Greek Notions of the Past in the Archaic and Classical Eras

DEBATING THE PAST IN EURIPIDES’
TROADES AND ORESTES AND IN
SOPHOCLES’ ELECTRA

This chapter will examine three passages from tragic agones that reflect on the problem of finding out what actually happened in the past. Tragedies present variant versions of events that supposedly happened, so when disagreement about facts is articulated within the world of the play itself, it is hard not to take it as a threat to the usual suspension of concern about the literal truth of any one presentation. Most tragic agones are not about facts, but about the evaluation of the facts, and there is at least basic agreement about what is relevant. In many other debates (in Euripides’Hippolytus, for example), one speaker is simply wrong. But some passages raise the question of whether what actually happened can be extracted from narratives that are usually self-serving or partisan or biased in other ways. Implicit in these passages are fundamental hermeneutic problems. All three at least approach a situation we would more readily associate with Hellenistic poetry; they raise the possibility that different characters actually belong to slightly different versions of the story. When different versions of a story seem to be active in one play, even in a context that recalls the everyday difficulty of deciding on the truth in a law-court, it becomes hard to avoid profounder questions of historical truth: the question becomes not which speaker is lying, but what historical truth is and how it could be known. Euripides and Sophocles sometimes seem to be reflecting on the difficulty of ascertaining historical truth, on its methodological problems.

In Orestes, the disagreement between Tyndareus and Orestes is not exactly about fact, but it is not exactly about how agreed facts should be evaluated, either. Tyndareus’ tirade against Orestes (491–539) presents the famous anachronism problem.1 As Porter has shown in detail, Tyndareus follows contemporary Athenian styles of forensic argument in his claim that Orestes ignored nomos (law or custom) in killing his mother:2

ὅστις τὸ μὲν δίκαιον οὐκ ἐσκέψατο

οὐδ’ ἦλθεν ἐπὶ τὸν κοινὸν Ἑλλήνων νόμον . . .

χρῆν αὐτὸν ἐπιθεῖναι μὲν αἵματος δίκην

ὁσίαν διώκοντ’, ἐκβαλεῖν τε δωμάτων

μητέρα· τὸ σῶφρόν τ’ ἔλαβ’ ἂν ἀντὶ συμφορᾶς

καὶ τοῦ νόμου τ’ ἃν εἲχετ’ εὐσεβής τ’ ἃν ἦν. (494–5, 500–3)

Who did not consider what was just, and did not have recourse to the common law of the Greeks . . . He should have imposed the pious penalty for bloodshed by prosecuting, and thrown his mother out of his house. Then he would have got [credit for] sophrosyne instead of a calamity, he would be keeping inside the law, and he would be reverent.

Tyndareus’ language is difficult and ambiguous when he directly addresses what Orestes should have done. West takes ὁσίαν as a noun rather than an adjective, and translates the phrase as ‘while aiming for religious correctness’.³ Still, even if we think that ὁσίαν is the adjective, and even if διώκοντ’could mean only ‘drive away’, in such a context it surely implies a prosecution (Biehl takes it as the noun, but still treats διώκοντ’ as a judicial term).⁴

Tyndareus’ argument, then, resembles the claims, common in Athenian forensic speeches, that the opponent despises and threatens to overturn the city’s laws. However, it is different from these because the audience cannot be certain that there is a clearly applicable nomos in Orestes’ situation, although Tyndareus’ argument makes no sense without such a nomos. He later states that the ancestors created the applicable rules:

καλῶς ἔθεντο ταῦτα πατέρες οἱ πάλαι·

ἐς ὀμμάτων μὲν ὄψιν οὐκ εἲων περᾶν

οὐδ’ εἰς ἀπάντημ’ ὅστις αἷμ’ ἔχων κυροῖ,

φυγαῖσι δ’ ὁσιοῦν, ἀνταποκτείνειν δὲ μή. (512–15)

Our fathers of long ago regulated these issues well: they did not allow anyone who had blood-taint to go where he would be seen or addressed, but to purify by exile, but not to kill in return.

However, he does not support exile in this case, but urges Menelaus to allow Orestes to be stoned by the citizens (536). So he has two distinct claims: that exile is the traditional penalty for homicide, and that civic procedures must preclude private vengeance. In any case, by giving no subjects to the infinitives that explain how the ancestors forbade contact with a polluted killer, he makes the subjects general and so implies a communal behaviour.

The difficulties that most commentators see in this speech are not the ones that concern me. For example, even if the Oresteia had established a tradition that no legal procedure for homicide existed until Athena created the Areopagus, Euripides was entirely within his poetic rights to invent a court at Argos – and it would not be unusual for him to disturb mythological tradition (indeed, he had already introduced unplacated Erinyes in Iphigenia in Tauris). Indeed, Homer and Hesiod presuppose the existence of legal institutions. Achilles’ shield, for all the difficulties of its trial scene, shows a procedure in the aftermath of a homicide. So when Euripides makes the Argive assembly debate the fate of Electra and Orestes, he could be seen as correcting the implications of Aeschylus’ version. In Choephori, Orestes leaves Argos very quickly after the matricide, so that the question of how the city would have reacted is easily ignored. Also, the play emphasises the illegitimacy of Aegisthus’ rule and defines his killing as a tyrannicide, and in Eumenides Orestes, once acquitted, speaks as ruler of Argos without hesitation. So the action avoids any messy problems about how Argos would have or should have responded. Still, the grandeur with which Athena endows the Areopagus implies that the heroic world has had no procedure other than blood-vengeance for handling homicide. Eumenides elides the differences between ordinary killings and murder within the family, for which legal institutions dependent on relatives to prosecute were not satisfactory. For murders committed by tyrants, no institutions could ever provide a satisfactory solution.

So Euripides, by having Orestes and Electra tried by the assembly, creates an Argos in which a legal procedure exists without denying that the Areopagus was the first real homicide court. The constitutional situation at Argos is vague: the assembly functions as if it were sovereign. Scholars disagree about whether the audience should believe that Menelaus is scheming to obtain rule, but Orestes certainly says that he is (1058–9).5 When Apollo tells Menelaus at 1660–1 to let Orestes hold power in Argos and himself to go rule Sparta, he evokes the stereotype that Spartans were unlikely to be contented with Sparta.⁶ In any case, Orestes will rule in Argos; it is not a democracy. Euripides may be assuming that even in the heroic world, the assembly would have full sovereignty during an interregnum, as the Ithacan assembly in the Odyssey apparently would if it could achieve any unity for action. Antinous, for example, is concerned that if Telemachus reports to the assembly how the suitors tried to murder him, the people will drive the suitors out (Od. 16.381–2).

So the difficulty is not anachronism as such. We should see the assembly, though deliberately contemporary in tone, as almost the opposite of an anachronism. It is more like a historical thought experiment: if Orestes had not fled immediately after the matricide, what would the Argives have done? The assembly debate is not so much a false intrusion of democracy into the heroic past as an attempt at creating a plausible scenario based on the assumption that human (that is, political) nature is always the same, as Thucydides believes (1.22, 3.84.2). Euripides works with an old, Homeric institution, the assembly; he has the model of Achilles’ shield for some kind of legal procedure for homicide in the remote past; he considers the likelihood that the assembly could assert itself in the absence of a ruler; and he examines the way politics work. The Argive assembly is more imagined history than anachronism, and if the assembly is possible, Tyndareus’ argument is not inherently anachronistic either.

Nor is the problem with Tyndareus’ speech that, in practice, Orestes had to kill Aegisthus, who held tyrannical power, and that he would have found it very difficult to expel his mother or initiate proceedings against her. Simply driving her out of the house before some legal process would not be a wise plan – we may remember her call for an axe at Choephori 889; she would not have gone quietly. It is far from certain that Orestes could control even the household with Clytemnestra alive and present, and in the world of the play, Aegisthus had his own faction. These may be reasons for the audience not to be convinced by him to abandon sympathy for Orestes. Tyndareus’ argument may well seem unfair in not recognising the extraordinary difficulties of Orestes’ situation. But that does not make the argument strange. A speaker in an agon can ignore practicalities of that kind, and it is open to the opposing speaker to introduce them in refutation.

Orestes, however, does not introduce the practical difficulties in following Tyndareus’ belated legalistic advice.7 Tyndareus’ argument is truly strange because Orestes’ utterly fails to respond to it. Orestes argues entirely as if he were in the world of Choephori. He argues, for example, that while he now suffers from his mother’s Erinyes, he would have been pursued by his father’s if he had failed to avenge him (580–4). His argument recasts Apollo’s complaint in Eumenides that the Furies did not pursue Clytemnestra, to which they reply that she had not spilled kindred blood. He speaks as if the only alternatives he had were to do nothing against Clytemnestra or to kill her. Orestes uses several lines to claim that he obeyed the command of Apollo (591–9), and he accuses Tyndareus of calling the god ἀνόσιον, but he does not say that Apollo told him to kill his mother rather than to try to exile her. His argument could have been just the same had Tyndareus never raised the possibility that Orestes could have acted to avenge his father without killing his mother. Precisely because a refutation speech in tragedy normally proceeds point for point, this is a striking oddity, and it is strange whether or not we think that Tyndareus’ language implies a judicial procedure.

The Argive assembly does not consider any such possibility either. The speakers disagree about whether Orestes was right to kill his mother and about how he should be punished if he was wrong, but nobody says that he had any choice between matricide and inaction. Talthybius, who praises Agamemnon but says that Orestes created a bad precedent (νόμους, 891), seems to be trying to define some middle ground in the debate, but does not specify what Orestes ought to have done. Tyndareus’ argument falls into the play as a foreign element that is utterly irrelevant to everyone else. That is the most significant problem. It feels out of place because no place is made for it. So what purpose does it serve? The main dramatic functions of Tyndareus are to intimidate Menelaus and to show on a small scale the pattern of Orestes’ behaviour: when he is received sympathetically, he acknowledges how guilty he feels; when he is attacked, he defends himself and counterattacks.8 But nothing requires that Tyndareus make an argument that dangles in air. The development of the play as a whole would not be different if Tyndareus simply criticised the matricide.

This isolation of Tyndareus’ argument is an invitation to interpret it at a different level. It is as if he were trying to contest the fictional world in which he has been placed. Although different characters judge the matricide very differently and consider very different precedents it could establish, nobody other than Tyndareus thinks that it could be evaluated by being compared to an alternative but effective action. Yet Tyndareus stresses that his alternative is the ancient nomos of the ancestors – which, from the perspective of Euripides’ audience, would make it profoundly ancient. Tyndareus seems to want Orestes to have followed a procedure nobody else ever considers, but he is an old man, and a Spartan – not a character from whom we would expect new ideas – who sees that procedure as belonging to hoary antiquity. Orestes was produced in 408, exactly at the time Draco’s homicide law was reinscribed (409/8) in the context of the revision and republication of Athenian law that followed the restoration of the democracy after the oligarchy. The surviving section concerns unpremeditated homicide. Scholars have extensively discussed the nature of the full law and whether Tyndareus’ speech reflects the old law. Holzhausen plausibly suggests that Tyndareus does not mean that a court could not have imposed the death penalty, but is referring only to the limits of self-help.⁹

This debate, however, does not consider the wider context of both the play and the reinscription of the laws. Whether Tyndareus’ argument refers to the substance of the old Athenian law may be less important than that he refers to ancient laws mentioned by nobody else. The reinscription of those old laws still in force surely indicates both the restored democracy’s concern to demonstrate its historical continuity with the past, and the democrats’ concern that claims about old laws would be problematic without an authoritative display of exactly what was valid. The infamous chapter 4 of the Ath. Pol., the ‘Constitution of Draco’, was probably forged around the time of the oligarchic revolution.10 Certainly democrats and oligarchs both claimed ancestral support. According to Ath. Pol. 29.3, Clitophon proposed that the probouloi examine the laws of Clisthenes, on the grounds that his constitution was not popular but close to Solon’s.¹¹ Thucydides 8.97.6 makes democrats in the fleet at Samos speak of πατρίους νόμους that the new government has (wrongly) dissolved. The expression ‘ancestral laws’ was evidently, as Mogens Hansen has pointed out, a ‘hurrah-word’; it indicated approval of the laws.¹²

In this political climate, it is not perhaps so strange that everyone else ignores what Tyndareus describes as ancient and ancestral rules. Someone may assert that the ancestors established a law long ago, but that does not mean that he is right, or that, even if the law is ancient, it has been in consistent use and had not been superseded. So having one character seem to live in a different past from the others is not pointless, since contemporary Athenians profoundly disagreed about what their past was.

We cannot know, however, exactly how this rough similarity between Tyndareus’ speech and contemporary political discourse would have impressed the contemporary audience. The play takes place far in the past, soon after the Trojan War, and Tyndareus refers to ‘the fathers, long ago’. Draco would be recent in comparison to Tyndareus, let alone to his ancestors. Maybe Tyndareus is simply modelled on an elderly Athenian oligarchic type: that he is old, Spartan and unyielding would cue the audience to see Tyndareus’ ancestors as the oligarchic Draco rather than the Draco the restored democracy could claim.¹³ In that case, the difference between Tyndareus’ long-ago ancestors and the long-ago ancestors of the audience was erased: he could be heard as if he were arguing in late fifth-century Athens, so that the historical argument was completely ahistorical in its framing. However, it would also be possible to see the speech as a deliberate transfer of a contemporary kind of argument about the past to a remote past, and so as a commentary on the futility of such arguments. From this perspective, the isolation of Tyndareus would implicitly support the decision to reinscribe laws that were still valid in contrast to attempts to re-establish an imagined earlier constitution as a whole.

Troades presents very different problems. The central issue for the agon is whether Helen came voluntarily to Troy. Helen treats the Judgement of Paris as a serious event and argues that Aphrodite was committed to her abduction, so that she was subject to superior force. Hecuba rejects Helen’s interpretation of the Judgement of Paris as a serious dispute and sees Aphrodite as a metaphor for Helen’s attraction to Paris. This leads to her curious argument that Helen cannot have been taken by force, since she did not cry for help (998–1001) – but Helen has not argued that she was forced in that sense. The Gorgianic defence of Helen is surely in play here in a complex way: Hecuba, though she rationalises the story by treating Aphrodite as metaphorical, rejects Gorgianic claims about the power of desire, while Helen, whose version is traditionally mythological, speaks with a sophistic edge.14 In the end, the entire debate is irrelevant to the outcome. Helen will survive because Menelaus succumbs to her erotically and so proves himself no stronger in resisting desire than Helen.

While this is the main debate in the agon, there are points that do not belong to the grand design but raise slightly different questions. Helen and Hecuba disagree about what happened after Paris was killed. Although this issue is relevant to the main argument, since if Helen tried to leave Troy it is more credible that she followed Paris only under divine influence, the disagreement also indicates another, subordinate problem in the debate. Helen claims that she then tried to escape by letting herself down from the walls with a rope, but was compelled to marry Deiphobus (955–60). She calls on the guards as witnesses – but of course they are all dead. Hecuba’s first counterargument, though, is very odd:

Where, then, were you caught fixing knots, or sharpening a sword, as a noble woman would do in longing for her previous husband? (1012–14)

The first alternative could describe preparations for the escape from the walls, but once the second appears it is clear that Hecuba is saying that if Helen had really regretted going with Paris she would have killed herself. She then quotes herself as offering to help Helen escape (1015–19).

Ὦ θύγατερ, ἔξελθ’· οἱ δ’ ἐμοὶ παῖδες γάμους

ἄλλους γαμοῦσι, σὲ δ’ ἐπὶ ναῦς Ἀχαιικὰς

πέμψω συνεκκλέψασα· καὶ παῦσον μάχης

Ἕλληνας ἡμᾶς τ’.

My daughter, leave. My sons will make other marriages, but I will help you slip away and convey you to the Greek ships. And make the Greek and ourselves cease from battle.

Helen is presumably lying about her attempts to escape from Troy, but Hecuba’s reply is strange and unreliable. Is the audience supposed to take Helen’s failure to commit suicide as proof that she did not regret her actions? Hecuba quotes herself as speaking to Helen with maternal affection, when everything else she says gives the impression that she has loathed Helen all along. (Even if Helen had wanted to escape, she would not have been likely to trust Hecuba.) Hecuba sees Helen as drawn to Paris not just by his good looks, but by the prospect of extraordinary luxury (991–7), and not wanting to leave because she enjoyed barbarian subservience (1020–2); she uses forms of ὑβρίζειν twice. In Hecuba’s account, Helen is a stereotypical Spartan, whose longing for Persian extravagance and opportunities for lording it over others becomes manifest once she leaves home (cf. Thuc. 1.77 and especially 1.130 on Pausanias’ luxurious and oppressive behaviour).

This section of the debate echoes the stories Helen and Menelaus tell at Odyssey 4.235–89, where the issue is, again, whether Helen returned her loyalties to Menelaus after the death of Paris. The question is thus traditionally unresolved. In Homer, Helen narrates how she helped the disguised Odysseus when he came to spy in Troy; Menelaus then tells how she tried to induce the men hidden in the wooden horse to betray themselves by imitating the voices of their wives. These opposing stories are told only after Helen has drugged the wine, and the problem of whether the two tales could be reconciled is not addressed. It does not need to be, not only because the participants are immunised against any real consequences of what is said, but also because the stories are in any case directed mainly at confusing Telemachus and the external audience about the extent to which Penelope should or will be trusted. But while Menelaus simply tells a story that requires a Helen who assists the Trojans, Hecuba’s speech in Euripides is very different: she has a full and completely negative interpretation of Helen’s character, and judges all her actions accordingly. In Hecuba’s view, Helen desires sexual satisfaction, luxury and opportunities to behave hubristically to others, and everything she does is solely and exclusively motivated by self-interested calculation of how these desires can be satisfied. The passage is an exceptionally rich instance of tragic theory of mind in operation. Whether one believes Helen’s claim that she tried to escape or Hecuba’s insistence that she did not cannot be a matter of weighing the external evidence. It is a matter of ἦθος.

Hecuba, furthermore, is self-interested. Helen has begun her speech by blaming Hecuba for giving birth to Paris and not killing him (919–22) – the first play in Euripides’ production of this year was Alexandros, which dramatised a murder attempt on Paris when he came to Troy after surviving exposure. So there is some obvious unfairness in Helen’s accusation – Paris seems unkillable – but she is right that the war cannot be blamed entirely on her. Her accusation, though, gives a particular edge to Hecuba’s response. Everything Hecuba says could be true, but it is also, we can assume, driven by her need to displace the responsibility of the Trojans themselves. In this light, her account of how she offered to help Helen flee is also self-justifying. So we are left in a typical Herodotean situation. We have two accounts. One may seem more credible than the other, but both come from self-interested reporters, and because the reporting situation is agonistic, there is no space for a nuanced approach.15 In any case, there is still a blunt question of fact. Either Helen tried to escape Troy or she did not. The interpretation will inevitably precede the judgement of fact instead of deriving from it: we will believe Helen or Hecuba depending mostly on whether we accept Hecuba’s understanding of Helen’s character. While the speeches are relatively balanced intellectually, Hecuba is sympathetic, while Helen certainly is not, so probably most members of the audience will be inclined to Hecuba’s side.¹⁶ Still, the scene could prompt reflection on the underlying problem, which is the danger of actor-observer bias or fundamental attribution error (also called ‘correspondence bias’).

Hecuba assumes that Helen has fixed characteristics that determine everything she does. Despite cultural variation, people generally overemphasise the role of individual character in the behaviour of others while placing far more weight on situational factors in explaining their own actions. Behaviour is generally less consistent than people assume when making judgements about others. For a historian, this fundamental attribution error is a constant danger in making historical assessments, especially because the process of understanding why historical figures acted as they did is inevitably a hermeneutic circle: character is inferred from actions, and then that inferred character is used not only to explain other actions, but to decide what a particular person did when other evidence is insufficient. Euripides obviously did not have the benefit of the vast research conducted on fundamental attribution error since 1967.17 But the problem is basic and universal – and one could base a credible reading of the Orestes, especially, as a critique of standard Greek assumptions about consistency of character.

In Sophocles’Electra, Clytemnestra argues that she killed Agamemnon justly, in vengeance for the sacrifice of Iphigenia. She offers arguments about why Agamemnon’s action cannot be excused. For example, he may have acted on behalf of the Greeks, or of Menelaus, but they had no right to kill her daughter (535–41). Menelaus had two children of his own, who were more obvious candidates for the sacrifice. Electra responds by claiming (on the authority of hearsay, 566) that Artemis demanded the sacrifice of Agamemnon because he had hunted a stag in her grove and boasted in a way that offended her. Artemis then demanded the sacrifice of Agamemnon’s daughter as ‘equal weight’ (ἀντίσταθμος) for the deer (570–1). This detail gives Artemis two slightly different motives: the boast, which goes back to the Cypria (p. 32, 55–7 Davies), and the killing of the sacred deer itself, which alone explains why the sacrifice is a fitting punishment (the two are combined in Dictys, FGrHist 49 F 5). Electra says that without the sacrifice the army was unable either to go to Troy or to disband, so with great reluctance (575) Agamemnon performed it.

This latter claim is a little odd, since the traditional story has the expedition held back by winds, and Aulis is not an island. If Electra’s version is to make sense, it requires a prophetic or oracular intervention that would warn Agamemnon against trying to avoid the sacrifice by abandoning the expedition. Parker, followed by Finglass, sees Electra’s assertion as unproblematic, comparing Lichas’ knowledge of why Zeus required Heracles to serve Omphale at Trach. 274–79.18 However, that passage is also not entirely straightforward, and tragic convention allows Lichas, as a messenger, to drift toward omniscience, a licence Electra cannot share.¹⁹ In any case, I do not think the parallel means that we must assume that Electra reliably knows of a prophecy that fully guarantees her account. Although, of course, a prophecy of Calchas is prominent in the tradition, we cannot simply fill in what is missing from that tradition, because there is no reason to think that in any earlier version Artemis forbade Agamemnon to abandon the war. In both Aeschylus’Agamemnon and Euripides’ Iphigeneia at Aulis, going home is a real choice. If Electra cited such a prophecy, we would have to accept her version, but here there are unclear boundaries dividing what Calchas said, what Electra has heard, and how Electra has understood what she heard.

It is just as likely that the spectator would guess that she has inferred that Agamemnon could not have abandoned the expedition, because she is subjectively certain that her father would not have sacrificed her sister if he had any choice at all. Still, Electra is the sympathetic character here, and the audience must be inclined to take her side. She must be telling the truth as she understands it, which means it is what she has heard, and therefore what has been reported. She proposes asking Artemis for the cause, before saying that such an inquiry would not be themis (563–5). Certainly, we should not treat it as Winnington-Ingram does: ‘Electra’s account . . . is the story she would like to believe; and we can hardly suppose that Sophocles wishes us to take it too seriously as an explanation of events.’²⁰ This reading of these lines is tendentious, governed by a larger interpretation of the play as condemning the matricide. Yet Electra’s story, with its essential but unreported prophecy, is not entirely transparent, and her version should not be treated unambiguously as Sophocles’ version.²¹ Since Electra is constructing her story from hearsay, we can see her as an amateur historian. She is trying to combine and make sense of all she has heard, and using her experience of her mother as further evidence.

So how are we to understand these speeches? One possibility is that Clytemnestra is simply lying. In Electra’s account, Agamemnon could certainly be blamed for the transgression that caused Artemis’ anger, but he had no choice about the sacrifice. Clytemnestra’s claim that she killed him in vengeance is then a further lie, and she killed him for her relationship with Aegisthus, using the sacrifice as a self-serving pretext.22 This is what Electra says at 561–2. In any case, if Agamemnon was at fault, her arguments about Menelaus’ children are a sham. The other alternatives are more complicated, but also more interesting: Clytemnestra is speaking in partial good faith (only partial, since she certainly has not acted solely for the motive she offers). If she knows Electra’s version, she does not believe it. She could be arguing from within the story of Aeschylus’Agamemnon, where Artemis’ anger does not have a direct and obvious motivation. She could also accept the story that Agamemnon angered Artemis by killing the stag (while deliberately ignoring it in her speech), but not believe that he had no choice in sacrificing his daughter. There is no way to tell from the text whether she was actually present at Aulis.

What we may have here, then, is again something very much like a Herodotean ‘he said/she said’. The disputants accept accounts of the past that justify their present beliefs, just as Herodotean figures present versions of the past that make their own nations appear guiltless. Clytemnestra implicitly denies that any prophecy or other authoritative source required that the sacrifice be Agamemnon’s child. The question of why the victim should be Agamemnon’s child is raised elsewhere in tragedy, and in Euripides’Iphigeneia at Aulis, Agamemnon at least implies that Calchas’ prophecy requires Iphigenia, both in the prologue (90) and in response to Clytemnestra (1262). Clytemnestra introduces a degree of choice that the tradition does not. Once Aeschylus’Agamemnon had told the story of the sacrifice without any particular motive for Artemis to punish Agamemnon, the question had to arise. Electra’s version, on the other hand, gives Agamemnon no effective choice at all. One version has an Agamemnon implausibly unconstrained, while the other creates so much constraint that he really has no dilemma. These interpretations of the past correspond to problems of interpretations of the present: Electra tells the chorus that her endless laments are entirely forced on her (256–7), at least in so far as she is εὐγενής, and she tells her mother that her abuse of her mother is also completely involuntary (619–21). On the other hand, she certainly does not see Clytemnestra’s behaviour as in any way forced on her, and she generally refuses to accept limits imposed by others on her own freedom of action.

Electra’s account leaves no moral ambiguities. If her version is correct, Agamemnon, apart from the folly of his boast, is guiltless, while Clytemnestra was not actually motivated by Iphigenia’s sacrifice at all – it is entirely an excuse. Clytemnestra, on the other side, (repellently) does not acknowledge any moral complexity in her own actions; she says that she is ‘not distressed by what was done’ (549–50), while she insists that Electra’s behaviour in speaking ill of her is blameworthy (523–4) and indeed regards it as hubris. In fact, she regards Electra’s constant references to the murder of Agamemnon as an excuse (πρόσχημα, 525), although she does not say what she believes Electra’s real motives are.

This agon, like the one in Troades, probes beyond the basic preference of any group for the story that makes it look better. People in antagonistic situations produce versions that dismiss the pressures and situational constraints of others while making themselves and those they support victims of circumstances. Tragic characters expect more consistency than reality typically presents. This is a danger for the historian who must interpret their accounts, and these tragedies put the spectator into the situation of such a historian. In Orestes, Tyndareus refers to an ancestral and panhellenic norm that nobody else recognises, so that an appeal to the past does not settle the controversies of the present. Instead, present controversies seem to be projected into the past, in potentially infinite regress.

1	P. E. Easterling, ‘Anachronism in Greek tragedy’, JHS 105 (1985), pp. 1–10, at p. 9.
2	J. Porter, Studies in Euripides’ Orestes (Mnemosyne Supp. 128; Leiden: Brill, 1994), pp. 99–172, esp. pp. 110–13.
3	C. W. Willink (ed.), Orestes (Oxford: Clarendon Press, 1986), p. 169 (on 500–1), points out that Orestes did impose a blood-penalty, so the adjective is essential. M. L. West (ed.), Euripides: Orestes. (Warminster: Aris & Phillips, 1987), p. 217 (on 501), thinks that the participle is redundant if we take ὁσίαν with δίκην, but I do not understand why. See M. Lloyd, The Agon in Euripides (Oxford: Clarendon Press, 1992), pp. 115–17, on the ambiguity of Tyndareus’ language.
4	W. Biehl (ed.), Euripides Orestes (Berlin: Akademie, 1965), p. 57 (on 501): ‘Eine gerechte, innerhalb der Grenzen des religös Erlaubten verlaufende Anklage führend.’ V. Di Benedetto (ed.), Euripidis Orestes (Florence: La Nuova Italia, 1965), p. 105 (on 501): ‘intendando una legittima azione legale’.
5	R. P. Winnington-Ingram, ‘Euripides: Poiêtês Sophos’, Arethusa 2.2 (1969), pp. 127–42, at p. 134, accepts Orestes’ interpretation of Menelaus; Willink, Orestes, pp. 191–2 (on 682–716), criticises this as an ‘illegitimate back-inference’.
6	Biehl, Orestes, p. 182 (on 1661), compares Andr. 582, fr. 723 Kannicht (Telephus), Sophocles Aj. 1102.
7	D. J. Conacher, Euripidean Drama: Myth, Theme and Structure (Toronto: University of Toronto Press, 1967), p. 219, argues (in my view unconvincingly) that Orestes’ failure to respond proves that the argument should be taken seriously.
8	I thus see Orestes as somewhat more consistent than many interpreters do. A good discussion of the inconsistencies of character as survival strategy is M. O’Brien, ‘Character in the agon of the Orestes’, in S. Boldrini (ed.), Filologia e forme letterarie: Studi offerti a Francesco Della Corte (Urbino: Università degli studi di Urbino, 1987), pp. 183–99.
9	J. Holzhausen, Euripides Politikos. Recht und Rache in ‘Orestes’ und ‘Bakchen’ (Beiträge zur Altertumskunde 185; Munich and Leipzig: Saur, 2003), pp. 52–67, has a full discussion.
10	P. J. Rhodes, A Commentary on the Aristotelian Athenaion Politeia (Oxford: Clarendon Press, 1981), pp. 53–6, 84–7.
11	M. H. Hansen, The Athenian Democracy in the Age of Demosthenes: Structure, Principles and Ideology (Oxford: Blackwell, 1991), pp. 162–3.
12	M. H. Hansen, ‘Solonian democracy in fourth-century Athens’, ClMed 40 (1989), pp. 71–99, at pp. 75–7.
13	Views of Tyndareus have differed widely. At one extreme, A. P. Burnett, Catastrophe Survived: Euripides’ Plays of Mixed Reversal (Oxford: Clarendon Press, 1971), p. 106, sees him as ‘a sensible old aristocrat, the Argive equivalent of a good Athenian dicast’. M. Heath, The Poetics of Greek Tragedy (London: Duckworth, 1987), pp. 58–60, points to the effect on audience sympathy of the contrast between Orestes’ respectful demeanour and Tyndareus’ harshness. Discussion and bibliography in Porter, Studies in Euripides’ Orestes, pp. 101–3.
14	R. Scodel, The Trojan Trilogy of Euripides (Göttingen: Vandenhoeck & Ruprecht, 1979), pp. 90–104.
15	Lloyd, Agon in Euripides, pp. 105–10.
16	M. Dubischar, Die Agonszenen bei Euripides: Untersuchungen zu ausgewählten Dramen (Stuttgart: Metzler, 2001), pp. 342–57. On the difficulty of judging between the arguments, see N. T. Croally, Euripidean Polemic: The Trojan Women and the Function of Tragedy (Cambridge: Cambridge University Press, 1994), pp. 151–3; A. M. van Erp Taalman Kip, ‘Truth in tragedy: When are we entitled to doubt a character’s words?’, AJP 117 (1996), pp. 517–36, at pp. 533–4.
17	E. E. Jones and V. A. Harris, ‘The attribution of attitudes’, Journal of Experimental Social Psychology 3 (1967), pp. 1–24.
18	R. Parker, ‘Through a glass darkly: Sophocles and the divine’, in J. Griffin (ed.), Sophocles Revisited: Essays Presented to Sir Hugh Lloyd-Jones (Oxford: Oxford University Press, 1999), pp. 11–30, at p. 17; P. Finglass, Sophocles: Electra (Cambridge: Cambridge University Press, 2007), pp. 267–8 (on 566–76).
19	J. Barrett, Staged Narrative: Poetics and the Messenger in Greek Tragedy (Berkeley: University of California Press, 2002), passim, esp. p. 96.
20	R. P. Winnington-Ingram, Sophocles: An Interpretation (Cambridge: Cambridge University Press, 1980), p. 220, following J. H. Kells (ed.), Sophocles: Electra (Cambridge: Cambridge University Press, 1973), on 566–633. So also M. Ringer, Electra and the Empty Urn: Metatheater and Role Playing in Sophocles (Chapel Hill: University of North Carolina Press, 1998), pp. 159–60.
21	As J. R. March, Sophocles: Electra (Warminster: Aris & Phillips, 2001), pp. 176–7, does, esp. on 564.
22	L. MacLeod, Dolos and Dike in Sophokles’ Elektra (Leiden: Brill, 2001), pp. 84–7; van Erp Taalman Kip, ‘Truth in tragedy’, p. 519.