy’. 7 8 9 Kraemer 1927, 667. Kraemer 1927, 665. Rand 1932, 58–9.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

THE ROLE OF THE PAINTING

31

characterization, it is not too hard to understand how Rand can refer to the rape as a ‘mad prank’.10 However, by the time of Kraemer and Rand another critical reaction was already afoot. In fact, they were both writing in response to the strong condemnation expressed in Gilbert Norwood’s book-length study of Terence’s oeuvre, a condemnation that would become typical of scholarship from the second half of the century. Norwood found fault with a tradition that preferred Plautus, ‘who wrote plays like a blacksmith mending a watch’, and was concerned at every turn to show Terence’s brilliant craftsmanship, truth to life, and good taste.11 He provides a canonical narrative of the young poet’s progress from the Andria, fine in its language but faulty in its construction, to that masterpiece of comedy and humanity, the Adelphoe. Along the way, he must it, the Eunuch ‘marks a faltering in our poet’s progress’. (57) The language is as elegant as ever, but certain ‘abjectly Plautine’ elements mar the plot, most clearly the rape of Pamphila.12 Where older critics had tended to ignore the rape and love the character, Norwood sets the tone for newer critics, who have generally started by condemning the rape and moved out from there to a more holistic ethical proscription, finding vices in all the virtues earlier readers had loved. The text itself provides no clear evidence that the rape was maliciously planned, so any absolute judgment of Chaerea has to be based on a composite sketch of his personality drawn from his behaviour in other scenes. For example, Norwood pursues his case by attempting to expose Chaerea’s character through his ‘raffish succulence of phrase’ (the tag mocked by Kraemer, as noted earlier). Examples drawn from various other scenes in the play prove that Chaerea is an oversexed young man, and we are left to draw our own conclusions. The focus is less on the act itself than on Chaerea’s attitude, ‘frankly, casually, and callously using fellow-creatures as mere material. Kant has bidden us always to treat a human being as an end, never as a means only’ (61). Norwood dismisses the well-worn comparison to Chérubin, for Chaerea lacks such charming vulnerability. Far from acting on a sudden impulse, Chaerea ‘deliberately and skillfully gratifies a physical appetite’ (62). Half a century later, in the introduction to his translation of the Eunuch, Douglass Parker ed a similar disapproval of Chaerea’s character, noting his ‘compulsive gregariousness’ and ‘smarmy disingenuousness’.13 Once again, the basis of condemnation is not found in the description of the rape, but rather in Chaerea’s behaviour in other scenes. Instead of Norwood’s explicitly Kantian ethical lens, Parker’s language reveals a focus on the diagnosis of Chaerea’s pathology of character. The debt to a

11 Rand 1932, 62. Norwood 1923, 1. Norwood 1923, 61. He also disapproves of the end, with its gross mistreatment of Thais, but singles out the rape as ‘perhaps the most objectionable feature of the whole’. 13 Parker 1974, 151. 10 12

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

32

MIMETIC CONTAGION

broadly psychoanalytic ethical and descriptive system is clear: sociability is automatically under suspicion of being compulsive, and few things are more contemptible than insincerity. In fact, this vaguely Freudian com plays the same role in his reading as that played in Norwood’s by Kant’s version of the golden rule.14 Writing today, at some remove from both Kantian and psychoanalytic ethical models, it is deceptively easy to historicize Norwood’s and Parker’s reading and to imagine that we see just the point at which an anachronistic moral system intrudes upon their reaction to the character. And yet their condemnation is so immediately intelligible that we are not surprised to see it repeated with ever greater vehemence as the twentieth century draws to a close. Several scholars have followed W. S. Anderson in finding the happy ending an insufficient absolution for a character so ‘utterly sober and deliberately underhanded in his attack’.15 The strongest statement of this position is from L. P. Smith, who takes issue not only with Chaerea’s defenders, but also with his comparatively indulgent earlier critics: ‘All these evaluations seem too cautious, lenient, and deferential; Chaerea is, after all, a soldier absent without leave who, on his way to a party, enters a private house under false pretences and rapes a young girl.’16 She argues that the Roman audience would be predisposed to a negative view of Chaerea first because of his dereliction of his post and second because of the shamefulness of his disguise. The legal consequences of rape were more serious in Rome than Athens, and Smith seems to be the first scholar to point out the very harsh punishments the Roman audience would be likely to anticipate for Chaerea (22–3). Rather than ignoring the role of the painting, as most scholars have done, Smith insists that we disbelieve Chaerea’s narration on this point as a lame excuse and not the least despicable aspect of his character: Chaerea justifies what he did by arguing that he imitated what he saw represented in a picture in Pamphila’s room. This is transparently as disingenuous and dishonest as his earlier specious justification of his entry into Thais’ house (382–87) as vengeance upon one meretrix on behalf of the sufferings of all men at the hands of any and all meretrices. That statement, justifying violence against meretrices as an appropriate response to their behavior, by a character who then commits rape, can be recognized as specious rhetoric by the audience, who have already heard Chaerea announce his real motive: he is in pursuit of the most beautiful girl he has ever seen. (293–97)17

K. F. Pierce sees it similarly: ‘Chaerea aligns himself with the king of the Roman gods—Jupiter—so as to provide himself with an excuse for raping a girl.’18 14 On the subject of Freud’s intellectual genealogy in Kant, see Fell 1976. More recently and for the more general psychoanalytic debt to Kantian ethics, see Gondek and Widmer 1994. 15 Anderson 1984, 131. For similar arguments see Hunter 1985, 94 and Forehand 1985, 75. Goldberg (1986, 121) does not go so far, but he does emphasize Chaerea’s ‘deceit’. 16 17 18 Smith 1994, 21. Smith 1994, 27. Pierce 1997, 175.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

THE ROLE OF THE PAINTING

33

V. J. Rosivach, in his survey of rape in New Comedy, focuses on the uniqueness of this violation: ‘The rape here is deliberately planned, not, as elsewhere, a spurof-the-moment affair.’19 Once again the painting enters the only to provide further evidence of the rapist’s depravity: ‘Chaerea’s language here shows clearly that in his mind what he and Jupiter have done, sneaking into a house and raping a young woman within, is nothing but a “game” (ludus), a lark’.20 Most recently A. Sharrock: ‘Chaerea is encouraged in his plan of rape by a picture on the wall’.21 Given this weight of critical opinion, perhaps it is not enough to argue ex negativo, merely insisting that there is no positive evidence the rape was planned. But we can go further. The text seems to hint very subtly at sexual contingencies and then emphatically exclude the possibility. When Chaerea learns that Parmeno is on his way to deliver Phaedria’s gifts, he breaks out in envious raptures (365–8): CH. o fortunatum istum eunuchum quiquidem in hanc detur domum! PA. quid ita? CH. rogitas? summa forma semper conservam domi videbit colloquetur aderit una in unis aedibus, cibum nonnumquam capiet cum ea, interdum propter dormiet. CH. O happy eunuch, to be handed over into this house! PH. Why? CH. You have to ask? His fellow slave at home, Beauty’s summit—he’ll always see her, chat with her, be together with her In the same house, sometimes he’ll take a meal with her, Occasionally sleep nearby.

Love’s power to debase, a theme already given explicit treatment by Phaedria and Parmeno in Act 1, is here again brought to the fore, this time with a comic turn: a virile, young Athenian citizen is reduced to envying a decrepit, castrated slave for the sole sake of his access to Pamphila. Only a few lines before, Chaerea had described the eunuch as ‘a vile person, fresh off the auction block, an old woman’ (inhonestum hominem, quem mercatus est heri, senem mulierem, 357). The five activities the lucky eunuch will share with Pamphila may seem to be taken at random from his day, but on closer inspection they turn out to make an ascending series, each term implying greater familiarity than the last. Any ascending polycolon of intimacy will probably generate an expectation of the ultimate closeness: actual sex. Of course, in the eunuch’s case this is an expectation that will have to be disappointed. He may make it through third base, but home plate will be anatomically off-limits. Even in his absurdly idealized portrait of a eunuch slave’s life, Chaerea veers at the end to

19 21

Rosivach 1998, 46. Sharrock 2009, 222.

20

Rosivach 1998, 47.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

34

MIMETIC CONTAGION

‘sleeping nearby’. Donatus’ comment confirms this reading and even indicates that the ancients had a fixed programme of ‘bases’: the five lines of love. VIDEBIT COLLOQUETUR ADERIT UNA: amatorie nimis quinque lineas amoris exsecutus est, adeo diligenter, ut etiam ordinem custodiret. [ . . . ] istae enim sunt amoris lineae, etsi non omnes est persecutus. HE WILL SEE HER, CHAT WITH HER, BE TOGETHER WITH HER: just like a lover, he has followed the five lines of love, and so carefully that he has even kept the order straight. [ . . . ] For these are the lines of love, even if he has not gone through all of them. Donatus assumes the system of five lines is not an anachronistic construction, but rather that Terence is alluding to them, both by keeping to them (custodiret) and by diverging from them (non omnes est persecutus). In Act 4 there is striking confirmation that he is right—Terence also knows the lines of love. Phaedria returns from the country ahead of schedule, and explains that if he cannot get to second base, he may as well be content with first (638–41): nil est. quid ‘nil’? si non tangendi copiast, eho ne videndi quidem erit? si illud non licet, saltem hoc licebit. certe extrema linea amare haud nil est. It’s nothing. Nothing? If he can’t cop a feel, Can’t a guy even look? If the one thing is out of reach, At least the other will be ok. Hey, to love from The outmost line is hardly nothing.

Donatus’ note is more explicit this time, and he spells out the five lines: CERTE EXTREMA LINEA: et hoc recte, quia quinque lineae perfectae sunt ad amorem: prima visus, secunda alloquii, tertia tactus, quarta osculi, quinta coitus. TO LOVE FROM THE OUTMOST LINE: just so, because five lines are traversed to love: first, seeing; second, talking; third, touching; fourth, kissing; fifth, sex. Returning to Chaerea’s fantasy with the quinque lineae in place, we note several things about the eunuch’s progress. The first two lines, looking and talking, are kept exactly and establish the paradigm. The third line, touching, will not quite do for the eunuch, who will have neither desire nor pretext to fondle Pamphila, so the of bodies has been weakened to proximity (aderit una in unis aedibus). The fourth line, kissing, would be even more out of bounds for the eunuch, but the analogy of kissing and eating is very widespread in Greek and Roman erotic

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

THE ROLE OF THE PAINTING

35

languages, and so the emphasis on shared oral activity is kept (cibum nonnumquam capiet cum ea).22 Finally, and perhaps most transparently, the fifth line has been bleached to what we might think was a euphemism for sex, if we did not know this to be impossible: ‘he will sleep hard by her’ (propter dormiet). The modification of the five lines of love is as significant as their invocation. In a pregnant ellipsis, the text adumbrates sex, only then to cut it out. It would be absurd to insist that Chaerea alludes to sex here to suggest the sinister things he would do in the eunuch’s place. Rather, the point is surely that love has reduced him to envying a truly unenviable character, and his desire merely to see Pamphila again and to be in her presence is so great that he can imagine full satisfaction in the limited pleasures the ‘fortunate’ eunuch will soon enjoy. Only now does Parmeno suggest impersonation and substitution, and Chaerea’s slowness on the uptake confirms that his thoughts have not been running to such things and emphasizes that Parmeno is the planner here. Indeed, this emphasis is another serious argument against taking Chaerea as a shrewd mastermind of anything. PA. quid si nunc tute fortunatu’ fias? CH. qua re, Parmeno? responde. PA. capias tu illi(u)s vestem. CH. vestem? quid tum postea? PA. pro illo te ducam. CH. audio. PA. te esse illum dicam. CH. intellego. PA. tu illis fruare commodis quibu’ tu illum dicebas modo: cibum una capias, adsis tangas ludas propter dormias; quandoquidem illarum neque te quisquam novit neque scit qui sies. PA. What if you were now to be ‘happy’? CH. How, Parmeno? Tell me. PA. You could take his outfit. CH. His outfit? What then? PA. Instead of him I’d deliver you. CH. I’m listening. PA. I’d say you’re him. CH. I see. PA. You’d be the one to enjoy those very pleasures You were just saying were his: You’d eat together, be with her, touch her, play with her, sleep nearby, All the while none of those women knows you or knows who you are.

From the first line it is clear that Parmeno is not proposing rape. His question looks back to the fortunatus of Chaerea’s exclamatory apostrophe four lines before, and implicit in the suggestion that Chaerea can be ‘happy’ (like the eunuch) is that he will enjoy just those pleasures he has said would be sufficient for envy. In case we miss the point, Chaerea’s confusion gives Parmeno the chance to make the reference to the eunuch’s pleasures explicit (illis . . . commodis), and to recapitulate the list of five stages, though he makes some substitutions and changes the order and the diction on all the except the last (propter dormias).

22

For the erotic uses of ‘eating’, see Adams 1982, 138–41.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

36

MIMETIC CONTAGION

Chaerea takes up the suggestion with characteristic enthusiasm, in spite of Parmeno’s second thoughts. When Parmeno warns that what they are doing is a crime, Chaerea protests that it is payback against Thais and prostitutes in general. Has the plan now changed from being modelled on the figure of the lucky eunuch to one of retribution, and if so, is rape part of what Chaerea means to justify? Smith says that Chaerea ‘makes a speech of specious justification for what he is doing as vengeance upon one meretrix for the sufferings of all young men at the hands of all meretrices (382–7), thus justifying the violent rape of one woman as revenge on others’.23 Let us look at this age more closely. PA. at enim istaec in me cudetur faba. CH. ah. PA. flagitium facimus. CH. an id flagitiumst si in domum meretriciam deducar et illis crucibu’, quae nos nostramque adulescentiam habent despicatam et quae nos semper omnibus cruciant modis, nunc referam gratiam atque eas itidem fallam, ut ab is fallimur? an potius haec patri aequomst fieri ut a me ludatur dolis? quod qui rescierint, culpent; illud merito factum omnes putent. PA. Well, your bean will be threshed on me. CH. Ha! PA. We’re doing wrong. CH. Is it wrong if I am led into a cathouse And pay back now those racks that despise us and our youth And that always torment us in every way, And if I deceive them like they deceive us? Or is it more just for this to happen to my father that he be tricked by me? Anyone who saw the latter would blame it, But they would all think the former was justified.

Smith herself has noted that the impersonation of the eunuch and entry into Thais’ house on a false pretext would already be enough to qualify as a flagitium. Is it possible that this is all they mean and rape is still not part of the plan? Parmeno’s concern is for his own culpability in what they are about to do (istaec in me cudetur faba). Given that he is the author of the plan to substitute Chaerea for the eunuch, he may well be worried about his direct responsibility for the travesty. If things go badly inside, as they do, he could also be considered responsible, as he will, but since they are both about to commit a flagitium together it is unnecessary to construe his concern at this point as focusing on a second as-yet unmentioned crime Chaerea may or may not commit once inside. This reading is confirmed in the plural that follows (flagitium facimus) and in Chaerea’s explicit question whether it can be a flagitium for him to be led into her house, as thanks for what adolescentes suffer at the hands of meretrices. Prostitutes pretend to be in love, and the young men who frequent them are duped into

23

Smith 1994, 23.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

THE ROLE OF THE PAINTING

37

believing false appearances. When Chaerea says he wants to ‘deceive them as we are deceived’ surely he means simply to turn the tables on trickery. Indeed, even if fallere could be used as a euphemism for rape (and such a colourful usage is totally unattested), Chaerea’s point of returning like for like would be lost. The next line makes this even clearer: an potius haec patri aequomst fieri ut a me ludatur dolis? If there were any hint of rape in the ‘deceptions’ of the previous lines, the anaphoric haec applied to his own father would be more monstrous than even Chaerea’s most hostile critics would suggest. Why does Parmeno not tell Chaerea what he has heard from Thais about Pamphila being an Athenian citizen? Even if he doubts its truth one might think he would on such important information. If we understood the plan, at the end of Act 2, still to be free of rape, we will gain important insight into Parmeno’s ineptitude as a servus callidus, a role that requires an active awareness of all possible contingencies. It is precisely this sudden and surprising failure of Parmeno’s callidity that Pythias makes explicit in Act 5 (1011). But if we believe that the plan now includes rape, Parmeno’s failure to mention to Chaerea what he has heard about Pamphila’s parentage becomes a slip so stupid as to strain credulity. Chaerea will return to the stage 83 lines later in his eunuch disguise, but like the Ethiopian slave girl and like Pamphila before, he will be introduced into Thais’ house without speaking a word. So after the end of Act 2 we have no glimpse of Chaerea’s intentions until he emerges from the house in 549, the rape committed and the play half over. Once again, in his narration to Antipho, he gives strong confirmation of the plan’s innocence (568–74): amare coepi. forte fortuna domi erat quidam eunuchu’ quem mercatu’ fuerat frater Thaidi, neque is deductus etiamdum ad eam. submonuit me Parmeno ibi servo’ quod ego arripui. AN. quid id est? CH. tacitu’ citius audies: ut vestem cum illo mutem et pro illo iubeam me illoc ducier. AN. pro eunuchon? CH. sic est. AN. quid ex ea re tandem ut caperes commodi? CH. rogas? viderem audirem essem una quacum cupiebam, Antipho. I fell in love. By great good luck there was A certain eunuch at home whom my brother had bought for Thais, And he hadn’t yet been given over to her. Our slave Parmeno gave me Some advice that I took up. AN. What was it? CH. Hush and I’ll tell you: That I trade clothes with him and order myself delivered in his place. AN. The eunuch? CH. Yeah. AN. What advantage could you hope to get From that? CH. You have to ask? I would see her, hear her, be together With the girl I wanted to be with, Antipho.

Antipho’s question picks up the earlier interchange about the eunuch’s ‘advantages’ (commodi), and Chaerea’s rogas? echoes his earlier rogitas? in response to

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

38

MIMETIC CONTAGION

Parmeno’s confusion about why the eunuch was so lucky. Rogitas? had served to introduce his modified five lines of love; now rogas? has the same function, but the last two lines have been stripped away and colloquetur has been replaced by audirem. Earlier he emphasized the innocence of the eunuch’s pleasure by making the last three lines increasingly divergent from the normative five. Here, with himself explicitly in the eunuch’s place, he has made the plan sound even more emasculated by cutting off the two boldest lines and weakening ‘conversation’ to the entirely ive ‘listening’. Now all that remains of the ‘real’ quinque lineae is vision, and as it turns out this is the sense that will get him into trouble. One recent scholar does advance a text-based argument that the rape was planned. Three times Chaerea uses the verb potior to talk about ‘getting’ Pamphila: mea nil refert, dum potiar modo. I don’t care, so long as I get her. obsecro hercle, Parmeno, fac ut potiar. I’m begging you, Parmeno, help me get to her. eamus; et de istac simul, quo pacto porro possim potiri, consilium volo capere una tecum. Let’s go; and meanwhile I want to get your advice On how I can have her for good.

320 362 613

Line 614 is, of course, after the rape, but if we focus on the first two occurrences of potior it may sound as though Chaerea is already planning something physical and violent. It is on this basis that Rosivach rejects Donatus’ reading. ‘“To win sexually, gain the submission of,” OLD s.v. potior 2.c. As Chaerea’s use of this verb makes clear, from the outset it was his intention to have sex with Pamphila (if not necessarily to rape her), and Donatus (ad Eun. 584) is wrong in saying that he had not entered Thais’ house with this intention but was only moved to assault when he saw a picture of Jupiter and Danaë in the house.’24 However, we may cite any of the other OLD entries for potior, especially 3 and 4, far less violent kinds of ‘getting’. The word certainly can connote ‘obtaining mastery’, but early Latin also provides plentiful examples of potior meaning ‘to reach’ (a destination) or ‘to come by’ (anything good or bad). Note, for example, the line in the younger Naevius: penetrat penitus thalamoque potitur (Cyp. Il. 2) ‘He goes within and gets to the bedchamber.’ Or the Ennian ille meae . . . potis pacis potiri (462 Vahlen) ‘He could have peace with me.’ Plautus provides numerous other examples, but the more important evidence is from Terence himself. Apart from Chaerea’s references to ‘getting to’ Pamphila, the verb potior appears seven times in the Terentian corpus. Four of these (Haut

24

Rosivach 1998, 169.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

THE ROLE OF THE PAINTING

39

322, 323; Ph 159, 830) have an amica as the implicit object, but in each case the context makes it clear that the issue is ‘getting the girl’ in an extended relationship with all the family and practical problems that arise from long-standing affairs, rather than anything specific to sex. Of course, as in English, ‘getting the girl’ may also euphemistically cover sexual intercourse, but that is not the focus in these cases. Surely this is also true of Chaerea’s use of potior in line 614. He can hardly be thinking of marriage, since he does not yet know that she is an Athenian citizen, but the use of porro (instead of, say, iterum) shows that Chaerea is imagining a long-term love affair, rather than another chance to rape her. The other three occurrences of potior in Terence (Ph 469; Ad 871, 876) are not directed to love objects at all, but to other good or bad things one can ‘obtain’. In summary, it is not clear that Terence ever uses potior in the specifically violent sense of OLD 2.c, ‘to gain the sexual submission of ’. Chaerea will ‘get the girl’ in the end, and in the middle of the play he will ‘get’ her in bed, and looking back on lines 320 and 362 from that perspective gives them rich dramatic irony. This may well be the point of the ambiguity that has led Rosivach astray. We are meant to hear obsecro hercle, Parmeno, fac ut potiar as the plea of a typically harmless comic lover who wants to ‘get to’ his girl, and the irony rests precisely on the gap between what Chaerea now intends and the more sinister fulfilment his prayer will take. Yet perhaps in offering a philological answer to Rosivach’s claim about potior we are not really going to the heart of the matter. After all, Rosivach comes at the end of the twentieth-century response to the play, and all the other scholars we have surveyed, who reached the same judgment about Chaerea’s malicious intentions, did so without the aid of any explicit text-based argument at all. If that makes their reaction to Chaerea less than scientific, the imperative to explain their universally shared revulsion becomes all the more pressing. It would be unforgivably arrogant to assume that we could sweep away such a strong hunch, common to so many learned and sensitive readers, simply by pointing out that there is no positive evidence for their view, and if we do not attempt somehow to for their reaction, we will have compounded arrogance with short-sightedness. Over the second half of the twentieth century, Western democracies underwent an historically unparalleled shift in the public consciousness of rape. A crude but objective index of this development may be seen in statutory changes, which have resulted in broader definitions of sexual assault, a proliferation of recognized types of rape (date, spousal, acquaintance, statutory, etc.), and perhaps most significantly, at least in the English-speaking world, a universal reclassification of rape as presumptively criminal.25 Parallel to these legal changes and sometimes in 25

The debate on the legal definition of rape has produced a vast literature. For a sample of some of the more influential work, see Estrich 1987, LaFree 1989, Bourque 1989, Schulhofer 1992, Berger, Searles, and Neuman 1995.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

40

MIMETIC CONTAGION

dialogue with them, the feminist debate on the nature and significance of rape has become more articulate with each successive decade.26 The revolutions of basic assumptions and goals within feminism have never taken attention away from the pressing necessity of formulating a way of thinking about rape, as a locus for the combination of problematic sex and power and as the most blatant form of gendered oppression. Changes in popular attitudes are, of course, much harder to track than legislative reforms and scholarly debates, but these discursive shifts have not transpired in a social vacuum, and for many of us there is perhaps no word in the English language so like fingernails on chalkboard as ‘rape’. By telling contrast with this lexical freighting, neither Greek nor Latin had a word to denote non-consensual sex. The crime was conveyed by more general words for ‘assault’ without sexual specificity, words that emphasize not the act but the resultative polluted state of the victim, and words for ‘adultery’ that completely ignore the issue of consent.27 This perceived paupertas linguarum antiquarum has provided a starting point for scholars interested in tracing out Greek and Roman constructions of sexual assault and the proper limitations on our retrojection of legal and moral categories.28 If there is any truth in Foucault’s claim of a modern self emergent in the discourses of sexual (self-)control, then the horror of rape, for a modern victim, rests not only in its disruption of autonomy, but in the fact that it operates within the very field of autonomy’s grounding.29 Since late antiquity, according to this

26

The beginning of the contemporary conversation and in many ways still the most prominent work is Brownmiller 1975. Brownmiller argued that rape is not sex, most fundamentally, but violence: ‘a deliberate, hostile, violent act of degradation and possession on the part of a would-be conqueror, designed to intimidate and inspire fear’ (391). The radical feminist thought on rape, best represented by MacKinnon 1989 and Dworkin 1989, essentially sought to invert Brownmiller’s formulation. Rape is different from ‘normal’ heterosexual sex only in the degree to which it unmasks the violence inherent in all intercourse, whether sexual or social, between men and women. ‘Considering rape as violence not sex evades, at the moment it seems most to confront, the issue of who controls women’s sexuality and the dominance/ submission dynamic that has defined it’ (MacKinnon 1989, 178). Of the more recent work on rape, I have found Cahill 2001 to be the most helpful. 27 The words for ‘assault’ in Greek are generally formed from bia; Latin uses the verb comprimere with or without the ablative compliment vi. The ‘shaming’ or ‘outrage’ words are mostly taken from hubris in Greek; Latin uses the verbs stuprare and vitiare. Adultery is, of course, moicheia, a classification of rape that makes spousal rape notionally impossible. For the Greek vocabulary of sexual assault, see Cole 1984, 98 and Cohen 1991. More recently and following this same tripartite division, see the comprehensive study in Omitowoju 2002. 28 Lape 2001, 84–5: ‘What the lack of a Greek lexical equivalent for rape does suggest is that “coerced sexual intercourse” in the Athenian context is not necessarily problematized in the same way or for the same reasons that we find today. For this reason, even to speak of rape in the Athenian context is, in some sense, to beg the question, importing an ideology and locus of legal and moral problematization that requires demonstration rather than mere assertion.’ 29 I am thinking primarily of Le Souci de soi, of course, but the basic idea is fundamental to much of what Foucault wrote. Rape is never an explicit subject in Foucault’s treatises, but in a roundtable discussion on

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

THE ROLE OF THE PAINTING

41

model, our cultural and institutional forms have constructed subjectivity out of sexuality, as the individualized agency for marshalling the deployment of sex. When sexual autonomy is instrumentalized for the articulation of selfhood, its brutal subversion can only be understood as a total dislocation of identity. It should be emphasized, since I have used his idiom, that this interpretation of the modern experience of rape is not strictly Foucauldian. Writing from an intellectual tradition often explicitly hostile to Foucault, Ann J. Cahill offers one of the most insightful and informed contemporary theorizations of rape. She says, ‘Rape cannot be defined or understood as theft, mere assault, or virtually identical to other forms of heterosexual intercourse. Rather, it is a sexually specific act that destroys (if only temporarily) the intersubjective, embodied agency and therefore personhood of a woman.’30 By making the physical and social trauma of rape secondary to a more fundamental dislocation of identity, Cahill provides a trenchant formulation of the atrocity of rape as it is experienced today, but she relies on a localization of agency which is itself arguably historicizable. If we posited a moment anterior to such localization, we would still expect the experience of sexual assault to be attended by the enormities of force and degradation, but without the cardinal shattering of self Cahill describes. Indeed, this would explain why ‘violence’ and ‘disgrace’ are the very semantic fields used in antiquity to denote sexual assault. This historical trajectory would also go some way towards explaining the difficulty modern critics have had reading the rape in the Eunuch. Their ethical abreaction is not superficially anachronistic, but tied to very deep, important allegiances. Even when they realize there is something distinctly modern in their response, this ission is usually made in a tone of despondency.31 Of course, this interpretive crux is not confined to the Eunuch. As strange as the ancients may seem in their forensic discourses of rape, more baffling still is its presence in New Comedy. First there is the overwhelming frequency of rape as a plot motif.32

penal reform in 1977 (published in Foucault 1988, 200–2), he advocated radical desexualization of the crime: ‘[T]here is no difference, in principle, between sticking one’s fist in someone’s face or one’s penis into their sex’. This suggestion seems to have won him few friends. As Teresa de Lauretis notes (1987, 37), misguided though it may be, Foucault’s proposal must be understood as a move of resistance against modernity’s ‘technoligization’ of sex. Cahill 2001, 13. Rosivach 1998, 169: ‘My own feeling is that he is still despicable, but that is not what the play, overall, seems to be saying.’ See also Hunter 1985, 94. 32 It is difficult to assess numbers for the badly fragmentary poets, but we are sure that at least eight plays by Menander (Georgos, Epitrepontes, Heros [two rapes!], Kitharistes, Plokion, Samia, Phasma, and the unidentified fragment on P. Cariensis 43227) and at least three plays by Plautus’ younger contemporary, Caecilius Statius (Davos, Plocium, and Titthe), involved rape. Plautus is rather restrained here, including rape in only three of the surviving twenty-one plays (Aulularia, Cistellaria, and Truculentus), but Terence, who apparently stays closer to Menander in this regard, employs rape in the plots of no fewer than four of his six comedies (Eunuchus, Phormio, Hecyra, and Adelphoe). 30 31

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

42

MIMETIC CONTAGION

But just as striking as its ubiquity is the formulaic regularity of its inclusion, where as Pat Easterling has noted, it seems to function as ‘a courtship practice’.33 In every play that relies on rape, except for the Eunuch, the crime occurs as part of the backstory, and is thus neatly sequestered from the action of the play itself. In most cases the rape was committed nine months before, and because of the victim’s untimely pregnancy, she has been abandoned by her new husband or dropped by her fiancé.34 After a series of miscommunications and revelations, and to the surprise of both parties, the assailant turns out to have been the lead, who unwittingly raped his own fiancée. If they had married before the separation, he may now return to his bride, and if they had not, the wedding is back on. Either way, a happy ending is made even happier by the birth of a son. In other cases, the rape occurred a generation before and resulted in the birth of a girl, who happens to be the ingénue.35 Here again the rape is essential to the happy ending: because of the surprising revelation of the girl’s citizen parentage, she is allowed to marry the lead, who loves her but has until now taken her for an unmarriageable prostitute. In either case, but especially in the latter, the audience is effectively shielded from the unpleasantness of the crime itself by its more or less remote occurrence. Another highly regularized feature of these rapes is the cluster of circumstances surrounding the crime. The rape almost always occurs in the context of a nighttime festival. As scholars have often noted, this provides an ‘excuse’ for the victim by explaining why a good citizen girl was away from home and vulnerable to attack.36 The dramatic interests of New Comedy simply do not revolve around undermining the ingénue’s chastity or impugning her eventual suitability to marry the lead and become a matrona. In addition to saving face for the girl, the nighttime festival context situates the event in an atmosphere of carnivalesque license and inversion of norms, and the cover of dark explains how rapist and victim could fail to recognize each other later, even after marriage. So this single contextual element provided a handy solution to several thorny technical problems involving ethical depiction of the victim and dramatic verisimilitude and it quickly became conventional. It is here, I would argue, in its conventionality, that the night-time festival context has an equally important narrative function, for the more formulaic the rape becomes, the more thoroughly it is denatured as a brutal crime and refashioned as a semiotic token, part of the sign system of New Comedy. The lack of ‘naturalism’ in this convention is therefore essential to its full utility, but such an

Easterling 1995. I am conflating two of the categories in Rosivach’s survey: ‘rapes which lead to marriage of rapist and victim’ (Rosivach 1998, 14–23) and ‘premarital rapes’ (27–35). 35 For ‘rapes whose revelation allows marriage of victim’s daughter’, see Rosivach 1998, 23–7. 36 Lape 2001, 93; Rosivach 1998, 36; Zagagi 1994, 115. 33 34

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

THE ROLE OF THE PAINTING

43

obvious departure from reality is also an irresistible incitement to self-conscious play. So, for example, Onesimos in the Epitrepontes says (451–3): ‘a night-time festival and women. It stands to reason (κατὰ λόγον), a girl was raped. She bore this child and exposed it—obviously! (δηλαδή).’ Parallel to the festival context, and clearly connected with it, is the equally universal motif of drunkenness. Here the function is plainly to mitigate the rapist’s guilt and shift the blame to the explosive compound of youthful ion and alcohol. As Susan Lape says, ‘[T]he rapist’s diminished capacity supplies a crucial exculpatory factor leading to the conventional wisdom that New Comic rapists cannot be held able for their actions because they act without thinking. They are always overcome by a potent combination of youth, ion and alcohol.’37 So in the Adelphoe, Aeschinus’ crime is rationalized by reference to the formula: ‘Night, love, wine, youth persuaded him—it’s only human’ (persuasit nox amor vinum adulescentia:/humanum est, 470–1). Most striking is that this excuse is made not by the rapist himself, but by Hegio, the victim’s representative. In the Aulularia, Lyconides, who was earlier said to be drunk when he raped Euclio’s daughter (vinolentum, 689), makes a claim similar to Chaerea’s, that it was a god that led him on (deus mihi impulsor fuit, is me ad illam inlexit, 737). He goes on to explain: ‘It was the fault of wine and love that I did it’ (vini vitio atque amoris feci, 745). Later he will pick up this reference to a divine impulsor and repeat his justification: ‘I it I wronged your daughter at the night-time festival of Ceres because of wine and the impulse of youth’ (ego me iniuriam fecisse filiae fateor tuae | Cereris vigiliis per vinum atque impulsu adulescentiae, 794–5). This combination of night-time festival, wine, and youthful ion recurs with a regularity that can only be called formulaic. In fact, there are only three comedies involving rape where any of these elements seem to be missing: the Samia, the Hecyra, and the Eunuch. In the Samia, Moskhion has impregnated the girl next door, Plangon, during a celebration of the Adonia, and his modesty prevents him from giving details about what happened. Perhaps we are meant to assume, for her chastity’s sake, that it was rape and not seduction, but his narration is exceptional in not making that explicit. Nowhere in the text we possess does he say that he was drunk, but the play is badly fragmentary and he may have made this excuse elsewhere. Furthermore, it may go without saying that since the Adonia was being celebrated in his own house, Moskhion will have been drinking too. The theme of Moskhion’s drunkenness and illicit sex does come up later in the play. Plangon gives birth, and the baby is given into the safekeeping of Khrysis, the concubine of Moskhion’s father. When the father

37

Lape 2001, 94.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

44

MIMETIC CONTAGION

returns home and finds Khrysis with a new baby, he does the maths and realizes it cannot be his, and he assumes that she got Moskhion drunk and seduced him (339–42): παρέλαβον αὐτόν που μεθύοντα δηλαδή, οὐκ ὄντ᾽ ἐν ἑαυτοῦ· πολλὰ δ᾽ ἐξεργάζεται ἀνόητ᾽ ἄκρατος καὶ νεότης, ὅταν λάβῃ τὸν συνεπιβουλεύσαντα τούτοις πλησίον. Well, she got hold of him when he was drunk, obviously, Not in control of himself. Unmixed wine Does lots of stupid things and also youth, When it comes into with someone scheming.

In the Hecyra the situation is reversed: there is no explicit mention of a holiday context for the rape, but the perpetrator is certainly drunk. Pamphilus raped an unknown woman in the street at night and in the struggle tore a ring from her hand. He went straight to his girlfriend Bacchis, confessed the crime, and made her a present of the ring. The victim will turn out to be none other than his estranged wife Philumena, and the ring will prove her faithfulness to her husband and save the day. Perhaps we are meant to understand that there was a festival, for there is no other obvious reason for Philumena to be on the street alone at night, but it is unusual that this is not made more explicit. In any case, Pamphilus was drunk (vini plenum, 823). So the Samia and the Hecyra do not spell out that the rape was attended by both religious festival and alcohol, but they certainly draw attention to one of the two and make it seem probable that the other factor was present as well. Aside from these partial exceptions, every other comedy involving rape makes fully explicit both the festival context of the crime and the drunkenness of the perpetrator—except for the Eunuch, of course, which has neither of these features. The twentieth-century reaction to Chaerea cannot be understood without reference to the apparent uniqueness of his crime. Instead of being neatly sequestered in the past, the rape is committed backstage in the middle of Act 3. This fact alone makes it seem less like a semiotic convenience of plot construction and more like a real part of the play. Indeed, as this book will show, that perception is well grounded—the painting and Chaerea’s reaction to it are far from incidental to the whole of the Eunuch. But for the judgment of Chaerea as a cool-headed, calculating rapist nothing has been more damning than his sobriety.38 Adele Scafuro and Susan Lape have argued (with respect to other rapists) that the Athenian legal discourse surrounding hubris made the combination of youthful ion and inebriation

38 Anderson 1984, 131: ‘utterly sober’; Smith 1994, 23: ‘completely sober, not drunk’; Rosivach 1998, 46: ‘not drunk but cold sober’.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

THE ROLE OF THE PAINTING

45

exculpatory, but this defence will not apply to Chaerea.39 However, if the rape is approached not as a real crime requiring evaluation of mens rea, whether by ancient standards or our own, but as an eccentric instantiation of a literary topos, the shift in our perspective has important consequences. Chaerea’s exuberant youthfulness and ion were immediately clear from his first entrance, and Parmeno made the comparison to Phaedria explicit: ‘You’ll say that other one was just fun and games, compared to what this one’s madness will do’ (ludum iocumque dices fuisse illum alterum, | praeut huiu’ rabies quae dabit, 300–1). Chaerea’s energy and his lust for Pamphila fit him perfectly for the first part of the rape formula, but when the crime occurs, the clear absence of both festival context and inebriation works as a kind of dramaturgical anacoluthon. The stylization of the formula is so iron-clad and this situation so unique that the audience listens to Chaerea’s narration in rapt expectation that it will disclose the missing elements. It is just here, I would argue, that Donatus’ reading of the painting’s role is most perspicacious, for the painting is not only the catalyst of the crime, but the site of the rape’s partial generic normalization. The missing element of religious context is, in fact, supplied by Chaerea’s unexpected encounter with the painting, which functions as an epiphany of Jupiter, henceforth to be invoked as tutelary god or dramatic genius (see p. 4). Drunkenness may be understood metaphorically elsewhere in Latin literature, for good fortune or love or other intangibles (Hor. C. 1.37.12; Cat. 45.11; etc.), but here the intoxicating agent is quite material. The sudden exposure to the painting sends Chaerea reeling from his senses, and his exultant Rausch leads straight to the rape. The ethical work accomplished by the night-time festival motif is twofold: it protects the victim’s reputation by giving her an excuse for being out and it helps for the young man’s loss of control. Similarly drunkenness has two functions: facilitating the subsequent lack of recognition and, again, explaining the young man’s loss of control. The Eunuch, however, does not need to defend Pamphila’s chastity, since she is raped at home, neither does it need to make room for future misrecognitions. But taking out these two inducements, the formulaic pattern offers a clarification still quite necessary in the Eunuch, namely that the perpetrator’s state of mind was jumbled to the edge of unability by something physically real and external. The painting fits into this niche exactly, accomplishing neither more nor less dramaturgical work than the situation requires. Unlike their modern counterparts, as I will argue in Chapters 2, 3, and 4, an ancient audience would have come to the theatre already alive to art’s potential for direct agency in inspiring imitation.

See Scafuro 1997, 246; Lape 2001, 94 and im. These scholars’ attempts to situate the mitigating factors of Menandrian rape in an Athenian legal context are very convincing and helpful, but perhaps they are less relevant for Terence as a Roman poet. In any case, the Eunuch is only cited as an exception. 39

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

46

MIMETIC CONTAGION

So Donatus’ ready assignment of efficacy to the painting seems prima facie much more likely for Terence’s audience than it has been for most modern readers, but the role of the painting becomes even clearer when we see how neatly it slips into the billet left empty by the absence of ‘religious event’ and ‘alcohol’. New Comedy’s tightly organized Erwartungshorizont for rape shunts the painting into a role that should be significantly exculpatory. As helpful as Donatus has been for this recovery of an ancient Roman response to the painting, there is another fourth-century reading of the same scene, where the confirmation of the painting’s role may be less explicit, but is essential to the way the age is deployed. In chapter 13 of the first book of the Confessions, Augustine repents of his foolish childhood love of false tales from Homer and Virgil, and in chapter 17 he laments precious time lost in declamatory exercises translating the wrath of Juno into prose. For a modern reader Augustine’s regrets may seem ill-placed, but between these diatribes on educational frivolity, we encounter a judgment to which our response is likely to be more sympathetic. In chapter 16, he apostrophizes the ‘hellish river of human custom’, in which the sons of Eve read about Jupiter’s dalliances, but Augustine es lightly over the already time-worn censures of the Homeric gods and their indiscretions and chooses a far more offensive text for his obloquy. The river of text calls out to the boys and their tuition-paying parents, advertising itself as the source of the rhetorical art, not only on a synthetic level (hinc adquiritur eloquentia), but on a lexical level (hinc verba discuntur). Perhaps ironically, Augustine answers the river’s ment with an erotema, a ‘rhetorical question’: ita vero non cognosceremus verba haec, ‘imbrem aureum’ et ‘gremium’ et ‘fucum’ et ‘templa caeli’ et alia verba quae in eo loco scripta sunt, nisi Terentius induceret nequam adulescentem proponentem sibi Iovem ad exemplum stupri, dum spectat ‘tabulam quandam pictam’ in pariete ‘ubi inerat pictura haec, Iovem quo pacto Danaë misisse aiunt in gremium quondam imbrem aureum, fucum factum mulieri?’ et vide quemodum se concitat ad libidinem quasi caelesti magisterio: ‘at quem deum!’ inquit, ‘qui templa caeli summo sonitu concutit. ego homuncio id non facerem? ego vero illud feci ac libens.’ non omnino per hanc turpitudinem verba ista commodius discuntur, sed per haec verba turpitudo ista confidentius perpetratur. non accuso verba quasi vasa electa atque pretiosa, sed vinum erroris quod in eis nobis propinabatur ab ebriis doctoribus, et nisi biberemus caedebamur, nec appellare ad aliquem iudicem sobrium licebat. et tamen ego, deus meus, in cuius conspectu iam secura est recordatio mea, libenter haec didici, et eis delectabar miser, et ob hoc bonae spei puer appellabar. Would we really not learn the words ‘golden shower’ and ‘lap’ and ‘trick’ and ‘quadrants of the sky’ and all the other words without Terence portraying a worthless youth who claims Jupiter as his model of vice, when he’s looking at a ‘certain painted ’ on a wall ‘where there was this picture, how Jove they say sent into the lap of Danaë a certain shower of gold to play a trick on a woman?’ And look how he whips himself up to concupiscence as if by a heavenly teacher: ‘But what a god!’ he says, ‘Who shakes the quadrants of the sky with his

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

THE ROLE OF THE PAINTING

47

lofty clap. And I, mere mortal, shouldn’t do it? You bet I did it, and I loved it! Those words are learned no more readily through this depravity, but through these words that depravity is committed more assuredly. I’m not blaming the words, which are like select, precious vessels, but the wine of error dispensed to us in them from drunken masters, and if we didn’t drink up, we were beaten, and there was no summoning a sober judge. But I, o Lord, in whose sight my memory is now sound, I learned these things gladly, and wretch that I was I took delight in them, and for this I was called a boy of good hope.

The reader coming to this age for the first time will probably be somewhat disoriented, as she encounters a dissociated list of words from Chaerea’s speech, without any preparatory indication of provenance. This is, of course, the point. The reader understands the words without knowing their literary context and thus enacts a proof of Augustine’s argument that the depravity of the scene has no pedagogical necessity. Indeed, by themselves the words are ‘precious vessels’, but because of their opprobrious deployment in this text they become the decanters of the ‘wine of error’. Augustine understands however that the words are not just a false pretext for our seduction, nor are they simply a medium of delivery for our destruction. They are, in their compromised state, a large part of the problem: per haec verba turpitudo ista confidentius perpetratur. Invoked merely as an example, out of what was obviously a very broad literary and rhetorical education, this scene from the Eunuch is taken as a kind of moral synecdoche for classical pagan literature. In our curriculum Terence has been comparatively marginalized, so Augustine’s choice may seem tendentiously quirky, like judging all English literature indecent on the basis of a scene from Wycherley. But we should that Terence was one of the most popular school authors from antiquity down to the nineteenth century, so Augustine’s representation of the Eunuch as foundational may well be true to his own educational experience and that of countless others. But the choice of this scene from this play is also clearly motivated by Chaerea’s combination of bad theology with outrageous behaviour. Indeed, there are probably few places in the classical canon that so neatly jumble the themes of the Confessions: human frailty, sex, divinity, the plunge into error. More suggestive still, Chaerea is the impressionable young man led astray by a work of art, just as the young Augustine is led astray by the dramatic work of art in which Chaerea appears. Augustine could have focused on the crime itself or even on Chaerea’s invocation of the myth of Jupiter and Danaë without the emphasis on the painting, but then his cameo of the hazards of pagan culture would have lost its inlaid mise en abîme, its programmatic inculpation of deleterious art. Corollary to this analogy is the partial displacement of blame from Chaerea’s character to the painting. Terence portrays a nequam adulescentem, but it is not clear whether Chaerea’s nequitia is ethically fundamental or the result of the

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

48

MIMETIC CONTAGION

painting’s bad influence.40 If this were a free-standing vignette we might unreservedly opt for the former, but considering its imbeddedness in the Confessions, specifically the Eunuch’s role in Augustine’s downward slide to perdition, the second, more sympathetic reading of Chaerea’s character seems more appropriate. When the same scene comes up in the City of God the language is very similar and the focus is even more explicitly on the play’s ruinous effect on the character of the young who consume it and are drawn into imitation of it. Of the crimes (flagitia) of the gods depicted on the comic stage Augustine laments: utinam solo risu, ac non etiam imitatione digna viderentur (‘Would that they had seemed worthy only of laughter, and not of imitation’, DCD 2.9). A bit later he asks: dignum fuit, ut Terentius vester flagitio Iovis optimi maximi adulescentium nequitiam concitaret? (‘Was it right for your Terence to stir up the mischief of young men by the crime of Jupiter Most High?’ 2.12). What was subtly suggested in the Confessions—that Terence’s introduction of a nequam adulescentem was corrupting for the young Augustine—is here made plain by Terence’s incitement of the (plural) adulescentium nequitiam.41 Not that we should imagine Chaerea before he saw the painting as morally spotless, any more than Augustine himself before he read the Eunuch. The point is rather that for a soul already enfeebled by sin, a wellplaced work of art can provide a strong or even irresistible incitement to imitation.

40

In the phrase induceret nequam adulescentem the adjective is probably to be understood as attributive, but inducere, in the sense of ‘portray as’, can also take a predicative compliment (OLD inducere 3.b). It is also tempting to read this nequam within the tradition of heroic ‘badness’ typical of the loveable scapegraces of Roman comedy. On this theme, see Anderson 1993, 88–106. 41 The danger of moral infection posed by the theatre is not, of course, limited to the Eunuch. The history of this pathogen goes all the way back to its institutional origins at Rome. In the following book’s catalogue of the failure of the gods to defend the city in her hour of need, Augustine mentions the plague that bore off M. Furius Camillus in 365 BCE as the occasion of the introduction of a far worse plague: in hac pestilentia scaenicos ludos aliam novam pestem non corporibus Romanorum, sed, quod est multo perniciosius, moribus intulerunt (‘In the midst of this pestilence they brought in theatrical shows, one more novel disease not for the bodies of the Romans, but, far more perilous, for their character’, DCD 3.17).

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

TWO

Quickening Images Mimetic Contagion in Cultic and Erotic Art

ὦ Μένανδρε καὶ βίε— πότερος ἄρ᾽ ὑμῶν πότερον ἀπεμιμήσατο; Aristophanes of Byzantium

The previous chapter has argued, from evidence internal to the play and from the recorded responses of its ancient readers, that modern critics have tended to underestimate the effectual power of the painting in the Eunuch. This discrepancy between ancient and modern distributions of fault may be partially explained by the differences in our conception of rape, but there is another difference that separates us even more sharply from Terence’s audience and has surely done much to occlude the painting’s role. Modern readers, to the extent that we are accustomed to thinking of artefacts as inert objects perceived by rational agents, are likely to be less alive to the possibility that the stream of agency may be reversed, that life may be induced to imitate art. My goal now is to explain what Chapter 1 merely noticed. By exploring some of the cultural background and discursive practices that feed into the conceit of artefacts shaping the world in their own image, I hope to for the influence of the painting in the Eunuch, locating its apparent effects first within a broad Greco-Roman context, then more specifically in mid-second-century BCE Rome, and finally within the genre of Roman comedy. I designate as ‘mimetic contagion’ the effectual power of images that replicate themselves in their viewers, not only because they conflate Frazer’s Law of Sympathy and Law of or Jakobson’s metaphor and metonymy, but because they implicitly suggest their own transmissibility. Whenever an image, notionally patterned on some real subject, inverts the presumed ontological order of similarity and becomes, itself, a subject for imitation, it proves it has the means of its own further propagation through its new ‘host’. More particularly, when the image is a representational artefact embedded in another work of art, whether plastic or theatrical, and it projects itself on to some figure in the larger work of art, the door is open for a vertiginous expansion of the image and of mimesis itself, as the mise en abîme becomes an emblem for the porous boundary between art and

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

50

MIMETIC CONTAGION

life. In our survey of the various forms of mimetic contagion prominent in Hellenistic culture, we shall meet several artefacts that show just this expansion and suggest a reading of the painting in the Eunuch as a figure of adhesive roleplaying within the play. It is Frazer’s fundamental confidence in the gulf between his own cognitive processes and ‘primitive thinking’ that authorizes his projection of the category ‘sympathetic magic’ on to the superstitious men of distant times and cultures and even, with some regret, the rural poor of his native Scotland.1 Subsequent anthropological scholarship has, of course, demolished the notion that there could ever have been a culture in which people routinely live in participatory psychic sympathy with the world around them, or in which adults are incapable of understanding causality except by matching similia similibus.2 But if Frazer’s ‘cunning and malignant savages’ (55) have been shown to be more like us than he assumed, another branch of twentieth-century critical theory was concerned to demonstrate that we are also more like them. Benjamin’s essay ‘Über das mimetische Vermögen’ (1933), advanced the view that even, or perhaps especially, our higher cognitive functions are shot through with the same mimetic impulse everywhere evident not only in primitive man, but in all of nature.3 Similarly for Adorno, the project of the Enlightenment is decried as a domination of the Other by objective ‘identification’. The antidote, he believes, is to be sought in a cultural poetics of self-modulation, of ‘identification with’ the Other.4 Without further demonstration, the vatic insights of Benjamin and Adorno might be dismissed as a blind calumny of the Enlightenment or at least an underestimation of the intellectual suppleness of modernity, but subsequent arthistorical investigation has provided striking confirmation of the basic validity of their views. For instance, David Freedberg has given an impressive catalogue of 1 ‘From the azure sky, the stately fanes, and the solemn ritual of ancient Egypt we have to travel far in space and time to the misty mountains of the humble cottages of the Scottish Highlands of to-day; but at our journey’s end we shall find our ignorant countrymen seeking to attain the same end by the same means and, unhappily, with the same malignity as the Egyptian of old’ (Frazer 1935, 68). 2 If this description of the primitive mentality sounds closer to Lévy-Bruhl, that is because I take LévyBruhl’s formulation to be an apt expression of an assumption implicit in Frazer. An adequate refutation of such prejudices may be found already in Durkheim (1912, 342), and by mid-century, as Lienhardt observes (1954, 95), no working ethnographers still believed in radically different modes of thought employed exclusively by primitive peoples. If anyone was still left, a decade later, labouring under Frazer’s influence on this point, surely he will have been set right by Lévi-Strauss 1962. 3 ‘Vielleicht besitzt [der Mensch] keine höhere Funktion, die nicht entscheidend durch mimetisches Vermögen mitbedingt ist’ (Benjamin 1977, 210). 4 ‘In den Kunstwerken ist der Geist nicht länger der alte Feind der Natur. Er sänftigt sich zum Versöhnenden. Nicht bedeutet sie nach klassizistischem Rezept Versöhnung: diese ist ihre eigene Verhaltensweise, die des Nichtidentischen innewird. Der Geist identifiziert es nicht: er identifiziert sich damit. Dadurch daß Kunst ihrer eigenen Identität mit sich folgt, macht sie dem Nichtidentischen sich gleich: das ist die gegenwärtige Stufe ihres mimetischen Wesens’ (Adorno 1997, 202).

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN CULTIC AND EROTIC ART

51

evidence that modern, Western, institutional ‘high art’ is indeed the exception to the rule that images are everywhere thought to have the power to shape reality.5 Even in our culture, the range of responses to art, from sexual arousal to rage at the prototype, gives the lie to aestheticist assumptions about images as ive objects. Most recently, Alfred Gell has articulated an anthropological theory of art grounded in a general taxonomy of agency associated with images and artefacts.6 Gell effectively invents a language for describing the nested complexities of social agency mediated by objects. Rather than assuming that the image must be ive and the viewer active or vice versa, Gell shows how artist, index, prototype, and recipient negotiate and renegotiate a tangled web of active and ive roles, partly prescribed by the broader culture and partly ad-libbed according to the particularities of each interaction. If the power of images, more specifically the capacity of certain artefacts to summon the viewer or participant into a mimetic role, has been recuperated to some extent in the twentieth century, the ancient Mediterranean world never needed to be reminded of this phenomenon for the simple reason that it had never forgotten such modes of thinking about images or tried to pretend them away. Among the Greek and Roman cultural practices of plastic representation that were organized around this principle of life imitating art, one fairly clear example is the ‘voodoo doll’.7 Because the doll is somehow like the victim, if only notionally and schematically, the victim is then supposed to suffer, if only metaphorically, the tortures and bindings inflicted on the doll. A relationship of similarity between figurine and subject is ritually reversed, and mimesis becomes a two-way street. Such practices were apparently ubiquitous in antiquity, but scholars are only now coming to appreciate how often allusions to these techniques lurk beneath the literary surface in Latin poetry.8 This process was not necessarily dependent on vision, since the dolls were often secreted away, but there were also practices in which the victim’s cognizance and vision of the image seem to have been important, such as representations of emphatically healthy people used by the infirm to effect a homoeopathic cure.9 There are also instances of the same formulation of Freedberg 1989. Gell 1998. Gell’s work has been especially influential in the study of non-Western material, but recently classical art historians have begun to explore his approach and adapt it to their subject; see now, Stewart 2007 and Osborne 2007. 7 See Faraone 1992 and now Collins 2003, 37–44. 8 Laurel Fulkerson (2002) argues convincingly that the erotic manikin of Protesilaus in Heroides 13 functions as a voodoo doll and that its burning coincides with his death. For sympathetic magic in Propertius, see O’Neill 1998. 9 The ritual deployment of votive body parts is well attested in the cult of Asclepius and beyond. For Minoan evidence, see Dickinson 1994, 271; for the classical period, see Forsén 1996 im; for fourthcentury Etruscan examples, see Turfa 2006, 96. The continuity of this practice in both Eastern and Western Christianity is well known, but other devotional Überlebsel of this ancient attitude to images are less likely to 5 6

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

52

MIMETIC CONTAGION

art and subject deployed on a public and even civic scale, where those implicated in imitation were clearly intended to see their models.10 In most defixiones, whether public or private, the emphasis is on the ions of the image, and the viewer or prototype is supposed to suffer changes somehow analogous to the changes visible in the index, but alongside such practices we may situate the very frequent ancient topos of an image that fascinates viewers into imitation of its inanimate fixity by freezing them in wonder, fear, or desire.11 To take only the clearest, most widely attested, and best studied of these phenomena, we may look to the mythological and representational traditions surrounding Medusa.12 The apotropaic gorgoneia affixed ubiquitously in antiquity, on temple façades, shields, ships, workshops, and common household ware, seem intended, in principle, to confront the victim head-on and induce panic and paralysis. They are truly ‘terror masks’, in both active and ive senses, in that their bugged-out eyes and distended rictus give them a terrified look but also, presumably, cause the same countenance in the hostile viewer.13 The victim who looks straight at that face, as it looks straight at him, finds himself frozen it its thrall, mirroring the image’s lifelessness as he mirrors its expression.14 But the monster herself was

be recognized, especially when the image is more obviously embedded in a narrative context. For an example of such an artefact, see Jacquelyn Tuerk’s essay (1999) on an early Byzantine amulet depicting the woman with the issue of blood from Mark, which would have been used by a woman suffering from the same ailment. By contemplating the two sides of the amulet, one showing the woman prostrating at Christ’s feet and the other showing her standing orant and evidently whole, the viewer would have attempted to analogize herself to the woman in the gospel and make the age from sickness to health after the narrative model prescribed by the sequential sides of the amulet. 10 The clearest example may be from the founding of Libyan Cyrene in the seventh century BCE. The colonists and those staying behind in Thera bound themselves to the conditions of their agreement by burning wax effigies, accompanied with the conditional self-imprecation that they themselves would ‘melt away and dissolve like the images’ should they violate the of their charter. See Faraone 1993. 11 Steiner 2001 has an excellent survey of this range of phenomena; for the ‘terrifying look’ theme, see especially 172–81, for hypnotized agalmatophilia, see especially 198–204, but the topic of thauma, as both object and response, is richly explored everywhere in her work. 12 For recent treatments of this aspect of the Medusa tradition, see Steiner 2001, 170, 173, 296–7; Faraone 1992, 38; and most thoroughly, Vernant 1991, 111–50. 13 By indicating a potential distinction between different types of viewers, welcome and hostile, I mean to offer a partial answer to Rainer Mack’s observation (2002) that the Gorgon’s very ubiquity is an indication that she was not necessarily so dangerous to look upon. He argues that the full mythological context for the disembodied Gorgon’s head implies that her representation in this form is emblematic not of her terror but of her defeat at the hands of Perseus. I would insist, however, that even after decapitation and death, Medusa’s head is terrible and proves lethal for some. Mack points to an interesting problem, but a simpler solution would be to say that, for any given social interaction mediated by such an image, there may have been some potential viewers who were indeed supposed to be warded off or terrified. 14 Already in the Iliad, the emphasis is on the icy fear produced in the enemy who views Medusa’s image. In the aegis (8.740–2), along with ‘chilling Rout’ (κρυόεσσα Ἰωκή), appears the ‘dreadful and terrible’ (δεινή τε σμερδνή τε) Gorgon’s head, and crowning Agamemnon’s shield (11.36–7), along with Fear and Dread (Δεῖμός τε Φόβος τε), the Gorgon appears ‘horrid of aspect’ (βλοσυρῶπις) and ‘fearsomely glaring’ (δεινὸν

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN CULTIC AND EROTIC ART

53

already a kind of living statue, with hands of bronze and wings of gold (Apollod. 2.4.3), so in this case the prototype is already a work of art, spawning copies both in her replicas and in the terrified faces of her victims. ‘When you stare at Gorgo, she turns you into a mirror where, by transforming you into stone, she gazes at her own terrible face and recognizes herself in the double, the phantom you become the moment you meet her eye.’15 If Medusa herself is already halfway to being an artefact, perhaps that explains why she works so well as an icon of this kind of power in images, but there are similar, if less insistent, stories associated with Dionysus and with the other gods and their images.16 Nor was this type of art confined to the Greeks. The Romans seem to have had a close equivalent: the apotropaic masks called oscilla.17 One might presume that the oscilla were a Roman adaptation of Greek terror masks, since any Roman with even a ing acquaintance with Greek culture would have noticed the gorgoneia, not so much because of their frequency on buildings and public art, which do not usually travel, as on inexpensive ware, which does. However, like voodoo and other forms of homoeopathic magic, apotropaic techniques were evidently constant across the historical and cultural continuum of the ancient Mediterranean world, and the essential similarity between gorgoneia and oscilla is more likely due to spontaneous polygenesis. In any case, the mesmerizing painting of Jupiter in Terence’s Eunuch would have been interpreted by an audience already operating under the assumption that images of the gods may be uncannily arresting. The widespread prevalence of such discourses in antiquity should never be lost from view as we attempt to understand how the painting is taken to affect Chaerea, but ultimately, of course, the painting in the Eunuch is not a volt sorcery talisman or a terror mask. Though the boundary between magic and art was porous in antiquity, or rather no boundary at all, perhaps there are other practices with images that can be considered more apposite contextual comparanda. Various sources testify to the Roman taste for architectural decoration on the principle of matching image to space.18 Vitruvius (7.5.1–2) describes the usual practice of δερκομένη). These descriptions based on implicit response are typical of the Medusa tradition, and the viewer’s petrifaction is usually, as here, metaphorical, but in a particularly literal version, relayed by Pausanias (9.34.2), a Boeotian priestess of Athena went into the sanctuary at night and, upon seeing Medusa’s head on the goddess’ tunic, became stone. Vernant 1991, 138. For Dionysus, see Frontisi-Ducroux 1991, 189–92; for a generalization to facial images of all divinities, see Steiner 2001, 174. For the eye cups as a form of ‘involved spectatorship’ in which the participant is made to look through the eyes of a satyr, see Hedreen 2007. 17 The standard work on the oscilla, as on all early Roman mask culture, is still Meuli 1955; see especially 214–18. For a more recent treatment, with comparative references to the Hellenistic crossovers in Magna Graecia, see Faraone 1992, 38. 18 For a recent survey of some aspects of this issue, see Perry 2005, 50–65. 15 16

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

54

MIMETIC CONTAGION

decorating interiors with painted construction materials and decorating walks with landscapes, and Pliny relates that the matching of painted representation to intended use could be more exact, for example, wrestling and anointing rooms adorned with portraits of athletes (HN 35.5). Of all such correspondences between art and space, surely none is so well attested in antiquity as the use of erotic images in cubicula, both in homes and in brothels.19 This phenomenon spanned all strata of Roman society, from the humblest Pompeian lupanaria to Tiberius’ infamous residence on Capri, where he kept in his bedroom (in cubiculo dedicavit, Suet. Tib. 44) a spectacularly valuable portrait, painted by none other than the old master Parrhasius, of Atalanta fellating Meleager. Such images might, of course, have many intended functions, from simple amusement to advertising the owner’s refinement or opulence, but their most obvious purpose, especially given their location, is to arouse the viewer and encourage the enactment of a scene like the one portrayed. If we are inclined to think of such an effect on the viewer as merely psychological and wholly separate from homoeopathic magic, perhaps this is, in part, because we do not generally ascribe a divine aetiology to sexual arousal. In antiquity, the image does not operate on the viewer with the meagre instruments of rational persuasion, and his (or her) involuntary reaction is met by a cultural tradition dictating imitation as an appropriate form of response. Alongside the raw power of erotic images to arouse viewers and thus indirectly replicate themselves, there is a more intentional form of imitation sometimes associated with sexually explicit art. The encyclopaedic fancy to catalogue and describe the various sexual positions was well known to the Greeks and Romans, an impulse that had found its fullest expression in the Hellenistic erotic handbooks traditionally ascribed to the courtesans Astyanassa, Philaenis, and Elephantis.20 These books may not have been illustrated guides to the art of sex, but corresponding pictures could always be supplied in an appropriate context.21 Ovid’s Ars amatoria is obviously much more than a poetic version of this tradition, since its didactics of love range well beyond a mere how-to of sexual positions, but the poem does end on just this note, surveying the various modi and figurae, explaining to the female reader why she might opt for each of the positions, and describing how to do each one (3.771–88). Here, at the end of the Ars, the praeceptor amoris is putting himself in the role of the older, experienced woman, not only because of the straightforward parallelism with the didactic tradition of The bibliography on figurae veneris and the significance of their placement is now quite vast. Myerowitz 1992, Jacobelli 1995, Clarke 1998, and McGinn 2002 have proved most germane to this project. 20 Parker 1992 has a clear survey of the evidence of these sex manuals, along with a perceptive attempt to situate them within larger cultural preoccupations. 21 Though Brendel (1970, 63–4) does believe there would have been both illustrations and a numbering system for the positions. See also Clarke 1998, 247. 19

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN CULTIC AND EROTIC ART

55

books by female experts on sexual positions, but because of his own claim towards the end of the previous book that old women make good lovers due to their incomparable savoir-faire.22 ‘However you like it, they make love in a thousand positions; no painting has found more ways’ (utque velis, Venerem iungunt per mille figuras: | invenit plures nulla tabella modos, 2.679–80). Clearly the tabella he has in mind is not a single-frame erotic painting, such as we generally have from Pompeii, but rather a visible index of positions. This kind of painting comes up again in Tristia 2, Ovid’s extended apologia for Ars amatoria. His strategy is to defend his erotic vademecum by pointing to all the other forms of didactic poetry on light themes (swimming, pottery, etc.) or on potentially pernicious themes (gambling), none of which are branded with censure like the Ars. Even Ennius and Lucretius allude to the adulteries of the gods, and from Tibullus one can be instructed in the high arts of erotic deception. Surely nothing Ovid has written could be as bad as the mimes. Chapter 5 will explore the full significance of mime for the rape scene in the Eunuch, but I will quote Tristia 2.497–524 here for its collocation of theatrical performance, which may or may not be supposed morally contagious, and didactic erotic painting: quid, si scripsissem mimos obscena iocantes, qui semper vetiti crimen amoris habent: in quibus assidue cultus procedit adulter, verbaque dat stulto callida nupta viro? nubilis hos virgo matronaque virque puerque spectat, et ex magna parte senatus adest. nec satis incestis temerari vocibus aures, adsuescunt oculi multa pudenda pati; cumque fefellit amans aliqua novitate maritum plauditur et magno palma favore datur; quoque minus prodest, scaena est lucrosa poetae, tantaque non parvo crimina praetor emit. inspice ludorum sumptus, Auguste, tuorum: empta tibi magno talia multa leges. haec tu spectasti spectandaque saepe dedisti (maiestas adeo comis ubique tua est) luminibusque tuis, totus quibus utitur orbis,

500

510

Ovid’s persona is often either explicitly female or implicitly feminized (see Sharrock 2002). There have been a number of recent studies on this aspect of the Heroides (see, for example, Spentzou 2003 and Fulkerson 2005), but to my knowledge no one has examined this gender slippage in the Ars amatoria. The praeceptor amoris has based his authority in experience (usus opus movet hoc, Ars 1.29), but book 3 will turn to the instruction of women, and it is the anus at the end of book 2 whose veteran status makes her relevant to this task. I am not, of course, arguing that the praeceptor is really a praeceptrix, but rather that the gender of erotic authority is destabilized by this move. 22

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

56

MIMETIC CONTAGION

scaenica vidisti lentus adulteria. scribere si fas est imitantes turpia mimos, materiae minor est debita poena meae. an genus hoc scripti faciunt sua pulpita tutum, quodque licet, mimis scaena licere dedit? et mea sunt populo saltata poemata saepe, saepe oculos etiam detinuere tuos. scilicet in domibus nostris ut prisca virorum artificis fulgent corpora picta manu, sic quae concubitus varios Venerisque figuras exprimat, est aliquo parva tabella loco.

520

What if I had written dirty jesting mimes, Which always contain the sin of forbidden love, In which a slick seducer is constantly strutting out And a clever bride tricking her stupid husband? Marriageable maiden, wife, husband, child—all watch these mimes, And a big part of the senate is there too. It’s not enough their ears are polluted with impure words; Their eyes get used to suffering many shameful sights, And when the lover has tricked the husband by some new contrivance, They cheer and bestow the palm with great praise. The less beneficial the stage is, the more profitable it is for the poet, And the praetor buys such outrages dear. Examine the books for your own games, Augustus, You’ll find many such things bought at great price in your name. You’ve watched them yourself and presented them often for others, So great is your generosity in all things, And with your eyes, twin lights to all the world, You’ve calmly watched theatrical adulteries. If it is permitted to write mimes that imitate shameful things, My stuff deserves a lighter penalty. Is it the stage that makes this type of writing safe? Does the theatre bestow all license to the mimes? My poems too have often been choreographed for the people, Often they have captured even your eyes. Of course in our houses, just as the venerable bodies of heroes Shine, painted by the hand of an artist, There is also a small painting in a certain place Representing diverse couplings and positions of love.

Ovid’s overt point is that mime is not normally considered liable for censure as indecent, and so a fortiori neither should the Ars amatoria. But throughout Tristia 2, Ovid’s defence of Ars amatoria is consistently undermined by playful misreadings that implicate himself and Augustus equally as corruptors of

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN CULTIC AND EROTIC ART

57

morals.23 The more mischievous reading of this age is suggested by Ovid’s indignant description of theatrical indecency: mime probably is a corrupting influence on its audience, and Ovid’s poetry too should be handled with care, but it is all likely to be misread by Augustus, who is a careless reader and as corrupt himself as the poems he pretends to judge. Mime is treated as morally contagious, not just because of its obscenity (497), but because it is a theatrical performance, which therefore aurally (503) and visually (504) assaults its spectators, including the young and impressionable (501). However, if mime is deleterious then Augustus is to blame for having presented so much of it (511). The eyes of the people (oculi, 504) may have become inured to the indecency of mime, but that is because the whole world uses Augustus’ eyes, which have themselves grown complacent at theatrical immorality (luminibusque tuis, totus quibus utitur orbis, | scaenica vidisti lentus adulteria, 513–14). If only Augustus would tear his eyes away from such depravity and look into (inspice, 509) his s, he would be a better reader (leges, 510) of the realities of literary entertainment at Rome.24 But whatever damage may be done by the strictly theatrical aspect of mime is shared by Ovid’s poetry, which has often been presented with accompanying dance (519). Once again Augustus’ eyes (oculos, 520) are at issue, but this time their captive gaze is trained on Ovid’s poetry in performance. None of this works as a straightforward defence of Ars amatoria or flattery of Augustus. It sets out to make mime seem more dangerous, but by describing the Roman people and Augustus as corrupted by mime, it makes Augustus both perpetrator and victim of mimetic contagion. Furthermore, by pointing to the exact parallel between mime and his own poetry, in thematic content, in performance context, and in its effect on the hapless emperor, Ovid makes Ars amatoria sound just as bad as the most pernicious mime. For the purposes of our study, however, the most interesting thing about this age is the abrupt shift from Augustus’ eyes held captive by a dramatic rendering of Ovid’s ittedly salacious poetry to a painting in a domestic setting presenting a panoply of sexual positions.25 Perhaps an array of erotic symplegmata collected together in one could be intended simply to arouse all the more, as if by visual polyptoton, but the emphasis on schematic variety makes this parva tabella sound more like a didactic illustration See Williams 1994, 201–9. Williams (1994, 179–93) argues that throughout Tristia 2 Ovid consistently hints that Augustus is too busy as emperor to have the otium required to be an adequately careful reader. 25 The text of line 521 is unclear; the painting is either in our (nostris) houses or in your (vestris) house. Myerowitz (1992, 155–6) has a full survey of the muddle of manuscript evidence and editorial indecision on this question. I have followed Owen 1924 and the majority of recent editions (pace Luck 1967 and Myerowitz) in reading nostris, but I consider neither reading implicitly better than the other. If we read nostris then Ovid is situating Augustus’ and the Romans’ viewing of sexually explicit theatre within typical Roman visual practice. If we read vestris then he is continuing the motif of Augustus’ docile gaze specifically into the imperial bedroom. 23 24

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

58

MIMETIC CONTAGION

of the various possible postures. The Ars amatoria is here defended by analogy to (or affirmed to be as inspirational as) a type of painting that it had mentioned itself at the end of book 2 and ekphrastically rendered at the end of book 3. There is a distinctly Ovidian irony in the use of the particle scilicet as the first word of line 521 to introduce the painting. In Metamorphoses 15 Ovid its that one might think all Caesar’s conquests and other achievements greater than simply fathering Augustus, but that assessment—so runs the logic of his panegyric—would obviously be wrong. The barb is the devilishly equivocal word scilicet, which begins line 752 and introduces Caesar’s res gestae. As Steven Hinds has argued for this age, Ovid mobilizes the ambiguity of scilicet to connote ‘evident absurdity; or evident truth’ and to frustrate a narrowly pro- or antiAugustan reading.26 The word scilicet appears four times in Tristia 2, always in the same metrical sedes, the first foot of a hexameter line, and every time, arguably, with this same delicately ambiguous force.27 It is pointedly impossible to determine, then, whether the erotic painting is introduced as a ubiquitous example of a visual practice parallel to and just as harmless as theatre, or whether it is ironically submitted as a form of art obviously intended to inspire direct imitation, making it an emblem for the itted danger of theatre but also an illustration of Rome’s current depravity, so that singling out the Ars amatoria is clearly unjust. Let me emphasize that there is no reason to think that the domestic use of erotic images was ever a novelty at Rome, any more than the use of figurative art more generally, but as we shall see, the myth of a fall from early Roman artistic restraint or even aniconism is recurrent and has cultural and political force at various times in Roman history. Propertius gives a contemporary and more explicitly nostalgic response to the same issue (2.6.25–36): templa Pudicitiae quid opus statuisse puellis, si cuivis nuptae quidlibet esse licet? quae manus obscenas depinxit prima tabellas et posuit casta turpia visa domo, illa puellarum ingenuos corrupit ocellos nequitiaeque suae noluit esse rudis. ah gemat in tenebris, ista qui protulit arte orgia sub tacita condita laetitia! non istis olim variabant tecta figuris: tum paries nullo crimine pictus erat. sed nunc immeritum velavit aranea fanum et mala desertos occupat herba deos.

30

Hinds 1987, 25. The other three instances are in lines 17, 107, and 219. Williams (1994, 176–7) has already suggested this reading of the scilicet in line 107. 26 27

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN CULTIC AND EROTIC ART

59

What’s the point of founding a temple to Chastity for the girls, If anything goes for any bride? The hand that first painted wanton pictures And put shameful things on display in a decent home— That hand corrupted the innocent eyes of girls, Not wanting them to be ignorant of its own naughtiness. Oh, let him groan in darkness, who brought forth by his art Mysteries hidden beneath silent bliss! Once houses were not decorated with those filthy illustrations, Back then no wall was painted with disgrace. But now spider webs enshroud the neglected temples And vile weeds overrun our forgotten gods.

Once again the didactic significance of erotic painting is clear.28 Its effect on society has been apocalyptic, revealing to tender eyes things hidden since the foundation of the world, but veiling from view (velavit, 25) things that should have been open, and meriting wretched obscurity (tenebris, 31) for its progenitor. Amid this play of visibility and invisibility, the joke is that none of this is strictly relevant to Cynthia, who is not the poet’s nupta and whose house was never a casta domus. As the opening lines of the poem reveal, Cynthia has more gentleman callers than any of the great Hellenistic hetaerae, including Menander’s Thais, ‘in whom all Athens took its pleasure’ (in qua populus lusit Ericthonius, 4). The irony here is just the opposite of that in Tristia 2. Ovid invokes the erotodidactic painting overtly to excuse himself, Augustus, and the Romans, but there was a criticism latent in his deployment of the picture. Propertius, on the other hand, mentions the erotic painting ostensibly for a stern reflection on moral decline, but his tirade on how dirty pictures corrupt good girls by teaching them things they should not know is totally irrelevant to his affair with Cynthia. In fact, Propertius’ amata is less like a virginal bride than she is like Menander’s Thais, for not only is she not innocent, she is also not real.29 Menander’s Thais is described as thronged by a crowd (turba, 3), but as the next line reveals, this crowd was none other than the Athenian people at play (populus lusit, 4). Not a mob of suitors, then, but a city at the theatre. Propertius invokes what was, as we shall see, a traditional moralistic 28 Francis Cairns has argued that there may be an allusion to Gallus’ erotodidaxis in many places in the Monobiblos, see especially Cairns 2006, 116–18. For this age alone, Cairns detects a Gallan flavour both for the vocabulary and the metrical sedes of nequitiae (line 30) and laetitia (line 32), see Cairns 2006, 95 and 205 respectively. 29 It is tempting to conjecture that Propertius may mean Menander’s Thais specifically as she appears in Terence’s Eunuch. Barsby 1999b argues that the post-comic exclusus amator tradition, as it emerges in Catullus and elegy, is due directly to Phaedria’s dialogue with Parmeno in Act 1, a scene adapted by Horace (Serm. 2.3.259–71) and Persius (5.161–75) and quoted several times by Cicero and Quintilian. Propertius certainly sounds like Phaedria here, but more particularly he accuses Cynthia of inventing false relations (7), and he suspects that she has among her entourage a man in drag (14).

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

60

MIMETIC CONTAGION

topos about the novelty of certain types of images and about the danger of mimetic contagion, but he does so to make fun of this attitude towards art or to make fun of his persona as elegiac amator or both. In any case, the irony of his treatment does not diminish his value as evidence of the didactic use of erotic painting or of the currency of Roman concern about the imitative responsiveness of young viewers. Roman brothels certainly had sexually explicit paintings and there may be material evidence of brothels decorated on a similar principle to Ovid’s parva tabella, presenting a catalogue of positions to inspire unimaginative clients or to enable communication between people who may not have spoken the same language, but this evidence is unclear and the interpretation of these images as didactic is highly conjectural.30 A closer comparandum for Ovid’s thumbnail sketch catalogue may be found by returning to Tiberius’ pleasure dome on Capri. According to Suetonius, Tiberius ‘decorated bedrooms variously arrayed with pictures and statuettes of the most provocative paintings and figures and stocked them with the books of Elephantis, so that his sex partners would never be stuck without a model of whatever position he had ordered’ (cubicula plurifariam disposita tabellis ac sigillis lascivissimarum picturarum et figurarum adornavit librisque Elephantidis instruxit, ne cui in opera edenda exemplar impe[t]ratae schemae deesset, Suet. Tib. 43). Where Ovid’s parva tabella was implicitly in a bedroom and intended for direct imitation, Tiberius’ images are explicitly so, and the connection to didactic written erotica is also made plain. The relevance of this material for the Eunuch may not be immediately apparent. The painting in the Eunuch, after all, is not part of a serial catalogue of sexual positions. It is, however, a sex scene painted above a bed in a brothel. By its placement if not by its nature alone it is clearly intended to arouse, and once aroused, the bedmates in such a cubiculum may be presumed to enact their own love scene. Yet Chaerea’s response is not simply one of arousal and incidental

30 Perhaps the best contender for erotodidactic painting in a brothel-like context is room 7 of the Suburban Baths in Pompeii. Discovered in 1986 and excavated by Luciana Jacobelli, the Suburban Baths are unusual in not having separate men’s and women’s sections. Most remarkably, the apodyterium contains a numbered series of sixteen boxes with a different sexual position represented above each box. Jacobelli (1995, 70) connected these images with the so called spinitriae, small bronze tokens of completely indeterminate purpose, with a number from one to sixteen on one side and on the other side a different sexual position for each number. Clarke (1998, 247) has suggested that these taxonomic numbering systems point to a shared source in the Hellenistic sex manuals, but that the boxes in the Suburban Baths are just a place to store clothes and the spinitriae are just game pieces, so the system is functioning here not to teach, but only to amuse viewers. I am inclined to agree with Clarke’s assessment, but even if patrons of the Suburban Baths could not be expected to order sexual positions by number and enact them there in the apodyterium (contra Guzzo and Scarano Ussani 2000, 17–25), it is quite possible that venal sex was available somewhere in the baths, especially if, as Jacobelli (1995, 94–7) argues, the Suburban Baths were used by men and women simultaneously. For baths doubling as brothels, see McGinn 2002, 12.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN CULTIC AND EROTIC ART

61

recapitulation, but one of avowed imitation. There is no way to know how far back in Roman culture the didactic use of erotic art may have gone or how widespread it was at any time. Suetonius’ condemnation of Tiberius seems to be directed not to the practice itself but to the outrageous scale at which it was executed, and Ovid’s reference to the modestly erotic painting does not seem to be intended as an indictment at all, nor is there any suggestion that it is a novelty. In fact, Ovid sets the parva tabella among a highbrow entourage of other Hellenistic types, predominantly images of heroes and gods (Tr. 2.521–8), and Tiberius’ inclusion of the books of Elephantis situates his porn collection as the composite visual aid to the specialist knowledge of the appropriate Hellenistic expert. Terence’s audience may well have known the didactic use of erotica from their own experiences in Roman brothels, but even if they did not, if this later testimony is any indication, they probably recognized it as especially appropriate to the house of a Greek hetaera. Our evidence of how and where paintings of any kind were used in the second century BCE is frustratingly sparse, but at least for later periods the use of erotic paintings in Roman bedrooms was evidently widespread enough to enable sophisticated play with the conceit of nested artistic response, reminiscent of some of the dynamics we will be tracing out from the painting in Terence’s Eunuch. For example, on the south wall of room 43 in the House of the Centenary (IX.8.6) in Pompeii there is a painting of a couple in the mulier equitans position with the woman facing her lover, who is propped up slightly on his left elbow.31 On the wall behind the couple, above them and to their right, is a dark rectangular patch of underpainting—all that remains of what was clearly a pinax on the fictional back wall of the lovers’ room. Removal from its original site for hasty storage in the locked Gabinetto degli Oggetti Osceni probably did more damage than time or volcanoes to the delicately applied secco that would have been used for this tiny picture-within-the-picture, so there is no way to be sure what was represented there, but scholars have generally concluded, from analogous images in other media, that it would have been an erotic painting, probably not too different from the scene within which it appears.32 The couple’s attention seems occupied at the moment, but if they were to look out of their painting and across the room, they would see an image strikingly similar to their own, for on the north wall of the same room there is another painting of a couple, also mulier equitans with the man leaning on his left elbow, though now the woman is turned around with her back towards her lover (adversa).33 Again there is a dark rectangle on their back wall, this time to the 31 32 33

Marcadé 1965, 79. See Myerowitz 1992, 145–6 and, more forcefully, Clarke 1998, 166–7. Marcadé 1965, 126.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

62

MIMETIC CONTAGION

left of them. Scholars have, of course, noted the similarity of these paintings, and some have supposed that two such pictures in one room are enough to designate it a camera d’amore.34 But such assignments of use are conjectural; there is no way to know whether room 43 was especially intended for sex. If it was a room in which people might legitimately be expected to copulate, then the mise en abîme in the paintings was literally projected into reality. If, on the other hand, the room was not apt for sex, then the paintings would seem to tease the viewer by showing enactment as a fitting response to the presence of painted sex and then playfully doubling this presence, flanking the viewer from north and south. However the response of real viewers in room 43 is meant to be constellated, there is a significant emphasis on the painted lovers, especially the women, as potential viewers of art. We do not fully understand the cultural valence of the mulier equitans position in antiquity, and there has been some recent disagreement about the attitude to women it conveys.35 But whether the position implies the woman’s subjugation or her liberation, it certainly requires a greater level of activity on her part than almost any other position. The two erotic paintings in room 43 illustrate the two possibilities for the woman straddling her partner: she may either face him or she may face away from him. Either way, the sex consists of her doing something, so it may be significant that in both images the paintingwithin-the-painting is situated in front of her. To be sure, in neither picture is she portrayed studying the painting in her room, but the shift in placement of the mise en abîme is one of the few differences between the two paintings, creating, I would argue, an emphasis on the visibility of these tiny paintings for the women who are perhaps in a position of needing instruction, inspiration, or encouragement. There are several other examples from Roman antiquity of tabellae nested in larger erotic images, but the most striking parallel to the paintings from the House of the Centenary is on the back of a first-century CE mirror found on

Dexter 1975, 150. Clarke (1998, 161–4) is critical of Dexter’s label, since there are many other, nonerotic paintings surrounding these s and the layout of the rooms does not particularly suggest an exclusively erotic use. 35 Paul Veyne (1978, 53–4) sees it as the position for lazy, chauvinistic men and deferential women, since the woman must do all the work and service her ive partner, who is left idle and free of the responsibilities of erotic labour. Catherine Johns (1982, 136–7), on the other hand, sees the prevalence of this position in Roman art as an indicator of the sexual emancipation of women, for it puts them metaphorically as well as literally on top. The woman who ‘rides’ her partner implicitly occupies the equestrian status, and she has unparalleled freedom of movement and control, since the position itself necessitates ‘her very active cooperation’. Clarke (1998, 217–19) criticizes both these readings for their anachronistic assumptions about this position. He notes that painting would be more likely to portray the woman on top, since this position more fully exposes the parts of the woman’s body he presumes most interesting to the viewer. If Clarke is right that the mulier equitans position naturally lends itself to representation because it lays open the most arousing aspects of the act for the viewer’s gaze, then we have all the more to for the apparent absence of this position from erotic images of some eras (see Dover 1978, 107). 34

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN CULTIC AND EROTIC ART

63

the Palatine.36 A man and woman are having sex and she twists around to kiss him, while on the wall behind them a shuttered portrays another erotic image. The sexual positions of the two couples are not identical, though they are both rear-entry. The woman’s hairstyle is the same, and in both images the woman looks back at her lover. What is most interesting is that the woman’s line of sight, if not obstructed by her lover’s head, runs straight to the painting on her wall. There is, of course, a difference in medium between wall painting and engraving, so the mise en abîme effect is not quite as seamless as in the paintings from the House of the Centenary, which may function for the people in room 43 just as the paintings within them do for the lovers. However, the mirror must be understood as an essential trapping of its owner’s toilette and can easily be imagined scattered somewhere within her room, just as on the mirror the woman’s intimate possessions (her slippers, footstool, puppy, bowl, pitcher, etc.) are shown occupying the same space as her lovemaking.37 In other words there may be within this image a hint that mirrors are just the sort of thing one might find lying around in a woman’s cubiculum, especially when she has been expecting her lover, in which case the mirror stands in the same relation to its owner as the painting does to the woman on the mirror.38 The shift in medium is also more suggestive than it may seem at first. The painting, by virtue of being included in an engraving, is only notionally a painting. By being brought into the engraving it has been upgraded from the two-dimensionality of a painted image to the semi-three-dimensionality of an engraving. If the mirror’s owner falls within its spell and enacts in her own cubiculum a scene somewhat like the one in the mirror, the projection to full three-dimensionality will be complete. This same possibility of the woman’s self-recognition, including the play with dimensional shift, informs the dynamic of the other side of the mirror, where the owner, who has been encouraged to identify with the woman on the mirror now becomes the woman in the mirror, this time in an image of depth rendered as surface. The mirror’s dimensional play may also explain the significance of the positions of the two pairs of lovers. As we noted earlier they are not identical, but they are quite similar (both rear-entry with the woman craning her neck around to look back at her lover), and there is an emphasis on the woman in the room looking back at the tabella, where her gaze would be met by that of her painted counterpart, looking back past her lover and out of the painting. One consequence of combining the similarity of their positions with the visual line between the two

A photograph of this mirror-back may be found at Clarke 1998, 44. For such images in ceramic media, see Marcadé 1965, 83 and 85C. 37 On the mirror’s associations with the feminine world, see Bartsch 2006, 29–30. 38 Goldhill (2001, 174) makes a similar point, as does Myerowitz (1992, 145–7). 36

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

64

MIMETIC CONTAGION

women is that they are near inversions of each other on the right/left axis or, as we say, ‘mirror images’. The two women seem to be looking back at each other, but the one is looking over her right shoulder and the other is looking over her left shoulder. The mirror’s capacity to distort by reversal of right and left was infamous in antiquity, and given the representational context of these lovers on a mirror, it is hard to ignore the reversal of polarity in their positions.39 Note, however, that if the woman is looking up at her lover and along the line of her body she cannot, within the fiction of the three-dimensional room, also be looking back at the tabella. The fact that she is in reality also facing the tabella is due to the flattening of her room into two dimensions for the engraving. If the tabella were actually conceived as a mirror on the back wall of her room, not only would there be no variation in their postures, but for the viewer standing in our notional position in front of the bed, the only reversal would be a Z-axis inversion. In other words, we would see the back of her in the tabella-mirror, but there would be no inversion of right and left. In order to see the tabella as a kind of mirror image we have to imagine that her room is (as it actually is) flat. The woman in the tabella is a mirror image of her only to the extent that we see it not from ‘our perspective’ in the reconstructed three-dimensional room, but from her perspective within the twodimensional space. The other side of the mirror renders us as flatlanders, whose three-dimensionality is an illusion, but on this side of the mirror we are induced to see mirroring from the perspective of a flatlander and simultaneously to allow the illusion of three-dimensionality to continue. The fact that we have no trouble intuitively seeing the tabella’s right/left inversion as a mirror image indicates how well the mirror succeeds in putting us in the woman’s place.40 The tiny painting at the top centre of this bronze mirror back, once noticed, alters the meaning of image and gaze, first for the lovers and then for the owner and her partner and finally for anyone who looks into the other side of the mirror. In this way, it is an excellent comparison for the painting at the centre of the Eunuch, for as we shall see, the painting is first replicated within Chaerea’s intrigue, then it becomes a model for how artifice will work throughout the play, and finally, teasingly, it threatens projection on to the audience.

39 On inversion in mirrors, see Pl. Tht. 193c–d and Ti. 46a–b, Lucr. 4.292–301, Ptol. Optica 3.96, Apul. Apol. 16.3; see also Bartsch 2006, 38. 40 In the Villa of the Mysteries there is a similar representation of a mirror image that bends the rules of optics to implicate the viewer. One of the last frames of the narrative series shows the now initiated woman at toilette with her hairdresser, while a cupid, standing ‘further back’ in the painting, holds a mirror up for her, and her mirror image, as one would expect, gazes out of the mirror back towards her. However, impossibly, the woman herself is not looking at the mirror. Instead, she gazes out into the room, apparently at the viewer (Herbig 1958, 37). So the viewer is transposed into the role of the mirror or perhaps of the initiated woman looking into the mirror. On the embedding of this dynamic in the narrative of the rest of the room, see John Henderson 1996, 260.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN CULTIC AND EROTIC ART

65

MIMETIC CONTAGION IN THE EPHESIACA

This chapter has surveyed a wide range of cultic and erotic objects of art that in one way or another work by recreating themselves in the behaviour of the notional viewer. Some of these have been real artworks that have survived to us or are widely attested types and others are singular literary creations, conjured with text and existing only in the imagination of author and reader. This latter category, the instances of mimetic contagion in Ovid and Propertius for example, we have considered primarily as isolated events, collateral evidence for the phenomenon, rather than asking what role mimetic contagion, thematized in these moments, plays in the Tristia or the Monobiblos more generally. Since this is precisely the approach I wish to take with the Eunuch, it will be worth considering another, remarkably similar literary moment to show how the logic of mimetic contagion may propagate out from an embedded artwork and suggest itself as a thematic principle in a work of literature. A boy and a girl gaze up together at an erotic mythical representation hanging above the bed in which they will soon consummate their union. This scene sketch could describe backstage events midway through Terence’s Eunuch, the unseen scene that provides the focus of this book, but it also fits a moment in the first book of the Ephesiaca by Xenophon of Ephesus.41 The lovers are the newly wedded Habrocomes and Anthia, and portrayed in embroidery on the canopy above their marriage bed are Ares and Aphrodite being brought together by Eros for their tryst. ἦν δὲ αὐτοῖς ὁ θάλαμος πεποιημένος· κλίνη χρυσῆ στρώμασιν ἔστρωτο πορφυροῖς καὶ ἐπὶ τῆς κλίνης Βαβυλωνία ἐπεποίκιλτο σκηνή· παίζοντες Ἔρωτες, οἱ μὲν Ἀφροδίτην θεραπεύοντες (ἦν δὲ καὶ Ἀφροδίτης εἰκών), οἱ δὲ ἱππεύοντες Ναβαταίαις στρουθοῖς, οἱ δὲ στεφάνους πλέκοντες, οἱ δὲ ἄνθη φέροντες· ταῦτα ἐν τῷ ἑτέρῳ μέρει τῆς σκηνῆς· ἐν δὲ τῷ ἑτέρῳ Ἄρης ἦν οὐχ ὡπλισμένος, ἀλλ᾽ ὡς πρὸς ἐρωμένην τὴν Ἀφροδίτην κεκοσμημένος, ἐστεφανωμένος, χλαμύδα ἔχων· Ἔρως αὐτὸν ὡδήγει, λαμπάδα ἔχων ἡμμένην. ὑπ᾽ αὐτῇ τῇ σκηνῇ κατέκλιναν τὴν Ἀνθίαν, ἀγαγόντες πρὸς τὸν Ἀβροκόμην, ἐπέκλεισάν τε τὰς θύρας. (1.8.2–3) The chamber had been prepared for them. A golden bed was spread with purple sheets and above the bed a Babylonian canopy had been embroidered. There were Erotes at play, some serving Aphrodite (there was also an image of Aphrodite), others riding Nabataean ostriches, others plaiting garlands, others bringing flowers. This was on one part of the canopy. On the other was Ares, not dressed for battle, but adorned to make love to Aphrodite, garlanded, wearing a cloak. Eros led him on the way holding a burning torch. It was beneath this canopy that they brought Anthia to Habrocomes and laid her out, and then they closed the doors.

41

The world of New Comedy is one of the most important sources for the ancient novel (for a compact survey of formal affinities and the recent scholarship, see Ruiz-Montero 2003, 52–4), but the similarities between this moment in the Ephesiaca and the Eunuch, along with these texts’ shared interest in mimetic contagion, need not be due to direct generic influence.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

66

MIMETIC CONTAGION

At the simplest level the image is suggestive of the sex the gods and the newlyweds are about to enjoy. Thus the canopy establishes a parity between the two couples, though it is not immediately clear whether we should understand its function visà-vis the human lovers as predictive or persuasive. As a specimen of anticipatory ekphrasis, this description of the canopy is hardly unique in the novels. Indeed, Bildeinsätze are both frequent and focal in almost all the surviving ancient novels, and they are usually somehow related to the plot. In some cases these works of art merely warn the characters of their impending future and thus are functionally akin to dreams and oracles, but in others they intervene more directly in the ‘real’ world by recreating themselves in the lives of their viewers.42 A particularly obvious example of this second type is the painting of Andromeda in Heliodorus’ Aethiopica 4.8. The Ethiopian queen was gazing at this painting as she and the king made love, so the image of the naked Andromeda, white skin and all, was impressed on the child they conceived.43 The queen’s response to the painting is not merely aesthetic or discursive. The image propagates itself in her body, and hints at a much broader pattern in the novel, which Tim Whitmarsh has called ‘ekphrastic “contagion”: the power of the visual icon to infect its surrounding discourse with ontological and perceptual uncertainty’.44 Descriptions of works of art that ‘seep’ into the story may be traced throughout the Aethiopica, so the queen’s interaction with the painting comes to seem programmatic for the more general dependence of life on art within the novel. This is especially clear since, as Shadi Bartsch has noted, this unusual conception is related in book 4, but it is actually the earliest event in the story and the cause of all future circumstances, making it directly parallel to the novelistic convention of the introductory painting ekphrasis, though in typical Heliodoran fashion it is playfully denatured by its transposition to the middle of the novel.45 Furthermore, the metapoetic aspect of this remarkable tale of art and response is made clear when we that the story is embroidered on a ribbon in the hard-to-read Ethiopian hieratic script and is being read, or rather recounted as read, by the arch-deceiver Calasiris. Artifactuality, narration, and interpretation converge here, so that the reader is no safer from ekphrastic contagion than any other character in the novel.46 42

Foundational for the study of ekphrasis in the novels as more than an ornament (or, as some would have it, a blemish!) is Bartsch 1989. More recently in this vein, but only for the Aethiopica, see Whitmarsh 2002. 43 The impression of an image on a foetus at conception has widespread ancient and modern attestation, but this example is vivid enough that Michael Reeve had dubbed the phenomenon more generally ‘the Andromeda Effect’ (Reeve 1989). This may indeed be one of the most frequently claimed forms of mimetic contagion, and the collocation of mythological painting, sex, and imitation is immediately suggestive for our reading of the Eunuch. 44 45 Whitmarsh 2002, 111. Bartsch 1989, 48. 46 The capacity of an image viewed by a woman to impress itself on the child she conceives need not always suggest less visual forms of imitation or become generally programmatic, but the Aethiopica was not the only time it did. See for example Dion. Hal. On Imitation fr. VI, pp. 203–4 U.-R. (Hunter 2009, 111–20).

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN CULTIC AND EROTIC ART

67

The Aethiopica is a highpoint of the ancient genre, and Heliodorus deploys art and its effects with great subtlety and sophistication. By contrast it is not surprising that the canopy ekphrasis in the Ephesiaca has been neglected, but even in this undeniably humble specimen similar dynamics may be recognized.47 The portrayal of the assignation of Ares and Aphrodite within an inset work of art has an obvious precedent in the second song of Demodocus (Od. 8.266–332). As the blind bard who sings of gods and men, Demodocus seems a fairly clear figure of the Homeric poet himself, so too in the Greek romance, where ekphrasis is marked as a touchstone of the novelist’s art, the description of the introductory erotic artwork functions as a heraldic miniature of what novelists do. Thus the canopy ekphrasis in the Ephesiaca is similar to Demodocus’ song both in mythological subject and in metapoetic function. This generically self-referential dynamic is even recognized within the novel by Habrocomes, who, lying beneath the canopy, exclaims that Anthia is ‘more fortunate than any woman ever narrated’ (τῶν πώποτε λαλουμένων εὐτυχεστέρα, 1.9.3). The novels are notoriously fond of ‘graphic analogue from myth’, as Grundy Steiner calls it, whereby a character in the story, particularly the ingénue, is assimilated to a figure from traditional mythology.48 However, the Ephesiaca is unusually restrained in this regard, with only one explicit comparison: at the beginning of the novel, before she meets Habrocomes, Anthia is compared to Artemis (1.2).49 Yet after tonight the maiden goddess of the hunt will no longer be an appropriate paradigm for Anthia. Even as the marriage bed itself despoils her of her old station in life, the image on the canopy above has the function of providing her with a new divine model. Anthia is brought in and presented to Habrocomes, and, as they lie together in hesitation, on the canopy above them Erotes prepare Aphrodite for her union with Ares. Some of them ride strouthoi, 47 The exception to this general neglect is Shea 1998, which takes the canopy ekphrasis as a starting point for understanding the novel, but goes in a very different direction than I do here. Her suggestion is that the Ephesiaca has been misfiled as a novel, when it is really a serial ekphrasis. She imagines a sympotic performance context in which the speaker goes from one artwork to another in the patron’s dining room, stitching together a narrative to unify the images into a single love story, a kind of Stations of the Romance. This explanation would provide an alternative to the epitome theory to for the vertiginous inconsistency in narrative pace and the intensely episodic character of the plot, since the text that has come down to us would be simply the script of one such performance. I find this suggestion of generic misidentification intriguing, and it is an attractive way of avoiding the conclusion that Xenophon is a bad novelist or that we only have an epitome of his novel, but without any further evidence that such a form of entertainment existed and some way of connecting the Ephesiaca to it, her theory must remain a speculation. 48 Steiner 1969. 49 Perhaps it is unwise to assume this was intended to be the only graphic analogue in the novel, since, as has sometimes been argued, the novel we now have may be an epitome or partially epitomated form of the original. The ‘epitome theory’ was introduced in Bürger 1892 and became dominant for a while, but since Hägg 1966, many scholars have been less convinced, and most now implicitly treat the work as a unity. For the complex arguments on both sides of this issue and a full bibliography, see O’Sullivan 1995.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

68

MIMETIC CONTAGION

the phallic bird par excellence, and some of them bring in flowers (anthe) in etymological anticipation of Anthia and her defloration.50 Anthia’s assimilation to Aphrodite goes beyond her loss of virginity. There is also a shift in her character, evident in the immediately following scene (1.9) when she becomes sexually aggressive and upbraids Habrocomes for his unmanly delay in making love to her. The replacement of Artemis with Aphrodite corresponds not just to Anthia’s objective change of station, but to an ethical makeover which began when she first saw Habrocomes and is enforced to its conclusion under the canopy.51 The novel has done much in the first half of book 1 to prepare us for the critical role this image plays. The story opens with an enthusiastic description of Habrocomes, whose beauty rivals the gods’ and emboldens him to spurn the power of Eros: Ἔρωτά γε μὴν οὐδὲ ἐνόμιζεν εἶναι θεόν, ἀλλὰ πάντη ἐξέβαλεν ὡς οὐδὲν ἡγούμενος, λέγων ὡς οὐκ ἄν ποτε οὐ<δὲ> εἷς ἐρασθείη οὐδὲ ὑποταγείη τῷ θεῷ μὴ θέλων· εἰ δέ που ἱερὸν ἢ ἄγαλμα Ἔρωτος εἶδε, κατεγέλα, ἀπέφαινέ τε ἑαυτὸν Ἔρωτος παντὸς καλλίονα καὶ κάλλει σώματος καὶ δυνάμει. καὶ εἶχεν οὕτως· ὅπου γὰρ Ἀβροκόμης ὀφθείη, οὔτε ἄγαλμα <καλὸν> κατεφαίνετο οὔτε εἰκὼν ἐπῃνεῖτο. (1.1.5–6) Eros he considered no god at all, but completely discounted him and reckoned him of no importance, saying no one would ever fall in love or be subordinated to the god unless he wanted to. Whenever he saw a temple or statue of Eros he would make fun of it and proclaim himself more handsome than any Eros, both in the loveliness of his body and in power. And it was true, for wherever Habrocomes was seen, no statue appeared lovely and no image was praised.

Even in statuary proxy Eros takes umbrage at being upstaged and, swearing vengeance on Habrocomes, finds his most irresistible weapon in Anthia. Immediately after the ekphrasis of Anthia in Artemisian form, we learn that some thought she actually was Artemis and others held her to be ‘someone else fashioned by the goddess’ (ἄλλην τινὰ ὑπὸ τῆς θεοῦ πεποιημένην, 1.2.7). Habrocomes, likewise, is proclaimed a ‘matchless representation of a handsome god’ (οἷος οὐδὲ εἷς καλοῦ μίμημα θεοῦ, 1.2.8). Hearing each other described as works of art, it is no wonder that each longed to see (ἐπεθύμει ἰδεῖν, 1.2.9) the other. When at last they do set eyes on each other, the optical infection is mutual and immediate: ἐνταῦθα ὁρῶσιν ἀλλήλους, καὶ ἁλίσκεται Ἀνθία ὑπὸ τοῦ Ἀβροκόμου, ἡττᾶται δὲ ὑπὸ Ἔρωτος Ἀβροκόμης καὶ ἐνεώρα τε συνεχέστερον τῇ κόρῃ καὶ ἀπαλλαγῆναι τῆς ὄψεως ἐθέλων οὐκ 50 On the strouthos as Aphrodite’s bird, see Pollard 1977, 29 and 147. On these anthe as figura etymologica, see Cueva 2004, 42. 51 Note that in her Artemis phase, Anthia’s eyes were timid or austere (φοβεροί ) as befits a woman in control of her desires (ὡς σώφρονος, 1.2.6). By the time she chides Habrocomes for his cowardly hesitation in bed she has a new tutelary deity now and her bearing has changed accordingly.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN CULTIC AND EROTIC ART

69

ἐδύνατο· κατεῖχε δὲ αὐτὸν ἐγκείμενος ὁ θεός. διέκειτο δὲ καὶ Ἀνθία πονήρως, ὅλοις μὲν καὶ ἀναπεπταμένοις τοῖς ὀφθαλμοῖς τὸ Ἀβροκόμου κάλλος εἰσρέον δεχομένη. (1.3.1–2) Then they see each other: Anthia is captivated by Habrocomes and Habrocomes is bested by Eros. He kept looking at the girl and, though he wanted to, was unable to tear himself away from the sight; the god lay heavy on him and held him in possession. It was bad for Anthia, too, as Habrocomes’ beauty flowed in at her enrapt, wide-open eyes and she received it.

They just want to keep looking at each other and they only leave off doing so with difficulty, but even then ‘reflection on the sight of each other infiltrates them’ (ἔννοια ἐκείνους ὑπῄει τῆς ὄψεως θατέρου, 1.3.4). Habrocomes recognizes that he is being punished for ridiculing Eros, and he tries, briefly, to separate himself from his ion by distinguishing between his eyes and himself: ‘To your eyes, Habrocomes, is Anthia beautiful, but not to yourself, if you wish it so’ (τοῖς σοῖς ὀφθαλμοῖς, Ἀβροκόμη, εὔμορφος Ἀνθία, ἀλλ᾽, ἐὰν θέλῃς, οὐχὶ σοί, 1.4.3). Such resistance only makes Eros redouble his efforts and Habrocomes is soon forced to it defeat. They both the night in this misery, suffering the lingering effects of the stamp they had inflicted on each other at first sight: ‘They had before their eyes the sight of each other, moulding their images on each other’s souls’ (εἶχον δὲ πρὸ ὀφθαλμῶν τὰς ὄψεις τὰς ἑαυτῶν, τὰς εἰκόνας ἐπὶ τῆς ψυχῆς ἀλλήλων ἀναπλάττοντες, 1.5.1). The description of characters, particularly beautiful women, in reminiscent of artworks, and the fascination with the mechanics of optical impression, are both typical of the ancient novel.52 In the Ephesiaca these topoi lead directly into the canopy ekphrasis and prepare the reader to expect both a high degree of alignment between art and life and a danger of infection through vision. Though this infection is explained in psychological , it is a strongly physicalist psychology, with constant reference to the organs and mechanical processes of perception, and when Habrocomes once tries to make a distinction between his eyes and himself, he is immediately proven wrong.53 The attentive reader will expect, therefore, that the image on the canopy is more than a pretty ornament for the wedding chamber and will not be surprised when it turns out to be broadly determinative of the roles Habrocomes and Anthia will play for the rest of the novel.

52 See, for example, the ekphrasis of Charicleia in Aethiopica 1.2, a description in many ways similar to the description of Anthia in Ephesiaca 1.2. For issues relating to vision in Achilles Tatius, see now Morales 2004. 53 The novels stand out for the consistency of their interest in vision, but the physicality of their of visual perception is hardly unique, and the consequences of such models of vision are common to many ancient texts; see Goldhill 1996 and 2001. Both in its final form and in various stages of preparation, Bartsch 2006 has been particularly important for my understanding of this issue.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

70

MIMETIC CONTAGION

An innocent reading of the canopy—surely the lovers’ reading as they gaze up at it—would only suggest an erotic context appropriate for the night of love they are about to enjoy (1.9.9), but this straightforward interpretation is frustrated by the prominence of Eros in the image, both in plural form thronging Aphrodite and in singular form conducting Ares. From the beginning of the novel, as we have seen, Eros sets the plot in motion as retribution for Habrocomes’ pride, and when Habrocomes finally capitulates and confesses his subjection, the god is not appeased: ‘This is what he said, but Eros was still angry and plotted to inflict a great punishment on Habrocomes for his arrogance’ (ταῦτα ἔλεγεν, ὁ δὲ Ἔρως ἔτι ὠργίζετο καὶ μεγάλην τῆς ὑπεροψίας ἐνενοεῖτο τιμωρίαν εἰσπράξασθαι τὸν Ἀβροκόμην, 1.4.5). The punishment is never explicitly stated, but as they are already in love, the god’s scheming must point to something beyond that. When their parents consult the oracle they are told that Habrocomes and Anthia are suffering from the same disease, which will also be the source of their cure, and that terrible suffering and endless toils await them afterwards, but eventually they will offer gifts to Isis as saviour and have a better lot in the end (1.6.2). Eros is not named here but it seems clear that the divine source of their suffering now and later is the same, as opposed to the goddess who will preside over their final salvation. Eros is the deity driving the plot of the novel onwards, and Isis is the deity who will allow the novel to end. The dominant role Eros plays in the canopy is therefore quite ominous. In the beginning Habrocomes was said to scoff at representations of Eros, but this is the only time in the narration that the lovers are confronted with their divine opponent in visible form. The canopy may have been intended only to set an aphrodisiac mood, but it reveals the truth about the force driving them together and determining the course of their life. Throughout the novel the image on the canopy threatens to come true in ways they could never have guessed on their wedding night. For instance, in the next book (2.13), Anthia has been taken prisoner by a band of robbers and comes dangerously close to being sacrificed to Ares. Rather than playing the role of Aphrodite opposite Habrocomes in the role of Ares, Anthia is to be given directly to Ares, a horrifically literal misinterpretation of Aphrodite’s role on the canopy. There is another sense in which a straightforward identification of Habrocomes and Anthia as Ares and Aphrodite is problematic. The scene on the canopy is not conjugal love, but rather adultery, and as Gareth Schmeling has noted, this image is inappropriate for a marriage bed.54 However, we may readily compare the coverlet in Catullus 64, where the sad story of Theseus and Ariadne is superficially inconsonant, yet in hindsight profoundly fitting, for the marriage of Peleus and Thetis.55 The resonance between ‘Why put such a scene over the marriage bed of especially chaste lovers? Does Xenophon know what he is doing? do dirty scenes produce fertility? ion? or is a little humor intended?’ (Schmeling 1980, 28). 55 On the significance of the myth on the coverlet for the framing myth, see Townend 1983. 54

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN CULTIC AND EROTIC ART

71

Habrocomes and Anthia’s fate and the image of Ares and Aphrodite is deep, but not as they could have known or wished. Shortly after their wedding night the newlyweds will be separated, and they will not come together again until the end of the story, after being tossed from one end of the Mediterranean to the other, but throughout the novel because of their exceptional beauty they will attract the attention of a seemingly unending series of lustful brigands, bandits, brothel owners, maharajas, military men, and infatuated women in the tradition of Potiphar’s wife. They will both, repeatedly, come within inches of being compelled to break their marriage vows. The threat of forced infidelity is common enough in the Greek novels, but only in the Ephesiaca and in Chariton’s Callirhoe could it be framed as adultery, for only in these two novels are the lovers married at the beginning, rather than at the end of their trials. So the theme of adultery on the canopy is not only apposite to Habrocomes and Anthia’s story, it is peculiarly right for them in a way it would not be for the hero and heroine of most of the other novels that have survived to us. In Demodocus’ song the bed in which Ares and Aphrodite couple has been fashioned into a trap, for hanging above their heads is a finely wrought web, indiscernible to them, which will fall upon them and constrain them with ineluctable bonds (Od. 8.274–5). In the same position, hanging above the lovers, the canopy in the Ephesiaca reveals the victory Eros can already celebrate over Habrocomes and Anthia, but it also hints, in ways still opaque to his hapless victims, at the future torments the god has in store for them. If we what the novel has already said about how vision, specifically vision of a beautiful work of art, can infect the soul with a lasting pattern, as well as the identical placement of the canopy and Hephaestus’ net in the story told on the canopy, we will recognize that the canopy both represents and enforces the god’s triumph, as it locks Habrocomes and Anthia into their future roles.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

THREE

Lifelike Likeness Mimetic Contagion in the Philosophical Tradition

Psychologie ist die Beschreibung der Spiegelung der irdischen Welt in der himmlischen Fläche oder richtiger: Die Beschreibung einer Spiegelung, wie wir, Vollgesogene der Erde, sie uns denken, denn eine Spiegelung erfolgt gar nicht, nur wir sehen Erde, wohin wir uns auch wenden. Franz Kafka, Oktavhefte (18 October 1917)

Some of the images surveyed in the previous chapter function persuasively on the viewer by virtue of their ritually correct deployment or aptness of contextual placement, and their power is not explicitly linked to any particular artistic skill with which they are wrought. In other cases, though, there seems to be an interest in developing the analogy between the image’s apparently miraculous likeness to its model and its ability to generate likeness in the viewer by mimetic contagion. This fascination with the power and danger of verisimilitude may be more prevalent in elite cultural discourses surrounding images, and where Chapter 2 looked at humble cultic talismans and brothel décor, this chapter will be more closely concerned with famous masterworks of lifelikeness and the Socratic response to the problem or potential of mimetic contagion. Xenophon’s Socrates comes to the topic of mimetic contagion through veristic elite painting, whereas Plato’s Socrates is less focused on the effect of this kind of artistic skill and more interested in the problem of representation itself, but Plato’s treatment of mimesis is at least partly framed in of the power of images to replicate themselves in viewers, so this chapter will move from mimetic contagion in ancient discourses of lifelikeness to explore the significance of mimetic contagion for Platonic mimesis. Unlike the wide range of examples surveyed in Chapter 2, this chapter’s focus on culturally elite art and philosophical thinking about images may seem less immediately relevant to the situation in the Eunuch, for neither Terence nor his audience may be assumed to be philosophers. The charge of philosophical influence may with greater plausibility be levelled at Terence’s model, for Menander was a student of Theophrastus, himself a student of Aristotle, but Greek philosophy had made considerable inroads in Terence’s Rome as well. The Epicureans were

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN THE PHILOSOPHICAL TRADITION

73

expelled from the city in 173 BCE and in 161, the year of the Eunuch’s staging, there was a general expulsion of philosophers and rhetoricians. Furthermore, there is explicit mention of the philosophers and their schools in the Eunuch itself, in a context immediately relevant to the theme of mimetic contagion (see Chapter 6). It is not my contention, however, that Plato’s formulation of mimesis as explored in this chapter had influence, direct or indirect, on Terence’s play, any more than the scene in the Ephesiaca that closed Chapter 2 need be understood as an imitation of the Eunuch. Rather, the texts and objects of Chapter 2 and the thinking about verisimilitude and mimesis in this chapter may all be viewed as parallel cultural phenomena, potentially independent responses to mimetic contagion, which may help to elucidate each other and the Eunuch. Lifelikeness may be mostly in the eye of the beholder. It is hard for a late modern, Western viewer to know how most of the images surveyed in Chapter 2 would appear to someone bred in a radically different visual culture, and absent ancient testimony for any given object, speculation on this score is probably unsound. Tiberius’ prized erotic painting by Parrhasius, however, (see p. 54) may point to an intersection of the aesthetic discourse of persuasive verisimilitude and mimetic contagion. In the famous contest between Zeuxis and Parrhasius, as Pliny relates it (HN 35.65), Zeuxis’ painted grapes are accurate enough (tanto successu) that birds are fooled into trying to eat them, whereas Parrhasius’ painting of a curtain is so very true to life (ita veritate repraesentata) that Zeuxis himself is tricked into trying to draw it aside. Both performances frame artistic success as verisimilitude so extreme that the viewer treats the image as reality.1 In this story, birds and artist take roles complementary to the object represented—grapes call for eating, curtain calls for lifting—but when Parrhasius paints a human being in the throes of some affect, his quasi-mythical ability to pull viewers into participation results in their literal sympathy with the painted subject. Another story about Parrhasius has him buying an Olynthian slave specifically to torture so as to provide himself a model for his painting of Prometheus.2 The slippery boundary between extremely lifelike art and life itself is here again at play, but knowing that there was a live model for the anguish of Prometheus somehow permits the suffering behind the painting to become real again in the viewers. So Seneca the Elder’s rhetorical exercises on the theme of Parrhasius’ terror art argue that the painter has harmed the public weal (accusatur rei publicae laesae, Con. 10.5). 1

Parrhasius is notoriously supreme in this game, not just over Zeuxis, but over all other artists. Athenaeus attributes an epigram to Parrhasius in which he claims to have ‘discovered the clear limits of art’ and ‘fixed a boundary which may not be ed’ (Ath. 12.543e). Pliny also says that Parrhasius called himself ‘prince of artists’ (HN 35.71), but he testifies that other artists concur in giving the palm to Parrhasius for the extreme skill (summa subtilitas) of his graphic line drawing (HN 35.67). 2 Morales 1996 is still the most perceptive treatment of this tale and of the responses to it in the Controversiae.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

74

MIMETIC CONTAGION

Several of the orators quoted by Seneca point to the threat that the horrific realism of Parrhasius’ painting must spill over into reality, but the orator Argentarius is most explicit: ‘An Olynthian was welcomed at Athens with such good cheer? Was it only the Olynthian that Parrhasius tortured? Does he not torture our eyes as well?’ (hoc hospitio Olynthius Athenis exceptus est? tantum porro Olynthium torsit Parrhasius? quid? non et oculos nostros torquet?, Con. 10.5.3). Given Parrhasius’ associations with mimetic contagion through exquisite verisimilitude, perhaps we are meant to understand that Tiberius’ painting was so charming precisely because of how effectively it communicated Meleager’s arousal to the viewer. However, naturalistic technique ‘convincing’ enough to entrance viewers into sympathetic response was not unique to Parrhasius or painting. For example, Pliny tells (HN 34.59) of Pythagoras of Rhegium’s sculpture of a lame man, which was so affecting that viewers experienced the subject’s sore feet. Pliny’s interest in Pythagoras of Rhegium is focused on his role in the development of verisimilitude, beginning with his ‘victory’ over Myron, the innovator of expanded realism (primus hic multiplicasse veritatem, HN 34.58), and ending with his distinction as the first artist to render sinews, veins, and elaborate hair. This focus on Pythagoras of Rhegium’s contest with other artists and innovative struggle to capture the visible world is consonant with Pliny’s broader fascination with the agon between nature and artistic skill.3 An adequate and coherent theorization of verisimilitude has proved totally elusive for modern visual theory, but if, as seems most probable, the perception of representational ‘lifelikeness’ is more cultural than natural, we can expect a permanent gulf between our own fractured discourses of verisimilitude and the ancient practices and experiences associated with the attempt to copy reality in art.4 Despite these difficulties, however, one of the few discernible points of consistency is that such forms of mimesis are so often framed as a summoning, a replication of the absent model.5 Pliny’s story about Pythagoras of Rhegium fits this much larger pattern, then, but it does so in an interesting way that amplifies the rhetorical effect of the topos, for the prototype, the model lame man, is not only evoked in the bronze of the statue, but in the living flesh of the viewer. Because of the successful realism of the artist’s technique, the claudicans is not merely represented to the viewers, nor indeed just ‘re-presented’, made present again before their eyes—rather they become him, if only temporarily, and perceive his pain as their own.

For Pliny’s investment in the ‘artifice of nature’, see Carey 2003, 102–37. The bibliography on the theoretical problem of verisimilitude is vast and beyond the scope of our study. Most recent formulations (see for example Mitchell 1995) are in some way a response to Gombrich 1960. For a nuanced discussion of how these issues apply to ancient painting, see Neer 1995. 5 For the place of verisimilitude in the larger discourse of ‘replacement and replication’, see Steiner 2001, 27–32. 3 4

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN THE PHILOSOPHICAL TRADITION

75

Precisely because such sympathetic response to naturalistic representation was generally held to be possible, or even inevitable, the philosophical tradition tended to approach these mimetic techniques with a certain suspicion, on the one hand, and, on the other, with a qualified interest in their educational value. So Xenophon’s Socrates, in his conversations with the three artists in Memorabilia 3.10, observes that it is the ‘quality of lifelikeness’ (τὸ ζωτικὸν φαίνεσθαι) that ‘principally manipulates men through the faculty of vision’ (ὃ δὲ μάλιστα ψυχαγωγεῖ διὰ τῆς ὄψεως τοὺς ἀνθρώπους, 3.10.6). The point of this ‘manipulation’ becomes clear in the larger context; the ‘character of the soul’ (τῆς ψυχῆς ἦθος) is argued to be imitable (μιμητόν, 3.10.3) because ‘friendly and hostile looking’ (τό τε φιλοφρόνως καὶ τὸ ἐχθρῶς βλέπειν) can be artistically copied in the eyes (μιμητὸν ἐν τοῖς ὄμμασι, 3.10.4). In a sympathetic exchange similar to that notionally expected with the Gorgon (see pp. 52–3), the ‘look’ in the image’s eyes determines the aspect of the viewer. Our expressions modulate to become ‘radiant’ (φαιδροί) at the joys of our friends and ‘gloomy’ (σκυθρωποί) at their misfortunes, so Socrates concludes by asking whether it is not ‘sweeter’ (ἥδιον, 3.10.5) to see exemplars of good character than bad. If feelings (τὰ πάθη) are not only artistically expressible, but visually contagious, then, as Socrates implies, art has a profoundly civic role to play.6 In a similar vein and even more explicitly, Aristotle discusses the ethical effect of art on its viewers in book 8 of the Politics (1340a). The broader investigation is on the infectious qualities of certain kinds of music, such as the melodies of the Phrygian composer Olympus, which ‘make souls inspired’ (ποιεῖ τὰς ψυχὰς ἐνθουσιαστικάς). ‘And moreover just listening to representations everyone comes to have the same feelings [as the representations], even apart from their rhythms and melodies’ (ἔτι δὲ ἀκροώμενοι τῶν μιμήσεων γίγνονται πάντες συμπαθεῖς, καὶ χωρὶς τῶν ῥυθμῶν καὶ τῶν μελῶν αὐτῶν, 1340a 12–14).7 More broadly, Aristotle tells us, certain types of music contain likenesses (ὁμοιώματα) of particular states of the soul, and these states, whether good or bad, are immediately transmitted to anyone hearing the music. The objects of touch and taste do not have this ethically communicative property, but the objects of sight do, to some extent, because they are figures (σχήματα). To be sure, not everyone is susceptible to the ethical content of images, nor is it strictly likenesses of character that are transmitted in this case, but rather the figures and colours are signs (σημεῖα) of character, the ‘Socrates’ conclusion is designed to introduce the ethical—the qualities of a good citizen—into the process of visual representation’ (Goldhill 1998, 111). Goldhill argues (109 and im) that this civic eye for the deployment of art explains Xenophon’s claim that Socrates was ‘useful’ (ὠφέλιμος, 3.10.1) to artists. 7 For the reading of this μιμήσεων as ‘representations’, in the same broad sense as in Poet. 1, rather than the more common translation as ‘imitations’, see Kraut 1998, 194. Kraut’s whole discussion of this age is invaluable for its careful exposition of the likely cathartic overtones of Lydian enthusiasm, what it means that the representations can have feelings, and how they can be conceived as essentially distinct from their melody and rhythm. 6

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

76

MIMETIC CONTAGION

external stamp of the body in an emotional state (1340a 32–5). After this elaborate circumscription of the ethical power of images compared to music, Aristotle its that pictorial art is a potent-enough influence on character to make some works dangerous for the young: ‘But no, since the contemplation of these things makes such a difference, the young should not contemplate the works of Pauson, but those of Polygnotus and any other painter or sculptor interested in character’ (οὐ μὴν ἀλλ᾽ ὅσον διαφέρει καὶ περὶ τὴν τούτων θεωρίαν, δεῖ μὴ τὰ Παύσωνος θεωρεῖν τοὺς νέους, ἀλλὰ τὰ Πολυγνώτου κἂν εἴ τις ἄλλος τῶν γραφέων ἢ τῶν ἀγαλματοποιῶν ἐστιν ἠθικός, 1340a 35–8). We know little of Pauson and Polygnotus, apart from what Aristotle himself tells us in Poet. 2: that the former painted men worse than we are, where the latter painted men better than we are. It does not seem, then, that Pauson’s fault was with his execution, but rather with his choice of subjects. It is perhaps not obvious how one could have known that the characters depicted were so vicious, unless, as Kraut has argued, Pauson’s subjects were drawn from mythology.8 In which case, there would already be an interesting point of comparison with the scenario in the Eunuch, in that a painted scene from one of the more unseemly myths has an immediately deleterious effect on the disposition of the viewer, but the more significant connection between the Politics and the Eunuch is that Aristotle’s concern is specifically for the young, whom he believes to be more vulnerable to this kind of suggestion. But perhaps ‘suggestion’ is the wrong word, at least if it is understood in a purely psychological sense, for one of the most striking things about Aristotle’s discussion of mimetic contagion in music and art is the insistent physicality of his language. He dismisses the relevance of touch and taste to this issue since he believes they have no ethical content, but he brings them up in the first place because this discussion of the implantation of auditory and visual stimuli in the body of the observer is obviously linked to his interest in the general mechanics of perception. Throughout the De anima Aristotle implies that perception somehow involves a physical change in the observer into conformity with the sensed object, as for instance the famous age 424a 17–24, which compares sensation (αἴσθησις) to a wax seal receiving an impression from a ring. The material of the wax is not changed to gold or bronze, but its form is now a copy of the ring (σημεῖον). If by its material recalcitrance the wax is thus saved from becoming itself a signet, the boundary between matter and form is often quite different for Aristotle than it would probably be for us. For instance, he repeatedly insists that the jelly of the eye actually changes colour to match the colour of the object of vision.9 Seeing

Kraut 1998, 197. For chameleon eye jelly and its relevance to the hylomorphic and functionalist reading of Aristotle’s philosophy of perception, see Sorabji 1978, 49. 8 9

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN THE PHILOSOPHICAL TRADITION

77

is thus becoming and the very act of perception is a partial corruption of the boundary between subject and object.

MIMETIC CONTAGION AND (PLATONIC) MIMESIS

The word mimēsis is central to Platonic and Aristotelian discussions of art and thence to all subsequent Greek (Hellenistic, Imperial, Byzantine) aesthetic discourses. Under its Latin calque, imitatio, the concept of mimesis, or rather the complex of concepts covered by this term, continued to play a pivotal role in Renaissance and Neoclassical thinking about what art can and should aspire to do. Romantic and post-Romantic aesthetic regimes have often understood themselves to be anti-mimeticist, and the words ‘mimesis’ and ‘imitation’ have not generally enjoyed currency in recent critical language, but the version of mimesis rejected by the late eighteenth- and early nineteenth-century revolution was only one aspect—and a particularly jejune one at that—of the conceptual field subtended by the term mimēsis in antiquity.10 Indeed even the most avowedly anti-mimetic theorizations of art in the twentieth century have often only succeeded in reanimating one of the less well-known corners of ancient mimesis under a different name. Thus mimeticism, in the broadest sense, has been the conceptual spine that runs through the history of Western aesthetics, and it is implicitly worth considering what its significance may be for our topic of mimetic contagion.11 If the word ‘imitation’ has been rendered irretrievably pejorative in modern English, connoting inauthenticity at best and at worst an outright counterfeit, this demotion of a once esteemed artistic ideal speaks volumes about the infiltration of Romantic aesthetic theory into our habits of speech and thought. The formulation of art as an ‘imitation of nature’ or ‘imitation of life’ was as commonplace from the sixteenth to the eighteenth century as any unquestionably true Aristotelian maxim.12 But the subsequent rise of industrialized mass production, among other late modern experiences of mechanized homogenization, has made this older understanding of art as an imitation of life aesthetically untenable for many of us. It may even be regarded as morally offensive, when the true vocation of art is

10 Auerbach 2003 is still essential for understanding the role of mimesis in critical discourse since World War II. Potolsky 2006 is a good recent survey of positions and issues, but it does not replace Gebauer and Wulf 1992 as an historical survey. For an excellent analysis of modern versions of mimesis as responses to serially isolated aspects of the ancient tradition, see Halliwell 2002, 344–81. 11 And not only aesthetics, but sometimes also anthropology. See, for example, Taussig 1993 for an eloquent exploration of ‘the mimetic faculty, the nature that culture uses to create second nature, the faculty to copy, imitate, make models, explore difference, yield into and become Other’ (xiii). 12 For the reign of art as imitation of life/nature in this period, see Hathaway 1962, 3–125; Halliwell 1992, 412–18.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

78

MIMETIC CONTAGION

taken to be expressive originality, the only legitimate resistance to a world of soulless reproductions.13 Thus for many of us sympathetic engagement with the older language of mimesis/imitation requires a leap of aesthetic imagination, but it also needs patient attention to the breadth of meaning these words enjoyed in a time less immediately suspicious of them. This philological task, however, has proved surprisingly difficult and productive of controversy. The Greek word mimēsis is plainly connected to the word mimos, which denotes both a particular type of generally lowbrow theatrical entertainment and the actor who plays in such a show, a polysemy that lives on in the Latin mimus and the English ‘mime’.14 The problems of mimesis, including mimetic contagion, are never too far from the associative field of mime, and I shall explore this connection further in Chapter 5. But from as far back as we have evidence the noun mimēsis and the related verb mimeisthai are employed in contexts much wider than the performance of theatrical mime, and it would be a mistake to assume that mime is universally determinative of the meaning of this Wortfeld, even in the earliest pre-Platonic uses.15 If anything, the unifying theme to these early attestations is not theatre, but a diffuse sense of ‘similarity’ or ‘likeness’, and perhaps it was the very breadth and abstractness of this concept that recommended mimēsis to Plato for such a central role in his thought. But to say that Plato, or for that matter Aristotle, has a unified doctrine of mimesis would be altogether wrong. Plato uses the word in various ways at different points in his career, sometimes apparently in revision of his own earlier ideas. Aristotle perpetuates this instability of meaning, both because he is pointedly responding to Plato by using the word differently and because, for him no less than for his teacher, mimesis is in itself a richly polyvalent concept. Still, as Halliwell has demonstrated, it is possible to class the various usages of the word, especially as it emerges from Plato and Aristotle, under two broad headings: mimesis as worldreflecting and mimesis as world-creating.16 For the sake of economy and clarity, and at risk of imposing an overly technical distinction where for the ancients there was ambiguous unity, I shall call these mimesis1 and mimesis2, respectively.17 The most eloquent exploration of these issues is still Walter Benjamin’s essay ‘Das Kunstwerk im Zeitalter seiner technischen Reproduzierbarkeit’ (1936). 14 The etymology of the mim- root is totally opaque. Schulze (1934, 53) once proposed a connection to the Skt. ma ˉya ˉ, ‘illusion’—a delightful suggestion which, if true, would amount to a genetic connection between mimesis and one of the central motifs in Hinduism and other eastern religions. But, alas, all recent Indo-Europeanists are agreed in dismissing this etymology and declaring a linguistic dead end. See, for example, Beekes 2010, 954–5. 15 For a careful examination of the limited textual evidence and the copious earlier scholarship, see Halliwell 1998, 109–16. 16 Halliwell 2002, 23 and im. 17 I intend no reference to Paul Ricœur’s similarly subscripted enumeration of different mimeses. His interest is in the mimetic events that occur prior to the act of aesthetic creation and subsequently in the 13

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN THE PHILOSOPHICAL TRADITION

79

Most of the pre-Platonic and Platonic attestations of the word are some version of mimesis1, references to the copying of a model, either in performative action or as an artistic product fashioned in matter. But alongside these instances of mimesis1 sometimes (and with increasing frequency in Aristotle and later writers) the emphasis on copying or mirroring seems to be absent, and the word refers to the production of an autonomous artistic creation, so that in some cases mimesis2 might virtually be rendered as ‘fiction’. It is mimesis1 that is most obviously relevant to our investigation of mimetic contagion, since the dynamic that interests us here is the imitation of a visible model. But if in most cases the model is assumed to be natural (i.e. not a product of art) and the copy a work of representation, then mimetic contagion effectively inverts this pattern by casting the artwork as the model and the viewing subject as the copy. To be clear, this is still a matter of mimesis1, only the causal directionality has been switched and the looking glass reversed, with the viewer now acting as the image of the artwork. There is, however, also a sense in which mimesis2 is relevant to mimetic contagion. When the world is refashioned in accordance with an artwork it can no longer be clear what the boundaries of the artwork are, since by drawing the viewer into recapitulation of its own ethos or action, it effectively exceeds its frame. Its fiction now includes the viewer, who is properly part of the fashioned world of the artwork in the sense that mimesis2 generally implies. However, this mimetically produced world is not safely autonomous from the ‘real’ world, but rather a replacement for it. Thus by inverting the usual sense of mimesis1, mimetic contagion offers an ironic twist on mimesis2. This inversioncum-irony points to a departure from the typical order of mimesis, but it also indicates where the two facets of mimesis, world-reflecting and world-creating, are fundamentally linked. If mimetic contagion represents a perversion of mimesis as normally understood, the threat posed by this perversion is always latent in the Greek understanding of mimesis, and Plato’s infamous reservations about the role of mimetic arts in the well-run city may be largely explained by the perception of just this danger. Plato’s critique of mimesis in the Republic is complex, but it emerges in ways that may be grouped under two broad headings, one concerned with the falsehood of mimesis and the other with its capacity for ethical corruption.18 The treatment of the bed in Republic 10 (596b–99a), for example, would be in the first category. The god

reader. For him these ‘upstream’ (mimesis1) and ‘downstream’ (mimesis3) events are no less mimesis than the creative act itself (mimesis2). His enumeration establishes a temporal order within what he considers a single process; mine is meant to distinguish between different types of mimesis. See Ricœur 1984. 18

These two complaints about mimesis may, of course, be linked to each other, especially by seeing the latter as a natural consequence of the former. Attempts to do so almost always focus on book 10 to the near exclusion of book 3. See, for example, Moss 2007.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

80

MIMETIC CONTAGION

cogitates an ideal bed, which is fully real; from this pattern the craftsman makes a phenomenal bed, removed from reality by one degree; and the painter imitates the mere appearance of this phenomenal bed, a type of creation which is now two degrees removed from truth and reality. For all the cogency this ontological critique has, both within Plato’s system and in the philosophical tradition more broadly, it is not immediately relevant to mimetic contagion. However, some of the Republic’s earlier attempts to associate mimesis with illusion and falsehood may be more germane to our topic. At the end of book 2 and the beginning of book 3, for instance, there is an extended critique of the presentation of the gods and heroes in traditional Greek poetry, where presumably they could have been accurately rendered as virtuous, but in fact they are often misrepresented as weak-kneed scoundrels.19 Adeimantus assumes that these stories have a bad effect on the young (365a), and Socrates agrees that since young people see the gods and heroes giving way to grief, for example, they will be influenced to do the same when they meet with adversity in their own lives (388d). More generally, such stories offer an excuse for bad behaviour and persuade us of the universality of turpitude, and so they ‘engender a proclivity (εὐχέρειαν ἐντίκτωσι) to wickedness in the young’ (392a). The danger that Adeimantus and Socrates expound here is certainly a version of mimetic contagion, but Plato’s agenda is manifold and his concern is not simply that impressionable people will take the representation as a model for imitation, but that the representation is in itself false and that the young will learn bad theology from it. If their bad behaviour is then the consequence of their erroneous ideas about what is desirable or even possible in a human life, rather than due to their direct imitation of the traditional stories, we may not wish to label this a straightforward instance of mimetic contagion. After expounding on the danger of the content (logoi) of these traditional stories, Socrates turns to address the issue of the style (lexis) in which they are mediated (392c). Here at last the word mimēsis enters the argument, in contrast to diēgēsis (‘narration’). We may imagine a type of poetry, for example dithyramb, consisting almost entirely of diēgēsis, while tragedy is obviously all mimēsis, and epic blends the two styles. This distinction between telling and showing—one of the most resilient elements of Plato’s thought in modern literary criticism—is as readily intelligible for us as for Adeimantus, and on the surface it seems to offer a neatly value-free spectrum for describing various types of storytelling.20 No 19 The criticism of the gods of Greek myth as overly anthropomorphic is, of course, a well-established tradition by the time of Plato, and it is tempting to see his hostility to mimesis more generally in this context as an extension and intensification of the pre-Socratic revision of received poetic wisdom. See Papadopoulou 2006. 20 For a thoughtful appraisal of this dichotomy and its critical Nachleben in Aristotle and in modern narratology, see Kirby 1991.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN THE PHILOSOPHICAL TRADITION

81

sooner is the distinction explained, however, than Socrates clarifies that the pressing question is now what kinds of mimēsis will be allowed in the city, if indeed it will be allowed at all (394d). Apparently diēgēsis will not be subjected to the same port scrutiny, though based on the preceding discussion of the danger of traditional poetry it is not immediately obvious why mimēsis should be singled out for interrogation, since theological errors may be conveyed by diēgēsis just as well. It bears remarking that this is mimesis1, imitation rather than imaginative representation in general, since the latter would certainly include narration and could not serve as its opposite. Socrates’ first concern is with the effect of mimesis on the integrity of the performer. Twice already in book 2 (370a–b; 374a–c) he had spoken of the need for specialization in the ideal city and he returns to this theme here. It is impossible to imitate many things as well as one, so we do not expect the same actor to perform comedy and tragedy with equal skill (395a). Likewise, it may be impossible both to imitate an action and to do that same action well, so if we want the guardians to be ‘craftsmen of the city’s freedom’ (395c), we must not wish them to pursue other crafts or imitations. This is a glancing preview of the kind of argument he will advance in book 10 in his ontological critique of mimesis, but here he immediately reverts to the moral argument and, without explanation, concedes that the young guardians may engage in some kinds of mimesis, but it must always be imitation of virtue and strength and never of vice or weakness. Whatever is imitated will eventually be naturalized and settle into character and second nature (εἰς ἔθη τε καὶ φύσιν, 395d). Fundamentally mimesis is likening oneself to another (ἀφομοιοῦν αὑτούς, 396a), but if we liken ourselves to men ultimately the madness will be our own, and we shall find ourselves actors no more. As we saw in the previous section, the danger of immoral representations of the gods and heroes might be construed as based in mimetic contagion, since the young would in fact behave like the characters, or their behaviour might have been understood as a ramification of bad theology, but now the danger is explicitly a matter of imitation. The mimesis of wickedness in any form, divine or human, false or true, is perilous because it tends to become unruly and outlast the performance by becoming the real character and nature of the performer. If we have defined mimetic contagion, however, as the viewer’s tendency to imitate or absorb behaviours or ethical qualities from an artwork, we may still note a possible discontinuity between the danger Socrates describes here and mimetic contagion. The mimetic poetry he has in mind is principally tragedy, which would of course have a visible public performance context. Drama in action is a visual art, and throughout antiquity the association between theatre and painting is close (see p. 107). It is primarily for the audience, however, that the drama is an object of spectation, whereas the actors can only be understood as ‘viewers’ of their roles in a metaphorical or reflexive sense. It may cause us to hesitate, therefore, in labelling

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

82

MIMETIC CONTAGION

the danger Socrates perceives as a form of mimetic contagion when we recognize that the concern he expresses in this section is entirely on behalf of the performers; the audience is not in view here. It will soon become clear, though, that this discussion has relevance for the visual arts as well as poetry and for the audience as well as the actors. Socrates shifts from dramatic performance to other poetic and musical styles, discussing the metres, modes, and instrumentation proper to each, and discovers that it is generally possible in all these cases to discern what ought to be allowed in the ideal city on the basis of whether it accords with harmony, grace, and intelligent order (398c–400e). However, such virtues and their opposites are not the exclusive province of poetry: Ἔστιν δέ γέ που πλήρης μὲν γραφικὴ αὐτῶν καὶ πᾶσα ἡ τοιαύτη δημιουργία, πλήρης δὲ ὑφαντικὴ καὶ ποικιλία καὶ οἰκοδομία καὶ πᾶσα αὖ ἡ τῶν ἄλλων σκευῶν ἐργασία, ἔτι δὲ ἡ τῶν σωμάτων φύσις καὶ ἡ τῶν ἄλλων φυτῶν· ἐν πᾶσι γὰρ τούτοις ἔνεστιν εὐσχημοσύνη ἢ ἀσχημοσύνη. καὶ ἡ μὲν ἀσχημοσύνη καὶ ἀρρυθμία καὶ ἀναρμοστία κακολογίας καὶ κακοηθείας ἀδελφά, τὰ δ᾽ ἐναντία τοῦ ἐναντίου, σώφρονός τε καὶ ἀγαθοῦ ἤθους, ἀδελφά τε καὶ μιμήματα. Παντελῶς μὲν οὖν, ἔφη. Ἆρ᾽ οὖν τοῖς ποιηταῖς ἡμῖν μόνον ἐπιστατητέον καὶ προσαναγκαστέον τὴν τοῦ ἀγαθοῦ εἰκόνα ἤθους ἐμποιεῖν τοῖς ποιήμασιν ἢ μὴ παρ᾽ ἡμῖν ποιεῖν, ἢ καὶ τοῖς ἄλλοις δημιουργοῖς ἐπιστατητέον καὶ διακωλυτέον τὸ κακόηθες τοῦτο καὶ ἀκόλαστον καὶ ἀνελεύθερον καὶ ἄσχημον μήτε ἐν εἰκόσι ζῴων μήτε ἐν οἰκοδομήμασι μήτε ἐν ἄλλῳ μηδενὶ δημιουργουμένῳ ἐμποιεῖν, ἢ ὁ μὴ οἷός τε ὢν οὐκ ἐατέος παρ᾽ ἡμῖν δημιουργεῖν, ἵνα μὴ ἐν κακίας εἰκόσι τρεφόμενοι ἡμῖν οἱ φύλακες ὥσπερ ἐν κακῇ βοτάνῃ, πολλὰ ἑκάστης ἡμέρας κατὰ σμικρὸν ἀπὸ πολλῶν δρεπόμενοί τε καὶ νεμόμενοι, ἕν τι συνιστάντες λανθάνωσιν κακὸν μέγα ἐν τῇ αὑτῶν ψυχῇ, ἀλλ᾽ ἐκείνους ζητητέον τοὺς δημιουργοὺς τοὺς εὐφυῶς δυναμένους ἰχνεύειν τὴν τοῦ καλοῦ τε καὶ εὐσχήμονος φύσιν, ἵνα ὥσπερ ἐν ὑγιεινῷ τόπῳ οἰκοῦντες οἱ νέοι ἀπὸ παντὸς ὠφελῶνται ὁπόθεν ἂν αὐτοῖς ἀπὸ τῶν καλῶν ἔργων ἢ πρὸς ὄψιν ἢ πρὸς ἀκοήν τι προσβάλῃ, ὥσπερ αὔρα φέρουσα ἀπὸ χρηστῶν τόπων ὑγίειαν, καὶ εὐθὺς ἐκ παίδων λανθάνῃ εἰς ὁμοιότητά τε καὶ φιλίαν καὶ συμφωνίαν τῷ καλῷ λόγῳ ἄγουσα; (401a–d3) But painting is full of these qualities, as is every craft of that sort, full too weaving, embroidery, architecture, and the production of other furnishings, and also the nature of our bodies and of other living things. In all of these there is gracefulness or gracelessness. And gracelessness, being out of step, and disharmony are kin to bad language and bad character, and the opposite of these things are kin to a prudent and noble character and are imitations (mimēmata) of it. [Glaucon:] Absolutely. So then is it only the poets we must regulate, compelling them to fashion an image of good character in their poems or else not to compose poetry in our midst? Or mustn’t we also regulate other craftsmen, forbidding them from fashioning a vicious character, unbridled, servile, or graceless, either in portraiture or in architecture or any other product of craftsmanship? Or, if someone is unable to comply, shouldn’t we bar him from practicing his craft around us, lest the guardians be raised among images of evil, as in an evil pasture, plucking and grazing much—a little every day from many such—and so without realizing it accumulate a great evil mass in their souls? Shouldn’t we look rather for those craftsmen

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN THE PHILOSOPHICAL TRADITION

83

cleverly capable of tracking down the nature of the beautiful and graceful, so that the young would abide, so to speak, in a healthy environment and be benefitted by everything around them, from which ever direction any of the beautiful works strikes their vision or hearing, like a wind wafting health from pleasant climes, and thus from their earliest youth it would subtly draw them to likeness to, affection for, and harmony with beautiful reason?

Now at last it becomes clear that Socrates’ earlier focus on the effect of dramatic poetry on the performer was an illustrative metonymy of what he sees as a more general problem implicating the visual arts and the effect of images (eikones), not only on those who produce them, but on all impressionable people who come into with them. The destructive influence of these images is not fundamentally due to bad theology or erroneous propositions of any kind, but rather an ingestive accumulation of wickedness transmitted directly from image to viewer/auditor.21 The soul’s rational assent in this process is as unnecessary as in the assimilation of food. Plato will return to the theme of the corrupting effect of dramatic mimesis on the audience in book 10. There he sees the danger looming not just for the young and unusually impressionable, but even for the respectable (τοὺς ἐπιεικεῖς) and the best among us (οἱ . . . βέλτιστοι ἡμῶν, 605c). We all harbour an irrational part in our souls, and though we know it would be shameful to yield to grief in adversity, for example, this irrational part of us would like to do just that. When we see a mimetic performance of unrestrained grief, ‘we enjoy it and, surrendering ourselves, follow it in sympathy’ (χαίρομέν τε καὶ ἐνδόντες ἡμᾶς αὐτοὺς ἑπόμεθα συμπάσχοντες, 605d). The combination of pleasure and pity we feel for those exhibiting behaviour we would, in principle, scorn nourishes the same irrational impulses in us, and ‘what we enjoy must from others to ourselves’ (ἀπολαύειν ἀνάγκη ἀπὸ τῶν ἀλλοτρίων εἰς τὰ οἰκεῖα, 606b). Older scholarship on mimesis in the Republic did not generally distinguish the performer’s identification in book 3 from the audience’s sympathy in book 10, regarding them as two facets of essentially the same mode of response to a dramatic character, but more recent work has usually tended to see these two ages as having different purposes and describing distinct phenomena.22 Plato is interested in the actor’s experience in book 3 and that of the audience in book 10, and the performer’s habituated self-likening to the role is clearly distinguishable from the theatregoer’s lack of self-control resulting from a habit of emotionalism and confusion about what is, in fact, praiseworthy.

21 Plato’s environmental materialism in this age may be situated within a larger cultural context and within his project more generally; see Ford 2002, 221. 22 Examples of the uncritical conflation of actor’s and audience’s response: Adam 1963 (vol. 1), 150; Cross and Woozley 1966, 272; Halliwell 1988, 4. Arguments against the application of ‘mimesis’ to the audience: Sörbom 1966, 122–4; Dyson 1988, 44–5; Halliwell 2002, 80–1 (in revision of his earlier position).

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

84

MIMETIC CONTAGION

While I would agree with the more recent line of scholarship that identification and sympathy are notionally distinguishable, Socrates’ claim in book 10 that what we enjoy in others necessarily becomes our own (oikeia) substantially conflates the effects of audience sympathy with those of the actor’s self-likening. Whether we encounter the morally questionable dramatic role as performers or spectators, we shall find ourselves not only weakened and worsened by the experience, but also ethically assimilated to the character. More importantly, though, the significance of the age I quoted earlier at length (401a–d3) for the connection of mimesis in books 3 and 10 seems usually to have been missed, perhaps because it is not strictly about dramatic mimesis but artistic mimesis in the most general sense.23 Indeed the focus there was, if anything, more on visual art than drama, so scholars keen to understand Plato’s view of poetry may understandably have skipped over a age that seemed less directly relevant.24 But of course drama is a visual art, and, as we saw earlier, Socrates describes his shift in focus in 401a–d3 as a summary generalization of the topic he had been treating before, so the actor/ audience distinction can hardly be as central to Plato’s understanding of this issue as we might make it, and we may well see the assimilating sympathy described in book 10 as another application of the general phenomenon described in 401a–d3. Plato’s purpose in the Republic is not to elaborate a grand unified theory of mimesis, still less is it to isolate and explain the dynamic we have dubbed mimetic contagion, but these ages of books 3 and 10 reveal that among other criticisms of artistic representation (as ontologically fallen, as traditionally erroneous, etc.), one of Plato’s primary concerns about art is its tendency to exceed its apparent boundaries and infect those who come into with it, even without their

Without this age clearly in view, one is likely to conclude that book 3 is only about performers and book 10 is only about the audience. See, for instance, Leszl 2006, 258: ‘The situations envisaged are then profoundly different’. Belfiore, by contrast, defends the unity of mimesis in books 3 and 10, but she excludes visual art from her definition of mimesis, even in the face of 401a–d3: ‘Thus, it is doubtful that Plato is discussing imitation at all at 3.401Aff . . . . no theoretical of imitation in the visual arts, good or bad, is given in Republic 3’ (1984, 134). 24 Plato is more absolute in his dismissal of poetic mimesis from the ideal state in book 10 than he was in book 3, and scholars have often defended him from the charge of inconsistency by arguing that he only means to banish bad, i.e. morally deleterious, mimesis. See, for example, Verdenius 1949, 21–3. Nehamas 1982, by contrast, takes Socrates at his word in book 10 and understands him to be banishing all poetic mimesis; the earlier more liberal expressions ultimately only apply to the visual arts. Nehamas notices that Plato’s most general model of mimesis seems to be painting, and he argues that pre-Platonic attestations of mimesis all concern aural phenomena, so this grounding of mimesis in painting must be a Platonic innovation, designed to make poetry (and all other art forms) share in painting’s inability to tell any deeper truth about things than their surface appearance. By itself, though, this limitation only makes painting trivial and harmless, whereas poetry is so compelling in its deceptiveness that it must be forbidden. Nehamas’ observations about the centrality of painting for Plato’s view of mimesis have interesting consequences for my thesis as well. However 401a–d3 shows that Plato does not think painting and other visual arts to be quite as harmless as Nehamas thinks they are. 23

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN THE PHILOSOPHICAL TRADITION

85

knowledge or assent. In these ages Socrates and his interlocutors draw the logical conclusion that mimesis should be banned from the ideal state or else regulated to an unrecognizably hygienic form. It would be rash, however, to take these ages in isolation from the rest of the Platonic corpus and label him, in any straightforward sense, a puritanical enemy of the arts. There are several other places where Plato speaks approvingly of mimesis, especially in his later works, so perhaps his views became less stringently hostile over time.25 Moreover, the dialogues themselves are essentially dramatic in form and often present characters we should not wish to emulate, so it is tempting to conclude that the same Plato who wrote tragedies in his youth lived on in the mature thinker and that he adapted these literary genres into his work as a way of situating his philosophy as a comparable social practice.26 Indeed, perhaps we were never meant to read the harsh judgments of mimesis in Republic 3 and 10 as literal proscriptions, but rather as logical conclusions in a thought experiment or agonistic bids in Plato’s ironic contest with a nexus of institutions that is, at least in its effects and its claim on our attention, surprisingly similar to philosophy.27 However literally we take these public policy recommendations and however permanent a view we believe them to represent in Plato’s thought, they are unsettling for most of us, not least because they are so difficult, even after two and a half millennia, to dismiss.28 Aristotle stands as the first inheritor of this critique, and his formulation of mimesis in the Poetics may be understood as an attempt to rehabilitate art from the vexed position in which the Republic had left it. He concedes that the mimetic artwork, at its most successful, does have a violent effect upon the emotions of the viewer, but rather than inspiring direct imitation, the twin forces of pity and fear are supposed to result in an experience of purgation (katharsis, 1449b27–8). Pity and fear are, in the first instance, feelings that arise in the presence of real calamities and dangers, so the audience’s response to mimesis mimics what their emotional response would be to a real encounter with the represented events. In itting that mimesis can call forth a ionate response in the audience and in implying that this response is predicated on their willingness to credit a kind of ‘as if ’ reality to the representation, Aristotle seems to come within reach of some of the issues that had interested Plato in Republic 3 and 10. But this similarity in topic is ultimately only apparent. Aristotle provides a positive for the beneficial role of the audience’s response to mimesis, as conducive to their 26 Halliwell 2002, 65–71. Nightingale 2000. ‘Next to philosophy, only art and specifically mimetic poetry have the power to create happenings, cultural and psychological, that subvert political, social, and ethical laws’ (Naddaff 2002, 4). 28 Few indeed would follow Plato all the way to the bitter end of his proscriptive conclusions, but one need not be a puritanical enemy of the open society to it that art can do harm, sometimes by remaking us in its pernicious image. For a refreshingly candid attempt to apply the Republic’s critique of mimesis to television (with apparent sympathy both to Plato and to the mass media), see Nehamas 1988. 25 27

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

86

MIMETIC CONTAGION

health and knowledge, and he is not obliged to confront and explain away the dangers Plato saw as inherent in mimesis, in large part because he means something subtly different by the word. As I said earlier, mimēsis covers a complex cluster of concepts, which may be broadly divided between versions of world-reflecting (mimesis1) and world-creating (mimesis2). In the ages of the Republic we have examined, Plato is using the word in the sense of mimesis1, but in the Poetics Aristotle has shifted almost entirely to mimesis2.29 Aristotle does not need to deny the power of mimesis to inspire imitation in the viewer, because he is ignoring the imitative aspect of mimesis altogether, or rather refocusing on the alternate sense of the word, as the creation of an autonomous fictional world. However far from view they may be in the Poetics, the possible benefits and risks of mimetic contagion are not lost on Aristotle more generally. As we have seen (pp. 75–6) in the Politics he explicitly addresses the ethically contagious effects of painting and sculpture on young viewers. In this age he might virtually be channelling the Socrates of Republic 3, but in the Poetics he never envisions the audience imitating the mimesis or even describes mimesis as having an imitative relationship to a model. The problem of mimetic contagion is not so much dispelled as side-stepped. Though the Poetics cannot be said to answer the Republic’s accusations against mimesis, Aristotle does succeed in reclaiming a positive space for (some version of) mimesis and thus justifying the ways of art to philosophy. Posterity has been grateful. But these ages in the Republic also mark a nadir in Plato’s assessment of mimesis, and whether explicitly or implicitly almost any other Platonic text using words from the mim- group is more affirming. Most of these do not bear in any direct way on the issue of mimetic contagion, but there is one worth examining for its relevance to our topic.30 The Timaeus is one of Plato’s last dialogues, probably written some twenty years after the Republic, but it immediately positions itself as a continuation of ‘yesterday’s speech’ about the ideal state (17c). Readers from Proclus to the twentieth century have often assumed that the dramatic date of the Timaeus was meant to be the day after that of the Republic, but of course Socrates, in fiction as in life, may have had many such conversations, and though the topics are similar, the Timaeus has different interlocutors than the Republic, so perhaps another occasion is meant.31 It has also been suggested that Plato intentionally made the dramatic date for these conversations ambiguous, so the dialogues’ 29 This shift allows Aristotle to treat mimesis as a cognitive process and thus, in the first instance, a means of knowing, rather than an adaptation in conformity with some external model. See Golden 1992, 63–103; Halliwell 2002, 151–233. 30 Conversely the Ion is plainly interested in mimetic contagion, but it avoids the word mimesis, so I do not treat it in this section. See p. 179. 31 Cornford 1937, 4–5.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN THE PHILOSOPHICAL TRADITION

87

setting is as timeless as their philosophy.32 As various recent scholars have noted, however, even if the dramatic dates of these dialogues are not successive days, it seems clear that the reader of the Timaeus is meant to have the Republic in mind.33 Socrates would like to extend the discussion of the ideal state to a narrative depiction of its conflict with some competing state, thus revealing the merits of the ideal at their full advantage by showing them in action, but he doubts his own ability to render the ideal dynamic, since the world he grew up in is so far from political perfection. Not that the poets would fare any better: ‘the mimetic clan (μιμητικόν ἔθνος, 19d) will most readily and best imitate (μιμήσεται) the things among which they were raised, but things outside their upbringing are hard for anyone to imitate (μιμεῖσθαι) in actions and even harder in words’. In this threefold invocation of the mim- root in one sentence, mimesis1 and mimesis2 are neatly blended, and Socrates hints that the poets are as useless to the rendering of the good city as they are to rendering the city good. Fortunately, Critias comes to the rescue with the remarkable story of the lost city of Atlantis and her contest with the ideal city, which turns out to be none other than Athens in the once-upon-a-time past. This story set in the distant antiquity of the human race requires its own cosmogony, so Critias soon interrupts himself and invites Timaeus to describe the constitution of the physical universe beginning with its origin and ending with the generation of mankind (27a–b). This is no small task, obviously, and Timaeus’ discourse runs to 65 Stephanus pages, but Critias has promised that he will continue with the Atlantis story once Timaeus has given the cosmic backdrop, and he does so in the immediately following dialogue, Critias. Thus the Timaeus-Critias consists principally of Timaeus’ long description of the universe and Critias’ shorter-frame narrative about ancient Athens and Atlantis, but Timaeus’ speech is not only set within Critias’ story, it is told at Critias’ request and follows the programme he gives it, so the whole complex may be characterized as Critias’ response to Socrates’ complaint about the inability of the ‘mimetic clan’ to represent the ideal polis in action. It is therefore no wonder that one of the things this response must do is give a new contour to mimesis and the phenomenon of mimetic contagion. The central conceit of Timaeus’ is that the world of our experience, the world of becoming and change, is a mimesis of the eternal, changeless world of the Forms. It is not a case of incidental likeness, where the phenomenal world merely happens to be similar to the ideal world, neither is the phenomenal world unconsciously propagated out of the ideal, rather the sensible world is an artistic copy, intentionally created by the Demiurge as he gazed on the eternal

32

Moors 1987.

33

Zeyl 2000, xxvii; Johansen 2004, 178.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

88

MIMETIC CONTAGION

(πρὸς τὸ ἀΐδιον ἔβλεπεν, 29a). The model (paradeigma) is not only eternal and real, but intelligible, uniform, and alive, so the Demiurge fashioned the sensible world into its likeness, a rational, spherical animal, an artwork/object of joy (ἄγαλμα, 37c). Thus he sent the stars endlessly wheeling around their celestial courses ‘in order that they might be as similar as possible to that perfect and rational animal after the imitation (μίμησιν, 39e) of its eternal nature’. Mimesis has a different colour here than Socrates had given it in the Republic and in Timaeus 19d, not because it means something fundamentally different, but because the results are so much happier. The ideal world, for all its notional invisibility, is described as the object of the Demiurge’s gaze, and his response is to imitate what he sees. But because what he contemplates is good and he is a good imitator, the mimesis is not destructive for him or for other viewers of the eikon (29b) he fashions. The universe of our experience is a ‘copy of the model’ (μίμημα . . . παραδείγματος, 48e), but if the heavens are therefore ‘secondary’ we can hardly complain about their ontological derivativeness. After all, as Timaeus says in the last sentence of the dialogue, this eikon of the intelligible is itself a perceptible god, greatest, best, fairest, most perfect, and only begotten (92c). Implicit in these descriptions of the created world as an image lies the possibility that the Demiurge’s mimesis of the ideal is not a dead end, but will itself inspire further imitation as a visible model, that mimetic contagion, in other words, will prove to be contagious. This is precisely the direction in which Timaeus develops mimesis as he continues his cosmogony. After fashioning the heavenly bodies and the other created but immortal gods familiar from Hesiod, the Demiurge ens these deities to follow his creative example. ‘Yourselves, according to nature, turn to the fashioning (δημιουργίαν, 41c) of animals, imitating (μιμούμενοι) my power in creating you.’ The children of the Demiurge respond with due obedience and fashion terrestrial life, ‘imitating their author’ (μιμούμενοι τὸν σφέτερον δημιουργόν, 42e).34 But they imitate not only his power or himself, but also the perfect celestial world he created. For example, they make man’s head a sphere ‘in imitation of the shape of the universe’ (τὸ τοῦ παντὸς σχῆμα ἀπομιμησάμενοι, 44d) and they ensconce this seat of higher faculties literally above the rest of the body. The structure of the human body thus mirrors the structure of the cosmos, even imitating its directionality in the fall from the spherical beauty of the head, with its organs of speech and sensation, to the upper torso, with its motions in breath and appended arms and hands, speechless but still capable of signification and reasoned control, to the lower torso and its grosser functions beyond the reach of rational volition. Even here, though, images from the intellect can penetrate, and the liver is designed specifically to mirror the mind’s state and on to the bowels a limited

34

See also 69c.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN THE PHILOSOPHICAL TRADITION

89

kind of participation in the higher life of the soul (71b–c). Our bodies are a mimesis of the cosmos, which means that we can resonate with its design even apart from the operation of the intellect. For example, sensation can, by definition, only give access to the sensible, but when low and high sounds combine according to geometrical principles of motion they generate delight ‘because of the imitation (μίμησιν, 80b) of the divine harmony’. So too the coursing of blood in the body recapitulates a more general structure, and the blood ‘is compelled to imitate (μιμεῖσθαι, 81b) the circulation of the universe’. All is not compulsion, however, and inasmuch as our thoughts and our behaviour may affect the degree to which our bodies and souls are mimetic of the cosmic order, we are under a corresponding moral duty to maximize that conformity. Precisely because we have a choice in how much attention we devote to the grand structure of the universe we should be as alert as we can to such matters, using our eyes to study the heavens and our minds to understand their order, so that ‘imitating (μιμούμενοι, 47c) the unvarying [cycles in heaven] we may stabilize the varying ones in ourselves’. We must take care of our diverse parts in imitation (ἀπομιμούμενον, 88d) of the form of the universe, where wet and dry, motion and stasis, are never allowed to get out of balance. The Demiurge gazed on the ideal world and made this world in imitation, and he ened the deathless created gods to imitate him in making us. They did so, copying not only his creative pattern but the pattern he created. Finally we are ened to take note of the order in our bodies and the cosmos and to live our lives in imitation of that order. Timaeus has described a descending chain of imitative order, each level looking up to the previous level for a paradigm of its own activity. It is mimetic contagion all the way down. One of the great interpretive problems for the Timaeus-Critias is understanding exactly how Timaeus’ speech is related to the Atlantis story.35 This is a difficult issue, significantly complicated by the fact that the Critias breaks off part way through the narration, so a comprehensive answer to this question is beyond our 35 A favourite interpretation of the story sees Atlantis as a figure for the Persian empire, which would make the victorious Athens in the fable a type for the Athens that won at Marathon. The tragic irony, of course, is the double signification: the aggressive imperialist regime would all too soon be Athens, which then must play the role of Atlantis in the Peloponnesian War. On this reading the story becomes a parable about the inevitable consequences of ill-organized, vice-ridden polities confronting chaste, disciplined neighbours. Vidal-Naquet 1986 is a standard formulation of this interpretation, and it has been refined and extended by others; see, for example, Morgan 1998. However, it is never exactly clear how this reading is supposed to connect to the cosmology of the Timaeus, if at all. Johansen 2004, 7–23 argues that both the cosmos of the Timaeus and the political history of the Critias are structured according to narratives in which there is a profound order in the world and all pleonexia is unnatural and must eventually be frustrated. This sounds right, but perhaps it does not get us very far. One would hardly expect Plato to tell a story or elaborate a cosmology in which the world turned out to be radically disordered or to reward vice. See also Pradeau 1997; Broadie 2012, 115–72.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

90

MIMETIC CONTAGION

scope. However, in tracing out the role of mimesis in Timaeus’ we may have stumbled on a clue. As we have seen, Timaeus’ speech is an extended interruption of the Atlantis story, told at Critias’ prompting and following his specifications. Its relationship to its surrounding frame is thus oddly similar to the terrestrial world’s relationship to the surrounding celestial cause/model. Critias begins his with an intriguing archaeology of the story itself: the Egyptians told it to Solon, who told it to Critias’ grandfather (also called Critias), who in turn ed it on to him. The quasi-mythical status of the Egyptians as timelessly ancient wizards is clear from his description, as is Solon’s role as wandering wise man, who must travel far from home but returns with an enriched understanding of the world and, particularly, of his home. The Atlantis story thus descends through time but also in a chain of narrators that carries it from the legendary disionate sages of Egypt to the semi-legendary lawgiver of Athens to the Athenian world of our acquaintance. It may be overly schematic to insist on a rigid corresponsion of levels (Egyptians = Demiurge; Solon = created gods; Critias = mankind), and it would probably be going too far to read the shared name of Critias as a marker of intergenerational mimesis, but it can hardly be missed that the Atlantis story follows an arc not at all unlike the paradigm’s descent from ideal to celestial to terrestrial, stamping itself on each stage successively. More striking still is Critias’ description of the story’s place in himself: Truly marvellous, as they say, how the things we learn as children hold our memory. I’m not sure I could call back to mind the things I heard yesterday, but of this story, which I heard so long ago, I’d be shocked if any part escaped me. I heard it with such pleasure and amusement, and the old man was eager to teach it to me, since I asked for it repeatedly, so that the story became fixed in me like the encaustic designs of an indelible painting (ὥστε οἷον ἐγκαύματα ἀνεκπλύτου γραφῆς ἔμμονά μοι γέγονε, 26c).

Timaeus’ speech seamlessly extends this motif of self-propagating mimesis from Critias’ introductory description of the Atlantis story to a theme undergirding the cosmos. When Critias takes up the sceptre again to continue the Atlantis story, he characterizes both Timaeus’ and his own as a necessary ‘imitation and copy’ (μίμησιν . . . καὶ ἀπεικασίαν, 107b), which, like paintings, must be evaluated for their faithfulness to the truth.36 Critias recognizes that mimesis may indeed fail to render the contour of reality, but he nevertheless lays claim to the word in reference to these speeches, a sanguine turn that seems fully justified after all Timaeus has said about the cosmic role of mimesis. Socrates’ dismissive attitude towards mimesis in Timaeus 19d and his anxiety about mimetic contagion in the Republic are thus answered by Critias and Timaeus together. Mimesis is not only

36 On the Atlantis story as an experiment in fiction (as pointedly distinct from falsehood), see Gill 1977 and 1979.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN THE PHILOSOPHICAL TRADITION

91

central to the operation of the world as we know it, it is so precisely because of mimetic contagion. Plato’s critique of mimesis in the Republic posed a serious question about how art affects us, a question the subsequent philosophical tradition has laboured to answer with complete satisfaction. Aristotle’s answer in the Poetics has doubtless been the most successful among readers and critics of practically all periods, but it works by taking advantage of the ambiguity in the word mimēsis. By focusing on mimesis2 Aristotle manages to highlight some of art’s very real psychological benefits, while quietly ignoring the problem of mimetic contagion, which only seems apparent when considering mimesis1. Plato’s own late return to this topic in the Timaeus-Critias is, in a sense, a more effective response to the Republic’s view of mimesis, because it squarely addresses the issue of mimetic contagion, but it presumes an entire cosmological system to do so, and if we cannot subscribe to his doctrine of the structure of the physical world, we may find the corresponding rehabilitation of mimesis somewhat precarious. The Timaeus has commanded its share of sympathetic readers, especially in late antiquity and the Middle Ages when its influence far overshadowed that of the Republic, and its view of the phenomenal world as a mimesis of the ideal world is never entirely lost from view in the later history of aesthetics.37 But wherever the Republic has been read it has disturbed complacent assumptions about the benign ivity of art, and it has often provided the conceptual tools for writers interested in exploring the unrulier side of mimesis. For example, Oscar Wilde’s 1891 essay ‘The Decay of Lying’ is written in the form of a Platonic dialogue and it makes repeated reference to Plato, and to the Republic more specifically, as a venerable corroboration of its thesis. Wilde’s Vivian presents an extended encomium of mendacity in art, partly on the grounds that creative lying provides relief from the grim monotony of reality, and partly because the basic principle that art and life should be similar will never ultimately be violated, for mimesis is a two-way street: ‘Life imitates Art far more than Art imitates Life.’38 He gives many examples of human behaviour brought into conformity with compelling artworks, sometimes willingly, sometimes not, and even posits that our natural environment, or at least our experience of it, is radically conditioned by its prior construction in art. Though Wilde’s valuation of mimesis is obviously a playfully ironic inversion of Plato’s, his basic understanding of the dynamics of mimetic contagion is similar.39 Vivian takes the Republic as a principal source for mendacity’s connection to art and for its educational utility,

37 38 39

For all aspects of the reception of the Timaeus in this period, see Reydams-Schils 2003. Wilde 1968, 186. On Wilde’s reuse and abuse of Plato in this essay, see Danson 1997, 58–9; Comfort 2008, 57–62.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

92

MIMETIC CONTAGION

and he even affects to regard Plato as a kind of personal alternative to truth.40 Most interestingly, in a age without overt reference to Plato, the image of the cave from Republic 7 becomes a figure for art’s capacity to transcend and transform the world: Remote from reality, and with her eyes turned away from the shadows of the cave, Art reveals her own perfection, and the wondering crowd that watches the opening of the marvellous, many-petalled rose fancies that it is its own history that is being told to it, its own spirit that is finding expression in a new form. But it is not so. The highest art rejects the burden of the human spirit, and gains more from a new medium or a fresh material than she does from any enthusiasm for art, or from any lofty ion, or from any great awakening of the human consciousness. She develops purely on her own lines. She is not symbolic of any age. It is the ages that are her symbols. (189)

Wilde inverts the cave’s metaphysical orientation. ‘Reality’ is now located not in the world outside, but in ‘the shadows of the cave’, and the role of the lone escapee is now played by Art herself. There is a ‘wondering crowd’ of spectators, as in Plato’s version, and they are just as confused as ever about the ontological status of what they see, but the object of their gaze is now Art’s perfect unfolding, a kind of striptease that lulls them into thinking that they are witnessing an imitation of their lives, when in fact their lives are only an imitation of Art’s performance.41 Plato never uses the word mimesis (or any words from the mim- group) in his Allegory of the Cave (514a–20a), but his description of the subterranean shadow theatre is as full of artistic image-making and theatrical performance as any part of the Republic. The images carried by the puppet masters are themselves wrought (εἰργασμένα, 515a) copies of other forms, even human forms (ἀνδριάντας, 514c), and their function is precisely to generate copies of themselves, shadow images on the back wall of the cave. Again, the language of mimesis is not used in this age, but the connection to the problems of mimesis in Republic 3 and 10 is clear enough, and the philosophical valuation of this imitative image-making is as 40 ‘Lying and poetry are arts—arts, as Plato saw, not unconnected with each other—and they require the most careful study, the most disinterested devotion’ (Wilde 1968, 168–9). ‘Lying for the sake of improvement of the young . . . still lingers amongst us, and its advantages are so irably set forth in the early books of Plato’s Republic that it is unnecessary to dwell upon them here’ (193). ‘Just as those who do not love Plato more than Truth cannot beyond the threshold of the Academe, so those who do not love Beauty more than Truth never know the inmost shrine of Art’ (194). 41 Wilde’s allusion to the Republic’s cave has not, to my knowledge, been noted in the scholarship on this text. No doubt this is mostly due to the fact that he does not mention Plato in this age, but it may also stem from readers’ assumption that the Allegory of the Cave would not have been particularly interesting to Wilde, because it is about epistemology and metaphysics, rather than art and the problem of mimesis. But this may not have been how Wilde read Plato’s cave. As he said in De Profundis, his famous letter written to Lord Alfred Douglas in 1897 while he himself was a prisoner in the cave of Reading Gaol, everything Plato ever said about metaphysics could ‘be transferred immediately into the sphere of Art, and there find its complete fulfillment’ (Wilde 2000, 741).

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN THE PHILOSOPHICAL TRADITION

93

negative as elsewhere in the Republic. Oscar Wilde was neither the first nor the last reader to mine Plato’s cave for an image apposite to the philosophical problems of mimesis, only to invert its value and use it to rehabilitate artistic imitation. Luce Irigaray has also found Plato’s cave a fertile symbol for the possibilities of strategic essentialism represented by mimesis. In her long essay ‘Plato’s Hystera’ (1974) she draws attention to the morphological similarity between Plato’s cave and the womb, both of which function as a matrix of mimetic production from which one must struggle to escape into the real world. But the cave is not only womb-like in its inscription of the copying faculty in a world of darkness and mystery, it is also ‘latter’ (hysteron), ontologically secondary to the real world of the Forms and truth.42 Like the womb (hystera) the cave generates imitations logically dependent on some masculinized originating agency of prior and superior reality. Each human life, however, begins with the womb (hystera) which is thus temporally if not ontologically prior (proteron). Similarly, for the denizens of the cave facing the back wall all their lives it is the real world that is behind (hysteron) and the shadowy images of images that are in front (proteron).43 This instability of hysteron and proteron in the space and time of the womb/cave expands into an instability of copy and original as the escapee moves from inside to outside and sees topographical features that can only remind him of the familiar furniture of the cave.44 Not only do the puppet masters turn out to have produced copies of individual items from the outside world, the whole architecture of the cave is designed to mimic the shape of that world, where vision, for example, is only possible because of a great light (fire inside the cave, sun outside).45 In both places masculine desire seems designed to induce a turning back towards what is physically behind, whether it be the yearning of the foetus or the philosopher to make his escape from the feminine cave and be born, or the nostalgic yearning of the philosopher to return to the entombed state of illusion or the lover’s desire to return to the womb in the oblivion of sex.46 Thus in the vertiginous spinning along the axis of hysteron and proteron there is a final confounding of the categories of original and copy, and the cave’s mimetic property turns out to have effected a kind of performative resistance to the tyranny of the masculine ideal, ever-reinscribing its features on a reality that ironically may already have been radically mimetic.47 Irigaray’s reading of Plato’s cave is perhaps the boldest reappropriation of this parable in the long philosophical and critical tradition, and if it is, in some sense necessarily, a hijacking of Plato’s myth for anti-Platonic purposes, it is also one that resonates deeply with elements genuinely latent in Plato’s . Her foregrounding of mimesis as the unstated but peculiar activity of the cave and her alertness to the unruly instability of mimetic contagion are profoundly Platonic. 42 45

Irigaray 1985, 244. Irigaray 1985, 294.

43 46

Irigaray 1985, 247. Irigaray 1985, 341.

44 47

Irigaray 1985, 286. Irigaray 1985, 363.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

94

MIMETIC CONTAGION

The aspect of mimesis that she elucidates more effectively than anyone else in this tradition is its radical unrepresentability and the aptness of the cave for symbolizing something that cannot in itself be shown.48 Indeed, an actor may copy a drunk and a painter may copy a flower, but how do either of them represent mimesis itself? The content of any given instance of mimesis is the representation of some object, but the form of that mimesis is representation itself, and how can representation itself be represented? The cave, on Irigaray’s reading, does just this, by figuring the hidden, blind matrix where likenesses are born and spawn new likenesses, only to serve as the origin of every possible vision of reality. Irigaray’s interest in Plato’s cave is thus neatly parallel to the philosophical side of our interest in mimetic contagion. As Aristotle said, man is mimetic by nature (1448b5–6), so it is not surprising when people imitate what they see in the world. But when what they see and imitate is itself a work of mimetic art, their behaviour serves as an icon not only of the object they see, but of the unrepresentable mode of existence of that object. Mimesis, if it can be represented at all, can only be represented transitively, in the contagious age from one imitation to the next.

‘The hystera, faceless, unseen, will never be presented, represented as such. But the representational scheme and sketch for the hystera—which can never be fulfilled—sub-tends, englobes, encircles, connotes, overdetermines every sight, every sighting, face, feature, figure, form, presentification, presence. Blindly’ (Irigaray 1985, 245). 48

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

FOUR

Mimetic Contagion in Terence’s Rome

Graecia capta ferum victorem cepit et artes intulit agresti Latio. Horace, Ep. 2.1.156–7

The arc of Chapters 2 and 3 has spanned ritual and other popular uses of images and the development of a philosophical discourse of the mimetically persuasive effect of art, particularly on the young. Some of this material will have been more directly familiar to Terence and his audience than other material, but even the more abstract philosophical reaction to mimetic contagion is rooted in a concrete concern with practical and universally pressing questions of pedagogy. At Rome, the educational role of images and their power as ethical exempla for the young took an institutional form more explicit than anything we know from Greek culture. Polybius, who was brought to Rome as a hostage in 168 BCE and might well have been present at the original production of the Eunuch, describes the singular role played by the imagines in Roman funerals of that time (6.53–54).1 After the burial, an image (εἰκών) of the deceased, a mask fashioned to meticulous likeness (εἰς ὁμοιότητα διαφερόντως ἐξειργασμένον), is displayed in the most conspicuous place in his house. Polybius indicates the ambiguous role of these likenesses as both artistic images and masks, noting that they are carefully displayed at public sacrifices, but that in subsequent family funerals they are actually worn by men who are as like the originals as possible (ὡς ὁμοιοτάτοις) in body type and who will enhance their performance by dressing the parts their subjects played in life.2 Throughout his description of the funeral, Polybius repeatedly emphasizes the importance of seeing. Twice he says that this kind of production is only undertaken for the distinguished or, more literally, ‘visible’ (ἐπιφανής, 6.53.1 and 6), whose noticeability in life only continues after death, and we can hardly miss the connection between the body’s conspicuous placement (ἐναργής) on the

Flower 1996, 91–127 has the best treatment of this age and its context within the broader evidence of Roman ancestor masks. 2 For the relationship between this semi-dramatic, masked performance practice and Roman theater, see Flower 1996, 114–15. 1

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

96

MIMETIC CONTAGION

Rostra and the mask’s prominent location (ἐπιφανέστατον) in the house. Not only does this ritual create a continuity of ive visibility from life to death and corpse to artefact, it also establishes a continuity of active spectation from one generation to the next. For the funeral eulogy is delivered ‘under the gaze’ (ὑπὸ τὴν ὄψιν) of all in attendance, and this will necessarily include, as the following description makes clear, the gaze from the hollow eyes of the ancestors’ masks. The old dead may be spectators of the new dead, but, according to Polybius, the parade of the old dead before the young is the most important part of the ceremony. The whole next chapter of book 6 is devoted to advertising the famous bravery of the Roman people and showing the relevance of their funeral rite for the Romans’ national character. Polybius understands the chief point of this manifestly theatrical spectacle (θέαμα) of dead men walking to the funerals of their descendants to be the effect of these images on the young. Young men hear the famous exploits of the men whose likenesses they are looking at, and word and image come together to inspire them to become ethical likenesses of their forebears. Polybius hints at the many examples he could cite of Roman piety stirred up by such means, stories of battles decided by single combat, heroic self-sacrifice, and fathers condemning their own sons to death for the good of the Republic, and we will of course recognize these types from the early books of Livy and other s of Rome’s legendary history. The example Polybius finally singles out as representative is the story of Horatius Cocles, an especially appropriate instance for several reasons. His deed is introduced as an example of the young being inspired to imitate their ancestors, so presumably we are to understand that such rituals as the Roman funeral have impelled Horatius to this act of bravery, but, within the story, he is explicitly inspired by the expectation of his own future renown (τὴν ἐσομένην μετὰ ταῦτα περὶ αὐτὸν εὔκλειαν, 6.55.3).3 In a way similar to the Roman funeral, then, where young men are spectators but are therefore expected to become ‘prominent’ (ἐπιφανής), and the recently deceased is the spectated but will subsequently be one of the old dead and play both roles at once, Horatius too finds himself in an essential continuity of inspirational watching and being watched. After taking up Horatius, Polybius does not return to the topic of the family imagines, but Horatius’ relevance for these inspirational modes of figuration becomes clear if we bear in mind Livy’s testimony (2.10.12) that there was an honorary statue of Horatius in the comitium, just as Cloelia, whom he constructs as an explicit imitator of Horatius (2.13.6–8), was honoured with a statue on the Sacred Way. If the statue of Horatius Cocles is supposed to be a civic-scale version of a family imago, inspiring the Roman people to imitative virtue, the reason for focusing on him in connection with the funeral rite becomes all

3

The richest treatment of Horatius Cocles and the role of vision in Roman exemplarity is Roller 2004.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN TERENCE’S ROME

97

the more clear. As usual, it is probably impossible to determine whether Polybius’ analysis of exemplarity in Roman culture is his own or adapted from his native sources, but it seems extremely likely that in this case Polybius is handing on his informants’ understanding of their funeral ritual.4 This would for his immediate turn to Rome’s early legends for exempla, as opposed to more recent history with which he was personally familiar, and it may also explain his failure to mention the public statue of Horatius, if this connection with the funeral imagines was obvious to his informants but not to him. As to Horatius Cocles’ role as paradigm of exemplarity, Matthew Roller has counted over thirty references to him in extant ancient literature.5 Sometimes the story is told afresh, but in other cases the name seems to be enough, and often he is invoked explicitly as an ethical standard against which contemporary figures are judged. Given this frequency and focus of deployment, it seems improbable that Polybius could be the single-handed author of this whole tradition, and it is much more likely that he is simply transmitting his informants’ commonplace understanding of the cultural centrality of these images. The significance of this conjecture is the implication that at least some portion of the original audience of the Eunuch would already have considered representational art to have an ethically contagious effect, particularly on the young, and that its operation on them was already being described in theatrical . This formulation of the popular Roman discourse of exemplarity is emphatically positive, as the focus here has been only on the effect of good images on the young, but just as Aristotle and Xenophon’s Socrates could point to both beneficial and destructive forms of mimetic contagion, the Romans had available cultural models for the detrimental effect of images, particularly in their understanding of the history of their own use of representations of the gods. Plutarch relates that Numa, as a good Pythagorean, forbade the use of painted or sculptural images of the gods and that for 170 years the Romans flourished in religious architecture but never compromised their piety by ‘likening the higher to the lower’ with statuary (Numa 8.7–8).6 The notion that Roman religion was ever aniconic is as fanciful as Numa’s very existence, but this story is traceable as far back as Varro.7 When the representational austerity of early Roman religion is invoked in these texts, it is always immediately coordinated to an indictment of subsequent corruption of piety after the introduction of images, so that the historicity of Roman religious aniconism is clearly subordinated to its rhetorical importance as a 4

The methodological problems inherent in trying to ascertain Polybius’ sources become even more intense for book 6. For the source issues, along with other problems unique to book 6, see Walbank 1972, 130–56. There is certainly no evidence for an earlier literary version of the Horatius Cocles tale, see Walbank 1957, 740–1. 5 Roller 2004, 2. 6 For the cultural background to the tradition of Numa’s Pythagoreanism, see Humm 2005, 541–54. 7 For Varro, see Aug. De civ. 4.31; see also Tert. Apol. 25.12–13.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

98

MIMETIC CONTAGION

moralistic topos. This myth has two inevitable corollaries: first, whenever it is expedient to do so, artistic representations, especially of the gods, can be construed as culturally imported and alien to ancestral Roman practice, and second, however pious the votary’s intentions, these images may be mysteriously harmful to public character. Closely related to this moralistic tradition is the conventional assertion that Rome’s expanding empire eventually brought in significant quantities of technically refined Greek art, which effected the corruption of the Roman people by spoiling their rugged simplicity and introducing Hellenic luxury. Here Marcellus’ sack of Syracuse in 211 BCE, followed by his elaborate ovatio, is often treated as the cardinal event. ‘For before that time’, says Plutarch, ‘Rome neither had nor knew anything of refinement or elegance’, but bristling with her barbaric arms and bloody trophies of war, she was ‘no sight for cowardly and luxurious spectators’ (οὐδὲ δειλῶν ἦν θέαμα καὶ τρυφώντων θεατῶν, Marc. 21.2). There is a nice irony here; every city gets the ‘spectators’ she deserves. Early Rome was not worth looking at (θέαμα), but then there was no one inclined to the idleness of looking at her (τρυφώντων θεατῶν), since this was before she became famous as the home of spectacles and mother of spectators! She was austere and sober and so were her citizens, until ‘Marcellus adorned the city with spectacles that had pleasure and Greek charm and persuasiveness’ (Μάρκελλος ἡδονὴν ἐχούσαις καὶ χάριν Ἑλληνικὴν καὶ πιθανότητα διαποικίλας ὄψεσι τὴν πόλιν, Marc. 21.3). The artworks’ first two subversions, their ‘pleasure’ and their ‘Greek charm’, will be familiar to anyone with a broader acquaintance with Roman moralizing on luxuria, but the third affliction, their ‘persuasiveness’, will be harder to understand apart from the discourse of mimetic contagion.8 Like the sculpture of Pythagoras of Rhegium, these artworks from Sicily are persuasive in two senses at once; they are deceptively lifelike, and precisely because they are convincing in this sense, they are also persuasive in that they somehow remake the Romans in their own image, as luxurious, idle Greeks. ‘He filled the people with idleness and chatter about art and artists, talking cleverly and fiddling with this a great part of the day’ (σχολῆς ἐνέπλησε καὶ λαλιᾶς περὶ τεχνῶν καὶ τεχνιτῶν, ἀστεϊζόμενον καὶ διατρίβοντα πρὸς τούτῳ πολὺ μέρος τῆς ἡμέρας, Marc. 21.5). In Livy the problem is presented somewhat differently: ‘This was the first beginning of iring works of the Greek arts and of this general license for despoiling everything sacred and profane, a freedom which eventually turned on the Roman gods, first of all on that very temple which was so lavishly adorned by Marcellus’ (inde primum initium mirandi Graecarum artium opera licentiaeque huius sacra profanaque omnia vulgo spoliandi factum est, quae postremo in 8 The best general treatment of Roman luxuria, seen through the lens of sumptuary legislation, is still Sauerwein 1970.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN TERENCE’S ROME

99

Romanos deos, templum id ipsum primum quod a Marcello eximie ornatum est, vertit. 25.40.2). The model that the Roman people are seduced into imitating this time is not associated with the images themselves, but with the whole context of their impious procurement. This shift in attention from the artworks to the conquering general is unique to this text, and it seems plain that Livy’s point in this turn is to set up the irony of the sack of Marcellus’ temple. The Romans have proven to be good students, imitating not only their teacher’s love of art, but his total disregard for provenance. Livy yokes the art’s captivating wondrousness (initium mirandi) to the public slide into Marcellus’ licentia without specifying the images’ direct dynamism, but the art’s implicit role in the Romans’ degradation is probably already clear. Later, however, when Cato refers to Marcellus’ haul, the images’ aggression is more explicit (34.4.3–5): haec ego, quo melior laetiorque in dies fortuna rei publicae est, quo magis imperium crescit—et iam in Graeciam Asiamque transcendimus omnibus libidinum inlecebris repletas et regias etiam adtrectamus gazas—, eo plus horreo, ne illae magis res nos ceperint quam nos illas. infesta, mihi credite, signa ab Syracusis inlata sunt huic urbi. iam nimis multos audio Corinthi et Athenarum ornamenta laudantes mirantesque et antefixa fictilia deorum Romanorum ridentes. ego hos malo propitios deos et ita spero futuros, si in suis manere sedibus patiemur. The better and brighter the fortune of the state becomes every day, the more the empire grows—and now we are crossing Greece and Asia, which are full of all the enticements of desires, and we are even laying our hands on royal treasures—, well, I shudder all the more, for fear those things will gain more control over us than we over them. Believe you me, those statues brought in from Syracuse were poison for this city. These days I hear too many people extolling and venerating the baubles of Corinth and Athens and laughing at the clay tiles of the Roman gods. For my part, these are the gods I’d rather see happy, and I hope they may remain so, if we let them stay in their own abodes.

Horace had subverted the roles of conquest at a national level—Greece captures her fierce captor (Ep. 2, 1, 156)—but Livy’s Cato locates the agency in the spoils themselves. The fear that Rome would be possessed by her possessions points directly back to the black day when the destructive (infesta) Syracusan artworks were brought into Rome, corrupting the Romans’ relationship with their old gods by giving them a taste for the wondrous (mirantes) imported images and making the old-fashioned representations seem laughable. ‘Cato’s’ view of the art from Syracuse and of the power of images is clearly of a piece with Livy’s; how much of this attitude can be reliably attributed to the real Cato is, of course, anyone’s guess.9 Livy does, however, give strong confirmation that this view of Letta (1984, 21 n.110) argues that this particular speech is either based on a real speech by Cato or assembled from genuinely Catonian material. Gruen (1992, 70) is more sceptical. For Livy’s access to actual speeches by Cato and his reliance on them, especially for his Cato’s speeches, see Astin 1978, 296–7. 9

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

100

MIMETIC CONTAGION

Marcellus’ ovatio and of Greek art in Rome more broadly goes back long before Plutarch, and by putting this sentiment in a speech made by Cato, thirty-four years before the Eunuch, he suggests that it is consistent with the culturally ‘traditionalist’ rhetoric of the mid-second century. Once again, though, Polybius provides the most direct evidence for views contemporary with the Eunuch, and he is as apprehensive of the art from Syracuse as Plutarch, Livy, and Livy’s Cato. If the Romans had always relied on the capture and display of such spoils, he says, that would be one thing, ‘but they were used to lives of simplicity, far from such luxury and magnificence, and yet they generally prevailed over those who had the greatest number and most beautiful of such works’ (ἁπλουστάτοις χρώμενοι βίοις καὶ πορρωτάτω τῆς ἐν τούτοις περιττότητος καὶ πολυτελείας ἀφεστῶτες ὅμως ἐπεκράτουν τούτων αἰεὶ παρ᾽ οἷς ὑπῆρχε πλεῖστα καὶ κάλλιστα τὰ τοιαῦτα, 9, 10, 5). How, he asks, can anyone doubt that the Romans’ change in practice after the sack of Syracuse was a mistake? For they abandoned the habits (ἔθη) of victors and took up emulation (ζῆλον) of the conquered. Along with the art itself came the envy (φθόνος) inevitably attending such objects and eventually an ardent hatred (ὀργή) of those successful enough to have taken it. Between Polybius, Livy, and Plutarch there are substantial differences in emphasis and in the of the precise way in which the art was so destructive for the Romans, but there is also clear agreement on the essential point: the Greek art introduced into Rome after the sack of Syracuse had a debilitating effect on the character of the Romans by teaching them to stare at art and, more subtly, by pulling them into ethical affinity with the art’s original owners and with the art itself. The older scholarship, even if it rejected the censorious judgment of Hellenism at Rome, tended to follow this moralistic tradition in its appraisal of the early Romans’ lack of sophistication and in seeing the sack of Syracuse as a turning point in Rome’s aesthetic history.10 But during the 1990s many scholars began to question the simple cultural boundaries implied by such as ‘Hellenization’ and even the very notion that Rome was ever meaningfully detachable from the greater Greek world.11 Regarding Marcellus’ ovatio and its effect on the Roman people, this more recent scholarly trend is best exemplified in the work of Erich Gruen, who has pointed out that already in the fourth century Rome was being decorated with imported statuary and paintings, including images of the gods brought in as spoils of war.12 Given this long-standing history of Greek art in Rome, Gruen argues, nothing about Marcellus’ ovatio would have seemed especially novel or problematic in 211 BCE, and the later tradition’s See, for example, Becatti 1951, 9–12; Pape 1975, 58–65; MacMullen 1991, 425–6. For example, Torelli 1990, 303–5. Along these lines, Gruen 1990 and 1992 together provide a thorough rethinking of Greek and Roman cultural interaction for this period. 12 Gruen 1992, 84–130. 10 11

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN TERENCE’S ROME

101

moralizing hindsight is anachronistic. Most recently, Myles McDonnell has argued, contra Gruen, that the Syracusan haul was indeed unprecedented, perhaps in its quality and more certainly in its quantity, and that a somewhat less sceptical reappraisal of the later tradition is in order.13 Happily, for the purposes of our investigation, the historicity of the traditional reading of Marcellus’ ovatio and the novelty of Greek art in Rome in 211 BCE may be left undecided. The only significant question for us is how Terence’s audience in the mid-second century understood the effect of Greek art on their own character. And it is precisely here, once again, that the testimony of Polybius is so valuable, for it confirms that the moralistic suspicion concerning Greek art and its influence on Romans can be traced back at least that far. As Gruen points out, similar claims (moral pollution, turning point in the history of Roman character) were also made for the triumphs of L. Cornelius Scipio Asiagenus in 189 and Cn. Manlius Vulso in 187, but where this may weaken the evidentiary value of the tradition for discerning what really happened in 211, it shows the continuing relevance of the moralistic trope in the first half of the second century.14 Considering the increasing frequency and splendour of triumphs in this period, Polybius’ claims about the spoils from Syracuse could hardly have a merely antiquarian point. The audience of the Eunuch needed look back no further than six years at the unprecedented haul of art and artists after the battle of Pydna.15

ART IN ROMAN COMEDY

These hopes and anxieties concerning imitative response are, of course, only one of the available ways of talking about the plastic arts in Middle Republican Rome. Accordingly, there is no reason to expect that the theme of art and mimetic contagion was a consistent interest of all Roman comedy, but before turning to the Eunuch, it is worth considering Plautus’ attitudes towards art, in order to establish a generic context for the dynamic evident in the Eunuch.16 Maurizio Bettini has convincingly argued that when images or the figurative arts are mentioned in Plautus’ plays, it is generally to indicate some perfected state of being. A striking illustration of this point is the adjective graphicus, which never has any of the meanings of the Greek γραφικός, all directly related to painting, but rather denotes perfection or unqualified success, especially at some form of

14 McDonnell 2006. Gruen 1992, 105–7. Livy 45.33.5–6; 45.39.5; Pliny HN 35.135; Plut. Aem. Paull. 6.5. 16 The most complete collection of evidence for painting in Roman comedy is Knapp 1917. More recently, Bettini 1999, 170–5 has a very compelling discussion of these issues (even touching on the Eunuch), the influence of which on my readings of Plautus will be apparent. 13 15

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

102

MIMETIC CONTAGION

trickery. For example, in Trinummus, Charmides calls one man a graphicus fur (1024) and another a graphicus nugator (936) and pergraphicus sycophanta (1139). A textbook tricky slave is a servus graphicus (Epid. 410), but even without explicit reference to a notionally crafty role, a character who fits his part to a T can be called a mortalis graphicus (Pseud. 519, 700; Stich. 570). The adverbial form, graphice, is used in the same way, to indicate some kind of picture-perfect rightness (Pers. 306, 464, 843; Trin. 767). The adjective graphicus holds up the painted image, if only in a recurring metaphor, as the standard of perfection against which ‘real life’ may be judged, but the comparison to painting is not limited to this metaphoric level. As Bettini says, ‘Plautus, in marked contrast with Plato’s rejection of the figurative arts, was convinced that images could be more true and perfect than truth itself.’17 For example, in Epidicus, the girl Telestis is as pretty as a picture: aspecta et contempla, Epidice: | usque ab ungiculo ad capillum summumst festivissuma. | estne consimilis quasi quom signum pictum pulchrum aspexeris? (‘Look her over, Epidicus; she’s a delight from tip to toe. Isn’t it like looking at a pretty painting?’ 622–4).18 There is no gap between her picture-perfect outward appearance and the idealized portrayal of feminine beauty one expects from a painting. But the qualities of a scene that make it worth comparing to a painting do not have to be strictly limited to the visible, nor must the painting be a recognizable type. In the Poenulus there is a scene so ‘picturesque’ that it suggests a non-existent archetype. The recognition is past and the play nearly over; Hanno has gained his long-lost daughters and everyone embraces in a group hug. At the sight, the young man Agorastocles breaks out: o Apelle, o Zeuxis pictor, | qur numero estis mortui, hoc exemplo ut pingeretis? | nam alios pictores nil moror huiusmodi tractare exempla (‘O Apelles, o Zeuxis the painter, why are you dead too soon to paint this subject? For I wouldn’t want other painters handling subjects of this sort’, 1271–3). Agorastocles can see the reunited family with his own eyes, but he yearns to see the scene painted: ‘The picture is not a bad imitation of reality: quite the opposite. It is manifest perfection, rendering the truth even more worthy of being true.’19 In Mercator, Lysimachus exclaims to the audience that his friend Demipho, despite his advanced age, is the very portrait of a lover: si umquam vidistis pictum amatorem, em illic est. | nam meo quidem animo vetulus, decrepitus senex | tantidemst quasi sit signum pictum in pariete (‘If you’ve ever seen a painted lover, behold the man. Anyway to my mind this broke down little oldster is just the same as a painted picture on the wall’, 313–15). The comparison here is a little more 17 18

Bettini 1999, 173. For signum in Plautus as generalized ‘representation’, rather than specifically ‘statue’, see Knapp 1917,

150. 19

Bettini 1999, 174.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN TERENCE’S ROME

103

complicated than that in Epidicus, because the essential identity asserted between life and art is not based in outward appearance, but rather on a hidden similarity of character. Demipho’s inside and outside do not match, and the irony is that he is likened, in spite of his appearance, to a painted image, which is, of course, all surface. The joke here is also metatheatrical, in that both senex and adulescens amator are stock character types, as visibly identifiable as the painted lover. We know that Demipho is a senex because he looks like one, as prescribed by the conventions of mask and costume. Lysimachus insists that Demipho’s character does not fit his appearance, but he does so by comparing his friend’s hidden heart to the two-dimensional painted lover, a conventional type just as ethically flat as the comic senex. He confronts the spectators as viewers (vidistis), hypothetically of a painted lover and then more immediately of ‘this man’ (em illic).20 When the fourth wall is so patently ignored and the audience, directly addressed as spectators, is asked to regard ‘this guy’, the referent of the deictic illic begins to slip between character and actor, especially in the context of this playful opposition of visible appearance and unexposed personality. The only secret self, hidden beneath the theatrical signifiers of character, is the actor, who probably is quite different from the senex he plays. The only other reference to painting in Roman comedy comparable in extent to the one in the Eunuch is in the Mostellaria. The slave Tranio has already proven his mettle as a servus callidus by pulling off several grand schemes to save his young master, and now, in his victory lap, he is simultaneously tricking the two old men Theopropides and Simo. Tranio has convinced Simo to let Theopropides tour his house, because Theopropides (he says) must make some home improvements and wants to use Simo’s house as a model. Theopropides, however, has been fooled into thinking that his son has bought Simo’s house, so the tour is an inspection of his own property. During the inspection, Tranio shamelessly plays to both old men at once, stroking the vanity of each by extolling the house each thinks he owns, while artfully ensuring that neither reveals his understanding of the situation to the other. As he goes on he gets bolder and more playful and cannot help pointing out a pair of posts, how amazingly stiff they are and how dense (quanta firmitate facti et quanta crassitudine, 819). Such loaded observations

20 The connection between the two acts of viewing in this sentence is subtler than it may at first seem. There is no logical dependence of the apodosis on the protasis; his being the quintessence of the lover is in no way affected by whether or not the audience has seen a painting. This is an example of a ‘speech-act conditional’ (for this category, see Sweetser 1990, 118–21). The full sense is something like: ‘If you’ve seen a painted lover before [then you’ll know what I mean]: look, that’s him’. The seeing implicit in the apodosis is due to the illocutionary force of em illic, which, like ecce, is essentially imperative. The discourse particle em was probably originally an imperative of emo, and in comedy generally expresses consternation or surprise at some visible person or thing, which the addressee is invited to consider. The deictic illic makes the bid for the spectators’ visual attention even stronger. See Hand 1969, vol. 3, 49.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

104

MIMETIC CONTAGION

accumulate and become more pointed, until they reach their climax in an ekphrasis of a painting (832–40): TR. viden pictum, ubi ludificat una cornix volturios duos? TH. non edepol video. TR. at ego video. nam inter volturios duos cornix astat, ea volturios duo vicissim vellicat. quaeso huc ad me specta, cornicem ut conspicere possies. iam vides? TH. profecto nullam equidem illic cornicem intuor. TR. at tu isto ad vos optuere, quoniam cornicem nequis conspicari, si volturios forte possis contui. TH. omnino, ut te apsolvam, nullam pictam conspicio hic avem. TR. age, iam mitto, ignosco: aetate non quis optuerier. TR. Do you see that painting of a crow tricking two vultures? TH. No, I don’t see that. TR. Well, I do. There’s a crow standing Between two vultures pecking at them one after the other. Look over here towards me and you’ll see the crow. Now do you see it? TH. Gosh, I don’t see any crow over there. TR. OK, look over towards each other, since you can’t catch the crow, Maybe you’ll be able to see the vultures. TH. You know, I’m just not seeing any painted bird at all. TR. Well, forget about it; it’s OK. You’re probably just too old.

Scholars are divided on the question of whether this painting is meant to be real or merely dreamed up by Tranio for his fantastic lie.21 I incline strongly to the latter view, but either way, the broad similarity between this situation and that in the Eunuch is clear: the characters take after the figures in a painting. Tranio, like the crow, makes a spectacle of the old men (ludificat), and they are helpless to stop him. Once again, the painting tells the truth about what is really going on, but the old men cannot see the painting, and this, paradoxically, proves the painting’s accuracy. For they are just as stupid as the vultures and as blind as Tranio says they are, otherwise they would surely catch him in his lie and reveal the painting to be simultaneously unreal and false. Apart from the Eunuch, there are two other references to mythological painting in Roman comedy, and unlike the painting in the Eunuch they are not ‘real’ in the play, but merely recollections of general types. One is from the Captivi, where Tyndarus returns from the quarries at the end of the play: Vidi ego multa saepe picta, quae Accherunti fierent | cruciamenta, verum enim vero nulla adaeque est Accheruns | atque ubi ego fui, in lapicidinis (‘I have seen the many tortures often painted, which happen in hell, but there’s no hell like where I was, in the quarries’, 998–1000). Here the painted scene is brought in as a comparandum for Tyndarus’ real sufferings and found inadequately horrific. As we have seen, this criticism of 21

See Leach 1969, 327–8.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN TERENCE’S ROME

105

painting’s failure to render the truth with sufficient vividness is atypical; in fact, it seems to be unique in all Roman comedy. The other reference to mythological painting has proven very difficult to interpret. When Menaechmus I emerges from his house, having filched his wife’s dress to give to his mistress, he crows happily to Peniculus (141–6): ME. vin tu facinus luculentum inspicere? PE. quis id coxit coquos? iam sciam si quid titubatumst, ubi reliquias videro. ME. dic mi, enumquam tu vidisti tabulam pictam in pariete ubi aquila Catameitum raperet aut ubi Venus Adoneum? PE. saepe. sed quid istae picturae ad me attinent? ME. age me aspice. ecquid adsimulo similiter? PE. qui istic est ornatus tuos? ME. You want to lay eyes on a brilliant accomplishment? PE. Who was the cook? I can tell if there was a slip-up just by looking at the leftovers. ME. Tell me, have you ever seen a wall painting Where the eagle carries off Ganymede or where Venus snatches Adonis? PE. Sure, but what do your pictures have to do with me? ME. Look at me. Am I the spitting image of anything? PE. What’s that you’re wearing?

Menaechmus has smuggled his wife’s palla out of the house by wearing it under his cloak, which he presumably parts or doffs in line 145 to reveal his unusual underwear and prompt the parasite’s reaction. He seems to think he is ‘just like’ something in an abduction painting.22 But what does he mean? It is unclear just what or whom he is supposed to resemble, and scholars have been divided on this question, without apparent acknowledgement of the diversity of likely interpretations. One possibility is that he is the wily abductor and the palla is the ravished beloved.23 Alternatively, there is no ‘abductor’ in this scenario, but dressed in drag, he is a champion pretty boy (luculentum), like one of the all-star effeminates of Greek myth.24 Or perhaps the figure he cuts recalls the painted ensemble, both aggressor and victim at once. If we imagine him holding his cloak out to reveal the dress, then his outspread pallium may recollect the eagle’s wings or Venus’ robes, while the dress underneath puts him in the role of erotic prey.25

For the colloquial use of figura etymologica for intensity in adsimulo similiter, see Hofmann 1951, 94–5. Knapp 1917, 152: ‘Menaechmus’ allusion is, to be sure, rather far-fetched; we have to suppose that he thinks of himself as the eagle or as Venus, of the cloak as Ganymede or Adonis.’ 24 Gratwick 1993, 151–2: ‘Men. is facetiously comparing himself with Ganymede and Adonis as types of champion “pretty boy” whom Zeus and Venus respectively as types of the superhuman found irresistibly attractive.’ 25 Moseley and Hammond 1962, 62: ‘Presumably he thinks that his outspread cloak represents the eagle’s wings or the robes of Venus, and he, in the woman’s mantle, one of the effeminate boys of Greek mythology (in 197 he is compared to a female impersonator). There is also implied the idea of a successful abduction.’ 22 23

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

106

MIMETIC CONTAGION

Whichever of these interpretations is adopted, none of them can easily be denied, so there is a prima facie ambiguity in the comparison to the painting, or perhaps better, an instability in the assignment of roles, akin to what we will be tracing out in the Eunuch in Chapter 5. I would suggest a progression, then, as follows. Given his manifest pride throughout this scene at successfully making off with his wife’s palla, Menaechmus casts himself predominantly in the role of divine plunderer and the dress as the worthy booty, but already ecquid adsimulo similiter? suggests a sardonic reply, which Peniculus does not quite make. As Gratwick says, ‘The suppressed answer to the question is “yes, a cinaedus.”’26 But this turns out to have been the right answer. The day’s events will soon come swooping down out of the blue on poor Menaechmus I, when his long-lost twin brother, Menaechmus II, shows up and creates all havoc. In the ensuing confusion the distribution of benefits is one-sided: Menaechmus II gets to enjoy manly delights and freedoms his brother can only dream of (sex with his brother’s mistress, followed by robbing her, then telling off his brother’s feared wife and father-in-law), while Menaechmus I suffers the consequences of his brother’s behaviour as a ive victim. Menaechmus I really is Fortune’s cinaedus, and his claim to be ‘just like’ the painting proves to true in ways he could not have predicted or wished.

EUNUCH

Where Chapter 1 provided substantial evidence of the painting’s persuasive role in the Eunuch, this chapter has endeavoured to show how it was ever possible for an image to be thought to have such power. We have attempted to situate this instance of mimetic contagion within a larger body of Greek and Roman discourses and practices involving works of art that impress themselves upon their viewers as models. More specifically, we have tried to indicate some of the ways in which this preoccupation with ethically persuasive art seems to have had a place in mid-second-century BCE Roman culture. More specifically still, we have examined earlier Roman comedy for evidence of attitudes to painting and found a range of suggestive precursors to the scenario in the Eunuch. In several of these Plautine examples the character’s averred similarity to a painting proved to be slipperier than has usually been noticed, perhaps due to the metatheatrical instability inherent in the enclosure, however brief, of a painting within a play. The remainder of this chapter will examine the Eunuch somewhat along the lines indicated by these Plautus plays, on the assumption that Terence’s audience would have been as ready to think with the painting as they were to notice its effect. 26

Gratwick 1993, 152.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN TERENCE’S ROME

107

If an artistic representation of a rape can incite a viewer to imitation, what about a play that contains a rape? This is the question implicit in Augustine’s metaphorical deployment of the painting.27 At this level of abstraction the painting is more than just a picture; it has already become a figure for the work of art that contains it, and its relationship to its dramatic setting is essentially emblematic.28 It is, of course, by no means surprising that a painting should be made to stand for a poem, least of all a dramatic poem. Horace’s famous dictum, ut pictura poesis (AP 361), is only the handiest example of an extremely widespread ancient tendency to locate the metaphorical ground of literary art in visual art.29 Rather than comparing the plastic to the poetic or rather than simply ignoring the possibility of transmedial analogy between the ‘mimetic’ arts, the Greeks and Romans, going back at least to Pindar and Simonides, usually explained verbal art by reference to pictorial art.30 The analogy to figurative representation is all the more irresistible with drama, for theatre is a visual art form in its own right and is, among literary forms, uniquely ‘made of ’ plastic art, from the masks, costumes, and other properties to the stage set, which may have been one of the most public displays of large-scale painting in the ancient city. As Alison Sharrock says, ‘Drama seems to stand on the boundary between verbal and visual art.’31 This approach to the narrated painting in the Eunuch as mise en abîme at the heraldic centre of the play suggests a type of self-awareness of theatre-as-art that is rather unlike any of the kinds of metatheatricality scholars have detected in Roman comedy. It is not simply a matter of a joking reference to the audience or a lexical wink at the language of the theatre, but rather a form of recursive autosimilarity. The painting is ‘fictional’, from the perspective of Chaerea and Pamphila, in that it tells someone else’s story for spectacular consumption and is obviously artificial, and yet the painting proves able spontaneously to overstep its frame and become ‘real’ in the lives of its viewers. If the painting’s relationship to the internal audience has implications for the play’s relationship to the external audience, there is another type of theatrical See pp. 46–8. By ‘emblem’ I mean to invoke, rather specifically, the Renaissance tradition of printing images with affixed captions. These pictures were often used in allegorical illustration of the accompanying text, but the inevitable gap between the two media could result in a relationship ranging from supplementary homology to irreconcilable divergence. My thinking about the painting in the Eunuch has been influenced by some of the scholarship on this aspect of early modern book culture. Especially helpful for the philosophical and historical background of emblems is Russell 1995; for a broader range of recent perspectives, see Saunders and Davidson 2005. 29 A recent survey of this theme across antiquity, with thorough references to earlier scholarship, is Benediktson 2000. 30 For Pindar see, e.g., N. 5; for Simonides, see Plu. Mor. 3.346F–347A. 31 Sharrock 1996, 104. 27 28

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

108

MIMETIC CONTAGION

self-consciousness at play in the same scene. The eunuch intrigue is quite clearly handled as a play-within-the-play, and scholars have long noted the prominence of theatrical vocabulary surrounding Chaerea’s ‘role’ as the eunuch.32 The most irrefutable example may be Parmeno’s boast, in Act 5, that his handiwork in scripting the day’s events deserves the Palme d’Or (palmarium, 930). Critics since Ashmore have read palmarium as a witty reference to the palma, the dramatic prize that would shortly be given to the Eunuch.33 But Parmeno is not the only character to use explicitly theatrical language in connection with this deception. Chaerea’s comparison of his own exploit with Jupiter’s emphasizes the spectacular and ludic aspect of both: egomet quoque id spectare coepi, et quia consimilem lat | iam olim ille ludum (‘I began watching it too, and because he had long ago played such a similar game’, 586–7). Later, when the indignant Pythias insists that the shrivelled old weasel whom Phaedria is calling the eunuch bears no resemblance to the attractive adulescentulus Parmeno delivered in Act 3, Phaedria responds ambiguously, hem quae haec est fabula? (‘Ah, what is this farce?’, 689). He may well doubt the veracity of her ‘story’, but he also begins to recognize the signs of comic subterfuge behind the mounting evidence of mistaken identity, for the word fabula, even more than ludus, is linked to the technical language of theatrical entertainment.34 Soon enough, by interrogating the real eunuch, he will learn of Chaerea’s involvement, and he must undertake damage control by getting himself and the eunuch offstage as quickly as possible. As he drives the poor creature off he pretends to threaten him: actumst, siquidem tu me hic etiam, nebulo, ludificabere (‘It’s all over if you’re going to make fun of me here too, you nebbish’, 717). But Pythias is not fooled by this attempt to shift the play-making (ludificabere) to the eunuch, and her recognition of Parmeno’s hand comes out in the next line: Parmenoni’ tam scio esse hanc technam quam me vivere (‘This is Parmeno’s artistry, sure as I’m alive’). Pythias recognizes Parmeno’s dramatic techina in the intrigue and knows him for the auctor (1013). Identity deception is a common-enough plot motif in comedy, but when coupled with dress-up, it takes on an unmistakably theatrical dimension.35

32 Earlier scholars (Saylor 1975, 304–5; Gilula 1989, 95) remarked on the general confluence of dramatic language and its appropriateness with respect to the intrigue, but Frangoulidis 1994a presents a coordinated argument for theatrical self-consciousness. 33 See Ashmore 1910, ad loc.; Duckworth 1952, 78; Fantham 1972, 33; Tromaras 1994, ad loc.; Frangoulidis 1994a, 125 n.11. We have no direct evidence that the Eunuch was awarded the palma, but Suetonius’ testimony (Vita Terenti 2) that it was performed twice in one day and was the highest remunerated comedy to date would seem to point in that direction. On the palm frond as a dramatic prize from at least the time of Plautus, see Manuwald 2011, 88. 34 35 See OLD, fabula 6. In Terence no less than in Plautus, pace Muecke 1986, 221.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN TERENCE’S ROME

109

This intrigue does involve a change of vestment, and the text repeatedly draws attention to Chaerea’s eunuch costume as such.36 Given such clear justification for a theatricalized reading of the intrigue, Stavros Frangoulidis argues that Chaerea’s successful performance in his play-within-the-play ensures a happy ending for the Eunuch: ‘Chaerea’s acting brings about his personal happiness as well as a positive outcome of Phaedria’s troublesome relationship with Thais.’37 He recognizes, however, that the rape does not figure in Parmeno’s plan: ‘In the context of the performance, this act must be considered as an ad hoc improvisation.’38 This approach to the rape as a kind of improvisational theatre is quite astute, and, as I shall explain in Chapter 5, I believe it points in the right direction, but there are difficulties with the straightforward ascription of the improvisational mode to this scene. The motif of unscripted improvisation ironically written into the texts of Plautine comedy is well documented, but there are some important differences from the Plautine specimens of this type.39 In Plautus, characters may contrive to dress up and affect new identities in order to accomplish a deception, and once under way in their new and overtly prescribed ‘roles’ they may spontaneously diverge from plan, either by necessity or just for fun. In the first case, something unexpected comes up and the trickster must adapt to the new contingency in order to keep the deception going; in the second, the frivolous departures from plan may come close to wrecking the deception, but they never actually do so. In both cases, the comedy comes both from the very visual pleasure of watching a character show off his virtuoso plasticity and also from the witty evocation of the truly unscripted forms of popular entertainment, the degree-zero of Italian folk theatre from Atellan farce down to the Commedia dell’arte. It is precisely this visual dimension, so essential to the comedy of the ad hoc, that is lost by placing Chaerea’s ‘improvisation’ backstage. An easy equation of this rape with improvisation becomes impossible once we recognize the fundamental denaturing of the improvisational mode when the antics happen not before our very eyes with the giddy delirium of the unplanned and indeterminate, but merely in our minds and after the fact. Another problem with reading the rape as an unscripted improvisational interlude is that the inspiration for the rape does not come from Chaerea’s masterful control of his performance or his genius for thespian invention, but directly from the painting.40 Indeed, as this chapter has prepared us to see, the painting 36

capias tu illi(u)s vestem. vestem? (370); qui hic ornatust? (546); quid sibi hic vestitu’ quaerit? (558)— note here the echo of the last line of the Prologue: quid sibi Eunuchus velit (45); vestem cum illo mutem (572); is dedit mi hanc (701); meam ipse induit (702); Chaerea tuam vestem detraxit tibi? (707); et east indutu’? (708). 37 38 Frangoulidis 1994a, 125. Frangoulidis 1994a, 123. 39 For a fuller treatment of the scholarship on improvisation in Plautus, see p. 121. 40 Though elsewhere Frangoulidis recognizes that the rape was not planned, his portrayal of the painting’s role demotes it to mere confirmation of Chaerea’s already formed intention: ‘Chaerea, as a

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

110

MIMETIC CONTAGION

functions precisely as a script. It would be giving Chaerea far too much credit to see him as the author of anything; he is an actor, who reads his new part written on the wall and then agrees to enact it. If we understand the rape to be theatricalized simply as an ad hoc improvised performance, not only must we totally elide the fixity and materiality of the painting as text, but more fundamentally we ignore the simple fact that there are two actors in this performance and that they come to their new script with the same essential ivity. No one would call Pamphila the authoress of anything, but it is her attention to the painting that conducts Chaerea’s eyes to it in the first place, and her ivity as reader and actress is ultimately parallel to Chaerea’s response to the painting. If it were simply a matter of Chaerea getting the idea from the painting, it would have been easy enough for him simply to describe what he saw and thought, without the dramaturgically unmotivated attention to Pamphila as viewer. Instead, his fixation travels along the line of her gaze to a new object and a new goal. During his ekphrasis we are invited to imagine a quiet scene of art appreciation in which rapist and victim contemplate the roles they are about to play. In what sense is Chaerea’s performance as the eunuch successful? Unlike a good improvisation, which complements and augments its framing performance, the theatricalized rape completely derails and vitiates all the careful verisimilitude of the performance in which it is embedded. The crime effectively blows his cover and renders him a lousy ‘eunuch’. Frangoulidis’ contention that the rape brings about a happy resolution for Phaedria and Thais would receive little from a critical tradition that finds the end of the play sour for both of them, but his claim that the rape makes Chaerea’s marriage possible seems less controversial.41 Notice, however, that Chaerea too gains nothing from the rape. If he had kept to the plan he would have spent part of the day captivated with Pamphila and left quietly, more in love than ever, to host the potluck with his friends. He would have returned to find that the girl he loves is actually a marriageable Athenian citizen and that her guardian Chremes is conveniently accessible and in a mood to dower his long-lost sister. As it is, his happy ending is not the result of his own performance, but of Thais’ goodwill and his father’s permissiveness in allowing him to marry without a dowry. Far from giving him control of his destiny, the rape puts Chaerea in the extremely degraded position of subjugation to a meretrix. When he returns to the stage in Act 5, mortified with shame at the eunuch outfit he still has been unable to doff and terrified that he will be recognized, he finds Thais there waiting, and he is stuck in his deception (ipsast. haereo | quid faciam, 848–9). viewer of the painting, discovers that the god, who secretly comes into Danaë’s bosom in the form of golden rain, plays the role he intends to play’ (1994a, 123). 41

On the problem of the play’s ‘happy ending’, see pp. 175–6.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN TERENCE’S ROME

111

He must meekly submit to his mistress’ (era, 851) indignation. By this point, of course, she knows his real identity, but she is now in a position to reveal Pamphila’s citizen status, so she has him at a great disadvantage. For fourteen lines (850–63) he must endure being treated like runaway chattel, and when he learns that Pamphila is a citizen he can only mutter that he thought she was a fellow slave (conservam esse credidi, 858). He realizes that he has committed a serious crime, but for the moment he is trapped in his part. Thais is in control here, and it is only when she sees fit to call off the teasing pretence, naming him as Chaerea and his crime as something unworthy of his real self, that he is finally let out of the eunuch role (missa haec faciamu’. non te dignum, Chaerea, | fecisti, 864–5). This does not mean, however, that he is free to resume the contemptuous attitude towards Thais he had expressed with such vehemence earlier (382–7). Rather, his gratitude to her is eternal (872), his love for her is profound (882), and when he comes to supplicate her help in securing Pamphila’s hand, he makes her his patroness: nunc ego te in hac re mi oro ut adiutrix sies, | ego me tuae commendo et committo fide, | te mihi patronam capio, Thai’, te opsecro (‘I hereby implore you to give me your in this matter, I consign and entrust myself to your protection, I take you as my patroness, Thais, I beg you’, 885–7).42 It is here, in Chaerea’s indignity, that the adhesive quality of role-playing in the Eunuch is first revealed. He has played the eunuch, presumably with the understanding that he could return to himself at will, but now, with nothing left to gain from the deception, he is inexplicably stuck in his part, until Thais decides she has punished him enough and releases him from the role. In taking the initiative to manumit him, she exhibits decisive magnanimity and effectively emasculates him. He responds by immediately suing to become her client and eternal protégé. Chaerea may have escaped the role of the eunuch at last, but he appears to be openly humiliated and unmanned.43 Since his emergence from the house the costume has been a persistent inconvenience for him, one he has tried in vain to escape, and now the part seems to have set in on an ethical level. Chaerea’s loss of virility is only symbolic, of course, and there is doubtless a playful quality in his grovelling. However, a few lines later we find an even more 42

The verb commendare has already been used once in this play, when Chaerea is narrating the rape scene and he comes to the moment when the housekeeper hands Pamphila over to his care (commendat virginem, 577). 43 Dessen 1995 sees several characters in the play as sexually ambivalent extensions of the figure of the eunuch, including Phaedria, Thais, Thraso, Gnatho, and above all Chaerea. She even reads Chaerea’s role in the rape as suggestive of the sexual monstrosity of this ‘controlling metaphor’. Though she recognizes the problem in Chaerea’s submission to Thais as patron, Dessen also sees it as a positive revaluation of the eunuch theme, since Thais’ forgiveness is ‘a better way of handling aggression’ (133). She goes on: ‘It is significant, I think, that the audience sees Chaerea acceding still dressed as a eunuch; he thus makes the eunuch a positive symbol for social change’ (134). Though I agree with her that symbolic castration is an important theme in the play, I strongly doubt there is anything positive about it.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

112

MIMETIC CONTAGION

striking, because more literal, reference to his castration. Pythias has been watching Parmeno from hiding, and she decides to take her vengeance by making him think things have gone even worse than they have. She exchanges the iambic senarii they have both been using for more agitated trochaic septenarii and launches into a bogus entrance monologue (943–58): PY. pro deum fidem, facinu’ foedum! o infelicem adulescentulum! o scelestum Parmenonem, qui istum huc adduxit! PA. quid est? PY. miseret me: itaque ut ne viderem, misera huc ecfugi foras. quae futura exempla dicunt in illum indigna! PA. o Iuppiter, quae illaec turbast? numnam ego perii? adibo. quid istuc, Pythias? quid ais? in quem exempla fient? PY. rogitas, audacissime? perdidisti istum quem adduxti pro eunucho adulescentulum, dum studes dare verba nobis. PA. quid ita? aut quid factumst? cedo. PY. dicam: virginem istam, Thaidi hodie quae dono datast, scis eam hinc civem esse? et fratrem ei(u)s adprime nobilem? PA. nescio. PY. atqui sic inventast: eam istic vitiavit miser. ille ubi id rescivit factum frater violentissimus PA. quidnam fecit? PY. conligavit primum eum miseris modis. PA. hem conligavit? PY. atque quidem orante ut ne id faceret Thaide. PA. quid ais? PY. nunc minatur porro sese id quod moechis solet: quod ego numquam vidi fieri neque velim. PY. With gods as my witness, what a foul deed! O you poor lad! O vile Parmeno, who brought him here! PA. What’s she on about? PY. It just kills me; I ran out here so I wouldn’t have to watch. What shocking punishment they say he’s in for! PA. O god, What’s this mess? Am I done for? Here goes. What’s up, Pythias? What are you saying? Punishments for whom? PY. You! You have to ask? You’ve destroyed that poor boy you delivered in place of the eunuch, Trying to trick us. PA. What? What’s happened? Tell me. PY. OK. The girl, the one given to Thais today as a gift, You know she’s a citizen? And her brother from one of the best families? PA. No. PY. Well, it’s been proven. That poor boy of yours raped her. When her brother found out about it, well, he’s a hot-tempered sort and . . . PA. What did he do? PY. First tied him up, rough like. PA. What? Tied him up? PY. Yes, and Thais all the time begging him not to. PA. What are you saying? PY. Now he’s threatening, you know, What happens to adulterers. Which I’ve never seen done and would just as soon not see.

Twice, in 945 and in 958, she indicates that she has fled outside to avoid witnessing his punishment. What it is exactly she does not say, but she feels sorry enough for him to use the pathetic diminutive (943, 949) and call him infelix (943) and miser (953). Her euphemistic reference to id quod moechis solet (‘the usual treatment for

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN TERENCE’S ROME

113

adulterers’, 957) seems frustratingly elliptical, but it is clear from Parmeno’s reaction that he understands what she means, so our confusion must be due to our own ignorance of second-century Roman punitive practice. In Athens, an adulterer could be killed on the spot if caught in the act by his partner’s husband, father, or brother (Dem. 23.55; Lys. 1.25–6), and under Augustan legal reforms Roman men apparently gained the same prerogative in some circumstances, but this provision is not securely attested for the pre-Augustan period.44 In any case, the situation described in Pythias’ narration is different. For one thing, she never says he was caught in the act, but more importantly, he has been bound and detained; the threatened punishment, the thing she has never seen done before, is clearly not simple killing in the heat of the moment. There were several inventive punishments for adulterers in antiquity, including depilation of pubic hair with hot ash and, infamously, anal penetration with radishes and/or mullets. Such discomforts are attested by Aristophanes (Nu. 1083; Pl. 168) and by later Roman poets (Catul. 15.19; Juv. 10.314–17), but there is no evidence from legal, historical, or even rhetorical writers that these torments were ever normal at Rome. Valerius Maximus cites flogging, castration, and rape by servants as typical punishments, but it is the second of these, castration, that is known to Roman comedy.45 ‘What are you doing?’ asks Milphio in the Poenulus. ‘Something red-handed adulterers hardly ever do’, answers Syncerastus (facio quod manufesti moechi hau ferme solent, 862). ‘What’s that?’ ‘I’m going home with my equipment intact’ (refero vasa salva, 863). At the beginning of Curculio Palinurus offers the young Phaedromus a circumspect warning about the consequences of illicit love (23–32): PA. numquid tu quod te aut genere indignum sit tuo facis aut inceptas facinus facere, Phaedrome? num tu pudicae cuipiam insidias locas aut quam pudicam esse oportet? PH. nemini; nec me ille sirit Iuppiter. PA. ego item volo. See Treggiari 1991, 270–5. VM 6.1.13. The Bacchides may be an outlier to this rule. In 860–9 Cleomachus threatens to kill his wife and her lover if he catches them together, so as to teach her a lesson, but there is an obvious joke here at the expense of the miles, as the quintessentially slow-witted character type. ‘What the soldier threatens in Bacchides is not evidence of what a Roman husband could legally do. He speaks of killing wife and lover, to make her it that he is not contemptible. This seems to be the only age in comedy which even mentions killing both. Plautus wants to make a joke, not to illustrate law’ (Treggiari 1991, 271). At first, Nicobulus, the lover’s father, seems ready to take Cleomachus at his word (867), but once he thinks his son safe, he rephrases the danger the boy has eluded: obtruncaret moechum manifestarium (‘he would have mutilated him as a red-handed adulterer’, 918). To be sure, obtruncare could be no more than a synonym of the earlier necare, but it also suggests the apparently more normal punishment: reduction to a trunk, i.e., the paring off of his son’s relevant extremity. 44 45

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

114

MIMETIC CONTAGION

ita tuom conferto amare semper, si sapis, ne id quod ames populus si sciat, tibi sit probro. semper curato ne sis intestabilis. PH. quid istuc est verbi? PA. caute ut incedas via: quod amas amato testibus praesentibus. PA. You aren’t doing anything unworthy of yourself or your family Or setting out to commit some crime, are you, Phaedromus? Laying a trap for an honest woman Or one that ought to be? PH. No, God forbid. PA. Good, that’s what I like to hear. If you’re smart, you’ll always check your love affairs, So if people find out who it is, it won’t be a mess. Just be careful you can always bear witness like a man. PH. Huh? PA. Walk the straight and narrow, you know. Love your love, but make sure you keep your witnesses.

The pun here is too straightforward to need much elaboration. If caught in adultery Phaedromus will not only be shunned as morally ‘detestable’ (OLD intestabilis 2), he will also be branded with infamia and therefore ineligible to participate in the legal process (OLD intestabilis 1). More immediately, though, he will be despoiled of the twin attestants of his manhood. This joke also comes into play in Plautus’ most extended treatment of the punishment of an adulterer, the entire final act of Miles Gloriosus. Periplectomenus has the stripped Pyrgopolynices thrown into the street, and then he calls to his cook to ready the instruments (1397–9): vide ut istic tibi sit acutus, Cario, culter probe. CA. quin iamdudum gestit moecho hoc abdomen adimere, ut faciam quasi puero in collo pendeant crepundia. Cario, make sure your knife is good and sharp. CA. Oh, it’s long been hopping to lop off this adulterer’s bag of appetites, So I can make him a nice toy to hang around his neck, like a little kid.

The lorarii rough him up for a few lines, and when he cries that he has been beaten enough, Cario is eager for the punishment to begin in earnest (1406–8): CA. quam mox seco? PE. ubi lubet: dispennite hominem divorsum et distendite. PY. opsecro hercle te ut mea verba audias priu’ quam secat. CA. How soon do I cut? PE. Whenever you like. Get the guy in position, spread eagle. PY. Oh god! I’m begging you, hear what I have to say before he cuts me.

Pyrgopolynices tries to explain that he was tricked and that he thought she was unmarried. They make him promise that if they do let him go in one piece (salvom,

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN TERENCE’S ROME

115

1413) he will never to try to take revenge for the beating he has received. He swears that he will never seek revenge and agrees that his beating has been just. He would consider it a privilege to be given witness protection, but he will also accept a life unencumbered: et si intestatus non abeo hinc, bene agitur pro noxia. | PE. quid si id non faxis? PY. ut vivam semper intestabilis (‘And if I don’t walk away from here unwitnessed, I’ll be getting off easy. PE. What if you don’t. PY. Let me live on, forever free of witness’, 1416–17). Cario says he should get another beating and then be set free, and Pyrgopolynices blesses the cook for this advocacy; then Cario demands a mina of gold, and Pyrgopolynices asks what for. ‘Well, lover boy, that’s the price of you getting out of here today with your testicles’ (salvis testibus, 1420). He agrees immediately, and they send him away stripped and beaten. ‘If I ever catch you around here again’, warns Periplectomenus, ‘you will lose your testicles’ (carebis testibus, 1426). Given these three ages from Poenulus, Curculio, and Miles Gloriosus, it seems incontestable that Pythias means castration. There is no evidence of any other standard punishment for the period, so when she says id quod moechis solet, the audience would probably identify castration as the customary default for this situation. But apart from the real world, considering the uniformity of the Plautine testimony, her solet probably also has a metatheatrical application. As we have already argued, Pythias is the consummately self-aware character in this play, the only one who consistently knows the conventions of the comic genre.46 Her reference here to what ‘typically’ is done to adulterers points to the punitive norm both on and off the Roman stage. Of course, it would be a vast departure from the comic genre for an adulescens actually to be castrated, and this I think is part of the humour of Parmeno’s send-up. If Parmeno were not so forgetful of his place in a fabula, and if he were paying closer attention to Pythias’ language, he might have noticed that she is not saying the castration has happened yet—nunc minatur porro sese id quod moechis solet: so far Chremes is only ‘threatening’ the usual for adulterers. Indeed, id quod moechis solet, in generic , is being threatened with castration, not actually suffering it.47 Surprisingly, the scholarship on this play has never pointed out the simple irony of Pythias’ suggestion: a character who has played a eunuch in a play called Eunuchus is threatened with castration.48 As apposite as Chaerea’s imagined punishment is, See pp. 14–15. In line 946 Pythias had described the castration as indigna. This word bothered Donatus enough to get the gloss per ‘indigna’ foeda crudeliaque significat (‘by indigna he means foul and cruel’). To be sure it is easy to find for this reading of indignus = ‘shocking’, ‘horrible’, vel sim. (cf. OLD indignus 6), but given the earlier argument, we may point out that castration is also indignus in the more proper sense of ‘unsuitable’ for a comic adulescens. 48 In some cases this seems to have been due to confusion about what punishment is intended. Surely the unforgettable radishes and mullets of Catullus 15 have played a role in blinding most readers to the more 46 47

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

116

MIMETIC CONTAGION

however, its real significance lies in its complementarity to the immediately preceding scene, in which Thais effectively unmanned Chaerea, first by teasing him as a runaway slave, and then by generously releasing him from his role as eunuch only to take him on as her client. Pythias observed all this and protested that he was getting off too easy, because in spite of his loss of social virility, he is still a sexual threat (883–4, 898–904). Now she effects his literal castration in Parmeno’s mind, the same mind that scripted the eunuch intrigue, as she herself realizes (Parmenoni’ tam scio esse hanc technam quam me vivere, 718). First symbolically at Thais’ hands and then fictionally at Pythias’ hands, the role of the eunuch transcends its boundary as play and is projected as reality.49 It is here that the painting’s role in the play as emblem is most clear. The repeated emphasis on the theatrical artifice of the eunuch intrigue, clustered around a ‘real’ work of art at the centre of the play, already suggests a connection between the travesty and the painting. Parmeno had crafted the role of the eunuch as a work of dramatic art, and he expected Chaerea would be able to step into it and out of it at will, but precisely because of the intervention of the painting in the plot, the role has turned out to have the same aggressive quality as the image. Chaerea’s performance as the eunuch, like the painting, wants to step outside its intended frame and become reality, and it is in this aspect of the role of the eunuch that the logic of the painting is most clearly discernible, but the painting is not only at the heart of the eunuch intrigue, it is also at the heart, indeed as we have said, at the very stichic centre of the whole play.50 If the painting is the key to understanding the dynamic of the intrigue, it is probably also pivotal to what the play has to say about art, including especially the dramatic art of role-playing. It is not surprising that we should be expected to think out from the eunuch intrigue to the rest of the Eunuch, both because of its implicitly metatheatrical status as play-within-the-play and since, as the title indicates, the role of the eunuch is somehow meant to be the play’s focus.51 As the ninth-century commentator Pseudo-Servius acutely observes:

contextually appropriate possibility. For example, Tromaras (1994, ad loc.) reads id quod moechis solet as a ‘gelungene Anspielung auf die ῥαφανίδωσις’. Barsby (1999a, ad loc.) acknowledges that ‘the reference here is probably to castration’, and Brothers (2000, ad loc.) likewise says ‘generally in comedy the punishment was castration’, but neither relate this observation to the character or the title of the play. 49

Note that the rape itself was framed, in part, as an act of separation from the eunuch role. If he had not raped her, he says, tum pol ego is essem vero qui simulabar, (‘then I would have been in truth the person I was pretending to be’, 606). The irony, according to my reading, is that he foreshadows the danger of the eunuch role becoming real, but this possibility only encourages him to do the very thing that will get him into trouble. 50 See p. 1. 51 There is also a real eunuch, Dorus, who could, at least in principle, be the play’s eponymous hero, but no one has been eager to attribute the play’s name to him. So Donatus’ commentary on the play begins: Haec masculini generis nomine nuncupata est Eunuchus fabula et est palliata Menandri vetus, quam ille auctor de facto adulescentis, qui se pro eunucho deduci ad meretricem us est, nominavit (‘This play is called Eunuchus (noun, masculine) and is an old New Comedy by Menander, who named it after the deed of a young man, who suffered himself to be led to a prostitute in a eunuch’s place’).

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN TERENCE’S ROME

117

Eunuchus dicta est fabula a quodam falso eunucho dicto Chaerea qui ductus in domum Thaidis meretricis pro eunucho vitiavit quandam virginem Atticam civem; iste talis eunuchus qui vitiavit illam virginem est principalis materia in hac comoedia; si aliae personae inducantur eunucho sviuntur, et omnes partes fabulae aliquo modo ad eunuchum tendunt.52 The play is called Eunuchus from a certain false eunuch named Chaerea who was led into the house of the prostitute Thais as a eunuch and raped a certain girl, an Athenian citizen. This ‘eunuch’ (so to speak) is the chief subject of the comedy. If other characters are brought on, they are auxiliary to the eunuch, and all the parts of the play in some way point back to the eunuch.

The theme of mimetic contagion around the painting is most manifest in Chaerea’s performance, but it is ultimately only realized here in a temporary way, first symbolically, in his inability to terminate the role without Thais’ permission, and then imaginatively, in Pythias’ lie. To find a more permanent embodiment of the theme of a performance taking on life of its own and becoming reality, we must look to Thais, the only other character around whom roleplaying is framed in such clearly metatheatrical . At the end of Act 1 she swears that she really loves Phaedria alone, and that her apparent affection for the soldier, Thraso, is all an act for the purpose of saving Pamphila from his clutches. She only needs one small favour from Phaedria: that he make himself scarce for a couple days and let Thraso take top billing. sine illum priores partis hosce aliquot dies | apud me habere (‘Let him have the leading role around here for a few days’, 151–2). This use of theatrical language to describe the lover’s role (priores partis) has long been noted by commentators.53 Later, when Chaerea asks who is so powerful as to give Pamphila to Thais as a gift, Parmeno identifies Thraso by stock comic labels: miles Thraso | Phaedriae rivali’ (‘The soldier Thraso, Phaedria’s rival’, 353–4). Chaerea picks up the theatrical metaphor and responds in kind: duras fratri’ partes praedicas (‘You’re talking about a hard role for my brother’). The theatrical dimension of this second reference to the lover’s partes has been noted by commentators since Donatus.54 Presented with Thais’ appeal, Phaedria is hardly inspired with confidence, but he grudgingly promises the two days’ absence she requests and departs with

52

Unfortunately there is no recent edition of Pseudo-Servius. His commentary was often printed in the sixteenth century and, along with Donatus’, had an incalculable influence on early modern readings of Terence, but once he was no longer confused with the fourth-century Servius his commentary generally lost its authority or at least its claim on modern attention. The age I have quoted is from 53R in the 1511 Venetian edition of Terentius cum quinque commentis. 53 See Ashmore 1910, Tromaras 1994, ad loc., Barsby 1999a, ad loc., and Frangoulidis 1994a, 127. 54 Donatus (ad 354): ‘partes duras’ μεταφορικῶς ab actoribus scaenicis. See also Tromaras 1994, ad loc., Barsby 1999a, ad loc., and Frangoulidis 1994a, 128.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

118

MIMETIC CONTAGION

Parmeno. Left onstage alone, Thais expresses her concern about her relationship with Phaedria in an intensely metatheatrical soliloquy (197–201): me miseram, fors[it]an hic mihi parvam habeat fidem atque ex aliarum ingeniis nunc me iudicet. ego pol, quae mihi sum conscia, hoc certo scio neque me finxisse falsi quicquam neque meo cordi esse quemquam cariorem hoc Phaedria. Oh dear, I’m afraid he doesn’t trust me And is judging me now based on other women’s characters. Listen, I know myself and I also know this for a fact: Neither have I fabricated any falsity, Nor is anyone dearer to my heart than Phaedria.

Her asseveration that she has not pretended anything false (neque me finxisse falsi quicquam) may be true with respect to what she has just said to Phaedria, but it is not literally true, for by her own ission she has been dissembling with Thraso.55 Moreover her trickery has taken the form of assuming the character traits of a comic meretrix. Thraso knows about her relationship with Phaedria (cf. 137–8), but she has pretended that her love is totally venal, so that he may hope to regain her affections with an expensive gift. She is, in effect, playing her own stock character. Thais recognizes that this is a dangerous game. Her very name gives her away as a meretrix in a Roman comedy, for it is as formulaic as her costume and other features of her stage identity.56 To these external characteristics she has added the Frangoulidis (1994a, 127) also notes the metatheatrical significance of this line, but his reading is somewhat different from mine: ‘Thais tries to assure Phaedria of the sincerity of her feelings for him and of the truth of her story with regard to Pamphila: neque me finxisse falsi (200). In a performance context, such an assertion is important, for it highlights Thais’ insincere feelings for the soldier and thus, underscores the fictional and, therefore, theatrical dimension of her intrigue.’ Thais cannot be trying to assure Phaedria of anything, since he is not present and this line is spoken to the audience alone. I agree that her protest of total honesty points back to her intrigue with the soldier and highlights its theatricality, but I do not think it points back only by contrast, but is itself performative and, strictly speaking, untrue (falsi quicquam). 56 Terence has pointedly changed her name to Thais. Menander’s original called the analogous character Chrysis (schol. on Pers. 5.161–75). The name Thais seems to connote more than just a hetaera; she is the archetypically devious, greedy man-trap of Greek New Comedy. Such at any rate is her function when she is invoked by Propertius, Ovid, and Juvenal. See Traill 2001. Menander’s play Thais was one of his most famous, apparently because of the unforgettable title character. In the prologue he gives a summary of her wiles, preserved for us by Plutarch (Mor. 19a): 55

ἐμοὶ μὲν οὖν ἄειδε τοιαύτην, θεά, θρασεῖαν, ὡραίαν δὲ καὶ πιθανὴν ἅμα, ἀδικοῦσαν, ἀποκλείουσαν, αἰτοῦσαν πυκνά, μηδενὸς ἐρῶσαν, προσποιουμένην δ᾽ ἀεί. Sing to me, goddess, of such a woman: Brazen, young and charming too, Breaking the rules, locking men out, always making requests, Loving no one, but ever pretending to.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

MIMETIC CONTAGION IN TERENCE’S ROME

119

pretence of being a woman of easier virtue than she really is, so she may well fear that her deception will prove more global than she intends. When she worries that Phaedria will judge her ‘from the characters of other women’ (ex aliarum ingeniis) she once again shows her keen awareness of the stock theatricality of her role, even as she insists that there is an ethical gap between her persona and herself.57 As Donatus notes here, the fascination of her character for the audience rests on her both being and not being representative of the standard meretrix.58 From Phaedria’s perspective (as clearly also from Parmeno’s), it is difficult to avoid the conclusion that she is simply a typical mercenary prostitute from Roman comedy, playing one man against another to get maximal lucre from both. By the end of the play, her fears have been realized. Phaedria has given up on true love and reached the conclusion that he cannot afford to keep her himself. He is essentially persuaded to pimp her out part-time to Thraso to help feed her appetite for gifts, which he has come to believe is insatiable. We shall explore how this happened in greater detail in Chapter 6. For now we only note the similarity between the hardening of her role into reality and the dynamic of the painting. She thought she could flirt with her own stock character, pretend, for a while, to have the venal disposition one would expect from a prostitute in a Roman comedy, and then step out of her pretence as soon as she got Pamphila safely away from Thraso. But by the end of the play the other characters have all agreed that she really is that mercenary meretrix, and they have made arrangements on her behalf to insure that she will be stuck in this role for good. What is merely suggested in the case of Chaerea, actually happens for her; she becomes the thing she has pretended to be. The artifice of her deception es outside its intended boundary and stops being make-believe, like the painted rape that steps out of its own frame to become real in the lives of the boy and girl who gaze up at it.

Note that our Thais is (or seems to be) guilty of every one of these things, indeed this last line is just what is at issue here. 57 I confess that this turn in Thais’ self-presentation always reminds me of Jessica Rabbit’s sultry protest to Valiant: ‘I’m not bad . . . I’m just drawn that way.’ Cartoon characters may not lay claim to the threedimensionality of ethical depth; there can be no real self, hidden beneath the flat surface of how they look and sound. Even trying to distance herself from her own appearance, Jessica is constrained to speak in Kathleen Turner’s voluptuous contralto, so that her protestation of innocence sounds as naughty as she inevitably looks. Similarly, despite Thais’ claim that she is not the stereotypical comic meretrix she appears to be, her language even in this scene with Phaedria and Parmeno, is notably in keeping with the conventions of her stock role (see Barsby 1999a, ad 87, 128, 130). 58 hic Terentius ostendit virtutis suae hoc esse, ut pervulgatas personas nove inducat et tamen a consuetudine non recedat, ut puta meretricem bonam cum facit, capiat tamen et delectet animum spectatoris (‘Here Terence shows it is within his capability to stage hackneyed characters in a novel way and yet not depart from the usual pattern, like making a “hooker with a heart of gold,” while capturing and delighting the mind of the spectator’).

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

FIVE

Mimic Rape Genre Switching and Role Confusion

induxit μίμησιν. dramatice more suo, non contentus dicere quod pollicitus sit tantum, sed quomodo etiam et quibus verbis.

CHAEREA ALIQVID INVENI

Donatus (ad Eun. 308)

Eduard Fraenkel’s Plautinisches im Plautus was, as any candid reckoning of the field today would probably it, ‘the most influential treatment of Plautus in the twentieth century, and is still the most important single book devoted to him’.1 Like most Plautus scholars coming out of the ninetenth-century German tradition, and more particularly as a student of Friedrich Leo, Fraenkel was fundamentally occupied with the question of the relationship between the Roman plays and their Greek originals. What most distinguishes Fraenkel’s project from the work of his teacher and colleagues is a simple difference in focus. As the title of his study suggests, Fraenkel was less interested in the elaborate reconstruction of Greek New Comedy for its own sake than in the appreciation of the peculiarly Plautine elements in Plautus’ plays. Of course, this line of inquiry still cast him back on the often hopeless task of divining details about lost comedies, and arguably Fraenkel’s most lasting contribution is to be found in his brilliant local readings and insights about individual plays, rather than in any grand scheme, but his reorientation of focus on to Plautus as a poet still continues to reverberate in the field.2 One aspect of how Fraenkel ed for Plautus’ departure from his Greek models was by reference to the largely forgotten world of pre-literary, improvisational Italian folk theatre. Fraenkel’s invocation of native Italian theatrical forms is global rather than specific; he uses broad notions of what these traditions were like in order to explain some of the recurrent features of Plautus’ comic landscape, tone, and style of invention.3 This somewhat vague suggestion of the probable influence of Italian folk theatre stood fallow for many years, but recently,

1 2 3

This quote is from the preface (xi) of Drevikovsy and Muecke’s new English translation (2007). For a fuller appraisal of the permanent significance of this shift, see Chapter 6. See, for example, Fraenkel 2007, 264.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

GENRE SWITCHING AND ROLE CONFUSION

121

following the orality turn of the 1990s, exponents of the ‘Freiburg school’ have begun to take up the challenge of reading individual ages through this lens, looking for scenes in Plautus’ plays where the influence of earlier performance traditions may be discernible.4 These various forms of non-literary Italian theatre all seem to have been based on unscripted improvisational entertainment. Plautus, too, loves to stage scenes where the characters are in confusion and we watch them respond on the fly to rapidly changing contingencies. Of course, Plautus’ plays are fully scripted literary comedies and can therefore be only pseudo-improvisational theatre, but the Roman audience’s taste for such scenes may well have been conditioned by their familiarity with pre-literary native traditions.5 It is generally assumed that Greek New Comedy was uninterested in drawing on the energies of improvisational theatre, so these moments in the Plautine corpus are readily taken to be his contribution. However valid this assumption may be for Plautus, it has erected a barrier against the importation of the same type of reading into Terence, who is understood to be far closer to his Menandrian model than to the native comic spirit.6 I will argue that the description of the rape in the Eunuch evokes a scene from mime, not for the audience’s visual entertainment, since indeed it only ‘happens’ backstage, but in order to comment on the generic boundary of Terence’s art.

COMEDY’S OTHER

The scholarship on Plautus’ debt to native Italian theatre generally draws indiscriminately on evidence of Fescennine, satura, Atellan farce, Phlyax, and the genre known properly as mime. An implicit assumption underlying this practice is that these theatrical types were all essentially regional variations of a single performance genre, but there is little reason to suppose that such diverse traditions constituted

4 Much of this scholarship has been published in the ScriptOralia series of the Gunter Narr Verlag in Tübingen. Most volumes in this series have been devoted to individual plays, for example, Menaechmi (ScriptOralia, 11), Captivi (ScriptOralia, 74), Pseudolus (ScriptOralia, 101), Amphitruo (ScriptOralia, 116), Persa (ScriptOralia, 121), Aulularia (ScriptOralia, 122), Epidicus (ScriptOralia, 125), Poenulus (ScriptOralia, 127), and Cistellaria (ScriptOralia, 128). But a few of them have had a more generally Plautine scope, for example, ScriptOralia, 25 (=Lefèvre, Stärk, and Vogt-Spira 1991), ScriptOralia, 73 (=Lefèvre 1995), ScriptOralia, 75 (=Benz, Stärk, and Vogt-Spira 1995). Outside the ScriptOralia series, see Vogt-Spira 1995. 5 ‘Hält man sich also vor Augen, daß diese Stegreifspiele keineswegs schlagartig verschwinden, wird die konkrete Rezeptionssituation klar, auf die die Palliatendichter treffen’ (Vogt-Spira 1995, 72). 6 Hunter 2002 is a notable exception. He argues for the presence of ‘mimic elements’ in Menander’s Dyscolus and Terence’s Adelphoe, and on some points there is clear similarity between our approaches. Eckard Lefèvre has also been consistently interested in the influence of Stegreiftheater on Terence, a theme that runs through his Zetemata series, now at last (as of 2013) covering all six plays.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

122

MIMETIC CONTAGION

a monolithic substrate of theatrical experience or that they formed a genuinely undeviating cultural backdrop for fabula palliata and togata. The barest survey of evidence demonstrates that these theatrical forms were as variegated as their provenance within Italy.7 Several factors do, however, justify allowing these various traditions to bleed into one conceptual category when attempting to estimate their impact on Roman comedy. First, despite the differences between these types of theatrical entertainment, there are undeniable similarities. They are all rough, bawdy, and apparently improvisational, and their humour consists largely of ridicule, either of specific people or of generalized rustic types. Second, they were not necessarily stable genres, immune from each other’s influence, especially in the south, where troupes of travelling performers could easily have cross-pollinated the Atellan farce of the Oscans and the Phlyax plays and Greek mime of Magna Graecia.8 This is just the kind of dissemination and confluence implied in Livy’s of the beginning of ludi scaenici at Rome. Formalized theatrical entertainment was, according to Livy, introduced in 363 BCE by a troupe of Etruscan dancers summoned to Rome, ‘a novelty for a warlike people’ (nova res bellicoso populo, 7.2.3). The exchange between foreign invention (peregrina res, 7.2.4) and domestic adaptation begins with the modified imitation of Etruscan dumb-show by the Roman youth (7.2.5–7): imitari deinde eos iuventus, simul inconditis inter se iocularia fundentes versibus, coepere; nec absoni a voce motus erant. accepta itaque res saepiusque usurpando excitata. vernaculis artificibus, quia ister Tusco verbo ludio vocabatur, nomen histrionibus inditum; qui non, sicut ante, Fescennino versu similem incompositum temere ac rudem alternis iaciebant sed impletas modis saturas descripto iam ad tibicinem cantu motuque congruenti peragebant. Then the youth began to imitate them, showering each other with wit in crude verses, and their movements were suited to their words. Thus was the custom adopted and kept up by frequent practice. The Etruscan word for player was ister, so the name histriones was given to native artists, who did not, as before, take turns tossing off something clumsy, random, and rough, like a Fescennine verse, instead they performed saturae full of metres, with singing now accompanied by flute and movements to match.

However accurate the particulars of Livy’s , he gives a clear sense of the fluidity of the early Italian theatrical traditions, in which genres could move from one culture to another and absorb or replace existing forms of entertainment, while undergoing significant modification themselves. 7 For a successful recent attempt to assemble a coherent picture of preliterary Italian theatrical entertainment, see Beacham 1992, 1–26. For a totally different view of the same material, Beare 1950, 10–23 is still worth considering. 8 Physical humour travels well and is not stopped by language barriers. Note the international popularity of ‘Charlot’, the export version of Charlie Chaplin’s Tramp.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

GENRE SWITCHING AND ROLE CONFUSION

123

Thus was satura conceived in Rome under the influence of an Etruscan style of theatre, very likely itself an adaptation of the Greek satyr play.9 However, satura was not born into a vacuum, rather it seems to have absorbed the social and performative functions of the Fescennine verses, a bawdy, rustic, and sometimes aggressive form of semi-theatrical festive poetry, easily compared with Greek phallica and iambic aischrologia.10 In taking over these functions, satura altered the economy of Roman poetic genres and helped render Fescennine obsolete, but it was not immune to alteration or replacement in its own turn. Livy’s narration goes on to explain that Livius Andronicus was ‘the first to dare depart from the saturae and compose a play with a plot’ (ab saturis ausus est primus argumento fabulam serere, 7.2.8). For some 120 years, then, saturae had been the theatrical standard, until 240 BCE, when Livius Andronicus first staged Greek tragedy and comedy in Latin translation (fabulae). To save his voice, Livius is supposed to have given the songs (cantica) to a singer, who stood in front of the flute player, keeping only the dialogues (diverbia) for himself. This left him free to dance and gesticulate more vigorously during the songs, but it could only have been done because fabula, highbrow theatre as he was inventing it for the Romans, was an entirely scripted form of entertainment.11 This literary transformation and professionalization of acting resulted in the withdrawal of the Roman youth from staged show business and a revival of aristocratic song (7.2.11–12):12 postquam lege hac fabularum ab risu ac soluto ioco res avocabatur et ludus in artem paulatim verterat, iuventus histrionibus fabellarum actu relicto ipsa inter se more antiquo ridicula intexta versibus iactitare coepit; unde exorta quae exodia postea appellata consertaque fabellis potissimum Atellanis sunt; quod genus ludorum ab Oscis acceptum tenuit iuventus nec ab histrionibus pollui a est; eo institutum manet, ut actores Atellanarum nec tribu moveantur et stipendia, tamquam expertes artis ludicrae, faciant. When, by this ordinance of the plays, show business came to be distinguished from clowning and broad humour, and drama had gradually turned into an art form, the

9 Wiseman 1988 presents a convincing case that the Etruscan form was a direct development from satyr plays. 10 On the general character of the Fescennines, see Horace Epist. 2.1.139–50. 11 Livy seems to be the only source for the claim that the jobs of actor/speaker and singer were ever split between two performers (unless Isid. Etym. 18.44 is to be taken as independent testimony, which it probably should not be), and most scholars have rejected the idea as unlikely. Beare (1950, 219) postulated that Livy was ‘thinking of some form of theatrical performance of his own day’, which may be right, or he may be faithfully ing on a tradition about Livius Andronicus that he learned from Varro. Whatever the historical veracity of this of Livius Andronicus, the salient point for our investigation is that Livy imagines the introduction of fabula as the moment when it became impossible for performance to be fully improvisational, if for no other reason than because the division of labour between two performers for the same part would now require the lines to be set. 12 For the broader context of this shift and its place in Republican constructions of writing and song, see Habinek 1998, 34–68.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

124

MIMETIC CONTAGION

youth abandoned the acting of playlets to professionals and returned to their old custom of bantering among themselves in unscripted poetic jests. From this arose the pieces later called exodia and generally worked up into Atellan farces, a type of drama derived from the Oscans, which the youth kept for themselves and did not allow to be polluted by professional actors. And so it remains an established rule that actors in Atellan farces are not kicked out of their tribe and they do military service as if innocent of the dramatic art.

Livy’s testimony here is fascinating, for it locates a bifurcation of Rome’s theatrical culture in the wake of the introduction of scripted plays. The new fabulae drive the older saturae offstage and presumably into productive retirement, as they will undergo an evolutionary course resulting in the later Roman genre of satire. The aristocratic Roman youth are presented as the guardians of an ancient tradition of unscripted blame poetry, which they preserve, paradoxically, by transforming it first into satura and then into ‘after pieces’, stitched together as Atellan farce. Once again the fluidity of these traditions, at least in their Roman reception, is impressive, for there is no reason to suppose that the Etruscan entertainment that inspired satura had much in common with the Oscan Atellan, but these diverse forms are framed as the successive vessels of an essentially uniform type of Roman poetic performance. Indeed, the only strong dichotomy in Livy’s is between fabula and everything else; fabulae are scripted, refined, and professionally acted, whereas Fescennine, satura, and Atellan are unscripted, crude, and amateur. Where exactly mime fits in this mix of Italian theatrical forms is difficult to determine, partly because of the obscurity of its early history and partly because it seems always to have been the most flexible style of entertainment. The genre known properly as mime came to Magna Graecia with the Dorians, where it not only flourished but proliferated in types.13 We do not know when or in what form mime first came to Rome. It may have been there long before the introduction of scripted plays in 240 BCE or it may have arrived later, but the earliest known Roman actor, who may be dated to the period just before the Eunuch (c.210–160 14 BCE), was a mime. Across its history, mime seems to have been generally improvisational but could lend itself to scripted performance.15 There were

13 For the early history of mime the fullest survey (more compact and less speculative than Reich) is Wüst 1932, cols. 1729–33. A good recent treatment may be found in Panayotakis 2010, 1–32. 14 Protogenes, slave of Cloulius; for inscription and analysis, see Rawson 1985, 111. 15 One of the early types of mime actors are called autokabdaloi, a name that emphasizes the improvisational nature of their performance (see Pickard-Cambridge 1962, 137–8), but they were by no means uniquely unrehearsed. Alongside this dominant practice of ad hoc performance, however, mimographoi or mimi scriptores are attested from an early date. Some important early names from this tradition are Epicharmus, Herodas, and Sophron. For a survey of the evidence on later mime writers, see Nicoll 1963, 110–15.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

GENRE SWITCHING AND ROLE CONFUSION

125

probably both amateur and professional forms.16 It included mythological burlesque but also portrayed everyday people.17 There are certain recurring character types (prostitutes, runaway slaves, adulterous couples, cuckolded husbands), but though names may sometimes have been repeated, mime was not limited to a fixed set of four named stock characters, as Atellan farce was.18 Also unlike Atellan, mime did not generally use masks, although Phlyax, which did use masks, was regarded by ancient writers as an Italian species of mime, and there is evidence of later mime occasionally using masks.19 Unlike any other form of ancient theatre the female parts could be played by actresses, but they could also be played by male actors.20 In short, mime was as anarchically free in its absorption of other styles of theatre as it was in its presentation of human folly, making it so capacious a category that it would eventually assimilate and supplant all other forms of theatrical entertainment.21 This displacement was far from immediate. Atellan farce dwindles but it is still attested as late as the second century CE, though given the evident plasticity of Atellan and the overwhelming popularity of mime in the Imperial period, it may be doubted whether by this time ‘Atellan’ was anything more than an exotic or archaic sounding name for mime.22 Cicero could still make a distinction between Atellan farce and mime, but only to point out that where once Atellan was performed, the norm was now mime (Epist. ad fam. 9.16.7). Just as in Livy’s of theatre in the third century BCE the important dichotomy was between scripted fabula and everything else, for Cicero the salient distinction is between fabula and mime. So, famously, in the Pro Caelio he summarizes Clodia’s crazy story about a plot to murder her, tricky slaves who agree to help Caelius poison their mistress and then double-cross him, a counter-plot to entrap Licinius, and a melodramatic scene at the baths with Licinius fumbling for the poison box and then mysteriously escaping: mimi ergo iam exitus, non fabulae; in quo cum clausula non invenitur, fugit aliquis e manibus, dein scabilla concrepant, aulaeum tollitur (‘So the denouement is not like in a proper play but in a mime, where when they can’t figure out how to end it, someone just slips away, then it’s cue the applause and . . . curtain!’, Pro Cael. 65). Clodia’s story is not only obvious fiction,

16 As with other species of theatre, the professionalization of mime is obscure and probably political, in the sense we have seen, that aristocratic privilege may insist on retaining amateurism for some forms. For example, Athenaeus (14.621f) testifies that there was a Theban type of mimic performers called ethelontai. 17 On the ‘literary realism’ of mime, see Reich 1903, 19–38. On burlesque, see pp. 145–6. 18 The four characters of Atellan were Pappus, Maccus, Bucco, and Dossennus (or Manducus). 19 On Phlyax as ‘Italian mime’, see Athenaeus (14.621f) and Nicoll 1963, 50. 20 The most compact collection of the testimony on mimae is Nicoll 1963, 92–9. 21 On mime in late antiquity (and its relationship to pantomime), see Webb 2008, 95–138. 22 Tacitus notes that under Tiberius Atellan farce had become so degraded and influential (eo flagitiorum et virium venisse, Ann. 4.14) that it had to be suppressed.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

126

MIMETIC CONTAGION

but tawdry pulp fiction as well, more like lowbrow mime than highbrow fabula.23 The two types of stage entertainment are implicitly opposed and mutually exclusive; there is apparently no significant tertium quid of theatrical genres. The standard practice, since Fraenkel, of treating the various native Italian theatrical traditions as a uniform backdrop for scripted comedy is, therefore, still broadly defensible, in part because there was a degree of homogeneity across these traditions, especially as they coalesced into mime, but more importantly because the Romans themselves evidently regarded the (perhaps artificially) consolidated category of mime as a stable Other for the polished world of scripted comedy and tragedy. In Bakhtinian , this dichotomization of generic space into a relatively simple highbrow/lowbrow binary would permit a rich dialogism to flourish within either mime or comedy, since a dramaturgically small gesture would be enough to suggest the excluded generic Other. Whether rightly or wrongly, the scholars who have investigated the sub-literary elements in Plautus’ plays have not generally allowed for the possibility of such a self-consciously bright line between these two modes of theatre. The analysis of Plautus’ plays into two strata, reserved Greek New Comedy and unscripted Italian screwball, may privilege either stratum, but it has invariably assumed that this analysis itself is a modern scholar’s game, rather than being part of the fun for the original audience. In Plautus’ case this assumption may be right, for he includes stock lazzi, heroic improvisation, and other routines from mime so frequently that they may not be perceptible as a break in the generic texture of his plays. Terence, however, precisely because he stays so much closer to the tone of his comparatively reserved Greek models, can count on the unexpected inclusion of an alien element from the world of mime to be felt as an irruption into his play.

UN-TERENCE IN THE PROLOGUES

Readers of Terence, from Cicero and Caesar to modern scholars, have always noted that he is more staid than other Roman comic poets.24 However, this 23

Cicero may be relying on recognizable associations appropriate to mime, such as the use of real meretrices to play female roles and the poisoning-scene as a standard motif. See Austin 1960, 128. 24 Because Plautus and Terence are so different, scholars sometimes tend to think of Roman comedy consisting of two broad types, but Wright 1974 shows convincingly that the fragmentary comic poets shared Plautus’ theatrical aesthetics. Considering the genre as a whole, Terence was the odd man. Suetonius’ Vita Terenti (7) reports both Cicero’s and Caesar’s epigrams. Cicero’s high praise of Terence does not seem to be undercut by his emphasis on the comic poet’s restraint (sedatis motibus) and his general ‘niceness’ (quiddam come loquens atque omnia dulcia dicens), but in Caesar’s epigram the same features significantly attenuate the commendation. In the first line, Terence is addressed as ‘half-pint Menander’ (o dimidiate Menander), and Caesar goes on to wish that greater ‘energy’ (vis) had been ed to Terence’s mild verse (lenibus . . . scriptis). ‘Then,’ says Caesar, ‘your comic vigor might have equal potency with the Greeks and you would not be

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

GENRE SWITCHING AND ROLE CONFUSION

127

universally held judgment is not, in the first instance, a product of disinterested literary history, but of poetic self-presentation, for it can be traced back to Terence himself. In the prologue of the Heautontimorumenos, Ambivius Turpio makes a distinction between Terence’s sedate play and a recent play by Luscius Lanuvinus that features a servus currens scene (31). He asks the audience to pay attention and keep quiet so he can perform this demure piece (statariam agere, 36).25 Ambivius is getting to be an old man, and if this one flops he will have to go back to playing running slaves, angry old men, hungry parasites, brazen tricksters, and greedy pimps, roles that require clamore summo and labore maxumo, as opposed to this Terence stuff which is ‘all about the script’ (in hac est pura oratio, 46). In this age Terence evokes, in some detail, a type of theatre notionally different from his own. The prologus does not refer to this alternative style of theatre as ‘mime’, but he does insist on a typological difference between Terentian comedy and other Roman comedy, exactly parallel to the difference we have seen other Roman writers drawing between fabula and mime. Mime’s place within the insistently dichotomous economy of Roman theatrical genres was such that anything identifiably other than New Comedy could be antithetically characterized as ‘mime’ or at least ‘mimic’. For example, when Aulus Gellius complains that Caecilius replaced the quiet subtlety of his Menandrian model with an alien farcical buffoonery (quae Menander praeclare et apposite et facete scripsit, ea Caecilius . . . quasi minime probanda praetermisit et alia nescio quae mimica inculcavit, NA 2.23.12), it is mime that provides Gellius his reference point for theatrical high jinks. As in the age from Pro Caelio above, mime and

scorned as a failure in this department! This is the one quality, Terence, I’m distressed and grieved that you lack’ (comica ut aequato virtus polleret honore | cum Graecis neve hac despectus parte iaceres! | unum hoc maceror ac doleo tibi deesse, Terenti). 25 In Brut. 116 Cicero uses the adjective statarius as a theatrical technical term for an actor in such a play, explicitly contrasted with the more agitated type of performance: ‘As on stage so also in the forum I don’t want to see only those men praised who display quick and taxing bustle, but also the ones they call statarii, whose acting has that straightforward and unaffected genuineness’ (volo enim ut in scaena sic etiam in foro non eos modo laudari, qui celeri motu et difficili utantur, sed eos etiam, quos statarios appellant, quorum sit illa simplex in agendo veritas, non molesta). Again in 239 he applies statarius to C. Piso, an orator not given to extreme gesture. Horace does not use this term, but the contrast is clear in Ep. 2.1.58–9, where Plautus scurries (properare) like Epicharmus, but Caecilius and Terence top their models in gravitas and ars, respectively. Evanthius explains that there are three types of comedies: motoria, stataria, and mixta. motoriae turbulentae, statariae quietiores, mixtae ex utroque actu consistentes (De com. 4.4 p. 22). Donatus uses the same constantly; see, ad Andr. praef. 1.2; ad Eun. praef. 1.2; ad Ad. praef. 1.2, 24.2, 299; ad Hec. praef. 1.2. Terence never uses the word motoria, and it seems unlikely that this juxtaposition of technical would go back as far as the second century BCE, but of course there is no way to know. It is more readily conceivable that this use of the word stataria in the Heautontimorumenos inspired the later terminological development of motoria for the other half of the dialogism. Even if stataria does not have a technical valence for Terence, it is striking that the later grammatical tradition would utilize his language, and the opposition implied in it, as the basis for its typology.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

128

MIMETIC CONTAGION

fabula constitute a binary generic space, but here it is not a question of classifying a whole story as fabula or mime, rather a drop in tone within a translated Greek New Comedy is described as an intrusion of ‘mimic elements’ (nescio quae mimica), though of course Caecilius’ play as a whole will remain a perfectly normal specimen of fabula. So the servus currens scenes and related stock physical elements of Plautine and other Roman comedy, which Ambivius contrasts with Terence’s type of theatre, may not be mime in the full sense, but they are alien to Terence’s aesthetic precisely as these mimic elements are alien to Menander. Terence is generally spare in his inclusion of such elements in his plays, except that he has, if only in praeteritio, injected them in narrative form (pura oratio) here in the prologue of the Heautontimorumenos. It would have been easy enough not to mention this other theatrical world; instead, it is made to seem perilously close. These same actors have acted in such plays and will do so again, if and when this production fails. The fragility of the production is probably the most consistent single theme across the prologues, and in every case it may be wondered what practical purpose was served by drawing attention to the possibility of the show being overwhelmed by disruptive events outside the play. As eager as we are for evidence of the performance conditions of these comedies, there is little reason to trust Terence’s prologues as a reliable record, for if these productions were really as vulnerable to interruption as he indicates, he would surely have little incentive to the fact. The most spectacular example, of course, is the prologue(s) of the Hecyra, which apprise us of two earlier aborted performances. The first prologue claims that an earlier showing fell apart when a tightrope walker bedevilled the attention of the dumbfounded audience (populu’ studio stupidus in funambulo | animum occuparat, 4–5). The second prologue returns to this event and renders its chaos in fuller colours (comitum conventu’, strepitu’, clamor mulierum, 35), and it goes on to tell that the second performance came off no better than the first. The play seemed to be going well (primo actu placeo, 39), when a rumour of gladiators resulted in general mayhem (populu’ convolat, | tumultuantur clamant, pugnant de loco, 40–1). Centuries of readers have pitied Terence having to work under such conditions and with such an unappreciative audience, but Erich Gruen has advanced a forceful argument that the prologues are a fiction; there were no aborted earlier performances.26 The joke here, for if Gruen is right the prologues Gruen 1992, 210–15. Even if Gruen’s explanation is rejected and Terence’s narration of events is taken at face value, we need not imagine (with Beare 1950, 161 or Duckworth 1952, 81) that the original audience simply lost interest and wandered off after something flashier. Rather, the prologus seems to suggest that a second group of entertainers and spectators showed up to crowd out the show in progress, perhaps implying that the disappointed original audience was thwarted from seeing the Hecyra; see Gilula 1978; Sandbach 1982; Parker 1996, 592–601. On the other hand, the description may suggest that the original audience ed in the tumult; see Lada-Richards 2004, 58–9. 26

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

GENRE SWITCHING AND ROLE CONFUSION

129

must be a kind of joke, draws on the same sense of humour as motivated L. Anicius Gallus’ outrageous production in 167 BCE, in which offstage Roman boorishness is carefully juxtaposed to, and allowed to trample on, highbrow Greek divertissement.27 Gallus entertains his audience, both by thumbing his nose at the cultivated artistry of the Greek musicians and by poking fun at a caricature of Roman crudeness so extreme as to assure the audience that it is on the right side of the laughter. To be sure, not all scholars have concurred with Gruen’s expansive reading of Roman ironic self-consciousness, either in the Hecyra or in Gallus’ follies.28 However, Gruen’s approach to the prologue of the Hecyra does conform to a recent trend towards reading Terence’s prologues as playfully ironic dramatic poetry, not totally divorced from the fictionality of the comedies they introduce.29 The prologue of the Phormio closes with an onition to the audience to attend in silence, otherwise it may be like the time the company got swept offstage by an upheaval (quom per tumultum noster grex motus locost, 33). This has generally been read as a reference to the failed first performance of the Hecyra in 165 BCE, but there is no explicit mention of the Hecyra here. If Gruen is right and there was no failed first performance of the Hecyra, then the prologue of the Phormio may simply be engaging in the same kind of humour, on a smaller scale, as the Hecyra prologues. Terence’s first transmitted play, from 166 BCE, is sufficient proof that

27

Our source for these events is Polybius (30.22), who sounds no less bewildered than dismayed at what he has to report. Not to be entirely outdone by Aemilius Paullus, who that same year celebrated a triumph of unprecedented magnificence, even staging elaborate Greek-style games at Amphipolis, Anicius reversed his rival’s strategy of cultural negotiation by bringing to Rome the finest musical performers Greece had to offer (τοὺς ἐκ τῆς Ἑλλάδος ἐπιφανεστάτους τεχνίτας, 30.22.2). He constructed a stage in the Circus and put all the flute players onstage playing together. Things seemed to be proceeding normally enough, the players moving sedately with their music, when Anicius demanded that they show more competitive spirit (ἀγονίζεσθαι μᾶλλον, 30.22.5). They were at a loss to understand, until one of the lictors explained to them that they should run at each other and make as if to fight. They did their best to comply, the chorus ed in and formed two armies, and before long the music had devolved into discord, as opposing sides advanced against each other and retreated across the stage. When one of the dancers took a boxer’s stance against an advancing flute player, the crowd went wild, and the whole production sank into general melee. Just then the prize fighters and buglers came in, along with more musicians and dancers, and Polybius’ powers of description desert him (ἄλεκτον ἦν τὸ συμβαῖον); he refuses even to tell us about the tragic actors, because he swears we would think he was joking (διαχλευάζειν, 30.22.12). Scholars have generally taken their cue for disgust from Polybius and have often assumed that this public brawl was a spontaneous departure from the musical program, perhaps due to the bored vulgarity of the Roman audience (see, for example, Crowther 1983, 270), but the boxers and buglers waiting conveniently in the wings are enough to suggest that this apparently riotous event was fairly well planned and remained under Anicius’ control. Gruen (1992, 215–18) has pointed out the production’s canny inversion of Aemilius’ Greek games in Greece earlier that same year. 28 Unconvinced about Gruen’s Hecyra: Parker 1996, 600; Moore 1998, 204–5; unconvinced about his Gallus: Forsythe 1994. These scholars express incredulity or assert that Gruen goes too far, but do not offer a specific explanation of their reservations. 29 Gilula 1989 is foundational, but see most recently Gowers 2004.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

130

MIMETIC CONTAGION

his interest in the theme of his own failure and expulsion from the stage predated any putative first flop of the Hecyra, for the prologue of the Andria ends with the open question for the audience, whether Terence’s future plays will have a hearing or be driven offstage before they are performed (spectandae an exigendae sint vobis prius, 27). The Eunuch comes closest to the Hecyra in actually narrating an earlier performance of the play that was begun and then interrupted. According to the prologue, Luscius Lanuvinus arranged for a preview performance of the Eunuch after it had already been bought by the aediles (perfecit sibi ut inspiciundi esset copia, 21). The official audience had arrived and the private showing had just got underway (magistratu’ quom ibi adesset occeptast agi, 22), when Luscius Lanuvinus broke in with theatrics of his own, crying (exclamat, 23) that it was a thief, not a poet, behind the boards. If Gruen’s claim that the Hecyra prologue is fictional rather than historical has left most critics unconvinced, perhaps it is imprudent to ignore their sense of the fundamental improbability of his thesis. Even if we dismiss the didascalic evidence as no more reliable (or independent of the prologues) than the Suetonian Vita, perhaps it does strain belief that a scenario as elaborate as this double-flop could be made up out of whole cloth and then delivered totally deadpan to the amusement of its original audience and the universal perplexity of future ages.30 It may be, however, that the chief problem with Gruen’s thesis is his reliance on a fictional/ historical dichotomy, as opposed to asking whether the inclusion of this narrative in the prologue of the Hecyra is rhetorically mobilized as part of Terence’s selfcharacterization vis-à-vis other forms of Roman entertainment, or whether, on the other hand, it is nothing more than an explanation for the play’s reappearance after a failed earlier production and a straightforward plea for attention at last. Reformatted in these , the question becomes both more readily answerable and more immediately significant, for even if there were earlier failed performances of the Hecyra, we may still ask why these events are given such focal prominence for the final performance. It is my contention that even if the story is completely true, it would not receive the emphasis it does unless its inclusion served Terence’s purpose of constructing 30

The Hecyra is unique in Terence’s plays in that it has two preserved didascaliae, one from the codex Bembinus (A) and a shorter version in the other manuscripts (Σ). Both versions report three performances, but they say slightly different things about them. The codex Bembinus does not mention any problems with the first performance but says the second performance was a failure (NON EST PLACITA, 8), whereas the Σ version says the first performance fell through completely (NON EST PERACTA, 3) but mentions no problems with the second performance. The two prologues purport to be from the second and third performances, and the A didascalia says the Hecyra was performed the first time without a prologue (5–6). This sounds suspiciously like an explanation for the missing prologue; in any case, we have no other mention of any Terence play being performed without a prologue. The same didascalia also tells us the Greek original was by Menander, where Donatus makes it clear that it was by Apollodorus. On the place of the Plautine and Terentian didascaliae in ancient scholarship on the plays, see Deufert 2002, 88–96.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

GENRE SWITCHING AND ROLE CONFUSION

131

a particular generic space for his plays, as pointedly vulnerable to displacement by rougher, more physical forms of entertainment from offstage.31 Still less would this theme recur without motivation in the Phormio, nor, as I have argued, would it be gratuitously adumbrated beforehand in the Andria, refashioned in a different narrative guise in the Eunuch, or framed as a fundamental opposition between two types of theatre in the Heautontimorumenos. In every one of these instances, of course, narration is its own form of inclusion. Terence’s claim of pura oratio does not mean monkey business has no place in his plays, rather it means that the rowdiness has all been reformatted as language. The prologus in the Heautontimorumenos, as ultimately also in the other plays, gives us a staid, discursive description of an alternative type of theatre not based on staid, discursive descriptions. This quarantined presence of the indecorous within the decorum of Terence’s theatre, as a scene conjured only with words and carefully located either temporally or spatially just offstage, is central to Terence’s aesthetic in the broadest sense.32 To be sure, these prologues never refer to the non-Terentian style of entertainment as ‘mime’, and I do not wish to claim that every instance of theatrical competition or interruption they describe should be understood as pointing to mime. I would rather situate the mimic elements of the rape scene within a larger Terentian pattern of play against a discretely framed lowbrow theatrical Other, which may at any moment overwhelm and displace the genteel comedy before our eyes.

IMITATIO VITAE AND VITALIS

Perhaps the feature of mime that relates most clearly to the rape scene in the Eunuch is obvious from the thesis of Chapter 4, that the painting’s effect on Chaerea is best understood as a form of mimetic contagion. The rape is framed as an imitation of an artwork embedded in a theatricalized play-within-the-play, and, as the word itself implies, imitation as play is the heart of ancient mime. The fourth-century CE grammarian Diomedes gives the fullest recorded definition of mime (Art. Gramm. Lib. III, p. 491 Kiel): For a similar approach to the Hecyra prologues, see Lada-Richards 2004. The contrast to the opening lines of Aristophanes’ Frogs is instructive. Xanthias wants to crack the vulgar old jokes that always get a laugh, a routine Dionysus is willing to permit to an extent, but he draws the line at the truly tired quips of Ameipsias, Lycis, and Phrynichus. The humour of this scene consists in the partial inclusion of these jokes, through ironic praeteritio. In any case, these poets are Aristophanes’ rivals with whom he was in direct and formal competition. In fact, in this very contest (the Lenaia in 405) the Frogs would take first place over Phrynichus’ Muses. This situation is obviously quite different from the challenges Terence describes facing from offstage physical entertainment that is totally different in kind from his theatre, not in formal competition with him, and, unlike a joke, impossible to incorporate partially in his play as Aristophanes does. 31 32

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

132

MIMETIC CONTAGION

Mimus est sermonis cuius libet imitatio et motus sine reverentia, vel factorum et dictorum turpium cum lascivia imitatio; a Graecis ita definitus, μῖμός ἐστιν μίμησις βίου τά τε συγκεχωρημένα καὶ ἀσυγχώρητα περιέχων. mimus dictus παρὰ τὸ μιμεῖσθαι, quasi solus imitetur, cum et alia poemata idem faciant; sed solus quasi privilegio quodam quod fuit commune possedit: similiter atque is qui versum facit dictus ποιητής, cum et artifices, cum aeque quid faciant, non dicantur poetae. Mime is an imitation of any kind of speech and irreverent movement or imitation with wantonness of shameful deeds and words. It is thus defined by the Greeks: ‘Mime is an imitation of life containing things both licit and illicit.’ The word mime is from the Greek word ‘imitate’, as if it alone were engaged in imitation, even though other genres do the same thing. Mime alone, however, has claimed as its special prerogative what was common to all. Likewise the man who makes verse is called ‘poet’, while other artists, though they are just as much makers, are not called poets.

Diomedes’ definition has been criticized for its failure to mention dance, songs, and music as important constituents of mime and its silence on other topics that interest us, such as mime’s origins in religious ritual.33 These omissions, however, can easily be explained by the fact that any kind of ancient theatre might have musical accompaniment and certainly had beginnings shrouded in early cult. In other words, these elements in mime were not constitutive of its distinctive code. The peculiar place of mime in the generic system of ancient theatre was determined by its name. Mime calls itself imitation, and though all poetry is in some way imitative, mime has a corner on the claim of preeminent imitation. As such it cannot be selective about the range of what it imitates, but must it all life as its subject. The embedded Greek definition, the source of which Diomedes unfortunately does not identify, makes mime’s imitation universal, accommodating both licit and illicit themes, but the Latin part of the definition makes it clear which type of material made a bigger impression (sine reverentia; turpium; lascivia).34 As we saw in Chapter 3, the Greek word mimēsis is notoriously difficult to pin down, and some of these ambiguities on to mime.35 One sense of mimēsis is as ‘representation’ (see p. 78), and it is clear from Diomedes’ definition that one of the ways in which mime is representative of life is in its range of s, a span from high to low no less diverse than the real world. However, mimēsis can also be closer to ‘mimicry’, and there are other ancient references to mime’s prodigious capacity for imitation. So Cicero, in De oratore, notes that there have been cases in Panayotakis 1995, xiii. An attractive possibility that has been put forward is that the Greek definition comes from Theophrastus, but there is no way to prove this and other suggestions have been made. See Giancotti 1967, 26–8. 35 The literature on the meaning of the word mimesis, as on everything else to do with the topic, is vast. Halliwell 2002 provides a good starting point and goes a long way towards bringing out the full complexity of the word in Plato and beyond. 33 34

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

GENRE SWITCHING AND ROLE CONFUSION

133

which vultus et vocis imitatio has been effective, but he advises the ideal orator to avoid visible imitation (De orat. 2.242): mimorum est enim et ethologorum, si nimia est imitatio, sicut obscenitas. orator surripiat oportet imitationem, ut is, qui audiet cogitet plura quam videat; praestet idem ingenuitatem et ruborem suum verborum turpitudine et rerum obscenitate vitanda. For excessive imitation, like obscenity, belongs to the mimes and character players. The orator should sneak imitation, so the audience imagines more than it actually sees. He should display his nobility and decency by avoiding shamefulness of speech and obscenity of gestures.

The ‘mimes’ and character players’ capacity for likeness’ (mimorum et ethologorum similitudo, 2.244) certainly has its place, he insists elsewhere, but there is a difference between laughable (ridiculum) and witty (facetum): quid enim potest esse tam ridiculum quam sannio est? sed ore, vultu, [imitandis moribus,] voce, denique corpore ridetur ipso; salsum hunc possum dicere atque ita, non ut eius modi oratorem esse velim, sed ut mimum. (2.251) For what can be as laughable as a buffoon? But he is ridiculed for his face, his expression, his imitation of character, his voice, in short for his very body. I can call him funny and he is, not of the sort as I would have an orator be, but as a mime.

Combining mime’s imitative quality with its improvisational freedom naturally means that mime could be responsive to its audience in ways scripted theatre could not. Many instances of this phenomenon could be cited; I single out here the often neglected epitaph of the mime Vitalis from the Latin Anthology (Riese 487a), which presents its subject as the mirror and vessel of his spectators (15–22): fingebam vultus, habitus ac verba loquentum, ut plures uno crederes ore loqui. ipse etiam, quem nostra oculis geminabat imago, horruit in vultus se magis isse meos. o quotiens imitata meos per femina gestus vidit et errubuit totaque muta fuit! ergo quot in nostro vivebant corpore formae, tot mecum raptas abstulit atra dies.36 I copied the faces, postures, and words of those talking, So you would think many were speaking from one mouth. Shackleton Bailey (1982, 116) has suggested emending crederes in 16 to crederer, thus preserving the metre and giving basically the same sense. However, the poem is presumably late enough that the length of this e is not important. Note, for example, the short i in the very same metrical position in line 2 (nescis amare iocos). Emending to the first person also diminishes the epitaph’s play with the shifting second person referent throughout. In lines 1–2 tu is death, in line 16 tu is the spectator, in lines 24–5 vos is the readers, and in lines 25–6 tu is Vitalis, thus balanced addresses to death and Vitalis at beginning and end of the poem frame invocations of his spectators while he was alive and then of his readers now that he is dead. 36

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

134

MIMETIC CONTAGION

The very man being twinned in his eyes by our image Shuddered at having ed more fully into my face. Often a woman, echoed by my body language, Saw and blushed and was struck dumb. So all the shapes who lived in our body Have been borne off with me by the black day.

Vitalis replicated his spectators onstage and in their own eyes (oculis geminabat imago), and when they realized they had been absorbed into his body (in vultus se magis isse meos), they fell into cold rigidity (horruit) and silence (muta), a foretaste of his death and the death they would die with him. His death is then an imitative response to their response to his imitations of them, and finally their death with him as formae is their ultimate imitation of his final performance. The dynamic of mimetic contagion is here mapped on to the hall of mirrors of an audience’s response to a mime. In Chapters 2–4, we examined the discursive habits of the Greeks and Romans surrounding the frequently recurring phenomenon of the inversion of mimesis, the reversal of ‘flow’ of imitation. The focus there was on works of art that inspire imitation, rather than meekly portraying objects, but if, as Diomedes’ definition implies, theatrical mime is mimesis’ true home, we may expect to find mimetic contagion here as well. Unfortunately we have so little evidence on ancient mime, beyond superficial descriptions; a thorough picture of contemporary constructions of sympathetic audience response to mime is beyond our present grasp. However, the Vitalis epitaph is not alone in suggesting that the audience could be drawn into mimesis of the mimic performance.

CONTAGIOUS MIME IN XENOPHON’S SYMPOSIUM

In chapter 2 of Xenophon’s Symposium an unnamed Syracusan entertainer arrives at Callias’ party with his mime troupe and has them perform a musical piece to the delight of the symposiasts. Socrates praises Callias for offering such pleasing sights and sounds to accompany the tastes of dinner, and Callias jokingly suggests that perfume be brought in to add one more sensory faculty, but Socrates replies that perfume drowns out the more attractive odours that distinguish free men from slaves. Far finer, he insists, is the residual smell of olive oil from the gymnasium (2.4). ‘Sure, on young men,’ responds Lycon, ‘but what should be the smell of a man too old for the gymnasia?’ ‘Nobility, surely’ (Καλοκἀγαθίας νὴ Δί’) replies Socrates succinctly. Lycon slyly asks where a man might come by this ointment, and Socrates agrees that it is not on sale in the perfumeries. In earnest answer to Lycon’s question, Socrates quotes a couplet from Theognis (35–6) indicating the ethically contagious quality of either good or bad associations: Ἐσθλῶν μὲν γὰρ ἀπ᾽ ἐσθλὰ διδάξεαι· ἢν δὲ κακοῖσι|συμμίσγῃς, ἀπολεῖς καὶ τὸν ἐόντα

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

GENRE SWITCHING AND ROLE CONFUSION

135

νόον (‘You will learn good things from good men, but if you mingle with the bad, you will destroy what good mind you have’).37 Socrates turns this line to praise Lycon’s son, Autolycus, who is also Callias’ eromenos, for his choice of friends, but not everyone present agrees that virtue can be learned. They are about to discuss the matter further, but Socrates notes that the dancing girl from the mime troupe is ready to perform a juggling feat, so they should delay this conversation. The girl’s performance is very impressive and Socrates uses it as proof that woman’s nature is not inferior to man’s and that a husband should confidently teach (διδασκέτω, 2.9) his wife whatever he would have her know. After a brief exchange about the challenges Socrates faces educating his own wife, the girl performs another dance, this time a dangerous one involving swords, and Socrates offers her performance as proof that even courage can be taught (καὶ ἡ ἀνδρεία διδακτόν, 2.12), for though she is female she was bold enough to jump into the swords. Antisthenes suggests that the Syracusan should exhibit the girl to the city and ask for money to teach the men of Athens to face combat. Amid such jokes, the boy in the mime troupe performs a graceful dance, and Socrates points out that as handsome as he is at rest, he appears even handsomer in his figures (σχήμασιν, 2.15). In all Socrates’ praise of the dancers it is not their natures or bodies that are emphasized, rather these are consistently subordinated to the skills or virtues they have been taught. Charmides notes this pattern and suggests that Socrates is really singing the praises of the dancing master (ὀρχηστοδιδάσκαλον). Socrates readily agrees and insists that anyone who wants his body to be more agile (εὐφορώτερον, 2.16) must dance this way. At this point, Socrates surprises everyone by turning to the impresario and applying to become his dancing student: ‘I too would be delighted, o Syracusan, to learn the figures from you’ (μάθοιμι τὰ σχήματα παρὰ σοῦ). The symposiasts laugh at Socrates’ suggestion that he learn to dance, so he goes on to defend the virtues of dancing for exercise and confesses that just the other day Charmides had caught him dancing early in the morning. Charmides plays along and Philip jokes about the absurdity of Socrates’ body. Then Philip orders flute music and renders a burlesque mimicry (μιμούμενος, 2.21) of the dances the troupe had just performed, exaggerating and misrepresenting every gesture. Ancient readers and modern scholars alike have generally taken this age as testimony that Socrates enjoyed dancing, but Bernhard Huss has argued that it is

37 Xenophon quotes these lines again, in his own voice, at Memorabilia 1.2.20. There too the theme is association with Socrates as a training in virtue (ἄσκησις τῆς ἀρετῆς). As long as Critias and Alcibiades were companions of Socrates his influence kept them good, but when they began spending their time with worse men, they deteriorated. For a reading of the Xenophon’s Symposium as a picture of the contagious influence of Socratic kalokagathia, see Huss 1999a, 405.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

136

MIMETIC CONTAGION

all a joke: ‘Socrates never danced’.38 Huss is surely right that we are meant to read the encomium of dancing’s salubrious benefits and the banter with Charmides as a jest, perfectly in keeping with the tone of the evening’s conversation, but as Wohl (2004) convincingly argues there is a serious kernel in Socrates’ proposal. He has been arguing throughout chapter 2 that virtue can be learned, specifically by associating with those who are virtuous and imitating them. He recognizes that for all the performers’ physical beauty, what is most impressive in them are the ethical qualities and the ‘figures’ they have been taught. The implication is that they have absorbed these qualities and forms of knowledge by ascetical exposure to and imitation of the impresario, and Socrates wishes to submit himself, as he wishes others to submit themselves, to the same kind of training in virtue. The notion of Socrates training in imitation of the mimes sounds funny to his companions, of course, and he plays to their laughter, but when Philip performs his coarse rendition (μιμούμενος) of the dances they have just seen, and exhausts himself in the process (ἀπειρήκει, 2.23), he is doing, if only at a superficial level, what Socrates had proposed. To find the imitation of the mimes enacted in full, we must look to the end of the Symposium. In chapter 7 a potter’s wheel is brought in as a prop for the dancing girl, who is about to use it in her juggling act, but Socrates suggests to the impresario that all the possible acrobatic feats the mimes could perform would not really be suitable for the occasion. Like scientific questions, such spectacles ‘are not conducive to the same goal as wine’ (οὐκ εἰς ταὐτὸν τῷ οἴνῳ ἐπισπεύδει, 7.5). He proposes that they leave and come back with a dance, perhaps portraying (γράφονται) the Graces, Horae, and Nymphs, as befits the charm of the symposium. The Syracusan agrees wholeheartedly and takes his troupe out. Socrates follows with his long Erotikos Logos (8.7–41) praising Callias’ love for Autolycus. Throughout, he frames their chaste, spiritual affection as not just beautiful and ennobling, but as a heavenly blessing from the gods.39 The man suffering under the vulgar love, by comparison, follows his beloved around like a beggar ‘pleading for a kiss or some other fondling’ (ἄλλου τινὸς ψηλαφήματος, 8.23). The course word ψηλάφημα and the rather abject image of the lover entreating for a grope are at odds with the polite urbanity of the party, so Socrates immediately offers an excuse for his bluntness: εἰ δὲ λαμυρώτερον λέγω, μὴ θαυμάζετε· ὅ τε γὰρ οἶνος συνεπαίρει καὶ ὁ ἀεὶ σύνοικος ἐμοὶ ἔρως κεντρίζει εἰς τὸν ἀντίπαλον ἔρωτα αὐτῷ παρρησιάζεσθαι (‘If I speak rather indecently, don’t be surprised, for the wine carries me along, and the love which

Huss 1999a, 387 (italics in original). Hindley points out (1999, 76–7) that we may be meant to understand that Callias and Autolycus’ relationship does have a physical dimension and that Socrates’ encomium is essentially protreptic. Such an assumption would explain Hermogenes’ observation in 8.12 that Socrates is onishing Callias as he praises him. 38 39

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

GENRE SWITCHING AND ROLE CONFUSION

137

ever abides in me goads me into frankness against its adversary love’, 8.24). Shortly before, Socrates claimed that a mimetic skit would be more in keeping with the trajectory of the sympotic wine than non-representational juggling and tricks would be. Now he reveals that the wine is also lifting him on its swell, and that its agency is shared by the heavenly love that comes from the gods and dwells inside him always. The heavenly love and the vulgar love may go under the same name and may sometimes look confusingly similar, but it must always be ed that they are really opposites. So too, if the wine and Socrates’ divine love have impelled him into some dictional indecency, it is only the candour born of sympotic friendship and must not be confused with real obscenity. Socrates’ speech ends and the Syracusan returns with his troupe and explains the scenario. The young woman will be playing Ariadne, waiting in her bridal chamber for Dionysus, who will arrive shortly, ‘having tippled a bit with the gods’ (ὑποπεπωκὼς παρὰ θεοῖς, 9.2). The Syracusan has succeeded better than Socrates could have hoped at matching the trajectory of the sympotic wine. Dionysus will be fresh from a symposium himself and in the same condition of mild inebriation as his audience. But more profoundly than merely matching his audience on this level, he will also portray the intersection of sympotic drinking with the epiphany of divine love coming to humanity. This was the experience of wine and love that Socrates was describing, but more importantly, if what Socrates says is true and there is a heavenly love dwelling in himself, then the symposiasts have, like Dionysus in the mime, been drinking in the presence of divinity.40 The characters in this mime are to mirror not only the superficial condition of the symposiasts, but their deeper story as well. Xenophon’s description of the mime closes the Symposium (9.3–7) and is worth quoting in its entirety, for its constant play on the slippage between reality and representation and its exemplification of mimetic contagion: To begin with, Ariadne, adorned as a bride, came in and took her seat on the chair. Dionysus was still nowhere to be seen, but there was Bacchic flute music. At this point everyone was amazed at the dancing master, for as soon as Ariadne heard it, she behaved in such a way that everyone would know she was delighted at what she heard (ἀκούσασα τοιοῦτόν τι ἐποίησεν ὡς πᾶς ἂν ἔγνω ὅτι ἀσμένη ἤκουσε). She didn’t go to meet him or even rise, but she was obviously struggling to keep calm. Finally, when Dionysus caught sight of her, he came dancing toward her as lovingly as could be, knelt down, put his arms around her, and gave her a kiss. She seemed like the picture of a modest woman (αἰδομένῃ μὲν ἐῴκει), but she lovingly returned his embraces nonetheless. As the symposiasts watched, they were clapping and shouting, ‘Encore!’ But when Dionysus stood up and helped

40

This seems to be what Autolycus’ father, Lycon, had intimated, when he closed the conversation with an interpretive last word: Νὴ τὴν Ἥραν, ὦ Σώκρατες, καλός γε κἀγαθὸς δοκεῖς μοι ἄνθρωπος εἶναι (‘By Hera, Socrates, I think you really are a noble human being’, 8.43).

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

138

MIMETIC CONTAGION

Ariadne to her feet, that was when one could see just what it looks like when people kiss and hug each other (ἐκ τούτου δὴ φιλούντων τε καὶ ἀσπαζομένων ἀλλήλους σχήματα παρῆν θεάσασθαι). This truly handsome (ὄντως καλὸν) Dionysus and beautiful Ariadne were not messing around but actually French kissing (οὐ σκώπτοντας δὲ ἀλλ᾽ ἀληθινῶς τοῖς στόμασι φιλοῦντας), and the spectators all got excited watching (οἱ ὁρῶντες . . . πάντες ἀνεπτερωμένοι ἐθεῶντο). For indeed they could hear Dionysus asking her if she loved him and her promising she did in such a way that not only Dionysus but all present would have taken an oath together that the boy and girl really were in love with each other. For they did not seem like people who had learned routines, but like those now permitted to do things long desired (ἐῴκεσαν γὰρ οὐ δεδιδαγμένοις τὰ σχήματα ἀλλ᾽ ἐφειμένοις πράττειν ἃ πάλαι ἐπεθύμουν). Finally, when the symposiasts saw them embracing and going off together, evidently to bed (ὡς εἰς εὐνὴν ἀπιόντας), the bachelors swore they would get married and the married men mounted their horses and rode home to their wives so they could enjoy them. But Socrates and those who remained went off to Lycon and his son on their walk. Thus was the dissolution of the symposium that night.

There is nothing surprising (except, of course, to the extent that all theatre is surprising) about Xenophon’s reference to the actor and actress as Dionysus and Ariadne. However, this fundamental principle of all theatrical mimesis, Coleridge’s ‘willing suspension of disbelief ’, becomes problematic when Xenophon causally connects the actress’ trained skill as a performer with the audience’s recognition that she was happy when she heard the music announcing her approaching bridegroom. Diderot’s famous ‘Paradox of the Actor’ is relevant here: the more powerfully an actor feels an emotion onstage the less successful he will be at conveying it. If the actor playing Ajax gives himself over to a genuine abyss of grief and embarrassment, he will be incapacitated as a performer and unable to project the signals appropriate to his character’s emotion. Xenophon points much the same direction by emphasizing the action of Ariadne (τοιοῦτόν τι ἐποίησεν) when she heard the music, its intended effect on the minds of the spectators, and above all their wonder at the expert skill of her coach. However, this objective focus on performance is undermined by the fact that the content of the spectators’ ‘knowledge’ is that, when Ariadne heard the music, she was happy. There is, of course, only one act of hearing, but it is stated twice, as a participle and then as a verb; the subject of ἀκούσασα is the expertly trained actress ‘Ariadne’, whereas the subject of ἀσμένη ἤκουσε, the woman who rejoices at Dionysus’ arrival, must be the character Ariadne. A similar problem arises when he contrasts her ‘looking like’ a modest woman, though she enthusiastically reciprocates her lover’s advances. Does he mean to refer to an ethically internal ambiguity, the disparity between how she seems and what she actually does? Or does he mean to suggest the disparity between the character and the actress, since the actress only simulates (ἐῴκει) the real modest woman, namely the mythological figure, while her profession itself is enough to disqualify her for real modesty? They are actors, providing the figures (σχήματα) of

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

GENRE SWITCHING AND ROLE CONFUSION

139

intimacy for a spectacle, but these are not empty appearances, as he hastens to explain. They were truly (ὄντως) a handsome pair and they were really (ἀληθινῶς) kissing, and the audience responds to all this reality on the stage by imitating it offstage: Dionysus arises (ἀνιστάμενος) and raises Ariadne up (συνανέστησε), and their kiss causes the spectators in turn to become ‘aroused’ (ἀνεπτερωμένοι). Dionysus asks Ariadne if she loves him and her oath (ἐπομνυούσης) of love is so believable that the spectators in turn would take an oath (συνομόσαι) that this is real romance, not between ‘Dionysus’ and ‘Ariadne’, but between the actors (τὸν παῖδα καὶ τὴν παῖδα). The distinction between represented ion and real ion has collapsed, as the realism of the performance reveals that the actors are not acting at all. But more interesting still is the inclusion of ‘Dionysus’ as the first of the potential oath-takers. Even if the audience is wrong and the boy has no feelings for the girl, he would of course, as ‘Dionysus’, still swear to the sincerity of his love.41 The audience’s earlier iration for the dancing coach (ὀρχηστοδιδάσκαλον) may be tempered now that it has become apparent that these are not actors who have learned their poses (δεδιδαγμένοις τὰ σχήματα) but simply lovers who have been given free rein, but here too, by using the verb ἐῴκεσαν, Xenophon subtly reminds us that the audience only reacted as it did because of how the actors seemed; their love may be apparent, but it is also therefore appearance. The spectators have now been drawn into fully sympathetic conformity with the mime, and when Dionysus and Ariadne leave together, presumably (ὡς) for the marriage couch, the married men take their cue and rush home to recapitulate the backstage denouement with their own wives. Socrates, Callias, and the rest also leave to Autolycus and his father on their walk and to enjoy a more spiritual interpretation of the mime’s hieros gamos: the exporting of noble sympotic friendship into the civic space, organized around the figure of divinely chaste homosexual love. Huss has noted the general parallelism of the erotic theme of Socrates’ speech in chapter 8 and the ‘highly risqué performance’ in chapter 9 and that this alternation of σπουδή and παιδιά informs the Symposium at a compositional level.42 However, the parallelism we have established here runs much deeper than the shared erotic theme. As we have argued, the Syracusan impresario

41 Various critics and scholars, beginning with Schenkl 1876, have noted the logical problem of Dionysus’ inclusion in the oath and sought to propose a lacuna after Διόνυσον or otherwise to emend the text; see most recently Bowen 1999, 125. Huss (1999b, 445) is more conservative: ‘Der Text is hier jedoch trotz leichter logischer Inkonzinnität tadellos’. I agree wholeheartedly with Huss, but I believe we can now offer an explanation for this ‘logical inconcinnity’, as one among many features of this ekphrasis that blur the distinction between mimesis and reality and point to this very blurring as still, paradoxically, an aspect of the mimesis. 42 Huss 1999a, 393 n.36. For a fuller presentation of the alternation of serious and light elements in the Symposium and in Xenophon more generally, see Huss 1999b, im and especially 30–7.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

140

MIMETIC CONTAGION

succeeds in matching the tone of the party and in representing, both at a superficial level and at a more profound level, the coordinated dynamics of sympotic drinking and divine love, which the symposiasts have just been hearing about from Socrates and experiencing through him. But the likeness of mime to symposium is no sooner established than its ontological order is reversed. The spectators are drawn first into affective imitation and then into full recapitulation of a performance which they no longer believe is a performance, and the boundary between representation and reality collapses, if not in the actors’ genuine arousal then certainly in the spectators’. In his Erotikos Logos, Socrates excused his slip into lexical indecency by displacing his agency on to the legitimating forces of wine and divine love; so also the symposiasts’ arousal in response to the mime may seem crude, but it is inspired by the presence of the very same forces within the mime, and its propriety is evident in its two resolutions, both of them civically productive: conjugal love and the public projection of sympotic friendship.43 For some literally, then, and for others metaphorically, the evening has ended with imitation of the mime. Socrates’ proposal of learning the ‘figures’ from the impresario in chapter 2, which quickly became a joke at the time, has turned out to have a more serious application, one which all the symposiasts would eventually follow to their own benefit, though at differing levels of abstraction.44 Socrates postponed discussion of the question whether virtue could be learned, suggesting instead that they watch the mime troupe perform. His answer to the question would seem to be implicit in his redirect: virtue, like vice, can indeed be learned—in fact, it is contagious. Wohl 2004 generally casts these two reactions (somatic, literal, and heterosexual vs spiritual, metaphoric, and homosocial) as incompatible, to the extent that she reads the impresario as being in implicit contest with Socrates throughout, but on pp. 356–7 she does consider the effective overlap in the two resolutions. 44 Scholarship on Xenophon’s Symposium has largely been devoted to comparing it (unfavourably) to Plato’s Symposium and to the relative dating of these two texts. Most scholars have come to the conclusion that Xenophon’s Symposium is later, but recently Danzig 2005 has provided fresh ammunition to the argument in Thesleff 1978, that Xenophon wrote the first seven chapters of his Symposium, presumably with a different ending, before Plato wrote his Symposium, and then later tacked on chapters 8 and 9, as a response to Plato’s text. Happily, for our purposes, the question who wrote first is as immaterial as it is inconclusive, and even if it could be demonstrated that Xenophon returned to his Symposium and added the last two chapters, it would not therefore follow that there must be a lack of unity in the text as he finally left it. The tendency to depreciate the unity of Xenophon’s ending as inferior to Plato’s is still very strong; see, for example, Bowen 1999, 126: ‘The clash between the heterosexual mime and Socrates’ words in VIII is left unresolved, and there are other loose ends. Plato wrote more tidily (Smp. 223b6-end).’ For Danzig, Xenophon’s relative artlessness is only the necessary cost of his greater ‘realism’. He points to Socrates’ speech at the end of Plato’s text, which subtly incorporates and answers the previous speeches: ‘This feature gives Plato’s Symposium a powerful literary unity that is, frankly, lacking in Xenophon’s’ (2005, 341). At the risk of diminishing Xenophon’s realism, I believe my reading shows the unity between chapters 8 and 9 and also between these last two chapters and the rest of the text. 43

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

GENRE SWITCHING AND ROLE CONFUSION

141

CHAEREA STUPIDUS

In both the Vitalis epitaph and Xenophon’s Symposium, mime’s imitative faculty is directed outwards to the audience, but there is an easier way to accomplish impromptu modulation of performance, namely between two characters within the mime. By keeping the game of reflective imitation within the skit, the actors can create a similar sense of improvisatory contingency, while maintaining somewhat more control of how the performance will develop. Because it does not involve the incalculabilities of audience interaction and can be practiced ahead of time by the actors, this effect can also be simulated in Plautine comedy. On the most basic level, this principle of competitive matching underlies one of Plautus’ favourite forms of pseudo-improvised emulative performance: the tit-for-tat duet, in which tight verbal echoes intensify the impression of an unscripted responsorial contest.45 The technical term for this theatrical motif, not only in modern scholarship but evidently also for Plautus, is par pari respondere.46 Alongside the general frequency of responsive copying in mime, the imitative function can also be consolidated on to one of the characters, who then becomes the generically constituted focus of mimicry. The mimic grex is often attested as having a leading actor, the archimimus, the head of the troupe and perhaps something like the vaudevillian ‘top banana’, but in addition to this figure the company would also need at least one other actor, a mimic fool variously called μῖμος δεύτερος, actor secundarum, μωρὸς φαλακρός, or simply stupidus.47 While the first mime pursued his heroic (or mock heroic) exploits, the second mime would follow along, aping his superior, picking up words and gestures for repetition in the wrong contexts, and generally finding himself in parallel, if ludicrously warped, situations. Horace, advising Lollius to steer clear of being an abject parasite, gives some idea how the part of the mimus secundus worked (Epist. 1.18.10–14): alter in obsequium plus aequo pronus, et imi derisor lecti, sic nutum divitis horret, sic iterat voces et verba cadentia tollit, ut puerum saevo credas dictata magistro reddere vel partes mimum tractare secundas; One man, flat on his back in flattery, A joker in the dunce chair, so trembles at the rich man’s nod, Parrots his phrases and picks up words as they tumble from his lips, You’d think he was a child repeating after his stern teacher Or a mime playing the second role. There is an excellent example in Act 2, scene 2 of Persa, appropriately called midway through by Paegnium (par pari respondes dicto, 223). 46 The fullest exploration of par pari respondere in Plautus is Wallochny 1992, 65f. 47 Nicoll 1963, 87–8. 45

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

142

MIMETIC CONTAGION

The character of the stupidus was one of mime’s most lovable and enduring inventions, from Sannio in antiquity to Harlequin in Commedia dell’arte to Papageno in Die Zauberflöte, whose love affair works as a parody of Tamino’s. Perhaps Papageno is a good illustration of another point about the second mime: he is ‘second’ in the sense that his is not the principal role and his performance is derivative of the one he is imitating, but that does not mean that he is less interesting as a character or that his part is technically easier. In fact, it is usually just the opposite on both counts. The archimimus is certainly a mime, but because the mimus secundus lives by direct and apparently improvised imitation, he is a mime’s mime. To quote (in a slightly different context) Diomedes’ definition of mime, solus quasi privilegio quodam quod fuit commune possedit (‘He alone has claimed as his special prerogative what was common to all’). The Eunuch, like most of Terence’s plays, is a duplex comoedia, and as usual one of the couples is better developed.48 Phaedria and Thais are both major characters in the play, and because of their prominence in Act 1, they emerge early on as the dominant love story. Unlike Thais, Pamphila is mostly talked about, comes onstage briefly in Act 3, says nothing, and is quickly ushered backstage never to appear again. So Chaerea’s love affair is fundamentally an auxiliary double of his brother’s, and when Chaerea first comes onstage in Act 2, Parmeno immediately subordinates his character to Phaedria’s by describing him as being like his brother, only more so (297–301): ecce autem alterum! nescioquid de amore loquitur: o infortunatum senem! hic vero est qui si occeperit, ludum iocumque dices fuisse illum alterum, praeut huiu’ rabies quae dabit. Great, the other one! Talking about love . . . Poor master! And this one is the type that, once he gets going, You’ll think that other one was fun and games, Compared to the mess this one’s madness will unleash.

Parmeno’s predictions are right, of course; Chaerea follows his brother in love by falling for Phaedria’s mistress’ adopted sister, and the effects of his rabies will soon

48

The only Terentian comedy with a single lover is Hecyra, one of the many ways this play is exceptional. In the other comedies the two love stories are never equally developed in interest, so that in most cases the apparent complexity of the plots is, as Sander Goldberg says, ‘a bit deceptive’ (1986, 144). In fact, the double-plot structure in Andria is so rudimentary that it is not always recognized as such (Levin 1967, 301), but the relative insignificance of Charinus and Philumena’s love story (and happy ending) for the story of the Andria is not fundamentally different from the disparity in other plays. It is only an extreme case of the usual asymmetry of the two lobes of the amatory double plot. On the Andria as barely (but still importantly!) duplex, see Gilula 1991; 2013, 239.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

GENRE SWITCHING AND ROLE CONFUSION

143

outstrip any problems his brother’s affair has caused. Phaedria is not the only one Chaerea copies, however. As I have argued extensively in Chapters 1 and 4, the eunuch intrigue originates with Parmeno, and until the intervention of the painting Chaerea is following Parmeno’s script. This complaisant reproduction of the role sketched by Parmeno is not the same thing as direct mimicry, but as we have seen, the play insistently points to the ethical consistency in Chaerea’s mindless conformity when faced with a model. For that matter Chaerea does mimic Parmeno, not in the eunuch intrigue but at the very beginning of their exchange. Parmeno has often rewarded Chaerea’s gifts by pledging to help if he is ever in love, and Chaerea calls in the favour now, but rather than simply reminding Parmeno of his promise, he parrots his actual words: Chaerea, aliquid inveni | modo quod ames: in ea re utilitatem ego faciam ut cognoscas meam (‘Chaerea, just find something to love; I’ll show you my usefulness’, 308–9). Donatus’ comment, which provides the epigraph for this chapter, draws attention to the mimetic excess of these lines. Chaerea’s performance of ‘Parmeno’ accomplishes nothing for the plot; its function is purely ethical. He reveals upon his first entrance that the tendency to mimesis is fundamental to his character (more suo).49 Chaerea fully adopts the qualities of mimus secundus only in his direct mimicry of the painting, or rather of Jupiter in the painting as archimimus. The later evidence for the second mime part, as regrettably thin as it is, does seem to fit Chaerea’s performance here perfectly, but was it normal to have the model, effectively the archimimus role, played by a work of art rather than a human actor? Because of the overall paucity of evidence for mime especially in this period, there is no way to know what connections this mimetic response to a work of plastic art might have had with improvisational routines specific to the stupidus type in second-century Italian folk theatre. There is, however, a tantalizing resemblance to the so-called Santia vase (Figure 5.1). A Campanian wine jug from the second half of the fourth century BCE, the Santia vase is usually classed as a ‘Phlyax vase’. Its form is unexceptional (oenochoe shape 3), and the overall manner of its secondary decoration is also typical of southern Italian Greek vases. It presents a figure standing next to a small, iconographically normal statue of Herakles. The figure’s mask is typical for a Greek comic slave, but his chlamys and chiton are different from those on other comic vases from Apulia and Paestum, and most notably he lacks the phallus.50 Above his Donatus has Chaerea in the ‘second part’ in another sense as well, following Parmeno as ing actor: in hac comoedia qui personam Parmenonis actor sustinet primas habet partes, secundae sunt Chaereae, tertiae ad Phaedriam spectant (‘The actor who plays the role of Parmeno in this play has the first part, the second is Chaerea’s, the third goes to Phaedria’, Don. Eun. praef. 1.4). 50 Taplin 1993, 40–1. Green 1995 argues that the costume is badly confused and that this is not a slave at all, but an old man misidentified as a comic slave, perhaps even by the purchaser of the pot rather than the painter: ‘[I]t seems clear that the painter has based his figure on other representations rather than direct 49

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

144

MIMETIC CONTAGION

F I G . 5.1. Santia vase, B.M. F233 Photo courtesy of the British Museum

head, in retrograde Oscan lettering appears the name SANTIA. Santia is easily explained as the Oscan form of Xanthias, which is cited by Aeschines as a typical name for a comic slave, and which is attested in labels for slaves on other southern Italian pots.51 In other words, the Santia vase is in itself strong confirmation of the adoption and modification of elements from older Greek comedy into similar, yet recognizably novel, Italian theatrical forms. observation . . . In any event, the vase does not deserve the place it has been given in modern scholarship as evidence for comic theatre. It is, rather, excellent evidence for what can happen on the margins of Greek culture’ (114–15). But precisely here, at the margins of Greek culture, hybridity of character types and costume conventions is more likely to be due to locally adapted forms of performance than to the painter’s confusion or ineptitude. 51 Aeschines 2.157. Other Xanthias pots: Museo Civico Archeologico, Milan, AO.9.284 and British Museum, London, F151.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

GENRE SWITCHING AND ROLE CONFUSION

145

Among its unique features, however, the most interesting thing about this vase for our inquiry is that Santia is engaged in direct mimicry of Herakles.52 He holds a staff with knots like those on Herakles’ club, and his left hand is raised in the same enigmatic gesture. Herakles’ left leg is dropped back in an attractive classical contrapost, a stance which is clearly meant to be echoed by Santia’s awkwardly crossed legs. Poor Santia, in short, is trying (and ludicrously failing) to copy the statue. There is, of course, no way to prove any kind of continuity between this isolated vase and Terence’s Eunuch a century and a half later, but specific lazzi can be remarkably persistent in folk theatre, and if Santia’s mimicry of the statue is part of a larger Italian expansion of the stupidus role to include the overt aping of artistic representations, Chaerea’s description of his own imitation of the painting may be reminiscent of a type of entertainment the audience would associate with native mimic traditions. Another similarity between the Santia vase and the rape scene in the Eunuch is that the figure being mimicked is a character from Greek myth. In both cases the dramatic re-enactment of an artwork takes the form of mythological burlesque. From its earliest beginnings, mime delighted in representing gods and heroes in humorously compromised circumstances.53 In fact, two Phlyax vases show situations very similar to the scenario Chaerea describes, in which an actor playing Zeus is creeping in at the window of a woman who may be Danaë.54 Mime here reveals one of its many affinities with Attic Old Comedy, as opposed to New Comedy, in which the gods appear rarely and only in prologues.55 In Roman comedy mythological burlesque is totally unknown, apart from the Amphitruo, which is probably not based on a Greek New Comic model.56 This pronounced difference between, on the one hand, Menandrian New Comedy, as adapted on the Roman stage, and, on the other hand, all the various older Greek, native Italian, and broadly ‘mimic’ traditions would make the rape scene in the Eunuch

52

In this regard Santia may be compared with Dionysus in Aristophanes’ Frogs (who is himself accompanied by a Xanthias). 53 Wüst 1932, 1752, cols. 9–22. 54 One vase is in the Vatican and the other is in the British Museum (F. 150). For images, see Nicoll 1963, 56–7. The woman is usually taken to be Alcmena, but in that case Zeus would presumably use the front door. For the argument that she is Danaë, see Stewart 1958, 369–70. 55 Pan delivers the prologue of Menander’s Dyscolus, the Aspis has Tyche, and the Perikeiromene has Agnoia. Five Plautus plays have prologues delivered by a god (Amphitruo, Aulularia, Casina, Cistellaria, and Rudens), but as Duckworth points out (1952, 296), these prologues differ little in content or function from those delivered by an impersonal prologus. In the Samia Demeas asks Nikeratos if he has not heard the tragic poets telling (λεγόντων, 589) how Zeus transformed himself to gold and paid a call on a girl, but this scene in Menander is itself far from mythological burlesque; it is not a comical re-enactment of Zeus and Danaë, but a reference to a telling. 56 The source may be a Middle Comedy (see Webster 1953, Hunter 1987) or perhaps from one of the native Italian traditions (see Chiarini 1980, 106–15).

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

146

MIMETIC CONTAGION

stand out as alien to its own generic context. Chaerea’s act is explicitly in imitation of Jupiter’s ‘play’ (ludus, 587), and he casts himself in the god’s role, not just for a momentary comparison, as in the Plautine examples we saw in Chapter 4, but for a recapitulative performance conscripting Pamphila in the role of Danaë. The clear thematic relevance of mythological burlesque to this narrated scene would, I believe, be felt as a generic rupture. The painted scaena is the only thing preserving the play’s smooth surface of Terentian decency from the mimic burlesque supposedly playing backstage.

MIME AND BODY

All drama can be located somewhere on a spectrum between pure declamation and mute action. The English word ‘mime’ generally implies a solely gestural performance, but mime in antiquity was by no means confined to dumb-show. There was, however, a striking preponderance of non-verbal entertainment in mime, especially compared with the overwhelmingly verbal aesthetic (pura oratio) of Terentian comedy.57 Although mimes with carefully scripted libretti were not unheard of, in most cases the language seems to have been simple and less important than the action.58 Dance and physical comedy were always part of the show, and from its earliest beginnings down through late antiquity, mime was capacious enough to contain acrobats, jugglers, and other thaumatopoioi.59 If only for its non-verbal physicality the rape in the Eunuch would look more like a routine from mime than a scene from New Comedy, but not only is it physical, it is explicitly sexual. Perhaps no feature of ancient mime is so consistently testified as its obscenity, and despite its frequent overlap in themes and scenarios with New Comedy and the Greek romance, mime is set apart from both these genres by the frankness with which it portrays sexual situations.60 Ancient descriptions of mime emphasize the prominence of adultery in the plots, the bawdiness of the language, and the sleazy movements and unbridled performance style.61 In the few mimic fragments 57

On the centrality of slapstick in mime, focusing on the apparently mimic associations of the word alapa, see Rawson 1993, 259. On sexual violence in mime more generally, see Fantham 1986. 58 ‘Die Sprache des Mimos ist die Umgangssprache des niederen Volkes’ (Wüst 1932, col. 1736). 59 Most of the scholarship on mime focuses exclusively on literary dramatic mime. For non-dramatic mime, see most recently W. J. Slater 2002. 60 As a genre only the Latin novel seems in any way comparable, though the Pseudo-Lucianic Ass and the Iolaus may also be adduced in this regard as Greek matches for the two Latin novels. 61 On the relative preponderance of primary obscenities and sexual metaphors in mime, compared to comedy, see Adams 1982, 219. In his list of the various types of mimic performers, Athenaeus mentions one called the kinaidologos, who apparently lived up to his name (14.620e–21b). Sotades, the champion of this particular style, ran afoul of Ptolemy Philadelphus by goading him for his incestuous marriage to his sister,

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

GENRE SWITCHING AND ROLE CONFUSION

147

that have come down to us, for example from Herodas (third century BCE), women coach each other on adultery (Mim. 1) and where to get a good dildo (Mim. 6). In another of Herodas’ mimes (Mim. 5), a mistress, Bitinna, who forces her slave, Gastron, to have sex with her has become jealous of his wandering attentions. Herodas, like Theocritus, is writing pseudo-mime for an educated elite, so the language is uncharacteristically artificial, literary, and clean.62 Even so, the first three lines show enough differences from the world of New Comedy to be worth quoting: λέγε μοι, σύ, Γάστρων, ἤδ᾽ ὐπερκορὴς οὔτω ὤστ᾽ οὐκέτ᾽ ἀρκεῖ τἀμά σοι σκέλεα κινεῖν ἀλλ᾽ Ἀμφυταίῃ τῇ Μένωνος ἔγκεισαι; Tell me, Gastron, is this so stuffed That jiggling my legs doesn’t cut it anymore But you’re wrapped up in Menon’s Amphytaea?

No one in Menander talks about sex this boldly, of course, least of all a woman, but just as striking, from the perspective of New Comedy, is the demonstrative pronoun in the first line. Bitinna would not fit in a Terence play because she is the kind of woman who will point to a man’s penis and ask rude questions about it, and Gastron would not fit because he has something to point at. One of the most conspicuous changes in the costume of comic actors across the history of the genre is the gradual lengthening of the jerkin and shortening of the phallus.63 The male characters in mime, as in Aristophanes, would have sported visible , whereas their New Comic and Roman cousins had discreetly hidden (which is surely to say, for purposes of costuming, non-existent) genitalia.64 Not only did Roman comic actors not wear a false phallus, they were meticulous about avoiding wardrobe mishaps that would give an unscheduled glimpse of their real privates. Cicero (Off. 1.129) explains, ‘Such is the modesty of stage actors (scaenicorum) by longstanding discipline (vetere disciplina) that none of them would set foot on stage without a subligaculum on, lest some accident befall and they make an Arsinoë. The line that got Sotades killed was: ‘You’re sticking your prick into an unholy hole’ (εἰς οὐχ ὁσίην τρυμαλιὴν τὸ κέντρον ὠθεῖς). Athenaeus also describes the magodos who: ‘has tambourines and cymbals and all his clothes are effeminate; he bumps and grinds (σχινίζεται) and does everything that’s over the top (τὰ ἔξω κόσμου), playing now adulteresses and pimps, now a drunkard in a revel going to meet his lady’ (14.621c). Compare, for example, the ‘popular mime’ fragment 7 (Cunningham), where the scenario is the same: mistress demanding sex from her slave, who has grown wayward since he fell in love with another woman. Her language is not just prosaic, but extremely coarse, her descriptions of the torments she will exact on her slave are brutal, and her sexual vocabulary would make Herodas’ Alexandrian heroine blush. 63 See Green 2002, 104–6. 64 For the phallus in mime, see Reich 1903, 258n. The scholiast for Juvenal (6.276) refers to phallum ut habent in mimo. 62

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

148

MIMETIC CONTAGION

improper spectacle by revealing certain parts of the body.’65 Given the sexual effusiveness of the males of Old Comedy and mime, compared to the relatively restrained gentlemen of New Comedy, it is not hard to see phallic size as a fairly obvious index of libido, especially for Hellenistic and Roman mime, since the persistence of the traditional large phallus is here inevitably in generic dialogue with the costuming conventions of New Comedy. If the males of the mimic stage were usually macrophallic, the females were often sexualized in another way. As we have already said (p. 125), the women’s roles in mime were generally played by actresses, while in highbrow theatre, whether tragedy or comedy, the cast was all male. In most cases the gender gap between male actor and female character may have been ‘invisible’ to an audience accustomed to this practice, but the gap was nonetheless always there, whenever playwrights or performers wanted to draw attention to it.66 Ancient theatrical cross-dressing probably did not indicate, as ‘drag’ does so readily for us, gender performativity at its most conspicuous, but along with masks, metrical language, and numerous other theatrical conventions, it contributed to the broadly nonillusionistic character of most ancient theatre. Conversely, the absence of any of these protocols in mime may have tended to push performance towards illusionism and audience response towards a sympathetic immediacy unlikely under the literary code systems of performed gender in tragedy and comedy. In Xenophon’s of the Dionysus and Ariadne mime, for instance, the unfeigned kisses and apparently genuine expressions of love blur the boundary between representation and reality. Does the audience become aroused because they see a convincing performance of Dionysus and Ariadne and come to a deeper understanding of divine love or because they watch an explicit make-out scene between an uncommonly attractive boy and girl, whose true love makes ‘Dionysus’ and ‘Ariadne’ irrelevant referents? I have argued above that Xenophon subtly avoids authorizing either interpretation to the full exclusion of the other,

65 For the subligaculum, see Olson 2003, 206 and Marshall 2006, 65–6. Marshall speculates that the comic actor may have worn some kind of token penis, over the subligaculum but under the tunic, and that this might explain the three attested Naevius plays with names pertaining to male genitalia. But if the actors in fabula palliata did wear a phallus, it was evidently so subtle as to be negligible compared to that of mime. Surely plays with titles like Triphallus, Testicularia, and Appella would be entertaining even without a costume phallus. 66 Jeffrey Henderson (1996) takes a strong position on the normativity, and therefore semiotic insignificance, of cross-dressing in Greek and Roman theatrical cultures: ‘male and female characters are at all times understood to be respectively men and women’ (17). Henderson’s reserve here may be prudent, as it is surely easy for us to overinterpret this feature of ancient theatre, but Greek tragedy and comedy (Bacchae and Thesmophoriazusae are perhaps the most obvious examples) do sometimes highlight the gender gap between male actor and female character as part of their exploration of sex and its relationship to performance, most explicitly by including cross-dressing as a plot element. The clearest example of this kind of play in Roman comedy is in the Casina; see Gold 1998.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

GENRE SWITCHING AND ROLE CONFUSION

149

but the collapse of performance into either illusion indistinguishable from reality or into reality itself would clearly be impossible without the gender parity between these performers and their roles.67 Given this tendency towards sexualized deployment of the female body, perhaps it is not surprising that ancient mimes sometimes maximized this form of realism by staging nude mimae. In fact, a bed might be placed onstage and sex consummated before the very eyes of the spectators.68 The direct testimony of sex acts in mime is considerably later than Terence, but there is no evidence that mime became more sexually explicit over time, nor is there any good reason to posit such a trajectory in its development. Indeed, mime’s very formlessness seems to frustrate all attempts to delineate a clear history of the genre. Whether or not Terence’s audience would have known live sex shows as a form of entertainment proper to mime, they certainly would have recognized the naked mima as a feature of the Floralia. The Eunuch was performed at the Ludi Megalenses (4–10 April) in 161 BCE, and the Floralia (28 April–10 May) was the next festival in the Roman calendar. The Floralia may have had a long prehistory in Roman religious life, but the Ludi Florales were officially incorporated into the calendar just twelve years before, in 173 BCE, and the theatrical centrepiece was evidently a mime performance with naked mimae (or naked meretrices functioning as mimae).69 For Pamphila’s brief, non-speaking appearance onstage in Act 2, she is, of course, played by a male actor, but if the audience were watching the backstage events, Pamphila could no longer be played by a man. She is summoned to her bath, leaves the room, comes back, and is placed on a bed, where Chaerea is instructed to fan her freshly bathed and presumably naked body. The other slaves leave the room to take their bath and Chaerea bolts the door behind them, so there will be no spectators for the subsequent dumb-show. The rape does not occur in some vague interior space, but in a room as discretely bounded as a stage, which Chaerea does not leave throughout the scene.70 Thais leaves the room with a pronouncement typical of characters leaving the stage (ego . . . ad cenam hinc eo, 580), and the bath is as hidden from Chearea as if it were securely ‘offstage’. The stage-like unity of this interior space fits the theatrical language Chaerea uses to

67 Plutarch says the mimes stage things not even slaves should be allowed to look at, and ‘they throw one’s soul into greater confusion than any amount of drink’ (Mor. 712E-F). 68 The bed: Ps.-John Chrysostom Patr. Graec. 56.543; real sex: SHA Elag. 25.4; Minuc. Felix 37.12; Val. Max. 2.6.7. 69 See Val. Max. 2.10.8 and Sen. Ep. 97.8. On the general character of the Floralia, see Ov. Fast. 5.329–68 (especially 347–8 for the scaena levis). 70 All Roman comedies with the exception of Rudens are set in the street in front of a house or pair of houses; see Duckworth 1952, 83. Mime’s setting is much more flexible, even including, as often in the case of adultery mime, interior spaces.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

150

MIMETIC CONTAGION

describe his performance, but the presence of a naked woman ‘onstage’, to say nothing of the rape itself, is enough to make the scene look more like a transplant from mime than a scene from Terentian comedy.

CODE SWITCH: MIME

We have surveyed the general features of the rape scene in the Eunuch that make it reminiscent of mime. It contains violent physicality, extreme sexuality, and female nudity, all of which are usual for mime and none of which would fit in the restrained world of New Comedy, and it is framed as an imitation either of another performance (a burlesque of Jupiter’s ludus) or of a work of art, making Chaerea’s response a type of the mimus secundus role. But perhaps it remains to show where Terence has more specifically intervened to draw attention to this generic rupture in his play. Because of the essential conservatism and continuity of mime throughout the Hellenistic period, much of what we have said here may be equally valid for Menander’s Eunouchos, but since we have lost that text, reflection on Terence’s relationship to his source must remain speculative. In one particular, however, we are on firmer ground: Donatus tells us (ad, 539) that in the Greek original for this scene the narration of backstage events was a soliloquy, and Terence has invented the character Antipho to break it up (ne unus diu loquatur, ut apud Menandrum).71 Antipho’s name essentially means ‘protatic character’, and Donatus’ claim that he was introduced by Terence for the sake of interruption seems credible enough. But of his dozen remarks most are very short and essentially phatic (ain tu?, 567; quid id est?, 571; miser!, 580). In fact, Antipho slips in only one real interruption of any substance. His comment is extremely interesting, however, for in it he conceives of the backstage events as a spectacle he would like to have seen. Chaerea had just been handed a fan and given instructions to aerate the freshly bathed Pamphila, with whom he would now finally be alone, and Antipho breaks in (597–8): AN. tum equidem istuc os tuom inpudens videre nimium vellem, qui esset status, flabellulum tenere te asinum tantum. AN. That’s when I would really have wanted to see your shameless face, and what your stance was, holding your little fan, you giant ass!

Antipho’s vellem introduces a past continuous unfulfilled condition, three things he would like to have seen. My translation can scarcely capture the force of his 71 I follow most scholars in taking Donatus at his word here, but Fraenkel (1968, 235–42) argues against doing so.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

GENRE SWITCHING AND ROLE CONFUSION

151

asyndetic tricolon, but of the three grammatically different objects of vision, it is the middle element, the indirect question, that seems hardest to render in English. Barsby (1999a, ad loc.) suggests ‘what posture you adopted’, where I have used an English cognate of status, the precise sense of which here is somewhat difficult to locate. Why should Antipho wish to see how his friend was standing? The answer becomes clear, I believe, if we consider the situation and pay close attention to the third element of the tricolon. Until the other slaves leave the room Chaerea must continue to play his part, but he is about to be left alone with Pamphila and his excitement is at its peak. Antipho understands that by this point his friend was struggling to hide an erection. Chaerea’s ‘stance’ would have been such as to conceal, as best he could, the evidence of his arousal. The ‘little fan’ is funny anyway, of course, as a ludicrous stage property for this virile adulescens to be ordered to hold, but it becomes even funnier when we realize, as Antipho does, that he must hold it in such a way as to make it an instrument of concealment.72 Animal insults are rare in Terence, so it is striking to hear Chaerea called a ‘huge ass’, but donkeys were notoriously oversexed and well-endowed, and Antipho’s point is not lost.73 If Chaerea’s erection were detected by the internal audience of his intrigue, it would vitiate his performance as the eunuch and blow his cover. But his visible phallus is also an indication that this scene is ‘costumed’ like a mime, rather than like a palliata, for as we have seen the archaically protruding male member is virtually a signature of mime in the Hellenistic and Roman periods. Chaerea’s erection is essentially a masculine analogue of Pamphila’s nudity. Another immediately visible difference in costuming between these two genres would probably have been the masks. Unfortunately, we can reconstruct very little about the masks of Roman comedy with any certainty, but it is at least fairly clear

As we have seen (p. 146), ancient mime was capacious enough to include performers we would call jugglers. These artists could be called simply mimi or they could be referred to by a more specialized term often based on the object they typically juggled. In a recent article W. J. Slater (2002) analyses the evidence pertaining to the ventilator, a type of juggler who danced with a toy shield, specifically arguing that although the term is not used, this is the type of mime Martial has in mind in 9.38, where nimble Agathinus is praised for balancing his little shield on various parts of his body without ever letting it fall. When Antipho broke in Chaerea was mimicking the slave who showed him how to ventilate Pamphila with the fan (ventulum huic sic facito, 595). If the toy shield juggler was a type Terence’s audience would have recognized, perhaps even under the name ventilator, part of the humour in this situation may rest in Chaerea’s deployment of the fan precisely as a shield to hide his erection. Chaerea’s pseudo-martial use of the fan is especially funny because he is a young man of military age; in fact, he is currently serving the ephebeia (miror quid ex Piraeo abierit; nam ibi custos publice est nunc, 290). If we can trust Livy’s of the capture of Heracleum in 169 BCE, just eight years before the Eunuch, the use of mimic shield play by young soldiers would already be familiar to Terence’s audience. Livy describes (44.9) how the wall was breached by a group of Roman youth (iuvenes etiam quidam Romani) who held their shields above their heads at an angle to form a slanted testudo, which could then be used by other soldiers as a ramp to the top of the wall. Livy says specifically that the young men had adapted this technique from a circus performance (ludicro circensi ad usum belli verso). 73 For an exhaustive bibliography on ancient prejudices about aroused donkeys, see Griffith 2006, 224. 72

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

152

MIMETIC CONTAGION

that the actors in Greek New and Roman comedy did wear masks, almost certainly regularized into tightly defined character types.74 Most mime, on the other hand, did not use masks, and though there may have been forms of mime throughout antiquity that did, it seems unlikely they were ever confined to the fixed masks of New Comedy’s schematic stock roles.75 In a theatrical culture that uses masks, references to a character’s facial expression can be used to convey information otherwise difficult to signify, or they may be used metatheatrically to point up the artificiality of the convention. Even if there were no play against the mimic genre here, Antipho’s reference to Chaerea’s face would still remind the audience that Chaerea is wearing a mask. But given Antipho’s evident recognition of the backstage generic shift, his wish that he could have seen his friend’s os inpudens signals the fact that this mime scene would have required a different face than the one Chaerea is wearing now on stage.

EKPHRASTIC ROLE CONFUSION

Mimus halucinatur, comoedus sermocinatur (‘the mime meanders, the comic actor converses’)—thus, at any rate, does Apuleius (Flor. 18) characterize the performance styles of the two genres we have been trying to distinguish in this chapter.76 If delirious lack of focus is typical of mime, perhaps it will not surprise that Chaerea’s description of the painting itself is so unclear. Indeed, there does seem to be an almost hallucinatory confusion in his interpretation of what exactly he saw depicted. He says Pamphila was looking up at a painted and (584–5): ibi inerat pictura haec, Iovem quo pacto Danaae misisse aiunt quondam in gremium imbrem aureum. on it was this picture, How they say Jupiter once sent a gold shower into the lap of Danaë.

Ekphrasis always replaces showing with telling, but here Chaerea has gone further, substituting mythological narrative (aiunt) for description.77 The version of the 74

Most scholars today agree that the actors in Roman comedy wore masks, but this has not always been the majority opinion. The tide seems to have turned in favour of masks after Beare’s masterful argument (1950, 303–9). See also Duckworth 1952, 92–4; Gratwick 1982, 83–4; Wiles 1991, 132–3. For a dissenting view, see Dupont 1985, 80–1. 75 For masks and no masks in mime, see p. 125. 76 If Apuleius’ implicit contrast between halucinatio and sermo is meant only to imply the difference between the fantastic and the ordinary, halucinari is a strange verb to choose. The examples OLD offers for (h)alucinor generally seem to imply delusional wandering of the mind. 77 It is possible that this invocation of what ‘they say’ should be read as an allusion to a specific text, an ‘Alexandrian footnote’. The Danaë story had provided the material for tragedies by Livius Andronicus and Naevius, so it is quite possible that Terence’s audience would think of a specific version. But, of course, such

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

GENRE SWITCHING AND ROLE CONFUSION

153

myth that Chaerea seems to mean is well known: Jupiter transforms himself into golden rain and directs himself to Danaë’s privy parts. But a few lines later he says something very strange. He rejoices (588–9): deum sese in hominem convortisse atque in alienas tegulas venisse clanculum per inpluvium fucum factum mulieri. that a god should turn himself into a man and creep onto someone else’s roof tiles to trick a woman through a skylight.

Just what is it then that the painting shows? Does Jupiter transform himself into a rain cloud or a man? In earlier ages, when editors were poets, several suggestions were offered to fix in hominem, but emendation is unnecessary here.78 There was another version of the myth, where Jupiter enters Danaë’s chamber through the roof as a shower, but once inside, rather than raining on her, he assumes human form for his epiphany.79 Donatus, who evidently has the same text we do here, finds several possible solutions. ATQVE IN ALIENAS TEGULAS hic apparet separatim Iovem, separatim auri fuisse picturam. DEUM SE IN HOMINEM CONVERTISSE utrum quia Iuppiter humana forma aurum infundens pictus erat in tabula, non pro Iove aurum? an ‘in hominem’, id est in hominis audaciam atque flagitia? AND ONTO SOMEONE ELSE’S ROOF TILES—It appears here that Jupiter and the painting of the gold were rendered separately. THATA GOD SHOULD TURN HIMSELF INTO A MAN—Is this because Jupiter in human form pouring gold was painted on the , rather than gold representing Jupiter? Or does ‘into a man’ mean into the impudence and vulgarity of a man? In the first comment Donatus considers the possibility that various moments in the myth may be discretely rendered, like a cartoon strip, in which case Jupiter can appear in human form here and in precipitate form there without logical inconsistency. Donatus’ deeper uncertainty about the age becomes clearer in the second comment, where he entertains both the possibility that Jupiter really does send (misisse) the rain down, while himself staying anthropomorphic, and the alternative possibility that he appears transfigured as rain, in which case the irony is

specificity would be hard for us to prove, since the Danaë myth had also been the subject of plays by each of the three Attic tragedians and was popular in other media as well. 78

Bentley liked in aurum or in pretium and Fabia suggested in imbrem. For excellent arguments against either of these emendations, see Tromaras 1985. 79 Pherecydes of Athens (FGrH 1.3.10) is our principal source: ἐρασθεὶς δὲ Ζεὺς τῆς παιδὸς ἐκ τοῦ ὀρόφου χρυσῷ παραπλήσιος ῥεῖ. ἡ δὲ ὑποδέχεται τῷ κόλπῳ, καὶ ἐκφήνας αὑτὸν ὁ Ζεὺς τῇ παιδὶ μίγνυται.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

154

MIMETIC CONTAGION

that what Chaerea describes is not Jupiter’s physical transformation, but his ethical transformation, his diminution to the all-too-human, in the same sense that the rape in the Adelphoe was called humanum (471). The conception of Perseus was a common motif for vase painters from the early fifth century on, as well as showing up on gems and other media. There are also several wall paintings of this scene that survive from Pompeii, enough to lend to the inference that this was a particularly well-loved theme in Roman domestic art. One is hesitant to draw strong conclusions based on frequency, since the Pompeian material is two centuries later than the Eunuch, but Pliny indicates that this scene was already painted on a tabula by Nicias in the fourth century.80 In the majority of these images Danaë languidly holds her himation aside, exposing her pudenda to the tiny, localized dribble of rain, but there are two paintings, one from the House of Regina Margherita (V.2.1) and the other from the so-called house of Gavius Rufus (VII.2.16), that show Jupiter, identifiable by his sceptre, in the form of a young man seated next to the typical Danaë-cum-drizzle configuration.81 Perhaps these images are intended to represent the alternative version of the myth, in which the transformation was only a means to get into her prison and Jupiter returned to his more usual anthropomorphic form to have sex, but without the iconographically standard golden shower it may have been difficult to indicate Danaë’s identity. In any case, these paintings show her being impregnated by rain (the more usual version of the myth), so Jupiter’s presence as a iuvenis is hard to explain. Indeed, he would seem not only to be in two places at once, but in two versions of the myth at once. I emphasize that these are not two discrete frames representing different points in time, like a cartoon strip, nor yet a jumbling of two moments in a single frame, neither of which would be unexampled in the mythological paintings from Pompeii, but rather a mythological catachresis, where the dominant version plays out as usual and the alternative version sits idly by and spectates.82

Pliny’s testimony (H.N. 35.131) is not explicit as to what kind of Danaë scene Nicias painted, but it was most likely some version of the conception of Perseus. Images of Danaë with the shower of gold are so much more frequent in all media than any other type of Danaë scene that when Pliny calls Nicias’ tabula simply a painting of Danaë, it is hard to imagine such an unmarked description fitting any other scene from the Danaë myth. Furthermore, Tiberius dedicated this , alongside another by Nicias portraying Hyacinthus, in the Temple of Augustus. Jean-Jacques Maffre (LIMC s.v. Danae, 330) notes the likely reason for this collocation: ‘Ce rapprochement donne à penser que les deux œuvres étaient consacrées à des mortels aimés par les dieux, et peut-être D. était-elle donc peinte en train de recevoir la pluie d’or.’ 81 For the painting from the House of Regina Margherita, see Schefold 1957, 69; for the one from the house of Gavius Rufus, see Schefold 1957, 170. The paintings are quite different from each other: in the Regina Margherita painting the setting is inside and Danaë is portrayed sitting; the Gavius Rufus painting is set outside with Danaë standing and receiving the golden shower poured from the horn of a Cupid. Corso (2004, 306–7) connects this Danaë with the Praxitelean tradition, by way of Nicias. 82 Elsner (2007, see esp. 32–3) explores a similar play in Pompeian paintings of Ariadne, where the internal gaze is refracted along the jagged fault lines of contradictory mythological traditions. 80

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

GENRE SWITCHING AND ROLE CONFUSION

155

Working back from Pompeii to Hellenistic painting is, of course, always dangerously speculative, but when two paintings of the same myth, from different houses and apparently from different hands, show such significant differences yet replicate the same essential form by duplicating Zeus in defiance of narrative logic, it seems reasonable to assume that there may be a standard Hellenistic type serving as model.83 Some of the of Terence’s audience could then very well have recognized Chaerea’s convoluted description of the painting as a fairly accurate rendering of this catachrestic type, which creates a narrative paradox by showing two contradictory versions of the myth in the same frame. In an iconographically standard pose Danaë looks meekly up at the imber streaming down on her, and she does not seem to notice the logically extraneous iuvenis who is gazing at her intently. Chaerea is obviously in just the same position as this iuvenis, watching his beloved as she watches something else. In the painting the imber and the iuvenis are somehow both Jupiter, but this essential identity with the painted object of Pamphila’s gaze is just what Chaerea claims and then enacts: consimilem lat | iam olim ille ludum (586–7). His ekphrasis sounds unclear because it makes a strange muddle of the roles of Jupiter/homo and Jupiter/imber, but if the audience recognizes this iconographic type in his description, they will also readily understand his claim of similarity to Jupiter, who is shown here as both scopophilic spectator and metamorphic sexual actor.84 Given the focal agency of the painting, the similarity of Pamphila’s role to Danaë’s is also clear: Danaë’s attention, like Pamphila’s, is directed not at the iuvenis, who is inexplicably in her room, but at an object which is normally inanimate, but is now exceeding its usual boundary, invading her body, and changing her life. The mime’s imitative faculty, as we have seen so many times, may be pointed either within or without the fictional world of play. If without, then he becomes like Vitalis, a mirror of life, drawing in his spectators and captivating them by rendering them as himself. But if the mime’s capacity for imitation is directed at some other character in his fictional world, perhaps even at a work of plastic art, then he takes on the mantle of the stupidus, who apes his superior in all gestures and actions, even (or especially) when he has no clear understanding of what he is doing. As a performance style, or rather as a way of approaching performance, mime instils an extreme flexibility of role. The model changes with the vagaries of the audience or the unpredictable grandeur of the troupe leader, and the performer has to remake himself with hypermobile and apparently thoughtless immediacy. When roles are felt to be labile and plastic to the degree mime

On the relationship of Roman -pictures to old masters, see Ling 1991, 128–35. Selene and Endymion are depicted on a comparable pair of Pompeian wall paintings, which likewise figure the deity as agent and primary viewer of the epiphany, though of course Selene is not presented in two forms simultaneously. On these paintings, see now Platt 2011, 372–6. 83 84

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

156

MIMETIC CONTAGION

inculcates, and the stupidus bumps into a painting where the roles are illegible because they are so ambiguous, he can only project this two-places-at-once quality in his description of the painting and in his dramatic response. Thus our reading of the generic rupture to mime in this scene is complementary to a full appreciation of Chaerea’s confusion both in his ekphrasis and in his imitative performance. If Chaerea and Pamphila were the only characters touched by mime, it would be a grim presence in the play. Happily, there is another character in the Eunuch who more perfectly submits to the logic of mimetic contagion and more successfully exemplifies its comic possibilities. We will explore his philosophy of performance in Chapter 6.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

SIX

The Poetics of Contamination

Verschiedene Weine zu mischen, mag falsch sein, aber alte und neue Weisheit mischen sich ausgezeichnet. Bertolt Brecht, Der kaukasische Kreidekreis

An etymology, as told by even the best lexicon, is a monologic trajectory, a onetrack evolution, and the meanderings subtended by a calque are fundamentally outside its scope. The Oxford English Dictionary’s entry for ‘contamination’ distinguishes between the general meaning as pollution or corruption (1.a and b), and the technical usage as a philological term of art for the blending of stories (c), lexical forms (d), or manuscripts (e). These technical meanings are obviously related; they are drawn from the same scholarly field, and they have in common with each other, and in contrast to the non-specialized uses, an ethically neutral focus on mixture, as opposed to ruining. The dates of first attestation (and hence the ordering in the entry) reveal that d and e are developments from c, and both the exempla offered for c are from texts discussing Roman comedy. What the OED does not tell us is that this technical usage is not a development from contamination’s general meaning, but rather an importation of the German Kontamination, which had steadily gained currency since Grauert’s magisterial treatment (1833) of contamination in Roman comedy.1 The scholars who introduced this homonymous double into English, eventually for circulation in contexts totally unconnected to classics, thought they were recovering the original meaning of the Latin word contaminare. If the more recent scholarship on the question is correct, they were in error, but I think it is likely that they were at least partly right.

I am not sure that Grauert coins the German word. He is the first scholar I can find using it, but that may only be because, unlike Westerhov, Drakenborch, and other early scholars who address the issue, he is not writing about contaminare in Latin. He uses the word as an established term, not a neologism. I note, too, in ing that since World War II and the cultural advent of radioactive contamination, German has imported the primary English sense as a secondary meaning of Kontamination—we have swapped calques! The 1960 Duden knows only the philological sense, but the 1997 Duden lists general and nuclear Verunreinigung as second and third definitions, respectively. 1

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

158

MIMETIC CONTAGION

It is the joy and frustration of studying the vocabulary of Roman comedy that so many words are recorded here for the first time. The three earliest surviving attestations of the verb contaminare are from Terence.2 The first of them comes in the prologue of his first play, the Andria (16): quae convenere in Andriam ex Perinthia fatetur transtulisse atque usum pro suis. id isti vituperant factum atque in eo disputant contaminari non decere fabulas. The bits from the Perinthia that fit into the Andria He its he lifted and used as his own. His enemies find fault with this practice, Namely they contend plays ought not be contaminated.

The prologus goes on to say that when they make this accusation, they are really accusing Naevius, Plautus, and Ennius, who did the same thing, but Terence would rather imitate the neglegentia of these poets than the ‘obscure diligence’ of his critics. The next occurrence of contaminare is in the prologue of the Heautontimorumenos (17): nam quod rumores distulerunt malevoli multas contaminasse Graecas dum facit paucas Latinas, factum id esse hic non negat neque se pigere et deinde facturum autumat. habet bonorum exemplum, quo exemplo sibi licere facere quod illi fecerunt putat. His enemies have spread rumors That he has contaminated many Greek plays Making few in Latin, which he doesn’t deny Nor is he sorry, and he says he’ll do it again. He has the model of good poets, by which model He supposes he can do what they did.

One certainly sees why Mme Dacier would be a little scandalized at how boldly Terence owns the charge of contamination. In her 1706 edition she wrote, ‘contaminare ne peut signifier ici gâter, et ceux qui l’ont traduit de même font dire à Térence une chose de fort mauvais sens; car qui doute que ce ne soit fort mal de gâter les comedies . . . ? Il faut donc que contaminare soit pris en bonne part’. But Mme Dacier is not sanitizing contamination in a vacuum—she will, of course, have known Donatus. In fact, most sixteenth- and seventeenth-century editions of Terence printed the commentary alongside the plays. His scholion for Andria 16 2 The nominal form, contaminatio, is not attested before the third century CE, and when it does occur it certainly is not in the literary technical sense.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

THE POETICS OF CONTAMINATION

159

(contaminari non decere) reads: id est: ex multis unam non decere facere. So an unambiguous definition of contamination as ‘making one from many’ can be traced from the fourth century.3 It might seem strange that contaminare would be so divorced from its ethically negative meaning elsewhere in Latin, but then these are the earliest attestations of the word. If this fact casts us back on the etymology, we note its relation to contagio and contingere.4 An important bit of evidence that seems to have dropped out of more recent treatments of the topic is the age in Gellius (12.9) that gives periculum, venenum, and contagium as examples of words that now are only used pro malis, but were once vocabula ancipitia. Perhaps it is not surprising, then, if in the second century BCE the word had more to do with than spoiling. The philological term of art rests on this ethically neutral sense, but often the word shades over almost imperceptibly to a more pejorative usage, implying hasty or clumsy workmanship.5 If blame is already implicit in the word contaminare, as a specification of the kind of mixing, that would explain its censorious use by Terence’s critics, and it would also fit well with the negative sense of the word elsewhere in Latin. We can answer the qualms of Mme Dacier (and the many others who have insisted that contamination must not be a bad thing if Terence its the practice) by calling his statement ironic; he throws back in his enemies’ teeth the same word they have used in their accusation as a blatant transvaluation of their literary code.6 Contaminare gets its rhetorical force here precisely by keeping its strongly negative connotation. Whether by contaminatio is meant merely ‘mixing’ or ‘mixing to bad effect’ (miscendo depravare, as the TLL has it) some notion of blending is fundamental to the academic study of contamination in Roman comedy as it developed in the nineteenth century. It was no less a titan than Friedrich Leo who first proposed taking Terence’s excuse at face value.7 If Terence is imitating Plautine practice, that ought to mean that some of Plautus’ plays are amalgamations of more than one Greek model. Perhaps if we press on just the right places in the Plautine text, the sutures will show themselves, and we can discern the botched cobbling of the 3

We will give less evidentiary weight to the later scholiast, who does not appear in the Terence MSS before the eleventh century, but his comment here is also apposite: NON DECERE id est non fas esse CONTAMINARI id est commisceri fabulas (Schlee 1893, 170). 4 In this etymology I follow Walde-Hofmann and the larger scholarly consensus, but there are those (Hartmann 1919, Pisani 1935) who take contaminare back to tabes. 5 Richter (1914, 40) surveys various nineteenth-century translations of contaminare before settling on Hauler’s translation as zusammanpatzen, which he then turns back into Latin as consarcire. Jachmann (1931, 147) translates the substantive as ein vermischendes Durcheinanderwerfen, Thierfelder (1935) likes rem corrumpere aliquid iscendo and Tredennick (1952) uses ‘spoil by combining’. 6 Cf. Fabia 1888, 179 n.2: ‘Térence défend la légitimité de procédé sans se préoccuper du mot dont ses ennemis le flétrissent; il avoue le fait et, sans lui chercher un autre nom, il l’apprécie tout autrement.’ 7 Leo 1912, 170.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

160

MIMETIC CONTAGION

Roman poet from the pristine artistry of the Greek original. With stunning rapidity, the science of Kontaminationsforschung emerges as a new sub-discipline of Plautus studies, complete with methodological orthodoxy, true believers, and hacks. Of course, this overnight development was predicated on the whole intellectual framework of classics in and to a lesser extent England and elsewhere. From its institutional beginnings in in the late eighteenth century, the modern science of classical philology had claimed the Homeric Higher Criticism not only as its crowning achievement, but as its foundational principle. A single text is refracted with an analytic sieve designed to separate the original from the epigonic. Reading, at its best, consists in the accurate delimitation of strata, invisible to the uninitiated eye. After the high-profile cases, like Homer and Moses, it was relatively easy to apply this technique to any other ancient author.8 What makes the study of contaminatio in Roman comedy interesting is that the text is not thought to be an aggregate product of a single cultural tradition, but rather a violent displacement from one language, culture, and artistic sensibility to another. In this way, the scholarly project resembles the source criticism for, say, an Augustan poet, but here too there are important differences. For Plautus and Terence are not supposed merely to draw on various dispersed literary sources for an obviously distinct artistic product, but rather to bastardize their otherwise maximally faithful translation of one play with a maximally faithful translation of a scene from another play. In such an intellectual culture, the distinction I have indicated between an understanding of contaminare as miscere vs male miscere is finally moot. When purity is a self-evident virtue, all mixing is bad. After Wagner there is a pervasive assumption that it is the peculiar calling of drama in any culture to supply the Gesamtkunstwerk. By being visual as well as auditory, by being kinetic as well as poetic, and ultimately by being communal, the theatre can bring us together and make us whole by being itself a unity. Menander’s plays (as far as anyone could tell) were sensitive and fine-drawn, bringing together the best of tragedy and comedy, tradition and novelty. Since the eighteenth century, Germans had been writing poems about how Menander must have been the original genius of his age. If the hallmark of genius is self-contained totality, how could anyone, above all a Roman, touch those plays without doing serious damage? The problems, both methodological and ideological, with this approach to classical texts are so obvious now that I think I can forgo rehearsing them. Apart from the general circularity and speculativeness that always plagues source criticism, the study of Plautine contaminatio rests on especially brittle ground. For For an illustration of how F. A. Wolf ’s analytic reading of Homer was immediately extended not only to other ancient authors, but also welcomed, at least in some circles, as a guiding principle for contemporary poetics, see 2007. 8

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

THE POETICS OF CONTAMINATION

161

one thing, as Beare never tired of pointing out, Terence may have lied.9 Plautus was certainly free with his models, and it may be that Terence wants us to associate his own practice of combining Greek plays with a more diffuse notion of Plautine neglegentia, even though Plautus himself only added scenes or deleted scenes for comic effect, but never actually combined two models. Apart from these two ages in Terence, there is no evidence that Plautus did anything of the kind; no single Plautine play can conclusively be shown to be ‘contaminated’. Around the same time as the search for Plautine contaminatio was called off, we encounter a more general suspicion of this scholarly category, even with respect to Terence. Already as early as Schwering (1916) most of the argument had been formulated, but it did not win widespread acceptance until much later. It has now often been claimed that there is no attestation of contaminare that absolutely cannot be translated with ‘spoil’ or ‘ruin’.10 The con- prefix may be intensive (OLD, con- 5), rather than collective (OLD, con- 1). Both prologues do in fact use the word contaminare with reference to combining plays, but that may be just a coincidence. Donatus’ comment may be taken as an explanation of the accusation, rather than as a definition of contaminare, or for that matter Donatus may be ignored altogether. But once the focus on textual interweaving is taken totally out of our definition, a new question presents itself: what exactly gets ‘ruined’ and in what sense? There is always a triangle of texts. First, there is the principal Greek model, such as Menander’s Andria; then there is the secondary Greek model, such as Menander’s Perinthia, which is mined for a small amount of material; finally, there is the Latin text produced by this combination, in this case Terence’s Andria. If contaminate means only spoil, which of these texts is contaminated and why? According to Duckworth (1952, 204), ‘A Greek play was defiled (in the eyes of literal translators) if alien matter was added to it from another source; perhaps also the second original was spoiled for later adaptation if a part of it was inserted into another play. It should be noted that to speak of a Latin comedy as being “contaminated” is highly inaccurate; it was always the Greek play that suffered injury.’ The exclusion of the third term, the Latin play, from consideration as ‘contaminated’ is really nothing but a corollary of the elimination of any essential notion of mixture from the definition of contaminare. Spoiling is a thing that can only happen to something that already exists, whereas mixture can be a useful designation both for the process and the result of mixing. The slippage of the perfect ive participle in some semantic fields between predicative and attributive uses is an ambiguity both English and Latin enjoy. We say, ‘Several single malt beverages are blended to produce this fine blended whisky’—the word ‘blended’ 9 10

Beare 1937, 108ff.; 1940, 33ff.; 1948, 70ff.; 1950, 90ff. Beare 1950, 90; Duckworth 1952, 204; Chalmers 1957, 12; Marti 1963, 25.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

162

MIMETIC CONTAGION

is verbal in ‘are blended’ and adjectival in ‘blended whisky’. So the resulting drink can be called ‘blended’ without itself having to suffer anything. If the contamination of Menander’s Andria consists only in the ixture of something from his Perinthia, then surely the blended product can be called ‘contaminated’. Indeed, there is no contamination of these two plays in any sense, apart from the existence of Terence’s Andria. The prologue says that the bits of the Perinthia that fit into the Andria Terence brought over (transtulisse). To be more precise, of course, he brought them over into his own Andria rather than Menander’s, but since the title is the same, he does not need to belabour that distinction. The three plays for which Terence avows combination of models, Andria, Eunuchus, and Adelphoe, all keep the title of their principal model, and there is a resultant blurring between the first and third points, as it were, on the contamination triangle. The prologue to the Eunuch even calls the following play quam nunc acturi sumus | Menandri Eunuchum (19–20). Ultimately this means that whatever is ‘suffered’ by Menander’s play is both inflicted and suffered by Terence’s play. As I say, this exclusion is only a necessary consequence of Duckworth’s axiomatic rejection of mixture from his definition of contaminare, but there is a more fundamental flaw in his argument. If Menander’s Andria and his Perinthia are both spoiled, it is, by Duckworth’s definition, in two very different senses. The first is cluttered by heterogeneous content, where the second is recklessly plucked, for the sake of a relatively small amount of material, from the roster of Greek plays still available for translation. This neat demarcation of a different kind of spoiling for the secondary model only highlights the reliance on hybridity as the sole explanation for the defacement of the principal model. He has effectively said that contaminatio is not ‘mixture’, but just ‘ruining’, except that for the primary model the ruining consists of nothing but mixture! The only way to fix this problem, without itting a definition focused on blending, is to locate the defilement only on the side of the excerpted model and deny that the principal model is corrupted at all. Such a refinement would deliver the most radically minimalist definition yet. This is exactly the strategy adopted by Goldberg 1986. The excerpted text, according to Goldberg, is spoiled for translation by any future Latin poet. He argues that contaminatio can only be understood with reference to the related charge of furtum. Terence is called a ‘thief ’ for mixing two characters from Menander’s Colax into the Eunuch, not because he somehow violated the artistic sanctity of the Eunuch, but because the Colax had already been translated by Naevius and Plautus. The prologus swears that Terence was ignorant of these earlier adaptations, but he points out that by strict application of this criterion you would have comedy without running slaves, good wives, bad mistresses, munchy parasites, braggart captains, long-lost children, old men tricked by slaves, love, hate, suspicion—in short, no comedy at all (Eun. 35–40). For the Adelphoe he mixed in a scene from the Synapothnescontes, a play by

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

THE POETICS OF CONTAMINATION

163

Diphilus which had already been translated by Plautus under the title Commorientes. Here his excuse is not ignorance, but the fact that this particular scene was left out of Plautus’ adaptation (Ad. 9–11). Goldberg argues that this intense focus on copyright explains the worry over contamination, since such a practice unfairly reduces the number of Greek plays available for translation. He says, ‘Terence is accused of violating operational, not aesthetic, rules. Only among modern critics has the charge of contamination become an aesthetic issue.’11 This has been tried before. Beare argued in 1937 for such a focus on copyright, but he changed his mind after Waltz (1938), and throughout the next two decades his argument becomes more and more clearly focused on the defacement of the principal original. His strongest statement appears in a 1959 article in CR. He says, ‘On the copyright theory, the most effective way of “spoiling” a Greek play for rival Latin dramatists would be to translate it outright, as Terence apparently did the Heautontimorumenos. This would mean that all translation was contaminatio.’12 Once again, Beare is quite right. In fact, by this definition, all of Terence is contaminatio, except for the Colax scenes in the Eunuch, which had already been sullied by translation and would presumably, therefore, be unharmed by being used again. Goldberg’s definition of contaminatio turns out to be effectively equivalent to ‘translation’ and almost exactly the opposite of what it needs to be. There are other reasons for doubting this exclusion of aesthetic concern from the definition of contaminatio. Goldberg’s quotation of the Andria prologue, like my quotation earlier (p. 158), begins with line 13. But the four lines previous are essential: Menander fecit Andriam et Perinthiam. qui utramvis recte norit ambas noverit: non ita dissimili sunt argumento et tamen dissimili oratione sunt facta ac stilo. Menander wrote the Andria and the Perinthia If you know one, you know both; That’s how close the plots are— Well, the diction and style are different.

The oratio and stilus will be changed in translation to Latin anyway, but with respect to plot, the part not lost in translation, the plays can be fitted together homogeneously. Given the charge of contaminatio immediately following, the point would seem to be that Terence may have violated a generally accepted aesthetic principle in combining these plays, but that there was no harm done in this case, since the plays were so similar. Another instance of disagreement over

11

Goldberg 1986, 95.

12

Beare 1959, 8.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

164

MIMETIC CONTAGION

translational practice expressed as aesthetic concern for the Greek model may be found in the prologue of the Eunuch. Here the tables are turned and it is Terence accusing Luscius Lanuvinus of spoiling Greek plays, not by contaminatio, but apparently by overliteral translation (7–10): qui bene vortendo et easdem scribendo male ex Graecis bonis fecit Latinas non bonas. idem Menandri Phasma nuper perdidit, atque in Thesauro . . . Who translating well but at the same time writing poorly Made bad Latin plays from good Greek ones. He recently ruined Menander’s Phasma, And in his Treasure . . .

And he goes on to explain how Luscius muddled the Treasure. Again the charge is not contaminatio, but the exalted Greek model can be ruined (perdidit) by clumsy handling in Latin, enough to suggest that aesthetic integrity is a vital concern for these translators.13 My suggestion would not be to return to an understanding of contaminare in the prologues as a purely neutral synonym of miscere, but to follow the TLL’s rendering as miscendo depravare or male miscere. Such a definition acknowledges that the word is always used in castigation, but also that there are times when the notion of mixture is essential to the reproach. As Hugh Tredennick remarked in his rather hostile review of Beare (1950), ‘Professor Beare’s view is that contaminatio in the traditional sense is a pure fiction due to Terence’s lack of candour and Donatus’ stupidity.’14 If we are willing to dismiss the evidence from Donatus on the grounds of his lateness and stupidity, perhaps it will not be difficult to extend our dismissal to his student Jerome, who wrote that he had been accused of merging together different works by Origen: dicunt Originis me volumina compilare et contaminari non decere veterum scripta.15 If Donatus is simply wrong about the meaning of the word, then we must convict Jerome of remarkable acquiescence in repeating his master’s mistake. Turning to earlier authors, we find many attestations of contaminare where it is difficult to decide between a rendering as ‘mix to bad effect’ and ‘ruin’. So, for example, in Pro Sulla 45 Cicero asks: per me ego veritatem patefactam contaminarem aliquo mendacio? (‘Would

13 The MSS for 9 have nunc nuper dedit, but I follow Bothe’s conjecture of nuper perdidit. As Barsby (1999a, ad loc.) says, ‘we seem to need an accusation that Luscius has ruined the play’. Even if the MSS reading is correct and Terence is not saying anything explicitly critical about Luscius’ version of the Phasma, the larger point still stands, since the bungled Treasure is obviously also a wreck. 14 15 Tredennick 1952, 28. Hier. comm. in Mich. 2 p. 480 Vall.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

THE POETICS OF CONTAMINATION

165

I willingly contaminate the obvious truth with some deception?’). One could certainly take the verb here as equivalent to miscendo depravare, but as in so many places, it is not clear that a reading as corrumpere would be wrong, so this text cannot be considered conclusive. We come much closer with a fragment of Accius’ Atreus (206–8 R.), the earliest attestation after Terence: quod re in summa summum esse arbitror periclum, matres coinquinari regias, contaminari stirpem ac misceri genus. What in the greatest house I hold to be The greatest threat: royal matrons corrupted, A contaminated stock, and a family alloyed.

If we take the two expressions in the last line as synonymous, then this is strong evidence of contamination as mixture, but perhaps the previous line’s coinquinari makes this synonymy less conclusive. Similarly, in Livy 4.1.1, the tribune of the people, C. Canuleius, brings a bill on the marriage of patricians and plebes, which is regarded by the patricians as being deleterious of the orders: qua contaminari sanguinem suum patres confundique iura gentium rebantur (‘by which the patricians thought their blood was being contaminated and the laws pertaining to families jumbled’). Jachmann (1931, 152) may be right that there is nothing morally despicable about plebeian blood as such (‘nicht um moralische, sondern um genetische Reinheit handelt es sich’), but separate is never equal—it is the patricians who object to the bill. This is still important testimony, since the contamination consists of the mixing of blood, but if we are looking for a case of contaminare where the mixture itself is clearly at fault and not the quality of the ixed material, this age will not do. In fact, we do have one such attestation. Here, too, the context is miscegenation, but the perspective is reversed, since it is a patrician transitio ad plebem. Cicero criticizes Clodius for his sham adoption by P. Fonteius (De domo 35): non aetas eius qui adoptabat est quaesita, ut Cn. Aufidio M. Pupio, quorum uterque nostra memoria summa senectute alter Oresten alter Pisonem adoptavit, quas adoptiones sicut alias innumerabiles hereditates nominis pecuniae sacrorum secutae sunt. tu neque Fonteius es qui esse debebas, neque patris heres, neque amissis sacris paternis in haec adoptiva venisti. ita perturbatis sacris, contaminatis gentibus, et quam deseruisti et quam polluisti, iure Quiritium legitimo tutelarum et hereditatium relicto factus es eius filius contra fas cuius per aetatem pater esse potuisti. He doesn’t have the requisite age, like Cn. Aufidius or M. Pupius, who, we , as old men adopted Orestes and Piso, respectively. And after those adoptions, as after countless others, there was an inheritance of name, estate, and cult. But you’re not ‘Fonteius’, as you ought to be, nor are you your father’s heir, nor have you taken up your adoptive family’s cult, though you’ve sure left your own. So the cults are a mess, the

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

166

MIMETIC CONTAGION

families are a jumble (the one you’ve dropped and the one you’ve polluted), and our established ancestral law regarding guardianship and inheritance lies forsaken—it’s unconstitutional, but you’ve become his son and you’re old enough to be his father!

It was just possible to read the Accius and Livy ages as describing miscegenation only from the dominant perspective, as pollution, but here the ixed agent is socially higher than his adoptive family. Cicero is not bothered by the change of gens as such, as the respectfully cited cases of Cn. Aufidius Orestes Aurelianus and M. Pupius Piso Calpurnianus show. A legitimately undertaken adoption should show three types of heredity: name, estate, and cult (hereditates nominis pecuniae sacrorum). The next sentence shows that Clodius’ adoption is bogus on all three counts. He has not taken Fonteius’ name, he is not his inheritor, and he is in a limbo of family cults. Finally, the ablative absolute of the last sentence repeats these three items, only changing the order to cult (perturbatis sacris), name (contaminatis gentibus), and estate (iure . . . hereditatium relicto). The thrice repeated tripartite structure reveals that by ‘jumbled families’ Cicero means Clodius’ problematic name. By being in the Fonteian gens without taking the nomen gentile, Clodius effects in his own person an unlawful blending of two families and becomes a monster of the clan system. For how can clan be divorced from name? As Cicero says elsewhere (Top. 29), gentiles sunt inter se qui eodem nomine sunt (‘clansmen are people who share the same name’). This age from the De domo is, I believe, by itself sufficient proof that contaminare could be used to cover cases of unhappy mixture, where neither of the two ingredients is bad but the manner of their combination violates etiquette. Taken together with the many other cases that suggest this reading without necessarily requiring it, and adding in the evidence of etymology, Gellius’ testimony on the early meaning of contagium, the direct witness of Donatus and Jerome, and, above all, the sense logically required by Terence’s prologues, it seems to me that recent scholarship has gone too far in taking mixture totally out of the definition of contaminare. Certainly the reductive understanding of contaminatio, that sees ‘ruining’ only in the prodigal consumption of Greek models, is insufficient and cannot be sustained. It is evident that the Latin poets are at least also concerned with damage to the principal original in the resulting dramatic product. On the other hand, the insistence of scholars like Duckworth that contaminare means ‘pollute’ or ‘spoil’ without any implicit notion of mixture is ultimately nonsensical, if the only thing the principal model suffers is hybridization. It would seem that Goldberg’s position is the logical, if extreme, conclusion of a line of argumentation that began as the salubrious corrective to the excesses of early twentieth-century German Altphilologie. We laugh now at their absurdly positivistic quest for Plautine contaminatio, and we stop laughing when we note their patently racist philhellenism and their eagerness to shred the texts we do still

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

THE POETICS OF CONTAMINATION

167

have, but our reaction to their abuses has left us with a definition of contaminare unable to for its use in the Terentian texts where the practice actually is attested.

CONTAMINATED EUNUCH

I wish to suggest that because of our ever more restricted and now inappropriately narrow understanding of contamination, we have failed to catch the Eunuch’s playful awareness of its own hybridity. We have examined the two attestations of contaminare in the prologues, but Terence uses the word one other time, in the scene that has been the focus of this book. Chaerea emerges from the house where he has just raped Pamphila, and though he is spied by his friend Antipho, he believes he is alone (549–62): CH. numquis hic est? nemost. numquis hinc me sequitur? nemo homost. iamne erumpere hoc licet mi gaudium? pro Iuppiter, nunc est profecto interfici quom perpeti me possum, ne hoc gaudium contaminet vita aegritudine aliqua. sed neminemne curiosum intervenire nunc mihi qui me sequatur quoquo eam, rogitando obtundat enicet quid gestiam aut quid laetu’ sum, quo pergam, unde emergam, ubi siem vestitum hunc nanctu’, quid mi quaeram, sanu’ sim anne insaniam! AN. adibo atque ab eo gratiam hanc, quam video velle, inibo. Chaerea, quid est quod sic gestis? quid sibi hic vestitu’ quaerit? quid est quod laetus es? quid tibi vis? satine sanu’s? quid me aspectas? quid taces? CH. o festu’ dies hominis! amice, salve: nemost hominum quem ego nunc magis cuperem videre quam te. AN. narra istuc quaeso quid sit. CH. immo ego te obsecro hercle ut audias. CH. Anyone here? Nope. No one following me? Not a soul. Can I let my joy out now? By god, I could stand to die at this moment before Life contaminates this joy with some distress. Amazing that there’s no busybody coming my way right now, Following me wherever I go, plaguing me, wearing me out with questions: Why I’m so ecstatic and happy, where I’m going, whose house this is, Where I got this get up, what I’m up to, am I sane or nuts! AN. I’ll go up and do him the favour I see he wants. Chaerea, why so ecstatic? What’s with the outfit? Why are you so happy? What are you up to? Are you OK? What are you looking at me for? Why the silence? CH. Oh, you human fiesta! Hello, friend! There’s no one in the world I’d rather see right now than you. AN. Tell me what it is, won’t you? CH. Oh no, I’m begging you to listen!

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

168

MIMETIC CONTAGION

Does Chaerea want an interlocutor or not? The exclamatory infinitive does not tell us whether he is relieved or disappointed at being alone, and scholars have disagreed on this point.16 Büchner, with his usual sensitivity to dramatic ambiguity, argues that Chaerea’s wishes are complex and not totally fixed.17 His surprise at being left alone certainly sounds grateful: curiosus is always a negative word and of course obtundere and enicare are also things to be avoided. But Büchner may be right to read a certain longing into this longwinded exclamation. As Donatus says (ad loc.): id cupit gaudens, quod aliis tristibus molestum est (‘He is happy and he wants something that is irritating for people when they are sad’). Antipho obviously understands the exclamation the same way, and, as far as we can tell from his welcome, he knows his friend’s mind. At first, though, Chaerea is only eager to confirm that he is alone so he can vent his joy. This is the summit of his life, the moment of perfect attainment, when, like Cleobis and Biton, he could die and be called happy, for no future changes could then alter his state. It is here that Terence uses contaminare for the last time. The second half of the play will, in fact, furnish Chaerea with plenty of opportunity for distress, but ‘life’ has already added something to surprise him: throughout his encomium of solitude he is not alone. Antipho was not in Menander’s Eunuch, as Donatus tells us (ad 539); Terence wove him in to break up Chaerea’s monologue (ne unus diu loquatur, ut apud Menandrum). Antipho’s name essentially just means ‘protatic character’ and his function is simply to motivate the exposition of the rape by asking all the questions Chaerea has just said would be so obnoxious. On the subject of redende Namen, Chaerea’s name suggests the gaudium he is afraid life will contaminate with some trouble. Terence has tinkered with Menander’s play and mixed in a new character to a scene that was ‘supposed’ to be all Chaerea. Of course, in order to catch the contamination joke the audience will have to know that Menander’s scene was a monologue. In general, Terence seems to assume that part of his audience has familiarity with Menander (cf. Haut. 7–9), but it would be unwise to suppose that many people in the audience of the Eunuch knew Menander’s text well enough to follow the irony here. That is unless their attention had already been drawn to this age. As it turns out, this line about contamination had already been contaminated. At line 959 of the Andria, the ecstatic adulescens Pamphilus thinks he is alone and so he lets out:

16 For the ‘relieved’ reading, see Pasquali 1936, 119 and more recently Barsby 1999a, ad loc.: ‘the immediate point of these questions is to establish that Chaerea has got away with his trick; there is no one to raise the alarm and he is free to give vent to his joy’. For the ‘disappointed’ reading, see Fabia 1895, ad loc. and more recently Tromaras 1994, 195: ‘sofort verspürt er das Bedürfnis, einer vertrauten Person sein Abenteuer mitzuteilen’. 17 Büchner 1974, 268.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

THE POETICS OF CONTAMINATION

169

ego deorum vitam propterea sempiternam esse arbitror quod voluptates eorum propriae sunt; nam mi immortalitas partast, si nulla aegritudo huic gaudio intercesserit. I hold the life of the gods to be eternal, inasmuch As their joys are abiding; for unto me immortality Is born, if no distress breaks into this joy.

This is just the sort of thing Terence got in trouble for with the Andria, as Donatus says (ad An. 959): hanc sententiam totam Menandri de Eunucho transtulit et hoc est quod dicitur contaminari non decere fabulas (‘He lifted this whole sentence from Menander’s Eunouchos; this is what they mean about, “plays ought not be contaminated”’). So you would not have to be a Menander expert or have an encyclopaedic knowledge of Terence’s earlier plays to understand the contamination allusion at Eunuch 552. He had taken this sentence out of a soliloquy in Menander’s Eunouchos spoken by a character who was happy to be alone and mixed it into his Andria for a character who was looking for someone else to tell his story (Davus), while being overheard by a third person (Charinus). Indeed, the next line (962) is: sed quem ego mihi potissumum optem, nunc quoi haec narrem, dari? (‘But who would be the ideal person to pop up right now for me to tell this stuff to?’). If this inconcinnity disconcerted overscrupulous critics, it will offend them even more when it is mixed back into the Eunuch with situational traces of the Andria still on it! All the while flouting their cavils by replacing intercesserit with the very word they had used in their accusations: contaminet. As the prologus had said in the Heautontimorumenos, he was not sorry he did it and he would do it again. The primary contamination in the Eunuch needs no discovery; there has never been any doubt about what does not belong. As the prologue says, Terence does not deny that he brought over the soldier and the parasite from Menander’s Colax (eas se non negat | personas transtulisse in Eunuchum suam, 31–2). But these characters are strangers in this play in more ways than one. The scenes in which they feature have a markedly different feel than other scenes in the play, with considerably more rowdiness and vulgarity, so they are alien not only to the text, but to the ethos of the Eunuch.18 They are also strangers in a more obvious sense. They are not Athenians or even resident aliens, and they are not lodging in either of the two onstage houses. They have apparently blown in from out of town and were supposed to be ing through, but they end up wanting to stick around. By the end of the play, they have successfully moved in and implicated themselves in the lives of the other characters.

18 This ethical discrepancy may go back to differences in tone between Menander’s Colax and Eunouchos. For the significance of these differences, as indicated by the fragments, see Büchner 1974, 287.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

170

MIMETIC CONTAGION

Perhaps the most important age, for our understanding of Thraso and Gnatho and their relation to the play’s themes, is the scene in Act 2 where we meet Gnatho for the first time. His monologue narrates an offstage encounter, like Chaerea’s narration of the backstage rape, and also like Chaerea’s report, it touches the theme of improvisational imitative role-playing, specifically the mime’s role of extempore mirroring that will be so important in Chaerea’s response to the painting (232–64): GN. Di inmortales, homini homo quid praestat? stulto intellegens quid inter est? hoc adeo ex hac re venit in mentem mihi: conveni hodie adveniens quendam mei loci hinc atque ordinis, hominem haud inpurum, itidem patria qui abligurrierat bona: video sentum squalidum aegrum, pannis annisque obsitum. ‘oh quid istuc’ inquam ‘ornatist?’ ‘quoniam miser quod habui perdidi, em quo redactu’ sum. omnes noti me atque amici deserunt.’ hic ego illum contempsi prae me: ‘quid homo’ inquam ‘ignavissime? itan parasti te ut spes nulla relicua in te siet tibi? simul consilium cum re amisti? viden me ex eodem ortum loco? qui color nitor vestitu’, quae habitudost corporis! omnia habeo neque quicquam habeo; nil quom est, nil defit tamen.’ ‘at ego infelix neque ridiculus esse neque plagas pati possum.’ ‘quid? tu his rebu’ credi’ fieri? tota erras via. olim isti fuit generi quondam quaestus apud saeclum prius: hoc novomst aucupium; ego adeo hanc primus inveni viam. est genus hominum qui esse primos se omnium rerum volunt nec sunt: hos consector; hisce ego non paro me ut rideant, sed eis ultro adrideo et eorum ingenia iror simul. quidquid dicunt laudo; id rursum si negant, laudo id quoque; negat quis: nego; ait: aio; postremo imperavi egomet mihi omnia adsentari. is quaestu’ nunc est multo uberrimus.’ PA. scitum hercle hominem! hic homines prorsum ex stultis insanos facit. GN. dum haec loquimur, interealoci ad macellum ubi advenimus, concurrunt laeti mi obviam cuppedenarii omnes, cetarii lanii coqui fartores piscatores, quibus et re salva et perdita profueram et prosum saepe: salutant, ad cenam vocant, adventum gratulantur. ille ubi miser famelicus videt mi esse tantum honorem et tam facile victum quaerere, ibi homo coepit me obsecrare ut sibi liceret discere id de me: sectari iussi, si potis est, tamquam philosophorum habent disciplinae ex ipsis vocabula, parasiti ita ut Gnathonici vocentur. GN. Gods above, how doth one man sur another! The difference Between a wise guy and a chump! Here’s what’s got me thinking: Coming here today I met a man of my rank and station from hereabouts,

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

THE POETICS OF CONTAMINATION

171

A basically decent guy who, like me, had guzzled away his inheritance. What I see is a shabby, filthy, rickety mess overgrown with rags and age. ‘Ugh,’ I say, ‘What’s up with the get up?’ ‘Well, I’ve lost all I had, so It’s come to this. Everyone I know, all my friends—they ditched me.’ At this point the contrast with myself was enough to make me detest him. ‘You sucker!’ I said, ‘Have you managed to lose every shred of hope? Have you misplaced your mind along with your money? Look at me; I hail from the same class. Note my complexion, My glossy sheen, my apparel, my physical condition! I have everything, yet I have nothing. There’s nothing there, but there’s nothing wanting.’ ‘Well, sap that I am, I can’t be a clown and take a beating.’ ‘What? You think that’s how it goes down? You’ve got it all wrong. Once upon a time, in another age, maybe that’s how they survived. It’s a whole new game these days; in fact, it’s a trail I blazed. There is a type of men who wish to be number one in everything, But they’re not. These are the ones I pursue, and I don’t make myself A butt for their jokes; no, I laugh at them, All the while wondering at their wit. Whatever they say I praise, and if they deny it I praise that too. He nays, I nay; he ayes, I aye. In short, I have imposed a rule on myself: Conformity in all things. This is by far the most profitable line.’ PA. The guy’s amazing! He makes nuts out of chumps. GN. While we’re talking like this we get to the market, And all the vendors run up, happy to see me— Fishmongers, butchers, cooks, sausage stuffers, sprat sellers, For whom, bull market or bear, I’ve been and always will be a windfall. They greet me, invite me to dinner, welcome me to town. When that poor starveling sees the esteem in which I am held and The ease with which I procure my living, then and there he starts pleading To be allowed to learn from me. I advise him to enrol as my pupil, So, just as philosophical schools take their names from their masters, Perhaps parasites will be called Gnathonists.

The philosophical overtones are obvious even to us, but they will have been even more glaring to Terence’s audience in 161 BCE, the very year in which all Greek philosophers and rhetoricians were expelled from Rome.19 Twelve years before, only the Epicureans had been banished, but in the year of the Eunuch’s staging there was, at least as a legal fiction, a total purge of these deleterious educators.20 Suet. Rhet. 1.2; Gell. NA 15.11.1. For an analysis of the sources, see Garbarino 1973, 370. Gruen (1990, 171–2) argues that the expulsion could not have been successful or even taken too seriously, but then he argues the same for sumptuary laws (170–3), the repressions of the Bacchanalia (34–78), and most other domestic legislation. But even if these actions only had symbolic value, that is enough for our purposes. 19 20

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

172

MIMETIC CONTAGION

This is not the only bit of legislation from 161 relevant to Gnatho’s monologue. The first half of the second century had seen occasional sumptuary laws, but the lex Fannia of 161 beats its predecessors, not only in of strictness, but also of specificity.21 It was directed to banquets given during the Ludi Megalenses, not only the year but the very occasion of the Eunuch’s production, and set very explicit limits on just how much money could be spent on the kind of sweetmeats and dainties Gnatho describes. We have fragments of a speech in of the law, saying that the purpose of these restrictions is to protect the Roman youth from their own prodigality, again a clear focus on the educational aspect of the issue.22 It is as a teacher that Gnatho emerges from this monologue. He mirrors the words and attitudes of his sugar daddy, but he also makes other men like himself, most obviously of course his new friend who will convert to professional bootlicking, but note also how the small tradesmen fawn on him and captate for his favour. The irony here is rich: his success at toadying puts him in a position where the vendors want to kiss up to him, and it is only upon seeing this adulation that his interlocutor wishes to emulate him. Their imitation of him is unconscious, his disciple’s imitation will be studied, but one way or the other, all the world is drawn into his ethical paradigm. More specifically, it is not Gnatho’s rhetoric that turns his friend’s mind, but rather witnessing his reception with the merchants. Donatus’ comment (ad 260) is essential: hic ostendit, quae res coegerit ad discendum . . . vide ut sententiose demonstret malos ex bonis contagione fieri, exemplis in pravum praevalentibus: ‘videt’ mihi hoc prodesse et discere optat, quod negabat se posse. Here he reveals what impels us to discipleship . . . Note how vividly he shows good men being made bad by contagion, since models for the worse are overwhelming: he sees that it is profitable for me and he longs to learn what he has said he could not do.

It is precisely in spectating (videt) the triumph of bad behaviour over good that men are corrupted, as Donatus says, ‘by contagion’. The relevance of all this to the themes of vision, role-playing, and mimetic contagion is too obvious to need stress. I point out here only that Gnatho is the element of contamination, a character alien to Menander’s Eunuch who has been, literally, ‘mixed in to bad effect’. From his first appearance in the play we see him as a spreading infection, and this is just the semantic field Donatus uses to describe him.

21 See Gell. NA 2.24.2–6 for the broadest survey of the restrictions. Athenaeus (6.274c) details some of the spending caps on food, both daily and annual, going into special detail for such delicacies as smoked meats and the finer vegetables. Pliny (NH 10.139) reports that there was a limit of one hen per meal and an outright ban on artificially fattened poultry. 22 Macrob. Sat. 3.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

THE POETICS OF CONTAMINATION

173

Altogether there are four scenes that feature the parasite with or without the soldier.23 It is striking how little they interact with other characters and how generally superfluous they are to the action of the play. Apart from delivering Pamphila to Thais’ house and causing Thais’ absence from her house at a key moment, they have no structural necessity to the plot, and both these effects could easily have been accomplished by other means. The hilarious scene where Thraso leads an assault on Thais’ house is oddly stuck on at the end of Act 4, and indeed the whole point of it is that the siege is easily aborted and comes to nothing.24 Critics who have prized Terence for his structural finesse and dramaturgical economy have tended to puzzle over this lapse and complain about the strangely prosthetic feeling of these scenes.25 One interesting consequence of this fact, however, is that back in the day of serious Kontaminationsforschung, critics could mentally delete these scenes and imagine with relative confidence that they were coming close to recovering Menander’s Eunouchos.26 But the scholars who attempted this surgery discovered that the operation founders on the last of the four scenes, which is also the last scene of the play. The resolution of the comedy is totally dependent on the two interpolated characters, and no efforts to remove them from the last scene have proved satisfactory in the least. In fact, Walther Ludwig goes so far as to say that the end of the play cannot, in any meaningful sense, have come from either of the Greek plays and must be Terence’s own invention.27 I certainly agree with this assessment, but I point out that it means a radical departure from the way he has combined the plays up to this point and also from what we were given to expect in the prologue, where he spoke only of the addition of fungible stock characters in modular generic scenes.

Gnatho is onstage, with or without Thraso, for lines 232–87; 391–506; 771–816; 1025–94. Ludwig (1959, 22–7) argues that there must have been something like this scene in the original. Thais has been hoping that Chremes would become her patron in gratitude for the restoration of his sister, but by the end of the play she looks towards Chaerea’s father for patronage. Ludwig sees a neat preparation for this shift in line 770: perii, huic ipsist opu’ patrono, quem defensorem paro (‘Drat, the man I’m snagging for a protector needs a patron himself!’). If this line does indeed go back to Menander then (according to Ludwig) there must have been some scene where ‘Chremes’ (the model for Terence’s Chremes) needs propping up, probably because he is afraid of ‘Thraso’, who must have been threatening legal action or physical assault, and the latter is more dramatically appealing, so Ludwig guesses that ‘Thraso’ was a soldier too. This is, to say the least, a tenuous thread of speculation. I agree that there is a nice irony in line 770, but hardly enough to provide sole motivation for this whole scene. If Menander (or Terence) just wanted the conceit behind this line, Chremes is marked throughout as a rusticus, and it is easy to imagine other possible situations that might have led Thais to lose confidence in his initiative or savvy. In any case, when Thais and Chremes depart the stage in line 810 they are ignoring Thraso (quaere qui respondeat) and going about the business they would be doing if he and his warrior band were not there at all. 25 See, for example, the spirited censure of Norwood 1923, 64–9, who attempts to show how these two characters are at every level a pointless blight on the construction of the play and not hard to remove. 26 The fullest synthesis of this type of work on the Eunuch was Meyerhöfer 1927. 27 Ludwig 1959, 36–8. 23 24

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

174

MIMETIC CONTAGION

Not only are the interpolated characters impossible to extricate from the dramatic structure of the denouement, but also, within the fiction, the end of the play implicates them in the lives of the other characters. Before this happens, there is a moment when the action of the play seems to have ed Thraso and Gnatho by and made them irrelevant. Phaedria, Chaerea, and Parmeno are all onstage and exultant because everything has played out to their advantage. ‘There’s another thing we can be happy about,’ says Parmeno in line 1041, ‘the soldier will be pushed out (miles pelletur foras).’ This comment is the first of several in this scene with theatrical double entendres. When Thraso approaches them, a few lines later, cap in hand, Phaedria thinks maybe he has missed Act 5 (1061–5): PH. tu fortasse quae facta hic sient nescis. TH. scio. PH. quor te ergo in his ego conspicor regionibus? TH. vobis fretu’. PH. scin quam fretu’? miles, edico tibi, si te in platea offendero hac post umquam, quod dicas mihi ‘alium quaerebam, iter hac habui’: periisti. PH. Maybe you don’t know what’s been happening. TH. I do. PH. So why am I still seeing you around here? TH. At your charity, sir. PH. Don’t you know it, soldier! Let me tell you something: if I ever run into you again on this street, I don’t care if you say, ‘I’m looking for someone else down this way’—you’re done for!

Phaedria repeats Thraso’s stock character name (miles), and ‘this street’ is, of course, the stage; the deictic references (his . . . regionibus; platea . . . hac; iter hac) indicate just what you see. This is the moment where Gnatho gets involved and begins brokering the deal that will close the play. He urges the boys, or perhaps specifically Phaedria, to accept the soldier as a rival (militem rivalem ego recipiundum censeo, 1072). This repetition of miles in apposition now with the role of the comic standing contender, rivalis, suggests that what Gnatho is imagining is not a resolution at all, but virtually a set-up for a new comedy. Phaedria is incensed, naturally; he has been trying all day to get rid of this creep, why would he willingly submit to such an arrangement? Gnatho reminds him how much he enjoys victitating with Thais, how expensive she is, and how little he has to give. If Thraso is kept on and given his share, he can be fleeced for all expenses, since he is rich, liberal, and stupid, and they can always get rid of him, if they want, because there is no chance Thais would fall in love with him. Indecent proposal or not, as Chaerea says, it sounds like Thraso is necessary anyway (quoquo pacto opust, 1083), and Phaedria agrees. ‘Right, then,’ says Gnatho, ‘but there’s one more thing I have to ask of you: that you accept me into your company (ut me in vostrum gregem | recipiati ’, 1084–5). I’ve been rolling this stone long enough.’ The word grex almost always has a contemptuous

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

THE POETICS OF CONTAMINATION

175

tone, in Latin more generally and certainly in Terence.28 However, in this case contempt is clearly precluded by the context, so we are virtually forced to understand the word metatheatrically, since it is also the term for an acting troupe.29 Note the repetition of the verb recipere from his first suggestion, a few lines before (1072), that Thraso be itted as a rival. The reception he is trying to win for Thraso and for himself is essentially the same: ission to the cast of the Eunuch. The comment about the rolling stone sounds cryptic enough to be proverbial, and Donatus (ad 1085) confirms that it is an idiom, drawn from the myth of Sisyphus, and used as he says for those ‘afflicted with inextricable work’ (proverbium in eos, qui inextricabili labore afflicti sunt). But the labour of Sisyphus is not only unending, it is also repetitive. The prologue, as we have seen above, insisted that Terence did not know the Colax had already been translated by Plautus and Naevius. Whatever we want to make of this protestation of innocence, one thing it certainly accomplishes is to inform the audience, if they did not already know it, that these two characters are now seeing the boards for at least the fourth time. By this point, one may well forgive the parasite for wanting a change of company. He gets his wish, naturally (recipimus, 1085), and all’s well that ends well. Thraso, too, is pleased things have worked out, but he cannot exactly be surprised; after all, the braggart soldier is always a hit. ‘I’ve never yet been anywhere,’ he says, ‘where everyone didn’t absolutely love me.’ ‘What did I tell you?’ says the parasite to his new hosts, ‘Attic charm.’ ‘Just what you promised,’ says Phaedria, ‘Come this way.’ And with this ite hac (1094) the interpolated characters are finally welcomed into Thais’ house and into the world of the Eunuch, and the play ends. The comment about Thraso’s Attica elegantia is a joke, of course; he will always be a boor. But he has now become indispensable to these Athenians and will have a home here, just as he and his toady have become structurally necessary to the end of the play. All this happens because the brothers agree to take the parasite’s advice, advice which consists in nothing other than becoming sponges like himself. ‘I give you this man,’ he says, ‘to feast on and ridicule’ (hunc comedendum vobis propino et deridendum, 1087). Among modern readers, Thais is one of the best-liked prostitutes in Roman comedy.30 We find her principled, loyal, humane, and strong. In scene after scene her remarkable qualities stand out, especially against the backdrop of all the weakwilled, egotistical men in the play. For this reason the final arrangement between Thraso and Phaedria has generally offended critics and seemed unjustly demeaning 28 The two other places Terence uses grex are Haut. 245 and Ad. 362, both plainly derisive. For comparison to Ciceronian and Horatian usage, see Fantham 1972, 56. 29 OLD s.v. grex 3b. See Frangoulidis 1994a, 130. 30 Norwood 1923, 60: ‘She is a splendid creation. To class her with the boisterous harlots of Plautus’ Bacchides and similar plays, merely because of the manner in which she happens perforce to earn her living, would be an outrage.’

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

176

MIMETIC CONTAGION

of this rara avis.31 But Thais is not the only one degraded by the ending. The play had opened with Phaedria’s profession of helpless love, a scene so striking for its language and its ethical intensity that it turns up in Cicero, Horace, and Persius, but here he is at close of day, effectively pimping her out.32 As Karl Büchner says, the only way this ending can be appreciated is if we forget everything that has come before.33 On the other hand, I would argue, this formulation is precisely wrong, for we can only appreciate the ironic artistry of the ending to the extent that we are conscious of this radical discontinuity of character. From the prologue on there is never a moment when we are allowed to forget the foreignness of the interpolated characters, and the end of the play can only integrate them by ethically denaturing everyone else. If this tale of moral infection is disturbing to us, perhaps it should be, and if it casts a shadow over the generically prescribed happy ending, then it only makes the Eunuch, like every other Terence play, something of a black comedy. One never sees the two-sidedness of contamination in this play, until one attempts, succeeds at first, and then finally fails to recover Menander’s Eunouchos from this text. The scholars from the early twentieth century who did just that and erected such an edifice of confusion have generally been disregarded for the last twenty if not fifty years as a prime example of bad methodology, bad aesthetics, and bad politics. But the thing that distinguishes this work on the Eunuch from Plautine Kontaminationsforschung, is that it is explicitly authorized by the text itself. Whatever misguided conclusions these scholars may have come to because of other convictions, in this regard they were only following the bouncing ball that Terence himself set in motion. Presumably the audience of 161 BCE may have included people who knew the Menandrian originals well, but I think we can be fairly sure that the vast majority of viewers did not. Perhaps they knew other plays or perhaps they had some vague awareness of the themes and style of Greek New Comedy, if only through its reworking as fabulae palliatae. That is to say, the average Roman viewer probably knew roughly as much about the sources of the Eunuch as we do. But when the text itself seems to suggest so loudly that we think about the Greek originals that lie just behind what we are seeing, one may well suppose that many viewers have walked similar hermeneutic paths to the ones trod

31 Pasquali 1936, 128: ‘quest’ ultima scena, cosi indegna di quella creatura gentile, non avrà forse offeso la maggior parte degli spettatori, ma avrà offeso i più fini anche dei lettori contemporanei di Terenzio; a noi moderni è insopportabile’. Sandbach (1977, 144–5) suggests that perhaps Terence did not fully understand the dignity with which a hetaira (as opposed to a porne ) should be treated, since Latin did not readily comprehend the distinction. Konstan (1986, 384–5) reads the moral ambiguity of the ending as a successful meditation on the complexities of commercial love. Not all readers have been so disturbed by the conclusion; see, for example, Brown 1990, 49–61; Goldberg 1986, 116–22. 32 Cicero: Tusc. 4.76, Nat. deor. 3.72; Horace: Serm. 2.3.259–71; Persius: 5.161–75. 33 Büchner 1974, 303–5.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

THE POETICS OF CONTAMINATION

177

by early twentieth-century scholarship. Our almost total ignorance of Menander before the 1960s would have made us close to what Wolfgang Iser calls ‘intended readers’. The irony is that these scholars thought of the text almost as an object under a microscope and their own project as one of scientific detachment, when it now seems conceivable that their attempt and failure to decontaminate the Eunuch was an extension of their own implication in the text.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

Epilogue

Die guten Abderiten waren so voll von dem, was sie gehört und gesehen hatten, daß sie sich genötigt fanden, ihrer Überfüllung noch auf andere Weise Luft zu machen. C. M. Wieland, Geschichte der Abderiten

On a hot summer’s day in Abdera, so the story goes, a production of Euripides’ Andromeda infected its audience with tragedy. Lucian tells this strange anecdote at the beginning of How to Write History (59) as a comparison to the epidemic of historiography that had spread since the inception of the Parthian War. The disease (νόσημα, 1.2) included seven days of fever, followed by either nosebleeds or sweats, but after the fever broke the Abderites were left in a ridiculous condition (1.6–21): ἐς γελοῖον δέ τι πάθος περιίστη τὰς γνώμας αὐτῶν· ἅπαντες γὰρ ἐς τραγῳδίαν παρεκίνουν καὶ ἰαμβεῖα ἐφθέγγοντο καὶ μέγα ἐβόων· μάλιστα δὲ τὴν Εὐριπίδου Ἀνδρομέδαν ἐμονῴδουν καὶ τὴν τοῦ Περσέως ῥῆσιν ἐν μέλει διεξῄεσαν, καὶ μεστὴ ἦν ἡ πόλις ὠχρῶν ἁπάντων καὶ λεπτῶν τῶν ἑβδομαίων ἐκείνων τραγῳδῶν, ‘σὺ δ᾽ ὦ θεῶν τύραννε κἀνθρώπων Ἔρως’, καὶ τὰ ἄλλα μεγάλῃ τῇ φωνῇ ἀναβοώντων καὶ τοῦτο ἐπὶ πολύ, ἄχρι δὴ χειμὼν καὶ κρύος δὲ μέγα γενόμενον ἔπαυσε ληροῦντας αὐτούς. αἰτίαν δέ μοι δοκεῖ τοῦ τοιούτου παρασχεῖν Ἀρχέλαος ὁ τραγῳδός, εὐδοκιμῶν τότε, μεσοῦντος θέρους ἐν πολλῷ τῷ φλογμῷ τραγῳδήσας αὐτοῖς τὴν Ἀνδρομέδαν, ὡς πυρέξαι τε ἀπὸ τοῦ θεάτρου τοὺς πολλοὺς καὶ ἀναστάντας ὕστερον ἐς τὴν τραγῳδίαν παρολισθαίνειν, ἐπὶ πολὺ ἐμφιλοχωρούσης τῆς Ἀνδρομέδας τῇ μνήμῃ αὐτῶν καὶ τοῦ Περσέως ἔτι σὺν τῇ Μεδούσῃ τὴν ἑκάστου γνώμην περιπετομένου. The funny thing was that some affliction befell their minds, for everyone went crazy with tragedy and they were squawking iambics and bellowing fortissimo. Most of all they sang solos from Euripides’ Andromeda and recited Perseus’ lines in metre, and the city was full of all these pale, thin, one-week tragedians braying, ‘O thou tyrant of gods and men, Eros,’ etc. in a loud voice. And so it was for a long time until winter, and there was a cold snap, which stopped their nonsense. The cause of all this, it seems to me, was furnished by the actor Archelaus, who was popular at that time. In the middle of summer, at the height of the heat, he performed the Andromeda, so that most people caught fever from the theatre and later, when they had arisen from their sickbeds, they slipped back into the tragedy. The Andromeda kept haunting their memory, and Perseus with his Medusa was flitting around each person’s mind.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

EPILOGUE

179

Lucian relates the story as traditional (φασί, 1.1), and perhaps the specificity of its placement in Abdera during the reign of Lysimachus and other particular details indicate a received legend, but given the vividness of his imagination, we cannot be sure he did not invent the story for this occasion.1 It matters little for our purposes whether the anecdote goes back to the fourth century BCE or comes straight from Lucian’s pen; what is significant for us is the story’s combination of theatrical performance and contagious imitation that spreads beyond the usual boundary of dramatic art. Recent scholarship on this anecdote has focused, quite rightly, on its relevance to the broader themes of How to Write History, specifically Lucian’s critique of rhetorical enthusiasm.2 Though such approaches have been very successful and are clearly necessary for a full appreciation of the significance of the anecdote within the treatise at large, our interest here is more narrowly circumscribed. The Abderite audience’s imitative response will remind the reader of many other sympathetic reactions to art and performance we have seen in this study. The Abderites were proverbially stupid, and perhaps we are to understand that greater sophistication would have made them immune to this extremity of mimetic contagion, but though the play may not have excited the same response from every audience, its effect here is obviously not a matter of merely rational persuasion.3 The heat of the day and the ardour of the performance combined to draw the impressionable Abderites into a sympathetic fever. As an explanation for being stuck in the Andromeda, this νόσημα is no more elective than the magnetism operating on the iron rings in Plato’s famous analogy for poetic inspiration in the Ion (533d1–536d3).4 Like a loadstone for an iron ring, the Muse draws the poet to herself, but along with the poetry, she bestows enough ‘divine power’ (θεία δύναμις, 533d3) on the poet that he can now transmit the same enthusiasm on to others. Ion its to having affective reactions consonant with the poetic scenario he is singing and says he can look out from the stage and see that the audience is experiencing the same ions, and Socrates explains that they are simply the last ring of the magnetized chain. Similarly, the Andromeda appears in Abdera as a communicable disease, which assimilates the audience willy-nilly, first to its heat and then to its language.

1

Another version of the same story has been transmitted in the fragments of the Historika hypomnemata, compiled by Eunapius of Sardis (345–c.420 CE). Eunapius’ is closely similar to, but apparently independent of, the version in How to Write History, so it seems most likely the story does not originate with Lucian. For the text, see Blockley 1983, fr. 48. There are many other broadly similar stories in the paradoxographic literature; see Anderson 1976, 61. 2 See especially Fox 2001 and Möllendorff 2001. 3 For a full survey of ancient prejudice on the stupidity of the Abderites, see Tschiedel 1986. 4 On Platonic enthusiasm, see Büttner 2000, 255–365.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

180

MIMETIC CONTAGION

The loss of Euripides’ Andromeda is surely one of the greatest misfortunes of the transmission history of Greek tragedy. First performed in 412 BCE and frequently revived thereafter, it was evidently among Euripides’ most moving works.5 Perhaps because of our ignorance of the specifics of this play, scholars who have worked on How to Write History generally seem to assume that the Andromeda is intended here simply as an icon for extreme Euripidean pathos and popularity. The exception is Peter von Möllendorff, who notes that among the extant fragments of the play, we happen to have the three lines immediately following the one line Lucian quotes (fr 136): σὺ δ᾽ ὦ θεῶν τύραννε κἀνθρώπων Ἔρως, ἢ μὴ δίδασκε τὰ καλὰ φαίνεσθαι καλά, ἢ τοῖς ἐρῶσιν εὐτυχῶς συνεκπόνει μοχθοῦσι μόχθοις ὧν σὺ δημιουργὸς εἶ. O thou tyrant of gods and men, Eros, Teach not the fair to seem fair, Or else propitiously suffer together with those in love, Who struggle with struggles of which thou art the source.

Möllendorff argues that the broader context of the one quoted line is relevant to a fuller understanding of Lucian’s game: When the Abderites rend the air in the streets with this particular verse, then they do so not simply because it is good, stirring stuff. Seen in its proper context, it is rather a sort of coded cry for help: as they are unable to talk any other ‘language’ than tragic verse, this is the one way the Abderites can beg to be cured of their ecstatic ailment (συνεκπόνει . . . ). Only the reader who knows how Euripides’ text continues will recognize this shift into the implicit mode, a move characteristic of Lucian’s method of allusion.6

Möllendorff ’s observation concerning this fragment is very suggestive and does, I believe, point in the right direction, but there is another connection we may draw between the available evidence on the Andromeda and the scenario in How to Write History. The Andromeda apparently opened with a scene in which the forlorn title character, chained in a desolate cave, engages in a dialogue with her only available interlocutor, the nymph Echo.7 After every line of Andromeda’s lament Echo would croon back her last word or two as a response. This technique of monologic dialogue could easily heighten the dramatic pathos beyond that attainable by a soliloquy: Echo is the only consort whose company signifies solitude. However, it is also possible to imagine this same element turned to other purposes; if Echo were not in sympathy with her interlocutor’s plight, her parroting would be intensely obnoxious. This comedic potential was not lost on 5 6

The scholiast on Frogs 53 is emphatic: τῶν καλλίστων Εὐριπίδου δρᾶμα ἡ Ἀνδρομέδα. 7 Möllendorff 2001, 128–9. The fullest treatment of the evidence is Bubel 1991.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

EPILOGUE

181

Aristophanes, whose Thesmophoriazusae, produced the year after the Andromeda, parodies Euripides’ ‘realism’ by introducing the tragedian himself in a hopeless ruse to free his kinsman from the renegade women.8 Aristophanes’ Euripides can think of no more effective means of salvation than to create a pastiche of rescue scenes from his own plays. He excerpts the Palamedes and the Helen before coming to the most ridiculous rehearsal of all, a re-enactment of useless re-enactment, so to speak. The kinsman sees Perseus in the distance and understands that he is now to play Andromeda, so he recalls the princess’ address to the vocal nymph of the cave, but when Echo takes the stage, the kinsman finds that he has summoned a slapstick gadfly: ‘Echo, cheeky repeater of words’ (Ἠχώ, λόγων ἀντῳδὸς ἐπικοκκάστρια, 1059). She explains how she served Euripides the previous year and how she will serve him again, and together they launch into the pseudodialogue from the beginning of the Andromeda. The irony is even richer if, as the scholia suggest, the actor who plays Echo is the actor who plays Euripides elsewhere in the play.9 Aristophanes’ parody indicates something of the popularity of Euripides’ Andromeda and the ease with which the conceit of its prominent Echo scene could be turned to comic effect. I would suggest that the opening of How to Write History is another form of play on the same dynamic. The Abderites’ obsessive repetition of lines from the Andromeda indicates that they are trapped in the play and dependent on its language, but within the play this behaviour is not unique: they are echoing Echo.10 Lucian’s story of the Abderites provides a convenient résumé for many of the issues that have concerned us in this book. The pattern of mimetic contagion that informs this anecdote is focused on a play, rather than a work of plastic art, except that, as we argued in Chapter 4, theatrical performance is itself a form of visual art and a prominent venue for painting. However, one of the things Lucian says is stuck in the spectators’ brains is Medusa’s head (1.21), presumably an artistic replica they would have seen as a property Perseus brought onstage with him in the Andromeda, and we recall from Chapter 2 (pp. 52–3) that the Gorgon’s head was one of the most widespread and consistent instances of mimetic contagion across Greco-Roman antiquity. Our inquiry in Chapter 5 also seems relevant here. The Andromeda is a tragedy, not a mime, but Aristophanes immediately realized the paratragic potential of the Echo scene, and the straightforward entertainment of watching one actor mindlessly ape another is strongly reminiscent of the stupidus role in mime. Similarly in Lucian’s anecdote, the Andromeda plays

9 N. Slater 2002, 175–7; Sommerstein 1994, 226–7. Rutherford 1896, 502. Returning to fr. 136, I note that there are a striking number of close verbal echoes (Ἔρως . . . ἐρῶσιν; καλὰ . . . καλά; μοχθοῦσι μόχθοις). The Abderites are choosing lines that will only intensify their repetitiveness. This kind of play with Echo as a figure of repetition, simultaneously intra- and intertextual, has a long tradition in Greek literature; see 2005. 8

10

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

182

MIMETIC CONTAGION

once in Abdera as a tragedy, but then it gets replayed ad nauseam as a farce. The foolish Abderites, like Chaerea, make a spectacle of themselves by clumsily imitating the roles in a work of art they do not fully understand. Finally, at the end of Chapter 2 (p. 66) we cast a glance at the scene in Aethiopica 4.8 where a painting of Perseus and Andromeda infects its viewer and causes her to replicate its image in the foetus she is conceiving. If it is not a coincidence that the Perseus myth provides the subject for both these stories, perhaps it is also significant that in the Eunuch it was a painting of Perseus’ conception that exceeded its frame to propagate itself in its viewers and reformat the play according to its own logic of art and life.11

For the widespread evidence of violent affect associated with the Andromeda myth, see Hilton 1998, 84–9. 11

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

BIBLIOGRAPHY ABEL, L. 1963. Metatheatre: A New View of Dramatic Form. New York. ADAM, J. 1963. The Republic of Plato. 2 vols. 2nd ed. Cambridge. ADAMS, J. N. 1982. The Latin Sexual Vocabulary. Baltimore. ADORNO, T. W. 1997. Gesammelte Schriften. Frankfurt. ANDERSON, G. 1976. Lucian: Theme and Variation in the Second Sophistic. Leiden. ANDERSON, W. S. 1984. ‘Love plots in Menander and his Roman adapters’, Ramus 13: 124–34. ANDERSON, W. S. 1993. Barbarian Play: Plautus’ Roman Comedy. Toronto. ASHMORE, S. G. 1910. P. Terenti Afri Comoediae. New York. ASTIN, A. E. 1978. Cato the Censor. Oxford. AUERBACH, E. 2003. Mimesis: The Representation of Reality in Western Literature. Princeton. AUSTIN, R. G. 1960. M. Tulli Ciceronis Pro M. Caelio Oratio. Oxford. AYGON, J.-P. 2004. ‘ “Imagination” et description chez les rhéteurs du Ier s. ap. J.-C.’, Latomus 63: 108–23. BARCHIESI, M. 1969. ‘Plauto e il “metateatro” antico’, Il Verri 31: 113–30. BARSBY, J. 1999a. Terence: Eunuchus. Cambridge. BARSBY, J. 1999b. ‘Love in Terence’, in ed. S. M. Braund and R. Mayer, Amor—Roma: Love and Latin Literature, pp. 5–29. Cambridge. BARTSCH, S. 1989. Decoding the Ancient Novel. Princeton. BARTSCH, S. 2006. The Mirror of the Self: Sexuality, Self-knowledge, and the Gaze in the Early Roman Empire. Chicago. BEACHAM, R. C. 1992. The Roman Theater and its Audience. Cambridge, MA. BEARE, W. 1937. ‘Recent work on the Roman theatre’, Classical Review 51, 105–11. BEARE, W. 1940. ‘Contamination in Plautus and Terence’, Revue de philologie, de littérature et d’histoire anciennes 66, 28–42. BEARE, W. 1948. ‘Terence, an original dramatist in Rome’, Hermathena 71: 64–82. BEARE, W. 1950. The Roman Stage. London. BEARE, W. 1959. ‘Contaminatio’, Classical Review 73, 7–11. BECATTI, G. 1951. Arte e gusto negli scrittori latini. Florence. BEEKES, R. 2010. Etymological Dictionary of Greek. 2 vols. Leiden. BELFIORE, E. 1984. ‘A theory of imitation in Plato’s Republic’, Transactions and proceedings of the American Philological Association 114: 121–46. BENEDIKTSON, D. T. 2000. Literature and the Visual Arts in Ancient Greece and Rome. Norman, OK. BENJAMIN, W. 1977. Gesammelte Schriften. Frankfurt. BENZ, L., E. STÄRK, and G. VOGT-SPIRA (eds). 1995. Plautus und die Tradition des Stegreifspiels. Tübingen. BERGER, R., P. SEARLES, and W. L. NEUMAN. 1995. ‘Rape-law reform: its nature, origins, and impact’, in ed. P. Searles and R. Berger, Rape and Society, pp. 223–32. Boulder.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

184

BIBLIOGRAPHY

BETTINI, M. 1999. The Portrait of the Lover. Berkeley. BLOCKLEY, R. C. 1983. The Fragmentary Classicizing Historians of the Later Roman Empire: Eunapius, Olympiodorus, Priscus and Malchus. Vol. 2. Liverpool. BOURQUE, L. B. 1989. Defining Rape. Durham. BOWEN, A. J. 1999. Xenophon: Symposium. London. BRENDEL, O. J. 1970. ‘The scope and temperament of erotic art in the Greco-Roman world’, in T. Bowie, O. J. Brendel, P. H. Gebhard, R. Rosenblum, and L. Steinberg, Studies in Erotic Art, pp. 3–107. New York. BROADIE, S. 2012. Nature and Divinity in Plato’s Timaeus. Cambridge. BROTHERS, A. J. (ed.) 2000. Terence: The Eunuch. Warminster. BROWN, P. G. MCC. 1990. ‘The Bodmer codex of Menander and the endings of Terence’s Eunuchus and other Roman comedies’, in ed. E. Handley and A. Hurst, Relire Ménandre, pp. 37–61. Geneva. BROWN, P. G. MCC. 2008. ‘The Eunuch castrated: Bowdlerization in the text of the Westminster Latin Play’, International Journal of the Classical Tradition 15: 16–28. BROWNMILLER, S. 1975. Against Our Will: Men, Women, and Rape. New York. BUBEL, F. 1991. Euripides: Andromeda. Stuttgart. BÜCHNER, K. 1974. Das Theater des Terenz. Heidelberg. BÜRGER, K. 1892. ‘Zu Xenophon von Ephesus’, Hermes 27: 36–67. BÜTTNER, S. 2000. Die Literaturtheorie bei Platon und ihre anthropologische Begründung. Tübingen. CAHILL, A. J. 2001. Rethinking Rape. Ithaca, NY. CAIRNS, F. 2006. Sextus Propertius the Augustan Elegist. Cambridge. CALBOLI MONTEFUSCO, L. 2005. ‘Ἐνάργεια et ἐνέργεια: l’evidence d’une démonstration qui signifie les choses en acte’, in ed. M. Armisen Marchetti, Demonstrare: voir et faire voire: forme de la démonstration à Rome, pp. 43–58. Toulouse. CAREY, S. 2003. Pliny’s Catalogue of Culture: Art and Empire in the Natural History. Oxford. CHALMERS, W. R. 1957. ‘Contaminatio’, Classical Review 71, 12–14. CHIARINI, G. 1979. La Recita. Plauto, la Farsa, la Festa. Bologna. CHIARINI, G. 1980. ‘Compresenza e conflittualità dei generi nel teatro latino arcaico (per una rillettura dell’Amphitruo)’, Materiali e discussioni per l’analisi dei testi classici 5: 87–124. CLARKE, J. R. 1998. Looking at Lovemaking: Constructions of Sexuality in Roman Art 100 BC–AD 250. Berkeley. COHEN, D. 1991. ‘Sexuality, violence, and the Athenian law of hubris’, Greece and Rome 38: 171–88. COLE, S. G. 1984. ‘Greek sanctions against sexual assault’, Classical Philology 79: 97–113. COLLINS, D. 2003. ‘Nature, cause, and agency in Greek magic’, Transactions and proceedings of the American Philological Association 133: 17–49. COMFORT, K. 2008. ‘The critic as artist and liar: the reuse and abuse of Plato and Aristotle by Wilde’, The Wildean 32: 57–70. CORNFORD, F. M. 1937. Plato’s Cosmology. London. CORSO, A. 2004. The Art of Praxiteles: The Development of Praxiteles’ Workshop and its Cultural Tradition until the Sculptor’s Acme (364–1 BC). Rome.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

BIBLIOGRAPHY

185

CROSS, R. C. and A. D. WOOZLEY. 1966. Plato’s Republic: A Philosophical Commentary. London. CROWTHER, N. B. 1983. ‘Greek games in republican Rome’, L’antiquité classique 52: 268–73. CUEVA, E. P. 2004. The Myths of Fiction: Studies in the Canonical Greek Novels. Ann Arbor. DANSON, L. 1997. Wilde’s Intentions: The Artist in his Criticism. Oxford. DANZIG, G. 2005. ‘Intra-Socratic polemics: the Symposia of Plato and Xenophon’, Greek, Roman, and Byzantine Studies 45: 331–57. DE LAURETIS, T. 1987. Technologies of Gender. Bloomington. DENHAM, R. D. 2010. Poets on Paintings: A Bibliography. Jefferson, NC. DESSEN, C. S. 1995. ‘The figure of the Eunuch in Terence’s Eunuchus’, Helios 22: 123–39. DEUFERT, M. 2002. Textgeschichte und Rezeption der plautinischen Komödien im Altertum. Berlin. DEXTER, C. E. 1975. ‘The Casa di L. Cecilio Giocondo in Pompeii.’ PhD dissertation, Duke University. DICKINSON, O. 1994. The Aegean Bronze Age. Cambridge. DOVER, K. J. 1978. Greek Homosexuality. Cambridge, MA. DUCKWORTH, G. E. 1952. The Nature of Roman Comedy: A Study in Popular Entertainment. Princeton. DUNSCH, B. 2005. ‘Sat habeo, si cras fero: Zur dramatischen Funktion der temporalen Deixis bei Plautus, Terenz und Menander’, Würzburger Jahrbücher für die Altertumswissenschaft N. F. 29: 123–50. DUPONT, F. 1985. L’Acteur-roi, ou, Le Théatre dans la Rome antique. Paris. DURKHEIM, E. 1912. Les formes élémentaires de la vie religieuse. Paris. DWORKIN, A. 1989. Pornography: Men Possessing Women. New York. DYSON, M. 1988. ‘Poetic imitation in Plato Republic 3’, Antichthon 22: 42–53. EASTERLING, P. E. 1995. ‘Menander: loss and survival’, in ed. A. Griffiths, Stage Direction: Essays in Honour of E. W. Handley, pp. 153–60. London. ELAM, K. 1996. ‘The fertile eunuch: Twelfth Night, early modern intercourse, and the fruits of castration’, Shakespeare Quarterly 47.1: 1–36. ELSNER, J. 1995. Art and the Roman Viewer: The Transformation of Art from the Pagan World to Christianity. Cambridge. ELSNER, J. 2007. ‘Viewing Ariadne: from ekphrasis to wall painting in the Roman world’, Classical Philology 102: 20–44. ESTRICH, S. 1987. Real Rape. Cambridge. FABIA, P. 1888. Les prologues de Térence. Paris. FABIA, P. 1895. Terenti Afri Eunuchus. Paris. FANTHAM, E. 1972. Comparative Studies in Republican Latin Imagery. Toronto. FANTHAM, E. 1986. ‘ΖΗΛΟΤΥΠΙΑ: a brief excursion into sex, violence, and literary history’, Phoenix 40: 45–57. FARAONE, C. A. 1992. Talismans and Trojan Horses: Guardian Statues in Ancient Greek Myth and Ritual. Oxford. FARAONE, C. A. 1993. ‘Molten wax, spilt wine and mutilated animals: sympathetic magic in Near Eastern and early Greek oath ceremonies’, Journal of Hellenic Studies 113: 60–80.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

186

BIBLIOGRAPHY

FELL, J. P. 1976. ‘Was Freud a follower of Kant?’ in U. Guzzoni, B. Rang, and L. Siep, Der Idealismus und seine Gegenwart: Festschrift für Werner Marx zum 65. Geburtstag, pp. 116–26. Hamburg. FLOWER, H. I. 1996. Ancestor Masks and Aristocratic Power in Roman Culture. Oxford. FORD, A. 2002. The Origins of Criticism: Literary Culture and Poetic Theory in Classical Greece. Princeton. FOREHAND, W. E. 1985. Terence. Boston. FORSÉN, B. 1996. Griechische Gliederweihungen: Eine Untersuchung zu ihrer Typologie und ihrer religions- und sozialgeschichtlichen Bedeutung. Helsinki. FORSYTHE, G. 1994. Review of Gruen 1992, Bryn Mawr Classical Review 94.02.11. FOUCAULT, M. 1988. Politics, Philosophy, Culture: Interviews and Other Writings, 1972–1977, ed. C. Gordon. New York. FOX, M. 2001. ‘Dionysius, Lucian, and the prejudice against rhetoric in history’, Journal of Roman Studies 91: 76–93. FRAENKEL, E. 1968. ‘Zur römischen Komödie’, Museum Helveticum 25: 231–42. FRAENKEL, E. 2007. Plautine Elements in Plautus. Oxford. FRANGOULIDIS, S. A. 1993. ‘Modes of metatheatre: theatricalization and detheatricalization in Terence, Eunuchus’, Liverpool Classical Monthly 18: 146–51. FRANGOULIDIS, S. A. 1994a. ‘Performance and improvisation in Terence’s Eunuchus’, Quaderni urbinati di cultura classica 48: 121–30. FRANGOULIDIS, S. A. 1994b. ‘The soldier as a storyteller in Terence’s Eunuchus’, Mnemosyne 47: 486–95. FRANGOULIDIS, S. A. 1997. Handlung und Nebenhandlung: Theater, Metatheater und Gattungsbewußtsein in der römischen Komödie. Stuttgart. FRAZER, J. G. 1935. The Golden Bough: A Study in Magic and Religion. New York. FREEDBERG, D. 1989. The Power of Images: Studies in the History and Theory of Response. Chicago. FRONTISI-DUCROUX, F. 1991. Le Dieu-masqué. Rome. FULKERSON, L. 2002. ‘(Un)sympathetic magic: a Study Of Heroides 13’, American Journal of Philology 123: 61–88. FULKERSON, L. 2005. The Ovidian Heroine as Author: Reading, Writing and Community in the Heroides. Cambridge. GARBARINO, G. 1973. Roma e la filosofia greca dalle origini alla fine del II secolo a.C. Turin. GEBAUER, G. and C. WULF. 1992. Mimesis. Hamburg. GELL, A. 1998. Art and Agency: An Anthropological Theory. Oxford. , R. 2005. ‘The figure of Echo in the Homeric Hymn to Pan’, American Journal of Philology 126: 187–208. , R. 2007. ‘Virgilian retrospection in Goethe’s Alexis und Dora’, Goethe Yearbook 15, 75–98. , R. 2013. ‘Andria’, in ed. A. Augoustakis and A. Traill, A Companion to Terence, pp. 225–42. Malden, MA. GIANCOTTI, F. 1967. Mimo e gnome: Studio su Decimo Laberio e Publilio Siro. Florence. GILL, C. 1977. ‘The genre of the Atlantis story’, Classical Philology 72: 287–304. GILL, C. 1979. ‘Plato’s Atlantis story and the birth of fiction’, Philosophy and Literature 3: 64–78.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

BIBLIOGRAPHY

187

GILULA, D. 1978. ‘Where did the audience go?’ Scripta Classica Israelica 4: 45–9. GILULA, D. 1989. ‘The first realistic roles in European theatre: Terence’s Prologues’, Quaderni urbinati di cultura classica 33: 95–106. GILULA, D. 1991. ‘Plots are not stories: the so-called duality method of Terence’, in ed. H. Scolnikov and P. Holland, Reading Plays: Interpretation and Reception, pp. 81–93. Cambridge. GOLD, B. 1998. ‘ “Vested interests” in Plautus’ Casina: cross-dressing in Roman comedy’, Helios 25: 17–29. GOLDBERG, S. 1986. Understanding Terence. Princeton. GOLDEN, L. 1992. Aristotle on Tragic and Comic Mimesis. Atlanta, GA. GOLDHILL, S. 1996. ‘Refracting classical vision: changing cultures of viewing’, in ed. T. Brennan and M. Jay, Vision in Context: Historical and Contemporary Perspectives on Sight, pp. 15–28. New York. GOLDHILL, S. 1998. ‘The seductions of the gaze: Socrates and his girlfriends’, in ed. P. Cartledge, P. Millett, and S. von Reden, KOSMOS: Essays in Order, Conflict, and Community in Classical Athens, pp. 105–24. Cambridge. GOLDHILL, S. 2001. ‘The erotic eye: visual stimulation and cultural conflict’, in ed. S. Goldhill, Being Greek under Rome: Cultural Identity, the Second Sophistic and the Development of Empire, pp. 154–94. Cambridge. GOMBRICH, E. H. 1960. Art and Illusion: A Study in the Psychology of Pictorial Representation. New York. GONDEK, H.-D. and P. WIDMER (eds). 1994. Ethik und Psychanalyse: vom kategorischen Imperativ zum Gesetz des Begehrens: Kant und Lacan. Frankfurt. GOWERS, E. 2004. ‘The plot thickens: hidden outlines in Terence’s Prologues’, Ramus 33: 150–66. GRATWICK, A. S. 1982. ‘Drama’, in ed. E. J. Kenney and W. V. Clausen, The Cambridge History of Classical Literature II: Latin Literature, pp. 77–137. Cambridge. GRATWICK, A. S. 1993. Plautus: Menaechmi. Cambridge. GRAUERT, W. H. 1833. ‘Über das Contaminiren der lateinischen Komiker’, in Historische und philologische Analekten, 116–207. Münster. GREEN, J. R. 1995. ‘Theatrical motifs in non-theatrical contexts on vases of the later fifth and fourth centuries’, in ed. A. Griffiths, Stage Directions: Essays in Honour of E. W. Handley (Bulletin of the Institute of Classical Studies Supplement 66), pp. 93–120. GREEN, R. 2002. ‘Towards a reconstruction of performance style’, in ed. P. Easterling and E. Hall, Greek and Roman Actors: Aspects of an Ancient Profession, pp. 93–126. Cambridge. GRIFFITH, M. 2006. ‘Horsepower and donkeywork: equids in the ancient Greek imagination’, Part One, Classical Philology 101: 185–246. GRUEN, E. 1990. Studies in Greek Culture and Roman Policy. Leiden. GRUEN, E. 1992. Culture and National Identity in Republican Rome. Ithaca, NY. GUZZO, P. G. and V. SCARANO USSANI. 2000. Veneris figurae: immagini di prostituzione e sfruttamento a Pompei. Naples. HABINEK, T. N. 1998. The Politics of Latin Literature: Writing, Identity, and Empire in Ancient Rome. Princeton.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

188

BIBLIOGRAPHY

HÄGG, T. 1966. ‘Die Ephesiaka des Xenophon Ephesios: original oder epitome?’ Classica et Mediaevalia 27: 118–61. HALLIWELL, S. 1988. Plato: Republic 10. Warminster. HALLIWELL, S. 1992. ‘Epilogue: The Poetics and its interpreters’, in ed. A. E. Rorty, Essays on Aristotle’s Poetics, pp. 409–24. Princeton. HALLIWELL, S. 1998. Aristotle’s Poetics. 2nd ed. Chicago. HALLIWELL, S. 2002. The Aesthetics of Mimesis: Ancient Texts and Modern Problems. Princeton. HAND, F. G. 1969. Tursellinus seu de particulis Latinis commentarii. Amsterdam. HARTMANN, F. 1919. ‘Contaminare’, Glotta 10, 266. HATHAWAY, B. 1962. The Age of Criticism: The Late Renaissance in Italy. Ithaca, NY. HEDREEN, G. 2007. ‘Involved spectatorship in Archaic Greek art’, Art History 30: 217–46. HENDERSON, JEFFREY. 1996. Three Plays by Aristophanes: Staging Women. New York. HENDERSON, JOHN. 1996. ‘Footnote: representation in the Villa of the Mysteries’, in ed. J. Elsner, Art and Text in Roman Culture, pp. 235–76. Cambridge. HERBIG, R. 1958. Neue Beobachtungen am Fries der Mysterien-Villa in Pompeji: ein Beitrag zur römischen Wandmalerei in Campanien. Baden-Baden. HILTON, J. 1998. ‘An Ethiopian paradox: Heliodorus, Aithiopika 4.8’, in ed. R. Hunter, Studies in Heliodorus, pp. 79–92. Cambridge. HINDLEY, C. 1999. ‘Xenophon on male love’, Classical Quarterly 49: 74–99. HINDS, S. 1987. ‘Generalizing about Ovid’, Ramus 16: 4–31. HOFMANN, J. B. 1951. Lateinische Umgangssprache. Heidelberg. HOLMES, B. 2010. The Symptom and the Subject: The Emergence of the Physical Body in Ancient Greece. Princeton. HUMM, M. 2005. Appius Claudius Caecus: La républic accomplie. Rome. HUNTER, R. 1985. The New Comedy of Greece and Rome. Cambridge. HUNTER, R. 1987. ‘Middle Comedy and the Amphitruo of Plautus’, Dioniso 57: 281–98. HUNTER, R. 2002. ‘ “Acting down”: the ideology of Hellenistic performance’, in ed. P. Easterling and E. Hall, Greek and Roman Actors: Aspects of an Ancient Profession, pp. 189–206. Cambridge. HUNTER, R. 2009. Critical Moments in Classical Literature: Studies in the Ancient View of Literature and Its Uses. Cambridge. HUSS, B. 1999a. ‘The dancing Sokrates and the laughing Xenophon, or the other Symposium’, American Journal of Philology 120: 381–409. HUSS, B. 1999b. Xenophons Symposion: ein Kommentar. Stuttgart. IRIGARAY, L. 1985. Speculum of the Other Woman, trans. G. C. Gill. Ithaca, NY. JACHMANN, G. 1931. Plautinisches und Attisches. Problemata 3. Berlin. JACOBELLI, L. 1995. Le pitture erotiche delle Terme Suburbane di Pompei. Rome. JOHANSEN, T. K. 2004. Plato’s Natural Philosophy: A Study of the Timaeus-Critias. Cambridge. JOHNS, C. 1982. Sex or Symbol: Erotic Images of Greece and Rome. Austin. KAPPELER, S. 1986. The Pornography of Representation. Minneapolis. KIRBY, J. T. 1991. ‘Mimesis and diegesis: foundations of aesthetic theory in Plato and Aristotle’, Helios 18: 113–28.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

BIBLIOGRAPHY

189

KNAPP, C. 1917. ‘References to painting in Plautus and Terence’, Classical Philology 12: 143–57. KNAPP, C. 1919. ‘References in Plautus and Terence to plays, players, and playwrights’, Classical Philology 14: 35–55. KNORR, O. 2007. ‘Metatheatrical humor in the comedies of Terence’, in ed. P. Kruschwitz, W.-W. Ehlers, and F. Felgentreu, Terentius Poeta: Zetemata Monographien zur klassischen Altertumswissenschaft 127, pp. 167–74. Munich. KONSTAN, D. 1986. ‘Love in Terence’s Eunuch: the origins of erotic subjectivity’, American Journal of Philology 107: 369–93. KRAEMER, C. J. 1927–8. ‘In defense of Chaerea in the Eunuch of Terence’, Classical Journal 23: 662–7. KRAUT, R. 1998. Aristotle Politics: Books VII and VIII. Oxford. KRUSCHWITZ, P. 2001. ‘Verszahlresponsionen bei Terenz’, Philologus 145: 312–23. LADA-RICHARDS, I. 2004. ‘Authorial voice and theatrical self-definition in Terence and beyond: the Hecyra Prologues in ancient and modern contexts’, Greece and Rome 51: 55–82. LAFREE, G. 1989. Rape and Criminal Justice: The Social Construction of Sexual Assault. Belmont. LAPE, S. 2001. ‘Democratic ideology and the poetics of rape in Menandrian comedy’, Classical Antiquity 20: 79–119. LEACH, E. W. 1969. ‘De exemplo meo ipse aedificato: an organizing idea in the Mostellaria’, Hermes 97: 318–32. LEFÈVRE, E. 1995. Plautus und Philemon. Tübingen. LEFÈVRE, E., E. STÄRK, and G. VOGT-SPIRA. 1991. Plautus barbarus: Sechs Kapitel zur Originalität des Plautus. Tübingen. LEO, F. 1912. Plautinische Forschungen. 2nd ed. Berlin. LESZL, W. G. 2006. ‘Plato’s attitude to poetry and the fine arts, and the origins of aesthetics’, Part III, Études platoniciennes III, 245–336. LETTA, C. 1984. ‘L’Italia dei mores Romani nelle Origines di Catone’, Athenaeum 62: 3–30 and 416–39. LEVEN, K.-H. 1993. ‘Miasma und Metadosis: antike Vorstellungen von Ansteckung’, Medizin, Gesellschaft und Geschichte 11: 44–73. LÉVI-STRAUSS, C. 1962. La Pensée sauvage. Paris. LEVIN, R. 1967. ‘The double plots of Terence’, Classical Journal 62: 301–5. LIENHARDT, G. 1954. ‘Modes of thought’, in ed. E. E. Evans-Pritchard, The Institutions of Primitive Society, pp. 95–107. Oxford. LING, R. 1991. Roman Painting. Cambridge. LONG, A. A. and D. N. SEDLEY. 1987. The Hellenistic Philosophers. 2 vols. Cambridge. LUCK, G. 1967. P. Ovidius Naso Tristia. Heidelberg. LUDWIG, W. 1959. ‘Von Terenz zu Menander’, Philologus 103, 1–38. MACK, R. 2002. ‘Facing down Medusa (an aetiology of the gaze)’, Art History 25: 571–604. MACKINNON, C. A. 1989. Toward a Feminist Theory of the State. Cambridge, MA. MACMULLEN, R. 1991. ‘Hellenizing the Romans (2nd century B.C.)’, Historia 40: 419–38.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

190

BIBLIOGRAPHY

MANUWALD, G. 2011. Roman Republican Theatre. Cambridge. MARCADÉ, J. 1965. Roma Amor: Essay on Erotic Elements in Etruscan and Roman Art. Geneva. MARSHALL, C. W. 2006. The Stagecraft and Performance of Roman Comedy. Cambridge. MARTI, H. 1963. ‘Die Deutungen des Wortes contaminare (An 16, Ht 17)’, Lustrum 8, 23–7. MCDONNELL, M. 2006. ‘Roman aesthetics and the spoils of Syracuse’, in ed. S. Dillon and K. E. Welch, Representations of War in Ancient Rome, pp. 68–90. Cambridge. MCGINN. T. A. J. 2002. ‘Pompeian brothels and social history’, in T. McGinn, P. Carafa, N. de Grummond, B. Bergmann, and T. Najbjerg, Pompeian Brothels, Pompeii’s Ancient History, Mirrors and Mysteries, Art and Nature at Oplontis, and the Herculaneum ‘Basilica’, JRA Supplementary Series No. 47: 7–46. MCKIM-SMITH, G. 2002. ‘The rhetoric of rape, the language of vandalism’, Woman’s Art Journal 23.1: 29–36. MEULI, K. 1955. ‘Altrömisher Maskenbrauch’, Museum Helveticum 12: 206–34. MEYERHÖFER, E. 1927. Der Aufbau des Terenzischen Eunuchus. Diss. Erlangen. MITCHELL, W. J. T. 1995. Picture Theory: Essays on Verbal and Visual Representation. Chicago. MÖLLENDORFF, P. V. 2001. ‘Frigid enthusiasts: Lucian on writing history’, Proceedings of the Cambridge Philological Society 47: 117–40. MOORE, T. J. 1998. The Theater of Plautus: Playing to the Audience. Austin. MOORS, K. 1987. ‘The argument against a dramatic date for Plato’s Republic’, Polis 7.1: 6–31. MORALES, H. 1996. ‘The torturer’s apprentice: Parrhasius and the limits of art’, in ed. J. Elsner, Art and Text in Roman Culture, pp. 182–209. Cambridge. MORALES, H. 2004. Vision and Narrative in Achilles Tatius’ Leucippe and Clitophon. Cambridge. MORGAN, K. A. 1998. ‘Designer history: Plato’s Atlantis story and fourth-century ideology’, Journal of Hellenic Studies 118: 101–18. MOSELEY, N. and M. HAMMOND. 1962. T. Macci Plauti Menaechmi. Cambridge, MA. MOSS, J. 2007. ‘What is imitative poetry and why is it bad?’ in G. R. F. Ferrari, ed. The Cambridge Companion to Plato’s Republic, pp. 415–44. Cambridge. MUECKE, F. 1986. ‘Plautus and the theater of disguise’, Classical Antiquity 5: 216–29. MYEROWITZ, M. 1992. ‘The domestication of desire: Ovid’s Parva Tabella and the theater of love’, in A. Richlin, ed. Pornography and Representation in Greece and Rome, pp. 131–57. Oxford. NADDAFF, R. A. 2002. Exiling the Poets: The Production of Censorship in Plato’s Republic. Chicago. NEER, R. 1995. ‘The lion’s eye: imitation and uncertainty in Attic red-figure’, Representations 51: 118–53. NEHAMAS, A. 1982. ‘Plato on imitation and poetry in Republic 10’, in ed. J. Moravcsik and P. Temko, Plato on Beauty, Wisdom, and the Arts, pp. 47–78. Totowa, NJ. NEHAMAS, A. 1988. ‘Plato and the mass media’, The Monist 71: 214–34. NICOLL, A. 1963. Masks, Mimes and Miracles: Studies in the Popular Theatre. New York.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

BIBLIOGRAPHY

191

NIGHTINGALE, A. W. 2000. Genres in Dialogue: Plato and the Construct of Philosophy. Cambridge. NORWOOD, G. 1923. The Art of Terence. Oxford. NUTTON, V. 2000. ‘Did the Greeks have a word for it? Contagion and contagion theory in classical antiquity’, in ed. L. I. Conrad and D. Wujastyk, Contagion: Perspectives from Pre-Modern Societies, pp. 137–62. Aldershot. O’NEILL, K. 1998. ‘Symbolism and sympathetic magic in Propertius 4.5’, Classical Journal 94.1: 49–80. O’SULLIVAN, J. N. 1995. Xenophon of Ephesus: His Compositional Technique and the Birth of the Novel. Berlin. OLSON, K. 2003. ‘Roman underwear revisited’, Classical World 96: 201–10. OMITOWOJU, R. 2002. Rape and the Politics of Consent in Classical Athens. Cambridge. ONG, W. J. 1988. Orality and Literacy: The Technologizing of the Word. London. OSBORNE, R. 2007. ‘Sex, agency, and history: the case of Athenian painted pottery’, in ed. R. Osborne and J. Tanner, Art’s Agency and Art History, pp. 179–98. Malden, MA. OWEN, S. G. 1924. P. Ovidi Nasonis Tristium liber secundus. Oxford. PANAYOTAKIS, C. 1995. Theatrum Arbitri: Theatrical Elements in the Satyrica of Petronius (Mnemosyne supp. 146). Leiden. PANAYOTAKIS, C. 2010. Decimus Laberius: The Fragments. Cambridge. PAPADOPOULOU, I. 2006. ‘Poètes et (philo)sophoi: pour une archéologie de la mimesis’, Revue de philosophie ancienne 24: 3–16. PAPE, M. 1975. Griechische Kunstwerke aus Kriegsbeute und ihre öffentliche Aufstellung in Rom. Hamburg. PARKER, D. 1974. ‘The Eunuch (Eunouchus)’, in P. Bovie, Terence: The Comedies, pp. 147–225. Baltimore. PARKER, H. 1992. ‘Love’s body anatomized: the ancient erotic handbooks and the rhetoric of sexuality’, in ed. A. Richlin, Pornography and Representation in Greece and Rome, pp. 90–111. Oxford. PARKER, H. 1996. ‘Plautus vs. Terence: audience popularity re-examined’, American Journal of Philology 117: 585–617. PASQUALI, G. 1936. ‘Studi terenziani III: Un personaggio e due scene dell’ Eunuco’, Studi italiani di filologia classica 13: 117–29. PERRY, E. 2005. The Aesthetics of Emulation in the Visual Arts of Ancient Rome. Cambridge. PETRONE, G. 1977. Morale e antimorale nelle commedie di Plauto: Ricerche sullo Stichus. Palermo. PETRONE, G. 1983. Teatro antico e inganno: Finzioni plautine. Palermo. PHILIPPIDES, K. 1995. ‘Terence’s Eunuchus: elements of the marriage ritual in the rape scene’, Mnemosyne 48: 272–84. PICKARD-CAMBRIDGE, A. W. 1962. Dithyramb, Tragedy and Comedy. Oxford. PIERCE, K. F. 1997. ‘The portrayal of rape in New Comedy’, in ed. S. Deacy and K. F. Pierce, Rape in Antiquity, pp. 163–84. London. PISANI, V. 1935. ‘Notulae Graeco-Latinae, 3: contaminare’, Indogermanische Forschungen 53, 25–8.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

192

BIBLIOGRAPHY

PLATT, V. 2011. Facing the Gods: Epiphany and Representation in Graeco-Roman Art, Literature and Religion. Cambridge. POLLARD, J. 1977. Birds in Greek Life and Myth. London. POTOLSKY, M. 2006. Mimesis. New York. PRADEAU, J.-F. 1997. Le monde de la politique. Sur le récit atlante de Platon. Timée (17a– 27b) et Critias. St Augustine. RAND, E. K. 1932. ‘The art of Terence’s Eunuchus’, Transactions and proceedings of the American Philological Association 63: 54–72. RAWSON, E. 1985. ‘Theatrical life in republican Rome and Italy’, Proceedings of the British School at Rome 53: 97–113. RAWSON, E. 1993. ‘The vulgarity of the Roman mime’, in ed. H. D. Jocelyn, Tria Lustra: Essays and Notes Presented to John Pinsent, pp. 255–60. Liverpool. REEVE, M. D. 1989. ‘Conceptions’, Proceedings of the Cambridge Philological Society 35: 81–112. REICH, H. 1903. Der Mimus: Ein literatur-entwickelungsgeschichtlicher Versuch. Berlin. REYDAMS-SCHILS, G. (ed.). 2003. Plato’s Timaeus as Cultural Icon. Notre Dame. RICHTER, M. 1914. Priscorum poetarum et scriptorium de se et aliis iudicia. Leipzig. RICŒUR, P. 1984. Time and Narrative. 3 vols. Chicago. ROLLER, M. B. 2004. ‘Exemplarity in Roman culture: the cases of Horatius Cocles and Cloelia’, Classical Philology 99: 1–56. ROSENMEYER, T. G. 2002. ‘ “Metatheater”: an essay on overload’, Arion 10: 87–119. ROSIVACH, V. J. 1998. When a Young Man Falls in Love: The Sexual Exploitation of Women in New Comedy. London. RUHNKENIUS, D. 1825. In Terentii Comoedias Dictata. Bonn. RUIZ-MONTERO, C. 2003. ‘The Rise of the Greek Novel’, in ed. G. L. Schmeling, The Novel in the Ancient World, pp. 29–85. Leiden. RUSSELL, D. S. 1995. Emblematic Structures in Renaissance French Culture. Toronto. RUSSELL, J. 1997. ‘Healing a disfigured Rembrandt’s wounds’, The New York Times, 31 August 1997. RUTHERFORD, W. 1896. Scholia Aristophanica: being such comments adscript to the text of Aristophanes as have been preserved in the Codex Ravennas. London. SANDBACH, F. H. 1977. The Comic Theatre of Greece and Rome. London. SANDBACH, F. H. 1982. ‘How Terence’s Hecyra failed’, Classical Quarterly 32: 134–5. SAUERWEIN, I. 1970. Die leges sumptuariae als römische Massnahme gegen den Sittenverfall. Hamburg. SAUNDERS, A. and P. Davidson, ed. 2005. Visual Words and Verbal Pictures: Essays in Honour of Michael Bath. Glasgow. SAYLOR, C. F. 1975. ‘The theme of planlessness in Terence’s Eunuchus’, Transactions and proceedings of the American Philological Association 105: 297–311. SCAFURO, A. 1997. The Forensic Stage. Cambridge. SCHEFOLD, K. 1957. Die Wände Pompejis: topographisches Verzeichnis der Bildmotive. Berlin. SCHENKL, K. 1876. Xenophontische Studien 3: Beiträge zur Kritik des Oikonomikos, des Symposion und der Apologie. Wien. SCHLEE, F. 1893. Scholia Terentiana. Leipzig.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

BIBLIOGRAPHY

193

SCHMELING, G. L. 1980. Xenophon of Ephesus. Boston. SCHULHOFER, S. 1992. ‘Taking sexual autonomy seriously: rape law and beyond’, Law and Philosophy 11: 35–94. SCHULZE, W. 1934. Kleine Schriften. Göttingen. SCHWERING, W. 1916. ‘Die sogenannte Kontamination in der lateinischen Komödie’, Neue Jahrbücher für das klassische Altertum 19, 167–85. SHACKLETON BAILEY, D. R. 1982. ‘Notes on Riese’s Anthologia Latina (Vol. 2)’, Classical Philology 77: 113–32. SHARROCK, A. 1996. ‘Representing metamorphosis’, in ed. J. Elsner, Art and Text in Roman Culture, pp. 103–30. Cambridge. SHARROCK, A. 2002. ‘Gender and sexuality’, in ed. P. Hardie, The Cambridge Companion to Ovid, pp. 95–107. Cambridge. SHARROCK, A. 2009. Reading Roman Comedy: Poetics and Playfulness in Plautus and Terence. Cambridge. SHEA, C. 1998. ‘Setting the stage for romances: Xenophon of Ephesus and the Ehrasis’, in ed. R. F. Hock, J. B. Chance, and J. Perkins, Ancient Fiction and Early Christian Narrative, pp. 61–76. Atlanta. SLATER, N. 1985. Plautus in Performance: The Theater of the Mind. Princeton. SLATER, N. 2002. Spectator Politics: Metatheater and Performance in Aristophanes. Philadelphia. SLATER, W. J. 2002. ‘Mime problems: Cicero ad Fam. 7.1 and Martial 9.38’, Phoenix 56: 315–29. SMITH, L. P. 1994. ‘Audience response to rape: Chaerea in Terence’s Eunuchus’, Helios 21: 21–38. SOMMERSTEIN, A. 1994. Thesmophoriazusae. Warminster. SORABJI, R. 1978. ‘Body and soul in Aristotle’, in ed. J. Barnes, M. Schofield, and R. Sorabji, Articles on Aristotle: 4. Psychology and Aesthetics. New York. SÖRBOM, G. 1966. Mimesis and Art: Studies in the Origin and Early Development of an Aesthetic Vocabulary. Uppsala. SPENTZOU, E. 2003. Readers and Writers in Ovid’s Heroides: Transgressions of Genre and Gender. Oxford. SPITZER, L. 1955. ‘The “Ode on a Grecian Urn,” or Content vs. Metagrammar’, Comparative Literature 7: 203–25. STEINER, D. T. 2001. Images in Mind: Statues in Archaic and Classical Greek Literature and Thought. Princeton. STEINER, G. 1969. ‘The graphic analogue from myth in Greek romance’, in Classical Studies Presented to Ben Edwin Perry by his Students and Colleagues at the University of Illinois, 1924–1960, pp. 123–37. Urbana. STEWART, P. 2007. ‘Gell’s idols and Roman cult’, in ed. R. Osborne and J. Tanner, Art’s Agency and Art History, pp. 158–78. Malden, MA. STEWART, Z. 1958. ‘The Amphitruo of Plautus and Euripides’ Bacchae’, Transactions and proceedings of the American Philological Association 89: 348–78. SWEETSER, E. 1990. From Etymology to Pragmatics: Metaphorical and Cultural Aspects of Semantic Structure. Cambridge.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

194

BIBLIOGRAPHY

SZAKOLCZAI, A. 2013. Comedy and the Public Sphere: The Rebirth of Theatre as Comedy and the Genealogy of the Modern Public Arena. New York. TANNER, J. 2006. The Invention of Art History in Ancient Greece: Religion, Society and Artistic Rationalisation. Cambridge. TAPLIN, O. 1993. Comic Angels and Other Approaches to Greek Drama through VasePaintings. Oxford. TAUSSIG, M. 1993. Mimesis and Alterity: A Particular History of the Senses. New York. THESLEFF, H. 1978. ‘The interrelation and date of the Symposia of Plato and Xenophon’, Bulletin of the Institute of Classical Studies 25: 157–70. THIERFELDER, A. 1935. Review of Jachmann 1931. Gnomon 11, 144–5. TORELLI, M. 1990. ‘Comments on Wallace’, in ed. W. Eder, Staat und Staatlichkeit in der frühen römischen Republik, pp. 303–5. Stuttgart. TOWNEND, G. B. 1983. ‘The unstated climax of C. 64’, Greece and Rome 30: 21–30. TRAILL, A. E. 2001. ‘Menander’s Thais and the Roman poets’, Phoenix 55.3–4: 284–303. TREDENNICK, H. 1952. Review of Beare 1950. Classical Review 66, 27–30. TREGGIARI, S. 1991. Roman Marriage: Iusti Coniuges from the Time of Cicero to the Time of Ulpian. Oxford. TROMARAS, L. M. 1985. ‘Ibi inerat pictura haec (Terent. Eunuch. 584–589)’, Hellenika 36: 268–77. TROMARAS, L. M. 1994. P. Terentius Afer: Eunuchus. Hildesheim. TSCHIEDEL, H. J. 1986. ‘Hic Abdera: Gedanken zur Narrheit eines Gemeinwesens im Altertum – oder: Wie dumm waren die Abderiten?’ in ed. P. Krafft and H. J. Tschiedel, Concentus hexachordus: Beiträge zum 10. Symposion der bayerischen Hochschullehrer für Klassische Philologie in Eichstätt (24.-25. Februar 1984), pp. 169–95. TUERK, J. 1999. ‘An early Byzantine inscribed amulet and its narratives’, Byzantine and Modern Greek Studies 23: 25–42. TURFA, J. M. 2006. ‘Votive offerings in Etruscan religion’, in ed. N. T. de Grummond and E. Simon, The Religion of the Etruscans, pp. 90–115. Austin. VASALY, A. 1993. Representations: Images of the World in Ciceronian Oratory. Berkeley. VERDENIUS, W. J. 1949. Mimesis: Plato’s Doctrine of Artistic Imitation and Its Meaning to Us. Leiden. VERNANT, J.-P. 1991. Mortals and Immortals: Collected Essays. Princeton. VEYNE, P. 1978. ‘La famille et l’amour sous le haut empire romain’, Annales: économies, sociétés, civilisations 33: 35–63. VIDAL-NAQUET, P. 1986. ‘Athens and Atlantis: structure and meaning of a Platonic myth’, in The Black Hunter: Forms of Thought and Forms of Society in the Greek World, pp. 263–84. Baltimore, MD. VOGT-SPIRA, G. 1995. ‘Traditionen improvisierten Theaters bei Plautus’, in ed. B. Zimmerman, Griechisch-römische Komödie und Tragödie, pp. 70–93. Stuttgart. WALBANK, F. W. 1957. A Historical Commentary on Polybius. Oxford. WALBANK, F. W. 1972. Polybius. Berkeley. WALLOCHNY, B. 1992. Streitszenen in der griechischen und römischen Komödie (ScriptOralia, 44). Tübingen. WALTZ, R. 1938. ‘Contaminare chez Térence’, Revue des études latines 16: 269–74.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

BIBLIOGRAPHY

195

WEBB, R. 2008. Demons and Dancers: Performance in Late Antiquity. Cambridge, MA. WEBB, R. 2009. Ekphrasis, Imagination and Persuasion in Ancient Rhetorical Theory and Practice. Burlington, VT. WEBSTER, T. B. L. 1953. Studies in Later Greek Comedy. Manchester. WHITMARSH, T. 2002. ‘Written on the body: ekphrasis, perception and deception in Heliodorus’ Aethiopica’, Ramus 31: 111–25. WILDE, O. 1968. Literary Criticism of Oscar Wilde, ed. S. Weintraub. Lincoln, NE. WILDE, O. 2000. The Complete Letters of Oscar Wilde, ed. M. Holland and R. Hart-Davis. New York. WILES, D. 1991. The Masks of Menander: Sign and Meaning in Greek and Roman Performance. Cambridge. WILLIAMS, G. D. 1994. Banished Voices: Readings in Ovid’s Exile Poetry. Cambridge. WISEMAN, T. P. 1988. ‘Satyrs in Rome?’ Journal of Roman Studies 78: 1–13. WOHL, V. 2004. ‘Dirty dancing: Xenophon’s Symposium’, in ed. P. Murray and P. Wilson, Music and the Muses: The Culture of ‘Mousike¯’ in the Classical Athenian City, pp. 337–63. Oxford. WRIGHT, J. 1974. Dancing in Chains: The Stylistic Unity of the Comoedia Palliata. Rome. WÜST, E. 1932. ‘Mimos’, in Byzantine and Modern Greek Studies, col. 1727–64. Stuttgart. ZAGAGI, N. 1994. The Comedy of Menander: Convention, Variation and Originality. London. ZEYL, D. J. 2000. Plato: Timaeus. Indianapolis, IN.

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

INDEX Accius 165 Adelphoe 10, 31, 43, 162 Adorno 50 Andria 10, 130, 158, 161–3, 168–9 Anonymous Seguerianus 18–19 Antipho 9, 150–2, 168–9 Aristotle 7, 18, 75–7, 85–6 Augustine 6, 46–8 Benjamin 50 Caecilius 127–8 castration 1, 8, 13–14, 111–16 Celsus 19 Chaerea 1–4, 8, 28–48, 109, 142–3 Chérubin (from Beaumarchais’ Le Mariage de Figaro) 30–1 Cloelia 96 Cicero 9, 19, 125, 132–3, 147, 164–6 contaminatio 10, 157–77 (im) Danaë 5, 154–5 defixiones, see similia similibus in cultic images de’ Nobili, sco “Cherea” 30 Diomedes 131–2, 142 Donatus 6, 29, 34, 45–6, 150 enargeia 18–25 ekphrasis 17–26, 66–7 emblem 6, 49–50, 107, 116 Euripides 178–81 exemplarity 95–101 Frazer 49–50 funeral ritual 95–6 Gellius 127, 159 Gnatho 2–3, 10, 169–75 Gorgon, see Medusa Heautontimorumenos 127, 131, 158, 169 Hecyra 9, 128–31 Heliodorus 66 Herodas 147

Horace 18, 107, 141 Horatius Cocles 96–7 Irigaray, Luce 93–4 Jerome 164 Jupiter 4–5, 10, 45, 153–5 Livy 96, 98–100, 122–4, 165 Lucian 178–82 Marcellus’ ovatio 8, 98–101 Medusa 52–3 Menander 10, 41, 59, 72, 118, 127, 145, 160–4, 168–9, 173, 177 metadosis/miasma 6 metatheatricality 4, 8, 11–17, 108–10 mime 8, 55–7, 78, 109, 120–56 (im) mirror 6, 52–3, 62–4 Numa 97 oscilla 53 Ovid 6, 54–8 Pamphila 1–4, 28, 30, 110, 149 par pari respondere 3, 141–5 Parmeno 1–3, 12–13, 33–6, 108, 142 Parrhasius 54, 73–4 Phaedria 1–3, 8, 13 phantasia 20–5 Plato 7, 77–94 Ion 179 Republic 79–85 Timaeus-Critias 86–91 Plautus 8, 10, 11, 101–6, 120–1, 160–1 Amphitruo 4, 145 Captivi 104 Curculio 113 Epidicus 102 Menaechmi 105–6 Mercator 102–3 Miles Gloriosus 114–15 Mostellaria 103–4

OUP CORRECTED PROOF – FINAL, 12/9/2016, SPi

198

INDEX

Plautus (Continued) Persa 102 Poenulus 102, 113 Pseudolus 102 Stichus 102 Trinummus 102 Pliny 74 Plutarch 97–8 Polybius 95–7, 100 Pompeian paintings 6 House of Gavius Rufus (VII.2.16) 154–5 House of Regina Margherita (V.2.1) 154–5 House of the Centenary (IX.8.6) 61–2 potior 38–9 Progymnasmata 17–18, 25 Propertius 6, 58–60 ps.-Longinus 20–1 ps.-Servius 116–17 Pythagoras of Rhegium 74 Pythias 3, 8, 13–15, 108, 112 quinque lineae amoris (five lines of love) 5, 34–5 Quintilian 19–22 rape 1, 4–6, 28–46 modern formulations 39–41 as plot motif in New Comedy 41–4 Rembrandt’s Danaë 26–7

Santia vase 9, 143–5 Seneca the Elder 73–4 sexually explicit painting as erotodidaxis 54–64 Shakespeare 30 similia similibus in cultic images 51–3 Solon 90 sympathetic magic, see similia similibus in cultic images Thais 1–4, 8, 110–11, 117–19, 175–6 Thraso 1–3, 8, 169–75 Tiberius 54, 60, 61, 73–4 Unity of Time 12–15 Varro 97 verisimilitude 7, 12, 24, 30, 42, 72–5 Vitalis epitaph (Latin Anthology, Riese 487a) 9, 133–4 voodoo dolls, see similia similibus in cultic images Wilde, Oscar 91–2 Xenophon of Athens 7, 9, 75, 134–40 Xenophon of Ephesus 6, 65–71