Greek Tragic Style: Form, Language and Interpretation 0521848903, 9780521848909

Citation preview

G R E E K T R AG I C S T Y L E Form, Language and Interpretation

Greek tragedy is widely read and performed, but outside the commentary tradition detailed study of the poetic style and language of Aeschylus, Sophocles and Euripides has been relatively neglected. This book seeks to fill that gap by providing an account of the poetics of the tragic genre. The author describes the varied handling of spoken dialogue and of lyric song; major topics such as vocabulary, rhetoric and imagery are considered in detail and illustrated from a broad range of plays. The contribution of the chorus to the dramas is also discussed. Characterisation, irony and generalising statements are treated in separate chapters, and these topics are illuminated by comparisons which show not only what is shared by the three major dramatists but also what distinguishes their practice. The book sheds light both on the genre as a whole and on many particular passages. r i c h a r d r u t h e r f o r d has been Tutor in Greek and Latin Literature at Christ Church, Oxford, since . His previous publications include The Meditations of Marcus Aurelius: A Study (), Homer, Odyssey 19 and 20 (), The Art of Plato (), Homer (Greece and Rome New Surveys in the Classics , ), and Classical Literature: A Concise History ().

GREEK TRAGIC STYLE Form, Language and Interpretation

R. B. RUTHERFORD

cambri dge uni versi ty p re s s Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, S˜ao Paulo, Delhi, Mexico City Cambridge University Press The Edinburgh Building, Cambridge cb ru, UK Published in the United States of America by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/ c R. B. Rutherford   This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published  Printed in the United Kingdom at the University Press, Cambridge A catalogue record for this publication is available from the British Library Library of Congress Cataloguing in Publication data Rutherford, R. B. Greek tragic style : form, language and interpretation / R. B. Rutherford. p. cm. isbn ---- (hardback) . Greek drama – History and criticism. . Poetics – History – To . I. Title. pa.r  ′ . – dc  isbn ---- Hardback

Cambridge University Press has no responsibility for the persistence or accuracy of URLs for external or third-party internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

For Emily and Nicholas

In my opinion we ought to refine, renew, and intensify the labours of the great scholars who have paved the way for future work, like them not shrinking from the minutest scrutiny while trying to see the wood as well as the trees. We may begin by observing words, their meaning, their structure and their order, and end with observing characteristic habits of the poet’s mind in shaping dramatic characters, bringing about a tragic tension, and revealing his religious convictions. Fraenkel, PBA  () 

The reader should be enabled to bathe in examples. Denniston, The Greek Particles vi

The prisoners sat in Poetry Appreciation chairs – strapped in. Vogons suffered no illusions as to the regard their works were generally held in. . . . The sweat stood out cold on Ford Prefect’s brow, and slid around the electrodes strapped to his temples. These were attached to a battery of electronic equipment – imagery intensifiers, rhythmic modulators, alliterative residulators and simile dumpers – all designed to heighten the experience of the poem and make sure that not a single nuance of the poet’s thought was lost. Douglas Adams, The Hitchhiker’s Guide to the Galaxy ch. 

Contents

Preface List of abbreviations Note on texts

page xi xvi xix

 Introduction I Starting points: three passages II Problems of definition (‘style’ and other terms) III Ancient and modern study of the subject

 Genre: form, structure and mode I II III IV V

Formal description Historical outline Generic appropriation and distortion Comedy and tragedy (paratragic and sub-tragic passages) Perceptions of the genre

 Words, themes and names

          

I Tragic diction, syntax and style in the prologue to Antigone II Setting the scene and establishing mood. Key words and dominant themes III Names: using, withholding and revealing them IV Forms of address; Sophocles, Philoctetes Appendix : Vocabulary Appendix : A note on alliteration and related phenomena

 The imagery of Greek tragedy I Introductory: the importance of the subject II Recurring or ‘thematic’ imagery in the Oresteia and elsewhere III Central fields of tragic imagery; border regions between god and man, bridged by imagery IV Personification, especially of divinised entities

vii

          

Contents

viii

 The dramatists at work: spoken verse I II III IV V

Preliminaries Long versus short; stichomythia Prologue Agon Messenger speech

 The dramatists at work: lyrics I Preliminaries: the tragic chorus II Three tragic choral songs (Aeschylus, Agamemnon, Sophocles, Electra, Euripides, Hippolytus) III Two songs on Athens (Euripides, Medea, Sophocles, Oedipus Coloneus) IV Song-speech combinations: a survey V Monody. Creusa’s monody in Euripides’ Ion VI New Music, new styles VII Late Sophoclean song: Oedipus Coloneus

 The characters of Greek tragedy I II III IV

Problems and approaches Character types, especially female The watchman in Aeschylus, Agamemnon Deceptive women: Aeschylus’ Clytemnestra; Sophocles’ Deianira V Neoptolemus in Sophocles, Philoctetes VI Changing perspectives: Euripides’ Hippolytus VII The drama of indecision: Medea

 The irony of Greek tragedy I Irony and related terms; the role of the gods; epic background II Malign irony: scenes of entrapment III Ignorance and recognition IV Divine manipulation and human downfall V Dionysus in the Bacchae Appendix: Metatextual irony

 The wisdom of Greek tragedy I II III IV

The general and the gnomic Novelty of thought and ideas Questions about deity Grandeur of expression: Aeschylus, Agamemnon, Sophocles, Oedipus Coloneus V General statements in Sophocles, Ajax

                                  

Contents

ix

 Epilogue



Bibliography



Select index of Greek words Index of topics Index of passages cited

  

Preface

Greek tragedy remains an inexhaustible source of enthralment and inspiration. Nevertheless, to add another book to shelves already crammed with valuable studies needs some justification. First, recent work on the tragedians outside the commentary tradition has focused on the life of the theatre, the social and political environment of the dramas, their relation to Athenian democracy and ideology, and the impact of the plays on later literature. My own emphasis is on the dramas as poetic texts. Second, although there are many valuable studies of the individual dramatists, some of which deal at some length with language and style (the works of Anne Lebeck on Aeschylean imagery or A. A. Long on abstractions in Sophocles come immediately to mind), there is no obvious work we can consult if we wish to understand more fully how the styles of Aeschylus, Sophocles and Euripides may be compared – what they have in common, how they differ, how far there exists a tragic koine or shared poetic language. My initial hope was that I might be able to provide such a book. The project has undergone considerable modification since then, but I have tried to keep the comparative aspect at the heart of it. It soon became evident that the absence of a work of the kind I was envisaging resulted in part from the richness of material and the difficulties of presentation. To analyse a passage of Greek, even one as short as ten or fifteen lines, in such a way as to go beyond superficialities while holding the reader’s attention, is difficult enough; to compare three passages (one from each major dramatist) is a cumbersome process, all the more so if one is to quote in both Greek and English in each case. Inevitably there have been compromises, but I hope to have struck a reasonable balance between quotation, discussion and comparison. At least it should be clear how much can be said, and how much is still to be done, in the critical study of tragic style and language. These terms, like others commonly used by myself and others (notably ‘rhetoric’), are hard to define and delimit: some of the issues are addressed in the first chapter. xi

xii

Preface

Here I mention only one point of general relevance. Although linguists naturally have other priorities, my own interest in the language of drama is that of an interpreter: my aim is not to describe the evolution of the Greek language or to locate particular features of the tragedians’ language in that development, but to understand how they used the linguistic and stylistic resources at their disposal, in conjunction with their other resources, for dramatic ends. Naturally this involves some detailed attention to particular linguistic features (such as poetic compounds, or oxymoron), but the goal is always a better understanding of their dramatic art. I am well aware that a specialist in linguistics would have produced a very different book. The best way to structure the work long remained a matter of perplexity. The introductory chapter paves the way for the enquiry, and describes some of the work which I have found most useful in earlier scholarship. The next two seek to describe first the formal aspects of the genre (form, indeed, will often be a preoccupation and sometimes threatens to oust language from the spotlight; but the two are hardly separable), then the main features of tragic vocabulary and expression, the so-called Kunstsprache. Thematic uses of key terms, and the diversity of forms of address, also receive attention. Chapter  deals with imagery and personification. Chapters  and , in some ways the heart of the book, attempt comparisons of similar or interestingly contrasting passages in the three dramatists, the first dealing with the spoken elements of tragedy (prologue-speeches, rhesis, stichomythia, agon, messenger speech), the second with lyric passages (both choral song and actors’ lyric). In each case a brief introduction outlines the typical features of these portions of the drama and suggests a number of contrasts between the three major dramatists. The remainder of the book consists of three essays in comparative interpretation, again making frequent use of quotations: these chapters address three topics which must be central to interpretation of Greek drama: characterisation, irony (particularly what is usually called dramatic irony, but other types are more briefly discussed), and the gnomic or ‘wisdom’ element. The epilogue draws some of the strands together and tries to define some of the key differences between the poets while also emphasising how much they have in common. The reader who finds it hard to see (or to remember) where a point is likely to be discussed is asked to make full use of the index. 

In view of this, it is a pleasure to anticipate the published version of Boas , a DPhil thesis, to be published as an Oxford Classical Monograph. I was able to read the thesis soon after sending in my own manuscript: his study will do much to bridge the gap between linguistics and literary criticism of tragedy.

Preface

xiii

Detailed though some of these discussions may be, the book is not and could not be exhaustive. My aim has been to approach the plays from many different angles, highlighting passages which are particularly rich and revealing; some scenes and central passages reappear in various contexts, considered from different points of view. Even the close readings of specific passages in chapters  and  are not meant to deal with every aspect; I count it a merit in a work that it leaves some of the thinking still to be done by the reader. Thirty-two plays survive from ancient times, as well as numerous fragments (in the case of Euripides the term ‘fragments’ is somewhat misleading, since we can form a very full assessment of several plays which still survive in substantial part). On the matter of authorship, I share the communis opinio that the Rhesus is not a work by Euripides, and think it likely to be later than the fifth century; the Prometheus Bound I take to be post-Aeschylean. This does not deprive these two plays of interest, but on the whole I have been less concerned with them: Prometheus has been the object of much detailed study, above all by Mark Griffith, while I am aware of a number of important commentaries in progress on Rhesus. As for the genuine plays, I have neither sought to provide equal coverage nor avoided the obvious high points of the genre. Aeschylus’ Oresteia, Sophocles’ Oedipus Tyrannus, Antigone and Philoctetes, Euripides’ Medea, Hippolytus, Bacchae regularly recur, and some other favourites, particularly the other Sophoclean plays and Alcestis, Ion, and Iphigenia in Tauris, run them close. Although all surviving plays are referred to somewhere or other, there are a few which are given little space, although they would doubtless repay more attention, which I hope others will give them. Translations are provided for all passages of quoted Greek; these are my own unless otherwise stated, but when I have borrowed those of others I have often modified them. Sometimes a translation alone is offered, when the point does not seem to depend on the detail of the Greek text. In certain cases I may have overemphasised one feature of the text at the expense of others in order to bring out the salient point in that context. It follows that the same passage may be translated rather differently in different parts of the book. I began thinking about the plan for the book as long ago as , and some sections derive from drafts of the late s; but the commissioning of my Classical Literature: A Concise History (published in ) deflected my research programme for some time, and the bulk of the writing of the present book took place between  and , much impeded by heavy administrative duties. The delay has however been beneficial in some

xiv

Preface

respects, in that I have been able to make use of some recent publications of central importance, notably Alan Sommerstein’s edition of Aeschylus, Donald Mastronarde’s long-awaited monograph on Euripides, and the monumental treatment of Euripides’ fragments by Richard Kannicht in the final volume of TrGF. Inevitably, other works have appeared too late for me to use them. I particularly regret that I have not been able to refer to the new commentary on Aeschylus’ Agamemnon by David Raeburn and Oliver Thomas (Oxford ). The Bibliography is long, but would have been many times the size if I had included everything that might be thought relevant to this enquiry. In particular I have been sparing in listing modern works on critical theory, stylistics and so forth, and studies of other poetic dramatists, above all Shakespeare. Omission of such works, therefore, should not necessarily be taken to imply ignorance. Although no part of this book has appeared in print before, some of the themes are addressed in an essay in a recent Companion to the Ancient Greek Language (Rutherford ), and a shorter account will be found in an entry on ‘Language’ which I have contributed to the forthcoming Wiley-Blackwell Encyclopedia of Greek Tragedy (edited by Hanna Roisman). I cannot do justice here to the many great scholars, some of them no longer living, whose lectures, teaching and publications have influenced my own work; but like all serious students of tragedy I stand on their shoulders. Several friends and colleagues have helped and encouraged me in the writing of this book. Robert Parker has probably forgotten the occasion, but a conversation about my future plans in  spurred me to think more seriously about this as a long-term project. Since then his cheerful confidence that a book would one day emerge has heartened me at difficult stages. Christopher Collard and Pat Easterling advised me in the formative stages and have always been ready with fresh advice. Peter Parsons has offered similar encouragement, suggesting possible lines of thought with typical modesty; and although Oliver Taplin has moved more in the realm of the reception and performance of drama, I hope he will find some things to give him pleasure in this volume. I have also benefited from conversations with Michael Silk and still more from his published work. For reading one or more chapters I am happy to thank Felix Budelmann, Anna Clark, Pat Easterling, Stephen Halliwell, Gregory Hutchinson, Christopher Pelling and Laura Swift. I am especially indebted to Patrick Finglass, who volunteered to read the entire book in draft and covered it with acute observations and helpful corrections. His forthright

Preface

xv

deletions have spared the reader much verbiage. He also generously sent me copies of a number of items old and new which had escaped my notice, as well as allowing me to consult substantial parts of his commentary on Ajax well before publication. I have subsequently been greatly assisted by the comments of the anonymous readers of Cambridge University Press (later identified as Donald Mastronarde and Michael Silk). At all stages Michael Sharp, Jo Lane and Christina Sarigiannidou at the Press have given support and assistance. Special thanks are due to Muriel Hall, whose acute and meticulous copy-editing has immensely improved the final product. Peter Parsons has given invaluable help with the proofs. Other debts are less specific and more personal. Catherine Whistler has dispelled my despondency on many occasions, as well as reminding me at the right moments that life is not all about research and publications. The volume is dedicated to two of my oldest friends, who have shared both life and scholarship since my first year (indeed, in Nicholas’s case my first week) as an undergraduate in Oxford, and whose friendship has given added zest and value to the study of tragedy along with much else: ‘For there is no man that imparteth his joys to his friend, but he joyeth the more; and no man that imparteth his griefs to his friend, but he grieveth the less.’ R. B. R. 

Bacon, Essays xxvii, ‘On Friendship’.

Abbreviations

Names of ancient authors and titles of their works are normally abbreviated as in OCD, except that A. = Aeschylus, S. = Sophocles, E. = Euripides (and note that ‘E. Her.’ refers to Euripides, Heracles, but ‘E. Heracl.’ to his Heraclidae). Periodicals are either as in OCD or given in full. Other abbreviated titles are as follows. Abrams, Glossary Bauformen Bruhn CAH Campbell, ‘Essay’ CHCL i Csapo-Slater Denniston DFA

M. H. Abrams, A Glossary of Literary Terms (th edn, New York ) W. Jens (ed.) Bauformen der griechischen Trag¨odie (Munich ) Anhang to Sophokles, ed. F. W. Schneidewin and A. Nauck, vol. viii (Berlin ) Cambridge Ancient History (Cambridge, nd edn –) L. Campbell, ‘Introductory essay on the language of Sophocles’, in Sophocles, the Plays and Fragments (Oxford –), vol. i, – The Cambridge History of Classical Literature i: Greek Literature, ed. P. E. Easterling and B. M. W. Knox (Cambridge ) E. Csapo and W. Slater, The Context of Ancient Drama (Ann Arbor ) J. D. Denniston, The Greek Particles (nd edn, Oxford ) A. W. Pickard-Cambridge, The Dramatic Festivals of Athens, nd edn revised by J. Gould and D. M. Lewis (Oxford , reissued with supplement ) xvi

List of abbreviations FGrH GGL GLP Guthrie HGP IG i Kranz LIMC LSAM LSCG LSJ

ML Musa Tragica OCD PCG PMG PMGF Preminger-Brogan SFP

xvii

F. Jacoby, Fragmente der griechischen Historiker (Berlin and Leipzig, –) W. Schmid and O. Stahlin, Geschichte der griechische Literatur i.– (Munich, –) D. L. Page, Greek Literary Papyri (Loeb series, Cambridge Mass. ) W. K. C. Guthrie, History of Greek Philosophy,  vols. (Cambridge –) Inscriptiones Atticae Euclidis anno anteriores, ed. D. M. Lewis et al.,  vols. (Berlin and New York –) W. Kranz, Stasimon (Berlin ) Lexicon Iconographicum Mythologiae Classicae,  vols. (Zurich and Munich, –) F. Sokolowski, Lois sacr´ees de L’Asie Mineure (Paris ) F. Sokolowski, Lois sacr´ees des cit´es grecques (Paris ) H. G. Liddell, R. Scott, H. S. Jones, A Greek English Lexicon (th edn, Oxford ), with revised supplement ed. P. G. W. Glare (Oxford ) R. Meiggs and D. Lewis, A Selection of Greek Historical Inscriptions (Oxford, nd edn ) Musa Tragica: Die griechische Trag¨odie von Thespis bis Ezechiel Unter Mitwirkung von R. Kannicht . . . ed. B. Gauly et al. (G¨ottingen ) S. Hornblower and A. J. Spawforth (eds.) The Oxford Classical Dictionary (rd edn, Oxford ) R. Kassel and C. Austin, Poetae Comici Graeci (Berlin –) D. L. Page, Poetae Melici Graeci (Oxford ) M. Davies, Poetarum Melicorum Graecorum Fragmenta,  vol. to date (Oxford –) A. Preminger and T. V. F. Brogan (eds.) New Princeton Encyclopaedia of Poetry and Poetics (Princeton ) C. Collard, M. J. Cropp and K. H. Lee, Euripides. Selected Fragmentary Plays i (Warminster ) and

xviii

Taplin TGFS TrGF Wales West, GM West, AGM

List of abbreviations C. Collard, M. J. Cropp and J. Gibert, Euripides. Selected Fragmentary Plays ii (Warminster ) O. Taplin, The Stagecraft of Aeschylus (Oxford ) Tragicorum Graecorum Fragmenta Selecta, ed. J. Diggle (Oxford Classical Texts, ) B. Snell, S. Radt, R. Kannicht, Tragicorum Graecorum Fragmenta i- v (G¨ottingen –) K. Wales, A Dictionary of Stylistics (London and NY ) M. L. West, Greek Metre (Oxford ) M. L.West, Ancient Greek Music (Oxford )

Note on texts

Citations from the tragedies are normally taken from the Oxford Classical Text. Aeschylus is quoted from Page (), Sophocles from Lloyd-Jones and Wilson (), and Euripides from Diggle (–). Hence I follow not only their text (and colometry) but their attributions in disputed passages unless explicitly stated. But I have of course constantly consulted the other standard editions, especially West for Aeschylus, and important variations are noted when this is significant for the argument. For Aristophanes I use Dover’s text of Clouds and Frogs, Austin and Olson’s of Thesmo., and Wilson’s OCT for the rest. Citations of tragic fragments are always from TrGF, of comedians from PCG. Fragments of authors other than the dramatists are cited from the editions generally considered standard: e.g. Merkelbach-West for Hesiod, Snell-Maehler for Pindar, Jacoby for historians. Where there might be uncertainty, I generally specify.

xix

c ha p te r 1

Introduction

i starting points: three passages The object of this book is to encourage readers to relate the particularities of specific passages to larger and general questions. We may then suitably begin with a few short quotations, one from each of the great tragedians, each stylistically remarkable in a different way and opening up different lines of enquiry. de± to© nin, Þv ›qreyen ›kpaglon t”rav, qane±n bia©wvá –kdrakontwqeªv d' –gÜ kte©nw nin, Þv toÎneiron –nn”pei t»de. (A. Cho. –)

Since she nurtured a terrible apparition, she must perish violently; and I, transformed into a snake, am her slayer, as this dream declares.

Orestes has just been told by the chorus-leader that Clytemnestra sent placatory offerings to Agamemnon’s tomb because she was frightened by an ominous dream. In that dream she gave birth to a snake which she wrapped in swaddling-clothes and which drew blood from her breast when she tried to feed it. As Orestes deduces, the dream prefigures his return and the bloodshed of the matricide. The immediately striking feature of these lines is the invented compound –kdrakontwqe©v, ‘transformed into a snake’. For obvious reasons this word does not recur anywhere else in Greek literature. But the language here is remarkable not just for the verbal coinage but because Orestes identifies himself with the snake of the dream. The matricidal act requires him to shed something of his humanity. Moreover, the use of this term associates Orestes’ intended act with a network of snake-references elsewhere in the trilogy. Clytemnestra was pictured in these terms earlier in the play (), when Orestes described Agamemnon’s offspring as ‘the brood bereft of their eagle father, killed in the twisted coils of a dreadful viper’, and he reuses this language when he 



Introduction

stands over her corpse (). More positive is the viewpoint of the chorus, who see Orestes as a Perseus-like figure who has lopped off the heads of twin snakes (Aegisthus and Clytemnestra) (). But immediately after that, the Erinyes become visible to Orestes, as monstrous forms with snakes entangled in their hair (–, cf. Eum. ). The imagery in all these cases is darkly sinister, but only in  does Orestes apply it willingly to himself. By accepting his role as his mother’s slayer he also takes a step towards sharing something of her monstrous character. Communication between god and man in tragedy is not restricted to dreams and omens, as a second extract, from Sophocles, will show. kale± g‡r aÉt¼n poll‡ pollaci qe»vá “å oÕtov oÕtov, O«d©pouv, t© m”llomen cwre±n; p†lai dŸ tˆp¼ soÓ bradÅnetai.” (S. OC –)

For a god summoned him many times and from many sides: ‘Hey you there, you, Oedipus! Why do we delay our journey? Long now is it held up thanks to you.’

In what is anything but a conventional messenger speech, the chorus and the audience are told of the mysterious end of Oedipus, brought about by supernatural forces which enter audibly into the human sphere. An unearthly summons is heard, loud and echoing (as poll‡ pollaci probably implies). What is important here is to try to judge the tone. The call oÕtov oÕtov seems peremptory, the repetition indicates impatience. But the words which follow moderate this with a gentler note: because the speaker uses the ‘sympathetic’ first person plural, a sense of comradeship or concern is introduced. The divinity is unidentified and unseen, yet establishes an immediate and specific contact with Oedipus, who is addressed by name. The poet creates a sense of the gulf between man and god but simultaneously creates a bond between them in this special case (‘we’): although the messenger did not see what follows and the audience are dependent on his report, a veil seems momentarily to have been lifted, though swiftly lowered again for those who survive Oedipus’ passing. The third passage, from Euripides’ Bacchae, also involves interaction between man and god, but here face to face, on stage; however, the god is   



See further Garvie’s commentary, pp. xxxvi–vii; Lebeck , ; Petrounias , ff. For an interesting parallel with the Bacchae see Dodds on lines – of that play. The locution is common from master to slave, and often expresses impatience. It is frequent in comedy; Aristophanes has over  examples, e.g. Ar. Ach. , , Vesp. , ; see Dickey , – and index; Stevens , –. Only here in tragedy is the term duplicated. Cf. Macleod on Iliad ., citing Wackernagel –, i. – (= Eng. tr. –).

I Starting points: three passages



in disguise and unrecognised by his mortal captor. What matters in this scene is that the apparent power-relations, between monarch and foreign captive, are in fact illusory: Dionysus is only feigning weakness, and already in this first encounter his words tantalise and provoke Pentheus while also hinting at his true strength. Pe. l†zusqeá katafrone± me kaª Qžbav Âde. Di. aÉdä me mŸ de±n, swfronän oÉ sÛfrosin. Pe. –gÜ d• de±n ge, kuriÛterov s”qen. Di. oÉk o²sq' Âti zv oÉd' Á driv oÉd' Âstiv e². Pe. PenqeÅv, %gauv pa±v, patr¼v d' ìEc©onov. Di. –ndustucsai toÎnom' –pitždeiov e². (E. Ba. –)

p e n t h e u s : (to guards) Get hold of him. This man treats me and Thebes with contempt. d i o n y s u s : I urge you not to bind me, speaking as a man of sense to those who lack it. P: Well, I say bind you, and I have more power than you have. D: You do not know what life you live, nor what you do, nor who you are. P: I am Pentheus, son of Agave, and my father was Echion. D: You have a name well suited to misfortune.

The momentous triple negative of Dionysus’ warning () makes clear that this is no ordinary rebuke, no mere claim that Pentheus is behaving foolishly. Rather, the god implies that his whole life is adrift from reality, that there is no respect in which he is acting correctly. Pentheus’ response answers only the third of the clauses, ‘who you are’, but on a wholly superficial level; he ignores the deeper implications of the phrase, as in the eloquent Delphic injunction ‘know thyself ’. The god’s stern challenge is met with self-assertive pretension. But even this inadequate response by Pentheus carries a seed of disaster, in the ominous utterance of his name, the potential negative force of which Dionysus immediately takes up (cf. already Tiresias at ). Penthos means sorrow, an obvious nom parlant of a type common in tragedy. Nomen omen: a name is not only a name but a sign, here one which forebodes disaster for its owner. Throughout the scene, ambiguity and double meanings give force and definition to a   

The text of  has been doubted, and Âti zv is obelised by Diggle, but I am satisfied with the defence by Seaford; cf. Fraenkel ap. Dodds: ‘you do not know what kind of life you lead’. Pl. Prt. b; ascribed to various of the seven sages by later authors. On the history of the concept see Rutherford ,  and n. . On these lines see also Silk a, – (bibl. in n. ), in the context of a valuable discussion of naming.



Introduction

clash between two different perspectives on the situation of Pentheus: the one partial, overconfident, blinkered and alarmingly naive; the other clear sighted, forward-planning, rightly confident in his power. In each of these three passages a linguistic or stylistic phenomenon has been singled out – in the first a bizarre verbal coinage, in the second the use of a particular form of address and of the first person plural present indicative rather than a second person imperative, and in the third the etymology of a character’s name. In each case the observation has inescapably led us on to larger questions affecting the interpretation of the passage in the context of the whole drama. This invites the question whether enquiries into language and style can practically or properly be separated from the criticism of a tragic drama as a whole in all its aspects. In order to clarify this issue we need first to consider how the terms ‘language’ and ‘style’ should be understood in this context, and also what contribution to their study has been made by previous scholarship, both ancient and modern. ii problems of definition (‘style’ and other terms) Since the Athenian tragic dramas of the fifth century bc were composed in Greek, they form part of the evidence studied by historians of the Greek language as a whole. But tragedy is a genre composed in verse, both spoken and sung, in an elevated poetic style remote from both formal prose of the period (represented chiefly by oratory and historiographic writings) and from the epigraphic language of decrees and treaties; still more remote, we can assume, from everyday speech. Tragedy owes much, in its vocabulary and imagery as in its cast of characters, to the epic poetry of earlier centuries, especially that of Homer. As with Homer, we may legitimately speak of a tragic Kunstsprache, an artistic or artificial language. The parodies and frivolous treatment of tragic situations and tragic diction in Aristophanic comedy make clear that the rival genre could be regarded as pompous, bombastic and pretentious. Greek tragedy, however, cannot be viewed as a homogeneous entity: several important distinctions need to be made at once. First, there is the chronological span. Our thirty-two surviving tragedies cover at least the period from  bc (Aeschylus’ Persians) to  (Sophocles’ Oedipus at Colonus, posthumously produced). If, as is likely, the Rhesus is misattributed to Euripides, it may well date from the first half of the fourth century. Second, we are dealing with the works of three very different poets (or five, if both the Prometheus Bound and the Rhesus are by unknown authors), who

II Problems of definition



enjoyed long lives and productive careers; it is natural to assume that they developed in technique and in their stylistic practice during their lifetimes. Indeed, Sophocles is said to have observed such a development in his own style, away from the Aeschylean manner. Third, even within the individual drama there are distinctions between the constituent parts, since tragedy is a hybrid genre: the linguistic texture varies in those different parts, and the variation is reflected also in the use of different metres. The contrasts are seen most plainly between the spoken and sung sections (less significant are the more infrequent passages of recitative or ‘chanted’ verse); but even within the spoken sections of the drama there is variation in form and manner. We note especially the difference between rhesis, stichomythia and irregular sequences of dialogue. Form here affects verbal expression: in rhesis the character’s speech is more expansive, in stichomythia more clipped and epigrammatic. Metrical variation is found even in the spoken sections: in some plays the regular iambic trimeters give place to trochaic tetrameters, normally indicating a more agitated mood and (we can often deduce) more rapid activity or animation on stage. It seems that we must speak not of the style of tragedy but of a plurality of styles. But what is style, and how does it relate to language, or to other indispensable terms such as rhetoric or discourse? Perhaps the most obvious starting-point is the principle which opens Dover’s study of the development of Greek prose: ‘It is generally, though not universally, agreed that if I am asked about the style of your performance of a certain act, I am being asked not what you did but how you did it.’ This draws attention to the fact that ‘style’ and cognates can be used with reference to many activities that do not involve language, or do so only incidentally – playing golf or tennis, playing the piano, cooking, eating or flower-arranging, to name a few. In the sphere of literature, it was normal in antiquity and later to oppose style and content, the medium and the message. Often 



 

Plut. De prof. virt., Mor. b = S. T: ãsper g‡r ¾ Sofoklv ›lege t¼n A«scÅlou diapepaicÜv Àgkon, e²ta t¼ pikr¼n kaª kat†tecnon tv aËtoÓ kataskeuv, tr©ton ¢dh t¼ tv l”xewv metab†llein e²dov, Âper  qikÛtat»n –sti kaª b”ltiston. Translation is controversial, but a possible rendering is ‘just as Sophocles said that after toying with Aeschylean pomposity, and next the sharpness and refinement of his own composition, he then thirdly shifted to the form of expression which is most characterful and best’. See further Bowra , Pinnoy , Pelling , with much earlier bibliography. Pelling argues that this passage tells us more about Plutarch than about Sophocles. See Murry , Wellek and Warren , ch. , Hough , Wellek , –, Turner , Fowler , Bradford . For general articles see Abrams, Glossary –, –, Preminger– Brogan s.v. Style, Stylistics. Dover , ; but note the important qualifications which follow. Dover , –, with numerous examples.



Introduction

the metaphor of ornamentation is used: style provides the cosmetics, the embellishments, the sugar on the cake. The notion that ‘language is the dress of thought’ is commonplace in ancient theory. It has a superficial plausibility when one thinks of stylistic ornamentation in terms of similes, expansion of the thought through comparisons, or other forms of amplification through examples, digressions and various figures of speech. On this model there is a core of ‘essential’ argument or narrative which cannot be dispensed with, whereas the ornament could be dropped without changing what is communicated. The weakness of this position is obvious: to dismiss any part of a text as no more than ornamental is dangerous, and especially so in poetry. Would Cicero’s speeches be undiminished if the historical exempla were removed, or Homer’s poems if deprived of the similes? In fact, the argument becomes impossible to sustain once one moves away from works with a clear narrative content. Here is a well-known poem by Tennyson. He clasps the crag with crooked hands; Close to the sun in lonely lands, Ringed with the azure world, he stands. The wrinkled sea beneath him crawls; He watches from his mountain walls, And, like a thunderbolt he falls.

How does one distinguish style and content in such an example? Whereas with the Odyssey, for example, it is at least feasible to give a paraphrase, however jejune, of the narrative content, here one is reduced to the banalities of a child’s early reading exercises: ‘See the bird sit. See the bird fly.’ Hence most studies since the nineteenth century have resisted the concept of style as ‘ornamentation’ and prefer a more organic model: style and content are interdependent, and the presence of similes, unusual words, bold or exotic turns of phrase, and so on have a part to play in the overall effect of the work and hence contribute to its meaning. The discussion is complicated further by another common opposition, between form and content. ‘Form’ does not seem to be quite the same as style, though they are evidently related. Form must clearly include metrical structure. Tennyson’s poem will again illustrate the point. It contains two short three-line stanzas marked as separate not only by the spacing but by the different rhymes in each: the first stanza describes the eagle in repose,  

For this and other metaphorical expressions employed in ancient criticism to describe style in ancient criticism see Van Hook ; see also Wiseman , –. Tennyson, ‘The Eagle’, Fragment, published : see Ricks , .

II Problems of definition



the second prepares for and then powerfully foregrounds his lightning-swift descent towards his prey. ‘He stands’ and ‘he falls’ are carefully placed in the terminal position in each stanza. Self-conscious use of metrical form is, as we have already seen, an important aspect of tragedy. But we are not concerned only with metrical form. The term ‘form’ is indeed employed in a number of different ways: to refer to the structural units of the play (ode or episode), or to the typical elements which recur from play to play (such as the agon), or more broadly to the overall construction of the dramatic plot in a sequence of scenes. In all these senses each of the dramatists has a highly-developed sense of form and plans the tragedy as a whole entity; scenes do not exist in isolation (and in a trilogy, significant formal parallels may exist between plays). Form in a sense is logically prior to language: when Euripides decides that the next scene will include a monody, he must choose appropriate metrical form for it, and will heighten the linguistic expression accordingly. Elaboration of imagery and looser, more associative sequences of thought characterise such passages of solo singing: emotionally fraught language is required to convey the passion of a Creusa or an Evadne. All of this goes far beyond merely choosing words which will fit into the lyric metres. We may legitimately treat form as an aspect of style which is strictly speaking independent of language, although in practice inseparable from it. Style, then, is a concept which overlaps with but cannot be identical with language. It is also possible to speak of musical and choreographical styles, and we know that music and dance were prominent in the performance of Greek drama and must have played a crucial part in the total theatrical experience; unfortunately we know little of the former and virtually nothing of the latter. Language, by contrast, is not dissociable from the verbal text. Vocabulary, word order, syntax, and the sound and shape of words (phonology and morphology) are among the central concerns of the linguist. They also form part of the province of the literary critic concerned with style. These border regions have sometimes been the subject of territorial wars, and much ink has been spilt over whether stylistics is a part of linguistics and what the relation of each is to poetics. In fact it is difficult to distinguish areas of language which will never under any circumstances be of interest to the linguist or to the critic, but the linguist and the critic do not necessarily have the same interest in every aspect. To adopt the terminology popularised by 

A famous trumpet-blast in these disputes was Jakobson’s essay ‘Linguistics and poetics’ (), repr. in Lodge , –. See further the essays in Freeman , Chatman , Carter .



Introduction

Saussure, the linguist in his capacity as a historian of language is perhaps more concerned with the langue, the overall map or rule-book of the language; whereas the literary critic is chiefly concerned with the parole or particular utterance, whether a specific line or a specific work. But the critic ignores the larger map of the language as represented inside and outside literature at his peril, for the individual usage of a particular text must be seen against the background of usage in other literary texts and also (where the knowledge is recoverable) in the context of non-literary usage. Here the concept of genre is of key importance. Even in the modern world, it is recognised that poetry and prose are different worlds: more specifically, most readers will recognise the metrical conventions of a limerick or a sonnet. In the ancient world, genre was a literary principle of overwhelming importance. One might indeed redeploy the Saussurean antithesis mentioned in the last paragraph: in a sense the tragic genre is the langue and the specific play the parole. In other words, anyone who reads or watches two Greek tragedies in succession will immediately see resemblances between them, not necessarily in the myth treated or the specific plot-pattern, but in the formal conventions governing verbal expression and overall structure. These conventions are partly the consequence of the origins and development of the genre, but they are retained because they continue to satisfy in aesthetic and functional terms. They are plainly distinct from the conventions which shape earlier genres, even those which strongly influence tragedy (epic, choral lyric), and also from the closely related satyric drama and the strongly contrasting genre of fifth-century comedy. As already indicated, within the overarching framework of the genre conventions also shape the particular formal units (stichomythia, choral ode, messenger speech and so forth). Innovation, modification, even subversion of these conventions are all possible, and it is hard to define the limits beyond which the tragedian could not step; but there is always a consciousness, shared between dramatist and audience, of the tradition and its expectations, and it is this consciousness which the modern reader needs to recreate in order to understand and appreciate the ancient forms. In particular tragedy cultivates an elevation of tone which is sustained through language distanced from the everyday. According to ‘Aeschylus’ in Aristophanes’ Frogs, ‘by a law of their nature sublime ideas and greatness 

Cf. Fraenkel’s celebrated dictum, ‘For Greek tragedy there exists also something like a grammar of dramatic technique’ (Agamemnon, vol. ii, , on lines f.); cf. Taplin : ‘While I should not like to press the “grammar” analogy too hard – the dramatists are constantly creating their usage, and a departure from the norm is not thereby “ungrammatical” . . . ’)

II Problems of definition



of thought are begetters of lofty expression; and again, demi-gods as of right should excel mere mortals in grandeur of phrasing, since greater magnificence is also the outward mark of their clothing’ (–). Just as royalty and deity wear grander clothes, so they should speak in a nobler vein. The principle is taken up in serious literary criticism, often in terms of t¼ pr”pon or decorum, ‘what is appropriate or fitting’. But although in the Frogs Aeschylus may win the contest, we must not neglect the other side of the debate: ‘Euripides’ rightly claims that he has brought the genre closer to the audience’s own concerns and extended their understanding, by admitting less exalted language and more everyday subject matter (–, –). Partly this is in order to make the tragedies more accessible, more comprehensible (Frogs –), but clarity taken to extremes can result in banality, as happened according to Aristotle to certain minor tragedians such as Cleophon and Sthenelus (Poet. .a–). The presentation of this debate in the Frogs brings out one of the most important points in this whole discussion: that stylistic discussion can only to a very limited degree be kept separate from content. If we are examining how the style of a passage is suited to the speaker or the situation or the larger context, or indeed if we wish to argue that the style is actually unsuitable, or that it sinks to a level below what is customary in tragedy, we are inevitably involved in consideration of content. Sharp or even gradual changes of style (or, as some prefer, of ‘register’) must be correlated with the nature of the events on stage. Otherwise the critic cannot possibly have an adequate basis on which to found discussion of characterisation through style, changes of pace and tone, creation of suspense or other moods, rhetorical interaction, and countless other features of the text. A definition of style, then, remains elusive. One possible formulation is that while language is what makes communication possible, style embraces all the techniques and devices which render the communicative act memorable and significant. This may however be thought to give too much emphasis to unusual or heightened style: neutral or plain words can also be noteworthy, particularly when set in contrast with something more unusual or exotic. We may draw the analogy with colour. Unless we are totally colour-blind, we all see the world as richly coloured, and it is hard to imagine any object, animate or inanimate, as wholly colourless (granting that black and white are special kinds of colour); many objects are  

E.g. Longinus, Subl. .: cf. Russell’s notes on Quint. Inst. ..–. Cf. Wales s.v., on this as on many other key terms; Dover ,  and –. Changes of register in tragedy normally involve some reduction of dignity or elevation, either through the introduction of colloquialism or some other incongruity (offensive language or ‘low’ comparison).



Introduction

variegated or have different colours in different parts. Similarly a text may have a very plain or monotonous style, but can hardly be style-free. It is the quality and variety of the style which make the work attractive or enjoyable to contemplate. We must next consider at what level the critic is to study style. Some of the ambiguities involved in the term arise from confusing analyses at different levels. On one level, perhaps the lowest, there is the analysis of a particular word, such as one of the numerous polysyllabic compounds coined by Aeschylus. Then there is the line or sentence within which that word occurs. Beyond that we may look at the whole speech in which the line figures, and then the scene in which the speech plays a part. At each level it is possible to ask questions about the stylistic qualities of the unit in question, though the analysis becomes more complex, the number of particular stylistic effects more numerous, and the task of producing useful generalisations more difficult at each stage. Then there is the play as a whole (as one may speak of the style of the Bacchae as being in general more archaic than that of the Orestes), and beyond that the oeuvre of the poet, and beyond that the collective oeuvre that constitutes the corpus of Greek tragedy. Lateral approaches are also possible: one can examine all the compound adjectives or all the prologues of a single tragedian, or of all three. All of these exercises are legitimate and interrelated. The essential point is that any tragedy or any part of a tragedy exists in a context: words and speeches cannot be kept in isolation, although for purposes of exposition or teaching they may be artificially wrenched from context. Even fragments such as those preserved in quotations and anthologies must originally have had a context. Once this sequence of concentric circles has begun, where can it end? Thus far the concern of the critic will be with style within the work or the genre, but it is also possible to range more widely. We have already mentioned the importance of comedy as a contemporary dramatic genre which follows different rules, stylistically as in other respects. In some ways, indeed, comedy defines itself by opposition to tragedy. Inter-generic influences form an important area of study: we can see that the sophist Gorgias in his prose emulated and even exaggerated poetic devices, and his mannerisms were transferred back into tragic verse by Euripides, Agathon and others. More difficult and speculative are enquiries into the style of a 

Taplin  and , Sommerstein ; see also below, pp. –, –. On the third dramatic genre, satyr-play, closely associated with tragedy but less ambitious and lighter in mood, see Seidensticker , Krumeich, Pechstein and Seidensticker , Harrison ; on the diction of the genre, Griffith .

II Problems of definition



period or a community. We can grasp what is meant by the suggestion that the Pre-Raphaelites or the English Augustan poets share a common style, but how meaningful is it to speak of the style of the Victorian period, far less the Periclean? A genius such as Auerbach could use short and shrewdlychosen extracts from classic texts as an index of the mentalit´e of past ages, but more specific and tightly-focused questions may provide more secure conclusions. It may be felt that an approach that is chiefly preoccupied with the verbal text (‘the words on the page’, the clarion call of the New Criticism) risks losing touch with the facts of performance, the drama of the play in the theatre. That is not a necessary consequence. Staging and dramatic effect remain essential to our understanding of tragic drama: awareness of the verbal and the visual need to be combined. Indeed, the text itself forces us to recreate, at least in the mind’s eye, some of the crucial movements, from supplication to physical assault. Aggressive action is often combined with vigorous and abusive language. Dress is normally suited to status and rank, so that occasions on which there is a mismatch are important (Telephus disguised as a beggar, Electra degraded and in rags). Equally important is the use of props on stage, which seems to have been relatively rare: props which do appear tend to be significant. The tapestries in the Agamemnon, the sword of Ajax, the urn that supposedly contains Orestes’ ashes, the bow of Philoctetes, are introduced on stage in order to be the subject of reflection or lamentation or heated dispute. The prominence of deictic terms (‘this’ and so forth) in such contexts serves to direct our urgent attention at what matters on stage. The study of deictics in a broader sense is currently a growth area in classical research. We have already referred more than once to the importance of interaction between characters. Even when one character is engaged in a lengthy     

 

See e.g. the well-known books by Houghton  and Newsome  on ‘the Victorian Frame of Mind’ and ‘the Victorian World-Picture.’ Auerbach . DFA –, Shisler , Spitzbarth , Kaimio , D. Cairns b. On the cultural history of gesture see the essays in Bremmer and Roodenburg . Relevant here is the discussion by Worman  (cf. my review in CR  () –). Wyles , on costume, appeared too late for me to take account of it. More generally see Taplin – and the more detailed treatment of S. Phil. in Taplin ; also Segal . Iphigenia’s letter to her brother is another case in point (in E. IT: see p. ); cf. the letter of Phaedra in E. Hipp., or of Agamemnon in IA. Deictics can, however, refer not to things visible on stage, but to someone or something vividly present in the speaker’s thoughts: see Finglass on S. El. . Cf. ‘The poetics of deixis in Alcman, Pindar, and other lyric’, Arethusa special issue . ().



Introduction

rhesis, it is necessary to consider the effect on the others present on stage. Here much can still be done within the categories laid down by ancient rhetorical theory. ‘Rhetoric’, an ambiguous term, is often used to describe the affective impact of any utterance on any audience: in this sense it would be possible to speak of the rhetoric of a Greek tragedy as a whole. More usefully specific is the definition of the term as the art of persuasion. Numerous scenes in tragedy dramatise the attempts of one character to persuade or dissuade another, and the techniques used are frequently paralleled in forensic oratory of the late fifth/early fourth century. Equally, the gap between the courtroom and the heroic world of violence and revenge is often highlighted, and failures to persuade can be as important as successes, as well as producing more intense scenes of pathos. Here the critic must look not only to the style but to the argumentation: weak or specious arguments abound, and not only in the mouths of ignoble figures. Sometimes style is an indicator of weakness in the argument: the nurse in Hippolytus is a suggestive case, because her most eloquent speech is inconsistent both in its lines of argument and in the stylistic register (her status as a slave seems here to explain her inability not only to understand Phaedra but to sustain her eloquence convincingly). Tiresias in his longest speech in the Bacchae (–) presents a bewildering tangle of contradictory views, reminiscent of a not-very-competent sophistic lecturer: this case is the more remarkable given the contrast with Tiresias’ impressive status in tragedy in general. Yet even here, Tiresias’ arguments may be startlingly defective but his conclusion is correct: the god must be worshipped. The concerns of rhetoric are taken further in modern work on sociolinguistics. The aim of this discipline is to establish more precisely the relationships between speakers, with particular reference to such determinants as their age, status, gender, and race. Since tragedy regularly highlights differences of this kind and allows conflict to emerge between young and old, father and son, husband and wife or concubine, and so forth, there is no shortage of material. Valuable work on forms of address, formulae of 





Note esp. the speech at Hipp. –, with Pelling , –. The fluctuation of style is especially clear at –: note the colloquial interjection at  (‘she takes him and you can’t think (päv doke±v) how badly she treats him’), followed by the shift to a loftier style exalting the power of the goddess in the next few lines (for the expression cf. Hec. , Ar. Ach. , , Stevens , ). Cf. the poorly-structured speech of ‘Lysias’ in Plato’s Phaedrus, and the later discussion of its weaknesses, which includes some reference to the dramatists’ concern with structuring their plays (c–e). See Coupland and Javorski  for a conspectus of these studies.

II Problems of definition



greeting, and so forth has already appeared, and more may be expected, especially in the field of ‘politeness’ studies. Investigations of this kind inevitably move beyond the raw material provided by one literary genre and indeed beyond the sphere of literature in general. Sociolinguistics in the modern era is at least as much concerned with real-life relationships, which provide the evidence that supports commentary on the secondary world of literature. The same relationship should theoretically obtain in classical studies, but the difficulty at once arises that most of what we have from the fifth century and many other periods is the material provided by literature. In effect scholars must endeavour to reconstruct the living language and its idioms and conventions from literary texts: thus comedy provides at least as much evidence for colloquial usage as graffiti or other ‘non-literary’ texts. Nevertheless, a recurring concern in many recent studies of tragedy is to reconstruct the historical context of the genre, and this includes the relationship of tragic language and diction to other prominent linguistic fields. In terms used by some modern linguists, stylistic analysis can be either textualist or contextualist, or, to rephrase, intrinsic or extrinsic. On the one hand the student of language may focus on the usage of tragedy; on the other, he may seek to demonstrate the interaction with other genres in both poetry and prose, and with non-literary forms of discourse such as a diplomatic treaty. Under the circumstances, however, there is greater scope for drawing connections with other literary texts than with the spoken language of the fifth century. One area where this kind of discourse analysis pays obvious dividends is in the language of political debate as conducted in the Athenian assembly. A variety of passages could be cited, but the most conspicuous is the scene between the Athenian king Theseus and the Theban herald in Euripides’ 

  



On forms of address see esp. Dickey ; on politeness theory see Brown and Levinson , Lloyd  and , and Brown  (on the Iliad ). Boas  discusses other useful approaches applying pragmatics to tragic dialogue, including relevance theory and discourse cohesion. Dover , and related studies. Cf. Goldhill a, discussing literary language (esp. the inheritance from Homer), ritual and religious language, the language of the courts and assembly, and formal rhetoric. See for this example A. Supp. – (cf. ML .–); E. Supp. – with Collard’s notes giving parallels from Staatsvertr¨age .– and elsewhere. Note Collard’s comment on p. : ‘it reflects the language of defensive alliances closely enough to give verisimilitude to the Exodos, loosely enough to accommodate the whole within Tragic idiom.’ Compare also S. OT – (formal cursing) with ML , esp. text A (Teos), .– (Draco on homicide); Parker , –. Another, not unrelated, may be summed up as laudes Athenarum. See further ch. , n. . These passages are especially notable when they have no justification in the plot, as when the Trojan women speculate about their future destination, and express a preference to be sent as prisoners to Athens rather than to Sparta (E. Tro. –).



Introduction

Suppliants (–). Although political disputes are frequent in tragedy, this passage is remarkable for the sustained prominence with which the catch-phrases and ideological terminology of the fifth century are brought into the mythical world of the heroic age. Within this passage we find references to annual succession of magistrates, equality between poor and rich, the value of a written constitution for the protection of the weaker citizen; all of this is opposed to the arbitrary and destructive domination by a tyrannical ruler. The vocabulary is redolent of democratic rhetoric: –leuq”ra p»liv (‘a free city’), ­son (‘equality’), n»moi koino© (‘laws in common’), d©khn ­shn (‘equal justice’), –v m”son (‘into the public arena’), dmov eÉquntžv (‘the people as regulator’). Even the characteristic proclamation of the democracy, t©v ˆgoreÅein boÅletai;, ‘Who wishes to speak?’, which we find quoted in historical contexts in the orators, is closely paraphrased by Theseus, who describes it as ‘that word of freedom’ (). On the opposing side the herald more briefly denounces the dangers of the system: unscrupulous demagogues with their deceptive rhetoric, self-interested slander and embezzlement. Even if we grant that Theseus, Athens’ favourite hero, is no ordinary king, the peculiarity of this debate, in which he champions the democratic principle against its opponents, is marked. Here at least the discourse of politics is not wholly assimilated to the tragic context. Even if we did not know that debates about constitutional theory were characteristic of this period, and played a part in the disputations of the sophists, it would be easy to guess this from the Supplices debate. It would seem that Euripides thought it more important to make Theseus a spokesman for democratic ideals, and to produce a powerful ideological clash that would give added spice to the conflict, than to avoid anachronism. Similarly in his Orestes a messenger describes a demagogue’s participation in an assemblyscene in terms which remind the reader of Cleon in Thucydides more than of the assemblies in Homer or earlier tragedy. Tragedy is a genre rich in variety, thanks to the diversity of formal structure and metre, the shifts between speech, recitative and song (united or communal song when the chorus is involved, but also individual song from particular actors). More will be said of this in the next chapter. Still more important, drama involves dialogue, thus inevitably introducing a   

For an account of the sources for our knowledge of Athenian democratic institutions and ideology, see Hansen , ch. . On the whole speech see Collard’s commentary. On anachronism in tragedy see esp. Easterling ; on the ideology of Athenian democracy see Jones , –; Hansen , –. Or. –, esp. –. See p. . Interpolation has been suspected in this passage, but if the suspicion is justified, this diminishes rather than eradicates the effect described.

II Problems of definition



diversity of viewpoints; moreover, there is no authorial presence as narrator, no master voice to guide our evaluation of the characters. This multiplicity of viewpoints has been characterised in recent criticism as polyphony, developing the influential contrast of monologic and dialogic or polyphonic forms put forward by Mikhail Bakhtin. That tragedy includes diversity of perspective is unquestionable. Doubtless some views are introduced only to be dismissed as unacceptable (for example, the tyrannical pronouncements of Lycus in Heracles); but more usually there is a conflict in which the right side may be hard to determine (archetypally so in such clashes as Creon versus Antigone). Dramatic dialogue certainly requires contrasting voices; whether these voices are also stylistically differentiated is one of the most delicate questions in the study of tragedy, but we shall see good reason to conclude that characters are given individual stylistic features in at least some cases. In any case, if we adopt sociolinguistic methods, it is not only features such as word-choice or metaphors or even recurring preoccupations or obsessions which define the characters and constitute the interpersonal relationship: we shall also be concerned with power-relations and styles of argument. The plots of tragedy, concerned with eminent figures facing catastrophic reversals of fortune, make it natural that the gap of power between characters will often be huge: king and slave, conqueror and concubine, god and man. Gender, class, race and age-differences may all play a part. Even between nominal equals, a weaker personality can be convincingly shown as submissive to but hesitantly resisting a dominant partner (Ismene versus Antigone). In the dialogues of tragedy, differences of viewpoint are often dramatised as building to an explosive climax, which may lead to violence or permanent estrangement (including physical separation, for instance through exile). The tempo of scenes of this type can sometimes be measured through the shift from longer dialogue to stichomythia (especially but not exclusively in the agon-scenes). Despite the eloquence and rhetorical sophistication of the participants, communication frequently breaks down: one character 

 



E.g. Hall , –. For Bakhtin see e.g. Morris , and for modern studies of his work, Clark and Holquist , Holquist , Lodge , Vice . The Dialogic Imagination (Eng. tr. ) has been especially influential. E.g. Pentheus’ harping on the sexuality of the maenads. This is not exactly a stylistic trait as commonly understood, but it is certainly indicative of his personality. Gender differentiation has received most attention: see esp. McClure ; also Mossman  (who promises a fuller study). On children see Golden ; on the speech of older men see Silk , in the same volume. See the outstanding paper by Griffith . Chrysothemis and Electra in S. El. may at a superficial reading seem just like the sisters in the Antigone, but more detailed study would certainly show a range of differences: for a summary, see Finglass on El. –.



Introduction

will not listen to the other, or cannot understand the other’s point of view, or cannot bear to remain in the other’s presence. Moments of reconciliation and concluding harmony do occur, but they are rarer, and in some cases follow upon the disastrous consequences of dissension (e.g. the final scene of Hippolytus); elsewhere reconciliation proves futile, since the outcome has become unavoidable (as with the ending of Menelaus’ and Agamemnon’s dispute in IA). In other cases, speakers seem not to be on the same wavelength: they are ‘talking past’ one another. The scene in Euripides’ Supplices is an example of this phenomenon, for Theseus fails to answer the herald’s argument about the power of demagogues in popular government; the herald in his turn has nothing to say about the compelling case made by Theseus against monarchy. Still more extreme are cases in which one speaker ‘breaks contact’, refusing to engage further with those on stage and instead addressing either him/herself or some other entity – the gods, the physical landscape, or persons absent or deceased. Besides speech, there is silence. We see from Aristophanes’ Frogs that Aeschylus was famous for his technique of bringing on characters and leaving them silent for long stretches, keeping the audience in suspense: Cassandra, Niobe, Achilles are the paradigm instances. Such occasions are given special force when others try to induce the silent one to speak, as appears to have been the case with all of these. Less conspicuous are the cases where a character lapses into silence while others speak around him: often this can be interpreted in various ways, but the crucial point is how and when the character speaks again. ‘Significant silences’ are found in all three dramatists. They are obviously a part of the tragic repertoire; at the same time they contribute to characterisation (defiance, resentment, the build-up to an explosive outburst). In a way this example clinches the argument that a study of tragic language need not, indeed must not, be confined to the words.

  

 

E. Phoen. –, especially the section in tetrameters from  to the end of the scene, is an excellent example. Many cases of this kind are examined by Mastronarde , ch. : see e.g. A. Ag. ff., E. Ion –., S. OC –. Ar. Ran. –, with Taplin , a classic discussion. Different is the handling of the figure of Pylades in A. Cho.: though in earlier scenes apparently a mute character, he breaks silence at the crucial moment (–). See further p. . E.g. E. Supp. , El. , Or. . Taplin , at –. For Sophocles see e.g. Orestes at El. –, Jocasta at OT –, Neoptolemus at Phil. b–. For Euripides see previous note; also E. Hipp. –, in an agon: Theseus first refuses to speak and then utters generalised utopian wishes rather than engaging with his offspring.

III Ancient and modern study of the subject



iii ancient and modern study of the subject I conclude this chapter with a brief and selective survey of the most notable discussions of our subject in ancient and modern times. Even in Homer we can see a kind of implicit poetics, and later writers, notably Pindar and Aristophanes, comment extensively on the nature of their art. It is reasonable to suppose that the tragedians also were conscious of the generic and stylistic choices they made: this is indeed supported by Sophocles’ alleged comment on his own development (already mentioned, p. ), and by the evidence that he composed a work on the chorus (T. Radt). There are also self-conscious allusions to formal features within the text (e.g. A. Eum. –, on the stichomythia about to commence, or E. Med. , on the agon in progress); and Euripides in particular seems to play with conventions and disrupt the audience’s expectations. Contemporaries such as Prodicus and Protagoras scrutinised poetic texts and analysed their weaknesses in terms of ½rqo”peia, ‘correct use of words’. Gorgias evidently held views about tragedy, being especially interested in its power over the emotions and its capacity to beguile the audience, bringing about what we might call a suspension of disbelief (frr. B and  DK). But by far the most important evidence from the fifth century is that of Aristophanes, especially in the Frogs. For all the comic exaggeration and distortion, this presentation of the rival tragedians must bear some relation to perceptions, however unfair, held by some of the audience. In the literary contest, Aeschylus is the spokesman for the good old days when men were men, and tragedy taught military and moral virtue; Euripides is the modernist, in style as in content: sophisticated, democratic, lucid in argument. Sophocles does not appear in the play: it is an old guess that he died shortly before it was put on, and the few references to him were inserted at the last minute. But it is clear that Aeschylus and Euripides stand for the two extremes, antique versus modern, older generation versus new: neither Aeschylus vs. Sophocles nor Sophocles vs. Euripides would have worked nearly as well. Aeschylus, the veteran of Marathon, who died fifty years before Frogs was performed, is a fitting symbol of the golden age of Athens, now long past. Negatively, his kind of tragedy seems remote, magnificent but archaic. The point is made more concisely in the Clouds:   

E.g. Winnington-Ingram . Pfeiffer , ch. ; see also Ford , with the valuable review of Halliwell . For discussion of the criticism in Frogs see Dover’s introduction, –; also Hunter , ch. , Griffith , Halliwell , ch. . For the other comic poets on tragedy, see the references collected by Olson , , , , ; also Silk b.



Introduction

when old Strepsiades asks his son to give him a rendering of something by Aeschylus, the sophist-trained youth dismisses the idea: ‘I regard Aeschylus as supreme among poets – at being full of noise, incoherent, a bombastic ranter and a creator of mountainous words’ (–). Instead he offers a rhesis of Euripides involving brother-sister incest, from the notorious play Aeolus. Reverting to the Frogs, we can see that Aristophanes has some awareness of critical terminology: the differentiation of prologue, lyrics and spoken verse is clear, and line , with its reference to ‘correctness in terms’, alludes to the sophistic interest in ½rqo”peia. The actual criticisms of the tragedians relate to both form (vocabulary, music) and content or moral impact (Aeschylus inspired men to heroism in war, Euripides corrupts them with decadent ideas). Lines are cited and criticised for swollen or extravagant compounds (–, –, ), redundancy of language (, ), monotony of construction (–, on the younger poet’s prologues). Form and content come together in the attack on ‘the Muse of Euripides’, for loose and effeminate metre and music are associated with ‘low’ or ignoble subject matter in the brilliant parody of monody (– and context, cf. ). The blend of technical and moral fault-finding is typical of much ancient criticism. Aristophanes has little to say about Sophocles: the adjective he does use of him, eÎkolov (‘relaxed, easy-going’ ), fits well enough with other testimonies about Sophocles’ charming and agreeable character, but gives us little insight into his art. In general critics have found it much harder to sum up Sophocles, whether describing his art or his thought: the other two playwrights have more pronounced characteristics, and their achievements seem to be in some ways easier to define as well as to parody. Later critics often treated Sophocles as the golden mean between the other two. Thus Dionysius of Halicarnassus describes Aeschylus as master of the austere style (like Thucydides in history, Pindar in lyric, Antiphon in oratory), Euripides of the polished (cf. Hesiod, Sappho, Isocrates), and Sophocles of the style which falls between these – the tempered or wellblended style (eÎkraton), in the same traditions as Homer, Herodotus, Demosthenes and Plato (Comp. Verb. –)! Dio in his essay on the three    

Cf. GGL i. .n.. Esp. Ion of Chios FGrH  F (= S. T Radt); see also Vita ; Lefkowitz , –, Olson , . Cf. Housman’s ‘Fragment of a Greek Tragedy’, which draws mainly on Aeschylean models. This masterpiece of parody has often been reprinted: see e.g. Ricks , –. For further information on ancient discussion see Trendelenburg ; De Propris ; Stanford , ch. .

III Ancient and modern study of the subject



tragedians gives more detail. Aeschylus he describes as possessing greatness of soul (megalofrosÅnh) and simplicity; his handling of the chorus is altogether more tragic and simpler (tragikÛteron kaª ‰ploÓsteron) while that of Euripides is more political and more precise, politikÛteron kaª ˆkrib”steron (.); Euripides writes clearly and precisely; he takes care not to introduce anything implausible and not to leave out explanations (.); his technique is highly political and rhetorical (politikwt†th kaª çhtorikwt†th), and most beneficial to readers (., cf. ; Quint. Inst. ..–). Again the critic seems to have found Sophocles the hardest to sum up: his lyrics are not notably didactic, and do not excite to virtue as Euripides’ do (.), but he ‘produces a poetry that is august and majestic, highly tragic and euphonious in its phrasing, so that there is the fullest pleasure coupled with elevation and solemnity’ (.). Aristophanes’ silence about Sophocles combined with Aristotle’s admiration for him have led to an over-simple model of tragedy’s development, with Aeschylus representing the primitive beginnings of tragedy, Sophocles its state of perfection, Euripides its decadent decline; but this not only defies chronology (Euripides died before Sophocles) but ignores the clear evidence of interaction between Sophocles and Euripides, with each learning from the other. Stylistically speaking, even the opposition between Aeschylus and Euripides is not straightforward, as two parallels quoted by ‘Longinus’ and Aristotle show. In describing the epiphany of Dionysus Aeschylus wrote –nqousii dŸ däma, bakceÅei st”gh (A. fr.  Radt)

The house is inspired, the building is raving.

while Euripides, imitating the same effect, wrote ‘less harshly’, according to Longinus: pn d• suneb†kceu' Àrov (Bacch. )

the whole of the mountain joined in the revelling (sc. with the Bacchants).

Here Longinus (.) seems to feel that Aeschylus’ metaphors are too daring, and that the Euripidean version reduces the bizarre effect (there is one clause rather than two, and it is nature, not a man-made structure,  

Dio Chrys. Or. , tr. in Russell and Winterbottom , –. Cf. Luzzatto . Cf. Long , .



Introduction

that participates in the bacchic experience). But in another case, cited by Arist. Poetics . b–, Euripides intensifies the image and, according to Aristotle, heightened the beauty of the line. Writing of Philoctetes, Aeschylus made him speak of fag”daina< >, ¤ mou s†rkav –sq©ei pod»v (A. fr.  Radt)

the cancer that eats the flesh of my foot,

but Euripides replaced the verb with qointai (‘that feasts upon’) (fr. ). Here we are guessing at the effect of particular verbs on the Greek audience, but although both verbs are found in prose, in this case it does seem that Euripides is striving for a more vivid, more highly coloured diction than his predecessor. These and many other comparisons make evident that a more exact and less evolutionary critical method is called for, which while not eschewing comparative judgements tries to assess each of the dramatists in their own terms. Yet the influence of the Frogs has permeated some of the most important critical writings of the modern era: the shadow of Aristophanes hangs over the important treatments by Schlegel in his Lectures on Dramatic Art and Literature and by Nietzsche in The Birth of Tragedy. Mention of Aristotle brings us closer to what we might regard as a more academic tradition of study. The Poetics counts lexis, diction, as an important element in tragedy, and both here (in ch. ) and in the third book of the Rhetoric the philosopher sets out some of the principles which underlie the use of poetic diction: ‘impressive and above the ordinary is the diction which uses exotic language (by “exotic” I mean loan words,  

  



See further Innes , . –sq©w is found only here in Aeschylus ( cases in Sophocles,  in Euripides, three of them in the Cyclops;  in Aristophanes, not including compounds. By contrast qo©nh and cognates are never found in Aristophanes, frequent in Euripides (less so in the work of his predecessors). Vigorously argued for Aeschylus by Lloyd-Jones a, – (= , i. –). For valuable essays on these see Behler  and Henrichs , in an issue devoted to the nineteenthcentury reception of Euripides; on Nietzsche see also Silk and Stern  and Henrichs . The omission of Plato from this discussion needs some justification. Tragedy is important for Plato chiefly as a social and cultural phenomenon – a corrupting one, in the main. His discussions of the genre include some sharp and suggestive formulations (e.g. Grg. d on tragedy as a form of demagogic rhetoric, chiefly or solely concerned to give pleasure to the audience; Laws ab on generic blending; d, on blasphemies and emotionality), but for the most part they look to tragedy’s impact on society rather than illuminate the inner form or language of the dramas. For Plato’s complex attitude to tragedy see esp. Halliwell , chs. – (summary of his position at pp. –). Halliwell  (reissued  with new preface) is the standard work: for our purposes see esp. ch. ,  and app. .

III Ancient and modern study of the subject



metaphors, lengthenings, and all divergence from standard usage)’ (Poet. .a–). Elsewhere he comments that dialect words are appropriate to epic, compound words to dithyramb, metaphor to tragedy: the idea is over-schematic, but at least it shows that such questions were under discussion. Aristotle’s passing references often seem to take up examples that were already controversial (as also in his chapter on Homeric problems): inconsistency of characterisation, as with Iphigenia; the awkwardness of the use of the deus ex machina, as in Medea; and, from the same play, the lack of preparation for Aegeus’ entrance (if this is what the rather cryptic reference means). So also with matters of diction: we are told that Ariphrades (probably the comic poet) made fun of the unnatural inversion of word order in tragedy (Poet. .b, dwm†twn Špo, hardly an extreme example). Elsewhere Aristotle remarks that the expression kÛphv ˆn†sswn, ‘being lord of the oar’ (i.e. rowing) is unsuitable, because the word ‘being lord of ’ is too elevated for the subject, so that the artificiality of diction is too obvious. Manner of speech and characterisation come together in the reference to ‘the rhesis of Melanippe’ as an example of inappropriate character (Poet. .a). The speech in question was obviously notorious (it is alluded to in Aristophanes’ Lysistrata) because it allowed a woman to show knowledge of cosmological speculation (ascribed to her mother). With specific criticisms such as these we come close to the concerns of the ancient commentators on tragedy. Study of the scholia on Homer has borne fruit in recent years not only for our understanding of ancient criticism but as a spur to modern discussion; the tragic scholia are less rich, but there is much of interest, especially in those on Euripides, ranging from the detection of irony or foreshadowing to puzzlement about the relevance of a choral ode. Since the scholiasts think in rhetorical terms, there is much comment on speeches and verbal detail. Ancient and modern perceptions find some common ground in the controversial scene in which Orestes mocks and bullies the terrified Phrygian. This scene has disturbed many readers who find Orestes a repugnant hero. On  (where Orestes demands that the eunuch declare that Helen deserved death) the scholia declare ‘the words here are unworthy of the tragedy and of Orestes’       

See further Isoc. .–, and Silk , , . Rh. .. a–, citing E. Telephus F; cf. e.g. A. Pers. –, E. Cyc. –. Melanippe the Wise F: cf. SFP i ad loc. See esp. Trendelenburg , Elsperger , Heath , e.g. , ,  and elsewhere; Heath , ; Meijering , Dickey , –, N¨unlist . See e.g. Richardson ; also Heath , ch. . Easterling b is a helpful essay on the Sophoclean scholia. See esp. the index analyticus in Schwartz’s edition of the Euripidean scholia, ii. –.



Introduction

misfortune’, and a few lines later, at  (a babbled line of self-exoneration by the Phrygian), they remark: ‘this is more in the manner of comedy, and prosaic’ (taÓta kwmikÛter† –sti kaª pez†). Another area of growing interest is the anthology tradition, which like many literary texts illustrates the range of quotations which were most popular, most apt to particular circumstances, or regarded as most worthy of preservation and study in late antiquity. This material not only preserves fragments of lost plays but provides valuable indications of ancient taste. The contribution of ancient criticism to our enquiry, then, is of limited use. First, with the analysis of Aristotle and his heirs, it provides a coherent conception of tragic diction and its distance from everyday speech, though this may need qualification. Second, more specific comments by critics, scholiasts and others often draw our attention to interesting details, and in some cases at least their response may help us see the text in question from a less anachronistic perspective (though we must remember that much of this body of criticism was compiled long after the heyday of tragedy). Third, the comparison of two or more of the three tragedians began already in the fifth century, and was developed with brilliant wit by Aristophanes in particular; but although this dramatic synkrisis operates more imaginatively and with a greater range of categories than the more particular comments of ancient scholarship, it is also seriously misleading as criticism, and despite frequent refutation, its influence still needs to be resisted. Any review of modern work on tragedy, even when focused on a particular aspect, risks swamping the reader with its bulk. Here I attempt a mere sketch of the main lines of work on the stylistics of the genre. Early modern work on tragedy was above all editorial in the sense that scholars were primarily concerned to establish the text; alongside this went the need to clarify the linguistic and metrical rules which the tragedians followed. It is a grave distortion to suggest that scholarship prior to the twentieth century never concerned itself with anything but 





Stobaeus’ Florilegium amply illustrates the greater popularity, or at any rate quotability, of Euripides. In the index volume to the edition of Wachsmuth and Hense Euripidean quotations outnumber those of any other author, easily outstripping even Homer and Plato. The listing of citations for Euripides covers over twelve closely-printed columns; Sophocles needs only four, Aeschylus does not even fill one. Cf. Morgan ; also various contributions by Easterling, including her chapter ‘From repertoire to canon’ (Easterling c). For papyrus-quotations see the database of Mertens-Pack (online at http://promethee.philo.ulg.ac.be/cedopal). For the larger picture see Sandys –, Pfeiffer , Wilamowitz , Highet ; a valuable outline by Collard ; also Michelini , part , Goldhill b, and the editors’ introduction to Goldhill and Hall .

III Ancient and modern study of the subject



textual criticism: the sympathetic survey of his predecessors by Fraenkel in the introduction to his Agamemnon is sufficient to refute this claim. Some editors extended their scope well beyond the strictly textual: Elmsley in  already gathered some material on the etymologising of names in a note on Bacchae , with reference to the same etymology of Pentheus’ name which was one of our opening examples. A few were alert not only to issues of staging and production but to the larger religious and historical dimensions of the drama. The famous quarrel between G. Hermann and K. O. M¨uller concerning the Eumenides illustrates the existence of differing approaches to the tragic texts: on the one hand, a scholar who treated textual and linguistic questions as central to his work, and on the other, one whose work on the text demands respect, but who also boldly drew on archaeological evidence and on material relating to Greek religion. Inevitably, as long as the attention to verbal detail was diffused across the body of the work, the picture of tragic form and style remained imperfect. Discussion of style has had a long-standing tendency to the atomistic: particular points are discussed in the notes on particular lines. This remained the case even with the expansion of the commentary form. Perhaps only with Wilamowitz’s edition of the Heracles did a model emerge for a commentary which never lost sight of the larger picture. In the English tradition Jebb’s series of commentaries on Sophocles has achieved a kind of classic status, but more for the editor’s humane judgement and numerous particular observations than for synthesis. The atomistic tendency was also observable in the many doctoral dissertations which appeared throughout the later nineteenth and early twentieth centuries. Many of these took up very particular questions, often focused on aspects of vocabulary, figures of speech or grammatical phenomena (alliteration, asyndeton and so forth), assembling examples from one or more of the dramatists but not always producing more than a catalogue. Even the useful compilation of material on grammar and stylistic usage in Sophocles by Bruhn, produced as an appendix to the commentary by Schneidewin and Nauck, was no more than an accumulation of data. In 

 

Wilamowitz , –, –; Fraenkel on A. Ag., , i. –, but note what is said of Hermann’s understanding of the broader picture on  (with n. ); Pfeiffer , –; H. LloydJones in the intro. to Wilamowitz , viii–xi. Jebb’s commentaries, originally published –, have been re-issued with new introductions in the early s; note esp. the general introduction by Easterling prefaced to each volume. Much of this work was cited and synthesised in the massive history of Greek literature by Schmid, GGL: see GGL i..– (Aesch.), – (Soph.), i..– (Eur.). I should also mention Smereka -, though the work was never completed and is not easily accessible. See also Budelmann , –, on the shifting emphases in Sophoclean studies.



Introduction

English Lewis Campbell’s essay on Sophoclean language performed some of the same work, but with more by way of critical comment. This last is still much-cited, and with justice. Attempts at a more synthesising approach to style were slow to appear. The pioneering approach of Wilamowitz, the champion of Altertumswissenschaft with its ideal of total scholarship, is paramount. Many particular notes in his Heracles illuminate not only the passage in hand but the genre and Greek mentalities as a whole. Equally important, the famous Einleitung which constituted the first volume amounted to a major monograph on tragedy. Although not concerned with style in detail in this volume, he gave an unparalleled survey of the genre rich in detailed aperc¸us, and also commented acutely on the traditions of criticism and scholarship. Many of those who subsequently contributed most to the study of tragic form and style among other aspects were pupils of Wilamowitz. Particularly important are W. Kranz’s study of the chorus, Stasimon (), and W. Schadewaldt’s penetrating account of soliloquies and related types of speech in Monolog und Selbstgespr¨ach (). These studies have few rivals in their combination of precise and sensitive scrutiny of passages and broad perspectives on the genre and its historical development. Further developments in the German tradition included much valuable form-criticism, examining the particular parts of tragedy with detailed documentation and comparison, sometimes in exhaustingly statistical detail: the prologue, the rhesis, stichomythia, and so forth. Much of this was synthesised in the collection Bauformen, edited by W. Jens, but this is probably most useful as a reference book. Statistics are also naturally prominent in works which seek to establish the authenticity or the reverse of disputed works: for instance, the monographs on 



 



The essay (Campbell ) is divided into two parts: the first on Grammar, the second on what he describes as ‘Peculiarities of Diction’, which includes sections on ‘sensuous and suggestive words’ and on metaphor; but even the first part has many comments which go beyond the purely grammatical. See e.g. the notes on  (on stichomythia),  (on personification),  (the Greek love of antithesis). Fraenkel’s commentary on the Agamemnon carried this method still further: see amongst others the nn. on ff., f. (on perverted imagery),  (on tricola of adjectives beginning with the privative alpha), and e.g. the index s.v. ‘pleonastic description of anything excessive’, ‘humble matter enobled by language’. Wilamowitz , esp. –, –. Nestle , Jens , Strohm . More recently narratological studies have taken formalism on to a more advanced theoretical level (e.g. De Jong , on the Euripidean messenger speech; Markantonatos , esp. ch. ). See also Schwinge , Erbse . Of the more recent commentaries on particular plays, Collard’s on Eur. Supplices is especially important in bringing these formal studies before an English-speaking public.

III Ancient and modern study of the subject



Prometheus by Griffith and on Rhesus by Ritchie amass much material on vocabulary and linguistic usage which can be drawn on with profit even by those not directly concerned with the particular controversies which they examine. The influence of the so-called New Criticism in Britain and America encouraged some scholars to be less dismissive of the practice of writing about poetry in a non-technical way. Many essays and articles, and some books, appeared in the mid-twentieth century and especially after the Second World War, though not all of them have stood the test of time. An attractive short book by the scholar-poet Richmond Lattimore, The Poetry of Greek Tragedy (), may still be read with pleasure, though it is meant for the general reader and all passages are quoted in his own verse translations. These, like the book in general, inevitably have a slightly dated feel, but it has not lost its value. The same may be said of a French book from the same period, Jacqueline de Romilly’s L’´evolution du path´etique (). More important for the trend of scholarship was the rising tide of work on poetic metaphor and imagery, stimulated by impressive studies of Shakespeare. Goheen, who also wrote rewardingly on the Oresteia, published an influential study of the imagery of Antigone, and Winnington-Ingram’s book on the Bacchae () made much use of the metaphorical language in that play, above all that of the hunt. Anne Lebeck’s later account of the language and imagery of the Oresteia has also been profoundly influential, and most scholars today would grant that the study of imagery, while sometimes overplayed, is an essential tool of the critic of tragedy. Images are easy to find, not always straightforward to interpret; some of the issues are discussed in a later chapter. Another important feature of tragedy, already touched on in the preceding section, is the prominence of rhetoric, in the sense of sustained debate and persuasive argument reminiscent of the lawcourts. Here again older studies often did little more than catalogue examples of particular 



 

Here too the atomistic approach is unduly prevalent. Special mention should be made of Clay  (the companion volume of  is less useful), which is an attempt to set out in categories the words used by each of the three tragedians, so that it is possible to see easily how much of the tragic diction is common to two or more of them. The work has serious defects but is nevertheless valuable. See the comments by Long , –, and by Griffith , ch.  (esp.  n. ). Griffith  (cf. earlier esp. Schmid ), Ritchie  (on which see Fraenkel ). Garvie’s important monograph on Aeschylus’ Suppliants (, recently reissued with new foreword) also assembles much valuable material on vocabulary and style in his second chapter. In English note e.g. Earp , ; better, Stanford . Winnington-Ingram a, Goheen , Lebeck . See further Barlow ; Hiltbrunner , Petrounias . Porter  is a valuable survey-article. See ch.  nn. –.



Introduction

parallels or figures of speech. Critics are now no longer content to set out the structure or formal attributes of a speech but also seek to understand its dynamic role in the drama. Characterisation, shifting of mood, and modification of audience-responses are part of what is involved, but other aspects also need to be considered. Rhetoric can be most illuminating when it does not succeed in persuading, or when the wrong moves are made, or speech-making itself may seem inappropriate. Chronology too is relevant here: not only do the tragedians develop more sparkling or showy rhetorical styles as the century advances, but increased disillusionment with or suspicion of rhetorical persuasion becomes more conspicuous, both inside and outside the tragic texts. These developments are not un-related. Commentaries continue to appear, and their authors have in general sought more and more to produce a synthesis in their introductions: the typical introduction to such a work now constitutes a substantial essay. Aspects of style often figure in these preliminary sections, whether treated seriatim or in more generalising descriptive terms. Other critics have followed the lead of Winnington-Ingram (as often, a pioneer) in publishing monographs on individual plays, in which aspects of form, style and rhetoric sometimes figure in separate chapters or sections. The more expository approach certainly makes easier reading than the catalogues in some commentaries, and the best books of this kind manage to make clear what is common coin in the genre and what is special (and often significantly so) about the play under discussion. The rise of structuralism as propagated by the Paris school of Vernant and Vidal-Naquet, Loraux and others encouraged an eclectic use of anthropological and socio-historical models which might have been expected to draw attention away from the poetic text. Nevertheless, seminal essays by the leading figures have illuminated some of the key conceptual terms and ideological use of language in the tragedies, seeing them as reflecting the values and social structures of classical Athens. There has been abundant output on religious structures and ritual language. The work      

E.g. GGL i. .–, with older citations; see also Duchemin . Lloyd  is a more up-to-date account of the agon, though still in many ways following the formalist tradition.  See further Buxton . Halliwell ; see further Pelling . E.g. Griffith on Antigone, pp. –, or Willink in his commentary on E. Orestes, lii–vii. E.g. Mossman , Allan . Vernant and Vidal-Naquet  and  with subsequent translations (combined volume in Eng., ); Loraux ,  and . See the notable review discussion by Hall . On religious language in tragedy see Citti ; also e.g. Faraone  on curse tablets in relation to the Binding-song of the Furies in A. Eum.; Easterling b; Pulleyn a on Greek prayers generally; Stehle in Murray and Wilson , –, on choral prayer in tragedy; Willi , –.

III Ancient and modern study of the subject



of John Gould brilliantly combined close reading of tragedy with anthropological insights into Greek religion, in the tradition of E. R. Dodds. Post-structuralist or deconstructive criticism made less of an impression on a sceptical (or conservative) discipline, although Simon Goldhill’s first book, a long and detailed reading of the Oresteia, created something of a stir and has been found stimulating by many readers of a very different outlook: the emphasis on ambiguity, obscurity, oracular use of language, and other difficulties or barriers to communication is at least acceptable in a moderate form even if one rejects the theoretical underpinnings. Current trends are more in the direction of audience-oriented criticism and readerresponse theory: the interesting monograph on Sophoclean language by Felix Budelmann draws on this tradition. More recently still, a new wave of studies inspired by linguistics has emerged especially from Holland: the book of Helma Dik on tragic word order will surely become a standard work. Many of the most rewarding and influential readings of tragedy have come from outside the academy, through essays and appreciations and above all the literary adaptations or recreations of the dramas: in recent years Tony Harrison, Ted Hughes, Seamus Heaney and others have enabled audiences and readers to experience particular dramas, above all the Oresteia, in powerfully innovative versions. As knowledge of the Greek language continues to decline, translations become ever more crucial as a means of awakening or maintaining the enthusiasm of the wider public. With the boom in reception studies, translation studies too can expect to flourish, and a detailed survey of the history of translation of Greek tragedy in English would be illuminating for the student of tragic style in the original as well. Apart from their intrinsic interest, these renderings or ‘re-readings’ of tragedy can help us see more than we saw at first in a play or passage; even if we decide in the end that the translator has ‘seen’ what is in fact not there, the process has sharpened our own attention. In more recent years there have been distinguished contributions both on specific passages and plays and on the tragic genre as a whole, which take    



His essays are collected in Gould : see esp. –, –, –. Goldhill , extensively reviewed and discussed: see esp. Moles  and Clark and Csapo . Budelmann : the sub-title, ‘Communality, communication and involvement’ is telling. On reader-response theory see esp. Iser ; Tompkins  is a useful anthology. Dik ; see also the essays in Rijksbaron  (e.g. G. Wakker on mžn, A. M. van Taalman Kip on § g†r), and in de Jong and Rijksbaron . Another valuable contribution is the revisionist view of the ‘instantaneous’ aorist by Lloyd : though not all are convinced, his discussion is important. France  is an indispensable reference work; see also now Walton , Lianeri and Zajko  (including an essay by Schein on translating Aeschylus).



Introduction

account of older approaches but bring a fresh sophistication. The best of these avoid dogmatism and critical superiority and seek to explore these extraordinary plays with scholarship not untinged with humility. Greek tragedy remains a perennial subject for debate and an inexhaustible source of inspiration. 

A personal selection: Silk b; Halliwell ; Easterling a and a; Schein ; Griffith  and . See also Henrichs –, which, though not primarily concerned with style, includes subtle readings of several important choral songs.

c ha p te r 2

Genre: form, structure and mode

i formal description Greek tragedy is a hybrid form, and the different parts of the drama are differentiated in form and style. Some account of the formal structure of the genre is therefore a necessary part of the background to more detailed analysis of specific passages. Most genres of Greek poetry are metrically consistent. Homer’s epics are entirely composed in dactylic hexameters; the elegists write in elegiac couplets throughout. Most elaborate are the lyric poets, who compose in stanzaic form, but even the most ambitious of these songs follow a recurring pattern (thus Pindar composes his odes in triads, each of which follows the same metrical scheme). Tragedy and comedy differ in that they incorporate a great variety of metres and modes of delivery. The variety also employed by satyr-drama is essentially imitative of (and arguably parasitic on) tragedy. The crucial distinction is between spoken and sung verse. Spoken verse is normally in iambic trimeters, the standard metre for dialogue in both tragedy and comedy, and the metre which according to Aristotle corresponds most closely with the rhythms of everyday speech (Poet. .a– ). In some passages a longer line is used, the trochaic tetrameter: Aeschylus makes some use of this, but it then seems to have fallen out of fashion until extensively revived in Euripides’ later work, where it is often employed in scenes of swift argument or agitated debate. Sung or lyric verse is far 



This section sets out some basic points which will be familiar to specialists. Those who already have a clear grasp of the structure of tragedy may wish to proceed directly to section ii. For other surveys which cover some of the same ground in comparable or more detail see Mastronarde on Medea, intro. pp. – (structure, syntax, style, metre); Easterling b and Burian ; Rosenmeyer , ch. . Dodds on E. Bach. –; DFA –: e.g. Her. –, Ion –, Hel. –, Phoen. – (with Mastronarde’s general note on this section), Or. –. Sophocles rarely uses this metre, and one surviving case (OT –) does not fit the tendency described. See also Drew-Bear .





Genre: form, structure and mode

more ambitious and elaborate, and the language of song is correspondingly more elevated, more remote from normal discourse. Furthermore, speech is normally more reflective, including description and analysis, whereas song conveys a greater sense of excitement and uses imagery and exclamations more freely. As a general rule we can state that the actors in tragedy speak, while the chorus of tragedy sing, but this principle needs extensive modification. Actors and chorus interact in spoken dialogue: normally the chorus-leader acts as spokesman and adopts the standard trimeter for these exchanges. More importantly, actors sometimes launch into lyrics, especially at moments of deep distress. Sometimes these songs are part of a sung dialogue with the chorus (‘amoebaic’ exchange), sometimes they form part of an exchange with another actor (whether singing or speaking), and sometimes they are solo performances (arias or monodies): the last is especially characteristic of Euripides. These shifts between speech and song are an important means by which to gauge the changing mood or emotional tempo of a scene. When an actor launches into song, it normally marks a significant moment; when one actor speaks and another sings, the contrast of modes will often be indicative of characterisation or contrasting response to good or ill fortune. Frequently this contrast will be related to difference in gender or rank. Tragic drama, therefore, is fundamentally constructed in terms of the alternation of speech and song. However, a more exact structural design, which goes back to Aristotle, is commonly used and with some refinements will be assumed in what follows. The relevant passage runs as follows: OE 8   P (8  H,  ",   H(,  8 ! ,!  & Q    " ,( ) ,,   ,R   F " ,  .' ! (8 ) ! 8 ,), (,   (8 J) -, H  8  +!  ,E  ((. ,  8  (8 ( 4   ! ! "- ',  ,  8 ( 4   ! (FS 4 " 6 (6, F 8 ( 4   (  :   , "- ( " - 8 ) (8 T 7E F  4E "-, ,), ( 8 ( "- ! M' +,'  "', ((! 8   ! "-  +! ,E. (Poet. .b.–) We spoke earlier of the components of tragedy that must be used as basic elements; but its formal and discrete sections are as follows: prologue, episode, exodos, choral unit (further divisible into parodos and stasimon). These are common to all plays, 

That the tragedians were regarded as artists who paid close attention to the structure of their works is indicated by Pl. Phdr. c–e; cf. the emphasis of Aristotle in the Poetics on the construction of the tragic plot, and Menander’s comment quoted at Plut. Mor. e–f.

I Formal description



but actors’ songs and kommoi [laments] are particular to some. The prologue is the whole portion of a tragedy prior to the chorus’ parodos; an episode is the whole portion of a tragedy between complete choral songs; the exodos is the whole portion of a tragedy following the final choral song. Of choral units, the parodos is the first complete utterance of the chorus; a stasimon is a choral song without anapaestic and trochaic rhythms; a kommos is a dirge shared between chorus and actors. (tr. Halliwell)

This scheme takes as basic the sequence prologue, choral entry song (‘parodos’), episode, choral song, episode, choral song . . . , concluding with the ‘exodos’ or closing sequence, at the end of which all players leave the stage. On this model choral song and dialogue episode alternate throughout, with the major choral songs (often called ‘odes’) marking pauses in the action (sometimes, but by no means always, these songs are performed when the stage is clear of actors). This scheme has some merit, but no tragedy is quite that simple. This model neglects or glides over some important points. In early tragedy the drama sometimes begins with the chorus’s entrance, so that the term ‘prologue’ is irrelevant. The scheme also ignores the varying length of episodes, and misleads if we assume that each episode is a uniform entity. In fact the interval between two odes may be hundreds of lines long, and can include some clearly separate sub-sections. Moderns familiar with Shakespeare or neoclassical drama will naturally think in terms of acts and scenes, but the parallel is not exact. The interval between odes could be designated an ‘act’, but action within that span is normally continuous, even if characters may leave or be joined by others. In fact the five-act structure seems to derive from a later, more rigid dramatic design: it was probably established in tragedy in the fourth century and is also used by New Comedy. In fifth-century tragedies choral odes may be as numerous as six, so that ‘acts’, if one includes the prologue and exodos, would number seven. A further problem with Aristotle’s scheme is that it takes little account of actor-song; yet there are clear cases where a lyric exchange between the chorus and an actor (or, in later Euripides, an actor solo) takes the place of an act-dividing song. A more flexible account of tragic structure has been offered by Taplin, who argues that Aristotle’s mode of thinking is over-rigid and insensitive    

A puzzling assertion, as both anapaestic and trochaic metres do occur in stasima; as to the former, Aristotle probably means ‘marching’ anapaests. Taplin ; Brink on Hor. Ars Poetica –. For the former see S. Phil. –, E. Hel. –, and Taplin –; for the latter, E. Phoen. –, IA –. Taplin ch.  passim, esp. –. He even questions the authenticity of chapter  of the Poetics (–).



Genre: form, structure and mode

to action in the theatre. His own schema focuses much more on stagemovement, in particular the closing of a sequence by the exit of a character. The action is articulated either by an act-dividing song or by a character exit: because the cast of characters in Greek tragedy is small and the action on stage more restricted, new arrivals and departures gain in significance. He makes the point most emphatically in the following passage: as I see it, then, the formal structure of Greek tragedy is founded on a basic pattern; enter actor(s) – actors’ dialogue – exeunt actor(s)/choral strophic song/enter new actor(s) – actors’ dialogue . . . Beneath the many complexities of the construction of the plays there lies, I suggest, this simple form . . . The earlier plays of Sophocles and Euripides approach it most nearly. S. Ant. is almost perfectly regular. But the basic pattern is seldom so near the surface; there are a multitude of variations on it.

Aristotle’s scheme, even with Taplin’s modifications, is essentially external, focusing on the formal structure in terms of song–speech alternation, and on stage movement in terms of exits and entries. Other aspects of the tragic repertoire can be seen in terms of recurring types of speech. In the dialogue sections of the drama there are two extremes: very long and argumentative speech (rhesis) and short one-line exchanges, whether in trimeters or tetrameters (stichomythia); a variation on the latter is a series of exchanges involving two lines each from speakers (distichomythia). Shorter and more freely structured dialogue does occur, but there is some tendency for speakers to move into one of these modes as a scene advances. Important scenes make use of both. Certain formal features of the genre were established early. The most obvious is the messenger speech, already present in Aeschylus’ Persians and conspicuous throughout the century. Almost every play has a messenger speech: some have more than one. This is a special type of rhesis, with its own conventions. If we accept that tragedy began as a predominantly choral form with only one actor, large parts of the play must have been occupied with the actor bringing news to the chorus. The messenger speech, though normally confined to a single scene of the play, is regularly the longest single speech, amounting to  or  lines. The other conspicuous use of the extended rhesis is in the so-called agon or contest of words, a full-scale debate normally involving speech and counter-speech by two   

Taplin . For a discussion of some of the issues raised by Taplin and others regarding the structure of tragedy, see Poe . See e.g. S. OT –, discussed in ch. , pp. – below. For the messenger speech see below, ch. , –. For a recent discussion of why tragedy needs messenger speeches see Sommerstein .

I Formal description



opposing characters. Like the messenger speech, this is discussed more fully in a later chapter. Besides dialogue there is also monologue, in which a character speaks without wishing or expecting to be overheard. The conventions of Greek tragedy make this process different from the canonical soliloquising of Shakespeare, in which the darkest secrets, hopes and fears of the character are exposed. In particular, the presence of the chorus makes total privacy hard to obtain except in the prologue (or, rarely, in ‘secondary prologues’ when the chorus has exceptionally left the stage: e.g. A. Eum. –, S. Aj. –, E. Hel. –). The prologue-monologue is, however, a rather special category, as the speaker is often describing his own ancestry and situation in a way which is intended to provide orientation for the audience: this practice brings these monologues close to audience address. Yet the presence of the chorus is less significant than at first appears. In many plays the chorus are ignored for long periods, and in some they are so closely identified with a principal figure that for him/her to speak to them of his dilemmas or discontent is almost like speaking to himself. The presence and absence of others on stage is not necessarily the decisive criterion: more important is whether the speaker addresses, interacts with, or seeks in any way to provoke a response from them. Even when characters do speak monologues, they are not necessarily delving into their own psyche. They more often look to the larger world, invoking the gods, the natural forces, significant features of the landscape (such as Mt Cithaeron, or the rocky tomb in which Antigone must die). Monologues often include prayers and curses (Ajax calling the Erinyes to punish the Atridae). Those who have been spurned by those who are present may call upon those who are far away or even upon the dead (lost parents, lost husband or wife, etc.). The range of emotions explored is wide: dread, misgiving, secret hope, long-enduring grief, malice and desire for revenge. Most memorable are the speeches in which a character vacillates between emotions: the long speech of Medea before the act of matricide is a high point in Greek drama, and here the presence of the chorus as silent, distraught and helpless witnesses adds greatly to the intensity of pathos and suspense.    

See ch. , pp. –. The classic treatments are by Leo  and Schadewaldt ; brief comments by Barrett on E. Hipp. . Note also Battezzato . Denied by Bain , Taplin –, but some softening of this position has been evident in more recent discussions, e.g. M. L. West on E. Or. – (p. ) and on , de Jong . See further Bain , Mastronarde .



Genre: form, structure and mode

An intermediate stage between spoken verse and song is occupied by recitative, but this involves only rather short portions of text and is generally less important. It most commonly occurs in passages in anapaestic metre, which is regularly used when the chorus first enter the stage, advancing along the eisodoi (side-passages leading into the acting space), before they begin their first song; the same metre is often used to introduce new arrivals, and it frequently recurs in the last  lines or so of the play, as an indication of imminent closure. It seems that these passages were normally chanted, and with some musical accompaniment. There is some testimony that the passages of actor-dialogue in trochaic tetrameters were also chanted and accompanied, though this is not accepted by all scholars or for all cases. We turn now to song. The majority of lyric utterances in tragedy, including most of the choral odes, are structured in pairs of stanzas which closely correspond metrically. The traditional terms for these stanzas are ‘strophe’ and ‘antistrophe’, meaning ‘turn’ and ‘counter-turn’; moderns assume that this refers to the original dancing movements of the chorus, but these cannot be recovered. Most choral odes have at least two strophic pairs, but in earlier tragedy the number can be much higher (the parodos of the Agamemnon, the longest lyric passage in extant tragedy, has seven pairs). Pindar and other earlier poets had composed their lyric songs in a different manner, using triads and preserving a consistent metrical pattern for each successive triad (AAB, AAB, AAB etc.). Triadic structure is occasionally found in tragedy (the third element is called an epode), but even when this pattern is followed, the sequence is non-recurrent; each new strophe is metrically distinct from the last, though some consistency of rhythmical type can usually be expected. Thus a normal choral ode will follow a pattern AA BB CC; the audience has less clear a sense of repetition and regularity than in Pindar. Connected with this metrical variation is the fact that individual stanzas are normally self-contained in terms of syntax and sense: that is, there is no overrun from one stanza to the next. Similar structures are used when there is lyric exchange involving actor and chorus, or actor and actor (or exceptionally three participants). An outstanding example is the Cassandra scene in the Agamemnon.   



DFA –. Possible further refinements are considered by Hall , – (= Hall , –). See the table in Griffith , , with the accompanying discussion. Epodes are found e.g. at A. Pers. –, PV –, S. Aj. –, E. Bacch. –. In Griffith’s table (last n.), any ode listed as having ‘/’ (or other fractional figures) strophic pairs may be assumed to have an epode. There are also occasional short stanzas known as ‘mesodes’, appearing between strophe and antistrophe: e.g. A. Supp. –, –. On the rare use of refrains see p. . Fraenkel on A. Ag., pp. –, and more generally Kranz –. The Cassandra scene is discussed in ch. , pp. –.

I Formal description



The metres of tragedy are bewilderingly varied, and are combined with extraordinary freedom. There are however certain clear tendencies. The most common metres in the other major genres, the epic hexameter and the elegiac couplet, are generally avoided (there is only one passage of elegiacs in the tragic corpus). Important in all the tragedians are verses based on the iambic, trochaic and dactylic metron. Dactylo-epitrite, favoured by Pindar, is sometimes used (never by Aeschylus, except in the doubtful PV ). The aeolic rhythms, which cannot be defined in terms of metra, are also important, especially in Euripides. Other species are found. One metre which appears to be the peculiar province of tragedy (and of comedy when it parodies tragedy) is the dochmiac, which is associated with great intensity of emotion, positive or negative: this is used extensively, for example, in the scene in which Cassandra raves prophetically in Aeschylus’ Agamemnon, and for the song the chorus of the Medea sing when they know that the mother is killing her children within the house (A. Ag. –, E. Med. –). The analysis of lyric metres, however complex, can be carried out to a high degree of exactness (though some uncertainties will always remain, partly due to textual corruption, partly because specific passages can be interpreted in different metrical terms). What is frustrating is that we can say so little about the effect or significance of the different metres. A few do seem to have a distinct character or particular associations. The anapaests accompanying entrances and some exits have already been mentioned (often referred to as ‘marching anapaests’, they are found in a Spartan soldiers’ song from the archaic period, PMG ). The ionic metre was associated with Asia and with cult songs, and may have conveyed eastern exoticism or enthusiasm. This would fit with its frequent use in the Bacchae; it is also common in Aeschylus’ Persae, set at the Persian court. The dochmiac, as already stated, is used as the vehicle of intense emotion in moments of crisis. Beyond this it is hard to go. Shifts of metre within stanzas or odes no doubt reflect some alteration of dance-step and musical 

  

Standard treatments are those of Dale  and West GM – (abridged as , ch. ). Wilamowitz b retains its classic status. See also Mastronarde on E. Med. –. A variety of metres are discussed by Herington , ch. . For a less cautious treament see the books by Scott , ; much more rigorous is L. P. E. Parker , ch. . For some comments on the significance of particular metrical effects, see L. Parker , – (on E. Her. –), Willink on E. Or. –, Allan on E. Hel. –. This occurs at E. Andr. –; for discussion see Page , Allan , –, –, –. Cf. Dale , –, esp. ; Conomis ; West GM –. Even here, however, not all scholars are confident about the overtones of the metre. Contrast Dodds’s Bacchae, p. – and West, GM , with the more cautious formulation of L. Parker , .



Genre: form, structure and mode

accompaniment, but how they affect the mood or the ethos of the song in question is largely a matter for speculation. We have spoken so far of strophic composition, but not all lyric song falls into that pattern. There are also ‘astrophic’ songs, whether by the chorus or by an individual actor, where no corresponding stanza follows. These were always in the minority, and there are relatively few cases in Aeschylus, but they become more prominent in the later part of the century. In Euripides astrophic composition becomes the norm for his famous monodies. The accompanying music to the songs was played on the aulos, an instrument variously rendered as a pipe or an oboe, a pair of which were played by a single accompanist, to judge by the vase-representations. Exceptionally, it appears, other instruments might be used if it suited a particular play or part. Orpheus in the Bassarids or Amphion in the Antiope may have played the lyre, accompanying their own song. We know that in Euripides’ Hypsipyle the heroine sang a lullaby to the infant in her care accompanying herself on the castanets (a curious choice); the scene is mocked in the competition of the Frogs. Were the chorus of the Bacchae accompanied by the drums and tambourines referred to in their entrance-song, or were these left to the imagination and evoked by the rhythms? A final point concerns the length of the dramas. No Greek tragedy exceeds  lines (the shortest is Rhesus, at , though it is highly probable that this play has lost a prologue). This contrasts with, for instance, Shakespearian drama (the shortest of Shakespeare’s tragedies, Macbeth, contains approximately  lines). The brevity of the plays is related to the structure of the dramatic festivals: at the City Dionysia a playwright normally put on a tetralogy, three tragedies and a satyr-play. But that only pushes the question further back: why did the tragedians prefer this format, rather than allowing themselves more scope for expansion and subplots? It seems clear that the genre favours concentration and intensity: the action is swift-moving, the emotions overwhelming. Aristotle’s comment on time in the tragic plot is relevant here: tragedy usually confines itself to events 

   

Listed by Rode in Bauformen – (a catalogue of strophic songs follows). In Aeschylus see Eum. –; also PV –. Whether the parodos of the Septem is partly astrophic (and if so, which parts) is disputed by editors. Ch. , pp. , –. DFA –; West AGM –;  n. ; Wilson ; Csapo-Slater –. E. F(f ). and context, Ar. Ran. –. For what it is worth, Senecan tragedy observes much the same limitations as the classical Greek plays: the spurious Herc. Oetaeus runs to a monstrous  lines, but otherwise longest is HF ().

II Historical outline



which take place within a single day. Aristotle nowhere comments on place, despite neoclassical pronouncements which supposedly follow his lead, but here too there is a strong tendency to restrict the action to a single location: in our surviving corpus, a single change of locale may be permitted in one play (as most notably in A. Eum. and S. Ajax), though more are alleged for certain lost plays. Intensity, tight focus, concentration on the crucial action and the moment of crisis: we shall see these priorities reflected on the level of tragic language. ii historical outline Thus far we have been describing in broad outline the shape and structure of Greek tragedy in general, as if there were a Platonic form of the genre. It is also necessary to consider the diachronic development of the drama. We may distinguish the ‘prehistory’ of the genre, its literary-historical origins in the late sixth century, and the period for which we have concrete evidence in the survival of complete dramas, a period extending from Aeschylus’ Persians in  to Sophocles’ OC posthumously produced in . (The Rhesus is probably later still.) I intend to say very little about the early evolution of tragedy. Here nothing is certain except that interpreters will never agree until all accept that the evidence does not permit confident assertion. It is customary to divide the critics into ritualists and literary historians, that is, into those who assume that the fundamental point about tragedy is that it evolves from a form of religious song in the context of a festival of Dionysus, and those who see the genre as more usefully explained through its subject matter, in terms of poetic inheritance, competitive performance, influence from epic and lyric poetry, and so forth. Aristotle’s Poetics is an essential source for both schools, though on the whole its emphasis falls on the latter side. The opposition is somewhat artificial, since each side must allow some importance to factors emphasised by the other; moreover, it 

 

See Arist. Poet. .b–.; Fraenkel on Ag. pp. –; Taplin –. There are exceptions, as already noted in ancient times (Dio Chrys. .), and choral odes often signify an unspecific interval of time, but the general tendency is as Aristotle describes. In particular Aeschylus’ Aetnaeae: see TrGF iii. –, citing P.Oxy. , fr. .–. According to this testimony the Aetnaeae involved multiple changes of scene. On the problem see Taplin –. I know of no saner short account than Winnington-Ingram . For fuller statements of the evidence see Lesky , chs. – (cf. Lesky , chs. –). For some specific problems see Lloyd-Jones  (more easily found in Lloyd-Jones , i. –). See further Else , Herington , and the strongly contrasting positions of Scullion  and Sourvinou-Inwood . A recent collection of papers, Csapo and Miller , offers up-to-date debate and some broader perspectives.



Genre: form, structure and mode

is not as though one side has the monopoly on the evidence, for both ritualists and literary historians are struggling to construct a jigsaw most parts of which are missing. Most scholars would agree that the origins of tragedy as an institutionalised form lie in the sixth century, though it has for some time been clear that the chronology is far less certain than is traditionally claimed. Tragedy may have been part of the extensive literary patronage by Pisistratus and his sons, but a case has been made for placing the first productions later, after the fall of the tyranny in . That the poet Thespis did produce some form of drama may still be accepted, but we cannot be sure that he did so in the Olympiad /, as most handbooks claim on the basis of epigraphic evidence; and although titles are preserved under his name (Games for Pelias, Priests, The Young Heroes and Pentheus are among them), the possibility that these are later fabrications cannot be ruled out. Aristotle maintained that tragedy had its origin in some form of dithyramb, but also connected it in some way to satyr-poetry: these remarks, though not wholly irreconcilable, are certainly in tension, and efforts to unite them result in highly speculative reconstructions. In the same obscure chapter he comments that the proper tragic grandeur was a later stage in the genre’s development: ‘after a period of slight plots and laughable diction, owing to development from a satyric ethos, it was at a late stage that tragedy acquired dignity, and its metre became the iambic trimeter instead of the trochaic tetrameter’ (Poet. .a–). If this judgement rests on first-hand knowledge of texts prior to Aeschylus, it is an important piece of evidence, but we may well doubt it. Aristotle’s interest even in Aeschylus is more limited than in his successors, and he cites no tragedian known to be earlier than Aeschylus. It is more important to stress that tragedy, as we said at the start, is a hybrid form, and that this combination of elements must have been a deliberate innovation by one poet or a group, whether we ascribe this to Thespis or some unknown. It is easy to lose oneself among a host of hypotheses. Whether we say that tragedy was initially choral and that an actor’s rhesis was added, or that tragedy came into being at the point when that addition took place is a question of definition, reminiscent of the debate on where Homer came in the oral tradition. Dithyramb is choral song, not drama, though it may involve quasi-dramatic episodes and mimesis of direct speech  

Connor ; for discussion of the uncertainties see West , Parker , –. The Parian Marble, an inscription recording chronology of early Athens: see Thespis T  Snell = FGrH A; but note Connor , – on the completely unreliable reading here.

II Historical outline



(as notably in Bacchylides  and ). What makes tragedy different is the combination of spoken iambic verse (familiar at Athens in a political and exhortatory context from the poetry of Solon) and lyric song (normally self-subsistent and associated with particular occasions – religious worship, weddings and funerals, social gatherings, sympotic celebrations, and so forth), and the use of this extraordinary mixed form to sustain a dramatic narrative. The consequence of the creation of this new form was that an audience could experience a diversity of pleasures from a single genre. Tragedy embraced mellifluous song, articulate argument, gnomic maxims, plangent emotion and hymns to the gods. The division into actors and chorus gave the opportunity for both solo and united song (other genres normally had to choose between them). The combination of male and female roles adds further to the rich mixture. Other genres were, if not fully absorbed, certainly echoed and emulated (section III). The compression of the form meant that these different effects were achieved within extraordinarily brief compass. Tragedy possessed both extensive range and unparalleled intensity. Although in some ways we step on to solid ground with the period from  onwards, to chart the genre’s development even in this period inevitably involves some fragile arguments, when we have so few plays compared with the vast number produced over three quarters of a century. Apart from a few stray details, it is not possible to put much flesh on any dramatist other than the big three. It is easy to take note of features which characterise the later plays of Euripides but unsafe to assume that these were mirrored in the works of his contemporaries. Even the chronological relationship of the plays we do have is often uncertain. The following account attempts to single out a few areas where we can speak with some confidence. First, it is obvious that the prominence of the chorus in earlier tragedy is not maintained: the actors gain pre-eminence. Evidence for the original form of Greek tragedy as developed in the sixth century is inadequate. We can neither confirm nor deny the theory that early tragedy was more concerned with the myths and worship of the god Dionysus. But it is at least consistent with some of Aristotle’s testimony to suppose that the genre had its origin in some form of celebratory cult song performed by a chorus. The choral element of tragedy long remained the most  

To stress this crucial point was the important contribution of Herington , following Else . On the emotional energy and intensity of the tragic genre see Hutchinson , –; Schauer .



Genre: form, structure and mode

important and prestigious. A crucial step was taken when the single actor was introduced, a separate figure from the chorus. Tradition, perhaps going back to Aristotle, ascribed this innovation to Thespis in the sixth century. More concrete is the explicit statement by Aristotle that Aeschylus introduced the second actor, Sophocles the third (Poet. . a–). When our evidence begins, the chorus and its songs are the most substantial and the most elaborate part of tragedy. In Aeschylus (excluding the PV) the choral contribution never occupies less than  per cent of the whole; in Sophocles, never more than  per cent. To a considerable degree the history of tragedy in the classical period is the story of the increasing dominance of the actors. Episodes between choral odes increase in length (one episode in the Helen lasts nearly  lines: Hel. –). In some of the later plays of Euripides in particular, the choral presence is marginal; the tragedian Agathon, who won his first victory in , is said to have introduced the practice of ‘interlude songs’, i.e. choral songs which were little more than intervals, with no necessary connection with the drama as a whole. Because of the convention that actors were masked, there were many changes of role and costume in mid-play. There is some tendency for the number of characters to multiply: thus the Persians employs only two actors and has only four characters (Atossa, a messenger, the ghost of Darius, and Xerxes); Euripides’ Medea has seven adults and two children with speaking roles; the same poet’s Orestes has ten characters other than the chorus, his Phoenissae eleven. There seems also to be a tendency for plays to grow longer as the century progresses. Both developments, together with the reduction of the chorus’s role, make for greater complexity of plot. We should also note the changing fashion as regards connected trilogies of plays. Each tragedian at the City Dionysia was obliged to present three tragedies and a satyr play: the three tragedies might be but need not be linked in a sequence. Aeschylus favoured the trilogy form, and from his hand comes the only complete trilogy to survive, the Oresteia. Sophocles and Euripides seem to have preferred to put on self-contained plays: of the extant corpus, only Euripides’ Trojan Women looks like part of a triad. This confinement of the action to a single drama (usually to a single brief   



Themistius . d (= TrGF i. F = Else , Appendix T ).  Arist. Poet. . a. Griffith , . The figure for PV is  %. On Aeschylean tetralogies or trilogies see further Gantz  and , Sommerstein , ch. . See more generally TrGF i. –, DFA –. M. Wright , – argues that Helen, Andromeda and IT may have formed a Euripidean escape-trilogy. With Alexandros and Palamedes in . For attempts at reconstruction see Scodel , and Barlow’s introduction to her edition of the Trojan Women.

II Historical outline



span of time) has important consequences for the audience’s perspective on the events dramatised. As the actors become more numerous, their relationships on stage become more flexible and are more skilfully handled. In Aeschylus, especially the earlier plays which employ only two actors, there is a certain formality in the actor-dialogue: an actor enters, speaks to the chorus or to another actor, departs and is replaced by another. Three-way dialogue is rare. When the messenger in the Persians arrives with the news of the disaster which has befallen Xerxes’ army (), he speaks first to the chorus, who respond in a series of strophic cries of distress; he then turns to the queen and engages in extended dialogue with her (), and the chorus play no further part until he has left the stage (). The same formality prevails in the earlier plays of Sophocles, but in the OT, at least, the dialogue is more fluid: in one scene Jocasta, Oedipus, Creon and the chorus are all involved in the dispute between the king and his brother-in-law (–); in another Oedipus’ questioning of the aged shepherd is backed up by the Corinthian (a telling combination, as between them these two men hold the key to Oedipus’ true identity) (–). Nevertheless, although techniques clearly become more sophisticated, there is still a strong tendency for dialogue to settle for extended periods into one-to-one exchange. This might be variously explained: on the one hand, it ensures that the audience in a large theatre could follow easily, on the other, it satisfies the evident desire of poets and audience for opposition and antagonism. There are other ways in which the structure of the dramas can be seen as ‘loosening up’ as the century goes on. The most objectively verifiable is in metrical practice in the handling of the trimeter. Long syllables in the tragic iambic trimeter may be resolved into two shorts, but the restrictions on this are more severe than in comedy. Aeschylus indeed became more severe as he grew older: in Persae resolution occurs in . per cent of the trimeters (excluding proper names), in the Oresteia . per cent. Sophocles’ figures do not show any obvious trend, except that the Philoctetes shows a higher frequency than the rest. Euripides’ willingness to use resolution rises markedly from Hippolytus onwards, reaching its peak in Orestes (. per cent). Aeschylus is also stricter than his successors in avoiding enjambement: in each of the authentic plays the number of lines endstopped by a pause or punctuation of some kind greatly outweigh those which run on into the next line (in Agamemnon only  of  trimeters are not



Cf. further Taplin – on the dramaturgy of Supp. –.



West GM .



Genre: form, structure and mode

endstopped). Sophocles and Euripides are readier to allow sentences to overrun the verse; Sophocles in particular seems to mimic the natural flow of speech. Another objective feature is the increased use of antilabe, division of a dialogue line between two or more speakers. This is not found in Aeschylus (there is a possible instance at PV , if the text is sound), but it figures in almost all the plays of Sophocles and Euripides, and is used with particular freedom in late Euripides (not least in trochaic passages). It renders dialogue less stately and more agitated: the technique is well suited to scenes of excitement, in which one speaker is repeatedly capping, countering or following up the ideas of another (esp. E. Phoen. –, Or. –, IA –). In Aeschylus the rich texture of the language is most evident in the formidably dense and difficult lyric passages, but the linguistic intricacy of the dialogue portions falls not far short of this level (‘Euripides’ in the Frogs finds obscurities and ambiguities in both). In his successors it is probably safe to say that dialogue becomes somewhat simpler in vocabulary and syntax, whereas sung portions remain on a higher plain. Euripides was praised for his ‘lucidity’. Related to this is the important development whereby tragedy increasingly drew on the resources of formal rhetorical prose. In the agon and elsewhere we can discern a more cerebral and intellectual element in the tragic repertoire. Aeschylus’ Pelasgus had declared that the people of Argos must decide; Euripides’ Theseus spends time expounding the theoretical justification of democracy. The dilemmas of Aeschylus’ characters are dramatised in image-laden lyric and deeply felt ruminations (Supp. –, Ag. –); the soul-searchings of Medea and Phaedra have a sharper intellectual edge (Med. –, Hipp. –). A novelty, as far as we can see, in Euripides is the use of child-actors who speak and even sing short parts – the sons of Medea, the son of Admetus and Alcestis. If this is a Euripidean innovation, it is in line with the general 



 

Griffith , –; for a useful note on Euripidean practice see Collard on E. Supp. –a, and for some general remarks see Mastronarde on Medea, pp. –. The fullest recent discussion of enjambement and emphasis is by Dik , ch. , esp. – on cases in Ajax. A valuable survey of this device is K¨ohler . For some notable Sophoclean examples see Ajax – (end of a scene), OT –, Phil.  (the only case in tragedy involving four utterances in one trimeter), OC –, –. It is most frequent in his late plays: in Ajax there are three passages involving one or more cases of antilabe, in Trach. only two; in Antigone it does not occur at all. GGL i. .–; cf. praise of clarity and straightforwardness in the actual plays, ibid. n., e.g. at Phoen. , F. For the legacy of Medea’s speech to later ancient debate on reason vs. passion, the divided soul, and so forth, see Mastronarde on Medea –, and appendix; also Gill .

II Historical outline



broadening or ‘democratisation’ of the genre of which he is made to boast in the Frogs (–). There are also new tendencies as the century advances in the handling of song. For the most part refrains, reminiscent of religious ritual, seem to disappear (though they are significantly revived in the Bacchae). The more strongly hymnic or ritualistic odes which we associate with Aeschylus are less common, and less protracted, in his successors. Choral odes become shorter: the norm in Sophocles and Euripides is two strophic pairs. After Aeschylus choral openings to dramas are virtually unknown: instead an actor or more than one actor takes the stage at the start. In Aeschylus, whenever the chorus enters, it dominates the stage with its initial song, but in later drama amoebaic exchanges with an actor are common even in the parodos. In several cases an actor has already ‘upstaged’ the chorus with anapaestic or lyric utterance prior to the parodos (most conspicuously in S. El. and E. Ion). More important still is the increased prominence of the monody or solo aria by an actor in Euripides’ work. Actors, as we have seen, already sing in his predecessors, but the extended operatic technique which we find in Euripides and especially in his later plays is something more spectacular, a tour de force of voice, music and metre. Their impact on the audience is shown by the references and extended parody in Aristophanes’ Frogs. Quite apart from the special case of monody, song was moving into different territory by around . The plays of Euripides in the latter part of his career include a number of choral odes in a more consciously ‘poetic’ and ornate style. These songs are often less closely tied to the action than choral song normally seems to be in Aeschylus and Sophocles; they are also more loosely structured, and composed in a more emotional style. Metrically too they are often more varied and harder to characterise than earlier choral lyric. It is hard to dissociate these developments from evidence (especially in comedy) for new developments in other genres, dithyramb 

  



Golden  discusses ‘baby-talk’ in ancient Greece; note Alc.  discussed on p. . See also Headlam-Knox on Herod. ., O. Thomas . For other passages highlighting children see E. Her. –, –, –, Or. –, GGL i .n.. Zeitlin  considers Euripides’ interest in children and childbearing. An actor-prologue is present in all extant plays of Sophocles and Euripides. In the Rhesus a prologue may well have been lost. On prologues see further ch. , pp. –. So already in PV; cf. Sophocles’ later plays, El., Phil. and esp. OC; E. Med., Heracl., Hec., El., Tro., IT, Ion, Helen, Or. Owen , –; W. Barner in Bauformen –; Collard on E. Supp. –; Beverley . See e.g. PV –, – (Io), E. Hec. – and –, Tro. –, Ion –, Or. –. See further ch. , pp. – below. Kranz –; Panagl ; Mastronarde on E. Phoen., pp. –.



Genre: form, structure and mode

included, developments normally referred to by the convenient label ‘the New Music’: much is obscure, but it would seem that performers were now allowing themselves a wider range of notes and sharper changes between musical modes and styles. At one stage it was common for critics to claim that Euripides took over the innovations of the lyric poet Timotheus, some of whose oeuvre survives; more recent accounts persuasively argue that Euripides himself is the chief innovator. The conventions of the genre are now handled in a rather more selfconscious and sophisticated way (unless this is an impression produced by the accident that we have more plays from the later part of the period). Established generic tendencies can be alluded to by the characters, creating a metatheatrical effect. In Euripides’ Electra, conscious that bad news is impending, the heroine asks ‘where are the messengers?’ (), the chorus assures her that they will swiftly arrive (‘for it is no small deed to slay a king’), and immediately a messenger appears. In the Orestes a plot is being hatched to kill Helen and Hermione, when Orestes belatedly realises that the presence of the chorus may be a security risk; Electra immediately reassures him that ‘they are our friends’ (). The assumption that the chorus keep secrets is so well-established that the matter is disposed of in two lines (earlier in the century they would no doubt have been asked to swear an oath, as e.g. in Medea and Ion). Even more perfunctory is the equivalent moment in the IA (). But the most obvious convention in late Euripides is the ‘deus ex machina’ scene which ends more than half of his plays. This also seems to have influenced Sophocles, who used it to very similar effect in the late Philoctetes. That this was a ‘stock’ ending by the close of the century is strongly suggested by the way in which the phrase ‘a god from the machine [crane]’ is evidently becoming a proverb. The Rhesus allows us a glimpse of further post-Euripidean development: in the fifth century the god or goddess spoke, but in the finale of the Rhesus the bereaved Muse sings, in suitable lamentation for the dead son whose corpse she cradles in her arms. 

     

See esp. West, AGM –; Csapo –. Murray and Wilson  is an important collection of papers: again see esp. Csapo’s contribution, ‘The politics of the New Music’, pp. –. See ch. , pp. – below. The older view e.g. in Webster , –; revised view Csapo –. On Timotheus see Herington , ch. ; West , –; Hordern . For more detail see appendix to ch. , pp. –. Barrett on E. Hipp. . For discussion of its significance see Schmidt , Mastronarde , –. On evidence for divine epiphanies in the lost plays of Sophocles see Parker , –. Plato, Crat. d, Clit. a, cf. Antiphanes F (=Olson , D), Dem. .. For a skilful sketch of other fourth-century developments see Hall ; more detail in XanthankisKaramanos .

III Generic appropriation and distortion



Throughout the fifth century it is conventional for the chorus to have the last word, usually making some gnomic pronouncement, but sometimes contributing something more substantial in song and thought. In Aeschylus’ Agamemnon Clytemnestra has the final say, but this is telling: the final section of the play has dramatised the reluctant submission of the chorus to the will of the usurpers. In the PV the chorus are apparently engulfed by the cataclysm along with Prometheus, but the intransigent hero must be given the last defiant words in the play; the Oceanids’ fate is secondary. Other cases where an actor may have the last word are controversial and textually questionable (S. Trach, OT). According to our manuscripts several plays of Euripides end with conventional choral tailpieces, some even repeated from play to play. Many scholars doubt the authenticity of these repeated passages; whatever the truth about the more conventional cases, Euripides is unlikely to have used the closing lines to voice a hope for victory in the contest, a plain breach of the dramatic illusion. In these bows to the audience we probably witness a later convention intruding on the older, classic texts (compare the concluding plaudite in Latin comedy). iii generic appropriation and distortion We have already seen that tragedy is a genre that imports freely from other sources, producing a complex compound. Further examples and fresh distinctions are needed. First and foremost comes the language of religious ritual. Whether tragedy is a ritual or originated in ritual remains controversial; that it dramatises many such rituals, some of which had crystallised in or been celebrated through particular literary genres, is beyond question. Invocations, prayers and full-bodied hymns occur regularly in the dramas, whether in spoken or sung form. In the parodos of the Oedipus Tyrannus the chorus entreat the gods to drive away the plague that afflicts the people of Thebes (an apotropaic hymn), in that of the Antigone they sing a song of celebration for the salvation of Thebes from foreign attack. Such ritual songs are of special importance in Aeschylus. Conspicuous among his many examples are the blessing-songs which figure in his Suppliants (–, in which the Danaids pray for the wellbeing of the city of Argos) 

 

As found in the manuscripts of Phoen. (–), Or., IA. The close of Rhesus is cleverer, expressing within the dramatic framework the chorus’s hope for victory on the new day of battle now dawning, but also permitting a double meaning if we think of the aspirations of the chorus extra personam. See esp. Easterling b, Lloyd-Jones . Griffith , –, with bibl.; see esp. Kranz –, Schadewaldt ; Fraenkel , i. – and on Ag., pp. –; Citti . For lamentation in particular see Porzig , ff.



Genre: form, structure and mode

and his Eumenides (–, where the Erinyes, finally appeased, bless the city of Athens). That these prayers owe something to actual cultic song is not strictly demonstrable, but hard to doubt. The repeated cries of the Bacchants ( B, ‘to the mountain!’) also probably echo authentic ritual exhortation. Naturally, some of the ritual situations in tragedy are remote from common experience. The second of the blessing-songs just mentioned is sung by divinities, not mortal worshippers. Few members of the Athenian audience would have experienced a ritual summons of a murdered father, as dramatised in the Choephori; none would have knowledge of the proper necromantic rites involved in the summoning of the ghost of a dead Persian monarch. Yet these events are given powerful ritual form and expressed in memorable but macabre language. In some cases the tragic version of a ritual presents something that might be familiar in a simpler or less horrific form to fifth-century Greeks. The terrifying ‘binding hymn’ of the Furies in Eumenides can be associated with curse magic of a kind documented for the historical period; the metaphor ‘I bind X with this spell’ is a frequent one. But familiar rituals are exaggerated and distorted to the point of perversion when tragedy treats of human sacrifice, with the human victim replacing the conventional beast, as when Aegisthus is savagely struck down in Euripides’ Electra or Polyxena slain to satisfy the vengeful ghost of Achilles in the Hecuba. Similarly, when in the Bacchae maenadic ecstasy is carried into homicidal madness, when Agave and her sisters tear her son limb from limb, when the hideous spectre of cannibalism (U(1) seems imminent (), we have moved far from the everyday world of the Athenians; yet the parallels and allusions to normal religious practice give these scenes a deeper purchase on the imagination.    





Solmsen , –; Furley and Bremer , i. –; caution in Johansen-Whittle on A. Supp., vol. iii, p. . E. Bacch. , ; see Seaford’s commentary, p. ; Henrichs , at –. On the debate as to whether the necromantic procedures in Aeschylus’ Persians draw on genuine knowledge of Persian religion, see the commentaries by Hall and Garvie on –. Faraone  on Eum. , and the song that follows, –, comparing Pl. Leg. a, and the so-called defixiones (curse-tablets); for these texts see further the works cited in Faraone n., and add Gager , Parker , ch.  (citing Eumenides on p. ). Similarly we can see signs of tension between generations in the fifth- and fourth-century evidence, for instance in oratory, and cursing of relatives may have occurred: ‘a parent can curse a child more effectively than anyone else’, comments Plato (Leg. b–c). But the potent cursing of his sons by Oedipus goes beyond normal experience. On curses see further Parker , –, Watson . On human sacrifice in Greek culture see Henrichs  and Hughes ; on its importance in the world of Greek tragedy as a ritual perversion, see Burkert , Zeitlin ; on the Bacchae, Seidensticker , Diggle , –; on sacrificial metaphor see Henrichs . For the general principle see Henrichs , at : ‘Greek ritual tends to mitigate where myth is cruel.’ Cf. Henrichs .

III Generic appropriation and distortion



One marker which suggests affinity with genuine cult song is the refrain. These are used freely by Aeschylus, not at all by Sophocles, and by Euripides in only three plays, and in significant contexts: as part of the hymn of Ion to Apollo, and in two choruses from the worshippers of Dionysus in the Bacchae. (A possible third case, the water-carrying song in Electra, has characteristics of a goos or dirge, a genre prone to repetition and reiteration, but the repetition here may also suggest the obsessiveness of Electra’s personality.) When we speak of tragedy exploiting other literary genres, various effects can be meant. Most obvious is the imitation of specific models, above all Homer (and no doubt the Cyclic epics, less accessible though these are to our enquiries). Quite apart from the general importance of the Trojan myths, specific scenes and episodes in the Iliad and the Odyssey are commonly exploited. The messenger speech in Sophocles’ Electra describing Orestes’ chariot race includes many reminiscences of the chariot race in Iliad , and the Iliadic Teichoskopia is clearly the model for the scene in E. Phoenissae in which the old paedagogus instructs Antigone in the identity of the warriors on the battlefield below Thebes. (Euripides reverses the roles: in Iliad  Helen answered old Priam’s questions; in Euripides the old man is the one who knows the answers.) On a lower level, there are many places where epic lines or phrases, memorable speeches or similes are recalled; there is also much epic vocabulary. Some comparable reminiscences of particular passages in lyric are also detectable: if we had more lyric, we would no doubt find others. We may also note passages in which the tragedian seems to echo no specific passage known to us, but nevertheless evokes the manner, the atmosphere, even the ethos of a different genre. It is perhaps reasonable to distinguish echoes of verse from those adapting prose: the former are doubtless less striking, given that poets have always felt free to absorb, appropriate or otherwise exploit their predecessors in verse. Homer already  





See further Fraenkel on A. Ag. , f.; Johansen and Whittle on A. Supp. –; West, GM ; Moritz . The subject has been addressed in detail by Swift , whose treatment will now be the first reference point (note esp. her appendix of linguistic markers identifying generic influence in particular passages). The account given here is only a sketch of the topic, and Swift’s discussion should be consulted on all examples. Sideras , Garner . Garvie , – notes potential problems in distinguishing epic borrowings in vocabulary from words common to the poetic diction more generally; for other problems see Davidson . E.g. S. Ant. –, echoing Pind. Paean  (F k = A Rutherford). S. Trach. – is strongly reminiscent of the style and mood of Sappho (cf. F  and Catullus , esp. –). For a more detailed study see Bagordo .



Genre: form, structure and mode

makes passing reference to other types of poetry – the paean, the wedding song, the lament for Linus. Tragedy, however, seems to go further, perhaps in the wake of some lyric poetry. The dramatists sometimes allude to, mimic or even parody other genres in order to undermine or distort them. The ironic handling of the wedding song in the Phaethon is a good example: this song is performed by the chorus as they celebrate Phaethon’s impending marriage (itself probably a Euripidean invention), but the audience already know that the boy has met his end in the cataclysm of the sun’s chariot. It is clearly not the case, then, that tragedy simply incorporates other types of poem in its hospitable embrace, though this may sometimes seem almost true (the cult hymn in the final stasimon of Antigone, for example, –, makes almost no direct reference to the framing drama or to the circumstances in which Dionysus is being invoked: only at f. is there a somewhat vague allusion). In general, however, the model genre is selectively handled or more seriously darkened, even subverted. A paean or an encomium looks different when assimilated to the tragic world. The subject has been worked out in detail for the paean in Ian Rutherford’s penetrating discussion. As he shows, there are many references to the paean in tragedy, some of which even use the characteristic form of invocation of Apollo, ie paian; but many of them allude to the genre through a distortion of its normal usage (e.g. E. Alc. –). Apollo is an Olympian deity, but tragedy often uses the term in chthonic contexts. Apollo is normally an ‘ephebic’ divinity, worshipped especially by young men; but in tragedy the paean is often associated with women. In the Choephori Electra orders the (female) chorus to sing a paean over the tomb of her father (–). In the IA Iphigenia tells the chorus to sing one while she is being sacrificed, a ghoulish transferral of the sacrificial paean from an animal to a human victim (–).  

 

 

Hom. Il. .–, ., . For another case, see Cassandra’s wedding song in E. Tro. (ch. , pp. – below). Cf. the various odes which celebrate some form of success, positive development or good news prematurely, after which a darker outcome follows: e.g. S. Aj. –; cf. Henrichs –,  n. . For another case see the subtle analysis of S. Trach. – as a prosphonetikon by Stinton , –. Phaethon F. –, with SFP i. –, –; see also Furley and Bremer , i. –, ii. –. Dale ,  made the basic point; more fully Furley and Bremer , i. –, esp.  ‘In a way, then, a dramatic hymn is an impostor: it is not a genuine cult hymn but rather an imitation of a cult hymn written for the fictional situation and performers of the play.’ I. Rutherford , –. See also Swift , –. But Parker , –, advises great caution in use of the Olympian-chthonian opposition mentioned here. I. Rutherford , .

III Generic appropriation and distortion



Two passages from Aeschylus’ Agamemnon make still clearer the generic allusiveness which gives tragedy a further ferocious edge. The first again concerns the paean. Iphigenia has been gagged and is being man-handled to the altar: ' 1   "',,  V, '3 + B((  1 , ', <   1, , 

',,  )  !  +6  0' (, J  +.   ! 1'  , (  6 1 (. (Ag. –) And with her robe of saffron dye streaming downwards She shot each of her sacrificers With a piteous dart from her eye, Standing out as from a picture, wishing To address each by name since often In her father’s hospitable halls She had sung, and virginal with pure voice Had lovingly honoured the paean Of felicity at the third libation Of her loving father. (tr. Lloyd-Jones)

The contrast is between Iphigenia free, loving and loved by her father, and able to sing a propitious response to a proper ritual, and Iphigenia gagged and helpless, herself the victim of a horrific ritual ordered however reluctantly by her father. The comfort and hospitality of a symposium in Agamemnon’s halls contrast with the atrocity being performed at Aulis. The paean plays its part in evoking what was positive and normal, in contrast with the negative and ominous present. The second passage concerns the arrival of Helen in Troy and the deluded celebrations that surround her ‘marriage’ to the adulterer Paris: again there is a contrast between past felicity and subsequent misfortune. A supernatural force, a divine Wrath (O ), has brought retribution.   

I. Rutherford , . Sympotic celebration is typically negated in tragedy, cf. e.g. S. Ajax –, and esp. E. Phoen. – (p.  below). Also note the adjective translated ‘virginal’, +., literally ‘not subjected to the bull’, suggesting the animal role taken by Iphigenia (and a sense of violation in the present circumstances). The word is plainly an Aeschylean coinage (elsewhere only Ar. Lys. –).



Genre: form, structure and mode ,,( ! '(1 ( ( 1) , C( , :   (,  + . (( )',  W( / )('     . E ( ' ,  3,', /)  ! & (Ag. –) It exacted vengeance from those who loudly Celebrated the song in the bride’s honour, The song which then fell to the bridegroom’s kin to sing; And learning a different tune Priam’s aged city, A tune of many sorrows, Loudly, I think, laments, calling upon Paris of the disastrous marriage

Here the genre alluded to is of course the marriage-song; but just as the wedding is one achieved by treachery and abduction, so the joys of the celebration must modulate into a different tune, a lamentation or dirge. A further example can be seen in Euripides’ Electra. There the heroine is isolated and determined to remain so. The chorus visit her on the undistinguished farm where she lives, the resentful wife of a worthy smallholder. They remind her of the impending festival of Hera, goddess of their native Argos, and encourage her to attend with them: ‘the maidens, one and all, will go in procession to Hera’. But Electra will have none of it.    +X , 1 ,

'(!    "',  4(  ( ) ,   Y,, "S ;  Z( .(1  Y ! .,  (. )', '"., .  ( (   !  N(. , ('   (  ."E ) (6 , &  ;(( .  ,  . (El. –)



See further Ag. –. Other relevant uses of musical and generic terminology are considered in Haldane , Wilson and Taplin .

III Generic appropriation and distortion



No fineries, my friends, No golden necklaces Give flight to my wretched Heart; nor setting dances Along with the brides of Argos Shall I pound out my whirling step. In tears I spend my nights, tears are my sorrowful Care day after day. Observe my sordid hair, This filthy clothing of mine; See if they are suitable for Agamemnon’s Royal child. (tr. M. Cropp)

Here the ‘foil’ to the tragic sequence of action is the Argive festival with all its song, dance and splendour of dress; the generic contrast is with maiden-songs such as Alcman’s choirs sang in Sparta. More broadly, the normal, positive rituals to which the chorus invite Electra contrast with the perversions of hospitality and sacrifice later in the play when Orestes and Pylades kill Aegisthus by the altar. The pattern, as we see, is for the more positive or constructive genre to become more disturbing or acquire negative overtones when it enters the orbit of tragedy. Some more positive examples might be found, but they are rare. The hymn of welcome and celebration by the Athenians escorting the Eumenides to their new home is perhaps the best instance. Dirges, formal lamentations by chorus or individual or both in unison, are at home in tragedy: the majority of plays contain examples, and some are dominated by these passionate songs of lament. Here there is little need to adapt the material to its tragic setting. In this case a different point needs to be stressed: tragedy appears to enhance the intensity of expressions of grief well beyond what would be tolerated in contemporary Greek communities. Legislation limited not only the expenditure on funeral celebrations but the outspokenness with which lamentation was expressed:   

 



See Calame , – (Eng. tr. , –). Zeitlin ; more generally on sacrifice perverted see Burkert . Another example is the epinikion: see e.g. the song of the chorus at E. El. -, –, acclaiming the success of Orestes in slaying Aegisthus – a celebration which seems at least off-key (Rehm , –). See also Rutherford a, –, esp. n.. See further Easterling b, who examines ways in which rituals (rather than literary forms) are adapted and significantly modified in tragedy. For a catalogue of major examples see Broadhead on A. Pers., pp. –; see also Collard on E. Supp. –, –; Alexiou ; more recent studies include Holst-Warhaft , Du´e , Suter a, Klinck , chs.  and , Swift , –. For data on Greek funerals see Kurtz and Boardman , Garland . On the substantive point see esp. Foley , ch.  (orig. ); see also Griffith , n..



Genre: form, structure and mode

the justification from social decorum and (in time of war) military morale is easy to discern. But in tragedy we find little inclination towards restraint, whether in public or private contexts. The violence of the emotions is mirrored in the actions of the mourners, frequently described in their songs – ripping of clothes, tearing of hair, beating and wounding of faces and breasts (e.g. E. Supp. –, –). Onlookers or advisers may advocate self-discipline and remind the mourner that ‘you are not the only one to have suffered’, but such sane counsel seldom has any effect. In a number of plays the lack of restraint is highlighted as a disturbing issue. Eteocles in Aeschylus’ Septem upbraids the female chorus for panicking and damaging morale by anticipating disaster before it is certain (–). If we turn to the debt owed by tragedy to prose literature, the most significant area is forensic oratory, which has already been mentioned in the first chapter. Speech and argument are important in all tragedy, as they had been in Homer. What is new in the fifth century is the significance of formal debate involving the whole citizen body, in the context of the Athenian radical democracy. Oratory became not only a necessary part of the deliberative process but a means of political self-promotion. The lawcourts had political significance as well as the assembly, for legal disputes often had political overtones; since Athenian juries were often very large, many citizens would become familiar with the typical arguments and deceptive moves of the courts. In the fifth century the Athenians were regularly seen as addicted to litigation. In the fourth, Athens could be described as a city of words. It was in the mid-fifth century that the sophists emerged, a group of intellectuals with widely varying interests and diverse origins, many of whom made a speciality of teaching the art of rhetoric: Protagoras, Gorgias, Hippias and Prodicus were the most significant, all of them familiar from the hostile presentation in various dialogues of Plato. There is relatively little sign of their influence in Athens prior to the s; Aeschylus is untouched by it. This does not mean that oratory and discussion are not prominent in his work; we need only think of the      

E.g. Plut. Solon . (Athens); LSAM  (Gambreia), LSCG  (Ceos). See further Seaford , –. Alexiou ,  n. ; see also the references in Denniston and Cropp on E. El. –. Examples at S. El. , E. Alc. , Or. –. The motif goes back to Homer (Macleod on Iliad .–). [Xen.] Ath. Pol. ., Ar. Nu. –, Vesp. passim, Thuc...; Bers , –. Dem. ., a phrase adapted as a chapter-title in Goldhill . Guthrie HGP iii, Kerferd a, b, Classen , CAH v () –, Rutherford , ch. .

III Generic appropriation and distortion



assembly described in the Suppliants (p.  below) and the trial scene in the Eumenides. Peitho, the personification of persuasion, is a potent figure in the thought-world of the Oresteia: malignant in the Agamemnon (), ambiguous in the Choephori (–), benevolent in the Eumenides (–, –). Although Aeschylus’ speakers are far from unsophisticated, there is a sense in which the rhetorical ‘turn’ becomes more evident in Sophocles and especially Euripides: the argumentation is more incisive and even legalistic. In both ancient and modern times the border areas between tragic verse and rhetorical prose have been explored. Plato already branded tragedy a form of demegoria, ‘public speaking’ (Grg. b–d). There are close affinities between the assembly and the stage, especially given the importance of public affairs in drama: mythical Argos as much as historical Athens must determine through public discussion whether to go to war or accept refugees from another state. The lawcourts also offer analogies. Euripides in particular favours the presentation of conflict through extended speeches in debate (the agon); the sophist Gorgias’ phrase ‘contest of words’ is echoed in the tragedian’s text. Standard rhetorical moves can be paralleled: the argument from probability, the generalisations on human nature, the systematic anticipation of an opponent’s argument. Rhetorical theory did not confine itself to structures and argumentation, fact and probability: the importance of emotional appeals also forms part of the teaching of the sophists and their successors. Particularly prominent is the appeal to pity and to anger (pity for the victim/defendant, anger at the outrages of their opponents and oppressors). The relevance to conflicts in drama is obvious. Tragic myth and prose rhetoric interact in many ways. Characters from epic and tragic myth can be used as spokesmen in model speeches (Gorgias’ Palamedes, Antisthenes’ Ajax and Odysseus), and the same situations are exploited on stage. Anachronistic material may be included in drama, including questions prominent in fifth-century moral and political, even philosophic, discussions: constitutional theory (E. Supp. –), the 



 

See esp. Buxton . For standard accounts of classical rhetoric see Kennedy , Worthington . For an excellent treatment of Euripidean rhetoric see now Mastronarde , ch.  (esp. – for the context and for what is new in Euripides). Arist. Poet. .b– comments that ‘the old poets (Y +" ) represented men as speaking in statesmanlike fashion (  6), but those of the present day in rhetorical fashion (KE 6) , ('. (Ant. –)

Yes, for it was not Zeus who made these proclamations, nor Justice, who shares her dwelling with the gods below, that ordained such laws as these among men.

Comparable are Philoctetes’ passionate speeches, when he discovers that (as Neoptolemus leads him to suppose) his very name has been forgotten in the period of his abandonment (–), or when, after a stichomythic sequence in which the shamefaced youth has confessed his deception, he breaks into a tirade of denunciation and abuse, a speech of almost  lines which must have taxed the stamina of an actor. # - ,S   (  '   "E( " ,, P) ( &),, P G)E (Phil. –) You fire, you total horror, you hateful masterpiece of utter villainy, how you have misused, how you have deceived me! (tr. Lloyd-Jones, adapted) 



Gould , n. identifies E. Her. – ( trimeters) as by far the longest iambic speech uttered by the chorus; second comes Hel. –. (A. Ag. – is a possible contender for second place, if all is correctly ascribed to the chorus: so Fraenkel, Taplin –, West in his edition). I print the text of the OCT; but see Austin  for powerful advocacy of the emendation  .,< in  (as he shows, Antigone repeatedly uses negative expressions, and this change would make four successive lines begin with a negative).

II Long versus short: stichomythia



The most memorable transitions tend to be from shorter speeches into rhesis, but the reverse can also be skilfully exploited. Thus in the Agamemnon the king addresses his subjects on entry in a long speech; Clytemnestra appears at the palace entrance and delivers an still longer rhesis of her own; Agamemnon responds with a further speech rejecting the suggestion that he enter the palace by treading on the tapestries, and evidently intends this declaration to settle the matter (note the gnomic conclusion); but rather than accepting defeat, Clytemnestra initiates a question-and-answer stichomythia through which she gains the upper hand. Agamemnon’s next rhesis is one in which he accepts defeat, and the scene closes with an exultant response by his wife in effect glorying in her success, and concluding with lines of ominous ambiguity. The interplay between longer and shorter speeches is best grasped through study of a particular scene, and I briefly outline here their use in the Tiresias-scene of the OT (–). First, as the prophet advances haltingly on to the stage, we have an initial appeal by Oedipus on behalf of the whole community, a rhesis full of the language of supplication (– ). Then comes short but irregular dialogue (some stichomythia, but more two-liners) as Tiresias responds grudgingly and uninformatively and Oedipus grows angrier and more impatient. The exchange becomes heated, with numerous questions and exclamations, vigorous forms of address by Oedipus (already  # 6 ) ,, ‘vilest of vile men’), all leading up to the king’s accusations and especially to his next rhesis, a denunciation which begins with an enormous sentence starting with outraged apostrophising. # -  '  "E "E C1', 6 '03  , 4,  C( ? 1  1'),, , & ,  +" W", ` (   E,   &E, &," ,, .E g ?  ,, CF +" 1, )  ( C   Y( , C1 ()   (E")1, 



A. Ag. –, with stichomythia at –; Clytemnestra’s final speech is –; ominous lines at – (see p. ). Cf. further S. OT – (Jocasta’s persuasive argument that oracles have no validity, based on the experience of herself and Laius), followed by a change of pace and tone, as Oedipus begins to doubt his own security and starts to question her about the past (–, opening a sequence of irregular stichomythia). This is an example of a ‘typical scene’ in tragedy. In four extant plays (S. OT, Ant., E. Phoen. and Bacch.) Tiresias encounters a monarch and advises him; in each of these plays the wise adviser is ignored and usually insultingly dismissed, but is proved to be right. In Antigone, indeed, the king capitulates to the chorus almost as soon as the prophet has withdrawn, though this does not save the situation.

The dramatists at work: spoken verse



  +.E, 4,    ,  ( , * "E  1' '1. (OT –)

O riches and kingship and skill surpassing skill in a life much-envied, how great is the hatred that you store up, if it is for the sake of this royal power, which the city placed in my hand as a gift, though I had not asked it, that Creon the trusty, my friend from the first, has crept up to me and longs to throw me out, setting upon me this wizard hatcher of plots, this crafty beggar, who has sight only when it comes to profit, but in his art is blind! (tr. Lloyd-Jones)

After his speech comes a brief and ineffectual choral attempt at peacemaking. Tiresias responds with an angry rhesis in which he launches his own enigmatic but menacing indictment. Oedipus’ next response, –, is a brief one, expressing his indignation and disgust at Tiresias’ claims, which he evidently does not take seriously. This utterance is marked by colloquialisms, which often figure in angry or emotional passages. A stretch of short exchanges follows, including stichomythia, in which Oedipus and Tiresias are as it were vying for control of the dialogue and the scene. Oedipus’ uneasy question at , a three-sentence line, suggests that Tiresias has thrown him off balance.  , ; (.   ( 1. 6; What parents? Wait! Who among mortals gave me birth?

The notion of the dominant and the subordinate partner in dialogue will often prove a helpful tool of interpretation, and perhaps above all in stichomythic scenes (we see this in the scene in which the Aeschylean Clytemnestra wins mastery over Agamemnon, and throughout the scenes involving Dionysus and Pentheus in the Bacchae). At the end of the short exchanges Tiresias says () ‘I shall be off. You boy, guide me on my way’; in the next line Oedipus tries to take control, to dismiss Tiresias and terminate the unsatisfactory dialogue (‘let him guide you, then; for while you are here you are an obstruction and a nuisance, and when you have left you will cause us no more grief ’). But although we may be assuming that closure is imminent, the scene ends with a momentous further rhesis from Tiresias, even more devastating than the previous one, fraught with prophetic ambiguities and at the end dismissing Oedipus: ‘Go inside and think this over, and if you find me to be mistaken, then you may say at once that I have no wisdom in my prophecies.’ No reply from Oedipus: we sense that he has lost control of the scene, and indeed of his whole 

See  (with Fraenkel on A. Ag. ), and cf. , ; Collard , .

II Long versus short: stichomythia



situation. He retreats into the palace without a word. (The Tiresias-scene in Antigone is comparable: again, a devastating rhesis at –, denouncing Creon’s folly and sacrilege and prophesying the consequences. After that the prophet makes his departure without waiting for a response; but there, the scene continues with the chorus’s awed cautionary voice and Creon’s dismay.) What I have tried to show is that the form of the scene is modified dynamically, in order to reflect changing mood and tempo, but also to convey a shift in power relationships, from the authoritative king dealing with his inferior, a blind and helpless old man, to the reverse situation, a menacing prophet laying out the terrible future (although cryptically) to a younger man who is by no means cowed but is certainly shocked and disconcerted. Oedipus, thrust on the defensive, is at least momentarily (at ) stricken with misgiving and doubt, and in the end fails to have the last word. With both rhesis and stichomythia, scholars have sought to group examples into categories. Some recurring types of rhesis can no doubt be distinguished, whether in more general terms (‘static, deliberative, dynamic’, or ‘informative, explanatory, self-revelatory’) or in terms of the type of speech act involved (supplication, denunciation, expression of despair, narration, etc.). In a few cases there are clear generic similarities with rhetorical forms, as in the psogos directed by Electra at the dead Aegisthus (E. El. –), or the funeral speech of Adrastus delivered over the corpses of the Seven against Thebes (E. Supp. –). But although some family resemblances can be observed, no such categories are really adequate substitutes for detailed analysis in context – a context that includes both interaction with the other characters involved and the cumulative picture created by this rhesis and other utterances by the same character. Similarly with stichomythia, it is possible to identify a number of standard types or uses of the form: to impart or extract information, for example, whether by courteous stages or in insistent interrogation. Stichomythia is also regular in episodes where two parties are intimately involved: the progress towards recognition, as when Cadmus brings Agave back to sanity; the formulation of plans or strategies, as when Orestes and Pylades conspire; or, of particular importance, the process of persuasion and deception.  

This effect would be nullified if Oedipus had already entered the house at the end of his own final speech, as some scholars have argued. Collard , , but he also cautions against excessive classification. For subtypes of rhesis see Mannsperger in Bauformen –.



The dramatists at work: spoken verse

‘Conflict-stichomythia’ is another important type: alternating utterances provide the ideal vehicle for escalating tension or increasing hostility. Rhesis of various kinds will be discussed elsewhere, but the rest of this section will focus on examples of stichomythia. We first consider recognition scenes, where the device serves to dramatise the swift or gradual approach to realisation, while also allowing scope for deferment, sceptical comment, doubts and misgivings. Thanks to the survival of plays by all three dramatists, we can see the family resemblance between the scenes in which Orestes is recognised by Electra. Aeschylus’ is the briefest and most perfunctory, reflecting the strictly subordinated role of the avenger’s sister in the Choephori (notoriously, she is dismissed from the action at f. and never mentioned again). %. "'   ),   ,1'  " ',, '")  6. b.   - V  ( '6; %. & B  L  k FE."' ) . b.   ,. , ) ( '(E 6; %. ,.   ( +7,. ',   +,",   +F( 06, " ! N(, " ! &,6 1). K ( 3, ,6(, +" 3,( ' -, ",   H( .



(Hipp. –)

P: I will grant it, since I feel respect and awe on account of your supplication. N: I’ll be silent now: from here on the word is yours. P: O wretched mother, what a passion you possessed! N: The one she felt for the bull, child? Or what’s this you’re saying? P: And you too, my poor sister, bride of Dionysus! N: Child, what’s the matter? Are you abusing your own family? P: And I, the third unfortunate one, how I am perishing! N: I’m alarmed. Where will this story end up? P: From those roots my misfortune springs, not from recent times. N: I am no closer to knowing what I want to hear. P: Ah! How might it be that you should say for me what I must say? N: I am not a prophet, who can know for certain what is hidden. P: What is it, this thing that they call ‘being in love’? N: It is both pleasant, child, and painful at the same time. P: The latter, then, would be my experience. N: What do you mean? Are you in love, my child? With whom? P: Whoever he is, that one, the son of the Amazon . . . N: You’re speaking of Hippolytus? P: You hear your own words, not mine. N: Oh no! What are you saying, my child? You’ve really destroyed me! Women, this is unendurable. I won’t put up with living any longer. Hateful is the day, hateful the light I see. I’ll hurl my body, throw it down, I will die and be rid of life. Farewell; I am no more. (Halleran’s tr. with some adjustments)

At  Phaedra takes the initiative, speaking cryptically and leading the nurse first to react and then to utter the key revelation herself. This transition of roles is clearly marked in the rather self-conscious line  (with apostrophe of her mother Pasiphae, dead or far away), and still more by  ‘how might it be that you should say for me what I must say?’ Phaedra is driven to reveal the truth, but does not want to utter it herself; instead she guides the direction of the exchange, first by citing the erotic misfortunes of her relations, then hinting at herself as ‘third’ in the pattern of disaster; when the nurse still fails to take the point, she is  

Quoted in Ar. Eq. , and so presumably notorious for its artificiality. The significance of Phaedra’a ancestry for her character and for the play as a whole is discussed by Winnington-Ingram  (= Mossman , –).

III Prologue



forced to offer an even more blatant hint, through the rhetorical question as to the meaning of ‘love’. At  the nurse finally catches her drift, but although she asks a plain question, Phaedra in her reply still tries to leave the identity imprecise with the evasive phrase 4,    $ ,  (‘whoever he is, that one’), and only the final word of the line, ‘Amazon’, makes all clear. The shock of the revelation is marked by antilabe and the breaking of stichomythia, as well as by the nurse’s frantic reaction – short clauses, asyndeton, no clear sequence of thought (–, five lines with no connectives). The revelation has been made in two stages. The first mention of Hippolytus made clear that he was a focus of concern, the second reveals the shocking truth of Phaedra’s immoral passion. Naming is crucial in each case. – is a significant echo of : in each place we meet the name Hippolytus, in the accusative, at the start of the line; in both lines there is antilabe; and the distressed cry of oimoi is repeated. But in – it is the nurse who utters the oimoi, now realising the purport of her mistress’s previous voicing of that cry. The reversal is reinforced by the echo of the verb +'( (‘destroy’): in , when the nurse had first used the name, Phaedra cries that she will be the death of her if she uses it (+7,). Now, it is the nurse who wails that her mistress has destroyed her ( > ( +7,). Thus the rhesis made Phaedra break silence, but the stichomythia first erodes her self-control still further, then drives her to reveal the truth and provides her with the means. Euripides could not have enabled her to make this gradual and indirect advance toward revelation through any other form. iii prologue Aristophanes makes clear that contemporaries had observed a certain formulaic quality in the Euripidean prologue, which characteristically introduces a single figure on stage – a figure who, in a kind of monologue, recounts his/her family history, describes the events leading up to



On asyndeton see Demetr. Eloc. , Long. Subl.  (with Russell’s notes), Schmid , , Bruhn §, GGL i.. n., i.. n., Lausberg , §–, Denniston , ch. ; Zucker . Two types should be distinguished. In one, as here, we have a series of sentences without connecting conjunctions or particles (for another case see S. Trach. –, quoted below, p. ). Often, as in these two cases, this indicates agitation or excitement. In the other, a number of words with parallel function in a sentence are combined without conjunctions: the effect is usually a sense of rapidity or urgency. For some common varieties see Bruhn §–, E. Her.  with Bond, Phoen.  with Mastronarde.



The dramatists at work: spoken verse

the present situation, and indicates what stage the action has reached. Sometimes the speaker is a central figure in the action (Andromache, Electra, Jocasta), but often Euripides uses a prologue-speaker who is less directly involved (Polydorus’ ghost in Hec., Medea’s nurse). The prologue-speaker is also sometimes a god, and in that case the deity almost always vanishes from the action after this initial scene-setting: Hermes in the Ion, Aphrodite in Hippolytus. Bacchae is therefore an important exception. Aristophanes of course oversimplifies, and even the preceding paragraph only begins to describe Euripidean practice. If we look at the tragic corpus as a whole, we see a wide variety but certain clear trends in each of the three dramatists. Two points are central: a prologue need not consist of a single ‘scene’, but may include two or even three in sequence, normally with no change of location but involving different characters and modes of delivery (thus song or recitative by an actor may precede the chorus’s entry). Second, there is a clear distinction between prologues which consist of monologue only, dialogue only, and a combination of these. The dialogue opening is typical of Sophocles. These points are related to the question of length. Not all of Aeschylus’ dramas have prologues, but when they do, the prologue is generally quite short ( lines only in Ag., remarkable given the vast length of the play). Later in the century they are typically between  and  lines long, but the length is usually achieved by an alteration of speakers and mode. Thus in IT there are two scenes, first the long monologue by Iphigenia outlining her situation, then after she has left the stage a dialogue between Orestes and Pylades. Once they have departed, Iphigenia re-enters and joins the chorus in a lyric exchange which stands in for the usual parodos. An example of a tripartite prologue occurs in Euripides’ Electra: first monologue by Electra’s husband, followed by dialogue between him and his wife, followed by a speech by Orestes to the silent Pylades (here, obviously, the first and second part are closely connected). Longest of all extant prologues is that of the Phoenissae, which has two scenes, a staggering -line monologue from Jocasta, followed by a lively dialogue between Antigone and her paedagogus, as they first ascend to the walls of the city, represented by the roof of the stage building, then observe the assembling warriors below.

 

Andromeda was evidently one exception, since the evidence suggests that the play began with anapaestic dialogue between the heroine and Echo (F). For Aristotle’s definition of the prologue as the part of the play preceding the chorus’s entry see p.  above. See further Nestle , Schmidt in Bauformen –, Erbse , Collard on E. Suppl. –, Roberts .

III Prologue



This second sequence brings the prologue to  lines. Ion and Helen are not far behind ( and  lines). As the prologue opens the play, the characters need to be introduced and the scene established. Important themes or aspects of the plot will immediately become apparent (sometimes even before a word is spoken, as when an individual or a group of suppliants are taking refuge at an altar). Language and choice of terms may have deeper resonances which will echo through the play, often going beyond the superficial sense in which the speaker uses the words in the opening lines: Sophocles is a master of this technique. The watchman alludes to the ‘man-minded heart’ of Clytemnestra; Odysseus emphasises the fact that Lemnos is uninhabited, a device which untraditionally enhances the isolation of the marooned Philoctetes. Aeschylus in two plays dispenses with the prologue, allowing the chorus to enter at the start of the play (Pers., Supp.); in two others the prologue is markedly short (Ag., Cho.). Most elaborate is the opening of the Eumenides, a remarkably rich and complex structure. The first part consists of two successive speeches by the Pythia, before and after she has entered the shrine. In the first she speaks with dignity of the history of holy Delphi. In the second her composure is gone, as she crawls in on hands and knees (–) to describe the hideous vision of the Furies whom she has found inhabiting the sacred space; at the end she flees from the stage. There follows a scene of dialogue between Apollo and Orestes: the god gives confident assurance, the mortal expresses the trusting dependency of a suppliant. Either during or after this scene the Furies become visible to the audience, revealed by Apollo to his prot´eg´e and the audience, or else called forth from the shrine by Clytemnestra. After Apollo, Orestes and possibly a silent Hermes have departed, the ghost of Clytemnestra appears and angrily urges the Furies to awaken and pursue her son, their escaping prey. One initial speech results only in restless mutterings or 

    

This figure ignores the probability of small-scale interpolation: Jocasta’s speech should probably be trimmed of at least  lines. Mastronarde sufficiently refutes the suggestion that the entire teichoskopia scene is spurious. I set aside IA ( lines in modern texts) as there is little likelihood that the extant prologue material is all Euripidean: probably two alternative prologues have been combined. So e.g. E. Supp., Hel.; cf. S. OT. See further pp. – above. Easterling a subtly analyses the verbal resonances in the opening speech by Oedipus in OC. A. Ag. , S. Phil.  (contrast Hom. Il. .–, Od. ., and the plots of Aeschylus’ and Euripides’ Philoctetes dramas: see Dio Chrys. Or. ). On the staging, see Taplin –; Brown ; Sommerstein on Eum., pp. –; West b, –.



The dramatists at work: spoken verse

growls from the daemonic horde; a series of further exhortations have more effect; then the chorus burst into wakefulness and seconds later into song, an extraordinary transition which ends the prologue with their angry realisation of Orestes’ escape (the song is predictably agitated, a mixture of dochmiacs and iambics). All this in less than  lines of verse. The Eumenides as a whole moves the trilogy from darkness into light, from conflict to resolution. The change of location, first to Delphi and subsequently to Athens, reflects this movement away from doom and blood-feud. Accordingly, the opening speech of the Pythia is one of piety and celebration. Her account of the succession myth of the Delphic oracle differs from others in that Apollo gains his place there peacefully, not by violence, and as a gift from the Titaness Phoebe, not by slaying a monstrous snake (–). The language of prayer is prominent in the first speech (–, , ,  ,, f. et al.). The ordered succession of holders of the oracle is mirrored in the orderly sequence of gods invoked, all of whom have some place in the history and rites of the holy place. The solemn mood and the expectation of order, of a day of customary prophetic consultations, are disrupted by the apparition which the priestess discovers when she enters the shrine. In an Athenian drama, the lines which connect the Athenians with Delphic Apollo deserve special notice. (',   !  ,0',  ( '    jb1,', "  +3(   T((E. (   ! )  (1 7, 51  "7 , '(3E MF. (Eum. –)

The sons of Hephaestus escorted him here with great reverence and made a road for him, taming an untamed land. After his arrival, the people magnify him with honour, as does Delphos, this land’s lord and helmsman.

That Athenians treat the gods and especially Apollo with reverence, that they are a civilising people (‘taming’), and that they provided escort to the god on his first arrival in Delphi not only does them credit but prepares for later events, in ways both obvious (they will pay respectful attention when Apollo bears witness) and unexpected (at the end of the play they 

Prayer opens all of Aeschylus’ authentic dramas except Septem (where an invocation figures later in the prologue at ).

III Prologue



will provide escort to gods of a different kind, the Erinyes, on their first arrival in Athens itself ). The prologue is designed to impress the audience with the authority of Apollo and the repulsiveness of the terrifying Furies: this paves the way for the clash between Olympian and chthonic, younger and older divinities, which dominates the rest of the play. The Furies’ effect on the Pythia is electrifying: she is reduced to panicky collapse. They are described before we see them, and in bewildered terms, as if no description is adequate: ‘a band of women seated – no, I don’t mean women, but Gorgons: but no, I can’t compare them to Gorgon figures . . . these have no wings to be seen. And they are black, utterly revolting in their manner, snoring out a breath which is unapproachable, while their eyes run with a loathsome fluid’ (–). The language is vivid but also ugly and straying beyond the usual limits of tragic diction. Nowhere else in tragedy do we meet the coarse word ' (‘revolting’) or its compounds, though the adjective is frequent in comedy; and K (‘snore, snort’) is also a rarity, occurring elsewhere only of Rhesus’ horses. Sleep may be ascribed to gods in high genres, but never snoring: here the Furies are treated as more like animals than immortal powers. A strong vein of invective and fierce criticism of the Furies – their looks, their character, their behaviour – runs through the earlier part of the play, particularly voiced by Apollo. In speaking to Orestes, as later in driving the Furies from his shrine, he spares no insult (–, –); in the trial scene he loses his temper and hurls further abuse (). The Pythia’s reaction prefigures her protector’s more authoritative contempt. The audience, already predisposed to sympathise with Orestes, naturally accepts these judgements; but as the play develops, it gradually becomes clear that outright dismissal of the Furies and their claims is inadequate: they do have a case, and a place in the scheme of things – eventually, a place in the Athenian community. (The ode beginning at  is especially important for this shift or broadening of sympathy.) Important here is the difference between Apollo and Athena. Apollo’s reaction to the Furies is abuse and denunciation; Athena is respectful and courteous (–, ), 



(, ‘escort’, recurs at the end of the play, at ; also the role of the propompoi, if correctly so called. This term is used by the scholia and the list of characters in the hypothesis, not in the text. Therefore the audience could not know that this secondary chorus was so designated. West gives them the title neokoroi (temple-servants). Rhes. . Sommerstein  discusses the ‘comic’ or grotesque features of language which characterise the Erinyes in this play. Snoring is of course at home in lower genres: e.g. Ar. Eq. , Nu. .

The dramatists at work: spoken verse



making every effort to placate and reassure the Furies after their frustration in the trial. There is a sense in which Athena, who presides over the court, is trumping Apollo, a partisan advocate; Athenian hospitality and piety turn out to surpass Delphic sanctity. The prologue of the Antigone follows a simpler structure, to characteristically Sophoclean effect. In all but one of his extant dramas the action begins with one character addressing another or (as here) entering in mid-dialogue, with recent events casually mentioned in the course of the exchange and the current situation swiftly becoming clear to the audience. In this case, as in Ajax and Philoctetes, one character needs enlightenment from the other and the audience also benefits (see especially the outline of the situation in –). One difference from these parallel scenes lies in the gender of the two characters. Since they are women of status, tragic convention requires some apology for their roaming freely outside the house. Antigone refers to this in her second speech: she has brought her sister out of the palace so that she can confide in her secretly. This stepping outside is a first symbolic step towards transgression: the idea of crossing boundaries is important in the play, and highlighted verbally in the confrontation between Antigone and Creon (, answered by ; cf. , ). ‘Do you know what’s happened? have you heard? . . . No, no fresh news has reached me . . . Well, I knew that, that is why I brought you outside with me . . .’ Antigone’s passionate opening questions, her urgent desire to share the bad news with Ismene, set the keynote: she is the one who takes the initiative, while Ismene responds more tentatively and from a position of inferior knowledge and courage. Sophocles uses this exchange to create a clear sense of a powerful, demanding personality confronting a more passive and timid figure. Yet Ismene does not lack a point of view of her own (–) and has at least some aspirations to integrity (f., ). Had the scene been composed by Euripides, we would surely have had an opening monologue by Antigone enumerating the previous sorrows of the house of Oedipus (cf. Jocasta in E. Phoen.). Instead Sophocles allows the situation to emerge dynamically, partly through Antigone’s speech reporting Creon’s action (‘this is what they say the admirable Creon has decreed’, ), partly through Ismene’s painful reminder to her sister of how much they have already endured (–, her only full rhesis). At first intimate and affectionate, the dialogue becomes more edgy after Ismene’s question: what has Antigone in mind? The vocatives used are revealing  

Gould , –, Easterling a; Finglass on S. El. . For a useful discussion see Taplin .

III Prologue



( # 1, ‘you bold-hearted person’,  ‘you rash creature – when Creon has forbidden it?’). Ismene’s counter-plea, seeking to restrain her sister from self-destructive action, leads to a breakdown of relations. Antigone resolves to proceed alone, declaring her allegiance to the dead (). A spasmodic passage of stichomythia concludes the scene: Ismene continues to dissuade and expresses her concern, Antigone dismisses both counsel and promises of secrecy. The distressed cry oimoi, previously used by Ismene in genuine dismay (, ), is sarcastically redeployed by Antigone in a line which spurns her sister’s promise of silence ( ‘Oh dear no, do say your piece. You’ll only make yourself far more hateful through silence’). In the prologue Antigone ends her first speech by declaring that ‘evils of enemies are coming against our friends’ (); but despite this initial assumption that their loyalties are as one, by the end of the scene she is speaking of her sister as an enemy (, –): it is her brother who is dear to her, philos: ‘I am his own, and I shall lie with him who is my own’ (). But as they part Ismene declares that Antigone, though misguided, is ‘truly dear to those who hold you dear’ (), an assurance that anticipates her futile effort to intervene and share the blame much later in the drama (–). Both the contrasting characters of the sisters and the increasing gulf between them are reflected linguistically. In an earlier chapter we considered the famous opening line of the play, spoken by Antigone, where both   and  )1 stress their closeness. Conspicuous too is the regular use of duals (‘we two’) in the early part of the scene, whereas later Antigone uses first and second person pronouns and verbs, frequently in opposition to each other (, , –; – mark a kind of transitional challenge). Syntactically, Antigone is more definite and decisive in her assertions (–, –, –); Ismene is more prone to doubtful questions, potential and conditional clauses (esp. –, , ), and generalising vagueness (e.g. the gnomic ending at –, and ; Antigone has no gnomai in this scene). Vocabulary also plays a part: Ismene, who approximates more to the Athenian stereotype of a woman, refers more than once to her or her sex’s feeble nature ( ‘you must recognise that  

 

The use of 1 in this play repays study: see esp. , , , , , , , , .  should be ascribed to Ismene: see Griffith ad loc. The language of sharing, especially the use of the prefix F'-, is noteworthy: see , , contrast ; later, , . , a famous line, should rather be connected to : Antigone’s loyalties are with the dead. See above, ch. , pp. , , and also Griffith . See p. . By contrast Ismene’s forms of address are either critical or, even in her fullest appeal, less emotional ( ‘o sister’).



The dramatists at work: spoken verse

we were born ( 1'() women, unfit to fight against men’, ; contrast Antigone’s use of the verb at , ‘whether you were born (1') noble, or base from honourable stock’). Ismene also repeatedly uses the word +(3" (incapable or impossible) (, , ) with regard to Antigone’s vain undertaking; moreover, she refers to ‘rulers’ and ‘being ruled by those more powerful’, whereas Antigone’s first reference to Creon is dismissive, as ‘the general’ (), another is sarcastic (, ‘the admirable Creon’), and nowhere in the play does she admit his title as ‘king’. In the prologue to Euripides’ Orestes Electra is present throughout, watching over the comatose form of her hapless brother. The first part is the characteristic monologue ( lines); next comes a brief dialogue with Helen, who emerges from the palace and makes an attempt at friendly overtures, then withdraws after dispatching the silent Hermione to Clytemnestra’s tomb ( lines); at the end Electra comments bitterly on Helen’s character, then observes with misgiving the arrival of the chorus, fearing that it will waken the sleeping Orestes ( lines). Electra’s monologue begins with a gnomic utterance, which is then illustrated by the example that follows (Tantalus), and can indeed be applied to his descendants. %  ,   8  ! k &   8 )   8 F'(1 3, u   \ M  M"  + 7' 1., . ?  ()  (  I 0 .") 5 ! 1'7,  ', , a) '1 C  (  +     .E E,

 (8 ', , 4  M  f,   0E +F( " H,, +, ," 6,,, &,",E ,. (Or. –)

There is nothing so terrible to relate, no suffering, no divinely-sent misfortune, the burden of which human nature may not be required to bear. For the man blessed with good fortune (I do not mock his experience), the offspring of Zeus, they say, Tantalus, dreading the rock that hovers above his head, floats suspended in air, and pays this penalty, so at least they say, because he, a man, enjoying equal status with gods at the shared table, was possessed of an ungoverned tongue, most shameful sickness. 

These figures allow no reduction for interpolations; of the deletions proposed, perhaps only  is certain, though  is likely. It remains obvious that the monologue is the chief part of the scene.

III Prologue



Despite Euripides’ well-known penchant for the gnomic, to begin a play with such a generalisation is unusual (Heracl. and S. Trach. provide parallels). It does not indicate Electra’s detachment from the ordeal she herself endures; more important are the recognition of human suffering as divinely-imposed, and the idea of the Tantalid house in particular as being subject to perpetual misfortune. True, Tantalus was ‘blessed’ at first, favoured by the gods, but his good fortune vanished. (Similarly she questions her father’s achievements later in the speech:  ‘the glorious Agamemnon – if indeed he was glorious – was his son’). But misfortune can be earned or self-inflicted, and the punishment of Tantalus is a paradigm of sin meeting retribution from the Odyssey onwards. The nature of his offence varies: here, the notion that he boasted or said too much seems a crime less horrendous than those of his descendants, Orestes included. The link with Orestes is swiftly made clear through the language of sickness ( ,), for Orestes’ madness is often so described. Does Orestes’ crime deserve still grimmer punishment? Almost at once, in line , we encounter a parenthesis, ‘excusing’ her account of the grim career of an ancestor. This is closely followed by the first of a number of references to tradition ( ‘they say’,  ‘so at least they say’). Reference to tradition is not in itself unusual: the Pythia in the Eumenides uses a similar expression to indicate that she has no firsthand knowledge of the shrine’s history (). Euripides, however, uses the phrases more freely and sometimes in ways which can be taken to imply scepticism (cf. Hel. –, , IA –). In other respects too Electra’s editorialising in this speech is foregrounded. A few lines later she draws attention to her own omissions of shocking detail:  ‘why should I give a full account of the unspeakable?’ (the Thyestean feast), and again at  ‘the son of Atreus (for I remain silent about the events in between)’ (passing over the crimes of her grandfather). Soon we draw closer to the central issues of the play, when she mentions her mother: ‘who cast a boundless woven cloth about her husband and killed him; but as for why she did this, it is dishonourable for a maiden to utter the reason; I leave this obscure, for general contemplation. And as for Phoebus’ injustice, why should I indict him?’ (–). With the shocking juxtaposition of Apollo’s name with +  Electra has voiced for the first time a complaint which runs through the play, that after convincing Orestes to kill his mother the god has abandoned him. 

Line , though corrupt, must say something about disease; see further Smith , Willink on Or., xlii–iii (who suggests that  F'(1) also has medical connotations).



The dramatists at work: spoken verse

All these self-conscious comments and interruptions produce a very spasmodic exposition. In some ways this suits the distraught Electra, exhausted and frantic with concern for her brother; and the chaste praeteritio with which she avoids reference to her mother’s adultery could create a positive impression of her modesty (an impression effaced by her later characterisation). But some at least of these comments seem almost to undermine or draw attention to the fictive quality of the narrative: events can be narrated or left out, characters can be praised or condemned. This suits the reading of Orestes, so late a product of Euripides’ pen, as a kind of palimpsest, in which numerous motifs from earlier dramas, his own and others’, on the Orestes legend and on other themes, are reworked in a highly selfconscious way, as part of a repertoire with which author, audience, and at least intermittently some of the characters are familiar. Despite the ostentatious omissions, it still takes Electra some  lines to bring the narrative up to the present situation: she and her brother are waiting for the decision on his fate by the Argive assembly (cf. pp. – below). F  c 6  (3  T( , , (* ' ",  , (E8 ,1   (E- '  L T(  u  , 1 ;  , & "*   ',( 7( . (Or. –)

This city of Argos has decreed that no one is to shelter us beneath his roof, no one to welcome us at their fireside, no-one is to speak to us, matricides that we are. And this is the appointed day when the community of Argives will divide its vote on whether the pair of us must die by stoning.

The emphatic negatives emphasise the strength of the prohibitions; the stress on ‘the appointed day’ brings home the critical stage they have reached. Tragedy, when it does anticipate the Aristotelian precept that the action should be restricted to a single day, naturally chooses a day of special importance. After six days of inaction, the crisis is now imminent. Electra’s speech gives the audience clues of characters and events yet to come: the madness of Orestes, the advent of Menelaus (their only hope in her view), the decision of the assembly. It is telling that in this version she entertains no hope of support from Orestes’ patron Apollo.  

See esp. Zeitlin . Taplin . Cf. e.g. S. Aj. –, OT  L< T( 1., ,   1  (‘this day will give you birth and will destroy you’), E. Alc. , ; Hipp. , Phoen. ,  with Mastronarde.

III Prologue



The subsequent interlude with Helen also serves to characterise Electra, whose hostility toward her is undisguised (–), but expressed with even more viciousness when she has left the stage ( ‘may the gods’ hatred pursue you . . .’). Again the dramatist is preparing for later events, this time unorthodox and unlikely to be foreseen by any audience: the assassination attempt upon Helen, proposed by Electra and eagerly accepted by Orestes and Pylades (, , etc.). The appearance of Hermione, a mute character thus far, is also functional: she too will be an object of their malice later, and will appear as a hostage in the final scene. Helen’s brief opening speech is a masterly presentation of a benign but shallow, tactless personality. [#  g' (3,  +((]   (! * ( , &" .,  , "0, - d( ,      + ' &((),  (   - 6   E ( , -  "  { ' 6 ,   ") . W ! C- '(! ?( ,7, +1', 6 IF d( ,1* ) ( ( ) 1  ,', "',  8 u,, _  , ) ,E! )'  ".(, .







(Ag. –)

I stand here where I struck, the deed is done. And so I achieved it – I will not deny it – that he could neither escape nor ward off his doom. An endless net, like a fish-net, I cast around him, an evil wealth of dress, and I struck him twice; and with two cries there on the spot he let his limbs go slack; and then, when he was down, I added a third stroke, a welcome offering of prayer to the Zeus who protects the dead, the one who reigns below the earth. So he belched out his life as he lay there, and blowing forth a sharp burst of blood, he struck me with a black shower of gory dew, and I rejoiced no less than the crop rejoices in the rich blessing of the rain of Zeus when the sheath is in labour with the ear of corn. (tr. Fraenkel, modified)

Whereas Clytemnestra earlier said she was not ashamed to protest her wifely devotion, here she declares herself unashamed to admit that she plotted revenge in secret. There is no apologetic note, no references to motives (Iphigenia comes only later): she speaks explicitly of ‘enemies’ (). The imagery of entrapment and hunting reverses the earlier descriptions of Troy ensnared by the nets of Zeus: now it is Troy’s conqueror that is captive. But above all the speech is conspicuous for its religious language.

IV Deceptive women



Clytemnestra turns her murder into a blasphemous religious sacrifice to Zeus Katachthonios, the protector of the dead; this makes the killing all the more fitting as revenge for the sacrifice of Iphigenia, another kin-killing (the ‘third offering’ was mentioned in that context, one which is recalled also at the end of the speech, with the sympotic metaphor, , cf. – ). Perverted religious ritual is compounded with perversion of nature: the blood of Agamemnon’s final choking gasps falls upon Clytemnestra (her own dress was presumably visibly stained throughout this scene), but instead of being repelled by this, she rejoices as the crops flourish in fresh rain or dew. Fertility and natural growth are juxtaposed with vicious, even sadistic destruction (some would even diagnose a sexual element). The queen’s perverse ‘joy’ is then sardonically picked up after her mockrespectful salutation to the elders, a vocative phrase exactly parallel with her words in the greeting-scene (): ‘be glad, if you will be glad’ echoes her own gladness in the rain of blood (" < ∼ "',). ‘I exult in it’ may raise qualms in those who remember Odysseus’ caution at exulting over the dead in book  of the Odyssey (an important source generally for this trilogy). Certainly the chorus’s reaction is to marvel at her boldness of speech in vaunting like this ‘over a man/husband’ (–). This is Clytemnestra’s moment of greatest triumph, but it marks the apex of her trajectory in the trilogy: now begins her gradual decline. Other scenes involving Clytemnestra are discussed in other contexts (see esp. pp. ,  on the confrontation in Choephori). What emerges from an overview is the importance of the point made already at Ag. , that she is a ‘man-minded woman’, ruling and managing the household in a man’s place, planning with resolution, resisting any accusations of feminine credulity. When the alarm is raised late in the Choephori, she calls for ‘a man-slaying weapon’ (). With this goes the idea of Aegisthus as a ‘woman’, playing an inferior role in the conspiracy and in their subsequent tyranny (Ag. , Cho. ). This notion of the usurper as a woman’s man is developed more fully in the later dramas on this theme. More broadly, the idea of ' , rule by a woman, already well-established in myth, becomes a powerful template in tragedy, no doubt because of the     

On the perversion of ritual language in Aeschylus see e.g. Ag. , –, , –. For imitation of this passage by Euripides see Bond on Her. . See Fraenkel’s superb note on , also citing the contrasting ‘source’-passage in Iliad .–. See Moles  on –; also probably ; Pulleyn b on –. On the erosion of Clytemnestra’s position in the lyric-recitative-dialogue which follows see Taplin – with older bibliography in n., Foley , –; ch. , n. . See Fraenkel on Ag. , and note the development of the motif in E. El. –, esp.  where Aegisthus is said to be regarded as ‘Clytemnestra’s husband’ rather than her being his wife.



The characters of Greek tragedy

Oresteia’s influence. Combined with this is the idea of Clytemnestra as wearing a mask, as arch-deceiver. In the first play she acts the part of the loyal wife to the point of exaggerated devotion: we have seen how the extravagant rhetoric of her speech-making strikes repeated false notes. At the climax of the Choephori, in desperation, she takes on the part of the affectionate mother: her simpler language and more intimate appeals must have a powerful impact on both the audience and Orestes (esp. –, , ), but we remain in doubt as to whether this too is an act. Deception too becomes a standard element in the tragic model of a woman: it is their m´etier, what they are best at (e.g. E. Andr. ). Finally there is the sexual factor. Greeks believed women to be more prone to sexual desire than men, and less able to resist it; this forms a motive for many dramas, not only in the tragic vein (Aristophanes’ Lysistrata comes to mind!). Clytemnestra perhaps hints at sexual deprivation in her speech to Agamemnon (Ag. –); she certainly appeals to it as a justification in the life-anddeath exchange with her son (Cho. ). In general she would seem to be superior to such weakness, but in the confrontation with the chorus after killing her husband she boldly declares that no fear will find its way into her house ‘as long as Aegisthus kindles the fire upon my hearth’ (– ), the first explicit use of her lover’s name in the play; and when she realises he is dead in Choephori she cries out in distress ( ‘dearest might of Aegisthus’), a cry which provokes an angry reply from her son. These three elements – masculinity of mind, readiness to deceive, and the motive of adulterous desire, seem to define the character of Clytemnestra, but the importance of deception to the plot and to her personality is that it makes her impossible to pin down: we cannot know, in particular, how seriously to take her anger over the death of Iphigenia, which she presents as justification: the question whether this or passion for Aegisthus was her dominating motive had already preoccupied Pindar (Pyth. ). 

 



The scene with the nurse is important here, since she has given an account of Clytemnestra’s behaviour which brands her as a hypocrite. In particular she claims to have nursed Orestes herself (), though that would not necessarily preclude his mother from doing so on occasion. For discussion see Garvie on Cho. –. Here is another case where we must remember that this is only a text: we cannot say whether one or both women are making true claims. [Hes.] F ; Dover , –; Foley , –. Cf. E. Med. , –. That sexual overtones are present is at least possible: see Henderson , –, –. For outspokenness in sexual matters in tragedy see also Ag. , E. El. – (Electra denounces Aegisthus; introduced with a self-righteous praeteritio). On the dating of the Pindaric poem see Finglass’s commentary on Pyth. , –, esp. –. The two possible dates are  and . The balance is in favour of , making Pindar’s version earlier than the Oresteia; many modern scholars have resisted this conclusion.

IV Deceptive women



Clytemnestra has been treated at such length because she is an archetypal figure. There is evidence that Aeschylus’ dramas were re-performed already in the late fifth century, and the fame of the Oresteia is attested by references in Aristophanes and still more by the numerous imitations in later tragedy: Sophocles’ Electra, Euripides’ Electra and Orestes, and central episodes in Euripides’ Hecuba and IA are shaped by or represent reactions to the magisterial trilogy. The influence of Clytemnestra on the image of the tragic female is still more widespread. To illustrate the point we can usefully turn next to Sophocles’ Trachiniae. Deianira is a character who parallels Clytemnestra – both are wives who destroy their husbands by devious means when they return home from campaigning with a young concubine. Verbal correspondences in the later part of the play reinforce the parallel with the Clytemnestra of Aeschylus in particular. That Deianira in earlier versions was a more malevolent figure, deliberately planning her husband’s death, is at least a tenable hypothesis. In Sophocles’ play she is a gentle and sympathetic character, who has suffered much as a result of Heracles’ frequent absences and infidelities (–, –, etc.), but who causes his death involuntarily, believing the Centaur’s lying promise that his preserved blood will act as a love-charm. Earlier in the play, however, we find a scene with Aeschylean pedigree: the episode in which the prisoners, including Iole, are escorted on to the stage and surveyed by Deianira, who is especially struck by the young woman’s nobility and beauty, while still unaware that she has been sent home to serve Heracles’ sexual needs. When she addresses the girl with compassion but is unable to make her speak, we cannot fail to see a similarity to the scene in which Clytemnestra tries to summon Cassandra indoors (Trach. –). This imitation characterises Deianira as different from Clytemnestra: her tone of pity and generous feeling makes for a delicate and touching moment, even though the audience may relish the irony of her ignorance. But later events change Deianira’s attitude: although she does not hate the girl, she is bitter and heartbroken; and in the end she takes steps that bring about her husband’s agonising death – so is she in fact so different from Clytemnestra after all? The scene involving Iole, however, deserves closer attention.    

 MacEwan , Kromar . TrGF iii, pp. –. Fraenkel on A. Ag. , Kapsomenos , ch. . [Hes.] F.– M-W, schol. Ap. Rh. ., Apollod. Bibl. ..; see further March , –, Gantz , –, –. Observed e.g. by Reinhardt , n..



The characters of Greek tragedy 4(  , ,  [ ,'(   ! [ ),, (* ,1 . (  9  ! &,E, 1 , . ?7,E ',('  FE "7 +' +))  +(, ~  (8 N, F '  H, +6, - 8 - H,"',  . # t- a, (3  &, ( , !  (! W ,( "3,)  , (E, H  ), , ,  07,E   W    ), ?(E. # ',) ,   9 ; M, _ -,,; ! (8  1.,  ) M  6,    .





(Trach. –)

But none the less, it is the way of those who consider things with care to fear for the man who is fortunate, in case he may one day come to grief. Yes, a strange pity comes upon me, dear women, when I see these unhappy ones homeless and fatherless, astray in a foreign land; perhaps they were formerly the children of free men, but now their life is one of slavery. Zeus, god of trophies, may I never see you go against my offspring in this fashion; if you do so, may it not be while I still live! Such is my fear as I look upon these women. Unhappy one, who among all young girls are you? Have you no husband, or are you a mother? You look as though you know nothing of all these things, but you are some noble person. (tr. Lloyd-Jones)

Deianira’s pity for the captives is expressed first to the chorus, her fear of similar disaster befalling her family or her house through an invocation of Zeus as patron of battles; finally her notice is drawn to one victim in particular, who is of course Iole. In her consciousness that fortune is fickle and success fragile, we discern a more prudent insight than Clytemnestra’s, and one typical of tragic wisdom. That awareness was lacking in Clytemnestra in the equivalent scene; although preening herself on the fact that her household had long been rich, she made no reference to the possibility that it might fall on harder times. Deianira’s insight leads on to pity, in an intense form ( ´o); the pathos of the captives is stressed through the negations of  (+' +))), and through the opposition of ‘free’ and ‘enslaved’. Pity is allied to fear, 

Cf. Odysseus in Aj –, Theseus in OC –; Cyrus in Hdt. .; Dover , . Konstan  argues for a distinction between pity, where distance exists between the two parties, and compassion, where the bond is closer (e.g. through kinship); but the trend of these passages seems to be to bridge the gap.

V Neoptolemus in Sophocles, Philoctetes



in a very Aristotelian way ( ,  9,   ;  ‘for her I pitied most’). After initially addressing the girl direct (–), Deianira questions Lichas, whose evasive reactions strongly suggest he has something to hide, then turns back to the girl (–), intensifying the request (‘speak to us, poor wretch, from your own lips’); remarkably, she even declares that it is a ‘misfortune’ ( F'(1)) to her not to know who the young woman is. Whereas Cassandra’s silence merely angered and frustrated Clytemnestra, we feel that Deianira is genuinely distressed at Iole’s reluctance. The captive’s silence is also emblematic of the helplessness of such victims in a context of war (in the Iliad Chryseis never speaks, Briseis only once, though memorably). The combination of ideas of pity, changing fortune and ironic ignorance seems characteristic of Sophocles. v neoptolemus in sophocles, philoctetes The anonymous Life of Sophocles praises his skill in characterisation: ‘He knows how to arrange the action with such a sense of timing that he creates an entire character out of a mere half-line or a single expression’ (). The praise is justified but deserves fuller illustration. The example of the Philoctetes also enables us to consider the representation of male relationships. The young Neoptolemus is under orders to follow Odysseus’ lead, and is suitably respectful of his authority in the opening scene ( MF; cf. ). But already there are signs of misgiving and of potential difference between them. Diction and syntax anticipate the dilemma to come: addressing him as ‘son of Achilles’, Odyssues tells him that he must be ‘noble’ () in what he is undertaking, ‘and not in body alone [i.e. it is more than just your strength that I need]; if you hear something novel, of which you had no knowledge till now, you must lend your aid, as you are here to give that support’ (–). The deferral in revealing what he has to say suggests the speaker’s uncertainty about Neoptolemus’ response. When asked to explain, he goes on: ‘What is needed is for you to steal away Philoctetes’ wits by tales told in conversation . . .’ (–). The verb   suggests rhetorical deception, but also hints at the actual theft of the physical object, the bow; and how well does the ‘nobility’ he mentioned a moment ago fit with ‘stealing away’? The prologue effectively prepares for the later rebellion 

Easterling , – offers as samples the whole line Ant. , and in shorter compass Phil. –, in which Philoctetes accusingly asks Neoptolemus  ( 1, , C1., (. + H( V ( !  ! M M   (*  ,  . . . . . . .( , MF, 6 6 F( ( _   6. %.