Language and character in Euripides' Electra 9780191835445, 0191835447, 9780192512208, 019251220X, 9780192512215, 0192512218

Citation preview

OXFORD CLASSICAL MONOGRAPHS Published under the supervision of a Committee of the Faculty of Classics in the University of Oxford

The aim of the Oxford Classical Monographs series (which replaces the Oxford Classical and Philosophical Monographs) is to publish books based on the best theses on Greek and Latin literature, ancient history, and ancient philosophy examined by the Faculty Board of Classics.

Language and Character in Euripides’ Electra EV ERT VAN EMDE BOA S

1

3

Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © Evert van Emde Boas 2017 The moral rights of the author have been asserted First Edition published in 2017 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2016942733 ISBN 978–0–19–879360–1 Printed in Great Britain by Clays Ltd, St Ives plc Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

To my parents

Preface I think of this book as the product of a marriage between linguistics and literary interpretation. To some, this may feel like a shotgun wedding; but if I can convince others that this can be a fruitful union between partners who give each other mutual reinforcement, then I will consider it a success. The book is based on my Oxford DPhil dissertation, submitted in 2010. The present text is considerably revised, even if the overall structure and a good deal of the argument are fundamentally unchanged. The revisions took much longer than hoped and foreseen: uninterrupted teaching obligations and several other projects left precious little time for the necessary work. If this unwelcome delay has one positive consequence, it is that I have been able to take into account several works on Greek linguistics, Greek tragedy, Euripides, and Electra specifically that appeared in the intervening years. Once the decision was made that I could not in good faith ignore these works, it became clear that my book would in fact be much improved by incorporating insights from three (!) new commentaries on the play,1 as well as from various other important works of scholarship.2 Had the book appeared sooner, some parts of it and points in it would probably have been outmoded almost at once (it is my hope that this will be less the case for the revised text). Many thanks are in order. My dual approach was reflected in a pair of dissertation supervisors (one for the literary aspect, one for the 1 Roisman and Luschnig 2011, Distilo 2012, Cropp 2013. I have reviewed two of these in van Emde Boas 2016. In fact almost nothing will be said in this book about Roisman and Luschnig 2011—a commentary geared towards a different audience (and see the justified reviews by Cropp 2011 and Kovacs 2012). The appearance of the second edition of Cropp’s commentary (2013, revised from the 1988 edition) gave rise to a particular quandary: the most straightforward solution—changing all my references to conform to the new edition, on the principle that only the more recent edition can be said to reflect Cropp’s thinking—turned out not always to be practicable (see my review for various nuanced changes of phrasing between the editions), and I have therefore retained some references to the older text. 2 Many works could be listed here (and several will be found in the bibliography), but I am thinking particularly of Bakker 2010a, Mastronarde 2010, Rutherford 2012, Roisman 2013, Giannakis 2014.

viii

Preface

linguistic), and twice as many in this case really meant twice as good. In Oxford, Bill Allan has taught me more about Euripides and tragedy than he might realize, and if I say anything of interest about the interpretation of Electra in this book, it is due in part to his constant call for ‘pay-off ’. He oversaw the project with patience (which I must at times have tried) and level-headedness (which I always appreciated). In Amsterdam I have known Albert Rijksbaron as a teacher, a supervisor twice over, and now as a collaborator (on a different book) and as a friend. In each of these guises, his vast knowledge of Greek has been an inspiration, his keenly discriminating mind a trusty guide. My examiners, Felix Budelmann and Judith Mossman, deserve gratitude not only for their comments and the most enjoyable twoand-a-half hours of talking about Greek tragedy that I’ve ever had, but also for recommending the book to the Oxford Classical Monograph committee. Both of them are now valued colleagues and collaborators on different projects, including my present work at the Calleva Centre (Magdalen College, Oxford), where I am in the fortunate position of working with Felix Budelmann on a daily basis. Richard Rutherford acted as OUP’s assigned adviser during the conversion from dissertation to book, and provided helpful notes during that process. The OUP reader was extremely generous with insightful comments, and suggested the new title (a great improvement over the dissertation’s bland ‘Linguistic Studies in Euripides’ Electra’). Earlier versions of two chapters were read by Philomen Probert and Angus Bowie, and by Scott Scullion and Adrian Kelly; talking through these chapters with them sharpened my thinking and prevented many errors. The questions and comments of audiences at various seminars and conferences where I presented material have also led to numerous improvements: pride of place belongs to that august institution, Oxford’s graduate work-in-progress seminars (long may WiP thrive!). In general, the graduate community in Oxford is not mentioned often enough in these prefaces; its members made my DPhil experience a richer and better one. Of my great teachers at school and university, I would like to mention Jan Krimp and Ton Jansen (Haarlem), Fred Naiden (Tulane, New Orleans), and Irene de Jong (Amsterdam). Teachers turn into colleagues, and for their support and friendship (while this book was lurking in the background), many thanks are due to my fellow classicists (too many to list by name) at the University of Amsterdam,

Preface

ix

the University of Groningen, Leiden University, VU University Amsterdam, and the University of Oxford. My time as a graduate in Oxford (first for the MSt, then for the DPhil) would not have been possible without the generous support of the VSB-Fonds, the ‘Talentenbeurs’ of the Netherlands Ministry of Education, the Prins Bernhard Cultuurfonds, the Arts and Humanities Research Council, Corpus Christi College, and the Charles Oldham Fund. In acquiring aid from the last of these sources, the assistance of my College Adviser at Corpus Christi College, Stephen Harrison, and of then College President, Sir Tim Lankester, was instrumental. At the Press, Georgina Leighton, Lisa Eaton, and Manuela Tecusan swiftly and expertly saw the book through to publication. I am grateful for their warm support. For sending me their work to read (and view), I thank Mary-Kay Gamel, Ann Suter, Margaret Kitzinger, and Patrick Finglass. Finally, for various reasons, I thank Glenn Lacki, Laura Bok, Tori McKee, Liz Lucas, Rob Cioffi, Juliane Kerkhecker, Luuk Huitink, and Mathieu de Bakker. Anouk Petersen was there with support in all the important ways at all the important moments. Optimis parentibus, who made it all possible, this work is gratefully dedicated.

Table of Contents List of Figures A Note on Citations, Abbreviations, and Cross-Referencing

xv xvi

Introduction: Modern Linguistics and Euripides’ Electra 1. Aims, approaches, outline 1.1. Reading, linguistically 1.2. Outline of the book 2. Linguistic approaches 2.1. Introduction: Bauformen and text types 2.2. Conversation analysis 2.3. Pragmatics

1 1 1 5 7 7 9 14

2.3.1. Speech acts 2.3.2. (Neo-)Gricean theories of meaning

17 20

2.4. Sociolinguistics 2.4.1. Gender 2.4.2. Politeness and power

2.5. Gnomic utterances in context: some aspects of modern paroemiology 2.6. Narrative and argumentative texts: discourse cohesion 3. Textual criticism 4. A view of the play 4.1. Characters and characterization 4.1.1. Conceptualization; characterization through style 4.1.2. Electra and Orestes

4.2. Themes and motifs 4.3. Tradition (and the recognition scene) 4.4. The roads not taken . . . I. Rustic Language: The Peasant 1. Introduction 2. A peasant’s tale (1–53) 3. Husband and wife (54–81, 341–63, 404–31) 3.1. A marriage under face threat 3.2. Getting water (54–81) 3.3. Welcoming guests (341–63) 3.4. Preparing food (404–31) 4. Further stylistic points; conclusion

26 27 31

40 47 50 51 52 52 54

56 57 58 60 60 62 70 70 71 74 76 78

xii

Table of Contents

II. Constancy amid Change: The Linguistic Characterization of Electra 1. Introduction 2. Resistance through lament: the early scenes (54–81, 112–214) 2.1. Electra as mourner 2.2. Electra as wife 2.3. Electra as Argive ‘maiden’ 2.4. The characterization of Electra 2.4.1. Patterns of miscommunication 2.4.2. Electra’s character: the debate

3. Electra and her unexpected guest (215–338) 3.1. The stichomythia (215–89) 3.2. The ‘message’ (300–38) 4. Recognition and planning (487–698) 5. Electra, Aegisthus, and Clytemnestra 5.1. A play of halves? 5.2. The ‘kakology’ (907–56)

5.2.1. Electra’s ‘undramatic’ generalizations 5.2.2. Analysis of the speech

5.3. ‘Into the boudoir’: Electra and Clytemnestra (998–1146) 5.3.1. 5.3.2. 5.3.3. 5.3.4. 5.3.5.

Opening exchanges (998–1010) Clytemnestra’s speech (1011–50) Parrhēsia (1055–9) Electra’s speech (1060–99) Mother and daughter (1102–46)

6. Exodos 6.1. The kommos (1177–232) 6.2. The deus ex machina (1233–358) III. Orestes’ Linguistic (Dis)guises 1. Introduction 2. Orestes incognito 2.1. Initial observations 2.2. The general reflections

2.2.1. ‘A man in exile is powerless’ (236) 2.2.2. Pity and intelligence (290–6) 2.2.3. Evaluating character (367–400)

2.3. The second disguise (774–858) 3. Conclusion

80 80 81 81 83 87 90 90 92

98 98 108 122 127 127 128 128 135

141 141 143 145 147 150

153 154 159 165 165 166 166 168 170 171 177

185 186

Table of Contents IV. Redrawing the Lines: Pragmatics and Gender in Textual Criticism 1. Introduction 1.1. Vexed passages 1.2. L and P 2. Divided (?) we pray (671–84) 3. The hesitation scene (959–87) 3.1. On giving orders and having fashion sense (959–66) 3.2. Diverging minds (967–87) 4. Conclusion

xiii 188 188 188 190 193 204 204 215 228 229 229 230 232 232

V. A Tense Affair: The Messenger Speech 1. Introduction 2. Not ‘what?’ but ‘how?’ 3. Analysis of the narrative 3.1. Setting the scene (774–8) 3.2. ‘A deliciously protracted game of cat and mouse’ (779–97) 3.3. A moment for prayer (798–810) 3.4. The sacrifice of Aegisthus (810–43) 3.5. Aftermath and resolution (844–55, 855–7, 857–8) 3.6. Evaluation 4. Conclusion

234 238 239 242 243 246

VI. The Language of Rhetoric: The agōn Revisited 1. Introduction 2. Exordium 2.1. Clytemnestra 2.2. Electra 3. Narratio 3.1. Clytemnestra 3.2. Electra 4. Argumentatio 4.1. Generalizations 4.2. Clytemnestra’s hypotheticals 4.3. Rhetoric and characterization 5. Peroratio 5.1. Clytemnestra 5.2. Electra 6. Peroratio (II)

248 248 250 250 251 253 253 255 258 259 260 262 265 265 266 267

xiv

Table of Contents

Conclusion: Approaching Tragic Language

269

Glossary of Linguistic Terms Bibliography Index Locorum Subject Index Index of Greek Words

273 280 303 308 318

List of Figures 0.1. The discourse structure of drama. 0.2. Preferred and dispreferred second parts.

5 11

0.3. Speech act types.

19

0.4. Politeness strategies.

33

4.1. Proposed line-to-speaker attributions in Electra 671–84.

195

4.2. Proposed line-to-speaker attributions in Electra 959–66.

206

A Note on Citations, Abbreviations, and Cross-referencing References to the works of ancient authors follow the conventions adopted in the Oxford Classical Dictionary (Hornblower et al. 2012), except that speeches by the orators are always referred to by number and Euripides’ Heracles is abbreviated Her. instead of HF. Fragments from tragedy are cited according to the relevant volumes of Tragicorum Graecorum Fragmenta. In the bibliography, journal titles follow the abbreviations in L’Année Philologique. Sigla for Euripides’ plays are usually not preceded by ‘Eur.’ when no other authors are discussed, unless this specification adds clarity. Commentaries and editions are normally cited by author name (and title of the play) only; thus ‘Basta Donzelli app. crit.’ refers to the apparatus criticus in her Teubner edition of the Electra, and ‘Fraenkel ad loc.’ or ‘Fraenkel on Aesch. Ag. 1’ refers to the relevant note in his Agamemnon. ‘Cropp’, unless otherwise indicated, refers to the second edition of his commentary (2013). All these works may be found in the bibliography. Occasionally I also mention the proposer of a textual emendation by name only: for full details in such cases, the reader is referred to Basta Donzelli’s edition. Otherwise, all works apart from standard reference works are cited using the author–year format. In keeping with OUP’s preferred house style (and with an eye to the book’s digital edition), I have inserted cross-references to other points in the book only by section number or footnote number (with the chapter added only if the reference is to a different one), rather than by page number. It is my hope that this will, on the whole, make the reader’s job in tracing my argument easier rather than more difficult. Translations of Greek and foreign languages throughout the text are my own. For the use of the asterisk (*), see the Introduction, §1.2.

Introduction Modern Linguistics and Euripides’ Electra

1. AIMS, APPROACHES, OUTLINE

1.1. Reading, linguistically This book is first and foremost a literary reading of Euripides’ Electra, a play that remains the subject of great interest and great controversy. In the course of my reading, I will often draw the same kind of conclusions (about the text and its interpretation, characters, themes, and so forth) as have been drawn by scholars who worked on the play previously. But the way in which I aim to reach those conclusions is different from what is usual in work on Greek tragedy: throughout, the basis of my analysis will be the play’s language. This may seem a vacuous statement concerning a play of which we have nothing left but its language,1 yet, in a discipline where ‘research into Greek 1 This statement requires some modification, of course—but not much. Almost all of the ‘evidence’ for what we ‘know’ about Electra is circumstantial: (1) its relation to works using the same mythical material, including Aeschylus’ Choephori and Sophocles’ Electra (whether the latter comes before or after it; see §4.4); (2) its dating in relation to that of other Euripidean works (which is insecure: see Basta Donzelli 1978: 27–71; long-held beliefs about allusions to contemporary events were disproven by Zuntz 1955: 63–71); (3) its place in the Euripidean oeuvre and within tragedy more generally (with regard to the development of this form, its literary techniques, etc.); and (4) acting and performance aspects of the first performance (see n. 8). There are very few—and largely inconsequential—scholia (see Keene 1893b), five papyrus fragments (gathered in Basta Donzelli 1995a: xxxvii–xxxviii), and Euripides’ version of the story does not appear to be represented in vase painting (the play is accordingly absent from Taplin 2007). The curious reference to this play in Plutarch’s Lysander (15.3) seems to suggest that Electra was known well enough in antiquity, disproving the view that it ‘has never been a terribly popular play’ (Whitehorne 1978: 5). For the play’s reception, see Bakogianni 2011 and Luschnig 2015.

2

Introduction: Modern Linguistics and Euripides’ Electra

tragedy’ has sometimes come to be synonymous with ‘research into the (sociopolitical, religious, and historical) context of Greek tragedy’, such a singular focus is worth making explicit upfront.2 More importantly, in approaching the play’s language, I plan to travel along some different avenues from those normally taken by classicists in linguistic enquiry: my close reading will be heavily informed by modern linguistic methodology, and it is a secondary aim of the work to show that this methodological apparatus, developed in the last half century or so in general linguistics, can teach us much about tragic language and how we should interpret it. In saying that I will follow ‘different avenues’ I do not wish to discount the rich body of work emerging in recent decades that applies modern linguistic theory to ancient languages.3 Yet not all of the linguistic approaches I will adopt have been utilized, or utilized as fully as they can be, even in this current of research; and no one (to my knowledge) has attempted to apply all of them, combined, to a single text in order to see what they can tell us about the interpretation of that text. A methodological problem immediately rears its head: can we hope to apply techniques that were developed in the study of modern languages (usually in their everyday, spoken form) fruitfully to ancient Greek (or to any ‘dead’ language), let alone to the highly stylized Greek of tragedy?4 The answer is a qualified ‘yes’: we can, and for various reasons. First of all, linguistic theory is often concerned with identifying ‘universal’ features of language usage. It must be 2 I do not mean to suggest that such a context is unimportant, only that it will not be central to my approach. There seems, in fact, to have been something of a pendulum swing towards work with a focus on language since the dissertation on which this book is based was submitted: one may point here, e.g. to Rutherford 2012; and note the review by Wright (2013). 3 Nor is it a coincidence that much of that body of work derives from what is sometimes called ‘Dutch scholarship’ (even if this is somewhat unfair to the numerous scholars of other nationalities who work in the field). Book titles such as Grammar as Interpretation (Bakker 1997a) and The Language of Literature (Allan and Buijs 2007) may be taken as emblematic of this tradition and explain my affinity with it. Outside classics, my approach coincides with work in the (modern) field of stylistics. The outstanding introduction to this field is Leech and Short 2007; good introductory matter may also be found in Toolan 1998. Within this tradition, I found the work of Toolan (1990) and that of Culpeper (e.g. 2001, 2002, 2009) on characterization especially instructive. 4 ‘Stylization’ itself is not a straightforward concept, of course. For some discussion, see Silk 1996 and Rutherford 2010, 2012.

Introduction: Modern Linguistics and Euripides’ Electra

3

admitted that, as soon as a linguist claims universality for a feature he/she has described, two others will usually come out of the woodwork to contest that claim.5 Yet many of the approaches I will describe have proved remarkably resilient in the face of such criticism, and have been fruitfully applied to a very considerable range of languages, even in literary contexts. Second, it has been argued that, in spite of some clear but superficial differences between literary and everyday language, all literature at a more fundamental level still uses the same rules and conventions as any other form of linguistic expression. Herman has argued, for example, that, in the case of drama, it is a question of mechanics, in the exploitation by dramatists of underlying speech conventions, principles and ‘rules’ of use, operative in speech exchanges in the many sorts, conditions and contexts of society which members are assumed to share and use in their interactions in day‐to‐day exchanges. The principles, norms and conventions of use which underlie spontaneous communication in everyday life are precisely those which are exploited and manipulated by dramatists in their constructions of speech types and forms in play. Thus ‘ordinary speech’ or, more accurately, the ‘rules’ underlying the orderly and meaningful exchange of speech in everyday contexts are the resource that dramatists use to construct dialogue in plays. (Herman 1995: 6)

In other words, for dramatic dialogue to be comprehensible to an audience, it still must use the same linguistic resources that are familiar to its members from their own daily conversations: naturally occurring conversation is, as it were, the template on which dramatic discourse is grafted. We find similar arguments used about ancient Greek literature, for example by Schuren (2014: ch. 1) on tragic stichomythia and by Bakker (1997b: 17), who argued that Homeric speech is a stylization of everyday discourse ‘departing from it and yet retaining, or even highlighting, its most characteristic forms’.6 The phrase ‘or even highlighting’ in Bakker’s formulation might be emphasized: literary language may be seen not so much as a deviation, but rather as a concentration of natural language. 5 For cultural differences in how language is used by speakers, see e.g. the essays in Blum-Kulka et al. 1989. 6 This is also a fundamental concern to Minchin 2007. Throughout that book, Minchin is concerned with the applicability of modern linguistic concepts to Homeric Greek.

4

Introduction: Modern Linguistics and Euripides’ Electra

An analogy may be drawn here with the debate concerning characterization in Greek tragedy (more about this in §4.1.1). In an influential article, Gould (1978: 55) argued that the stylized nature of Greek tragic language was an obstacle to any straightforward reading of (realistic) character in that language and pointed out for instance that, if ‘we describe [stichomythia] to ourselves as “conversation” or “dialogue” in its everyday sense, we are likely to be misled as to its role in Greek drama’. In her response to Gould, Easterling (1990: 89) points out that what we get in Greek tragedy is in fact not quite like real-life character but a more ‘concentrated’ phenomenon: ‘what is important is that drama is much more intensely concentrated and meaningfully shaped, “purer” than ordinary unscripted experience, so that everything the stage figures do and say . . . has to be taken as significant’. This view of characterization in tragedy seems to me justified, and I would argue that what holds for character holds— mutatis mutandis, of course—for language as well: dialogue in tragedy, for instance, may not be quite like conversation ‘in its everyday sense’, but it is nevertheless significantly related to it, a ‘purer’ form of it perhaps. To be able to understand that purer form, tools that help us make sense of everyday conversation are vital. In short, a better understanding of how language works will help us to understand how ancient Greek literary language works. All this is not to say that we should not constantly be sensitive to differences between literary and everyday language,7 between Greek and other languages, and between Greek society (as a determinant of language use) and societies in our present day (this is especially true for the sociolinguistic approaches discussed in §2.4). One further caveat involves the different levels at which literary language, especially that of drama, ‘means’. As audience members or readers of a play, we overhear communication taking place between characters, but at the same time the playwright is communicating with us. Mick Short has represented this ‘embedded discourse structure’ as shown in Figure 0.1.

7 This sensitivity requires that we are not surprised in poetry by some linguistic phenomena that would surprise us in everyday language—for example the fact that characters in Greek tragedy consistently speak and sing in metre.

Introduction: Modern Linguistics and Euripides’ Electra

5

Figure 0.1. The discourse structure of drama. Short 1981: 149.

In applying linguistic approaches to the language exchanged between characters in a play—addresser 2 and addressee 28—we should be continually aware that such language will be influenced by the aims of addresser 1 (in our case, Euripides).9 In the end, in the matter of the value of linguistic methodology for literary interpretation, the proof is in the eating (and hopefully the pudding will taste good): if the approaches I use can be shown to lead to plausible readings,10 a case might be made for some methodological leniency, and we may give those approaches the benefit of the doubt.

1.2. Outline of the book Most of what remains of this introduction is in essence a ‘crash course’ in selected subdisciplines of modern linguistics that will inform my analysis of Electra throughout the book. Since many or all of these approaches may be unfamiliar to classicists, I will devote a good deal of space to them. For the reader’s ease, key terms are explained in a separate glossary at the end of the book that makes 8 The model ought really to be supplemented with an additional level: that of the actors representing ‘addresser 2’ and ‘addressee 2’ (cf. the similar diagrams in Mahler 2013: 153). Much has been written, of course, about acting and performance in antiquity (e.g. Easterling and Hall 2002, Csapo 2010), and it must be noted that direction and acting choices will have influenced heavily the way Electra’s first audience perceived the play. Yet any speculation about such aspects must remain just that: speculation. Accordingly, actors and acting will not play a significant role in this book. 9 I do not wish to imply that such aims are retrievable (though I have few qualms about approaching the Electra as a work written by an author who intended it to have meaning in some ways but not in others), only that we should be constantly aware that there is more going on in a play than communication between characters. 10 As a first step towards showing that this is the case, I will as often as possible use examples from Electra in my description of linguistic approaches in this chapter.

6

Introduction: Modern Linguistics and Euripides’ Electra

references to the relevant discussions in this section; and these terms are accompanied by an asterisk (*) at their first occurrence here and then throughout the later chapters. This typographical practice is meant to identify terms whose meaning may not be evident and terms that are used in an unfamiliar, technical manner. Armed with the glossary, the reader may thus work through my actual reading of the play in the subsequent chapters without (constant) recourse to the discussions in this introductory chapter. Still, it will be necessary to provide a relatively full account of the approaches first, since they underpin my analyses even when I do not explicitly say so. I will also briefly outline (in §3) my principles in practising textual criticism (roughly speaking, I do this only where it matters). Finally, I end the introduction with a very brief view of the play in which I give a preview of the major issues of interpretation in Electra that will concern me later in the book. In the first three chapters that make up the body of this book, I will apply the theories described here to the language of three important characters in the play, in each case in order to show that their characterization is driven in part by their linguistic behaviour. Chapter I deals with Electra’s husband, the Peasant, and serves both as a concise exemplification of my approach and as a reading of the first few scenes of the play. Chapter II is concerned with the play’s central character, Electra herself: the main thrust of this chapter (by necessity the longest of the book) will be, again, to show how Electra’s linguistic behaviour helps to establish her overall characterization, although many other issues are treated along the way. Chapter III deals with Orestes in a similar, if somewhat more limited fashion (I focus on his language in the first half of the play): my reading argues for a new way of looking at Orestes’ moralizing passages in the early scenes and against the excessive distrust that these passages (especially 367–400) have so frequently aroused. Chapter IV takes a somewhat different tack. In it I zoom in on two scenes of (presumed) stichomythia, 671–93 and 959–87, both notorious for problems concerning the division of lines and their attribution to speakers. My aim is still to show how an analysis of Orestes’ and Electra’s language in these scenes helps us to make sense of their characters; yet I also hope to demonstrate that some of the linguistic approaches I have introduced have a bearing on textual–critical problems, specifically on issues of line-to-speaker attribution (that is, the attribution of a line to one out of several possible speakers).

Introduction: Modern Linguistics and Euripides’ Electra

7

By focusing on stichomythia, chapter IV also becomes the first in a series of chapters concerned with particular ‘forms’. Chapters V and VI, which complete the book, continue on this path by looking at two other forms: the messenger speech (chapter V) and the agōn (chapter VI). I hope that my analyses in these chapters can contribute to our understanding of the workings of narrative in Euripides’ tragedies as well as to our understanding of his use of rhetorical techniques, although similar issues of interpretation (especially characterization) also remain central to my purpose.

2. LINGUISTIC APPROACHES

2.1. Introduction: Bauformen and text types Any Greek tragedy—and one by Euripides perhaps most of all— consists of various, fairly well circumscribed forms (rhēsis, stichomythia, choral ode, amoibaion, monody, etc.), and each of these forms should be accorded a somewhat different approach. The methodological tools that will help us understand the internal workings of a long messenger speech are not the same as those that elucidate the finer nuances of a fast-flowing stichomythic exchange. This is one of the reasons why advances in modern linguistics may be so helpful to an analysis of Greek tragedy: even though many admirable works of classical scholarship have explored each of the tragic subgenres in great detail,11 the idea that language ‘works’ differently in these different forms is rarely made concrete, nor are those differences explored fully. My own approach will be guided by divisions along somewhat different lines from the traditional Bauformen mentioned above, namely those of what is sometimes called ‘text types’: narrative, argument, description, dialogue, and so forth. The categories covered

11 See the collected discussions in Jens 1971. On individual forms and some more particular applications, see Schwinge 1968, Collard 1980, and Schuren 2014 on stichomythia; Schadewaldt 1926 and Battezzato 1995 on monologues; Erbse 1984 and van Wolferen 2003 on prologues; de Jong 1991, Barrett 2002, and Dickin 2009 on messenger speeches; Duchemin 1968 and Lloyd 1992 on the agōn; Popp 1968 on the amoibaion. I have left choral odes out of consideration (see §4.4).

8

Introduction: Modern Linguistics and Euripides’ Electra

by these headers are not necessarily discrete, but rather frequently recurring networks of form, function and content.12 It will be clear from the terms themselves—‘narrative’, ‘argument’, and the like— that such categories do not neatly correlate to the different building blocks of tragedy, even though some overlap may be expected (messenger speeches are mostly narrative, agōn speeches are largely argumentative, and so forth). Modern linguistic theory has become more and more attuned to differences in the workings of language in such different forms and has developed different analytical techniques for each. Fundamental to all of this is the realization that the sentence, traditionally the unit of analysis for linguistic study (something that is demonstrably still the case in all grammars of ancient Greek), is not a very helpful ‘level’ to zoom in on, or at least not unless sentences (or, better, utterances) are considered part of a larger discourse, which coheres in significant ways and is situated in a particular communicative context. The text types mentioned above display telling differences in how they cohere internally (which affects how certain structural features such as particles, pronouns, and tenses are used) and in how they interact with their extralinguistic context. Accordingly, in my outline of linguistic approaches, my initial and main focus will be on the workings of conversation, understood not as a collection of self-contained sentences but as an intricate complex of interrelating utterances. From this discussion of the structure of conversation I move on to other aspects of linguistics that demand that language be examined in its communicative context: that is to say, I will maintain a focus on the fact that tragedy features language used by particular speakers for particular addressees in particular situations. Finally, I will briefly mention some aspects that will prove important particularly to my analysis of larger stretches of narrative text and argumentative text—and here I will pay special attention to the use of particles and tenses.

12 For an introduction to text types (or ‘modes of discourse’), see Smith 2003. Plato’s distinction between διήγησις and μίμησις could be seen as a rudimentary precursor to text types (though whether he meant exactly the same as we do when we distinguish between ‘narrative’ and ‘non‐narrative’ is not a straightforward question).

Introduction: Modern Linguistics and Euripides’ Electra

9

2.2. Conversation analysis In the last half century or so, studies of naturally occurring conversation have taught us much about the fundamental workings of spoken communication between two or more interactants. Many of those studies were conducted in the name of conversation analysis (CA)*.13 CA sets out from the observation that conversation is characterized by turn-taking*: one participant speaks and stops, another starts and stops again, and so forth. With two participants, this gives an A–B–A–B–A–B . . . pattern of distribution, which is complicated in the case of three or more speakers. Speakers will normally hold the floor*, that is, they will maintain the ‘right’ to speak, until they come to a point where another speaker is meant to take over. At such moments, the present speaker either selects* the next speaker to take a turn (for instance by directing a question at a particular addressee) or may allow anyone to self-select*. If no one self-selects, the first speaker may opt to continue speaking and take another turn him- or herself. Self-selection, of course, also takes place when a speaker first initiates conversation. CA deals extensively with many particulars of turn-taking and of the selection process that are less relevant for Greek tragic discourse: these include overlaps (remarkably rare in natural conversation, but presumably much rarer still in fifth-century BC theatre), pauses, gaps, and so forth (silences, however, are of great interest in tragedy as well, though less so in the case of Electra).14 More relevant to my purposes, however, are the notions of adjacency pairs* and preference organization*. An adjacency pair is an ‘automated’ pattern in the structure of conversation that consists of a first part* produced by one speaker and a second part* produced by a different speaker. The utterance of a

13

I give a fuller introduction to the theory in van Emde Boas in press-a. The foundational works are Sacks et al. 1974 and Sacks 1992, now supplemented by Schegloff 2007. Good introductions to the field may be found in Levinson 1983: ch. 6, Mey 2001: ch. 6, Sidnell 2010, and (very fully) in Sidnell and Stivers 2012. CA has been applied to drama, e.g. by Herman (1991, 1995, 1998a, 1998b), Bennison (1993), Culpeper (2001: 172–80), and Mandala (2007). It has very rarely been used in classical studies; the only exceptions (as far as I know) are Minchin 2007, Drummen 2009 (an article that owes much to Basset 1997), Smith 2012, and Schuren 2014: ch. 1. CA will also feature prominently in Bonifazi et al. in press. Mastronarde 1979 is not influenced by CA, yet is concerned with similar issues. 14 On silences in tragedy, see Taplin 1972, Montiglio 2000: ch. 2, Chong-Gossard 2008: chs 3–4.

10

Introduction: Modern Linguistics and Euripides’ Electra

first part immediately creates the expectation of a second part designed to complete the adjacency pair. Here is a simple example of an adjacency pair from Electra (1116–17): ΗΛ. ΚΛ.

τί δ᾽ αὖ πόσιν σὸν ἄγριον εἰς ἡμᾶς ἔχεις; τρόποι τοιοῦτοι.

El. Cl.

And why do you let your husband treat us so savagely? Such are his ways.

QUESTION ANSWER

Questions invite answers. By uttering a question, a speaker (Electra) not only selects another speaker (Clytemnestra; at this point the first speaker normally must give up the floor), but also raises the expectation that that new speaker will use his/her speaking turn to answer the question. Apart from question–answer, prototypical adjacency pairs include greeting–greeting, offer–acceptance, and apology– minimization. More complex patterns have also been identified, such as the adjacency triplet complaint–apology–forgiveness.15 More than one possible type of second part can complete a first part. For example, answers are not the only possible responses elicited by questions: a second speaker may well refuse to provide an answer or may protest his/her ignorance. What CA has shown convincingly, however, is that not all second parts are equal. There is a ranking, a preference organization*, operating over the alternative responses to first parts, such that there is at least one preferred* and one dispreferred* category of response. These notions do not necessarily equate to psychological ones (in that they do not refer to speakers’ and hearers’ personal preferences), but are rather structural features suggesting that one type of response (the dispreferred one) is more marked and consequential for the flow of the conversation. Dispreferred responses will usually be performed with more linguistic complexity and will be accompanied by hesitations, hedges, explanations, and so forth:16 Speaker A: Speaker B:

Can you help me carry this box? Well, ehm, I’m not supposed to do heavy lifting.

REQUEST REFUSAL

Some general patterns of preferred and dispreferred responses are laid out in Figure 0.2. It is not certain that these patterns are uniformly applicable to tragic dialogue (the preferred response in one 15 16

See Edmondson 1981 and Culpeper 2001: 152–3. For such linguistic elaborations in the light of politeness theory, see §2.4.2.

Introduction: Modern Linguistics and Euripides’ Electra

11

Figure 0.2. Preferred and dispreferred second parts. Following Levinson 1983: 336.

culture might be the dispreferred one in another);17 yet, for the first four categories at least, I would contend that the patterns they describe exist in Greek tragic discourse.18 The expectation raised by a first part is powerful: once an adjacency pair is initiated, a second part is normally expected, even if it takes a while to arrive; otherwise at least an account for the absence of a second part is required. While such disruptions of adjacency pairs are less common and typically significant, delays in the completion of an adjacency pair are fairly common, and thus strict ‘adjacency’ is not a condition for the fulfilment of a pair. Rather, a first part gives rise to ‘conditional relevance’: if some other move occurs instead of the required second part, it will be interpreted, where possible, as an act preliminary to the performing of the second part; the relevance of the overarching adjacency pair normally remains in place until its first part is attended to. An example of delayed fulfilment of an adjacency pair in Electra is found a few lines down from the example cited earlier in this section (1124–32):

17 There is also great variation in preference patterns depending on individual contexts. Preference is, in fact, a highly complex topic in CA: good discussions are Schegloff 2007: ch. 4 and Pomerantz and Heritage 2012. 18 The fact that, as indicated in this figure, denials are the ‘preferred’ (= unmarked) response to expressions of blame (for modern English, this is borne out by empirical research: see Atkinson and Drew 1979: 80; other languages may behave differently) again makes it clear that preference organization cannot be equated with the speakers’ desires, personal preferences, and so on.

12

Introduction: Modern Linguistics and Euripides’ Electra

ΗΛ. ἤκουσας, οἶμαι, τῶν ἐμῶν λοχευμάτων· τούτων ὕπερ μοι θῦσον (οὐ γὰρ οἶδ᾽ ἐγώ) δεκάτῃ σελήνῃ παιδὸς ὡς νομίζεται· τρίβων γὰρ οὐκ εἴμ᾽, ἄτοκος οὖσ᾽ ἐν τῷ πάρος. ΚΛ. ἄλλης τόδ᾽ ἔργον, ἥ σ᾽ ἔλυσεν ἐκ τόκων. ΗΛ. αὐτὴ ᾽λόχευον κἄτεκον μόνη βρέφος. ΚΛ. οὕτως ἀγείτων οἶκος ἵδρυται φίλων; ΗΛ. πένητας οὐδεὶς βούλεται κτᾶσθαι φίλους. ΚΛ. ἀλλ᾽ εἶμι, παιδὸς ἀριθμὸν ὡς τελεσφόρον θύσω θεοῖσι. . . . 20 El.

Cl. El. Cl. El. Cl.

g

REQUEST

QUESTION ANSWER



QUESTION ANSWER



 19 

ACCEPTANCE

You have heard, I think, that I have given birth? Please make the customary tenth-night sacrifice for me: I do not know how to myself, as I am inexperienced and was childless until now. This is the duty of someone else, the woman who delivered your child! I was my own midwife, and bore the child on my own. Is your house so bereft of friendly neighbours? No one wants to acquire friends who are poor. Well then, I will go and sacrifice to the gods for the child’s completed term.

Clytemnestra does not immediately accede to Electra’s request, but rather uses two insert expansions* to sort out preliminary issues. The first part of the original adjacency pair (the request) retains its conditional relevance over the stretch of intervening question–answer sequences, which, because they intervene before the second part of the original pair, will be taken as somehow meaningful for the fulfilment of that pair. In other words, the structure steers interpretation of Clytemnestra’s questions as preliminaries she wants dealt with before making a decision about Electra’s request. That she is not entirely satisfied with Electra’s answers is clear from her use of the particle ἀλλ(ά) in 1132, which suggests that Clytemnestra breaks off the series of questions (implying that a sufficient answer has not been provided) and instead reverts to the original issue of the request.21 In addition to insert expansions, other expansions* are possible as well: adjacency pairs may be elaborated by pre-expansions* (which prepare the ground for the production of the adjacency pair) and 19 Readers may object that Clytemnestra’s first insertion is not a question: this is true in the grammatical sense, but I nonetheless think that her utterance is meant to elicit information from Electra and thus functions, at least in part, as a question (it also functions as a kind of objection). For such ‘indirect speech acts’, see §2.3.1. 20 For the text (many editors transpose 1107–8 after 1131), see ch. II, §5.3.5. 21 See van Emde Boas in press-a for my treatment of this ‘assentient’ use of the particle (and cf. the treatment in Denniston 1954: 16–20).

Introduction: Modern Linguistics and Euripides’ Electra

13

post-expansions* (which somehow modify, challenge, or expand on the closed adjacency pair). In fact the pair marked question 2–answer 2 in the example above (1129–30) might well be seen as a postexpansion to the first question–answer pair (1127–8): Clytemnestra does not simply move on to the next question but asks Electra to elaborate on her first answer. As for pre-expansions, the classic example is the summons–answer sequence, which serves no function other than to enable subsequent conversational turns. In a significantly elaborated form, Electra offers a neat pair of examples in the deus ex machina scene (1292–8): ΟΡ.

ὦ παῖδε Διός, θέμις ἐς φθογγὰς τὰς ὑμετέρας ἡμῖν πελάθειν; ΚΑ. θέμις, οὐ μυσαροῖς τοῖσδε σφαγίοις. ΗΛ. κἀμοὶ μύθου μέτα, Τυνδαρίδαι; ΚΑ. καὶ σοί· Φοίβῳ τήνδ’ ἀναθήσω πρᾶξιν φονίαν. ΟΡ. πῶς . . . Or. Ca. El. Ca. Or.

SUMMONS ANSWER



SUMMONS ANSWER



 

22 QUESTION

Sons of Zeus, is it lawful for me to enter into conversation with you? It is lawful: you are not polluted by this slaughter. May I too take part in this conversation, sons of Tyndareus? You too. I will hold Phoebus responsible for this act of bloodshed. How . . . ?

Such pre-expansions typically function to ensure that the more central conversational moves can be executed successfully, in this case by verifying that the siblings actually have a willing conversational partner in the Dioscuri. The fact that such summon–answer sequences are relatively rare in tragic dialogue suggests something about the significance of moments such as this one, where the siblings’ position is particularly delicate.23 The use of expansion sequences and dispreferred responses may, as these few examples have already suggested, reveal much about the dynamics of a particular conversation and about speakers’ attitudes and behaviour.24 Dispreferred and failed responses in particular may 22 For the text and attribution of lines to speakers (both of which have suffered from needless confusion), see ch. II, §6.2. 23 For discussion of a similar pre-sequence at Eur. Tro. 48–53 (Athena and Poseidon, in an equally delicate situation), see Lloyd 2009. Lloyd uses politeness theory, not CA, to approach the scene (see also §2.4.2). 24 For a suggestive discussion of the use of conversational features in characterization, see Culpeper 2001: 172–80.

14

Introduction: Modern Linguistics and Euripides’ Electra

be a sign that communication between interactants is not proceeding fluently, and I will argue in chapters to follow that such conversational features underline Electra’s isolation and lack of connection to her adoptive surroundings in the play. Moreover, the distribution of turns and the selection of speakers can reveal much about differing levels of power and involvement. Observing which speakers tend to initiate conversational sequences, self-select, ‘hog’ the floor, and so forth can suggest which characters are dominating a certain scene, functioning as instigators, and so forth (differences of gender also play a role here; see §2.4.1). Related to this is a final notion that has been extensively studied in CA, namely that of topic management*: CA has shown that, as well as involving certain devices that speakers may use in order to introduce, maintain, or change a discourse topic (what a conversation is ‘about’), topic management is related to power distribution no less than to speakers’ attitudes, different degrees of engagement with the interaction, and elements of this sort.25 Topic management, too, will play a part in my analysis of the language of Electra’s characters.

2.3. Pragmatics In my treatment of the structural mechanics of the turn-taking system in §2.2, I used such terms as ‘question’, ‘greeting’, ‘request’, or ‘complaint’ as if they were self-evident categories that require no further comment. The question how certain utterances come to express these particular communicative intentions and effects, in other words what the relationship is between the form of utterances and their meaning, was left unanalysed. Yet such issues are anything but simple. Meaning as construed by participants in communication is not meaning in the abstract: in other words, the meaning that an utterance has in a particular interaction is not simply a sum of the meanings of the words that make up that utterance, organized by some syntactical rules. In fact those constituent words are often no more than ‘cues’ for the interactants to engage in all kinds of interpretative processes, which eventually yield a meaning much richer and much more complex than what is expressed by the linguistic 25 See Culpeper 2001: 174. Culpeper refers to ample empirical support for these claims.

Introduction: Modern Linguistics and Euripides’ Electra

15

tokens uttered.26 Consider the following lines from the parodos of Electra (171–80): ΧΟ.

νῦν τριταί‐ αν καρύσσουσιν θυσίαν. Ἀργεῖοι, πᾶσαι δὲ παρ᾽ Ἥ‐ ραν μέλλουσιν παρθενικαὶ στείχειν. ΗΛ. οὐκ ἐπ᾽ ἀγλαΐαις, φίλαι, θυμὸν οὐδ᾽ ἐπὶ χρυσέοις ὅρμαις ἐκπεπόταμαι. τάλαιν᾽, οὐδ᾽ ἱστᾶσα χοροὺς Ἀργείαις ἅμα νύμφαις εἱλικτὸν κρούσω πόδ᾽ ἐμόν. Ch. El.

175

The Argives are proclaiming a sacrifice two days from now, and the maidens are all going to go in procession to Hera. Not for fineries, my friends, nor for golden necklaces has my heart taken flight, me wretched one, nor will I stamp my whirling foot setting dances together with the brides of Argos.

The propositional content of the Chorus’ and Electra’s utterances might be paraphrased as follows: ‘All the virginal maidens are going to a festival of Hera two days from now’, and ‘I have no desire for fineries or jewellery and I am not going to dance’. But with this description of ‘literal’ meanings27 we have in no way described what the corresponding utterances actually mean in the context of their interaction. The women of the Chorus are inviting Electra to come along to the festival (an inference Electra has to draw, since the women never explicitly say something along the lines of ‘We invite you’). Electra’s response, in turn, should be seen as a rejection of the Chorus’ invitation, even though she never explicitly says so.28 26 The notion involved here is that of ‘linguistic underdeterminacy’, which is the driving force behind a massive body of linguistic research, including all work in pragmatics. For a good discussion of this concept, see Carston 2002: ch. 1. 27 I use scare quotes around the word ‘literal’ deliberately: it is a highly problematic term (see e.g. Recanati 2004). 28 There is in fact a lot more inferencing and processing going on in the interpretation of these utterances than is hinted at here. Electra and the audience must infer, for example, that πᾶσαι παρθενικαί refers to all the maidens in Argos (rather than in Sparta, Troy, the whole world, etc.), that παρ᾽ Ἥραν refers to the festival and sacrifice mentioned one line earlier (θυσία, 173), that στείχειν involves the particular kind of ‘going’ that maidens do on the occasion of festivals, and so on. All this information requires interpretative work beyond what can be gathered only from the words of the sentence: it depends, one may say, on ‘pragmatic intrusion’ into the ‘truth‐ conditional’ meaning of the utterance.

16

Introduction: Modern Linguistics and Euripides’ Electra

Meaning in communication thus depends on a variety of factors: although it is somehow based upon the propositional content of utterances (the Chorus will normally not, for instance, infer from Electra’s words that she likes cheese), this content does not determine it fully; the ‘end product’ is rather the result of complex inferences, assumptions, knowledge, and desires of the participants to the interaction and depends greatly on the linguistic and extralinguistic context of that interaction. Nor is such meaning necessarily univocal: different people can understand the same utterance in different ways and the same person can understand a single utterance in more ways than one; miscommunication and non-communication occur alongside successful interaction. As I will argue in chapter II, the Chorus’ word παρθενικαί is suggestive of why communication between the women and Electra is not very effective). The branch of linguistics that deals with meaning in communication (or ‘in use’, ‘in context’, etc.) is known as pragmatics*. It is, notoriously, not a well-delineated discipline: it has been aptly described as a ‘wastebasket’,29 which has come to contain all the aspects of language study that do not fit into the traditional fields of semantics and syntax—in other words, all those aspects that are not easily described by the formal systems of analysis (akin to logic and mathematics) that are the tools of most semanticists and syntacticians. But linguists do not agree on what exactly is in the wastebasket:30 depending on one’s definition, pragmatics studies either only a few topics, fairly well defined (deixis, reference, speech acts*, For the ‘wastebasket’ metaphor, see Mey 2001: 19–21. There is a rough distinction between a (narrower) Anglo-American conception of the field and a (wider) continental European one; see Mey 2001: 6–10 or Huang 2007: 4–5 for a discussion of problems in delimiting the use of the term. Mey’s book is an especially readable introduction to the discipline; Huang’s a slightly more formal one. Yule 1996 is a decent introduction on a smaller scale. Levinson’s (1983) classic textbook is out of date in some respects but remains one of the finest treatments of any field in linguistics, and it contains an extensive discussion of problems of definition (pp. 5–35; Levinson fits into the Anglo-American tradition). The literature on the pragmatics of ancient Greek is very rich, but—unsurprisingly for this diffuse area of linguistics—there is no single generalizing work that may serve as a starting point for the uninitiated (not even in a Companion: Bakker 2010b offers two valuable case studies rather than an overview). A selection of some works with ‘pragmatic(s)’ in their title will give an indication of the range of topics covered: Bonifazi et al. in press, Slings 1992, 1997b, Dik 1995, Brown 2006, de Jong and Rijksbaron 2006, Nordgren 2015. Using the search term ‘pragmatics’ in the online version of Giannakis 2014 will also give a good idea of various areas of current research. 29 30

Introduction: Modern Linguistics and Euripides’ Electra

17

inference and implicature*, and presupposition), or a much wider range of issues, which encompass all the approaches discussed in this chapter, including conversation analysis and sociolinguistics.31 At the same time, recent trends in semantics and syntax have aimed at reclaiming certain topics from what is normally considered the purview of pragmatics (or in some cases have denied that those topics should be part of linguistic enquiry at all), so that defining what counts as what is harder than ever. Still, some parts of this fluctuating territory are stable and fundamental, and they include the two strands of exploration that will be most relevant to my analysis of Electra: speech acts and meaning. Not by coincidence, these two strands can be traced back to the ‘founding fathers’ of pragmatics (it is fair to call them that): the former to Austin and Searle, the latter to Grice.32

2.3.1. Speech acts Simply put, speech act theory helps us understand what speakers do with utterances: the women of the Chorus invite Electra, Electra rejects the invitation. The notion that people do anything at all with their words was for a long time not recognized in—or at least not central to—linguistic theory: the traditional focus of language philosophers was the declarative sentence and the conditions under which such sentences were true or false (their ‘truth‐conditional meaning’). In this tradition there was a distrust for ordinary, everyday language use, which was considered at best inaccurate, at worst positively meaningless. The work of J. L. Austin was instrumental in causing a move away from the strictures of truth-conditional analysis. In a series of lectures published posthumously as How to Do Things with Words (Austin 1962), Austin pointed to such sentences as ‘I declare war on Zanzibar’ and ‘I dub thee Sir Walter’ and argued that the corresponding utterances have a force that changes the world in substantial ways, rather than having a truth-conditional meaning. The utterances do rather than describe something: the saying of the sentence itself I am personally sympathetic to a ‘wide tent’ approach to pragmatics: my (sub)section divisions should accordingly not be taken to express a theoretical persuasion. 32 The key, foundational publications by these three have been usefully reprinted together in Martinich 2010: part II. 31

18

Introduction: Modern Linguistics and Euripides’ Electra

results in a state of war being in effect, or in Walter’s having the title. Austin described utterances such as these as performatives*, which he contrasted initially with constatives (assertions, statements, i.e. utterances that describe states of affairs). He went on to argue that every utterance should in fact be seen as some kind of performative, in the sense that they all have one force or another: they are used to make requests, give orders, place bets, ask questions, and so on. Even apparent constatives are best treated as ‘doing’ something, namely asserting or stating. Austin and other speech act theorists distinguish between three different levels at which utterances are acts: the locutionary act (the act of saying the utterance), the illocutionary* act (the act performed by making the utterance, such as requesting, stating, warning, etc.), and the perlocutionary act or effect (the bringing about of certain effects by means of the utterance). Unsurprisingly, speech act theory has focused on illocutionary force, in other words on what utterances ‘count as’—or even ‘are intended as’: their function in communication.33 Speech act theory replaces truth conditions with certain conditions of appropriateness that must obtain before an utterance can be said to ‘count as’ an example of a speech act with a certain illocutionary force. In order for the performative ‘I dub thee . . . ’ to actually be successful at making Walter a Sir, the speaker has to be in a position of sufficient authority to do so. If such felicity conditions* are not met, the utterance misfires*. Misfiring can also happen at the other end: a speech act of betting, for example, is not really felicitous until the addressee responds with ‘You’re on’ (or a similar expression) to signify his ‘uptake’ of the utterance as a bet. Various classifications of speech act types have been proposed. The most influential taxonomy is that of John Searle, a student of Austin and the towering figure of speech act theory.34 His classification

33 When studying the pragmatics of a ‘dead’ language, one must necessarily rely heavily on perlocutionary effects (what kind of reactions does a certain speech act elicit?) in determining how various utterances function in terms of their illocutionary force. 34 Many other taxonomies have been proposed; see, e.g. Bach and Harnish 1979. Within classics, the classification in Risselada 1993: ch. 1 (on Latin directives) offers considerable advantages over Searle’s; I have retained Searle’s list only because it is still widely accepted as the standard and because it is sufficiently suitable for my own purposes. Some doubts must remain about the potential for any such classificatory exercise to capture all possible kinds of speech act (see Levinson 1983: 240–2).

Introduction: Modern Linguistics and Euripides’ Electra

19

Figure 0.3. Speech act types. Following Searle 1979.

(Searle 1969, revised in 1979) depends on what kind of illocutionary point is expressed by an utterance and on how the speech act relates to the world (the ‘direction of fit’). It is summarized in Figure 0.3. This classification already makes it clear that there is no one-to-one fit between different speech act types and the syntactic structure or propositional content of utterances. In other words, although there are ways of making the illocutionary force of an utterance explicit, for example by using a phrase like ‘I promise [to be back tomorrow]’, no such indication is required in order for an utterance to count as a certain type of speech act. An utterance like ‘Congratulations!’ seems in fact to have little syntactic structure or propositional content, if at all. More importantly, there is no necessary correlation between the three different syntactic sentence types that are present in most languages (declarative, interrogative, and imperative) and what speakers do with sentences of those types. To give two of the most frequently cited examples, both of the sentences below would normally function as requests: Can you pass the salt? [Interrogative sentence type] It’s cold in here. [Said to someone sitting by an open window; declarative sentence type]

20

Introduction: Modern Linguistics and Euripides’ Electra

Searle called utterances like these, which do not express their illocutionary force by ‘wearing it on their sleeve’, indirect speech acts*. It should be noted straightaway that this distinction between direct and indirect speech acts has been called into question, generally for good reasons.35 Yet, even if we question the validity of the concept of indirect speech acts, still a more general notion of indirectness has proven to be incredibly helpful to linguists by making sense of a whole range of other issues, such as metaphorical language, irony, or politeness (see §2.3.2). We will see shortly how indirectness can be explained in terms of Gricean pragmatics. Looking again at the interaction between Electra and the Chorus, speech act theory as described thus far proves quite helpful in analysing what is going on. The Chorus’ statement is an indirect speech act inviting Electra to the festival, and Electra similarly responds with an indirect speech act turning the Chorus down. The notion of felicity conditions is useful in determining why communication between them appears to be flawed: one of the conditions for inviting someone to something would normally be that the invitee is able and allowed to attend the occasion. As we will see in chapter II, it appears that Electra and the Chorus are not exactly in agreement on whether these conditions are really met: for Electra, who has a fundamentally different view of the state of things, the Chorus’ invitation ‘misfires’ (but then so too does Electra’s response: more about this in chapter II). Left unanswered, however, is the question how Electra (and the audience) can know that the Chorus’ statement is really an invitation: how do we assess what is really meant beyond bare propositional meaning? To answer these questions, I now turn to the other founding father of pragmatics, the British language philosopher Paul Grice.

2.3.2. (Neo-)Gricean theories of meaning Grice’s crucial suggestion, expounded in lectures later published under the title ‘Logic and Conversation’ (Grice 1975),36 is deceptively simple: human conversation is by nature a cooperative enterprise. 35 See Levinson 1983: 263–76 (with compelling arguments against what Levinson calls the ‘literal force hypothesis’) and Huang 2007: 111. Another problem is that some ‘indirect’ speech acts become conventionalized to the extent that they seem to ‘wear their illocutionary force on their sleeve’ after all: the ‘normal’ interpretation of a question beginning with ‘Could you . . . ?’ is as a request. 36 Grice pursued his theory through publications in 1978, 1981 and 1989.

Introduction: Modern Linguistics and Euripides’ Electra

21

He argued that conversation is guided by certain unwritten and unspoken rules, which all participants to an interaction expect all others to follow. These rules are part of a general cooperative principle*, which Grice formulated as the following prescription for speakers: Make your conversational contribution such as is required, at the stage at which it occurs, by the accepted purpose or direction of the talk exchange in which you are engaged. (Grice 1975: 45)

This principle subsumes several more specific maxims* of conversation: a maxim of quantity* (make your contribution no more and no less informative than required), one of quality* (try to make your contribution one that is true), one of relation* (be relevant), and one of manner* (be perspicuous, avoid obscurity and ambiguity). According to Grice, a speaker in a conversation may normally be expected to observe the overall cooperative principle: s/he will therefore normally observe the maxims. Yet speakers frequently deliberately fail to comply with the rules, blatantly flouting one or more maxims. In such a case, the addressee will attempt to reconcile this exploitation of a maxim with his/her expectation that the speaker is in compliance with the cooperative principle by trying to draw from the utterance the kind of meaning that does suggest that the speaker is cooperating with the principle. In this way an utterance will give rise to what Grice called a conversational implicature*. Grice offers many examples of such conversational implicatures—for example the following two, where the maxims of quantity and relation are respectively exploited: A is writing a testimonial about a pupil who is a candidate for a philosophy job, and his letter reads as follows: ‘Dear Sir, Mr. X’s command of English is excellent, and his attendance at tutorials has been regular. Yours, etc.’ (Gloss: A cannot be opting out [of the cooperative principle], since if he wished to be uncooperative, why write at all? He cannot be unable, through ignorance, to say more, since the man is his pupil; moreover, he knows that more information than this is wanted. He must, therefore, be wishing to impart information that he is reluctant to write down. This supposition is tenable only on the assumption that he thinks Mr. X is no good at philosophy. This, then, is what he is implicating.) (Grice 1975: 52) At a genteel tea party, A says Mrs. X is an old bag. There is a moment of appalled silence, and then B says The weather has been quite delightful

22

Introduction: Modern Linguistics and Euripides’ Electra

this summer, hasn’t it? B has blatantly refused to make what HE says relevant to A’s preceding remark. He thereby implicates that A’s remark should not be discussed and, perhaps more specifically, that A has committed a social gaffe. (Grice 1975: 54)

If we now look again at our earlier example from Electra, it is possible to explain how the Chorus’ invitation works as a conversational implicature, using a similar gloss: The women of the Chorus have said that all the maidens are going to the festival, which as a mere informative statement seems not to conform to the maxims of quantity (they are saying more than Electra needs to know) and relation (what’s it to Electra?). But there is no reason to suppose that they are violating these maxims. They must, then, be trying to impart more than what they have literally said, a supposition that is tenable on the assumption that they want Electra to join them. This, then, is what they must be implicating. Using the cooperative principle and the theory of conversational implicatures, Grice’s theory thus begins to allow us to explain how meaning can be inferred beyond that of (or can replace that of) the bare propositional content of an utterance, in other words to understand how more is communicated than what is explicitly said. This, in turn, helps in explaining a whole host of ‘indirect’ uses of language, including not only indirect speech acts (which get their ‘true’ illocutionary force as an implicature) but also such literary techniques such as irony, metaphor, tautology, and so forth.37 It is, then, unsurprising that such a powerful model has had a profound and lasting impact on pragmatic theory, to the extent that Grice’s writings have rightly been claimed to form the basis of ‘virtually all current work in linguistic pragmatics’ (Chierchia and McConnellGinet 2000: 239). Much of this ‘Neo‐Gricean’ work has been concerned with reinterpreting, revising, or reconstructing the cooperative principle and its constituent maxims: all of those maxims, and the various submaxims that Grice proposed, have been subjected to intense 37 Irony/sarcasm: violation of the maxim of quality (e.g. ‘Medea treated her children wonderfully!’—not true), therefore something else must be implicated (Medea was very cruel to them). Metaphor: violations of all kinds of maxims (quantity, relation), leading to reinterpretation. Tautology: violation of the maxim of quantity (e.g. ‘War is war’—says nothing that is informative), again leading to an implicature (see ch. II, nn. 57, 67, and 188 for the implicatures of various tautologies in the Electra).

Introduction: Modern Linguistics and Euripides’ Electra

23

scrutiny and debate. Yet the whole body of work suffers from some of the same flaws that can be imputed to Grice’s original proposals: • Some of Grice’s postulates (and those of many Neo-Griceans) do not appear to hold up in different cultural settings, and the extent to which any of them is ‘universal’ has therefore been questioned.38 • A more fundamental theoretical problem lies in the description of conversation as by nature and universally cooperative: in certain situations (often situations of conflict), conversation clearly neither is nor is expected to be cooperative.39 Greek tragedy, in fact, very often features contexts where interactants explicitly do not expect their interlocutors to be truthful or informative. Rather, as Leezenberg has pointed out: tragic utterances are driven by something like a linguistic habitus, that is, by an expected way of speaking and acting that is differentiated according to age, gender, and social status, rather than by the conscious deliberation of an autonomous and rational speaker. One cannot speak here of any neutral or ‘normal’ form of conversation that is violated or exploited, but rather of a socially constituted and differentiated power to speak in specific ways, a power which moreover is always open to negotiation and challenge. (Leezenberg 2005: 200–1)

• At a more practical level, the Gricean model does not give an explicit account of how conversational implicatures are derived (or how some are, but others are not). On the speaker’s side of things, it is left unexplained why a speaker would produce utterances that flout maxims in the first place (if one intends to be cooperative anyway, why not simply obey all the maxims?) What we need, then, is a theory that still allows us to maintain a saying–implicating distinction—in other words to understand how more can be communicated than is said without promoting cooperation to an inviolable rule of linguistic behaviour—and that explains why speakers produce utterances in certain forms and hearers derive certain inferences from them as opposed to others. The best and most

38 39

See Huang 2007: 34–5 with n. 10 for discussion. See Mey 2001: 76, with further references.

24

Introduction: Modern Linguistics and Euripides’ Electra

influential attempt at such a theory has become known as relevance theory*.40 Relevance theory is a theory not merely of communication, but of human cognition in general. It claims that human cognition tends to seek optimal relevance*, where ‘relevance’ is defined in a technical sense, as a balance between the processing effort required to interpret a stimulus (an input) and the amount of cognitive effects achieved: the greater the effort, the lower the relevance; the greater the cognitive effects, the greater the relevance. For Sperber and Wilson, the proponents of this theory, ‘meaning’ is a change in a person’s ‘cognitive environment’. Applied to communication, relevance theory proposes that both the production and the interpretation of utterances are governed by a single principle of relevance*, which is reminiscent of Grice’s maxim of relevance but is seen to be an overarching law of all cognition rather than a principle with various constituent maxims. Sperber and Wilson formulate their principle as follows: Every ostensive stimulus conveys a presumption of its own optimal relevance. (Wilson and Sperber 2004: 612)

What this means is that every communicative utterance will be presumed to be relevant enough to be worth the receiver’s processing effort, and that it is the most relevant one that the communicator can produce given his/her abilities and preferences (which need not include being cooperative, informative, etc.). In other words, a speaker will produce an utterance as relevant as s/he is willing and able to produce, and the addressee will try to derive as much relevance from that utterance as s/he can. Thus this process of deriving meaning from an utterance is geared towards deriving as much meaning from it as possible while exerting as little processing effort as possible (hence an addressee will not go on deriving unlimited numbers of implicatures). The processing effort decodes the propositional meaning of an utterance and also tackles all the necessary pragmatic inferences:

40

The theory originates in the work of Dan Sperber and Deirdre Wilson (1986), who updated it in a second edition in 1995 and continued it in Sperber and Wilson 2002 and in Wilson and Sperber 2004 and 2012. It has been usefully elaborated upon by Carston (Carston and Uchida 1998, Carston 2002).

Introduction: Modern Linguistics and Euripides’ Electra

25

There may be ambiguities and referential ambivalences to resolve, ellipses to interpret, and other underdeterminacies of explicit content to deal with. There may be implicatures to identify, illocutionary indeterminacies to resolve, metaphors and ironies to interpret. All this requires an appropriate set of contextual assumptions, which the hearer must also supply. (Wilson and Sperber 2004: 613)

The interpretation arrived at may include various implicatures of varying strength: stronger ones, which are vital to satisfy the expectation of relevance, and weaker ones, which help with satisfying that expectation but are not essential because the utterance suggests a range of similar possible implicatures, any one of which would do. The co-occurrence of various weaker implicatures allows for ambiguity. Looking at the Chorus’ invitation to Electra one last time, we might imagine Electra’s (and the audience’s) processing of their utterance along the following lines (not intended as a chronological sequence, but as an ongoing process with multiple aspects that operate in parallel): • decoding the linguistic tokens of the utterance and performing disambiguation and reference resolution so that the proposition derived is maximally relevant: πᾶσαι . . . αἱ παρθενικαί refers to all Argive maidens; παρ᾽ Ἥραν to the festival just mentioned, στείχειν to a procession—and so on (see n. 28); • supplying various contextual assumptions about the situation (including whether or not the women might intend merely to inform Electra about the festival happening, whether it is the kind of festival they might think Electra should attend, etc.); • constructing an appropriate hypothesis about the intended meaning or force of the utterance (which will be considered optimally relevant as an invitation, as lesser interpretations do not sufficiently change Electra’s cognitive environment to be worth the processing effort). Finally, after this discussion I should briefly point out how I believe Gricean pragmatics, particularly its most developed offshoot, relevance theory, may prove helpful to my reading of Electra. First, as a way of looking at meaning in communication, such theories of inference provide us with ways of explaining how speakers may mean more and different things from what they ‘literally’ say, while allowing (in the case of relevance theory at least) for ambiguity, multiple meanings, and

26

Introduction: Modern Linguistics and Euripides’ Electra

communicative misfires. At numerous points in the following pages I hope to make clear how such theories of inference may make clear what is going on in the communication between characters and between Euripides and his audience. Relevance theory in particular also suggests something else, namely that we should pay attention to every word on the page. Every word a character utters demands that other characters put in the required processing effort to make some sense of it, and we are to assume that every utterance and every part of an utterance are significant for the message that a speaker wants to convey as well as (and sometimes even more) for the ‘message’ that Euripides is conveying to us. In a way, relevance theory thus confirms that, in order for us to approach questions of how an original audience member might have reacted to a play, close textual reading is an absolute minimum requirement.

2.4. Sociolinguistics The pragmatic theories described so far focus on what speakers do with utterances and how addressees make sense of them given that the sentence form and the propositional content of most utterances are insufficient or even irrelevant to what is really meant. But thus far we have treated speakers and addressees themselves mostly as abstractions: in reality, of course, the speech of participants in conversation is determined in part by all kinds of social factors, such as status, race, gender, age, religion, cultural norms of social behaviour, and so on. Each of these social determinants brings with it a complex of rights, responsibilities, and taboos with regard to speech. If we hear a school pupil on his first day greet his teacher with ‘What’s up, Jill?’, we are surprised not because the ‘meaning’ of that utterance is hard to work out or because a greeting is an unsuitable kind of utterance for the occasion, but because we expect speech between school pupils and teachers to be regulated by certain social pressures that this particular student appears not to heed. Though it may easily be subsumed under the broad heading of pragmatics (and often is),41 the branch of linguistics that studies the

41

See §2.3 (with n. 30).

Introduction: Modern Linguistics and Euripides’ Electra

27

effects of such societal aspects on language use is widely known as sociolinguistics*. Despite the difficulties inherent in studying ancient Greek through this lens (lack of native speakers, meagre evidence for the social roles of some groups, such as women and slaves), there has been a steady flow of work in this area in the past 20 years; Eleanor Dickey’s (1995, 1996) work on forms of address and Andreas Willi’s (2003) book on sociolinguistic variation in Aristophanes are significant examples. Including these contributions, most work has centred on the effects of gender (especially female gender) on language use,42 while some work has also begun to emerge in the study of politeness phenomena, a burgeoning subfield of sociolinguistics.43 These two aspects will prove most important for my discussion of Electra as well (though see §2.2 for the relation between power and conversational behaviour and §2.5 on social aspects of gnomic utterances), and I will briefly discuss the state of research in both of them more fully.44

2.4.1. Gender ‘The perception that women use different speech from men, and employ it differently, is much older than the science of linguistics’, notes Mossman (2001: 374). This is very true, but the science of linguistics has done much to match that perception with an understanding of how exactly women’s speech is different from men’s and how they employ it differently. The bibliography on language and gender (in modern languages) is terrifyingly vast,45 and differences 42 On female (and male) speech in Greek, see Gilleland 1980, Bain 1984, McClure 1995, 1999, Sommerstein 1995, Dickey 1995, 1996, McClure and Lardinois 2001, Mossman 2001, Alexiou 2002: esp. 102–4, Willi 2003, Duhoux 2004, Fögen 2004, 2010, Minchin 2007: part II, Chong-Gossard 2008, Mastronarde 2010: ch. 7, Rutherford 2012: ch. 7. 43 On politeness, see Brown 2003, 2006, Willi 2003, Lloyd 2004, 2006, 2009, 2013, Minchin 2007, and Scodel 2008, all with regard to Greek. There are some remarks also in Dickey 1996, and one will find a careful discussion in Denizot 2011: esp. 484–91. For Latin, see e.g. Hall 2009. Lloyd’s contributions have been especially helpful, although in §2.4.2 I will register some issues with the Brown–Levinson model, which he uses without much criticism. 44 Gender and politeness, incidentally, are increasingly interrelated fields in linguistics: a small library could be filled with books on how gender influences linguistic politeness (for an overview, see Mills 2003). 45 The modern debate begins with Jesperson 1922: 237–54; recent influential contributors are, among many others, Lakoff 1975, Tannen 1993, 1996, Cameron and Kulick

28

Introduction: Modern Linguistics and Euripides’ Electra

between male and female speech have been mapped to a great level of detail, their interpretation relentlessly and furiously fought over (two factions dominate: the explicitly feminist ‘theory of deficit/dominance’ and a more benign ‘theory of difference’). As for the application of such research to Greek literature, this is by no means a straightforward exercise: complications include the fact that women are given a prominent voice only in a few genres of Greek literature (tragedy and comedy above all others) and that the works in which we find them were mostly written by and for men. Women themselves, of course, had a largely marginal role in Greek society.46 Thus, even if we find differences between male and female speech, it is probable, as Adams (1984: 43) put it for Latin, that we are dealing with ‘popular stereotypes of female behaviour rather than . . . objective observations’. Still, much fruitful work has been done on genderspecific language in Greek literature,47 and there are more insights to be gained. The two-part question above (‘how women’s speech is different . . . how they employ it differently’) is an indication of how the issue of women’s speech in Greek tragedy might be examined at two different levels—what we might call a micro and a macro level. At the micro level, we might look at what women ‘sounded like’, in other words at points of (quantifiable) linguistic detail, such as the use of certain interjections, particles,48 syntactical constructions,49 word order, and so forth. Laura McClure (1995) has shown, for example, that certain expressions are exclusive to or preferred by either male or female 2003, and Coates 2004. Some recent introductory textbooks and textbook chapters are Eckert and McConnell-Ginet 2003, Julé 2008, and Holmes 2013: chs 7, 12. Numerous bibliographies and other resources may be found online, for example through http://www.lancs.ac.uk/fass/organisations/igala/Index.html (accessed 8 June 2016). 46 This is a topic too complex and vast to go into in any detail, though it must be inextricably linked to a sociolinguistic enquiry into female and male speech (in this case more than usually, the ‘socio‐’ part of ‘sociolinguistic’ is a real can of worms). For a concise summary with further bibliography, see McClure 1999: 19–24. On the role of women in tragedy, foundational work may be found in various chapters of Zeitlin 1996; for a brief introduction, see Mossman 2005. 47 See n. 42. 48 Denniston (1954: lxxiii) suggested that women were perhaps ‘peculiarly addicted to the use of particles, just as women to‐day are fond of underlining words in their letters’. This suggestion (the part about ancient Greek women, at least) is to some extent confirmed for Aristophanes by Willi (2003: 182–4). 49 In a review, Gildersleeve (1907: 208) suggested that irregular syntax in the speech of Homeric women was ‘intended to be feminine syntax’.

Introduction: Modern Linguistics and Euripides’ Electra

29

speakers in Euripides. These include interjections such as οἲ ᾽γώ (exclusively female), ἒ ἔ (preferentially female), and εἶα (preferentially male), certain forms of address, certain kinds of oaths, and so on. Still, McClure’s research at this level has not revealed a great many differences, and, even if we assume that she has not exhausted all the features we might find, the stylized nature of tragic language makes it likely that what gains can be made in this area will be on the margins at best.50 At the macro level we may ask broader questions: What do women talk about? What kind of rhetorical techniques do they use? What kind of speech activity are they likely to perform? McClure, again, has contributed much to our understanding of these aspects of female speech through her book Spoken Like a Woman (McClure 1999), where she deals with ‘verbal genres’ such as lamentation, gossip, and sexual persuasion. Mossman’s (2001) article cited above applies similar methods to the Electra, particularly to the characters of Electra and Clytemnestra. The gendered use of language in Euripidean song (and its ‘non‐use’ in silences) have also been examined in a book by Chong-Gossard (2008). All of these works will inform in particular my analysis of Electra’s language in chapter II. Though it is useful to keep in mind that gender variations occur in these different ways (both in form and in content), it may be superfluous to point out that no clear dividing line between my two levels can or should be drawn—in fact I will argue in chapter IV that the whole range of the issues involved needs to be considered in practicing textual criticism. Moreover, a crucial aspect of what modern research into gender-conditioned language has taught us falls somewhat outside the scope of my two levels: women and men ‘sound’ differently and use language differently depending on whether they are in single-sex* or mixed groups*.51 Thus, such differences as there are between male and female talk cannot be analysed without looking 50 See Bain 1984: 27. More research may yield unexpected findings, and areas such as particle usage remain un(der)explored. Matters are different in (Aristophanic) comedy, where Willi (2003: 176–92, 193) has isolated a significant number of female preferential features, many of which may make female speakers appear more polite than male speakers. The features he has identified for Aristophanes do not appear to show the same distributional variations in Euripidean tragedy (if they occur there at all), although more research on this subject is needed. 51 The principle was first explored, among others, by Kramer (1975) and West (1979) and has been further developed by Tannen (1993) and Coates (1996).

30

Introduction: Modern Linguistics and Euripides’ Electra

at contextual setting: female linguistic behaviour in all-female contexts differs from female behaviour in mixed groups as well as from male behaviour in all-male groups, and so forth. Very significant too are interruptions, either into a gender-exclusive conversation from someone of the other gender, or within mixed-gender interactions (where behaviour such as ‘grabbing the floor’ and ‘hogging the floor’ may reflect dominance). This realization that the make-up of the group matters, now well established in modern research, is slowly making its way into classical studies as well.52 I hope to show that it is highly significant in the case of Euripidean tragedy, where real differences can be demonstrated between women’s linguistic behaviour in female-only situations and in mixed groups, and where ‘interruptions’53 are significant as well. A neat example can be found early in our play: immediately after the parodos, Electra, who up to this point was on stage only with the female Chorus, or so she believed, is suddenly confronted by two men (215–17): οἴμοι· γυναῖκες, ἐξέβην θρηνημάτων. ξένοι τινὲς παρ᾽ οἶκον οἵδ᾽ ἐφεστίους εὐνὰς ἔχοντες ἐξανίστανται λόχου. Oh no! Women, I have left my lamentations: some strangers here, lying in wait at the altar by the house, are coming out of their ambush.

Electra’s first line virtually symbolizes the issue: Electra, up to this point only in the company of women (γυναῖκες is a significant form of address here), has to abandon her typically female linguistic behaviour (θρηνημάτων; lamentation is one of McClure’s ‘verbal genres’)54 as the female-exclusive setting is interrupted by male intrusion (ξένοι τινές). I hope to make clear in chapter II that this is only one of many points in the play where Electra’s language is fundamentally

52 The point is fundamental to Mossman 2001; it is also made, if in different terms (‘space’) and without the same grounding in current linguistic enquiry, by ChongGossard (2008: 6–21). 53 I use scare quotes here because interruptions are par excellence a feature of naturally occurring dialogue and we should be careful before applying insights from it to dramatic dialogue, especially dialogue as regimented as that of tragedy; still, interruptions do occur in tragedy and often seem significant (see ch. II, §4). 54 For work on male lamentation (which recent scholarship has begun to explore in its own right), see the references listed in ch. II n. 2.

Introduction: Modern Linguistics and Euripides’ Electra

31

determined by her gender, and that her linguistic behaviour varies depending on whether men are present. In chapter IV I will revisit the issue of gender-specific language in the context of some moments of contested line-to-speaker attribution in the play.

2.4.2. Politeness and power Attention to linguistic politeness phenomena had long been of a purely prescriptive nature (i.e. it had only produced manuals on how to speak politely), until sociological and anthropological interest in the topic seeped into linguistics in the 1970s. Since then, politeness research, both linguistic and other, has undergone an incredible expansion; and it continues to grow at breathtaking pace.55 Politeness as an object of research has been described as ‘a many‐headed hydra’ (Watts 2003: xi), not only because the concept is so incredibly complex, but also because many and sometimes diametrically opposed models have been developed to describe and explain it. Even if we focus merely on polite language use rather than on polite behaviour more generally (but should we?), dozens of competing approaches mean that it will be impossible to provide here more than the barest outline of the field. Still, what work has been done on linguistic politeness in Greek and Latin literature56 shows a fairly strong uniformity, since a single text underpins the approach of all those offerings. This classic work—to date the most influential text on linguistic politeness,57 even though it has been challenged by numerous and often justified critiques—is a long article by Penelope Brown and Stephen Levinson (1978), republished in 1987 as a book with a new introduction that proposes some ‘universals in language usage’.

55 A recent overview is Kádár and Haugh 2013. A bibliography of important works (‘by no means a comprehensive bibliography’) of the Linguistic Politeness Research Group lists some 200 titles (http://research.shu.ac.uk/politeness/bibliography.html, accessed 8 June 2016). There is also a quickly growing literature on impoliteness, for which see Bousfield 2008 and Culpeper 2011 (with further references). 56 See n. 43. 57 The major model competing with that of Brown and Levinson before the publication of Watt’s (2003) Politeness was that of Leech (1983, updated in 2014). Other models have not entirely lost all currency either: approaches by Robin Lakoff (1973, 1975), Fraser and Nolen (1981), Arndt and Janney (1987), and Blum-Kulka et al. (1989) are still frequently cited in the literature.

32

Introduction: Modern Linguistics and Euripides’ Electra

Brown and Levinson’s model revolves around Erving Goffman’s concept of face*, which Goffman (1955: 213) defined as ‘the positive social value a person effectively claims for himself by the line others assume he has taken during a particular contact [i.e. interaction]’ and as ‘an image of self, delineated in terms of approved social attributes’. Brown and Levinson modified this concept and split it into two parts, which correspond to the two universal human desires to be approved and admired (‘positive face’) and not to be imposed upon (‘negative face’): [positive face:] the positive consistent self‐image or ‘personality’ (crucially including the desire that this self‐image be appreciated and approved of) claimed by interactants [negative face:] the basic claim to territories, personal preserves, rights to non‐distraction—i.e. to freedom of action and freedom from imposition. (Brown and Levinson 1987: 61)

As far as possible, these desires should be maintained by participants in a social interaction for the entire duration of that interaction, and it is thus in the interest of all the participants to reduce possible threats to each other’s face (like Grice, whose model they utilize extensively, Brown and Levinson assume that conversation is by nature cooperative). Interactants in conversation will thus adopt communicative strategies designed to maintain each other’s face as best they can. In Brown and Levinson’s (1987: 65–8) account, face threats caused by conversational utterances such as criticisms and insults, which damage one’s positive face, or by commands, requests, and threats, which damage one’s negative face, are described as face-threatening acts (FTAs)*. FTAs litter everyday conversation and range from the very marginal to the quite severe. ‘Politeness’ is then defined as any measure undertaken by a speaker to reduce the threat of a FTA by supporting or enhancing the positive face of the addressee (‘positive politeness’) or by avoiding transgression of the addressee’s freedom of action and freedom from imposition (‘negative politeness’). Brown and Levinson (1987) suggest that a speaker will select one of various politeness strategies* to negate or minimize the threat that may be caused by a contribution he or she wishes to make to the conversation. The simplest strategy to avoid any face threat is not to make the intended contribution at all. At the other end of the spectrum, the ‘worst case’ is to perform the FTA with no redressive action, or baldly*. In-between these extremes, speakers can use various

Introduction: Modern Linguistics and Euripides’ Electra

33

linguistic instruments to offer redress; such instruments are oriented either towards the addressee’s positive face (positive politeness strategies, e.g. familiar forms of address, expressions of interest or agreement, 1987: 101–29) or towards his/her negative face (negative politeness strategies, e.g. expressions of deference, hedges, indirectness, apologies, etc., 1987: 129–211). Alternatively, and as an extension of negative politeness, speakers may phrase their utterances in an ambiguous way, so that it is possible to ignore the threat to face. Brown and Levinson call this last strategy going off the record, as opposed to the performance of unmistakable FTAs that are on the record. Such off-record* strategies are usually couched in the form of indirect speech acts (see §2.3.1). Which strategy from this list is selected will depend largely on the amount of face threat the speaker calculates that his FTA will carry. The greater the expected face threat, the more likely it is that a speaker will choose not to do the FTA or do it in a circumspect way. Figure 0.4 presents Brown and Levinson’s different strategies diagrammatically, with the different strategies ranked (from 5 to 1) according to how much compensation they can offer.

Figure 0.4. Politeness strategies. Brown and Levinson 1987: 60.

As for the estimation of risk of face loss, the seriousness of an intended FTA depends, according to Brown and Levinson, on the social distance between speaker and hearer, the power that the hearer has over the speaker, and the extent to which the act constitutes an imposition on the hearer. Most of Brown and Levinson’s book is concerned with a discussion of individual ‘substrategies’ that serve to pursue one or more of the ‘superstrategies’ laid out in the diagram. To give a few examples:

34

Introduction: Modern Linguistics and Euripides’ Electra [Positive politeness strategy:] Notice, attend to H (her/his interests, wants, needs, goods, etc.): Jim, you’re really good at solving computer problems. ! (FTA) I wonder if you could just help me with a little formatting problem I’ve got. [Negative politeness strategy:] Give deference: Excuse me, officer. I think I might have parked in the wrong place. [Off‐record strategy:] Use rhetorical questions: I’m sorry, but how was I to know that? (First two examples from Watts 2003: 89–90, third adapted from Brown and Levinson 1987: 223)58

Each of these examples in fact includes more strategies than one: in the second, for example, we find, apart from the deferential address ‘officer’, the conventionalized apology ‘Excuse me’, as well as various hedges in ‘I think . . . might’. Similarly, in the first example, the positive politeness strategy is combined with negative ones (hedging in ‘just’ and ‘little’ and the conventional phrasing as an indirect question ‘I wonder if . . . ’). According to Brown and Levinson, these strategies are used crosslinguistically, if not universally: the same strategy is found in different languages to serve similar FTA-redressing goals. It is this supposed universality that has allowed classical scholars to apply Brown and Levinson’s model to Greek texts (and I too will make use of their repertory of strategies, many of which appear operative in Greek). Yet the universality Brown and Levinson claim for their model has come under severe criticism from other sociolinguists—as have other aspects of their work.59 Many of these criticisms have been gathered together in a textbook on politeness by Richard Watts, whose own model has been gaining some ground in the field. Many of the points Watts (2003: 85–116) makes are familiar from earlier criticisms of Brown and Levinson: that they focus too much on politeness rather than on impoliteness (thereby ignoring the more salient behaviour);60

58 For a more recent taxonomy of politeness structures, see Watts 2003: 182–200, with further references. 59 The most extensive and scathing (if largely accurate) critique may be found in Werkhofer 1992. 60 This complaint parallels similar criticisms of Grice’s cooperative principle (see §2.3.2): for attempts to move towards a ‘conflict model’ of communication based on

Introduction: Modern Linguistics and Euripides’ Electra

35

that they fail to properly account for the role of the addressee and of bystanders in a conversation;61 that their dualistic account of face— the negative part—does not apply to many (mostly Far Eastern) societies that are more collectivist in nature; that the ‘variables’ in their formula cannot predict conversational behaviour (a criticism borne out by empirical studies).62 But Watts’ main point of criticism is much more fundamental in nature: he argues that Brown and Levinson (and others) improperly conflate linguistic politeness strategies (‘second‐order politeness’ or ‘politeness2’) with ‘folk’ or ‘lay’ interpretations of politeness, that is, with how polite a certain behaviour actually is perceived to be in conversations (‘first‐order politeness’ or ‘politeness1’). It is in fact very possible for someone’s use of Brown and Levinson’s politeness strategies to be considered not polite at all, and they can even be used with a deliberately impolite effect. We need only think of an example like ‘Would you ^mind closing the window?’63 to see that a mere change in the prosodic contour of a sentence can dramatically alter the way it is going to be interpreted (under Brown and Levinson’s theory, the indirect question is inherently a politeness device). Politeness1—‘real’ politeness—is, according to Watts, not merely a matter of the linguistic expressions that [one] uses, but rather depends on the interpretation of that behaviour in the overall social interaction. . . . A theory of politeness2 should concern itself with the discursive struggle over politeness1, i.e. over the ways in which (im)polite behaviour is evaluated and commented by lay members and not

power relations, see Leezenberg 2002 and 2005 (on Greek tragedy); Watts’ own model is in fact in this vein. 61 The addressee is involved only inasmuch as s/he factors into the speaker’s calculations of the weightiness of a FTA. As Werkhofer (1992: 157) phrased his criticism, Brown and Levinson’s model ‘is biased towards a one‐sided individualism, a bias that is not only due to the role ascribed to the speaker’s initial face‐threatening intention, but to other individualistic premises’. 62 Especially questionable in the light of these empirical studies is the notion that the weightiness of a FTA determines the selection of a politeness strategy. Nor is social distance as accurate a predictor of polite behaviour as, e.g., the affective relationship between speaker and hearer. More importantly, as Werkhofer (1992: 176) points out, because Brown and Levinson defined their variables ‘distance’, ‘power’, and ‘imposition’ as ‘static entities that determine polite meanings, these variables represent a narrow approach to social realities . . . that neglects the dynamic aspects of social language, aspects that . . . should be at the heart of a [proper account of politeness]’. 63 Where ^ represents a rising pitch (emphasis) on the verb ‘mind’.

36

Introduction: Modern Linguistics and Euripides’ Electra

with ways in which social scientists lift the term ‘(im)politeness’ out of the realm of everyday discourse and elevate it to the status of a theoretical concept. (Watts 2003: 8–9)

Watts proposes to improve on Brown and Levinson’s model by going back to the original Goffmanian conceptualization of face, which he thinks Brown and Levinson wrongly appropriated and misconstrued as a fixed property of individuals. The key, Watts argues, is that a social concept like face is open to constant change and renegotiation and that, in order to understand first-order politeness (i.e. behaviour that people actually consider to be polite), we need to be attuned to the dynamic nature of conversational interaction as social practice: no two conversations are the same, and the situation as mutually constructed by interactants changes constantly during a conversation. Watts’ shift of focus to this more contextually sensitive interpretation of politeness—his focusing on interlocutors’ perceptions of polite and impolite behaviour rather than on abstractions—leads him to reclassify much of what had been considered politeness as just supportive facework* with no function but to maintain politic behaviour*. Watts defines politic behaviour as ‘that behaviour, linguistic and non‐linguistic, which the participants construct as being appropriate to the ongoing social interaction’ (2003: 276): politic behaviour circumscribes the normal parameters in which an interaction of a particular kind between particular speakers operates (though the baseline may itself shift during the course of the conversation). On this account, many conventionalized or ritualized utterances such as ‘Could you pass me . . . ?’ and ‘Would you please . . . ?’, which Brown and Levinson considered politeness devices, are usually no more than what is expected in the circumstances and therefore do not have to give rise to an interpretation of polite behaviour (though they may). Only when speakers use linguistic facework in excess of what is required by the situation can this be interpreted as linguistic politeness;64 conversely, when interactants clearly do not do enough, or when they deliberately exacerbate a FTA (which is aggressive facework), such behaviour is interpretable as impolite. To complicate matters even further, politeness and impoliteness may be evaluated in

64 Watts uses relevance theory* to underpin his model: excess facework* is, then, maximally relevant in context if it is interpreted as politeness (which is an implicature* of an utterance).

Introduction: Modern Linguistics and Euripides’ Electra

37

different ways by different interactants: sometimes behaviour that is interpreted as polite is still evaluated negatively (for example when someone is being ‘too polite’ for the situation). A helpful feature of Watts’ treatment is his focus on the relationship between politeness and power, which he defines as ‘the freedom of action to achieve one’s goals, regardless of whether or not this involves the potential to impose one’s will on others to carry out actions that are in one’s interests’ (Watts 2003: 276). Rather than a fixed variable that factors into the calculation of how threatening a FTA will be (as in Brown and Levinson’s model), Watts sees power, too, as something over which people involved in social interaction continually struggle. The exercise of power—one person trying to affect another in a manner contrary to his/her initially perceived interests—can often be interpreted as impolite, although facework can be used in an attempt to prevent such an inference. Watts (2003: 213–15, 218–32) amply proves that the exercise of power and (im)politeness are closely interrelated. To round off this section, I will attempt to show how politeness theory might fruitfully be applied to Greek texts,65 by looking ahead at a few more lines of the Electra, immediately following the moment discussed in the section on gender (§2.4.1). As Orestes and Pylades come forward out of their hiding place, the following exchange between Electra and Orestes (in disguise) takes place (215–29): ΗΛ. οἴμοι· γυναῖκες, ἐξέβην θρηνημάτων. ξένοι τινὲς παρ᾽ οἶκον οἵδ᾽ ἐφεστίους εὐνὰς ἔχοντες ἐξανίστανται λόχου· φυγῇ σὺ μὲν κατ᾽ οἶμον, ἐς δόμους δ᾽ ἐγὼ φῶτας κακούργους ἐξαλύξωμεν ποδί. ΟΡ. μέν᾽, ὦ τάλαινα· μὴ τρέσῃς ἐμὴν χέρα. ΗΛ. ὦ Φοῖβ᾽ Ἄπολλον, προσπίτνω σε μὴ θανεῖν. ΟΡ. ἄλλους κτάνοιμι μᾶλλον ἐχθίους σέθεν. ΗΛ. ἄπελθε, μὴ ψαῦ᾽ ὧν σε μὴ ψαύειν χρεών. ΟΡ. οὐκ ἔσθ᾽ ὅτου θίγοιμ᾽ ἂν ἐνδικώτερον. ΗΛ. καὶ πῶς ξιφήρης πρὸς δόμοις λοχᾷς ἐμοῖς; ΟΡ. μείνασ᾽ ἄκουσον, καὶ τάχ᾽ οὐκ ἄλλως ἐρεῖς.

65

215

220

225

As with all sociolinguistic approaches, there are inherent risks in applying this methodology to a culture at great historical remove and to a language without living native speakers. For ‘historical’ politeness research, see Culpeper and Kádár 2010 (particularly the introduction, a discussion of methodological pitfalls and possibilities).

38

Introduction: Modern Linguistics and Euripides’ Electra

ΗΛ. ἕστηκα· πάντως δ᾽ εἰμὶ σή· κρείσσων γὰρ εἶ. ΟΡ. ἥκω φέρων σοι σοῦ κασιγνήτου λόγους. ΗΛ. ὦ φίλτατ᾽, ἆρα ζῶντος ἢ τεθνηκότος; El.

Or. El. Or. El. Or. El. Or. El. Or. El.

Oh no! Women, I have left my lamentations: some strangers here, lying in wait at the altar by the house, are coming out of their ambush. Let us run away in flight from these criminals: you go down the track, I will go into the house. Wait, poor lady! Do not shrink from my hand! Apollo, I plead to you, do not let me die! May I kill others more hated than you. Get away! Keep your hands off what you are not supposed to handle! There is no one whom I could touch with more right. And how is it then that you are lying in ambush, sword in hand, at my house? Stay, listen, and you may change what you say. I am still; I am yours anyway, since you are stronger. I am here bringing word to you from your brother. My dear man! Is he alive, then, or dead?

Brown and Levinson’s model cannot help much in making sense of Electra’s first few contributions to this exchange. She ignores completely Orestes’ first plea (μέν᾽ 220, ‘stay’) and directs her utterance at Apollo instead (221);66 in her next turn*, without any form of greeting or redress, she tells Orestes to get away and accuses him of touching someone whom he should not (223); her turn after that contains another accusation, this time of violent posturing (ξιφήρης) and laying an ambush (λοχᾷς, 225). All of these turns thus constitute threats to Orestes’ face, and Electra performs her FTAs ‘baldly’ and ‘on the record’. This could really only be explained in Brown and Levinson’s model as an estimation on Electra’s part that the FTAs are not that severe (see Figure 0.4). But the key, of course, is to realize that this is not the kind of situation where politeness has any place for Electra. She believes that she is being assaulted: not the kind of ‘interaction’ where one has to give deference or attempt to do any supportive facework. Brown and Levinson’s model, with its reliance on rational actors who constantly seek to maintain and support each other’s face, thus cannot really explain what is going on here. If we attempt to apply some of Watts’ terminology, the politic behaviour in such an extremely conflictual situation allows for any 66

Electra’s σε in this line is not addressed to Orestes; see Cropp ad loc.

Introduction: Modern Linguistics and Euripides’ Electra

39

amount of face threat without any compensating facework, and we should not be surprised that Electra remains, up to 227, entirely unfriendly. When we now look at Orestes’ contributions to the exchange, it is interesting to see how he seems to be negotiating with Electra over the very nature of the interaction they are having: he is constantly trying to modify how Electra views the situation (and thus what the politic behaviour of the situation is). His own utterances contain several instances of supportive facework, which we could interpret as polite: his use of ὦ τάλαινα (220), his assurance that she is not his enemy (222). At 222, and especially at 224 (τάχ’ οὐκ ἄλλως ἐρεῖς), his comments can be seen as attempts to steer the conversation towards cooperative rather than conflictual discourse. Electra remains unconvinced, however: her πάντως in line 227 suggests that she rejects Orestes’ suggestion,67 even though she has found it necessary to let go of her own perceived interests (running away) because of the enormous power that he has over her (armed man over unarmed woman). This rejection also indicates that, in spite of Orestes’ efforts at politeness, Electra herself has not yet evaluated any of his utterances positively. Only after the stranger has revealed that he speaks ‘on behalf of Orestes’ does Electra accept the new status of the conversation, and her sudden change of tone is remarkable: she calls him ὦ φίλτατ(ε) (229). This inaugurates a scene that, as I hope to show in my full analysis of the stichomythia in chapter II (§3.1), is the locus of a great many interesting features from the viewpoint of politeness theory. Applying such theory can, I believe, add much to our understanding of the interaction not only between Electra and Orestes, but between Electra and other characters (especially the Peasant) as well. As mentioned above, I will make use in my analysis of the repertory of linguistic politeness strategies proposed by Brown and Levinson, but I will also try to remain sensitive to context and interactants’ reactions (à la Watts), to avoid coming to untenable conclusions about (im)politeness in the play.

67

By highlighting that the current proposition (‘I am yours’) is true regardless of the truth value of preceding ones (Orestes’ reassurances), Electra signals that she rejects the validity of Orestes’ earlier statements. Cf. Watts’ (2003: 228) discussion of the similar English phrase ‘in any event’.

40

Introduction: Modern Linguistics and Euripides’ Electra

2.5. Gnomic utterances in context: some aspects of modern paroemiology General reflections on the vicissitudes of life, on the relationship between men and gods, and on proper moral behaviour are a feature found in Greek literature from its earliest incarnations. Both the Homeric narrator and his characters express such gnōmai in the Iliad and Odyssey, and after Homer they remained a popular feature of literature—as well as becoming increasingly important in rhetorical training and education.68 Tragedy is no exception; what is somewhat exceptional, however, is the way in which tragic gnōmai—especially Euripidean gnōmai—have been studied by scholars. The key term here is ‘relevance’69 (an issue that is germane, in fact, to many discussions of Euripides’ technique, particularly his choral odes). If we trace the development of scholars’ treatment of Euripidean gnōmai, we find early efforts to excise from our texts any instance that was felt to be irrelevant to the preceding and following lines. Many such generalizations were considered to be either intruding marginal comments or the interpolations of overzealous actors, trained in a culture that enjoyed rhetorical grandstanding. In recent decades this trend has been reversed70—if not fully71—as scholars reevaluate how general reflections (usually viewed chiefly as a feature of rhetoric) fit into the structure of the speeches that contain them and how they relate to the themes, motifs, and characters of plays. Still, considering this preoccupation among scholars with the contextual relevance of gnomic utterances, it is surprising to note the limited scope they have allowed, even recently, for the term ‘context’. What many fail to take into account is that every gnōmē uttered is in itself a communicative act situated in a particular communicative context: every general reflection is uttered by a particular speaker, speaking to a particular (set of ) addressee(s), in a particular spatio68 See RE (Suppl. vol. 6, 74–90), s.v. ‘Gnome, Gnomendichtung, Gnomologien’ (by Horna). 69 The term is not used here to refer to relevance theory* (on which see §2.3.2). 70 For a good discussion and further bibliography, see Rutherford 2012: ch. 9. Most (2003) looks specifically at gnōmai in Electra, comparing them with those in Aeschylus’ Choephori and Sophocles’ Electra. 71 Indeed one of the more serious demolition enterprises in the text of Electra was a deletion of nearly all of the gnomic passage 367–400 by Reeve (1973), who was followed to a large extent by Distilo (2012). I shall register my disagreements with Reeve more fully in ch. III (§2.2.3).

Introduction: Modern Linguistics and Euripides’ Electra

41

temporal setting, with particular aims. Therefore the questions we should ask are not only ‘How does this gnomic utterance fit in with the rest of the speech?’, but also ‘Why does this speaker use this gnomic utterance here?’, ‘Why does he/she do so in this company?’, and—perhaps the one most important of all—‘Why does the speaker use a gnomic utterance at all (as opposed to another form of expression)?’ When it comes to a closely related feature, the use of proverbs, questions very similar to these have been asked for some time by anthropologists and ethno- and sociolinguists working on language use in present-day societies. Indeed, ‘context’—in a broad sense—has become central to the field of paroemiology (the study of proverbs).72 Some of the findings and theories of modern paroemiologists were applied to Greek gnomic utterances in a dissertation and a series of articles by André Lardinois.73 His discussion of Homeric gnōmai (and, in one case, of Sophoclean ones)74 is instructive, and I shall follow his approach to some extent in my own treatment of gnomic passages in Electra. Readers may wonder whether work on present-day proverbs can be readily applied to ancient gnomic language. Indeed, as Lardinois (1997: 213–14) notes, ‘Greek gnōmai are not the same as modern proverbs’;75 yet, he adds, ‘they can be effectively studied in the same way’. Proverbs, in the strictest sense, are in fact a rather limited subset of a larger category of ‘wisdom sayings’: what makes proverbs exceptional within this larger category is the fact that they are normally anonymous and passed on in (almost) identical form from one

72 A recent handbook surveying the field is Hrisztova-Gotthardt and Varga 2014, useful collections of essays are Mieder and Dundes 1981 and Mieder 1994. I have also profited much from Norrick 1985. 73 Lardinois 1995, 1997, 2000, 2001, and 2006. Some other works have looked at how gnomic language characterizes speakers in Greek literature (e.g. Tompkins 1993, Griffith 1999: index s.v. gnōmē, Budelmann 2000: 74–80, Mossman 2001 and 2011: index s.v. ‘wisdom’). 74 Lardinois 2006. 75 Nor, to be fair, are ancient gnōmai the same as ancient proverbs: see Morgan 2007: 26 n. 19. Distinguishing between γνώμη, παροιμία, ἀπόφθεγμα, ὑποθήκη, and the like is not straightforward; for discussion, see chapter 2 on ancient proverbs in Morgan 2007, as well as Kindstrand 1978, Lardinois 1995: 7–20, and Russo 1997. I will use gnom- words as catch-all terms (alongside ‘general reflection’, ‘sententious statement’, etc.).

42

Introduction: Modern Linguistics and Euripides’ Electra

generation to the next.76 We do not find such formal stability in (early) Greek gnōmai: Instead, what we find are novel renditions of the same basic thoughts; for example, Odyssey 7.294: αἰεὶ γάρ τε νεώτεροι ἀφραδέουσιν (‘For always thoughtless are youngsters’) expresses the same sentiment as Iliad 3.108: αἰεὶ δ᾽ ὁπλοτέρων ἀνδρῶν φρένες ἠερέθονται (‘Always flighty are the minds of young men’) but in different words. I submit that Greek gnomai were, at least until the fourth century BCE, part of a living tradition in which every performance was a re‐creation, very much like epic verse. They are, like epic verses, ‘coined’ with the help of traditional formulae and themes. (Lardinois 1997: 215)

Thus, in a sense, gnōmai can be both an expression and a source of popular wisdom.77 Nor is this unique to Greek gnōmai: some presentday cultures show a similar ‘living tradition’ in the production of wisdom sayings.78 As for the ‘traditional formulae’ Lardinois mentions, there are certain linguistic and structural features of gnōmai (and proverbs) that make them stand out: these features include an absence of specific referents and deictics, tense usage (normally present or future),79 and certain recurring phrases (many Greek gnōmai begin with χρή, ‘it is necessary’, for example). More important for my purposes than differences between proverbs, Greek gnōmai, and other wisdom sayings, is what they have in common. This commonality resides first of all in the type of content that such expressions present: ‘wisdom, truths, morals, and traditional views’ (Mieder 2004: 4). Second, wisdom sayings are frequently metaphorical, and this holds for many Greek gnōmai (though certainly not

For a fuller discussion of ‘proverbiality’ (a vexed concept itself), see Arora 1984, Kirschenblatt-Gimblett 1973, Norrick 1985, 2014, and Mieder 2004: 4–20. 77 It is not for nothing that gnomic expressions are an important source for scholarship on ancient moral thinking and that such works as Dover 1974 refer to them frequently (see also Morgan 2007: ch. 2). Yet gnōmai are not necessarily acceptable to all addressees: Electra’s musings on wealth, status, and moral worth, for example, have been seen to undermine established moral thinking rather than to affirm it, and this has even been held to be the whole point of the play (see especially Arnott 1981: 179–81). In any case, Electra is very much concerned with the applicability of received notions of status and nobility, all of which must have been highly relevant at the time of the play’s first performance (see §4.2). 78 Lardinois 1997: 217. The origin of proverbs is well discussed more generally in Mieder 2014. 79 Greek also uses the gnomic aorist, as well as the potential optative construction. For a survey of structural features of (English) proverbs, see Mac Coinnigh 2014. 76

Introduction: Modern Linguistics and Euripides’ Electra

43

for all). Third and most importantly, they share the particular manner in which their meaning is ‘produced’. The very meaning of a proverb in normal usage is determined by context: without context, for example, we cannot know whether ‘A friend in need is a friend in deed’ is to be taken as an expression of gratitude (by the recipient of friendship), of loyalty (by the giver of friendship), or of something else. Similarly, in the case of the Greek gnōmai cited from Homer by Lardinois, we need to know that Menelaus is producing this gnōmē when he is trying to get Priam rather than one of his sons to take an oath (Il. 3.108): only then can we see how it functions as an argument in support of his view that oaths sworn by young men are unreliable. And Odysseus’ gnōmē (Od. 7.294) only gets its full meaning if we realize that it said in praise of Nausicaa, to her parents, because she is unusually wise for her age. General reflections are thus by nature an indirect* way of speaking (see §2.3.1), sometimes doubly indirect. If someone uses ‘A bird in the hand . . . ’ in a conversation, the literal meaning of his metaphorical expression (having to do with birds, hands, and bushes) is translated into some expression of a general truth (the preferability of something certain over something that is greater but uncertain),80 but it only derives its final, contextualized meaning from application to the particular situation at hand (e.g. advice to stay in one’s job). The last of these steps—from general meaning to specific application, which is known in paroemiology as the explanation*—is present in the interpretation of all forms of wisdom sayings and, inasmuch as modern paroemiology can shed light on this kind of production of meaning, it can be fruitfully applied to Greek gnōmai as well. Incidentally, in drama this indirectness of gnomic expressions makes it particularly suitable to be manipulated for purposes of dramatic irony. Because the ‘applied’ meaning of a general reflection is determined by the particular context in which it is used, it allows for different interpretations by different addressees.81 In the case of 80 Whether this first step of ‘translation’ is prominent, or even present at all, depends on various factors, including the extent to which a metaphorical expression has been conventionalized ‘(not everyone will fully activate concepts and images of buckets and flying legs when someone speaks of ‘kicking the bucket’). Figurative meaning and metaphor have been a favourite topic in cognitive linguistic research for some time, most influentially in Lakoff and Johnson 1980 and Lakoff 1987. For more resources, visit http://www.conceptualmetaphor.net (accessed 8 June 2016). 81 In short, proverbial expressions are ‘hetero‐situational, poly‐functional and poly‐semantic’ (Mieder 2004: 132).

44

Introduction: Modern Linguistics and Euripides’ Electra

generic expressions embedded in dramatic conversation, this context dependency also encompasses the ‘contextual’ gap between the characters (internal addressees) and the audience (external addressees). Thus, as Lardinois has argued for gnomic expressions in the ‘Deception Speech’ of Sophocles’ Ajax: The polysemous nature of such expressions increases the different layers of interpretation. In this case, as so often in tragedy, there is a discrepancy between the things the text‐internal addressees (Tecmessa and the chorus) know and those which the text‐external addressee (the Athenian audience) knows, creating the well‐known effect of dramatic irony. (Lardinois 2006: 217)

We shall see that in Electra, too, general reflections are used to manipulate the varying levels of knowledge between different characters and between the characters and the audience. Another aspect of the usage of gnomic speech relates closely to this notion of indirectness as well. It has been shown, for indirect utterances in general and for proverbs in particular, that such expressions may serve as a hedging technique—as a way in which a speaker can obscure or decrease the personal commitment s/he expresses towards his/her utterance. Gnomic utterances can allow a speaker to deny responsibility for the content of his/her utterance, exactly by focusing not so much on the particular as on the generic (regardless of the particular application it will always have). The purposes of utilizing such distancing techniques vary: a speaker may be trying to be polite or to avoid having to give his/her own opinion. In the case of proverbs in particular, Neal Norrick has demonstrated that they allow the speaker ‘to disguise his true feelings, to leave himself an escape route’, and that ‘speakers cite proverbs to avoid personal commitment’ (Norrick 1985: 27).82 With this last point I have already entered the social, interactional dimensions of proverbial speech,83 and it is in this area that modern paroemiology will prove to be particularly helpful for my discussion of Electra. As well as allowing an avoidance of personal commitment to an utterance, proverbs also serve as an invocation of authority:

82

See also Taylor 1962: 169 and Green 1975. For the social context of proverb usage, see Mieder 2004: 133–4 (with further references). The pioneering study in this field is Arewa and Dundes 1964; an exemplary study is Briggs 1985. 83

Introduction: Modern Linguistics and Euripides’ Electra

45

in uttering a proverb a speaker quotes a traditional item of the folklore of the community. As items quoted from this stock, proverbs carry the force of time‐tested wisdom, and the speaker can draw on this authority. This correlates strongly with the observation . . . that proverbs most commonly have evaluative function and a didactic tone in free conversation. Essentially the same effect can be achieved by citing scripture, famous authors or recognized authorities in the relevant field. In each case the speaker adds authority and credibility to his utterance by identifying himself with traditional wisdom, beliefs and prejudices of the community at large. (Norrick 1985: 28)

Although Greek gnōmai are not so much a stock of traditional items of folklore or ‘time‐tested’ per se, they nonetheless may be said to make a similar appeal to generally accepted (or acceptable) wisdom. Thus they, too, serve as a method of invoking an authority outside, and more widespread than, that of the speaker alone. Readers of Euripidean tragedy will also have little trouble relating the description ‘most commonly have evaluative function and a didactic tone in free conversation’ to gnōmai as they find them used by Euripides’ characters. The evaluative and didactic tone of generalizing expressions also entails that their use in conversational contexts carries with it some complications. In the negotiations of face* that constantly go on in conversations, it can be damaging to make an evaluative utterance about an addressee, or to adopt a didactic tone. Proverbial expressions thus prove to be a particularly complex and subtle instrument for negotiating speaker–addressee relations in a conversation: they may constitute a threat to someone’s face while at the same time softening that threat. For example, when I say ‘No man is an island’ to a friend who I believe is unreasonably refusing my help, my statement can well be considered to threaten my friend’s face (I am berating him for his behaviour), yet I have made some effort to soften the blow of my criticism by avoiding a more outright expression of disapproval. Whether my expression is considered to be polite depends, of course, on the nature of the conversation that I was having with my friend, our relative social status, and so forth.84 These social complexities of proverb usage were analysed by Peter Seitel (1969). Seitel categorized proverbial utterances by their applicability to the speaker (first-person sayings*), the addressee (secondperson sayings*), or a person other than the speaker or addressee 84

For proverbs as a mechanism of politeness, see Obeng 1996.

46

Introduction: Modern Linguistics and Euripides’ Electra

(third-person sayings*)—these designations are independent of the use of grammatical person and number in the ‘proverb text’. My ‘No man is an island’ example, in this classification, would be a secondperson proverb (applied to the addressee, my friend), even though the verb in the proverb text is in the third person. Singular or plural persons are possible (depending on the number of people the proverb is applied to), and proverbs may apply to more than one person at one time, directly or indirectly. As Seitel observed in his empirical research, there are important differences between the conversational uses of first, second- and third-person sayings: second-person sayings, applying to the addressee, tend to be restricted to speakers in positions of relative authority, while first and third-person sayings are employed more freely. Lardinois, who applied Seitel’s model to Homeric gnōmai, confirmed that the trends identified by Seitel hold true there as well, but added that second-person sayings are found to be used among social equals in the Iliad when they serve as insults and taunts, or in the case of compliments and praise. Though neither Seitel nor Lardinois fully explored the reasons for this distribution, it is, I hope, clear from my preceding discussion why their findings should be the way they were: second-person proverbs, when they serve as an evaluative or didactic statement, constitute a possible threat to the addressee’s face* (while also softening that threat). Whether any facework* is required to compensate for such a threat, and how much, depends on the relative power and authority of the speaker and of the addressee and on the nature of the conversation they find themselves in. In short, the use of general reflections in conversation can serve a complex function in the intricate negotiation of face and power between speakers and addressees. Another brief example from Electra may be instructive here. Shortly before the moments between Clytemnestra and Electra treated earlier (§2.2), the queen makes the following remark to her daughter (1102–5): ὦ παῖ, πέφυκας πατέρα σὸν στέργειν ἀεί. ἔστιν δὲ καὶ τόδ᾽· οἱ μέν εἰσιν ἀρσένων οἱ δ᾽ αὖ φιλοῦσι μητέρας μᾶλλον πατρός. συγγνώσομαί σοι· My child, it is in your nature to keep feeling affection for your father. This, too, is something that happens: some belong to the male side, others are more devoted to their mother than to their father. I’ll be patient with you.

Introduction: Modern Linguistics and Euripides’ Electra

47

Clytemnestra’s general sentiment, expressed in the sentence οἱ μέν . . . πατρός (1103–4), which the introductory phrase ἔστιν δὲ καὶ τόδ᾽ signposts as a generalization, is applied to the particular case of Electra; thus we are dealing with a second-person saying*. In the context (Electra has just completed an angry speech directed at her mother, 1060–96), Clytemnestra’s lines seem designed to deflect some of her daughter’s criticisms by portraying them merely as the result of Electra’s affection for her father rather than as reflecting an objective assessment of the case at hand. Clytemnestra’s gnōmē serves to portray Electra’s emotions as a typical trait of some children. The example involves many social intricacies: on the one hand, Clytemnestra’s gnōmē seems to strike a conciliatory note, since it suggests that she does not blame her daughter personally for her bias—that is merely something to be expected from some children. Her seemingly placatory tone is made explicit in the following line, συγγνώσομαί σοι (1105). On the other hand, however, the fact that Clytemnestra can use a secondperson saying (which is not obviously meant to insult, criticize, or admonish) is suggestive of the power differential that she believes still exists between herself and her daughter: equals and people of inferior status are less likely to use second-person gnōmai, unless they mean to be hurtful (in which case the constraints of status are ignored). Moreover, the very fact that Clytemnestra uses a generalization to make her point seems to preclude debate on it: it is not only Clytemnestra who thinks that Electra is biased; commonly accepted wisdom says so. We may thus detect, perhaps, a note of impatience and superiority in Clytemnestra’s words. As I will argue in chapter II (§5.3.1, 5.3.5), this is a prevailing feature of this scene. Throughout this book I hope to show that the aspects of gnomic language described above are vital for a full interpretation of the gnōmai used by the characters of Electra. As a related point, I hope to show that the textual suspicion that many of these general reflections have elicited is unfounded.

2.6. Narrative and argumentative texts: discourse cohesion All the approaches discussed thus far have focused on aspects of language in interaction, aspects that prove important especially in dialogical sections of plays. But, as I have noted (§2.1), some parts of Greek tragedy do not lend themselves well to analysis through such

48

Introduction: Modern Linguistics and Euripides’ Electra

methods. Longer rhēseis in a narrative or argumentative mode should still, I hasten to add, very much be seen as situated in their communicative context (and my discussion of gender and gnomic language will prove relevant to these longer speeches as well). But the internal structure of such discourses has less to do with the interactive relationship with other characters and more with the organization of thought and argument (in the case of argumentative speeches) or of story and plot (in the case of narrative passages). An analysis of internal structure requires that we consider a number of linguistic devices that lend coherence to a text. Recent work in ancient Greek linguistics has devoted much attention to the cohesive properties of many of these devices and instruments: particles, tense usage, pronouns, word order, sentence complexity, and so on.85 Much of what I want to say about these features is best reserved for individual discussions in the chapters to follow, but I would like to note a number of general points of recurring significance here. First, a few words about particles: it is now generally recognized that these devices should be described in terms of ‘functional’ rather than ‘referential’ meaning: particles ‘do’ rather than ‘mean’, if the latter is taken as referring to some entity, property, or relation in the world described by the text. Such functions occupy different levels. Some particles primarily determine relationships between different parts of a discourse (these are traditionally known as ‘connective’ particles, e.g. καί, δέ, γάρ); others (‘modal’ or ‘interactional’, e.g. δή, τοι, μήν) deal more with the relationship between interactants (speaker, addressee) and between them and the content of an utterance.86 A function at both these levels is possible as well—and there are often diachronic shifts of function. The second category of particles—the ‘modal’ kind—is interesting in the light of the nuances of conversational interaction I have discussed earlier in this Introduction; and I will comment on their use

85 See in general the contributions in Bakker and Wakker 2009, which cover all these topics, as well as the relevant articles in Giannakis 2014 (e.g. Drummen 2014a, 2014b, Perdicoyianni-Paleologou 2014, and Wakker 2014). Further references may be found there and will be given in this book where relevant. 86 In a study of Latin particles, Kroon (1995) has influentially distinguished between a ‘representational’, a ‘presentational’, and an ‘interactional’ level of discourse. Most particles function either at the presentational or at the interactional level. For a direct application of Kroon’s model to Greek particles, see Wakker 1997a. Revuelta 2014 is a helpful encyclopedia article.

Introduction: Modern Linguistics and Euripides’ Electra

49

where appropriate. The first—the discourse-structuring kind— deserves some further comment here, especially with regard to the ‘scope’ of these particles. One of the crucial gains made in recent scholarship87 is the realization that connective particles do not necessarily connect one sentence to another, but rather serve to structure a discourse into constituent ‘acts’ and ‘moves’,88 which may stretch over a series of sentences, even whole parts of speeches. Acts and moves form a hierarchy: a move may consist of a main act and a subsidiary act, and different particles are used to indicate this structure. To give a few examples: Bakker (1993a) has shown that the particle δέ marks a (neutral) boundary between sections of discourse; the particle γάρ has been shown to mark ‘subsidiary’ segments of discourse (e.g. Sicking 1993, Slings 1997a) and may introduce entire embedded narratives (de Jong 1997), while οὖν has been shown to introduce newly relevant sections of a discourse of whatever length (rather than, for instance, merely sentences that serve as an inferential conclusion: see my discussion of οὖν at §2.2.2 in ch. III). These discourse-structuring functions are important for an analysis of argumentative discourse, in order to follow the precise development of an argument; and for an analysis of narrative discourse, in order to see how narrative segments interrelate, to find breaks in the flow of a narrative, and to trace the speed and ‘mode’ of narration. A sequence of sentences introduced by δέ is very different in terms of narrative style from a similar sequence of sentences introduced by καί.89 When it comes to the style and speed of a narrative—issues that have been examined especially by narratologists90—another crucial factor is the use of verbal tenses. The basic narrative tenses are the

87 Particularly vis-à-vis the treatment of these particles in Denniston 1954 and 1952. The former work, Greek Particles, justifiably remains a classic but cannot be taken today as the definitive treatment (see Rijksbaron 1997a). 88 For this terminology, see e.g. Filliettaz and Roulet 2002. An earlier version of their model is followed by Kroon (1995). 89 On ‘narrative modes’, see Allan 2007, 2009, Kroon 2007, and Adema 2007. The linguistic and narratological studies underlying these works include Fleischman 1990, Chafe 1994, Roulet et al. 2001, and Smith 2003. 90 There has been a confluence in recent work between linguistics and narratology—fields that were compatible from the start—in that more attention has been paid to the linguistic phenomena that create certain effects in narrative texts. For a general discussion, see e.g. Mey 2001: ch. 9, and, for good examples of this kind of research, see the works mentioned in the previous note, along with de Jong 1997, 2007 and Allan et al. 2014.

50

Introduction: Modern Linguistics and Euripides’ Electra

imperfect (or pluperfect), aorist and the historic present, and the distribution of forms in these various tenses reveals a great deal about the way a narrative is structured and paced. Again, I reserve most of what I have to say on these issues for the chapters to follow (particularly chapters I, V, and VI), but I do wish to make one brief comment here on the function of the historic present (which plays a significant role in the two longest narrative sections in Electra, the prologue and the messenger speech). Recent research on the historic present in Greek91 has attempted to give a more precise description of the value of this phenomenon than the rather nebulous term ‘vividness’, which still has great currency in classical scholarship. According to this research, the primary function of the historic present is to mark decisive, pivotal events in a narrative—events that unalterably change the state of the narrated world.92 This function is not an objective one: we do not find every ‘objectively crucial’ event in a narrative marked by the historic present. Rather the historic present is a device that can be exploited by a narrator for his/her own rhetorical purposes: it is the narrator who chooses to mark certain events as more significant than others, highlighting those he considers to have a high degree of ‘tellability’.93 The subjective purposes of a narrator may sometimes even clash with what would objectively be considered a crucial event (for a striking example of this in the messenger speech of Electra, see ch. V, §3.4). Issues such as tense usage, particles, and other points relating to cohesion in texts will be of some significance throughout the following chapters but will come to the fore especially in chapters V and VI, where I discuss the messenger speech and the rhēseis of the agōn, respectively.

3. TEXTUAL CRITICISM It remains for me to very briefly outline my principles in practicing a less ‘modern’ type of analysis of Greek tragic texts: determining what 91 My description is based mostly on Rijksbaron 2002: §7, 2006, and Allan 2007, 2009. Some other recent contributions are Sicking and Stork 1997, Lamers and Rademaker 2007, Lallot et al. 2011, Boter 2012, Nijk 2013, and Rijksbaron 2015. 92 This use is sometimes extended into a ‘punctuating’ use that marks the start of a new episode in a narrative; see Rijksbaron 2002: §7.3 and Allan 2009: 193 with n. 57. 93 I adopt this term from Allan 2009: 195.

Introduction: Modern Linguistics and Euripides’ Electra

51

those texts (should) actually say. No work that foregrounds the language of a classical text can ignore problems of textual transmission, and the Electra has its share of difficult cruces. My approach has been not to avoid these, yet not to seek them out either. In other words, in what follows I will practise detailed textual criticism when my reading hits upon places where the constitution of the text is an issue of debate, but only when what is read actually has some bearing upon my analysis of the play. Thus, for example, I will not treat the famous crux in the first verse of the play in detail (see ch. I, §2 n. 8), but I will discuss the textual problems in 311–13 in depth (ch. II, §3.2); similarly, I have no desire to wade too deep into the quicksand surrounding the authenticity of the Aeschylean tokens in the recognition scene (see §4.3), but I will argue against treating too much of 367–400 as an interpolation (see ch. III, §2.2.3). Discussions of textual issues that are relevant but not crucial for interpretation are normally relegated to the footnotes. A pair of cases is given special attention, to the extent that the passages in question are treated in a separate chapter (ch. IV). The problems in 671–93 and 959–87 concern not so much what is read as which characters should be assigned which lines to perform. I believe that the linguistic approaches described in this chapter can shed some light on these notorious issues of line-to-speaker attribution, and I hope to show that, in this way too, these approaches can help us to interpret the play (the question ‘who says what’ is of more than simply textual–critical interest, after all). My treatment of these cruces (and some others) is also designed, more generally, to contribute to our understanding of the status of the manuscript evidence in Euripides’ so-called ‘alphabetic’ plays, particularly when it comes to the issue of line-to-speaker attribution.

4. A VIEW OF THE PLAY ‘Controversy about the evaluation of Euripides’ Electra still rages as fiercely as empty tigers or the roaring sea’, wrote Arnott in 1981 (p. 179)—a statement that is hardly less true today.94 To conclude 94 This section deliberately shares its title with a similar, more extensive one in Cropp’s excellent commentary (Cropp 2013: 1–12): comparison will show that it shares more than its title. My debt to Cropp throughout this book is sizeable.

52

Introduction: Modern Linguistics and Euripides’ Electra

my introduction, I provide here a very brief overview of some of the interpretative issues that have given rise to such controversy, and a first indication of my own views concerning the issues involved.95

4.1. Characters and characterization 4.1.1. Conceptualization; characterization through style Given that ‘character’ is one of the terms in my book’s title, questions of conceptualization inevitably arise. Character and characterization are among the most fundamental (and vexed) topics in literary criticism generally; and they have a particularly complex and chequered history in classical scholarship.96 Seen against this background, what I will have to say about such conceptual issues here may seem surprisingly brief. There are two reasons for this: first, I will deal with this topic more fully in another publication, and I can refer to several other excellent discussions;97 second, in my view, the best way to talk about Euripidean characterization is to discuss examples of his practice, and much of this book can in fact be seen as a series of case studies in Euripidean characterization.98 Rather than rehashing here the many arguments swirling around characterization in tragedy, then, I will observe merely that scholars appear to have abandoned—rightly—the extreme positions of the debate for a more measured approach. Neither the notion that tragedy represents ‘real people’, with lives that extend beyond what we 95 As with any of Euripides’ plays, especially the later ones, the debate surrounding Electra is deeply intertwined with controversy over how to understand Euripides’ works more generally. On such general issues, see Mastronarde 2010, and for a history of the scholarship on them, see Michelini 1987: ch. 1, though this does not go past the early 1980s, and has lacunae: see Diggle’s (1989) critical review. 96 For some overviews of the debate in literary criticism generally, see Margolin 2007, Eder et al. 2010, and Jannidis 2013 (each with further references). Specifically on drama and with a focus, like mine, on characterization through language is Culpeper 2001. Culpeper’s work on this topic has been formative for my own thinking (see also Culpeper 2002, 2009). 97 Van Emde Boas in press-b. Among the many works that discuss characterization in tragedy, I have found the following particularly useful: Easterling 1990, Mossman 1995: ch. 4, Allan 2000: ch. 3, Seidensticker 2008, Thumiger 2007, 2013, Budelmann and Easterling 2010, Mastronarde 2010: chs 6–8, and Rutherford 2012: ch. 7 (This list excludes work predating Pelling 1990a; see also n. 102). 98 That is to say, his practice in one of his plays: as has often been seen, Euripides’ practice is not uniform.

Introduction: Modern Linguistics and Euripides’ Electra

53

see on stage,99 nor the idea that the tragedians were not interested in characterization at all100 has much going for it today. Instead more balanced views have emerged, which suggest that characters on the Greek stage will at least have some degree of ‘human intelligibility’,101 and which—crucially—show an awareness that our terms ‘character’, ‘personality’, ‘individuality’, cannot be unproblematically translated into equivalent Greek concepts.102 A question much less discussed, but vital to my own approach is whether the tragedians used linguistic or stylistic means to characterize their figures, in other words whether speakers in Greek tragedy are characterized by their own language. This is an understudied subject in Greek literature tout court: as Tompkins rightly noted almost 40 years ago, the ‘widespread reluctance to consider character important in Greek literature’ led to a ‘consequent failure to observe characteristic language’ (Tompkins 1972: 182).103 Fraenkel indeed explicitly denied the existence of any linguistic characterization in Euripides: ‘In Euripides, characterization through language does not

99 A frequent (if not entirely fair) target for scholars who criticize this ‘humanizing’ position is Vickers, who writes, for example: ‘Tragedy is about people, and what they do to each other . . . The plays translate the clash of will and motive into forms, which although obeying complex literary conventions, still represent human actions, and convey them with intensity, if we are prepared to accept the conventions’ (Vickers 1973: 3). 100 This ‘dehumanizing’ view was first expounded in relation to Sophocles by Tycho von Wilamowitz (1917) and is sometimes referred to as ‘Tychoism’; so far as Euripides goes, the most influential restatements are those of Zürcher (1947) and Dale (1954). 101 Easterling’s (1973) term, influentially criticized by Gould (1978). 102 Most (if not all) of the essays in Pelling 1990a represent a definitive shift away from the extremes mentioned. For discussion of the Greek concepts of character and personality, see Gill 1990, Halliwell 1990, Pelling 1990b, Gill 1996, Blundell 2002: 53–67 (on Plato), Worman 2002, Thumiger 2007, and De Temmerman 2014: Introduction (on the novel). 103 The exception is Homer’s Achilles, whose use of language has received more attention than that of any other character in Greek literature, from Adam Parry’s (1956) famous ‘The Language of Achilles’ onwards (see e.g. ch. 4 in Martin 1989 for a survey of pre-1990 scholarship). Noteworthy in particular is the debate in the journal Language around 1980: initial article by Friedrich and Redfield (1978), reaction by Messing (1981), response Friedrich and Redfield (1981). Without Achilles, the list of publications dealing principally with this topic is much shorter. It includes Tompkins (1972, 1993), Katsouris (1975), and Kitzinger (1976); of course, numerous further discussions may be found in individual articles, monographs, and especially commentaries (good examples are Griffith 1999 and Mossman 2011). For Latin, see e.g. Abbott (1907) and Maltby (1979).

54

Introduction: Modern Linguistics and Euripides’ Electra

occur, although it does in Aeschylus and Sophocles’ (Fraenkel 1977: 36).104 It will be my contention that Euripides does adapt his language to the individuals who speak it and that close examination of the speaking styles used by the characters in Electra (through the linguistic approaches discussed in §2) will in fact reveal a great deal about their characterization.

4.1.2. Electra and Orestes Electra and Orestes are the main characters of interest in the play, and a brief preview of their treatment is in order. It is safe to say that the modern consensus on both siblings, in spite of a few dissenting voices (too often ignored), is overwhelmingly negative. The entry on our play in The Encyclopedia of Greek Tragedy presents things thus: Electra: a monomaniacal harridan, victim of sexual jealousy of her mother, who feels more sorry for herself than her dead father or exiled brother and bullies her reluctant brother into committing matricide . . . Orestes: a snobbish youth, uncertain of his identity, undecided about his mission, too cowardly to enter the city and face his father’s killers, in need of his sister’s help and stronger personality. (Luschnig 2013: 381)

Such extremes of opprobrious criticism are to my mind misguided. To begin with Electra herself (the central figure of the play, and the most controversial one), I will argue in chapter II that the early scenes of the play present us not with a woman wallowing in self-inflicted misery (as many scholars would have her), but with a figure torn from her proper surroundings and placed in incompatible, impossible circumstances. She is a ‘heroic’ figure trapped in an entirely mundane universe, an Electra who wants to play her traditional—and legitimate—part but is unable to do so because of the very untraditional setting she finds herself in. This adds an interesting layer to her motivation for the matricide: her distress at her personal displacement (and at Orestes’ exile) fuses with her desire to avenge Agamemnon’s death, and this complex of issues engenders in Electra the vindictive hatred that inspires her—and, through her, Orestes—to matricide. As the play progresses and the matricide approaches, the balance between these motivations seems to shift somewhat, so that, 104 In Euripide non si trova caratterizzazione per mezzo della lingua, ma in Eschilo e Sofocle c’è.

Introduction: Modern Linguistics and Euripides’ Electra

55

in the agōn between Electra and her mother, the issue of her own and her brother’s mistreatment is the focus of her speech (see chapter VI). This calibrates audience sympathy and casts another troubling light on the matricide. Electra’s motivations and behaviour are also coloured greatly by her gender, which Euripides foregrounds in varied and subtle ways (to be discussed throughout chapter II). Orestes, too, has been much maligned. Two aspects of his behaviour have irked scholars: the long delay in the revelation of his true identity and the way in which he kills Aegisthus in the messenger speech. The ‘ugly’ aspects of the latter scene, I will argue in chapter V, are not designed to convey a simple moral, a good-versus-bad condemnation of Orestes and rehabilitation of Aegisthus (who is certainly not as sympathetic a character as some scholars suggest); nor is the messenger speech, with its exciting and suspenseful narrative, really the place for such character evaluations in the first place. Rather the troubling undertones of the tyrannicide foreshadow the much more serious problems of the matricide (which has some important thematic parallels to the messenger narrative: an unsuspecting victim attending a sacrifice, being sacrificed him-/herself). In consequence, the messenger scene represents one stage in a gradual shift in tone, from the more jubilant atmosphere surrounding the tyrannicide to the deeply tragic horror of the matricide. As for Orestes’ delayed recognition, I cannot improve on Cropp’s discussion: This prolonged concealment is in part a reflection of the Odyssey, where Odysseus reveals himself to his son only when they are alone, withholds his identity from his wife even after he is assured of her loyalty, and later does the same to his aged father . . . Euripides uses Orestes’ disguise similarly to create dramatic suspense and surprise, while also reworking motifs from the equivalent scenes of Aeschylus and Sophocles. The delay also fits his own design, for the play’s early scenes are not so much about Orestes as about Electra. The portrayal of her grief and isolation requires a postponement of the recognition . . . Meanwhile the prolonged disguise limits the characterization of Orestes . . . (Cropp 2013: 7)105

105 Schuren (2014: 84), following Lloyd (1986: 12–13), adds the important point that the presence of a ‘stranger’ to respond sympathetically to Electra provides another perspective on her situation—one that guides audience response to it. Schuren’s discussion gives a helpful overview of the debate on Orestes’ secrecy (with further references). See also Mastronarde 2010: 287–8.

56

Introduction: Modern Linguistics and Euripides’ Electra

In chapter III, I will focus on the ingenious ways in which Euripides ‘limits the characterization of Orestes’ in the early scenes by modulating his linguistic behaviour, and how he creates the effects of ‘dramatic suspense and surprise’ that Cropp mentions.

4.2. Themes and motifs The language of Electra features some recurring visual and thematic motifs, which calibrate the moral and emotional tone of the play. For my purposes, identifying these themes is important because they can help us distinguish between different ‘layers’ of meaning and communication, between characters, and between playwright and audience (introduced in §1.1). The visual and thematic strands that run through the play are principally to be seen as part of the latter ‘message’, and it may be counterproductive to read too much psychological subtlety into moments that reinforce that message. We should perhaps not dwell too long, for example, on the question why the Messenger uses a simile drawn from the athletic sphere (824–5), except to reflect on the prominent role of this motif in a large part of the play. Similarly, when Denniston comments, on 959–66, that Orestes ‘could not use so brutally unequivocal a phrase as σφαγῆς πάροιθε’, this may be making too much of a phrase that seems designed to underline the motifs of hunting and sacrifice at a crucial point in the play (of course, Euripides is very capable of crafting such moments to have local as well as broader thematic significance).106 The examples just given—hunting, sacrifice and athletics—are the dominant metaphorical motifs in the play: the first two of these, in particular, are pervasive in Greek tragedy in general and in the Oresteia tradition in particular,107 and they serve a prominent role surrounding the two killings. It is significant that the athletic imagery, with its triumphant overtones, disappears entirely after Orestes

106

For my own reading of 959–66, see ch. IV, §3.1. For imagery in Euripides, see Barlow 1986b; in tragedy more generally, see e.g. Porter 1986 (with further references). For sacrifice in tragedy and in the Oresteia myth of the house of Atreus (including the Electra), see Zeitlin 1965, Burkert 1966, 1983, Lebeck 1971, Mirto 1980, Porter 1990, and Henrichs 2000. For hunting, see LloydJones 1972, Vernant and Vidal-Naquet 1981, Burkert 1983, and Vidal-Naquet 1986. 107

Introduction: Modern Linguistics and Euripides’ Electra

57

comes on stage with Aegisthus’ corpse: this is part of a marked shift of tone surrounding the matricide.108 Thematic material of a somewhat different nature is prominent in the earlier scenes of the play. Euripides explores issues of wealth, nobility, and character and has the Peasant, Orestes, and the Old Man all expound a ‘new ethic’, in which good conduct is a more reliable test of worth than external signs or hereditary status. Such issues were surely highly relevant in Euripides’ contemporary democratic society; in the context of the play only, these new standards seem to underline the gulf between the traditional (‘heroic’) role that Electra wants to adopt for herself and her incompatible new surroundings; they are also indicative of Euripides’ innovative approach to tradition, more on which presently.

4.3. Tradition (and the recognition scene) I have described Electra as someone ‘who wants to play her traditional part’ but cannot do so because of the ‘untraditional’ nature of her circumstances (see §4.1.2). This is typical of Euripides’ treatment of tradition in this play (and in his later oeuvre more generally),109 particularly of his engagement with Aeschylus’ Choephori (for Sophocles’ Electra, see §4.4). Euripides’ approach in this respect is complex and highly innovative, and Electra draws attention to its novelty from the start through the unexpected setting the protagonist is placed in. I do not think such features are, at their core, ‘polemical’ towards tradition, as various scholars have seen them: Euripides’ modifications and reinterpretations of Aeschylus (and others) advertise the creativity inherent in his own approach rather than ‘flaws’ in his predecessors. 108 On the athletic imagery, see Arnott 1981: 187–9, Kubo 1967: 23–4, Myrick 1994: 138–41, and Swift 2010: 156–72. Swift’s discussion helpfully draws out the many points of contact with epinician lyric. Imagery related to the sun, light, and related elements also disappears entirely from the play just before the matricide (see Cropp 2013: 18). 109 On the play with tradition in this play, see e.g. Goldhill 1986a, Michelini 1987, Goff 1999–2000 (as well as various other essays in Sansone et al. 1999–2000). Goldhill’s and Michelini’s books cover Euripides’ oeuvre more generally, but have sections and chapters on Electra. Another valuable general account of Euripides’ use of tradition is Allan 2000: ch. 1 (on Andromache). On tragedy more generally, see e.g. Anderson 2005.

58

Introduction: Modern Linguistics and Euripides’ Electra

It is in this way, too, that I think the recognition scene, with its much discussed ‘criticisms’ of the tokens in Choephori, should be interpreted (rather than as ‘polemic’, ‘parody’, or ‘satire’).110 To make my position clear from the outset, I take the entire scene to be genuine and do not think that any convincing structural or linguistic argument has been offered to cast suspicion on it;111 thus the debate surrounding its authenticity is not as relevant for my purposes as many other textual questions in which such arguments do have a role to play (other kinds of objections against the scene are inherently subjective and, to my mind, less reliable in determining authenticity).112 The scene should be seen as a deeply self-conscious statement of Euripides’ innovative approach, the most explicit example of such playful intertextuality in his extant work.113 But, like other similar cases, the scene has local significance apart from its ‘look how original!’ function. As I will discuss in chapter II, the scene is as much about Electra’s lack of insight as it is about playing with Aeschylus’ Choephori, and deleting the scene would actually spoil much of the effect.

4.4. The roads not taken . . . Finally, I should mention a few things that I will not discuss in the chapters to follow. A first general point to make is that this is a reading of Euripides’ Electra (and one with a specific bent at that), not a line-by-line commentary, so that some scenes will receive more attention than others, while some will receive none at all. The most glaring omission is that of the choral odes, but the absence of detailed 110

For an assessment of the different terms applied to Euripides’ reaction to Aeschylus in the recognition scene, see Davies 1998: 401–2. 111 See Basta Donzelli 1980b and Cropp on 518–44 (‘Text and language’) for a discussion of the ‘objective’ arguments against authenticity—which both see, quite rightly, as very weak. 112 The debate over authenticity began in earnest with a note in Appendix D of Fraenkel’s commentary on the Agamemnon (Fraenkel was following Mau 1877) and was taken up in Paduano 1970, Bond 1974, Ronnet 1975, Bain 1977b, Basta Donzelli 1980b, West 1980, Tarkow 1981, Halporn 1983, Kovacs 1989, Goff 1991, Jouanna 1997, Davies 1998, Müller 2000, Gallagher 2003—as well as commentaries ad loc., of course. These are only the most directly relevant works. 113 But not, in the end, fundamentally different from other examples of the same technique (which have sometimes caused less suspicion). See Mastronarde ad loc. on the similar case at Phoen. 751–2 and the discussion of several other cases in Davies 1998.

Introduction: Modern Linguistics and Euripides’ Electra

59

discussions of other parts of the play will also be noticed (Orestes’ speech at 82–111 and Castor’s at 1238–91, for example; the recognition scene will also receive relatively limited attention by comparison to the preoccupation with the scene in scholarship on the play). The omission of the choral odes deserves some additional comment. Apart from limitations of space and scope, it will be seen that the type of linguistic analysis that I aim to perform is inherently less applicable to the odes. Without a clear communicative setting, the odes resist interpretation through the methodological approaches I aim to apply to the rest of the play. For interpretation of the thematic and emotional contribution the odes make to the play (they are certainly much more than the irrelevant embolima they are sometimes made out to be), I would refer to a number of articles discussing the individual odes,114 as well as to Cropp’s commentary. Lastly, one of the perennial issues for scholars dealing with Euripides’ Electra is the chronological priority between it and Sophocles’ play of the same name. Although I suspect that Sophocles’ play is earlier (for arguments, see Cropp 2013: 27–8), I do not think the question can be settled unless some new evidence is presented,115 and I would prefer to analyse Euripides’ play in its own right. Sophocles’ Electra can be a helpful point of comparison, but this kind of approach carries with it the risk that one play is seen only in relation to the other. In any case, from the point of view of linguistic analysis, the rest of Euripides’ oeuvre is a more helpful comparandum than any Sophoclean tragedy.

114 See Walsh 1977, Rosivach 1978, King 1980, Morwood 1981, Basta Donzelli 1992, 1995b, Jouanna 1998, Csapo 2009, Willink 2009, and Allan 2013a. 115 Not much that is new has been said about the issue recently, and the history of the question (with bibliography) in Matthiessen 1964: 81–8 has not really been superseded.

I Rustic Language The Peasant

μορφῇ μὲν οὐκ εὐωπός, ἀνδρεῖος δ᾽ ἀνήρ, ὀλιγάκις ἄστυ κἀγορᾶς χραίνων κύκλον, αὐτουργός (οἵπερ καὶ μόνοι σῴζουσι γῆν) ξυνετὸς δέ, χωρεῖν ὁμόσε τοῖς λόγοις θέλων, ἀκέραιος, ἀνεπίπληκτον ἠσκηκὼς βίον. —Euripides, Orestes 918–221

1. INTRODUCTION In a ‘student‑friendly’ study guide for first-time readers of Euripides’ Electra, the first topic suggested for further discussion is this: Examine the speeches of the character of the Farmer, both in his prologue to the audience and his words to Electra, his wife. What clues to his character do we find there? How does he speak of his wife, and to her? Why might Euripides have chosen to portray this humble figure, new to a legend concerning the troubles of a royal house? (Underiner 2005: 69)

Some of these questions, apparently so immediately relevant to the reader of the Electra, have not received the attention they deserve in ‘Not a man handsome to look at, but brave, one who rarely had anything to do with the city or the market circle, a man who farmed with his own hands, the sort who alone keep the land from destruction, yet clever enough to grapple in argument when he wanted: having lived a life of integrity, above reproach.’ 1

Rustic Language: The Peasant

61

scholarship on the play. This is not to say that the Peasant is generally ignored altogether: he has in fact been the subject of several studies.2 What we might take issue with, however, is that scholars have sometimes answered the third question quoted above before sufficiently considering the first two—that they have come, in other words, to sweeping conclusions about the Peasant’s function in the play as a whole, without relating those conclusions closely to his ipsissima verba. Consequently the Peasant is sometimes explicitly portrayed as little more than a dramatic device, be it as a moral foil for Electra and Orestes (e.g. do Deserto 1994),3 as a detail illustrating Electra’s squalid living circumstances (e.g. Gellie 1981: 2), or more generally as part of Euripides’ project of ‘deglamorization’ of the Electra myth (e.g. Arnott 1981: 179–81). Of course, Euripides’ remarkable choices in shaping the structure and plot of the play impose, in a way, a need for some dramatic innovations: the Peasant may thus be all those things. But to view him exclusively in this light, as merely an instrument employed by the poet to perform the dramaturgical function required of him, is in my view an oversimplification of a subtle piece of Euripidean artistry. There is, I suggest, more to the Peasant than that: he is imbued with a distinct character that springs very much to life in the early scenes of the play. In this first chapter I wish to take a closer look at that character, specifically as it is manifested through the Peasant’s own words in the opening monologue (1–53) and in his three moments of interaction with Electra (54–81, 341–63, 404–31). This stated aim at once reveals where I stand on two related questions, which will occupy me throughout the present chapter and the subsequent two: first, whether Euripides was concerned with (psychological) characterization; and, second, in case he was, whether or not that characterization is expressed, at least in part, through stylistic means. I have argued in the Introduction (§4.1.1) for

2 Basta Donzelli (1978: ch. 7), who herself could rightly complain about the ‘rather modest attention’ paid to this character (227, ‘un’attenzione piuttosto modesta’), offers an extensive treatment of his sociopolitical and ethical dimensions and the historical inferences that may be drawn from them. Michelini (1987: 194–9) devotes a substantial section of her chapter on Electra to the character. See also do Deserto 1994 and Yoon 2012: 99–105 (with some further references). 3 Cropp suggests that his main dramatic purpose in the play is not so much to discredit Electra and her brother as it is to highlight the ethical complexities of the situation they find themselves in (Cropp 1988: xl, in slightly weakened form at 2013: 14). For a more negative view of the Peasant’s character, see Schottländer 1982.

62

Rustic Language: The Peasant

a balanced, but on the whole positive answer to these questions: I hope to show that Euripides did indeed tailor his speeches to the characters speaking them—including the very first character to come on stage.

2. A PEASANT’S TALE (1–53) When attempting to read the opening monologue of Electra as a stylistically individualized text, one is confronted with an immediate difficulty, for the Peasant here assumes the additional role of prologizomenos. Does the highly formalized nature of Euripidean prologues, so effectively parodied by Aristophanes (Ar. Ran. 1175–1248), leave any room for individual traits? Erbse (1984), in what has become the locus classicus for insights on the Euripidean prologue, presents an even bigger problem to anyone looking for individual characterization in the openings to Euripides’ plays: he depicts prologue speakers as objective reporters, truthfully narrating (only) the relevant facts in chronological order.4 It is, then, hard to see how we could find any characterization in passages so frigid. Though Erbse’s account is widely corroborated by other scholars, there are serious problems with such a categorical view of the prologues. Many Euripidean prologue characters can hardly be called objective, and even their ‘factual’ expositions are not as clear-cut as they seem.5 To this I would add that generalizations that take the opening monologues of Euripides’ plays as uniform constructs

4 ‘The statements in the prologue are not only meaningful, they are also true. Which is to say: they are absolutely reliable in the sense that the audience gets from them the requisite background information of the play’ (Die Aussagen des Prologs sind . . . nicht nur sinnvoll, sie sind auch wahr. Das besagt: Sie sind unbedingt zuverlässig in dem Sinne, daß der Zuschauer in ihnen die zureichenden Bedingungen des Spieles erfährt: Erbse 1984: 289). Erbse (158) goes so far as to claim that Electra herself could never have delivered the opening monologue in our play, because she wouldn’t be sufficiently objective. 5 See e.g. Hamilton 1974, 1978, Segal 1992, Lowe 2004: 270–3, de Jong 2010. A good example given by Lowe is the Electra of Orestes, who is explicitly less than forthcoming when it suits her, at 16 and 26–7. For our present passage, Goldhill (1986a: 246) has similarly argued that ‘it would be wrong to regard this prologue as entirely or distinctly separate from the dramatic action’.

Rustic Language: The Peasant

63

without internal variation are greatly misleading. This point was well made already by Grube: Where critics err is that they apply [Aristophanes’] criticism to the monologues as a whole, whereas it can be justified only of the first few verses. . . . Nearly every monologue refers very briefly to the more remote past, then passes on to a much fuller, more vivid and more dramatic presentation of the present circumstances and the immediate past which brought them about. . . . [T]he remote past rarely takes up more than ten or fifteen verses . . . After this the speaker often rises to a high pitch of emotion and the poetry, pedestrian at first, rises too. (Grube 1941: 66)

Moreover, even in those first few verses, ‘Euripides usually mitigates the formality of his narrative prologues’ (Denniston on line 1) by having his characters begin with a brief gnomic statement and/or a personal revelation (e.g. Iolaus at Heracl. 1–5, 5–9, Electra at Or. 1–3), apostrophize present or distant places or gods (e.g. Alc. 1, Andr. 1, El. 1, Phoen. 1), lament past misfortunes (e.g. Med. 1–6,6 Phoen. 4–6), and so forth. All in all, there is more going on in these monologues than Erbse would allow. The prologue of Electra provides an effective example. The first thing that may be noticed is how the Peasant organizes his narration, with various changes of narrative pace and detail.7 He begins with a brief apostrophe of Inachus (in a textually baffling verse),8 before embarking on a concise exposition of the antecedents leading up to the current situation (1–10). The first seven lines are indeed straightforward (Grube might call them ‘pedestrian’): they consist of an uninterrupted string of short clauses with aorist indicatives (ἔπλευσε 3, ἀφίκετ’ 6, ἔθηκε 7 9), preceded in most cases by one or more aorist

6

Med. 1 is conspicuously featured at Ar. Ran. 1382, where Dionysus and the tragedians are no longer concerned with prologues but with weighing their best ῥήματα (1367) in the scales. 7 In the following analysis I accept much from the note on this passage in Slings 1997c: 132 and from the relevant sections in van Wolferen 2003 (esp. 5–7). 8 Many scholars have tried, through emendation or explanation, to make sense of the paradosis ὦ γῆς παλαιὸν ἄργος: in addition to the notes in the commentaries of Keene, Denniston, Cropp, Roisman and Lushnig, and Distilo, see Zuntz 1970, Haslam 1976, Basta Donzelli 1980a, Kovacs 1996: 95–7, Willink 2009: 205 n. 1. No one has done enough to put the matter to rest (and I can offer nothing new on the matter). 9 Haupt’s conjecture ἔθηκε for L’s τέθεικε is well defended by Slings (1997c: 132); contra Kamerbeek (1987: 276).

64

Rustic Language: The Peasant

participles (ἄρας 2, κτείνας 4, ἑλών 5)10 and linked by the particle δέ. As Buijs well discusses, in clauses like these—with a nominative aorist participle preceding an (aorist) main verb—the verbal forms typically relate ‘a series of sequential actions within a referentially continuous thematic sequence’ (Buijs 2005: 187); in other words, both the participle and the main verb present closely related information that pushes the narrative forward, even if one (the indicative) may be conceptually more significant than the other (the participle). This is thus narration at its most bare-bone: every verb form—one quickly following after another—contributes to an outline of events in the remoter past, and no lengthy adverbial modifiers or subordinate clauses11 intervene to complicate the story. In 8–10, the Peasant shifts gears12 and the narration moves into territory that is more directly relevant for the characters of the play. The historic present θνῄσκει (9) plays an important part in this:13 Agamemnon’s death is presented as a pivotal moment, which changes the state of the narrated world. Its lasting relevance becomes clear immediately afterwards: the murder has given rise to a completely new balance of power at Argos. The regicide thus forms an easy transition14 into a description of the new political situation, which has lasted until the time of the play (11–13; the perfect and present indicatives both refer to present conditions). This rounds off the brief sketch of the more remote γένος τοῦ δράματος,15 and the Peasant now comes to a detailed 10 With one humorous exception (Ar. Ran. 1237), ‘Aeschylus’ always attaches his ληκύθιον ἀπώλεσεν to a preceding participle. If Aristophanes’ parody is directed at syntax as well as at content and metre, this is certainly a feature under attack. One almost imagines him trying to sneak in the words somewhere in the first seven lines of the Electra (but not later): in Euripides’ defence, he would have some difficulty doing so. 11 Strictly speaking, of course, lines 2–3 (introduced by ὅθεν) are a subordinate clause, and it could be argued that this is an element of syntactic complexity. In terms of narrative complexity, however, this is hardly relevant: the narration begins in line 2 (line 1, an apostrophe, cannot be narrative), and ὅθεν is a convenient device to link that narration to Inachus. 12 Slings (1997c: 131) uses the term volta for this type of shift in Euripides (he gives several examples), which frequently occurs in the prologues after about 10–20 lines. 13 See my Introduction, §2.6 for discussion of the function of the historic present. 14 The use of the particles καί (in χὠ) and μέν together (11) suggests that the description of the current situation is still part of the first sequence (καί), while at the same time serving as a transition to the next and longer narration (by contrasting Agamemnon’s fate to that of his children, μέν . . . δέ 14). Similarly at 8 (κἀκεῖ μὲν εὐτύχησεν. ἐν δὲ δώμασι). See van Wolferen 2003: 28–9. 15 Ar. Ran. 946–7.

Rustic Language: The Peasant

65

description of what happened to Orestes and Electra, the principal actors of the story and, significantly, those most directly relevant to the Peasant’s own situation (14–42). From this point onwards, the narration gets more complex both in form and in content: rather than a bald summary of events, we get a discussion of causation and motives, and these added layers of content are mirrored in an increased syntactic variation and complexity throughout. This increased syntactic complexity is apparent from the outset: the new topic is introduced by an elaborate, two-verse-long expression consisting of an ‘autonomous’ relative clause (οὓς . . . ἔπλει 14)16 with an apposition (ἄρσενά . . . θάλος 15). The length of this introduction and its placement at the beginning of the sentence, outside the actual syntactic construction (‘left dislocation’), are significant: the siblings will function as the theme of the entire subsequent exposition.17 For both siblings, the crucial events that have determined their current situation are again expressed in the historic present: Orestes’ exile is initiated by the Old Man’s stealing him away (ἐκκλέπτει 16), and Electra’s social and political elimination is accomplished by Aegisthus’ marrying her off (δίδωσιν 34).18 There is some distance in the text between the two events: the Peasant gives only a brief account of Orestes (16–18) but dwells on Electra’s fate. Before the pivotal verb δίδωσιν informs the audience of what eventually happened to her, there is a blow-by-blow account of the build-up to that event, with a prominent role in it for its instigator, Aegisthus. 16

For autonomous relative clauses, see KG II §554.4 and Rijksbaron 2002: §28. For left dislocation (a feature of many languages) and ‘theme’ status in tragedy, see Dik 2007; more generally in ancient Greek, see Dik 1995 passim, Matić 2003: 581–2. Relative clauses used in this way are not uncommon in Euripides (cf. 19 and further examples from prologues, e.g. Hipp. 3–4, 21, Andr. 47, Her. 31–2); the present instance is perhaps somewhat unusual in the amount of ‘dislocated’ material. 18 This δίδωσιν has been seen as a ‘perfective’ or ‘registering’ present rather than a ‘historic’ present (Rijksbaron 1991: 3 n. 6), and a case could be made with the other two present indicative forms in the prologue which do not refer to present time (θνῄσκει 9, ἐκκλέπτει 16) for interpreting them as ‘perfective’ as well; for the distinction, see KG I §382.a2/4 and Rijksbaron 1991: 1–4 (‘denote . . . some state of affairs that was carried out in the past but is still relevant at the moment of speech’). Unlike Rijksbaron, however, I think that these forms are all part of narrative rather than ‘expository’ discourse, and should therefore be taken as historic presents after all. In any case, under the conception of the historic present outlined in my Introduction (§2.6), ‘historic’ and ‘perfective’ uses may in fact be closely related (this is suggested also by Rijksbaron 1991: 2–3 and 2002: 23). 17

66

Rustic Language: The Peasant

Aegisthus’ role in the prologue is worthy of further consideration, in particular when viewed in relation to Clytemnestra’s. In comparison with the rest of the play, where Clytemnestra will come to play a crucial role, she is given somewhat short shrift in the Peasant’s opening monologue. Apart from what could be described as a ‘cameo’ in 27–30, where she interferes with Aegisthus’ plans (ἐξέσωσεν Αἰγίσθου χερός 28, ‘she saved her from Aegisthus’ hand’), the Peasant has little concern for Clytemnestra. His focus is clearly on Aegisthus’ motives: it is his fear that brings about Electra’s mock marriage (δείσας 22, φόβου 25). He alone is described as plotting Orestes’ death and Electra’s removal (ἐμηχανήσατο 31). Even in the description of Agamemnon’s murder (9–10), Aegisthus’ part is equal to that of Clytemnestra, if not larger: he is given the more elaborate introduction—with the added τοῦ Θυέστου παιδός, ‘son of Thyestes’— and it is suggested that he performed the actual deed (χερί) as opposed to Clytemnestra’s scheming role (δόλῳ). Forms of the name Αἴγισθος appear seven times in the opening monologue, against a single Κλυταιμήστρας (9; but also Τυνδαρίδα κόρην in 13). All this suggests that Aegisthus, rather than Clytemnestra, is foremost in the Peasant’s mind.19 This brings me to a first general point about the prologue: it appears that the elements of the story that are of personal significance to the Peasant are presented by him with specific attention. The three characters with the most prominent presence in his speech are those who are most relevant to him personally: Aegisthus as the person directly responsible for his marriage to Electra, Electra herself, and Orestes (Electra’s kurios, for whom the Peasant will show consideration at 47–9; his stance is confirmed by Electra at 261). The elaborate introduction of Electra and Orestes focuses his narrative on them, and the aspect of the tale that the Peasant proceeds to relate in most detail is Electra’s marital situation, which is of course the one that concerns him most. Thus the prologue is far from an objective ‘playbill’ designed only to provide the necessary background information to the audience. It sets up the rest of the play to be sure, but it does so in a way that is very specific to the person narrating it. 19

The prominence of Aegisthus in the prologue has alternatively been explained as a method of removing guilt for Agamemnon’s murder from Clytemnestra by painting her in a more sympathetic light. Even if this is true (but see also Cropp’s good discussion of Clytemnestra’s character, 2013: 4–5), it does not invalidate my point.

Rustic Language: The Peasant

67

I have thus far left out of discussion the parts of the prologue that are most clearly an expression of the Peasant’s personal involvement: his first mention of himself and of his family background (34–8), and his surprising revelation that he has not touched Electra (43–6) and subsequent ardent defence of that choice (50–3). It is here that the Peasant comes most to the fore as an individual, and I would maintain that it is here, too, that we may find stylistically personalized touches. That we are in different territory from line 34 onwards is clear from the first word, ἡμῖν. In the preceding 33 lines, uniquely in Euripidean prologues,20 there is not a single first-person pronoun, verb, or anything else to draw attention to the person delivering the opening speech.21 After ἡμῖν, we find ἐξελέγχομαι (36), ἁνὴρ ὅδε/μοι (43), ᾔσχυν(α) (44),22 αἰσχύνομαι (45), στένω/ἐμοί (47), μ᾽ (50), and θιγγάνω (51): the point of view has clearly changed. The delayed revelation of the Peasant’s identity has been rightly seen as a way to engage the audience through curiosity,23 but what concerns me here is his almost overpowering presence when he at last becomes the centre of attention: it is (if I may be allowed an anachronistic analogy) as if the spotlight is directed for a long time at a projection of images on a screen, before suddenly hitting with full intensity the person in the middle of the stage who was providing the narration with those images—and only then do we find out why that person is telling this story, and why he does so in this way. The intensity of the ‘spotlight’ lies not only in first-person pronouns, verb forms, and deictic expressions. Another striking aspect of these passages is the use of non-connective particles, that is, particles that deal (by and large) with the attitudes and expectations of speaker and addressee.24 Such particles are wholly absent up to line 30, but in

20 The only comparable delay occurs in the prologue of Helen, where the title character waits until line 15 before first referring to herself, and seven more before revealing her identity. 21 The closest thing is a single demonstrative expression that has the speaker as its ‘deictic centre’: τόδ’ (6). 22 An uncertain reading, but ᾔσχυνεν too would have the Peasant (ἁνὴρ ὅδε) as subject. 23 See Cropp 1988 on lines 1–212 with Slings 1997c: 131—who rightly compares the similar opening of Helen (see n. 20 above). 24 See Introduction, §2.6.

68

Rustic Language: The Peasant

the lines 31–44 they proliferate, especially δή (31, 34, 36, 44) and γε (36, 37 bis).25 Assigning stylistic significance to these particles is problematic, especially since a satisfying definition of the function of δή remains, in my opinion, a desideratum.26 A full discussion of the particle is outside the scope of this work, but I will succinctly state my own view: in tragedy at least, δή is normally used to highlight the personal commitment of the speaker to the utterance he/she is making (or to a specific element of that utterance, depending on its scope). It is something for which he/she can vouch, or which he/she considers evident and important; and the speaker wishes to draw the hearer’s attention specifically to this.27 This function is extended into a frequent use (often combined with anaphoric elements such as τῶνδε in 31), in which δή marks the transition to a new step in a discourse that is central to the speaker’s case. To return to the issue at hand, what makes the use of these particles stylistically significant? First, it is worth noting a statistical peculiarity: no other character in extant Greek tragedy uses the particle δή as often within such a short space as the Peasant does here.28 This is interesting in itself, but meaningless if we cannot determine why the 25 To be precise, γε is a particle of scope, limiting the applicability of an utterance to a certain element, whereas δή belongs to the category of ‘modal’ particles. In my count I have read Stobaeus’ (97.5) γε μήν for L’s δὲ δή at 37. Although the latter is printed by most editors and commentators, I have a slight preference for Stobaeus’ combination (adopted by Kirchhoff, Nauck, Wecklein, and Keene), where γε limits the family’s privation to the material aspect (χρημάτων) and μήν goes against any expectations raised by λαμπροί (for μήν as an affirmative particle anticipating disbelief on the hearer’s part, see Wakker 1997a): ‘illustrious by descent, yet when it comes to means underprivileged’ (pace Denniston ad loc.: ‘γε μήν, “aye, but”, would stress the poverty too much’: why too much?). A scribal error in L is conceivable in view of δὲ δή two lines earlier. 26 For contrasting views, see Bäumlein 1861: 98, KG I §500, Denniston 1954: 203–79, Sicking and van Ophuijsen 1993: 52, 82, passim, Wakker 1994: 351–7, 1997a, 1997b. My own views on the particle’s function coincide most closely with those of Wakker; I do not think that ‘evidentiality’ (so Sicking and van Ophuijsen) is a necessary and inherent part of δή’s force. 27 This definition is only slightly different from that given by Wakker (see previous note). 28 Even when not counting the possible instance in 37 (see n. 25). The only other character in Euripides to display a similar heavy-handed use of the particle is Electra, in the prologue of the Orestes (7 occurrences in 70 lines). Two other δή-dense passages in Euripides, but less so, are Med. 1021–67 (5 occurrences, Medea speaking) and Andr. 319–33 (3 occurrences, Andromache). In Sophocles, see Aj. 992–5 (3 occurrences, Teucer).

Rustic Language: The Peasant

69

particle is used so often.29 With regard to this question, even if we assume for the particle Denniston’s (1954: xxxviii) commonly accepted value of ‘emphasis’,30 it is hard to escape the conclusion that the Peasant is particularly forceful in this passage. When the crucial twist in the narrative (Aegisthus’ marital scheming) comes about (31), the Peasant ensures, through δή, that the audience is paying attention. He does so again when he mentions himself and the marriage (34). The next two instances of δή (36, 44)31 both appear in places where the Peasant is almost needlessly elaborating his points: that he is from Mycenean stock (πατέρων . . . Μυκηναίων ἄπο 35) implies that he is above reproach in terms of his birth,32 but the Peasant emphatically spells it out (οὐ δὴ τοῦτό γ᾽ ἐξελέγχομαι 36, ‘on this point I cannot be faulted’); the fact that he has not humiliated Electra by consummating the marriage (οὔποθ᾽ . . . ᾔσχυν᾽ ἐν εὐνῇ 43–4, ‘I have not brought shame on her in bed’) logically entails her status as παρθένος, but he stresses the issue (παρθένος δ᾽ ἔτ᾽ ἐστὶ δή 44, ‘she is still a maiden’). About γε, similar things may be said. The Peasant employs it in 36–7 to concentrate the audience’s attention on the particular contradiction in his situation that he wants made clear: he is of good birth (above reproach in that respect), but simply penniless. All this suggests that the Peasant is determined, when discussing his own social status and his ensuing conduct towards Electra, that he should not be misunderstood. At this point I can venture to use the word ‘character’ again: when it comes to questions of social status and norms of behaviour, the Peasant is vigorous in his beliefs, and his speaking style changes 29 I have wondered if the accumulation of instances of δή perhaps reflects an affinity for the particle in everyday speech, or even in the speech of ‘people from the country’. There appears to be, however, little evidence for this, at least in tragedy (where I have looked at all cases of δή); see also the study of particles in different types of texts by Duhoux (1997). The particle appears in Stevens’ diligent collection of colloquialisms in Euripides only in the combination δὲ δή used in questions (Stevens 1976: 46). The only feature of the Peasant’s language that Stevens (33) might have called colloquial is ἔα in 341; Collard (2005: 368, 375) might perhaps add τί δ’; in 406 and εἰ δοκεῖ σοι in 77 and 420. But see my comments on the ‘register’ of the Peasant’s language in §4. 30 Denniston 1954: xxxviii. I here pass over Denniston’s (1954: 204) refinements on the term ‘emphasis’ (‘δή denotes that a thing really and truly is so’), which actually turn out to be one of the main problems of his description. A simple matter of distribution, the frequent appearance of the particle in questions and commands (which cannot be ‘true’), virtually rules out a truth-associated meaning. 31 32 Not counting 37 (see n. 25). See Denniston ad loc.

70

Rustic Language: The Peasant

accordingly: it becomes energetic and forceful. The Peasant clearly finds his values important and is firm in his expression of them. There is one more stylistic point that adds relief to this conclusion, and it is to be found in the Peasant’s condemnation of anyone who might mock his sexual restraint (50–3). Again, the Peasant is particularly forceful in the defence of his conduct. One indication is his use of words of fairly strong ethical colouring, particularly πονηροῖς (52) and τὸ σῶφρον (53).33 A less obvious clue, one that is best viewed in conjunction with the Peasant’s interaction with other characters in the play, is his use of the imperative ἴστω (‘he should know’) without any form of redress. As I will discuss, this expression is as bald* as the Peasant will express: the same posture of unadulterated moral righteousness is not something he can sustain when talking to his wife.

3. HUSBAND AND WIFE (54–81, 341–63, 404–31)

3.1. A marriage under face threat We should expect the language of the Peasant to take on some different qualities in his conversations with Electra. Instead of a long speech by a lone actor, here we have dialogue where two characters interact (or more: see 357–62). But some of the same aspects of the Peasant’s character I have outlined are visible in these passages as well, together with a new but related feature: the Peasant is supremely polite in his conduct towards his wife. Interestingly, we can detect in some of his expressions a hint that these two impulses (being polite to his wife and being true to his ideas of good conduct) conflict, causing the Peasant to strike an occasionally awkward balance. For my analysis of the interaction in these passages, I now turn to some of the methodological approaches discussed in my Introduction, especially politeness theory* (§2.4.2). The conversations between the Peasant and Electra are teeming from the start with face-threatening acts (FTAs)*, that is, conversational turns that pose a threat to the roles the two characters seem to want 33 For πονηρός (and πονηρία, etc.), see LSJ s.v., III and Sluiter 2008: index s.v.; for σώφρων (and σωφροσύνη, etc.), see Rademaker 2005, with discussion of this passage at 151–2 and 162–3.

Rustic Language: The Peasant

71

to adopt for themselves in their mutual interaction. What we see going on between the two is a constant struggle and negotiation over authority and proper behaviour (in fact the two find it hard to agree on anything at all). At the same time, both husband and wife carefully couch their disagreements in polite and affectionate terms, and the Peasant in particular seems to prioritize maintaining social equilibrium between himself and his wife over the fulfilment of his own moral expectations. His conduct towards her is characterized throughout by what Brown and Levinson have called negative politeness*, especially deference (often at his own expense), which is particularly suitable in dealing with someone of superior social standing.34 Considering the fact that most of what the two argue about concerns rather homely and everyday activities (fetching water, preparing food), the polite posture of the Peasant is all the more striking, since we might expect the politic behaviour* in such interactions not to require great amounts of supportive facework*. A closer view of their conversations will clarify these points.

3.2. Getting water (54–81) The first thing that the Peasant says to Electra follows on her entrance and her address to Night, in which she laments her situation and expounds her reasons for going off to get water (54–63).35 Electra, in addressing this speech to a divine audience (ὦ νὺξ μέλαινα 54, ‘O dark Night’), has failed to note the Peasant. The Peasant, however, has seen and heard her,36 and asks her about her water-carrying venture (64–6): τί γὰρ τάδ’, ὦ δύστην’, ἐμὴν μοχθεῖς χάριν πόνους ἔχουσα, πρόσθεν εὖ τεθραμμένη, καὶ ταῦτ’ ἐμοῦ λέγοντος οὐκ ἀφίστασαι;

34

See Brown and Levinson 1987: 178–9. I discuss this speech in more detail in the next chapter, §2.2. 36 So Bain 1977a: 33, Mastronarde 1979: 27, Cropp ad loc. The fact that the Peasant seems, in his question, to ignore the reasons offered by Electra at 57–9 has caused scholars to grasp for solutions ranging from the deletion of 57–9 (e.g. Kirchhoff, Wecklein ad loc.) to a staging that sees the Peasant depart for the duration of Electra’s speech (Denniston on 64–6). Such solutions are unlikely in themselves, and entirely disproven by the Peasant’s use of γάρ at 64 (see next note). For the apparent inconsistency between the Peasant’s question and 57–9, see also my discussion in ch. II, §2.2. 35

72

Rustic Language: The Peasant

Why, then, unhappy woman, do you toil at these tasks on my account, enduring labours when before you were finely brought up, and do you not refrain from them even though I tell you to?

This turn is seemingly a request for clarification (γάρ 64),37 but beneath the surface of the question much more is going on. By adding ἐμοῦ λέγοντος (‘even though I tell you to’),38 the Peasant ensures that it cannot be read as a completely neutral request for information. In terms of Gricean pragmatics*, the addition would be irrelevant if the Peasant truly wanted merely to know why Electra is fetching water. It steers the interpretation of the question as something other than just a question (it gives rise to an implicature*): in other words, simply by mentioning that Electra is going against his wishes, the Peasant betrays that he is not very pleased to see her do so. ‘His question is in fact a kind of gentle rebuke’ (Bain 1977a: 34).39 His conviction seems to be that women of high class (εὖ τεθραμμένη 65, ‘finely brought up’) should not demean themselves by doing outdoors manual labour, or at least that there is no need for Electra to do so: his criticism can be read as an attempt to assert this conviction as well as his authority, as a husband, to steer his wife’s conduct in such matters. 37 To my mind, no one to date has satisfactorily explained this use of γάρ; Denniston ad loc. calls it ‘difficult’, Bain (1977a: 34 n. 1) ‘hard to classify’. As I discuss more fully in van Emde Boas in press-a, γάρ in questions opening an expansion sequence* (something that might be said of 64–6) often indicates that the speaker is not wholly satisfied with an earlier turn of the interlocutor and needs further clarification. Here it should be taken as a signal that the Peasant wants to know more about what Electra has just said (which presupposes, of course, that he has heard it). The cases of γάρ in questions listed by Denniston (1954) under V.(6).(iii), including El. 64 (listed at p. 82), and under VI (‘progressive’) should in my view be taken together. 38 It is tempting to take καὶ ταῦτ’ as adverbial, modifying the genitive absolute ἐμοῦ λέγοντος (‘and that too when I tell you (to stop)’; for the construction καὶ ταῦτα with participle see K-G II §486 Anm. 8 (= pp. 85–6), citing e.g. Xen. An. 6.2.10 and Pl. Resp. 404b). But this necessitates Dobree’s πόνους in 65, which needlessly divorces the participles in that line from μοχθεῖς in 64. It is preferable, then, to take καί as connective between μοχθεῖς and ἀφίστασαι. See Denniston ad loc. 39 So too, e.g. Mastronarde 1979: 27 n. 34 and Distilo ad loc.; and see ch. II, §2.2. Denniston seems to miss this tone in his note on 64–6: ‘Electra has given the answer herself at 57–9. But the Farmer was not on stage then, and he is not a good enough psychologist to guess the solution’; yet the Peasant is only half-interested in the answer (for Denniston’s staging, see n. 36; for the apparent inconsistency between 57–9 and 64–6, see ch. II §2.2). The interpretation as criticism is facilitated by the negation in οὐκ ἀφίστασαι: such rhetorical questions with a negation suggest that the positive form (in this case, Electra ceasing her water venture) is the desired, expected, or believed state of affairs (on this ‘polarity reversal’, see Romero and Han 2004).

Rustic Language: The Peasant

73

Yet if there is reproach in his words, the Peasant is careful to express it in a circumlocutory and conciliatory way: the address ὦ δύστην᾽ (‘unhappy woman’) is an expression of genuine sympathy40 (which can mitigate the threat of a FTA*),41 and the very fact that the Peasant has disguised his comment in the form of a question makes it so indirect that it is possible to ignore or miss it—the Peasant gives his wife an ‘out’ (his phrasing constitutes an off-record* strategy).42 Ignoring it is, in fact, exactly what Electra appears to do in her response, an interestingly crafted turn (67–76).43 Not only does she refuse to change her behaviour in accordance with her husband’s criticism, she also fails to really engage with that criticism at all. She makes use of the ‘out’ her husband had offered her by treating his question as if it were just that: a question, which she answers by explaining her reasons (71–6). Yet Electra, too, is concerned with maintaining equilibrium between herself and her husband: her extensive praise of him (67–70) constitutes a significant amount of effort exerted to support his face*, in spite of her failure to go along with his wishes. Whether it is due to Electra’s high praise or his own leniency, the Peasant’s response constitutes retreat. Again, his turn is marked by a great deal of facework*, particularly in the unmistakable expression of deference and acquiescence εἴ τοι δοκεῖ σοι (77, ‘if you think so’)—the first of two occasions (cf. 420) on which the Peasant uses this phrase.44 The deferential expression does not conceal the Peasant’s 40 Dickey 1996: 164: ‘in epic and tragedy δύστηνε generally indicates real sympathy’. 41 See Brown and Levinson 1987: 106–7. 42 See Brown and Levinson 1987: 223–5, Watts 2003: 184. A way of putting this in conversation analysis* terminology is to say that the way in which the Peasant constructs his turn allows for the avoidance of a dispreferred turn* by activating several preference organization* structures at once: he allows the dispreferred reaction* at one level (rejecting the Peasant’s implicit request to change behaviour) to be packaged as a preferred reaction* at another, more explicit level (answering a question). For such conflicting preference structures, see Schegloff 2007: 73–8. 43 I will have more to say about these lines in the next chapter, §2.2. 44 See Barrett on Hipp. 507–8 for excellent discussion of the formula (and the way in which it is there manipulated). Katsouris (1975: 75) rightly considers the phrase emblematic of the Peasant. His imperative στεῖχε will not be felt as a large FTA*, in part because it is exactly what Electra wants to do, in part because the water expedition has already been the topic of discussion: the imperative does not come out of the blue (this is known as ‘task‑orientation’; see Brown and Levinson 1987: 97, Lloyd 2006: 226; in Watts’ terms, the linguistic politic behaviour* is such that it requires less facework*). There are, it appears, many conditions under which an imperative in

74

Rustic Language: The Peasant

displeasure at the course of events: by delegating the decision to Electra’s judgement—something the Peasant underlines with his particle τοι—he also suggests that his own judgement is still unchanged. Thus, humouring Electra’s wishes means conceding his own position, and his exaggerated explanation for agreeing to let her go and for backing away from his previous objections (καὶ γὰρ οὐ πρόσω πηγαὶ μελάθρων τῶνδ᾽ 77–8, ‘in fact the spring is not far from this house’) seems directed at this concession as much as anything else. In this first brief exchange we can thus see the tricky negotiations of face* and authority that the Peasant and Electra have to perform in their interaction, the Peasant drawing the short straw (voluntarily). The conversation is also suggestive of a fundamental inability of the two characters to see each other’s point of view, which I will revisit more fully in the next chapter.

3.3. Welcoming guests (341–63) The Peasant departs with a general reflection about idleness and toil (80–1),45 and is then absent for more than 250 lines. On his return (339), he is unpleasantly surprised (ἔα)46 to find his wife conversing with two strange men. He wants to know who they are and why they have come (341–3), of course, but he also cannot refrain from expressing his displeasure. Electra’s mingling with men outside the house does not conform to his conception of good womanly conduct and, as before, he takes a strong stance when he sees something that ancient Greek was not considered impolite (or in any case not as impolite as in English): see Probert and Dickey 2005, Lloyd 2006: 227, Denizot 2011: esp. 246–50. But much on this issue is still unclear: in spite of Denizot 2011 (and, specifically on Homer, Minchin 2007: ch. 8), there is still a need for a large-scale study of the politeness of various directive forms in Greek, and face-threat politeness theory would be a promising avenue of approach. 45 This is, according to the categorization outlined in the Introduction (§2.5) a first-person saying* (gnōmē), by which the Peasant embeds his own hard work in a larger ethical framework in which being idle (ἀργός 80) is a particular vice. The generalization serves at once as an explanation for the Peasant’s departure and as a neat characterizing touch. 46 Fraenkel on Aesch. Ag. 1256 (n. 4) gives the most accurate description of ἔα (more than simply ‘surprise’, as, e.g. LSJ s.v.): ‘Without exception in Euripides it expresses the surprise of the speaker at some novel, often unwelcome impression on his senses, the visual sense or another’ (my emphasis). Cf. its use attendant upon a surprising smell at Hipp. 1391.

Rustic Language: The Peasant

75

does not square with his values: γυναικί τοι | αἰσχρὸν μετ᾽ ἀνδρῶν ἑστάναι νεανιῶν (343–4, ‘for a woman, as you know, it is shameful to stand with young men’; note the strong ethical connotations of the word αἰσχρός).47 This generalizing statement plays a nuanced role in the Peasant’s communicative strategy. Following Seitel’s classification outlined in my Introduction (§2.5), this is a second-person saying*, whose content is applicable to the addressee (Electra, a woman).48 The Peasant makes this explicit through the particle τοι, used ‘to point the applicability of a universal truth to the special matter in hand: it forces the general truth upon the consciousness of the individual addressed’ (Denniston 1954: 542). Such second-person sayings* constitute some threat to the face* of the addressee, and τοι heightens the face threat for Electra, as does the strong term αἰσχρόν. The Peasant’s appropriation of conventional morality here can thus be seen as another, more forceful attempt to assert his husbandly authority over Electra. Yet, again, his reproach is not without a placatory note: simply phrasing it in general terms somewhat limits the damage done. Electra is not explicitly implicated in the second person (this is another case of an off-record* politeness strategy*).49 Still, in this case the Peasant’s gnōmē can barely be felt by Electra as anything other than a reproach, and this time the husband prevails. Electra immediately strikes a conciliatory note (ὦ φίλτατ᾽, εἰς ὕποπτα μὴ μόλῃς ἐμοί 345, ‘dearest husband, do not get suspicious of me’) and offers an explanation. Electra’s entire turn is in fact concerned with maintaining a social equilibrium between all the interactants in the conversation, and this requires her to divide her attention between the various men surrounding her: after placating her husband, she also apologizes to Orestes and Pylades (348) in case they felt maligned by the Peasant’s suspicions.

47

See LSJ s.v. II.2 and Sluiter 2008: index s.v. The word order, with γυναικί placed initially, makes it clear that women’s behaviour is what this saying is about (it can be taken as ‘topic’ or, possibly, as ‘theme’; see Dik 2007: ch. 2). 49 See Brown and Levinson 1987: 226. In terms of Gricean pragmatics*, the proverb, when understood completely literally, would be an irrelevant contribution (it is not meant to ‘inform’ Electra about general behavioural guidelines for women, nor is their discussion about that topic): this forces the indirect interpretation as criticism, which does have relevance in the specific context. The particle τοι steers the interpretation forcefully in this direction. 48

76

Rustic Language: The Peasant

The Peasant is brought up to speed concerning Orestes and his ‘messengers’ (349–56). As soon as he learns that the necessary information has already been exchanged between Electra and the strangers (355–6),50 he reverts to his practical attitude and once again asserts his view of proper behaviour, in this case on the subject of being a good host: οὔκουν πάλαι χρῆν τοῖσδ’ ἀνεπτύχθαι πύλας; (357, ‘then shouldn’t our doors have been opened to them by now?’). A note of criticism of Electra’s behaviour is again inescapable: she has not displayed as much guest-friendship as she should have. But, once again, the Peasant’s formulation dampens the FTA* by way of negative politeness*: the impersonal χρῆν effaces the Peasant’s role as ‘deontic source’, and Electra herself is not ‘nailed down’ in the second person as the one who should have opened the doors.51 The Peasant asserts his husbandly authority, but only indirectly. He is more direct in his words to the strangers at 358–63 (and possibly to their servants at 360),52 using the imperative χωρεῖτ’ (358, ‘go in’) and the equivalent construction μή with subjunctive in μηδὲν ἀντείπητε (361, ‘don’t object’), even if the directness here need not be seen as impoliteness. His directives serve as a welcome (358) and a reassurance (361–2), and thus imply little face threat (it will take Orestes some time, however, before he can recover from his shock and accept the offer: see ch. III, §2.2.3).

3.4. Preparing food (404–31) I treat one last example in detail, from the final interaction between the Peasant and Electra (404–31). In his very last words to her (420–5), there is once more a delicate balance between FTAs* and 50 The particle που (‘I suppose’) in 355 is a neat touch, which shows how well the Peasant knows his wife. His surmise that Electra will have gone into detail about her situation is clearly justified. 51 See Denizot 2011: 406 (on expressions of deontic modality such as δεῖ/χρή + inf. or -τέος gerunds): ‘le rôle énonciatif du locuteur est effacé dans les énoncés déontiques’; for the notion of ‘deontic source’, see, for Latin, Risselada 1993: 183. On impersonal constructions and the avoidance of the second person as possible politeness devices in Greek, see e.g. Probert and Dickey 2005: 3 and Lloyd 2009: 187–8. To be clear, most of the works cited in this footnote refer specifically to directives, and χρῆν . . . ἀνεπτύχθαι, expressing unfulfilled obligation (see e.g. Rijksbaron 2002: §8.1), cannot really count as that. Nevertheless, for the purposes of analysing politeness, an analysis along these lines appears valid. 52 On the authenticity of line 360, see ch. III, §2.2.3 n. 40.

Rustic Language: The Peasant

77

politeness strategies*, as well as more signs that the Peasant is conflicted between what he feels is sensible and his consideration for Electra’s wishes. The first word of the speech, the particle ἀλλ᾽, is important in this last respect. Denniston (1954: 16–17) describes this use of the particle very well, I think, when he categorizes it under this header: ‘Agreement is presented . . . as wrung from the speaker malgré lui. ἀλλά then points the contrast between the assent given and the considerations which have militated against the giving of it.’53 The Peasant will go, but it is not by his own volition that he does so. The protestation is quickly swallowed, however, as he again advertises his deference (εἰ δοκεῖ σοι 420, ‘if you think so’) and agrees to do her bidding (τούσδ᾽ ἀπαγγελῶ λόγους | γέροντι 420–1, ‘I’ll report these words to the old man’). We thus see, again, how the Peasant prioritizes comity with his wife over any objections he may feel to her request, and responds with a preferred turn* rather than complicating the conversation. Yet, even after the Peasant concedes defeat, the marital struggle over authority is not completely over. He finishes with a request of his own, ‘in which his practical attitude is revealed, and in which there is a trace of criticism of Elektra according to his own values’ (Smith 1948: 94): she should take care of the guests (421–5). Again, the imperatives χώρει (421, ‘go’) and ἐξάρτυε (422, ‘prepare’) are not necessarily very threatening,54 yet they are also not as profusely polite as his earlier display of concern over her water-carrying mission. His request is followed by another generalizing account of the woman’s role (γυνή) in such situations (422–3): it is carefully phrased (note the ‘cautious’ potential optative ἂν εὔροι),55 but also leaves no doubt (τοι) that the Peasant feels that Electra should behave like a proper wife in this regard. This analysis could be extended further, but the general picture is by now clear. The Peasant is still the same person as he was in the prologue, characterized specifically by his vocal attitude when it comes to his value system. Yet, when speaking to Electra, he finds himself torn between two things he finds important: his wife and his 53 Cf. the use of ἀλλά cutting off a series of insert expansions* by Clytemnestra at 1131, discussed in the Introduction (§2.2); in the present case, ἀλλά suggests that the Peasant foregoes any such sequences (foregoes, in other words, voicing any objections) and simply assents to Electra’s request. For this use of ἀλλά in ‘corrections of discourse topics’, see also Drummen 2009: 151–2. 54 55 See n. 44. See Rijksbaron 2002: §14.2.1.

78

Rustic Language: The Peasant

morality. In the end, his wife carries the day: towards her he cannot adopt the same posture of absolute moral superiority as he does in 52–3, calling his hypothetical detractors πονηροί and giving them a bald* directive (ἴστω).56 Husband and wife bicker and struggle over proper behaviour, but the Peasant remains understanding and polite, never letting their conversation devolve into actual conflict and eventually giving in to his wife’s wishes easily. There is, we may conclude, some complexity to this character. What my discussions so far have shown, I hope, is that his characterization lies not only in what Euripides has him say, but also in how he makes him say it.

4. FURTHER STYLISTIC POINTS; CONCLUSION The Peasant’s conversational behaviour is thus a crucial aspect of his characterization, and it should be clear at this point that such aspects of language in communication are as much part of a speaker’s ‘style’ as more commonly acknowledged stylistic features such as vocabulary, sentence complexity, and so forth. This is not to say that the Peasant is not characterized by his language in those respects as well: Katsouris (1975: 75) has pointed to his use of ‘plain vocabulary’ in such lines as ἐγὼ δ᾽ ἅμ᾽ ἡμέρᾳ | βοῦς εἰς ἀρούρας ἐσβαλὼν σπερῶ γύας (78–9, ‘at daybreak I will drive the cattle to the field and sow the furrows’), and it is worth noting in this respect that the Peasant’s vocabulary is largely unexotic throughout. This is clear, for example, from the low frequency of hapax legomena in his speech: apart from Castor, who utters none, the Peasant has, in the play, the lowest frequency of words that do not recur in Euripides. The same holds if we count only words that do not recur in Greek literature at all, that is, ‘absolute’ hapax legomena. His one absolute hapax is in fact typical of the ‘everydayness’ of his vocabulary: it is προσφορήματα (423, ‘supplies’), describing food. 56 The communicative context of this imperative is, of course, very different from the Peasant’s interactions with Electra: it is phrased generally to no particular addressee present on stage (though one might argue that 50–3 are addressed, at least indirectly, to the audience; for different views on the issue of audience address in Euripidean prologues, see Bain 1975, Taplin 1977: 129–34, Erbse 1984: 64, Goldhill 1986a: 246, de Jong 2007: 20–1).

Rustic Language: The Peasant

79

Another possible stylistic marker is the Peasant’s repeated use of gnōmai (50–3, 80–1, 343–4, 422–3, 426–31), some of which have already been discussed in §3. The use of generalizations or proverbial expressions is sometimes seen as itself a stylistic indicator of lower class characters in tragedy (the Watchman in Aeschylus’ Agamemnon being the most cited example).57 As I have made clear in my Introduction (§2.5) and as will become evident throughout this book, my own view is that the use of generalizations may be very differently significant in different cases, depending on how and when they are used, and that it is best analysed character by character and context by context. In this particular case, however, there is certainly good reason to see the Peasant’s use of gnōmai as a characterizing touch. Cropp, in his commentary to lines 35–8, detects a ‘tendency to moralize in a self‑conscious and slightly naïve fashion’: whether we should call the Peasant ‘naïve’ or not, his generalizations do seem indicative of his humble, simple status, not only because they hint at a moralizing ‘bent’ (which could indeed be linked to lower class status), but particularly because of the content of his gnōmai, which build up an ethical worldview through references to labour (80–1), food (422–3, 430–1), and poverty (426–31). The combined picture, then, is striking: the Peasant—Euripides’ opening gambit in the play—is drawn with meticulous care. The contours of his characterization are found not only in what he says, but also in how he says it. This detail of characterization mirrors that of the person he goes up against: Electra herself. It is to this controversial figure that I now turn.

57

See, e.g. Rutherford 2012: 296–7 and Seidensticker 2009: 238.

II Constancy amid Change The Linguistic Characterization of Electra

Elektra and her motivations are the battleground for critics of the play. —Ann Norris Michelini (1987: 187)

1. INTRODUCTION In the previous chapter I examined the language of the Peasant in order to show that Euripides imbued that character with a nuanced individuality and that this characterization is achieved not only by what the Peasant is made to say (and do), but also by how he is made to say it. Regardless of the intricacy of his portrayal, however, the Peasant is undeniably a character of narrow scope in the play, and the contexts in which he appears are limited. Things could not be more different in the case of Electra, the play’s most extensively used character (on stage for about six sevenths of the whole) and the one given the widest range of situations and emotions to contend with. In this chapter I will argue that for Electra, too, the way in which she speaks is a consistent part of her overall characterization. In the process I will revisit the question of how that character should be viewed—which is a source of considerable controversy in scholarship on the play: as a delusional and bitter, ‘anti-tragic’ and ‘anti-heroic’ version of Aeschylus’ (and Sophocles’?) heroine, or perhaps in a more complex fashion. The obvious difficulty confronting anyone who looks for constants in Electra’s ‘style’ is the fact that she utters speech and song in a

The Linguistic Characterization of Electra

81

variety of different modes: in the course of the play she sings two laments (her monody and amoibaion with the Chorus, 112–214, and her amoibaion with Orestes and the Chorus, 1177–232), pontificates in two rhetorical set pieces (the invective against Aegisthus, 907–56, and her half of the agōn, 1060–99), utters prayers and pleas, commands and admonitions—in short, she takes part in all kinds of communicative activities. These variations in situational and communicative context inevitably lead to variations in Electra’s language, so that distilling a ‘pattern’ from these scenes is not a straightforward exercise. To complicate matters even further, Electra’s role in the play undergoes various changes itself, as she evolves from a powerless, marginalized woman devoid of any effective help to a scheming mastermind reunited with her brother, an implacable instigator of an impious killing, and finally a remorseful co-matricide confronted with the injustice of Apollo’s divine commands. In spite of these variables, I believe that it is possible to identify several constants—continuing strands in Electra’s language that make it specifically characteristic of her. In the following sections I will go through most of her scenes in the play one by one (though some are discussed more fully in chapters IV and VI), analysing in each case how Electra’s use of language is shaped by her underlying motives, by her past, and—crucially—by her gender.

2. RESI STANCE THROUGH LAMENT: THE EARLY SCENES (54–81, 112–214)

2.1. Electra as mourner Electra leaves little doubt in the early scenes of the play that she is unhappy with her personal situation: not only does she outline quite explicitly what she feels is wrong with her circumstances (58–64, 118–21, 133–4, 175–88, 188–9, 207–10, 300–13), she also repeatedly uses terminology of lamentation to describe her own activities: terms such as γόος (‘wailing’), κατακλαίω (‘(be)wail’, ‘lament’), θρήνημα (‘dirge’),1 and so forth are frequent in the early parts of the play 1 γόος: 59, 125, 141, 144; κατακλαίω: 113, 128, 156; θρήνημα: 215; δάκρυον (‘tear’): 126 (πολύδακρυν, ‘full of tears’), 181 (bis); note also the terms for lyrical lamentation in

82

The Linguistic Characterization of Electra

and are nearly always used by Electra about herself. It is thus not for nothing that the first few scenes have been dubbed ‘Electra κατακλαίουσα’ (Katsouris 1975: 75), or that the character has been seen to claim ‘an ownership of grief ’ (Chong-Gossard 2008: 68). In this regard, Electra is part of a group of women in Euripides’ plays who react to misfortune and marginalization by adopting a posture of lamentation, often in song.2 To give but two examples: Helen in her eponymous play is occupied for a significant part of the first half of that play with the expression of grief, marked by a similar explicit use of lyric and grief vocabulary (cf. esp. Hel. 164–78); similarly, Andromache extensively laments her wretched circumstances at the start of her play (Andr. 91–116; note again the lyrical mode and the explicit use of terminology, especially in 92–3).3 It is instructive to compare the language of these other women in Euripides,4 because the comparison suggests that the wilful adoption of such a disposition is to some extent an expected reaction for a (Euripidean) woman who has suffered a reversal of fortunes. At the same time, the reaction of each individual woman in Euripides will also have its own unique dimensions. In the context of the character of Andromache, just mentioned, Allan similarly warns against a generalization of women’s behaviour: Their marginal status, often compounded by foreignness and slavery, enhances the impact of their moral and intellectual challenge to ‘the dominant orderings of patriarchal society’.5 If there is one feature

142–3 and the references to physical signs of lamentation in 146–50. Electra’s use of these terms and expressions is, as I will argue, not unusual for women in dire straits, though the frequency with which she uses them is high. 2 On the similar role of Sophocles’ Electra as a perpetual mourner and her use of the θρῆνος genre, see Swift 2010: 337–44, 366. See Swift’s ch. 7 for the use of θρῆνοι in Greek tragedy more generally; for the ritual of lamentation and the role of women and men in that ritual, see Foley 1993, Sultan 1993, McClure 1999: 40–7, Alexiou 2002, and Suter 2008a. 3 Both Electra and Andromache refer to the ‘pleasure’ that comes with female lamentation: Andr. 93–5: ἐμπέφυκε γὰρ γυναιξὶ τέρψις τῶν παρεστώτων κακῶν ἀνὰ στόμ’ ἀεὶ καὶ διὰ γλώσσης ἔχειν (‘there is a pleasure for women in having their present troubles constantly on their lips’); El. 126: ἄναγε πολύδακρυν ἁδονάν (‘raise the delight full of tears’). 4 Not all of whom, incidentally, have been condemned for their behaviour as much as Electra has been. 5 Allan here quotes Goldhill 1986a: 115.

The Linguistic Characterization of Electra

83

which Euripides’ work may be said to communicate most penetratingly in this area, it is to stress the distinctive tragic potential of women’s constrained experience. However, our appreciation of such distinctiveness needs to be carried through to the level of individual characters. In the Troades, for example, Cassandra, Hecuba and Andromache all face the same bleak backdrop of Troy in ruins, but each of them interprets her situation and reacts to it in a unique way. (Allan 2000: 194–5)

What, then, are the features that make Electra’s reaction to her fall from grace unique to her? And to what extent do such distinct features characterize Electra’s language? To answer these questions, it is useful to identify first the unique set of circumstances in which Euripides has situated his Electra. Unlike Andromache, whose misfortune also consists of the death of members of her oikos, banishment, and a new husband (of sorts), Electra is placed by Euripides in new surroundings that are mostly friendly to her. The central figure here, of course, is the Peasant, a husband who is as appreciative of Electra’s plight as he can be. Nevertheless, Electra interprets her new circumstances as social ‘death’ (θανάσιμος γάμος 247, ‘a deathly marriage’) and refuses to accept them. In fact resistance might be seen as the defining feature of Electra’s conduct in the first few scenes of the play. It is also this feature, I suggest, that defines her linguistic characterization in these scenes: it prevents her from having successful communication with those around her and determines, in her amoibaion with the Chorus, the lyric mode in which she operates.

2.2. Electra as wife The initial instance of the resistance pattern can be found in Electra’s first exchanges with her husband. In the previous chapter I have already touched on Electra’s behaviour in her scenes with the Peasant (54–81, 341–63, 404–31). As outlined there, she remains polite and even affectionate towards the Peasant (see esp. 67–70, 345), in spite of his careful grumbles about her behaviour. Euripides puts it beyond doubt that Electra’s affection for her husband is genuine: when the Peasant is nowhere in sight, she comes to his defence against the suspicious questions of the ‘stranger’ (253–61). What I aim to show here is that, in spite of this affection, there is in fact

84

The Linguistic Characterization of Electra

a fundamental lack of understanding between Electra and her husband. When she comes out of the house at her first entrance, Electra indicates exactly what she is doing (55–9): τόδ᾽ ἄγγος τῷδ᾽ ἐφεδρεῦον κάρᾳ φέρουσα πηγὰς ποταμίας μετέρχομαι— οὐ δή τι χρείας ἐς τοσόνδ᾽ ἀφιγμένη, ἀλλ᾽ ὡς ὕβριν δείξωμεν Αἰγίσθου θεοῖς— γόους τ᾽ ἀφίημ’ αἰθέρ᾽ ἐς μέγαν πατρί.6 carrying this pitcher set on my head I go to fetch water from the stream— not because I am lacking to such an extent, but in order to exhibit the brutality of Aegisthus to the gods—and I send laments for my father to the broad heaven.

This is the initial pronouncement of two closely linked concerns for Electra, to which she will return several times in the play: on the one hand her personal degraded estate, brought on by Aegisthus’ marital scheming, and on the other the unresolved, unavenged death of her father, for which she will hold her mother principally responsible. She notes with some force (δή 47) that her display has not so much to do with extreme deprivation as with the reason why she is in the circumstances that she finds herself in (Aegisthus’ hubris). This, she claims, is the reason for her display, together with the unavenged murder of her father. The last few lines of Electra’s opening speech name the other culprit responsible for her expulsion from the palace (60–3): ἡ γὰρ πανώλης Τυνδαρίς, μήτηρ ἐμή, ἐξέβαλέ μ᾽ οἴκων, χάριτα τιθεμένη πόσει. τεκοῦσα δ᾽ ἄλλους παῖδας Αἰγίσθῳ πάρα πάρεργ᾽ Ὀρέστην κἀμὲ ποιεῖται δόμων. 6 Line 59 should not be moved to follow 56 (pace Diggle app. crit.), nor should it be taken as inauthentic (pace Bain 1977a: 33 n. 3, Sansone 1984: 33, Distilo ad loc., and see also Collard 1984: 13). No probable motive for interpolation has been offered, and Diggle’s line order disturbs the sense (so Sansone 1984: 337 and Cropp ad loc.). Concerning the argument that 57–9 should all be deleted (so Kirchhoff, Wecklein), see ch. I, n. 36. Simple emendation will fix the paradosis ἀφίην: Reiske’s ἀφίημ’, combined with Radermacher’s punctuation (Radermacher 1899: 701–2, taking 57–8 as parenthesis), is probably to be preferred over Portus’ optative ἀφείην (defended by Slings 1997c: 134–5; but the combination of a shift from subjunctive to optative and an optative in primary sequence—both unusual, if not exceptional features—is perhaps too awkward to accept).

The Linguistic Characterization of Electra

85

The baleful daughter of Tyndareus, my mother, cast me out of the house as a favour to her husband. And by bearing other children to Aegisthus she ensures that Orestes and I have inferior status in the palace.

Electra explains (γάρ 60)7 how her loss of status was effected and how it is still being made worse. Although the hubris she has suffered originates in the first instance with Aegisthus, Electra points out that it is her mother who has facilitated it and who is compounding it by producing heirs and thereby securing Aegisthus’ claim to the kingdom. Finally, Electra does not fail to mention that her mother’s political gambit affects her brother Orestes as much as it does herself (Ὀρέστην κἀμέ 63). Thus, in her first few lines, Electra has explained why she behaves in the way she does, what her exact concerns are (the loss of her status and that of Orestes, the unavenged death of her father), and who the people are whom she holds responsible. This initial expression of concerns will prove to be a remarkably accurate guide in explaining Electra’s conduct throughout the play. As we saw in the previous chapter (§3.2), the Peasant’s reaction to Electra’s speech (64–6) is in essence a very gentle rebuke, an attempt to get his wife to change her behaviour, politely packaged as a question. Electra’s refusal to comply with her husband’s wishes (67–76) constitutes a dispreferred second part* in an adjacency pair*.8 Accordingly, Electra’s turn is highly elaborate: it is introduced by four lines of extremely polite high praise (67–70) and develops into an extensive explanation for her insistence on water carrying (71–6). These are the first words that Electra addresses to her husband, and they reveal much about her sympathetic attitude towards him. Still, not everything seems right. For one thing, notwithstanding the extent to which Electra cushions her words through praise and hedging, her turn is in some ways not a response at all. The Peasant has offered her an ‘out’ by using an off-record* strategy, and Electra makes full advantage of it by ‘answering’ only the Peasant’s ‘question’, while doing nothing to respond to the reproach encoded in that question. A key phrase, κἀκέλευστον (71, ‘even when not asked to’), underscores the strangeness of her answer. With it, Electra seems to If line 59 is sound (see previous note), γάρ here has a general explanatory function rather than explaining ὕβριν in 58 alone. Such a reading is unproblematic and gives no reason to question the place or authenticity of 59. 8 But see ch. I, n. 42. 7

86

The Linguistic Characterization of Electra

suggest that the Peasant is merely arguing that she should not have to do work unless requested to do so; but the Peasant’s ἐμοῦ λέγοντος (46) is indicative of a stronger stance on his part. More remarkable still is the fact that Electra makes no mention at all to the Peasant of the reasons she had given earlier (57–9), but instead says that she feels compelled to lighten his labour (ὡς ῥᾷον φέρῃς 72, ‘so that you may bear it more lightly’). Scholars have been troubled by the apparent inconsistencies in this dialogue: the Peasant is asking Electra about her motivations (64–6) even though she has already listed them (57–9), and Electra first gives one explanation for her actions, then an entirely different one (57–9, 71–6).9 In fact these inconsistencies are highly significant and suggestive of a fundamental lack of understanding between husband and wife. The Peasant, even though he has heard Electra state her reasons, can only interpret her action at the level of the mundane and fails to understand why, for a woman of Electra’s stature (εὖ τεθραμμένη 65), it is necessary to do this kind of work on his behalf (ἐμὴν . . . χάριν 64).10 Paraphrasing his turn (and his use of γάρ at 64)11 very elaborately, we might interpret his question as: ‘That’s all well and good, but I still don’t see why someone like you feels the need to do this for me.’ This is an opening, for Electra, both to ignore (as we have seen) the reproach inherent in the Peasant’s words and to formulate her response at the same level of the everyday, as if she undertakes her water-carrying expedition simply to help her husband in his everyday tasks. Cropp (on 54–81) rightly argues that Electra’s reasons are ‘not so much contradictory as on different levels’.12 The crucial point, 9 For proposed ‘solutions’ involving interventions in the text or staging, see ch. I, n. 36. 10 Without wanting to read too much either into the hyperbaton of ἐμήν and χάριν (on which see Devine and Stephens 2000: 108–11) or into the fact that the possessive adjective ἐμήν precedes its head χάριν (on which see Dik 2007: 116–19), I think it is plausible that ἐμήν carries emphasis here. 11 On γάρ here, see ch. I, n. 37. 12 For Denniston (in commentary on 71–3), ‘Electra’s motives are, in fact, mixed. She wishes to pour out her emotions in solitude: but she also wants to feel that she is doing something to help her husband.’ For Michelini (1987: 191), Electra ‘wishes, quite naturally, to equal his moral nobility by helping him, rather than playing the aristocrat at his expense’. Lloyd (1986: 7) calls her justification ‘convincing’. Of course we cannot know whether Electra is actually motivated by a desire to lighten the Peasant’s labours (these words, unlike those addressed to Night at 57–9, are to be seen in the context of the interaction between husband and wife), but the genuineness of her affection for him is guaranteed by her later conversations with Orestes.

The Linguistic Characterization of Electra

87

however, is that the Peasant does not appear to have access to one of those levels. For Electra, the disparity between her birth and her current predicament, a point raised by the Peasant as something that should dissuade her from going out, is in fact the very reason (since it is the result of Aegisthus’ hubris) why she feels she must: but she does not say this to her husband. Electra, in short, not only resists the Peasant’s admonitions, but refuses to engage with them at a more fundamental level. The two characters are communicating, they are even communicating affectionately, but not really about the same thing. This communicative muddle is suggestive of a fundamental gulf between the two characters: they operate from irreconcilably different perspectives on what Electra’s role, given her circumstances, should be, and neither side is able or willing to adopt the other’s view. As kind as he is to her (and as much as she appreciates him), the Peasant never understands Electra’s motivations. The gap is only heightened when the Peasant consents (εἴ τοι δοκεῖ σοι 77, ‘if you think so’), noting—perhaps somewhat naively—that the stream is not far anyway. Cropp’s (1988) comment on lines 77–8 is sensitive: There is some humour as the Peasant brings El[ectra]’s hardships into perspective, but this reflects on his unimaginativeness too, rather than simply undercutting her view of the situation . . . The nature of her labour is more important than the amount of it, both to her (exhibiting affliction . . . ) and to the Peasant (objecting to her demeaning herself; his objection shows that her own attitude is not frivolous).

Moreover, as I will argue presently, it would hardly have been suitable for Electra to be in full lamentation mode around her husband: that is something she reserves for the company of women.

2.3. Electra as Argive ‘maiden’ The ‘unimaginativeness’ of the Peasant is very closely mirrored by that of the Chorus 100 lines later, in the parodos. The women of the Chorus have suggested that Electra accompany them to a festival of Hera (167–74), but she shows herself unwilling to join them (175–87):13

13

See my discussion of (171–80) in the Introduction, §2.3.

88

The Linguistic Characterization of Electra οὐκ ἐπ᾽ ἀγλαΐαις, φίλαι θυμὸν οὐδ᾽ ἐπὶ χρυσέοις ὅρμοις ἐκπεπόταμαι τάλαιν᾽ ... σκέψαι μου πιναρὰν κόμαν καὶ τρύχη τάδ᾽ ἐμῶν πέπλων, εἰ πρέποντ᾽ Ἀγαμέμνονος κούρᾳ . . .

Not for fineries or golden necklaces, my friends, have I, wretched as I am, taken flight in my heart. . . . Look at my squalid hair, and these tattered clothes, see if they befit a daughter of Agamemnon.

Again, Electra refuses to go along with someone who is trying to disrupt her pattern of lamentation: again a dispreferred turn*— rejecting an invitation—couched in affectionate terms (φίλαι 175) marks the moment. And, once again, Electra provides a justification for her refusal, a justification not fully understood by the Chorus. It is significant, in this regard, that Electra is technically not complaining about a ‘lack of raiment for the feast to which the Chorus bids her’ (Conacher 1967: 205, my italics), but stating that she does not care for such accoutrements; this is surely how ἐκπέτομαι ἐπί with the dative is to be interpreted.14 Together with the remainder of the strophe, this is again a broad reflection by Electra upon her wretched estate and loss of royal status through her sham marriage; and it is this situation, as well as her lament over her father—in other words, a complex of reasons—that prevents her from attending the festival (as she herself will repeat to Orestes at 310–11; see §3.2).15 All this is to say that, as noted in my Introduction (§2.3.1), the Chorus’ invitation to Electra misfires*: the felicity conditions* in which such an 14

Pace also Arnott 1981: 185 and Kovacs 1985: 309. Both argue along the same lines as Conacher (though in Kovacs’ case not in order to attack Electra). The ‘lack of clothing’ argument was effectively deconstructed by Seaford (1985: 319 n. 38): ‘It is not true that “lack of clothing was precisely the reason she gave the Chorus in 175–9 for declining their invitation” [Kovacs] to the festival. The reason she gave there was lamentation . . . And so the chorus’ offer of fine clothes (189ff.) is of course not taken up.’ 15 See Michelini 1987: 192, where the idea that Electra’s refusal is ‘evidence of bad faith’ on her part is rightly rejected, as is a comparison with Sophocles’ Electra: ‘Would Sophokles’ Elektra have accepted the loan of a dress from a friendly chorus? Even to ask the question is to answer it: such offers do not come to the heroines of heroic tragedies.’

The Linguistic Characterization of Electra

89

invitation could ever be successful have, according to Electra at least, not been met. The women of the Chorus, however, display as narrow an understanding of Electra’s behaviour as the Peasant has done before, and the failure of their communication becomes all the more apparent in the next exchange. First the women offer to lend Electra a dress and jewellery (190–2): ‘as if that was the real trouble!’, Grube (1941: 301) rightly exclaims. When they next produce the advice to stop groaning and to turn to prayers and reverence instead (193–7), Electra is able to pivot the topic of conversation away from what the women had intended: none of the gods is listening to her laments (οὐδεὶς θεῶν . . . κλύει 198), and none pays attention to the misfortunes of Agamemnon and Orestes. Electra’s failure to achieve effective mutual communication with the Chorus is also reflected in a more fundamental aspect of her language, namely in the fact that she sings. In a recent book dealing with female lyric in Euripides, Chong-Gossard has argued that the lyric mode can express various communicative ‘semantics’, among which he identifies the notion of resistance:16 The lyrical voicing of [a woman’s] powerlessness gives her a strange kind of control, in that Euripides’ heroines act with authority when they resist the predominant expecations of how they should behave— whether that means giving in to someone else’s understanding of an event, or giving up lamentations and moving forward with their lives. (Chong-Gossard 2008: 68)

Besides Electra herself, Chong-Gossard (2008: 70) offers three instructive parallels for what he calls ‘an independent tragic verbal genre’: Hypsipyle in her fragmentary play (especially Hyps. frr 752f.29–31 and 752g.3–17), Alcestis in the eponymous play (Alc. 244–79), and Hermione in Andromache (Andr. 825–65). All these women, Chong-Gossard argues, attempt ‘to hold their visions of the world against the dominant opinion’ and refuse to be comforted or distracted from their thoughts (Chong-Gossard 2008: 88–9). Though we might argue with Chong-Gossard’s (2008: 69) apparent assertion that a refusal to be comforted is the only reaction to 16 I should note that, although I find Chong-Gossard’s analysis of the functions of female lyric often convincing, I disagree with his interpretation of Electra’s character. I will return to this point in §2.4.2.

90

The Linguistic Characterization of Electra

potential comforters that ‘Euripides chose to convey in song’,17 his analysis is useful to the extent that the lyric mode, when used in exchanges (epirrhematic and lyrical amoibaia, as distinct from monodies and choral odes), does regularly seem to portray female characters in Euripides as operating on a somewhat distinct plane, as not fully ‘connected’ in communication.18 Thus Electra’s lyrics in the early scenes of the play seem to serve two distinct, if related, functions: in her monody (before the entrance of the Chorus, 112–66), the traditional identification of solo song with lamentation explains her use of lyrics;19 after the Chorus’ entrance, however, the fact that she continues to sing in the parodos—which is thus ‘linked metrically with the monody’ (Cropp on lines 167–212)—suggests that she has not left her posture of lamentation behind, and in fact does not want to do so.

2.4. The characterization of Electra 2.4.1. Patterns of miscommunication Thus a pattern emerges in the way in which Electra communicates— or rather fails to communicate—with her husband and the Chorus. Neither of her conversational partners is able to engage fully with what she is saying, or to have her engage fully with what they are saying. Yet, in spite of these snags in their communication, there is a remarkable continuity in the mutual expression of sympathy between Electra and her husband and between Electra and the Chorus. This sympathy can be seen clearly in the various forms of address employed by them, even if each of these addresses has a more specific significance in its context as well. When the Peasant calls Electra ὦ δύστην᾽ (64, ‘unhappy woman’), it is a gesture of true sympathy, though also one particularly suitable here because it signals the Peasant’s awareness of Electra’s fall from grace and thus underpins his attempts to exempt her from menial labour. When the women of 17

Helen accepts the advice of the Chorus in song at Eur. Hel. 330. This feature of lyric is not limited to Euripides: for example, when Cassandra sings in Aeschylus’ Agamemnon (1072–177), this also marks a lack of connection between herself and the Chorus, who fail repeatedly to comprehend her fully (see Fraenkel ad locc.). 19 See ch. II, n. 2 for references on this issue. 18

The Linguistic Characterization of Electra

91

the Chorus end their part of the amoibaion with ὦ παῖ (197, ‘child’), this is surely a term of affection, though we may also detect in it a mild assertion of superiority, adding weight to the advice.20 The women remain sympathetic even after they are rebuffed, and for good measure throw in a swipe at Helen (213–14), which amounts to an endorsement of Electra. Opinion has varied about the motivation for the sudden mention of Helen here, but, whatever her relevance to Electra’s predicament,21 the Chorus’ sudden change of topic may in fact be suggestive, at a different level, of the extent to which the women are willing to humour Electra: by saying something that is irrelevant to the immediately preceding dialogue about the festival, the women make it clear that they are willing to let that issue (and their invitation) rest.22 Again, Electra’s stubbornness prevails. For her part, Electra answers the Peasant’s question only after offering him extensive praise (67–70), and calls the women φίλαι even when rejecting them (175):23 in both these cases it is possible to argue that the affection offered is designed to compensate for the fact that Electra is ignoring their wishes (positive politeness*), but it is affection nonetheless. A difference between the two situations should be noted as well: whereas Electra’s response to the Peasant pivots the topic of discussion away from Electra’s misery (to her wifely helpfulness), when talking to the Chorus Electra exerts a great deal of topic management* to keep her misery and lamentation at the centre of attention. That she does so is no doubt related to the fact that the Chorus is a group of women: lamentation, after all, is behaviour apparently reserved for

20 It is hard to say whether the superiority implied here is one of age as well as of wisdom: Denniston (1939: xxxi–xxxii, with n. 1) believes that the Chorus consists of women older than Electra (this allows them to use the address ὦ παῖ). But Zeitlin rightly points to the Messenger’s address ὦ καλλίνικοι παρθένοι Μυκηνίδες (761, ‘Mycenaean maidens, glorious in victory’), which would seem to rule out married women (note also παρθενικαί, ‘maidenly’, used by the women themselves at 174). 21 Helen’s relevance to Electra’s predicament is in fact pervasive. As the ‘original adulteress’, Helen is relevant to her sister’s adultery with Aegisthus, mentioned just previously by Electra (211–12); through her elopement she is responsible for the ills that beset the house of Atreus. This moment also prepares for the revelation, at 1283, that Helen did not go to Troy at all. All these points are mentioned by Cropp ad loc. and by Slings (1997c: 139). 22 Cf. Grice’s ‘tea party’ example discussed in the Introduction, §2.3.2. 23 Again at 272, 747, 751, 871, 1230, and 1230.

92

The Linguistic Characterization of Electra

single-sex groups* (female only), whether they have shown a particular desire to listen to it or not. We are now able to draw a fuller picture of these scenes: the Electra–Peasant and Electra–Chorus pairs engage in communication that, in spite of consistent affection from all sides,24 fails to be fully effective, largely because of Electra’s refusal to come to terms with her new situation. The communicative gulf works both ways, however: the fact that the Peasant and the Chorus fail not only to effect any change in Electra’s behaviour, but also to comprehend it fully is crucial. In the diverging perspectives of Electra on the one hand and the Peasant and the Chorus on the other, we see reflected the inexorable divide between Electra’s proper social status as a princess and her actual life in what amounts to near slavery.25 It is significant, in this regard, that the Peasant and the Chorus effectively represent Electra’s misfortunes: her faux husband is the symbol of her social banishment; the members of the Chorus—virginal maidens (παρθενικαί 174) attending a festival that Electra, as a ‘married’ woman, even if secretly still virginal, cannot attend—are a strong reminder of the extent to which that banishment affects her.

2.4.2. Electra’s character: the debate With this we come to one of the major points of scholarly contention about Electra: how are we to interpret Electra’s stance of resistance here? A persistent reading would see Electra as a delusional woman, revelling in her self-imposed, exaggerated misery. Some damning comments about her have already been cited (see Introduction §4.1.2); a litany of further ones could be added, but let it suffice to hint at the chronological range of such views by citing Wilamowitz’s (1883: 230) verdict—‘a rotten soul’ (ein[e] verdorben[e] Seele)—and that of Chong-Gossard (2008: 72) in the contribution already discussed (§2.3)—‘it is very difficult for anyone to feel sorry for her. Her miseries are self-imposed, part of her own strange and skewed vision

24 There is no indication whatsoever in the play that Electra ‘alienates . . . her chorus of women’ (pace Chong-Gossard 2008: 72): even when they blame her in the kommos for egging on her brother (1201–5), they remain affectionate (see §6.1). 25 It is significant that Orestes, upon first seeing Electra, describes her as nothing more than πρόσπολός τις (107, ‘some servant’) and a δούλη γυνή (110, ‘a slave woman’), on the basis of her physical appearance.

The Linguistic Characterization of Electra

93

of the world’.26 The two moments in the play just discussed serve as some of the prime exhibits in the case against Electra. Conacher, for example, sees things as follows: Electra’s complaint about carrying water (55 ff.) is immediately shown up by her gentle peasant-husband who says, first, that he has told her not to do it (64–66), and second, that it’s a very slight labour in any case (77–78). Electra’s complaint of lack of raiment for the feast to which the Chorus bids her is immediately shown up by her refusal of glorious clothes the Chorus offers. In both cases, it is the chance of self-martyrdom which Electra is seizing . . . In the first episode, Electra repeats each of these complaints and hypocrisies in her ‘message’ to Orestes (300–38): to the specific repetitions about the water-carrying and the lack of raiment (304–9), she now adds the exaggeration, ‘Banned from the festal rites . . . !’ (310) for which we have been well prepared. (Conacher 1967: 205)

But the notion of self-inflicted misery (and, to a lesser extent, that of self-martyrdom) is misplaced, and Electra’s vision of the world is not so much ‘skewed’ as it is incompatible with that of the Peasant and the Chorus.27 Electra’s problems, on the contrary, are very real: she is a princess who was destined to be married to a demigod (312–13) but has now been deprived of palace life (306) and of a successful marriage (being joined instead to a poor man in what she considers ‘a deathly marriage’, θανάσιμος γάμος 247).28 All this is the direct result of Aegisthus’ hubris against her, and that hubris is undeniably

26 Further examples include Steiger 1897: 22–3, Masqueray 1908: 139, Steiger 1912: 573–4, Sheppard 1918: 139, Hunger 1936: 11, Grube 1941: 303, Blaiklock 1952: passim, Stoessl 1956: 61, Kitto 1961: 333–4, O’Brien 1964, Conacher 1967: 204–6, Solmsen 1967: 40, Ronnet 1975: 69, Knox 1970: 72, Arnott 1981: 185–6, Tarkow 1981: 151–2, Burian 1997: 180, Raeburn 2000, and Yoon 2012: 74–7, 99–105. Decidedly contra is Lloyd 1986: 2–10. More nuanced defences can be found in Zürcher 1947: 121–2, Steidle 1968: 66–8, Zeitlin 1970, Cropp 2013: 8–11 and passim, Mossman 2001: 376–7, Papadimitropoulos 2008, and Ormand 2009. An outstanding analysis, in my view, is that of Michelini (1987: 187–94), who places Electra squarely in the context of both the tragic tradition and contemporary Athenian attitudes, while exposing the dangers of over-applying modern psychological concepts to her personality. 27 See Mossman 2001: 377: ‘What is wrong with Electra and Orestes is not that they are not heroic, but that their heroism is horribly misdirected.’ 28 See Steidle 1968: 66, Zeitlin 1970: 650 n. 21, Michelini 1987: 193, and Cropp on 247. For a discussion of Electra’s exile in the context of Athenian legal and religious practice, see Ormand 2009: 250–8.

94

The Linguistic Characterization of Electra

real,29 whatever we may think about her insistence on exhibiting it. That ostentatious display, moreover, has been argued to fit right into Greek cultural norms: It was appropriate for someone suffering ὕβρις . . . not to suffer in silence but to make a show of distress. The point of this was . . . to demonstrate that a crime was actually taking place; if one does not complain, then no one can tell that one is being wronged. (Lloyd 1986: 3)30

As for Conacher’s ‘hypocrisies’, I will discuss Electra’s rhēsis at 300–38 in more detail in §3.2, but a few points are worth making here already. First, Electra has never said that she lacks clothing, and there is no need to doubt that her current circumstances force her to weave her own (which is her only claim at 307–9). As for her water carrying, Electra can hardly be accused of hypocrisy after she has shown herself well aware that she is not doing it out of ‘need’ (χρεία; 57–8): if she carries water to exhibit Aegisthus’ hubris, then it is surely one of the things she ‘should’ mention in a message designed to portray the severity of her troubles. And when it comes to her lack of participation in festivals, we may note that Conacher in his skewering of Electra ignores the complexity of her situation as a ‘liminal figure caught between the worlds of parthenos and gunē’ (Swift 2010: 32):31 Electra herself elaborates on this issue in the next 29 This is virtually guaranteed by the reception of Aegisthus’ behaviour within the play. No other character contradicts Electra’s position: the Peasant himself calls his marriage δυστυχής (49, ‘unfortunate’), Orestes sees the need to groan over it (ᾤμωξ᾽ 248) and describes it as an act of hubris (ὕβρισ᾽ 266), and even Clytemnestra does not deny that Aegisthus has mistreated Electra (she ascribes it to his ‘manner’: τρόποι τοιοῦτοι 1117). 30 Lloyd (1986: 3) rightly refers to Fraenkel on Aesch. Ag. 1317. See also Michelini 1987: 188 and Steidle 1968: 67; the latter argues that ‘this behaviour on the part of a woman is fully comprehensible given the degradation of the royal daughter, at least in the eyes of an ancient spectator’ (dies Verhalten von Seiten einer Frau [ist] zumindest in den Augen eines Antiken Beurteilers angesichts der Degradation der Königstochter wohl begreiflich). 31 On this point—Electra’s ‘painfully anomalous status as a virginal matron’ (Michelini 1987: 192)—see the fuller discussion in Swift 2010: 189–91, and Zeitlin 1970: 650, with nn 21–3. Part of the problem of Conacher’s representation is his translation ‘Banned from the festal rites’, which makes too much, I think, of Electra’s ἀνέορτος ἱερῶν καὶ χορῶν τητωμένη (310): τητάομαι can refer to deprivation caused by human intervention (e.g. Soph. Phil. 383, Eur. Heracl. 31; in such cases ‘banishment’ is an appropriate term), but also to deprivation caused more generally by circumstances, fate, or gods (e.g. Eur. Hel. 274). Since no agent is expressed in the present case, and considering the next line, it seems likely that Electra is referring here to her indeterminate marital status as the reason for being ‘deprived’ of the festivals, and not to

The Linguistic Characterization of Electra

95

line, where she outlines the reason why she cannot join the adult women either (ἀναίνομαι γυναῖκας οὖσα παρθένος 311, ‘as a maiden, I shun the women’). Moreover, each of these ‘hypocrisies’ forms part of a message that is specifically aimed at inciting Electra’s long-lost brother into making his glorious return, and should be seen in the light of this rhetorical purpose. Another accusation that has often been levelled against Electra is that she is selfish and egotistical, less focused on Agamemnon and her brother than on her personal mistreatment. Some scholars attempt to distil this feature even from the order in which Electra sings about her problems, and in which she mentions them later in her speech at 300–38. Grube goes so far as to label that order ‘disgraceful’: The order ‘My misery . . . my father’s murder’ seems wrong. . . . First her own misery and degradation (far more profound than she realizes), then her mother’s luxurious life, and last of all Agamemnon in his grave. (Grube 1941: 300, 303)

But this interpretation32 seems to me to be based on an unacceptable amount of selective reading. First of all, though Electra’s monody (112–66) begins with an ‘introduction’ of herself (115–19),33 the remainder of it is entirely devoted to her father and brother (she refers to herself only once, noting that Orestes’ absence left her ‘in a most terrible condition’, ἐπὶ συμφοραῖς ἀλγίσταισιν 133–4). Second, as I have argued on 55–63 (see §2.2), Electra’s concerns about her (and her brother’s) loss of status are, throughout the early scenes of

some non-existent human decree (which is what Conacher seems to read in τητωμένη; the ‘exaggeration’ is his own). 32 Expounded also, e.g. by Conacher (1967: 205) and O’Brien (1964: 29). ChongGossard (2008: 72–3) sees similar, but inverted, problems in the order of topics in the amoibaion: ‘Her final lyric response to the chorus—her finale—culminates in what she considers the worst aspect of her misfortune: not that she cannot avenge her father because Orestes has not returned, but . . . [quoting 207–12].’ One might ask here, which is it: is extra significance lent to what comes first or to what comes last? (The latter, ‘climactic’ reading of ordering would be more usual, but it would—as Lloyd (1986: 4–5) points out—lend emphasis to Electra’s mention of Agamemnon’s murder (!) in the monody and in the rhēsis at 300–38.) On the whole, I would suggest that not too much interpretative weight should be assigned to the order of topics in any of these passages. 33 The idea that Electra is characterized as self-absorbed simply because she begins her monody by singing about herself has been rightly dismissed by Lloyd (1986: 4), who points out, with ample parallels, the prologue-like quality of these lines, and by Cropp (on 112–66).

96

The Linguistic Characterization of Electra

the play, always wedded to the unresolved murder of her father: in her ‘message’ to Orestes (300–38), too, she gives equal time to both of these aspects (see 304–13, on her own situation; 318–31, on Agamemnon and Orestes). Considering Electra’s language throughout the play, she seems in fact to be almost excessively preoccupied with her father: to give but one indication, Electra uses the word πατήρ to refer to Agamemnon about once every 12 lines, as compared to Orestes (both in and out of disguise), who uses it about once every 22 lines (Sophocles’ Electra uses the word about once every 19 lines, his Orestes once every 48 lines).34 In addition, there is a more fundamental problem with such interpretations, which was laid bare by Michelini: A psychological interpretation of Elektra that rests upon the identification of ‘selfishness’ or ‘egotism’ in her personality is unlikely to be valid. There are no terms available in Greek culture to translate these notions, and analysis of Greek moral values has made clear the very important role of self-aggrandizement in almost all areas of the culture. Elektra has suffered a wound to her pride, and such a wound is a legitimate and honorable cause for revenge. (Michelini 1987: 188)

Electra’s problems, in short, are neither invented nor self-inflicted, nor is she—initially, at least—unreasonably concerned with one aspect of them over the other. All this is not to say that she is therefore automatically and straightforwardly ‘sympathetic’—as Lloyd (1986: 19) would have it. Rather, the continued merging of the issues of personal degradation with the fortunes of Agamemnon and Orestes means that, as Cropp puts it, the personal grievances come to be seen as an integral element in [the claims of Agamemnon and of justice], feeding her vengefulness and inducing that single-minded extremity of hatred which leads to matricide. (Cropp 1988: xxxvi)

As we will see, the balance between Electra’s motivations begins to shift as the play progresses and her unadulterated hatred for Aegisthus and Clytemnestra becomes more and more central. This calibrates our

‘Electra emphasizes home and father; these are the terms through which she understands her social identity and her current state’ (Ormand 2009: 253). For Electra’s preoccupation with her father, see also 934–5, 1102–4, and my discussion of 974–8 in ch. IV, §3.2. 34

The Linguistic Characterization of Electra

97

sympathy, until the matricide reveals the terrible consequences of Electra’s rancour. What we see on stage in these early scenes is, in short, a woman who is condemned to an estate that is not her own but has been thrust upon her illegitimately. She violently resists conforming to her new role—in fact she does all she can to present an image of herself that does not conform35—and sympathetic offers of help and consolation from the Peasant and the Chorus women cannot alleviate her situation because they are, for Electra, besides the point. Her refusal to accept her new status combines with the other items of ‘unfinished business’ (Agamemnon going unavenged, Orestes’ exile) to determine her behaviour—including her linguistic behaviour—throughout the play, culminating in the matricide and her own realization that her hatred has taken her too far. To sum up the points made thus far: Electra is individualized in her language by the inability and refusal to communicate fully with those who effectively represent her loss of status, much though there is mutual sympathy between her and these characters. Her resistance to adapting to her new life is continually visible in a series of misunderstandings and communicative misfires, as well as in the use of the lyric mode in her exchange with the Chorus. An additional characteristic of Electra’s language, which brands it in these scenes and will play a significant role throughout the play, is that it is marked as typically female. This is, of course, hardly a bold claim, given the concerns and activities that occupy Electra in the scenes treated thus far: lamentation, easily the most prototypical form of gendered communication in Greek culture,36 and her social status as determined by marriage, which ‘gives a woman her place in social existence’ (Michelini 1987: 193). As I will show throughout the 35

Michelini has helpfully analysed the disparity between Electra’s life in the country and her former life in the palace in terms of genre and tradition: ‘we might say that Elektra sees herself as a conventional “serious” figure; but her circumstances thwart her, by forcing the audience to be aware of the concreteness and normality of Elektra’s daily life . . . social determinants have placed her in the wrong class’: thus Electra is ‘a tragic play that includes and assimilates the comic universe’ (Michelini 1987: 192–3, 230). Similar in approach is a reading by Thévenet (2009: 315–20), who argues that Electra is concerned with presenting herself as ‘ “la malheureuse Électre” de la tradition’, though she is not in the right play for that. All this accords well with the innovative approach to tradition Euripides adopts in this play, for which see Introduction, §4.3. 36 See Alexiou 2002; but see also Suter 2008b on the importance of male lament.

98

The Linguistic Characterization of Electra

remainder of this chapter, however, Electra’s gender also conditions her language in more subtle ways in the later parts of the play.

3. ELECTRA AND HER UNEXPECTED GUEST (215–338) I go on to examine the conversation between Electra and Orestes’ ‘messenger’ more closely, to see how Electra’s linguistic behaviour in it bears out her preoccupations and the mismatch between her expected status and the reality of her situation.

3.1. The stichomythia (215–89) The extended stichomythia between Electra and Orestes—an ‘interrogation’, whose careful organization and many ironies have been analysed extensively by Schwinge (1968: 252–61, 295–317)37—begins with a charged dramatic moment (215–27)38 and then moves from one topic of discussion to the next, in fairly discrete sections of about ten lines. The first of these sections (228–36) deals with Orestes, who is wandering (according to his ‘messenger’) from one city to the next (234). Orestes’ significance to his sister as her long-hoped-for saviour is evident from various points. First, there is the complete reversal of Electra’s attitude towards the stranger39 upon receiving the news that he bears a message from Orestes.40 While she initially perceives the 37 See also Solmsen 1967: 10–14, Basta Donzelli 1978: 73–102, Lloyd 1986: 11–13, Michelini 1987: 207–8, Schuren 2014: 82–4, 192–3, and Cropp ad locc. 38 I have discussed this moment in the Introduction, §2.4.2. 39 Electra is ‘suddenly as if wholly changed’ (Schwinge 1968: 304, plötzlich wie umgewandelt). 40 Although the phrase σοῦ κασιγνήτου λόγους would naturally be taken to mean ‘a message from your brother’ (genitive of the subject), and not ‘word about your brother’ (so Denniston ad loc.), I am not convinced that the phrase precludes ambiguity, as is argued by Slings (1997c: 140) against Denniston’s reading; such ambiguity between ‘from’ and ‘about’ is detected by Schwinge (1968: 300 n. 58) and seems to me supported by Slings’ own parallel at Eur. IA 842. In any case, Slings is right to point out that a more pertinent ambiguity resides in the ironical fact that the formulator of the message (λόγους) is at the same time the speaker of these specific words (λόγους).

The Linguistic Characterization of Electra

99

two men as φωταὶ κακοῦργοι (219, ‘villains’), dismisses Orestes’ advances41 as illegitimate (μὴ . . . χρεών 223), and accuses him of lying in ambush (λοχᾷς 225), she changes tone dramatically immediately after his revelation. Electra’s greeting, ὦ φίλτατ᾽ (229, ‘dearest’), is significant in this respect.42 The superlative φίλτατε is, in tragedy,43 mostly reserved for immediate family members44 (it is thus ironically ‘truer than she knows’ here, Cropp ad loc.), and is occasionally used for body parts, important friends or protecting gods.45 However, as Gregor (1957) has shown, when φίλτατε is used to address a character who should be relatively unknown or insignificant to the speaker, the friendship term seems in every case to suggest that the speaker expects from the addressee some form of delivery from a dire situation (in such cases it is more the message that is φίλος than the messenger): thus Iolaus asks of Hyllus’ servant (Heracl. 640): ὦ φίλταθ᾽, ἥκεις ἆρα σωτὴρ νῷν βλάβης; Dearest one, have you come, then, to save us from injury?

And Electra herself, in a state of desperation, fearing that the revenge plot has failed, later in our play addresses the messenger who brings relief as φίλτατε (767).46 In short, the use of the address here, immediately after the word ‘brother’ (κασιγνήτου 227) has been uttered, underlines the importance of Orestes to Electra as a ‘liberator from hardships’ (λυτὴρ πόνων 135–6).47 After this strong reversal, Electra is 41 It is impossible to say, solely on the basis of the imperfective aspect of Electra’s imperative (μὴ ψαῦ’ 223), whether Orestes is actually touching Electra here, or even making a move to do so (both would allow the present imperative to have a ‘discontinuative’ sense): prohibitions with the present imperative can, but do not have to, have this meaning; see Slings 1997c: 140, with further references. 42 This is not an apostrophe, but an actual address to the stranger, pace Schwinge 1968: 308. 43 For the use of φίλτατε outside tragedy, see Dickey 1996 (index s.v.). 44 e.g. in Euripides, Theseus to Hippolytus (Hipp. 1452), one of Medea’s children to another (Med. 1272), Electra to Orestes (Or. 1045). 45 e.g. in Euripides, Medea to her hands and mouth (Med. 1071), Hippolytus to Artemis (Hipp. 1092), Orestes to Pylades (Or. 1100), Cadmus to Tiresias (Bacch. 178). 46 Cf. also, e.g. Eur. Heracl. 788, Hipp. 1452, Andr. 64, Hec. 505, 1114–15, Supp. 641, Phoen. 1072, Bacch. 178. 47 The address is similarly employed, also with ironical undertones, by Iphigenia in speaking to Orestes at IT 815—an instructive parallel, as 12 lines later (after recognition) she herself ‘reinterprets’ the term in her embrace: ὦ φίλτατ’, οὐδὲν ἄλλο, φίλτατος γὰρ εἶ (827, ‘dearest, and nothing else, for you are most dear to me’). Familial proximity and expectation of delivery seem to go hand in hand also in cases such as at Her. 531, where Megara addresses Heracles who has come to rescue her.

100

The Linguistic Characterization of Electra

for a few lines extremely polite (note especially the wish ‘may you prosper’, εὐδαιμονοίης 231) and remains friendly throughout their conversation, although a few of his remarks and questions seem to rankle her. In terms of conversational structure and turn-taking*, Electra retains the initiative throughout the brief section on Orestes’ circumstances by following up each answer with a new question (there are question–answer adjacency pairs* at 233–4, 235–6).48 At line 236, the stranger uses a general reflection (ἀσθενὴς . . . φεύγων ἀνήρ, ‘a man in exile is powerless’) to effectively cap Electra’s first line of questioning, which is a first indication that he is trying to effect some topic management*.49 Indeed, from this point in the stichomythia onwards, the stranger gains control of the discussion, so that it is here that the ‘interrogation’ properly starts. Electra, accordingly, yields the floor* in the next line (237) by inviting the stranger to deliver Orestes’ message: λόγον δὲ δὴ τίν᾽ ἦλθες ἐκ κείνου φέρων; What message, then, have you come to bring from him?

The stranger’s reply (238) does not contain much in the way of the requested λόγος but makes use of the offered initiative to query, on behalf of ‘Orestes’, Electra’s condition: εἰ ζῇς, ὅπως τε ζῶσα συμφορᾶς ἔχεις. Whether you live, and if alive, how you are faring.

To this query, the preferred*—and polite—conversational reaction would be for Electra simply to provide the messenger with information about her συμφορά. Electra’s response, however, is to suggest that the answer to his question should be evident to him without an explanation. Electra makes her point with two insert expansions*, a pair of questions50 whose answers should suggest to the stranger all 48 Of these, 235–6 are best seen as a post-expansion* on the sequence in 234–5, while 237 starts a properly new sequence (see also next note). 49 On this line, see ch. III, §2.2.1. The notion that 237 initiates a new sequence is borne out by the use of the particles δὲ δή (which indicates a shift to a new, pertinent topic). 50 I fail to see why editors print a full stop rather than a question mark in 241 (Murray did so already in 239): the connective καί and the lack of a verb make it clear that Electra picks up the syntax begun by οὔκουν ὁρᾷς, which cannot be anything but a question (even if it is a rhetorical one).

The Linguistic Characterization of Electra

101

he needs to know about Electra’s condition. The move proceeds in two parts: Electra points first to her degraded physical condition (οὔκουν ὁρᾷς μου πρῶτον ὡς ξηρὸν δέμας; 239, ‘Don’t you see, first, how withered my body is?’)51 and, second, to the outward signs of her lamentation (καὶ κρᾶτα πλόκαμόν τ᾽ ἐσκυθισμένον ξυρῷ; 241, ‘And how my head and hair are shorn by the razor blade?’). Her rhetorical questions do thus offer the information required to answer Orestes’ inquiry, but their phrasing (οὔκουν ὁρᾷς ὡς . . . ;) is not fully polite: by suggesting that the stranger should really be seeing something that, as it appears from his question, he isn’t seeing, Electra threatens his face* more than she needs to, and she offers little redress for this imposition. Questions introduced by οὐχ/οὔκουν ὁρᾷς tend to have this tone in Euripides and in Greek literature more generally.52 In short, Electra does not offer a restrained exposition of her circumstances here—those circumstances sting her, and the fact that the messenger even has to ask elicits a momentary flash of annoyance. What is there to make of Electra’s prickliness here? Clearly her sensitivity about her expulsion from the palace is on display again. This aspect provokes, as it were, a brief flare in her linguistic behaviour.53 It is also noteworthy that Electra’s reaction to a request for information about her condition (συμφορά 238) focuses on physical

51 The connotations of ‘dry’ (ξηρός) here are difficult to determine (Denniston ad loc. suspects that it means ‘lacking unguents’; Keene ad loc., Slings 1997c: 141, and Cropp ad loc. all prefer ‘withered’, and Cropp additionally suggests an intriguing allusion to contemporary medical thought about dryness associated with sexual inactivity). In any case, it seems best to take the following line as an explanation offered by the stranger rather than as a ‘new point’ (Denniston), in which case Heath’s γε is preferable (L has τε, but γε is ‘more in keeping with stichomythic style as a whole and the run of this part of the stichomythia’: Slings 1997c: 141). 52 The most immediate parallel is Eur. Heracl. 734 (Iolaus to the Servant, a mildly humorous instance)—the only other occurrence in tragedy with οὔκουν (as for the accentuation of this word, see Denniston ad loc., who is probably right to argue for οὔκουν, though the accentuation of οὔκουν/οὐκοῦν strikes me as an intractable problem). For οὐχ ὁρᾷς, cf. e.g. Eur. Andr. 349, Phoen. 713, Aesch. Ag. 1623, Soph. El. 973, and Ar. Plut. 257. One further Euripidean example, Phoen. 131 (Tutor to Antigone, pointing out another warrior), seems milder in tone. 53 One might claim that the lady doth protest too much here and is in fact overeager to point out her physical state: more generously, it is no doubt true that Electra’s emphasis on her situation serves her as a tool ‘to prove how necessary it is for her brother to come’ (Schwinge 1968: 306–7, mit dem sie beweist, wie notwendig es ist, daß ihr Bruder kommt).

102

The Linguistic Characterization of Electra

aspects,54 which may be linked to her gender: the physical display of grief was very much a woman’s role, and the loss of status that Electra faces is expressed—for a woman—above all in her physical circumstances. The stranger, meanwhile, correctly draws inferences from Electra’s appearance: she is consumed with grief and mourns her exiled brother and murdered father (240, 242). To both points, the stranger adds touches of politeness: with the first he expresses his sympathy (ὥστε με στένειν 240, ‘which makes me groan’; this is also, of course, a touch of dramatic irony), and with the second point he is careful not to phrase his conclusions too definitively (ἴσως 242, ‘perhaps’).55 Electra quickly confirms, with a sigh, that these are in fact the reasons for her appearance (243): οἴμοι· τί γάρ μοι τῶνδέ γ᾽ ἐστὶ φίλτερον; Alas! Yes, for what is dearer to me than these?

At this point, Euripides has his two protagonists engage in the the first of three exchanges that punctuate a ‘game full of surprises and frustrations’ of audience expectation (Solmsen 1967: 13). Electra’s expression of affection for her brother inadvertently stirs powerful emotions in Orestes, who gasps (φεῦ φεῦ 244; cf. 262, 282), but composes himself, and thus thwarts the resolution of the recognition that Euripides has tempted the audience to expect. I will have more to say about these ‘near misses’ in the next chapter (§2.2.2), where I will deal with Orestes’ linguistic behaviour in disguise. As for Electra, her repeated function in these exchanges is to stress the foiled recognition (and to facilitate a momentary receding of suspense), which she does by emphasizing that Orestes is far away (ἀπὼν ἐκεῖνος, οὐ παρών 245, ‘he is absent, not present’; cf. 263, 283 and 331). Some critics complain that these moments serve no purpose apart from toying with audience expectation and that they are poorly motivated.56 But Electra’s remarks at these points tell us something 54 Cropp ad loc. suggests that a medical connotation of the word συμφορά (for which he finds parallels at Alc. 1047–8 and Or. 153–4, with Willink ad loc.) steers Electra’s response (see n. 51). 55 See Denniston ad loc.: ‘He delicately hesitates to intrude overboldly on her feelings.’ 56 e.g. Parmentier and Grégoire 1925: 183 n. 4, Pohlenz 1930: 128, Wuhrmann 1940: 49, Grube 1941: 302–3, Garzya 1962: 85, and O’Brien 1964: 28. The criticism focuses mainly on Orestes: I have discussed the issue in the Introduction, §4.1.2.

The Linguistic Characterization of Electra

103

about her as well: the tautological phrasing of 245 (ἀπὼν . . . οὐ παρών) very probably conveys a tinge of ‘bitterness’ (Denniston ad loc.)57 and has been rightly seen to function as ‘an implied rebuke’ (Michelini 1987: 208) designed to spur Orestes into action through his messenger (this will be a recurring feature of Electra’s behaviour in the scene). Moreover, Electra’s repeated stress on Orestes’ absence reveals a gap not only between her knowledge and that of the audience, but also between her expectations and what the audience can see. In an influential article, Arnott (1981) argued that this ‘double vision’—the discrepancy between Electra’s own stated view of characters and things and the way in which they are presented elsewhere and by others—is one of the key themes of the Electra. Although I disagree with many of Arnott’s conclusions and with his general treatment of the siblings,58 the notion of ‘double vision’ seems to me useful in discussing Electra’s stance on Orestes in this scene and immediately before the recognition (esp. 524–8; see §4). Throughout these pre-recognition scenes, Electra’s language about her absent

57 The use of tautologies is a topic that has stirred a remarkable amount of controversy among linguists (see, e.g. Gimbel and Bulhof 2004, Meibauer 2008), in part because they do not always translate well into different languages. Some care should therefore be taken before simply believing our ‘intuitions’ about the force of a tautology in Greek. In this case, however, it seems not a stretch to say that Electra uses the formulation to emphasize her displeasure with Orestes’ absence (about which she has complained before, 130–4). Gricean pragmatics* sheds light here: by saying more than she strictly has to—ἀπών already logically entails οὐ παρών—Electra invites her utterance to be interpreted as relevant in some other way. The implication of bitterness is then easily available in this context, especially since a negation often suggests that the alternative (παρών) is the desired state of affairs. 58 Besides Orestes, Arnott extensively discusses Electra’s portrayal of herself, Clytemnestra, and Aegisthus. He argues that the ‘double vision’ regarding each of these characters effects the ‘deglamorization’ of Electra and Orestes (Arnott 1981: 181) and thus undermines the heroic moral code that they stand for—the Iliadic moral values that Electra expounds are undermined as her assessment of characters and situations is shown to be false. Although I believe that Euripides in this play explores the incompatibility of the heroic value system with democratic society (see Introduction, §4.2), I think that Arnott goes too far in his condemnation of the siblings. My objections are as follows. (1) Arnott’s notion of a ‘double vision’ is applicable much better to Electra’s expectations about Orestes than to her assessments of Aegisthus, Clytemnestra, and herself. I reject Arnott’s (1981: 184) claim that ‘Euripides unambiguously presents [Electra] as a liar’ about these characters, which cannot be substantiated (see also n. 154 below). (2) I do not believe that Euripides is very different from Aeschylus and Sophocles in treating the matricide as problematic, though he is different from Aeschylus in his focus on Electra’s situation and emotions. See Cropp (2013: 23–5).

104

The Linguistic Characterization of Electra

brother envisages an athletic young saviour, who will be quick and decisive in taking down Clytemnestra and Aegisthus. This image— which will return in full force in the part of the play dealing with Aegisthus’ death—arises most vividly at three moments: first at 275, where Electra rebukes the stranger for doubting Orestes’ next move upon arrival in Argos (ἤρου τόδ᾽; αἰσχρόν γ᾽ εἶπας· οὐ γὰρ νῦν ἀκμή; ‘You ask this? That’s a shameful thing to say! Isn’t now the right time?’—the strongest outburst of the stichomythia);59 next at 336–8, where Electra incites Orestes into action by pointing out the advantages of his youth and lineage (αἰσχρὸν . . . εἰ | . . . ἄνδρ᾽ ἕν᾽ . . . οὐ δυνήσεται κτανεῖν | νέος πεφυκὼς κἀξ ἀμείνονος πατρός, ‘it’s a shameful thing if he is going to be incapable of killing a single man, even though he is young and from a better father’); and lastly at 524–8, where Electra similarly rejects the Old Man’s suggestion that Orestes would enter Argos furtively and remarks in passing on Orestes’ nobility and athletic training (οὐκ ἄξι᾽ ἀνδρός . . . σοφοῦ λέγεις, | εἰ κρυπτὸν ἐς γῆν τήνδ᾽ . . . | δοκεῖς ἀδελφὸν τὸν ἐμὸν εὐθαρσῆ μολεῖν, ‘what you say is not seemly for a wise man, if you think that my brother, brave as he is, would come to this country in secret’, and πλόκος | . . . παλαίστραις ἀνδρὸς εὐγενοῦς τραφείς, ‘a lock, tended in a nobleman’s wrestling halls’). These moments bring into the spotlight the poor fit between Electra’s expectation of delivery from Orestes and what he has actually delivered in the early parts of the play. As a result, the moment has been taken by Arnott and others as the ultimate proof that Electra glamourizes what should not be glamourized and that ‘Orestes is a coward’ (Adams 1935: 120). I have laid out my objections to such views in the Introduction (§4.1.2); here I would suggest merely that these three moments should be seen in conjunction with Electra’s utterances at the near misses in the stichomythia. Taken together, these moments serve not only to create powerful touches of dramatic irony (making the eventual recognition all the more exciting), but 59 More precisely, the only such outburst after 228; up to that point the tone of the stichomythia is, of course, very different. Together with line 227, 275 is the only line that consists of three separate utterances, which may here be suggestive of a harsh tone (each utterance is as bald* as can be). There can be no question that the words αἰσχρόν γ’ εἶπας are strongly impolite: Electra offers no redress for this outright attack on the stranger’s face* (see Brown and Levinson 1987: 66 and Culpeper 2011, esp. ch. 4), and the key criticism implicit in αἰσχρόν is intensified by γε. Electra’s similar rebuke of the Old Man at 524–6, if still strong, is significantly more polite: see §4.

The Linguistic Characterization of Electra

105

also to reveal something about Electra’s character: her language is ‘bitter’ when it comes to Orestes’ absence, and gets a sharper edge (impolite moments at 275, 336–8 and 524–6) when the mere possibility of Orestes’ not meeting Electra’s expectations is suggested. In short, Electra’s anxiety about her brother’s return and her idealization of that return are noticeable in the quality as well as in the content of her language. All this underlines Electra’s discontent with her current situation and provides further motivation for her hatred of Aegisthus and Clytemnestra. Returning to the stichomythia, Electra’s downcast complaint about Orestes’ absence (245) leaves the floor* once again to the stranger. The fact that initiative is left to him at this point is tantalizing: A situation has now arisen in which Orestes has only two possible ways of continuing the conversation: he must either reveal himself to his sister right away, or move on to an entirely new topic. (Schwinge 1968: 308)60

Orestes, assured by Electra of his own absence, chooses the latter course (forcing the audience to reassess the likelihood of imminent recognition), and now fully takes the reins of the conversation by engaging in a consistent line of questioning. After this point the stichomythia consists almost entirely of adjacency pairs* in which Orestes asks a question and Electra gives an answer—a pattern that is interrupted only for the two remaining near misses, at 262–3 and 282–5. Orestes then reclaims the interrogator’s position at 286, and gets to fire off two more questions before the stichomythia is interrupted by his final emotional outburst. The ‘entirely new topic’ for discussion in the next section of the stichomythia (246–62) is that of Electra’s marriage; and, again, a few elements that have characterized Electra previously in the play recur here. Her strong feelings about being deprived of the marriage she was supposed to enter into (οὐχ ᾧ πατήρ μ᾽ ἤλπιζεν ἐκδώσειν 249, ‘not the man to whom my father betrothed me’)61 and being 60 Damit . . . ist . . . eine Situation entstanden, in der es für [Orestes] nur zwei Möglichkeiten gibt, das Gespräch fortzusetsen: Entweder gibt er sich seiner Schwester jetzt sofort zu erkennen oder er beginnt mit etwas völlig Neuem. 61 My translation ‘my father’ effaces the ambiguity of πατήρ in the Greek: Agamemnon was also the father of Electra’s interlocutor, even though she does not know this. For the ambiguity in the use of familial terms in this passage, see Schuren 2014: 83.

106

The Linguistic Characterization of Electra

thrown instead into a marriage that equals social death (θανάσιμον γάμον 247) are coupled with consistent signs of affection and appreciation for the husband she does have (γενναῖος . . . εὐσεβής 253, ‘noble . . . respectful’, 255, 257, 259, σώφρων 261, ‘right-minded’). Electra vouches for the Peasant even in the face of the stranger’s tendentious questioning: Orestes here seems to be taking his role as Electra’s hereditary kurios very seriously and focuses again and again on what the marriage means to him (albeit in the third person) rather than to Electra (ᾤμωξ᾽ ἀδελφὸν σόν 248, ‘I groan for your brother’; εἴφ᾽, ὡς . . . σῷ κασιγνήτῳ λέγω 250, ‘tell me, so that I can report it to your brother’; Ὀρέστῃ μή ποτ’ ἐκτείσῃ δίκην 260, ‘afraid that he might face punishment from Orestes’). The stranger is careful to maintain the reserve that keeps his identity hidden,62 though he is not loath to express his surprise and indignation at the poverty of Electra’s husband.63 Amazed, at last, at Electra’s assurances of the Peasant’s moral standing, Orestes again briefly loses his composure at 262, inaugurating the second near miss. The final sections of the stichomythia shift attention to the issues that will drive the rest of the play: an initial orientation on the revenge plot and its targets (264–81, capped by the last near miss, 282–4); and, finally, a few lines that may alert the audience to Euripides’ use of the Aeschylean recognition pattern by mentioning the Old Man and Agamemnon’s grave (285–9). The first of these sections contains a few more moments that are instructive with respect to Electra’s linguistic characterization. The intensity of her reaction at 275, revealing her high expectations from Orestes, has already been noted above. What also emerges from the lines is the fervour of Electra’s hatred for her mother: first she discounts any possibility of remaining affection for her on Clytemnestra’s part (265): γυναῖκες ἀνδρῶν, ὦ ξέν᾽, οὐ παίδων φίλαι. Women are loyal to their men, not their children, stranger.

62

See my discussion of this feature of Orestes’ language in the next chapter. For the tone of Orestes’ questions here, see Denniston on 252 and 254; Denniston seems to me right to suggest that Orestes evinces ‘haughty scorn’ by saying that a σκαφεύς or βουφορβός would suit the house, and perhaps in his use of the phrase σῷ πόσει (254). Orestes will come to deplore his own lack of insight about the Peasant later, in his moralizing speech at 367–400; see my discussion in ch. III, §2.2.3. 63

The Linguistic Characterization of Electra

107

Next, Electra puts her hatred on full display in a bloodthirsty series of lines (276–81): Ορ. Ηλ. Ορ. Ηλ. Ορ. Ηλ.

ἐλθὼν δὲ δὴ πῶς φονέας ἂν κτάνοι πατρός; τολμῶν ὑπ᾽ ἐχθρῶν οἷ᾽ ἐτολμήθη πατήρ.64 ἦ καὶ μετ᾽ αὐτοῦ μητέρ᾽ ἂν τλαίης κτανεῖν; ταὐτῷ γε πελέκει τῷ πατὴρ ἀπώλετο. λέγω τάδ᾽ αὐτῷ, καὶ βέβαια τἀπὸ σοῦ; θάνοιμι μητρὸς αἷμ᾽ ἐπισφάξασ᾽ ἐμῆς.

Or. El.

Well then: if he comes, how might he kill his father’s murderers? By treating them with the same impudence with which our father was treated by his enemies. Would you really dare kill your mother together with him? Yes, with the same axe by which our father perished. Shall I tell him this? Is it sure on your part? May I die once I’ve shed my mother’s blood!

Or. El. Or. El.

Electra confirms her intention without a moment’s hesitation, picking up Orestes’ syntax without interruption in line 279 through her use of γε and answering his broader question (‘Would you dare?’) with a very specific ‘yes’: ‘with the same axe’.65 She presents partaking in her mother’s murder as her ultimate wish (281).66 In these lines, then, Electra’s vengefulness against her mother seems to erupt: ‘Nothing Electra has said so far has foreshadowed these bloodthirsty responses. This is important for her characterization’ (Cropp on 279–82). The stichomythia ends with the final near miss and a foreshadowing of the devices that will complete the recognition (the Old Man and the grave); and Electra rounds it off with another bitter tautology (Agamemnon ‘got what he got’ as his grave, ἔκυρσεν ὡς

64 I am persuaded (if not without doubts) by Slings’ (1997c: 141–2) defence of οἷ᾽ ἐτολμήθη πατήρ as a possible phrase: even though an active counterpart with this verb (e.g. οἱ ἐχθροὶ τοιαῦτα ἐτόλμησαν τὸν πατέρα) would be impossible, the use of a passive with a personal subject is not inconceivable (for arguments and examples, see Slings’ discussion). Of the proposed emendations, Winnington-Ingram’s πατρί strikes me as preferable to πατρός (Broadhead) or ποτέ (Nauck). 65 For this use of γε, see ch. IV, §2 (on 672). 66 The optative forms θάνοιμι and θάνοιμεν are infrequently used in this hyperbolic way (‘May I die once only I . . . ’); such use is found only in tragedy and a few lyric fragments and is always related to the trope of a wish to kill someone as a singular ambition that would make one’s life complete: cf. 663 (Old Man about the revenge), Aesch. Cho. 438 (Orestes about Clytemnestra, a likely intertext for the current use; see Sier ad loc.), Soph. Aj. 391 and Eur. Or. 1100 (with West ad loc.).

108

The Linguistic Characterization of Electra

ἔκυρσεν 289),67 whose implications alarm Orestes enough to make him erupt into a fuller instance of the near-miss pattern (290–6).68 Having examined the entire stichomythia, we may now draw some conclusions about Electra’s linguistic behaviour in it and how it continues some patterns we saw earlier. First, Electra is prone to disrupting the conversational equilibrium with an impolite utterance or a dispreferred turn*, at moments where her own situation and fall from grace are at issue (239, 241) or where the notion of Orestes returning to remedy that situation is called into question (275; her distress about Orestes’ absence is clear also from 245). What is evident from such moments is the extent to which Electra detests the situation she finds herself in (in spite of her genuine affection for her husband, which is unchanged in the stichomythia, 251–61). What also emerges from the scene is the severity of her hatred for her mother (265, 276–81), while repeated references to her dead father (243, 249, 276, 279) underline her urge to settle the score on that account. Together, these moments motivate Electra’s desire to commit the matricide. Lloyd (1986: 13) is thus right to see the whole stichomythia ‘as part of Electra’s account of her sorrows’: they show us a woman who is so outraged about her displacement and the unavenged death of her father that she will seek the ultimate revenge. About this motivation Electra herself is explicit enough, but Euripides’ characterization is more subtle than simply having Electra say out loud what drives her; her preoccupations affect her conversational behaviour itself.

3.2. The ‘message’ (300–38) In §2.4.2 I commented briefly on Electra’s rhēsis at 300–38, arguing against readings that insist on its ‘hypocrisy’. The speech, which rounds off the instruction of the stranger with an explicit appeal 67 Although both 245 and 289 express two things that are logically identical (‘absent . . . not present’; ‘he got what he got’), the effect of the tautology is slightly different. The negation in 245 (οὐ παρών) implies, as negations commonly do, that the reverse (Orestes’ presence) would be preferred (see n. 57). The present instance is an example of a pankoinon (on which see Johnstone 1980, who treats all instances in tragedy): a literal tautology, which often has a dismissive effect (the issue is so selfevident or immutable that there is no point in discussing it further). 68 I treat these lines extensively in the next chapter (§2.2.2).

The Linguistic Characterization of Electra

109

to Orestes,69 is designed explicitly to spur her absent brother into action. Electra underpins her appeal by contrasting the siblings’ exile with the shamefully comfortable life of their father’s murderers and by underlining the extent to which Agamemnon’s death remains unavenged (again, she weds her own circumstances to the unresolved murder). The opening of the speech is an interesting example of how gender differences and relative positions of power affect conversational behaviour. Electra, having been urged by the stranger and, again, by the Chorus to speak (λέξον 292, ‘speak’; βούλομαι . . . μαθεῖν 299, ‘I wish to hear’), is cautious in taking the floor* (λέγοιμ᾽ ἄν, εἰ χρή 300, ‘I’ll tell you, if it is proper’) and makes sure to point out why she feels justified in making her speech (χρὴ δὲ πρὸς φίλον λέγειν | τύχας βαρείας 300–1, ‘and it is proper to speak of one’s heavy misfortunes to a friend’; ἐπεὶ δὲ κινεῖς μῦθον 302, ‘since you urge me to tell my story’).70 There is some debate among scholars about whether Electra’s tone here is reluctant or deferential, or perhaps both: Denniston (on 300) translates λέγοιμ’ ἄν as ‘I had better speak’, adding (on 302) that Electra is ‘naturally reluctant to talk of τύχας βαρείας’ and that κινεῖς μῦθον must stand for the provocation of reluctant speech (see also Keene ad loc.). Cropp (ad loc.) and Slings (1997c: 144) seem to read these lines differently: they translate εἰ χρή ‘if it’s proper’ and ‘if I may’, respectively. I would argue that the hesitation that Electra expresses even as she begins her speaking turn has to do with her status as a woman as well as with the content of her speech. Because Electra is a woman in a mixed group* and her speech will express complaints, distress, and incitement to action, her freedom to make a longer speech of this kind is not at all self-evident. It is only because she has been invited to do so by a φίλος that she can proceed without shame; this is what has made it appropriate (χρή).71

69 Cropp ad loc. usefully compares similar appeals to recently arrived allies at Med. 708–18, Andr. 920–53, Supp. 163–92, Or. 449–55, and IA 900–16. 70 Battezzato 2001 has rightly restored the punctuation of 300–1 to that of the editio princeps (ignored by many editors and commentators): χρὴ δέ . . . λέγειν is not a parenthesis, and τύχας βαρείας is the object of λέγειν, not of λέγοιμ’ ἄν. 71 For this use of χρή, Denniston himself provides a very instructive parallel, which somewhat undermines his argument: at Her. 141, Lycus claims that χρή (‘it is right’) for him to ask whatever questions he wants of Amphitryon, since he is the ruler of the land—he needs no permission to speak with παρρησία: εἰ χρή μ᾽, ἐρωτῶ· χρὴ δ᾽, ἐπεί γε δεσπότης | ὑμῶν καθέστηχ᾽, ἱστορεῖν ἃ βούλομαι. (‘If it is right for me, I have

110

The Linguistic Characterization of Electra

Electra’s awareness of the delicacy of her situation is made explicit by her use of the first-person potential optative construction (λέγοιμ᾽ ἄν), which often indicates polite acceptance of an interlocutor’s suggestion or permission to do something.72 Electra is aware of similar restrictions on her speech at various other points in the play (at 900–6 this comes from fear of φθόνος: see §5.2.1; at 945–6 certain topics would be off limits: see §5.2.2; and at 1055–60 she speaks only after having been offered παρρησία by Clytemnestra at 1049: see §5.3.3).73 Such restrictions on Electra’s speech add some significance to her brief outbursts, which I identified in the stichomythia (§3.1). There, not yet able to give full expression to the problems that plague her or to explicitly demand Orestes’ glorious return, Electra responded with some urgency when the stranger tangentially raised those issues. But those moments showed us (and Orestes) only glimpses of Electra’s outlook: now that she has been urged to speak at length, those glimpses are developed into a full treatment, providing a clear portrayal of her view of the situation. There is another subtle difference between the stichomythia and the rhēsis: whereas the former deals in large part with the causes of the current situation (we accordingly find mostly a mix of imperfects and aorists in Electra’s responses to the ‘interrogation’ that begins at 238), the latter is a description of the current situation (and thus consists almost exclusively of present tense main verbs, i.e. presents and perfects).74 For these reasons, Denniston’s complaint (on 292ff.) that Electra ‘has not much that is new to tell’ in the speech is not greatly to the point, even if she does cover some of the same ground as in the stichomythia.75 The speech is something to ask you—and it is right, since I have been installed as your ruler, for me to ask whatever I want.’) 72 See Lattimore 1979 and KG I §396.4. Other examples of λέγοιμ’ ἄν in tragedy (often suggesting similar power differentials) are Soph. OT 95 (Creon to Oedipus), Eur. Hec. 1132 (Polymestor to Agamemnon), Supp. 465 (Herald to Theseus), and Or. 640 (Orestes to Menelaus). 73 Examples of other passages in Euripides where gender seems to play a role as a check on speech are Hec. 234–6, Her. 279, Tro. 906–13, Hel. 1049 (in some of these passages other power differentials are also at work). Such restrictions are made very explicit at Heracl. 474–83. They play, of course, a major role also in Aeschylus’ Agamemnon (e.g. 348–51, 592, 918, 1399). See also, e.g. McClure (1999: 24–9). 74 The exception is ἔλαβε at 325, which, combined with οὔπω, clearly refers to a stretch of time up to the present (‘he has not received’; this is a negative constative aorist, on which see Rijksbaron 2002: §8.3.1). 75 Lloyd (1986: 7–8) rightly points out that there are structural reasons for this as well: both the stichomythia and the rhēsis are of conventional ‘types’—interrogation,

The Linguistic Characterization of Electra

111

a new and distinct part of Electra’s exposition of her troubles, and the repeated emphasis on her own humiliation (the ‘same ground’ covered) is deeply suggestive of the way in which Electra interprets those troubles. The mention of Agamemnon’s grave at 289 led the stranger to request a more complete description of the situation at Argos. Electra obliges by providing that description in the only terms she can: by reference to her own personal degradation as well as to the impudent behaviour of Agamemnon’s yet to be punished killers. For Electra, after all, the former issue (her personal situation) is an inseparable component of the latter (the insolence of Clytemnestra and Aegisthus). The opening four lines underline this equation of her own troubles with those of the rest of her family (a recurring feature of Electra’s language, as we have already seen). Electra speaks of her own fate as being closely linked to that of her father (τύχας βαρείας τὰς ἐμὰς κἀμοῦ πατρός 301, ‘my own heavy misfortunes, and my father’s’), and of her own evil circumstances as being linked to those of Orestes (τἀμὰ κἀκείνου κακὰ 303, ‘troubles, both mine and his’).76 The latter conflation of κακά is designed to spur her brother into action, and that is indeed what she goes on to do, very explicitly. Electra underlines the urgency of her appeal by characterizing it as a supplication, both at its opening and at the end: her first speech act* in the rhēsis (after accepting permission to speak) is the performative* ἱκετεύω, ‘I beseech you’ (302), and the same formulation is used in the peroration (332).77 appeal—demarcated clearly by the transitional lines of Orestes and the Chorus (290–9). The necessarily separate presentation of these types makes some overlap of content inevitable. In other words, this is to some extent a matter of Euripidean formalism. 76 ἐκείνου refers to Orestes, not to Agamemnon—a reading that Slings (1997c: 144) wrongly imputes to Cropp (but see Cropp 1988 on 303) and that Battezzato (2001: 731) adopted. There is, as Denniston ad loc. says, some ‘rhetorical force in the echo of 301’: all unresolved family affairs are characterized as part of the same overarching problem. 77 Cf. in similar ‘appeal scenes’ ἄντομαι at Med. 709 (Medea to Aegeus) and Andr. 921 (Hermione to Orestes), and the posture of supplication adopted by Adrastus at Supp. 165. The verb ἱκετεύω seems to be used outside contexts of genuine supplication, however (see Naiden 2006: 44 with n. 86, pointing out that using a word or gesture alone does not ‘suffice to establish that supplication is occurring’—though it might be argued that each of Naiden’s ‘four steps’ of supplication—approach, gesture/ word, request, response—is in fact taking place here). We may in any case wonder whether we should view the current situation as ‘true’ supplication.

112

The Linguistic Characterization of Electra

The intervening parts of the speech, which form the core of Electra’s message, are neatly structured into four different sections through a πρῶτον μέν . . . δέ . . . δέ . . . δέ sequence (304, 314, 318, 323)78 where the topic of each of the last three sections stands in initial position in its corresponding δέ clause: μήτηρ δ᾽ ἐμή (314–18, ‘my mother’, on Clytemnestra), αἷμα δ᾽ . . . πατρός . . . , ὃς δ᾽ ἐκεῖνον ἔκτανεν (318–22, ‘the blood of my father . . . and the man who killed him’, on Aegisthus’ usurpation of Agamemnon’s power) and Ἀγαμέμνονος δὲ τύμβος (323–31, ‘and Agamemnon’s grave’, on his defilement of Agamemnon’s grave and his taunts to Orestes). The first of the four sections focuses again on Electra’s state, and this section above all others in the speech has elicited much condemnation of her character.79 Compounding the problems of interpretation is a set of thorny textual issues surrounding lines 307–13: much uncertainty about punctuation exists (depending on their texts, some editors are forced to accept asyndeton; full stops have been printed after 306, 309, 310, or nowhere at all), and there are contested readings in 311–13.80 Since these problems are directly relevant to

Denniston is wrong to claim (in his note on 304) that ‘μέν is not formally answered’: the particle often has scope over a long stretch of discourse before being picked up by δέ. (Denniston proceeds to say that Electra ‘passes to her spiritual distress’ at 314, without noting that there is in fact a δέ to mark that change of topic.) Martina (1979: 26) similarly misconstrues the structure, I think. This has some bearing on the issue of punctuation in 306–13 (on which more will be said later). 79 See, in addition to the comments of Conacher cited before (§2.4.2), e.g. Kitto 1961: 334 n. 1, Tarkow 1981: 151, Electra ‘whines about her predicament’, O’Brien 1964: 28–9, and Barlow 1986b: 92–4. An extensive discussion of lines 304–13 may be found in Martina 1979. 80 In addition, there is line 308, deleted by Camper and condemned as a ridiculus versus by Wilamowitz (1875: 72): when Kovacs (1985: 306) says that the line ‘looks very much like a member of Wilamowitz’ ‘family’ of ἤ-interpolations’, he fails to notice that Wilamowitz (1875: 205) had actually listed the case among that family’s number. Apart from Keene, all recent editors and commentators either delete the line or add a lacuna after it. I still have some doubts: (1) πέπλων is very easily supplied as complement from the previous line, and στερήσομαι likely has the durative sense ‘go without’, ‘be apart from’ (see Allan 2003: ch. 4; 2006: 124–6), all of which diminishes the need for an explicit complement in the genitive (it also makes Nauck’s future perfect κἀστερήσομαι unnecessary); (2) the anaphora αὐτὴ μέν . . . αὐτὴ δέ (307–9) loses little of its force with the intervening line. Wilamowitz’s ἤ interpolations strike me as a group that might well represent a Euripidean (and sometimes Sophoclean) mannerism. I also find the theory that the corruption γυμνάς at 311 may be explained by γυμνόν here attractive (so Keene app. crit. and Seaford 1985: 319 n. 38)—though this hypothesis does not rule out an interpolation, if the latter occurred before the corruption of 311. 78

The Linguistic Characterization of Electra

113

the interpretation of Electra’s speech, it may be prudent to spend some time on them. To begin with the problems in 311–13, the manuscript evidence for these lines gives the following: ἀναίνομαι δὲ γυμνὰς οὖσα παρθένος ἀναίνομαι δὲ Κάστορ᾽ ὧ πρὶν ἐς θεοὺς ἐλθεῖν ἐμὲ μνήστευον οὖσαν ἐγγενη 311: L and P have δὲ γυμνὰς, Triclinius corrected to γυναῖκας (without deleting δὲ).

The correction of γυμνάς to γυναῖκας is assigned by Zuntz (1965: 170) to the second stage of Triclinius’ corrections on the basis of ink colour, which would also explain the presence of δὲ γυμνάς in P.81 Triclinius did not correct the arising metrical error—deletion of δέ is required if we accept γυναῖκας (thus, if we follow Triclinius, there is no connective between 310 and 311—more on which shortly). Zuntz argues convincingly that Triclinius’ correction was probably based on evidence from another manuscript, which would lend significant weight to γυναῖκας.82 However, several scholars, most extensively Kovacs (1985: 306–10), have argued for retaining γυμνάς and changing παρθένος to παρθένους in order to draw out the sense ‘I shun the (choruses of ) maidens since I am naked’—a figurative sense for ‘naked’ being implied. But, in addition to Triclinius’ evidence, there are several insurmountable problems with this text (all concisely summarized in Seaford 1985: 319 n. 38). First, no matter how much Kovacs would like the modern ‘I haven’t a thing to wear’ to be an appropriate equivalent (1985: 309 n. 10), there is simply no way for γυμνάς to be taken to mean ‘without proper festal attire’.83 Second, as observed earlier (§2.4.2), a lack of clothing is not what Electra complains about: she earlier spurned the Chorus’ offer of a suitable dress because she was in mourning (175–89). Reading γυναῖκας, on the other hand, gives us a line that precisely portrays Electra’s liminal status, which has been well established earlier in the play: she is a woman married in name only, who (through her husband’s kindness)

81

For the relation between L and P, see ch. IV, n. 5. Contra Kovacs (1985: 307 n. 3) and Distilo ad loc.; but I do not think that their objections are sound. 83 No parallel can be found to support this use, and none of the parallels offered by Kovacs (Hes. Op. 391, Ar. Nub. 498, Xen. An. 4.4.12, and Pl. Resp. 474a) comes close. 82

114

The Linguistic Characterization of Electra

remains a virgin—a play on Electra’s very name. Such a woman can well say ‘I shun the women, as I am (still) a maiden’, especially one line after complaining that she cannot participate in choruses (‘of maidens’ no doubt being implied): χορῶν τητωμένη (310). In short, γυναῖκας is easily the preferable reading. The next two lines provide the next set of problems: the manuscript reading of the relative clause (introduced, the accentuation suggests, by the dative pronoun ᾧ) has an unnamed ‘they’ betrothing Electra to Castor (ᾧ) before he is apotheosized (the subject having to be understood in the latter case as ‘my parents’ or as a generic ‘they’; taking ἐμνήστευον as first person seems implausible). But there are problems with this reading: first, the lack of subject for (ἐ)μνήστευον is awkward; second, it has been recognized (e.g. by Denniston ad loc., or by Kovacs 1985: 306–7) that a double construction of μνηστεύω with an accusative and a dative, for ‘betrothing someone to someone else’, is not established outside this instance and that the verb appears never to mean ‘betroth’ in any of its uses.84 A third problem (also mentioned by Kovacs) is that, in order for the subject of the infinitive in πρὶν ἐς θέους ἐλθεῖν to be Castor, that same referent must also be the subject of the relative clause (otherwise it should be mentioned separately, in the accusative). Two possible solutions have been proposed: • Kovacs (1985: 308–9), following Scaliger, wants to read the text, with minimal changes, as a series of duals (with a plural verb): Κάστορ(ε), ὥ . . . ἔμ᾽ ἐμνήστευον (‘the Dioscuri, who were my suitors’), with the dual Κάστορε standing for both Castor and Polydeuces (a usage of the dual that Kovacs sees paralleled in the

84 Pace LSJ s.v. II (citing only this instance). This sense (along with the ms. text) is defended by Martina (1979: 44–8), who needs to transpose Eur. IT 208 after 221 in order to find what he considers a sufficient parallel for this sense of μνηστεύω (I still find it unconvincing). For a better parallel (not signalled in LSJ and unnoticed by recent commentators and Martina) we might look at Stes. fr. 93 (= S148), 14–15 (Davies and Finglass 2015), where what remains of the text indicates a mother on a journey to find a wife for a man (possibly Eriphyle for her son Alcmaeon): ἔβα παράκοιτι[ν] . . . μναστεύσοισα ματή[ρ] (‘the mother went to woo a wife’). An indirect object similar to our ᾧ is not expressed, or perhaps was lost (παῖδ’ in the next line is probably accusative), but the passage is an example of what we need, in that the husband-to-be is not the subject. Considering the other objections to the paradosis of our present lines, I do not think, however, that the Stesichorus fragment offers sufficient grounds for preserving the ms. reading.

The Linguistic Characterization of Electra

115

use of Αἴαντε for Ajax and Teucer, and in Latin Castores). Kovacs is followed by Distilo (ad loc.). • Most recent editors (including Denniston, Diggle, Basta Donzelli, and Cropp) prefer Nauck’s further changes Κάστορ(α), ὅς . . . ἔμ᾽ ἐμνήστευεν (‘Castor, who was my suitor’). Though Kovacs’ solution has significant appeal in that it requires fewer changes to the transmitted text than Nauck’s three palaeographically untrivial emendations, I do not, in the end, find it very plausible, if only because it would involve both Castor and Polydeuces vying for the same girl. We have been told at 249 that Agamemnon once hoped to betroth Electra ‘to someone’ (just one person): οὐχ ᾧ πατὴρ μ᾽ ἤλπιζεν ἐκδώσειν ποτέ. It seems likely that we are here told that that person was in fact Castor. But it must be said that either reading yields sense and that even the transmitted text cannot be ruled out entirely (see n. 84). In any case, whatever we read in 312–13, the question remains what it is exactly that Electra is shunning:85 scholars’ suggestions have ranged from ‘the thought of Castor’ to ‘the worship of Castor/the Dioscuri’.86 In truth, the Greek (which merely mentions the name Κάστορ᾽) is underdetermined and might allow either interpretation; however, the connotations of the verb ἀναίνομαι87 and the preceding mention of ἱερά (310, ‘festive rites’), may steer us towards the latter option, as is suggested by Martina: Line 310 is in some sense elaborated by the following three, which complete the picture of Electra’s situation. . . . Electra . . . tells [Orestes] that it is a consequence of her condition that she cannot participate in certain festivals and that she is deprived of choral dances . . . She

85 Or ‘feels shame towards’, depending on whether we read ἀναίνομαι (MSS) or αἰσχυνομαι in 312 (see n. 87). 86 Martina (1979: 28–42) provides an extensive survey of evidence pertaining to the cult of the Dioscuri and suggests—appealingly—that there was a specifically ‘female’ cult of Castor, connected to the Eleusinian mysteries. 87 Page (as recorded by Denniston ad loc.) argued for replacing ἀναίνομαι in 312 with αἰσχύνομαι, which avoids Electra’s ‘spurning’ Castor; this text is adopted by Diggle, Basta Donzelli, and Cropp. But the ‘problem’ in 312 is equally applicable to 311 (ἀναίνομαι γυναῖκας), where there is no reason to assume that Electra is hostile to gatherings of women. As Denniston ad loc. points out, corruption from ἀναίνομαι to αἰσχύνομαι in both lines is ‘improbable’; a better solution, in my view, is to accept that a slightly extended sense of ἀναίνομαι is at play here (‘abstain from’, ‘shun’) and to retain the anaphora between 311–12.

116

The Linguistic Characterization of Electra

mentions one of the ἱερά in particular, perhaps the one which is most significant for her. She says, that is, that ‘she abstains from Castor’, that is to say, from the cultic and sacrificial practices in honour of Castor. (Martina 1979: 25)88

Nonetheless, we need perhaps conclude no more than that ἀναίνομαι . . . Κάστορ(α) is phrased generally and should be taken simply to mean that Electra has withdrawn from any contact with her famous uncle, in contexts of worship or any other. The final relative clause (312–13) then adds the painful reason why: a former husband-to-be is, in her current circumstances, not an appropriate person to seek out. Remaining unsolved (and perhaps insoluble) is a difficult problem of punctuation: if we adopt Triclinius’ γυναῖκας—without δέ—in 311, there seems to be asyndeton either at 309–10 or at 310–11, and this has vexed scholars greatly.89 It is possible to argue that, thematically, the material covered in 304–9 goes together (physical toil), and that the same holds for 310–13 (Electra’s seclusion from certain groups and occasions as a result of her marital status). If such thematic unity is considered paramount, a full stop would need to be printed after 309,90 and there would in fact be a somewhat unusual asyndeton at 309–10. Kovacs, as mentioned, prefers keeping L’s δέ γυμνάς (ante correctionem), which would provide a particle for connecting 310–11.91 The solution I would adopt is one that was already offered by Wilamowitz (1875: 64), if I interpret his minimal comment on these lines correctly: rather than trying to find a place to print a full stop somewhere in 309–11, we may rather print one after 306. All this gives the following text for 303–13: ἄγγελλ᾿ Ὀρέστῃ τἀμὰ κἀκείνου κακά, πρῶτον μὲν οἵοις ἐν πέπλοις αὐλίζομαι,

88

Il v. 310 viene in un certo sense chiarito dai tre successive, con i quali si completa il quadro della situazione di Elettra. . . . Elettra . . . gli dice che non partecipare a determinate feste ed essere priva di cori è una conseguenze della sua condizione . . . Dei ἱερά ne ricorda uno, forse il più significativo per lei. Dice cioè che ‘si astiene da Castore’, cioè dal culto e dal sacrificio in onore di Castore. 89 Denniston on 310–11 offers an overview of the solutions offered. 90 So Keene ad loc. and Martina 1979: 25–6. 91 This disturbs the flow of thought much more than Kovacs would admit (given his insistence that his text restores ‘an Electra whose logic and grammar are faultless’, Kovacs 1985: 308): Electra would still ‘wander without discernible plan from poverty and toil to religious festivals’ (1985: 307) and then elaborate only on the part about religious festivals.

The Linguistic Characterization of Electra

117

πίνῳ θ᾿ ὅσῳ βέβριθ᾿ , ὑπὸ στέγαισί τε οἵαισι ναίω βασιλικῶν ἐκ δωμάτων. αὐτὴ μὲν ἐκμοχθοῦσα κερκίσιν πέπλους— ἢ γυμνὸν ἕξω σῶμα καὶ στερήσομαι— αὐτὴ δὲ πηγὰς ποταμίους φορουμένη, ἀνέορτος ἱερῶν καὶ χορῶν τητωμένη, ἀναίνομαι γυναῖκας οὖσα παρθένος, ἀναίνομαι δὲ Κάστορ᾿ , ὅς πρὶν ἐς θεοὺς ἐλθεῖν ἔμ᾿ ἐμνήστευεν, οὖσαν ἐγγενῆ.

305

310

Report to Orestes the misfortunes that are mine and his: first, what kind of clothes I lie in, how much dirt I am laden with, and beneath what kind of roof I live, cast out of the royal halls. Toiling at the shuttle to make clothes myself (else I will have my body unclothed and go without), and carrying water from the stream myself, missing festival dances and cut off from choruses, I shun the women as I am a maiden, and I shun Castor, who, before taking his place among the gods, wooed me, his kinswoman.

Electra now begins her message with a summary of her problems in three points, each marked by an indirect question: the state of her attire (οἵοις ἐν πέπλοις αὐλίζομαι 304, ‘what kind of clothes I lie in’), her malnourishment (πίνῳ θ᾽ ὅσῳ βέβριθ᾽ 305, ‘how much dirt I am loaded with’), and her decrepit place of residence (ὑπὸ στέγαισί τε | οἵαισι . . . ναίω . . . δωμάτων 305–6, ‘and under what kind of roof I live’). After the indirect questions follows a natural break: the next sentence then begins with αὐτὴ μέν and runs through a series of modifiers—a loosely strung together series of three circumstantial participles and an adjective (ἐκμοχθοῦσα 307; φορουμένη 309; ἀνέορτος and τητωμένη 310)—before reaching its (first) main verb at ἀναίνομαι in 311.92 Reading the text this way gives us a long second sentence that elaborates on the summary content of Electra’s indirect questions, even though it mentions a new aspect as well: her nonparticipation in various female groups. We might summarize the meaning like this: ‘tell him what a state I am in without proper clothing or food, and far removed from palace life’ (304–6): ‘I have to make my own clothes, get my own water, and cannot take part in female activities or in the marriage that I was supposed to have’ (307–13).

92 This solution is not very different from that adopted by Denniston (on 310–11), except that he would not print a full stop, not even after 306.

118

The Linguistic Characterization of Electra

The textual issues in these lines remain difficult, and we may have to accept that more is corrupt here than can be mended. What this should not obscure, however, and what I hope my discussion has made clear, is that Electra paints here a picture of her life that is one of bleak and impossible circumstances. As she sees it, the murder of Agamemnon and her subsequent banishment and sham marriage have left her poor and unable to benefit from her rightful estate, a virgin who cannot be married to an appropriate spouse but (to top it all off) is ‘virtually’ married nonetheless. Electra now escalates her case in the remaining sections (314–18, 318–22, 323–331) by contrasting her miserable life with that of her father’s murderers. These descriptions deftly move from the female realm to the male: the initial contrast Electra draws between herself and Clytemnestra (314–18) pivots on an axis that is distinctly female, namely outward appearance (Clytemnestra’s riches and well-dressed slaves being the opposite of Electra’s outfit); after that, the description of Aegisthus—primarily set against Agamemnon, male against male—incorporates violence, lack of burial, and taunts. Not coincidentally, at the end of this sequence Orestes himself is brought in to suffer Aegisthus’ abuse (330–1), and this leads straight into the repeated appeal for his arrival and a display of manhood (336–8). This organization allows Electra once again to link her own situation intimately to the unresolved murder of Agamemnon and to Orestes’ exile. A point of detail worth mentioning in this regard is Electra’s use of the pronoun ἡμᾶς (329, ‘us’): she considers herself as much the object of Aegisthus’ insults as Orestes himself. These latter sections of the speech are packed with subtle verbal details designed specifically to malign Clytemnestra and Aegisthus: the geographical details in the description of Clytemnestra (a Phrygian throne, Asian slaves wearing Idaean robes, 314–17), animal metaphor in the use of ζεύγνυμι (‘yoke’) for Clytemnestra’s slaves (318), and a well-placed epithet here and there (e.g. Aegisthus’ ‘bloodstained hands’, μιαιφόνοισι χερσί 322).93

93 On these and other uses of imagery and metaphor in the rhēsis, see Barlow 1986b: 92–4. I agree with Barlow that such details reveal ‘Electra’s anger’ (94); I do not agree, however, that they are there in the first place to show how ‘warped’ she is. The rhetorical purpose of Electra’s rhēsis (inciting Orestes into action) requires that she present the situation in as dire colours as she can.

The Linguistic Characterization of Electra

119

Some other linguistic minutiae in these sections that have less often been noticed are worth mentioning as well. First, although the speech consists of only main verbs with present tense reference (the perfect σέσηπεν at 319, of Agamemnon’s blood that ‘lies rotten’, is particularly effective in sketching the ongoing outrage of her father’s murder), Electra uses a few past tenses in relative clauses to intersperse strategic references to Agamemnon’s former glory as supreme commander (ἔπερσ(ε) 316, ‘plundered’; ἐστρατηλάτει 321, ‘commanded’). These relative clauses clearly serve to provide a contrast with, and thus to emphasize the scandalousness of, Clytmnestra’s and Aegisthus’ current behaviour. Second, Electra blackens Aegisthus’ name further still by way of the phrases she uses to refer to him. Her initial mention of Aegisthus leaves him nameless,94 identified only through the incriminating relative clause ὃς δ᾽ ἐκεῖνον ἔκτανεν (319, ‘the man who killed him’). The next phrase that refers to Aegisthus in the speech, τῆς ἐμῆς μητρὸς πόσις | ὁ κλεινός (326–7, ‘my mother’s husband, the glorious man’) is even more damning, layering sarcasm on sarcasm. Word order and metre, which have a complex interaction here, do much to heighten the effect. The possessive τῆς ἐμῆς μητρός is preposed to πόσις, an ordering that suggests that the modifier is (for some reason) more ‘salient’ than the noun itself.95 This nuance brings out the depth of Electra’s contempt: Aegisthus is to her, above all else, an extension of her mother. But she does not stop there: in enjambed position after the line break, she adds the deeply sarcastic ὁ κλεινός. Such ‘adding’ enjambments (where the end-stopped element is not syntactically 94 So is Clytemnestra, who is referred to only by μήτηρ δ᾽ ἐμή, the tone of which might well be indignant; interestingly, Electra addresses Clytemnestra with μῆτερ in the agōn (1006, 1055, 1058), something that Sophocles’ heroine never does (see Vahlen 1891: 361 and Michelini 1987: 218 n. 162). Electra refrains from using Aegisthus’ name in other places as well, using phrases such as ἄλλῳ (212, ‘another man’) and τὸν δόντα (259, ‘the one who gave me’): such circumlocutions are not unusual in tragedy, but they are deliberately employed in this speech at least. 95 This has been demonstrated by Dik (e.g. 2007: ch. 4, esp. 107–12) and, more fully (though only for Herodotus), by Bakker (2009b: 38–52). On genitives, Bakker (2009b: 46) writes that ‘the preposition of the modifier may indicate that the exact nature of the relation between the referent of genitive/possessive and the referent of the head noun is less relevant than the fact that a relation exists’. In other words, Electra puts less emphasis on the exact relationship between Aegisthus and Clytemnestra than on the fact that it is her that he somehow belongs to. By the same token, Electra seems to emphasize the fact that Clytemnestra is her own mother by preposing ἐμῆς to μητρός.

120

The Linguistic Characterization of Electra

required to complete the sense) have been shown by Dik to receive some additional emphasis,96 and the pause before ὁ κλεινός here surely contributes to the outrage: Aegisthus, scandalously drunk and little more than his wife’s pet, is called ‘glorious’!97 A final point of note in this part of the speech is Electra’s employment of direct speech in reproducing Aegisthus’ taunts (329–31): καὶ τοῦτο τολμᾷ τοὔπος εἰς ἡμᾶς λέγειν· Ποῦ παῖς Ὀρέστης; ἆρά σοι τύμβῳ καλῶς παρὼν ἀμύνει; ταῦτ᾽ ἀπὼν ὑβρίζεται. And he dares to say this, directed at us: ‘Where is your boy Orestes? Doing a good job, then, of being here and defending your tomb for you?’ Thus Orestes is abused while he is away.

As I will discuss more extensively in chapter V (§3.2), the use of direct speech invites the addressee to imagine him-/herself ‘experiencing’ the words as they were uttered (whether they were actually uttered or not, which is largely besides the point).98 The fact that Electra invokes Aegisthus in quotation marks allows her to cast him in the most ignominious light: he comes to life, as it were, equipped with a sneering tone that should sting Orestes right where it hurts most. The language of the insults is keenly designed: short, direct questions, addressed to a murdered victim through a direct second-person reference (σοι); an unceremoniously colloquial tone, visible especially in the use of καλῶς (see Stevens 1976: 55 and Cropp ad loc.); a strong

96 Dik (2007: ch. 6) shows that line-initial position is not intrinsically a position of emphasis (and line-end position is even less so) and that end-stopped words tend to have no added significance if they are ‘necessary enjambments’, i.e. elements that are required syntactically. 97 Denniston and Cropp are right in taking ὡς λέγουσιν in 326 with the sentence as a whole (as an ‘evidential’ marker specifying the basis of Electra’s claims about Aegisthus behaviour) rather than specifically with ὁ κλεινός. The latter option would ‘greatly weaken the irony’ (Denniston ad loc.), and I do not think that ὡς λέγουσιν can mean ‘as they call X’ anyway (we might expect καλοῦσιν). Incidentally, it is not an obvious interpretation of ὡς λέγουσιν to say that it makes Electra’s claims look ‘suspiciously like an invention of her own’ (Raeburn 2000: 154): such evidential markings serve as a hedge, placing the authority for making a claim with an external source—the phrase is interpreted here quite naturally as a way for Electra to overcome the objection that she could not have seen Aegisthus perform his grave-hopping escapades herself. 98 It is pointless to accuse Electra of ‘lying’ here because Aegisthus may not have actually spoken the words as she produces them: this is to miss the rhetorical point of the entire speech (see Lloyd 1986: 8).

The Linguistic Characterization of Electra

121

note of sarcasm in the use of παρών,99 and a hint of it in the use of the particle ἆρα (which may have a tone of mock surprise, suggesting that Aegisthus is brought to his question by external evidence that he did not expect).100 These violent face* assaults are superficially directed at Agamemnon; but, through Aegisthus’ chosen tactic of haranguing the father by reference to the failed son, through Electra’s explicit framing (εἰς ἡμᾶς 329; ἀπὼν ὑβρίζεται 331), if not through the inherent pain of a murdered father being maligned,101 the offence is to be felt by Orestes just as strongly. Leading immediately to her repeated appeal to Orestes, this portrayal of Aegisthus’ taunts could not be more powerful as the climax of her speech (the next word in the speech, ἀλλ(ά), marks the break and the beginning of Electra’s peroration). In the peroration itself (332–8), Electra formalizes her appeal once more by repeating the performative* ἱκετεύω (332) and adds weight to her message by multiplying its ‘senders’ (333–5): Electra’s hands, tongue, mind, and shorn head combine with Agamemnon himself (ὅ τ᾽ ἐκεῖνον τεκών 335, ‘the father who begot him’) in longing for Orestes’ return.102 Agamemnon, the last sponsor of the message, is the one who brings Electra to conclude her speech by relating Orestes’ manhood to his father’s (πατὴρ μὲν . . . , ὁ δ’ . . . 336–8).103 The sternness of her warning to Orestes is evident from the construction used, a conditional with a future indicative (εἰ . . . | οὐ δυνήσεται 336–7, ‘if he is to prove incapable’): such conditionals tend to be used in ominous predictions, warnings, and threats.104

99 To make matters more complex still, Euripides inserts a dose of dramatic irony directed more at the audience in the theatre, by contrasting Aegisthus’ παρών with Electra’s ἀπών (Orestes is less ἀπών than Electra knows, of course). 100 For this function of the particle (generally seen to be a combination of ἦ and ἄρα), see ch. IV, n. 67 on ἄρα. 101 For family and other social groups as a formative component of face* (and of attacks on face), see e.g. Culpeper 2011: 28–9. 102 For the conceit, Cropp ad loc. rightly compares Hec. 836–40. 103 Electra will appeal directly to Orestes’ manhood again at 693 (in the planning scene) and at 982 (Orestes’ moment of hesitation; see my discussion in ch. IV, §3.2). 104 See Wakker 1994: 167–73 (with justified warnings against seeing this as an inherent value of the construction). Incidentally, we may also note the use of οὐ rather than μή in this conditional clause, which is typical of its function as an obligatory subject constituent (as subject of αἰσχρόν (ἐστι)), syntactically similar to ὅτι clauses more than to ‘true’ conditionals. For this type of εἰ clause and the use of the negation, see Wakker 1994: 286–94, esp. 287–8. As Wakker (1994: 194) notes about a similar instance (Hdt. 7.9.2), such clauses leave ‘open the possibility that the terrible situation will not be realized’ and are therefore somewhat ‘more cautious’ than outright

122

The Linguistic Characterization of Electra

In concluding our earlier survey of Electra’s interaction with the disguised Orestes (§3.1), we saw that, in their stichomythia, Electra’s linguistic behaviour was marked especially when her own exile or Orestes’ absence were at stake—issues that were consistently linked with the continuing outrage of Agamemnon’s unavenged murder. In addition, the first clear signs of her vehement anger at her mother and at Aegisthus—both as murderers of her father and as authors of her own degradation—became apparent. In her rhēsis each of these elements recurs, as Electra integrates them into a formidably crafted bulletin designed to inspire her lost brother to make his glorious return. That she adapts her complaints to serve this single purpose is crucial for our interpretation not only of the speech (which, as a piece of targeted rhetoric, cannot be seen as a mere tissue of lies and delusions), but of Electra’s character as well. In addition to further exploring her motivations for the matricide, Euripides shows us how his protagonist—a female protagonist—goes about achieving her aims. ‘As a woman,’ Cropp (2013: 10) points out, ‘Electra uses words as her chief weapon’: if she is to inspire Orestes to return, they are in fact her only weapon.

4. RECOGNITION AND PLANNING (487–698) Few scenes in Greek tragedy have been the topic of more, and more vehement, discussion than the second episode of Electra, which centres on the moment of recognition between the siblings and deals with the subsequent planning of the murders. Set against the mountain of scholarship on this one scene,105 what I will have to say about it may come across as cursory. There are two reasons for this. First, the principal issues that scholars have confronted in dealing with the scene—the authenticity of lines 518–44 and the question negative judgements: Electra gives Orestes the opportunity to prove himself as manly as she would like him to be. 105 This applies more to the first part of the episode than to the second, postrecognition part. ὄνομα δ᾽ ἑκάστου διατριβὴ πολλὴ λέγειν (‘to tell the name of each would be a great waste of time’, Phoen. 751), so rather than provide a full doxography I will simply list those works that I have found most instructive, apart from Cropp ad locc.: Basta Donzelli 1980b, Bain 1981a, Goldhill 1986a: 247–59, Davies 1998, Gallagher 2003, and Thévenet 2009: 149–56. See also Introduction, n. 112.

The Linguistic Characterization of Electra

123

how the intertextuality between this passage and the corresponding scene in Aeschylus’ Choephori is to be interpreted—are not matters that are naturally tackled using the linguistic approaches I have adopted,106 other than to note that no convincing linguistic or structural argument has been offered against the authenticity of 518–44.107 Second, even with regard to aspects of the scene that are more directly relevant to the characterization of Electra, which is the topic of this chapter, I have little to add to a consensus that seems to have formed around the idea that the scene highlights Electra’s authadeia, the wilfulness that will drive her to matricide, and her particular style of sophia, a cleverness with argument that will mark her behaviour again in the ‘hesitation scene’ (on which see ch. IV, §3).108 It remains for me to make a few observations on points that clarify Electra’s characterization in the scene and that have not, I think, received as much attention as they deserve. First, there is a structural feature of Electra’s rejection of the tokens that has gone largely unnoticed. Electra’s use of discourse-structuring devices and of moods will allow us to trace this structure. The lock of hair, both in its introduction by the Old Man and in its rejection by Electra, is combined with a broader discussion of the very possibility of Orestes’ return. The Old Man first very tentatively (ἴσως που 518, ‘perhaps’) raises the general notion (518–19), and then asks Electra to verify it through use of the first token (520–3).109 Electra’s rebuttal proceeds in two stages: first comes a rejection of the general idea that Orestes has returned in secret (οὐκ ἄξι᾽ ἀνδρός . . . σοφοῦ λέγεις, | εἰ κρυπτὸν . . . | δοκεῖς ἀδελφὸν τὸν ἐμὸν . . . μολεῖν 524–6, ‘what you say is not seemly for a wise man, if you think that my brother would come in secret’), and then a more specific rejection of the token. The separation 106

I have briefly stated my view on these questions in the Introduction, §4.3. See Cropp on 518–44, Basta Donzelli 1980b, and Davies 1998, in response to arguments put forward by Kovacs 1989. 108 For the links between the recognition scene and the hesitation scene, see Gallagher 2003 (I think Gallagher takes the notion of Electra’s sophistry too far, however, and places too much emphasis on the alienation of sympathy from Electra, which he sees taking place in the scene). For the characterization of Electra in the scene, see, apart from Gallagher, e.g. Cropp on 487–584, Halporn 1983, Davies 1998: 402 n. 57. 109 We cannot know for certain whether the Old Man actually presents Electra with some trimmings at this point, but the lack of demonstratives (no τήνδε in 520, for example) and the rest of the scene strongly imply that this token is just as hypothetical as the other two (see Cropp on 518–44, under ‘The lock (515–31)’). 107

124

The Linguistic Characterization of Electra

between these two points is made explicit by ἔπειτα (527, ‘moreover’), which indicates that Electra is moving on to a distinct new point. Lines 524–6 should therefore be taken not as a first argument against the first token, but as a rebuttal of the Old Man’s basic suggestion that Orestes is back; as such, these lines form the overarching framework against which each of the more specific rejections of the tokens should be read. Electra will ‘close’ that framework after the rejection of the last token, by dismissing the entire discussion (ἀλλ᾽ 545)110 and by revisiting the general question of Orestes’ return (545–6).111 When we now look at the dismissal of the lock in Electra’s remaining lines of her first rejoinder (527–31) and compare it with her other two rebuttals (534–7 and 541–4), a strong parallelism of argumentative structure becomes clear. Each of Electra’s rejections consists of an argument against the possibility of the token working in the particular case at hand, followed by a general dismissal of the reliability of the token in general. In each case, Electra moves from specific to general with δέ (530, 535, 542) and indicates by her use of verbal mood (optative with ἄν) that she is rejecting the very possibility of the token’s being a reliable criterion. To take the rejection of the fabric as a representative example, Electra first discounts the notion that such a token could work in the case of the siblings (541–2): οὐκ οἶσθ᾽, Ὀρέστης ἡνίκ᾽ ἐκπίπτει χθονός, νέαν μ᾽ ἔτ᾽ οὖσαν; Don’t you know that when Orestes left the land I was still a young girl?

And she goes on to dismiss the general possibility that such a token could work, even if it did exist (542–4): εἰ δὲ κἄκρεκον πέπλους, πῶς ἂν τότ᾽ ὢν παῖς ταὐτὰ νῦν ἔχοι φάρη, εἰ μὴ ξυναύξοινθ᾽ οἱ πέπλοι τῷ σώματι; And even if I had been weaving clothes,112 how could someone who was a boy then have the same garments now, unless clothing were to grow up along with the body?

110

See Introduction, §2.2 on this use of the particle. I do not have a solution for the textual crux in 546, but this does not affect the argument. 112 A very neat use of the imperfect in a counterfactual conditional in the past (the common notion that imperfects in such conditions must refer to the present and aorists to the past is false; see Wakker 1994: 146–50). 111

The Linguistic Characterization of Electra

125

What does this repeated structure tell us about Electra? Above all it indicates, I think, a certain sophistication in her rhetoric, which—as it turns out—she will repeatedly misdirect in the play. The sense that Electra always has a follow-up argument ready to undermine an opponent’s case will return in much grimmer fashion in the discussion between Electra and Orestes just before the matricide (there it will eventually sway Orestes; see ch. IV, §3.2), and in the agōn (particularly lines 1086–93; see ch. VI, §5.2). A second observation concerns the tone of the interaction between the Old Man and Electra. Much has been made of the ‘arrogance’ with which Electra treats the loyal old retainer (so, e.g. Arnott 1981: 185–6). There is no question that Electra’s disbelief is played up in part by the way in which she criticizes the Old Man for suggesting what turns out to be the truth. But the notion that she is ‘hostile’ to the Old Man strikes me as off the mark. From the moment of this character’s entrance, Euripides stresses two characteristics of the Old Man that militate against this. First, the Old Man himself constantly underlines the social distance between himself and Electra (πότνι᾽ ἐμὴ δέσποινά τε 487, ‘my lady and mistress’; ὦ πότνι᾽ 563, ‘lady’), but combines this with consistent expressions of affection (ὦ θύγατερ 493, ‘daughter’; ὦ παῖ 516, ‘child’; τέκνον 533, ‘child’; θύγατερ 563). Second, the old age and feebleness of the man are repeatedly emphasized (489–92, 503, 524, 554, 566)—traits that, incidentally, also serve to make Electra’s rejection of the Old Man’s suggestions somewhat plausible. These aspects shed some light on Electra’s treatment of the Old Man and seem to indicate, more than anything else, that their interaction is portrayed with a generous helping of humour. As Cropp (on 487–92) points out, such irreverently treated old men are amply paralleled elsewhere (e.g. Peleus in Andromache, Iolaus in Heraclidae), and the humour with which they are treated tends to lead to ‘a surprising reversal when they assert themselves effectively’. The two moments that have given rise to the accusation of hostility more than any others are 524–6 (cited above) and 568 (μὴ σὺ γ᾽ οὐκέτ᾽ εὖ φρονεῖς; ‘are you perhaps no longer in your right mind?’); but neither is straightforward. I have already observed (§3.1) that 524–6 does seem to indicate a brief ‘spike’ in Electra’s linguistic behaviour, but this has much to do with the subject matter: Electra’s idealization of Orestes’ return has been emphasized in this way before (275), and her reaction is instigated by a suggestion that runs directly counter to that idealization. In any case, the impoliteness of this particular

126

The Linguistic Characterization of Electra

moment seems to be relatively limited. As just remarked, the Old Man himself has explicitly stressed the social differential between himself and his mistress, and this has consequences for the way in which his face* needs to be attended to. The politic behaviour* in a conversation between these two should not require a great deal of deference from Electra, and the degree to which her criticisms constitute serious FTAs* is thus decreased. Electra uses some hedging in her criticism: she circumscribes his ‘mistake’ with the litotic and somewhat vague οὐκ ἄξια σοφοῦ . . . ἀνδρός (the negative formulation seems to suggest that the Old Man does not live up to a default expectation of his wisdom); another hedge is the formulation with an εἰ clause.113 In short, I agree with Cropp in his first edition that the tone here ‘need not be hostile’ (Cropp 1988 ad loc.). Similarly, Electra’s question in line 568 about the Old Man’s senses is phrased very carefully: the detail οὐκέτ᾽ seems again to refer to the retainer’s age; the formulation μή with the indicative is used for hesitant suggestions,114 and thus introduces something of a hedge. We may also note how, at this point in the scene, Euripides is mustering every possible touch of dramatic irony (note, for example, the pairing of ἀπών/ὤν in 564): Electra’s idea that the Old Man is the one who isn’t seeing clearly should be regarded in that light as well. Moving on to the planning scene, a vital moment for Electra’s characterization there can be elucidated through conversation analysis*. She has been silent for over 60 lines, and Orestes and the Old Man have been planning the murder of Aegisthus. When the topic turns to Clytemnestra, Orestes shows himself flummoxed as to how to proceed (646–7): ΟΡ. πῶς οὖν ἐκείνην τόνδε τ᾽ ἐν ταὐτῷ κτενῶ;115 ΗΛ. ἐγὼ φόνον γε μητρός ἐξαρτύσομαι.

113 For the politeness of such features, see Brown and Levinson (1987: e.g. 144–5, 217–18, 226) and Leech 2014: 193. 114 See, e.g. Rijksbaron 2002: §5 and Moorhouse 1982: 320. 115 Denniston’s suggestion that a question mark should be printed after πῶς οὖν, adopted by Diggle, is surely wrong: as an independent question, πῶς οὖν; in tragedy (found only in Euripides, at Med. 1376, Hipp. 598, 1261, Hec. 675, and Hel. 587, 1228, 1266) never seems to work in the elliptical way suggested by Denniston, but always to interrogate an immediately preceding statement. Moreover, the fact that Clytemnestra and Aegisthus are not together is exactly what Orestes and the Old Man have been discussing. The connotation might be that, if Orestes takes out Aegisthus first, he may lose his shot at Clytemnestra.

The Linguistic Characterization of Electra Or. El.

127

How then am I to kill her and him in the same instance? I will arrange our mother’s murder.

Considering the flow of the conversation, there can be little doubt that Orestes selects* the Old Man as the next speaker in 646. Three question–answer adjacency pairs* between Orestes and the Old Man have preceded and Electra has been quiet for much longer than that, so that, when Orestes poses his next question, the floor* should normally pass to the Old Man.116 Rather Electra self-selects*, interrupts, and even takes over the reins of the conversation (she starts distributing speaking turns, and can select* the Old Man at 651).117 It has been noted that this ‘sudden breaking in of Electra with her terrible resolve is highly dramatic’ (Denniston ad loc.): the effect of this, I think, is not only to highlight the idea that Electra will be personally very involved with the murder of her mother (which may have been new in the tradition: see Cropp on 647), but also to show the urgency and determination she feels in that matter. As such, this moment follows well on lines 276–81 (see §3.1) and foreshadows Electra’s purposeful and strong-willed posture in the ‘hesitation scene’ (see ch. IV, §3).

5. ELECTRA, AEGISTHUS, AND CLYTEMNESTRA

5.1. A play of halves? I will soon go on to discuss Electra’s language in the ‘second half ’ of the play, picking up at the moment Aegisthus’ corpse is brought onstage (the intervening messenger scene will be treated separately in chapter V). By way of a brief interlude, it is worth noting that, in various accounts of Electra’s character, references to her behaviour in the post-recognition scenes are either lacking or much less extensive than the intricate dissections of her behaviour in the early scenes.118 116 There is, of course, no way in which we can be entirely sure that this was the case in the original performance, where Orestes may have indicated selecting* Electra by a gesture. But this seems unlikely. 117 Electra’s control over the conversation and her sense of urgency will return later on, in the scene at 684, if that line is indeed hers, as I will argue in chapter IV, §2. 118 e.g. Conacher 1967 and Lloyd 1986.

128

The Linguistic Characterization of Electra

This may have much to do with a structural aspect of the play, which is discussed by Michelini: The initial delay of the recognition permits a sustained focus on the country setting and its dominant figures [and as such on ‘Elektra’s personality and her sham marriage’]. After the recognition, the vengeance plot proceeds relatively directly to its ends, with the focus shifting to confrontations between the siblings and their two adversaries. (Michelini 1987: 181)

However, as Michelini points out immediately afterwards, the play itself is in no real sense more bipartite than other dramas of recognition (anagnōrisis) and conspiracy (mēchanēma). The themes of the second half are already present in the first, just as the themes of the first half find their interpretation and development in the events of the second.

Indeed, when it comes to Electra’s language, many of the dominant features identified so far can be seen in the post-recognition scenes as much as in the first half of the play, though the fact that Electra is now pitted against Aegisthus and her mother results in a more central role for her resentment towards these two. I will focus, in the remainder of this chapter, on the material in and around her two major rhetorical set pieces: first her speech over Aegisthus’ corpse (907–56), and then her contribution to the agōn (1060–96). I will end with a brief discussion of Electra’s language after the matricide (1172–359).

5.2. The ‘kakology’ (907–56) 5.2.1. Electra’s ‘undramatic’ generalizations As Orestes and Pylades enter leading Aegisthus’ body119 on stage, Electra welcomes them (880–9) in terms that continue the athletic imagery that has become prominent in and immediately after the messenger speech (ὦ καλλίνικε 880, ‘glorious in victory’).120 The male overtones of the welcome are enhanced through references to fighting 119 Not his decapitated head, as was suggested by Paley on 894 (followed by Denniston on 856, 894–5 and by Sider 1977: 16–17; Raeburn 2000: 161–2 is ambivalent): this is a silly notion, unsupported by the Greek, and it should have been laid to rest by Kovacs 1987b, if not before, by Gellie 1981: 11 n. 2. 120 See Introduction, §4.2, and Swift 2010: 156–70.

The Linguistic Characterization of Electra

129

and conquest, as Orestes is compared to his father, the sacker of Troy (880–1). Orestes accepts his sister’s praise121 and offers her the opportunity to do with Aegisthus’ body as she pleases. Electra hesitates, needing a few lines of stichomythia before she can set aside her anxiety over popular envy and contempt—probably a particularly female anxiety.122 The course of action that worries her is not exposing Aegisthus to wild animals or impaling him (as Orestes has suggested, 896–7),123 but reviling him with words.124 In the end, Electra obliges and begins a long speech (907–56), which serves as a sort of inverted funeral eulogy (aptly referred to by various scholars as a ‘kakology’)125—inverted both in terms of moral appraisal (negative rather than positive) and in terms of the gender of its speaker (funerary speeches were the province of men).126 121 Although not before transferring it first to the gods (890), in language reminiscent of Agamemnon’s at his entrance at Aesch. Ag. 810–13 (also a victorious return). The second half of Electra is full of such intertextual moments, especially around entrances and exits, and often with interestingly reversed roles, which paint Electra as an ‘heiress’ of Aeschylus’ Clytemnestra: Electra welcomes her mother as she arrives in a chariot (988, cf. Aesch. Ag. 907), sends her off into the house with a note of sarcasm (1139–46, cf. Aesch. Ag. 958–74), etc. For these and other intertextual references, see Cropp 2013: 10–11 and ad locc. and Goldhill 1986a: 250. 122 For a similar moment at 300–2, see §3.2. Mossman (2001: 377) has argued that Electra’s trepidation here about popular censure (900–4) is typical of female speech (for gossip as a female ‘verbal genre’, see McClure 1999: 56–62); the speech to follow, as we shall see, is similarly patterned on gender. It may be significant in this respect that this scene is reminiscent of the stichomythia in the ‘Tapestry Scene’ at Aesch. Ag. 931–43 (and see previous note): both scenes feature characters (Agamemnon and Electra) afraid of committing hubris and afraid of popular φθόνος and ψόγος, who are nonetheless persuaded to proceed. As has been argued (see, e.g. Goldhill 1984: esp. 74–9), the scene in Agamemnon has its own complexities of gender. 123 And, according to Nestor in the Odyssey, exactly what Menelaus would have done had he gotten his hands on Aegisthus while still alive (Od. 3.259–61). Much of this passage seems to look back at Aegisthus as he appears in the third book of the Odyssey especially. Throwing Aegisthus’ corpse to the dogs is also what Electra tells Orestes to do at Soph. El. 1487–90 (‘which might be remembered by those who think Euripides’ Electra a less pleasant character than Sophocles’, notes Lloyd 1986: 16). 124 Schwinge (1968: 83–4) and Cropp (1988: on 900–6) are right to reject Denniston’s claim (on 900–1) that βούλομαι εἰπεῖν cannot mean ‘I want to speak’ (‘which would be λέγειν’, says Denniston): Electra does in fact refer to a speech all along, and Orestes’ reactions in the stichomythia refer to it as well. The aorist infinitive εἰπεῖν suggests that Electra is talking about the performance of the speech as a whole, rather than about the process of performing it; the difference is clear also from Orestes’ two imperatives, λέξον (901, ‘make your speech’) and λέγε (905, ‘start speaking’). For these aspectual values, see Rijksbaron 2002: §16.2. 125 Cropp (1988 on 907–56) seems to be the first to have used the term. 126 See Alexiou 2002: 104–8.

130

The Linguistic Characterization of Electra

The similarities between this speech and accusations in agōn sequences have often been noted,127 and it is therefore instructive to compare it to Electra’s agōn speech later on in the play. To start with, there are some obvious formal similarities in structure and in the rhetorical devices used, although, strictly speaking, the speeches belong to different oratorical genres, the first epideictic and the second forensic. Both speeches also provide further insight into Electra’s motivation for the killings of Aegisthus and Clytemnestra, explicitly (in the reasons that Electra gives) as well as implicitly (in the way she selects and filters information to present her argument); the speeches concentrate heavily on problems within the palace, particularly sexual issues, which is indicative of the extent to which the ‘issues central to the agonistic speeches . . . are . . . to do with . . . a self-destructive oikos’ (Mossman 2001: 377). Yet differences between the two speeches exist as well, and a crucial one was identified by Mossman: of the fifty lines of this speech, twenty and a half lines are general reflection (twenty-two if one counts 947–8, which have some generalizing force: 41 or 44 per cent). In Electra’s thirty-seven-line speech to Clytemnestra (deleting 1097–9 with Hartung), there are only five and one-third (14.4 per cent). Why is one speech so much more sententious than the other? (Mossman 2001: 377)

Indeed, if we look at the structure of the kakology, which is again very neat (each successive point marked by δέ), it turns out that nearly all of Electra’s overarching points are rubbed in with sententious generalizations, taking up a significant (and increasing) amount of rhetorical space: • • • • •

907–13: Preface 914–17: Orphaning of Electra and Orestes, killing of Agamemnon 918–24: Clytemnestra’s infidelity—gnōmai at 921–4 925–9: Mutual corruption of Clytemnestra and Aegisthus 930–7: Aegisthus as the lesser party in the marriage—gnōmai at 932–7 • 938–44: Wealth and nobility—gnōmai at 940–4 • 945–51: Aegisthus’ sexual hubris—gnōmai at 948–51 • 952–6: Conclusion—gnōmai at 953–6 127 e.g. Strohm 1957: 14, 82, Basta Donzelli 1978: 157, Lloyd 1992: 56, and Mossman 2001: 377 with n. 13.

The Linguistic Characterization of Electra

131

The speech has been condemned for this heavy-handed use of generalizations: Denniston (ad loc.) called it ‘[n]ot, as a whole, a very powerful speech,’ because of its ‘weak’ and ‘undramatic’ gnomic passages; Grube (1941: 309) thought Electra’s words ‘stilted, full of generalities, artificial’. Yet it is worth asking if there are explanations other than ‘stiltedness’ for why we find so many gnomic passages bunched together. Mossman’s own explanation for her figures is worthy of careful consideration: after rejecting differences of oratorical genre,128 subject matter, or interpolations129 as adequate reasons, she argues: in this first rhesis Electra is couching her speech in particularly careful terms because she is in a mixed group [i.e. male and female], and so finds it desirable to underline her conformity with generally accepted moral thinking by using platitudes; whereas in the second speech she is in an all-female group (comprising her mother, the Trojan slaves, and the female chorus), and so can speak more trenchantly in the absence of men. (Mossman 2001: 378) 128 Mossman (2001: 377–8 n. 14) argues that ‘[e]pideictic speeches are not by nature necessarily more sententious than agon speeches’, but without providing a great deal of support for that claim and without going into the peculiar subgenre of epideictic speech that we are dealing with—funeral orations (even if the present speech is a doubly inverted case, being delivered by a female, in scorn). What evidence is provided by extant funeral orations is inconclusive (if only because comparing one author to another, particularly if the first writes oratory and the second tragedy, is a treacherous exercise; for tropes common to epideictic oratory in general and to funeral orations specifically, see Carey 2007). The closest parallel for this kind of speech in Euripides is probably the funeral oration at Supp. 857–917, which has only a few generalizing lines at 892–5 and at the end, in lines 911–17 (about 15 per cent of the total speech); popular morality is, however, expressed earlier in the speech, although focalized through the honorands, namely Capaneus at 865–6 (note ἔφη) and Eteoclus at 879–80 (these are not generalizations by the orator). Lysias’ Epitaphios is similarly devoid of generalizations, featuring only a few in the last few sections (2.54, 2.73, 2.77–8, 2.79), even though, again, the wise opinions of the dedicatees and their ancestors are often phrased in general terms (e.g. 2.8, 2.9, 2.10, 2.23). Pericles’ funeral oration at Thuc. 2.35–46 is, however, relatively rich in generalizations by the speaker. In all, I agree with Mossman that the clustering of gnōmai in Electra’s speech is striking and requires explanation. 129 Lines 921–4 were deleted by Hartung, who also sought to delete 936–7; all of 932–7 were suspected by Wecklein; Bruhn deleted 941–4 (942–4 had been previously deleted by Vitelli, and 943–4 appear in Stob. 4.31.99, where they are attributed to Eur. Phoen.; but see ch. III, n. 25); and Kovacs (1996: 110–17) leaves very little of the speech intact. For defences, see Denniston ad loc. (grudgingly), Basta Donzelli 1991b: 107–13, Mossman 2001: 378 n. 15, Distilo ad locc., and Cropp ad locc. I too prefer to keep the speech intact, and more generally would take issue with the excessive tendency to athetize gnomic passages, as I will discuss in the next chapter.

132

The Linguistic Characterization of Electra

Indeed, as I have discussed in my Introduction (§2.4.1), sociolinguistic research has been very successful at showing that linguistic behaviour in mixed groups differs from that in single-sex groups*, and various scholars have argued that the speech of women in classical Greek drama follows similar patterns (although this does not hold equally for male speech, which means that, if Mossman is correct about Electra here, ‘this alone characterizes her as female, since men have no need to change their discourse in a mixed group’: 2001: 378). Yet it is useful to put Mossman’s findings on a somewhat more solid footing, by showing how Electra’s ‘platitudes’ help to ‘underline her conformity with generally accepted moral thinking’, and to what extent this indeed constitutes ‘couching her speech in particularly careful terms’. It is worth revisiting here a few aspects of gnomic language discussed in my Introduction (§2.5). As outlined there, Greek gnōmai (identifiable as such on the basis of linguistic and structural features) are often different from (modern) proverbs in the sense that they are not necessarily fixed utterances deriving from a stable corpus of wisdom sayings, but may rather serve as an ‘instant addition’ to that body of traditional and folkloristic wisdom, which is in constant flux—though such gnōmai would be ineffective if their content did not to some extent reflect moral ideas acceptable to the addressees. Thus Mossman can rightly say that ‘generally accepted moral thinking’ underlies Electra’s sententious generalizations,130 even if they were not necessarily established sayings. In a sense, therefore, Electra’s gnōmai function as an appeal to a greater authority (generally accepted thinking) than that of the speaker alone.131 This notion of authority is also instructive in analysing the social component of her generalizations, that is, their use in the specific ‘conversational’ context. Electra’s generalizations in the kakology are all second-person generalizations* in that they apply to Aegisthus; they are, however, a peculiar kind of second-person sayings in that the addressee is a corpse.132 Lardinois, following Seitel, 130 All of Electra’s sententious utterances in the play appear to be actual commonplaces, i.e. to reflect actually accepted principles of Greek morality: each notion expressed is amply paralleled in other texts. For good lists of sources regarding the gnōmai in this particular speech, see Cropp ad locc. 131 See Norrick 1985: 28–9. 132 Electra is well aware of these special circumstances (on which more will be said in §5.2.2).

The Linguistic Characterization of Electra

133

showed that second-person sayings* are (in Homer at least) generally reserved for speakers of superior authority, except in some contexts. Seen in this light, Electra’s gnomic utterances reveal much about the unusual context in which she utters her speech. Orestes’ final phrase before handing over the body to be abused, σὸς γὰρ ἐστι νῦν | δοῦλος, πάροιθε δεσπότης κεκλημένος (898–9, ‘he is now your slave, where before he was called your master’),133 proves important: Electra is at last in a position of superiority over Aegisthus, and her consistent use of generalizations to rub in her accusations bears this out. Each of the gnōmai functions as an assertion of moral authority deriving from the prevailing ethical norms, which Electra uses to underline the notion that Aegisthus’ conduct was in contravention of those norms. The very fact that Electra is able to condemn Aegisthus’ behaviour to his face—even if only by way of platitudes—is proof of the new power relationship: only an emasculated, powerless male could be subjected to this treatment by a woman. Aegisthus, being dead, is just that. Still, the question remains why Electra feels the need to follow each of her explicit accusations with a supporting gnōmē, and this is where Mossman’s notion of ‘couching’ comes in. As noted in the Introduction (§2.5), generalizations are by their very nature an indirect way of speaking: such indirectness in conversation may serve a range of functions, including linguistic politeness, ‘allowing the speaker to disguise his true feelings, to leave himself an escape route . . . avoiding double binds . . . e.g. when they are called upon for a judgment that might hurt another’s feelings or reveal their own private preferences’ (Norrick 1985: 27). It may seem that there is a paradox between the authoritative nature of gnomic speech and this noncommittal nature of indirect language use, but Norrick has argued that this paradox is in fact unreal: By invoking tradition and the community as a whole, the speaker not only disappears as an individual directive agent, he also imposes the weight of social sanctions. So the traditional character of proverbs imbues their ideational meanings with authority and lends their directive interactional meanings force, while allowing the speaker himself to fade into general community opinion. (Norrick 1985: 27–8)

133 Orestes’ use of κεκλημένος here contains, I think, an indignant note: Aegisthus was called Electra’s master merely by virtue of his power, not due to any inherent claim to such a role.

134

The Linguistic Characterization of Electra

I believe that this is exactly what is going on in our speech. It is unlikely that Electra’s strategy is to be polite to Aegisthus in order to mitigate the impact of her revilings: in fact each of her generalizations follows on a more direct (bald*, on-record) accusation or insult. The gnōmai underpin rather than mitigate the revilings,134 and sometimes even expand on them. Yet Aegisthus is not the only person on stage, and Electra has shown herself fully aware that her position in making this speech is precarious: she is ashamed (αἰσχύνομαι 900) to revile a dead man and fears public censure and envy in the πόλις (902, 904). In such a position Electra may seek, unsurprisingly, to signal her conformity with traditional norms. In other words, Electra’s need to suppress her responsibility for her own claims by appealing to traditional morality is brought about by the precariousness of the kind of speech she wants to make. I also believe, however, that Mossman is right to see the presence of males on stage as a factor, since the very act of making a speech on an occasion like this is contrary to female propriety. It is also noteworthy in this regard that most of Electra’s gnōmai evaluate the moral standards of proper male behaviour (and concerns about female excess),135 another reason why Electra, as a female, may want to embed her accusations in a framework of popular (male) morality. Both these aspects—the hazardous nature of her speech and the mixed company in which she utters it, neither of which is present in the case of Electra’s agōn speech—may help to explain the heavy use of generalizing speech in the kakology: by resorting to the prevailing axioms of Greek (male) society, Electra can behave like ‘one of the boys’. Going through the speech and lifting out a few examples may illuminate these points.

134 The relationship between the individual gnōmai and the accusations preceding them is in each case made explicit by a connecting particle: ἴστω δ᾽ (921, a contrast between what Aegisthus believed and what someone should believe); καίτοι (932, a particle that adds a point which reflects back on the previous, contrasting point: see Slings 1997a); τὰ δ’ οὐδέν (940, again a contrast between what Aegisthus thought and what is sound belief); ἀλλ᾽ ἔμοιγ᾽ (948, opposition between Aegisthus’ behaviour and Electra’s ideal husband). The final gnōmē is a concluding evaluation that generalizes from Aegisthus’ ignorance (ὧδε 953, ‘thus’) the established notion that criminals will face justice in the end. 135 The speech ‘neatly expresses fifth-century Athenian concerns about women usurping control of the household and especially the production of heirs’ (Ormand 2009: 261, comparing relevant passages from pseudo-Demosthenes’ (= Apollodorus’) speech Against Neaera).

The Linguistic Characterization of Electra

135

5.2.2. Analysis of the speech After accepting the granted permission to speak (with εἶἑν),136 Electra begins with the common rhetorical technique of noting the wealth of material available to her to use in her accusation as well as the magnitude of her task (907–8).137 But if her rhetorical questions (‘Where to start? Where to end?’) give rise to an expectation that she will actually have difficulty putting together a speech, she immediately forestalls any such expectation (καὶ μήν)138 by noting that she has often rehearsed exactly what she wants to say (909–13): καὶ μὴν δι᾽ ὄρθρων γ᾽ οὔποτ᾽ ἐξελίμπανον θρυλοῦσ᾽ ἅ γ᾽ εἰπεῖν ἤθελον κατ᾽ ὄμμα σόν, εἰ δὴ γενοίμην δειμάτων ἐλευθέρα τῶν πρόσθε. νῦν οὖν ἐσμεν· ἀποδώσω δέ σοι ἐκεῖν᾽ ἅ σε ζῶντ᾽ ἤθελον λέξαι κακά. But be sure that I never stopped rehearsing, in the twilight hours, exactly what I wanted to say to your face, if I should actually become free of my former fears. Well, now I am free; and I will settle my score with you, reviling you as I wanted to while you were alive.

Electra here shows herself well aware of the peculiarity of addressing a speech to a corpse. She will instead treat the speech as one addressed to a live Aegisthus. In a way, these lines also suggests that she is setting out to make a forensic speech—to formulate what, during his lifetime, could have served as a criminal accusation. As it turns out, however, the forensic ‘flavour’ she imparts to her speech lasts only for a very short time: the shift to the main body of the speech is marked by asyndeton, and in four lines (914–17) Electra briefly lists the crimes Aegisthus has committed against her family—orphaning her and Orestes, adultery, and the murder of Agamemnon. As she did in her appeal to Orestes (300–38), Electra points out the contrast between Aegisthus’ victim Agamemnon, a paragon of manliness (ἄνδρα . . . στρατηλατοῦνθ᾽ Ἕλλησιν 916–17, ‘a man who commanded the Greeks’), and his own cowardice (the next three words, οὐκ ἐλθὼν

For the function of εἶἑν, see my discussion in ch. IV, §3.1. Parmentier compares Eur. IA 1124–6 as well as Hom. Od. 9.14–15 and Gorg. 82 B11a4 DK. 138 For the counter-expectational value of μήν, see Wakker 1997a and van Emde Boas in press-a. For the collocation καὶ μήν, see also van Erp Taalman Kip 2009. 136 137

136

The Linguistic Characterization of Electra

Φρύγας, ‘not having gone to Troy’, carry great scorn).139 This theme of unmasculine behaviour is what Electra then goes on to develop in the remainder of the speech, painting an intimate picture of Aegisthus’ disgraceful lifestyle in the palace. The shift away from the brief forensic tone (by which I mean the listing of crimes in what looks like a criminal accusation) towards a more epideictic one (that is, one that provides a portrait of Aegisthus’ life and character) is evident from Electra’s use of tenses. The summation of his misdeeds at 914–17 is given in the aorist (ἀπώλεσας 914, ἔθηκας 915, (ἔ)γημας and ἔκτανες 916): this tense often serves in introductions and summations in forensic oratory, where it describes actions from the viewpoint of the orator, as he is delivering his speech. The ‘deictic centre’ of these aorists lies in the present of the performance,140 while the past actions described in this way reflect upon the present determination of guilt or innocence. One more sentence in the aorist follows (ἐς τοῦτο δ᾽ ἦλθες ἀμαθίας ὥστ᾽ ἤλπισας 918, ‘you reached such depths of ignorance’), but then the mention of Aegisthus’ ἀμαθία sets Electra off,141 and the main thrust of the rest of her speech is to characterize Aegisthus through a long string of imperfects (ᾤκεις 925, ᾔδησθα 926,142 ἀνῃρεῖσθον 928, ἤκουες 930, ἤπατα 930, ηὔχεις 939, ὕβριζες 947). These imperfects take us, as it were, into the palace and sketch Aegisthus’ debauched lifestyle as a series of ongoing or repeated behavioural tendencies. The general picture that Electra presents is one of a fool (thus the terminology of intellectual capacities, or rather of their lack, is rampant),143 and of a man severely wanting in proper masculine qualities.144 139 The theme of Aegisthus’ cowardice is traditional, cf. Hom. Od. 3.262–4 (where Nestor is equally scornful about him staying behind in ‘a quiet little corner of Argos’ while other Greek men went to Troy), 3.310, Aesch. Ag. 1625–7, 1633–5, Soph. El. 300–2. 140 For this value of the aorist, see Bakker 2010b (with further references). 141 For the moral component of this term, see my discussion of 294–6 in ch. III, §2.2.2. 142 Morphologically a pluperfect, but in the case of the verb οἶδα this is a meaningless distinction. 143 e.g. ἀμαθίας 918, ‘ignorance’, δύστηνος . . . εἰ δοκεῖ . . . 923, ‘a poor fool to think that’, ἠπάτα σε . . . οὐκ ἐγνωκότα 938, ‘deceived you, though you did not know it’, σκαιῶν 943, ‘fools’, οὐδὲν εἰδώς 952, ‘completely ignorant’. The phrase οὐ δοκῶν at 925 probably refers not to what Aegisthus thought but to the impression he tried to give (with δοκέω in the sense of ‘seem’ rather than ‘think’; this cannot be determined from the syntax, but is strongly suggested by the following ᾔδησθα γάρ . . . ἀνόσιον γήμας γάμον ‘for you knew that you had entered into a sinful marriage’: few translators see this). 144 Thus we find a great many suggestive collocations of men/women and masculinity/feminity (Ὁ τῆς γυναικός, οὐχὶ τἀνδρὸς ἡ γυνή 931, see n. 146; γυναῖκα . . . ἄνδρα

The Linguistic Characterization of Electra

137

Electra’s first two points about Aegisthus’ folly deal with his marriage to Clytemnestra, and this allows her to cast aspersions on her mother (the principal object of her hatred) as well as to insult Aegisthus. It also casts Electra once again as a female speaker, concerned primarily with issues of marriage and sexuality. Aegisthus was, according to Electra, a fool for thinking that he would have a trustworthy wife in Clytemnestra (918–20), a point she rubs in by noting that an adulteress never makes for a good wife (921–4); incidentally, this gnōmē not only buttresses Electra’s point, but has the added effect of insinuating that Aegisthus was forced to marry Clytemnestra (ἀναγκασθῇ λαβεῖν 922).145 This is a clever manipulation of the indirectness of gnomic speech: by expanding the content of the gnōmē beyond what is immediately applicable—that is, beyond the explanation*—Electra can enhance her account with a particularly devious suggestion, without having to be explicit about it. Next she accuses Aegisthus of being the lesser party in the marriage (930–1)146 and invokes popular morality to condemn this as something shameful (932–8); again, her gnōmai both underline her point and expand it with innuendo, this time about the children that Clytemnestra has borne to Aegisthus (933–5; we have been told about these children at 62–3). The remaining two points of Electra’s sketch highlight aspects of Aegisthus’ character that were traditionally associated with tyrants and usurpers. Electra first turns to his reliance on wealth: Aegisthus ‘prided’ himself ‘on being someone’ (ηὔχεις τις εἶναι 939) on the basis of possessions, but Electra’s insistence that he was deceived by this notion (ἠπάτα σε 938) implies not only that this connection between wealth and worth is false, but also that Aegisthus was a non-entity all along. In her gnomic reinforcement, Electra combines a number of ethical notions related to the contrast between wealth (especially 932, τοῦ μὲν ἄρσενος πατρός . . . τῆς δὲ μητρὸς 934–5, τἄνδρος μὲν οὐδείς, τῶν δὲ θηλειῶν λόγος 937, μὴ παρθενωπός ἀλλὰ τἀνδρείου τρόπου 949). 145 This is an unexpected twist, but ‘a good insult’ and ‘no ground for deletion’ (Cropp ad loc.). 146 The phrase Ὁ τῆς γυναικός, οὐχὶ τἀνδρὸς ἡ γυνή (931) is far from simple, although it is clear what is meant: as Slings (1997c: 158) points out, the phrase ὁ τῆς γυναικός cannot mean ‘the man’s his wife’s’ (Cropp’s translation, explained in his note ad loc., but I doubt that ‘ὁ τῆς γυναικός ’ can be ‘easily understood’, as Cropp argues). The line is probably to be construed along the following lines: ‘[you were called] “the woman’s husband” ’ (with substantivizing article), and ‘the woman was not called “the man’s” (with ἤκουε implied as verb with ἡ γυνή)’.

138

The Linguistic Characterization of Electra

unjustly acquired wealth) and nature—ideas that have taken on a special significance in this play due to previous musings about wealth and nature by Orestes (367–90) and the Peasant (426–31). Electra’s clinching insult (945–9) is the most emasculating of all, and equips Aegisthus with the typically tyrannical quality of sexual hubris and with the typically adulterous one of pretty looks.147 The striking praeteritio with which she introduces this topic (ἅ δ᾽ ἐς γυναῖκας—παρθένῳ γὰρ οὐ καλὸν | λέγειν—σιωπῶ 945–6, ‘as for your dealings with women—as it is unseemly for a maiden to speak of such things—I’ll keep my tongue’) is a subtle device. It again aligns Electra with standards of proper behaviour, in this case proper speech; but it also allows her to refer to herself as παρθένος—a status that she can now fully reclaim and that Aegisthus tried to destroy (31–9). As Mossman (2001: 379) notes, this serves to ‘reinvent Electra’ as a woman released from her ‘deathly’ marriage to the Peasant. In this light, it is unsurprising that the following generalizing lines about good husbands (948–9) are phrased in a more personal fashion (ἔμοιγ᾽ εἴη . . . ): the lines function as a second-person saying* that indicts Aegisthus (he is not the type of husband and father one needs), but they can also be taken literally (Electra is ‘available’ again).148 Asyndeton and Electra’s coup de grâce ἔρρ᾽ (‘off with you’) round off the body of the speech.149 Her conclusion—Aegisthus didn’t know what hit him (952–3)—serves both as a final jibe and as a neat summation of the ἀμαθία that Electra has been criticizing throughout the speeech.150 This final reflection on ignorance brings Electra to 147 Sexual outrages: cf. e.g. Hom. Od. 22.37 and Aesch. Ag. 1438, Eur, Supp. 452–4 (with Collard ad loc.); girlish looks: cf. e.g. Hom. Il. 3.54–5 and Fraenkel on Aesch. Ag. 1625ff. 148 The line may anticipate Electra’s eventual marriage to Pylades, cf. 1249, 1284, 1311 (a constant feature of the Euripidean treatment, cf. IT 695–6, 915, Or. 1092–3). 149 The exclamation is addressed to people only at Med. 1346 (Jason to Medea) and at Soph. OC 1383 (Oedipus to Polynices); Hermione uses it when she casts off her veil (Andr. 830, in lyrics): this low frequency may suggest a particularly violent tone. Asyndeton is regular with exclamations, and indeed no connection is ever found with ἔρρε in Euripides. 150 The sense depends on how we interpret οὐδὲν εἰδὼς ὧν . . . δίκην δέδωκας. I agree with Denniston ad loc. that ὧν is difficult, but I would prefer to keep it (rather than changing to ὡς), interpreting it (with Cropp ad loc.) as ‘knowing nothing of what . . . you have paid the penalty for’, i.e. Aegisthus is presented as unaware of his crimes rather than of his punishers: syntactically this is indeed the most likely reading (δίκην δίδωμι + gen.), and Electra is here presenting an abstraction (χρόνῳ 952), not

The Linguistic Characterization of Electra

139

a concluding evaluation on Aegisthus’ life (953–6), which again reflects a notion deeply engrained in Greek thought: no mortal should consider him-/herself safe or happy until he/she reaches the end of his/her life. Again, Electra is careful to suggest her own conformity to such ideas: the ‘ethical’ dative μοι at 954 makes the alignment explicit. As we have seen, each of Electra’s generalizations serves to underline an aspect of the perversity of Aegisthus’ character. Her constant appeals to popular morality allow her to drive her points home, at times even enlarge on them with clever innuendo. At the same time, they signal her membership of that part of society that has its values straight, and this is directed as much at Orestes, Pylades, and the Chorus as it is at Aegisthus. Electra shrouds herself in the cloak of popular authority: she is, she would have us believe, not a woman speaking inappropriately at an inappropriate occasion, but rather a voice for the same morality by which they all abide. Finally, a few remarks on Electra’s ‘selection’ of arguments are in order. We have seen that the approach she takes in the speech changes very quickly from a listing of crimes to a more intimate sketch of Aegisthus’ contemptible character. Several scholars have taken issue with this structure and consider it unsympathetic. Conacher, for example, impugns Electra as follows: we hear not a furious indictment of her father’s murderer (this occupies but two of some fifty lines), but a sarcastic tirade on the folly of those who marry their adulterous mistresses and on the plight of those who marry wives more illustrious than themselves. . . . as elsewhere in the play, the main impetus to Electra’s bitterness stems from her own outcast situation . . . Thus . . . the poet once again refuses his Electra the opportunity of making a ‘sympathetic’ impression on the audience . . . (Conacher 1967: 207–8)

Others have gone one step further and considered the whole speech a web of outright lies (Arnott 1981: 183–4). Certainly Conacher is right in linking the organization of Electra’s speech to her character: the centrality, to the speech, of issues of marriage and sexuality (female issues) is reflective of the personal injustice Electra feels she has

Orestes, as the agency behind Aegisthus’ fall. Still, I suspect that the lines are phrased generally enough to allow the notion that Aegisthus didn’t know who got him to come to mind; he was, after all, struck from behind (839).

140

The Linguistic Characterization of Electra

endured in exactly these areas151 and, when she accuses Aegisthus of luxuriating in wealth, one should bear in mind that it is her and Orestes’ inheritance he was usurping.152 I have also mentioned above that the focus on palace life allows her to implicate Clytemnestra along with Aegisthus.153 In short, the speech has an intensely personal flavour (no matter how generalizing it gets), and in this respect Electra’s language in it continues to show the same features as we have seen throughout the play. Yet, much as Electra’s account reflects her own personality, it also says something about that of Aegisthus: each of the characteristics for which Electra condemns him is part of his traditional persona (as found in the Odyssey and in Aeschylus’ Agamemnon; see Cropp on lines 907–56), and typical of tyrants and adulterers more generally. There is no reason to believe that Electra is inventing his crimes,154 or that he would not have been seen by the audience as a despicable character indeed. We do not, then, have to be sceptical of the Chorus’ evaluation, which caps Electra’s speech (957–8): ἔπραξε δεινά, δεινὰ δ᾽ ἀντέδωκέ σοι καὶ τῷδ᾽· ἔχει γὰρ ἡ Δίκη μέγα σθένος.

151

See Michelini 1987: 214. See Mossman 2001: 378–9: ‘The compression of the arguments [in 940–4], and perhaps too the number of end-stopped lines here, gives the impression of a surge of resentment (the possessions of which she speaks were presumably Agamemnon’s and should have been Orestes’—and her own dowry—so this section has affected her more personally than what she has been saying since 917).’ Cf. also 1088–9, Electra to Clytemnestra: πῶς οὐ πόσιν κτείνασα πατρῴους δόμους | ἡμῖν προσῆψας; ‘How is it that you didn’t, after you killed your husband, attach the house of our father to us?’ 153 For Grube (1941: 309), this is in fact the only reason why Electra touches on these subjects at all: ‘Aegisthus was always incidental’. 154 Arnott (1981: 183) thinks that ‘[t]he very fact that Electra’s view of him accords so well with the traditional view should in itself give us pause’. Other views of him in the play (in the prologue and messenger speech), Arnott claims, belie Electra’s account and cast Aegisthus as a forceful and at times sympathetic character. But all this is predicated principally on a belief in Euripides’ antagonistic attitude to tradition, a notion that Arnott takes much too far (no view which sees the relationship between Electra and tradition as anything but complex captures the whole truth; see Introduction, §4.3). Moreover, when we consider the reaction of the internal audience to Aegisthus’ death, it is surely significant that no one at all (not the Chorus, not Castor) shares Arnott’s reading. Finally, as I will argue again in ch. V (§3.4), Arnott’s idea of an antithetical ‘double vision’ about Aegisthus does not hold up to scrutiny: in the prologue, too, he can be dominated by Clytemnestra (27–8), and both there and in the messenger speech he has typically tyrannical features (fear of being deposed, illegitimacy). 152

The Linguistic Characterization of Electra

141

He did terrible things, and a terrible payment he has rendered to you and Orestes here. Justice, after all, possesses great power.

5.3. ‘Into the boudoir’: Electra and Clytemnestra (998–1146) After having bullied Orestes into going through with the matricide (for the ‘hesitation scene’ at 959–87, see ch. IV, §3), Electra comes at last face to face with her mother. The Chorus women welcome Clytemnestra with a greeting in anapaests (988–97), which is typical of similar scenes featuring prominent figures arriving in chariots (see Cropp ad loc.). In particular, the fact that it recalls a comparable scene in Aeschylus’ Agamemnon—Agamemnon’s arrival with his Trojan spoils (783–809)—heightens the dramatic irony, which the women are already laying on thick with their hyperbolic honorifics emphasizing Clytemnestra’s wealth (πλούτου 995) and vaunted status, and with their comment that ‘the right time has arrived for attending to your fortunes’ (τὰς σὰς δὲ τύχας θεραπεύεσθαι | καιρός 996–7).155 The parallelism with the death of Agamemnon also raises questions about the killing to follow: ‘the opprobrium of the first murder is transferred to the second’ (Zeitlin 1970: 657).

5.3.1. Opening exchanges (998–1010) Clytemnestra, in her opening words (998–1003), expresses some pride in the luxury of her Phrygian entourage, though she is also quick to cast it as merely a small recompense for the loss of her daughter. Electra insists, however, on keeping her mother’s riches and status in the spotlight: her ‘greeting’ (1004–6) makes much of the queen’s luxurious circumstances and more still of the contrast with her own situation. Electra’s turn is a clever manipulation of the conventions of ‘commissive’ speech acts* such as offers (§2.3.1). Before her mother’s command to her Trojan slaves to give her a helping hand can be fulfilled, Electra intervenes and offers her own hand as well (οὔκουν ἐγώ . . . λάβωμαι μακαρίας τῆς σῆς χερός; 1004–6, ‘shall I not take 155 Line 997 is corrupt (the MSS give καιρός, ὦ βασίλεια, but a paroemiac is needed; see e.g. Diggle, app. crit., Distilo ad loc.), but καιρός is probably sound.

142

The Linguistic Characterization of Electra

your blessed hand’).156 The felicity conditions* for an offer of this kind would include first the existence of a differential of power and status that makes it appropriate for Electra to perform the service in the first place and, second, an intention to carry it out. Yet it is not obvious that especially the first of those conditions is met, and it is exactly this aspect that Electra exploits, by making explicit that the condition is in fact fulfilled: δούλη γὰρ ἐκβεβλημένη | δόμων πατρῴων δυστυχεῖς οἰκῶ δόμους (1004–5, ‘as a slave I live in this hapless house, having been cast out of my ancestral home’; γάρ explains the fact that she is making the offer in the first place). Electra’s intentions now become clear: the point of her gesture is not to make it, but to demonstrate that she can make it. Her words, superficially an offer, are merely the packaging of a bitter swipe at her mother,157 whose oriental spoils contrast painfully (and, no doubt, visually: see Zeitlin 1970: 647) with her own poverty. The sarcasm is heightened by the juxtaposition of her own ‘hapless house’ (δυστυχεῖς . . . δόμους 1005) and Clytemnestra’s ‘blessed hand’ (μακαρίας τῆς σῆς χερός 1006).158 Electra’s sarcasm is bitter, but Clytemnestra either misses or ignores it.159 Her answer (1007) picks up her daughter’s δούλη and points instead to her own δοῦλαι, as she remarks that Electra need spare her no trouble. The answer is noticeably curt, especially for a dispreferred* move (turning down an offer): the single line contains two abrupt utterances, including an unmitigated imperative with the bald* address σύ.160

156 For the use of οὔκουν with the first person in what is ‘virtually a command, delivered to herself ’, see Denniston ad loc. 157 A reading in terms of politeness theory* yields similar results: Electra’s phrasing of the οὔκουν question and her deference (calling herself a ‘slave’ and her mother ‘blessed’) constitute facework* well in excess of politic behaviour*, which may give rise to an interpretation of politeness, but in this case is taken so far that it may well be negatively interpreted as impolite (for ‘over-politeness’ and sarcastic politeness, see Culpeper 2011: 100–3, 178–80). 158 Denniston ad loc. is right to note that the position of the adjective μακαρίας, a ‘title of royalty’, gives it ‘increased importance’: see Dik 2007: ch. 4 and Bakker 2009b: 109–13. 159 I am more tempted by the latter option, for which see Denniston ad loc. (‘pointedly ignoring’). Michelini (1987: 218) favours the former (‘fails to notice’). Of course, Clytemnestra’s interpretation of Electra’s words need not necessarily be shared by the audience (which will surely sense the sarcasm). 160 The careful modulation of forms of address in this scene (transitioning from bald σύ to more affectionate τέκνον) has been noted by Griffin (in Mossman 2001: 382 n. 25). See also Kubo 1967: 26.

The Linguistic Characterization of Electra

143

Yet Electra will not let go of the point so easily, and is spoiling for a fight (1008–10). She rejects the validity of her mother’s objections (τί δ᾽;),161 and restates more explicitly the point that she had left somewhat implicit in her earlier turn:162 Clytemnestra’s treatment has left Electra in no different state from that of the Trojan slaves, which she prides herself on. Electra reinforces the parallelism with exaggeration, describing the palace at Argos as if it were sacked in war (ᾑρημένων . . . δωμάτων 1009) and herself as booty (αἰχμάλωτον 1008). The particle τοι, an explicit sign that what Electra says has bearing on Clytemnestra’s refusal, drives the point home. Once again, we see how Electra’s loss of status motivates her anger; she ends her turn, however, with another reference to Agamemnon (πατρὸς ὀρφανοὶ λελειμμένοι 1010, ‘bereft of a father’). These opening salvos between mother and daughter thus provide, in short order, a revealing look at the dynamics of their relationship: Electra’s resentment and Clytemnestra’s impatience. We shall find that, throughout this entire scene, Euripides needs only a few light brushstrokes to paint an interaction that is suggestive of the deeply charged history between the two women. The opening exchange also leads effortlessly to the agōn proper: ‘Electra’s mention of her lost father . . . stings the queen enough to provoke a reply’ (Michelini 1987: 219), and Clytemnestra’s defence speech then begins with two lines that ‘arise naturally from the dialogue, but also serve as a headline to the speech’ (Lloyd 1992: 61). Clytemnestra responds directly to her daughter’s complaints: Agamemnon himself is to blame for his own demise and for Electra’s other troubles (1011–12).

5.3.2. Clytemnestra’s speech (1011–50) I will analyse the language and rhetorical organization of the two agōn speeches in detail in chapter VI. For the purpose of this chapter, I merely wish to point to a few features of the speeches that illustrate the relationship between mother and daughter, and Electra’s characterization in particular. 161

See Denniston 1954: 175. That Electra is forced to repeat herself is a sublime touch in the exchange between mother and daughter, and certainly does not call for deletion (as proposed by Wilamowitz, 1883: 222 n. 1; and see the appropriate rejoinder in Vahlen 1891: 357–8). It is also not an ‘aside’ (Kubo 1967: 26)—an awkward suggestion in light of τοι (1008). 162

144

The Linguistic Characterization of Electra

First, it is noteworthy that in Clytemnestra, the ‘defendant’, speaks first in the agōn—an order that reverses forensic practice, but not without tragic effect. As Lloyd (1992: 101) puts it (comparing Helen in Troades), Clytemnestra ‘is so much on the defensive when the scene begins that a further prosecution speech would be superfluous’; and, like Helen, Clytemnestra in a sense ‘must plead for her life against a charge which has not been formally expressed’, even if the situation in our play is different in that Clytemnestra does not know that she has already been condemned to death (this makes the order of speeches itself ironical).163 The feeling that Clytemnestra, in arguing her case, is fighting against entrenched opinion is evident from the start.164 The first rhetorical move she makes after the formal opening of her speech (λέξω δέ 1013, ‘I will make my case’) is an elaborate captatio benevolentiae (1013–17), in which she employs the stock technique of mentioning the prejudices and disadvantages she faces in making her case:165 public opinion is against her, but she should be given a fair hearing.166 Apart from being a common rhetorical trope, Clytemnestra’s pre-emptive technique is very appropriate for the speaker and the occasion, as Lloyd points out: This rhetorical concern for reputation is also in keeping with Clytaemestra’s character. Her concern for public opinion has already been mentioned in the play (30, 643), and it contrasts with her truculence in Aeschylus (e.g. Ag. 1401–6). Even this rather formal agon reflects character to some extent, and the argument from prejudice also has a tragic point in that the case is indeed prejudged. (Lloyd 1992: 61)

163

For the unusual ordering of the speeches and Clytemnestra’s position of weakness, see also Dubischar 2001: 67, 83–96, 124–5. 164 The particle μέντοι (1011) contributes to this. Slings (1997a: 114–22) has well described this particle as functioning as a ‘denial-of-expectation adversative’: that is to say, the discursive move in which μέντοι is embedded modifies or controverts an expectation raised by the previous move. In conversation, it may mark a turn that counters implications just raised by the previous speaker; and, accordingly, Clytemnestra uses it here to signal that she will argue a point contrary to what Electra has just suggested: ‘I’m like a slave because I lost my father and patrimony’ / ‘Well, actually, it is your father who is to blame.’ 165 See ch. VI, n. 4. 166 For the interpretation of 1014, see Cropp 1982: 52–4 and 1988: on 1013–18. I agree with Cropp that Diggle’s κακῶς (for L’s καλῶς) is a most unfortunate change. (Cropp has exercised damnatio memoriae in his 2013 edition, never mentioning κακῶς either in his critical apparatus or in his commentary.)

The Linguistic Characterization of Electra

145

In short, Clytemnestra shows herself well aware that she is going against the grain, and even that her daughter and the wider public are liable to ‘abhor’ and ‘hate’ her (μισεῖν, στυγεῖν 1016–17). It is, in this light, not surprising that she has been seen to ‘make the most of a weak case’ (Cropp on 1011–50), largely on the basis of hypothetical arguments and appeals to lesser evils (for the style of Clytemnestra’s argumentation, see ch. VI). The two lines that conclude Clytemnestra’s speech (1049–50) are also indicative of the fractious relationship between mother and daughter. She explicitly gives Electra permission to speak (λέγ᾽ . . . κἀντίθες παρρησίᾳ 1049, ‘speak and make your rebuttal freely’), but also subtly tries to shield herself from attack by delineating the question to which Electra should attend in her speech (ὅπως τέθνηκε σὸς πατὴρ οὐκ ἐνδίκως 1050, ‘in what way it is not right that your father is dead’)167—perhaps the one topic that she feels somewhat confident she can defend herself on. Electra, of course, will not oblige and will attack her mother along entirely different lines.

5.3.3. Parrhēsia (1055–9) The material between the two speeches is more elaborate than in most other Euripidean agōn scenes: we would normally expect Electra to begin her half immediately after the Chorus’ ‘punctuating comment’ at 1051–4 (which is regular, if also somewhat longer than usual).168 Underlying this ‘disruption’ of the formal pattern is a delay in the conversational structure of the scene: Electra uses two insert expansions* (1055–9), in which she asks for extensive and repeated verification of her parrhēsia before acceding to her mother’s request to speak.169 Electra’s reluctance recalls earlier moments in the The negation in οὐκ ἐνδίκως is perfect for Clytemnestra’s goals: the fact that Electra has to explain how Agamemnon’s death was not just implies that the ‘default option’ (the one that will be upheld if Electra fails to persuade) is that the killing was just. 168 See Lloyd 1992: 65. The fact that the justice of Clytemnestra’s words is immediately questioned by the Chorus at 1051 ‘does not come as a surprise and is in itself an indication of the perspective the audience is meant to adopt’ (Dubischar 2001: 324, kommt . . . nicht überraschend und ist auch selbst ein Hinweis auf die auktorial intendierte Rezeptionsperspektive). 169 Clytemnestra’s command at 1049–50 in a sense turns a formal requirement (that one agōn speech is immediately followed by the second, apart from a choral comment) into a conversational one (by making the first speech the first part* of an adjacency pair*). 167

146

The Linguistic Characterization of Electra

play where she hesitated before speaking (300–2, see §3.2; 900–5, see §5.2.1);170 but, unlike before, her trepidation here cannot be ascribed to the fact that she is in the company of men. In fact the very use of the word παρρησία (not used in the other two scenes) is suggestive of the female-only context: [Parrhesia] is the freedom of speech associated with a healthy (male) political system and specifically with democracy as opposed to tyranny. That difference in expression, then, not only defines the contrasting relationships she has with Clytemnestra, her mother, but also a powerful queen whom she has alienated, and with Orestes, her loving and beloved long-lost brother: it also suggests a different gender atmosphere. How could either woman speak of parrhesia in front of a man? (Mossman 2001: 381)171

Electra’s reluctance here must thus be related to the superior power of her regal mother. Her insistence that her parrhēsia be confirmed has been variously explained: Mossman (cited above) and Kubo (1967: 26) seem to suggest that Electra is actually held back by the power differential between herself and Clytemnestra. Lloyd (1992: 66), conversely, thinks that Electra ‘is relishing the irony of the situation’. To me it seems clear, in any case, that Electra’s emphasis on the difference between her mother and herself is (as before) a deliberate reminder of her mistreatment, whether she is actually still afraid of her mother or not (later on in the scene, at 1122, Electra will play with the irony of her professed fear of Aegisthus).172 The repeated verification is also a signal on Electra’s part that she plans her rejoinder to be an all-out attack (she would, after all, have little to fear otherwise). As for Clytemnestra, her assurances show us the first signs of the more conciliatory posture she will adopt throughout much of the rest of the scene: she uses the more affectionate address τέκνον (1057, Note also her opening with the cautious potential optative λέγοιμ᾽ ἄν (1060), and cf. line 300 (with my discussion at §3.2, including n. 72). 171 For parrhēsia as a privilege of men (but occasionally conferred to women), see Sluiter and Rosen 2004, esp. Roisman’s contribution (ch. 5, with further references). Roisman (2004: 112) says about Electra that parrhēsia ‘is associated with her ruthlessness and bloody-mindedness’ and points out that Electra is one of the few women in tragedy who use parrhēsia for ends other than political (indeed, Electra’s speech focuses entirely on the oikos, not the polis). 172 O’Brien (1964: 21 n. 12) similarly suggested that Electra’s ‘confused’ and ‘stilted’ words in the kakology are an indication of her residual terror of Aegisthus. O’Brien (1964: 18) observes more generally that fear is ‘the prevailing mood of the play’. There is certainly some truth to this. 170

The Linguistic Characterization of Electra

147

‘child’) and rejects any suggestion that she might treat Electra harshly if she speaks too freely (1059).173

5.3.4. Electra’s speech (1060–99) The rhetorical organization and tactics of Electra’s speech will, again, be examined more fully in chapter VI, but a few points are worth making here. Electra’s insults in the opening of her speech (1060–8)174 do more than directly contradict points made by Clytemnestra—who has distanced herself from Helen (1027–9 and 1041–4) and adduced the loss of Iphigenia as a righteous cause: they show us some flashes of Electra’s preoccupations as well. Her point that Helen and Clytemnestra are siblings ‘unworthy of Castor’ (Κάστορος . . . οὐκ ἀξίω 1064) is a striking touch for someone who has made much of her former betrothal to that very man: is Electra— who, as we have seen (§5.2.2), has ‘reinvented’ herself as an unmarried παρθένος—perhaps implying that she herself is worthy of Castor? Finally, the description of her father as ‘the noblest man’ (ἄνδρ᾽ ἄριστον 1066, rephrased at 1081–2) is another indication of her attachment to Agamemnon (a point that Clytemnestra will pick up on at 1102–4). The next few sections of the speech (1068–85) are similar in important ways to Electra’s invective against Aegisthus’ corpse: they serve as characterizing vignettes of Clytemnestra’s behaviour in the palace (for the repeated use of imperfects to perform this role— another point of similarity to the kakology—see §5.2.2), and Electra’s main points are followed up by second-person generalizations* (1072–5, 1084–5). But there is an interesting difference in the way Electra frames her descriptions: whereas her statements about Aegisthus were sometimes explicitly based on second-hand or public knowledge (ἐν Ἀργείοισιν ἤκουες 930, ‘you had a reputation among the Argives’; cf. ὡς λέγουσιν 327, ‘as the story goes’), Electra here lays claim to authoritative first-hand knowledge of Clytemnestra’s behaviour (οὐ γὰρ σ᾽ ὡς ἔγωγ᾽ ἴσασιν εὖ 1068, ‘they do not know you as well 173 That this is the general sense here is clear, even if the text of 1058–9 is problematic, though not nearly as problematic as to justify what some have done to it (e.g. Distilo, absurdly, takes Broadhead’s ἔρξαις as ‘forma di participio aoristo dorica’). 174 For the meaning of ἀρχὴ . . . προοιμίου and the length of the prooimion, see ch. VI, §2.2.

148

The Linguistic Characterization of Electra

as I do’, οἶδ᾽ ἐγώ σ᾽ . . . 1076, ‘I know that you . . . ’). This makes the speech all the more personal, and the intimacy of Electra’s knowledge characterizes her once again as female, a role that for once she uses to her advantage: Making Clytemnestra’s preening in the mirror indicate her faithlessness is not just conventional: it takes us into the boudoir and seems to justify her claim to special knowledge, as does her account of Clytemnestra’s reaction to news from Troy. (Mossman 2001: 381–2)

Again, the preoccupation with sexual topics is typical of Electra, and typical of her gender. As for Electra’s generalizations, they again serve as a source of moral authority, underpinning her evaluation of Clytemnestra’s behaviour. Since this seems to be the pattern of Electra’s argumentation, we might by this point expect the rest of the speech to develop in similar fashion, through individual points followed by supporting gnōmai. What we get, however, is something altogether different. At 1086 the tone of the speech changes dramatically. Electra here leaves aside the whole question of whether Agamemnon’s murder was justified (the question that Clytemnestra wanted Electra to limit herself to, 1050), and focuses instead on what happened afterwards— the siblings’ expulsion from the palace. For the sake of argument, Electra even suggests that she is willing to concede the former issue (1086; she is careful, however, to distance herself from the argument by ascribing it explicitly to Clytemnestra: ὡς λέγεις, ‘as you claim’), but then makes her complaint about the latter with all the more vehemence. She argues her point with a striking sequence of rhetorical questions (1087–93), which are designed to point out the flaws in Clytemnestra’s reasoning.175 The questions follow each other in an uninterrupted flow, which gives the impression of a breathless litany: 175 This effect is an important reason for the use of the question form in rhetorical questions (whose illocutionary force* is not to acquire information, but to assert something): such questions force the addressee to consider possible answers but to come up short (they induce, in other words, a sense of aporia), which leads to the conversational implicature* that the answer is negative (‘Why did you do that?’ implicates ‘For no good reason’). By inviting addressees to reach an interpretation through a failure to answer pertinent questions, a speaker can point out the flaws in their reasoning. An additional effect, on audience members, is that they too are invited to ponder a range of answers to the questions: as Budelmann (2000: 71) puts it for a different agōn, the questions ‘are rhetorical . . . but they are still questions . . . and thus fully accessible to the spectators’ probing’.

The Linguistic Characterization of Electra

149

there is no connective particle between the first two (1087–8),176 and the others are strung together by κοὔτ’ and οὔτ᾽ (1091–2), which provide no internal ‘structure’ other than adding the complaints to the list. Electra’s torrent of questions may be interpreted as a sign of spontaneity and heightened emotion, as may the relative lack of generalizations that would have underpinned her arguments.177 Electra’s fervour here may easily be linked to the content of her rhetorical questions: again, as soon as the issue of her exile from the palace is raised, her linguistic behaviour ‘spikes’. She portrays her expulsion as ‘death’ (double death, even: δὶς τόσως ἐμὲ | κτείνας ἀδελφῆς ζῶσαν 1092–3, ‘having killed me twice as much as my sister, though I am alive’). It is noteworthy that these are the issues that Electra and Clytemnestra revisit in their dialogue after the speech (see §5.3.5) and about which Clytemnestra will express some regret. These issues, which have vexed Electra throughout the play, lead her once more to display some bloodlust in her resentful peroration. The lex talionis is phrased in a threatening conditional clause with the future indicative,178 and its ‘logical’ conclusion—Electra and Orestes will kill Clytemnestra—brought out with the equally (if not more) threatening future indicative in ἀποκτενῶ σ’ ἐγώ (1094, ‘I will kill you’),179 which is made all the more personal by the use of the first person singular pronoun and the postponement of Orestes’ participation to the next line. For an audience that knows that Clytemnestra is about to meet her doom, this is a grim closing to the speech: they will attach a more ‘factual’ (and thus more dire) interpretation to ἀποκτενῶ than Clytemnestra. Electra ends with a simplistic restatement of the ‘justice’ of tit-for-tat violence (1096): by this time, the audience is bound to have its doubts and, as Mossman (2001: 382) points out, the fact that the lex talionis is twice phrased in the form of

176 The asyndeton may also be explained as changing the second question into a restatement, in more specific terms, of the first. 177 See also Mossman (2001: 381). 178 See §3.2 (on 336–7), with n. 104. 179 For the use of the future indicative in threats, see Rijksbaron 2002: §9. Also see his Note 2 (following Bakker 2002) on the use of the future indicative in logical conclusions, particularly in the apodosis of conditional periods: ‘it expresses that, in the hypothetical world under discussion, the state of affairs in question with near certainty follows from the preceding line of reasoning’. This use is not limited to Plato, as Rijksbaron seems to suggest.

150

The Linguistic Characterization of Electra

a condition (1093–4, 1096) ‘implicitly raises the idea that [it] leaves much to be desired’.180

5.3.5. Mother and daughter (1102–46) After a brief choral comment (1100–1)181 follows a masterfully crafted passage of dialogue. In the face of Electra’s onslaught, Clytemnestra adopts an unexpectedly mollifying tone, which raises more troubling implications for the matricide to follow: Euripides makes Clytemnestra, if not sympathetic, at least vulnerable and ordinary in her anxiety to please both Electra and Aegisthus, while Electra continues to toy with her. She is very different from Aeschylus’s or Sophocles’ Clytemnestra (or even Pindar’s ‘ruthless woman’, Pyth. 11.22) and from hybristic victims such as Aegisthus, Lycus in Heracles, or Polymestor in Hecuba. Her regrets foreshadow her children’s later remorse. (Cropp on 1102–46)

Clytemnestra’s first words (1102–10) bring into relief the more conciliatory figure first hinted at in 1055–9. After the violence of Electra’s accusation, Clytemnestra pardons her adherence to her father with a second-person saying* about family relationships (1103–4),182 and promises restraint (συγγνώσομαί σοι 1105, ‘I’ll be patient with you’): both these moves are suggestive of the position of superiority that Clytemnestra deems herself to be in, but they are also clear attempts at a peace offering. Clytemnestra even goes on to express some regret for her actions and her excessive wrath against her husband (1105–6; Electra fails to learn the implicit lesson about

180

See my discussion in ch. VI, §5.2. Although I am on the whole sceptical of the excision of concluding general reflections (see ch. III, §2.2), I must in the end agree with Hartung’s deletion of 1097–9: the lines spoil the climax of Electra’s speech and seem to have little bearing on Clytemnestra (they are more applicable to Aegisthus). They were defended by Vahlen (1891: 362), Basta Donzelli (1978: 175), who later changed her mind in her Teubner edition (Basta Donzelli 1995a), de Romilly (1983: 416), Michelini (1987: 220–1), and Foley (2001: 237). That the lines are ascribed by Stobaeus (4.22) to Cretan Women is no good argument for deletion (see Luria 1929: 85–7), but they may perhaps be explained as a parallel for 1100–1 from elsewhere in Euripides intruding into the text. Lines 1100–1 themselves should be allowed to stand (see Denniston ad loc., Lloyd 1992: 68–9, and Willink on Eur. Or. 602–4). 182 I have discussed these lines in my Introduction, §2.2. 181

The Linguistic Characterization of Electra

151

anger management).183 Clytemnestra’s forms of address in this section (ὦ παῖ 1102, ‘child’; τέκνον 1106, ‘child’) betray her new stance, too. If lines 1107–8 are kept in place,184 Clytemnestra also indicates here her sympathy with Electra’s plight as a new mother in impoverished circumstances (just as Electra predicted she would: 658). But Electra will have none of it (1111–13): for her, Clytemnestra’s regrets come ‘too late’ (ὀψέ 1111), as they do nothing to remedy the situation she and Orestes are in. As earlier (at 1086–7), the indications are that this issue has come to take precedence over the question of justice for Agamemnon, and Electra again concedes the latter point for the sake of argument (πατὴρ μὲν οὖν τέθνηκε· τὸν δ᾽ ἔξω χθονὸς | πῶς οὐ κομίζῃ παῖδ’ ἀλητεύοντα σόν; 1112–13, ‘my father is dead, yes: but why do you not restore your son, who wanders in exile abroad?’).185 As for Clytemnestra, she lets her daughter take control of the conversation for the moment, and is forthcoming with answers to Electra’s complaining questions (preferred* responses in adjacency pairs* at 1113–14, 1116–17, 1118–19). She again shows some consideration for the plight of her children: she does not deny that they indeed suffer, and offers her own fear (δέδοικα 1114, ‘I am afraid’) and Aegisthus’ temper (τρόποι τοιοῦτοι 1117, ‘that’s his way’) as explanations: yet she also notes that no resolution for the situation is to hand, since Orestes’ anger (θυμοῦται 1115, ‘he is angry’) and Electra’s stubbornness (αὐθάδης ἔφυς 1117, ‘you are strong-willed’) prevent a reconciliation. 183 L’s πόσιν in 1110 must be emended for the reasons given by Denniston ad loc.; contra Steidle 1968: 66 n. 24. Gomperz’s πόσει works well. 184 A vexed issue. I am convinced that the lines do not belong after 1131, where they are usually transposed (following Weil). They cause more damage there than where they are found in the MSS (between 1106 and 1109), as they are ‘thoroughly unconnected with grammar or context’ after 1131 (Michelini 1987: 222 n. 179). Slings’ (1997c: 160) assertion that Clytemnestra’s turn could not begin with ἀλλ’ εἶμι (1131) is unsupported: cf. Aesch. Ag. 1313, Cho. 781, Soph. Trach. 86, 389, Ar. Eq. 488 (see also Distilo ad loc. and Mastronarde on Phoen. 753), and see my discussion of the force of ἀλλά in the Introduction, §2.2. I would opt in the end for keeping the lines in place, though I am not entirely comfortable with the flow of 1102–10 (defended as ‘naturalistic’ by Mastronarde 1979: 91–2), nor with that fact that 1124 is being uttered after 1107–8 (but see Mastronarde’s explanation). Other options (deletion, different transpositions) seem to be dead ends. 185 In the combination μὲν οὖν . . . δ᾽, the new point that falls under the scope of οὖν is introduced in the δέ clause (see Wakker 2009: 70–1). This suggests that Electra transitions from a less relevant point (in the μέν clause) to the point that is most crucial to her (in the δέ clause). Thus Electra again seems to elevate the issue of her and Orestes’ exile above all others.

152

The Linguistic Characterization of Electra

In a deeply ironical move, Electra even makes the suggestion of a rapprochement (παύσομαι θυμουμένη 1118, ‘I’ll stop being angry’)—a suggestion that Clytemnestra grasps with both hands, by assuring her daughter (μήν 1119)186 that Aegisthus will temper his anger as well. But the hint of goodwill swiftly disappears as Electra refuses to let the issue of Aegisthus’ behaviour rest, and Clytemnestra reacts in piqued fashion (1120–1: post-expansions* that reflect the women’s unwillingness to let each other’s statements go unchallenged). Thus any faint hope that Clytemnestra’s regrets might avert the matricide come to nothing,187 as the impasse reached between mother and daughter spells an irreversible march towards a deadly resolution. The inevitable is underlined by a series of grisly ironies in Electra’s language: οὐκ ἔχεις ἄκη 1111, ‘you’ve no remedy’ (indeed she does not); τὸν δ᾿ ἔξω χθονός . . . παῖδ’ ἀλητεύοντα σόν 1112, ‘your son, who wanders in exile abroad’ (or does he?); παύσομαι θυμουμένη 1118, ‘I’ll stop being angry’ (when both Clytemnestra and Aegisthus are dead); δέδοικα γὰρ νιν ὡς δέδοικ᾽ ἐγώ 1122, ‘I fear him as I fear him’ (not at all).188 All this leads to the supremely ironical send-off in 1139–46: the whole scene is a powerful example of the motif of mocking and luring a doomed victim,189 merged with the exchange of arguments that normally follows Euripidean agōnes (see Cropp on 1102–46). Clytemnestra confirms that the two women have reached an impasse when she brusquely caps the discussion at 1123 (παῦσαι λόγων τῶνδ᾽, ‘enough of this talk’). The impatience seen earlier in the scene (see §5.3.1) takes the upper hand again: she has indulged Electra long enough (note the bald* imperative παῦσαι), and she takes control of the conversation by shifting the topic (ἀλλά) to the reason why she was called (a forceful display of topic management*). The motif of sacrifice, prominent in the messenger speech (see ch. V), is now reintroduced, and adds another troubling touch to the

186

For the particle, see Wakker 1997a and van Emde Boas in press-a. Not an actual hope, of course, since the audience knows how the myth must end; but even the suggestion is an effective touch of tragic pathos. 188 This last case is a neat manipulation of the fact that tautologies inherently require interpretation: the audience’s interpretation will be quite different from Clytemnestra’s (for an explanation of tautologies in terms of Gricean pragmatics*, see n. 57). For this kind of malign irony in scenes of entrapment, see Rutherford 2012: 329–34. 189 See, e.g. Eur. Her. 701–33, with Bond ad loc. 187

The Linguistic Characterization of Electra

153

impending matricide. That the killing is imminent becomes palpable as the argument between the two women reaches its end, but Euripides allows suspense to linger by dragging out Clytemnestra’s exit into the house with her insert expansions* at 1128–9 and 1130–1 (see Introduction, §2.2). But then she is off, her last words on stage a significant and ironic indication of the priority she accords to pleasing her husband (see Mossman 2001: 382). Before Electra follows Clytemnestra in, her last words to her mother (1139–40) provide a perfect merging of dramatic irony and characterization. As Denniston (ad loc.) puts it: Once again, Electra’s hatred of the fashionable lady (cf. 966, 1070–1) breaks out. She never once forgets the squalor of her home and the penury to which she has been reduced. The politeness of μοι (‘please’) sharpens the sting.

In addition, Cropp notes (on 1139–41, comparing Aesch. Ag. 774–5) that the poverty of Electra’s house promises the victory of justice over Clytemnestra’s sinful luxury. More ironies abound in Electra’s final five lines (1142–6), which must be performed as an aside and cannot be heard by Clytemnestra.190 The sacrificial language continues the motif given prominence in the messenger speech, while the description of Aegisthus as a bull (1143) recalls similar imaginings of Agamemnon (Hom. Od. 11.411, Aesch. Ag. 1126), again linking the two murders. Her parting words (1145–6) reprise Clytemnestra’s: she will do her a ‘favour’ by allowing her to lie next to Aegisthus in Hades as in life; Clytemnestra, for her part, will pay the penalty for Agamemnon’s murder—and Electra exits into the house to personally ensure that she does.

6. EXODOS I have been occupied throughout this chapter with finding ‘constants’ in Electra’s use of language: her notable preoccupation with certain issues (her personal degradation, Orestes’ exile, the death of Agamemnon), which leads to ‘flares’ in her linguistic behaviour; her violent resentment of Clytemnestra and Aegisthus; the fact that her 190

See Bain 1977a: 34–5 and Mastronarde 1979: 30.

154

The Linguistic Characterization of Electra

language is conditioned repeatedly by her gender and by the gendered make-up of the company she is in. I hope to have shown that these aspects contribute to a consistent ‘linguistic characterization’ of the figure throughout the play. However, in the aftermath of the matricide—the dramatic climax of the play, which lies ‘at the exact pivot point between the enraged Electra and the lamenting Electra’191— we may reasonably expect Electra’s language to be irrevocably transformed,192 and indeed much is different when Electra and Orestes come out of the house again at 1172. To conclude this chapter, I will look briefly at Electra’s language in the exodos in order to show how, through its differences and similarities with what we have seen up to now, it reflects her new role.

6.1. The kommos (1177–232) In order to analyse Electra’s singing part in the lyrical amoibaion that follows the siblings’ appearance on stage with their mother’s corpse, it is crucial to determine first what that part is. The attribution of lines to singers in our manuscript L is somewhat jumbled,193 but the demands of strophic responsion, if we accept that it entails a

191 Tra l’Elettra θυμουμένη e l’Elettra che piange c’è di mezzo appunto il matricidio (Basta Donzelli 1978: 195). 192 That such an expectation is reasonable is not an opinion which has been shared by all. Zürcher (1947: 136–42), for instance, argued that Euripides abandons any notion of ēthos for Electra after the matricide, at which point the telos of the tragedy supersedes any interest in Electra’s character. But matricide is no small thing, and that Electra is very different after it should not be seen as evidence for an absence of interest in plausible characterization. 193 L has 1177–81 Or., 1182–9 El., 1190–7 Or., 1198–200 El., 1201–9 Or., 1210–12 El., 1213–17 Or., 1218–20 Cho., 1221–3 Or., 1224–5 El., 1226–9 Cho., 1230–2 El. Triclinius corrected the manuscript by assigning 1206–9 to Orestes, which implies that he believed 1201–5 to belong to someone else, evidently the Chorus. Any remaining confusion resides above all in the identity of singers (I will argue that 1201–5 and 1210–12 should be assigned to the Chorus, 1227–9 to Orestes), less so in the moments where a new singer is required. Changes at 1185 (Chorus) and 1227 (instead of 1226, Orestes) are needed. On the reliability of indications of speaker changes as opposed to speaker names, see ch. IV, §1.1. Lack of responsion shows that L also has a lacuna between 1181 and 1182. Thus, on the whole, L’s evidence is not tremendously far off the mark (for an unconvincing defence of the entire transmitted distribution, see Stoessl 1956: 75–81); its flaws can be attributed to the fact that MSS in general have more difficulty with strophic songs than with other parts of plays.

The Linguistic Characterization of Electra

155

symmetry in the division of parts,194 allow for a fairly straightforward reconstruction. What I believe to be the correct division was first proposed by Kirchhoff; arguments for this distribution were comprehensively marshalled by Steidle (1968: 90 nn. 151–4a; see also Cropp ad locc.). I repeat some crucial ones here: • A change of singer at 1185 is required. It is made probable by ἰώ, and the singer of αἰτία δ᾽ ἐγώ (1182, ‘I am to blame’) can hardly be the same as that of πατρὸς δ᾽ ἔτεισας φόνον δικαίως (1189, ‘you have atoned for their (?) father’s murder justly’). Most importantly, this creates symmetry with the antistrophe (1201–5, see next point). This symmetry and σῶν τέκνων ὑπαί (1188, ‘at the hands of your children’) suggest that lines 1185–9 are sung by the Chorus, which implies in turn that what has fallen out after μᾶτερ τεκοῦσ᾽ (1186) must be an object of τεκοῦσ(α). This would create a neat chiastic structure with ἄλαστα . . . παθοῦσα. Grotefeld’s suggestion that ἄλαστα should be duplicated, yielding ‘who bore accursed children and suffered an accursed fate at their hands’, is very attractive. As a neat parallel, Diggle (app. crit.) offers Stesichorus’ fragment 17, lines 2–3 (Davies and Finglass 2015: ἀλασ[τοτόκος κ]αὶ ἄλ̣[ασ]τ̣α̣ π̣α̣θοῖσα ‘bearing grievous offspring and suffering grievous ills’). • Lines 1201–5 must belong to the Chorus: Orestes could hardly say κασίγνητον οὐ θέλοντα (‘your unwilling brother’) about himself, and φίλα is, as an address to Electra, only suited to the women of the Chorus. • Lines 1206–9 were ascribed by Triclinius to Orestes (see n. 193), and this cannot be doubted: the singer must be one of the siblings (only they saw their mother’s supplication), he must be the one supplicated by Clytemnestra, and τακόμαν δ᾽ ἐγώ (1209, ‘my strength melted away’) settles the issue: Orestes is the one who got weak in the knees.195 Symmetry then suggests that lines 1213–17 are also sung by Orestes, and this is confirmed by γένυν

194

I believe it does; see West 1982: 79, Page 1937, and McDevitt 1981. Symmetry is not a secure argument outside of strophic song, however (see ch. IV, n. 36). 195 Admittedly τακόμαν is an emendation (Seidler; L has τὰν κόμαν), but this is universally accepted.

156

The Linguistic Characterization of Electra

ἐμάν (‘my chin’), since only men are supplicated by touching the chin.196 • Lines 1210–12 and 1218–20 belong to the Chorus (L gives 1210–12 to Electra): symmetry demands the same speaker, and neither ματρὸς ἅ σ᾽ ἔτικτεν (1212, ‘the mother who bore you’) nor the question in 1218–20 should be ascribed to Electra. • Orestes should be the one to initiate the covering of Clytemnestra’s corpse in 1227–9 (the members of the Chorus can perform no such action). • L ascribes to the Chorus 1226, but not 1232: symmetry requires both or neither (i.e. giving them to Electra; so Diggle (following Seidler); see arguments for the Chorus in Steidle 1968: 90 nn. 151–4a and Slings 1997c: 160–1).197 Decision is difficult, but I would in the end opt for following Diggle (and Seidler) because of the syntax of 1230–2: there τέρμα (1232) stands in apposition to what is said in the preceding two lines (1230–1), which seems more probable if the speaker of all three lines are the same. This option seems to require the additional change of ἔρεξας to ἔρεξα in 1226.198 Once this emendation is adopted, the otiose speaker change in L at 1226 is perhaps less surprising if we note that the manuscript is missing a speaker in 1227; perhaps the indication of speaker change was misplaced and 1226 subsequently misattributed, which may also have led to the change to ἔρεξας. What is the upshot of adopting this division of parts? If it is right, Electra’s contribution to the amoibaion is remarkably limited (only 1182–4, 1198–200, 1224–6 and 1230–2 are hers). Orestes, on the 196 This argument, in the end, swayed Denniston (addenda), though he had argued for Electra in the original text of his commentary (ad loc.). For the gender preference of the gesture of supplicating someone while touching their γένυς, see Foti 2005: the discussion there of πρὸς γενείου at Med. 65 has bearing on the present issue. 197 Cropp ad loc. assigns the line to the Chorus, apparently having changed his position (cf. Cropp 1988 on 1177–232; also Cropp 1996). 198 Cropp (1988 ad loc.) argued that ἔρεξας is possible even if Electra sings the line, but this strikes me as unlikely. Slings’ (1997c: 160–1) objection to the attribution of 1226 to Electra on the grounds that a connective particle is in that case required but cannot be inserted is undermined by his own point that the period ends at 1225 (as is confirmed by the fact that the second syllable of ἅμα in 1225 is brevis in longo): this makes asyndeton more palatable (cf. 1182–3); and, despite what Slings seems to suggest, period end does not require a change of speaker (cf. 1202–3). Reading ἔρεξα, I would print a colon after 1225 rather than a full stop.

The Linguistic Characterization of Electra

157

other hand, is given a much greater role, initiating each strophe and antistrophe and singing far more lines than his sister. As Mossman has pointed out, this seems to indicate something of a reversal in Electra (and in her brother): Given that lamentation is often seen as the province of women, this is striking. It seems to represent a readjustment of their relationship: Orestes is no longer being dictated to by Electra’s emotions, but is expressing his own. . . . The limiting of her song here is perhaps the most effective way of showing a chastened and subdued Electra—and, of course, an Electra once more on stage in the presence of her male relatives. (Mossman 2001: 383)

Indeed, the siblings’ language is still patterned along lines of gender even in this moment of supreme emotion (which might have elided such differences): Orestes laments that he will be unwelcome in cities and with ξένοι (1195–7; the male, political sphere), while Electra bewails her probable exclusion from choruses and marriage (1198–200; the female sphere).199 But in other respects Electra’s language is very much changed: not only does she take full responsibility for inciting the matricide, she also recognizes its causes and implications. Her first words after coming out of the house are emblematic (1182–3): δακρύτ᾽ ἄγαν, ὦ σύγγον’, αἰτία δ᾽ ἐγώ. διὰ πυρὸς ἔμολον ἁ τάλαινα ματρὶ τᾷδ᾽, ἅ μ’ ἔτικτε κούραν. All too lamentable, brother, and I am to blame. I, unhappy woman, burned with hatred against our mother here, who bore me.

Electra’s fire metaphor (on which see LSJ s.v. πῦρ, II) conveys her awareness of the unnatural depth of her earlier hatred. The pleonastic collocation of μήτηρ (‘mother’) and a form of τίκτω (‘give birth’), which earlier in the play had expressed sarcasm (264 and 964, with my discussion at ch. IV, §3.1), now deepens the regret (it is repeated in the amoibaion at 1165, 1186, and 1212).200 The women of the Chorus confirm Electra’s reading of events and agree that her earlier mental state was depraved (φρονεῖς γὰρ ὅσια νῦν, 199

See Mossman 2001: 376, 382 and Ormand 2009: 251–2. As Cropp rightly notes (on 1182–4), references to mother, childbearing, and children recur repeatedly—and tellingly—in the song (1165, 1168, 1172, 1188, 1197, 1215, 1220, 1223, 1227, 1229). 200

158

The Linguistic Characterization of Electra

τότ᾽ οὐ | φρονοῦσα 1203–4, ‘your thinking is pure now, but then it was unholy’). They also condemn her for egging her brother on (δεινὰ δ᾽ εἰργάσω . . . κασίγνητον οὐ θέλοντα 1204–5, ‘you wrought dreadful things on your unwilling brother’)—all too true, as the audience knows. It is worth noting, however, that the women have not lost their sympathy for Electra, as is clear from their address φίλα (1205). The tragic pathos of the kommos rises to its highest levels when Orestes takes us inside the house, to experience the matricide as it unfolds (1206–26). The key first word of this passage, κατεῖδες (1204, ‘did you see?’), invites the audience to visualize the murder in all its detail; Clytemnestra is vividly brought to life gesturing, supplicating her son, her pleas quoted in direct speech (τέκος ἐμόν, λιταίνω 1215, ‘I beseech you, my child’).201 Assuming that the division of parts outlined here is correct, Electra’s only contribution to this ‘re-enactment’ comes at its end and delivers an unexpectedly horrific climax: Electra drove Orestes to the deed when his strength failed him, and even had her hand on the sword (ἐγὼ δὲ γ᾽ ἐπεκέλευσά σοι | ξίφους τ᾽ ἐφηψάμαν ἅμα 1224–5, ‘and I urged you on, and grasped the sword together with you’).202 If right, this is an ingenious manipulation of the structure of the song: by extending the ‘narration’ of the murder over one strophic pair—not featuring Electra—but only over half of the next, Euripides limits her role in it to the shocking admission that she physically took part in the matricide. Electra then concludes the description of the murder by again assuming blame (δεινότατον παθέων ἔρεξα 1226, ‘I brought about the most terrible of sufferings’);203 the ‘startling oxymoron’ (Cropp 1988 ad loc.) in the phrasing—πάθος and ῥέζω are normally opposing concepts—is indicative of the singer’s newfound understanding: doing has led to suffering. The remainder of the exodos will repeatedly stress the shared responsibility of the two siblings (see esp. 1305–7), and this is emphasized at the end of the amoibaion, when they jointly cover their mother’s corpse (1227–31).204 Meanwhile Electra’s newfound insight is evident from her words φίλᾳ τε κοὐ φίλᾳ (1230): it is impossible to imagine Electra using this word about her mother earlier in the play, 201

For the effects of direct speech, see §3.2 and ch. V, §3.2. The use of contrastive ἐγώ (with γε, if the emendation is correct) is a further reason to accept the division of parts outlined here. This seems more natural on the assumption that Electra has been quiet for a while (particularly since 1206). 203 Again, supposing that the attribution of this line to Electra is correct. 204 So also Mossman 2001: 383. 202

The Linguistic Characterization of Electra

159

but she now realizes the horrible paradox of having as a (political) enemy someone who is connected by blood.205 Only now does the φιλία of blood relation take precedence over Electra’s prior enmity, as she dutifully performs her role as daughter in Clytemnestra’s burial (the ‘act of covering the body is one that should be performed for a parent by a child’, Michelini 1987: 223 n. 184). As Denniston (ad loc.) points out, even though expressions of the type ‘X but not X’ are something of a Euripidean mannerism, ‘here its pathos is most moving’.

6.2. The deus ex machina (1233–358) Apart from showing the siblings at their most emotional and a changed, chastened Electra, the kommos has implicitly raised the question of what will happen to them next: both Orestes and Electra despair about what the future holds for them and seem aware that they face exile (1194–200).206 Their uncertainty does not last long, however: the Dioscuri soon appear over the Peasant’s hut (1233), and Castor assigns the siblings to their rightful place in (Euripides’ version of) the extended myth (1238–91). This theophany and the ending of the play have confounded scholars greatly: they struggle with the ‘neatness’ of the resolution provided by the Dioscuri, the tone (does the play end on an optimistic or a bleak note?), and the relationship between gods (especially Apollo) and humans.207 Such problems of interpretation have prompted some to condemn the ending (if not the whole) of the play in disgust: Vickers (1973: 565), for example, believes everything For the evolving role of φιλία in Electra, see Konstan 1985 (181–2 on the present passage). 206 The fact that these concerns are raised in the form of (rhetorical) questions may again (see n. 175) be seen as a way of inviting audience members to ponder the question of what the future does hold for the siblings. (Their knowledge of the myth will of course provide part of the answer, even if some aspects of Euripides’ treatment, for instance the marriage of Electra and Pylades, may have been innovations.) 207 For interpretations of the ending, see Grube 1941: 313–14, Zürcher 1947: 138–40, Stoessl 1956: 74–92, Spira 1960: 101–12, Conacher 1967: 210, Vögler 1967: 69–70, Steidle 1968: 87–9, Vickers 1973: 565–6, Basta Donzelli 1978: ch. 6, Whitehorne 1978, Michelini 1987: 223–7, Ormand 2009: 265–8, Mastronarde 2010: 181–95 (valuable on deus ex machina scenes generally, with comments on Electra), and Cropp 2013: 11–12—but the list is not exhaustive. These scholars continue to disagree about the extent to which Electra and Orestes deserve our sympathy. 205

160

The Linguistic Characterization of Electra

after the kommos to be ‘an unsatisfying deflection of the action’, the effect of which is ‘to wrench the play away from tragedy into polemic against Apollo’. Yet the difficulty of drawing simple conclusions about such larger questions at the play’s end may be reflective of Euripidean design: Basta Donzelli (1978: 184–5), after usefully surveying some earlier, wildly divergent opinions about the ending, suggests that ‘the text itself seems to lend itself to such contradictory conclusions’.208 Indeed, the sense that no easy answers are forthcoming—that Euripides is in fact posing questions—is inherent in much of his work: ‘the plays provoke thought about large issues beyond the confines of the drama. They offer stimulation and interest rather than theories and ultimate views’ (Allan 2000: 270). The ending of Electra stimulates such contemplation, not least with regard to the relation between gods and humans, and divine motivation. In no play is Apollo’s role in supporting the matricide questioned more overtly and repeatedly (979–81, 1245–6,209 1296, 1302), yet Apollo’s reasons remain opaque. Even other gods, the Dioscuri, criticize him, but they have no real answers to provide either, and seem themselves strongly disconnected from human experience. It is in this last respect that Electra’s final contributions to the play reveal much, and looking at the structure of her conversation with her brother and the Dioscuri may aid our interpretation. Again, analysis of the dialogue must be preceded by an attempt to settle the division and attribution of speaking parts in it. Confusion in the manuscript tradition is more profound than usual in this case, as L and its apograph P themselves disagree in various places and a corrector has altered the speaker indications in P.210 Sadly, most 208

Il testo stesso sembrerebbe prestarsi a conclusioni cosi contraddittorie. Note in these lines the rare aposiopesis (Φοῖβος δέ, Φοῖβος—ἀλλ’ ἄναξ γὰρ ἐστ᾽ ἐμός, | σιγῶ, ‘and Phoebus, Phoebus—but he’s my master, so I’ll say nothing’), which underlines both the criticism of Apollo and the limits on the Dioscuri’s ability to criticize him forcefully. 210 For the relation between these MSS, see ch. IV, n. 5. The distribution in them is as follows: L has 1292–3 Or., 1294 Ca., 1295 El., 1296–7 no indication, 1298–300 paragraphos, 1301–2 Ca., 1303–4 no indication; P has 1292–3 Or., 1294 Ca., 1295 Or. (the corrector adds Ἠλ ἢ before Ὀρ, giving ‘El. or Or.’), 1296–7 Ca., 1298–300 no indication (corrector: El.), 1301–2 paragraphos., 1303–4 paragraphos. Taken together, the MSS indicate speaker changes in all the required places; the indications of names in the MSS are without much authority (see ch. IV, §1.1), but I believe not very far off here again. 209

The Linguistic Characterization of Electra

161

modern editions, from Victorius’ editio princeps to Diggle’s standard text, have done more to add to the confusion in this passage than to remedy it. Recently, however, a solid case has been made, by a number of scholars cumulatively,211 for what is surely the right distribution: • Victorius assigned 1292–3 and 1298–300 to the Chorus (the manuscripts give the first pair to Orestes and are inconclusive on the second), and he is followed in this by most editors. But choruses never address the deus ex machina in (extant) Greek tragedy; it would make their role inexplicably prominent here if they did, and Orestes must be the first to speak to the Dioscuri after their entire speech was addressed to him. By using secondperson verb forms and pronouns throughout his speech, Castor selects* Orestes as the next speaker (see Kovacs 1985: 313 n. 20 for convincing parallels). Moreover, if the Chorus is given 1292–3 and one of the siblings 1295, this creates a parallelism between them that is highly unsuitable. • If we accept that lines 1292–3 belong to Orestes, the attribution of 1295 to Electra follows logically: if one of the siblings asks for permission to speak first, the second should do so next.212 It is also plausible that Electra asks permission to converse with the deus before asking her question at 1303–4 (which can only be hers: feminine φονίᾱν is required by the metre): scholars have leapt at the chance to explain an ‘unannounced’ question by Electra as indicative of her intemperate character,213 but this seems hard to reconcile with her regret after the matricide.

211 Stoessl (1956: 82–5), Steidle (1968: 85–6), Basta Donzelli (1978: 210–13), Kovacs (1985: 310–14), Basta Donzelli (1991b: 29–34), and Cropp on 1292–359. 212 Slings (1997c: 161–3) and Distilo ad loc. want to assign both 1292–3 and 1295 to Orestes, but their interpretation—Orestes asking for confirmation that Castor’s οὐ μυσαροῖς indeed applies to him as well—is very hard to swallow, and is rendered moot if we note that the speaker of 1295 uses a form of address to the Dioscuri (Τυνδαρίδαι). This virtually guarantees that he/she was not already talking to them: the two requests for permission to enter into contact cannot be spoken by the same person. 213 So, e.g. Winnington-Ingram (1937: 52, citing Murray): ‘Electra (ut solet, audacior) interposes’; Denniston on 1295: ‘Electra’s “boldness” comes out most strikingly in the fact that she, unlike the Chorus and Orestes, speaks to the gods without obtaining their permission first’.

162

The Linguistic Characterization of Electra

• In several proposed configurations, one of the petitioners in 1292–3 and 1295 does not proceed to ask one of the questions in 1298–300 and 1303–4. This is deeply unsatisfying: Kovacs (1985: 312) rightly wonders whether it is ‘idle to enquire what Euripides had in mind when he made [someone] ask permission to speak’ only not to use that permission. • If Orestes is given 1292–3, he should be the one to ask the question in 1298–300. This creates a neat structure, in which each of the siblings asks permission to speak, is granted it, and proceeds to ask a question, which is answered. • Finally, Arnoldt (1878: 331 n. 1) and Winnington-Ingram (1937) proposed transposing 1295–7 after 1302.214 But (regardless of who says what) the request in 1295 appears elicited directly by, and complementary to, the question and answer in 1292–4 and should not be separated from them (note κἀμοί in 1295, ‘for me as well’). Moreover, the δ᾽ in 1303 seems wrong immediately after 1296–7 but is completely expected after 1301–2 (Electra shifts the topic with her question). There is thus little reason to change the transmitted order and distributions (particularly that of ms. L) more than needed: the correct division, I believe, is: 1292–3 Orestes, 1294 Castor, 1295 Electra, 1296–7 Castor, 1298–300 Orestes, 1301–2 Castor, 1303–4 Electra, 1305–7 Castor. Both siblings preface their questioning of the Dioscuri with pre-expansions*, ensuring that the subsequent dialogue will take place on a proper footing.215 The elaborateness of this conversational structure, unusual in tragic dialogue, is indicative of the care the siblings are taking in this interaction: given their recent actions, they find themselves in a precariously sensitive situation. Another interesting aspect of the interaction between Electra and the Dioscuri is revealed if this distribution is right. Castor’s replies to Electra (1296–7 and 1305–7) consistently deal with matters on the divine plane: Electra may converse with them, since Apollo is to blame; ‘shared fates’ (κοινοὶ . . . πότμοι 1305) and a paternal ‘doom’ (ἄτη 1307) have driven her to share in Orestes’ matricide. Conversely, Electra, who in the kommos viewed her part in the matricide as a result of her hatred (1183), makes explicit that she does not 214 215

Winnington-Ingram appears not to have known Arnoldt’s earlier proposal. See my discussion in the Introduction, §2.2.

The Linguistic Characterization of Electra

163

understand the role higher powers played in the unfolding of events (τίς . . . Ἀπόλλων, ποῖοι χρησμοί 1303, ‘what Apollo? what kind of oracles?’): after all she had not, like her brother, received an oracle from Apollo. The question at 1295 (if it is Electra’s) might also be seen as indicative of Electra’s difficulty with understanding the relation between gods and humans. We may wonder why she asks for verification of her right to address the gods after it has already been conferred on Orestes. She must have, that is to say, a sense that her situation is not exactly the same as that of her brother. The difference might be perceived to be one of gender (Electra wonders whether the Dioscuri’s permission extends also to women), but it is also possible to read Electra’s question as her wondering whether Apollo’s commands absolve Orestes, but not her. Thus the exchanges between the Dioscuri and Electra are indicative to some extent of a difference in their thinking: Electra’s instinct is to interpret events as being caused by human factors, and she does not immediately comprehend causes that do not operate on the human plane; but Castor can reply only in terms of divine factors. This discrepancy—which is inherent in Electra’s and Castor’s words even if my preferred distribution is not accepted (though it is reinforced by it)—is all the more painfully made clear shortly afterwards, in the exchange at 1308–15. Orestes laments that, after their recent reunion, he and Electra will no longer see each other. By addressing his sister directly (ὦ σύγγονέ μοι 1308) and by using second-person pronouns (σ᾽ 1308; σῶν 1309; σ᾽, σοῦ 1310), Orestes seems to select* her to take the next speaking turn*, but Castor self-selects* and interrupts with a comment about Electra’s fate (1311–13): it is not so pitiable, he argues, because she will have a husband and a house, and will merely have to leave her city. Castor’s interference in the conversational structure reinforces the blunt lack of comprehension of the human experience, which is evident in his words. This lack of comprehension is immediately pointed out by Electra, who attaches a disagreeing post-expansion* (1314–15): καὶ τίνες ἄλλαι στοναχαὶ μείζους ἢ γῆς πατρίας ὅρον ἐκλείπειν; And what other pains are greater than leaving the borders of one’s native country?

For the remainder of the exodos, Orestes and Electra are concerned entirely with the all too human emotions that their imminent

164

The Linguistic Characterization of Electra

separation evokes. Their shared expressions of grief bring even the aloof god to show some emotion: Castor groans (φεῦ φεῦ 1327) and remarks that even gods can be impressed by such human suffering (δεινὸν τόδ᾽ ἐγηρύσω | καὶ θεοῖσι κλύειν 1327–8, ‘this cry of yours is dreadful even for gods to hear’; that Castor feels the need to explain his groans is suggestive of the gulf between the divine sphere and the human). The exchanges between Electra and the Dioscuri thus prove to be something of a microcosm of a much wider issue—an issue of some significance in Euripides’ work: the inability of humans and gods to understand each other’s motivations and emotions. Electra’s language, and even the very structure of her conversation with her brother and the deus ex machina, seem designed to point up this problem. It is perhaps no coincidence that the character whose very human emotions have driven her use of language throughout the play is here made almost to symbolize humanity’s difficult struggle with the divine.

III Orestes’ Linguistic (Dis)guises A man of fashion never has recourse to proverbs and vulgar aphorisms. —Philip Dormer Stanhope, Earl of Chesterfield, in a letter to his son (27 September 1749)

1. INTRODUCTION This chapter, like the previous two, will focus on how a particular character in the play, in this case Orestes, is characterized by his language. Unlike in the preceding chapters, however, here I will focus my discussion on a particular aspect of his linguistic behaviour, namely the way Orestes uses language to disguise his identity in various parts of the play. Particular attention will be given to Orestes’ use of generalizing speech: it has not escaped scholars’ notice that he is, in the scenes leading up to the recognition, a sententious speaker, prone to (what some have thought excessive)1 moralizing—a feature of his language that fades as soon as his disguised persona makes way for the revealed brother–avenger himself. This proves in fact to be only one feature of a noticeable discontinuity between Orestes’ speech in the first and second half of Electra. It is this discontinuity, and the reasons for it, that I hope to examine in this chapter. To be precise, Orestes maintains two separate disguises in the course of the play: first when pretending (to Electra, the Peasant, 1 It is perhaps unsurprising that one of the most vexed passages of the play, subject to a great deal of critical suspicion, is a prolonged stretch of moralizing by Orestes (367–400). I will discuss the (perceived) problems of this passage in §2.2.3.

166

Orestes’ Linguistic (Dis)guises

and—unsuccessfully—the Old Man) to be his own emissary (215–576), and again in the messenger speech, when confronting Aegisthus (779–843). In both these scenes Orestes’ disguise is—at least insofar far as it has been preserved for modern readers—above all a linguistic one: it is by virtue of what he says that Orestes manages to keep his true identity hidden.2 The dramatic uses of this linguistic camouflage are obvious: Euripides exploits the knowledge gap between his characters and the audience so as to imbue every word that Orestes speaks (or fails to speak) with dramatic irony. What I want to argue here is that this irony runs deeper than mere superficialities, just as the nature of Orestes’ disguise is not limited to obfuscation and lying: in each of his different personas Orestes is a different kind of speaker, and with these changes Euripides manages to produce a whole range of effects. In this chapter I will take a detailed look particularly at the first of these disguises, which Orestes sustains from the moment he approaches Electra (215) until she is finally brought to see through it by the Old Man (577). I conclude with a few observations on Orestes’ language in his second disguise (the messenger speech) and in other parts of the play.

2. ORESTES INCOGNITO

2.1. Initial observations Before delving into Orestes’ use of sententious speech, it is useful to start with a few other, general observations on the (usually deeply ironical) ways in which Orestes’ language is designed to keep his identity hidden. We may sum up the following instruments in his deceptive arsenal: 2 I am aware that I am greatly oversimplifying things when talking about ‘Orestes’ speech’ in the messenger scene: his words there are reported by an internal narrator (the Messenger), who will have his own purposes in selecting (and phrasing) Orestes’ words, and can moreover provide narratorial comment on what the characters in his narrative say. At the same time, we can be sure that the disguise (and the suspense arising from it) is one of the reasons why direct speech is so extraordinarily dense in the messenger’s narrative, as I will argue in ch. V. The only character for whom Orestes’ linguistic disguise does not work (because physical signs prove a stronger indication of identity) is the Old Man, who immediately sees through it.

Orestes’ Linguistic (Dis)guises

167

• The technique Orestes uses most frequently to deflect attention and maintain his cover is to talk about himself in the third person. In fact he uses his own name or words such as κασίγνητος (‘brother’) and ἀδελφός (‘brother’) frequently, and mentions ‘himself ’ to his sister much more regularly than she does to him.3 • Orestes in disguise is characterized by suppression of emotion. We sometimes ‘see the real Orestes’ emotions in conflict with the courtesy and reserve of his assumed persona’ (Cropp 2013: 7; more on this point in §2.2). Yet in spite of these moments the full extent of Orestes’ emotional range becomes apparent only when his true identity is revealed. The undisguised Orestes is given two passionate embraces with his sister (577–95 and 1321–33) and a moving lyric exchange with her and the Chorus (1177–232, see ch. II, §6.1), revealing a depth of feeling that the disguised Orestes never allows himself to display. • It appears to me that the reserve of the disguised Orestes may also be seen in his use of moods and modal expressions. The evidence is too slim for us to draw any secure conclusions, but, in disguise, Orestes proves statistically somewhat more likely to use potential optatives (with ἄν), counterfactual indicatives (with ἄν), and the optative and secondary indicative in (realizable and unrealizable) wishes.4 This is especially the case when Electra is the addressee (most of Orestes’ later uses of such expressions come in planning the murder of Aegisthus with the Old Man). Much of this has to do, of course, with the particular situational context of each instance in which one of these expressions is used: yet we may perhaps make the generalization that the disguised Orestes is more likely, in this way as well, to use language that implies a ‘non-committal’ style or a certain ‘distance’.

3 Ὀρέστης (4 occurrences): 260, 274, 282, 365; κασίγνητος (5 occurrences): 228, 244, 250, 292, 397; ἀδελφός (2 occurrences): 242, 248. Electra uses the name Ὀρέστης four times with Orestes present: at 303, ironically juxtaposed with ξένε (‘stranger’) in 302; at 330, using Aegisthus‘ voice (see ch. II, §3.2); at 347, again close to ξένοι; and finally at 560, just before the recognition. 4 In disguise (93 lines): 222, 224, 276, 278, 282, 378, 395, 397, 399; not in disguise (152 lines): 599, 612, 620, 632, 634, 669, 903, 981. Note also his use of the modal adverbs τάχα (226, ‘perhaps’), and ἴσως (399, ‘perhaps’).

168

Orestes’ Linguistic (Dis)guises

For the purposes of this chapter, however, another aspect of Orestes’ language in disguise interests me most. That aspect is his use of general reflections, to which I now turn.

2.2. The general reflections It is only when Orestes is disguised as a messenger sent by himself that he displays the moralizing tendencies for which he has been maligned by a number of scholars.5 By a crude count of lines, roughly one third of Orestes’ lines in this part could be described as sententious.6 Perhaps even more strikingly, none of his lines outside these scenes fits that description particularly well.7 This discrepancy is not coincidental, of course; but the reasons for it are complex. I would like to expand here on a suggestion made by Cropp, that Orestes’ moralizing in these scenes does not merely coincide with his disguise but is in fact an integral part of it: The moralizing speeches about pity, sensibility, and the criteria of nobility sound snobbish and superficial, but they come from the disguised persona, and the first, at least, seems designed to divert attention from Orestes’ identity. (Cropp 2013: 7)

We may take this claim one step further and argue that the other gnomic utterances delivered by Orestes also serve as a diversionary tactic. I have discussed some of the forces at work here in my Introduction (§2.5) and in the previous chapter (§5.2.1): first, the very act of uttering speech with a generic content entails that the focus is not on the particular, on the individual situation at hand. For Orestes, in other 5 See e.g. Denniston on 294–6 (‘Orestes prides himself, rather priggishly, on his sensitiveness and fine feeling’); Goldhill 1986b, esp. 163–71. 6 Orestes speaks in 238 lines in total (including those reported by the Messenger, but see n. 2), of which 93 are uttered by his first disguised persona. Of these 93, some 29 (35%) could be considered sententious, on the basis of such linguistic markers as third-person generic presents (e.g. δάκνει 291; ἔχουσι 368), nouns without an individual/specific referent (e.g. βροτούς 291; κριτῇ 374) or abstract nouns (e.g. αἴσθησις 291; εὐανδρίαν 367), and certain quantifiers (e.g. οὐκ . . . οὐδέν 367; τις 373), etc. In some cases the determination is harder to make on purely linguistic grounds (e.g. 369–72, lines that I have counted as sententious even though they are constructed around the first-person forms εἶδον and ἔλθω): see Friis Johansen 1959: 97. For the problems of delimiting gnomic speech, see also Introduction, §2.5, and Norrick 2014. 7 The one exception, perhaps, is the conditional clause in 795, as reported by the Messenger: but the generic content there is limited to the subordinate clause.

Orestes’ Linguistic (Dis)guises

169

words, talking about general truths is a way of not talking about himself, to obscure the man underneath the disguise. It is vital to note in this regard that Orestes’ moralizing speeches occur at moments where something Electra has unwittingly said leaves him on the verge of being overcome by emotion (οἴμοι, τόδ᾽ οἷον εἶπας 290, ‘ah! what a thing you say there!’; φεῦ 367) and blowing his cover. The general statements thus serve as a way for Orestes to avoid saying something about himself at moments when his disguise is most vulnerable. A related reason for the concentration of gnomic expressions in the early scenes featuring Orestes incognito is the enhanced potential that they offer for dramatic irony, in a situation that particularly solicits it. Recent paroemiological research has emphasized ‘that proverbial expressions are, in normal usage, always applied to a particular situation, but also that the particular situation helps to determine the meaning of the proverb’ (Lardinois 2006: 215). Because different addressees (whether internal or external) find themselves in different particular situations, generalized expressions can ‘mean’ differently to different addressees, and this ambiguity opens up the possibility of using such expressions for dramatic irony. With regard to this last point, it is worth pointing out that all of Orestes’ generalizations may be classified, according to Seitel’s taxonomy (on which see Introduction, §2.5), as either first-person sayings* or third-person sayings*, but in some cases straddle the line between the two. The intricacy of Euripides’ ironies is that what he makes the ‘stranger’ say about others (‘Orestes’ or the Peasant) can also be taken to apply to the speaker himself. These classifications also mark an important difference between Orestes’ gnomic utterances and those of Electra discussed in the last chapter, which were primarily second-person sayings* and were designed to ground her accusations against Aegisthus and Clytemnestra in established moral thinking. Orestes, on the other hand, invokes popular notions to frame his own behaviour as part of a larger socioethical fabric. To sum up, the density of generalizing speech in Orestes’ scenes with Electra before the recognition may be ascribed both to an internal motivation on the part of Orestes (who, in his disguise, uses it as a distancing technique, to divert attention from himself) and to the instability of the meaning of such language, which is exploited by Euripides as part of a range of literary techniques intended to create suspense and irony. With all this in mind, let us examine the three instances of sententious speech uttered by Orestes.

170

Orestes’ Linguistic (Dis)guises

2.2.1. ‘A man in exile is powerless’ (236) Orestes’ first generalizing contribution is a mere half-line (234–6): ΟΡ. οὐχ ἕνα νομίζων φθείρεται πόλεως νόμον. ΗΛ. οὔ που σπανίζων τοῦ καθ᾽ ἡμέραν βίου; ΟΡ. ἔχει μέν, ἀσθενὴς δὲ δὴ φεύγων ἀνήρ. Or. El. Or.

He languishes, moving from city to city. Please don’t tell me8 he lacks daily sustenance! That he has, but a man in exile is powerless.

The short phrase ἀσθενὴς δὲ δὴ φεύγων ἀνήρ differs from the two longer sententia-filled speeches at 290–6 and 367–400 in that it does not have a particularly moral content and is not sustained. It is also different in that it is quite explicitly designed to describe (the ‘absent’) Orestes, as is implicit in the contrast between ἔχει μέν (with him as its particular subject) and the generic statement introduced by δὲ δή.9 Yet the principles outlined above apply here as well: the generalizing tone enhances the irony (Electra does not realize that the phrase is directly applicable to the man in front of her,10 whereas the audience knows that the speaker’s general claim is in fact a description of his own specific situation), and is at the same time, by virtue of its indirectness, a distancing technique for Orestes. It is noteworthy in this respect that his gnōmē is followed immediately by a change of topic (Electra inquires about the message the ‘stranger’ bears, a change to a more pressing point; δὲ δή, again in 237). Orestes has successfully deflected Electra’s interest in the circumstances of ‘Orestes’ without letting on too much, and his gnōmē effectively caps the discussion of this point. Orestes is thus able to use his generalization as a way of topic 8 See Caspers 2010: 342 on οὔ που questions: ‘οὔ που questions typically distil a proposition (p) from information that is being offered, and ask the interlocutor to disaffirm p—because p is too terrible to be believed, or because the information offered previously leaves room for doubt’. 9 It is possible to take ἀσθενής as a predicate adjective with the subject of ἔχει, and read φεύγων ἀνήρ as a predicative appositive: ‘That he has, but he is helpless as a man in exile’, in which case the expression would not be a generalization at all. Similarly (though I think this unlikely) we could read ἁνήρ instead of ἀνήρ, also eliminating the generalizing force. On balance, I think the reading as a generalization is preferable, and Electra seems to repeat it that way to the Peasant at 352. 10 It is interesting to note here that Herwerden emended to ἀσθενὴς δέ τοι, with τοι ‘to point the applicability of a universal truth to the special matter in hand’ (Denniston 1954: 542). But that is, I think, something that Orestes should not be doing here.

Orestes’ Linguistic (Dis)guises

171

management*; and, as I have argued in the previous chapter (§3.1), he does not let go of control over the topic of the stichomythia again.

2.2.2. Pity and intelligence (290–6) Some 50 lines later, the stichomythia draws to a close and Orestes delivers the first of his two longer moralizing speeches (290–6). Euripides has constructed the finale of the stichomythia carefully. Following directly upon the mention of the Old Man in 285–7 (which takes up a full line at 287: πατρός γε παιδαγωγὸς ἀρχαῖος γέρων ‘our father’s tutor, a very aged man’), the discussion of Agamemnon’s grave (288–9), which induces Orestes’ speech, appears specifically designed to foreshadow the resolution of the recognition plot (in which the Old Man and Agamemnon’s grave play crucial roles). It is worth mentioning in this regard that at 288 (ὁ κατθανὼν δὲ σὸς πατὴρ τύμβου κυρεῖ; ‘And your father who died, does he have a tomb?’) Orestes asks, for the first and only time, a question to which the audience can expect him to already know the answer. They have heard Orestes say that he has visited his father’s grave and left offerings at it (90–2); his enquiring after its very existence is thus particularly noticeable. Again, the effects of this gambit are multilayered: we may see Orestes here trying hazardously to maintain his disguise while at the same time utilizing the freedom it gives him to feign ignorance (he wants to know how Electra feels about the grave, and he probes her by pretending to know less than he actually does). Again, this is also a deliberate manipulation by Euripides of the differing levels of knowledge between characters and audience, for which the question is suggestive of a literary tradition that will come to the forefront explicitly a few hundred lines later. It might even be argued that the nod to Aeschylus’ Choephori will have tempted the audience to expect the recognition to follow immediately (an ‘intertextual’ variant of the game Euripides has been playing throughout this scene); but this expectation is—again—thwarted for quite some time. It is significant, then, that this foreshadowing of things to come concludes the stichomythia. As so often in Euripides, the scene is now capped by a general reflection. The beginning of Orestes’ speech shows him overcome by emotion after a comment made by Electra (289). It is not the first time that something like this has happened: on three earlier occasions, some unwitting remark by Electra has caused Orestes to lose restraint and gasp (244 φεῦ φεῦ 262 φεῦ 282 φεῦ). In

172

Orestes’ Linguistic (Dis)guises

each case Orestes covers up the ‘near miss’ with a statement designed to explain his reaction: • 244–5: As Electra voices her love for Orestes and Agamemnon, he almost reciprocates, but manages to say instead that ‘her brother’ feels the same way (Orestes averts disguise disaster by using the third person). • 261–2: When he learns about the Peasant’s right-mindedness, Orestes (Electra’s kurios) is, again, almost overwhelmed by emotion but manages to comment on it in general terms rather than go into the meaning it has for him personally. That the ‘messenger’ has been preoccupied here with Orestes’ role as kurios is evident from his previous line of questioning (248, 250, 260; see ch. II, §3.1), but he has thus far managed to make no show of his personal involvement.11 At 262, however, Orestes cannot avoid gasping (φεῦ) but repairs the damage by using an impersonal, passive construction: εὖ δραστέον.12 • 281–2: finally, as Electra proves her willingness to cooperate in the execution of the vengeance plot, he gasps but then pretends to long for ‘Orestes’ to hear her words (εἴθ᾽ ἦν Ὀρέστης πλησίον κλύων τάδε), using his own name to divert attention. The present instance is the culmination of this pattern, featuring a fuller emotional exclamation (οἴμοι, τόδ᾽ οἷον εἶπας 290) and a more elaborate cover-up by Orestes. His first one-and-a-half line of generalizing is telling (290–1): αἴσθησις γὰρ οὖν καὶ τῶν θυραίων πημάτων δάκνει βροτούς. I mean, awareness of suffering, even that of others, is painful to mortals. It is possible to see a hint of reserve even in the ‘exclamation’ ᾤμωξ᾽ ἀδελφὸν σόν (248, ‘I groan for your brother’), since ᾤμωξ(α) is a ‘tragic’ aorist. In his (rightly) standard article on this use Lloyd (1999: 28) argues that the tragic aorist indicates politeness or reserve. This particular aspect of Lloyd’s analysis seems, however, much more open to doubt than his central claim that tragic aorists are performatives. For some recent modifications to Lloyd’s views, see Bary 2012, Mossman on Med. 791; Nijk (in press) largely supports Lloyd’s take on the effects of the usage, though on the basis of a very different linguistic explanation: ‘the aorist strictly binds the designated event to the utterance itself. As soon as the utterance has been made, the matter is dropped.’ 12 The impersonal use of the gerund is another way of speaking in general than in specific terms. On the significance of the passive voice in diverting focus from the agent performing an action, see George 2005, esp. 19–42. 11

Orestes’ Linguistic (Dis)guises

173

The generalization allows Orestes to justify his strong reaction (οἴμοι) to Electra’s disgruntled comment in 289 while maintaining his disguise.13 The function of the expression is made explicit through the particle combination γὰρ οὖν, a neat little touch in Orestes’ language that suggests that he is actively trying to ameliorate the ‘damage’ done to his disguise by his emotional outburst. Denniston (1954: 446) explains that οὖν, in this combination, ‘adds to γάρ the idea of importance or essentiality’.14 This has much to do with the basic function of οὖν, which has been described as ‘indicating that what went before need occupy [the] listeners only in so far as it may assist them in grasping what follows’ (van Ophuijsen 1993: 84): οὖν signals that the text segment in which it is contained is part of the principal message of the speaker, and that whatever precedes is preliminary, introductory, or explanatory to it. The combination γὰρ οὖν, then, serves at once to explain or elaborate on what precedes and to diminish it in status, the new contribution being portrayed as the essential point. The combination is very suitable in mildly selfcorrective contexts, where a speaker explains (γάρ) his preceding utterance by adding one that is more to the point (οὖν); the force could then be paraphrased by ‘what I meant by that is . . . ’, or even by the collocation ‘because anyway’, as is sometimes found in colloquial English. Some parallels from tragedy may be helpful here.15 Compare for example this exchange from Sophocles’ Antigone (740–1): Κρ. Αἱ.

ὅδ᾽, ὡς ἔοικε, τῇ γυναικὶ συμμαχεῖ. εἴπερ γυνὴ σύ. σοῦ γὰρ οὖν προκήδομαι.

Cr. This boy is fighting, it seems, on the side of the woman. Hae. Only if you are a woman: I mean, it is your best interests that I have at heart. 13

See Basta Donzelli 1978: 93–4 and Cropp ad loc. The very fact that these particles combine is not trivial: many students, going by their default translations of particles, will believe the two to have diametrically opposed ‘meanings’, γάρ introducing explanations (‘for’) and οὖν inferences (‘therefore’). But, especially in the case of οὖν, these default translations are misleading: οὖν need not have inferential force, though it may have such a nuance (it is worth examining the illuminating discussion of this point in Sicking 1993: 25). 15 Apart from the cases discussed here, γὰρ οὖν occurs in tragedy at Aesch. Ag. 524, 674, Eum. 372; Soph. Ant. 96 (text disputed), 489, 771, 1255, Phil. 298, 766, OC 980, 985; Eur. Cyc. 251, Med. 533, Heracl. 202, Hipp. 666, Or. 1147. The collocation is frequent in Plato (181 occurrences), especially in short answers; on this use and the combination more generally, see Bakker 2009a (my reading of the particles here diverges from her general argument, however). 14

174

Orestes’ Linguistic (Dis)guises

Here σοῦ γὰρ οὖν προκήδομαι explains Haemon’s strange conditional, and is at the same time what he really means by it. Slightly differently, a bewitched Pentheus asks Dionysus in Euripides’ Bacchae (921–2): [ὁρᾶν μοι δοκῶ] καὶ σῷ κέρατα κρατὶ προσπεφυκέναι. ἀλλ᾽ ἦ ποτ᾽ ἦσθα θήρ; τεταύρωσαι γὰρ οὖν. I seem to be seeing horns attached to your head . . . were you an animal at any point before? . . . because anyway, you are a bull now.

Pentheus adds τεταύρωσαι γὰρ οὖν to account for his interruption (ἀλλ᾽) of his own observations with a confused question, and resumes the point he was making. Returning, then, to the present instance, γὰρ οὖν seems to suggest that Orestes feels the need not only to explain why he had an emotional outburst, but almost to ‘retract’ it, to recast it in generic terms more pertinent and appropriate to his ‘stranger’ persona.16 Orestes deliberately employs sententiousness as an instrument to extract himself from too personal an involvement, the generalizing tone and the carefully applied word θυραῖος (‘external’) providing the distance he needs to secure his identity (the audience can see, of course, that the πήματα causing the sting are in this case anything but θυραῖα; the word is used again, to similar ironical effect, at 832). Let us now turn to the next five lines (292–6): λέξον δ᾽, ἵν᾽ εἰδὼς σῷ κασιγνήτῳ φέρω λόγους ἀτερπεῖς, ἀλλ᾽ ἀναγκαίους κλυεῖν.17 ἔνεστι δ᾽ οἶκτος ἀμαθίᾳ μὲν οὐδαμοῦ, σοφοῖσι δ᾽ ἀνδρῶν· καὶ γὰρ οὐδ᾽ ἀζήμιον γνώμην ἐνεῖναι τοῖς σοφοῖς λίαν σοφήν. But tell me your story, so that I may learn and relay to your brother words that are without joy but must be heard. There is no pity to be found in ignorance, but in men with discernment—it is, in fact, not without penalty that people with discernment have a mind that is too discerning.

The last three of these lines have suffered from the suspicion so often cast by scholars on Euripidean generalizing speeches, on which I will have more to say in §2.2.3. Lines 294–6 were seen to be ‘unsuitable in 16 See Keene ad loc., who perceptively notes: ‘The words αἴσθησις γὰρ κ.τ.λ. are added apologetically for the interest Orestes, though an apparent stranger, feels in the treatment of Agamemnon.’ 17 For the accentuation (probably aorist infinitive κλυεῖν rather than present infinitive κλύειν here), see West 1984: 179.

Orestes’ Linguistic (Dis)guises

175

this place’ (versus hoc loco inepti, Nauck, app. crit.) and have been either deleted by some or (more frequently) transposed by others after 291.18 Denniston ad loc. formulates the objections as follows: These lines continue the train of thought begun in 290–1, and the demand expressed at 292–3 stands awkwardly in between. The arrangement is possible, but hardly probable; . . . 294–6 would come much more appropriately after 291.

Little needs to be added, however, to the defence of the lines and ordering given by Basta Donzelli (1978: 93–102). The key, as she has argued, is to recognize the polysemy of the words ἀμαθία (the abstract noun standing for the plural ἀμαθεῖς, abstractum pro concreto) and σοφός in Orestes’ generalizations. Detached from this particular context, Orestes’ expression would most naturally mean that those without an innate intelligence, even ‘moral wisdom’,19 are incapable of pity. But ἀμαθία can also designate a lack of schooling, training, or specific information (while σοφία means the opposite),20 which opens up a different interpretation, namely that someone must know about a person’s suffering before he/she can feel pity for that person (sometimes agonisingly so: καὶ γὰρ οὐδ᾽ ἀζήμιον, ‘and in fact it’s not painfree’). What Orestes meant by λόγους ἀτερπεῖς, ἀλλ᾽ ἀναγκαίους κλυεῖν (293, ‘words that are unpleasant, but necessary to hear’) becomes clear only now: in order for Electra’s brother to have the appropriate response to her plight, he must learn about it fully. Thus, far from being irrelevant to the current context, Orestes’ general reflections

18 Deleted: Steinberg, Wecklein 1898. Transposed: e.g. Wecklein 1906, Denniston. Michelini (1987: 208 n. 113) voices the objections of many in saying that ‘the gnōmē . . . hangs a bit detached from its setting’. Much of the uncertainty concerning lines 294–6 is due to the fact that Stobaeus (3.27) seems to quote them as part of Antiope (that is, he combines 294–6 with two lines from that play, Eur. fr. 202); the two quotations are clearly unrelated, however, and it is more likely that an attribution has simply been omitted by Stobaeus, as happens often (see Luria 1929: 85–7). 19 For this aspect of σοφία in these lines, see Dover 1974: 120 and Pucci 1980: 28, 78, 174. Bain (1982: 273) thinks that this makes Basta Donzelli’s interpretation of 294–6 ‘far‐fetched’, but Dover points out that we find several shades of meaning of the term ‘side by side’, and it strikes me as therefore quite possible that two shades overlap here. As I have argued, generalizations often ‘mean’ differently to different addressees, and this would be enhanced in this case by the instability of the terms ἀμαθία and σοφία. 20 Basta Donzelli (1978: 95 n. 44) offers Ar. Ran. 1109 (cf. σοφῶν at 1118), Eur. Supp. 421: both are adequate. Note also that ἀμαθής mean ‘ignorant of something’ (construed with a genitive or a prepositional phrase with πρός or περί; see LSJ s.v. A.1).

176

Orestes’ Linguistic (Dis)guises

have a very specific application for Electra. Again, the listener’s specific context conditions the interpretation of gnomic speech. Finally, for the audience (with its surplus knowledge of what is going on on stage), the generalization may suggest one final point: the ‘real’ Orestes turns out to be a glutton for punishment and risks saying so almost explicitly. By opening himself up to the full weight of Electra’s suffering, he invites suffering of his own: his knowledge will be too much—λίαν.21 In sum, the problem with Wecklein’s ordering (294–6 placed before 292–3) is not only that it flows not nearly as neatly as he and Denniston suggest, but also that the change destroys the pattern that we have seen emerging from Orestes’ language throughout the scene: as was the case a few lines before, Orestes is at pains to explain why a ‘messenger’ would say a certain thing (or, in this case, make a certain request). Electra must tell her story so that ‘Orestes’ can be informed of it, and from this new information he will derive the sense of pity— or, still more, pain—that he should feel towards his sister. But the generalizations that follow ‘mean’ differently to the speaker and his internal and external audiences. With one blow, the ‘stranger’ manages to prevail upon Electra to relate the account that the man underneath wants to hear, and at the same time to insulate himself from suspicion about his identity. The mechanisms employed before do the trick again: a third-person reference to Electra’s ‘brother’ (σῷ κασιγνήτῳ)22 and a dose of unspecific moralizing deflect attention; Orestes’ disguise is secure once more. 21 It is possible to read a slightly more dangerous, almost reckless streak in Orestes’ constant hints and ironies. Just as Clytemnestra can be seen to ‘want to have said’ that Δίκη will lead Agamemnon into a house that is ἄελπτος (Aesch. Ag. 911; Fraenkel ad loc. rightly remarks that the fact that Clytemnestra ‘entrusts her true desire to the κρείττονες’ openly ‘is of the highest importance’), Orestes can be seen to consciously hint at his true identity, deliberately walking a fine line between deception and revelation (Electra is consistently deaf to the clues, however). This seems to be the interpretation of Solmsen (1967: 39–42); it is certainly true of Orestes’ ironies in the messenger speech (see §2.3). It would go too far to claim that such psychological subtleties are an inevitable interpretation of the text, however, and I subscribe to Cropp’s (2013: 7–8) general point that Orestes ‘need not be . . . supplied with alternative motives for keeping his disguise’. 22 In itself, the purpose clause ἵν’ εἰδὼς σῷ κασιγνήτῳ φέρω λόγους (‘so that I may know and convey the message to your brother’) seems specifically designed to motivate the stranger’s request—he wants to know not for himself but in order to provide ‘Orestes’ with information. The situation is identical at 250: εἴφ᾽, ὡς ἀκούσας σῷ κασιγνήτῳ λέγω (‘speak, so I may hear it and relate it to your brother’).

Orestes’ Linguistic (Dis)guises

177

2.2.3. Evaluating character (367–400) The final, longest, and most controversial instance of Orestean moralizing occurs some 70 lines later (367–400): it is a sustained series of reflections on the impossibility of judging a person’s worth by simple, external signs. Few scholars, while discussing this passage in the past century or so, have left it entirely intact,23 and the interventions proposed are imposing and varied. Reeve would leave only lines 383–5 in place and others made different deletions:24 • 368–72: deleted by Reeve (371–2 Schenkl, 369–72 Vitelli); • 373–9: deleted by Wilamowitz, Vitelli, Wecklein, Page, Friis Johansen, Reeve, Diggle, Distilo; • 383–5: considered doubtful by Murray, Reeve; deleted by Distilo; • 386–90: deleted by Wilamowitz, Vitelli, Page, Friis Johansen, Reeve, Diggle; • 396–400: deleted by Reeve. The arguments against the passage offered by these scholars pertain in some cases to technical and linguistic issues (staging, structure and style; e.g. the use of γάρ in 380, which, according to Reeve (1973: 152), ‘makes no sense anywhere except after 367’), but above all to the perceived lack of ‘relevance’ of the various reflections (e.g. Reeve 1973: 152, again, on 386–90, with caution typically thrown to the wind: ‘irrelevant, and no more words need to be wasted on them’).25

23 Denniston on 367–72, 373–9, 386–90, Basta Donzelli 1978: 229–42 and 1991b: 113–19, de Romilly 1983: 416, 7 n. 24, Goldhill 1986b, Egli 2003: 225–9. Cropp on 367–400 is inconclusive, but appears to argue in favour of leaving the whole speech intact (this contrasts with Cropp’s earlier position, at least on 386–90; see his commentary ad loc.). 24 Reeve 1973: 151–3, Wilamowitz 1875: 190–3, Schenkl 1874: 88–9, Vitelli 1880: 512, Page 1934: 74–5, Friis Johansen 1959: 95–8. My list is based on Basta Donzelli 1991b: 113 and Distilo ad locc. 25 The passages deleted are usually explained as histrionic interpolations rather than marginal adscripts that intrude into in the text: the latter possibility is ruled out, at least in the case of 367–79, by the appearance of those lines in a papyrus dated between 250 and 210 BC (P.Hibeh. 7; see Page 1934: 74). It is, as most scholars agree, hardly important that Diogenes Laertius (2.33) cites 379 as a line from the Auge. Even if that is true (the MSS have ἐν τῇ αὐτῇ and ἐν τῇ αὐτοῦ, hence ἐν τῇ Αὔγῃ is merely a feasible emendation; see Distilo 2012: 645–8), we could either accept it as a repeated line, or more likely as a mistake. De Romilly (1983: 413) put it mildly when she said that ‘the authors of our citations sometimes make mistakes’ (nos auteurs de citations se trompent parfois).

178

Orestes’ Linguistic (Dis)guises

The latter point concerning relevance is attributable to a type of reasoning that (with respect to our present case) has been rightly deconstructed early on, by scholars such as Grube (1941: 98),26 and more recently, for example by De Romilly, Basta Donzelli, and Goldhill (whose arguments lead him to somewhat different conclusions, however).27 The point, trivial but apparently worth underlining again, is that the literary Stilgefühl of modern critics cannot by itself suffice as an argument for deleting entire passages from Greek tragedy, especially when those passages are on the one hand a clear mark of an author’s style (and generalizations are a mark of Greek authors in general and of Euripides in particular), and on the other hand alien to what most modern critics would consider agreeable. In other words, when Wilamowitz (1875: 192) argues that his trimmed version of 367–400 is genuina et optima Orestae oratio, quae ne verbum quidem addi patitur (‘the genuine and best version of Orestes’ speech, which brooks not a single additional word’), the equation of what he considers to be optimum and genuinum should be open to considerable doubt. But perhaps even more important, in our present case, is the fact that many of the objections concerning the irrelevance of (certain segments) of Orestes’ speech do not hold up to close scrutiny.28 Before looking at the individual deletions proposed by scholars, it is perhaps best to begin with a reading of the speech as a whole, as I see it. The problem treated by Orestes is thematized at the very start in 368: he discusses the difficulty of judging εὐανδρία.29 But, as becomes clear in the course of the speech, Orestes is concerned not 26 For Grube (1941: 98), though, the target is critics who interpret ‘irrelevant’ generalizations as a biographical expression of the poet’s own experience (i.e. fall into the biographical fallacy): ‘The picture of Euripides as a poet who expresses his own opinions with little care for dramatic relevance is so common that a vigorous protest should be made against it.’ This criticism is in fact apposite in the case of Denniston, who, on 386–90, cites Aulus Gellius and argues that the ‘outburst against athletes’ is a ‘violent reaction’ against the early training that Euripides is said to have received in his youth. In any case, though Grube’s critique is perhaps no longer as relevant as it was in 1941, the points he makes in the process have lost none of their lucidity. 27 For references, see n. 23. Goldhill’s reading is discussed more fully in what follows. Similar arguments, on scenes in Alcestis and Hippolytus, may also be found in Conacher 1981. 28 I am again in substantial agreement with Basta Donzelli (1991b: 113–19) and borrow much of her argument in what follows. 29 It is safe to say, I think, that the terms χρηστά (370), γνώμην μεγάλην (372), ἀγαθός/ἄριστος (378, 382) and εὐγενεῖς (385) all relate to the same concept of εὐανδρία for the purpose of this passage.

Orestes’ Linguistic (Dis)guises

179

so much that there are no signs whatsoever that can help in ascertaining εὐανδρία, but (and this is a crucial difference) that there are no outward and easily discernible signs.30 This problem is a recurring Euripidean theme: at Medea 516–19 Medea complains about the lack of a physical ‘mark’ (χαρακτήρ) that identifies persons lacking in character; at Hippolytus 925–31 Theseus, too, wishes that there were an ‘obvious sign’ (τεκμήριον σαφές) to identify base men, and he suggests a second voice, which always speaks the truth; at Heracles 655–68, the old men of the Chorus suggest that a second life awarded to good men might be a ‘clear mark of virtue’ (φανερὸν χαρακτῆρ᾽ ἀρετᾶς), which would make it possible for people to ‘recognize the base and the good’ (τούς τε κακοὺς . . . γνῶναι καὶ τοὺς ἀγαθούς 665–6). The counterfactual, indeed utopian nature of such tokens makes it quite clear that the speakers involved are entirely pessimistic about the possibility of judging a person’s character by ‘easy’ standards, and Orestes in our present passage has come to the same conclusion. What links these passages is that the speaker is driven to this despair by the realization that he, she, or they have completely misjudged a person in the past (Medea feels misled by Jason, Theseus by Hippolytus, Orestes has underestimated the Peasant—the men of the Chorus in Heracles are somewhat different in that they implicitly foreshadow the dramatic reversal in Heracles’ ἀρετή):31 consequently, it is hard to deny these speeches any relevance whatsoever as regards plot and character. This is, I think, categorically true of most general reflections in Euripides; what we are then left with, as Grube (1941: 93) has seen, is ‘only a few lines here and a few lines there’ to question. Orestes, bewildered by his own lack of insight (φεῦ 367), couches his disappointment in general terms, just like Medea and Theseus; but it is nonetheless this disappointment (or, more generously, mildly surprised bemusement) that brings on the reflections. The speech itself opens with an articulation of the problem in broad terms (367–72) and then moves on to a detailed rejection of possible external criteria for judging εὐανδρία (373–9); this is followed by the key proof: the Peasant confounds any expectations that might be derived from external factors (380–2). What follows (383–90) can be described as an extended summation; Orestes rejects the search for 30 See Basta Donzelli 1991b: 114, esteriori ed evidenti a tutti (‘external and visible to all’). 31 See Bond ad loc. and Papadopoulou 2005: 31–4.

180

Orestes’ Linguistic (Dis)guises

external criteria, and does so quite firmly: note the emphatic parainesis addressed to an imaginary audience (383–5)32 and the ‘passionate asyndeton’ at 383 (Denniston ad loc.). He then affirms that the only way to judge someone is by his or her behaviour. This point is elaborated upon through a concluding justification for the parainetic exhortation, with an added political component:33 people of good ἦθος (such as the Peasant) should be judged good because they will govern well in both the public and the private sphere, whereas seemingly powerful, physically strong people are really worthless without good ἤθη, as is shown also by the fact that physical strength does not matter in battle. At 391–400 Orestes has concluded his reflections and takes the actions that are appropriate in light of them: he states that he will go into the house, he tells his slaves to bring the baggage (393–5), and (rounding off his discussion with Electra, 396–400) he thanks the Peasant for his hospitality and adds a final ironical touch in the service of his disguise (397–9). The generalizing part of the speech is perhaps not a textbook case of a perfectly structured, logical argument, but complaints about its ‘clumsiness and incoherence’ (Reeve 1973: 152 n. 16) are, I think, exaggerated. Clumsiness and incoherence, moreover, even if we admit them, do not equal irrelevance per se: ‘the awkwardness, where it exists, is always in the manner rather than the matter’ (Grube 1941: 98). All this may become clear as we examine each of the proposed deletions individually. Lines 373–9 are usually deleted because, in Reeve’s words (1973: 152 n. 16), ‘the notion of employing wealth as a criterion of εὐανδρία has no business to be entertained after 371’. But the perceived irregularity between a rejection of possible criteria and the subsequent discussion of such criteria is not really solved even if all of Reeve’s deletions are accepted: Orestes would still claim (367) that nothing is clear in people’s natures, only to argue subsequently (383–5) that there is a workable method of judging someone, namely by that person’s ἦθος. Rather than trying to do away with the irregularity,

32 The ‘segment’ of humanity targeted by such parainetic repudiations is usually expressed explicitly, as it is here through the relative clause οἱ . . . πλανᾶσθε (383–4, ‘you who err’)—e.g. at Andr. 622, Supp. 744, 949. See Bain 1975: 20, where the type of expression is called a ‘Euripidean mannerism’. 33 For the political implications of the Peasant, see Basta Donzelli 1978: 227–69; for different views, see Goossens 1962: 551–2 and Schottländer 1982.

Orestes’ Linguistic (Dis)guises

181

we can easily explain it as emanating from the structure of Orestes’ argumentation. Thus also Denniston ad loc.: [T]he irregularity of the exposition is more apparent than real. 367–72 merely state the problem. The formal search for a solution begins at 373. The first suggestion, πλούτῳ, is already out of court, but it is included for the sake of completeness and lucidity, only to be summarily dismissed.

Reeve’s (1973: 152) other argument against 373–9, that the γάρ of 380 does not follow well on 379 (see also Wecklein app. crit.), is unsustainable: Denniston ad loc. and Goldhill (1986b: 158–9) both give a perfectly natural reading. (Goldhill’s runs: ‘It’s best to let such criteria lie. For (γάρ) here is an example of a man who confounds such stereotypes.’) Since the ‘misuse’ of γάρ is the only argument that Reeve can offer against 368–72 too, the case against those lines seems very weak.34 The case against lines 386–90 is perhaps easier to make, and the lines have been seen as ‘certainly very irrelevant’ even by Denniston (who goes on to defend them, if somewhat half‐heartedly).35 Cropp (1988 ad loc.) summarized the problems as follows: The preceding lines appear to be climactic and conclusive and do not provide much of a basis for [γάρ] or [οἱ τοιοῦτοι] in 386. It is not clear why brainless athletes are being put in contrast with morally sound poor men . . . In isolation the passage would seem to make a well‐known contrast between intelligence and physical strength.36

34 As for the ‘clumsiness’ of the phrase τοῖς ἔχουσι μηδέν (375, lit. ‘those who have nothing’, i.e. ‘the poor’), I agree with Slings (1997c: 145–6) that taking this phrase as a concretum pro abstracto (i.e. τοῖς ἔχουσι μηδέν  τῷ ἔχειν μηδέν, ‘having nothing’, i.e. ‘poverty’) is something of a stretch; but I disagree that, as a result, the only way to take the Greek is to construe the dative with χρήσεται, rendering ‘he will be using the poor (as judge)’. To my mind, the dative is rather still on a par with πλούτῳ (as an adverbial modifier with κρινεῖ), even if rather loosely deployed: ‘will someone judge by (using) the poor (as the standard)?’ 35 Reeve (1973: 152 n. 17) dismisses Denniston’s careful handling of the lines (‘it does not follow with certainty that Euripides could not have put it in’) as follows: ‘So much for οὐκ ἐλήρουν ὅτι τύχοιμ᾽ οὐδ’ ἐμπεσὼν ἔφυρον (Euripides at Ar. Ran. 945)’. But it surely will not do to take what the ‘Euripides’ of Frogs says about himself as an argument for textual decisions in Euripidean drama (never mind the fact that the quotation is taken out of context: Ar. Ran. 946–7 makes clear that ‘Euripides’ is talking about his prologues). 36 Cropp has revised his position in his 2013 edition (ad loc.), defending the lines’ relevance and keeping them in the text.

182

Orestes’ Linguistic (Dis)guises

For Goldhill (1986b: 169), however, the apparent clumsiness of the lines is deliberate, and doubly relevant: It is relevant first in the characterisation of Orestes as a ‘penny philosopher’: his rhetorical argument, ill‐grounded and ill‐argued, tells us much about the awareness and approach of this man who is faced with a god‐ordered matricide. . . . Second, his (mis)use of the mannerisms and style of contemporary rhetoric . . . is relevant to the recognition that misplaced confidence in moral wisdom can lead to the horrors of tragedy, and that the σοφία of rhetoric can be an instrument to advance such horrors.

It appears to me, first, that the clumsiness is, once again, not as great as it is generally made out to be. Although the train of thought up to 385 could well be said to be complete, there is in principle nothing amiss with Orestes’ adding a final motivation (γάρ) for the parainetic command given in 383–5 (or with the use of the particle). The phrase οἱ τοιοῦτοι, moreover, is quite appropriate in its lack of specificity, which allows the listener to use complex notions to ‘fill in’ the category expressed by a condensed phrase. I would argue that οἱ τοιοῦτοι refers not simply to the εὐγενεῖς of the previous line, but to those (including the Peasant) who have proven themselves εὐγενεῖς by their conduct. The gist of 386–7, then, is this: ‘it is vital that you distinguish good people by their ἤθη, because people who are found to have good ἤθη can be trusted in the political and private spheres’. All this is fully relevant to a discussion of what makes a good man and who should be distinguished as one. The only real surprise, then, lies in the disavowal of αἱ σάρκες αἱ κεναὶ φρενῶν (387), who seem to appear out of nowhere. But the train of thought is, again, not very hard to plot: ‘judge people by their conduct, because good conduct, not brute strength, is the mark of someone who will run states and families well’. It is important to recall in this light that the term under discussion, εὐανδρία, tends to have a physical component as much as (if not more than) an ethical one:37 thus the fact that Orestes remarks on physical strength (as opposed to moral virtue) is in itself not surprising—quite the opposite, in fact. It is the automatic equation of the physical and ethical excellence that Orestes is arguing against here, thus pointing out the fact that the ‘heroic’ value system of

37

See Basta Donzelli 1991b: 116 n. 81 and LSJ s.v. εὐανδρία II.1.

Orestes’ Linguistic (Dis)guises

183

Homeric myth is no longer applicable in Orestes’ more complex reality (see Introduction, §4.2). All this is not to say that Goldhill is wrong to suppose that this part of Orestes speech reflects somewhat poorly upon himself. His parainetic exhortation is certainly ironic, considering the fact that he himself was ‘wandering’ (cf. πλανᾶσθε 385) in his view of the Peasant until just a few lines ago.38 But I cannot agree with Goldhill that the structure and rhetorical techniques of Orestes’ speech reflect ironically on him as well, and more generally on the dangers of rhetoric: what Goldhill is saying, in effect, is ‘yes, the speech is mostly irrelevant, but that is deliberate and designed to paint Orestes as snobbish and dangerously misguided,’ whereas I hope to have shown that the irrelevance of the speech has been unduly inflated (I also think that such a characterization of Orestes is somewhat off the mark). Reeve’s (1973: 153) proposed deletion of 396–400 is based on the idea that the speech must be at an end after 395, because Orestes has signalled his intention to go in (χωρεῖν χρεών 393) and has explained that intention (ὡς . . . ξένος 395). But what is physically going on on stage amply explains the resumption of the speech at 396. Orestes has addressed his slaves in order to deal with the baggage (393–5), and then resumes and concludes (οὖν 396) his conversation with Electra (396–400), adding one last tantalizing reference to her brother. Given the absence of any plausible reason for an interpolation, Reeve’s deletion strikes me as very wide of the mark.39 There is, finally, one other controversy regarding this scene, and my discussion of it will lead to a concluding assessment of 367–400. At 357–63, the Peasant invites Orestes and Pylades into his house. Regardless of whether or not he also tells Orestes’ servants to take the 38 It is quite possible that Orestes’ irony at 383–5 is not to be seen as unintentional at all, but that he includes himself among ‘those who err full of meaningless fancies’ and is thus outright apologetic and self-deprecating here. Similarly, in the parainetic passages at Supp. 744, 949 (see n. 32), Adrastus is certainly thinking about himself as one of the mortals addressed (note the proliferation there of first-person forms surrounding the address in the second person). 39 See also Goldhill 1986b: 196–7. Reeve also objects to the idea that Orestes could mention the oracle to Electra (399–400) when she has not been told of it before. Yet ‘Euripides may have overlooked her ignorance for the sake of the continuing dramatic irony’ (Cropp on 396–400): such irony seems in fact the very point of these lines. The idea that lines 399–400 are addressed to Pylades (Wecklein ad loc.) is implausible: the first half of 399 (ἴσως δ’ ἂν ἔλθοι, ‘he may perhaps come’) must be aimed at Electra, and the rest of 399–400 is added to explain those words (note γάρ).

184

Orestes’ Linguistic (Dis)guises

baggage in at 360,40 it is remarkable that his hospitality is accepted only some 30 lines later. This ‘problem’ of staging is, in fact, one of the reasons why scholars have looked askance at parts of 367–90 (see, e.g. Reeve 1973: 151). But the Peasant’s offer of hospitality does not only lead to an apparent violation of staging principles: there are textinternal indications of its ‘abnormality’ as well. First, the Peasant himself sees fit to explain his offer as a rejection of ‘ill‐bred behaviour’ (ἦθος δυσγενές), even in the face of poverty. Second, Orestes’ initial reaction to the offer is strong (πρὸς θεῶν 364, ‘by the gods!’), and he is profuse in his later acceptance of it (390–2, 395). Finally, the Peasant is roundly rebuked by his wife for offering hospitality in the first place (404–5). Thus the Peasant’s gesture is clearly meaningful, and not even the elimination of the staging ‘problem’ by means of extensive deletions can obscure this fact. Rather, considering it to be a ‘problem’ in itself demonstrates an utter failure to understand the niceties of this scene. I quote Goldhill once more: This does not . . . constitute a difficulty in the staging, but is rather an important element of characterisation that is quite misunderstood by Reeve . . . The awkwardness in this scene is the careful and witty dramatization of an awkward situation in which a young man of noble birth is, to his evident surprise and Electra’s distress, offered hospitality by a man of extreme poverty and (as far as Orestes is concerned: cf. 252, 267f.) considerably inferior birth. This difficult social encounter prompts Orestes’ reflections on wealth and social status. (Goldhill 1986b: 161–2)

Goldhill is exactly right: it is the offer of hospitality from an entirely unexpected corner that brings Orestes to his reflections. I would, however, prefer again to slightly rephrase Goldhill’s subsequent 40

A much debated question: scholars have doubted that the Peasant could give orders to the servants of noblemen, and find it strange that the order is not carried out until 393 (see Bain 1981b: 36–7, Basta Donzelli 1991b: 110–11, Distilo ad loc., Cropp ad loc.); Mastronarde (1979: 106–7), however, defends the action, arguing that μηδὲν ἀντείπητε (‘don’t object’) may be a reaction to the servants’ non-compliance. To my mind, condemning this line as interpolated (what was the interpolator’s aim?) takes away a nice touch from the Peasant’s character (‘the officiousness perhaps suits the Peasant’, Cropp ad loc.) and deprives Euripides of any leeway for modifying the ‘rules of the stage’ (Bain) even when that is dramatically meaningful (as I believe it is here). Distilo’s additional objections to the sense of τεύχη as bagagli seem to me besides the point: why not simply assume that the ‘baggage’ to be brought in is in fact equipment/ armour (a regular sense of τεύχη)?

Orestes’ Linguistic (Dis)guises

185

conclusion that ‘[t]he hollow rhetoric of Orestes’ reflections is . . . essential to this depiction of a snobbish nobleman in an unexpected and difficult social situation’: unexpected and difficult, certainly; but perhaps not so much snobbish.41 It strikes me rather that the Peasant’s display of an admirable ἦθος brings Orestes to realize the extent to which he has underestimated this person merely on the basis of status and wealth—a fact that he deplores extensively, but also a flaw of judgement that he wants to ascribe not merely to himself but to wider problems in recognizing character. In sum, the generalizations uttered by Orestes in 367–400 are very much ‘in character’: confronted with an unexpected paradigm of virtue, Orestes is forced to admit that the preconceptions he previously held left him unprepared to find this type of excellence in this type of person. He ascribes his own lapse of judgement to general problems in the evaluation of character, which explains why he turns to prolonged general reflection. The sententious speech at the same time performs similar functions as described earlier for Orestes’ other reflections: they allow for different levels of interpretation (as they can be applied variously to the Peasant or Orestes himself), and they contribute to Orestes’ disguise in that they provide him with the opportunity to speak in general terms about things to which the hidden Orestes has a deeply personal reaction. In the conclusion of the speech (391–400), Orestes brings out the rest of his deceptive arsenal in the form of references to himself in the third person (Ἀγαμέμνονος παῖς 392, ‘Agamemnon’s son’; κασίγνητος . . . σός 397, ‘your brother’) and ironical ambiguity (ὅ τε παρὼν ὅ τ᾽ οὐ παρών 391, ‘the one who is here and the one who is not here’).

2.3. The second disguise (774–858) I will deal with the language Orestes is made to use by the Messenger in more detail in chapter V. But it is worth pointing out here a few ways in which Orestes’ second linguistic disguise differs from his first one. There is, to begin with, an obvious change in that Orestes’ second persona is much further ‘removed’ from his true self: he and Pylades pretend to be Thessalians on their way to a sacrifice (781–2) rather 41 What, for example, is snobbish about οὗτος . . . ἄριστος ηὑρέθη (380–2, ‘this man has been found to be most excellent’)?

186

Orestes’ Linguistic (Dis)guises

than messengers from Orestes himself. This requires Orestes to utter lies that are more fundamentally untrue (in his first disguise, he manages to say very little that is patently false), and we see him confronting some interesting consequences, as Aegisthus invites him to wash (791–2) and challenges him to prove his good Thessalian stock (815–18). These challenges to Orestes’ second, disguised persona lend great suspense to the Messenger’s narrative, as I will argue in chapter V. Even the similarities between the two scenes point to their differences: again, the conversational partner of the disguised Orestes brings up the topic of the real Orestes (Aegisthus refers to him with τοὺς . . . ἐμοὺς ἐχθρούς 807, ‘my enemies’; Ἀγαμέμνονος παῖς 833, ‘the son of Agamemnon’), but this time he never comes close, in his reactions, to revealing himself in an emotional outburst. Rather he plays with the references, injecting some ironies of his own (see especially φυγάδος . . . δειμαίνεις δόλον, | πόλεως ἀνάσσων; 834–5, ‘do you fear an exile’s ploy, when you rule the city?’), which appear much more devious and intentional than those uttered from under his first disguise. In sum, the ways in which Orestes disguises himself on the two occasions are very different from each other. Orestes’ second mask is much less transparent—his assumed persona has little relation to the ‘real’ Orestes. He does not take things personally and there is less need for him to suppress his personal reactions: it is, perhaps, partly for these reasons that in the messenger speech Orestes uses no generalizing language, although the small number of lines we ‘hear’ Orestes speak under this second disguise makes such conclusions highly tentative at best.

3. CONCLUSI ON A few conclusions may be drawn from the preceding discussions. First, as already argued in my Introduction (§2.5), an analysis of gnomic speech in Greek tragedy should consider the ‘relevance’ of a particular general reflection not only in the context of the material surrounding it, but also in the context of the communicative setting (particular speakers, particular addressees, particular aims) in which it is embedded: unless we ask why a particular character would use

Orestes’ Linguistic (Dis)guises

187

gnōmai in the first place and what it says about that character that he or she does, we run the risk of ignoring a range of subtle literary effects. In the case of Orestes, I believe that his use of general reflections is one in a number of features of his ‘style’ that is to be related to his disguise, to his need to present to Electra a persona without much personality of his own. At the same time, the multiplicity of (possible) meanings that such generalizations license provides Euripides with an ideal opportunity to lace his conversations with dramatic irony. Second, I hope to have shown, in this chapter and the preceding two, that any study of tragic characters needs to examine the kind of speakers they are, that is, how they behave in their use of language (I will return to this point in the conclusion of the book). In the next few chapters issues of characterization through language will remain central to my purpose, but some other questions will rise to prominence. Chapters V and VI take the structure of narrative and of argumentative text, respectively, as their central object of research. First, however, chapter IV looks at stichomythia and at the same time offers an exercise in a particular kind of textual criticism: the distribution of parts among speakers.

IV Redrawing the Lines Pragmatics and Gender in Textual Criticism

in fatto di distribuzione di battute nei testi teatrali la testimonianza dei mss. non sempre è attendibile, spesso è fuorviante, in nessun caso decisiva (‘in the case of the distribution of lines in theatrical texts, the manuscript evidence is not always reliable, often deceptive, and in no case decisive’). —Giuseppina Basta Donzelli (1991b: 5)

1. INTRODUCTION

1.1. Vexed passages The bleak appraisal of manuscript evidence quoted above will not inspire much confidence in those who hope to determine who says what in a tragic text, especially when that manuscript evidence is itself scanty at best.1 The author of this epigraph, whose scholarly output consists almost entirely of work on Electra2 (an ‘alphabetic’ play with a textual tradition that goes back to a single manuscript), will have been confronted with this particular problem more than most. Electra, in fact, features several of the most vexed passages of Euripidean

1 Parts of this chapter, particularly §3.1, have been published in a slightly different form in van Emde Boas 2015 (which also discusses Med. 49–95 and Hel. 625–59). 2 This work consists of the Teubner text (Basta Donzelli 1995a), a major monograph (Basta Donzelli 1978), and 10 other articles on various textual and metrical issues, including one on line-to-speaker attribution (Basta Donzelli 1991a).

Redrawing the Lines

189

tragedy when it comes to line-to-speaker attribution. Two (1177–232 and 1292–305) have already been discussed in chapter II (§6.1, 6.2); in this chapter I will focus on two other, perhaps even more controversial ones: 671–93 and 959–87. Both are passages of stichomythia in most editions (although the first is not marked as a stichomythia in the manuscripts), and the present chapter is meant to contribute to our understanding of that form as well. Much has been written on these two passages, yet a reappraisal of the issues involved is, I think, not only desirable but also a necessary prerequisite for a cogent interpretation of the play. In other words, determining ‘who says what’ is not merely a matter of concern for textual critics striving to arrive at the best possible text; it is also (especially in the case of 959–87) of pivotal importance for our conception of plot and character. In fact, much of the scholarly discussion surrounding these passages can be traced back to differing conceptions of Electra’s and Orestes’ characterization.3 When Kubo (1967: 25) argues that ‘[d]ramatic characters can be construed in any way you like, since the same line can be delivered in many different ways’, the point can also be reversed: the imagined method of delivery of a series of lines and, by extension, even their division among speakers can depend entirely on a scholar’s construction of a particular dramatic character. ‘X is not something that character Y would say at this point’ is often what the entire argument boils down to—to which another scholar may object that X will have been performed by Y in an entirely different fashion from what the first critic supposes and is therefore not inappropriate at all. Such subjective arguments are unavoidable: all that remains for the critic to do is to base his judgements as much as possible on ‘firm’ evidence in support of his or her views, wherever such evidence is available. It is in the discovery of such evidence that, in these particular cases, scholars have been perhaps less inventive than possible. In pursuing my own solutions, I feel it is necessary to widen the textual critic’s traditional purview as far as possible, beyond matters of grammar and narrowly defined style, by integrating two alternative lines of attack. First, a proper appreciation of the structure of the exchanges between Orestes and Electra (and possibly, in the case of 671–93, the

3

For similar arguments, see Kovacs 1987a and Finglass 2009.

190

Redrawing the Lines

Old Man) requires that a critic be fully aware of the way in which their utterances relate to each other and how they function in their communicative context. In other words, a thorough linguistic analysis of Orestes and Electra’s interactions, using the approaches outlined in my Introduction, may reveal some indications for, and place limitations on, line-to-speaker attribution. Two understudied aspects of the scenes will prove of importance in this respect: first, the use of particles and, second, the degree of ‘contact’ between the two speakers (the terminology is that of Donald Mastronarde’s important Contact and Discontinuity, 1979). The second approach I will adopt in this chapter is to follow up on a suggestion made by Judith Mossman (2001: 379), who raised the possibility that lines such as 963, 965 and 966 ‘might be seen as linked to Electra’s gender’.4 In other words, I wish to examine the possibility of using linguistic gender specificity (see Introduction, §2.4.1) as an indicator of the likelihood that a particular line is spoken by a particular (male or female) character.

1.2. L and P As it is my aim to arrive at a text and line distribution ‘from scratch’, as it were, it is appropriate to print here the text of the two passages as found in manuscript L, as well as the Triclinian corrections of that manuscript and the variant readings found in its close relative P.5 In order to reflect the manuscript as closely as possible,6 I have given speaker indications only where the manuscripts have them, and such 4 Mossman mentions 963, 965 and 967, but by 967 she must mean 966 (she follows Diggle’s text, with a lacuna before 966, which would explain the slip). There are precursors to Mossman’s suggestion about 966 (e.g. Slings 1997c: 159), but it is significant that she reasserts it in an article specifically about gendered speech. 5 With Zuntz (1965), Basta Donzelli (1991b: 5 n. 2) accepts that P è verosimilmente apografo di L (‘P is probably an apograph of L’), copied after Triclinius did the first of his three rounds of corrections in L. Ioannes Katrares independently added P’s paragraphoi and notae. For a recent critique of Zuntz’s view, see Distilo 2012: xix–xlvi. In any case, the few deviations between the two MSS are not without significance for our present purposes. 6 I have been able to consult facsimiles of L and P and use the wealth of information found in the apparatus criticus accompanying Basta Donzelli’s (1995a) Teubner text and her article on line-to-speaker attribution (Basta Donzelli 1991a). With the exception of added adscript and subscript iotas and a few uncontested corrections (681 ἀνήλωσαν Barnes, 682 ὦ Reiske, 684 οἶδ᾽ Victorius; 959 ἔσω Brubach, 960 τε Reiske, 961 μὴ ᾽σίδῃ Scaliger, 966 γε Schaefer), I will discuss all changes to the MSS text in what follows.

Redrawing the Lines

191

indications are given precisely in the form found in the manuscripts, including the use of the paragraphos (—) for a speaker change without indication of the speaker’s name; ‘— P’ represents a paragraphos in P only. As we will see, punctuation in the manuscripts can provide some vital clues, so this, too, I have recreated as precisely as possible, including the use of the ‘full dot’ (τελεία στιγμή, ·) and ‘medial dot’ (μέση στιγμή, ·)—the punctuation marks used in the manuscripts to indicate fully and partly completed phrases, respectively.7 I have also used the lunate sigma ϲ (instead of internal -σ- and final -ς) as the manuscripts have it, but I have silently added and corrected accentuation in a few cases (except in 672, for which see note 8). A small critical apparatus given after each passage lists Triclinian corrections (Tr) and variants found in P. 671–85 Ὀρ: Ἠλ: Ὀρ: Ἠλ: Ὀρ:

Ἠλ:

ὦ Ζεῦ πατρῷε καὶ τροπαῖ᾽ ἐχθρῶν ἐμῶν οἰκτειρεθ᾽ ἡμᾶϲ· οἰκτρὰ γὰρ πεπόνθαμεν· οἴκτειρε δῆτα ϲούϲ γε φύνταϲ ἐκγόνουϲ· Ἥρα τε βωμῶν ἣ Μυκηναίων κρατεῖϲ· νίκην δὸϲ ἡμῖν εἰ δίκαι᾽ αἰτούμεθα· δὸϲ δῆτα πατρὸϲ τοῖϲδε τιμωρὸν δίκην· ϲὺ τ᾽ ὦ κάτω γῆϲ ἀνοϲίωϲ οἰκῶν πάτερ· καὶ γῆ τ᾽ ἄναϲϲα, χεῖραϲ ᾗ δίδωμ᾿ ἐμάϲ ἄμυν᾽ ἄμυνε τοῖϲδε φίλτατοιϲ τέκνοιϲ· νῦν πάντα νεκρὸν ἐλθὲ ϲύμμαχον λαβών, οἵπερ γε ϲὺν ϲοὶ Φρύγαϲ ἀνάλωϲαν δορί· ἤκουϲαϲ; ὡϲ δείν᾽ ἐξ ἐμῆϲ μητρὸϲ παθών· χὤϲοι ϲτυγοῦϲιν ἀνοϲίουϲ μιάϲτοραϲ πάντ᾽, οἶδεν· ἀκούει τάδε πατήρ· ϲτείχειν δ᾿ ἀκμή· καί ϲοι . . .

671 672 673 674 675 676 677 678 679 680 681 682 683 684 685

672: οἰκτείρεθ᾽ Tr P8 7 Editors and commentators generally pay insufficient attention to punctuation in the MSS in such cases. Punctuation has, of course, about as much evidential weight as speakers’ names (it will have been largely absent from texts in antiquity; for the use of punctuation in medieval MSS, see, e.g. Rijksbaron 2007: 68–71), but, where it is consistent with the presence or absence of speaker changes, it can at least show that what is at stake is not a mere accident by the copyist of ‘our’ manuscript but an intentional division of lines (which may, of course, go back to an earlier, flawed text). 8 L seems not to have had an accent mark on οἰκτειρεθ᾽ before Triclinius’ correction. The form as it stands in L after correction and in P is plural (οἰκτείρετ(ε), not οἴκτειρε τ᾽), possibly due to a mistaken reading of καί in the previous line (which of course links two titles of Zeus, not two different gods).

192

Redrawing the Lines

959–87 Ὀρ: Ἠλ: Ὀρ: — — — — — — — — — — — — — — — — (— P) (— P) — — — (— P)

εἶἑν· κομίζειν τοῦδε ϲῶμ᾽ εἴϲω χρεὼν ϲκότω γε δοῦναι δμῶεϲ ὡϲ ὅταν μόλη μήτηρ ϲφαγῆϲ πάροιθε μ᾽ εἰϲίδη νεκρόν· ἐπίϲχεϲ ἐμβάλωμεν εἰϲ ἄλλον λόγον· τί δ᾽ ἐκ Μυκηνῶν μῶν βοηδρόμουϲ ὁρῶ; οὔκ· ἀλλὰ τὴν τεκοῦϲαν ἥ μ᾽ ἐγείνατο· καλῶϲ ἄρ᾽ ἄρκυν ἐϲ μέϲην πορεύεται· καὶ μὴν ὄχοιϲ τὲ καὶ ϲτολῆ λαμπρύνεται· τί δῆτα δρῶμεν μητέρ᾽ ἦ φονεύϲομεν; μῶν ϲ᾽ οἶκτοϲ εἷλε μητρὸϲ ὡϲ εἶδεϲ δέμαϲ; φεῦ: πῶϲ γὰρ κτάνω νιν ἥ μ᾽ ἔθρεψε κἄτεκεν· ὥϲπερ πατέρα ϲὸν ἥδε κἀμὸν ὤλεϲεν· ὦ Φοῖβε, πολλήν γ᾽ ἀμαθίαν ἐθέϲπιϲαϲ ὅπου δ᾽ Ἀπόλλων ϲκαιὸϲ ἦ, τίνεϲ ϲοφοί· ὅϲτιϲ μ᾽ ἔχρηϲαϲ μητέρ᾽ ἣν οὐ χρῆν, κτανεῖν· βλάπτη δὲ δὴ τί πατρὶ τιμωρῶν ϲέθεν· μητροκτόνοϲ νῦν φεύξομαι τόθ᾽ ἁγνὸϲ ὤν· καὶ μήν ἀμύνων πατρὶ δυϲϲεβὴϲ ἔϲη· ἐγὼ δὲ μητρὸϲ9 τοῦ φόνου δώϲω δίκαϲ; τῷ; δ* πατρώαν διαμεθίηϲ τιμωρίαν· ἆρ᾽ αὔτ᾽ ἀλάϲτωρ εἶπ᾽ ἀπεικαϲθεὶϲ θεῶ ἱερὸν καθίζων τρίποδ᾽; ἐγὼ μὲν οὐ δοκῶ οὐδ᾽ ἂν πιθοίμην εὖ μεμαντεῦϲθαι τάδε· οὐ μὴ κακιϲθεὶϲ εἰϲ ἀνανδρίαν πέϲηϲ· ἀλλ᾽ εἰϲ τὸν αὐτὸν τῆδ᾽ ὑποϲτήϲω δόλον; ᾧ καὶ πόϲιν καθεῖλεν Αἴγιϲθον κτανών· ἔϲειμι· δεινοῦ δ᾽ ἄρχομαι προβλήματοϲ καὶ δεινὰ δράϲω· εἰ θεοῖϲ δοκεῖ τάδε ἔϲτω· πικρὸν δὲ καὶ ἡδὺ τἀγώνιϲμά μοι·

959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987

961: μὴ εἰϲίδη L (in the margin) 976: μή Tr γ᾽ 978: δὲ or δὴ, αὶ (in superscript), δαὶ P 979: ; (question mark) P 980: · (high dot) P 981: οὐκ ἂν P 984: καθεῖλεϲ Tr 986: δράϲω γ᾽ Tr P 987: δὲ χ᾿ ἡδὺ P

9 With Keene (app. crit.), I am not entirely convinced that L actually has μητρὸϲ (P certainly does) rather than μητρὶ. The difference between μητρὸϲ and μητρὶ in the ms. resides solely in the final letter (the rest of the ending stands in ligature), and in this case the curvature of that final letter could, to my eye, well indicate an iota (as, e.g. with πατρὶ in the line before) rather than a sigma (as, e.g. at 968). μητρὶ is also the reading of two later apographs of L (see Basta Donzelli 1991a: 26 n. 72, Curnis 2005); P has μητρὸϲ. Most collators (e.g. Murray, Basta Donzelli, Distilo) have read the word

Redrawing the Lines

193

2. DIVIDED (?) WE PRAY (671–84) As a rule, manuscript evidence for the attribution of lines to speakers becomes even less trustworthy when more than two speaking character share the stage at the same time. This is unsurprising: it is usually assumed that no indications of speakers’ names, at least, appeared in the ancestors of our manuscripts, even if speaker changes were marked.10 In the present scene, which features Electra, Orestes, and the Old Man on stage, we may therefore forgive editors for being, on the whole, rather intolerant of the division among speakers presented in L and P. But in this case editors have gone much further than simply switching names, as an overview of the most important editions will show: • Kirchhoff (followed closely by Nauck) first suggested that 671–83 should be divided as a stichomythia between Orestes and Electra (in addition, Kirchhoff and Nauck transposed 672–3 after 676: this is manifestly impossible, for the reasons given by Weil and Keene in their apparatus). • Kvičala subsequently proposed that the stichomythia should be spoken not by two, but by three characters, the Old Man coming in to speak each of the last lines of four successive triplets, so that the order in each triplet would be Orestes– Electra–Old Man. Kvičala is followed throughout by Diggle, by Parmentier and Grégoire, and by Cropp in his first edition (see Cropp 1988 on 671–84).11 The order of speakers within some of these triplets was further changed by Murray to Or.–El.–OM, El.–Or.–OM, Or.–El.–OM, Or.–El.–OM and by Denniston to Or.–El.–OM, El.–Or.–OM, Or.–El.–OM, El.–Or.–OM.

in L as μητρὸϲ, however, and Basta Donzelli calls the reading of the apographs un errore. I follow this consensus here; I will discuss the line more fully in §3.2. 10 The classic discussion of this problem remains Lowe 1962; other useful works include Olson 1996, 2001, and Finglass 2014. A good example of a text with speaker changes marked, but speaker names infrequently or not at all marked is the Bodmer papyrus of Menander: see, e.g. the editio princeps of Dyscolus by Martin (1958). 11 See also, e.g. Schadewaldt 1926: 103 n. 3, Zürcher 1947: 110 n. 7, 156 n. 11, Mastronarde 1979: 62 n. 33, and Slings 1997c: 153–4.

194

Redrawing the Lines

• Kvičala’s attribution of lines 673 and 676 to the Old Man has remained popular with editors, but several have since rejected the idea of stichomythia in these lines: following Di Benedetto (1961: 320–1), Basta Donzelli attributes 671–2 to Orestes, 673 to the Old Man; 674–5 to Electra, 676 to the Old Man; and 677–83 to Orestes (with 682–3 transposed).12 • Kovacs, having defended L’s original attributions in various articles (Kovacs 1984, 1987a), preserves them in his Loeb edition exactly as transmitted (except for the transposition of 682–3); so does Distilo.13 • Finally, Cropp, in his second edition (Cropp 2013), follows Di Benedetto in keeping 671–2 and 674–5 together (as lines spoken by Orestes and Electra, respectively), but accepts stichomythia in 677–84, which he divides into two triplets in the order Or.–El.–O.M., followed by one line (682, transposed with 683) given to Orestes and one (684) given to the Old Man. • Apart from the debate about whether or not 671–84 should be divided as a stichomythia, various scholars have followed Broadhead in transposing 693 (πάντ’ οἶδα . . . ) after 684 (by inserting a comma: πάντ’, οἶδα), typically also deleting 685–9 (with Nauck), and giving 684 to the Old Man (with Murray). For the reader’s ease, the suggested divisions are set out in Figure 4.1 (which leaves the text itself, for now, unaltered from L), except for the changes mentioned in note 6. Faced with this jarring number of possible divisions and attributions, it seems superfluous (but proves surprisingly necessary) to point out that the introduction of stichomythia in 671–84 is by no means an insignificant alteration of the transmitted text. However unreliable L’s speaker attributions are thought to be on the grounds that paragraphoi and speaker notations may have fallen out, it is not so much ‘improbable’ (Kovacs 1987a: 264) as impossible that

12 She explains her division in Basta Donzelli 1991a: 9–18; her discussion there overlaps to some extent with mine. 13 See also Stoessl 1956: 61–3.

Redrawing the Lines TEXT 671 ὦ Ζεῦ πατρῷε καὶ τροπαῖ᾽ ἐχθρῶν ἐμῶν

A Or

672 οἰκτείρεθ᾽ ἡμᾶϲ· οἰκτρὰ γὰρ πεπόνθαμεν·

195

B Or

C Or

D Or

E Or

F Or

674 El

El

El

El

El

673 οἴκτειρε δῆτα ϲούϲ γε φύνταϲ ἐκγόνουϲ·

El

674 Ἥρα τε βωμῶν ἣ Μυκηναίων κρατεῖϲ·

Or 676 El

Or

El

El

Or

672 Or

El

Or

Or

El

675 νίκην δὸϲ ἡμῖν εἰ δίκαι᾽ αἰτούμεθα·

G Or

H Or

675 OM OM OM OM OM OM Or El

El

676 δὸϲ δῆτα πατρὸϲ τοῖϲδε τιμωρὸν δίκην·

El

673 OM OM OM OM OM OM El

677 ϲὺ τ᾽ ὦ κάτω γῆϲ ἀνοϲίωϲ οἰκῶν πάτερ·

Or

Or

Or

Or

Or

Or

678 καὶ γῆ τ᾽ ἄναϲϲα, χεῖραϲ ᾗ δίδωμ᾿ ἐμάϲ

El

El

El

El

El

679 ἄμυν᾽ ἄμυνε τοῖϲδε φίλτατοιϲ τέκνοιϲ·

Or

OM OM OM OM

OM

680 νῦν πάντα νεκρὸν ἐλθὲ ϲύμμαχον λαβών,

El

Or

Or

El

Or

Or

681 οἵπερ γε ϲὺν ϲοὶ Φρύγανϲ ἀνήλωϲαν δορί·

Or

El

El

Or

El

El

Or

Or El

682 ἤκουϲαϲ; ὦ δείν᾽ ἐξ ἐμῆϲ μητρὸϲ παθών·

683

El OM OM OM OM 683 OM 683 683 683 683 683 683

683 χὤϲοι ϲτυγοῦϲιν ἀνοϲίουϲ μιάϲτοραϲ

682

Or Or Or Or Or 682 Or 682 682 682 682 682 682

684 πάντ᾽, οἶδ’, ἀκούει τάδε πατήρ· ϲτείχειν δ᾿ ἀκμή· ...

El

El

El

OM OM OM El

El

El

693 El

OM 693 El

A = MSS. LP; with 682–3 transposed: Stoessl 1956, Kovacs, Distilo; B = Kirchhoff, Nauck; C = Kvičala 1879; D = Murray; E = Denniston; F = Diggle; G = Di Benedetto 1961, Basta Donzelli; H = Cropp 2013

Figure 4.1. Proposed line-to-speaker attributions in Electra 671–84.

L’s text is the result of mere accident. For not only would this presuppose the unfortunate disappearance of eight paragraphoi or speaker notations within twelve lines; scholars would also have to account for the pattern of punctuation at the end of lines 671, 674 and 677, which in each case suggests that the same speaker will continue. The division in L and P is without question intentional, and we can only assume that it is in agreement with the tradition upon which L is based (or, possibly but implausibly, upon the tradition that the scribe of L started here himself). There is no way of knowing, of course, whether that tradition is itself the result of previous incorrect readings (perhaps going back to an ancestor

196

Redrawing the Lines

without any indications of speakers). Yet the manuscript evidence we have is unequivocal.14 Still, we may ignore L’s speaker notations for a moment and see whether the text itself provides any indications for a division. It has often been noted that the text requires a change of speaker in at least two places: lines 673, with the repetition of an imperative of οἰκτίρω (‘pity’) and particle δῆτα; and line 676, with the repetition of the imperative δός (‘give’) and again the particle δῆτα.15 There is no doubt that both of these lines are spoken by a new speaker, and L indeed has speaker changes there. The other place where a change of speaker has been seen as required is 672, if Fix’s conjecture οἴκτ(ε)ιρέ γ᾽ is accepted for the manuscripts’ οἰκτείρεθ᾽ (or, if we read L with a change of accent, οἴκτειρέ θ᾽). I hasten to point out that basing speaker arrangements on uncertain emendations is shaky textual criticism to begin with,16 but let us follow the argument. That the manuscripts’ οἰκτείρεθ᾽ cannot stand here is clear;17 we must then wonder, first, whether οἴκτιρέ γ᾽ is the best change and, second, whether that reading really requires a change of speaker. As to the first question, γ᾽ certainly is the easiest emendation (supposing the common confusion of uncial Γ and Τ, with τ subsequently corrected to θ), and other imaginable changes of just the one letter (οἴκτιρε δ᾽ ἡμᾶς18 or οἴκτιρε χἠμᾶς) are in no way better alternatives. The other change that has been suggested (by Dobree), to the aorist imperative οἴκτιρον, is palaeographically less convincing,19 and objections have rightly been raised to the change 14 It is thus questionable, to say the least, to append aesthetic judgements to the use of stichomythia by ‘Euripides’ in this passage, as Parmentier and Grégoire (1925: 187) do in the preface to their edition. The effect of the stichomythic composition here may be saisissant (‘thrilling’), but whose composition is it exactly? 15 For δῆτα, see Denniston 1954: 276–7. 16 Basta Donzelli (1991b: 10–11), who rejects the introduction of stichomythia, also rejects Fix’s emendation for that reason. 17 Pace Stoessl 1956: 61 and Weil ad loc.: both of them would connect τε with Ἥρα τε in 674. Denniston rightly rejects this. 18 This is what we find in the identical line at IA 985 (universally considered to be part of an interpolation): δέ in the present line does not work, however. 19 Pace Kovacs 1987a: 264, and see Slings 1997c: 153. It is true that (as Basta Donzelli points out, 1991b: 10) if οἰκτείρεθ᾽ (so accented) is a false ‘correction’ to the plural in light of Zeus’ two titles in 671, the palaeographical argument loses force. But I would suggest that the near repetition of the line at IA 985 (see previous note), whether it is the interpolation of an actor who appropriated this line or a case of recycling by Euripides himself, speaks as circumstantial evidence in favour of οἴκτιρε.

Redrawing the Lines

197

of aspect stem in the imperatives (λῆγε . . . λῆξον in Hipp. 473–4, mentioned by Kovacs, is not a good parallel, because there is a semantic difference between the two imperatives there).20 Therefore οἴκτιρέ γ᾽ seems to be, palaeographically, our best bet. The next question, however, is whether Fix’s reading really requires a change of speaker, as is nearly universally claimed. When reading οἴκτιρέ γ᾽, scholars see an example of the use of γε that is common when one speaker syntactically completes or modifies another speaker’s sentence (see Denniston 1954: 137–8). The most often cited parallel is the threefold use of the particle (once added by conjecture) in Aristophanes’ Peace, at 446, 449, and 452, one of which I cite here (Ar. Pax 444–6): ΕΡΜΗΣ ΤΡΥΓΑΙΟΣ He. Tr.

κεἴ τις ἐπιθυμῶν ταξιαρχεῖν σοὶ φθονεῖ εἰς φῶς ἀνελθεῖν, ὦ πότνι᾽, ἐν ταῖσιν μάχαις— πάσχοι γε τοιαῦθ᾽ οἷάπερ Κλεώνυμος. And if anyone with a desire to have a command begrudges you your coming to light again, my Lady, in his battles . . . may he suffer a similar fate to Cleonymus!

But the use of the particle here and in the other passages listed by Denniston under the same header deserves some further attention. I would suggest that the particle in such cases does not merely ‘mark the continuity of syntax across the change of speaker’ (Mastronarde 1979: 54),21 nor should it be described simply as a particle that lends ‘emphasis’ to the preceding word. (Denniston’s classification into ‘emphatic’ and ‘limitative’ is vague and the distinction he makes between these two kinds of use is somewhat artificial.) As Denniston (1954: 114–16) himself outlines, the basic function of γε throughout its (admittedly very wide) range of uses might be described as ‘concentration’: γε focuses attention upon a single element or phrase.22 Thus, 20 See Slings 1997c: 153, although I disagree with Slings’ explanation of that semantic difference. In my view, λῆγε κακῶν φρενῶν (‘leave off these evil thoughts’) should be read as a command to Phaedra to cease from her current, ongoing thoughts: λῆξον ὑβρίζουσα (‘leave off your pride’) is a command to refrain from committing ὕβρις altogether (the Nurse goes on to explain how the term ὕβρις applies to the situation, 474–5). See also Barrett ad loc. 21 See also Mastronarde 2002: 90–1. 22 For a more recent, synthetic description of γε as a ‘demarcating scope particle’, see Wakker 1994: 308–15. More work is needed on γε, ‘one of the subtlest and most elusive of particles’ (Denniston 1954: 116): progress may be expected from Bonifazi et al. in press.

198

Redrawing the Lines

in the example from Peace (and all others mentioned under the header), where speakers pick up their predecessors’ syntax and adopt it to their own purposes, γε helps to ‘refocus’ that syntax and steer it in a new and often unexpected direction. The particle itself, however, does not ‘mark’ a change of speaker as much as it is simply very suitable for such contexts (indeed there are many examples of speakers completing or amending each other’s syntax without γε). Coming now to El. 672, it is certainly not impossible that a similar use of γε is to be read here: a change of speaker cannot be ruled out. But I do not think that it is required, and several arguments weigh against it. First, this would be, as far as I can discover, the only place in Greek literature where γε is used in this fashion after only a form of address: that is to say, there is not really any preceding syntax to ‘redirect’ here.23 Second (although this is not a point I wish to press), this would be the only time when γε is so used with an imperative (the Aristophanic examples that come closest, such as the ones from Peace mentioned above, have cupitive optatives). Moreover, as Kovacs (1987a: 264) points out, there is no parallel in tragedy for γε being used to divide up a prayer among various speakers in this fashion;24 comic examples should not unthinkingly be accepted as sufficient parallels. Is there, then, a way to read οἴκτιρέ γ᾽ without resorting to a change of speaker? It is interesting to note that Fix himself seems to have thought so: he prints both 671 and 672 as spoken by Orestes, as in L. In support of his emendation, he cites in his apparatus criticus (p. lxix) a few other passages in Euripides (Supp. 842 εἰπέ γ᾽, IA 817 δρᾶ γ᾽, Ion 518 εὖ φρόνει γ᾽) and one from Sophocles (El. 411), all mentioned by Denniston (1954: 125–6) under a different header for

23

Neither Ion 561–2 (where Ion’s syntax is suspended from 559, across Xuthus’ intervening 560, to an exclamatory apostrophe address at 561—see Mastronarde 1979: 53—and where the subsequent phrase with γε is also not an imperative or an optative) nor Soph. OC 1108–9 (where Oedipus’ exclamation is modified by Antigone) are good parallels. 24 The divided prayer at Or. 1225–45, often cited in this context and no doubt to be seen as part of the ‘clear line of descent’ (Cropp on 671–84) that runs from the great antiphonal kommos at Aesch. Cho. 315–478 via the present Electra passage, is still not a good parallel: speakers there append their own directives to forms of address, and only in the second part (1235–40), where there are no forms of address, is the syntax suspended across various speakers (not distributed between them: each speaker utters complete phrases).

Redrawing the Lines

199

γε: its ‘emphatic’ use in commands and wishes.25 The parallel with Sophocles’ Electra 411 is particularly telling: ὦ θεοὶ πατρῷοι, συγγένεσθε γ᾽ ἀλλὰ νῦν. Paternal gods, now at last come to my aid (whatever else you do—γε).26

It is true that Diggle (1981b: 22) has cast doubts on most of the cases listed by Denniston for this use. But Diggle lets this passage (Soph. El. 411) stand,27 as well as several others, including a few where a change of speaker is not in question (Soph. El. 345 ἕλου γε, [Eur.] Rh. 623 πάρες γε). All in all, the matter is uncertain: none of the examples of imperative with γε is entirely above textual suspicion, but together they provide, perhaps, sufficient evidence to establish the usage. I would conclude, then, that we should probably accept οἴκτιρέ γ᾽ as a palaeographically plausible (and metrical) emendation. And, although γε certainly does not preclude the possibility of a change of speaker, I am not convinced that it requires one. Fix’s original fix, with the speakers maintained as in L, deserves full consideration, and I believe it is at least possible to read and translate the first two lines as follows:28 ὦ Ζεῦ πατρῷε καὶ τροπαῖ᾽ ἐχθρῶν ἐμῶν, οἴκτιρέ γ᾽ ἡμᾶς. οἰκτρὰ γὰρ πεπόνθαμεν. Zeus, god of my fathers and router of my enemies, pity us (if nothing else). For our sufferings are pitiable.

Finally, I should touch on the two other lines where the grammar has been thought to require a speaker change: 980 (because of the asyndeton) and 981 (because of the collocation οἵπερ γε).29 As for the asyndeton in 980, there are numerous examples in tragedy of asyndetically connected sentences starting with νῦν (‘now’; cf. e.g. in Euripides According to Denniston (1954: 125), γε in this use ‘sharpens the tone of an imperative’. 26 The gloss ‘whatever else’ or ‘if nothing else’ for γε, which I have borrowed from Rijksbaron 2007 (see his Index of Greek Words s.v. γε), is often appropriate, and a helpful way of thinking about the particle. 27 It should be added that Diggle feels that γε is possible at Soph. El. 411 only because ἀλλὰ νῦν ‘makes all the difference’ (1981: 22); I am not sure why he thinks this is so. 28 This is also the solution adopted by Cropp in the revised edition of his commentary (on 672), where he calls γε ‘intensive’ and remarks that ‘it need not imply a change of speaker’. 29 See Kovacs 1987a: 264 n. 9 and Slings 1997c: 153–4. 25

200

Redrawing the Lines

Phoen. 1252–3, El. 867, Tro. 761, Hel. 722). I wonder whether we can really speak of full asyndeton in these cases, and at any rate νῦν is a common word at the start of prayers and exclamations (cf. e.g. Cyc. 351, Med. 767, Heracl. 867), which have a tendency to be less integrated into the flow of a discourse. As for the collocation ὅσπερ γε, this is the only place in tragedy where it occurs,30 but relative clauses more generally can include γε (see Denniston 1954: 123–4, with several Euripidean examples). Aristophanes has two instances of the collocation, one with speaker change (Vesp. 146) and one without (Eccl. 275). None of the five extant Platonic instances suggests that a change of speaker is required. In short, οἵπερ γε at 681 proves very little. Up to this point, I have argued that a change of speaker is required only at 673 and 676, that the matter is inconclusive at 672, and that no such change is required at 980 or 981. In short, nothing thus far requires us to introduce stichomythia, and the bulk of palaeographical and stylistic argument weighs against it. I would add one final point here. As Mastronarde (1979: 73) has exhaustively shown, ‘truly incomplete utterances . . . are exceedingly rare in Greek tragedy’. When one speaker is interrupted by another, the latter will as a rule either complete the former’s syntax or the former will continue his or her own syntax after the interruption (this is what Mastronarde calls ‘suspended’ syntax; see §3.2 in this chapter on El. 971–3). If we suppose that this is an instance of the first option—(cooperative) completion of syntax by the interrupting speaker—it would be, as already mentioned, an unparalleled example of a prayer split among characters such that one provides only the address and one or two others step in to utter the actual request(s). As Kovacs (1987a: 264) points out, matters would become even stranger at 677–9, where two speakers (Orestes, Electra) would provide separate addresses and the third (the Old Man) the actual prayer. I suspect that it is for this reason that Mastronarde (1979: 62 n. 33) prefers to see this passage as an example of suspended syntax (similar to Or. 1235–7), so that ‘Orestes’ prayer would consist of 671 + 674 + 677 + 680’. But this reading would have Orestes address, consecutively, Zeus, Hera, and Agamemnon before he directs his request only to Agamemnon (ἐλθέ 680, ‘come’). And to whom are Electra and the Old Man addressing their prayers in 672/675 and 673/676, respectively? And, again, what

30

Eur. Cyc. 692 is usually emended to ὅπερ μ᾽.

Redrawing the Lines

201

is the point, at 680–1 and 683 (adopting the transposition of 682–3), of having first Electra and then the Old Man elaborate on Orestes’ πάντα νεκρόν by contributing a relative clause? All this is a dead end. All in all, no matter how we try to construe the syntax, the introduction of stichomythia in these lines is not in line with tragic practice. The only manuscript evidence we possess speaks strongly against it, and I do not think that any scholar has presented a sufficient case to overturn that evidence. Editors have rendered Euripides needlessly artificial in this respect. Now, if we keep speaker changes only where L has them (and these are also the only places where the text requires them), the question remains which speaker utters which line. In this matter, L’s evidence carries virtually no weight, as speaker names were probably not included in its ancestors even if speaker changes were. Thus as a final step towards an attribution of the lines, we may wonder whether the language or the content (if not both) at any place suggests that the line in question should be attributed to a particular speaker. Several points may be noted here: • Considering that in Euripides, ‘male characters usually swear by male deities, and females by female deities’ (McClure 1995: 49), there is a case for giving 671–2 to Orestes and 674–5 to Electra, as proposed by Reiske. Indeed, Zeus is addressed much more often by men in Euripides (and in tragedy in general);31 together with the ‘masculine’ epithet τροπαῖ᾽ ἐχθρῶν ἐμῶν (‘router of my enemies’), the address to Zeus strongly points towards attribution to Orestes (who at 1177 will lament his fate to Zeus and Ga, cf. 679). Hera is very infrequently addressed by name in tragedy,32 so statistics will not help here. It is, however, worth pointing out that the goddess has appeared before in the play

31 The percentages for the form Ζεῦ in Euripides are 64% male, 36% female (when choruses are discounted, this becomes a more pronounced 70–30% split). Women speak roughly 48% of all lines in Euripides (39% if choruses are excluded); see McClure 1995: 40. It is important to add the proviso that ‘it is questionable whether a statistical analysis of the language of Tragedy can yield any definite conclusions’ (McClure 1995: 41). 32 There appear to be only three other instances in all of tragedy: Aesch. Sept. 152 (Chorus), Eur. Hel. 1094 (Helen), and Eur. Phoen. 1365 (Polynices).

202

Redrawing the Lines

(173), in relation to Electra.33 Hera may have been chosen here for that reason.34 • An additional argument for giving at least 675 to Electra is the use of the word νίκην (‘victory’). Especially in the second half of the play, the language of victory is typical of Electra: she uses νικ‐ words at 697, 759, 872, 880 (twice) and 955, whereas Orestes uses none in the rest of the play (although he does speak of getting a ‘crown’ at 614). It seems congruous with Electra’s character to pray for victory here. Similarly, Electra appears to me more suited to add the confident εἰ δικαί᾽ αἰτούμεθα (‘if our request is just’) to her prayer—‘the if-clause rhetorically justifies the demand for victory’ (Cropp ad loc.):35 she, more than anyone, is convinced of the justice of her own actions. • Several scholars have pointed out that the word τοῖσδε in 676 cannot be equivalent to ἡμῖν (there is no example of the plural demonstrative pronoun used by itself, standing for the first person personal pronoun). If this word is Electra’s, as in L, it must refer to Orestes and the Old Man (or, an option that goes overlooked by scholars, to Orestes and Pylades, who is certainly on stage with Orestes at this moment: see 548–50). But it would be strange for Electra not to include herself here, especially in light of the expression πατρὸς . . . τιμωρὸν δίκην (‘in vengeance for our/their father’). Thus the Old Man, who is not given any of the lines 671–84 in L, is the ideal candidate to deliver 676. By extension, especially if 671–2 are given to Orestes and 674–5 to Electra, the Old Man is also the most likely candidate to be the speaker in 673.36 • If, then, we do not introduce a further speaker change until 684, 677–83 must be delivered by Orestes (as the lines from 685 on are unquestionably Electra’s). There is no problem with ascribing

33 See Zeitlin 1970 and Ormand 2009: 268–9 for the importance of Hera’s festival in the Electra. Without discussion, Zeitlin (1970: 645) attributes 674 to Electra. 34 See Allan on Hel. 1093–106 for the ‘tailoring’ of invocations to the specific context by characters in Euripides. 35 Invocations often include such ‘pseudo‐conditionals’, cf. e.g. Aesch. Ag. 520 and Soph. OT 164–7 with Wakker 1994: 190–1. 36 This seems called for by considerations of symmetry, although, as Kovacs (1987a: 263) rightly points out, such considerations ‘prove very little except in literary forms, like the stasimon, that are inherently symmetrical’.

Redrawing the Lines

203

τοῖσδε φίλτατοις τέκνοις at 679 to Orestes to refer to himself and Electra: Basta Donzelli (app. crit.) provides convincing parallels (Aesch. Cho. 501, Soph. OT 41). • Line 684 could possibly belong to the Old Man, but there is no stringent need to give the line to a ‘third’ character merely in order to maintain a parallel between στείχειν δ’ ἀκμή (‘it is high time to go’) and ἔρδοις ἂν ἤδη (‘now do the deed’, Aesch. Cho. 510–13: Chorus), ἀπηλλάχθαι δ᾽ ἀκμή (‘it is high time to depart’, Soph. El. 1338: Tutor), or πρὸς ἔργον ἐξορμώμεθα (‘let us set off towards the deed’, Eur. Or. 1240: Pylades), as has been suggested by various scholars. In this play, Electra is constantly the one to urge to action (cf. especially 275 and 668, and see §3 on 959–87). Moreover, the phrase καί σοι, which starts the next line (685), is inappropriate for a new speaker (καί cannot mean ‘also’ here) and indeed is never so used in tragedy.37 I believe, then, that the text should be printed as follows: ΟΡ. ΠΡ. ΗΛ. ΠΡ. ΟΡ.

ΗΛ. Or.

ὦ Ζεῦ πατρῷε καὶ τροπαῖ᾽ ἐχθρῶν ἐμῶν, οἴκτιρέ γ᾽ ἡμᾶς. οἰκτρὰ γὰρ πεπόνθαμεν. οἴκτιρε δῆτα σούς γε φύντας ἐκγόνους. Ἥρα τε βωμῶν ἣ Μυκηναίων κρατεῖς, νίκην δὸς ἡμῖν, εἰ δίκαι᾽ αἰτούμεθα. δὸς δῆτα πατρὸς τοῖσδε τιμωρὸν δίκην. σὺ τ᾽ ὦ κάτω γῆς ἀνοσίως οἰκῶν πάτερ, καὶ Γαῖ᾽ ἄνασσα, χεῖρας ᾗ δίδωμ᾿ ἐμάς, ἄμυν᾽ ἄμυνε τοῖσδε φιλτάτοις τέκνοις. νῦν πάντα νεκρὸν ἐλθὲ σύμμαχον λαβών, οἵπερ γε σὺν σοὶ Φρύγας ἀνήλωσαν δορί, χὤσοι στυγοῦσιν ἀνοσίους μιάστορας. ἤκουσας, ὦ δείν᾽ ἐξ ἐμῆς μητρὸς παθών; πάντ᾽, οἶδ᾽, ἀκούει τάδε πατήρ· στείχειν δ᾿ ἀκμή. καί σοι πρωφωνῶ . . .

671 672 673 674 675 676 677 678 679 680 681 683 682 684 685

Zeus Paternal, Router of my enemies, have pity on us. For pitiful are the things we have endured.

37 It is true, to complicate matters further, that several scholars would insert 693 (to be spoken by Electra) between 684 (giving to the Old Man) and 685, which would negate this problem. Moreover, many would delete 685–9 (lines which are, admittedly, not free of problems: for discussion see Cropp, Distilo ad loc.). But the transposition of 693 is not easily explained, and the case against 685–9 not strong enough, to my mind, to justify deletion. I would prefer, with Basta Donzelli (1991a: 12–18), to keep these lines in place (with 693 spoken by Orestes and Electra in antilabe, for which see Kovacs 1984: 239).

204 OM El. OM Or.

El.

Redrawing the Lines Have pity on them, indeed; they are your descendants. And Hera, you who govern the Mycenaean altars, grant us victory, if what we ask for is just. Grant them, indeed, justice in requital for their father. And you, father, who dwell below the earth, and mistress earth, on whom I strike my hands, defend, defend us, your dearest children. Come now, bringing as your allies all the dead who together with you conquered the Phrygians by the spear, and all who abhor impious desecrators. Have you heard, you who suffered so at my mother’s hands? He hears it all, I know it. But it is time to go. And to you I say . . .

This text, close to the manuscript evidence as regards the moments of speaker change, presents us with an interesting division of labour between the siblings. The ‘perfunctory fashion’ that Cropp detected in the invocations (on 671–84) is, in this configuration, first and foremost Electra’s. The scene shows her ‘attention to religious concerns to be cursory by contrast’—not only with the parallel scene in Aeschylus’ Choephori, but even with Orestes: after two brief prayers comes Orestes’ more extended invocation of the dead Agamemnon, but upon its conclusion, rather than joining in calling upon Agamemnon (as might be expected from the literary tradition), Electra caps off the prayers and pushes for action (going on, in 685–98, to threaten suicide). Were she given, say, 672 or parts of 677–84, her role in uttering the prayers would be more extensive. As it is, her one contribution to the series of invocations is an almost haughty call on Hera to grant the siblings victory. This Electra, then, is someone who is completely intent upon the destruction of her mother and Aegisthus. As we shall see presently, this is not the last time in the play she displays these characteristics.

3. THE HESITATION SCENE (959–87)

3.1. On giving orders and having fashion sense (959–66) Ever since Camper first proposed reversing the transmitted distribution of lines 959–65 (959–61, 963, 965 El.; 962, 964 Or.; Camper gave 966 to the Chorus), as many defenders of L’s distribution have sprung up as have supporters of Camper’s changes (though no one follows him in the attribution of 966). Again, I sketch the lay of the land:

Redrawing the Lines

205

• Murray and Denniston both follow Camper, but give 966 to Electra, accepting a break in the stichomythia at 965–6.38 • Kirchhoff, in the second edition of his text, followed Camper up to 965 as well, but with a lacuna after 965, to keep the stichomythia intact (in this he was followed by Diggle; Nauck posited the lacuna after 966).39 • In his first edition, Kirchhoff had kept the transmitted divisions of L up to 964, but then reversed the order of 965 (spoken by El.) and 966 (spoken by Or.). This solution was adopted by Weil, Wecklein, Parmentier and Grégoire, Keene, and Kovacs.40 • L’s divisions are left unaltered by Basta Donzelli in her Teubner edition.41 Again, a table may help us to keep track of the different options; these are laid out in Figure 4.2 (again, no changes to L’s text have been made in this figure, apart from the minor ones listed in note 6). The reasons for departing from the transmitted line-to-speaker attributions are summarized by Denniston as follows (on 959–66): 959–61 must be spoken by Electra. For, though Orestes had arrived in Argos intending to kill his mother, his resolution now falters as the execution of the task draws near. He could not use so brutally unequivocal a phrase as σφαγῆς πάροιθε at 961, and then, immediately afterwards, at 967, 969 express his reluctance. Conversely, the horrified dwelling on the maternal tie at 964 suits Orestes, not Electra.

But Kovacs objects (mostly repeating points that had been raised before by others): There is no need to assume that Orestes cannot speak the words σφαγῆς πάροιθε (961) because his nerve is already failing. On the contrary, it is only the sight of his mother later that unnerves him (968–69: γάρ = yes, for). Nor is 964 a ‘horrified dwelling on the maternal tie’ (Denniston)

38

See also, e.g. Mastronarde 1979: 107 n. 41 and Bain 1981b: 13 n. 10, 1982: 273. Nauck’s suggestion was taken up by Cropp (1988 ad loc.), who subsequently (1996) abandoned it in favour of Diggle’s text, before taking it up again in his revised commentary (2013 ad loc.). 40 See also, e.g. Schwinge 1968: 85–9; Kovacs 1984: 239; Lloyd 1986: 17. Wilamowitz 1875: 68 gives 959–61 to Electra, then posits a lacuna, keeps 962–4 as in L, and reverses 965–6 (a palaeographical mess if ever there was one). 41 See also, e.g. Stoessl 1956: 66–8; Steidle 1968: 75–6; Basta Donzelli 1978: 159–60, 1991b: 18–20; Slings 1997c: 158–9; Burnett 1998: 237 n. 48. 39

206

Redrawing the Lines

TEXT 959 εἶἑν· κομίζειν τοῦδε ϲῶμ᾽ ἔϲω χρεὼν

A Or

B El

C El

D El

E Or

F El

G El

962 ἐπίϲχεϲ ἐμβάλωμεν εἰς ἄλλον λόγον·

El

Or

Or

Or

El

El

Or

963 τί δ᾽ ἐκ Μυκηνῶν μῶν βοηδρόμουϲ ὁρῶ;

Or

El

El

El (ὁρᾷϲ)

Or

Or

El (ὁρᾷϲ)

964 οὔκ· ἀλλὰ τὴν τεκοῦϲαν ἥ μ᾽ ἐγείνατο·

El

Or

Or

Or

El

El

Or

965 καλῶϲ ἄρ᾽ ἄρκυν ἐϲ μέϲην πορεύεται·

Or

El

El

960 ϲκότῳ τε δοῦναι δμῶεϲ ὡς ὅταν μόλη 961 μήτηρ ϲφαγῆϲ πάροιθε μὴ ’ϲίδῃ νεκρόν·

El

966 Or 966 Or

El

966 καὶ μὴν ὄχοιϲ γε καὶ ϲτολῇ λαμπρύνεται·

El

Chor

El

965 El 965 El

Or

967 τί δῆτα δρῶμεν . . .

Or

El

Or

Or

Or

Or

Or

A = MSS. LP, Basta Donzelli, Distilo B = Camper; C = Murray, Denniston; D = Kirchhoff 1867, Diggle; E = Kirchhoff 1855, Weil, Parmentier-Grégoire, Kovacs; F = Wilamowitz 1875; G = Cropp 2013

Figure 4.2. Proposed line-to-speaker attributions in Electra 959–66. but rather typical stichomythic padding. Against change: first, Orestes is the natural one to give orders to his slaves. His εἶεν, likewise, is a proper resumption of the business at hand after the harangue of Electra, for which he himself had given permission. Second, only Electra can identify Clytaemnestra in 964 since Orestes has not seen her since he was a baby. (Kovacs 1984: 239)

In weighing these arguments, let us set out from lines 959–61, beginning with the first word, εἶἑν. The point Kovacs makes about this interjection42 is not without merit: it is usually described as a word marking a shift ‘to the next point or the next step’ (Barrett on Hipp. 297; see LSJ s.v. εἶἑν) or a ‘new stage in the argument’ (Parker on Alc. 299). Such definitions could benefit from some greater precision: εἶἑν may be said to mark the transition to a new conversational move, 42 The word is hard to classify, but this problem resides more in established systems of classification of word types than in any inherent property of εἶεν. For problems of classifying interjections, see Biraud 2010 and Nordgren 2015. Nordgren’s (2015: 180–5) treatment of the εἶἑν is helpful, but understates the extent to which it has a transitional function (it does not merely indicate ‘compliance’ with what precedes, but also prepares for what follows, as my discussion bears out).

Redrawing the Lines

207

but more specifically it signals that all the preconditions for executing that conversational move have been met: permission to make a speech has been given (e.g. the first word of Electra’s invective against Aegisthus at 907), supporting arguments have been provided (e.g. Odysseus’ use at Hec. 313), the appropriate preparations for an action have been made (e.g. Megara’s coming back on stage after dressing herself and her children in funeral attire, Her. 451), and so on. Thus Kovacs is right to point to the resumptive function of the particle here: it does indeed signal the return to the business of dealing with Aegisthus’ corpse after concluding Electra’s speech (‘now that is out of the way, let’s go inside’). Unfortunately, this yields no conclusive answer to our problem: εἶἑν could be a sign that Orestes is getting back to the matter at hand, but it could also be interpreted as an indication from Electra that she is ready to move on. It is worth noting, however, that in the only two other instances where the particle is used immediately after a choral comment, it is a different speaker who takes over (IT 342: Herdsman–Chorus–Iphigenia; Hel. 761: Servant–Chorus–Helen), which would suggest Orestes here. Yet my present focus is on Kovacs’ other point, that it is more appropriate for Orestes to give orders to his own slaves (that they are his is certain: δμῶες is always masculine in Greek tragedy, see Bain 1981b: 42–3 n. 25). Steidle takes this argument one step further, adding a gender-related point: It is entirely inappropriate, not to say impossible, that at 959–61 Electra gives a command in the presence of two men, Orestes and Pylades, especially since it concerns a servant of Orestes. (Steidle 1968: 75, my emphasis)43

A twofold question thus arises: first, is it unlikely for characters in Greek tragedy to give commands to slaves who are not their own? Second, is this the case for women when (free) men are present? (On the issue of female speech in mixed groups*, see Introduction, §2.4.1.) As for the first question, it turns out that, although there are a few cases of characters giving commands to the slaves of others, such orders always have a dramatic significance and are by no means ‘normal’.44 43 Zunächst ist es durchaus unpassend, ja unmöglich, daß in den V.959/61 Elektra in Gegenwart zweier Männer, des Orest und des Pylades, einen Befehl gibt, zumal es sich um Diener Orests handelt. 44 In fact we do not have to look very far for an example. At El. 360, the Peasant orders Orestes’ retinue to take up the baggage and go inside. It is striking, however, that the order is not executed until Orestes confirms it (393; see ch. III, §2.2.3, where

208

Redrawing the Lines

There seems to be no clear-cut parallel where such an order is simply carried out without further ado.45 But the problems for those wishing to ascribe the lines to Electra become even greater when Steidle’s gender-based point is taken into consideration: from what we can surmise from the rest of extant tragedy, it would be unusual indeed for a woman to give orders to a man’s slaves in the presence of their master. Apart from the present instance, there are 36 commands in Greek tragedy to δμῶες/δμωαί, ὀπαδοί, ὁπάονες, πρόσπολοι, ἀμφίπολοι, and λάτρεις (singular and plural, masculine and feminine), of which eight are given by women.46 None of these eight are addressed to slaves other than the woman’s own, and only one is spoken in the presence of a free man (unsurprisingly, the command in question is given by Aeschylus’ Clytemnestra to her δμῳαί at Aesch. Ag. 908, with Agamemnon on stage).47 Of course, even if Electra’s giving a command is irregular, we could try to link this ‘insolence’ to the characterization of a woman described by some as a ‘sharp‐tongued virago’ (Conacher 1967: 205) or a ‘monomaniacal harridan’ (Luschnig 2013: 381); but should not Orestes’ servants, then, be at least as hesitant to carry out these orders as when they were given commands by the Peasant?48

the authenticity of the line is also discussed, n. 40). In the messenger speech, Orestes dissuades his ‘father’s former servants’ (πατρὸς παλαιοὶ δμῶες, 851) from attacking him after he has slain their ‘new’ master Aegisthus: but his regained authority to command them is exactly what he is asserting there. Another set of examples may be found at Andr. 577 and 715–16: Peleus here twice orders Menelaus’ attendants to release Andromache (successfully on the second attempt, when Menelaus is not there to countermand his orders)—but these commands too are not quite ordinary, in fact they are rather like threats. At Heracl. 1050–3 both Alcmene and the Chorus give orders to attendants who are ‘presumably servants of Hyllus’ (Allan ad loc.), who is not present; it would seem that both have the authority to do so. I have left out of consideration instances of a character giving orders to have gates or doors opened (gates: e.g. Phoen. 1067–8, Bacch. 170–4; doors: e.g. Med. 1314, Hipp. 808, IA 1340). 45 Mastronarde (1979: 107 n. 41) and Bain (1981b: 13 n. 10) are surely right to assume that the servants go on into the house even though the person leading them is stopped by ἐπίσχες (the metrically identical plural imperative ἐπίσχετ’ would have been available, cf. Hipp. 567, Hel. 1184). 46 Totalling 22%. The percentage for Euripides (6 out of 27) is identical (22%; the percentage drops to 16% if choruses are excluded). As noted earlier (n. 31), women speak 48% of all lines in Euripides (39% if choruses are excluded). 47 At Heracl. 1050 (Alcmene to δμῶες, discussed in n. 44), a captive Eurystheus is on stage. 48 That is, of course, assuming that El. 360 is kept. On the behaviour of the servants, see n. 44.

Redrawing the Lines

209

On balance, it appears to me that the arguments against attributing 959–61 to Electra are stronger than Denniston’s objections to the ‘unequivocal brutality’ of σφαγῆς πάροιθε. Moreover, as has been frequently pointed out,49 Orestes’ nerve to kill his mother has been holding up quite well thus far (cf. 89, 276–82, 582, 589–600, 613–14, 646–8): if anything, he seems to have taken the killing for granted—a token of his lack of foresight that will contrast starkly with his realization at the end of the play.50 In fact the pragmatism of 959–61 is congruous with the character of Orestes developed throughout the play thus far: he should be allowed to keep the lines. ἐσβάλωμεν εἰς lines 962–4.51 If our determination here were to be made only on the basis of what is dramatically realistic, it is hard to see how anyone could argue with Wilamowitz’s analysis (in which he takes Nauck to task): He [Nauck], after all, makes Orestes the one who announces the arrival of Clytemnestra, and Electra the one who wonders whether they are Argive troops coming to support Aegisthus. This is, in itself, unsatisfactory (Orestes cannot be the one to recognize his mother, it has to be Electra), and is entirely disproven by ὁρῶ. Therefore the manuscript attributes 962–4 to the speakers correctly.52

Having Orestes recognize his mother at 964 necessitates that Electra is unable to recognize her at 963 (unless one insists on changing ὁρῶ to ὁρᾷς and imagines Electra as having poor eyesight or her back turned).53 If Orestes delivers line 963, his mistaking the approaching band for reinforcements is, dramatically, quite acceptable.

49

See, e.g. Basta Donzelli 1991a: 19 (with further references). So too Cropp 2013: 5. As for Orestes’ σφαγῆς (‘slaying’), rather than overly psychologizing it, we might interpret it as a continuation of the sacrifice motif that pervades this part of the play (see Introduction, §4.2). 51 If Denniston ad loc. is right to suspect ἐμβάλωμεν (962), Diggle’s (1994: 162–3) ἐσβάλωμεν is certainly the most attractive change (the parallel at Supp. 92, καινὰς ἐσβολὰς . . . λόγων, ‘[things that require] a change of topic’ is convincing). 52 is [Nauck] enim Orestam facit Clytaemnestrae advenientis nuntium, Electram utrum mater sit an Argivi Aegistho opitulantes dubiam. quod et per se parum placet, (matrem enim Orestes novisse nequit, debet Electra), et per ὁρῶ omnino refellitur. itaque codex 962–64 personas recte disponit (Wilamowitz-Moellendorf 1875: 68). 53 The change to ὁρᾷς seems necessary only if Electra speaks this line (in itself this is not a great boon to the proponents of reversing the speakers), though Denniston ad loc. attempts to explain how Electra could speak the line with ὁρῶ. If Orestes speaks the line, ὁρῶ will do quite nicely here (pace Cropp ad loc. and 1996): there are perfectly good parallels for questions with μῶν plus a first-person verb form (the best is 50

210

Redrawing the Lines

But there are reasons other than mere dramatic realism to retain L’s distribution here. Steidle (1968: 75) argues that the question about Mycenaean reinforcements is much more appropriate for Orestes than it is for Electra. This point deserves consideration, and we should again wonder whether gender can be used as an argument here—that is, whether or not the line is more likely to be spoken by a man because it deals with military matters; alternatively, we might detect, with Mossman (2001: 379), a degree of typically female ‘alarmism’. It is impossible to make a ‘statistical’ case here, for example on the basis of lexical selection.54 Yet I would argue that the question in 963 does square nicely with Orestes’ portrayal earlier in the play, where he has repeatedly shown concern for discreet strategy and for remaining undetected.55 The tenor of this line is thus not congruous with ‘male’ characters per se, but very appropriate for this particular man. As for 964, Denniston’s view that the line expresses ‘horrified dwelling on the maternal tie’ should perhaps not be dismissed as quickly as some have thought. To be sure, ‘it is dangerous to read psychological subtleties into pleonasms in stichomythia’ (Lloyd 1986: 17), but to condemn ἥ μ᾽ ἐγείνατο as mere ‘stichomythic padding’ (Kovacs) takes things too far (after all, Euripides could have ‘padded’ the line with anything).56 There are, however, other interpretations that one could assign to the ‘dwelling on the maternal tie’ than Denniston’s reading of pathos. Commentators have rightly pointed to the parallel at Sophocles, Electra 261, τὰ μητρὸς, ἥ μ᾽ ἐγείνατο, | ἔχθιστα συμβέβηκεν (‘the affairs of my mother, who bore me, are Andr. 896, μῶν ἐσφάλμεθ’ ἢ σαφῶς ὁρῶ δόμων ἄνασσαν τῶνδε Μενέλεω κόρην; ‘am I mistaken or do I see, clearly, the daughter of Menelaus, mistress of this house?’). Cropp (on 959–66) suggests that changing the speakers is necessitated by the nature of the extended exchange between an ‘observer’ and a ‘reactor’, but this is circular: that relationship exists only by virtue of ὁρᾷς (which is only needed when the speakers are reversed). 54 Forms of βοηδρόμος (‘reinforcement’) and the verb βοηδρομέω (‘come to support’) occur in tragedy only at Aesch. fr. 46c and in Euripides (particularly in his Orestes): they are used by female as well as male speakers (cf. e.g. Electra’s not dissimilar use at Or. 1290, and Hipp. 776, Phoen. 1432). 55 Cf. 95–7, 616, and his behaviour throughout the messenger speech, where he is careful to move closer and closer to Aegisthus while avoiding his armed guards (see ch. V). 56 Moreover, the exact same phrase is used outside of the stichomythia as well, e.g. Phoen. 996, πατρίδος ἥ μ᾽ ἐγείνατο. Are we to assume that it has ‘meaning’ there, but not in the stichomythia?

Redrawing the Lines

211

most hateful to me’, Electra speaking), where the phrase ἥ μ᾽ ἐγείνατο ‘sharpens the paradox of the following ἔχθιστα’ (Finglass ad loc.): a similar irony might be read into our line.57 But perhaps we need not read more into the fullness of expression than a way for Electra to make perfectly clear whom she sees approaching.58 On balance, dramatic realism would seem to require that 962–4 stand as they are in L; and, even if we ignore dramatic realism, the objections raised against the manuscript distribution do not appear to me strong enough to outweigh the arguments in favour of it. The lines as they stand in L are easily reconcilable with the characters of Electra and Orestes as we have seen them throughout the play: an Orestes who is prudent and concerned with strategy, an Electra whose troubled relationship with her mother bubbles to the surface easily. We may now turn to the most controversial pair of lines in the stichomythia: 965–6. As noted above, a vast array of different solutions has been proposed here. However, no matter what scholars do with 966 (whether they have it spoken by Electra, by means of a lacuna or break in the stichomythia, or give it to Orestes), most of them agree that 965 must be Electra’s.59 I will discuss the issues surrounding 965 after focusing first on line 966. Who speaks 966? Most of those who attribute the line to Orestes generally tend to do so only in order to avoid giving him 965. But Schwinge (who endorses Kirchhoff ’s transposition of 965–6) has given an intensely psychological reading of the scene: Orestes finally recognizes his mother by her pomp (to which he was alerted by Electra, 314–18),60 and the sight of her awakens in him the doubt that will plague him for the rest of the scene. Electra puts in a typically cruel comment (965), but Orestes has begun to question the mission. 57 See Wecklein ad loc., Stoessl 1956: 67, Lloyd 1986: 18, Basta Donzelli 1991a: 19. Finglass’s description of the words’ effect in Sophocles is particularly apt because it does not suggest too much about Electra ‘herself ’ as being contemptuously ironical (one of those dangerous psychological subtleties). 58 There is no way, I think, to determine whether or not we should read a comma between τὴν τεκοῦσαν and ἥ μ᾽ ἐγείνατο (the difference between digressive ‘my mother, who bore me’ and restrictive ‘the mother who bore me’—the former perhaps slightly more poignant, the latter more appropriate in the context of one person identifying another). 59 For the exceptions, see Figure 4.2. 60 This idea applies well also if Orestes speaks 962 and 964: he has recognized his mother by her appearance. So Cropp ad loc., maintaining the ‘observer’ and ‘reactor’ distinction he emphasizes (see n. 53).

212

Redrawing the Lines

His sister is surprised by this but realizes that the actual sight of Clytemnestra has caused Orestes to doubt (ὡς εἶδες δέμας 968, ‘now that you have seen her in person’). This is a perfectly coherent reading, and the strongest case to make for the transposition of 965–6. But a different reading is possible, and there are reasons to believe that 966 should be Electra’s. Again, her gender comes into play: it is Electra who comments here for the first time on her mother’s splendid turnout . . . [this] might be seen as linked to Electra’s gender. (Mossman 2001: 379) 966 is a futile comment for Orestes, but highly to the point for Electra, who is a female and a poor female. (Slings 1997c: 152)

It would be, again, quite difficult to argue for a gender preference here on strictly linguistic grounds: the statistics of the ‘language of dress’ are certainly no help.61 But there may be a case to make for taking the line as part of a specifically female ‘verbal genre’ (see McClure 1999), namely one related, through concern for the appearance of others, to another typically female genre: that of gossip. In any case, the line seems particularly ‘in character’ for this particular woman. As Denniston (on 959–66) puts it: All through the play her poverty galls her, and the sight of her mother’s pomp and state adds fuel to her hatred here, as it does when she flings her parting taunt at 1139–40.

All through the play indeed: Electra’s own appearance, and that of her mother and her mother’s consort, are a consistent source of distress to her.62 On the whole, taking considerations of gender into account, the evidence of 966 again appears to speak in favour of maintaining L’s assignations of speakers. As for 965, few scholars have dared mount a defence of L’s attribution of this line to Orestes. If it, too, should be spoken by Electra, the only reasonable solution would be the acceptance of a 61 The only part of the line that we might suspect of being gender-preferential is the word στολή, which (just like forms of the verb στέλλω, meaning ‘to dress’) is in fact slightly more likely to be uttered by or about male characters in tragedy, though any statistics here are clearly distorted by the frequent use of these words just before and in the ‘dressing scene’ at Eur. Bacch. 821–36, which is hardly without its own complications of gender. λαμπρύνω is too rare a word to ‘measure’. 62 Cf. e.g. 175–87, 239–41, 304–18, 947–8, 1069–76, with my discussions in ch. II.

Redrawing the Lines

213

lacuna after it.63 But I wonder to what extent we should really be disturbed by 965’s coming from Orestes’ mouth and how much it really voices ‘satisfaction’ (Cropp ad loc.). Several scholars have pointed out that the image of the hunting net (ἄρκυν ἐς μέσην) conforms to imagery used by Orestes before (582), so its use here is not ‘out of character’.64 Nor does the metaphor, in its continued use, strike me as distinctly brutal: that Clytemnestra is a prey in a trap is, after all, not in question among the siblings. Still, the perceived offensive tone of the line perhaps resides more in the word καλῶς than in the hunting imagery.65 It would be imprudent, however, to simply equate Orestes’ use of καλῶς here with his ‘personal’ view. First, if the adverb means something like ‘at the opportune moment’ or even ‘thoroughly’ (i.e. ‘directly into our net’) these meanings would not exclude a degree of hesitancy on Orestes’ part. There are sufficient parallels for the use of the adverb (in the sense of ‘thoroughly’) in contexts where the speaker does not appear to assent to the content of his utterance, for example ὦ παῖ, καλῶς εἶ δῆλος οὐκ εἰδὼς τί δρᾷς (‘my child, it is plainly clear that you do not know what you are doing’, Soph. OT 1008).66 Even the more evaluative senses of the adverb can be used ironically (cf. e.g. Soph. Ant. 739, Eur. Med. 588). Second, apart from the precise meaning of καλῶς, there is another 63

Murray, followed by Denniston, solved the problem by printing a break in the stichomythia and made Electra deliver both 965 and, after a pause, 966: but this is very weak. Not only would such practice be an unusual disturbance of stichomythic convention (Denniston’s remark, on 959–66, that ‘the break in the stichomythia needs no defence’ is a neat rhetorical flourish but a poor defence of the break in the stichomythia); it seems to me highly unlikely in view of the particle combination καὶ μήν . . . γε, which tends to be used at the start of a speaker’s turn (in reaction to something that has come up in the interlocutor’s preceding turn). 64 Since hunting was the province of men, one might wonder whether the use of the word ἄρκυς could be taken as another reason for attributing 966 to Orestes. Indeed, that word is used almost exclusively by male characters in Euripides (the exceptions being its use by the Chorus at Bacch. 870 and by a child at Med. 1278), and more generally in tragedy. But other hunting terms (e.g. δίκτυον, ἀρκύστατα, forms in κυνηγ-; for an overview see van de Wijnpersse 1929, on Sophocles) are used without a clear gender preference in Euripides and tragedy in general, and hunting terminology is used by women in both prose and poetry as a metaphor for sexual courtship. 65 So also Basta Donzelli 1978: 160. Her own tentative suggestion that καλῶς here means magnificamente, splendidamente (a suggestion designed clearly to soften the tone) is unconvincing: the adverb does not seem to have this sense (see LSJ s.v. C.II). 66 See Finglass in press ad loc.; I am grateful to Patrick Finglass for allowing me to see portions of his commentary before publication. The tone at Soph. OT 1008 is certainly slightly ironical, not only because of καλῶς but also because of the address ὦ παῖ to the king of Thebes (who is, to be fair, the messenger’s junior); see Dawe ad loc.

214

Redrawing the Lines

aspect of the line that prevents us from simply equating it with Orestes’ personal opinion. The particle ἄρα signals that he is compelled by some external form of evidence to come to this conclusion (in this case, the information just provided in 964 that it is Clytemnestra who is approaching).67 This may suggest, as often with ἄρα, a certain discomfort on the speaker’s part with the content of his utterance. The best English equivalent, ‘apparently’, can similarly express tentativeness. If the exchange runs as follows: Speaker A: Speaker B:

No, in fact it is my mother, she who bore me. She is, apparently, making her way neatly into our trap.

—it seems to me that speaker B can be identified with Orestes without trouble, even an Orestes who is on the verge of having doubts about his mission. When those doubts, then, come to full expression at 967, and not before, that line is all the more surprising and dramatically effective. L’s line-to-speaker attribution up to 966 is, in sum, nowhere indefensible, and in places vastly to be preferred. Scholars who consider the lines incompatible with the characterization of Electra and Orestes have in my view either misconstrued the import of the lines or ignored clues earlier in the play as to what exactly those characters are like. In fact, 963 is very much ‘in character’ for Orestes, and 966 for Electra (this remains, of course, a subjective point), and the lines as they stand in L conform to a number of gendered patterns that can be detected in this play and in Greek tragedy more generally. What deserves emphasis, moreover, is that reversing L’s attributions is not only a matter of going against the manuscript’s ‘authority’ with regard to speaker names (which, as noted, is negligible): this move requires further serious surgery on the text (one or more lacunae, transpositions, or a break in the stichomythia). I cannot agree that this patient is sufficiently ill to require such drastic measures.

67 ἄρα is an ‘evidential’ marker, signalling that the speaker makes a statement (or, with ἆρα, asks a question) on the basis of evidence with which he or she has just been confronted (in drama, often a visual or other input; in Platonic dialogue, usually a particular line of reasoning). This meaning, which is expounded in Bakker 1993b and (more convincingly) in the relevant sections of Sicking and van Ophuijsen 1993, covers all the uses of the particle distinguished in Denniston 1954. The use of the particle in 965 is another reason not to follow Kirchhoff ’s transposition of 965–6: 965 is linked directly to 964 by ἄρα.

Redrawing the Lines

215

3.2. Diverging minds (967–87) Any doubts about line-to-speaker attribution cease at 967: Orestes here for the first time utters explicit doubts about the slaying of his mother: τί δῆτα δρῶμεν; μητέρ᾽ ἦ φονεύσομεν;68 (‘What must we do? Will we really slay our mother?’) Electra now needs to readjust, and asks with some trepidation (μῶν 968) whether Orestes has been seized by pity upon the sight of his mother. The extra lineam interjection φεῦ that follows is in my view smartly placed to mark a crucial moment in the scene, if not in the whole play.69 The exclamation marks Orestes’ withdrawal from ‘contact’ with his sister, the moment where their emotional paths separate (as their physical paths will, 1331–3). This aspect of the scene has not been examined as fully as it deserves, and it is worth pointing to the various features of the exchange that mark the lack of connection between the siblings. The particle γάρ in 969 (πῶς γὰρ κτάνω νιν . . . ;) serves simultaneously to do two things: first, it implicitly answers Electra’s question of 968 affirmatively (γάρ = ‘yes, for’, as Kovacs, cited in §3.1, correctly points out); second, together with πῶς and the deliberative subjunctive κτάνω, it signals that Orestes’ own question (‘How can I kill her?’) cannot be interpreted as a genuine, information-seeking interrogation. An explanatory particle could not be used in such a request for information, and γάρ thus forces 969 to be read as a rather despairing rhetorical question.70 Yet, in the first of a series of crossed wires, 68 As for the punctuation of the line, the parallel with Aesch. Cho. 899 Πυλάδη, τί δράσω; μητέρ᾽ αἰδεσθῶ κτανεῖν; (‘Pylades, what should I do? Should I shrink from killing my mother?’) speaks in favour of punctuating after δρῶμεν, as does word order. Having the question mark after μήτερ’ would place that word in unmarked ‘pragmatically expected’ postverbal position, whereas starting the question with μητέρ᾽ means it stands in a pragmatically marked preverbal position (see Dik 2007: 125–8); note also that the metrical configuration of the next line is identical, with a syntactical break at the caesura, immediately followed by a form of μητήρ (⏒ ‒ ⏑ ‒ ⏒ ⁞ μητ-), which in turn is followed by a postponed form that is normally prepositive (ἦ, ὡς), and then the verb. It seems clear that the forms of μητήρ are deliberately emphasized in this way (see Radt 1985: 114). The position of ἦ is sufficiently paralleled (see Rijksbaron 1991: 180). 69 For a similar instance of an extra lineam interjection at a turning point in a play, compare the famous use of ἆ at Bacch. 810 (with Seaford’s note ad loc.). 70 A rhetorical specifying question (or ‘wh-question’) with γάρ will normally steer the addressee to answer (not verbally, but interpretatively) with the ‘null set’ (that is to say, the ‘answer’ to a question introduced by τίς γάρ is ‘no one’, to ποῦ γάρ ‘nowhere’, to πῶς γάρ ‘in no way’, etc.). The force of πῶς γὰρ κτάνω νιν, ἥ μ᾽ ἔθρεψε κἄτεκεν, then,

216

Redrawing the Lines

Electra chooses not to accept this illocutionary force* for the question but to give a direct answer, as to a genuine question (picking up on Orestes’ πῶς with ὥσπερ). And this communicative mismatch extends to content: whereas Orestes now begins to focus on the matricide and its implications, Electra keeps trying to bring the topic of conversation back to the murder of her father. The detachment between Orestes and Electra becomes even more severe in the next triplet of lines, as Orestes embarks on what Basta Donzelli has rightly called ‘a veritable soliloquy up until 973’.71 He breaks off his conversation with Electra by addressing Apollo directly (971), using a second-person verb form (ἐθέσπισας, ‘you decreed’) that is not directed at his conversational partner up to this point. Electra, passed over, forcibly inserts herself back into the conversation: she self-selects* and interrupts with a rhetorical question of her own.72 But Orestes ignores her and continues his own complaint (973), with another second-person verb directed at Apollo (ἔχρησας, you bade’)—an unusual suspension of syntax across turns.73 Orestes is now so engrossed in the problematic aspects of matricide that he has no regard for Electra’s objections. This pattern continues after 973, and if contact is restored in that line—in that one of the siblings at least appears to react to what the other says—it is only partial. The gulf between their frames of mind is visible in the siblings’ ongoing struggle over topic management*, which is poignantly visible in the alternation, for six successive lines, of μητ(ε)ρ- and πατρ- words. Every time Orestes raises a concern about the matricide (μητέρ᾽, ἣν οὐ χρῆν, κτανεῖν 973, μητροκτόνος νῦν φεύξομαι 975, μητρί . . . φόνου δώσω δίκας 977), Electra hammers away at the need to avenge Agamemnon (πατρὶ

is something like: ‘There is no way I could kill her, since she nourished and bore me.’ The question here, of a type that Mastronarde (1979: 9) has called ‘aporetic questions’, expresses a high degree of desperation. 71 [U]n vero soliloquio sino al v. 973 (Basta Donzelli 1978: 161). 72 Electra’s question functions as a compact reductio ad absurdum by way of a one–two punch: first she qualifies the context in which her question should be interpreted by positing the premise that Apollo speaks unwisely, then she quickly follows it with a question that cannot be answered in that new context (hence pointing out the absurdity of the premise). The ὅπου clause is preposed in order to facilitate the argument. 73 See Mastronarde 1979: 62.

Redrawing the Lines

217

τιμωρῶν 974—note δὲ δή marking a forcible change of topic, μή γ᾽ ἀμύνων πατρί 976, πατρῴαν . . . τιμωρίαν 978). Matching, or rather overlapping with, this tug of war over the terms of the debate are disruptions in the conversational sequence. The turn pairs in the stichomythia at this point do not map onto neat adjacency pairs* but seem to devolve into a series of expansions* and restarts, which suggests that neither sibling is willing or able to ‘work with’ the other’s statements or questions as they come. When Orestes answers Electra’s question as to what harm he will face from avenging his father (βλάπτῃ . . . τί 974) by pointing out that this will lead to his exile and pollution as a matricide (μητροκτόνος . . . φεύξομαι 975), Electra latches on a post-expansion* to adjust his claim: it will be equally shameful not to defend his father (976, note καί, ‘too’, indicating that this line adds directly to the preceding one). Orestes, in turn, does not answer this point directly but attempts to begin another sequence (ἐγὼ δὲ μητρί . . . 977; note δέ), reasserting the topic of his own status as a matricide and its consequences. For her part, Electra initiates another new sequence herself (978, δέ again): the consequences of Agamemnon’s murder are at stake.74 Yet again, Orestes does not answer his sister’s question but tries an entirely new, unrelated tack—perhaps something went wrong with the oracle? (979). This time Electra does answer the question (980), but the answer leaves Orestes unsatisfied, and he attaches another postexpansion* (981). At this point Electra has finally had enough, cuts the dialogue off and goes for the jugular (982–5). The overarching sense, then, is of conversational sequences that either are left without completion or have completing second parts* that are challenged or subverted: the siblings’ failure to see each other’s point of view is reflected in their failure to construct conversation cooperatively. Such notions of detachment and conversational disarray may be significant in resolving the textual problems thrown up by the troublesome pair 977–8.75 At 977 (L: ἐγὼ δὲ μητρὸς τοῦ φόνου δώσω δίκας) most critics follow Denniston, whose extensive discussion brings him 74 In terms of conversation analysis*, 978 could well be coded as a ‘counter’, a first part* that replaces another first part (cancelling its ‘conditional relevance’; this contrasts with insert expansions*, which leave that relevance intact). For counters, see Schegloff 2007: 16–19. 75 In 976, Triclinius’ correction μή must be accepted. See Curnis 2005 for a review of the numerous emendations suggested for 977; both there and at 978 the problems may be impossible to solve convincingly.

218

Redrawing the Lines

to adopt Herwerden’s ἐγᾦδα· μητρὸς δ᾽ οὐ φόνου δώσω δίκας; (‘I know. But won’t I be punished for killing my mother?’). Denniston is brought to his changes by two considerations: (1) the ‘flatness’ of the line as a statement with the article τοῦ, and (2) the ‘unnatural stress’ on ἐγώ. But ἐγᾦδα as a completely self-standing phrase in the sense ‘I’m aware of that’ is without parallel,76 and its tone would certainly be out of place here;77 nor does it, incidentally, do much to lessen the ‘stress’ on ἐγώ. As for that ‘stress’—which, to be more precise, is a case of contrasting topic78—it can be explained as an attempt by Orestes to get his sister ‘on the same page’ with respect to what it is exactly that they are debating and what is involved. As much as it shifts between loyalties to a father and a mother, the siblings’ debate alternates between two different perspectives on the role Orestes must play. Electra, concentrating on a past crime and the appropriate retribution, conceives of her brother only as an avenger of Agamemnon;79 Orestes, now thinking of his future crime and its moral and legal repercussions,80 wants to talk about his would-be status as a matricide. With ἐγὼ δέ, he seems to be asking Electra to consider his future if he commits the crime he now no longer dares to commit. The force implicit in ἐγὼ δέ, then, is something like: ‘But consider my situation: . . . ’ Denniston’s other point, that the line is bland as a statement rather than a question, is in my view more pertinent, and there is then much to say for Weil’s palaeographically attractive ἐγὼ δὲ μητρός γ᾽ οὐ φόνου δώσω δίκας;, the combination δέ . . . γε being common in retorts and rejoinders.81 Finally, it is preferable to read μητρί: L may 76 The phrase, which occurs only about 20 times in Greek literature, is used by itself in comedy (not elsewhere in tragedy), but there it normally marks a eureka moment (‘I’ve got it!’), as at Ar. Thesm. 850. Only in a single comic fragment (Strattis fr. 35.2 PCG) is the use of ἐγᾦδα comparable to what Denniston needs it to be here— but there it answers a question with οἶσθ(α). The phrase οἶδ᾽ ἐγώ is used parenthetically, but never independently (nor should we unhesitatingly assume that the two phrases are used in the same way). In prose, ἐγὼ οἶδα or οἶδα ἐγώ never stand by themselves as a complete sentence. All in all, ἐγᾦδα as an independent sentence should not be introduced into our passage lightly. 77 Schwinge (1968: 94 n. 68) and Curnis (2005: 110 n. 3), too, object to ἐγᾦδα on this ground. 78 See Dik 2007: ch. 2. 79 She drives the point home, as Schwinge (1968: 93) points out, by underlining the father–son relationship with the possessive σέθεν (974, ‘your’). 80 See Denniston on 975 for the legal terminology φεύξομαι and δώσω δίκας. 81 See Fraenkel on Aesch. Ag. 939 and Denniston 1954: 152–4.

Redrawing the Lines

219

in fact have this text (and two copies of L certainly do have it),82 it gives better syntax (δίκην δίδοναι + dative + genitive) than μητρός, and it aptly evokes the notion of Clytemnestra’s furies.83 All this gives, then, ἐγὼ δὲ μητρί γ’ οὐ φόνου δώσω δίκας; (‘And I, won’t I pay the penalty to my mother for her killing?’).84 The dative μητρί, if accepted, also correlates nicely with L’s dative τῷ in the next line, which I would let stand. Most editors read τί δ᾽ ἤν . . . διαμεθῇς (‘but what if you forego . . . ?’), a possible colloquialism (Diggle 1974: 17 n. 1, Stevens 1976: 30–1); but the contrast between maternal and paternal furies (who are no doubt meant to come to mind) is stronger if the masculine τῷ, suggesting Agamemnon (and his furies), is kept and interpreted as an interrogative.85 In this regard, it has been rightly noted that Euripides has Electra play the part of the Orestes of Aeschylus’ Choephori, whereas Orestes uses the arguments of that play’s Clytemnestra.86 Changing L’s τῷ in 978 actually obscures this parallelism, and this is a good reason why we should hesitate to do so. Electra’s question counters Orestes’ argument that he’ll have to face his mother’s wrath, by implying (she need do no more than that) that, if he does not go through with the murder plot, he will have to face a different set of equally disastrous consequences. Next, I would (reluctantly) keep Triclinius’ δαί, which requires Porson’s διαμεθείς later on in the line rather than Barnes’s διαμεθῇς.87 This gives τῷ δαὶ πατρῷαν διαμεθεὶς τιμωρίαν; (‘To whom then [will you pay the penalty] if you forgo paternal vengeance?’).88 The question is

82

See n. 9. This is argued more fully (and with ample parallels) by Curnis (2005), who keeps, however, the article τοῦ. 84 The text I argue for is also adopted by Distilo and Cropp ad loc. 85 Porson reads τῷ as a demonstrative (this underscores the contrast, but is a more exceptional locution; see Denniston’s objections ad loc. to the demonstrative, and also Slings 1997c: 132–4). 86 See, e.g. Keene ad loc., Schwinge 1968: 94, Lloyd 1986: 18, and Cropp ad loc. The relevant lines from Choephori are 924–5: ΚΛ. ὅρα, φύλαξαι μητρὸς ἐγκότους κύνας. / ΟΡ. τὰς τοῦ πατρὸς δὲ πῶς φύγω παρεὶς τάδε; (Cl. Look out, be careful of your mother’s vengeful hounds! / Or. But how will I escape my father’s if I forego this?) 87 The MSS have διαμεθίης, which is unmetrical (and the present indicative gives poor sense). 88 All in all, my reading of 977–8 is identical to that of Parmentier and Grégoire; but some doubts about the particle δαί in tragedy must remain (see Denniston 1954: 262–4). If δαί is rejected, δή instead, or else τῷ δ᾽ ἄν . . . διαμεθῇς (Distilo; Camper had this, but with ἤν), or Porson’s τῷ δ᾽ αὖ . . . διαμεθείς (but taken as a question; see n. 85 and cf. 1116, with Diggle app. crit.) are all plausible alternatives. 83

220

Redrawing the Lines

rhetorical, of course, and has as an implicature*: the dead Agamemnon will haunt Orestes unless he exacts punishment on Clytemnestra. The final textual problems of the scene are thrown up by the triplets 979–81 and 982–4, which L, going by its use (and non-use) of paragraphoi, attributes to Orestes (979–81) and to Electra– Orestes–Electra (982–4) respectively. L also lacks a paragraphos at 985—the clearest demonstration that something has gone wrong with speaker indications here (the speaker of 985–7 cannot, of course, be Electra). Things are, somewhat surprisingly, better in P, which has paragraphoi in 980 and 981 (as well as punctuation at 979 and 980, which is lacking from L; this suggests that the differences between the two MSS are not coincidental). Scholars nowadays typically side with P, where Orestes–Electra–Orestes alternate in 979–81, but they diverge from both manuscripts in giving Electra all of 982–4. Cropp (1988: on 979–81, 982) formulates the case concisely: Undoubtedly 985–7 belong to Or. (who goes in) . . . and 982–4 all to El. (with Weil’s corrections in 983) . . . ; L’s reading of 983 looks like a weak attempt to make sense of it after the false attribution to Or.; . . . 979–81 are much more forceful as an exchange between the two of them, and again the misassignment of lines may have led to the false reading οὐδ᾽ ἄν, ‘nor would I’.

To my knowledge, no one has argued recently for having Orestes as the speaker of all of 979–81 (as in L), and Cropp is surely right about οὐδ᾽ in 981, which despite scholars’ best efforts has never been satisfactorily construed. We may then read P’s οὐκ ἄν or Hermann’s conjecture οὐ τἄν (= οὔ τοι ἄν).89 In any case, the import of these three lines is clear. Orestes first tries another desperate line of argument to justify his doubts: perhaps a demon made the prophecy? The question does not really seem addressed at anyone in particular, and, as we have seen, it is unconnected to the preceding line: here again communication between the siblings seems to break down.90 Electra 89 For the impossibility of maintaining οὐδ’, see Denniston ad loc., and pace Kells 1966: 53. Of the other options, I have a preference for οὐκ ἄν over οὐ τἄν: τοι (in τἂν) would emphasize the relevance of Orestes’ statement for Electra and so mark it clearly as an answer to her rejection, whereas, with οὐκ ἄν, Orestes’ statement stands wholly on its own (this makes the asyndeton more harsh). The latter option seems suitable for this context of reduced contact and makes Orestes’ expression more forceful (pace Cropp ad loc.). 90 There is less textual evidence for the lack of contact between the speakers here than earlier in the scene (but see previous note). However, as Mastronarde (1979: 79)

Redrawing the Lines

221

immediately points out the absurdity of Orestes’ question with a penetrating question of her own (ἱερὸν καθίζων τρίποδ᾽; 980, ‘sitting at the sacred tripod?’), and dismisses it (ἐγὼ μὲν οὐ δοκῶ, ‘I for one doubt that’).91 But Orestes continues to question the validity of the oracle, since accepting it would be to agree implicitly to carry out the matricide (981).92 Faced with an Orestes who is uncooperative and unresponsive to her logic, Electra abandons rational argument. It is significant that her next turn is not directly relevant to Orestes’ one in 981: this lack of relevance in itself gives rise, by implicature*, to the suggestion that Electra is dismissing the preceding line of the debate.93 Her new line of attack is to make an aggressive appeal to his manliness—a very serious threat to his face*: οὐ μὴ κακισθεὶς εἰς ἀνανδρίαν πέσῃ;94 (‘Won’t you refrain from cravenly falling into cowardice?’). This approach jerks Orestes back into the here and now, and it is the one that finally sways him: at 985 he states his resigned intentions with ἔσειμι (‘I will go in’). The lines that intervene, however, are extremely difficult. No feasible sense can be extracted from the manuscripts’ ἀλλ᾽ εἰς τὸν αὐτὸν τῇδ᾽ ὑποστήσω δόλον (983). Moreover, the paragraphoi in L’s text at 983, and especially at 984, are suspect: though it is not unusual for one speaker to complement another’s syntax with a subordinate clause,95 Electra’s completion (with her relative ᾧ clause in 984) of points out, it is in such cases often ‘left to the modern reader to posit some sort of preoccupation in one of the dialogue‐partners, and it is necessary to consider to what extent the theater‐audience was required to make on its own the same sort of assumption’. 91 For this use of ἐγὼ μέν (‘I, for one’), with μέν on its own implying a contrast with others’ possible opinions (even if that contrast is never expressed by a counterbalancing δέ), cf. e.g. Her. 1014, Hel. 496, and the phrase μὰ Δί’ ἐγὼ μὲν οὔ (‘By Zeus, not me!’), common in Aristophanes. 92 The negative and potential optative (οὐκ ἂν πιθοίμην) make for a strong denial (‘I couldn’t possibly believe’): see Rijksbaron 2002: §14.2.1. The perfect passive phrasing of εὖ μεμαντεῦσθαι τάδε (‘that this has been well prophesied’) focuses attention on the unreliability of the oracle itself, allowing perhaps for a mistake in the oracle’s utterance (see Cropp ad loc. on the ‘impersonal phrasing’; for the nuances of the perfect passive, see Rijksbaron 1984 and 2002: §10.1). Orestes has phrased doubts earlier about mortal prophesying (399–400). 93 See Grice’s tea party example discussed in the Introduction, §2.3.2. 94 I follow Denniston ad loc. in opting for the future indicative πέσῃ rather than the subjunctive πέσῃς (for the problem, see Rijksbaron 1991: 167–74, esp. 171–2); in any case, the force of the expression is that of a strong prohibition. 95 See Mastronarde 1979: 53.

222

Redrawing the Lines

Orestes’ τὸν αὐτὸν . . . δόλον (983) would be highly unusual.96 Further doubts are raised by the Triclinian correction καθεῖλες (‘you killed’) for L’s καθεῖλεν (‘she killed’) in 984. καθεῖλες is a necessary change if Αἴγισθον κτανών is retained, but these words, too, have been challenged (see below). All in all, it seems to me that textual corruption may have done more damage here than anywhere else in the scene and that efforts to salvage a text must remain tentative at best. Two solutions yield readable texts: • The first solution is to retain the alternation of speakers and adopt Victorius’ ἦ for εἰς.97 Basta Donzelli’s text, accordingly, reads: ΗΛ. οὐ μὴ κακισθεὶς εἰς ἀνανδρίαν πέσῃ; ΟΡ. ἀλλ᾽ ἦ τὸν αὐτὸν τῇδ᾽ ὑποστήσω δόλον; ΗΛ. ᾧ . . . El. Or. El.

Won’t you refrain from cravenly falling into cowardice? But am I really to lay the same trap for her? As the one with which . . .

• The second solution is to give all three lines to Electra and to adopt Weil’s combination of two well parallelled but relatively infrequent idioms: a prohibition with οὐ μή and the future indicative followed by ἀλλά and a positive command,98 and (for that positive command) the phrase οὐκ εἶ (ibis) and the future participle.99 This gives the text read by Denniston, Diggle, and others: Ἠλ. οὐ μὴ κακισθεὶς εἰς ἀνανδρίαν πέσῃ ἀλλ᾽ εἶ τὸν αὐτὸν τῇδ᾽ ὑποστήσων δόλον ᾧ...; El.

Won’t you refrain from cravenly falling into cowardice, but go to lay the same trap for her as the one with which . . .

96 I have been unable to find a parallel for the completion of a form of ὁ αὐτός by another speaker. τῇδε cannot be taken with τὸν αὐτόν δόλον but must go with ὑποστήσω(ν) (i.e. ‘preparing for her’, not ‘the same trick as her’). 97 The reading is actually Matthiae’s (Victorius read ἤ). Denniston has strong objections to the use of the particle combination ἀλλ’ ἦ here, but it seems to me rather apt: ἀλλά protests against the suggestion of ἀνανδρία and changes the topic one last time to the nature of the crime Orestes would need to commit, and ἦ in questions is supremely appropriate in stichomythia when a speaker is about to be persuaded to yield (it serves as the last gasp of protest as well at Aesch. Ag. 942). See also the refutation of Denniston by Basta Donzelli (1991b: 22–4). Still, I believe ἀλλ’ ἦ is not the best reading. 98 See KG I §387.7 (p. 177), with esp. Ar. Nub. 505, Ran. 462, 524. 99 Elsewhere only at Soph. Trach. 83, Eur. Hec. 579, and Supp. 326; the idiom εἶμι + future participle is in itself unexceptional.

Redrawing the Lines

223

Although the phrasing of the second solution is complex, it is palaeographically straightforward,100 and I prefer it for the reason that τὸν αὐτὸν . . . δόλον | ᾧ distributed among two speakers seems unsound, and also because, as Denniston (ad loc.) says, ‘Electra’s final appeal, at the end of the stichomythia, should broaden out into something more extended than a single line. The three lines, if we give them all to her, are well balanced by Orestes’ three.’ Although the need for ‘balance’ is never a sound argument, Denniston is right that Electra’s coaxing works better as a three-line speech and that there is no need to keep the stichomythia intact at the end of the dialogue.101 It would work well for Orestes, moreover, to respond with a terse ἔσειμι immediately after Electra has played the ‘cowardice card’. As for the last words in line 984, the textual alternatives here yield two fundamentally different scenarios. Electra is drawing a parallel between the upcoming killing and a previous one, and which previous killing that is depends on minor variations in the text. Basta Donzelli accepts Triclinius’ ᾧ καὶ πόσιν καθεῖλες Αἴγισθον κτανών (‘by which you laid low her husband, killing Aegisthus’: Orestes killing Aegisthus by δόλος), whereas most editors would opt for a much more significant intervention in the text, ᾧ καὶ πόσιν καθεῖλεν Αἴγισθου μέτα (‘by which she killed her husband together with Aegisthus’, or some words to this effect102—Clytemnestra killing Agamemnon by δόλος). It is in truth possible to construct a strong case for either of these solutions. Denniston lays out the case for καθεῖλεν and against Αἴγισθον κτανών as follows: πόσιν is rather too honourable a word . . . to be applied by Electra to such a skunk as Aegisthus. And Euripides clearly had the Choephori so much in mind when he wrote his Electra that an echo of Ch. 888 δόλοις ὀλούμεθ᾽, ὥσπερ οὖν ἐκτείναμεν naturally suggests itself here (cf. ib. 274, 556–7), in which case the point is that Clytemnestra is to perish by the 100

It is also, without question, the lectio difficilior. It may be relevant that L, in 980–1, leaves out two paragraphoi where I am certain that they should stand, and subsequently inserts them in 983–4, where I think they should not. I have wondered whether it is possible that L’s scribe, as he was putting in paragraphoi, was anticipating a three-line break in the stichomythia and simply put it in the wrong place. The lack of punctuation at the end of 979 and 980 (suggesting a conscious decision to give 979–81 to one speaker) and the question mark at the end of 983 would speak against this, however. 102 See Denniston ad loc., Parmentier and Grégoire, Diggle, Cropp ad loc., and Distilo ad loc. The last three all dagger Αἴγισθον κτάνων. Αἰγίσθου μέτα is Wilamowitz’ emendation; Parmentier and Grégoire suggest Αἰγίσθου χέρι, Cropp Αἰγίσθου χάριν. 101

224

Redrawing the Lines

same guile as she used against Agamemnon. Finally, while it is true enough that Orestes, having carried out one δολοφονία, is now to embark upon another, is that a point that Electra is likely to emphasize here? . . . [O]ne expects Electra to emphasize here, not that Orestes has just used it against Aegisthus, but that he is to use it now in retaliation against a woman who killed her husband by δόλος.

At first sight, this seems conclusive, and could even be strengthened: the theme of a dose of her own medicine is prominent not only in Choephori, but in this play as well (Denniston might have also referred back to 276–9 and 969–70, and for the theme of Clytemnestra’s deception to 8–10). This reading also gives good sense to ‘the same trap’ (τὸν αὐτὸν . . . δόλον 983), as it is not merely trickery in the abstract, but the specific trick of luring someone into a house where he/she will be killed, which would connect the two killings. Yet scholars have underestimated the opposing case, which rests on much more than palaeography (i.e. on the fact that Αἴγισθον κτάνων can be left intact). For one, πόσις is, in spite of Denniston’s misgivings, used by Electra and Orestes six times elsewhere in the play to refer to Aegisthus,103 and, even if we think the word to be full of sarcasm in each of those cases, there is no reason why sarcasm should not be at play here as well. One might even argue that it would be strange for Electra to refer to her father using this word here. As for the connection between the killings of Clytemnestra and Agamemnon, that is, as mentioned, a prominent theme; but the connection between the killings of Clytemnestra and Aegisthus is, in the Electra, just as prominent. The two murders are planned together (596–670), and Clytemnestra and Aegisthus are mentioned in one breath as joint victims repeatedly (276, 599–600, 613, 646); a particularly telling parallel may be found at 1142–4, where Electra sends Clytemnestra off with the words κανοῦν δ᾽ ἐνῆρκται καὶ τεθηγμένη σφαγίς, | ἥπερ καθεῖλε ταῦρον, οὗ πέλας πεσῇ | πληγεῖσα (‘the basket is ready and the knife whetted, the very one which killed the bull, next to whom you will lie, struck down’). There is a parallelism, too, between the specific nature of the δόλος in both cases: both Aegisthus and Clytemnestra are killed at a sacrifice (feigned or not), both are

103 Orestes: 270, 641; Electra: 61, 326, 1091, 1116. See Basta Donzelli 1991a: 20–1 and Schwinge 1968: 97 n. 75.

Redrawing the Lines

225

portrayed as sacrificial victims (778, 839–43, 1142–5),104 and both unsuspectingly go into the house where they will meet their fate. The matter is settled, to my mind, by a particle. Those who would emend Αἴγισθον generally do not discuss καί in 984. But that καί points in the direction of Orestes as the killer and Aegisthus as the victim. Taking Clytemnestra as the killer, the sense of addition expressed by καί would yield something like ‘lay the same trap for her as the one with which she also killed her husband, with Aegisthus’ help’; or, less naturally (given the position of πόσιν), ‘the same trap that was also the one with which . . . ’. Both these options are ill warranted. By contrast, a reading that keeps Αἴγισθον κτάνων and adopts Triclinius’ καθεῖλες gives the perfectly plausible ‘lay the same trap for her as the one by which you killed her husband too’. In other words, the link between τῇδε (983) and πόσιν (984) that καί seems to highlight falls best into place if the other ‘variable’ (the killer, Orestes) stays the same.105 What, then, of Denniston’s final point, that Electra’s rhetoric here should be to remind Orestes of the way in which Agamemnon died, rather than of his own earlier δολοφονία? Here, too, I think that a different reading is possible. Electra’s approach in these lines, as we have seen, is to accuse her brother of—or rather steel him against— weakness. In making that case, it is a strong argument for her to point out that he has already started on this path and that he should stay the course (rather than ‘falling into unmanliness’). The rhetoric, then, would be: ‘you’ve already done the exact same thing before; not to follow through now would be cowardice’. In terms of Orestes’ motivations, Electra’s juxtaposition of the two murders could also explain 104 Of course, the image of a bull at 1143 can itself be seen as a play on similar portrayals of the doomed Agamemnon; see ch. II, §5.3.5. 105 In all this, it matters somewhat whether we interpret the scope of καί as ranging over the relative clause as a whole (‘also the one with which’) or over πόσιν only, as a (narrow) additive focus particle (‘her husband, too’). In his Greek Particles, Denniston (1954: 294) apparently opts for the first possibility by listing this instance as one where, ‘in general’, ‘καί emphasizes the fact that the relative clause contains an addition to the information contained in the main clause’ (my paraphrase ‘which was also the one with which . . . ’ attempts to do justice to that categorization, as does, I suspect, Cropp’s ‘the very same trap with which . . . ’). I prefer, however, the latter (‘narrow’) option, not least because of the position of πόσιν immediately preceding καθεῖλεν (the position for ‘narrow focus’ constituents; see Matić 2003). In fact a number of the cases Denniston lists under the same header as this one are, to my mind, more properly explained in this way. In the end, καί points towards my preferred reading of the lines even if we do think that it has scope over the relative clause as a whole.

226

Redrawing the Lines

why he finally consents to go in: his divine mission consists of two parts, and only one has been executed thus far. In the Electra, the murders of Aegisthus and Clytemnestra play out as a diptych, much of the tragic horror of the play residing in the fact that the siblings, especially Electra, treat both murders as on a par when they should not. At this point in the play, then, it is entirely ‘in character’ for Electra to emphasise the parallelism between the two killings. She equates them precisely because this means viewing Clytemnestra not as her mother, but as merely an equal partner in the murder of Agamemnon. This is how Electra persists in viewing her mother throughout, even when she is finally faced with her in the agōn (see ch. II, §5.3.1): to Electra, Clytemnestra is ‘nothing but a stranger’.106 Orestes, by this point, no longer shares his sister’s outlook but is in the end compelled, both by his sister’s overweening willpower and by divine decree, to go ahead anyway.107 Again, I reprint the text as I think it should stand: ΟΡ. ΗΛ. ΟΡ. ΗΛ. ΟΡ. ΗΛ. ΟΡ. ΗΛ. ΟΡ. ΗΛ. ΟΡ. ΗΛ. ΟΡ. ΗΛ. ΟΡ. ΗΛ.

106

εἶἑν, κομίζειν τοῦδε σῶμ᾽ ἔσω χρεὼν σκότῳ τε δοῦναι, δμῶες, ὡς, ὅταν μόλῃ μήτηρ, σφαγῆς πάροιθε μὴ ᾽σίδῃ νεκρόν. ἐπίσχες, ἐσβάλωμεν εἰς ἄλλον λόγον. τί δ᾽; ἐκ Μυκηνῶν μῶν βοηδρόμους ὁρῶ; οὔκ, ἀλλὰ τὴν τεκοῦσαν ἥ μ᾽ ἐγείνατο. καλῶς ἄρ᾽ ἄρκυν ἐς μέσην πορεύεται. καὶ μὴν ὄχοις γε καὶ στολῇ λαμπρύνεται. τί δῆτα δρῶμεν; μητέρ᾽ ἦ φονεύσομεν; μῶν σ᾽ οἶκτος εἷλε, μητρὸς ὡς εἶδες δέμας; φεῦ· πῶς γὰρ κτάνω νιν, ἥ μ᾽ ἔθρεψε κἄτεκεν; ὥσπερ πατέρα σὸν ἥδε κἀμὸν ὤλεσεν. ὦ Φοῖβε, πολλήν γ᾽ ἀμαθίαν ἐθέσπισας, ὅπου δ᾽ Ἀπόλλων σκαιὸς ᾖ, τίνες σοφοί; ὅστις μ᾽ ἔχρησας μητέρ᾽, ἣν οὐ χρῆν, κτανεῖν. βλάπτῃ δὲ δὴ τί πατρὶ τιμωρῶν σέθεν; μητροκτόνος νῦν φεύξομαι, τόθ᾽ ἁγνὸς ὤν. καὶ μή γ᾽ ἀμύνων πατρὶ δυσσεβὴς ἔσῃ.

959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976

[N]ient’ altro che un’ estranea (Basta Donzelli 1991a: 21). My text of 982–4 is identical to Weil’s. In the last three lines, 983–5—which, in spite of the lack of a paragraphos in L (P has one), obviously belong to Orestes—two further changes are necessary: for L’s προβλήματος, read Denniston’s προβήματος, and for L’s δὲ καὶ ἠδύ (P and Triclinius: δὲ χ᾿ ἡδύ), read δ’ οὐχ ἡδύ. In both cases, I follow Denniston’s argumentation ad locc. 107

Redrawing the Lines ΟΡ. ΗΛ. ΟΡ. ΗΛ. ΟΡ. ΗΛ. ΟΡ.

Or.

El. Or. El. Or. El. Or. El. Or. El. Or. El. Or El. Or. El. Or. El. Or. El. Or. El.

Or.

ἐγὼ δὲ μητρί γ’ οὐ φόνου δώσω δίκας; τῷ δαὶ πατρῴαν διαμεθεὶς τιμωρίαν; ἆρ᾽ αὔτ᾽ ἀλάστωρ εἶπ᾽ ἀπεικασθεὶς θεῷ; ἱερὸν καθίζων τρίποδ᾽; ἐγὼ μὲν οὐ δοκῶ. οὐκ ἂν πιθοίμην εὖ μεμαντεῦσθαι τάδε. οὐ μὴ κακισθεὶς εἰς ἀνανδρίαν πέσῃ, ἀλλ᾽ εἶ τὸν αὐτὸν τῇδ᾽ ὑποστήσων δόλον ᾧ καὶ πόσιν καθεῖλες Αἰγίσθον κτάνων; ἔσειμι· δεινοῦ δ᾽ ἄρχομαι προβήματος καὶ δεινὰ δράσω γ᾽. εἰ θεοῖς δοκεῖ τάδε, ἔστω· πικρὸν δ᾽ οὐχ ἡδὺ τἀγώνισμά μοι.

227 977 978 979 980 981 982 983 984 985 986 987

Well then; servants, bring this man’s body indoors and hide it in the shadows, so that when my mother comes, she does not lay eyes on the corpse before her slaying. Hold on: here’s another matter we must attend to. What is it? Those are not reinforcements from Mycenae I see, are they? No, it’s my mother, who bore me. She is, apparently, coming neatly into our trap. Yet look how splendid she is in her carriage and fine clothing! What must we do, then? Will we truly kill our mother? Don’t tell me that pity has taken hold on you now that you’ve seen your mother in person. Ah! Yes, for how can I kill my mother, who bore and nurtured me? Just like she killed your father, and mine. Phoebus, your oracle was wholly unwise . . . Where Apollo is blind, who can be wise? . . . .when you ordained, as you shouldn’t have, that I am to kill my mother. What harm is there for you in avenging your father? I will go in exile as a matricide now, though I was pure then. If you do not defend your father, too, you will be impious. But I, won’t I pay the penalty to my mother for her death? And to whom will you pay the penalty if you leave your father unavenged? Did some avenging spirit perhaps utter these things, impersonating the god? Sitting on the sacred tripod? I, for one, doubt that. I cannot accept that this has been properly prophesied! Won’t you stop cowardly falling into unmanliness! Go in to prepare for her the same trap by which you also laid low her husband Aegisthus, when you killed him. I will go inside. I’m starting on a terrible path, and will do terrible things. If the gods think this right, so be it; but the trial is bitter, not sweet, for me.

228

Redrawing the Lines 4. CONCLUSION

It is presumably clear from the discussions in this chapter that I believe that L and P together give us a text that is more reliable than many editors give them credit for; the manuscripts are also generally more reliable in these passages than in, say, the amoibaion of 1177–232 (discussed in ch. II, §6.1), where the lyrical mode and the strophic structure have caused much more confusion in the transmission (this is a common phenomenon). L and P present us, both in 671–85 and in 959–87, with an Electra who is unreservedly intent on her mother’s destruction. In the latter scene we also see an Orestes who—before his sister does so—suddenly understands the horrible nature of what he is about to undertake. Euripides, dramatically, allows Orestes’ realization to take place within a passage of tense stichomythia, at the moment he first sees his mother. And because Electra is not yet at that point of understanding, the siblings’ communication suddenly and completely breaks down. They are, in the latter half of the stichomythia, only partly in contact, both focused on different things, and at times quite literally talking past each other. Electra’s feverish efforts to persuade Orestes flounder, yet cannot but succeed in the end: Orestes is under divine orders to do his terrible deed, and his resistance must fail. Resigned to his tragic fate, he departs, placing responsibility squarely with the gods: εἰ θεοῖς δοκεῖ τάδε, | ἔστω (986–7).

V A Tense Affair The Messenger Speech

1. INTRODUCTION If the characters of Electra and Orestes in Euripides’ Electra have become ‘something of a test case in the discussion of characterization in Greek tragedy’ (Cropp 2013: 6), the messenger speech halfway through the play (774–858) could well be considered a prime exhibit.1 Due to the very nature of the scene, the evidence is circumstantial, and one might object that it derives from hearsay; nonetheless, the testimony of the Messenger2 has carried some weight in scholars’ interpretations of the characters (particularly of Orestes) and of the play as a whole. The Messenger is most often called upon as a witness for the prosecution: by those, in other words, who find Orestes’ character more or less distasteful. A typical reading is that of Erdmann: The messenger speech shows Aegisthus, the sacrificial victim, as an unsuspecting and hospitable man, the victorious Orestes, on the other hand, as a wily, dangerously clever and entirely unscrupulous murderer.

1 The bibliography on the scene is sizeable: works of importance for this chapter are Keene, Denniston and Cropp ad locc.; Grube 1941: 308–9, Stoessl 1956: 88–9, Kitto 1961: 336, Erdmann 1964, O’Brien 1964: 34, Conacher 1967: 206–7, Steidle 1968: 63–91, Zeitlin 1970 passim, Arnott 1973: 50–2, 55–6, 1981: 186–8, Vickers 1973: 560–1, Mirto 1980, Gellie 1981: 6, Lloyd 1986: 15–16, Michelini 1987: 214, Easterling 1988: 101–4, Mikalson 1989: 87–9, de Jong 1990: 382–9, 1991 passim, Porter 1990, Bers 1997: 82–3, Barrett 2002: 61–2, 91–2, Gärtner 2005: 9–10, Allan 2013b: 609–10. 2 Throughout this chapter, I use ‘messenger’ for the type of character in general and ‘Messenger’ for the particular figure portrayed in Electra.

230

A Tense Affair: The Messenger Speech

Without a doubt, Euripides does not want to depict Aegisthus as a villain and to let him end as such. (Erdmann 1964: 142–3)3

But, as Cropp (1988 on 774–858) was aware, ‘the case is not simple’. If we are to judge the characters of Orestes and Aegisthus on the basis of this particular scene, we must be mindful not only of how the attitudes and sensibilities of the play’s original Greek audience may have differed from our own, but also of the intricacies of the messenger speech ‘genre’.4 In this chapter I aim to show that the portrayal of Aegisthus and Orestes in the scene is much more complex than Erdmann and others would allow, but also that the nature of the scene—and that of messenger scenes more generally—affects how we should think about such issues in the first place. As I will argue, scholars who interpret the messenger speech as simply an exercise in characterization fail to make sufficient allowance for what that scene is designed to be: a gripping, exhilarating, absorbing narrative. My own analysis will therefore look closely at how suspense develops as the scene progresses and how Euripides modulates his use of narrative tenses and other compositional techniques to plot that suspense. In this respect, the chapter also aims to contribute more generally to our understanding of Euripidean narrative, particularly as it is found in messenger speeches.

2. NOT ‘WHAT?’ BUT ‘HOW?’ Immediately upon arrival, the Messenger’s first words remove any lingering doubt over the outcome of Orestes’ mission (761–4): ὦ καλλίνικοι παρθένοι Μυκηνίδες, νικῶντ᾽ Ὀρέστην πᾶσιν ἀγγέλλω φίλοις,

3 Der Botenbericht zeigt Aegisth, das Opfer, als einen arglosen gastfreundlichen Mann, den siegreichen Orest dagegen als einen verschlagenen, gefährlich geschickten und gänzlich skrupellosen Mörder . . . Ohne Zweifel . . . will Euripides den Aegisth nicht als einen Bösewicht zeichnen und als solchen enden lassen. 4 For the problems of delimiting this ‘genre’, see Perris 2011, with further references; for a general treatment of messenger speeches, see Rutherford 2012: 200–16, with further references in n. 78.

A Tense Affair: The Messenger Speech

231

Ἀγαμέμνονος δὲ φονέα κείμενον πέδῳ Αἴγισθον· Mycenaean maidens, glorious in victory, I bring news to all our friends that Orestes is victorious, and that Agamemnon’s murderer Aegisthus is laid low.

In this respect, this messenger speech is no different from any other in Euripides: messengers normally begin by succinctly stating the outcome of what has happened offstage, and it is only at the instigation of their interlocutors that they provide more details about the actual course of events.5 As Kannicht points out, this technique has an important consequence: as a result, the basic suspense on the part of the addressee (and the audience) concerning the What is resolved, and transformed into a distinct suspense concerning the How of the sequence of events. (Kannicht 1969: 169)6

In other words, the development of suspense in messenger speeches, in fact the very nature of that suspense, depends entirely on the manner in which the messengers’ narratives unfold, rather than on the outcome of that narrative. Yet this does not necessarily mean that messengers’ narratives are any less exciting for the audience to see and hear performed; it means only that the source of that excitement is of a different kind. With Electra, what the audience wants to hear is ‘in what manner’ Aegisthus died (ποίῳ τρόπῳ 772) and what was the sequence of events (the ‘pattern of the killing’, τίνι ῥυθμῷ φόνου 772).7 The effectiveness of the Messenger’s narrative hinges on the details he paints to flesh out that manner and on the suspense he builds into that pattern.

5 See, e.g. Med. 1125–6 and Heracl. 786–7. The only exception to this rule is the second messenger speech in Or. 1395–502; see de Jong 1991: 32. 6 [D]adurch wird in dem Adressaten (und im Zuschauer) die elementare Spannung auf das Was gelöst und in eine differenzierte Spannung auf das Wie des Ereignisverlaufs verwandelt. See also de Jong 1991: 32–4 and Finglass 2007 on Soph. El. 679. 7 Although the connection of the word ῥυθμός with φόνου (which in any case is to be taken apo koinou with τρόπῳ as well) is bold, I do not think that LSJ is right to classify this instance under a separate heading with the abstract meaning ‘manner’, ‘fashion’ (s.v. ῥυθμός VI; the other examples there are better). Electra’s two questions in 772 are not tautologous.

232

A Tense Affair: The Messenger Speech 3. ANALYSIS OF THE NARRATIVE

3.1. Setting the scene (774–8) As to that pattern, the Messenger provides Electra and the Chorus with a linear narrative: his modulations rely not on changing the order of events, but on his slowing down and speeding up when his tale (or rather the way he wants to tell it) calls for it. His first step is to briefly contextualize the narrative within the chronology and setting of the play (774): ἐπεὶ μελάθρων τῶνδ᾽ ἀπήραμεν πόδα . . . When we had left this dwelling here . . .

The ἐπεί clause is typical of Euripidean messenger speeches.8 It refers back to the moment when Orestes and his retinue set off on their mission (669–92); μελάθρων τῶνδ᾽ supplies the point of departure. Next, the short trip from the Peasant’s cottage to Aegisthus’ horse pastures (ἀγρῶν πέλας τῶνδ᾽, ἱπποφορβίων ἔπι 623, ‘out on his horse pastures, not far from these fields’) is quickly dealt with in a single line (ἐσβάντες ᾖμεν δίκροτον εἰς ἁμαξιτόν 775, ‘we came onto a twintracked wagon path and followed it’) and, without further ado, Aegisthus is introduced into the narrative (776): ἔνθ᾽ ἦν ὁ καινὸς τῶν Μυκηναίων ἄναξ. to the place where the new lord of the Mycenaeans was.

The adjective καινός9 (‘new’) is one of the many clues in the scene that betray the Messenger’s personal stance towards Aegisthus and his death: other indications may be found in the sarcastic μητρὸς εὐνέτης σέθεν (803, ‘your mother’s bedfellow’) and his comment on Aegisthus’ use of the word ἐχθρούς (λέγων Ὀρέστην καὶ σέ 807–8, ‘enemies— referring to you and Orestes’).10 Altogether ‘the Messenger is 8 Roughly two thirds of Euripidean messenger speeches open with an ἐπεί clause: see Rijksbaron 1976 and de Jong 1991: 34. 9 Kvíčala’s emendation: it is difficult to choose between this and the MSS reading κλεινός (‘glorious’), but the latter is perhaps too ironical, and corruption between the two words is common and easy to explain (see Denniston ad loc., Cropp ad loc., and Basta Donzelli, app. crit.). 10 This is, I think, correctly seen by de Jong (1990: 385) as ‘indignant’, not naïve. The Messenger twice (here and at 788–9) interrupts his tale with a ‘narratorial intervention’: both strike me as highly effective. This Messenger appears to me to be an expert narrator, far beyond ‘naïveté’, pace Grube 1941: 308 and Bers 1997: 83 n. 103. Bers

A Tense Affair: The Messenger Speech

233

convinced of the righteousness of Orestes’ deed and his report fully savours the irony of the situation’ (de Jong 1990: 384).11 He presents the narrative from his own, biased perspective, even though he quickly ‘fades into the background’ as a participant in the events (after this point, his only use of a first-person verb form is ἦμεν at 790).12 Having dealt in a cursory manner with his party’s journey, the Messenger slows down when he arrives at the scene of the sacrificial ceremony. A two-line description full of graphic detail shows Aegisthus preparing for the sacrifice to the Nymphs, in a lovely setting (777–8): κύρει δὲ κήποις ἐν καταρρύτοις βεβώς, δρέπων τερείνης μυρσίνης κάρᾳ πλόκους. He happened to be strolling around in a watered orchard, cutting off sprigs of tender myrtle for his head.

The verb form κύρει (‘he happened’) used in this description deserves additional comment. All editions print this form accented as κυρεῖ (present), but the historic present seems very out of place in this orienting, descriptive moment (the imperfect would be expected, cf. e.g. Hipp. 1173–5, Andr. 1086–91, Bacch. 677–9). Is the accentuation κυρεῖ correct, however? Rijksbaron (2006) has shown that a number of third-person singular forms of -έω verbs—long thought to be (and almost universally printed as) historic presents—are more likely unaugmented imperfects. The only change involved is one of accentuation, and the manuscripts in fact often have imperfects (e.g. κύρει at Soph. Phil. 371, where editors print κυρεῖ—a noteworthy example for my purposes). The fact that our form, contrary to what we normally find with historic presents, is ‘atelic’ (together with βεβώς, it clearly refers to a continuous process or state) strongly suggests that we should in fact read κύρει, imperfect.13 That imperfect, in any case,

typically but unjustly associates frequent quotation of direct speech with a naïve style of narration; Huitink 2013 argues well against this. The intervention at 788–9 appears to me to be another instance of the Messenger savouring the irony. 11 On the subjectivity of Euripidean messengers, see de Jong 1991: ch. 2. 12 See Barrett 2002: 61–2 n. 65, 75–6. The first person is needed here in order for the Messenger to remain a reliable eyewitness, in view of the change of scenery that occurs at 790. For the satisfaction of such restrictions on messengers’ knowledge, see de Jong 1991: 12–30. 13 Telicity—the presence of an inherent ‘end point’ in the semantic content of a verb (phrase)—is one of the features that, according to Rijksbaron, belong to nearly

234

A Tense Affair: The Messenger Speech

seems much more suitable in the context of the description of the locus amoenus where Aegisthus will meet his end. The details in that description are not merely decorative. The image of Aegisthus crowning his head with myrtle adds to the dramatic irony of the scene: myrtle had a general decorative function, but also a specific ritual function in funerary adornments.14 Aegisthus’ self-crowning may have significance, too, if we think of him as the true ‘victim’ of this particular sacrifice:15 one of the first stages of sacrificial rituals was the adorning of victims with crowns or other decorations (see Denniston on 791ff.).

3.2. ‘A deliciously protracted game of cat and mouse’ (779–97) The next part of the scene has been subjected to an unnecessarily onesided treatment by most scholars. Before the actual sacrifice, where knives, cleavers, and the like dominate the scene (810–43), there is an extended sequence (taking up 19 lines) of exchanges between Orestes and Aegisthus. This part of the scene has been read as ‘preliminary’ material (so, e.g. Cropp on 774–858), and for many it does little more all unambiguous historic presents: they occur (1) in narrative, (2) in declarative sentences, (3) in the first and third person, (4) with telic verbs, (5) not normally in the passive, (6) not normally in subordinate clauses, (7) not normally with a negative, and, finally, (8) not normally with γάρ (δέ and καί are common). Rijksbaron (2006: 147 n. 31) himself mentions our example in a footnote (see also Allan 2009: 193 n. 58). Various objections have been raised against Rijksbaron’s views: there may be cases of non-telic historic presents (see Boter 2012, with Rijksbaron’s response, 2015); more generally, one might argue that Rijksbaron is ‘replacing a rare feature with a much rarer one’ (Battezzato 2007: 159). But the list of unambiguous examples of unaugmented past tense forms in messenger speeches alone includes several dozen examples (Rijksbaron 2006: 136), and I am persuaded in this case that κύρει (imperfect) is right. (As for the MSS, L has no accent on κυρει, which suggests -εῖ as the diphthong stands in ligature; P has -εῖ; but manuscript ‘support’ is, in the case of accentuation, a virtually meaningless criterion.) 14 For ancient uses of myrtle, see A. Steier’s entry ‘Myrtos’, RE XVI.1: 1171–83, El. 324 and 512, and Alc. 172. For its significance here, see Cropp ad loc. and Michelini 1987: 213. 15 Mikalson (1989: 88 n. 45) objects to this reading, remarking that, ‘had Euripides intended this, he would have written lines 838–43 quite differently’. But seeing Aegisthus as das Opfer (Erdmann; see n. 3 above) actually matches very well with the description in 842–3, πᾶν δὲ σῶμ᾽ ἄνω κάτω | ἤσπαιρεν ἠλέλιζε (‘with his entire body he struggled and shuddered’), which could easily describe a sacrificial victim on the altar (see Mirto 1980: 312–13).

A Tense Affair: The Messenger Speech

235

than to highlight Aegisthus’ welcoming nature and Orestes’ extreme duplicity.16 It is, in that case, little more than an exposé, a study in character preparing us for the bitter taste that the second half of the scene will give us. What this neglects to take into account is that this first confrontation between Aegisthus and Orestes is actually of vital importance for the creation of suspense and for the resolution of the mēchanēma plot as a whole. Several formal aspects of the passage are worth pointing out in this respect: • The incidence of direct speech in this messenger scene as a whole is ‘especially dense’ (Bers 1997: 82; see de Jong 1991: 131–9). It is in fact the highest in extant Euripidean tragedy: 11 quoted speeches make up some 35 out of the 85 lines of the narrative. Direct speech is particularly frequent in these ‘preliminary’ exchanges. It is interrupted only by the Messenger’s brief speech introductions (779, 781, 783, 790, 793), which sometimes take up less than a line of text. • These brief speech introductions by the Messenger (i.e. the verba dicendi) follow a very precise pattern: Aegisthus’ verb of speaking is always in the (historic) present tense (ἀυτεῖ 779; ἐννέπει 783; ἐννέπει 790; cf. also λέγει in 814), and Orestes’ responses are marked by the aorist εἶπε (781, 793).17 • The dialogue concludes with a strongly marked break in the text: τοῦτον μὲν οὖν μεθεῖσαν ἐκ μέσου λόγον (797, ‘this was their conversation amidst the company’), where the particle combination μὲν οὖν (together with the following δέ in 798), the anaphoric τοῦτον, and the concluding content of the description function as a rounding off of this part of the narrative, before the Messenger moves on to the next stage. Why does the Messenger spend such a long time on this exchange, and why does he make such elaborate use of the instrument of direct speech? A judicious answer to these questions should, I believe, take note of several things. First, the audience knows nothing at this point 16

See, e.g. Grube 1941: 308–9, Erdmann 1964, and Arnott 1981: 186–7. These speech introductions, particularly the instances of ἐννέπει, also do something to give epic ‘flavour’ to the narrative; this epic mode will continue in the detailed description of preparations for the sacrifice (798–810). The influence of epic on Euripidean messenger speeches is well established (see, e.g. Cropp on 774–858, Barrett 2002). 17

236

A Tense Affair: The Messenger Speech

except that Orestes will eventually succeed in waylaying Aegisthus. The latter’s hospitality has been predicted by the Old Man (ἰδών σε δαιτὶ κοινωνὸν καλεῖ 637, ‘when he sees you he will call you to share in his feast’), who did not fail to add, however, that, from the moment of Aegisthus’ invitation, Orestes would be left to his own devices (τοὐνθένδε πρὸς τὸ πῖπτον αὐτὸς ἐννόει 639, ‘from that moment you yourself must devise your plan as the dice fall’). What happens at this point in the story is thus precisely what the audience cannot expect: it is here that the Messenger begins to reveal the ‘how?’ that the audience is after. Second, any possible expectation in the audience that Orestes will simply walk up to Aegisthus and strike him down is about to be thwarted repeatedly by Euripides: the Messenger in fact goes to great lengths to show that there are obstacles between the two, which are gradually removed (λόγχας δὲ θέντες δεσπότου φρουρήματα | δμῶες . . . ἵεσαν χέρας 798–9, ‘the servants put down the spears that protected their master, and put their hands to work’; δμῶας δ᾽ ἀπωθεῖ 822, ‘he pushed the servants away’). How Orestes manages to get closer and closer to Aegisthus—which is exactly what he is doing in this part of the scene—is thus of great significance. All this leads me to conclude that, for the audience, this point in the narrative is not at all ‘preliminary’. The first few exchanges between Aegisthus and Orestes are in fact a crucial part of the mēchanēma plot. Seen in this light, the high incidence of direct speech and the peculiar pattern of verba dicendi introducing that speech (presents for Aegisthus, aorists for Orestes) can be seen to serve a specific function in the construction of the narrative. In her analysis of the use of ‘character speech’ in messenger scenes, Irene de Jong (1991: 138–9) asserts that a Euripidean messenger ‘selects those speeches which he himself considers important . . . indispensible . . . pathetic, or vital to his own “message”’. Victor Bers (1997: 82), in a book about the use of direct speech in Greek literature, adds on the present episode that ‘Euripides clearly saw the verbal exchanges as making up a deliciously protracted game of cat and mouse’. I think this is right, but there is reason to complicate the straightforward division between ‘cat’ (Orestes) and ‘mouse’ (Aegisthus). In fact Aegisthus may unwittingly be more ‘catlike’ than commonly accepted. To an audience for which the ‘how’ of the scene is the primary source of excitement, every question that Aegisthus poses to Orestes has the potential to blow the latter’s cover. Each time Aegisthus puts Orestes on the spot, the

A Tense Affair: The Messenger Speech

237

audience is left wondering: ‘How will he get out of this one?’, ‘What will his reply to this one be?’—and the suspense rises. This effect is reinforced by the use of direct speech, which invites the audience to imagine itself ‘experiencing’ the conversation between Aegisthus and Orestes as it takes place (cf. de Jong’s term ‘pathetic’).18 It is also reinforced by the use of narrative tenses: regardless of the precise function we wish to ascribe to the historic present,19 it is a ‘marked’ tense, meaning that, every time it is used here, attention is (re)focused on Aegisthus’ speeches—each of which presents a new test for Orestes’ plot. Thus what emanates from the present–aorist alternation could be described as an ongoing game of ‘blow and parry’. Orestes has to provide the correct riposte to each new challenge to his scheme (all highlighted by the historic present), in order not only to preserve his secrecy, but also to bring himself closer, step by step, to his prey, until he is within striking distance.20

18 This description is not very different from (and is indebted to) the characterization of the functions of direct speech in Luuk Huitink’s (2013) Oxford dissertation on reported discourse in classical Greek. Huitink draws upon the work of, among others, Clark and Gerrig 1990 and Collins 2001 to argue that Greek direct discourse, rather than being suggestive of ‘literalness’ (pace especially Bers 1997), serves to depict selective aspects of an original discourse situation, in order to make the addressee feel what it was or would be like to experience these aspects. Such aspects may concern the wording of the original discourse, but more usually relate to the manner or tone of delivery, the emotional state in which it was produced or received, and other aspects (see Huitink 2013: 179–201). I am grateful to Huitink for helpful discussions about these issues. See also Allan 2009: 177 (‘immediacy’) and 183 n. 30. 19 The discussion about the precise function of the historic present is ongoing: see Introduction, n. 91. For discussion of the historic present in messenger speeches, see de Jong 1991: 38–45 and Allan 2009. Both de Jong and Allan observe that there is a correlation between the use of direct speech and the use of the historic present in messenger speeches. 20 This interpretation goes very well with Allan’s (2009: 174–5) definition of the narrative function of historic presents in messenger speeches. Building on the works of Labov and Fleischman, Allan analyses the structural organization of narratives in general (as a rule, they flow through various phases: ‘orientation’, ‘complication’, ‘peak’, ‘resolution’, etc.). According to Allan, historic presents (which he considers to have a high degree of ‘tellability’) (1) occur specifically in peaks (climaxes), (2) serve to mark ‘a visually or verbally dramatic “shot” in the Complication’, or (3) initiate such a complication. The last of these descriptions, I think, is precisely applicable here; Allan (2009: 193 n. 58) lists them as ‘dramatic shots’, but I would argue that each of Aegisthus’ speeches opens up a new ‘mini-episode’ in the narrative, requiring resolution. Allan himself (2009: 195 n. 60) remarks on the ‘blow-and-parry’ or ‘action– reaction’ effect of the alternation between historic presents and aorists in this scene; he rightly compares Or. 887–917.

238

A Tense Affair: The Messenger Speech

To be clear, I do not wish to claim that Aegisthus is consciously trying to break through Orestes’ disguise, nor that Orestes is ever in serious danger of being discovered: de Jong (1991: 385 n. 47) is right to point out that this would ‘spoil the irony of the whole scene’. Rather, each of Aegisthus’ unsuspecting questions requires Orestes to come up with a measured response, while each hurdle that Orestes passes is an ironic reflection upon Aegisthus. For the Messenger, this is reason to dwell on the exchange extensively; for the audience, it is a source of ever-increasing suspense. The significance of the initial exchange is made clearer still as soon as it ends. Upon conclusion of the first dialogue, Orestes’ ‘reward’ for successfully maintaining his disguise is given: Aegisthus’ first layer of protection is peeled away, as the slaves put away their weapons (798–9). It is a recurring feature of the scene, in fact, that by shrewdly playing his part Orestes moves closer and closer to his opportunity to deliver the death blow.

3.3. A moment for prayer (798–810) Yet Orestes’ initial success does not immediately lead to the climax of the scene. The discarding of weapons in fact proves to be the first detail in a lengthy description. As well as an updated picture of Aegisthus’ defences, this description details the preparations for the sacrifice, which include two contrasting prayers by Aegisthus and Orestes; Aegisthus speaks his prayer aloud (804), while Orestes utters his under his breath (809). The entire section is narrated in the imperfect, suggesting that time is static: there is thus a notable pause in the forward flow of the narrative.21 The introduction of both prayers by verbs in the imperfect is especially noteworthy: for once, Aegisthus’ direct speech is not introduced by a historic present—the τοιάδ᾽ ἐννέπων ἔπη (804, ‘uttering such words’) is coincident with the imperfect ἔβαλλε, ‘he sprinkled’. Rather than pushing the story forward, therefore, the contrasted prayers provide the background 21 This is typical of what Allan (2009: 179–81) calls the ‘descriptive mode’. The imperfects ἔβαλλε (804, ‘he sprinkled’) and ηὔχετο (809, ‘he prayed’) make it clear that this mode continues on to 810 rather than to 802 (pace Allan 2009: 181 n. 25), even if there is a shift after 802 from the servants’ preparations to the prayer sequence (see Cropp on 747–858, 799–802). The fact that these prayers are also narrated in the imperfect is in fact crucial, as I will argue.

A Tense Affair: The Messenger Speech

239

against which the rest of the scene will develop.22 As a way of setting the stage for the rest of the narrative, this could not be more significant: once the prayers are set as the framework within which the actual sacrifice takes place, the bad omens with which Aegisthus will be confronted later (827–9) are bound to be seen in light of those prayers. This facilitates an important conclusion drawn by Mikalson, worth quoting in full: [T]he messenger describes how, after the prayers, Aegisthus’ sacrificial victim proved to be defective and ill-omened. The unacceptability of Aegisthus’ victim was an unmistakeable sign to an ancient audience that the sacrifice had been unsuccessful, that the prayer accompanying it would also be unsuccessful. By then assassinating Aegisthus at this sacrifice Orestes may be, as Denniston (p. 149) puts it, ‘sailing very near the wind’. But he is not guilty of sacrilege, I think, not because he does not receive the χέρνιψ or the meat (Denniston, pp. 148–9), but because the gods (more specifically the Nymphs) have, by rejection of Aegisthus’ sacrifice, removed him from themselves and their protection. The gods have clearly thrown in their lot with Orestes and they choose to answer his prayer. (Mikalson 1989: 88–9)23

3.4. The sacrifice of Aegisthus (810–43) The Messenger shifts gears again and resumes his narrative. Aegisthus’ ritual acts are now presented in story-furthering aorists (ἔθηκε 812; ἔσφαξε 813), before a new speech introduced by another historic present (λέγει 814) presents another challenge (or opportunity) to Orestes. Again, the Messenger shows how Orestes’ reaction gets him one step closer to the execution of his mission: in complying with Aegisthus’ request, he manages to remove the attendants from the picture (δμῶας δ᾽ ἀπωθεῖ 821, ‘he pushed away the servants’)24 and to lay hands on his first weapon. With another success under his belt, Orestes undertakes the flaying of the calf. The imperfect ἐγύμνου (823, ‘he bared’) is probably a case of 22 For the use of the imperfect to paint the background against which (or the ‘framework’ within which) other actions are to be viewed, see Rijksbaron 1988. 23 See Michelini 1987: 214 on ‘the ominous entrails that themselves signify, predict, and in part justify his death’. 24 This historic present has a function somewhat distinct from that of the earlier verba dicendi. Allan (2009: 193 n. 58) sees it as another example of a ‘visually or verbally dramatic “shot” in the Complication’.

240

A Tense Affair: The Messenger Speech

what Rijksbaron (2002: §6.2.3) calls the ‘immediative’ use: ‘no sooner had he taken the calf ’s foot than he was baring its white flesh’. The aorists ἐξέδειρεν (824) and ἀνεῖτο (826) are then appropriate for expressing the successful butchering as a whole, an action that moves the Messenger to draw an admiring comparison (824–5):25 θᾶσσον δὲ βύρσαν ἐξέδειρεν ἢ δρομεὺς δισσοὺς διαύλους ἱππίους διήνυσεν. And he skinned off the hide quicker than a runner completes two lengths of the horse-track.

The comparison seems out of place at first sight but is designed to fit in with the athletic imagery that recurs consistently in this part of the play (614, 761–2, 781–2, 854, 862–3, 872, 880–9;26 διήνυσεν is a gnomic aorist); the continued portrayal of Orestes as a victorious athlete is another indication of the Messenger’s partisanship (though not necessarily of Euripides’).27 The Messenger now slows down again, to paint a more elaborate picture of the examination of the entrails (imperfects ἤθρει 827; προσῆν 827 and ἔφαινον 829), which will give rise to the remarkable exchange of speeches at 830–7. The description is rich in detail, and the details again have significance beyond lending a graphic quality to the scene. Euripides makes the most of the dramatic irony inherent in the failed sacrifice and plays it up through his use of the substantivized participle τῷ σκοποῦντι (829). On this form, Barrett comments as follows: [W]e see, for the moment, with Aigisthos’ eyes. The syntactical ambiguity of the participle in the dative case (τῷ σκοποῦντι) underscores this: as a complement of the verb (ἔφαινον), the participle may indicate anyone who looked at the innards; construed as the object of the coming hostile attack (κακὰς προσβολάς), it clearly defines Aigisthos as author of the gaze. (Barrett 2002: 91–2)

Although Barrett’s term ‘construed’ is not entirely felicitous (that Aegisthus is object of the ‘hostile attack’ is an implication of the context, not a syntactic phenomenon), he is right to point to the ambiguity between the general sense of the participle phrase and its specific 25 On the subjectivity of comparisons in messenger speeches, see de Jong 1991: 87–94. 26 27 See Introduction, §4.2. See Swift 2010: 156–72.

A Tense Affair: The Messenger Speech

241

referent: the article, which is naturally interpreted with participles as having a ‘generic’ function (‘whoever was looking on’), here also clearly identifies the person it describes (‘the man who was looking on’). That person’s gaze is not a happy one. The Messenger now allows the irony of the scene to rise to new levels, and this time the initiative lies with Orestes (830–5): χὣ μὲν σκυθράζει, δεσπότης δ᾽ ἀνιστορεῖ· τί χρῆμ᾽ ἀθυμεῖς;—ὦ ξέν᾽, ὀρρωδῶ τινα δόλον θυραῖον. ἔστι δ᾽ ἔχθιστος βροτῶν Ἀγαμέμνονος παῖς πολέμιός τ᾽ ἐμοῖς δόμοις. ὃ δ᾽ εἶπε· φυγάδος δῆτα δειμαίνεις δόλον, πόλεως ἀνάσσων; He scowled, and my master asked, ‘What concerns you?’—‘Stranger, I fear some deceit from outside. Of all men, Agamemnon’s son is most hostile to me, an enemy of my house.’ ‘So you are afraid of deceit by a fugitive’, Orestes said, ‘you, who rule over the city?’

As if spurred by Aegisthus’ distress, Orestes here for the first time self-selects* and initiates a conversational sequence—he takes matters into his own hands. The historic presents (σκυθράζει, ἀνιστορεῖ) again focus attention on the quoted speeches, which are rife with dramatic irony, unintentional in the case of Aegisthus’ θυραῖον and ἔστι . . . Ἀγαμέμνονος παῖς, intentional in the case of Orestes’ δῆτα,28 φυγάδος and πόλεως ἀνάσσων.29 At last Orestes asks for his murder weapon and strikes . . . the calf. Euripides’ deft play with the expectations of the audience at this point has been well analysed by Arnott: Surely the murder now? No, not quite yet; . . . The audience may have been expecting a designation of the king to follow the word ‘smashes’ as its object, but if so, their expectations are . . . cheated. Orestes simply smashes the breast-bone. The murder of Aegisthus follows three lines later . . . The reason why Euripides lays this false trail in his Electra is not hard to seek. By making his audience wonder at each mention of a weapon whether this is going to be the climactic moment of murder, he sets everybody on tenterhooks and builds up tension. (Arnott 1973: 56)

28 The particle adds a ‘note of surprise or indignation’ (Denniston 1954: 272) to Orestes’ question, here ironically. 29 For the overlapping levels of irony here, see de Jong 1990: 386.

242

A Tense Affair: The Messenger Speech

Arnott’s analysis of building tension may be reinforced if we look once again at the use of tenses, particularly the use of the historic present κόπτει in line 838. As Allan noted, there is a discrepancy between, on the one hand, the crucial event in the plot structure and, on the other hand, the linguistic Peak (signalled by the historical present) . . . the blow to the calf ’s head (κόπτει) is marked by a historical present, whereas the fatal blow to Aegisthus’ head is expressed by an aorist form (ἔπαισε). The first action is rhetorically highlighted by means of the present tense, perhaps as a dramatic foreshadowing of the killing of Aegisthus. (Allan 2009: 194)

We can strike Allan’s ‘perhaps’ in the last sentence. The blow is not one ‘to the calf ’s head’, since the calf has been dead for 25 lines (κἄσφαξ᾽ 813, ‘he cut its throat’), and thus the historic present cannot be interpreted as the climactic moment in the slaughter of the calf. κόπτει, as ‘linguistic Peak’, must have given rise in the audience to the expectation that Aegisthus was the recipient of this blow. But Euripides surprises by having Orestes strike the breastbone instead. Arnott (1973: 56) is certainly right, then, in suggesting that Euripides very deliberately lays out ‘a trail of red herrings’. But I cannot agree with him that this is necessarily a reflection on Orestes’ poor character (Arnott 1981: 186). On the contrary, I think the ‘tenterhooks’ again force the audience to ‘experience’ the events, as they unfold, with the Messenger (through whom the narrative here is clearly focalized; see, e.g. the use of δεσπότης in 830), and therefore with Orestes (on whose side the Messenger obviously is). The Messenger takes a deep breath, sets Aegisthus to work at entrails examining—the imperfect ἤθρει (839, ‘gazed’) marks the ‘final positions’ before the event, as it were—and finally lets Orestes dispatch him, in particularly gruesome fashion (839–43).

3.5. Aftermath and resolution (844–55, 855–7, 857–8) The messenger speech concludes with a brief, somewhat separate account of the confrontation between the attendants and Orestes and Pylades, which leads into the triumphant atmosphere of the following choral celebration (860–79). The narration is in the aorist (844–53), until the final historic present στέφουσι (854, ‘they crowned’: again, an epinician term) marks Orestes’ status as victor.

A Tense Affair: The Messenger Speech

243

There follows a description of the current situation (Orestes is on his way: 855–7, present) and a concluding evaluation by the Messenger (857–8, with the ‘constative’ aorist ἦλθε, ‘he has come’;30 note νῦν, ‘now’). Little else needs to be added about these passages except that they show again how literally everyone in the play—including Aegisthus’ own servants—is on Orestes’ side. No one seems to mind Aegisthus’ death: I am tempted to believe that the audience is not invited to do so either, although audience members may well have been troubled by the brutal form that event takes. It is this issue, the evaluation of Orestes’ actions, that I address in the remainder of this chapter.

3.6. Evaluation Up to this point I have tried to show that the narrative is carefully constructed to engage the audience’s attention and that an analysis of any part of it as merely a sort of ‘character tableau’ overlooks the powerful forces of narrative suspense that Euripides unleashes on the audience in this scene. This is not to say that there are no indications at all that the murder is presented as problematic. There has been much debate over notions of ‘ungentlemanly behaviour’ (killing someone when his back is turned) and ‘sacrilege’ (killing someone at a religious occasion) on Orestes’ part. Scholars have variously held one or both of these offences to be at issue here; others have objected that Greek morality would have no real problem with either behaviour.31 There is, it is true, little to suggest that Greek ethics condemned trickery or back-stabbing when it came to the rightful killing of a 30

For this use of the aorist, see Rijksbaron 2002: §8.3.1. Cropp (on 774–858) rightly warns against anachronistic views. But even those on guard against such views can tumble in the same pitfall. A striking example is Grube (1941: 309): ‘The unpleasantness of this murder is not due to Aegisthus’ being stabbed in the back. To regard the murder as “unsportsmanlike” on this account is to introduce modern, and probably false, values. Trickery was clearly essential. The unpleasantness to a Greek was due to the king being killed while offering sacrifice. Orestes, in fact, does what Hamlet failed to do.’ But Shakespeare is no more applicable than our present-day morality. The Electra, as Gellie (1981: 6) points out, is ‘not the Hamlet kind of play’. Nonetheless, the fact that Aegisthus is killed like an animal at a sacrifice does complicate the moral picture, as I will argue presently. 31

244

A Tense Affair: The Messenger Speech

proven foe, particularly where adulterers were concerned.32 But the second issue—murder at a sacrifice—is more complex. Various scholars point to Parker’s (1983: 159–60) suggestion that religious sacrifices did not offer participants protection such as refuge at an altar did,33 but this argument has met with some opposition,34 and each of the examples Parker mentions in its support seems to have problematic undertones of its own.35 There are, moreover, some undeniably troubling notes in the build up to the death blow, most notably the way in which Orestes manages to avoid ritual cleansing (793–5). Finally, the scene presents us not just with a murder at a sacrifice but with the disturbing equation of animal and human victim.36 It is hard to escape the conclusion that the setting in which the murder takes place casts it in something of an ugly light. What we are left with, then, is a picture of some complexity, and sympathy is not easy to assign. The narrative does something to suggest that the gods are not on Aegisthus’ side (805–7, 826–30), and there are other reasons why we should not feel too sorry for him. Much has been made, especially by Arnott (1981), of the ‘double vision’ to which the character of Aegisthus is subjected: as many scholars see it, the character flaws that are ascribed to him by all

32

This case is well made, e.g. by MacDowell 1966: 77. But cf. Soph. Trach. 269–80 for a passage that offers a somewhat different view. 33 Lloyd (1986: 16) endorses this point with several literary parallels, which suggest, according to him, that ‘such circumstances can reveal the will of the gods with particular clarity’: at Soph. Trach. 749–812, Heracles is struck down while sacrificing to Zeus after an unholy act of aggression (the suggestion is that Zeus did not look kindly upon this particular sacrifice); Harmodius and Aristogiton killed Hipparchus at the Panathenaea (with Athena as an approving witness, suggests Lloyd 1986: 16, on the basis of a scolion). But see n. 35. 34 See Porter 1990, Appendix, 278–80; but note also his cautions at 279 and n. 69; the body of Porter’s article posits, among other things, the attractive suggestion that the sacrifice in the messenger scene is to be compared to the ritual at the festival of Bouphonia: however, although Porter convincingly establishes the presence of this allusion, I still would not agree that the link necessarily favours ‘traditional’ (wholly negative) readings of the scene. 35 To be fair, Parker does seem to imply throughout that murder at a sacrifice might be regarded as an impiety (but that there is no explicit evidence for ritual pollution). But his examples are problematic: the famous case of Hipparchus’ death at a sacrifice may not be a death at a sacrifice after all (in the scolion in question, ‘Harmodius’ Song’, PMG 895, Ἀθηναίης ἐν θυσίαις may mean no more than ‘at the festival’); at Plut. Dion 58.1 Callipus is described as τηλικούτου μύσους ἁψάμενος (‘tainted by this much pollution’). I thank Scott Scullion for discussion on this point. 36 See Henrichs 2000: 187.

A Tense Affair: The Messenger Speech

245

the other characters in the play disappear in this scene. I wonder, however, to what extent this is really so. If there is one characteristic of Aegisthus that stands out in the prologue delivered by the Peasant (who is considered by some to be another ‘neutral observer’), it is his near-paranoid fear of being overthrown (δείσας . . . ποινάτορα 22–3, φόβου . . . τέκοι 25–6, φόβον 39)—a fear typical of tyrants.37 This characteristic of Aegisthus prevails in the messenger speech as well, and is evident from his ‘own words’ (807, 831–3). I have argued (§3.2), moreover, that Aegisthus’ oft-vaunted hospitality serves a complex function in the build up of suspense in the narrative; in any case it certainly contributes to the irony with which the Messenger laces his tale, and, depending on how ‘conventional’ such hospitality was considered to be, it may not say much about Aegisthus’ character at all. It has also been seen as a poetic reversal of the tradition that Aegisthus was Agamemnon’s host when he killed him, which we find in the Odyssey (4.529–37, 11.388–9 = 24.21–2, 11.409–11; see Cropp on 774–858). All in all, I believe we should be careful before interpreting Aegisthus’ character as unproblematically or surprisingly pleasant. He is not ‘the good guy’, nor is he an innocent and undeserving victim. Yet the motif of corrupted sacrifice suggests that the tyrannicide cannot be seen as a clean and unproblematic act either. For the play as a whole, the effect of these ugly undertones is, I think, a subtle but definitive modulation of tone—a problematization of retributive violence (as a man on a mission of personalized vengeance inside the family, Orestes was always going to be a problem case),38 and as such a foreshadowing of the even more destructive matricide (where the theme of sacrifice will return with a vengeance). In this way the messenger scene prepares us for the tragic resolution towards which the play is hurtling. In any case, the messenger scene is not a moment where prolonged reflection on such issues is the most obvious response. What this exhilarating narrative does, above all, is grab the audience by the throat and haul it through the tense ups and downs of Aegisthus’ final moments.

37

This is not the only time in the play that Aegisthus is characterized as a typical tyrant: see Cropp on 907–56. 38 Allan 2013b is a valuable discussion of the problematic status of retributive violence in Greek tragedy generally.

246

A Tense Affair: The Messenger Speech 4. CONCLUSI ON

The implications of the preceding analysis may be summed up as follows: • As for the ‘preliminary’ part of the scene (774–97), I believe that, rather than simply providing a bleak picture of Orestes’ character, it actually presents the first, exciting stage of a delicately crafted plot. I suspect that, for the audience, this passage did not invite— at least not in the first place—sustained reflection upon Orestes’ flaws, but rather an ‘experiencing’, along with the Messenger (and thus, implicitly, along with Orestes), of the events as they unfold. The Messenger crafts his narrative to show how Orestes has to navigate Aegisthus’ questions and invitations carefully, moving ever closer to an opportunity to strike. The balance is delicate, and the Messenger piles on the suspense through a deliberate use of narrative tenses and direct speech. These effects are in no way diminished by the ample doses of dramatic irony that the Messenger interlaces into the tale (to Aegisthus’ detriment): there is no reason why an audience cannot at the same time be excited by the developing plot and still crack a grin at Aegisthus’ failure to understand what is going on. • In the second half of the scene, Orestes is allowed to take the initiative completely, which results in ever-increasing dramatic irony. Assigning sympathy becomes progressively complex. Still, the scene does not turn into an exposé of character: it is above all the resolution of a gripping narrative. • There is no question that the Messenger is firmly on Orestes’ side throughout the scene. Any notion of ‘objectivity’—that ‘the messenger is an outsider, a third-person objective witness who records events in an unbiased way’ (Barlow 1986a: 14)—is untenable.39 But the Messenger is not alone on this score: no one in the play reacts unfavourably to the killing of Aegisthus, as they will very clearly do when it comes to Clytemnestra’s murder later on. • Still, the murder remains an ugly event, and the human–animal confusion plays up the problems: by shedding blood at a

39 A similar, equally untenable view of messengers is held by Bremer (1976: 64); it is effectively debunked by de Jong (1991: ch. 2).

A Tense Affair: The Messenger Speech

247

sacrifice, at the altar, Orestes renders his religious status ambiguous at best. To conclude, then, what does the messenger scene say about Orestes? That is, how should we decide between Erdmann’s view, quoted at the start of this chapter, and that of Lloyd and Cropp? The latter has been aptly summarized (though not endorsed) by Porter: [I]n their view, this exciting and innovative account of the tyrant’s demise is calculated to serve as a foil for the ensuing matricide, which (as in Aeschylus’ Choephori) is to be regarded as providing the true crux of the play and which serves as the focus of its angst-ridden concluding scenes. (Porter 1990: 260)

The truth, I think, lies somewhere in the middle, and it is not by coincidence that the messenger scene lies somewhere in the middle (of the play) as well. It is not a matter of simple condemnation of Orestes and rehabilitation of Aegisthus (who is loved by no one),40 and yet the troubling aspects of the scene cannot be discounted altogether. In itself, this is above all a suspenseful narrative, a deliciously crafted romp filled with many tense moments, visually effective shots, and the constant manipulation of audience expectations.41 But, seen in the context of the play as a whole, it marks a first step in a gradual development towards the painful dénouement of the matricide, which is—in this respect, the view summarized above is, I think, exactly right—‘the true crux of the play’.

40

The fact that Aegisthus is offered burial at the end of the play (1276–7) has little to do with sympathy and more with breaking the cycle of violence, part of the comprehensive resolution offered by the Dioscuri. 41 This maps in part onto Heath’s (1987: 11) description of the tragedian’s task: ‘to evoke an emotional response from his audience, while the audience for their part value his work because of the pleasure that accompanies such emotional excitation under the controlled conditions of a theatrical fiction’. Heath has since developed his thinking (e.g. in Heath 1999, 2006). All this is not to claim, of course, that the Electra has no ‘meaning’ beyond that of hedonistic entertainment, nor that Orestes is a character without problems (nor, again, is Electra simply an accumulation of deplorable characters): the play is a ‘true’ tragedy, the tragic heart of which is the matricide and its powerful consequences.

VI The Language of Rhetoric The agōn Revisited

1. INTRODUCTION If the previous chapter showed that such problems as we may detect in Orestes’ killing of Aegisthus are moderated by the pace and suspense of a thrilling messenger narrative, the present chapter aims to show that, during the next confrontation between a conspirator and a victim—the agōn between Electra and Clytemnestra (998–1146)—full attention is called to complexities in the younger woman’s character that have become increasingly evident throughout the play and are now given prominence immediately before the matricide. If, moreover, my analysis of the messenger scene showed Euripides to be a masterful craftsman of narratives—a virtuoso with resources of narrative composition inherited from a tradition that goes back to Homeric epic—this chapter will hopefully show him as an equally brilliant logographer, adopting and modulating the tropes and techniques of rhetoric that became popular in fifth-century Athens and crafting them into rhetorical set pieces supremely suitable to the characters who utter them. Thus this chapter is meant to contribute to our understanding not only of this particular scene and the characters in it but also of Euripides’ use of rhetoric more generally.1 1 For the intellectual background underlying Euripides’ employment of rhetoric, specifically in his agōnes, see Lloyd 1992: 19–36. Lloyd’s treatment of this particular agōn scene (1992: 55–70) has informed much of my discussion in this chapter, as have those of Michelini (1987: 217–23), Dubischar (2001: passim, but esp. 322–33), Mossman (2001: 380–3), and Cropp ad locc. The scene has received comparatively

The Language of Rhetoric: The agōn Revisited

249

I have already discussed in chapter II (§5.3) how various moments throughout the extended agōn scene (998–1146) illuminate the relationship between Electra and her mother and how the content of the speeches and the surrounding dialogues reflects certain aspects of their characterization. The focus in this chapter is different and more narrow: I am interested here especially in the internal organization of the two set piece speeches (1011–50, 1060–99) and in the specific oratorical devices that the two speakers use. The aim is, once again, to show how Euripides carefully modulates the language of his speakers to suit their background, their aims, and their preoccupations. The nuances of the two rhēseis are best brought out through comparison, and I therefore propose to analyse them side by side. In order to isolate the most striking similarities and differences between the two speeches, I will treat one pars orationis at a time, following the traditional division of (forensic) orations into exordium (προοίμιον), narratio (διήγησις), argumentatio/probatio (πίστις), and peroratio (ἐπίλογος).2 My divisions of the speeches into sections corresponding to these categories will be based on a variety of linguistic and non-linguistic criteria (position in the speech, tense usage, particles, content, etc.). Very roughly speaking, the opening part of a speech will normally be the exordium; a narratio should be marked by the use of narrative tenses (aorist, imperfect, pluperfect, and historic present) and will be found somewhere after the beginning of a speech; argumentationes will similarly be found ‘in the middle’ but marked by non-narrative tense usage; and perorationes close speeches. This is a vastly oversimplified account, of course, and modifications will be introduced as the chapter progresses. I should emphasize that my organization is not meant to suggest that either speech in the agōn of Electra is neatly structured along such traditional lines (or, for that matter, that speeches in this period, whether in tragedy or elsewhere, were normally so structured).3 In fact it will become clear that Electra’s oration, in particular, is little (sustained) attention elsewhere. The role that rhetoric plays in Euripidean drama, on the other hand, is a massive topic: for a valuable discussion with further bibliography, see Mastronarde 2010: ch. 6. For agōn scenes, see also Rutherford 2012: 190–200, with further references in n. 57. 2 For the traditional four-part division of speeches, see de Brauw 2007, with further references. 3 The traditional division may not have been established until the fourth century; see Schiappa 1999 and de Brauw 2007: 187–91.

250

The Language of Rhetoric: The agōn Revisited

hard to analyse as an orderly sequence of discrete parts—and I hope to show that the blurring of structural lines in her speech has a deliberate effect, while some irregularity in the structure of Clytemnestra’s speech, too, can be seen to serve rhetorical purposes.

2. EXORDIUM

2.1. Clytemnestra The first two lines of Clytemnestra’s agōn speech (1011–12) serve as a headline: her defence will rest on an indictment of Agamemnon and his behaviour towards his family. By selecting this as her topic, Clytemnestra makes clear from the outset that she will meet charges relating to the murder of her husband (including the charge of adultery). These charges, however, accord only in part with Electra’s complaint immediately before the speech (1008–10): . . . αἰχμάλωτόν τοί μ᾽ ἀπῴκισας δόμων, ᾑρημένων δὲ δωμάτων ᾑρήμεθα, . . . πατρὸς ὀρφανοὶ λελειμμένοι. You have, after all, displaced me from my home as a prisoner: my home captured, I am in captivity, and left deprived of my father.

Electra, as we will see, will devote an important part of her speech to her exiled condition; but Clytemnestra’s case will deal only with the death of Agamemnon. The vague τοιαῦτα (1011), which Clytemnestra uses to refer to Electra’s accusation against her, allows her to pick and choose in this fashion; at the end of the speech, too, she will attempt to limit the topic under discussion (see §5.1). Euripides signals that an exchange of set speeches is about to begin by having Clytemnestra make a formal announcement of her oration, λέξω δέ (1013, ‘I will make my case’). Such first-person announcements of intent with the future tense are found frequently in Attic oratory, often to introduce a new pars orationis; the verb used is usually in the family of λέγω (‘speak’), as here, διδάσκω (‘explain’), or δείκνυμι (‘demonstrate’). In the light of the preceding two lines, Clytemnestra’s announcement could be taken to indicate that she will delve straight into the details of Agamemnon’s crimes. But Clytemnestra delays her narratio and first discusses the inherent

The Language of Rhetoric: The agōn Revisited

251

disadvantages she faces in making her case (1013–17). This defensive captatio benevolentiae is a stock technique, entirely in accordance with good rhetorical practice4 but also particularly suited to Clytemnestra, who is battling with more disadvantages than she knows.5

2.2. Electra Having rounded off her proem in this fashion, Clytemnestra embarks at 1018 on her account of the past. Her speech thus gives the impression of being an organized piece of rhetoric from the start, one that uses established oratorical techniques and progresses through a clear structure. The beginning of Electra’s speech, on the other hand, has a much more impenetrable organization. This is perhaps surprising in view of her opening: like Clytemnestra, she begins with a formal announcement; and she continues with even more ‘signposting’, which is suggestive of a highly structured speech (1060): λέγοιμ᾽ ἄν· ἀρχὴ δ᾽ ἥδε μοι προοιμίου· I will speak, then. And this is the beginning of my introduction: . . .

Yet, even though Electra uses technical terminology (προοίμιον) to indicate her rhetorical organization, the exordium is not a clearly delineated section of her speech, and it is not easy to see which lines constitute the ἀρχή and which the προοίμιον as a whole. Scholars have accordingly struggled to determine exactly where Electra’s exordium ends. Denniston ad loc. argues that it lasts until 1085, covering the entire sketch of Clytemnestra’s character. Diggle (1969: 53–5) took issue—rightly, I think—with an exordium taking up two thirds of a speech, but this led him to suggest altering προοιμίου to παρρησίας in

4 For this kind of captatio benevolentiae, see [Arist.] Rh. Al. 1442a–c, Arist. Rh. 1415a–b, and Lloyd 1992: 26–7. Other Euripidean examples include Andr. 186–91, Tro. 914–17, and Or. 544–50. The particle used to introduce the captatio, καίτοι, is found in similar cases in Attic oratory (e.g. Andoc. 1.72, Dem. 58.7) and is ideally suited to introduce a constraint that prevents the speaker from proceeding, before overcoming that constraint (in the δέ clause, 1015–17). For the function of the particle as a ‘reverse denial of expectation’, which is to introduce a point whose implications seem to clash with the preceding discursive move (in this case, λέξω δέ), see Slings 1997a. 5 See my discussion in ch. II, §5.3.2.

252

The Language of Rhetoric: The agōn Revisited

his Oxford Classical Text (app. crit.)—a needless conjecture.6 Collard (1984: 13) suggested that ἀρχή . . . προοιμίου should be taken as a sort of tautology: ‘This is my beginning (ἁρχή) by way of introduction (προοιμίου)’. But this strikes me as an unnatural reading of the genitive προοιμίου, and the notion that all of 1060–85 constitutes Electra’s proem as infelicitous. Lloyd is, I think, very close to the truth when he says that ‘the concise insult in 1061 is the beginning of a proem which extends until 1068’ and that ‘[t]he rest of the speech is introduced by ἥτις (1069)’, an instance of ‘the more general use of ὅστις to introduce a proof ’ (Lloyd 1992: 66, followed by Cropp on 1060). But perhaps the crux is that, in spite of Electra’s suggestion of a clear organization, her speech in fact lacks such structure. Rather than seeing a clear break at 1069, I would suggest that her exordium there (or perhaps a few lines earlier) ‘bleeds into’ the central section of her speech, which itself lacks organization in interesting ways (more on this in §3.2). Thus Mossman (2001: 381) hits the nail on its head when she remarks that Electra’s speech ‘carefully gives the impression of being not very structured’: this suggests perhaps an emotional and spontaneous speech, and we will certainly see Electra’s temper flare up later on. If we take 1060–8, then, as ‘more or less’ Electra’s exordium, it consists of a neat insult (1061), followed by a telling comparison with Helen (1062–8), which in turn contains Electra’s main accusation against her mother (1066–8). The counterfactual wish at 1061 (εἶθ᾽ εἶχες . . . βελτίους φρένας, ‘would that you had better sense’) conveys exactly what the line of attack will be: Electra will go after her mother’s character and ‘sense’ (her φρένες), rather than defending Agamemnon, as Clytemnestra has asked her to do. It is, indeed, Clytemnestra’s character that Electra next expands upon: she is pretty, like her sister, but bad (the two are ‘vain’, ματαίω 1064), and unworthy of their brother Castor.7 The main charges against both women are presented in 1065–8: Helen was a shameless adulteress, Clytemnestra killed ‘the foremost man of Greece’ (ἄνδρ᾽ ἄριστον Ἑλλάδος 1066)—and it was not because of Iphigenia. This last accusation also serves as a headline for the rest of the speech: Electra will Kovacs’ text in the Loeb, εὐχὴ δ’ ἥδε μοι προοίμιον (‘the following wish shall be my introduction’, referring to the unrealizable wish in 1061) is very neat, but I would maintain that no emendation is necessary. 7 For the significance of Castor at this point, see ch. II, §5.3.4. 6

The Language of Rhetoric: The agōn Revisited

253

pierce through her mother’s ‘pretext’ (σκῆψιν 1067, an effective word) and show that she was not at all inspired by Iphigenia’s death and was as much of an adulteress as her sister. Thus we have already seen some telling differences between the two women’s speeches: both begin with a headline that indicates their overall topic, but, whereas in Clytemnestra’s case that topic is the causes of her actions, in other words the events leading up to Agamemnon’s death, Electra for her part will zoom in above all on her mother’s character. The two speeches also reveal divergent styles: Clytemnestra’s comes across as organized and sophisticated, while Electra’s leaves the impression of being more unstructured, and thereby more spontaneous.

3. NARRATIO Some of the most interesting and significant differences between the two women’s speeches may be found in their narrationes. Here as in chapter V, paying close attention to Euripides’ employment of verbal tenses will prove to be highly instructive: such an analysis will reveal that Clytemnestra and Electra, in describing (their interpretation of) the past, have not only divergent views but also divergent styles and strategies, which in turn suggest much about their characterization. Neither woman’s speech has any section that can count as a continuous and clearly distinguishable narratio. Rather, as is common in early Attic oratory (see de Brauw 2007: 189), narratio and argumentatio are mixed in a long central part of the speech: both speakers alternate their reporting of events with a rhetorical ‘interpretation’ of those events.

3.1. Clytemnestra To begin, again, with Clytemnestra’s rhēsis, the central part of this speech (1018–45) is just such a mix of narrative and argumentative text. Clytemnestra begins by recounting the events leading up to Iphigenia’s sacrifice (1018–23: narrative) and subsequently discusses the conditions under which such a sacrifice would be excusable— conditions that did not hold (1024–9: argument); next she relates her

254

The Language of Rhetoric: The agōn Revisited

initially benign reaction and her change of heart when faced with Cassandra (1030–4: narrative) and goes on to paint her resulting adultery as a general behavioural tendency of women (1035–40: argument); a separate argument follows that returns to the issue of Iphigenia (1041–5), before Clytemnestra ends with her peroration. The narrative portions of Clytemnestra’s speech (1018–23 and 1030–4) are carefully constructed mini stories. A series of main verbs in the aorist marks the main events in her tale, which is a summary of her marriage to Agamemnon and of how it went downhill: their betrothal (ἔδωκε 1018), the killing of Iphigenia (διήμησ(ε) 1023), and, as the proverbial last straw, the concubinage with Cassandra (ἦλθ(ε) 1032, ἐπεισέφρηκε 1033).8 This basic narrative structure is elaborated upon through various additions, with which Clytemnestra frames the main events and steers their interpretation. First, Clytemnestra invokes a prenuptial condition of the marriage via the suggestive result clause οὐχ ὤστε θνῄσκειν οὐδ’ ἃ γεναίμην ἐγώ (1019, ‘not so that I or my offspring would die’): her marriage was not supposed to be of the kind that leads to her own death, or that of her children;9 next, a participle phrase and an imperfect provide the damning background to Agamemnon’s sacrifice of Iphigenia, κεῖνος . . . παῖδα τὴν ἐμὴν . . . πείσας ᾤχετ(ο) . . . ἄγων . . . Αὖλιν (1020–2, ‘he persuaded my daughter and brought her to Aulis’).10 Clytemnestra’s vivid imagery in this sequence, particularly her bold use of the word παρήϊς (1023, ‘cheek’),11 is designed to add to the severity of Agamemnon’s guilt. After a stretch of argument, Clytemnestra resumes her narrative at 1030 and, again, frames Agamemnon’s actions against a background that is meant to make them seem more damning. The frame is built with the help of another participle and another imperfect: ἐπὶ τοῖσδε . . . καίπερ ἠδικημένη | οὐκ ἠγριώμην (1030–1, ‘even though I had been wronged, I was not angry’). 8 ἐπεισέφρηκε (‘he imported’), so found in the MSS, is an aorist and needs no emendation (-φρηκε presumably derives from -προηκε (προίημι), rapidly pronounced; these and other forms were then regularized as coming from the verb that LSJ lists as ἐπεισφρέω): see Slings 1997c: 159. 9 For this value of result clauses with the infinitive (an action ‘of the sort that leads to’), see Rijksbaron 2002: §23. 10 Of course, there is no aorist equivalent form for ᾤχετο (at least not of the same verb), so that not too much significance should be accorded to the fact that it is imperfect. 11 For the use of παρήϊς here (difficult, but not to be emended), see Kells 1966: 53–4 and Maxwell-Stuart 1975.

The Language of Rhetoric: The agōn Revisited

255

The last few ‘main events’ that we might have expected to be expressed by aorists, Clytemnestra’s adultery and her killing of Agamemnon, are not explicitly mentioned until her concluding peroration (ἔκτειν(α), ἐτρέφθην . . . πρὸς τοὺς ἐκείνῳ πολεμίους 1046–7). Only implicit hints at these actions are given in the narrative sections, particularly the second (1030–4), which has the suggestive counterfactual οὐδ᾽ ἂν ἔκτανον πόσιν (1031, ‘nor would I have killed my husband’) and the somewhat menacing imperfect κατεῖχ’ (1034, ‘he tried to keep two wives’). The context forces on that imperfect a conative flavour (Agamemnon’s attempt obviously did not end well); in terms of narrative structure, the use of the imperfect suggests that κατεῖχ’ was not the end of the story, but Clytemnestra leaves the actual end unexpressed, for now at least. The argumentative sections in 1035–40 (adultery) and 1041–5 (killing) similarly give implicit hints. Yet it is surely significant that these events are absent from the narratives proper. On the whole, Clytemnestra’s narrative portions are highly appropriate for a defence speech: she relates the principal causes of the actions of which she stands accused, leaving those actions themselves implicit until the end and filtering the course of events through her own interpretative lens. The structure of her narrative—aorists alternating with imperfects and subordinate constructions—shows that she is interested in causes and effects, in events as they unfolded (particularly those that discredit Agamemnon), and in placing them in their proper context (which makes them even more damning).

3.2. Electra At a more abstract level, Clytemnestra’s narration consists of fairly ‘typical’ narrative text, in that we find there an alternation of tenses in which aorists play a central role by marking the events that push the story forward.12 Once Electra’s rebuttal speech is seen in this light, something very interesting emerges about it: if a narrative with aorists is the norm, Electra’s narrative portions are highly atypical indeed. This unusual nature of Electra’s narratio may explain why, again, it is not easy to pin down precisely which lines of her speech fall under the header; it is, in fact, possible to argue that Electra’s rhēsis has no 12

See my analysis of the aorists in the messenger speech in the previous chapter.

256

The Language of Rhetoric: The agōn Revisited

distinguishable narrative section at all. Yet if, as argued above (§2.2), her prooimion ‘bleeds into’ a new section at 1068–9, we may identify a central segment of her speech running from 1069 to 1085, which oscillates (like Clytemnestra’s speech) between narrative and argument. It begins with Clytemnestra preening her locks in the mirror (1069–71: narrative), followed by a generalizing comment on such behaviour (1072–5: argument); the next section recounts Clytemnestra’s emotions on hearing news from the front (1076–9: narrative), followed by reasons why Clytemnestra did not need, and ought not, to have behaved so treacherously (1080–5: argument, again capped by a general reflection in 1084–5). The narrative portions thus isolated cover lines 1069–71 and 1076–9. Yet, if we take a closer look at these lines, it turns out that Electra’s narrative is very different from Clytemnestra’s, particularly with respect to her use of tenses. In all of these lines there is only one finite verb referring to the past: the imperfect ἐξήσκεις (1071, ‘you were cultivating’). This verb is the main predicate in the first narrative section, yet in terms of syntax, strictly speaking, it is not even a ‘main verb’, as it heads a relative clause (introduced by ἥτις, 1069), elaborating on σ᾽13 in the preceding line (οὐ γὰρ σ᾽ ὡς ἔγωγ᾽ ἴσασιν εὖ, ‘for people do now know you as well as I do’).14 What all this suggests is, first, that Electra is not relating a sequence of events (there are no aorists to indicate the main points in a plot) but describing a repeatedly occurring scene in the palace and, second, that Electra takes care to present her sketch as based on personal, first-hand knowledge (1068). This is consistent with the way Electra headlined her speech in her exordium (see §2.2): there she focused on Clytemnestra’s nature rather than on her actions, and this first description, too, is more concerned with giving a polemical character sketch of Dobree’s σ᾽ in 1068 (absent from the MSS) seems to me required, though more than one of the locations that have been suggested for its insertion in the line (see Diggle or Basta Donzelli, app. crit.) would be possible. 14 The syntactic connection between the two phrases is relatively weak, however: because the syntax of οὐ . . . εὖ (1068) is complete in itself, because the relative pronoun begins a new line (and a new section of the speech), and because the relative clause is of the ‘digressive’ (or non‑restrictive) kind (see Rijksbaron 2002: §28), lines 1069–71 function more as an independent main clause than as a subordinate clause. Indeed, editors rightly print a full stop after 1068, to indicate that ἥτις begins a new sentence (‘relative connection’). Still, as I will argue, it is not completely without significance that Electra ‘subordinates’ her narrative material to an expression of authoritative knowledge (οὐ . . . σ᾽ ὡς ἔγωγ᾽ ἴσασιν εὖ). 13

The Language of Rhetoric: The agōn Revisited

257

Clytemnestra (through the imperfect ἐξήσκεις, which paints a picture of her repeated behaviour) than with relating actions and events. After Electra’s generalization at 1072–5, the second brief narrative section, 1076–9, continues in the same vein. Here there are no finite verbs referring to the past. Instead we find the participles κεχαρμένην (1077) and συννέφουσαν (1078) in an accusative and participle construction governed by οἶδ᾽ ἐγώ (1076). The tense and the aspect of these participles are significant: both the perfect and the present participle express repeated actions, as is clear especially from the subordinate conditional clauses with iterative optatives (εἰ . . . εὐτυχοῖ 1077, εἰ . . . εἴη 1078: ‘And I know that you, alone among all Greek women, were overjoyed whenever things fared well for the Trojans, but wore a gloomy look if they fared less well’).15 Again, we may note the absence of any verbal forms in the aorist: Electra is concerned not with relating a sequence of events, but with characterizing Clytemnestra’s repeated behaviour in the palace. Also, by subordinating the ‘narrative’ verb forms to a form of οἶδα (‘know’), Electra’s syntax is designed, as before, to present her sketch as an authoritative depiction of first-hand knowledge. Both Clytemnestra’s and Electra’s narrationes thus take us back to the time around Iphigenia’s death and to the Trojan War, but the nature of their narratives is very different. To use a modern analogy, if these sections were made into films, Clytemnestra’s would have a chronologically progressing plot and would feature ‘action sequences’, whereas Electra’s would consist only of a montage sequence with no significant progression through time. This difference is indicative of the divergent rhetorical strategies that the women adopt in the agōn: Clytemnestra’s narratio, as we have seen (§3.1), is highly suited to a defence speech intended to justify the murder of Agamemnon and her adultery. But Electra, for her part, is not interested in discussing the events that led up to Agamemnon’s murder (in spite of Clytemnestra’s insistence that Electra should concern herself with the same material as she herself had done, 1049–50); instead she chooses to rebut Clytemnestra’s points

15 For this use of the optative, see, e.g. KG II §560.1 and Rijksbaron 2002: §24.4.2. Had Electra expressed in a main clause the actions denoted by the participles here, the corresponding verbs would be in the pluperfect and imperfect, respectively. The perfect κεχαρμένην is of the ‘intensive’ kind, expressing a ‘stronger’ form of the state of ‘being happy’ than the present χαίρουσαν would; for this use of the perfect, see, e.g. Rijksbaron 2002: §10.2, with some modifications in García Ramón 2006.

258

The Language of Rhetoric: The agōn Revisited

implicitly, by way of character assassination, on the basis of her personal experience of that character. This first-hand knowledge, Electra suggests, makes her impervious to Clytemnestra’s ‘pretence’ (σκῆψιν 1067) about retributive justice— an argument that, in this scene, has almost persuaded the Chorus (δίκαι᾽ ἔλεξας 1051, ‘there is justice in what you say’). And indeed, much of what Electra aims to do in her brief narrative sections is to expose incongruities between Clytemnestra’s public face and her private demeanour: though Clytemnestra presents her adultery to the outside world as the result of Agamemnon’s actions, Electra ‘knows’ that its origins are to be found in character traits that she displayed earlier (1069–71); and though outwardly a loyal Greek wife,16 Clytemnestra was secretly anything but loyal to the Greek cause (1076–9). Many of the traits of Electra’s narrative sections discussed here— the absence of aorists, the focus on private behaviour—may have struck a familiar note in the audience, since everyone has heard Electra give a very similar speech over Aegisthus’ corpse. The parallels between these two speeches—noticed, but not treated in detail in scholarship on the play17—will be discussed shortly, after we take a look at the argumentative parts of the women’s speeches (which, as this section has hopefully made clear, are closely integrated with the narrative parts).

4. ARGUMENTATIO Each of the narrative sections discussed in §3 is followed by portions of argumentative text, and in the way Clytemnestra and Electra phrase these arguments they have a markedly different approach. The sections I have identified as ‘argument’ are, for Clytemnestra, 1024–9, 1035–40 and 1041–5 and, for Electra, 1072–5 and 1080–5. I will also treat Electra’s sequence of rhetorical questions in 1086–93 under this header, though it is not easy to classify these lines (an intentional feature, as 16 Clytemnestra’s professed loyalty to the Greek cause is nowhere mentioned explicitly in this play but is implicit in Electra’s argument. Aeschylus’ version of Clytemnestra goes to some lengths to project the image of a loyal Greek wife (see, e.g. Aesch. Ag. 587–97, 855–913). 17 See, e.g. Lloyd 1992: 68 and Mossman 2001: 381.

The Language of Rhetoric: The agōn Revisited

259

I will argue). Two aspects of these argumentative portions are of particular interest: the hypothetical nature of many of Clytemnestra’s arguments; and both women’s use of generalizations.

4.1. Generalizations To begin with the latter aspect, as Mossman (2001: 381) points out, ‘[g]eneral reflections cluster in Clytemnestra’s speech when she is at her most defensive’, namely in the captatio benevolentiae (see §2.1), and in 1035–40, a section intended to ‘gloss over her adultery’. It is not by coincidence that these generalizations are, both, first-person sayings*, which serve to embed Clytemnestra’s own behaviour in ethical norms that she must hope are shared by her audience. And, considering the shape of that audience—Electra and the Chorus, both female—some sympathy might have been expected: both general reflections deal with women’s reputations and express an apologetic (even revisionist) view of women’s behaviour: women’s bad reputations should not automatically be believed, 1013–18; and, although women are a ‘foolish thing’ (μῶρον 1035), they cannot be held responsible if men make the first mistake (1035–40). But it is in this last respect that Clytemnestra’s generalizations fall on deaf ears: Electra has aligned herself with prevailing male standards before (see ch. II, §5.2) and will continue to do so throughout the agōn (this adds significance to her later remark that Electra is loyal to her father and to the masculine, 1102–4). Electra’s generalizations, on the other hand, are second-person sayings* that reinforce her accusations: Clytemnestra tended towards adultery even before the sacrifice of Iphigenia (there is no other explanation for her cosmetic extravagance, 1072–5), and her betrayal was all the more shameful in light of her sister’s (1084–5).18 And, whereas Clytemnestra’s gnōmai expressed a female-oriented morality, Electra substitutes her mother’s revisionism with a more ‘traditional’19 18 παράδειγμα in 1085 is presumably nominative (ἐστιν is to be understood, which gives τὰ γὰρ κακὰ | παράδειγμα τοῖς ἐσθλοῖσιν in 1084–5: ‘bad deeds serve as an example to good people’, or perhaps ‘as a pattern for good ones [deeds]’). Thus τ’ connects two clauses rather than two objects of ἔχει, which governs only εἴσοψιν (‘and hold attention’); see Cropp ad loc., and also Stieber 2011: 372–3. 19 See Denniston on 1035–40 (‘Greek morality [on adultery] was more severe on woman’s frailty than on man’s’); also Pomeroy 1976: 86–7 and Michelini 1987: 219.

260

The Language of Rhetoric: The agōn Revisited

view of female adultery: women left at home while the husband is away should behave properly.

4.2. Clytemnestra’s hypotheticals Thus the difference between the two women’s use of general reflections is indicative of their different approaches and viewpoints. The fact that Clytemnestra has to resort to innovative morality may also be taken as a sign that she is, as Cropp puts it (on 1011–50) making ‘the most of a weak case’. Cropp bases this assessment on the organization of Clytemnestra’s speech, but there are, I think, other signs which point in the same direction. Clytemnestra’s argumentatio leans heavily on ‘artificial’ proofs (ἔντεχνοι πίστεις).20 Although Euripides’ agōnes display similarly sophisticated techniques quite often, the density of such arguments in this particular speech is remarkable (see Lloyd 1992: 28–34). Clytemnestra’s case leans heavily on manipulated hypothetical situations, and in consequence we find a sustained series of counterfactual constructions (ἄν with secondary indicative). Two ‘hypothetical syllogisms’ begin this sequence:21 at 1024–9 she argues that Agamemnon’s sacrifice of Iphigenia ‘might have been forgivable’ (συγγνώστ᾽ ἂν ἦν 1026) under different circumstances; and at 1030 her narrative suggests that she ‘might not have killed’ her husband (οὐδ᾽ ἂν ἔκτανον πόσιν 1031), but for his introduction of Cassandra. In her peroration, too, Clytemnestra resorts to hypotheticals (1048, τίς ἂν . . . ἐκοινώνησέ μοι; ‘who would have shared with me?’). But the most elaborate instance is Clytemnestra᾽s wildly innovative reductio ad absurdum at 1041–5, the only example in Euripides of a ‘hypothetical role reversal’ (see Lloyd 1992: 32, 64): What if Menelaus had been abducted instead of Helen? (1041). Having posited this outlandish counterfactual, Clytemnestra attaches a series of rhetorical questions: ‘Would I have had to kill Orestes?’ (κτανεῖν μ’ Ὀρέστην χρῆν, 1042); ‘How would your father have taken that?’ (σὸς δὲ πῶς πατὴρ | ἠνέσχετ’ ἂν ταῦτ’; 1043–4). By forcing Electra (and the audience) to come up with answers, Clytemnestra’s rhetorical questions are designed to highlight the absurdity of Agamemnon’s 20 Also known as ‘invented’ or ‘artistic’ proofs: the distinction from πίστεις ἄτεχνοι (‘artless’ proofs, i.e. witness testimony, laws, oaths, etc.) is due to Aristotle (Rh. 1355b–6a). 21 For this type of proof, see Lloyd 1992: 32–3.

The Language of Rhetoric: The agōn Revisited

261

actual behaviour. Yet the distinct remoteness of the hypothetical (Menelaus’ abduction) does little to help Clytemnestra’s case. Taken together, Clytemnestra’s hypothetical arguments serve a complex function: they seem designed to show that Clytemnestra was not herself inclined to the murder, but forced into it by circumstances that could have been prevented by Agamemnon (thus the blame is shifted on him). But the sustained use of hypotheticals also makes it clear that Clytemnestra is straining to formulate a strong case. We see this also reflected in her constant focus on things that didn’t happen: Agamemnon didn’t treat the marriage the way it was originally intended by Tyndareus, he didn’t actually marry Iphigenia to Achilles as he suggested, he didn’t return from Troy in a reasonable fashion (sans concubine). It is easy to see, then, why Electra, for most of her speech, focuses on things that did happen—as we have seen, she stresses the fact that she herself saw them happen: the unadulterated facts of the case indict Clytemnestra. One final point may be made concerning the organization of Clytemnestra’s long central narrative–argumentative section. It is at first sight curious that she returns to the issue of Iphigenia at 1041 after she has discussed Cassandra and hinted at her own adultery (1030–40). Her highly rhetorical argument in 1041–5 disrupts the pattern of narratio followed by accompanying argumentatio, and thus seems to disturb the flow of the whole speech. For some, this was sufficient reason to suspect 1030–40 or to transpose 1041–5 before 1030,22 but Cropp (on 1011–50) has rightly pointed out that such interventions are misguided: the way in which Clytemnestra introduces Aegisthus into her speech glosses over their adultery and obscures its timing: did it begin with Cassandra (1036–8, unspecific) or with Iphigenia (1046–8)? This design is spoiled if 1030–40 or 1041–5 are deleted, or if 1041–5 is placed before 1030 so as to combine the arguments about Iphigenia. (Cropp ad loc.)

Apart from leaving her alliance with Aegisthus very vague, the structure of Clytemnestra’s rhetoric allows her to have her cake and eat it. 22 The whole of 1030–40 was deleted by Vitelli and Wecklein, and Wilamowitz deleted 1041–5; their arguments are summarized by Denniston (on 1030–48). One reason adduced for deleting 1030–40 is that Electra does not answer the point about Cassandra in her speech. But this is entirely appropriate, as Cropp (on 1060–96) saw: ‘[t]he embarrassing subject of Cassandra can be ignored’.

262

The Language of Rhetoric: The agōn Revisited

Her hypothetical οὐδ᾽ ἂν ἔκτανον (1031) briefly suggests that she was generous enough not to kill Agamemnon over the issue of Iphigenia alone; but then Cassandra is used only as a reason for Clytemnestra to defend her own adultery and Iphigenia is reintroduced as the driving motivation behind the murder, lest Clytemnestra be accused of killing her husband over a concubine.

4.3. Rhetoric and characterization Electra’s rhetorical tactics in her narrative and argumentative sections, discussed in §3.2 and 4.1, find a direct parallel in her speech over Aegisthus’ corpse (discussed in ch. II, §5.2). There, too, Electra built her speech around imperfects which served primarily to describe the typical behaviour of her intended target, and she used generalizations to back up her points. There, too, she was concerned with Aegisthus’ behaviour in private, including his sexual conduct (945–8). The fact that her rhetorical style is so similar across the two speeches suggests that Euripides is doing more in the present scene than drafting the best orations he can come up with for his characters in this particular context. These are not detached rhetorical exercises governed merely by ‘the rhetoric of the situation’;23 rather they tell us much about the women who deliver them them. Indeed, the style and organization of both speeches do much to characterize their speakers. In Clytemnestra’s case, concern for her public reputation—which is twice the focus of general reflections of a defensive nature—characterizes her as female24 but also confirms earlier remarks about exactly this aspect of her disposition: the Peasant in his prologue has commented on her avoidance of φθόνος (30, ‘resentment’), and the Old Man has similarly described her fear of popular condemnation (ψόγον τρέμουσα δημοτῶν 643, ‘fearing the people’s censure’).25 It is also integral to Clytemnestra’s characterization—particularly by contrast with that of her daughter—that she focuses in the central part of her speech on (failed) marriage alliances: between herself and Agamemnon, between Iphigenia and Achilles, and between Helen 23 Using Dale’s (1954: xxv) famous phrase. Dale’s approach to rhetoric in Euripidean tragedy is discussed fairly in Conacher 1981; see also Mastronarde 2010: ch. 6. 24 25 See Mossman 2001: 377, with n. 12. See Lloyd 1992: 61.

The Language of Rhetoric: The agōn Revisited

263

and Menelaus. As Mossman has pointed out, this orientation is indicative of Clytemnestra’s loyalties: It makes her move from thinking in terms of vertical, cross‑generation alliances . . . which had the potential to benefit the oikos . . . to horizontal, matrimonial alliances, which are dysfunctional and damage the oikos . . . This contrasts . . . her with Electra, who, as we have seen, after a confused period when she has been unable to define herself fully in either vertical or horizontal terms, has recently redefined herself in terms of her blood family. (Mossman 2001: 380)

Indeed, Electra’s main grievance, as we will soon see, concerns thwarted vertical relations within the family. As for her own characterization in the first half of her speech (1060–86), just as in the kakology (see ch. II, §5.2), Electra’s focus on palace life and private (particularly sexual) issues reflects both her gender and her personal grievances (she was expelled from the palace and cut off from normal marital life). Electra’s use of second-person gnōmai* to rub in her accusations of misbehaviour also adds to her characterization, showing the audience how she lays claim to the moral high ground and associates herself with standards of proper behaviour. Yet the part of Electra’s speech that is most revealing of her emotions and motivations comes after any discussion of Clytemnestra’s behaviour during the Trojan War, in lines 1086–93. ‘After’ in a chronological sense too: these lines deal with events that occurred when that period had reached its conclusion, that is, after Agamemnon’s death—and in this way Electra refuses to engage with the case as it was explicitly defined by her mother (1049–50) and ends up arguing a different one altogether, namely her own and that of her brother. In fact, as Electra moves to this new part of her speech (δ᾽ 1086), the first thing she does is to demote Clytemnestra’s entire case to little more than a side issue. Electra’s conditional in 1086, εἰ δ᾽, ὡς λέγεις, σὴν θυγατέρ᾽ ἔκτεινεν πατήρ (‘if, as you say, my father killed your daughter’), concedes to Clytemnestra her main claim for the sake of argument,26 and thus suggests that Electra’s case remains intact regardless of whether or not Clytemnestra’s accusations are accepted. 26 The relationship between the conditional clause in 1086 and its main clause is ‘propositional’, meaning that the εἰ clause specifies a condition for the truth or relevance of the statement in the main clause (cf. e.g. ‘If I am not mistaken, he is in Athens now’) rather than a condition for the realization of the action expressed in the main clause (‘predicational’ conditions: e.g. ‘If it rains, he will take an umbrella’): for

264

The Language of Rhetoric: The agōn Revisited

The phrasing of this conditional clause, a ‘fleeting and evasive reference’ to Iphigenia’s killing (Lloyd 1992: 68), is dense with ingenious points of detail: the parenthetical ὡς λέγεις suggests that Electra remains unconvinced of the truth of Clytemnestra’s claim, which is in any case left open through the use of a conditional; with σὴν θυγατέρ᾽ she distances herself from her mother even further by describing Iphigenia only as her daughter (she forgets for a moment that Iphigenia was her own sister, and thus she glosses over any injury suffered by herself as a result of Agamemnon’s actions).27 Electra moves on to what is, for her, the heart of the matter: the vertical relationships that Clytemnestra lost sight of in her defence speech. These return now to prominence, as Electra blames her mother above all for ignoring her obligations to her offspring and for disrupting the orderly transference of patrimony from Agamemnon to his children by ‘buying’ another husband (1090). The shift from attacking Clytemnestra’s character to talking about Electra’s own situation is emphasized by the word order of 1087, with ἐγώ standing before the postponed interrogative τί. As Dik (2007: 157–66) has shown, this ordering lends prominence to ἐγώ in its pragmatic role as topic (or perhaps ‘theme’);28 Electra thus ensures that this question and the next few are seen to be about her. Paraphrasing, the sense is something like: ‘If what you say is right, what about me? What did I do to you to deserve such treatment, and Orestes?’ Electra’s complaints take the form of a series of angry rhetorical questions, strung together in a sequence that gives the impression of a breathless litany (as discussed in ch. II, §5.3.4). This form makes it difficult to say whether this section qualifies as narratio or as argumentatio, or even belongs already to Electra’s peroration. She touches on actions and events in the past that have not been mentioned in the speech before, but this is not an ‘account’ of those actions (and there

this distinction, see Wakker 1994: ch. 5. Here, of course, the main clause is a (rhetorical) question, and the subordinate clause presumably has a concessive flavour (even if the default concessive καὶ εἰ/εἰ καί is not used. There is no need to think that Electra questions whether Agamemnon actually slew Iphigenia: the conditional clause is rather a rhetorical device meant to cast some doubt on the relevance of Clytemnestra’s argument. 27 She will call Iphigenia ‘sister’ at 1093 (ἀδελφῆς), but only in noting that her own fate is twice as bad as hers. 28 For ‘themes’ (and ‘left dislocation’), see ch. II, n. 17.

The Language of Rhetoric: The agōn Revisited

265

is no indication of a narrative structure or an alternation of narrative tenses). And, although some sophisticated reasoning is implicit in Electra’s complaints (if Clytemnestra applied her notions of retributive justice evenly, Electra and her brother should not be in exile and Aegisthus should be punished; similar logic is used in her peroratio),29 this section comes across most of all as an emotional, accusatory outburst, not primarily meant to fulfil a specific role in the formal structure of the speech. Again, this blurring of structural lines seems designed to give the speech a spontaneous and emotive feel, which— considering its subject matter—is a striking touch of character: Electra is set off by the topic of her own and her brother’s treatment. A hint of insidious irony caps Electra’s sequence of rhetorical questions (1092–3): [πῶς οὐ πόσις] ἀντ᾽ ἐμοῦ τέθνηκε, δὶς τόσως ἐμὲ κτείνας ἀδελφῆς ζῶσαν; Why is your husband not dead in retribution for me, since he has inflicted on me twice the death in life that my sister suffered?

Aegisthus lies dead inside the house, of course. This vicious touch seamlessly leads into Electra’s peroration, which will exploit a similar irony, but this time with regard to the death to come—Clytemnestra’s.

5. PERORATIO

5.1. Clytemnestra Clytemnestra begins her peroration (1046–50) by finally making explicit, as a grim and inevitable conclusion, the charges against her as she sees them: the murder of Agamemnon (ἔκτειν᾽ 1046, ‘I killed him’) and her alliance with Aegisthus (ἐτρέφθην ἥνπερ ἦν πορεύσιμον | πρὸς τοὺς ἐκείνῳ πολεμίους 1046–7, ‘I took the path that was viable, turning to his enemies’). The one-word statement ἔκτειν᾽ almost emblematizes Clytemnestra’s point, that she had no other It is thus not entirely correct to say that ‘there is no real logical link between the first and second halves of 1093’ (Mossman 2001: 382), though I agree with Mossman that Electra’s ‘announcement’ that the lex talionis will result in Clytemnestra’s death (1093–5) comes about abruptly, and that this adds to the effect (see §5.2). 29

266

The Language of Rhetoric: The agōn Revisited

options left:30 this point is then repeated in the metaphor of a path that is available or can be ‘passed through’ (πορεύσιμον). The other inevitable conclusion, that that path led to Aegisthus, follows in the next line. Clytemnestra’s subsequent hypothetical question (τίς ἂν . . . ἐκοινώνησέ μοι; 1048) is one final suggestion that alternative courses of action were not open to her, and it portrays her adultery as merely a means to an end: killing Agamemnon. Clytemnestra’s speech thus closes with a brief summation of her own view of past events, and this is borne out by the use of two simple concluding aorists, ἔκτεινα and ἐτρέφθην (1046). What Clytemnestra, unaware that her life is at risk, now expects from her daughter is a similar evaluation of past actions. Her challenge to Electra (1049–50) is to show how anything about the fact that Agamemnon is dead can be held against her—a limitation of subject matter by which, as we have seen, Electra will not abide.

5.2. Electra In Electra’s own peroration (1093–6),31 which follows suddenly (and in mid-line) on her series of angry questions,32 the focus is not on a summary analysis of past actions but on a general view of justice and its specific consequences. Indeed, the way in which Euripides manipulates the general and the specific in these lines is remarkable. Electra presents the lex talionis as a starting point and applies the law to a specific case, following her own logic to its startling conclusion: if Clytemnestra’s view of justice (the concept is inherent in δικάζων, 1094)33 is to be upheld, she should die (and Electra would be the one to carry out the law, with her brother).34 The simplistic way in which Electra phrases this conclusion belies the significance of what she is saying and thus problematizes the whole notion of retributive 30 Asyndeton brackets the peroratio from the central section of Clytemnestra’s speech and is immediately followed by another asyndeton (ἔκτεινε, ἐτρέφθην . . . 1046): this is striking, and surely designed to enhance the abruptness of ἔκτειν’. 31 For lines 1097–9, see ch. II, n. 181. 32 ‘[A]s though Electra were impelled to say this out of anger at the situation described in the previous lines’, suggests Mossman (2001: 382), appealingly. 33 For the construction with δικάζω here (the syntax is difficult), see Kells 1960. 34 The future indicative in the conditional clause, ἀμείψεται (1093), is suited for such dire warnings: see my discussion of 336–7 in ch. II, §3.2 (with n. 104) and Wakker 1994: 167–8.

The Language of Rhetoric: The agōn Revisited

267

violence: if the application is so gruesome, can the general principle be right? The ‘neutral’ phrasing also allows for the deeply ironic use of the future indicative ἀποκτενῶ (1094): to Clytemnestra, this may sound merely like an analytical conclusion, as if Electra signals only where her logic leads her, rather than an actual intention.35 For the audience, of course, the meaning of ἀποκτενῶ is very different, and not at all analytical.

6. PERORATIO (II) The agōn in Electra is a ‘portrayal, through the medium of formal debate, of a central conflict in the play’ (Lloyd 1992: 69). I have tried to show in this chapter that this portrayal is subtle and layered: the two women’s speeches reflect their aims, motivations, and characters in various ways. Clytemnestra’s case is built wholly on proving that she had been wronged by Agamemnon and left with no other choice but to respond. To prove this, she recounts what happened in a way that paints Agamemnon in as bleak a light as possible; and she uses sophisticated rhetorical proofs to reject alternatives that might have resulted in a different outcome. In the process, Clytemnestra minimizes her alliance with Aegisthus and portrays it as a means to an end and a mere quid pro quo for Agamemnon’s own prior violation of the marriage (Cassandra). Apart from rhetorical strategy, Clytemnestra’s speech also reveals facets of her character: her concern for her public reputation and her innovative outlook on women’s role in society. In her speech, Electra refuses to engage with the very foundation of her mother’s case: she is not interested in discussing Agamemnon’s crimes, merely in treating those of her mother. Personal experience allows her to contradict Clytemnestra’s claims, and she proceeds to do so in a prolonged character assassination. Electra’s speech also shows us hints of the temper and preoccupations with which we have become familiar: her somewhat disorganized speech comes across as emotional and spontaneous, and this aspect is all the more pronounced in the most powerful part of the speech, 1086–93, where 35

For the interpretation of this future indicative, see ch. II, §5.3.4 (with n. 179).

268

The Language of Rhetoric: The agōn Revisited

Electra deals with what, for her, is the heart of the matter: her own exile and that of Orestes. It is, then, unsurprising that this agōn (like most others in Euripides, incidentally)36 ends without a clear resolution: Electra and Clytemnestra argue from different perspectives and in different ways, but, perhaps most importantly, they argue two entirely different cases to begin with.

36

See Collard 1975: 61–2 and Lloyd 1992: 15–16.

Conclusion Approaching Tragic Language

Considering the central role that language, in many different forms, plays in our discipline, and considering, too, that the science of linguistics has seen dramatic innovations and changes take place in the last few decades, it is remarkable that many classicists, when they seek to better understand an aspect of Latin or Greek, turn first to works dating back to the first half of the twentieth century, if not further back still. Admittedly, it is not easy to surpass works like Kühner’s grammars or Denniston’s Greek Particles: but, in the light of new and constantly progressing insights about language, we may ask whether those classic works should be the end point or the point of departure for our enquiries. The modern linguistic approaches described and adopted in this book offer more than terminological gimmickry: it is not, in other words, just a question of giving new names to familiar concepts. Rather, terms such as ‘speech act’ and ‘pragmatics’1 reveal a fundamentally different way of looking at language: instead of a collection of symbols that bear meaning according to certain rules, language is a social activity, a doing; it acquires its meaning in a process of communication between interactants, each with his/her own background and knowledge, aims, and desires. Especially in the first three chapters of this book, it has been my aim to show that precisely these aspects— which amount to what I have called linguistic (or conversational)

1 Though the term does not have a felicitous derivation, as Bakker (2010b: 151) points out: the discipline looks at the doing of language (πρᾶξις) rather than at the product of this doing (πρᾶγμα).

270

Conclusion: Approaching Tragic Language

behaviour—contribute to the characterization of the Peasant, Electra, and Orestes (and of the other characters) just as much as their ‘actions’ do. The theoretical approaches outlined in my Introduction in fact suggest that a distinction between words and deeds is bound to mislead: words are deeds, and conversation is a type of social interaction which provides, perhaps, more insight into what makes people tick than anything else. If I am right to think that modern linguistics can help us analyse the use of ancient Greek language in ways in which traditional descriptions of Greek cannot, then the study of Greek literature can be much enriched by insights developed in this field. I hope to have shown that, at least in the case of Greek tragedy, looking at language in this way can shed some light on all kinds of interpretative questions. I suspect that my analyses will also have made clear—implicitly or explicitly—some of the limitations inherent in my approach. Many of these limitations are due to the fact that this book was never meant to be a collection of linguistic observations but intended as an attempt to study a single work of literature and to say something about that work. Given that aim, some methodological techniques from modern linguistics quickly turn out to be much more useful than others (and I have come nowhere near to exhausting its store). Nor would I want to suggest that my approach is a ‘better’ or an ‘exclusive’ way to reach certain conclusions about the meaning of a play. Indeed, I began this book with the point that many of my findings would be similar (in kind and content) to those of other scholars, though I would reach them by a different route. In the end, the point is that linguistic analysis can and should aid and steer literary interpretation, not that it can function as a replacement for it. Four areas seem, to my mind, most promising for further work. First, the idea that speakers are characterized by their speaking styles has not been a very popular one in our discipline—or at least it has not given rise to much focused exploration. But this says more about a pervasive narrowness in the accepted definition of ‘style’ (a narrowness that can be overcome by adopting the perspective on language sketched above) and about the troubled history of the very concept of characterization in our field (see Introduction, §4.1.1) than about what, as I believe, we actually find in our texts. Especially in the case of tragedy (a genre in which, as far as our evidence suggests, the spoken word was of greater significance to the resolution of plots than any physical actions portrayed), it is hard to imagine how

Conclusion: Approaching Tragic Language

271

a playwright could affect characterization without modulating his characters’ speech—unless, of course, we want to give up on the notion of characterization altogether, which I do not.2 Second, I believe that more can be said about gendered language in Greek tragedy. There has been no lack of attention to this subject in recent decades (see Introduction, §2.4.1), yet I am confident that much is still to be discovered. In Electra, it is striking to find the extent to which the make-up of the group (mixed* versus single-sex*) seems to determine the linguistic behaviour of Electra, and I suggest that looking at other plays and texts in this light may yield similar (yet different, since at least in Euripides no two women are quite the same) findings. Third, I feel that general reflections in Greek tragedy deserve to be fundamentally re-evaluated and positioned, as I have tried to do throughout this book, squarely within their communicative context. Conservatism in textual criticism is, apparently, a vice of youth; yet I have never been satisfied with the knee-jerk reach for the interpolation scalpel in the case of Euripidean gnōmai that do not sit as well in their surrounding lines as some scholars think they should. Finally, in arguing, in my Introduction (§1.1), for the applicability to tragic language of modern linguistic theories designed to describe naturally occurring language, I claimed that the former (literary, stylized) kind of language could be seen as a more ‘concentrated’ version of the latter (everyday) one. A few tentative observations aside, I have not made a consistent effort in this book to describe how precisely such concentration works. In other words, there is still scope for work that investigates what it is that makes tragic language ‘tragic’,3 not only in terms of vocabulary, dialect, syntax, and metaphor but also in terms of conversational structure, pragmatics, politeness phenomena, and so on. Significantly, each of the three tragedians, but perhaps Euripides in particular, is also still in need of integrative, modern work describing his own stylistic range and peculiarities.4 2 I hope to further explore the issue of characterization through language in future work. 3 See Rutherford 2012: 22–8 for an overview of modern scholarship on this point; Rutherford’s book itself does much to further the effort. 4 Of course, there is significant scholarship on the style of the three tragedians (for references, see the overview in Rutherford 2012). But, as Rutherford points out, much of this work is atomistic rather than synthetic, and works of the latter type tend to be on Aeschylus and Sophocles rather than on Euripides—Aeschylus has to his

272

Conclusion: Approaching Tragic Language

Each of these projects would require a very careful analysis of Euripidean (or Aeschylean or Sophoclean) language. If I have shown that such an analysis could profit from the application of modern linguistic methodology, that shall be a useful conclusion in itself. If, on top of that, I have said anything worth saying about Electra—a difficult, but to my mind powerfully tragic play—then my job is done.

service, e.g. Bruhn 1899, Stanford 1942, Earp 1948, Rosenmeyer 1982: ch. 4–5, while Sophocles has, e.g. Earp 1944 and Moorhouse 1982 (as well as Budelmann 2000, de Jong and Rijksbaron 2006, Goldhill 2012). No equivalent book about Euripides (nothing like ‘The Language of Euripides’) seems to exist, although the contribution of works such as Barlow 1986b or Stieber 2011 is not to be discounted. None of the three tragedians has his own version of Willi 2003—the finest example of the kind of book I have in mind.

Glossary of Linguistic Terms The number in parenthesis after each headword represents the section in the Introduction where the concept in question is first discussed. Here as elsewhere in the book, cross-references to other headwords are marked by asterisk. In formulating these entries I have benefitted from Allott 2010 (an introduction to pragmatics organized by ‘key terms’) and from the glossary in Watts 2003. Adjacency pair (2.2): In CA*, an adjacency pair is an automated sequence of conversational turns* that consists of a first part* and a corresponding second part*, the first giving rise to the expectation that the second will follow. Examples are question–answer, request–acceptance/denial, greeting–greeting. A preference organization* determines which variety of second part is the more regular and unmarked response to a first part (e.g. an acceptance is the unmarked response to a request; a denial is the more marked variety). Pairs may be interrupted, preceded, or elaborated upon by expansions*. Such patterns may vary across languages and cultures. Bald (2.4.2): In politeness theory*, a FTA* that is performed without any redressive facework is characterized as ‘bald’. Bald FTAs may, but do not have to, give rise to interpretations of impoliteness. Conversation analysis (CA) (2.2): Also known as the study of ‘talk in interaction’. CA involves detailed examination of naturally produced samples of language in use, in order to catalogue and understand the repertoire of members of a speech community for the organization of talk. Topics include the system by which speakers divide speaking turns (turn-taking*), regularly occurring patterns of sequential turns (adjacency pairs*), and the ways in which speakers agree on and maintain or change a topic of conversation (topic management* or maintenance). CA is sometimes seen as part of pragmatics*. The theoretical vocabulary and method developed in CA has been applied to literary conversation as well. (Conversational) implicature (2.3.2): In Gricean pragmatics*, an implicature is what an utterance made by a speaker means beyond what it actually (‘literally’) says. Take the following conversation: Pylades: Would you like to go see a film tonight? Electra: I have a lot of work to do.

Here Electra clearly intends to communicate that she cannot accept Pylades’ invitation, even though such a meaning is not part of what the

274

Glossary of Linguistic Terms

words of her utterance mean. The communicated meaning has to be derived through a process of inference. In Grice’s work, such inferences depend on the cooperative principle* and on maxims of conversation*; other pragmatic theorists have proposed different systems (e.g. the principle of relevance* in relevance theory*). Cooperative principle (2.3.2): This is the notion, first proposed by H. P. Grice, that in conversation participants try to make their contributions suitable to the shared purpose of the ‘talk exchange’ that they are engaged in: speakers follow a set of maxims of conversation*, principles with which it is reasonable to comply in the pursuit of cooperation in communication. The cooperative principle also allows hearers to draw inferences: if a speaker seems to violate a maxim, the hearer will reinterpret that speaker’s utterance in a way that ensures that its meaning does comply with the principle (such an inference is called a conversational implicature*). The cooperative principle has been heavily criticized, for example by the proponents of relevance theory*. Dispreferred (turn, response, second part, etc.): See preference organization*. Expansion (2.2): In CA*, an adjacency pair* may be expanded through conversational turns or sequences that precede the first part* of the ‘base pair’ (pre-expansions*), intervene between the first part* and the second part* (insert expansions*), or follow the second part (post-expansions*). The expansions serve to support, facilitate, modify, and so on the base pair, often with a view to the prevention of dispreferred turns*. The following example features one instance of each type of expansion: Old Man: Peasant: Old Man: Peasant: Old Man: Peasant: Old Man: Peasant:

Could you help me? Of course. How much does that iPad cost? Pounds or euros? Pounds, please. It’s three-nine-nine, then. Sorry, how much? Three-nine-nine.

(pre-expansion) (base pair first part) (insert expansion) (base pair second part) (post-expansion)

The pre-expansion initiated by the Old Man clears the way for the transaction conducted in the base pair. The Peasant, before giving his answer in that base pair, initiates an insert expansion that facilitates the production of that answer. Finally, the Old Man’s post-expansion opens up the base pair again because of a lack of comprehension of the second part offered by the Peasant (an instance of ‘repair’). Explanation (2.5): In modern paroemiology, the meaning of a proverb or wisdom saying as applied to the particular case at hand. Take the following conversation:

Glossary of Linguistic Terms

275

Clytemnestra: Will you ever finish fixing that infernal radiator? Aegisthus: Rome wasn’t built in a day, darling.

Here Aegisthus deflects Clytemnestra’s implied criticism about the pace of his work by suggesting that the task is a considerable one and therefore takes a significant amount of time to complete. This kind of application of a general truth (here, that it takes long to perform large tasks) to a particular case (here, to the radiator) is called ‘explanation’. Face (2.4.2): In politeness theory*, ‘face’ is the image that a person wants to adopt for him-/herself in social interaction, in other words, the way in which an individual sees him-/herself or would like to be seen by others. As a technical term, the concept derives from the work of Goffman. Politeness can be understood as an attempt to avoid damage to another’s face. Brown and Levinson divided the concept into ‘positive face’ (the desire to be liked and appreciated) and ‘negative face’ (the desire to be free from constraint); others have criticized this division. Face-threatening act (FTA) (2.4.2): In politeness theory*, this is any act (verbal or otherwise) that threatens the face* of one of the participants in an interaction or conversation. According to Brown and Levinson, all expressions of politeness are oriented towards a particular FTA and serve to limit the damage to face done by that FTA. Facework (2.4.2): In politeness theory*, ‘facework’ is any effort made by participants in a verbal interaction that is designed to preserve (or damage, in the case of aggressive or negative facework) their own face* or that of others. Felicity conditions (2.3.1): In speech act* theory, felicity conditions are the conditions that must be satisfied for a speech act to be performed successfully. If such conditions are not satisfied, the speech act misfires*. For example, if a person presiding over a wedding ceremony says ‘I now pronounce you husband and wife’ but is not qualified to officiate, then no marriage has taken place (a case of misfiring*). Felicity conditions generally concern the types of participants that must be involved in a certain speech act, the attitudes and intentions of those participants, and the procedures and conventions that must be followed in performing the act. First part: See adjacency pair*. First-person (saying, generalization, etc.) (2.5): In modern paroemiology, proverbs and other wisdom sayings are sometimes classified according to the person(s) or thing(s) to whom or to which the general sentiment expressed is applied in a particular case (see also explanation*). If the generalization is applied to the speaker him-/herself, the saying is a first-person saying; if it is applied to the addressee, it is a second-person saying; if it is applied to someone (or something) else, it is a third-person saying. Singular and plural are both possible, as are multiple persons (when a sentiment applies to more

276

Glossary of Linguistic Terms

than one group or person). This is not to be confused with grammatical person and number. For example, take the following conversation: Castor: What did you think of Agamemnon’s PowerPoint presentation? Polydeuces: Well, clearly, you can’t teach an old dog new tricks.

Here Polydeuces’ generalization is a third-person saying, as his general sentiment (a person who is used to doing things in a certain way will not normally change) is applied to Agamemnon (who probably still uses slides). Floor (2.2): In CA*, ‘floor’ is the metaphorical space (i.e. the ‘right’) to speak. A speaker has the floor for as long as his/her speaking turn* lasts, at which point the floor is either taken over by another speaker or, if no other speaker is selected* or self-selects*, the first speaker may continue. Gricean pragmatics (2.3.2): This is a branch of pragmatics originating in the work of H. P. Grice and deals with meaning as it is inferred by hearers and intended to be communicated by speakers, beyond the meaning encoded in the actual words used. In a sense, relevance theory* is also part of this tradition. Key terms include (conversational) implicature*, the cooperative principle*, and maxims of conversation*. Illocution, illocutionary force: See illocutionary act*. Illocutionary act (2.3.1): in speech act* theory, the act performed through making an utterance—for instance promising, asserting, requesting—is an illocutionary act. The illocutionary act is traditionally contrasted with the locutionary act (the physical act of saying something) and with the perlocutionary act or effects (the resulting effects that are caused by the utterance). For example, when considering the sentence: ‘New recruits have to polish their boots twice a day’, we may wonder about its illocutionary force even if we know the ‘literal meaning’ of the sentence. If spoken by a reporter in a news story on in the army, it will probably have the force of an assertion, but if it is uttered by a commanding officer to a group of recruits, it will have the force of a command. Implicature: See (Conversational) implicature. Indirect speech act (2.3.1): In speech act* theory, this is an utterance that achieves a certain illocutionary force without ‘wearing it on its sleeve’—in other words any utterance where the intended force does not correspond to the sentence type (declarative, interrogative, imperative) or to an illocutionary ‘tag’ (‘I request that . . . ’, ‘I promise that . . . ’). For example, someone can make an assertion in the form of a question: thus the utterance ‘Who wouldn’t want to win a million pounds?’ would normally function as an assertion, namely the assertion that everyone wants to win such a prize. Deriving the ‘real’ illocutionary force from an indirect speech act depends on inference, in some cases on linguistic and cultural conventions; for example, in English requests are conventionally uttered in indirect

Glossary of Linguistic Terms

277

form (‘Could you . . . ?’) rather than in direct form (‘I request that . . . ’ or the imperative). Insert expansion: See expansion*. Manner: See maxim of conversation*. Maxims of conversation (2.3.2): Conversational maxims are a central part of Gricean pragmatics*: they are rules or principles that speakers observe in conversation and that, together, make up the cooperative principle*. If they are intentionally violated (‘flouted’), this will give rise to a conversational implicature*. The four maxims are that of quantity* (make your contribution no more and no less informative than required), quality* (try to make your contribution one that is true), relation* (be relevant), and manner* (be perspicuous, avoid obscurity and ambiguity). Misfire: See felicity conditions*. Mixed group (2.4.1): This is an interaction consisting of male and female participants, as opposed to a single-sex group*. The make up of a group is an important factor in the analysis of the influence of gender on language use. Negative politeness: See politeness strategy*. Off-record (2.4.2): In politeness theory*, an off-record politeness strategy* consists of formulating a FTA* in such a way that the face threat can be ignored, often in the form of an indirect speech act*. The speaker will ‘implicate’ (see conversational implicature*) that part of his/her message that entails the face threat without literally saying it. The hearer is thus offered an ‘out’, which he/she can take by ignoring the implicature. Performative (2.3.1): A speech act* that ‘changes the world’ merely by virtue of the utterance being made is ‘performative’. For example, merely by uttering the sentence ‘I hereby declare this exhibition opened’, a speaker may change the status (opened or closed) of the exhibition in question. Performatives ‘create’ rather than ‘describe’ facts; and they cannot be true or false (they do, however, have felicity conditions*). Politeness strategy (2.4.2): In politeness theory, this is one in any number of possible forms of expression that prevents damage to the hearer’s (and/or speaker’s own) face*. According to Brown and Levinson, politeness strategies are oriented either towards positive face* (positive politeness*) or towards negative face* (negative politeness*). Politeness theory (2.4.2): Politeness theory is the study of how speakers use linguistic expressions in order to be polite to others and maintain social equilibrium. The central concept involved in politeness theory is that of face*. The most influential theory is that of Brown and Levinson, but it has been frequently criticized, among others by Watts. Politic behaviour (2.4.2): In Watts’ version of politeness theory*, ‘politic behaviour’ is the behaviour (linguistic and non-linguistic) that participants construct as being appropriate to the ongoing social interaction.

278

Glossary of Linguistic Terms

Positive politeness: See politeness strategy*. Post-expansion: See expansion*. Pragmatics (2.3): ‘Pragmatics’ is the study of language as it is used by speakers in a context. Pragmatics studies the relation between what sentences mean and what speakers mean when they utter them, or (in another conception of the discipline) it attempts to describe in psychologically realistic terms how human communication works. Pre-expansion: See expansion*. Preference organization (2.2): In CA*, ‘preference organization’ is the system that regulates which second part is the more marked response to the first part* of an adjacency pair*. For example, both a rejection and an acceptance are possible second parts to complement an invitation as first part, but rejections normally constitute a more complex conversational turn (one that includes pauses, hedges, etc.) and are therefore more marked than acceptances. In this case, rejection is the dispreferred* second part*, acceptance the preferred* second part*. These terms should not be taken to suggest psychological/emotional preference, only linguistic markedness. Preference organization may vary between languages and cultures. Preferred (turn, response, second part, etc.): See preference organization*. Principle of relevance: See relevance theory*. Quality: See maxims of conversation*. Quantity: See maxims of conversation*. Relation: See maxims of conversation*. Relevance theory (2.3.2): This is a theory of communication and cognition developed by Sperber and Wilson. The theory is built around the technical concept of relevance*, defined as the balance of ‘cognitive effects’ and the processing effort required for those effects. Relevance theory claims that human cognition generally seeks to achieve the greatest gain in accurate representations of the world for a minimal processing effort. As part of linguistic pragmatics* (in the tradition of Grice), relevance theory provides an account of how utterances are interpreted through a process of inference to the best explanation. This inferential process is guided by the communicative principle of relevance*, which states that every utterance contains an assumption of its own optimal relevance (i.e. that it is worth expending processing effort on interpreting the utterance until the meaning derived from it satisfies expectations of meaningfulness). Relevance (2.3.2): In relevance theory*, ‘relevance’ is a property of stimuli (inputs). The amount of relevance depends on how much ‘cognitive effect’ (crudely speaking, how much ‘meaning’) can be derived from an input and how much processing effort is required to derive those effects. The greater the effort, the lower the relevance. The greater the effect, the greater the relevance.

Glossary of Linguistic Terms

279

Second part: See adjacency pair*. Second-person (saying, generalization, etc.): See first-person saying*. Select: See turn*. Self-select: See turn*. Single-sex group: See mixed group*. Sociolinguistics (2.4): Sociolinguistics is the study of the effects of social variables such as status, gender, age, and so on on the use of language. Politeness theory* is normally considered part of sociolinguistics, which is itself sometimes seen as part of pragmatics*. Another important topic of research in sociolinguistics is gendered language. Speech act (2.3.1): Speech act* theory, originating in the work of Austin and Searle, examines the things that speakers do with their utterances (such as making a request, placing a bet, giving an order, or describing a state of affairs). Apart from classifying utterances by their illocutionary force*, into categories such as ‘commissive’, ‘performative’, and so on, speech act theorists have looked at the felicity conditions* that must be fulfilled for the successful utterance of a speech act. Speech act* theory is considered a part of pragmatics*. Third-person (saying, generalization, etc.): See first-person saying*. Topic management (2.2): In CA*, topic management refers to the ways in which interactants agree on, maintain, or change the topic of conversation. CA* studies the linguistic devices that speakers can use in order to keep a topic going or in order to change it, and also the relation between topic management and social and cultural factors such as the relative authority of speakers. Turn, turn-taking (2.2): In CA*, the mechanisms that govern the turntaking system form one of the central topics of analysis. Interactants in a conversation alternate their speaking turns, each of which lasts up to the point where it would be natural for a different speaker to take over (in other words, speakers normally have the floor* until they complete their turn). At such ‘transition relevance places’, the first speaker either selects* (explicitly or implicitly) another speaker to take over, or allows any other speaker to self-select*. In case no one self-selects*, the first speaker can continue speaking. There are surprisingly few overlaps and pauses between speaking turns in everyday exchanges, which indicates a highly effective system of organization. The power to apportion turns (selection), to ‘hog the floor’, or to self-select* may depend on the relative authority of the speakers or on other social and contextual factors.

Bibliography Abbott, F. F. 1907. ‘The Use of Language as a Means of Characterization in Petronius’, CPh 2.1: 43–50. Adams, J. N. 1984. ‘Female Speech in Latin Comedy’, Antichthon 18: 43–77. Adams, S. M. 1935. ‘Two Plays of Euripides’, CR 49.4: 118–22. Adema, S. 2007. ‘Discourse Modes and Bases in Vergil’s Aeneid’, in Allan and Buijs 2007, pp. 42–64. Alexiou, M. 2002. The Ritual Lament in Greek Tradition, 2nd edn, ed. by D. Yatromanolakis and P. Roilos. Lanham. Allan, R. J. 2003. The Middle Voice in Ancient Greek: A Study in Polysemy. Amsterdam. Allan, R. J. 2006. ‘Sophocles’ Voice: Active, Middle, and Passive in the Plays of Sophocles’, in de Jong and Rijksbaron 2006, pp. 111–26. Allan, R. J. 2007. ‘Sense and Sentence Complexity: Sentence Structure, Sentence Connection and Tense–Aspect as Indicators of Narrative Modes in Thucydides’ Histories’, in Allan and Buijs 2007, pp. 93–121. Allan, R. J. 2009. ‘Towards a Typology of the Narrative Modes in Ancient Greek: Text Types and Narrative Structure in the Messenger-Speeches of Euripides’, in Bakker and Wakker 2009, pp. 93–121. Allan, R. J. and M. Buijs, eds. 2007. The Language of Literature: Linguistic Approaches to Classical Texts. Leiden. Allan, R. J., I. J. F. de Jong, and C. C. de Jonge. 2014. ‘Homerus’ narratieve stijl: Enargeia en immersion’, Lampas 47.3: 202–23. Allan, W. 2000. The Andromache and Euripidean Tragedy. Oxford. Allan, W., ed. 2001. Euripides: The Children of Heracles. Warminster. Allan, W., ed. 2008. Euripides: Helen. Cambridge. Allan, W. 2013a. ‘The Armour of Achilles in Euripides’ Electra’, Lampas 46.2: 166–82. Allan, W. 2013b. ‘The Ethics of Retaliatory Violence in Athenian Tragedy’, Mnemosyne 66.4/5: 593–615. Allott, N. 2010. Key Terms in Pragmatics. London. Anderson, M. J. 2005. ‘Myth’, in Gregory 2005, pp. 121–35. Arewa, O. and A. Dundes. 1964. ‘Proverbs and the Ethnography of Speaking Folklore’, American Anthropologist 66.6/2: 70–84. Arndt, H. and R. Janney. 1987. InterGrammar: Toward an Integrative Model of Verbal, Prosodic and Kinesic Choices in Speech. Berlin. Arnoldt, R. 1878. Die chorische Technik des Euripides. Halle. Arnott, W. G. 1973. ‘Euripides and the Unexpected’, G&R 20.1: 49–64.

Bibliography

281

Arnott, W. G. 1981. ‘Double the Vision: A Reading of Euripides’ Electra’, G&R 28.2: 179–92. Arora, S. L. 1984. ‘The Perception of Proverbiality’, Proverbium 1: 1–38. [Repr. Mieder 1994: 3–30.] Atkinson, J. M. and P. Drew. 1979. Order in Court. London. Austin, J. L. 1962. How to Do Things with Words. Oxford. Bach, K. and R. M. Harnish. 1979. Linguistic Communication and Speech Acts. Cambridge, MA. Bain, D. 1975. ‘Audience Address in Greek Tragedy’, CQ 25.1: 13–25. Bain, D. 1977a. Actors and Audience: A Study of Asides and Related Conventions in Greek Drama. Oxford. Bain, D. 1977b. ‘Euripides, Electra 518–544’, BICS 24: 104–16. Bain, D. 1981a. ‘Euripides, Elektra 515ff.’, LCM 6: 137. Bain, D. 1981b. Masters, Servants and Orders in Greek Tragedy: A Study of Some Aspects of Dramatic Technique and Convention. Manchester. Bain, D. 1982. Review of Basta Donzelli 1978, CR 32.2: 272–3. Bain, D. 1984. ‘Female Speech in Menander’, Antichthon 18: 24–42. Bakker, E. J. 1993a. ‘Boundaries, Topics and the Structure of Discourse: An Investigation of the Ancient Greek Particle δέ’, StudLang 17.2: 275–311. Bakker, E. J. 1993b. ‘Discourse and Performance: Involvement, Visualization and “Presence” in Homeric Poetry’, ClAnt 12: 1–29. Bakker, E. J., ed. 1997a. Grammar as Interpretation: Greek Literature in Its Linguistic Contexts. Leiden. Bakker, E. J. 1997b. Poetry in Speech: Orality and Homeric Discourse. Ithaca. Bakker, E. J., ed. 2010a. A Companion to the Ancient Greek Language. Malden. Bakker, E. J. 2010b. ‘Pragmatics: Speech and Text’, in Bakker 2010a, pp. 151–68. Bakker, S. J. 2002. ‘Futura Zonder Toekomst’, Lampas 35.3: 199–214. Bakker, S. J. 2009a. ‘On the Curious Combination of the Particles γάρ and οὖν’, in Bakker and Wakker 2009, pp. 41–62. Bakker, S. J. 2009b. The Noun Phrase in Ancient Greek: A Functional Analysis of the Order and Articulation of NP Constituents in Herodotus. Leiden. Bakker, S. J. and G. C. Wakker, eds. 2009. Discourse Cohesion in Ancient Greek. Leiden. Bakogianni, A. 2011. Electra, Ancient and Modern: Aspects of the Reception of the Tragic Heroine. London. Barlow, S. A., ed. 1986a. Euripides: Trojan Women. Warminster. Barlow, S. A. 1986b. The Imagery of Euripides: A Study in the Dramatic Use of Pictorial Lanugage, 2nd edn. Bristol. Barrett, J. 2002. Staged Narrative: Poetics and the Messenger in Greek Tragedy. Berkeley. Barrett, W. S., ed. 1964. Euripides: Hippolytos. Oxford. Bary, C. 2012. ‘The Ancient Greek Tragic Aorist Revisited’, Glotta 88: 31–53.

282

Bibliography

Basset, L. 1997. ‘Ἀλλ’ ἐξόλοισθ᾽ αὐτῷ κοαξ: Réexamen des emplois de ἀλλά a la lumière de l’énonciation dans les Grenouilles d’Aristophane’, in Rijksbaron 1997b, pp. 75–99. Basta Donzelli, G. 1978. Studio sull’Elettra di Euripide. Catania. Basta Donzelli, G. 1980a. ‘Euripide, Elettra 1’, RFIC 108: 385–403. Basta Donzelli, G. 1980b. ‘Euripide, Elettra 518–44’, BICS 27: 109–19. Basta Donzelli, G. 1991a. ‘Sulla distribuzione delle battute nell’Elettra di Euripide’, BollClass 12: 5–35. Basta Donzelli, G. 1991b. ‘Sulle interpolazioni nell’Elettra di Euripide’, Eikasmos 2: 107–22. Basta Donzelli, G. 1992. ‘Osservazioni sul primo stasimo dell’Elettra di Euripide’, Eikasmos 3: 111–22. Basta Donzelli, G., ed. 1995a. Euripides: Electra. Leipzig. Basta Donzelli, G. 1995b. ‘Osservazioni sul II stasimo dell’Elettra di Euripide’, in G. Tarditi and L. Belloni, eds, Studia classica Iohanni Tarditi oblata, pp. 883–8. Milan. Battezzato, L. 1995. Il monologo nel teatro di Euripide. Pisa. Battezzato, L. 2001. ‘Euripides, Electra 300–301’, Mnemosyne 54.6: 731–3. Battezzato, L. 2007. Review of de Jong and Rijksbaron 2006, JHS 127: 159–60. Bäumlein, W. 1861. Untersuchungen über griechische Partikeln. Stuttgart. di Benedetto, V. 1961. ‘Responsione strofica e distribuizione delle battute in Euripide’, Hermes 89: 298–321. Bennison, N. 1993. ‘Discourse Analysis, Pragmatics and the Dramatic Character: Tom Stoppard’s Professional Foul’, Language and Literature 2.2: 79–99. Bers, V. 1997. Speech in Speech: Studies in Incorporated Oratio Recta in Attic Drama and Oratory. Lanham. Biraud, M. 2010. Les Interjections du théâtre grec antique: Étude sémantique et pragmatique. Louvain-la-Neuve. Blaiklock, E. M. 1952. The Male Characters in Euripides: A Study in Realism. Wellington. Blum-Kulka, S., J. House and G. Kasper. 1989. Cross-Cultural Pragmatics: Requests and Apologies. Norwood. Blundell, R. 2002. The Play of Character in Plato’s Dialogues. Cambridge. Bond, G. W. 1974. ‘Euripides’ Parody of Aeschylus’, Hermathena 118: 1–14. Bond, G. W., ed. 1981. Euripides: Heracles. Oxford. Bonifazi, A., M. de Kreij, and A. Drummen. in press. Particles in Ancient Greek Discourse: Pragmatic Studies of Particle Use across Genres. Washington, DC. Boter, G. J. 2012. ‘The Historical Present of Atelic and Durative Verbs in Greek Tragedy’, Philologus 156.2: 207–33. Bousfield, D. 2008. Impoliteness in Interaction. Amsterdam. de Brauw, M. 2007. ‘The Parts of the Speech’, in Worthington 2007, pp. 187–203.

Bibliography

283

Bremer, J. M. 1976. ‘Why Messenger-Speeches?’, in J. M. Bremer, S. L. Radt, and C. J. Ruijgh, eds, Miscellanea tragica in honorem J. C. Kamerbeek, pp. 29–48. Amsterdam. Briggs, C. L. 1985. ‘The Pragmatics of Proverb Performances in New Mexican Spanish’, American Anthropologist 87: 793–810. [Repr. Mieder 1994: 317–49.] Brown, H. P. 2003. The Pragmatics of Direct Address in the Iliad: A Study in Linguistic Politeness. Dissertation, Ohio State University. Brown, H. P. 2006. ‘Addressing Agamemnon: A Pilot Study of Politeness and Pragmatics in the Iliad’, TAPhA 136.1: 1–46. Brown, P. and S. C. Levinson. 1978. ‘Universals in Language Usage: Politeness Phenomena’, in E. Goody, ed., Questions and Politeness, pp. 56–289. Cambridge. Brown, P. and S. C. Levinson. 1987. Politeness: Some Universals in Language Use, 2nd edn. Cambridge. Bruhn, E. 1899. Sophokles, with commentary by F. W. Schneidewin and A. Nauck, vol. 8: Anhang. Berlin. Budelmann, F. 2000. The Language of Sophocles: Communality, Communication and Involvement. Cambridge. Budelmann, F. and P. E. Easterling. 2010. ‘Reading Minds in Greek Tragedy’, G&R 57.2: 289–303. Buijs, M. 2005. Clause Combining in Ancient Greek Narrative Discourse: The Distribution of Subclauses and Participial Clauses in Xenophon’s Hellenica and Anabasis. Leiden. Burian, P. 1997. ‘Myth into Muthos: The Shaping of Tragic Plot’, in P. E. Easterling, ed., The Cambridge Companion to Greek Tragedy, pp. 178–208. Cambridge. Burkert, W. 1966. ‘Greek Tragedy and Sacrificial Ritual’, GRBS 7.2: 87–121. [Rev. and repr. W. Burkert, 2001, Savage Energies: Lessons of Myth and Ritual in Ancient Greece. Chicago, pp. 1–36. Chicago.] Burkert, W. 1983. Homo Necans: The Anthropology of Ancient Greek Sacrificial Ritual and Myth. Berkeley. Burnett, A. P. 1998. Revenge in Attic and Later Tragedy. Berkeley. Cameron, D. and D. Kulick. 2003. Language and Sexuality. Cambridge. Camper, P., ed. 1831. Euripidis Electra. Leiden. Carey, C. 2007. ‘Epideictic Oratory’, in Worthington 2007, pp. 236–52. Carston, R. 2002. Thought and Utterances: The Pragmatics of Explicit Communication. Oxford. Carston, R. and S. Uchida. 1998. Relevance Theory: Applications and Implications. Amsterdam. Caspers, C. 2010. ‘The Pragmatic Function and Textual Status of Euripidean οὔ που and ἦ που’, CQ 60.2: 327–44. Chafe, W. 1994. Discourse, Consciousness, and Time: The Flow and Displacement of Conscous Experience in Speaking and Writing. Chicago.

284

Bibliography

Chierchia, G. and S. McConnell-Ginet. 2000. Meaning and Grammar: An Introduction to Semantics, 2nd edn. Cambridge, MA. Chong-Gossard, J. H. K. O. 2008. Gender and Communication in Euripides’ Plays: Between Song and Silence. Leiden. Clark, H. H. and R. J. Gerrig. 1990. ‘Quotations as Demonstrations’, Language 66: 764–805. Coates, J. 1996. Women Talk: Conversations between Women Friends. Oxford. Coates, J. 2004. Women, Men, and Language: A Sociolinguistic Account of Gender Differences in Language, 3rd edn. Harlow. Collard, C. 1975. ‘Formal Debates in Euripides’ Drama’, G&R 22: 58–71. Collard, C. 1980. ‘On Stichomythia’, LCM 5: 77–85. Collard, C. 1984. ‘The Re-Editing of Euripides’ (Review of Diggle 1981a, 1981b), CR 34.1: 9–15. Collard, C. 2005. ‘Colloquial Language in Tragedy: A Supplement to the Work of P. T. Stevens’, CQ 55.2: 350–86. Collins, D. E. 2001. Reanimated Voices: Speech Reporting in a Historical– Pragmatic Perspective. Amsterdam. Conacher, D. J. 1967. Euripidean Drama. Toronto. Conacher, D. J. 1981. ‘Rhetoric and Relevance in Euripidean Drama’, AJPh 102.1: 3–25. Cousland, J. R. C. and J. R. Hume, eds. 2009. The Play of Texts and Fragments: Essays in Honour of Martin Cropp. Leiden. Cropp, M. J. 1982. ‘Euripides Elektra 1013–7 and 1041–4’, LCM 7: 51–4. Cropp, M. J., ed. 1988. Euripides: Electra. Warminster. Cropp, M. J. 1996. Review of Basta Donzelli 1995a, BMCRev 96.7.3. http://bmcr.brynmawr.edu/1996/96.07.03.html (accessed 12 June 2016). Cropp, M. J. 2011. Review of Roisman and Luschnig 2011, BMCRev 2011.08.28. http://bmcr.brynmawr.edu/2011/2011-08-28.html (accessed 12 June 2016). Cropp, M. J., ed. 2013. Euripides: Electra, 2nd edn. Oxford. Csapo, E. 2009. ‘New Music’s Gallery of Images: The “Dithyrambic” First Stasimon Of Euripides’ Electra’, in Cousland and Hume 2009, pp. 93–110. Csapo, E. 2010. Actors and Icons of the Ancient Theater. Chichester. Culpeper, J. 2001. Language and Characterisation: People in Plays and Other Texts. Harlow. Culpeper, J. 2002. ‘A Cognitive Stylistic Approach to Characterisation’, in E. Semino and J. Culpeper, eds, Cognitive Stylistics: Language and Cognition in Text Analysis, pp. 251–78. Amsterdam. Culpeper, J. 2009. ‘Reflections on a Cognitive Stylistic Approach to Characterisation’, in J. Vandaele and G. Brône, eds, Cognitive Poetics: Goals, Gains and Gaps, pp. 125–59. Berlin. Culpeper, J. 2011. Impoliteness: Using Language to Cause Offence. Cambridge. Culpeper, J. and D. Z. Kádár, eds. 2010. Historical (Im)politeness. Bern.

Bibliography

285

Curnis, M. 2005. ‘Eur. El. 977’, SIFC 98.1: 109–16. Dale, A. M., ed. 1954. Euripides: Alcestis. Oxford. Davies, M. 1998. ‘Euripides’ Electra: The Recognition Scene Again’, CQ 48.2: 389–403. Davies, M. and P. J. Finglass, eds. 2015. Stesichorus: The Poems. Cambridge. Dawe, R. D. 2007. Sophocles: Oedipus Rex, 2nd edn. Cambridge. Denizot, C. 2011. Donner des ordres en grec ancien: Étude linguistique des formes de l’injonction. Mont-Saint-Aignan. Denniston, J. D., ed. 1939. Euripides: Electra. Oxford. Denniston, J. D. 1952. Greek Prose Style. Oxford. Denniston, J. D. 1954. The Greek Particles, 2nd edn, ed. by K. J. Dover. Oxford. do Deserto, J. 1994. ‘O Agricultor na Electra de Eurípides: Um homem simples no reino do ódio’, Humanitas 46: 111–21. Devine, A. M. and L. D. Stephens. 2000. Discontinuous Syntax: Hyperbaton in Greek. New York. Dickey, E. 1995. ‘Forms of Address and Conversational Language in Aristophanes and Menander’, Mnemosyne 48: 257–71. Dickey, E. 1996. Greek Forms of Address: From Herodotus to Lucian. Oxford. Dickin, M. 2009. A Vehicle for Performance: Acting the Messenger in Greek Tragedy. Lanham. Diggle, J. 1969. ‘Marginalia Euripidea’, PCPhS 15: 30–59. [Repr. 1994: 5–33.] Diggle, J. 1974. ‘On the Heracles and Ion of Euripides’, PCPhS 20: 3–36. [Repr. 1994: 90–136.] Diggle, J., 1981a. Euripidis fabulae, vol. 2. Oxford. Diggle, J. 1981b. Studies on the Text of Euripides: Supplices, Electra, Heracles, Troades, Iphigenia in Tauris, Ion. Oxford. Diggle, J. 1989. Review of Michelini 1987, AJPh 110.2: 357–62. Diggle, J. 1994. Euripidea: Collected Essays. Oxford. Dik, H. J. M. 1995. Word Order in Ancient Greek: A Pragmatic Account of Word Order Variation in Herodotus. Amsterdam. Dik, H. J. M. 2007. Word Order in Greek Tragic Dialogue. Oxford. Distilo, N. 2012. Commento critico–testuale all’Elettra di Euripide. Padova. Dover, K. J. 1974. Greek Popular Morality in the Time of Plato and Aristotle. Oxford. Drummen, A. 2009. ‘Discourse Cohesion in Dialogue: Turn-Initial ἀλλά in Greek Drama’, in Bakker and Wakker 2009, pp. 135–54. Drummen, A. 2014a. ‘Coherence’, in Giannakis 2014. http://referenceworks. brillonline.com/entries/encyclopedia-of-ancient-greek-language-and-li nguistics/coherence-SIM_00000428 (accessed 12 June 2016). Drummen, A. 2014b. ‘Cohesion’, in Giannakis 2014. http://referenceworks. brillonline.com/entries/encyclopedia-of-ancient-greek-language-and-li nguistics/cohesion-SIM_00000429 (accessed 12 June 2016).

286

Bibliography

Dubischar, M. 2001. Die Agonszenen bei Euripides: Untersuchungen zu ausgewählten Dramen. Stuttgart. Duchemin, J. 1968. L’ ΑΓΩΝ dans la tragédie grecque, 2nd edn. Paris. Duhoux, Y. 1997. ‘Grec écrit et grec parlé: Une étude contrastive des particules aux Ve-IVe siècles’, in Rijksbaron 1997b, pp. 15–48. Duhoux, Y. 2004. ‘Langage de femmes et d’hommes en grec ancien: L’example de Lysistrata’, in J. H. W. Penney, ed., Indo-European Perspectives: Studies in Honour of Anna Morpurgo Davies, pp. 131–45. Oxford. Earp, F. R. 1944. The Style of Sophocles. Cambridge. Earp, F. R. 1948. The Style of Aeschylus. Cambridge. Easterling, P. E. 1973. ‘Presentation of Character in Aeschylus’, G&R 20.1: 3–19. Easterling, P. E. 1988. ‘Tragedy and Ritual: “Cry ‘Woe, woe’ but may the good prevail!” ’, Métis 3: 87–109. Easterling, P. E. 1990. ‘Constructing Character in Greek Tragedy’, in Pelling 1990a, pp. 83–99. Easterling, P. E. and E. Hall, eds. 2002. Greek and Roman Actors: Aspects of an Ancient Profession. Cambridge. Eckert, P. and S. McConnell-Ginet. 2003. Language and Gender. Cambridge. Eder, J., F. Jannidis, and R. Schneider. 2010. ‘Characters in Fictional Worlds: An Introduction’, in J. Eder, F. Jannidis, and R. Schneider, eds, Characters in Fictional Worlds: Understanding Imaginary Beings in Literature, Film, and Other Media, pp. 3–65. Berlin. Edmondson, W. J. 1981. ‘On Saying You’re Sorry’, in F. Coulmas, ed., Conversational Routine: Explorations in Standardised Communication Situations and Prepatterned Speech, pp. 273–87. The Hague. Egli, F. 2003. Euripides im Kontext zeitgenössischer intellektueller Strömungen: Analyse der Funktion philosophischer Themen in den Tragödien und Fragmenten. Munich. van Emde Boas, E. H. 2015. ‘The Tutor’s Beard: Gender-Specific Communication and Speaker-Line Attribution in Greek Tragedy’, Mnemosyne 68.4: 543–68. van Emde Boas, E. H. 2016. Review of Cropp 2013, Distilo 2012, Mnemosyne. 69.2: 341–8. van Emde Boas, E. H. in press-a. ‘Analysing Agamemnon: Conversation Analysis and Particles in Greek Tragic Dialogue’, CPh. van Emde Boas, E. H. in press-b. ‘Euripides’, in K. de Temmerman and E. H. van Emde Boas, eds, Characterization in Ancient Greek Literature. Leiden. Erbse, H. 1984. Studien zum Prolog der euripideischen Tragödie. Berlin. Erdmann, G. 1964. Der Botenbericht bei Euripides: Struktur und dramatische Funktion. Dissertation, Kiel. van Erp Taalman Kip, A. M. 2009. ‘Kαὶ μήν, καὶ δή and ἤδη in Tragedy and Comedy’, in Bakker and Wakker 2009, pp. 111–34.

Bibliography

287

Filliettaz, L. and E. Roulet. 2002. ‘The Geneva Model of Discourse Analysis: An Interactionist and Modular Approach to Discourse Organization’, Discourse Studies 4: 369–94. Finglass, P. J., ed. 2007. Sophocles: Electra. Cambridge. Finglass, P. J. 2009. ‘Sophocles’ Tecmessa: Characterisation and Textual Criticism’, Eikasmos 20: 85–96. Finglass, P. J. 2014. ‘Sophocles Philoctetes 671–673: A Reconsideration Reconsidered’, Mnemosyne 67.3: 443–9. Finglass, P. J., ed. in press. Sophocles: Oedipus Rex. Cambridge. Fleischman, S. 1990. Tense and Narrativity: From Medieval Performance to Modern Fiction. Austin. Fögen, T. 2004. ‘Gender-Specific Communication in Graeco-Roman Antiquity: With a Research Bibliography’, Historiographia Linguistica 31: 199–276. Fögen, T. 2010. ‘Female Speech’, in Bakker 2010a, pp. 312–26. Foley, H. P. 1993. ‘The Politics of Tragic Lamentation’, in A. H. Sommerstein, S. J. Halliwell, J. Henderson, and B. Zimmerman, eds, Tragedy, Comedy and the Polis, pp. 33–47. Bari. Foley, H. P. 2001. Female Acts in Greek Tragedy. Princeton. Foti, F. 2005. ‘Attribuzioni di battute nella Medea di Euripide’, in S. Grandolini, ed., Lirica e teatro in Grecia: Il testo e la sua ricezione, pp. 189–99. Napoli. Fraenkel, E., ed. 1950. Aeschylus: Agamemnon, 3 vols. Oxford. Fraenkel, E. 1977. Due seminari romani di Eduard Fraenkel: Aiace e Filottete di Sofocle, ed. by L. E. Rossi. Roma. Fraser, B. and W. Nolen. 1981. ‘The Association of Deference with Linguistic Form’, International Journal of the Sociology of Language 27: 93–111. Friedrich, P. and J. Redfield. 1978. ‘Speech as a Personality Symbol: The Case of Achilles’, Language 54.2: 263–88. Friedrich, P. and J. Redfield. 1981. ‘Contra Messing’, Language 57.4: 901–3. Friis Johansen, H. 1959. General Reflection in Tragic Rhesis: A Study of Form. Copenhagen. Gärtner, T. 2005. ‘Verantwortung und Schuld in der Elektra des Euripides’, MH 62: 1–29. Gallagher, R. L. 2003. ‘Making the Stronger Argument the Weaker: Euripides, Electra 518–44’, CQ 53.2: 401–15. García Ramón, J. L. 2006. ‘Expresión del estado y tipos de lexema en griego homérico’, in E. Crespo, J. de la Villa, and A. R. Revuelta, eds, Word Classes and Related Topics in Ancient Greek, pp. 193–218. Louvain-laNeuve. Garzya, A. 1962. Pensiero e tecnica drammatica in Euripide: Saggio sul motivo della salvazione nei suoi drammi. Napoli. Gellie, G. 1981. ‘Tragedy and Euripides’ Electra’, BICS 28: 1–12. George, C. H. 2005. Expressions of Agency in Ancient Greek. Cambridge.

288

Bibliography

Giannakis, G. K., ed. 2014. Encyclopedia of Ancient Greek Language and Linguistics, 3 vols. Leiden. http://referenceworks.brillonline.com/browse/ encyclopedia-of-ancient-greek-language-and-linguistics (accessed 19 July 2016). Gildersleeve, B. L. 1907. Review of M. Bréal, 1906. Pour mieux connaître Homère, AJPh 28.2: 208–17. Gill, C. 1990. ‘The Character-Personality Distinction’, in Pelling 1990a, pp. 1–31. Gill, C. 1996. Personality in Greek Epic, Tragedy, and Philosophy: The Self in Dialogue. Oxford. Gilleland, M. E. 1980. ‘Female Speech in Greek and Latin’, AJPh 101.2: 180–3. Gimbel, S. and J. Bulhof. 2004. ‘A Tautology Is a Tautology (or Is It?)’, Journal of Pragmatics 36: 1003–5. Goff, B. E. 1991. ‘The Sign of the Fall: The Scars of Orestes and Odysseus’, ClAnt 10: 259–67. Goff, B. E. 1999–2000. ‘Try to Make It Real Compared to What? Euripides’ Electra and the Play of Genres’, in Sansone, Cropp, and Lee 1999–2000, pp. 93–105. Goffman, E. 1955. ‘On Face Work: An Analysis of Ritual Elements in Social Interaction’, Psychiatry 18: 213–31. Goldhill, S. 1984. Language, Sexuality, Narrative: The Oresteia. Cambridge. Goldhill, S. 1986a. Reading Greek Tragedy. Cambridge. Goldhill, S. 1986b. ‘Rhetoric and Relevance: Interpolation at Euripides Electra 367–400’, GRBS 27.2: 157–71. Goldhill, S. 2012. Sophocles and the Language of Tragedy. New York. Goossens, R. 1962. Euripide et Athènes. Brussels. Gould, J. 1978. ‘Dramatic Character and “Human Intelligibility” in Greek Tragedy’, PCPhS 24: 43–67. [Repr. 2001, Myth, Ritual, Memory, and Exchange: Essays in Greek Literature and Culture, pp. 78–111. Oxford.] Green, G. 1975. ‘Nonsense and Reference; Or, the Conversational Use of Proverbs’, in R. E. Grossman, L. J. San, and T. J. Vance, eds, Papers from the Seventh Regional Meeting of the Chicago Linguistic Society, pp. 226–39. Chicago. Gregor, D. B. 1957. ‘Ὦ φίλτατ’’, CR 7.1: 14–15. Gregory, J., ed. 2005. A Companion to Greek Tragedy. Malden. Grice, H. P. 1975. ‘Logic and Conversation’, in P. Cole and J. L. Morgan, eds, Speech Acts, pp. 41–58. New York. Grice, H. P. 1978. ‘Further Notes on Logic and Conversation’, in P. Cole, ed., Pragmatics, pp. 113–28. New York. Grice, H. P. 1981. ‘Presupposition and Conversational Implicature’, in P. Cole, ed., Radical Pragmatics, pp. 183–98. London. Grice, H. P. 1989. Studies in the Way of Words. Cambridge, MA. Griffith, M. 1999. Sophocles: Antigone. Cambridge.

Bibliography

289

Grube, G. M. A. 1941. The Drama of Euripides. London. Hall, J. 2009. Politeness and Politics in Cicero’s Letters. Oxford. Halliwell, S. J. 1990. ‘Traditional Greek Conceptions of Character’, in Pelling 1990a, pp. 32–59. Halporn, J. W. 1983. ‘The Skeptical Electra’, HSCPh 87: 101–18. Hamilton, R. 1974. ‘Bacchae 47–52: Dionysus’ Plan’, TAPhA 104: 139–49. Hamilton, R. 1978. ‘Prologue, Prophecy and Plot in Four Plays of Euripides’, AJPh 99: 277–302. Haslam, M. W. 1976. ‘ “O Ancient Argos of the Land”: Euripides, Electra 1’, CQ 26.1: 1–2. Heath, M. 1987. The Poetics of Greek Tragedy. London. Heath, M. 1999. ‘Sophocles’ Philoctetes: A Problem Play?’, in J. Griffin, ed., Sophocles Revisited: Essays Presented to Sir Hugh Lloyd-Jones, pp. 137–60. Oxford. Heath, M. 2006. ‘The “Social Function of Tragedy”: Clarifications and Questions’, in D. L. Cairns and V. Liapis, eds, Dionysalexandros: Essays on Aeschylus and His Fellow Tragedians in Honour of Alexander F. Garvie, pp. 253–82. Swansea. Henrichs, A. 2000. ‘Drama and Dromena: Bloodshed, Violence, and Sacrificial Metaphor in Euripides’, HSCPh 100: 173–88. Herman, V. 1991. ‘Dramatic Dialogue and the Systematics of Turn-Taking’, Semiotica 83: 97–122. Herman, V. 1995. Dramatic Discourse: Dialogue as Interaction in Plays. London. Herman, V. 1998a. ‘Multiple “Floors” and “Turns” in Drama’, Journal of Literary Semantics 27: 110–27. Herman, V. 1998b. ‘Turn Management in Drama’, in J. Culpeper, M. H. Short, and P. Verdonk, eds, Exploring the Language of Drama: From Text to Context, pp. 19–33. London. Holmes, J. 2013. An Introduction to Sociolinguistics, 4th edn. Harlow. Hornblower, S., A. Spawforth, and E. Eidinow, eds. 2012. Oxford Classical Dictionary, 4th edn. Oxford. Hrisztova-Gotthardt, H. and M. A. Varga, eds. 2014. Introduction to Paremiology: A Comprehensive Guide to Proverb Studies. Berlin. Huang, Y. 2007. Pragmatics. Oxford. Huitink, L. 2013. Reported Discourse in Classical Greek: Grammatical Form and Communicative Function. Dissertation, Oxford. Hunger, H. 1936. ‘Realistiche Charakterdarstellung in den Spätwerken des Euripides’, Commentationes Vindobonenses 2: 5–28. Jannidis, F. 2013. ‘Character’, in P. Hühn, J. Pier, W. Schmid, and J. Schönert, eds., The Living Handbook of Narratology. Hamburg: Hamburg University Press. http://www.lhn.uni-hamburg.de/article/character (accessed 12 June 2016).

290

Bibliography

Jens, W., ed. 1971. Die Bauformen der griechischen Tragödie. Munich. Jesperson, O. 1922. Language: Its Nature Development and Origin. London. Johnstone, H. W. 1980. ‘Pankoinon as a Rhetorical Figure in Greek Tragedy’, Glotta 56: 49–62. de Jong, I. J. F. 1990. ‘Three Off-Stage Characters in Euripides’, Mnemosyne 43.1: 1–21. [Repr. Mossman 2003: 369–89.] de Jong, I. J. F. 1991. Narrative in Drama: The Art of the Euripidean Messenger-Speech. Leiden. de Jong, I. J. F. 1997. ‘Γάρ Introducing Embedded Narratives’, in Rijksbaron 1997b, pp. 175–86. de Jong, I. J. F. 2007. ‘Sophocles Trachiniae 1–48, Euripidean Prologues, and Their Audiences’, in Allan and Buijs 2007, pp. 7–28. de Jong, I. J. F. 2010. ‘Euripides and his Prologues: A Reappraisal’, Pharos: Journal of the Netherlands Institute in Athens 17.1: 21–34. de Jong, I. J. F. and A. Rijksbaron, eds. 2006. Sophocles and the Greek Language: Aspects of Diction, Syntax and Pragmatics. Leiden. Jouanna, J. 1997. ‘Notes sur la scène de la reconnaissance dans les Choéphores d’Eschyle (v. 205–211) et sa parodie dans l’Électre d’Euripide (v. 532–537)’, CGITA 10: 69–85. Jouanna, J. 1998. ‘Qui est la fille de Tyndare? Texte (v. 479 sq.) et interprétation du premier stasimon de l’Électre d’Euripide (v. 432–486)’, REA 100: 525–32. Julé, A. 2008. A Beginner’s Guide to Language and Gender. Clevedon. Kádár, D. Z. and M. Haugh. 2013. Understanding Politeness. Cambridge. Kamerbeek, J. C. 1987. ‘Some Notes on Euripides’ Electra’, Mnemosyne 40.3/4: 276–85. Kannicht, R., ed. 1969. Euripides: Helena, 2 vols. Heidelberg. Kannicht, R., ed. 2004. Tragicorum graecorum fragmenta, vol. 5: Euripides. Göttingen. Katsouris, A. G. 1975. Linguistic and Stylistic Characterization: Tragedy and Menander. Ioannina. Keene, C. H., ed. 1893a. The Electra of Euripides. London. Keene, C. H. 1893b. ‘Scholia on the Electra of Euripides’, CR 5.9: 432–3. Kells, J. H. 1960. ‘Euripides, Electra 1093–5, and Some Uses of δικάζειν’, CQ 10.1: 129–34. Kells, J. H. 1966. ‘More Notes on Euripides’ Electra’, CQ 16.1: 51–4. Kindstrand, J. F. 1978. ‘The Greek Concept of Proverbs’, Eranos 76: 71–85. King, K. C. 1980. ‘The Force of Tradition: The Achilles Ode in Euripides’ Electra’, TAPhA 110: 195–212. Kirchhoff, A., ed. 1855. Euripidis Tragoediae, vol. 2. Berlin. Kirchhoff, A., ed. 1867. Euripidis Fabulae, 2nd edn, vol. 1. Berlin. Kirschenblatt-Gimblett, B. 1973. ‘Towards a Theory of Proverb Meaning’, Proverbium 22: 821–7. [Repr. Mieder and Dundes 1981: 111–21.] Kitto, H. D. F. 1961. Greek Tragedy: A Literary Study, 3rd edn. London.

Bibliography

291

Kitzinger, M. R. 1976. Stylistic Methods of Characterization in Sophocles’ Antigone. Dissertation, Stanford. Knox, B. M. W. 1970. ‘Euripidean Comedy’, in F. Fergusson, A. Cheuse, and R. Koffler, eds, The Rarer Action: Essays in Honor of Francis Fergusson, pp. 68–96. New Brunswick. [Repr. 1979, Word and Action: Essays on the Ancient Theater, pp. 250–74. Baltimore.] Konstan, D. 1985. ‘Philia in Euripides’ Electra’, Philologus 129.2: 176–85. Kovacs, D. 1984. Review of Diggle 1981a, 1981b, AJPh 105.2: 236–41. Kovacs, D. 1985. ‘Castor in Euripides’ Electra (El. 307–13 and 1292–1307)’, CQ 35.2: 306–14. Kovacs, D. 1987a. ‘Treading the Circle Warily: Literary Criticism and the Text of Euripides’, TAPhA 117: 257–70. Kovacs, D. 1987b. ‘Where Is Aegisthus’ Head?’, CPh 82.2: 139–41. Kovacs, D. 1989. ‘Euripides, Electra 518–44: Further Doubts about Genuineness’, BICS 36: 67–78. Kovacs, D. 1996. Euripidea Altera. Leiden. Kovacs, D., ed. 2002. Euripides, vol. 5. Cambridge, MA. Kovacs, D. 2012. Review of Roisman and Luschnig 2011, JHS 132: 186–7. Kramer, C. 1975. ‘Women’s Speech: Separate But Unequal’, in B. Thorne and N. Henley, eds, Language and Sex: Difference and Dominance, pp. 43–56. Rowley. Kroon, C. H. M. 1995. Discourse Particles in Latin: A Study of nam, enim, autem, vero, and at. Amsterdam. Kroon, C. H. M. 2007. ‘Discourse Modes and the Use of Tenses in Ovid’s Metamorphoses’, in Allan and Buijs 2007, pp. 65–92. Kubo, M. 1967. ‘The Norm of Myth: Euripides’ Electra’, HSCPh 71: 15–31. Kühner, R. and B. Gerth. 1898–1904. Ausführliche Grammatik der griechischen Sprache, Part 2: Satzlehre, 2 vols. Hannover. [= KG]. Kvičala, J. 1879. Studien zu Euripides. Mit einem Anhang Sophokleischer Analekta, 2 vols. Vienna. Lakoff, G. 1987. Women, Fire, and Dangerous Things: What Categories Reveal About the Mind. Chicago. Lakoff, G. and M. Johnson. 1980. Metaphors We Live By. Chicago. Lakoff, R. T. 1973. ‘The Logic of Politeness: Or Minding Your P’s and Q’s’, Chicago Linguistics Society 8: 292–305. Lakoff, R. T. 1975. Language and Women’s Place. New York. Lallot, J., A. Rijksbaron, B. Jacquinod, and M. Buijs, eds. 2011. The Historical Present in Thucydides: Semantics and Narrative Function. Leiden. Lamers, H. and A. Rademaker. 2007. ‘Talking About Myself: A Pragmatic Approach to the Use of Aspect Forms in Lysias 12.4–19’, CQ 57.02: 458–76. Lardinois, A. 1995. Wisdom in Context: The Use of Gnomic Statements in Archaic Greek Poetry. Dissertation, Princeton.

292

Bibliography

Lardinois, A. 1997. ‘Modern Paroemiology and the Use of Gnomai in Homer’s Iliad’, CPh 92: 213–34. Lardinois, A. 2000. ‘Characterization through Gnomai in Homer’s Iliad ’, Mnemosyne 53.6: 641–61. Lardinois, A. 2001. ‘The Wisdom and Wit of Many: The Orality of Greek Proverbial Expressions’, in J. Watson, ed., Speaking Volumes: Orality and Literacy in the Greek and Roman World, pp. 93–107. Leiden. Lardinois, A. 2006. ‘The Polysemy of Gnomic Expressions and Ajax’ Deception Speech’, in de Jong and Rijksbaron 2006, pp. 213–24. Lattimore, R. 1979. ‘Optatives of Consent and Refusal’, in G. W. Bowersock, W. Burkert, and M. C. J. Putnam, eds, Arktouros: Hellenic Studies Presented to Bernard M.W. Knox on the Occasion of His 65th Birthday, pp. 209–16. Berlin. Lebeck, A. 1971. The Oresteia: A Study in Language and Structure. Cambridge, MA. Leech, G. N. 1983. Principles of Pragmatics. London. Leech, G. N. 2014. The Pragmatics of Politeness. New York. Leech, G. N. and M. Short. 2007. Style in Fiction: A Linguistic Introduction to English Fictional Prose. Harlow. Leezenberg, M. 2002. ‘Power in Communication: Implications for the Semantics–Pragmatics Interface’, Journal of Pragmatics 34.7: 893–908. Leezenberg, M. 2005. ‘Greek Tragedy as Impolite Conversation: Towards a Practice Approach in Linguistic Theory’, in S. Marmaridou, N. Kiki, and E. Antonopoulou, eds, Reviewing Linguistic Thought: Converging Trends for the 21st Century, pp. 191–208. Berlin. Levinson, S. C. 1983. Pragmatics. Cambridge. Lloyd, M. A. 1986. ‘Realism and Character in Euripides’ Electra’, Phoenix 40.1: 1–19. Lloyd, M. A. 1992. The Agon in Euripides. Oxford. Lloyd, M. A. 1999. ‘The Tragic Aorist’, CQ 49.1: 24–45. Lloyd, M. A. 2004. ‘The Politeness of Achilles: Off-Record Conversation Strategies in Homer and the Meaning of kertomia’, JHS 124: 75–89. Lloyd, M. A. 2006. ‘Sophocles in the Light of Face-Threat Politeness Theory’, in de Jong and Rijksbaron 2006, pp. 225–40. Lloyd, M. A. 2009. ‘The Language of the Gods: Politeness in the Prologue of the Troades’, in Cousland and Hume 2009, pp. 183–92. Lloyd, M. A. 2013. ‘Politeness/Impoliteness’, in Roisman 2013, p. 987. Lloyd-Jones, H. 1972. Review of Lebeck 1971, JHS 92: 193–5. [Repr. 1990, Greek Epic, Lyric, and Tragedy: The Academic Papers of Sir Hugh LloydJones, vol. 1, pp. 278–82. Oxford.] Lowe, J. C. B. 1962. ‘The Manuscript Evidence for Changes of Speaker in Aristophanes’, BICS 9.1: 27–42.

Bibliography

293

Lowe, N. 2004. ‘Euripides’, in I. J. F. De Jong, A. Bowie, and R. Nunlist, eds, Narrators, Narratees, and Narratives in Ancient Greek Literature, pp. 269–80. Leiden. Luria, S. 1929. ‘Entstellungen des Klassiker Textes bei Stobaios’, RhM 78: 81–248. Luschnig, C. A. E. 2013. ‘Euripides: Electra’, in Roisman 2013, pp. 378–84. Luschnig, C. A. E. 2015. ‘Electra’, in R. Lauriola and K. N. Demetriou, eds, Brill’s Companion to the Reception of Euripides, pp. 199–237. Leiden. Mac Coinnigh, M. 2014. ‘Structural Aspects of Proverbs’, in HrisztovaGotthardt and Varga 2014, pp. 112–32. MacDowell, D. M. 1966. Athenian Homicide Law in the Age of the Orators. Manchester. Mahler, A. 2013. ‘Aesthetic Illusion in Theatre and Drama: An Attempt at Application’, in W. Wolf, W. Bernhart, and A. Mahler, eds, Immersion and Distance: Aesthetic Illusion in Literature and Other Media, pp. 151–82. Amsterdam. Maltby, R. 1979. ‘Linguistic Characterization of Old Men in Terence’, CPh 74.2: 136–47. Mandala, S. 2007. Twentieth-Century Drama Dialogue as Ordinary Talk: Speaking between the Lines. London. Margolin, U. 2007. ‘Character’, in D. Herman, ed., The Cambridge Companion to Narrative, pp. 66–79. Cambridge. Martin, R. P. 1989. The Language of Heroes: Speech and Performance in the Iliad. Ithaca. Martin, V. ed. 1958. Ménandre: Le Dyscolos. Geneva. Martina, A. 1979. ‘Euripide, Elettra 304–13’, AION (Filol) 1: 23–50. Martinich, A. P. 2010. The Philosophy of Language. New York. de Martino, F. and A. H. Sommerstein, eds. 1995. Lo spettacolo delle voci, 2 vols. Bari. Masqueray, P. 1908. Euripide et ses idées. Paris. Mastronarde, D. J. 1979. Contact and Discontinuity: Some Conventions of Speech and Action on the Greek Tragic Stage. Berkeley. Mastronarde, D. J., ed. 1994. Euripides: Phoenissae. Cambridge. Mastronarde, D. J., ed. 2002. Euripides: Medea. Cambridge. Mastronarde, D. J. 2010. The Art of Euripides: Dramatic Technique and Social Context. Cambridge. Matić, D. 2003. ‘Topic, Focus, and Discourse Structure: Ancient Greek Word Order’, StudLang 27.3: 573–633. Matthiessen, K. 1964. Elektra, Taurische Iphigenie und Helena: Untersuchungen zur Chronologie und zur dramatischen Form im Spätwerk des Euripides. Göttingen. Mau, A. 1877. ‘Zu Euripides Elektra’, in Commentationes philologae in honorem Theodori Mommseni, scripserunt amici, pp. 291–301. Berlin.

294

Bibliography

Maxwell-Stuart, P. G. 1975. ‘Παρήϊς: A Note on Euripides Electra 1023’, CQ 25.2: 312–13. McClure, L. 1995. ‘Female Speech and Characterization in Euripides’, in de Martino and Sommerstein 1995, vol. 2, pp. 35–60. Bari. McClure, L. 1999. Spoken Like a Woman: Speech and Gender in Athenian Drama. Princeton. McClure, L. and A. Lardinois, eds. 2001. Making Silence Speak: Women’s Voices in Greek Literature and Society. Princeton. McDevitt, A. S. 1981. ‘Antilabe in Sophoclean Kommoi’, RhM 124.1: 19–28. Meibauer, J. 2008. ‘Tautology as Presumptive Meaning’, Pragmatics and Cognition 16.3: 439–70. Messing, G. M. 1981. ‘On Weighing Achilles’ Winged Words’, Language 57.4: 888–900. Mey, J. L. 2001. Pragmatics: An Introduction, 2nd edn. Oxford. Michelini, A. N. 1987. Euripides and the Tragic Tradition. Madison. Mieder, W., ed. 1994. Wise Words: Essays on the Proverb. New York. Mieder, W. 2004. Proverbs: A Handbook. Westport, CT. Mieder, W. 2014. ‘The Origin of Proverbs’, in Hrisztova-Gotthardt and Varga 2014, pp. 28–48. Mieder, W. and A. Dundes, eds. 1981. The Wisdom of Many: Essays on the Proverb. New York. Mikalson, J. D. 1989. ‘Unanswered Prayers in Greek Tragedy’, JHS 109: 81–98. Mills, S. 2003. Gender and Politeness. Cambridge. Minchin, E. 2007. Homeric Voices: Discourse, Memory, Gender. Oxford. Mirto, M. S. 1980. ‘Il sacrificio tra metafora e mechanema nell’Elettra di Euripide’, CCC 1: 299–329. Montiglio, S. 2000. Silence in the Land of Logos. Princeton. Moorhouse, A. C. 1982. The Syntax of Sophocles. Leiden. Morgan, T. 2007. Popular Morality in the Early Roman Empire. Oxford. Morwood, J. H. W. 1981. ‘The Pattern of the Euripides Electra’, AJPh 102.4: 362–70. Mossman, J. 1995. Wild Justice: A Study of Euripides’ Hecuba. Oxford. Mossman, J. 2001. ‘Women’s Speech in Greek Tragedy: The Case of Electra and Clytemnestra in Euripides’ Electra’, CQ 51.2: 374–84. Mossman, J., ed. 2003. Oxford Readings in Classical Studies: Euripides. Oxford. Mossman, J. 2005. ‘Women’s Voices’, in Gregory 2005, pp. 352–65. Mossman, J., ed. 2011. Euripides: Medea. Oxford. Most, G. W. 2003. ‘Euripide ὁ γνωμολογικώτατος’, in M. S. Funghi, ed., Aspetti di letteratura gnomica del mondo antico, pp. 141–66. Florence. Müller, C. W. 2000. ‘Elektras Erkenntnisproblem: Zu Eur. El. 503ff ’, RhM 143: 251–5.

Bibliography

295

Myrick, L. D. 1994. ‘The Way up and Down: Tracehorse and Turning Imagery in the Orestes Plays’, CJ 89.2: 131–48. Naiden, F. S. 2006. Ancient Supplication. Oxford. Nauck, A., ed. 1876. Euripidis Tragoediae, 3rd edn, vol. 1. Leipzig. Nijk, A. 2013. ‘A Pragmatic Account of the Use of the Historic Present in De Corona’, Mnemosyne 66.3: 365–98. Nijk, A. in press. ‘A Unified Account of the Nonpast Uses of the Aorist’, JHS. Nordgren, L. 2015. Greek Interjections: Syntax, Semantics and Pragmatics. Berlin. Norrick, N. R. 1985. How Proverbs Mean: Semantic Studies in English Proverbs. Berlin. Norrick, N. R. 2014. ‘Subject Area, Terminology, Proverb Definitions, Proverb Features’, in Hrisztova-Gotthardt and Varga 2014, pp. 1–27. O’Brien, M. J. 1964. ‘Orestes and the Gorgon: Euripides’ Electra’, AJPh 85.1: 13–39. Obeng, S. G. 1996. ‘The Proverb as a Mitigating and Politeness Strategy in Akan Discourse’, Anthropological Linguistics 38.3: 521–49. Olson, S. D. 1996. ‘Manuscript Indications of Change of Speaker in Aristophanes’ Peace’, ICS 21: 5–34. Olson, S. D. 2001. ‘Manuscript Indications of Change of Speaker in Aristophanes’ Acharnians’, ICS 26: 1–36. van Ophuijsen, J. M. 1993. ‘Οὖν, Ἄρα, Δή, Τοίνυν: The Linguistic Articulation of Arguments in Plato’s Phaedo’, in Sicking and van Ophuijsen 1993, pp. 67–164. Ormand, K. 2009. ‘Electra the Exile’, in D. McCoskey and E. Zakin, eds, Bound by the City: Greek Tragedy, Sexual Difference, and the Formation of the Polis, pp. 247–74. Albany. Paduano, G. 1970. ‘La scena del riconoscimento nell’Elettra e la critica razionalistica alle Coefore’, RFIC 98: 385–405. Page, D. 1934. Actors’ Interpolations in Greek Tragedy. Oxford. Page, D. 1937. ‘The Chorus of Alcman’s Partheneion’, CQ 31.2: 94–101. Papadimitropoulos, L. 2008. ‘Causality and Innovation in Euripides’ Electra’, RhM 151.2: 113–26. Papadopoulou, T. 2005. The Heracles and Euripidean Tragedy. Cambridge. Parker, L. P. E., ed. 2007. Euripides: Alcestis. Oxford. Parker, R. 1983. Miasma: Pollution and Purification in Early Greek Religion. Oxford. Parmentier, L. and H. Grégoire, eds. 1925. Euripide, vol. 4. Paris. Parry, A. 1956. ‘The Language of Achilles’, TAPhA 87: 1–7. Pelling, C. B. R., ed. 1990a. Characterization and Individuality in Greek Literature. Oxford. Pelling, C. B. R. 1990b. ‘Conclusion’, in Pelling 1990a, pp. 245–62.

296

Bibliography

Perdicoyianni-Paleologou, H. 2014. ‘Discourse Analysis and Greek’, in Giannakis 2014. http://referenceworks.brillonline.com/entries/encyclopediaof-ancient-greek-language-and-linguistics/discourse-analysis-and-greekCOM_00000112 (accessed 12 June 2016). Perris, S. 2011. ‘What Maketh the Messenger: Reportage in Greek Tragedy’, in A. Mackay, ed., ASCS 32: Selected Proceedings. www.ascs.org.au/news/ ascs32/Perris.pdf (accessed 12 June 2016). Pohlenz, M. 1930. Die griechische Tragödie, vol. 2. Leipzig. Pomerantz, A. and J. Heritage. 2012. ‘Preference’, in Sidnell and Stivers 2012, pp. 210–28. Pomeroy, S. B. 1976. Goddesses, Whores, Wives, and Slaves: Women in Classical Antiquity. New York. Popp, H. 1968. Amoibaion: Zur Geschichte einer Dialogform der griechischen Tragödie. Dissertation, Tübingen. Porter, D. H. 1986. ‘The Imagery of Greek Tragedy: Three Characteristics’, SO 61.1: 19–42. Porter, J. R. 1990. ‘Tiptoeing through the Corpses: Euripides’ Electra, Apollonius, and the Bouphonia’, GRBS 31.3: 255–80. Probert, P. and E. Dickey. 2005. ‘Giving Directions in Euripides’ Hecuba’, Omnibus 49: 3–4. Pucci, P. 1980. The Violence of Pity in Euripides’ Medea. Ithaca. Rademaker, A. 2005. Sophrosyne and the Rhetoric of Self-Restraint: Polysemy and Persuasive Use of an Ancient Greek Value Term. Leiden. Radermacher, L. 1899. Review of Wecklein 1898, GGA 161: 691–719. Radt, S. L. 1985. ‘Aristophanica’, in W. J. Aerts, J. H. A. Lokin, S. L. Radt, and N. van der Wal, eds, ΣΧΩΛΙΑ: Studia ad criticam interpretationemque textuum Graecorum et ad historiam iuris Graeco-Romani pertinentia viro doctissimo D. Holwerda oblata, pp. 103–18. Groningen. Raeburn, D. 2000. ‘The Significance of Stage Properties in Euripides’ Electra’, G&R 47.2: 149–68. Recanati, F. 2004. Literal Meaning. Cambridge. Reeve, M. D. 1973. ‘Interpolation in Greek Tragedy III’, GRBS 14.2: 145–71. Revuelta, A. R. 2014. ‘Particles (Syntactic Features)’, in Giannakis 2014, pp. 31–41. Leiden. http://referenceworks.brillonline.com/entries/encyclope dia-of-ancient-greek-language-and-linguistics/particles-syntactic-featuresCOM_00000272 (accessed 19 July 2016). Rijksbaron, A. 1976. ‘How Does a Messenger Begin His Speech? Some Observations on the Opening Lines of Euripidean Messenger Speeches’, in J. M. Bremer, S. L. Radt, and C. J. Ruijgh, eds, Miscellanea Tragica in Honorem J. C. Kamerbeek, pp. 293–308. Amsterdam. Rijksbaron, A. 1984. ‘Het Griekse perfectum: Subject contra object’, Lampas 17.5: 403–19.

Bibliography

297

Rijksbaron, A. 1988. ‘The Discourse Function of the Imperfect’, in A. Rijksbaron, H. A. Mulder, and G. C. Wakker, eds, In the Footsteps of Raphael Kühner, pp. 237–54. Amsterdam. Rijksbaron, A. 1991. Grammatical Observations on Euripides’ Bacchae. Amsterdam. Rijksbaron, A. 1997a. ‘Introduction’, in Rijksbaron 1997b, pp. 1–14. Rijksbaron, A., ed. 1997b. New Approaches to Greek Particles. Amsterdam. Rijksbaron, A. 2002. Syntax and Semantics of the Verb in Classical Greek: An Introduction, 3rd edn. Amsterdam. Rijksbaron, A. 2006. ‘On False Historic Presents in Sophocles (and Euripides)’, in de Jong and Rijksbaron 2006, pp. 127–50. Rijksbaron, A., ed. 2007. Plato: Ion; or, On the Iliad. Leiden. Rijksbaron, A. 2015. ‘Stative Historical Presents in Greek Tragedy: Are They Real?’, Philologus 159.2: 224–50. Risselada, R. 1993. Imperatives and Other Directive Expressions in Latin: A Study in the Pragmatics of a Dead Language. Amsterdam. Roisman, H. M. 2004. ‘Women’s Free Speech in Greek Tragedy’, in Sluiter and Rosen 2004, pp. 91–114. Leiden. Roisman, H. M., ed. 2013. The Encyclopedia of Greek Tragedy. Oxford. Roisman, H. M. and C. A. E. Luschnig, eds. 2011. Euripides’ Electra: A Commentary. Norman. Romero, M. and C.-H. Han. 2004. ‘On Negative Yes/No Questions’, Linguistics and Philosophy 27.5: 609–58. de Romilly, J. 1983. ‘Les réflexions générales d’Euripide: Analyse littéraire et critique textuelle’, CRAI 127.2: 405–18. Ronnet, G. 1975. ‘L’Ironie d’Euripide dans Électre (vers 513 à 546)’, REG 88: 63–70. Rosenmeyer, T. G. 1982. The Art of Aeschylus. Berkeley. Rosivach, V. J. 1978. ‘The “Golden Lamb” Ode in Euripides’ Electra’, CPh 73.3: 189–99. Roulet, E., L. Filliettaz, and A. Grobet. 2001. Un modèle et un instrument d’analyse de l’organisation du discours. Bern. Russo, J. 1997. ‘Prose Genres for the Performance of Traditional Wisdom in Ancient Greece: Proverb, Maxim, Apophtegm’, in L. Edmunds and R. W. Wallace, eds, Poet, Public, and Performance in Ancient Greece, pp. 49–64. Baltimore. Rutherford, R. B. R. 2010. ‘The Greek of Athenian Tragedy’, in Bakker 2010a, pp. 441–54. Rutherford, R. B. R. 2012. Greek Tragic Style: Form, Language, and Interpretation. Cambridge. Sacks, H. 1992. Lectures on Conversation, ed. by G. Jefferson. Oxford. Sacks, H., E. A. Schegloff, and G. Jefferson. 1974. ‘A Simplest Systematics for the Organization of Turn-Taking for Conversation’, Language 50: 696–735.

298

Bibliography

Sansone, D. 1984. Review of Diggle 1981a, 1981b, CPh 79.4: 335–40. Sansone, D., M. J. Cropp, and K. Lee, eds. 1999–2000. Euripides and Tragic Theatre in the Late Fifth Century (Special issue of ICS 24/25). Champaign. Schadewaldt, W. 1926. Monolog und Selbstgespräch: Untersuchungen zur Formengeschichte der griechischen Tragödie. Berlin. Schegloff, E. A. 2007. Sequence Organization in Interaction: A Primer in Conversation Analysis I. Cambridge. Schenkl, K. 1874. ‘Kritische Studien zu Euripides Elektra’, Zeitschrift fur die Österreichischen Gymnasien 25: 81–96. Schiappa, E. 1999. The Beginnings of Rhetorical Theory in Classical Greece. New Haven. Schottländer, R. 1982. ‘Fortschrittsverkündung oder Adelsrhetorik? Zu einer euripideschen Maxime’, Hermes 110.4: 490–4. Schuren, L. 2014. Shared Storytelling in Euripidean Stichomythia. Leiden. Schwinge, E. R. 1968. Die Verwendung der Stichomythie in den Dramen des Euripides. Heidelberg. Scodel, R. 2008. Epic Facework: Self-Presentation and Social Interaction in Homer. Swansea. Seaford, R. 1985. ‘The Destruction of Limits in Sophokles’ Elektra’, CQ 35.2: 315–23. Seaford, R., ed. 1996. Euripides: Bacchae. Warminster. Searle, J. R. 1969. Speech Acts. Cambridge. Searle, J. R. 1979. Expression and Meaning: Studies in the Theory of Speech Acts. Cambridge. Segal, C. P. 1992. ‘Tragic Beginnings: Narration, Voice, and Authority in the Prologues of Greek Drama’, YClS 29: 85–112. Seidensticker, B. 2008. ‘Character and Characterization in Greek Tragedy’, in M. Revermann and P. Wilson, eds, Performance, Iconography, Reception: Studies in Honour of Oliver Taplin, pp. 333–46. Oxford. Seidensticker, B. 2009. ‘Charakter und Charakterisierung bei Aischylos’, in J. Jouanna and F. Montanari, eds, Eschyle à l’aube du théâtre occidental, pp. 205–56. Vandoeuvres-Genève. Seitel, P. 1969. ‘Proverbs: A Social Use of Metaphor’, Genre 2: 143–61. [Repr. Mieder and Dundes 1981: 122–39.] Sheppard, J. T. 1918. ‘The Electra of Euripides’, CR 32: 137–41. Short, M. H. 1981. ‘Discourse Analysis and the Analysis of Drama’, Applied Linguistics 2.2: 180–202. Sicking, C. M. J. 1993. ‘Devices for Text Articulation in Lysias I and XII’, in Sicking and van Ophuijsen 1993, pp. 1–66. Leiden. Sicking, C. M. J. and P. Stork. 1997. ‘The Grammar of the So-Called Historical Present in Ancient Greek’, in Bakker 1997a, pp. 131–68. Sicking, C. M. J. and J. M. van Ophuijsen. 1993. Two Studies in Attic Particle Usage: Lysias and Plato. Leiden.

Bibliography

299

Sider, D. 1977. ‘Two Stage Directions for Euripides’, AJPh 98.1: 16–19. Sidnell, J. 2010. Conversation Analysis: An Introduction. Malden. Sidnell, J. and T. Stivers, eds. 2012. The Handbook of Conversation Analysis. Chichester. Sier, K. ed. 1988. Die lyrischen Partien der Choephoren des Aischylos: Text, Übersetzung, Kommentar. Stuttgart. Silk, M. S. 1996. ‘Tragic Language: The Greek Tragedians and Shakespeare’, in M. S. Silk, ed., Tragedy and the Tragic: Greek Theatre and Beyond, pp. 458–96. Oxford. Slings, S. R. 1992. ‘Written and Spoken Language: An Exercise in the Pragmatics of the Greek Sentence’, CPh 87: 95–109. Slings, S. R. 1997a. ‘Adversative Relators between PUSH and POP’, in Rijksbaron 1997b, pp. 101–30. Slings, S. R. 1997b. ‘Figures of Speech and Their Lookalikes: Two Further Exercises in the Pragmatics of the Greek Sentence’, in Bakker 1997a, pp. 169–74. Slings, S. R. 1997c. ‘Notes on Euripides’ Electra’, Mnemosyne 50.2: 131–64. Sluiter, I., ed. 2008. KAKOS: Badness and Anti-Value in Classical Antiquity. Leiden. Sluiter, I. and R. M. Rosen, eds. 2004. Free Speech in Classical Antiquity. Leiden. Smith, C. S. 2003. Modes of Discourse: The Local Structure of Texts. Cambridge. Smith, D. L. 2012. The Rhetoric of Interruption: Speech-Making, TurnTaking, and Rule-Breaking in Luke-Acts and Ancient Greek Narrative. Berlin. Smith, L. P. 1948. Studies of Characterization in Euripides: The Medeia, Elektra, and Orestes. Dissertation, Princeton. Solmsen, F. 1967. ‘Electra and Orestes: Three Recognitions in Greek Tragedy’, Mededelingen der Koninklijke Nederlandse Akademie van Wetenschappen (Afdeling Letterkunde) 30.2: 30–62. Sommerstein, A. H. 1995. ‘The Language of Athenian Women’, in de Martino and Sommerstein 1995, vol. 2, pp. 61–85. Bari. Sperber, D. and D. Wilson. 1986. Relevance: Communication and Cognition. Oxford. Sperber, D. and D. Wilson. 1995. Relevance: Communication and Cognition, 2nd edn. Oxford. Sperber, D. and D. Wilson. 2002. ‘Pragmatics, Modularity and MindReading’, Mind and Language 17: 3–23. Spira, A. 1960. Untersuchungen zum deus ex machina bei Sophokles und Euripides. Kallmünz. Stanford, W. B. 1942. Aeschylus in His Style: A Study in Language and Personality. Dublin.

300

Bibliography

Steidle, W. 1968. Studien zum antiken Drama. Munich. Steiger, H. 1897. ‘Warum schrieb Euripides seine Elektra?’, Philologus 10: 561–600. Steiger, H. 1912. Euripides: Seine Dichtung und seine Persönlichkeit. Leipzig. Stevens, P. T. 1976. Colloquial Expressions in Euripides. Wiesbaden. Stieber, M. 2011. Euripides and the Language of Craft. Leiden. Stoessl, F. 1956. ‘Die Elektra des Euripides’, RhM 99: 47–92. Strohm, H. 1957. Euripides: Interpretationen zur dramatischen Form. Munich. Sultan, N. 1993. ‘Private Speech, Public Pain: The Power of Women’s Laments in Ancient Greek Poetry and Tragedy’, in K. Marshall, ed., Rediscovering the Muses: Women’s Musical Traditions, pp. 92–110. Boston. Suter, A. 2008a. Lament: Studies in the Ancient Mediterranean and Beyond. Oxford. Suter, A. 2008b. ‘Male Lament in Greek Tragedy’, in Suter 2008a, pp. 156–80. Swift, L. 2010. The Hidden Chorus: Echoes of Genre in Tragic Lyric. Oxford. Tannen, D. 1993. Gender and Conversational Interaction. New York. Tannen, D. 1996. Gender and Discourse, 2nd edn. New York. Taplin, O. 1972. ‘Aeschylean Silences and Silences in Aeschylus’, HSCPh 76: 57–97. Taplin, O. 1977. The Stagecraft of Aeschylus: The Dramatic Use of Exits and Entrances in Greek Tragedy. Oxford. Taplin, O. 2007. Pots and Plays: Interactions Between Tragedy and Greek Vase-Painting of the Fourth Century BC. Los Angeles. Tarkow, T. A. 1981. ‘The Scar of Orestes: Observations on a Euripidean Innovation’, RhM 124: 143–52. Taylor, A. 1962. The Proverb; And an Index to ‘the Proverb’. Hatboro. De Temmerman, K. 2014. Crafting Characters: Heroes and Heroines in the Ancient Greek Novel. Oxford. Thévenet, L. 2009. Le Personnage: Du mythe au théâtre: La question de l’identité dans la tragédie grecque. Paris. Thumiger, C. 2007. Hidden Paths: Self and Characterization in Greek Tragedy: Euripides’ Bacchae. London. Thumiger, C. 2013. ‘Character and Characterization in Greek Tragedy’, in Roisman 2013, pp. 206–12. Tompkins, D. P. 1972. ‘Stylistic Characterization in Thucydides: Nicias and Alcibiades’, YClS 22: 181–214. Tompkins, D. P. 1993. ‘Thucydides Constructs His Speakers: The Case of Diodotus’, Electronic Antiquity 1.1. https://scholar.lib.vt.edu/ejournals/ ElAnt/V1N1/tompkins.html (accessed 12 June 2016). Toolan, M. J. 1990. The Stylistics of Fiction: A Literary–Linguistic Approach. London. Toolan, M. J. 1998. Language in Literature: An Introduction to Stylistics. London.

Bibliography

301

Underiner, T. L. 2005. MAXnotes: Euripides’ Medea and Electra. Piscataway. Vahlen, J. 1891. ‘Zu Sophokles und Euripides Elektra’, Hermes 26: 351–65. Vernant, J.-P. and P. Vidal-Naquet. 1981. Tragedy and Myth in Ancient Greece, trans. J. Lloyd. Sussex. Vickers, B. 1973. Towards Greek Tragedy. London. Victorius, P., ed. 1545. ΕΥΡΙΠΙΔΟΥ ΗΛΕΚΤΡΑ: Euripidis Electra, nunc primum in lucem edita. Rome. Vidal-Naquet, P. 1986. The Black Hunter: Forms of Thought and Forms of Society in the Greek World, trans. A. Szegedy-Maszak. Baltimore. Vitelli, G. 1880. ‘Appunti critici sull’Elettra di Euripide’, RFIC 8: 401–516. Vögler, A. 1967. Vergleichende Studien zur sophokleischen und euripideischen Elektra. Heidelberg. Wakker, G. C. 1994. Conditions and Conditionals: An Investigation of Ancient Greek. Amsterdam. Wakker, G. C. 1997a. ‘Emphasis and Affirmation: Some Aspects of μήν in Tragedy’, in Rijksbaron 1997b, pp. 209–32. Wakker, G. C. 1997b. ‘Modal Particles and Different Points of View in Herodotus and Thucydides’, in Bakker 1997a, pp. 215–50. Wakker, G. C. 2009. ‘ “Well I Will Now Present My Arguments”: Discourse Cohesion Marked by οὖν and τοίνυν in Lysias’, in Bakker and Wakker 2009, pp. 63–82. Wakker, G. C. 2014. ‘Text Linguistics and Greek’, in Giannakis 2014. http:// referenceworks.brillonline.com/entries/encyclopedia-of-ancient-greeklanguage-and-linguistics/text-linguistics-and-greek-COM_00000344 (accessed 12 June 2016). Walsh, G. B. 1977. ‘The First Stasimon of Euripides’ Electra’, YClS 25: 277–89. Watts, R. J. 2003. Politeness. Cambridge. Wecklein, N., ed. 1898. Euripidis fabulae, vol. 1. Leipzig. Wecklein, N., ed. 1906. Euripides: Elektra, mit erklärenden Anmerkungen. Leipzig. Weil, H., ed. 1905. Sept tragédies d’Euripide. Paris. Werkhofer, K. 1992. ‘Traditional and Modern Views: The Social Constitution and the Power of Politeness’, in R. J. Watts, S. Ide, and K. Ehlich, eds, Politeness in Language: Studies in Its History, Theory and Practice, pp. 155–99. Berlin. West, C. 1979. ‘Against Our Will: Male Interruptions of Females in Crosssex Conversation’, in J. Orasanu, M. K. Slater, and L. L. Adler, eds, Language, Sex, and Gender: Does la différence Make a Difference? pp. 81–96. New York. West, M. L. 1980. ‘Tragica IV’, BICS 27: 17–22. West, M. L. 1982. Greek Metre. Oxford. West, M. L. 1984. ‘Tragica VII’, BICS 31: 171–92.

302

Bibliography

Whitehorne, J. E. G. 1978. ‘The Ending of Euripides’ Electra’, RBPh 56.1: 5–14. van de Wijnpersse, W. M. A. 1929. De Terminologie van het Jachtwezen bij Sophocles. Dissertation, Utrecht. von Wilamowitz-Moellendorf, T. J. W. 1917. Die dramatische Technik des Sophokles. Berlin. von Wilamowitz-Moellendorf, U. 1875. Analectea Euripidea. Berlin. von Wilamowitz-Moellendorf, U. 1883. ‘Die beiden Elektren’, Hermes 18: 214–63. Willi, A. 2003. The Languages of Aristophanes: Aspects of Linguistic Variation in Classical Attic Greek. Oxford. Willink, C., ed. 1986. Euripides: Orestes. Oxford. Willink, C. 2009. ‘Euripides, Electra 432–486 and Iphigenia In Tauris 827–899’, in Cousland and Hume 2009, pp. 205–18. Leiden. [Repr. 2010, Collected Papers on Greek Tragedy, ed. by W. Benjamin Henry, pp. 754–66. Leiden.] Wilson, D. and D. Sperber. 2004. ‘Relevance Theory’, in L. R. Horn and G. Ward, eds, The Handbook of Pragmatics, pp. 607–32. Malden. Wilson, D. and D. Sperber., eds. 2012. Meaning and Relevance. Cambridge. Winnington-Ingram, R. P. 1937. ‘Euripides, Electra 1292–1307’, CR 51.2: 51–2. van Wolferen, A. J. 2003. Vertelde en Gespeelde Wereld: Taalkundige Notities bij de Openingen van Euripides’ Tragedies. Dissertation, Vrije Universiteit Amsterdam. Worman, N. 2002. The Cast of Character: Style in Greek Literature. Austin. Worthington, I., ed. 2007. A Companion to Greek Rhetoric. Malden. Wright, M. 2013. Review of Rutherford 2012, CR 63.1: 33–4. Wuhrmann, W. 1940. Strukturelle Untersuchungen zu den beiden Elektren und zum euripideischen Orestes. Dissertation, Zürich. Yoon, F. 2012. The Use of Anonymous Characters in Greek Tragedy: The Shaping of Heroes. Leiden. Yule, G. 1996. Pragmatics. Oxford. Zeitlin, F. I. 1965. ‘The Motif of the Corrupted Sacrifice in Aeschylus’ Oresteia’, TAPhA 96: 463–508. Zeitlin, F. I. 1970. ‘The Argive Festival of Hera and Euripides’ Electra’, AJPh 101: 645–69. [Repr. Mossman 2003, 261–84.] Zeitlin, F. I. 1996. Playing the Other: Gender and Society in Classical Greek Literature. Chicago. Zürcher, W. 1947. Die Darstellung des Menschen im Drama des Euripides. Basel. Zuntz, G. 1955. The Political Plays of Euripides. Manchester. Zuntz, G. 1965. An Inquiry into the Transmission of the Plays of Euripides. Cambridge. Zuntz, G. 1970. ‘On the First Verse of Euripides’ Elektra’, RhM 113: 276–81.

Index Locorum Index of passages from Electra discussed Passing mentions are ignored. Discussions of individual lines or shorter passages that arise in the context of a discussion of a longer passage (e.g. discussion of line 1 as part of the discussion of lines 1–53) are omitted, unless they are discussed separately in particular detail. Textual-critical discussions are listed separately in italics. 1–53 62–70 1 63 n. 8 6 63 n. 9 37 68 n. 25 44 67 n. 22 57–9 71 n. 36 65 72 n. 37 78–9 78 112–66 90 167–214 87–9 171–80 15–16, 17, 20, 22, 25 213–14 91 215–89 98–108 215–17 30 215–29 37–9 228–36 98–100 234–6 170–1 237–45 100–5 241 100 n. 50 244–5 172 246–63 105–6 248 172 n. 11 261–2 172 264–84 106–7 275 104 277 107 n. 64 281–2 172 285–96 107–8 287–9 171 290–6 171–6 294–6 174–6 300–38 93–6, 108–22 300–1 109 300–3 109–111 304–13 112–18 307–13 112–18 314–31 118–21 332–8 121 336–8 104

341–63 74–6 360 184 n. 40 367–400 177–85 368–72 181 373–9 180–1 386–90 181–3 396–400 183 404–31 76–8 487–698 122–7 518–23 123 518–44 123 518–44 123–5 524–6 123–4, 125–6 524–8 103–4 527–31 124 534–7 124 54–81 71–4, 84–7 541–4 124 568 125–6 646 126 n. 115 646–7 126–7 671–93 193–204 671–93 190–1, 193–204 761–4 230–1 774–8 232–4 774–858 185–6, 229–47 776 232 n. 9 777 233–4 779–97 234–8 798–810 238–9 810–43 239–42 824–5 56 844–55 242 855–7 242–3 857–8 243 880–906 128–9 898–9 133 900–4 134 907–56 131 n. 129

OUP CORRECTED PROOF – FINAL, 4/1/2017, SPi

304 (cont.) 907–56 128–41, 131 n. 129 907–8 135 909–13 135 914–17 135–6 918–29 137 930–8 137 939–44 137–8 945–51 138 952–6 138–9 957–8 140–1 959–87 192, 204–28 959–66 56, 204–14 967–87 215–28 997 141 n. 155 998–1010 141–3 998–1146 141–53 1008–10 250 1008–10 143 n. 162 1011–50 143–5, 246–68 1011–17 250–1 1014 144 n. 166 1018–23 253–5 1024–9 260–2 1030–40 261 1030–4 253–5 1035–40 259–60, 260–2

Index Locorum 1041–5 260–2 1041–5 261 1046–50 265–6 1051–4 145 1055–9 145–7 1060–99 147–50, 246–68 1060 251–2 1060–8 251–3 1069–71 255–8 1072–5 259–60 1076–9 255–8 1080–5 259–60 1086–93 263–4 1086 256 n. 13 1093–6 266–7 1102–46 150–53 1102–5 46–7 1107–8 151 1107–8 151 n. 184 1110 151 n. 183 1116–17 10 1124–32 11–13 1177–232 154–9, 154–6 1233–358 159–64 1276–7 247 n. 40 1292–307 160–2 1292–8 13–14

Index of passages from other works cited or discussed See also Subject Index for general discussions. Sigla in square brackets refer to works generally believed to be spurious. For the abbreviations used, see ‘A Note on Citations, Abbreviations, and Cross-referencing’ in the book’s front matter. Aeschylus (Aesch.) Ag. 348–51 110 n. 73 Ag. 520 202 n. 35 Ag. 524 173 n. 15 Ag. 592 110 n. 73 Ag. 674 173 n. 15 Ag. 587–97 258 n. 16 Ag. 774–5 153 Ag. 810–13 129 n. 121 Ag. 907 129 n. 121 Ag. 908 208 Ag. 911 176 n. 21 Ag. 918 110 n. 73 Ag. 939 218 n. 81 Ag. 942 222 n. 97 Ag. 931–43 129 n. 122 Ag. 958–74 129 n. 121 Ag. 1072–7 90 n. 18 Ag. 1126 153

Ag. 1256 74 n. 46 Ag. 1313 151 n. 184 Ag. 1317 94 n. 30 Ag. 1399 110 n. 73 Ag. 1401–6 144 Ag. 1438 138 n. 147 Ag. 738–809 141 Ag. 1623 101 n. 52 Ag. 1625–7 136 n. 139, 138 n. 147 Ag. 1633–5 136 n. 139 Ag. 855–913 258 n. 16 Cho. 438 107 n. 66 Cho. 501 203 Cho. 510–13 203 Cho. 781 184 Cho. 315–478 198 n. 24 Cho. 899 215 n. 68 Eum. 372 173 n. 15

OUP CORRECTED PROOF – FINAL, 4/1/2017, SPi

Index Locorum Sept. 152 201 n. 32 fr. 46c 210 n. 54 Andocides (Andoc.) 1.72 251 n. 4 Aristophanes (Ar.) Eccl. 275 200 Eq. 488 151 n. 184 Nub. 498 113 n. 83 Nub. 505 222 n. 98 Pax 449 197 Pax 444–6 197 Pax 452 197 Plut. 257 101 n. 52 Ran. 462 222 n. 98 Ran. 524 222 n.98 Ran. 945–7 181 n. 35 Ran. 946–7 64 n. 15 Ran. 1109 175 n. 20 Ran. 1237 64 n. 10 Ran. 1367 63 n. 6 Ran. 1382 63 n. 6 Ran. 1175–248 62 Thesm. 850 218 n. 76 Vesp. 146 200 Aristotle (Arist.) Rh. 1355b–6a 260 n. 20 Rh. 1415a–b 251 n. 4 [Rh. Al.] 1442a–c 251 n. 4 Demosthenes (Dem.) 58.7 251 n. 4 Euripides (Eur.) Alc. 1 63 Alc. 299 206 Alc. 244–79 89 Alc. 1047–8 102 n. 54 Andr. 1 63 Andr. 47 65 n. 17 Andr. 64 99 n. 46 Andr. 93–5 82 n. 3 Andr. 91–116 82 Andr. 186–91 251 n. 4 Andr. 349 101 n. 52 Andr. 319–33 68 n. 28 Andr. 577 208 n. 44 Andr. 622 180 n. 32 Andr. 715–16 208 n. 44 Andr. 830 138 n. 149 Andr. 825–65 89 Andr. 896 209–10 n. 53 Andr. 921 111 n. 77

Andr. 920–53 109 n. 69 Andr. 1086–91 233 Bacch. 170–4 208 n. 44 Bacch. 178 99 nn. 45–6 Bacch. 677–9 233 Bacch. 821–36 212 n. 61 Bacch. 870 213 n. 64 Bacch. 921–2 174 Cyc. 251 173 n. 15 Cyc. 351 200 Cyc. 692 200 n. 30 Hec. 234–6 110 n. 73 Hec. 313 207 Hec. 505 99 n. 46 Hec. 579 222 n. 99 Hec. 675 126 n. 115 Hec. 836–40 121 n. 102 Hec. 1114–15 99 n. 46 Hec. 1132 110 n. 72 Hel. 15 67 n. 20 Hel. 22 67 n. 20 Hel. 94 94 n. 31 Hel. 164–78 82 Hel. 330 90 n. 17 Hel. 496 221 n. 91 Hel. 587 126 n. 115 Hel. 625–59 188 n. 1 Hel. 722 200 Hel. 761 207 Hel. 1049 110 n. 73 Hel. 1094 201 n. 32 Hel. 1184 208 n. 45 Hel. 1093–106 202 n. 34 Hel. 1228 126 n. 115 Hel. 1266 126 n. 115 Her. 31–2 65 n. 17 Her. 141 109 n. 71 Her. 279 110 n. 73 Her. 451 207 Her. 531 99 n. 47 Her. 655–68 179 Her. 701–33 152 n. 189 Her. 1014 221 n. 91 Heracl. 1–5 63 Heracl. 5–9 63 Heracl. 31 94 n. 31 Heracl. 202 173 n. 15 Heracl. 474–83 110 n. 73 Heracl. 640 99 Heracl. 734 101 n. 52 Heracl. 788 99 n. 46 Heracl. 786–7 231 n. 5

305

306

Index Locorum

Euripides (Eur.) (cont.) Heracl. 867 200 Heracl. 1050–3 208 nn. 44, 47 Hipp. 3–4 65 n. 17 Hipp. 21 65 n. 17 Hipp. 297 206 Hipp. 473–4 197 Hipp. 507–8 73 n. 44 Hipp. 567 208 n. 45 Hipp. 598 126 n. 115 Hipp. 666 173 n. 15 Hipp. 776 210 n. 54 Hipp. 808 208 n. 44 Hipp. 925–31 179 Hipp. 1092 99 n. 45 Hipp. 1173–5 233 Hipp. 1391 74 n. 46 Hipp. 1452 99 nn. 44, 46 IA 817 198–9 IA 842 98 n. 40 IA 900–16 109 n. 69 IA 985 196 nn. 18–19 IA 1124–6 135 n. 137 IA 1340 208 n. 44 Ion 418 198 Ion 561–2 198 n. 23 IT 208 114 n. 84 IT 342 207 IT 695–6 138 n. 148 IT 815 99 n. 47 IT 827 99 n. 47 IT 915 138 n. 148 Med. 1 63 n. 6 Med. 1–6 63 Med. 65 156 n. 196 Med. 49–95 188 n. 1 Med. 533 173 n. 15 Med. 516–19 179 Med. 588 213 Med. 709 111 n. 77 Med. 708–18 109 n. 69 Med. 767 200 Med. 791 172 n. 11 Med. 1071 99 n. 45 Med. 1021–67 68 n. 28 Med. 1125–6 231 n. 5 Med. 1272 99 n. 44 Med. 1278 213 n. 64 Med. 1314 208 n. 44 Med. 1346 138 n. 149 Med. 1376 126 n. 115 Or. 1–3 63

Or. 16 62 n. 5 Or. 26–7 62 n. 5 Or. 1–70 68 n. 28 Or. 153–4 102 n. 54 Or. 449–55 109 n. 69 Or. 544–50 251 n. 4 Or. 602–4 150 n. 181 Or. 640 110 n. 72 Or. 918–22 60 Or. 918–22 60 Or. 1045 99 n. 44 Or. 1092–3 138 n. 148 Or. 1100 99 n. 45, 107 n. 66 Or. 1147 173 n. 14 Or. 1240 203 Or. 1235–7 200 Or. 1225–45 198 n. 24 Or. 1290 210 n. 54 Or. 887–917 237 n. 20 Or. 1395–502 231 n. 5 Phoen. 1 63 Phoen. 4–6 63 Phoen. 131 101 n. 52 Phoen. 713 101 n. 52 Phoen. 751–2 58 n. 113 Phoen. 753 151 n. 184 Phoen. 996 210 n. 56 Phoen. 1072 99 n. 46 Phoen. 1067–8 208 n. 44 Phoen. 1252–3 199–200 Phoen. 1365 201 n. 32 Phoen. 1432 210 n. 54 [Rh.] 623 199 Supp. 92 209 n. 51 Supp. 165 111 n. 77 Supp. 163–92 109 n. 69 Supp. 326 222 n. 99 Supp. 421 175 n. 20 Supp. 452–4 138 n. 147 Supp. 465 110 n. 72 Supp. 641 99 n. 46 Supp. 744 180 n. 32, 183 n. 38 Supp. 842 198 Supp. 949 180 n. 32, 183 n. 38 Supp. 857–917 131 n. 128 Tro. 48–53 13 n. 23 Tro. 761 200 Tro. 906–13 110 n. 73 Tro. 914–17 251 n. 4 fr. 202 (Antiope) 175 n. 18 fr. 752g 3–17 (Hyps.) 89 fr. 752f 29–31 (Hyps.) 89

Index Locorum Gorgias (Gorg.) fr. 82 B11 a4 135 n. 137 Herodotus (Hdt.) 7.9.2 121 n. 104 Hesiod (Hes.) Op. 391 113 n. 83 Homer (Hom.) Il. 3.54–5 138 n. 1477 Il. 3.108 42–3 Od. 9.14–15 135 n. 137 Od. 24.21–2 245 Od. 22.37 138 n. 147 Od. 3.262–4 136 n. 139 Od. 3.259–61 129 n. 123 Od. 4.529–37 245 Od. 7.294 42–3 Od. 11.388–9 245 Od. 11.411 153 Od. 11.409–11 245 Lysias (Lys.) 2 (various passages) 131 n. 128 Pind (Pind.) Pyth. 11.22 150 Plato (Pl.) Resp. 404b 72 n. 39 Resp. 474a 113 n. 83 Plutarch (Plut.) Dion 58.1 244 n. 35 Lys. 15.3 1 n. 1 Sophocles (Soph.) Aj. 391 107 n. 66 Aj. 992 68 n. 28 Ant 740–1 173–4 Ant. 739 213

El. 261 210–11 El. 300–2 136 n. 139 El. 345 199 El. 411 198–9 El. 679 231 n. 6 El. 973 101 n. 52 El. 1338 203 El. 1487–90 129 n. 123 OC 1383 138 n. 149 OT 41 203 OT 95 110 n. 72 OT 164–7 202 n. 35 OT 1008 213 Phil. 371 233 Phil. 383 94 n. 31 Trach. 83 222 n. 99 Trach. 86 151 n. 184 Trach. 269–80 244 n. 32 Trach. 389 151 n. 184 Trach. 749–812 244 n. 33 Stesichorus (Stes.) fr. 17. 23 155 fr. 93 (=S148) 14–15 114 n. 84 Stobaeus (Stob.) 4.22 150 n. 181 3.72 175 n. 18 97.5 68 Strattis fr. 35.2 218 n. 76 Thucydides (Thuc.) 2.235–46 131 n. 128 Xenophon (Xen.) An. 6.2.10 72 n. 39 An. 4.4.12 113 n. 83

307

Subject Index accentuation 101 n. 52, 114, 174 n. 17, 191, 233 Achilles 53 n. 103, 261, 262 acting 5 n. 8 address, forms of 27, 30, 33, 34, 39, 71, 73, 90–1, 99, 119 n. 94, 125, 142, 146, 151, 155, 158, 161 n. 212, 198, 201, 208, 213 n. 66; see also Index of Greek Words adjacency pair, see conversation analysis (CA) Aegisthus: burial 247 n. 40 characterization (general) 118–21, 135–6, 140, 229–30, 244–5 cowardice 135–6 effeminate 136–8 hospitality 230, 236, 245 hubris 84–5, 93–4, 118–19, 138 ignorance 136–8, 238, 241 lesser than Clytemnestra 119–20, 137, 140 n. 154 referring phrases used to identify 66, 119–20, 232 role in the prologue 65–6 seen as rehabilited/sympathetic 229–30, 244–5, 246–7 tyrannical qualities 137–8, 140 n. 154, 245 see also messenger scene: sacrifice/ religious aspects in; Orestes: ‘ugly’ murder of Aegisthus Aeschylus 54, 271–2 Agamemnon 79, 90 n. 18, 110 n. 73, 129 n. 121, 140, 141, 144, 150, 208, 258 n. 16 Choephori 1 n. 1, 40 n. 70, 55, 57–8, 103 n. 58, 106, 123, 171, 204, 219, 223–4, 247; see also Index Locorum Agamemnon: grave 106–7, 111, 112, 120 n. 97, 171; see also Clytemnestra: indictment of Agamemnon; Electra: distress over father’s death/preoccupation with

father; matricide: link to the murder of Agamemnon age (as determinant of language use) 26, 91 n. 20, 125–6 agōn 143–50, 248–68 relation to oratorical practice/ contemporary background 144, 248–9 structure of Clytemnestra’s speech 250–1, 253–5, 261, 265 structure of Electra’s speech 251–2, 255–6, 263, 266 use of generalizations in 259–60, 263; see also Clytemnestra: rhetoric in the agōn; Electra: agōn speech similar to kakology Alcestis, see Euripides: Alcestis ambiguity 16, 25, 33, 43–4, 98 n. 40, 105 n. 61, 169, 185, 240–1; see also dramatic irony amoibaion 87–90, 95, 154–9 anapaest 141 Andromache, see Euripides: Andromache animal, see hunting and animal imagery ‘anti-heroic’ tone, see ‘heroic’ (or ‘anti-heroic’) tone/figures aorist: constative (indicative) 243 counterfactual (indicative), see indicative: counterfactual gnomic (indicative) 42 n. 79, 240 imperative, see imperative: present vs. aorist indicative 50, 63–4, 110, 136, 235–7, 239, 240, 242, 254–5, 256–8 (absence of), 266 infinitive, see infinitive: present vs. aorist participle 63–4 tragic (indicative) 172 n. 11 Apollo 38, 159–60, 162–3, 216 aposiopesis 160 n 209 apposition 65, 156, 170 n. 9

OUP CORRECTED PROOF – FINAL, 4/1/2017, SPi

Subject Index argument (text type) 47–50, 258–65; see also agōn; kakology Aristophanes 27, 28 n. 48, 29 n. 50 Frogs 62–3, 64 n. 10, 181 n. 35 see also female and male speech: in Aristophanes; Index Locorum article 137 n. 146, 218, 240–1 aside 143 n. 162, 153 aspect, see imperative: present vs. aorist; infinitive: present vs. aorist asyndeton 112, 116–17, 135, 138, 149, 156 n. 198, 180, 199–200, 220 n. 89, 266 n. 30 athletic imagery 56–7, 104, 128–9, 240, 242 attribution (of lines to speakers), see textual criticism: line-to-speaker attribution audience expectation 102, 152 n. 187, 171, 236, 241–2 authority, see generalization: as appeal to authority; power differences between interactants background (in narrative), see narrative: foreground and background bald, see politeness theory Bauformen 7–8; see also agōn; amoibaion; choral ode; Euripides: formality; kommos; messenger scene; monody; prologue monologue; rhēsis; stichomythia biographical fallacy 178 n. 26 CA, see conversation analysis (CA) captatio benevolentiae, see rhetoric case (grammatical): accusative 114, 257 dative 139, 181 n. 34, 219, 240 genitive 98 n. 40, 119 n. 94, 138 n. 150, 252 nominative 259 n. 18 vocative, see address, forms of Cassandra 254, 260–2, 267 in Aeschylus’ Agamemnon 90 n. 18 Castor 78, 114–16, 147, 159–64, 247 n. 40, 252 characterization: interface with textual criticism 189–90 as topic in literary criticism 52–4; see also Aegisthus; Castor;

309

Clytemnestra; Electra; Orestes; Peasant choral ode 59 Chorus: affectionate to Electra 90–2, 158 evaluation and treatment of Clytemnestra 141, 145 evaluation of Aegisthus 140–1 failure to understand Electra 87–9, 90–2 punctuating comments 140–1, 145, 150, 258 class, see social status and class Clytemnestra: in Aeschylus’ Agamemnon, see Electra: modelled on Aeschylus’ Clytemnestra conciliatory 47, 146–7, 150 fear of popular censure 144–5, 262, 267 impatience/superiority over Electra 12, 47, 142–3, 150, 152 indictment of Agamemnon 254, 260–1, 265, 267 language determined by gender 146, 259, 262 luxury, wealth, appearance 118, 141–2, 153, 212, 256 rhetoric in the agōn 143–5, 250–68; see also agōn: structure of Clytemnestra’s speech role in the prologue 66 seen as sympathetic 66 n. 19 use of generalizations 46–7, 150, 259; see also Electra: hatred for Clytemnestra and Aegisthus colloquialism 69 n. 29, 120, 219 comedy 97 n. 35, 198, 218 n. 76; see also Aristophanes command, see directive commissive, see speech act theory conditional clause 121, 124 n. 112, 126, 149, 202, 257, 263–4, 266 conditional relevance, see conversation analysis (CA) connective particle, see particle; asyndeton contact 190, 215–16, 220, 228; see also syntax: picked up or completed by other speaker; syntax: suspended conversation analysis (CA) 9–14 adjacency pair 9–13, 85, 100, 105, 127, 145 n. 169, 151, 217

310

Subject Index

conversation analysis (CA) (cont.) conditional relevance 11, 217 n. 74 counter 217 n. 74 expansion 11–14, 72 n. 37, 77 n. 53, 100–1, 145, 152, 153, 162, 163, 217 floor 9, 100, 105, 109, 127 preference organization 10–11, 73 n. 42, 77, 85, 88, 100, 108, 143, 151 selection 9–10, 14, 127, 161, 163, 216, 241 topic management 14, 91, 100, 152, 170–1, 216 turn-taking 9, 100 conversational implicature, see implicature counterfactual, see indicative: counterfactual culture and cultural norms as determinant of language use 4, 10–11, 26; see also sociolinguistics ‘deglamorization’, see ‘heroic’ (or ‘anti-heroic’) tone/figures deictic: deictic centre 67 n. 21, 136 see also person (grammatical); pronoun: personal; pronoun: demonstrative; tense usage democracy 57, 103 n. 58, 146; see also ethics description (text type) 7, 110–11, 136, 233–4, 238, 240, 256–7 deus ex machine, see Castor Dioscuri, see Castor direct speech 120, 158, 232–3 n. 10, 235–7, 238, 241 directive (commands, requests, etc.) 72–3, 76, 77, 78, 99, 142 n. 156, 182, 184 n. 40, 207–8, 221–2 as speech act type, see speech act theory see also future: indicative with οὐ μή; imperative; subjunctive: prohibitive (with μή) discourse cohesion 47–50 discourse modes, see text types discourse structure of drama 4–5 dispreferred, see conversation analysis (CA): preference organization ‘double vision’ 103–4, 140, 244–5 dramatic irony 43–4, 98 n. 40, 99, 102, 103–4, 121 n. 99, 126, 141, 144, 146,

152, 153, 166, 167 n. 3, 169, 170, 176 n. 21, 180, 183, 185, 186, 187, 211, 233–4, 238, 240, 241, 245, 246, 265, 267; see also ambiguity; audience expectation Electra 80–164; affection for the Chorus 91–2 affection for the Peasant and behaviour as wife 73, 75, 83, 85–7, 91–2, 106 agōn speech similar to kakology 147, 258, 262–3 appeals to Orestes’ manhood 118, 120–1, 221, 225 appropriation of male standards 134, 139, 259–60 balance between concerns over siblings’ exile and concerns over father’s murder 54, 84–5, 88, 95–7, 109, 111, 118, 122, 148, 151, 263–4, 267–8 characterization (general) 54–5, 85, 90–8, 108, 122, 123, 139–40, 153–4, 204, 226, 262–3, 267–8 cleverness with argument (sophia) 123, 125; see also Electra: rhetoric consistency vs. variation in speaking style 80–1, 128, 153–4 distress over exile and degradation 54, 84–5, 87, 88, 95–6, 101, 108, 108–9, 118, 142–3, 148–9, 151, 250, 264–5, 268 distress over father’s death/ preoccupation with father 47, 54, 84–5, 88, 95–6, 108, 108–9, 147, 216–18 distress over Orestes’ exile/ expectations of Orestes as saviour 85, 95–6, 98–100, 102–5, 108, 108–9, 118, 148–9, 151, 264–5, 268 in Euripides’ Orestes 62 n. 5 focus on sexual topics 130, 137–40, 148, 262–3 hatred for Clytemnestra and Aegisthus 54, 96–7, 106–7, 108, 118–19, 128, 153, 157, 204, 226 as ‘heroic’ (or ‘anti-heroic’) figure, see ‘heroic’ (or ‘anti-heroic’) tone/ figures

OUP CORRECTED PROOF – FINAL, 4/1/2017, SPi

Subject Index isolation and incompatibility of circumstances 54, 93–7, 118 and lamentation 30, 81–3, 88–90, 97, 101, 157 language determined by gender 30–1, 39, 90, 91–2, 97–8, 101–2, 109–10, 122, 129, 131–4, 137, 139–40, 146, 148, 157, 163, 201, 207–8, 210, 212, 263 and lyric 89–90, 154–9, 163 and marriage 93–5, 97, 105–6, 113–15, 117–18, 138, 147, 157 message/appeal to Orestes 108–22 miscommunication with Peasant and Chorus 83, 86–7, 88–9, 90–2, 97 modelled on Aeschylus’ Clytemnestra 141 negative scholarly consensus on character 54, 92–7, 112, 125, 161 personal involvement in planning and executing matricide 127, 157–8 regret after matricide 154, 157–8 rhetoric 95, 108–22, 130–41, 147–50, 251–3, 255–8, 259–60, 262–5, 266–8 in Sophocles’ Electra 88 n. 15, 96 spontaneous ‘flares’ or ‘surges’ in language 101–5, 108, 122, 140 n. 152, 148–9, 252, 265, 267–8 treatment of the Old Man 125–6 use of generalizations 130–4, 139, 147–8, 259–60, 263 wilfulness (authadeia) 123, 127, 151, 226 epinician, see athletic imagery ethical dative, see case (grammatical): dative ethics 42 n. 77, 57, 96, 131–4, 137, 139, 148, 177, 178–80, 182–3, 243–4, 259–60; see also generalization: as appeal to authority; Peasant: moral convictions ēthos, see characterization; Index of Greek Words: ἦθος Euripides: Alcestis 89 Andromache 82, 89, 125 Electra within his wider oeuvre 1 n. 1, 52 n. 95, 82 formality 62–3, 110–11 n. 75, 145 Helen 82 Heracles 179 Heraclidae 125

311

Hippolytus 179 Hypsipyle 89 interrogatory nature of work 160 Medea 179 Supplices 131 n. 128 and tradition, see tradition (literary) Troades 83, 144; see also Index Locorum everyday language 2–4, 17; see also colloquialism evidentiality 68 n. 26, 120 n. 97 expansion, see conversation analysis (CA): expansion explanation (in paroemiology), see generalization face-threatening act (FTA), see politeness theory face, see politeness theory facework, see politeness theory felicity conditions, see speech act theory female and male speech 26, 27–31, 271 in Aristophanes 28 n. 48, 29 n. 50 differences at micro vs. macro level 28–9 female anxiety over popular contempt 129, 134, 262 gendered spheres 118, 128–9, 134, 137, 139, 146, 157, 213 n. 64; see also female and male speech: verbal genre in lamentation/display of grief 29, 30, 91–2, 97, 102, 157 in lyric 29 in prayer 201 single-sex vs. mixed groups 29–30, 91–2, 109–10, 132–4, 157, 271; see also power differences between interactants in supplication 156 verbal genres 29, 30, 129, 212 see also Clytemnestra: language determined by gender; Electra: language determined by gender; Orestes: language determined by gender female sphere, see female and male speech: gendered spheres first part, see conversation analysis (CA): adjacency pair

312

Subject Index

first-person generalization (saying, gnōmē, etc.), see generalization focalization 242 foreground (in narrative), see narrative: foreground and background forms of address, see address, forms of forms, see Bauformen four-part division of speeches, see rhetoric free speech (parrhēsia) 145–7 FTA, see politeness theory: facethreatening act (FTA) future: indicative 42, 121, 149, 221 n. 94, 222, 250, 266 n. 34, 267 indicative with οὐ μή 221–2 participle 222 Ga 201 gender, see female and male speech general reflection, see generalization generalization 40–7, 186–7, 271 as appeal to authority 44–5, 132, 134, 139, 148, 259–60 as closural device to cap discussion 47, 100, 170 distancing 44, 133, 168–9, 170, 173 and dramatic irony 43–4, 169, 176, 183 explanation 43, 137 first-person/second-person/thirdperson 45–7, 74 n. 45, 75, 132–3, 138, 147, 150, 169, 259, 263 indirect 43–4, 133, 137 linguistic markers of 42, 47, 168 n. 6 meaning determined by context 43–4, 169, 176 and politeness 45–7, 75, 133 social dimensions of use 44–7, 132–3; see also agōn: use of generalizations in; Clytemnestra: use of generalizations; Electra: use of generalizations; kakology: use of generalizations in genitive absolute, see participle genitive, see case (grammatical) gerund 172 gnōmē, see generalization gods (relation to humans) 159–60, 162–4 gossip, see female and male speech: verbal genres

grave, see Agamemnon: grave Gricean pragmatics 20–6, 72, 75 n. 49, 91, 103 n. 57 conversational implicature, see implicature cooperative principle 20–1, 23 maxim of manner 21 maxim of quality 21 maxim of quantity 21, 103 n. 57 maxim of relation 21, 72, 75 n. 49, 91, 221 neo-Gricean pragmatics 22–3; see also relevance theory hapax legomenon 78 hedge 10, 33–4, 44, 85, 102, 120 n. 97, 126 Helen 91, 147, 252, 262–3; see also Euripides: Helen; Euripides: Troades Hera 201–2 ‘heroic’ (or ‘anti-heroic’) tone/ figures 54, 61, 88 n. 15, 93 n. 27, 97 n. 35, 103 n. 58, 104, 182 hesitation scene 204–28 historic present, see present: historic (indicative) Homer 3, 28 n. 49, 40, 41–3, 46, 133, 183 Iliad 46, 103 Odyssey 55, 129 n. 123, 140, 245 see also Index Locorum humour 87, 125 hunting and animal imagery 56, 118, 153, 213, 244, 246 hyperbaton, see word order: in noun phrases hypothetical syllogism 260 Hypsipyle, see Euripides: Hypsipyle illocution/illocutionary act, see speech act theory imagery, see athletic imagery; hunting and animal imagery; sacrifice imagery imperative 70, 73 n. 44, 76, 77, 78, 142, 152, 196–9 present vs. aorist 99 n. 41, 129 n. 124, 196–7 imperfect 50, 110, 136, 147, 233–4, 238–9, 240, 242, 254, 256–7 conative 255 immediative 239–40 see also indicative: counterfactual

Subject Index impersonal 76, 172, 221 n. 92 implicature, 21–2, 24–5, 36 n. 64, 72, 75 n. 49, 148 n. 175, 220, 221; see also Gricean pragmatics; relevance theory impoliteness, see politeness theory indicative: counterfactual 124 n. 112, 167, 252, 255, 260 in hesitant suggestions (with μή) 126; see also aorist; future; imperfect; perfect; pluperfect; present indirect speech acts, see speech act theory indirectness 19–20, 22, 33, 43–4, 73, 75 n. 49, 133; see also implicature infinitive: expression of subject 114 present vs. aorist 129 n. 124, 174 n. 17 insert expansion, see conversation analysis (CA): expansion interactional particle, see particle interjection 206 n. 42; see also Index of Greek Words interpolation, see textual criticism: interpolation interrogative, see question interruption 30, 127, 163, 200, 216 Iphigenia 147, 252–3, 253–4, 257, 260–2, 264 irony 22; see also dramatic irony kakology 128–141 similarity to agōn speeches 130, 147; see also Electra: agōn speech similar to kakology structure 130 use of generalizations in 130–4, 139 kommos 154–9 lamentation, see Electra: and lamentation; female and male speech: in lamentation left dislocation, see word order: in clauses linguistic underdeterminacy 15 n. 26 confluence with narratology 49 n. 90 linguistics (modern): application to literary texts 2–5 attuned to differences between text types 8 universality of features described, see universals

313

literal meaning, see meaning locus amoenus 234 locutionary act, see speech act theory lyric, see amoibaion; choral ode; Electra: and lyric; monody; kommos Lysias, Epitaphios (oration 2) 131 n. 128 male speech, see female and male speech male sphere, see female and male speech: gendered spheres manuscripts, see textual criticism marriage, see Electra: and marriage; Electra: affection for the Peasant and behaviour as wife matricide: as climax/crux of the play 154, 247 link to the murder of Aegisthus 223–6, 245 link to the murder of Agamemnon 141, 153, 223–6 troubling notes in build-up to 55, 56–7, 149–50, 152–3, 245 see also Electra: personal involvement in planning and executing matricide maxim, see Gricean pragmatics meaning: (neo-)Gricean theories of, see Gricean pragmatics literal 15, 20 n. 35, 25 referential vs. functional 48 messenger scene 185–6, 229–47 epic colouring 235 n. 17 modulation of suspense 231 sacrifice/religious aspects in 234, 243–5, 246–7 use of tenses in 232–43 Messenger: as eyewitness 233 n. 12 partisanship/ lack of objectivity 232–3, 240, 243, 246 see also messenger scene metaphor 22, 42–3; see also athletic imagery; hunting and animal imagery; sacrifice imagery metre 141 n. 155, 156 n. 198, 161, 215 n. 68; see also word order: enjambement misfire, see speech act theory: felicity conditions monody 90, 95

314

Subject Index

morality, see ethics myrtle 234 myth, see tradition (literary) narrative 47–50, 62–70, 229–47, 253–8 foreground and background 238–9, 254–5 near-miss pattern 102, 104–8, 171–2; see also suspense negation 72 n. 39, 103 n. 57, 121, 145 n. 167, 220, 221 n. 92; see also indicative: in hesitant suggestions (with μή); subjunctive: prohibitive (with μή); subjunctive: with οὐ μή negative face, see politeness theory negative politeness, see politeness theory nominative, see case (grammatical) off-record politeness, see politeness theory Old Man 107, 123–7, 171, 202–3, 236 optative: iterative 257 potential (with ἄν) 42 n. 79, 77, 109–10, 124, 146, 167, 221 n. 92 in primary sequence (purpose clause) 84 n. 6 in wishes 99–100, 107 n. 66, 167, 198 oracle 163, 183 n. 39, 216–17, 221, 225–6, 228 oratorical genre, see rhetoric order, see directive (commands, requests, etc.) Orestes 165–87 care for strategy 209, 236–7 characterization of (general) 55–6, 165–6, 247 divine compulsion, see oracle as Electra’s kurios 66, 105, 172 as ‘heroic’ (or ‘anti-heroic’) figure, see ‘heroic’ (or ‘anti-heroic’) tone/ figures hesitation over matricide 211–14, 215–26, 228 in the kommos 156–7 language determined by gender 157, 201, 210 manhood, see Electra: appeals to Orestes’ manhood negative scholarly consensus on character 55–6, 229–30

prolonged disguise 55–6, 165–6, 166–85 reserved demeanour and suppression of emotion in disguise 106, 166–85 significance to Electra, see Electra: distress over Orestes’ exile/ expectations of Orestes as saviour in Sophocles’ Electra 96 surprise at Peasant’s character 106, 172, 184–5 ‘ugly’ murder of Aegisthus 55, 229–30, 243–7 use of generalizations 156–7, 168–85, 186 use of grammatical features (person, mood, voice, etc.) 167, 172, 176, 185 ‘out’, see politeness theory: off-record politeness pankoinon, see tautology papyrus 1 n. 1, 177 n. 25, 193 n. 10 paragraphos 160 n. 210, 191, 193–5, 220, 221, 223 n. 101, 226 n. 107 parainesis 180, 183 parodos 87–90 paroemiology 40–7; see also generalization parrhēsia, see free speech (parrhēsia) participle 72 n. 38, 117, 240–1, 254, 257; see also aorist: participle; future: participle; perfect: participle; present: participle particle 28, 48–9; see also Index of Greek Words passive 107 n. 64, 172, 221 n. 92 Peasant 60–79, 177–85 conflict of politeness and moral stance 70, 77–8 failure to understand Electra 86–7, 92 moral convictions 69–70, 72, 74–5, 77–8 overall function in the play 60–1 personal tone of prologue narrative 66, 69–70 polite/affectionate to his wife 70, 75, 77–8, 90 register of speech 69 n. 29, 78–9 use of generalizations 74 n. 45, 75, 77, 79; see also Electra: affectionate to the Peasant and behaviour as wife; Orestes: surprise at Peasant’s character

OUP CORRECTED PROOF – FINAL, 4/1/2017, SPi

Subject Index perfect: indicative 64, 110, 119 intensive interpretation 257 n. 15 interpretation of perfect passive 221 n. 92 participle 257 performative, see speech act theory perlocutionary act/effect, see speech act theory person (grammatical): first 67, 110, 118, 142 n. 156, 149, 156, 183 n. 38, 202, 209 233, 233–4 n. 13, 250 second 75, 76, 120, 156, 163, 183 n. 38, 209, 216, 222 third 106, 167, 168, 172, 176, 185, 222, 233–4 n. 13 person (of generalizations), see generalization: first-person/secondperson/third-person Pindar, see Index Locorum planning scene 126–7, 236 Plato, see Index Locorum pluperfect 50, 136 n. 142 Plutarch, see Index Locorum politeness theory 31–9 bald 32–3, 38, 70, 78, 104 n. 59, 134, 142, 152 face 32, 35, 45–6, 73–5, 101, 104 n. 59, 121, 126, 221 face-threatening act (FTA) 32–8, 45, 70–1, 73, 75, 76, 76–7, 126, 221 facework 36, 38–9, 46, 71, 73, 142 n. 157 first-order vs. second-order politeness 35–6 impoliteness 31 n. 55, 34, 35–7, 76, 101, 104–5, 108, 125–6, 142 n. 157 negative politeness 33–4, 71, 76 off-record politeness 33–4, 73, 75, 85 politeness strategy 32–4; see also politeness theory: negative politeness; politeness theory: positive politeness politic behaviour 36, 38–9, 71, 73 n. 44, 126, 142 n. 157 positive politeness 33–4, 91 task-orientation 73 n. 44 politic behaviour, see politeness theory positive face, see politeness theory positive politeness, see politeness theory post-expansion, see conversation analysis (CA): expansion

315

power differences between interactants 14, 29–30, 33, 37, 39, 46–7, 109–10, 132–3, 146, 162, 208; see also age (as determinant of language use); female and male speech: single-sex vs. mixed groups; politeness theory; social status and class pragmatics 14–26, 269–70 delineation of the field 16–17, 26; see also speech act theory; Gricean pragmatics; relevance theory prayer 193–204 pre-expansion, see conversation analysis (CA): expansion predicative adjective 170 n. 9 preference organization, see conversation analysis (CA) preferred, see conversation analysis (CA): preference organization present: historic (indicative) 50, 64, 65, 233–4, 235–7, 239, 241, 242 imperative, see imperative: present vs. aorist indicative 64, 110 infinitive, see infinitive: present vs. aorist participle 257 prohibition, see directive (commands, requests, etc.) prologue monologue: personal tone 67–70 structure 63–6 as audience address 78 n. 56 in Euripides generally 62–3 pronoun: demonstrative 202 interrogative, see question personal 67, 118, 149, 161, 163, 202 relative, see relative clause see also Index of Greek Words proofs, see rhetoric proverb, see generalization pseudo-Demosthenes, Against Neaera, see Index Locorum punctuation 112, 116–17, 191, 215 purpose clause 176 n. 22 Pylades 138 n. 148, 139, 159 n. 206, 183 n. 39, 202, 207

316

Subject Index

question 71–2, 85–6, 107, 120–1, 126, 127, 156, 163, 170, 171, 209–10, 215 n. 68, 219, 220–1, 222, 238, 241 n. 28, 264, 266 indirect 117 question–answer pair 10, 12–13, 100, 104–5, 127, 151, 162 rhetorical 34, 50–1, 72–3, 101, 135, 148–9, 159 n. 206, 215–16, 219–20, 260–1, 264–5 recognition scene 57–8, 122–6 linked to the hesitation scene 123, 125 prefigured or delayed 102, 106, 171–2 reductio ad absurdum 216 n. 72, 260 relative clause 65, 114, 116, 119, 180 n. 32, 199–200, 210–11, 222–5, 252, 256 relative connection 256 n. 14 relevance (contextual) 40–1, 59, 174–5, 177–85, 186–7 relevance (linguistic), see relevance theory; Gricean pragmatics: maxim of relation relevance theory 23–6, 36 n. 64 request, see directive (commands, requests, etc.) result clause 254 revenge, problematization of 149–50, 245, 266–7; see also Electra: hatred for Clytemnestra and Aegisthus rhēsis, see agōn; Electra: message/appeal to Orestes; kakology; messenger scene rhetoric: captatio benevolentiae/introductory technique 137, 144–5, 251, 259 four-part division of speeches 249–50 oratorical genre (forensic, epideictic, etc.) 130–1, 135–6 proofs (artificial vs. artless) 260 see also agōn; Clytemnestra: rhetoric in agōn; Electra: rhetoric rhetorical question, see question sacrifice imagery 55, 56, 152–3, 209 n. 50, 224–5, 234, 244; see also messenger speech: sacrifice/ religious aspects in sarcasm 22 n. 37, 119, 121, 129 n. 121, 142, 224 saying, see generalization scholia 1 n. 1

scolion 244 n. 33, 244 n. 35 second part, see conversation analysis (CA): adjacency pair second-person generalization (saying, gnōmē, etc.), see generalization: first-person/second-person/thirdperson selection, see conversation analysis (CA) self-selection, see conversation analysis (CA): selection sentence type 19–20 sententiousness, see generalization Shakespeare, Hamlet 243 n. 31 silence 9, 29 social status and class 26, 33, 45–7, 69, 72, 79, 92, 97, 184–5; see also power differences between interactants sociolinguistics 26–39 difficulty of applying to ancient Greek 27, 28 see also age (as determinant of language use); female and male speech; generalization: social dimensions of use; politeness theory; power differences between interactants; social status and class song, see lyric Sophocles 53 n. 100, 54, 271–2 Ajax 44 Electra 1 n. 1, 40 n. 70, 55, 59, 80, 82 n. 2, 88 n. 15, 96, 103 n. 58, 119 n. 94, 129 n. 123, 150, 199, 210–11 see also Index Locorum speech act theory 17–20, 269–70 classification of speech act types 18–19 commissive 19, 141–2 directive 19–20, 73 n. 44; see also directive (commands, requests, etc.) felicity conditions 18, 20, 88–9, 142 illocution, illocutionary act 18–19, 148 n. 175, 216 indirect speech acts 19–20, 22, 33; see also indirectness locutionary act 18 performative 18, 111, 121 perlocutionary act/effect 18 speech, see direct speech staging 71 n. 36, 183–4 stichomythia 98–108, 188–228; see also syntax: picked up or completed by other speaker; syntax: suspended

Subject Index stigme, see punctuation Stobaeus, see Index Locorum subjunctive: deliberative 215 prohibitive (with μή) 76 with οὐ μή 221 n. 94 supplication 111, 155–5 suspense 55, 102, 153, 169, 230–1, 235–8, 241–2; see also audience expectation symmetry (in division of parts) 155–6, 202 n. 36 syntax: picked up or completed by other speaker 107, 197–9, 199, 216, 221–2 suspended (continued by same speaker after turn by other speaker) 100 n. 50, 198 n. 23, 200–1, 216 see also apposition; asyndeton task-orientation, see politeness theory tautology 22, 103, 107–8, 152 n. 188 telicity 233 tense usage: in generalizations 42 in narrative 49–50, 63–5, 232–43, 253–8 present vs. past tenses 110, 119, 257 see also aorist; future; imperfect; perfect; present; pluperfect text types 7–8; see also narrative; argument (text type); description (text type) textual criticism: interpolation 40, 84 n. 6, 112 n. 80, 131, 177–83, 184 n. 40, 196 n. 18, 271 line-to-speaker attribution 51, 154–6, 160–2, 188–228; see also paragraphos manuscripts 113, 154, 160, 188–93 principles for use 50–1

317

transposition 84 n. 6, 114 n. 84, 151 n. 184, 162, 174–6, 193–4, 203 n. 37, 205, 211–12, 214 n. 67, 261 see also characterization: interface with textual criticism; Index Locorum theme, see word order: in clauses themes and motifs 56–7; see also athletic imagery; ethics; hunting and animal imagery; sacrifice imagery third-person generalization (saying, gnōmē, etc.), see generalization: first-person/second-person/thirdperson tokens, see recognition scene topic (in CA), see conversation analysis (CA): topic management topic (word order), see word order: in clauses tradition (literary) 54, 56, 57–8, 97 n. 35, 127, 136 n. 139, 140, 159 n. 206, 171, 198 n. 24, 204, 245 transposition, see textual criticism turn-taking, see conversation analysis (CA) universals 2–3, 10–11, 23, 34–5 vocative, see address, forms of wisdom 45, 175; see also generalization; Index of Greek Words: σοφός/σοφία wish, see optative: in wishes; indicative: counterfactual word order: in clauses 65, 75 n. 48, 112, 215 n. 68, 216 n. 72, 218, 264 enjambment 119–20 in noun phrases (hyperbaton; adjective–noun ordering) 86, 119 Zeus 191 n. 8, 201

Index of Greek Words Words are given in their dictionary form, except in the case of some vocatives. δμώς 207–8 (ὁ) αὐτός 221–2 δόλος 222–6 ἀδελφός/ἀδελφή 106, 167, 264 n. 27 δύστηνε 73, 90 αἰσχρός 75, 104 n. 59 αἰσχύνομαι 115 n. 87 ἀλλά 12, 77, 121, 124, 134 n. 134, 151 n. ἔα 69 n. 29, 74 184 152, 222 (ἀλλ’ ἤ) ἐγώ 139 (μοι), 149, 153 (μοι), 158, ἀμαθής/ἀμαθία 175; see also Aegisthus: 217–18, 221 n. 91, 264 εἰ (τοι) δοκεῖ ignorance σοι 69 n. 29, 73–4, 77 ἄν, see indicative: counterfactual; εἰ, see conditional clause optative: potential (with ἄν) εἶἑν 135, 206–7 ἀναίνομαι 115–16 εἶμι 222 ἄπειμι/πάρειμι 102–3, 120–1, 126, 108 ἐπεί 232 n. 67, 185 ἐπεισφρέω 254 n. 8 ἆρα 121 ἔπειτα 123–4 ἄρα 214 ἔρρε 138 ἄρκυς 213 εὐανδρία 178–83 αὐθάδης/αὐθάδεια 151; see also Electra: ἐχθρός 232 wilfulness (authadeia) ζεύγνυμι 118 βοηδρόμος/βοηδρομέω 210 n. 54 ἤ 112 n. 80 γάρ 49, 72, 85, 86, 142, 181, 182, 183 n. ἦ 222 39, 215; see also γὰρ οὖν (ἀλλ’ ἦ); see also ἆρα γὰρ οὖν 173–4 ἦθος 180–1, 184 γε 68 n. 25 (γε μήν), 69, 101 n. 51, 104 n. ἡμεῖς 118 59, 107, 158, 196–9, 199–200, 213 n. θρήνημα 30, 81 63 (καὶ μήν . . . γε), 218 (δέ . . . γε) θύγατερ 125 γείνομαι 210–11 θυραῖος 174, 241 γόος 81 γυμνάς 113 ἱερά 115–16 γύναι/γυναῖκες (voc.) 30 γυνή 113 ἱκετεύω 111, 121 ἴσως 102 δαί 219 δέ 49, 46, 113, 116, 124, 130, 134 n. 134, καί 49, 64 n. 14 (καί . . . μέν) 100 n. 50, 162, 217, 218 (δέ . . . γε), 251 n. 4, 203, 217, 225; see also καὶ μήν 263; see also δὲ δή; μέν . . . δέ καὶ μήν 135, 213 n. 63 (καὶ μήν . . . γε) ( . . . δέ); μέν οὖν . . . δέ καὶ ταῦτα 72 n. 38 δὲ δή 69 n. 29, 100, 170 καινός 232–3 δέσποινα 125 καίτοι 134 n. 134, 251 n. 4 δή 68–9, 84; see also δὲ δή καλέω 133 n. 133 δῆτα 196, 241 καλῶς 120, 213 δικάζω 266 κασίγνητος 98 n. 40, 99, 155, 167, δίκην δίδωμι 138 n. 150, 219 176, 185

OUP CORRECTED PROOF – FINAL, 4/1/2017, SPi

Index of Greek Words κατακλαίω 81 κλεινός 119–20, 232 n. 9 κυρέω 233–4 λόγος/λέγω 98 n. 40, 129 n. 124, 250 μέν . . . δέ ( . . . δέ) 64 n. 14, 112; see also μὲν οὖν . . . δέ μέν 64 n. 14 (καί . . . μέν), 117, 221 n. 91; see also μέν . . . δέ ( . . . δέ); μὲν οὖν . . . δέ μὲν οὖν . . . δέ 151 n. 185, 235 μέντοι 144 n. 164 μή, see future: indicative with οὐ μή; negation μήν 68 n. 25 (γε μήν), 152; see also καὶ μήν; μῶν μήτηρ 119 n. 94, 157, 210–11, 215 n. 68, 216–17, 217–19 μνηστεύω 114–15 μοι, see ἐγώ μῶν 209 n. 53, 215 νίκη/νικάω 202 ξένος 167 n. 3 ξηρός 101 n. 51 ὁ (ἡ, etc.), see article ὅδε (ἥδε, etc.) 202, 225 οἶδα 217–18 (ἐγῷδα), 257 οἴμοι 172 οἰμώζω 172 n. 11 ὅς (ἡ, etc.), ὅστις (ἥτις, etc.), see relative clause οὔ που 170 n. 8 οὐ(κ), see negation; future: indicative with οὐ μή; οὔ που οὐκοῦν/οὔκουν 101, 141–2 οὖν 49, 183; see also γὰρ οὖν; μὲν οὖν . . . δέ οὔτε 149 πάθος 158 παῖ (voc.) 91, 125, 213 n. 66 πάρειμι, see ἄπειμι/πάρειμι παρήϊς 254 παρθένος/παρθενικός 69, 91 n. 20, 92, 95, 138

319

παρρησία, see free speech (parrhēsia) πατήρ 96, 216–17 πονηρός 70 πόσις 106 n. 63, 119, 223–4 πότνια 125 που 76 n. 50; see also οὔ που προοίμιον 251–2 πρὸς θεῶν 184 πῦρ 157 πῶς 215–16 ῥέζω 158 ῥυθμός 231 σκῆψις 253 σοφός/σοφία 175; see also Electra: cleverness with argument (sophia) στολή 212 n. 61 σύ 142, 163, 218 n. 79 (σέθεν) συμφορά 101–2 σφαγή 205, 209 σώφρων 70 τάλαινα (voc.) 39 τέκνον 125, 142 n. 160, 146–7 τητάομαι 94 n. 31 τί δέ; 69 n. 29, 143 τίκτω 157 τίς/τί, see question τοι 74, 75, 77, 143, 143 n. 162, 220 n. 89 (τἄν) τοιοῦτος 182, 250 τολμάω 107 n. 64 φεῦ 171–2, 179, 215 φθόνος 129 n. 122, 262 φίλη/φίλαι (voc.) 88, 155, 158 φιλία/φίλος 158–9 φίλτατε (voc.) 39, 99 φρήν 252 χαίρω 257 n. 15 χρή 42, 76, 109 ψόγος 129 n. 122, 262 ὧδε 134 n. 134 ὥσπερ 216 ὥστε, see result clause