By Robert J. Cardullo

I came to the title of this essay not long ago by way of sitting in a theatre which had the words "a mirror up to nature" inscribed on the centre of its proscenium arch high above the stage. Like all compulsive readers, I read the words again and again while the house lights were up. At the same time, I let my gaze follow along the line of the arch, which was made up of moulded and gilded flowers, among which gilded putti appeared in frolicsome attitudes. After the house lights had come down and the play began, I continued to glance from time to time at a particularly aggressive-looking putto who was picked out in silhouette by the stage lights. His body was turned away from the events I had come to witness and, as I followed the line of his steadfast gaze into the shadows, I was diverted. It was thus that I began to wander down the path of my introspection.

My thoughts turned to the mirror and nature and whence their juxtaposition had come, and I remembered Hamlet saying that playing's end "both at the first and now, was and is, to hold, as t'were, the mirror up to nature." When Shakespeare has Hamlet say "both at the first and now," he is going back, whether he knows it or not, through Horace's Ars Poetica (circa 18 BC), through Philodemos, and finally to Aristotle and that much-discussed literary treatise, theoretical document or teacher's lecture notes – whichever it may be. I mean, of course, the Poetics (334 BC), which, in turn, may or may not be an accurate representation of the thrust of dramaturgy in Athens of the fifth century BC

Many scholars have interpreted the Poetics. I like the remarks of Roy Hack writing in 1916 in the Harvard Studies in Classical Philology. Hack is mostly concerned to discuss the Ars Poetica, but he has perceptive remarks to offer about Aristotle and realism, about the nature and origin of the theory of mimesis. Aristotle, argues Hack, wanting to save both the baby and the bath water, wilfully perverted Plato's fulminations against poetry in the Republic (360 BC). And the essence of Aristotle's transformations of his predecessor was to establish the moral imperative that tragic drama must absolutely mimic reality. Plato, you will remember, objected to poetic creation because it was one remove too many from the grand reality of Ideal Forms, which are laid up in Heaven. There was the earthly imitation of the table, which Plato could barely tolerate, and then there was the poet's bastard creation of it, which he certainly could not. This, we must remember, was Plato's formal, logical objection to poetry, which in the end may be no more than metaphor, one philosopher-poet's contrivance of words for contrivance's sake. But Eric Havelock has touched upon a more serious objection that Plato must have had, a natural objection for anyone intent on establishing an intellectualist way of handling reality – that is, poetry's, or in this case drama's, lack of objectivity. And one can imagine others: for instance, the aristocratic Plato's fear of an educative art form's getting out of control, which tragic drama clearly had done by the close of the fifth century.

Nevertheless, Plato's image of a bastard or faulty reality gave impetus to Aristotle's theory of perfect imitations, which in turn became enshrined in the minds of western Christendom's dramaturges. It is not clear, however, that Greek tragic drama was ever intended to conjure up reality, to mirror nature, to be "true," and therefore one contemplates other metaphors. There is always the phrase pompa diabolic, which was used by the Church fathers to describe Attic tragedy. But the distillery metaphor comes particularly to mind, since it seems to me that the real effort of ancient tragic drama was to make over events in such a fashion as to make them safe, distant, no longer palpable and menacing – in sum, to shatter or to shroud any mirror in sight.

Consider, by way of analogy, the last scene and the ultimate scene of violence, in Arthur Penn's movie Bonnie and Clyde (1967), when the hero and heroine drive into an ambush and are machine-gunned to death. It is a long scene, showing the two characters riddled with bullets, blood spurting out of dozens of punctures, their bodies writhing in death-agony as they are cast up by the force of the repeated bullet impacts. And yet, and yet… it is all so beautiful, shot as it is in italicising, aestheticising slow-motion, and featuring two beautiful people, the young actors Faye Dunaway and Warren Beatty, in the roles of Bonnie and Clyde. There is a dance-like quality to the action and, besides that, a sensual rhythm of intercourse – of the two bodies in their coupled rising and falling. Here are the grace, the sexual release and the lyricism that our heroes were really aiming for as they committed criminal mayhem across the American Southwest.

This sensual, choreographed, almost beatific scene does not exactly match up, however, with contemporary photographs of the event (which took place in May of 1934) or with the homely looks, let alone the psychopathic natures, of the historic figures of Bonnie Parker and Clyde Barrow. It is certainly so much a violation of the moral implications of the film's earlier scenes – in which innocent people are killed and their money or property stolen – that it can only be called an instance of supreme, not to say divine, decadence. And decadence, I might add, that by being absorbed almost immediately into popular American culture through memorialising or iconising song and fashion, as well as material spin-off of every other conceivable kind – Bonnie and Clyde was one of the first movies to be "merchandised" – bore witness to Oscar Wilde's witticism that the United States is the only country in history to have passed from barbarism to decadence without ever stopping for civilisation in between. In any event, the final scene of Bonnie and Clyde shows the way in which you can fabricate an overall moral statement for your narrative in the very aesthetics of your own filming of it – particularly so momentous an event as the execution-by-overkill of our hero and heroine.

Something similar seems to be at issue in ancient Greek tragedy as well, though it's true that most of its scenes of violence were kept offstage or viewed only in their aftermath. I'm thinking especially of that most famous example of Greek tragedy, Oedipus the King (430 BC), the whole of which is as unrealistically dramatised as the final scene of Bonnie and Clyde and, for its own moral purposes, thus shatters or shrouds any mirror in sight. As we all know, Greek tragic drama of this kind – though Oedipus the King seems to be filled with more improbabilities than any other – took place during a religious festival. The Greater Dionysia was essentially an open, state-cult holiday with very little of the ritualistic or mysterious aspect of other religious days. Some students of ancient religion are therefore moved to say that it was not truly a holiday – a holy day, that is, in any serious sense of the term. But then again, we may ask ourselves if perhaps the ritual or mystery that satisfied the religious instincts of the ancient Greeks at other times was not put aside on the days of the Greater Dionysia for something else, for a celebration that equally, if indeed not more so, spoke to their spiritual needs – I mean the enactment of the dramas. As Bronislaw Malinowski has remarked, primitive religion is not to be believed, it is to be danced.

Greek tragic drama thus seems to be a special literary development, one that we rarely find elsewhere than in ancient Greece. One notes in this context the great number of early cultures that created epic or heroic narrative, which appears to be the spontaneous reaction of humankind to events of a military or political kind – real or imagined – at a particular period in time. Tragic drama, by contrast, seems to be essentially the Greeks' creation, as though it had sprung up to fill some need peculiar to their culture. It never appeared in any important way among the Romans, who in almost everything else enthusiastically imported and adapted Greek cultural forms to their own requirements.

I should come right out at this point and say that this special need for which the Greeks created and sustained tragic drama was one that in other cultures religion satisfies. We may note that in the fourth century BC when Isocrates, Epicurus and Zeno created philosophical systems that provided ethical modes and quasi-theological notions, tragedy died, as though it were no longer needed when set side by side with these emergent species of religion. At the same time, ancient tragedy's heir, new comedy – into which tragedy evolved – became exactly that mirror of nature which Shakespeare's Hamlet seemed to desire. Remember, Aristophanes of Byzantium said of the comic poet Menander: "Which is life, which is Menander?"

I have already used the term "Greek tragedy" quite a few times in this essay, as though I knew what it meant, or that you, the reader, did. Dangerous business, this, and I ought to be honest and admit it. To generalise about ancient Greek drama requires considerable temerity, for we know so little, and so few plays remain as a statistical sample for our generalities. Moreover, the plays are so different – each of them from the others. Ulrich von Wilamowitz-Moellendorff was right to say that the only honest definition of tragedy is something like a dramatic narrative habitually enacted on a certain set of days in ancient Athens. Be that as it may, the one generalisation that holds true for more plays than not is that ancient tragic drama tends to depict a line of action that goes from good to bad or sometimes from bad to worse. Put another way, we all must die and life doesn't work. I think these are the two truths – religious truths, or perhaps religious problems – with which Greek tragedy grapples.

The problems of death, of evil, of deserved or undeserved suffering are, as we all know, topics that pervade the Judaeo-Christian religious teachings. But where would we find such teachings in ancient Greek religion? From what little we know, the Greek religion was a collection of systems designed to offer immediate psychological transformation through ritual – I think of the Eleusinian Mysteries – or to ensure future, beneficial action through what amount to contractual arrangements, of the kind found in votive and sacrificial practices. But a body of theology we do not find. We all recall that there was neither any dogma in ancient Greece nor any priestly caste to enforce it, and that therefore the poets were free to contrive the Godhead as they chose. Greek religious practice itself seems to have evolved principally to help achieve an empathetic response from nature, but not to try to make meaning of nature.

It fell to the poets, or so it seems – and whether by design or not we cannot tell – the task of making meaning. Martin Pierrson Nilsson once said that religion is man's protest against the ultimate meaninglessness of life. And perhaps we might say, by analogy, that for the ancient Greeks, art is man's protest against the ultimate meaninglessness of life, that this was the function of art right from the start. (We can see this function in action in two great pieces of lamentation: one from the Greeks and the other from Old Testament – Hesiod's Works and Days [circa 700 BC] and the Biblical Ecclesiastes.) Greek tragic drama itself seems, by its very form – its structure, organisation and finally artistry – to mute the fundamental horror and despair of human existence.

But first, a few words on the social value of Greek tragedy may prove helpful, at least on that tragedy as the Athenians perceived it. We must remember that tragedy in ancient Greece was part of a state festival, funded by the state entirely, and that the competing playwrights were selected by a city official who himself had been chosen by lot. In other words, we are not dealing with a literary clique or an elite social class when we talk about the production of ancient Greek drama, but instead, an art form that was popularly satisfying. Furthermore, we are dealing with an art form that remained remarkably static over almost a century – a period of time during which Greek thought itself changed quite a bit. It is as if the Athenians had a compulsive need for the tragic form, then; and this, again, is analogous to the need for religion, which is notorious for the way in which it remains true to form over long stretches of time.

Pursuing this line of thought, I would like to posit the idea that tragedy evolved as it did principally to deny the truth of what it seemed to be saying, or at least to make it more palatable. In a way, one could even say that Greek tragedy is like the paintings of René Magritte, where the conventional representation of thoroughly preposterous scenes is immensely reassuring in that it sanctifies or solemnises madness – or of course the reverse, that it makes the found world mad. (I suppose this all depends upon the viewer's own grip on sanity!) And we can sometimes get a better idea of what tragedy is all about when we are shocked by its violation of its own conventions of representation.

Euripides is particularly good at this, pulling us up short with the unexpected so as to let us see how much, in fact, the tragic form conceals or represses or denies. Indeed, there are moments in the latter part of Euripides' career when it becomes clear that the form cannot be held much longer to portray what it was meant to portray. I'm thinking particularly of the Orestes (408 BC), in which there are two or three of these unsettling moments. The most dramatically successful one concerns the character of the Phrygian slave, who speaks the kind of broken Greek we might imagine a newly-arrived immigrant to speak – but which, of course, no other tragic character has ever done. The shock of this fracture in the otherwise smooth façade of dramatic contrivance provided by Euripides is made the more profound by the very mockery of convention tragic diction contained in the slave's words. He is trying for the grand tragic tone, but he can't get it. He does manage to fill his broken syntax with a number of literary or mythical allusions – the "bull-horned ocean" comes to mind – but in the end the Phrygian slave offers up no more than a mangled messenger speech that breaks down when it attempts to strike a high note (as in "All right, ladies, if you want the details, then come into this pride of lions" [Orestes and Pylades]).

This language is ludicrous, naturally, but perhaps no more so than the subsequent arrival of Apollo in his role as deus ex machine, who forces the characters – some of them at each other's throats – back into the pattern of marriage-and-alliance that the myth demands. The improbable moves to the absurd here as a kind of advertisement that we need no longer suspend our disbelief. And all the more so because Euripides has already called the myth into question in this play, when he has Orestes and Electra decide to murder Helen, and with this bold stroke abstracts from the myth the idea that these two are not so much the tragic, star-crossed avengers of their father as they are morally delinquent murderers. Everything about the Orestes is like this: dangerous, subversive, as Euripides frightens us with his taunting and mockery. The tragic form is thus made to be suspect of a great lie.

What I have said about Orestes might also be applied to Euripides' Electra (413 BC), principally the scene where Electra discounts the value of the identification clues that her slave has found at Agamemnon's grave. Whether or not Euripides expects us to remember the same identification scene in Aeschylus's Choephoroi (458 BC), we may note that, as Aeschylus presented the clues that Orestes was still alive, they had an inexorable, fairy-tale quality. They are to be accepted because, like the three questions in Gozzi's Turandot (1762), they provide the magical turn in the lock that at this moment is meant to be opened. Euripides' heroine, however, dismisses the grave-site clues for logical reasons which in their probing, logic and reason – even their intellectualism – deny everything that tragedy is about.

Two tragic attitudes are being attacked by Euripides here, attitudes that go to make ancient tragedy the comfortable art form I imagine it to have been. One has to do with distancing (what Bertolt Brecht later called "Verfremdung", which could also be translated as "defamiliarisation") and the other with the dramatisation of inevitability. Let's begin with distancing and the absence of it in the Phrygian slave's speech from the Orestes. For this speech demands that we recognise the slave's speech as peculiar to him, that he is therefore a specific person situated in time and space, that he is therefore an accident – not a universal – and exists before our eyes at the moment we see him. In other words, we are forced to get too close. Similarly, the speech of the deus ex machine, by its shocking absurdity, asks us to judge its words against preceding events and in turn to invest ourselves in the action – again, bringing us too close to it. But we are brought closest to the action when Orestes and Electra walk out of the myth and perform, or attempt to perform, the very same action from it – namely, the murder of a mother figure – before us without, as it were, their costumes and make-up. I am sure that this play, in production, occasioned wave after wave of nervous titters. As Walter Kaufman says in Tragedy and Philosophy (1968), "Sophocles' heroic figures are inspiring; his perfection comforts. Euripides makes his audience squirm."

Electra's dismissal of the means of confirming Orestes' continued existence, on the other hand, is far more nerve-wrecking, because it argues for the character's taking control and moving into the crazy no man's land of decision-making and change, where everything is possible and nothing is known. Suddenly every action matters and is freighted with potential. Clearly enough, there is no momentous change to the myth in this play, but the suggestion of such change is always there, and is a tantalising alternative to the otherwise fixed train of tragic action. In this way, Euripides was attacking – as he often did – the notion that the structure of reality must be tragic. Contrary to E.R. Dodds's belief, as expressed in his 1929 Classical Review essay titled 'Euripides the Irrationalist', the dramatist only sometimes opposed the claim that the structure of reality must be rational.

I should say here that the principal mode of inquiry in the fifth century BC was through the tragic form, hence that the principal structure of reality was indeed tragic – which means an aesthetic juxtaposition of realms, not a logical one. The sophistic movement applied an increasingly rational structure to reality, which cause the tragic structure to grow irrelevant until finally Euripides became bored with it. But, of course, he was a great artist, more an artist than some early version of George Bernard Shaw, as some would have it, and he could always return to the tragic structure, as he did most successfully at the end of his career with the Bacchae (405 BC).

Let us now examine the matter of inevitability and the tragic form. Inevitability is the most obvious feature of ancient Greek tragedy and an often discussed one. Perhaps it has never been put more brilliantly than by Jean Anouilh in the mouth of the chorus figure from his Antigone (1944). I give this speech to you in Lewis Galantiere's English translation; it comes after Antigone has fatally defied Creon and the die is cast, or her fate is sealed:

Tragedy is clean, it is restful, it is flawless. It has nothing to do with melodrama, with wicked villains, persecuted maidens, avengers, sudden revelations, and eleventh-hour repentances. Death, in a melodrama, is really horrible because it is never inevitable. The dear old father might so easily have been saved; the honest young man might so easily have brought in the police five minutes earlier. In a tragedy, nothing is in doubt and everyone's destiny is known. That makes for tranquility. There is a sort of fellow-feeling among characters in a tragedy: he who kills is as innocent as he who gets killed; it is all a matter of what part you are playing. Tragedy is restful; and the reason is that hope, that foul, deceitful thing, has no part in it. There isn't any hope. You're trapped. The whole sky has fallen on you, and all you can do about it is to shout.

The surviving Greek tragedies themselves are all so different from one another that it is both hazardous and presumptuous to generalise about them. But our evidence allows us to declare at least one general truth about Greek tragedy: Athenians preferred traditional, well-known stories as the subject matter upon which the dramas were built. This fact is crucial to what Anouilh is saying about the tranquility of tragic drama and, by extension, the mood of the ancient Athenian spectator. Here, then, is the stuff of tragic irony, that foreknowledge of events which spectator and god possess but which the characters in the drama do not. Sometimes, this foreknowledge is reinforced in a prologue speech that outlines the plot to come, such as what Aphrodite delivers in Euripides' Hippolytus (428 BC). But even less directly expository prologues are invested with an air of knowing expectation.

Coming from a culture that values surprise – a value stemming from life viewed as an act of free will – we have trouble accommodating our criticism to a dramatic form like ancient Greek tragedy, which is the constant reiteration of the commonplace and the commonly known. The use of familiar stories with established endings sets the tone for viewing subsequent events; and the initial, premonitory choruses serve the same end: they tell us that everything is already known, that no new information will be forthcoming. Consider the chorus of Euripides' Alcestis (438 BC), which enters the stage ready to mourn the dead or dying Alcestis, even though there is no way they could have known that this is the day on which she will meet her death. The implicit sense of déjà vu or entendu that the spectator thus brings with him allays all fears, dulls suspicion, palliates horror and terror by taking the tension or suspense from the drama. The turn of events, no matter how horrible, will be satisfying because those events are the essential fulfillment of what is already known or agreed upon. This is the formal aspect of the religious solution for which Greek tragedy was created in the first place. No wonder that the character whom Melina Mercouri played in the film Never on Sunday (1960) was able to describe the dénouement of every tragedy she remembered as "And then they all went to the seaside."

The tragedians' disinclination to use original fictions, their use instead of repeated stories, makes the spectator safe in his knowledge, even as the protagonist is threatened in his ignorance. In a sense, we, the audience, are the angels, for whatever happens will not be unanticipated by us. We are thus forearmed. Consider, by contrast, the absolutely devastating shock to the spectators' sensibilities at the very end of King Lear (1605). In a play that depicts, it seems, every imaginable form of brutality and suffering, Shakespeare reserves his most horrific coup de théâtre for the dénouement. At precisely the moment when things appear for once to be going well – when the evil characters have been vanquished and some sort of justice, however partial or provisional, seems to govern the universe – "enter Lear, with Cordelia dead in his arms." Not only does this strike us a cruel, arbitrary stroke of fate – for we identify so much with Lear and by extension his only loyal daughter, Cordelia – but it jolts our sense of aesthetic justice, too.

It's easy to see why the 20th century, with all its political turbulence and turbulent experimentation in the arts, took Lear to heart, while other centuries, most notably the 18th, have rejected or rewritten it. Samuel Johnson, a stickler for form if ever there was one, said in the preface to his 1765 edition of King Lear that he could not endure the play, as its grim view of the world offended him so much. But human existence is horrible and horrifying, pain is undeserved and life doesn't work – these are tragic truths. One can get to them, however, and deal with them if there are no surprises, no sudden lurches from the track, which imply "might have been". There is the true horror: the idea that events could have been otherwise. Ancient Greek tragedy protects us from looking into the deep, dark, frightening well of historical as well as artistic alternative by constantly insisting that whatever is, is, and it must be so.

Another remarkable generalisation we are permitted to make from our scanty evidence about ancient tragedy is that the Athenians preferred that the playwright never depart from an absolutely rigid sequence of choral odes and dialogue passages. I have to insist on the matter of preference because sometimes critics work from the notion that the fixed tragic form was a tiresome encumbrance around which the adroit dramatist cleverly manoeuvred. Not at all, however. The alternation of choral odes and dialogue, the one-liner ping pong of stichomythia, the interchange between actor and chorus in the kommos – these formal elements are the essence of tragedy. They reinforce the sense of inevitability that I have been discussing.

Like the great identification scene in the Choephoroi, where the clues fall into place despite their unlikelihood – fall into place, that is, because the drama's unbroken circle of events triumphs over the fragmentary nature of the perceived world of experience – the dramatic world engendered by the dialogue scenes will inevitably be discharged in the choral odes. In this way, no dialogue will ever run its course, ever go freewheeling and blundering into darkened rooms or musty closets hitherto unopened. The choral ode is always there at the fail-safe line, waiting to block such progress. The inevitability of the juxtaposition of ode and dialogue itself is reinforced within the odes by their arrangement into strophes and antistrophes that have exactly repeated metres. Thus what is said in the ode is caught and returned formally, as minor resolutions are achieved through rigid symmetries.

I have remarked that the spectator is made safe while the protagonist is threatened. An important feature of Greek tragedy is that the spectator is kept just that: a spectator, freed from the beguiling seduction of empathy. This has to do with distancing. How often the chorus of ancient tragedy takes the spectator away from the play, calms him, defuses whatever emotion the spectacle onstage might ordinarily inspire. One thinks of the choral response to Denaneira's agony in Sophocles' Trachinian Women (413 BC), in which the chorus sings an accepting ode on the inevitability of the events they have witnessed, on the fact that things have happened as they must, and then quite helplessly ask, "What can we do?" to be followed a few lines later by the wish, "May a wind carry me away –" The chorus, which is a social group, functions here as a paradigm for the spectator group, effectively distancing themselves and us from the action through acceptance, passivity and finally thoughts of flight.

Or one thinks of Medea's agonised, even tortured, plotting in the Euripidean play (431 BC) where her wickedest thoughts are spoken aloud to the chorus, who placidly hear them out, remonstrate in the lowest key and then continue into an ode which treats the killing inside with so much tangential lyricism that the tragic event gets away from the spectator. The textbooks tell us that choruses such as this one became an increasing embarrassment to the playwrights, who were growing more interested in realistic dramatic situations. But the embarrassment is more in the eye of the modern beholder. Certainly the Athenian community demanded choruses and the dramatists not only wrote them in but insisted upon their presence through the brilliance of the lyrics that they put into choral mouths. Euripides was a maestro of the choral ode. And in the instance I have cited, the chorus's tolerance of Medea's plans takes the sting out of the evil action, denies the seriousness in reality of its consequences. Nothing here is severe enough, it seems, to get a reaction from the chorus. And we the spectators are instructed to react with equanimity as well. The alternative is to imagine the chorus to be a monstrous, complicit band, or as inconsequential as those choruses that are incongruously witness to love-song duet scenes in the Ziegfeld Follies.

It might also be argued that the choral presence at moments like this sufficiently defuses the actual horror one would feel at what might be the real event, so that the symbolic value of the moment, the overriding, more general truth of the event, may dominate. As, for instance, in the scene from the Medea, where a woman confronts two facts: that her husband will abandon her emotionally and physically quite easily in order to pursue the course to what he believes is greater well-being and security; and that the only way she can any longer establish herself in his mind as someone worth considering is to destroy those children for whom he (like any man) initially married, thus at the same time destroying the very creatures that alone give to women their maternal identity. The moral and religious dilemma here, the domestic tragedy, becomes clearer when the sheer terror of the event is made to subside – to retreat into the background, as it were. Through the medium of the neutral choral presence, that is, the whole problem becomes socialised, or socially assimilated, in thus being made more manifest.

We can say, then, that the ever-present chorus is never an embarrassment, but actually a necessity. It is significant in this context that the one time in extant Greek tragedy when the chorus leaves a character alone – when Ajax commits suicide in Sophocles' play (440 BC) – it is because the rest of the drama deals with the socialisation and acceptance of Ajax. In order to achieve this end, the poet is required to place him outside the boundaries of society – of the social presence of the chorus – and rational thought, only then and there, to retrieve him.

Elsewhere, the chorus functions like a proscenium arch, effectively holding the spectator back from too great an involvement in the terrible events he must discover or witness. For example, when Agamemnon calls out that he is being murdered in Aeschylus's play of the same name (458 BC), the chorus members shift their feet, so to speak, and deliberate at what seems an absurdly slow pace, ultimately declining to enter the palace to find out what horror is taking place. One is reminded of the York Crucifixion, a medieval mystery play of the Corpus Christi cycle. In it, the realistic facts of the physical agony of the Christ are muted in the slapstick comedy being played out before us by the four simple-minded, loutish Roman soldiers who, because they are clumsy, just can't seem to nail Jesus to the cross. The distancing that the grotesque humour creates – distancing from Christ's agony – functions here, as in ancient Greek tragedy, like a proscenium arch. Aeschylus did not have to make Agamemnon call out, we must remember; but he chose to have us hear Agamemnon's cries only through the arch, or the lens, of the chorus's grossly comic hesitation.

Similarly, often when we are in danger of feeling too much empathy, we are prevented from doing so – we are pulled back from over-identification with the main characters – by the commonplace choral observation that they, the members of the chorus, do not strive for higher things, hence do not suffer the pangs of the protagonists. Choral odes are platitudinous in this way when they are not downright clichéd, and the effect is that they tend to dampen, to make fuzzy, what in the dialogue has been expressed in terms of sharp contrast. In this sense, the influence of Pindar on tragic choruses cannot be overestimated. For his preserved oeuvre reveals a tendency to create ideas as an aesthetic projection of the needs of the occasion: hence ideas as adornment, as embellishment, even as a kind of form. In other words, nothing Pindar says really matters, apart from the occasion that inspired it or the place in the structure of the poem to which it is assigned. One thinks of the opening of the celebrated first Olympian ode, which begins, "Water is best," a vaguely aphoristic statement that has only to do with the nature of contests and winners and, like so many of the aphoristic remarks in the poem, really has no intrinsic, consequential meaning.

Something which has always startled me is that Euripides uses the same tag ending for the Alcestis, the Medea and the Bacchae:

Gods are here in many shapes,
Gods bring to pass the unforeseen,
What we anticipated comes to naught,
For the unexpected God finds a way,
And so it happened in this play.

Although it is true that great artists of every sort recycle their material – one thinks especially of Bach – the words here are so innocuous that, while they are a perfectly reasonable conclusion to the mildly amusing, if possibly unsettling, Alcestis, they seem utterly flat and out of place following the profound devastation of the Bacchae. Flat and out of place, that is, until one concedes that they essentially mean no more than a falling curtain or the words "The End" looming before us on a movie screen. Like so much of Pindar, then, the thin generality of Euripides' tag lines betokens their function as form rather than content.

In place of tag lines, of course, Jews and Christians had a developed theology. The book of Genesis, for instance, contains a number of stories of betrayal that have the potential for tragedy. I think of Adam and Eve's fall, Cain's murder of Abel, Esau's sale of his birthright, Jacob's deception of Isaac on his deathbed, Abraham's attempted sacrifice of Isaac or Joseph's betrayal of his brother – and we could add other examples. As Erich Auerbach demonstrated in Mimesis (1946), the reader of the sparse, unyielding Biblical narrative must fill up the interstices in the action from the treasure-house of his own spirit. Just remember what Kierkegaard was able to make of the Abraham-and-Isaac confrontation in Fear and Trembling (1843)! Still, the Biblical narrator or narrators take the horror, desolation and depression out of these familial traumas by casting them in the light of love and concern for the people of Israel. Consider Cain, who is described in terms of his progeny and his future in general: his role in Yahweh's plan for the Israelites' own future redeems him.

Greek saga and mythology, for their part, are filled with the same familial horror stories as the Bible, but no theology is on hand to redeem them. It is tragedy that must do so. Aristotle speaks of tragedy as engendering pity and fear in the hearts of spectators and somehow thereby purging them of these presumably excessive emotions. Being the son of a physician, Aristotle easily turned to medical terminology; being a student of Plato, he shared his teacher's penchant for casting ideas in metaphor. But the celebrated purgation of pity and fear may be nothing more than the demystification or even disenfranchisement of nature through the formal denial of issue to any tragic dialogue by means of the intervention of the chorus.

As evidence, consider the structure of Greek tragedy – a structure that, as I have said, changed remarkably little over a period of nearly 100 years. The tragic drama consisted of dialogues interspersed with choral odes. The chorus sang groups of metrically coordinated lyrical lines at the same time as they danced. The choreography of these odes is lost to us, but what we can honestly imagine is something spectacular combining words, music and dance. What we must ask ourselves, however, is how this series of choral interludes worked in tandem with the dialogue sections of the tragic drama – that is, what the relationship was between choral ode and dramatic action.

In the introduction to his Bride of Messina (1803), Friedrich Schiller goes so far as to remark that the choral odes effectively cancel out the realism of Greek tragedy, but there is more to it than this. Like Hesiod in his artful construction of meaning through form in the Works and Days, the Greek tragedians were able to impose a sense of resolution – not mere cancellation or negation – where none obtained through the device of the chorus. In this way the open-ended, hence frightening and mysterious, dynamic of the dialectical collision of the protagonists is itself effectively stopped in its tracks. The inherent failure or fatality of life is thus denied, and we are freed from our own fear of failing as well as dying.

In the Oedipus the King, for example, the ode to the vanity of human aspiration, which precedes the messenger's graphic description of Jocasta's suicide and Oedipus's self-blinding – from which actions we are already safely at one remove, since they occur offstage – dulls the horror of these actions, even as they are occurring, by engulfing us in a calming sea of lyrical generalities. What is more, choral odes in general deny the validity of horrible events, in art as in life. That is, erga (deeds) are always related to logoi (words), the intellectual process whereby man intervenes in the mythological or the historical process. It is man's ability to transform the realism-cum-reality, to make it over to serve his own ends, which is significant in this regard. Moreover, the very beauty of choral lyrics, their art, their complexity, is an act of redemption in the face of the tragic dilemma posed by the action. Such self-conscious artistry itself blunts the tension of the drama in the dialogue, even as the group, through the chorus, takes over the private anguish of Oedipus and recasts it into one platitude or another. In so doing, the group robs the individual human experience of its particularity, its specialness or uniqueness, swallowing it up into a socialised response on the part of the community.

Related to this, it has been said that when tragedy flourished, in the fifth century BC, the breakdown of the oikos family structure caused every person to realise that he dies alone. It has also often been remarked that prominent in the definition of the tragic hero is his isolation. The chorus of Greek tragedy, by contrast, is a constant visual reminder that all human experience is social, that private values and actions are eventually cancelled out or at least transformed in society's assimilation of them. Hence individual mortality is rendered immaterial or insignificant, and the tragic hero himself is cast aside like a well-chewed bone after the chorus has digested the meat of his actions. So it follows that we must be careful in our interpretation of choral passages, trying not to see them as direct commentary upon the action or even as deriving immediately from the action.

For example, again from the Oedipus, the choral ode that speaks of arrogance and ends by noticing the dangers of impiety in rejecting oracular pronouncements – at first blush, this ode may seem to be the poet's commentary on Jocasta's sophistic rationalising of the oracles, which came before it. Indeed, this is frequently held to be the case. But instead we may say that the choral ode works contrapuntally in this instance. Let me explain. Jocasta wants to comfort Oedipus here; furthermore, as can be seen more clearly later on, she passionately resists the revelation to come. Her speech about oracles displays her way of prevaricating or equivocating in order to confront an ominously materialising reality.

Sophocles is showing us here, among other things, how three human beings react before the imperative to discover a hideous truth: Jocasta rationalises the evidence away, Creon ignores it supinely; wile Oedipus is the only one prepared to go forth to meet it head-on. Jocasta seizes upon the presumably verifiable piece of data that there were three highway robbers and argues logically from this that the oracle must be altogether false. But she is not really taking on oracles – the Oracle of Apollo at Delphi or any other. And when her analysis only makes Oedipus the more certain that he may have been Laius's murderer, Jocasta desperately insists upon the plurality of the highwaymen, going so far in her insistence as to deny oracles in their entirety. But, again, this is rhetoric, for the woman feels cornered and understandably so.

She and Oedipus have edged far out on thin ice, where the cracks are sounding louder and louder; yet, courageously, they both hold on. "Get me that peasant witness," demands Oedipus. And Jocasta rushes out as she says, "I'll do anything to make you happy." This has been a terrifying and immensely sad dialogue; we witness two people moving on a now inexorable path or train toward self-revelation, in the process gaining inklings of what hideous stain besmirches their lives together. Victims both, they nonetheless move with dignity, trying all the exits but finding every door, one by one, locked. After all this, the chorus sings of arrogance and impiety, crime and punishment. The contrast with the preceding dialogue is almost shocking. The chorus is in fact quite off the subject, which may be in character, of course; like le people in general, they simply don't get the point. But by talking in easy moral terms, tidying up the ominous irresolutions with their high-blown words (especially on the theme of crime and punishment), the chorus comforts the spectator, deflects him from the passion of Oedipus and Jocasta, protects him from the horror of the spectacle of arbitrary and irrational evil slowly closing in on the royal couple.

Similarly in the Medea, when Medea's plan to kill her children is so far advanced that their fate is sealed, the chorus sings a remarkable ode quite off the subject of this woman's murder of her progeny or the sacrifice of her motherhood. They sing instead about the problems of parenthood and the work of rearing children, saying that in fact it is better never to have any. The chorus translated Medea's tragic agony, in other words, into commonplace domestic terms that dull the sharp and terrifying pain of her dilemma.

Notice by contrast the sarcastic, ironic tone of the brother, Tom Wingfield, in Tennessee Williams's The Glass Menagerie (1944), who functions somewhat like a choral figure in this domestic play. However, Tom is also very much an actor in the drama – involved, almost too much, in the lives of his sister and mother. Moreover, his ironic tone sometimes provokes the audience to laughter, to which they find themselves reacting with embarrassment. And the ambivalence, even confusion, of their response comes from Tom's undertone of bitterness – the actual token of his commitment – which darkens or diminishes his attempts at distancing himself, and us, from the action. This has the effect of forcing the audience to commit themselves utterly to the terms of the drama, even if Tom the individual drops his own commitment at the end by abandoning the stage and his family in a way that the socially-minded Greek chorus is never permitted to do to its fellow citizens, the heroes and heroines of ancient tragedy.

To return to that drama: a particularly interesting phenomenon that we may follow throughout the fifth century BC is the tendency on the part of playwrights to violate or transcend the form or convention of tragedy. Earlier in this essay, I mentioned some examples from Euripides' Electra and his Orestes. The latter play especially stands out as Euripides' protest, finally, against the growing meaninglessness of Greek mythology and thus of tragic drama itself. The way out for him lay with Iphigenia in Tauris (414-412 BC), Helen (412 BC) and Ion (421-408 BC), works of a distinctly comic and even life-affirming tone that pointed toward the realism of the later comic theatre, the beginnings of romance and the eventual birth of the novel.

As for Sophocles, we have so little left of his work by which to judge. But in what we do have, there is one notable instance of dramatic tension that destroys the proscenium-arch effect of which I have been speaking. I mean the moment in Philoctetes (409 BC) when the chorus, in secret alliance with Odysseus and Neoptolemus, lies to the central figure, Philoctetes himself. Nowhere else in extant Greek tragedy, except for the Bacchae, is there so harsh and direct an entry of the chorus into the dramatic action; we in our complicity are brought in as well – and this in Sophocles' last or next-to-last play, mind you. One thinks also of Euripides' Trojan Women (415 BC), which develops into one long threnody in which the unceasing woe takes on a momentum that overwhelms or subverts the whole of the dramatic structure that precedes it – chorus and dialogues both. There are many other examples, to be sure, but what seems important in this context is that we might consider two processes simultaneously at work and at war with each other: the logic of drama or conflict, and the logic of religion or resolution.

"Play something like the murder of my father," says Hamlet, hoping to witness Claudius's reaction to the special truth of usurpation: of the previous king's untimely demise and the present ruler's own hand in it. But there are no special truths in ancient Greek tragedy, there is no topicality. The Oresteia (458 BC) is not an ode to civic justice; the Oedipus is not a put-down of liberal thinkers on the subject of prophecy, the sophists in the face of a religious crisis; Euripides' Suppliants (421 BC) is not a poetic tract on the merits of an allegiance with Argus. No, the tragedians – pace Werner Jaeger – were not really the teachers of ancient Athens. A.J.A. Waldock, in the witty introductory chapter to his book on Sophocles (1951), himself asks, "What do we learn from Sophocles?" Not to kill our father, nor marry our mother? That excessive hate breeds morbidity? But, as Waldock asks, didn't we know all this before?

And of course we did. The Greeks took to their tragic performances, then, not to be educated and informed, but to be reminded and reassured. For, as Alvin Gouldner has shown, the Greeks led profoundly insecure lives, living in a society based on a win-or-loss contest system, and seeing around them constantly the slaves whose fate could on any day become their own. Gilbert Murray, for his part, has described how the anthropomorphisation of natural forces unleashed a wave of paranoia in ancient Greece, as its citizens realised they were contemplating a universe of wilful, uncaring, arbitrary, capricious deities whose interest in mankind was exploitative, retaliatory or sportive. The Greeks went to the theatre to confront this fact. And they saw the enactment of stories that reinforced the central or universal truths which I much earlier identified: life doesn't work in the end, and death cancels all meaning.

Again and again, the tragedies reiterate these ideas in a variety of ways. In the Oresteia, the conflict between man and woman is inherent and never to be resolved; in the Philoctetes, we see man as society's inevitable and necessary victim; and in the Ajax, man has descended to being a mere historical fault. The Oedipus at Colonus (401 BC) shows us the impotence and fury of ageing, while in Euripides' Hippolytus (428 BC) the impossible demands we place on ourselves and others destroy everything; and one could go on. Hamlet may want to catch the conscience of the king through drama, but ancient Greek drama wants to lay the conscience to rest. Throughout the tragedies, the form saves the spectator, pulling him back, making him safe, throwing up horrid truths for his inspection but placing them at a distance. When this does not happen, we are all the more aware of form's potential role as a distancing device. I think in particular of the Bacchae, where choral ode and dialogue work together to provide a relentless, seamless piece of frightening theatre.

Christianity, I believe, allows for more engulfing, more immediate, theatre than the kind we find in ancient Greece. One has only to compare Shakespeare with the Greek tragic dramatists to see the enormous difference. And this, I think, is due to the fact that redemption and reassurance came from the Church, not the theatre; that whatever happened on the stage could be absorbed, its shock could be cushioned, by the security provided through the teachings of the Christian fathers. Consider what Horatio says at the end of Hamlet: "So shall you hear / Of carnal, bloody, and unnatural acts, / Of accidental judgements, casual slaughters, / Of deaths put on by cunning and forced cause." We hear this and know that God's plan for human salvation lies elsewhere than in the theatre or on the stage. There's the comfort. Just so, the horrors of ancient Greece – the blood and carnage, the suffering and slaughter, the inhumanity and injustice – were themselves mitigated only by being cast into the concreteness, hence the comfort, of dramatic form.

_____