SOURCE: "Mythistoria and the Drama," in Thucydides Mythistoricus, 1907. Reprint by Routledge and Kegan Paul, 1965, pp. 129-52.

[Cornford was an English classicist whose books include From Religion to Philosophy (1912), Greek Religious Thought (1923), and Before and after Socrates (1932). In the following excerpt from a work originally published in 1907, Cornford argues that— despite the historian's intentions to "exclude the mythical"—Thucydides "unconciously" fitted his History to the structure of Greek drama.]

MYTHISTORIA AND THE DRAMA

The epithet 'dramatic' has often been applied to Thucydides' work; but usually nothing more is meant than that he allows his persons to speak for themselves, and presents their character with vividness. The dramatization which we have pointed out in the treatment of Cleon is a very different thing; it is a principle of construction which, wherever it operates, determines the selection of incidents to be recorded, and the proportions and perspective assigned them. In this chapter we shall attempt to describe and analyse the type of drama that we have to do with, and to trace the literary influence under which Thucydides worked.

We ought first, perhaps, to meet a possible objection. It may be urged that Thucydides in his preface expressly excludes anything of the nature of poetical construction from his literal record of what was said and what was done. He criticizes the methods of poets and story-writers, and warns us that, at the cost of making his story 'somewhat unattractive', he intends to exclude 'the mythical'.…

He cannot, therefore, it might be inferred, have done what we have thought we found him doing. But we would ask for a careful examination of the passage in question. What was in Thucydides' thoughts when he wrote it, and above all, what precisely did he mean to exclude when he banished 'the mythical'?

The words occur towards the end of the introduction, which is designed to establish Thucydides' belief that the Peloponnesian war was the most memorable of all that had ever been in Greece. The possible rivals, he points out, are the Trojan war and the Persian invasion. For the first of these events the only literary evidence we have is that of the epic poets, and chiefly of Homer, whose record cannot be checked by direct observation, while much of his theme through the lapse of time has passed, or 'won over', into the region of the mythical and incredible. The only tests we have are certain indications in the existing condition of Greece which seem inconsistent with the past state of things as represented by the literary authorities. With these indications we must be content; and they suffice to show that the epic poets embellished their tale by exaggeration. The story-writers, again, on whom we depend for the history of the Persian wars, were not bent upon accurate statement of truth;—witness the carelessness of Herodotus about points of detail. Their object was rather to make their recitations attractive and amusing to their audience; and if we discount their evidence accordingly, we shall find, going by ascertained facts alone, that the Peloponnesian war was the greatest ever seen.

Thucydides next passes abruptly to the formulation of his own method; he intends to record what was said and what was done as accurately and literally as possible. The result, he then remarks, will probably be somewhat unattractive to an audience at a recitation, because the facts recorded will have nothing 'mythical' about them; he will be content, however, if they are judged useful by people who wish to know the plain truth of what happened.

The phrase 'winning over into the mythical' is illuminating. It suggests the transformation which begins to steal over all events from the moment of their occurrence, unless they are arrested and pinned down in writing by an alert and trained observer. Even then some selection cannot be avoided—a selection, moreover, determined by irrelevant psychological factors, by the accidents of interest and attention. Moment by moment the whole fabric of events dissolves in ruins and melts into the past; and all that survives of the thing done passes into the custody of a shifting, capricious, imperfect, human memory. Nor is the mutilated fragment allowed to rest there, as on a shelf in a museum; imagination seizes on it and builds it with other fragments into some ideal construction, which may have a plan and outline laid out long before this fresh bit of material came to the craftsman's hand to be worked into it, as the drums of fallen columns are built into the rampart of an Acropolis. Add to this the cumulative effects of oral tradition. One ideal edifice falls into ruin; pieces of it, conglomerates of those illassorted and haphazard fragments, are carried to another site and worked into a structure of, perhaps, a quite different model. Thus fact shifts into legend, and legend into myth. The facts work loose; they are detached from their roots in time and space and shaped into a story. The story is moulded and remoulded by imagination, by passion and prejudice, by religious preconception or aesthetic instinct, by the delight in the marvellous, by the itch for a moral, by the love of a good story; and the thing becomes a legend. A few irreducible facts will remain; no more, perhaps, than the names of persons and places—Arthur, Caerleon, Camelot; but even these may at last drop out or be turned by a poet into symbols. 'By Arthur,' said Tennyson, 'I always meant the soul, and by the Round Table the passions and capacities of man.' The history has now all but won over into the mythical. Change the names, and every trace of literal fact will have vanished; the story will have escaped from time into eternity.

When we study this process, we seem to make out two phases of it, which, for the criticism of Thucydides, it is necessary to distinguish. The more important and pervasive of the two is the moulding of fact into types of myth contributed by traditional habits of thought. This process of infiguration if we may coin the word may be carried to any degree. Sometimes the facts happen to fit the mould, and require hardly any modification; mere unconscious selection is enough. In other cases they have to be stretched a little here, and patted down there, and given a twist before they will fit. In extreme instances, where a piece is missing, it is supplied by mythological inference from the interrupted portions which call for completion; and here we reach the other phase of the process, namely invention. This is no longer a matter of imparting a form to raw material; it is the creation of fresh material when the supply of fact is not sufficient to fill the mould. It leads further to the embroidery of fabulous anecdote, which not only has no basis in fact, but is a superfluous addition, related to fact as illustrations in a book are related to the text.

The process, in both its phases, can be illustrated from the version preserved by Thucydides of the legend of Harmodius and Aristogeiton, the tyrant-slayers. Harmodius' sister, whom the tyrant insults, makes her first appearance in this account. She is superfluous, since the murderers had already a sufficient private motive arising out of the love-quarrel. That is not in itself an argument against her historical character, for superfluous people sometimes do exist; but other circumstances make it not improbable that she owes her existence to the mythical type which normally appears in legend when tyrants have to be slain. The two brothers, or lovers, and the injured sister, or wife—the relationships vary—are the standing dramatis personae on such occasions. Collatinus, Brutus, and Lucretia are another example from legend; while the purely mythical type which shapes such legends is seen in the Dioscuri and Helen. The suggestion is that Harmodius and Aristogeiton were identified with the Heavenly Twins. If there is any truth in the story of how Peisistratus was conducted back to Athens by a woman dressed as Athena and accepted by the citizens as the goddess in person, it is not surprising that the next generation of Athenians should have recognized the Dioscuri in Harmodius and his friend. Given that identification, the injured sister is felt to be a desirable, if not indispensable, accessory; she is filled in by inference, and she becomes a candidate for the place of 'basket-bearer' in the Panathenaic procession, at which the murder took place. Thus, the legend of Harmodius illustrates both the phases of the process we described: first, it is moulded on the mythical type of the Heavenly Twins, and then invention supplies the missing third figure.

Mythical types of this sort can be discovered and classified only after a wide survey of comparative Mythistoria; for we all take our own habits of thought for granted, and we cannot perceive their bias except by contrast. The Greek who knew only Greek legend could not possibly disengage the substance from the form; all he could do was to prune away the fabulous and supernatural overgrowths, and cut down poetry into prose. It is thus that Thucydides treats myths like the story of Tereus, Procne, and Philomela; he rationalizes them, thinking that he has reduced them to history when he has removed unattested and improbable accretions, such as the transformation of Tereus into a hoopoe. But history cannot be made by this process (which is still in use); all that we get is, not the original facts, but a mutilated legend; and this may very well be so mutilated that it is no longer possible to distinguish the informing element of fiction, which was discernible till we effaced the clues.

The phenomenon that especially concerns us now is some thing much wider than the mythical infiguration of a single incident here or there, such as the legend of the Tyrantslayers. It is the moulding of a long series of events into a plan determined by an art form. When we set the Persians of Aeschylus beside the history of Herodotus, we see at once that the tragedian in dramatizing the events of Xerxes' invasion, some of which he had personally witnessed, has also worked them into a theological scheme, preconceived and contributed by his own mind. Further we remark that Herodot-us, although he is operating in a different medium and writing a saga about the glory of Athens, uses the same theological train of thought as a groundwork, and falls in with the dramatic conception of Aeschylus. This is a case of the infiguration of a whole train of events by a form which is mythical, in so far as it involves a theological theory of sinful pride punished by jealous divinity, and is also an art form, by which the action is shaped on dramatic principles of construction, involving such features as climax, reversal, catastrophe. The theory and the form together provide the setting of the whole story—the element which makes it a work of art. This element is so structural that it cannot be removed without the whole fabric falling to pieces, and at the same time so latent and pervasive, as not to be perceptible until the entire work is reviewed in its large outline. Even then it can be detected only by a critic who is on his guard and has not the same scheme inwrought into the substance of his own mind; for if he is himself disposed to see the events in conformity with the scheme, then the story will answer his expectation and look to him perfectly natural.

When Thucydides speaks of 'the mythical', it seems probable from the context that he is thinking chiefly of inventive 'embellishment'. The accretions of fabulous anecdote are comparatively easy to detect; they often bring in the supernatural in the forms of vulgar superstition, and being for this reason improbable, they require better evidence than is forthcoming. Also, poets tend to magnify their theme for purposes of panegyric, flattering to their audience; they will, for instance, represent Agamemnon's expedition as much larger than it probably was. It is on these grounds that Thucydides objects to the evidence of Ionian Epos and Herodotean story-telling. He warns us against the faults which struck his notice; and he was on his guard against them, even more than against the popular superstition and dogmatic philosophy of the day, which he tacitly repudiates. But there was one thing against which he does not warn us, precisely because it was the framework of his own thought, not one among the objects of reflection,—a scheme contributed, like the Kantian categories of space and time, by the mind itself to whatever was presented from outside. Thucydides, like Descartes, thought he had stripped himself bare of every preconception; but, as happened also with Descartes, his work shows that there was after all a residuum wrought into the substance of his mind and ineradicable because unperceived. This residuum was his philosophy of human nature, as it is set forth in the speech of Diodotus,—a theory of the passions and of their working which carried with it a principle of dramatic construction presently to be described. That he was not forearmed against this, he himself shows when, in attacking Herodotus, he accuses him of trivial errors of fact, and does not bring the one sweeping and valid indictment which is perfectly relevant to his own point about the embellishment of the Persian War. The dramatic construction of Herodotus' work, which stares a modern reader in the face, apparently escaped the observation of his severest ancient critic.

Another proof can be drawn from Thucydides' own account of a series of events which he evidently believed to be historical, the closing incidents, namely, of Pausanias' career. He shows us the Spartan king intriguing with the Persian, and 'bent upon the empire of Hellas'. Pausanias commits certain treacherous acts; boasts of his power to the Great King; 'intends, if the king please, to marry his daughter'; is so 'uplifted' by the king's answer that he can no longer live like ordinary men; behaves like an oriental; cannot keep silence about his larger designs; makes himself difficult of access, and displays a harsh temper. We know all these symptoms well enough, and we foresee the end. Pausanias is recalled, but the evidence against him is insufficient. He writes a letter betraying his designs and ending with an order for the execution of the bearer. The messenger, whose suspicions are aroused, opens the letter and shows it to the authorities at Sparta. The ephors arrange that they shall be concealed behind a partition and overhear a conversation between the king and his treacherous messenger, who contrives to draw from Pausanias a full and damning avowal. The end follows in the Brazen House.

This is not the sort of thing that Thucydides objects to as 'mythical'; it is not 'fabulous', not the embroidery of mere poetical invention; and so he reports it all in perfect good faith. What does not strike him, and what does strike us, is that the story is a drama, framed on familiar lines, and ready to be transferred to the stage without the alteration of a detail. The earlier part is a complete presentation of the 'insolent' type of character. The climax is reached by a perfect example of 'Recoil' …, where the hero gives the fatal letter to the messenger, and thus by his own action precipitates the catastrophe. The last scene is staged by means of a theatrical property now so cheapened by use as to be barely respectable—a screen! The manner of the hero's death involved sacrilege, and was believed to bring a curse upon his executioners. Could we have better proof that Thucydides was not on his guard against dramatic construction, and was predisposed to see in the working of events a train of 'causes' which tragedy had made familiar?

When we are alive to the dramatic setting, we can infer with some certainty the stages through which the Thucydidean story of Pausanias has passed. The original stratum of fact must have been that Pausanias somehow misconducted himself, was recalled, and put to death in circumstances which were capable of being used by superstition and policy against the ephors. These facts worked loose into a legend, shaped by imagination on the model of preconceived morality and views of human nature. The mould is supplied by drama; and meanwhile fabulous invention is busy in many minds, embroidering the tale with illustrative anecdotes. Thucydides brushes away these extravagant and unattested accretions, and reduces the legend again to what seemed to him a natural series of events. It is only we who can perceive that what he has left is the dramatized legend, not the historical facts out of which it was worked up. It is not wildly paradoxical to think that the historian who accepted the legend of Pausanias might frame on the same pattern the legend of Cleon. Not that Thucydides invented anything; all that was needed was to select, half unconsciously, those parts of his life which of themselves composed the pattern.

We must now come to closer quarters with the epithet 'dramatic'. It is worth noting, at the outset, that in the mere matter of external form, the history seems to show the influence of tragedy,—a fact which need not surprise us, if we remember that Thucydides had no model for historical writing. The brief abstract of the annalist was a scaffold, not a building; and Thucydides was an architect, not a carpenter. Chroniclers and story-writers like Herodotus had chosen the lax form of epic, congenial to ramblers; but whatever the history was to be, it was not to be like Herodotus, and it was to draw no inspiration from the tradition of Ionian Epos. So Thucydides turned to drama—the only other developed form of literature then existing which could furnish a hint for the new type to be created. The severe outline and scrupulous limitations of this form satisfied his instinct for self-suppression. The epic poet stands before his audience and tells his own tale; but the dramatist never appears at all: the 'thing done' … works itself out before the spectators' eyes; the thing said comes straight from the lips of the actors.

Best of all, to Thucydides' thinking, if we, of after times, could ourselves have watched every battle as it was won and lost, and ourselves have heard every speech of envoy and statesman; we should then have known all, and much more than all, this history was designed to tell. But as this cannot be, we are to have the next thing to it; we shall sit as in a theatre, where the historian will erect his mimic stage and hold the mirror up to Nature. Himself will play the part of 'messenger' and narrate 'what was actually done' with just so much of vividness as the extent of his own information warrants. For the rest, the actors shall tell their own tale, as near as may be, in the very words they used, 'as I heard them myself, or as others reported them.'

Speeches are much more prominent in Thucydides' history than they are in that of Herodotus. The change seems partly due to the later historian's preference for setting forth motives in the form of 'pretexts', instead of giving his own opinion; but it is also due to his being an Athenian. Plato similarly chose to cast his speculations in the dramatic form of dialogue, allowing various points of view to be expressed by typical representatives, without committing himself to any of them. Even oratory at Athens was dramatically conceived; the speech-writer did not appear as advocate in court; he wrote speeches in character to be delivered by his clients. It has often been remarked that the debates in Thucydides resemble in some points of technique the debates in a Euripidean play. There is moreover in one respect an intellectual kinship between Thucydides and the dramatist who was contemporaneously moulding the form of tragedy to the strange uses of realism, and working away from Aeschylus as Thucydides had to work away from Herodotus. The two men are of very different temperaments; but in both we seem to find the same sombre spirit of renunciation, the same conscious resolve nowhere to overstep the actual, but to present the naked thoughts and actions of humanity, just as they saw them. No matter how crude the light, how harsh the outline, so that the thing done and the thing said shall stand out as they were, in isolated sharpness, though

Mist is under and mist above,…
And we drift on legends for ever.

These considerations, however, touch only the question of external form: they show why so much that we should state directly is stated indirectly by Thucydides, in speeches. The choice of this form is consistent with a complete absence of plot or of dramatic construction: otherwise Thucydides could not have chosen it at starting; for at that moment the plot lay in the unknown future. We mention the point only because evidently it was somewhat easier for an historian who consciously borrowed the outward form of tragedy, to take unconsciously the further step, and fall in with its inward form and principle of design. It is this which we now wish to define more closely. The type of drama we have detected in the history is not the Euripidean type; it will be found, on examination, to show an analogy with the older form existing in the tragedies of Aeschylus.

The resemblances are reducible to two main points. The first is an analogy of technical construction, seen in the use and correlation of different parts of the work. The second is a community of psychological conceptions: a mode of presenting character, and also a theory of the passions which has a place not only in psychology, but in ethics. We shall begin by studying the structure; but we may bear in mind that this structure is closely involved with the psychological theory.

An art form, such as the Aeschylean drama, shapes itself as a sort of crust over certain beliefs which harden into that outline. When this has happened, the beliefs themselves—the content of the mould—may gradually be modified and transmuted in many ways. Finally, they may melt and almost fade away, leaving the type, which is preserved as a traditional form of art. This survival of an element of technical construc-tion may be illustrated by the instance of 'reversal'.… A 'reversal of fortune' is the cardinal point of primitive tragedy; and it originally means an overthrow caused by an external supernatural agency—Fate or an angry god. When the belief in such agencies fades, 'reversal' remains as a feature in drama; but the change of situation is now caused by the hero's own act. The notion of 'recoil' comes in: that is to say, the fatal action itself produces results just the opposite of those intended—a perfectly natural occurrence. In this way a piece of technique outlasts the belief which gave rise to it.

The Aeschylean drama appears to us to have gone through a process of this kind. The structure, as we find it, seems to imply an original content of beliefs in some respects more primitive than those explicitly held by Aeschylus himself, but surviving in his mind with sufficient strength to influence his work. Similarly, as we hope to show, in transmission from Aeschylus to Thucydides, the dramatic type has again outlasted much of the belief which informed it in the Aeschylean stage. It is the artistic structure which is permanent; the content changes with the advance of thought. Hence, if we point to Aeschylean technique in Thucydides, we are not necessarily attributing to him the creed of Aeschylus.

We must first attempt to describe the structure of Aeschylean tragedy. In order to understand it we must try to imagine a yet more primitive stage in the development of the drama than any represented in extant Greek literature, a stage which the earliest of Aeschylus' plays has already left some way behind. A glance at the development of modern drama may help us.

Certain features which survived in Greek tragedy suggest that we should look back to a type somewhat resembling the mediaeval mystery and some of the earliest modern dramas, such as Everyman, which are like the mystery in being religious performances and in the element of allegorical abstraction. Their effect, due in part to each of these features, may be described as symbolic. Everyman is a sermon made visible. To watch it is like watching the pastime called 'living chess', in which the pieces are men and women, but the man who is dressed like a bishop is nothing more than a chessman who happens to be automatic. He has not the episcopal character; his dress is a disguise with nothing behind it; his words, if he spoke, would be the speech of a parrot. And so it is with Everyman. The persons are not persons at all, but personae, masks, symbols, the vehicles of abstract ideas. They do not exist, and could not be conceived as existing, in real space and time. They have no human characters, no inward motives, no life of their own. Everyman, as his name is meant to show, is in fact not a man, but Man, the universal.

The main development of modern drama shows, in one of its aspects, the process by which this symbolic method gives way to the realistic. The process consists in the gradual filling in of the human being behind the mask, till the humanity is sufficiently concrete and vital to burst the shell and step forth in solid flesh and blood. The symbol comes to contain a type of character; the type is particularized into a unique individual. The creature now has an independent status and behaviour of its own. Every gesture and every word must be such as would be used by an ordinary human being with the given character in the given situation. Once created, the personality is an original centre; it cannot be made to do what we please or to utter our thoughts. In some such terms as these a modern novelist or playwright will speak of his characters; and it is thus that they appear to us.

Now we can observe a certain intermediate stage in which these two methods, the symbolic and the realistic, are balanced in antagonism, so as to produce a curious effect of tension and incoherency. A good instance is Marlowe's Faustus. Faustus himself occupies the central plane; he is a living man, but still imprisoned in a symbolical type. The intrusion of humanity has gone far enough to disturb the abstract effect, and it reacts on some of the persons in the play who ought to be purely symbolic. Lucifer, it is true, is kept apart and remains non-human; but Mephistophilis oscillates in our imagination between the ideal and reality, with a distressing result. Again, on a lower level than Faustus there is yet another grade of persons, in contrast with whom he shows up as heroic and ideal. These are the vintner, the horse-courser, and other pieces of common clay picked out of a London alley; they belong to a different world, and we feel that they could no more communicate with the tragic characters than men can talk with angels. Thus there are in this one play four sets or orders of persons: (1) the purely abstract and symbolic, such as Lucifer, who only appears on an upper stage at certain moments, and takes no part in the action; (2) the intermediate, for instance Mephistophilis, who ought to be symbolic, but treads the lower stage, a cowled enigma, horrible because at moments he ceases to be symbolic without becoming human; (3) the heroic or tragic: Faustus, who is an ideal half realized, hanging together on its own plane; (4) the real: common mortals who would attract no attention in Fleet Street.

The Greek drama, although in the detail of historical development it started at a different point from the modern, and followed another course, seems, nevertheless, to pass through a phase analogous to that which we have just described. The original substance of the drama was the choral lyric; the actors as they afterwards became began as an excrescence. At a certain stage the actors are assimilated to the chorus and move in the same atmosphere. Thus in the earliest play of Aeschylus, the Suppliants, we find that the chorus of Danaids are actually the heroines of the action, which centres round them, so that they are not merely on the same plane with the actors, but themselves a complex actor, and the effect is simple, coherent, and uniform. In the Prometheus, again, the chorus belong to the same ideal world as the Titan hero, a world in which abstract symbols like Mastery and Violence can move without showing as unreal against the other persons. The whole drama is on the symbolic plane, the life in it being due to anthropomorphic imagination, not to the intrusion of realism.

But in the latest plays of Aeschylus, the beginning of a change is clearly marked: the actors are becoming human, while the lyric is rising above them, or else remains suspended in a rarer atmosphere from which they are sinking. This is a natural stage in the passage from pure symbolism to realism. The advance shows itself externally in the drifting apart of the lyrical element from the dialogue,—a separation which, of course, widens in the later tragedians, till the choral ode, though still an indispensable and very beautiful feature, becomes in point of construction little more than an interlude, which relieves the concentrated intensity of the action. This change is commonly taken as a phenomenon which needs no explanation; but really it is caused inevitably by the coming to life of the persons in the drama. In proportion as these become more real, the lyric becomes more ideal and further removed from the action.

In the stage observable in Aeschylus' latest plays, the choral part is still dramatic, and of equal importance with the dialogue. The two elements are evenly balanced; but at the same time they have begun to occupy different worlds, so that we are sensible of the transition from one to the other. The result is a curious duplication of the drama which now has two aspects, the one universal and timeless, the other particular and temporal.

The nature of this phenomenon will, we hope, become clear, if we take as an illustration the Agamemnon. In this play, the visible presentation shows how the conqueror of Troy came home and was murdered by the queen. The events that go forward on the stage are particular events, located at a point of legendary time and of real space. The characters are certain individuals, legendary or historic—there is to Aeschylus no difference here—who lived at that moment and trod that sport of earth. But in the choral odes the action is lifted out of time and place on to the plane of the universal. When the stage is clear and the visible presentation is for the time suspended, then, above and beyond the transient spectacle of a few suffering mortals caught, just there and then, in the net of crime, loom up in majestic distance and awful outline the truths established, more unchangeably than the mountains, in the eternal counsels of Zeus. The pulse of momentary passion dies down; the clash and conflict of human wills, which just now had held us in breathless concentration, sink and dwindle to the scale of a puppet-show; while the enduring song of Destiny unrolls the theme of blood-haunted Insolence lured by insistent Temptation into the toils of Doom. As though on a higher stage, uncurtained in the choral part, another company of actors concurrently plays out a more majestic and symbolic drama. On this invisible scene walk the figures of Hybris and Peitho, of Nemesis and Ate—not the bloodless abstractions of later allegory, but still clothed in the glowing lineaments of supernatural reality. The curtain lifts for a timeless moment on the spectacle of human life in an aspect known to the all-seeing eyes of Zeus; and when it drops again, we turn back to the mortal tragedy of Agamemnon and Clytemnestra, enlightened, purified, uplifted, calm.

Thus we find in Aeschylus something analogous to the hierarchy of persons we noted in Faustus; although, for various reasons, there is not the same crude effect of incoherency and tension. The supernatural characters—Zeus, supreme above all, and the demonic figures of Hybris, Nemesis, Ate, and the rest, are not seen, as Lucifer is seen on the upper stage of the Elizabethan theatre, but remain in the spiritual world to which lyrical emotion exalts the inward eye—the world where metaphor (as we call it) is the very stuff of reality, where Cassandra quickens and breathes, and whence she strays among mortal men like a fallen spirit, sweet-voiced, mad, and broken-winged. Hence the effect is far more awful and solemn than the actual apparition of Lucifer; and when Apollo and Athene and the spirits of vengeance take human shape in the Eumenides, a spell is broken, a veil rent, an impression shattered, for which not the most splendid symphony of poetical language can atone.

Here, however, we would confine our attention to the Agamemnon. At the lower end of the scale we find a further advance of realism in some minor characters, the watchman and the herald; the nurse in the Choephori is of the same order. These are allowed some wonderful touches of common humanity, below the heroic level; for they are not directly concerned in the central action, and a little irrelevant naturalism does no harm, if it is not carried far. But they are only just below the heroic standard, and are certainly not the sort of people you would have met in a walk to the Piraeus.

Thus, the two planes in the Agamemnon are divided by an interval less wide and less abrupt than the divisions in Faustus. In psychological conception also the union is very close, since the heroic characters are still so abstract and symbolic that they are barely distinguishable from the pure abstractions of the lyrical world. Agamemnon, for instance, is simply Hybris typified in a legendary person. He is a hero flown with 'insolence' (the pride and elation of victory), and that is all that can be said of him. He is not, like a character in Ibsen, a complete human being with a complex personality,—a centre from which relations radiate to innumerable points of contact in a universe of indifferent fact. He has not a continuous history: nothing has ever happened to him except the conquest of Troy and the sacrifice of Iphigenia; nothing ever could happen to him except Pride's fall and the stroke of the axe. As we see him, he is not a man, but a single state of mind, which has never been preceded by other states of mind (except one, at the sacrifice in Aulis), but is isolated, without context, margin, or atmosphere. Every word he says, in so far as he speaks for himself and not for the poet, comes straight out of that state of mind and expresses some phase of it. He has a definite relation to Cassandra, a definite relation to Clytemnestra; but no relation to anything else. If he can be said to have a character at all, it consists solely of certain defects which make him liable to Insolence; if he has any circumstances, they are only those which prompt him to his besetting passion.

Now it is in some such way as this that Thucydides presents his principal characters. Cleon is a good instance. He is allowed no individuality, no past history, no atmosphere, no irrelevant relations. He enters the story abruptly from nowhere. A single phrase fixes his type, as though on a play-bill: 'Cleon, the most violent of the citizens and first in the people's confidence'; that is all we know of him. There follows a speech in which the type reveals itself in a state of mind,—Violence in its several phases. Then he vanishes, to reappear, before Sphacteria, as Violence with one of its aspects ('covetousness') emphasized, and a sudden passion of ambitious self-confidence … added thereto. Finally, we see him wrecked by this passion at Amphipolis. Pericles is introduced in the same way, with a single epithet: 'Pericles, the son of Xanthippos, a man at that time first among the Athenians, and most powerful … in action and in speech.' His characteristic quality is wise foresight ( … —the opening word of his first speech); and he stands also, in the Funeral Oration, for the glory … of Athens. Alcibiades we shall study later. In every case the principal characters are nearly as far removed from realism, nearly as abstract and impersonal as the heroic characters in Aeschylus. Thucydides, in fact, learnt his psychology from the drama, just as we moderns (whether historians or not) learn ours, not by direct observation, but from the drama and the novel.

But we can carry the analogy further; it extends to minor points of Aeschylean technical construction, which follow naturally upon the drifting apart of lyric and dialogue. In the Agamemnon we note that the separation of the two planes has gone far enough to make it impossible for the members of the chorus to interfere with the action at its crisis. The elders, when they hear the death-cry, cannot enter the palace; not because the door is locked, nor yet because they are feeble old men. Rather they are old men because an impassable barrier of convention is forming between chorus and actors, and their age gives colour to their powerlessness. The need of a separate stage for the actors, though tradition may cling to the old orchestra, is already felt. The poet is half aware of the imaginative separation, and he bridges it by links of two kinds— formal links of technical device, and internal connexions of a psychological sort, which will occupy us in the next chapter.

The formal links are provided by what is called 'tragic irony'. The dialogue is so contrived that, instructed by the lyric, we can catch in it allusions to grander themes than any of which the speakers are conscious, and follow the action with eyes opened to a universal significance, hidden from the agents themselves. Tragic irony, however, is not a deliberately invented artifice; it arises of itself in the advance from the purely symbolic stage of drama. In that earliest stage the whole dialogue might be called 'ironical', in the sense that it is the poet's message to the audience, not the expression of the persons' characters, for they have none. But it becomes ironical in the strict sense only when the persons begin to have elementary characters and minds, and so to be conscious of one meaning of their words, which is not the whole meaning or the most important. The effect is now no longer merely symbolic, but hypnotic; the speaker on the stage is like a somnambulist—alive, but controlled and occupied by an external personality, the playwright.

Tragic irony is used by Aeschylus with great freedom; because his persons are still so near to the symbolic, they have so little character and psychology of their own, that they do not mind serving as mouthpieces. Here and there we find instances of perfect irony, where the speaker's words bear both constructions equally well, and are at once the natural expression of the appropriate state of mind and also a message from the poet to the spectator, applying one of the lyrical themes. This is the only sort of irony admitted by Sophocles, whose characters have become so human that they will not speak merely for another. In Aeschylus, however, there are whole speeches which are hypnotic, and hardly in character at all. The effect is so unfamiliar to readers schooled in realism that it is often missed.

The first two speeches of Clytemnestra, for instance, seem to be of this kind; notably, the beacon speech. If we try to interpret this as a realistic revelation of Clytemnestra's character and thoughts, we shall not find that it helps us to much insight, because its main function has nothing to do with her character. The poet is speaking through her, and the thoughts are his. The early part of the play, down to the entrance of Agamemnon, is an overture, in which Aeschylus musters and marshals the abstract themes which are to be the framework of the trilogy. One of them is expressed in the beacon speech; and it is this. The fire of Idaean Zeus has fallen upon Troy, 'neither before its season nor striking as an idle glancing shaft beyond the stars'; but that same fire, the symbol of Justice, speeds now to 'strike the roof of the Atreidae'. From mountain top it leaps and hastens across the sea to mountain top; and like the torch passed from hand to hand in the race, it is itself a runner and the only one which 'running first and last reaches the goal'. This description of the symbolic fire conducted along the beacon chain is given to Clytemnestra because it can be given to no one else, not because it is the best means of illustrating her psychology. The speech, by the way, also exhibits another artifice employed to link the two planes—the allusive verbal echo between dialogue and lyric. The symbol of the fire, in a slightly varied form, recurs at the beginning of the next chorus, and the keyword … is reiterated to mark the correspondence.

Now the speeches in Thucydides can be roughly classed under four heads. There are, first, a few realistic speeches by minor characters; for instance, the short, sharp utterance of the Spartan ephor, which has the trick of the laconic practical man. Next, there are idealistic speeches, designed as direct expressions of character or of national ideals; the Funeral Oration will serve as an example. These shade off, through a class in which sketches of national character are introduced indirectly, with some strain upon dramatic probability, into a class where irony is openly employed in the tragic manner. Cleon's Mytilenean speech, for instance, is nearly all of the character-revealing sort, but it contains a passage about the evil results of exceptional prosperity which is without any true application to the position of Lesbos or to the history of the revolt. It runs as follows:

Conceiving a reckless confidence in the future, and hopes that outran their strength though they fell short of their desires, they went to war; and they thought fit to prefer might to right, for where they thought they saw a chance of success, they set upon us when we were doing them no wrong. It is always so: when exceptional prosperity comes sudden and unexpected to a city, it turns to insolence: and, in general, good fortune is safer for mankind when it answers to calculation than when it surpasses expectation, and one might almost say that men find it easier to drive away adversity than to preserve prosperity. We were wrong from the first. We ought never to have put the Mytileneans above the rest by exceptional treatment; then their insolence would not have come to this height. It is a general rule that human nature despises flattery, and respects unyielding strength.

These words are patently inapplicable to the revolted island, whose exceptional position was notoriously a survival of the status originally enjoyed by every one of the allies, but now forfeited by all but a few; to speak of it as a sudden access of prosperity is simply meaningless. We are driven to see in the passage a use of tragic irony; Thucydides puts into Cleon's mouth the very moral which his own career is to illustrate. The device is unskilfully employed, since dramatic probability is too completely sacrificed. Sophocles would not have passed these sentences, which on the speaker's lips have not even a plausible meaning; but Aeschylus would have passed them, and after all Thucydides was only an amateur tragedian.

A fourth use of speeches is illustrated by the Spartan envoys' homily before Sphacteria. This is still further removed from realism, and resembles the beacon speech, which is but one degree below the lyric plane. The historian, reluctant to break silence in his own person, sets forth the theme and framework of his drama in the form of a solemn warning. He has already described the Athenians at Pylos as 'wishing to follow up their present good fortune to the furthest point'. This is a dangerous frame of mind, against which Themistocles had warned the Athenians after Salamis, when they wished to press forward and destroy the Persians' bridges over the Hellespont. 'I have often,' says Themistocles, 'myself witnessed occasions, and I have heard of many from others, where men who had been conquered by an enemy, having been driven quite to desperation, have renewed the fight and retrieved their former disasters. We have now had the great good luck … to save both ourselves and all Greece by the repulse of this vast cloud of men; let us then be content and not press them too hard, now that they have begun to fly. Be sure that we have not done this by our own might. It is the work of gods and heroes, who were jealous that one man should be king at once of Europe and Asia.… At present all is well with us—let us then abide in Greece, and look to ourselves and to our families.'

The warning of the Spartan envoys is conceived in the same spirit; but it is unheeded and unanswered. No answer, indeed, was possible; the speech is not an argument, but a prophecy. A reply from Cleon, a statement of the war party's policy, such as modern critics desiderate, would be as inappropriate as a reply from Clytemnestra to the Second Chorus in the Agamemnon. The stage is clear while this prophecy, addressed not to the actors but to the spectators, passes unheard by those who, could they have heard it, might have been saved.

One further point of formal resemblance between Aeschylus and Thucydides is the allusive echoing of significant phrases, which sustain the moral motive dominant in the plot. We have seen an instance of this device in the repetition of the words 'coveting more' … , which reappear at critical moments after the use of them in the envoys' speech; and we shall note other examples later. This completes the analogy with Aeschylean form, so far as concerns external peculiarities.