The politics of Aeschylus' Eumenides

Keith Sidwell

A pair of young lads from Belfast are shown talking to tramp in a London park. At the exact time this conversation is happening, a pub bombing takes place in far away Guildford. Later they experience racial jibes at a hippy squat. We subsequently see a scene in which the anti-Irish hippy shops one of the boys to the police. In both scenes, the drama underlines the part played by prejudice and ignorance in the tragedy of these young Irish rovers.

Most of you will have recognised by now that my snippets come from Jim Sheridan's film In the Name of the Father, which was released here to great applause just after Christmas 1993 and in Britain on February 12th 1994. I have chosen to begin here, carefully working the word "tragedy" into my opening scenario, so as to outline some generic problems about the function and manner of dramatic storytelling. These will give you from familiar surroundings, I hope, a better grip on the unfamiliar way in which I want to approach similar questions in relation to Athenian tragedy of the fifth century.

The first point to make is the obvious one. The film is made out of and about real events and real people. I suppose that when it is seen in Beijing dubbed into Cantonese it will acquire a certain universality ,and this is, apparently, what the film's financial backers, Hollywood's Universal pictures, intended should be the case. However, the unfamiliar -these days - experience of hearing a Dublin audience applaud a movie suggests to me that few here do not consider its specificity to be pertinent. I suppose mutatis mutandis the same will be true of audiences in Britain.

The second point is connected. Whatever the director's intention (for which see Ronan Bennett's article in The Observer for Sunday 6th February 1994), most people see the film as political. It is pretty obvious why. The two lads in my opening scenario are Gerry Conlon and Paul Hill, two of the Guildford Four. They were arrested in 1974 and convicted of the pub bombings and released only in 1989 after an appeal uncontested by the British Crown. Lord Lane, the Lord Chief Justice who let the Four out said of the Surrey detectives who conducted the inquiry: "they must have lied". The film rests at an important lacuna in the judicial process, and its political effect, as opposed to its intention, could well be to force answers to questions such as: "why have the policemen who forced the confessions from the Four not been brought to trial?".

I need to add that - quite fortuitously - the film appeared in the public arena at a point just after the Downing Street Declaration, but before any answer was forthcoming from Sinn Fein about the response of the IRA. So, in a way which cannot have been foreseen by the filmmakers, events in the real world and the commentary on them inevitably created broad political readings of the film relating to Britain's failure in Ireland. The third point is this. Despite the fact that the story is what we might like to call fact, rather than fiction, its dramatisation contains many deviations from truth, both in respect of the facts of the story and in terms of such matters as legal procedure. The two items from my scenario are examples. The furore concerning this "dancing with facts" will be well known to you. It arose out of the same sort of political reading I have just discussed, because it matters to the people involved in the specific political issue that X happened rather than Y.

What I want you to carry forward from this brief discussion are the following ideas:

(1) that fictional drama can be made out of real events and have potential or real effects in the world to which it is presented;

(2) that the meanings of drama are constructed by an audience in interaction not with an author's intention, but with

(i) underlying assumptions in their culture and

(ii) what is going on around them in what modern jargon now handily calls "the discourse" - i.e. all the things which are happening and being talked about, and the ways in which they are being discussed;

(3) that even when the drama is being made out of real events, the pressure of the creative process or some other factor, such as, for example, political ideology, can dictate changes to the story or its details. The same will apply, of course, to fictional tales which are retold. So mutatis mutandis these changes, if they can be detected (as they can in the film I've given as an example), can both tell us what mattered to the storytellers and carry messages to the audience. Sometimes, if the process and inner aesthetic of creation is divorced geographically, temporally or culturally from the discourse, these alterations will carry the wrong message, as in my example, and the artefact will fortuitously create meanings it was not specifically designed to contain.

II

It is pretty clear that we tend to read Greek tragedies as fictions. The characters for us belong to a world of myth which we unconsciously bracket with Goldilocks and Rumplestiltskin. The events of their lives retain some set actions, but in many ways they are as changeable as the retellings of any fairy-tale. The world in which they move is a sort of Middle Earth, removed not only in time, but also in some indefinable way in space, from the geographical and political world of the audience. However, we are wrong to think that Athenians would have had the same thoughts. There are two main points to make.

First of all, our view depends on defining the material of Greek tragedy as "mythical" or from "heroic saga". But it is as well to remember that our earliest extant Greek tragedy, Aeschylus' Persians, first produced in 472, deals with a very recent historical topic, the battle of Salamis of 480. In Herodotus (VI.21) we hear of a play called The Capture of Miletus by Phrynichus which dealt with an event so recent and painful to the Athenians that its author was fined and his play refused any subsequent showing.

Secondly, there was no distinction for an Athenian between "myth" and "history". The stories of Oedipus, Orestes, Theseus and Herakles told early tribal histories and were embedded in a quite specific geographical/political framework (Thebes, Argos, Athens, Thebes). It follows that it matters very much indeed who it is whose story is being told. This emerges from Aristotle's discussion of names in the Poetics, even though he is arguing for a different view. We ignore this correlation at our peril, especially when we see that the titles of the tragedies Euripides wrote in Macedon reflect the early myth/history of his adoptive home.

We also have to face up to a penetration of anachronistic matter into the "mythical" world of tragedy. The institutions and language of the polis often intrude upon and articulate these pre-polis stories. But there is a more serious problem. Occasionally, issues and events which clearly belong in the world of the audience are highlighted so strongly that it is hard not to feel that there is some political axe being ground. Yet because they are only occasional and are embedded in mythical material, it is also difficult to argue unequivocally for a political interpretation.

Hence, despite the presence of such material as the Areopagus and Argive alliance in Aeschylus' Eumenides, it is possible for some scholars to hold that Greek tragedy is never political in any narrow, partisan sense. Others hold the diametrically opposite view that plays are directly political arguments and contributed to debate on current issues in Athens. Some scholars have argued that they use allegory to make these points. A more recent approach (see especially the work of Simon Goldhill) contends that the political function of tragedy was to test (as he puts it to "set at risk") the dominant "civic ideology" of the democratic polis, that is, those basic assumptions about roles and duties of citizens shared by everyone in the city.

In my view, none of these explanations covers all aspects of the phenomena completely or solves the basic problems set by them. This is because we have ignored two pertinent facts, one about the use of "myth/history" and the other about the textual tradition of tragedy in the fifth and fourth centuries.

Let us take "myth/history" first. It is not merely that our distinction between "myth" and "history" does not apply to the way Athenians saw this material. It is that these stories were used by politicians in actual political debate as a means of persuasion towards certain courses of action. An Athenian politician argued in Arcadia in the fourth century against the Arcadians allying with Thebes and Argos on the grounds that the first city had produced parricides (Oedipus, Alcmaeon) and the second the matricide Orestes. In book 2 of his history (ch 29), Thucydides takes great pains to disabuse his readers of the notion that Teres, father of the then king of the Odrysian Thracians, Sitalkes, whose son Sadocus was granted Athenian citizenship in the early 420s, has anything genealogically to do with Tereus, the "mythical" Thracian king who married Pandion's daughter Prokne and raped her sister Philomela. He can only be doing this because the connection was current in Athens and had probably even been used in the ekklesia debate by opponents of the grant.

Where did Athenians get to know these stories? In the theatre of Dionysus, of course (Sophocles' Tereus could certainly have been produced at this period). On this view, then, the divergent tellings of these tales which we encounter in the surviving and fragmentary material would express different attitudes to the cities whose heroes were put centre-stage. The purified Orestes of Aeschylus' Oresteia who returns to rule in his native land represents a desire to tell the pre-history of the Argives sympathetically, while the guilt-ridden and permanently exiled Orestes of Euripides represents the opposite aim.

Now we can move on to the textual tradition. As the example of Phrynichus' Capture of Miletus demonstrates, the Athenian polis took an interest in what went on in the tragic theatre. From this moment until the 330s, we have a fair amount of evidence which show that this concern continued. Legislation that we know of after the Phrynichus episode concerns the right to produce plays by dead dramatists at the major Dionysiac festivals. Soon after Aeschylus' death comes a decree allowing a chorus to anyone who wishes to produce an Aeschylus play and we know that many dramatists won prizes with such revivals. In 387/6 we have evidence for the inception of a competition with "old plays" and thereafter several concrete instances of especially Euripidean drama being revived at the festivals. Most important of all, we know that probably in the 330s Lycurgus passed a decree which ordered the composition of a state edition of the plays of Aeschylus, Sophocles and Euripides and forbade actors to deviate from the texts in the official collection. This tells us that there had been a great deal of interference with these plays in revivals and that the city had a good reason for wishing to restrict it. Other sources confirm this picture.

So far, scholars have tended to see this decree as ecologically inspired, a sort of 4th century tragic conservation measure, designed to protect ancient monuments against aesthetic pollution. But I have another interpretation to offer. If tragedies were in the first place connected with the use of "myth/history" in political debate, then they may have continued to be so used when revived. After all, the geography of Greece had not changed and the major new player in the fourth century, Macedon, made sure that it connected its origins deep into this myth-historical past by tracing its ancestry back to Argos. The advantage of using a classic dramatist's play to tell "how it really happened" (according to your own political standpoint) was precisely that it conveyed authority upon the version. But in order to make sure it really was your point of view (your ideology) which shaped the story, you might need to do a pretty large touch of rewriting. It is as well to recall that tragedy was alive and well as a craft in 4th century Athens and that it ran, like carpentry and masonry, in families. A direct descendant of Aeschylus was writing tragedies in the mid 4th century.

What I'm saying is quite radical, then, and I had better be quite clear about it before we continue. Because of the use of myth/history in political debate and the possibility that such debate was linked to the divergent tellings of these stories in the tragic theatre, Greek tragedies may very well have been subject to "ideological interpolation", even complete revision, from the mid fifth century onwards. By the time the rot was stopped, the "damage" had been done. The purpose of Lycurgus' decree will have been to stop the use of "old" tragedies as pawns in the increasingly dangerous internal political conflict within the city as it faced the prospect of direct interference from the great power to the North, Philip the II's Macedon. For the purity of the textual tradition, I guess they did not care two figs, if they had not minded about such revision before Lycurgus. Hence, versions transcribed into the state copy may very well still contain the interpolated material. Some obviously do.

The next step is to test the hypothesis. To do this we will need to find a play with political material over whose interpretation there has been unresolvable debate. We will need to have evidence that the play had been revived after the dramatist's death. Then we will need to look around that date or dates to see if the political debate has anything in common with the problematic material. If its application can be made to illuminate the play and its removal to reveal a different set of aims in the story's construction, then we may have isolated an "ideological interpolation". Let us see what we can do.

III

My title is a giveaway, of course. The Eumenides conforms to my criteria in all three respects.

(1) there is internal evidence of reperformance in revised form at the very least in the alternative stagings of lines 404 and 405; and, in my opinion, there is external evidence for two reperformances; (2) there is prominent ideological material in it, in particular the Argive alliance and the Areopagus, but also a claim to Athenian territory in the Troad and an appeal against civil war;

(3) there is a venerable and quite unresolved debate about the position adopted in the play in relation to these ideological matters.

The main difficulties are these:

1) In 462/1 the Areopagus had been deprived of its wider powers by Ephialtes and remained merely a homicide court. Yet the establishment speech of Athena appears to ascribe to it powers which are far wider than would fit into that remit. In particular, lines 700-706 draw parallels with Scythians and Spartans, seemingly on the basis of similar councils. Since both were renowned for eujnomiva ("obeying the laws"), this looks uncomfortably like a reference (expressed in the "heroically vague" manner alluded to by Prof. Easterling as characteristic of tragedy) to the function of nomophylakia "care for the laws" taken away by Ephialtes and given to a special and short-lived board of nomophylakes "guardians of the law". Did Aeschylus oppose the reforms, or can the lines be interpreted as implying that he supported them. Or was he trying to reconcile the opposing factions?

(2) The Argive alliance of 462/1, which replaced a traditional tie to Sparta, was a policy of the demos and to some this means that the attitude it implies to the Ephialtic reforms of the Areopagus was positive. Others say that there is no need for agreement on internal and external politics to be joined in this way. However, the main difficulty with the alliance lies in the details of Orestes' prophetic farewell speech at 763-74. For this has all the hallmarks of being constructed around an already past historical scenario in which alliance existed, invasion of Athens by descendants of Orestes was nonetheless possible, but the alliance prevailed and the invasion never took place. It is possible that there was doubt in Athens in 458 about the firmness of the new alliance. We do hear of some anti-democratic Athenians trying to upset matters, among them the alliance, by inviting a Spartan army to travel through Attica, and that the Athenians went out against this army in full force to fight the battle of Tanagra (Thuc. 1.107-8). But

(a) we do not know the date of this battle (458 or 457 is all we can manage). So we can't be sure whether it had happened before or after the first performance of the Oresteia;

(b) anyway, 1,000 Argives went with the Athenian army. This does not betoken any sort of current feeling of distrust of the alliance in either city at around the time of the play's first production;

(c) in any case, our evidence from Thucydides relates to Athenian dissenters, not Argive dissenters.

(3) The claim to a "great portion" of land for Athens in the Troad is taken to refer to Sigeum. But (1) Sigeum was not an Athenian possession at this time; (2) in any case, the claim seems to be for a larger chunk of territory than merely Sigeum.

(4) Civil war. The evidence of Thuc. 1. 107-8 does suggest that a small anti-democratic group was operating in Athens at around this time. I'm not certain that on its own this is enough to justify warning against stasis, especially since the passage has certain technical problems attached to it which suggest it may be interpolated.

This play, then, makes an excellent test-case. So let's look at the two important categories in turn: (i) reperformance dates (ii) compatible ideologies at those dates.

(i) At Aristophanes Frogs 1124, Euripides asks Aeschylus for the prologue from Oresteia, which turns out to be the first few lines of Libation Bearers, the second play of what we call the Oresteia. At Clouds 534, what appears to be the recognition scene from the same play is recalled. I am among those who disbelieve in a wide reading-public for tragedies at this period. It seems to me that the only way in which such references could be used as the basis for humour is if the audience of comedy can be assumed to have seen them on stage within at least a reasonable time-span. What that would be can be judged by the gap of 21 years between Euripides' Telephus of 438 and the Aristophanic parody in Thesmophoriazusae of 411. So it would not be untoward to think of a reperformance of Choephoroi in the period before c. 415, and, if Burkert is right to redate Euripides' Electra to 420, before 420. There seem to me to be three reasons why one should include Eumenides in this revival: (a) Libation Bearers is not complete by itself; (b) the title Oresteia seems more adequately to cover the action of the two plays, rather than only that of Choephoroi; (g) Euripides' Electra contains rebuttals of the version of events which appear in Eumenides.

The argument for a fourth-century reperformance has a similar basis. The comedy of Timokles entitled Orestautokleides has a character describing the way the pederast Autokleides is sitting surrounded by a group of prostitutes (fr. 27 PCG). This is a parody of Eumenides which relies specifically upon the audience's ability to recall the play's opening. A well-known 4th century Italian vase shows this very scene (Apulian Kelchkrater, Leningrad St. 349). Webster dates the play to the 320s, but recent work I have done on the changes from Old to Middle and New Comedy convince me that the play cannot post-date the early 330s, if, as seems likely, Autokleides was represented on stage as Orestes. A passing reference in Aeschines (I.190-1) to the Poinai chasing wrongdoers across the stage with lighted torches could possibly relate to the reperformance implied by the Timokles play. Another in the same speech, delivered in 345, to Autokleides' pederasty shows that we are in the right period. In any case the reperformance ought to lie somewhere in the 350s or 40s.

(ii) Of the four ideological foci of the play, only one is of real relevance to the first revival date in the 420s. That is the Argive alliance. Sundry scholars have in fact in the past assigned certain of the passages which bring this into prominence in the play to a revival at this time. However, the question of the Areopagus as "guardian of the laws", which seems to be what lies behind 700-6, was not relevant during this period. This question only comes into prominence again after the restoration of the democracy in 403. In the 420s, the Areopagus is merely a homicide court. Likewise, in the 420s, during the Archidamian War, Sigeum and the Troad generally were not an area which caused problems. Nor is the question of civil war high on the agenda of a city already at war with powerful external enemies, though of course, one might say that stasis was always a possible scenario for any Greek state.

When we turn to the 350s and 340s, however, we are confronted by a different picture.

(1) The Areopagus has now once more become the centre of ideological debate. Indeed, in Isocrates' Areopagiticus we have a full presentation of a view of the Areopagus as the main institution of the original democracy. As Bock has shown, this is so close to the thesmos speech of Athena in Eumenides as to be virtually a commentary on it. Removal of the elements in the speech which relate to this ideology cannot be done neatly. But it would be a crucial finding of this hypothesis that such large-scale remodelling was in fact the norm. However, that said, when one sets aside the elements which correlate with the Isocratean ideology, one discovers the outlines of a speech whose purpose in situ is clearly to persuade half of the jury to ignore the plain facts of the case - which has been made for Orestes by the evidence and exegesis of Zeus' spokesman Apollo - and vote to keep within the city the element of fear, viz. the automatic punishment for certain offences whose agents are the Erinyes.

In my view, this whole side to the play, including the metamorphosis of Erinyes into Semnai Theai, is communicated to the audience through an existing religious/mythical schema. According to this, it is Horkos, the realm of "oath", but personified elsewhere and in my view here too, which unites the Erinyes and the Semnai in a common system. Thus the broader powers of the Erinyes claimed already in the second stasimon and enclosed by Athena within the terms of the thesmos of the Areopagus, are visibly related to the areas normally associated with "oath" in Greek literary sources. The hypothesis of ideological interpolation helps us, then, to re-establish according to available religious/mythical schemata, a disrupted dramatic strategy of the play.

(2) Argos is not in itself of importance in the political discourse of this period. But I have already mentioned the way in which the changing political geography of the fourth century Greek world is accommodated within the political use of myth. The growing power of Macedon, the central feature of this period, was so accommodated. The Macedonians had always traced their origins to Argos (as Isocrates reminds Philip in a letter: "Argos is your native land"). Of course, it is true that normally they gave their ancestry via the Temenids and Herakles, who was easily presented as a positive figure, "the cleanser of Greece". Notice how Speusippus points out to Philip that an incident from the life of Herakles gives him an excellent claim of Ambracia, which Philip had attacked (FGrH 69 F2 [Antipatrus]). However, there was in Macedonian territory a place called Oresteion. And there is some evidence, difficult to date unfortunately, but possibly reflecting an ideological battle of the mid-fourth century argued out in mythical terms, which shows that Orestes could be thought of as the mythical progenitor of the Macedonians (Theagenes v. Hegesippus of Mecyberna).

So it is entirely possible, given the model we have constructed, that in the discourse about Macedonian power being conducted at this time in Athens, some were saying that it was not Herakles, but Orestes who was the progenitor of Macedon. This will originally have been an attempt to blacken the Macedonians and to prevent the citizens from voting to ally with them. Revival of this play, then, will have served as a part of the discourse of defence of such an alliance, conducted in the enemy's terms.

There are two external reasons to think that this is a reasonable account of the matter.

(A) During the period between the fall of Olynthus and the Peace of Philokrates (348-6), the Athenians waited with apprehension to see whether Philip would press home his advantage against Greece itself, and Philip himself made overtures and offers of alliance to Athens. In the political discourse of the time, two opposing factions formed. One (Demosthenes' crowd) violently opposed making peace and alliance with Philip. The other (Aeschines' lot) were keen on peace and alliance. The Athenians voted for peace and a perpetual alliance (cf. Eumenides) in 346. But very soon afterwards, events led to a major panic in Athens, during which it was believed that Philip, despite the alliance, was poised to invade Athens. He chose instead to honour the alliance.

It is easy to imagine how in such circumstances a threat made in the mythical world of drama, couched in terms applying to the present time of the audience, could take the form it does in Eumenides. The Athenians have the assurance of Orestes himself to set against the antagonistic voices of Macedon's enemies, that the natural state of affairs is for his descendants to support Athens and they have the evidence of recent history that Orestes had so used his influence. The play's reperformance, then, could hardly be before Dionysia 345.

(B) We have the solid evidence of the didaskalic inscriptions for the years 342-340 (IG ii².2320) that two plays of Euripides relating to the Orestes story were produced in the same period we require to locate the reproduction of Eumenides. They are Iphigeneia (which one we don't know, but probably in Tauris) and Orestes. Furthermore, one of the new plays of 342-1 was an Orestes. Our model suggests that the choice of mythical subject had political import. This tends to confirm the circle of interpretation according to which Eumenides was revived at this period with ideological interpolations.

(3) The Athenian claim to Sigeum on the basis of their participation in the Trojan War dates back some way (Hdt. 5.94.2). Sigeum itself is of no importance at this period. However Ilium was captured and claimed for Athens c. 360 by Charidemus of Oreus (Tod GHI no. 148, pp. 147-8). More importantly, perhaps, Isocrates, whose ideological claims about the Areopagus we have already seen emerging from Athena's thesmos speech, was very keen that the Greeks, including Macedon, should unite to fight Persia (Letter to Philip, after 338). Within that contemporary ideology, the age-old Athenian claim to a slice of territory in the Troad (of which Ilium would only be a part) makes excellent sense.

(4) About the theme of civil war, it is perhaps enough to recall the violence of the two groups who opposed one another in Athens at this time over the city's policy towards Philip. It is worth noting in addition that Isocrates argued often for homonoia ("unity of purpose") and "harmony" as political principles.

It is also worth adding that we can name at least two contemporary tragic poets in the circle of Isocrates who might have been competent to produce pseudo-Aeschylean reworkings. Astydamas, winner at the City Dionysia of 347, was not only a pupil of Isocrates, but also a descendant of Aeschylus. The other poet, Aphareus, also a politician, was Isocrates' son-in-law. It cannot go without saying, however, that this whole hypothesis not only challenges the conventional view of the inviolability of the texts of the fifth century dramatists but also requires that poets could do passable imitations of the styles of the earlier tragedies. The detailed literary parodies of Frogs assure us that this would not have been impossible in principle or practice.

We can now return briefly to the ideology of the 420s version. Once we wipe away the accretions, though again I emphasise that textually speaking, this cannot be done with neatness, we find at the centre of the play's ideological concerns the Argive alliance. For even when we take into account the fact that at least two of the passages which specifically mention this are suspect, one of them for ideological reasons - which may of itself bring the other two with it - we must accept that the ideological purpose of emphasising mutual links between Athens and Argos, and in particular Argos' debt to Athens, are deeply inscribed on the play. To mention the major movement of the plot only, I have argue in a forthcoming paper that the peculiar treatment of pollution and purification in the play can be understood only on the premise that a central goal of the drama is to have Orestes reinstated as King of Argos with clean hands. In terms of the sort of argument used by Kallistratus in Arcadia, and those we have conjectured were being used about Tereus, this corresponds with an attempt to sanitise one of the possible mythic progenies for historical Argos.

IV

I don't have the space here to deal with the detailed results of the application of my overall hypothesis to Eumenides. But it seems to me that they are promising, even though they are far-reaching. Just how disruptive of standard views the model would be I can briefly reveal by reflecting on the 420s version revealed by my ideological archaeology. For, as I have already said, this version ignores the Ephialtic issue in relation to the Areopagus and is constructed around the specific object of vindicating Orestes as unpolluted King of Argos. How fortuitous that an Aeschylean play written, one supposes, for a very different set of political circumstances in relation to Argos, should so exactly fit the needs of its producer in the 420s.

I will not try your patience any further with the major heresy implicit in my last statement. I want to end with a brief summary of the implications of this hypothesis. At least some of these plays have come down to us in their ideologically interpolated versions, including, of course, Eumenides.

We do not know whether this is true of all. Nor, I add here to comfort those who believe still in a non-political tragedy, do we know whether this was the only function served by tragedy. After all, in comedy, Krates was completely out of kilter with contemporary uses of his genre, as Aristotle makes clear. Nor do we know whether there were other means by which tragedies could produce political meanings, perhaps more directly. For example, Euripides' Troades is often thought to relate in a critical way to the Peloponnesian War and specifically to Athens' recent treatment of Melos. This would imply not so much a genealogical link with the past, as a kind of allegory. I merely throw in here that it had been a comedy, the Dionysalexandros of Cratinus, produced in 430, which had first made the link between the Trojan War and the conflict with Sparta in the theatre of Dionysus.

It will be necessary to subject whatever plays we can assign even an approximate revival date to the sort of examination offered here. In view of the apparent importance of the Orestes myth, it seems it might be best to begin with Euripides' Iphigeneia in Tauris and Orestes, both of which have interesting relationships with Eumenides, were revived in 342 and 341 respectively, and both of which are extraordinarily close to the style of New Comedy which so obviously grew out of Euripidean tragedy.

Note