Split-vision: Secondary Action in Greek Tragedy William J. Slater N THE TECHNICAL ASPECTS of tragic performance the Swiss scholar Anna Spitzbarth wrote a thesis inO1946, and one of her chapters was dedicated to what she called Doppelvorgänge, i.e. passages where two discrete actions are happening on the stage at the same time.1 After listing about twenty passages where it is obvious from the text that two separate actions are indeed being performed simul- taneously on the stage, she recognized a main hindrance posed to such an investigation. We simply do not often know beyond all doubt to what extent a character is visible on the stage or not, especially during choral songs, let alone what they are doing, and in such cases the various conventions and mech- anisms by which we try to determine the presence or absence of a character, or define their exits and entrances,2 are all too often subject to exceptions. But she concluded correctly (77) that 1 A. Spitzbarth, Untersuchungen zur Spieltechnik der griechischen Tragödie (Zurich 1946: hereafter SPITZBARTH) 73–78. She includes a number of scenarios that I exclude, e.g. chorus responding to off-stage cries, and omits accidentally some of the examples I offer, e.g. the old seaman in Euripides’ Helen. 2 M. R. Halleran, Stagecraft in Euripides (London/Sydney 1985) 50ff, gives an introduction. Examples of this kind of deduction of normative dramaturgy are G. R. Manton, “Identification of Speakers in Greek Drama,” Antichthon 16 (1982) 1–16; R. Hamilton, “Announced Entrances in Greek Tragedy,” HSCP 82 (1978) 63–82; J. P. Poe, “Entrance-Announcements and Entrance-Speeches in Greek Tragedy,” HSCP 94 (1992) 121–115. D. J. Mastronarde, Euripides Phoe- nissae (Cambridge 1994) ad 690, analyses passages where an actor remains on stage during a stasimon. But the insuperable problem of “exceptions” to hypoth- esized dramaturgical norms is raised by W. Allan, Andromache and Euripidean Tragedy (Oxford 2000) 75. Greek, Roman, and Byzantine Studies 43 (2002/3) 341–372 © 2003 GRBS 342 SPLIT-VISION there were undeniably scenes where the poet simply allowed a character to be on the stage for some time without apparently anything to do. This echoes the famous criticism of Aeschylus at Ranae 911ff where this silent stage-presence is claimed to be both something typical of older drama and also indicative of gauche dramaturgy.3 Another example would be Pylades in Cho- ephori, who is present for hundreds of lines throughout the play but speaks only three lines. Technical limitations can prevent speech. Sometimes, with three actors on stage, there may have been further technical considerations against allowing all three to speak together, since presumably it was more difficult for an audience to follow such dialogue among masked actors in a theatre with spectators on three sides. Undoubtedly, then, there were times when non-speaking actors of some importance were on stage for periods longer than an effective dramaturgy would now approve. Spitzbarth concluded that true double-action was rare, and that normally an actor simply waited until his turn came to speak, much like the soloist in an orchestra. During this time, he was doing nothing. On the other hand, reflection suggests that no actor on a stage is doing nothing. Merely by being visible, he is having some effect. Crowd scenes are textually silent in tragedy—so much so that we cannot determine what they looked like, or even if there was a crowd—but do have an effect, as does the on-stage presence of the children of Heracles in the Heraclidae. In this wider sense, an audience often could have several objects for its attention at one time. In Old Comedy of course there was much comic business engaging the audience directly, but which cannot now be reconstructed; there, silence in our comic texts notor- iously conceals much visual appeal to the audience. A recent 3 O. Taplin, “Aeschylean Silences and Silences in Aeschylus,” HSCP 76 (1972) 57–97, discusses the long silences of Aeschylus. WILLIAM J. SLATER 343 valuable treatment4 of such silences and their consequences was that of W. G. Arnott, who limited himself to scenes in Menan- der. He usefully draws attention (77) to the amount of move- ment and preparation necessary before anything is actually said on stage, e.g. in Misoumenos, even though there is no known example in New Comedy of an extended silence such as that of Cassandra in Agamemnon or Orestes in Euripides’ Orestes. He concludes that “virtually all of the stage action has to be inferred from the words spoken by the characters” (75). He suggests that “from these remarks the play’s producer is expected to work out the stage actions involved.” A further in- ference could easily be drawn that one could from the text work out the stage action of New Comedy. This would surely not be justified. We do not know what Daos is doing in his 63 lines of silence (Dyscolus 301–363) nor do we know how Gorgias eaves- dropped before 821,5 or when exactly he would have been visible to the audience. In fact, as Arnott himself says, there are silent actions that “have to be inferred from the context” (77). In short there are actions not indicated by the text but to be hypothesised on more fragile grounds. Tragedy, whose texts notoriously do not openly acknowledge the auditorium, will be even more difficult.6 Some examples It cannot be denied that there are tragic passages where the prolonged silence of a character seems awkward, and his or her 4 W. G. Arnott, “Visible Silence in Menander,” in S. Jäkel and A. Timonen, edd., The Language of Silence I (Turku 2001) 71–85. S. Montiglio, Silence in the Land of Logos (Princeton 2000), surprisingly does not deal with these issues; but she argues against silences during performance, where no actors speak. J. J. Farber, “Preplay Ceremony and Euripides,” Text and Presentation 12 (1992) 23–27, draws attention to initial stage directions in Euripides [reference from Prof. C. W. Marshall]. 5 All these passages are noted by Arnott. 6 This is not to deny that in a sense “il pubblico sia sempre inscritto all’ in- terno del testo dramatico,” as D. Bertolaso reminds me. 344 SPLIT-VISION presence prima facie unnecessary, but usually capable of some explanation, e.g. in the one play Eur. HF we have Lycus at 252– 332, where his attendants probably were involved in some action,7 Heracles at 1164–1229 (but he is veiled and mourning), Amphitryon at 1214–1404. Antigone’s long silences at several places in the difficult OC must be due to her practical role as a guide for the blind Oedipus; she cannot leave. Actors too who remain during choral songs sometimes have some obvious reason to do so, such as being suppliant at an altar or unable to move. For other passages8 one may offer similar solutions with some confidence. Adrastus in Eur. Supp. 262–734 is almost totally silent, but then he is a mourning suppliant; moreover, he is forcefully told to be quiet by Theseus at 512ff, and he mean- while silently learns some wisdom.9 Clytemnestra’s silence10 at IA 1368–1433 is noticed by Iphigeneia at 1433, i.e. is drawn (in strict logic, unnecessarily) to the attention of an audience, and should presumably be construed as an actantial sign of despair. At Eur. Phoen. 1355ff the long silence of Creon is more difficult to explain; but he may have gone out at some point, or the text is corrupt. On this passage Mastronarde notes justly, “The im- portant question is whether Creon’s silent presence is awkward, noticeable or distracting,” but concludes none too satisfactorily, “not, I would say, to an audience focusing its attention where the dramatist wanted it to be.”11 A competent dramatist would normally try to ensure that a potential distraction did not 7 Speculations by Wilamowitz and Bond on this possibility are noted by G. W. Bond, Euripides: Heracles (Oxford 1981) ad 268ff. 8 F. L. Shishler, “The Use of Stage Business to Portray Emotion in Greek Tragedy,” AJP 66 (1945) 377–397, at 388–389 collected several passages, where she felt that silence indicated strong emotion, including apprehension. 9 C. Collard, Euripides: Supplices (Groningen 1975) II 209. 10 So Shishler (supra n.8) 388. 11 Mastronarde (supra n.2)) 514. Shishler (supra n.8) 388 thought apprehen- sion was indicated. WILLIAM J. SLATER 345 exist,12 and would not leave his play at the mercy of an Athen- ian audience’s dubious ability to focus, which depends precisely on not having distractions. Explanations of texts that suggest performance solutions are to be found even in Alexandrian scholia13 as in modern com- mentaries, and they are derived largely from reading the texts, and occasionally perhaps from reminiscences of performances, though presumably never the originals, and as such must be re- garded as practically worthless. An excellent example is given by pseudo-Demetrius, On Style 190, who offers suggestions about the possible actions of Ion with his imaginary(?) bow and imaginary arrows14 at Eur. Ion 15–81; and while the bow and arrows of Euripides’ hallucinating Orestes are likely imaginary, Philoctetes’ weapon in Sophocles is not. Such explanations of performance are explications de texte, to use Handley’s phrase,15 but they are, or should be, at the same time performance-based, in the sense that a performance must be visually imagined, as indeed Aristotle (Poet. 1455a25) demanded of the poet. Ancient drama is helpfully full of conspicuous “implicit stage direc- tions,” many of which would now be omitted by anyone writing for stage performance and adding extraneous stage-directions; and it was not unreasonable that a more text-based and less 12 The dangers of distraction of audience attention was known to Aristotle, e.g. Rhet. 3.1408b21ff. 13 E.g. Aristophanes of Byzantium fr.390 Slater. The discussion of T. Falkner, “Scholars versus Actors: Text and Performance in the Greek Tragic Scholia,” in P. Easterling and E. Hall, edd., Greek and Roman Actors (Cambridge 2000) 342–361, notes rightly that occasionally scholiasts were influenced by what they saw on their contemporary stages, but he underestimates their inventive- ness and readiness to assert what they did not know. 14 D. Kovacs, “Rationalism, Naive and Malign in Euripides’ Orestes,” in J. F. Miller, C. Damon, K. S. Myers, edd., Vertis in Usum: Studies in Honour of E. Courtney (Munich 2002) 275–286, makes the distinction. 15 E. Handley, “Acting Action and Words in New Comedy,” in Easterling/ Hall (supra n.13) 165–188, at 169. 346 SPLIT-VISION performance-based commentary16 on ancient tragedy has been usual. Nonetheless, I believe that the judgement of Spitzbarth that actors simply “wait their turn,” and so sometimes do not require to have any actions allotted to them, underestimates the growing sophistication of the stage,17 and I prefer to argue that at least from the time of the Oresteia, when three actors had be- come possible, there was an awareness of the dramatic power to be derived first from immobile silence, and then increasingly from action without speech. That an actor’s silence can be a sign of “archaische Unbeweg- lichkeit” as Spitzbarth claimed (77) is not really a helpful categorization, even though it is very true that there was little naturalism about ancient performance in speech or acting, and much that was artificial and dictated by convention.18 It is of course obvious that the chorus at any time could respond to the words of the actors by gestures,19 which would emphasize or contradict the words heard, and indicate anger, sympathy, 16 It is not my intention here to address this wider issue. See P. Pavis’ Diction- ary of the Theatre, transl. C. Schantz (Toronto 1998) s.v. “Semiotics.” Much theoretical discussion can be found in L. Edmunds, Theatrical Space and Historical Place in Sophocles’ Oedipus at Colonus (Lanham 1996), esp. 15–38, “Theorizing theatrical space.” 17 A good survey of performance criticism of tragedy will be found in H. Altena, “Text and Performance: On Significant Actions in Euripides’ Phoe- nissae,” ICS 24–25 (1999–2000) 303–321. 18 See K. Valakas, “The Use of the Body by Actors in Tragedy and Satyr Play,” in Easterling/Hall (supra n.13) 69–92. 19 Gesture in ancient acting, rhetoric, and art, all related, has its own large bibliography. I note H. Ortkemper, Szenische Techniken des Euripides: Unter- suchungen zur Gebärdensprache im antiken Theater (diss. Berlin 1969); A. Boegehold, When a Gesture Was Expected (Princeton 1999); some articles in J. Bremmer and H. Roodenburg, A Cultural History of Gesture from Antiquity to the Present Day (Oxford 1991); and esp. R. Green, “Towards a Reconstruction of Performance Style,” 93–126, and E. Fantham, “Orator and/et Actor,” 362– 376, in Easterling/Hall (supra n.13). M. Pedrina, I gesti del dolore nella ceramica attica (VI–V secolo a.C.) (Venice 2001), is the most recent study in art, with large bibliography. S. Bertelli and M. Centanni, Il gesto, nel rito e nel cerimoniale dal mondo antico ad oggi (Florence 1995), is a useful complement, and P. Pavis, “Problems of a Semiology of Theatrical Gesture,” Poetics Today 2 (1981) 65–93, deals with the semiotics. [My thanks to Daria Bertolaso for much help.] WILLIAM J. SLATER 347 grief, agreement, or surprise, and in a wider sense the chorus, and its reactions, have often been considered an extension of the audience.20 These gestures would form a moving backdrop to any speech, especially if they were synchronized. Likewise, in a long messenger speech, one will expect that chorus and actor(s) will react by gesture or body movement to the details of the description, usually of horrible deaths, that the messenger brings, as when Hecuba has to endure Talthybius’s description of her daughter’s sacrifice. Here I leave aside such undoubted reactive gesture, to consider other aspects of secondary per- formance. Nor is it always useful to ignore the silent actor, by arguing that this situation is enabled by the technical two- or three-actor rule which forbids an actor under certain circum- stances to speak; a poet is after all under no compulsion gen- erally to have a potential speaker on stage doing nothing; he must choose to do so and live with the consequences. In the same way, it has long been observed that Euripides tends to avoid a triloquium, precisely when there are three actors on stage;21 but that still leaves the question open why the third person would be on stage at all. Rather, instead of explaining away or ignoring the silent perfomer, one may usefully ask if the presence of a non-speaking actor could have a dramaturgical function, and it is argued below that in some cases, at least, it was indeed intended to produce a deliberate effect. It cannot be denied that mere silence can of course be effective as performance, and even an actor’s lack of reponse can on oc- casion be significant action for an audience,22 which expects one. Andromache’s presumed23 presence with her son on stage 20 R. Barthes, Écrits sur le théâtre (Paris 2002) 44, “une sorte de prolongement spatial,” etc. 21 See n.58 infra. 22 The point is made to me by Prof. C. W. Marshall. 23 See Bond (supra n.7) 219 and 249; D. J. Mastronarde, Contact and Discon- tinuity (Berkeley 1979) 100; Allan (supra n.2) 75 and 227 for the controversy. 348 SPLIT-VISION after Andr. 1041 can be argued to increase the pathos. Electra (Soph. El. 817) provides a reason why she will remain silently outside the house;24 as with Clytemnestra in Agamemnon, the dramatic effect of her silent and immobile presence beside the door was enough justification; she is after all a dominating and menacing character. This “distracting effect”25 can be em- phasised. The most notable example, Cassandra, is on stage in Agamemnon from the time she arrives with Agamemnon; she is silent and nameless in the orchestra, visible to all raised up in the cart, and not easily considered unobtrusive, while Agamem- non is invited to enter. The audience do not know her name, and eventually it may forget to look directly at this blackclad woman as it concentrates on the fateful events before the door and the deception of Agamemnon. She is named when Clytem- nestra reappears to call her in. But she is still silent, and now attention has been deliberately focussed on her silence, when she is left alone with the chorus. At this point, after 260 lines, she bursts out with the loud screams of wailing invocation and prophecy that mark her slow and indirect way towards the door where by contrast with Agamemnon, she knows that her death awaits. No one can doubt then that Aeschylus has used the long silence for the pity-and-terror effect produced by this Kontrastfigur when she does eventually sing out. The very refusal to speak, to which the poet draws the audience’s attention, means that the shock is all the greater when she does; but it is very possible that Cassandra shows increasing signs of violent agitation just before she sings, an extreme example of the “distracting effect.” We can be reasonably sure that the dramatist wanted the emotional §kplhktikÒn produced by this 24 Spitzbarth (75) calls these examples “blosses Verharren.” Note that Electra notably by contrast goes off at the end of the fourth epeisodion to accompany the murderers, and so can report directly on their progress and eventual success when she re-enters after the short choral ode. 25 More on Cassandra’s acting in Valakas (supra n.18) 80–81; O. Taplin, Greek Tragedy in Action (London/Berkeley 1978) 59–60. WILLIAM J. SLATER 349 dramaturgy. But he also knew that the audience are intrigued by Cassandra even when she says nothing and when she is ig- nored.26 Why is she not named? Who is she? Will she speak? The poet intends these questions to be asked. We cannot now calculate sight lines meaningfully, but since Cassandra is in the cart all the time, and Agamemnon cannot have descended any- where but in front of his own palace, it follows that she was easily visible, and so presumably deliberately left in the or- chestra as a silent secondary enigma, just as Clytemnestra is left by the door to emphasize its importance in the play and her role as its keeper. Silence in Aeschylean dramaturgy can already produce a nuanced and subtle overtone and deliberate back- drop to the primary action. It is therefore reasonable to assume that the necessarily dynamic progress of later stage business would improve on his efforts in this direction. Secondary background But there are also passages where two actions occur, and where we are informed of the secondary action—e.g. the judges in Aeschylus’ Eum. vote, while the chorus and Apollo debate for twenty lines. While Orestes in IT (77–103) informs the audience of the situation, Pylades presumably explores the area as he says he will do (75).27 If however a dramatist can gain an effect with the relatively sparse means available in the theatre of Dionysus, the natural dynamics of stage production require that such resources be exploited in ever more complex ways. Normally this doubling of stage action would be avoided, since it may result in lack of audience concentration and become an 26 Prof. C. M. Marshall asks why e.g. Pylades in Electra would not be just as intriguing. I should answer that Pylades, like many others, would be—or could be made—less obtrusive beside the stage building than Cassanda in a cart in the orchestra. The shadow of the north facing skene-building would in the circum- stance of Athenian spring sunshine allow for a virtual disappearance, which raises interesting possibilities. 27 “Searching” is often a primary but also a secondary action, Valakas (supra n.18) 74. 350 SPLIT-VISION annoying distraction; non-speaking actor(s) or attendants will usually avoid attracting such attention for that reason, and seek to be unobtrusive. A chorus may or may not provide back- ground gesture. But suppose that the character is not un- obtrusive; then the effect of making the audience keep its eyes on two separate acting locations instead of one can be to puzzle and perhaps unsettle; and occasionally this may be the aim of a dramatist. But any investigation of such “split-vision” soon runs into the problem of what Taplin calls “inferred stage directions,”28 where we infer action which is not immediately proven by the words of the text. First, however, one must concede that not only characters or chorus can provide this unsettling background. The two statues on the stage in Hippolytus, one of Aphrodite ungarlanded and the other of Artemis with a garland, are permanent reminders of the central combatants of the play, and that the characters are playing against a background of divine forces.29 More telling is the backdrop of the Iphigeneia in Tauris. A chorus of captive Greek girls form an attractive chorus before the temple/palace of the king. What is unsettling here is not just the human sacrifice that in her opening speech she claims that she has to practice on strangers; this we might forget or pardon from some- one who is herself normally sacrificed by her father. But we are not allowed to forget the human skulls that decorate the façade, and which Pylades and Orestes point out to the audience as soon as they enter (75ff). These can be taken to have existed, even if only painted,30 though one cannot prove this to the 28 O. Taplin, “Opening Performance: Closing Texts?” Essays in Criticism 45 (1995) 102–105; criticism in Valakas (supra n.18) 74. 29 Cf. Allan (supra n.2) 243 on Andromache: “The statue of Thetis is men- tioned at various points: the goddess’s presence is suggested by Andromache’s appeal to Hermione (‘Do you see the statue of Thetis looking closely at you?’ 246 cf. 260); Wiles [=Tragedy in Athens] 1997: 201 remarks ‘… it is a crucial determinant of how the audience will respond.’” 30 The parallel of the masks in Aeschylus’ Isthmiastae (fr.78a) suggests that such stage-properties were not entirely imaginary. M. Cropp, Euripides: Iphi- WILLIAM J. SLATER 351 satisfaction of those that prefer a minimalist position; if we accept their existence, these tokens of ungreek savagery, now explained and highlighted, remain throughout the play both to produce a sensation that we can well feel subverts the claims of Greek civilization that encompass Iphigeneia, and to lend a piquant touch to Thoas’ astonishment (1170ff) at the barbarity of Greeks who (like her brother) would kill their own mother. We are not allowed to forget what Iphigeneia has been doing. And yet, there was no need whatsoever for Euripides to put those skulls there to remind us—or, with a minimalist position, why Euripides should have mentioned them at all—and it may be already suspected that it is a quirk or indeed a principle of his dramaturgy to utilize more effectively the background to the primary action. Silence, as was remarked earlier, may indicate many emotions such as grief31 or anger32 or madness,33 or different kinds of ill- ness, as when the sick Phaedra is snapped out of her veiled and immobile silence on her couch when the nurse mentions the name Hippolytus (309). But it more interestingly can indicate appre- hension, in which case the audience will likely be studying the character who is silent but not necessarily immobile, as well as ——— geneia in Tauris (Warminster 2000) ad loc., suggests a display of heads had parallels in Sophocles’ and Euripides’ Oenomaus. But various degrees of min- imalism have also been generally urged, see e.g. D. Wiles, Tragedy in Athens: Performance Space and Theatrical Meaning (Cambridge 1997) 161–162, with further references. 31 Hecuba’s distressed state of mind is shown long before she speaks when Troades opens with her sorrowful figure stretched upon the ground (1–98); later in the play too after sinking to the ground she lies outstretched throughout a long speech of her own (466ff), and after being led to a low pallet (506–508), throughout a long choral song. In Hecuba 438–501 she lies grief-stricken on the ground throughout the choral song and until Talthybius enquires of her where- abouts (486–487) and rouses her. Adrastus’ distress is shown as Supp. opens by his lying on the ground; Iolaus in Heracl. 603ff is led to a hedra and lies there in sorrow. 32 Hipp. 911, where afterwards Theseus still refuses to address or look at Hippolytus; or Oedipus to Polynices at Soph. OC 1271. 33 Bond (supra n.7) ad 930–1009. 352 SPLIT-VISION keeping an eye on the speaker. Certainly there are several passages where silences may reasonably be taken to indicate apprehension, and imply action.34 A good example can be found in the Oedipus Tyrannus. At 1072 Jocasta exits after des- perately begging Oedipus to stop asking for information; she already knows the truth. But she already knew at 1050, when she was forced to break her long silence by a question from her husband, and must therefore have known even earlier while the messenger was being interrogated and when she was saying nothing. But was she doing nothing for about fifty lines? That is here unlikely. She will have had to indicate during her silence in some way her gradual apprehension of the truth, as the mes- senger revealed the details she would recognize;35 logic seems to require that the audience must be apprized of her realization before she speaks, and since no words are recorded, she must have used body language—veiling her head, raising her arms, putting her hands to her mouth, or simply retreating back from Oedipus. These gestures act as a visual counterpoint to the words of Oedipus, who does not see her; but the audience do, and her gestures, revealing her new knowledge, subvert for them his demands to know. The alternative is to suppose with Spitz- barth that Jocasta did nothing at all, save to wait unobtrusively until Oedipus spoke to her. 34 Creon in Phoen. 960–961, hearing the prophecy that means his son’s death, maintains a silence that helps to show his apprehension; the silence of Orestes shows apprehension also in Eur. Electra (295–363). Euripides’ Ion at line 582 probably indicates by his silence that he is not keen on Xuthus’ ideas for his future. Perhaps silence means shame for Phaedra in Hipp. 297ff, or Heracles be- fore Theseus in HF 1173ff. Polyxena at Hec. 216–340 watches the debate over her life. It should be observed that “silence” is not something that will figure in Altena’s “speech acts” (supra n.17). 35 D. Bertolaso argues (by letter) that there were three distinct signs, at 1026, 1034, and 1042, to which Jocasta could react. Eric Csapo reminds me that this is what is revealed by the Capodarso painter’s depiction of exactly this scene, see E. Csapo and W. Slater, The Context of Ancient Drama (Ann Arbor 1995) 63 and plate 4A. WILLIAM J. SLATER 353 We face a similar dilemma in the departures of Deaneira (Trach. 813) after the long speech of Hyllus, and that of Euryd- ice (Ant. 1245), who silently leaves after the messenger speech, though her silence is notably an object of speculation to the others. (It is a pity that we do not know if Phaedra was on stage or not at Hipp. 600–680; she speaks as though she is going off to kill herself, but then unexpectedly reappears, and we cannot use this example.) There is no way to prove absolutely that gesture was used, since this secondary action, apart from leaving, cannot easily for reasons of plot be indicated in the text, a consideration that often applies. We do find a similar scene in Hippolytus (866–873) in miniature.36 As Theseus opens and reads the letter Phaedra has left, the chorus seems to indi- cate that they recognize its contents will be deadly; this is turn suggests that his gestures will have alerted them to its content. Teucer also says nothing as Odysseus and Agamemnon battle over the body of Ajax (Aj. 1315–1381), but he would, one as- sumes, be supporting Odysseus against Agamemnon with ap- propriate gesture. A similar situation appears during the silence of Neoptolemus at Soph. Phil. 974–1074, when Odysseus enters brutally to remove the bow from Philoctetes. Neoptolemus can be suspected of demonstrating pity for Philoctetes, as he admits in 1074, and Philoctetes even remarks on his silence (1066), something that should alert us and the audience to its importance. During such a dramatic scene Neoptolemus, not after all an unobtrusive figure, ought not be displaying dis- interest. But just how Neoptolemus showed pity, concern, or indignation cannot now be determined, though it certainly will not do to argue that we need not consider the possible sec- ondary actions just because he does not speak, or because his actions are not mentioned in the text. 36 The text is regrettably very uncertain. The comparable scene is in IA 34ff, where Agamemnon is trying to write his letter. 354 SPLIT-VISION But there are more interesting possibilities, where the effects are more profound.37 In Trachiniae, the first messenger/Old Man remains silent on stage for over one hundred lines while Lichas gives his spurious account of the history of Heracles. Only after Lichas departs does the messenger/Old Man come forward again to denounce Lichas as a liar. The text does not imply that the messenger did anything, and Reinhardt38 wondered why we needed this unnecessary play of truth and falsehood at all, im- plying that Sophocles’ technique was at an early stage. Webster suggested that the Old Man was moved to indignation by Lichas’ speech,39 and Parlavantza-Friedrich thought it “denk- bar” that the Old Man might be using gesture to show im- patience.40 Taplin, while admitting that the silence is “peculiar,” turns this view into a parody: “that the Old Man has been making frantic gestures throughout the intervening time,” and so is enabled to dismiss it; he thinks that the Old Man “stayed unobtrusively.”41 His only argument however is the circular one that the intervention of the Old Man at 335 “must be a surprise.” Seale is more open to the suggestion of gesture, and points out rightly that Taplin can hardly argue for the scene being peculiar at the same time as he maintains that the Old Man is unobtrusive.42 In fact, since the Old Man is listening to a version of the story, which he will be shown to reject, it would 37 Obviously I cannot deal in detail with the literature on all the passages mentioned here; but the example I give from Trachiniae suggests that more care needs to be taken in commenting on the issue. 38 K. Reinhardt, Sophokles4 (Frankfurt am Main 1976) 52. 39 T. B. L. Webster, Introduction to Sophocles2 (London 1969) 91ff. 40 U. Parlavatza-Friedrich, Täuschungszenen in den Tragödien des Sophokles (Berlin 1969) 29 and 80–81. 41 O. Taplin, The Stagecraft of Aeschylus: The Dramatic Use of Entrances and Exits in Greek Tragedy (Oxford 1977) 89 n.2. 42 D. Seale, Vision and Stagecraft in Sophocles (Chicago 1982) 194–195 and 213 n.21, the longest discussion known to me. “One extreme is to visualize him almost hopping from one foot to another [he footnotes Parlavantza-Friedrich, who says nothing like this] during the evasions of Lichas and the enquiries of Deianeira.” WILLIAM J. SLATER 355 indeed be peculiar stage-action if he were obtrusively doing nothing, and not in some way indicating to the audience that he disbelieves Lichas, and, for that, no “frantic gestures” are re- quired. It might be tempting to argue that the audience would be alerted to this situation by the mere presence of two messengers, and by the reputation of heralds, but this kind of reasoning that reads the mind of an audience becomes too fragile to support an interpretation. Nonetheless the question is very worthy of dis- cussion in any reading of the play: are the audience alerted or not to the mendacity of Lichas as he speaks? Very different dramaturgies then arise. A general and important objection to the notion of silent gesture was made by Easterling as a rhetorical question in her review of Parlavantza-Friedrich: “is this likely on the Greek stage, or even, given that it is possible, would it be desirable?”43 The answer to the second question is positive, because the scene is then more dramatic; but the answer to the first question is also unequivocally positive, as we shall see. Towards a solution It is impossible to uphold the extreme position that all actions are somehow indicated in the text44 and that consequently, if there is no mention in the text, there is no action; if this were so, we would not need to speculate when Clytemnestra enters in 43 CR N.S. 22 (1972) 19–21, at 21, echoed by Seale (supra n.42) 195: “it is a good question whether this kind of dumb show in which someone is gesturing while another is speaking, was possible in the Greek theatre.” 44 Falkner (supra n.13) 357 writes, though it is unclear to me if he approves of these views: “One [= the ancient scholiast?] can extrapolate to the range of stage business: ‘Stage-directions’ are implicit in the text, and there is no need to go ‘beyond’ the text to determine them. This recalls the position elaborated by Taplin and others that significant stage action is always implicit in or sanctioned by the text; that a thing done is always a thing said.” Perhaps this kind of assertion derived ultimately from the old—and equally unjustifiable— philological claim that (some) poems need no external information, as famously in E. Fränkel, Horace (Oxford 1957) index s.v. “riddles”; but we should rather put the blame on what Allan (supra n.2) 42 calls the endemic “text-bounded- ness” of philologists. 356 SPLIT-VISION Agamemnon and to debate many other questions of perfor- mance. More commonly the comforting fall-back view is pro- posed that all significant action is so indicated. But since the entry of Clytemnestra is significant, this too will not stand with- out exception. In a recent paper Altena has recently denied the validity of this generalization,45 and sought to mediate the sometimes rarified disagreements between Wiles, Taplin, and Goldhill and others on the matter of “significant action,” and readers may be referred to his study for a history of this largely unhelpful controversy. Altena himself regrettably disables his own case with the massive petitio principii (318): “I do not want to suggest the kind of ‘extra unsignalled stage business’ to which Taplin rightly objects.” The use of both “extra” and “rightly” is without justification or definition. There was undoubtedly “un- signalled stage business” that was “extra” to the text, but integral to performance; it remains to be seen if we are justified in postulating such business in any given scene. Perhaps Aeschylus’ original ideas on performance were trans- mitted by the acting profession for a time, though even this is unsure,46 but realistically it would be left up to the actors, as now to us, to decide what was effective with any particular stage and audience, and to make appropriate changes. First, therefore, we have to demonstrate that split-vision dramaturgy was intended and practised, and not just possible; and this must start from those passages where the text actually tells us what the secondary action was. There are many minor and well- known examples indicated by implicit stage direction, such as Ion 967: “why do you hide your head and weep, old man” in 45 See supra n.17. 46 If a tragedy from the Dionysia was reperformed at some deme theatres, it would presumably not be able to afford the same expensive chorus, training, and apparatus as at the Great Dionysia. But I recognize that the evidence for this assumption is practically non-existent in the classical period. WILLIAM J. SLATER 357 stichomythia; this tells us what the old man has done47 and is doing after he has said the one line before. More interestingly, when Jason in the middle of his speech tells someone to “keep quiet” (Med. 550), we are obligated to assume that Medea has uttered a noise of outrage, which is not recorded in our text, and we could infer that she has probably thoughout the speech48 indicated to the audience by gesture her increasingly outraged reactions. When Phaedra (Hipp. 325) asks, “are you using force in seizing my hand,” we can be sure that the nurse is doing exactly that, but we can also assume49 that in addition she has fallen to her knees and adopted some suppliant posture.50 In the monologue Medea 1019–1080, in which Medea determines to kill her children, the speech is marked by changes of emotion connected to the unrecorded movement of the children on stage,51 though of course they do not speak at all. What we need however are incontrovertible examples that cover a longer acting period, and fortunately two well-known 47 I cannot follow Montiglio’s reiterated view (supra n.4) that the ancient stage did not occasionally fall silent. 48 D. J. Mastronarde, Euripides Medea (Cambridge 2002) ad loc., comments: “occasionally when a gesture is highly significant, it is referred to in the words of the speaker, as here.” This is true enough, but cannot logically be used to conclude that all significant gesture is so indicated. 49 On the other hand, the wild assumptions of J. H. Kells, Sophocles: Electra (Cambridge 1973), about unspoken action at Electra 1354ff are an invention designed to buttress an impossible psychological interpretation, an example of what T. Slezak, “Sophokles’ Elektra und das Problem des ironischen Dramas,” MusHelv 38 (1981) 1–21, calls the “Beliebigkeit und Willkür, den die Inter- pretationen zur Elektra nicht selten bieten” (2). 50 More examples of this in the study by J. Gould, “Hiketeia,” JHS 93 (1973) 74–103, at 85–86. 51 “It is not appropriate to inquire too closely into what they [the children] are making of Medea’s cryptic words and shifting moods.” So Mastronarde’s puzzling comment (supra n.11). It would seem necessary rather to inquire whether the “shifting moods” of Euripidean emotional heroines are to be as- sumed to be accompanied by dramatic actions and gesture, which would affect the “reasonableness” of their speech, so often the only philological criterion applied to them. 358 SPLIT-VISION passages in late Euripides come to our rescue,52 and put the matter beyond doubt; they have to be read carefully. In Electra (558–562) the Old Man recognizes Orestes, but he does so by staring at Orestes, who comments on it because it is not some- thing the audience can easily see, and they had best be directed towards it; but the Old Man does do something the audience can easily see for the scene to work, viz. pacing round him in silence for a time until he has satisfied himself completely, and both Orestes and Electra comment on it: OR. t¤ d¢ kukle› p°rij pÒda; EL. kaÈtØ tÒd' efisor«sa yaumãzv, j°ne. Here there is plenty of room for telling stage action. The audi- ence must appreciate what the pacing and peering mean, viz. incipient recognition, even while the Old Man is silent; they would have been expecting it anyway. A more original and sophisticated use of the same technique is found in Orestes, where Menelaus’ silent pacing accompanies a fiery exchange be- tween Tyndareus and Orestes. This time the audience can un- derstand the pacing to be a sign of profound pondering (and apprehension) by Menelaus, and the characters afterwards ex- plain it in this sense (632–635): OR. Men°lae, po› sÚn pÒd' §p‹ sunno¤& kukle›w, dipl!w mer¤mnhw diptÊxouw fi"n ıdoÊw; ME. ¶ason: §n §maut“ ti sunnooÊmenow ˜p˙ trãpvmai t!w tÊxhw émhxan«. In this way it becomes easy for the observers to conclude that Orestes’ trust in his uncle’s help is likely to be misplaced; they will after all have hints of Menelaus’ character from other plays, 52 M. West, Euripides: Orestes (Warminster 1990) on line 632. Shisler (supra n.8) 396 says that pacing to and fro is indicated only by Euripides. It is also noted as unique by J. Porter, Studies in Euripides’ Orestes (Leiden 1994) 91 and 169. WILLIAM J. SLATER 359 and correctly see that the hopeful and defiant words that Orestes is uttering are undermined by the action of Menelaus. In both these passages we are told that the characters are pacing because other characters have remarked on this as an oddity. We have therefore in these passages proof that Euripidean dramaturgy could utilize silent background action by one character—secondary action—to contrast with, undermine, counterpoint the spoken words of other characters—primary action. Two actions are evidently sometimes construed to be better than one. Whether Sophocles did exactly the same thing in Trachiniae we shall never know for certain, but there is every reason to consider the possibility. It must be noted that these two examples are not truly im- plicit stage directions, as in the example cited from Ion above, since the audience cannot see weeping or peering, and may have to be informed of it—though there are gestures that could be used to suggest weeping—but they do immediately recognize such a distraction as pacing, and therefore do not need to be alerted to what they have been able to see. These are moments therefore when the primary player(s) are said to recognize the secondary action which the audience have already witnessed, and when the audience are thereby deliberately informed by the dramaturgy that the primary player(s) are possibly coming to the same conclusion that the audience themselves have already reached. In these cases therefore, the dramatist wished to do more than just create secondary action; he wanted to do some- thing more complex, and this is why we now know accidentally that there was secondary action at all! It now becomes reasonable, pace Spitzbarth, to examine whether this conclusion can be applied to the performance of those passages where there is nothing explicitly said in the text about the secondary action, i.e. when the dramatist had no interest in the explicit recognition 360 SPLIT-VISION of the secondary action by the primary players. I offer three examples. A good test case is Helen, where the message by the old seaman, who stops speaking at 621, makes it possible for the complex recognition scene between Menelaus and Helen to follow. Afterwards—eighty lines later (700)—he comes forward again to ask if this woman was really Helen.53 “Menelaus, share your joy with me too… Do you mean we toiled in vain for a cloud?” He had in fact come to this conclusion already at 604, but what has he been doing during this long recognition scene? No answer is provided by the commentaries, because the question is not asked. But it should be. The pragmatic question, with which the seaman breaks his silence, brings the exaggerated theatrical joy of the newly reunited couple swiftly back to earth, by recontextualizing it in the greater framework of the Trojan War. It seems reasonable, then, to suggest that in the time lead- ing up to his renewed participation in the text he was not doing nothing, but rather revealing somehow the attitude that his earlier reaction and eventual question express.54 One possibility open to a director or interpreter is that he paced up and down pondering, so suggesting apprehension and anxiety, perhaps twisting his cap in his hands, and creating a subversive back- drop for the lyrical reunion he has promoted. A powerful argu- ment here is that this explanation answers the question why after all Euripides kept him on stage for eighty lines, in order to extract precisely this special effect, which Euripides clearly sought.55 Otherwise it would have been simple and desirable to 53 The role of this Therapon as a tertius iudex in the scene is well explained by R. Kannicht, Euripides: Helena (Heidelberg 1969) II 201–202. 54 Prof. C. W. Marshall suggests that a frozen silence on the part of the old seaman would be another option available to a director. I merely affirm that Euripides could not reasonably leave him on the stage in order to be invisible. 55 Allan (supra n.2) on Andromache 60 and Mastronarde (supra n.11) on Phoen. 690 note examples of actors on stage during choruses, but there are many disputed passages. WILLIAM J. SLATER 361 have him disappear after his initial speech. He is there for the sole purpose of counterpointing the joy of recognition, and in a sense the real recognition is the one he alone expresses. That he stays to comment as a simple decent man on the deceptions of gods and prophets and the vanity of human suffering should scarcely be construed as comedy,56 though as often in Euripides, it could be played that way; it should be better seen as a tragically moving moment of disillusion. Again, one can prove nothing beyond doubt; but the interpretation is much more than a possibility, and if so, the dramaturgy of Euripides is more complex than we are encouraged to assume. In the same play, Menelaus, dressed in (comically?) wretched rags, is on stage from 1085 to 1251, all through a choral ode, then, unseen, endures threats by Theoclymenus to kill him, and the subsequent stichomythia between Theoclymenus and Helen. All three scenes are of immediate personal concern to him. What is he doing? During the choral ode he can be sitting, as he proposes (1084), by the primitive tomb on centre stage, but he must then, as Theoclymenus and his group enter from the side (1165), be in hiding apparently behind it, though perhaps the audience can see him. During his silence in hiding, he must first endure, like Odysseus among the Phaeacians (Od. 8.83ff), the long story of his own travails as sung by the chorus, and then the death threats of Theoclymenus, then after his discovery (1203) the announcement by Helen that he is dead. It seems difficult to think that he does not react to all of these. He can 56 So notably W. G. Arnott, “Euripides’ Newfangled Helen,” Antichthon 24 (1990) 1–18, at 16–17, who treats the Old Man as a Polonius, a “garrulous windbag,” whose sole raison d’être on the stage is apparently to amuse. Much of his speech has been deleted, on the same basic grounds of irrelevance. As I argued in “Gnomology and Criticism,” GRBS 41 (2001) 99–12, I hold this to be a profound error, the consequence of misplaced academic disdain for such gnomic musings, which were greatly popular with ordinary people, like the sea- man himself. One need only consider how many of the audience had suffered on Athenian naval expeditions under commanders of dubious competence and even more dubious motivation, to realize the effect of such words. 362 SPLIT-VISION scarcely ignore a choral ode which is all about his own story; he can accompany it with appropriate gestures of sorrow. How- ever, he can certainly not ignore the unexpected entrance and threats of Theoclymenus; he must display alarm and conceal himself, for Helen describes him as “cowering” (1203), which he does not need to do as the chorus sing. If he is visible to the audience, the scene could be played in any register from serious to comic, but how he reacts after his discovery is a more deli- cate question, for there is plenty of room here for stagy business as Helen in front of her silent husband explains the death of Menelaus and her new readiness to marry Theoclymenus. Two points are worth making here. Once again there was no need to leave Menelaus on stage all this time, save to show his different reactions to the primary action. Euripides seems al- most to anticipate this possibility by an implicit stage direction, for Helen commands Menelaus to stay on stage while she goes off (1085); future directors will be glad to know this, for the couple could quite easily have gone off together, and entered together. They do not, because Euripides wanted the drama- turgical effect of the silent reacting Menelaus. The second observation to be made regards the question of tone. It is objectively possible to play this scene, like a number of others in Euripides,57 as comic or serious or somewhere in between, and it 57 B. Seidensticker, Palintonos Harmonia (Göttingen 1982); A. N. Michelini, Euripides and the Tragic Tradition (Madison 1987); J. Gregory, “Comic Ele- ments in Euripides,” ICS 24–25 (1999–2000) 59–74, and many others deal with Euripides’ apparent defiance of tragic propriety. Arnott (supra n.56: 13) notes that scenes could be played “straight and seriously or for laughs,” but con- cludes with dubious logic for Helen 420ff that the “ambience is unmistakably humorous.” There are too many gradations between high tragedy and “laughs” to assert this. More circumspectly he had (“Red Herrings and Other Baits,” MphL 3 [1978] 1–24, at 4) memorably spoken of passages “where Euripides seems deliberately to be aiming a brick at the brittle window of dramatic illusion.” The word “seems” is to be emphasised. A. P. Burnett, Revenge in Attic and Later Tragedy (Berkeley 1998) 247–272, misreads the Orestes as a farce, an extreme case. Good remarks on the problem of genre are in D. J. Mastronarde, “Euripidean Tragedy and Genre: The Terminology and its Problems,” ICS 24– 25 (1999–2000) 23–39. WILLIAM J. SLATER 363 is an enduring puzzle that this should be so. There is however nothing in the text that alerts us as to how Euripides might have wanted it to be played, and so it is left entirely up to any actor or director to determine the tone, a matter one would have thought of the greatest concern to a playwright. Implicit stage direction would here have been most helpful as guidance, but the conclusion must be that neither significant stage action nor significant tone is competely readable from our text. As in this Helen scene, Orestes 1013–1245 has three actors on stage:58 and the scene is divided into two sections where only two speak, followed by a brief interlacing triloquium. In each sec- tion we can ask what the third non-speaking person (Pylades, Electra) was doing, even though no indication is furnished by the text. An answer is suggested by the observation that at the end of each section the non-speaking person comes up with a sudden bright idea, Pylades at 1069, Electra at 1177. We seem to need some indication of body language that a bright idea is forthcoming, and once again pondering indicated by pacing is the likely candidate,59 the primary action then broken by an excited indication that some plan has been excogitated by the silent character. But we have to be careful: not every sudden intervention of this sort would need to be signalled. A similar passage is found in the Electra of Euripides, where Electra is silent while the Old Man and Orestes discuss how to attack Aegisthus, but, when they start discussing Clytemnestra, she 58 C. W. Willink, Euripides: Orestes (Oxford 1986) 259, says that this kind of scene—where two speak and a third says nothing—is not uncommon and cites H. Strohm, Euripides: Interpretationen zur dramatischen Form (Munich 1957) 43, and criticizes Taplin for suggesting that it is. 59 Fantham (supra n.19) 370 translates the terms at Quintilian 1.11.3 vultus manus excursio as “voice [a slip], arms [but only hands will be meant] and pacing around”; but in a comparison with comic acting this is perhaps not the meaning of excursionibus, but rather (comic) “digressions,” its usual meaning, and so “ad-libbing”; at Cic. Orat. 59 it is a military metaphor, and means a move- ment to the front by the speaker. The term ambulatio (Quint. 11.3.126) could be applied to Menelaus. 364 SPLIT-VISION suddenly bursts in (647) with the statement that she will attend to her mother’s murder, and then explains her stratagem. In this case we do not expect her to show any sign that she has been doing anything more than listening intently, and she has not even been addressed. Her intervention is most effective60 if her remark is delivered as spontaneous. This criterion of dramatic effectiveness, of course, is not with- out its problems, and, when applied to performance questions, usually brings dramaturgy into collision with philology. What, for example, actually happened at Orestes 380? There Menelaus, with his companions, has entered on the stage, and for twenty- four lines he has addressed the stagebuilding and the chorus, without seeing the sick Orestes on the stage, and indeed he informs the audience that he would not recognize him anyway. According to Martin West’s stage direction,61 Orestes then, as Menelaus finishes speaking, “runs forward from his bed and crouches before Menelaus.” Why? The philologist’s eye has no need to imagine Doppelvorgänge, and assumes as a default position, as did Spitzbarth, that a character acts when he has to speak, i.e. at the point where the textual critic has to take notice of the character’s words. But would any director dealing with this text not demand a different scenario? that the sick Orestes gradually wake up at the sound of Menelaus’ entry, spot Menelaus as he is talking, recognize him, and crawl slowly (he is an invalid!) up behind him, and grab the unsuspecting Menelaus’ knees? For twenty glorious lines, the audience can see what is going to happen, and knows already the effect that Orestes’ frightening mask is going to have in this pseudo- epiphany. The dramatic effectiveness of such a scene surely de- serves consideration, if we wish to understand why Euripides’ dramaturgy had such an effect on the future of drama. 60 M. Cropp, Euripides Electra (Warminster 1998), comments ad loc. : “strik- ing after 62 lines of silence.” 61 West (supra n.52) 85. WILLIAM J. SLATER 365 Proposal: the Electra of Euripides The admitted danger62 therefore is that when we are without textual support there is nothing to delimit clearly the invention of historically possible dramaturgy. On the other hand it has been argued above that a refusal to consider the possibility of important secondary dramatic action is unjustified, especially if made on the claim that all important action is noted in the text. Consequently it would be the duty of a commentator on the text to consider the possibility of actions that are not directly indicated but which could have a bearing on the overall inter- pretation of the text. What limitations can be imposed on such an enquiry? how far should the search for dramaturgical action go? One play that seems to me to raise these issues in a way that is exemplary is Euripides’ Electra, which, as a Euripidean melodrama, can be expected to be among the more sophisti- cated and inventive products of the fifth-century stage; and I conclude with an interpretation on the lines I have been follow- ing, to suggest that a full interpretation of the text should and perhaps must take into account aspects of performance. First: what we do not know. Any dramatic audience in an- tiquity immediately knew something significant that is denied us: the age and character traits of the principal exponents of the story. This was obvious from the masks and clothing chosen by the poet as director, which well before the end of the fifth century would have been increasingly realistic, as vases show them, though this is not of course a completely certain de- duction. Just how much an audience could see depended on their distance from the actor, and the quality of their vision; we could certainly agree that the primary traits of age and sex would be priorities,63 but also poverty, class, squalor, and much 62 I have ventured to run this risk in an article in Dioniso, forthcoming 2003, “Orestes as Hamlet?” 63 C. W. Marshall, “Some Fifth-century Acting Conventions,” G&R 46 (1999) 188–202, is a valuable overview of recent work. He argues strongly for min- imal if important information such as age and sex being derived from the mask 366 SPLIT-VISION else that could be signalled with no great need of sophistication. Their central importance for performance is recognized by Aristotle (Poet. 1450b17). Hermione in Andromache is visibly a teenager,64 even if a married woman, and that immediate aware- ness is vital for appreciating the dramatic logic of her appar- ently inconsistent behaviour. We know from the text much of the appearance of Electra and the teenage Orestes in Orestes, but we do not know what led the chorus to describe Menelaus’ appearance as oriental luxury (350). In the Electra of Euripides we know a good deal about Electra’s appearance but little of Orestes’. Vase painting is little help. Artists like the Dokimasia painter portray Orestes heroically as an armed ephebe, but the dramatic story itself requires him to be in the guise of an in- nocent and apparently unarmed traveller. Likewise artists perhaps surprisingly do not like to portray the filthy mourning Electra of the stage.65 Yet if we have recourse in desperation to Pollux’s list of later tragic masks of young men (Onom. 4.133–142) we find ourselves unable even to decide whether Orestes should be the “delicate”—“blond with ——— itself, and that lesser variations were “not theatrically significant” (191). I regret that I cannot altogether agree: I should argue that the notably frightening mask of the “mad” Orestes in Eur. Or. is a focus of the theatricality of the first quarter of the play, and that it was unique. If the Hellenistic theatre needed the surprising variants of Pollux’s tragic masks—alleged to be “latently” mini- malist (200)—one cannot convincingly argue on theatrical grounds that such variants were not significant for a classical audience or playwright. 64 Allan (supra n.2) 105 describes her rightly—“youthful brashness,” 107 “unstable young woman,” “insecure young wife”—all of which makes it un- reasonable to describe (99) Wilamowitz’s very appropriate description “child- ish vanity” as “too dismissive,” especially when Andromache deliberately points up, as he says, this juvenile aspect of Hermione’s behaviour (at 184, 192, 238, 326). This is clearly how young aristocratic brides were thought generally to behave. Orestes’ posturing in Orestes (esp. 640ff, 1016ff.) is a first-rate portrait of an insecure teenager striking inconsistent “attitudes” which are nonetheless perfectly comprehensible within a stereotypic teenage character. This is how teenage youths were imagined to behave. M. Heath, “Sophocles’ Philoctetes: A Problem Play?” in J. Griffin, ed., Sophocles Revisited (Oxford 1999) 137–160, at 144 has particularly good remarks on the youthful Neoptole- mus’ character in the same vein. Consistency and logic are not the best means to analyze character portrayal. 65 See LIMC s.vv. “Electra,” “Orestes.” WILLIAM J. SLATER 367 ringlets, white complexioned, cheery, a model of a handsome god”—or the “second squalid”—“much thinner and younger” (than the first squalid). As for Electra, we have several choices available from the nasty selection of shaven and yellowish masks, enough certainly to remind us how carefully the audience would watch for indications of the type of moves to be expected of the characters. All of this basic guidance is denied us, and so left to our imagination, as controlled by our inter- pretation of the text of the play. Second: what we are not told. In an interpretation, when we ask specific and central questions about character—did Orestes really want to kill his mother? or: was Electra sorry that she killed her mother?—the text is much less than clear, and the commentators not in agreement. This lack of clarity is itself extraordinary, for why one asks could an audience be left in any doubt about such issues? To take only one commentator: Den- niston writes on line 600: “Even if phoneusi in v.89 is a poetic plural, and refers to Aegisthus only, [but it could be simply a generalizing plural!] 614 and 646 make it clear that Orestes is prepared to kill his mother.”66 But of course “prepared to kill” is not “wanting to kill,” and in addition: (a) 614 is a reply to the Old Man’s suggestion that Orestes kill both Clytemnestra and Aegisthus, and all Orestes in fact says is the ambiguous “I have come for this stephanos = prize [i.e. lord- ship of Argos not just the killings], but how shall I get it?” The discussion after this is about Aegisthus. (b) 646 has a better punctuation in other editions, e.g. Cropp’s “How then? Am I to kill him and her at the same time?” where- upon Electra intervenes. Neither of the passages proves what Denniston claims, and 66 J. D. Denniston, Euripides’ Electra (Oxford 1939). Far too much has been made of the non-specific plural “murderers” in 89, and it should be observed that even Denniston set no store by it. If Euripides had wanted to make Orestes eager to kill his mother, he would have said so. He simply did not, and no amount of commentary will prove otherwise. 368 SPLIT-VISION there is no passage that unequivocally does. There is no passage therefore in which Euripides tells us that Orestes wanted to kill his mother; i.e. Euripides deliberately chose not to tell us. The passage at 580ff is now corrupt and much emended, but it too does nothing in its present state to justify an interpretation that Orestes is eager to kill his mother. The text simply does not tell us clearly; there is a certain anomaly in suggesting67 that a text tells us all we need to know for the dramaturgy but also that tragic texts tend to be open-ended, and leave the meaning of that dramaturgy in doubt. On this particular issue we can compare and contrast the Electra of Sophocles, where Stevens could sensibly describe Orestes as “calm, resolute, matter of fact,”68 and where no one, following the text, can or should be in doubt about his inten- tions. When we turn to ask the important question, whether Electra shows any remorse69 for murdering her mother, we find commentators agreed that she does. But in fact she does not, for she says only at the textually difficult passage (1183) that she is or was to blame, “responsible”—afit¤a , mistranslated often as “guilty”—in answer clearly to Orestes’ claim (1181) that it was “by his hand” that the deaths occurred. She says this to Orestes to console him, i.e. she and not he was responsible. This is simply “taking the blame,” not showing remorse70 in any normal 67 D. H. Roberts, “Sophoclean Endings: Another Story,” Arethusa 21 (1988) 177–196, at 189, recommended by Taplin, “Comedy and the Tragic,” in M. Silk, ed., Tragedy and the Tragic: Greek Theatre and Beyond (Oxford 1996) 188–202, at 197. Compare the remarks of Slezak (supra n.49), esp. 20: “an ironically open end to a drama, whereby the audience would first have to supplement what was really intended, does not correspond to the Sophoclean use of irony.” 68 P. T. Stevens, “Sophocles: Electra, Doom or Triumph?” G&R 25 (1978) 111–120, at 116: “In Sophocles’ play no one could pretend that the calm re- solute matter of fact Orestes shows any sign of mental disturbance.” Some readers, e.g. H. Foley, Female Acts in Greek Tragedy (Princeton 2001) 33, claim to detect in Orestes in Choephori “some signs of hesitancy towards committing the crime”; but even that is debatable. 69 So Cropp (supra n.60) xxlv. 70 Cf. the real remorse at Eur. Andr. 814. At 1226, in Cropp’s text, Electra does say that she has committed deinÒtaton pay°vn—this too is not remorse. WILLIAM J. SLATER 369 sense, and one could with equal justification argue that it was pride in her role. Neither in Sophocles nor in Euripides is it de- monstrable that Electra shows remorse, though the chorus may criticize her in Euripides (1175ff). Nor is it at all demonstrable that an Athenian audience expected her or someone in her position to show remorse. Once again modern commentators have read their own moralising into the characters, and then without justification into the text. It is Orestes who is treated very differently in the two Electra plays, while Electra remains the same. Notably then masks and text leave us in the dark, with commentators eager and perhaps desperate to extract character from unforthcoming dialogue, for without clear indications of character the play seems to lack a coherent interpretation. It may therefore be that Euripides answered these questions in more subtle ways that are not to be found directly in our texts, though such solutions must be compatible with the text and also solve problems posed by the text. I should argue, if forced to choose a mask to make the play coherent, that Orestes is intro- duced as the “naive young man” who as Cropp puts it “has not envisaged the implications of matricide.”71 Electra on the other hand could be an older “squalid” type, whose role is the tradi- tional mourning self-abasing woman, openly seeking revenge, that many recent studies have now explicated and applied successfully to the dramatic picture of Electra,72 in contrast to the moralizing disapproval of the past.73 Such dangerous female 71 Cropp (supra n.60) xxxiv. 72 Foley (above n.68) 24 with bibliography; C. N. Seremetakis, The Last Word: Women, Death and Divination in the Inner Mani (Chicago 1991); G. Holst- Warhaft, Dangerous Voices: Women’s Lament and Greek Literature (London 1992); J. Mossmann, Wild Justice (Oxford 1995) 169–192; Allan (supra n.2) 174–175; Burnett (supra n.57). But long ago Prosper Merimée in his novel Colombe had seen the true relationship between Electra and Orestes in a vendetta-driven society; cf. E. de Martino, Morte e pianto rituale nel mondo antico (Turin 1958) 63–67. 73 B. Knox, “Euripidean Comedy,” in E. Segal, ed., Oxford Readings in Menan- der, Plautus, and Terence (Oxford 2001) 3–24, at 7 can still write of Electra’s 370 SPLIT-VISION public behaviour could bring sympathy74 and even prestige,7 5 but the juvenile Orestes is unready to be the instrument of his sister’s will. He therefore does not reveal himself, so creating the single greatest problem in the dramatic unfolding of the plot, because nothing in the text tells us clearly why he does not do so. The now silent masks would then be one clue to understand- ing what follows. But it would also be necessary to invoke the kind of “secon- dary action” that I have pleaded for above. The undisguised Orestes does not know what awaits him; but the disguised Orestes cannot speak openly to the audience of his reluctance, precisely because he is disguised; the poet has not utilized “asides” about his feelings or wished to have him soliloquize on an empty stage. Euripides has thereby created a most interest- ing version of disguise-drama which ends in exposure, not recognition or revelation; but the only mechanisms left to the poet to indicate to an audience the inner feelings of Orestes are indirect. Recently I have argued that there are indeed hints in the ——— murder of her mother as “an unnecessary act of paranoiac jealous hatred,” a judgement that has little to do with Greek tragedy. 74 Slezak (supra n.49) 11 convincingly shows the chorus of Sophocles’ Electra siding eventually with Electra’s position, and criticizes rightly those scholars (Segal, Gellie) who have tried to suggest that the position of Chryso- themis is treated as morally equal to that of Electra. He is wrong however in one central point in assuming that Electra is breaking with “social conven- tions” (13); in her profound mourning for her father, she is precisely following such conventions. 75 Rather than add to the extensive bibliography about Greece, note rather the unjustly neglected examples in Latin. Tacitus Hist. 13.32 tells of the woman who after the murder of her relative Livia Julia, wore mourning and grieved unceasingly for forty years (from A.D. 43 onwards). This escaped punishment under Claudius and, notes Tacitus, gave her prestige. Less fortunate is the old lady Vitia executed in 32 for lamenting the death of her son, for as the historian notes, women could not be charged with aiming at power, so they were charged with lamenting (Ann. 6.10). The wife of Rubellius Plautus (Ann. 16.10) had maintained mourning since her husband was killed, an unkempt widow eating barely enough to stay alive. She waited at Nero’s door in Naples, where she ululated (eiulatu) and screamed. But Nero showed himself unmoved by precibus or invidia, which I understand, perhaps wrongly, to be the threat of the Evil Eye. WILLIAM J. SLATER 371 text itself,76 and I do not repeat those arguments here. There ought to have been indications in the masks, as we saw, though there will be no agreement how detailed these were. Here I argue for secondary action of some type, easily imagined when Electra is demanding matricide of Orestes, while the audience, like modern readers, are increasingly curious why the young man does not reveal himself. There seems to be only one explanation for his prolonged disguise and refusal to reveal himself to his sister: a simple reluctance to take off his disguise; and the only explanation for this reluctance: silent unwillingness, based on apprehension, which can, as we have seen, be rendered visible by dramaturgy. The play was all much easier to understand if the audience could derive his apprehension on the stage from a performance that brought out his reactions and body language along with ambiguous and evasive expressions as he is con- fronted by his sister’s implacable determination to be revenged on her mother. In that case there would be no doubt about the answer to the questions that the text itself does not appear to answer satisfactorily. I hope to have shown here that the mature Euripides was certainly capable of creating such relatively complex dramaturgy. Beyond that, one can only appeal to the coherence of interpretation it enables. Scientists can postulate the existence of an unobserved planet from perturbations in the orbits of those they can observe. In much the same way, if less scientifically, this paper argues for the existence of action that is now beyond the direct reach of our evidence, on the basis admittedly of probabilities and possibilities, but also in the belief that such hypothetized dramaturgy makes for a more interesting and sophisticated play and even occasionally more coherent interpretation. Perhaps it will be alleged that this is a petitio principii;77 but this would be, 76 Slater (supra n.62). 77 See Slezak (supra n.49) 119, who animadverts correctly on errors of this kind to be found in R. P. Winnington-Ingram, “The ‘Electra’ of Sophocles: Pro- 372 SPLIT-VISION I think, unfair. There is without any doubt a problem in the Electra of Euripides that causes perturbation: we do have to explain why it is reasonable that Orestes remains in disguise, and that in turn means that an audience also had to be able to guess why this was so, for no audience will tolerate for long a performance that has no dramatic logic. We must then postulate how and what an ancient audience could know; there is no alternative. The social anthropology of revenge they knew instinctively, unlike all scholars; they saw masks, unlike all scholars; they saw a performance whose gestures they under- stood, and we cannot. And they heard immediately the hints that we can only try to puzzle out. The text is not always enough.78 June, 2003 Department of Classics McMaster University Hamilton, Ontario Canada L8S 4M2 slaterw@mcmail.cis.mcmaster.ca ——— legomena to an Interpretation,” PCPS 183 (1954/5) 20–27, reprinted in E. Segal, ed., Oxford Readings in Greek Tragedy (Oxford 1983) 210–216. 78 My thanks are due to Prof. C. W. Marshall for astute criticism; his knowl- edge of practical aspects of performance has been invaluable. Prof. Eric Csapo has once again corrected many of my generalizations. Professor W. Allan pointed out weaknesses. Ms. Daria Bertolaso (Paris and Ferrara) has very kindly provided me with much detailed criticism based on her superior knowl- edge of continental dramatic theory. But I have been forced to cut out a good deal of bibliography which did not directly add to the central argument; it was my aim to make a specific point about interpretation, not to review theories of performance.