Poetry Is Truer Than History

I propose to treat of poetry itself and of its kinds; to say what each kind can do; to enquire how a plot must be put together if the poem is to be a good one; to say how many parts a poem has and what sort of parts they are.¹

Following the order of nature, let us begin with the things that come first.

Let Us Begin With What Comes First

Epic and tragedy, comedy, dithyramb, most music of the flute and the lyre — all of these are, in their general conception, modes of imitation. They differ from one another in three respects only: in the medium they imitate with, in the objects they imitate, and in the manner of the imitating.

Some men imitate through color and shape. Some through the voice. In the arts I am concerned with, the imitating is done by rhythm, by language, or by tune — singly or in combination.

That is the medium.

As for the objects: what is being imitated is men in action, and men in action are either better than we are, or worse, or about the same. It is the same in painting — Polygnotus painted men nobler than they are, Pauson less noble, Dionysius exactly true to life. And this is precisely what divides tragedy from comedy. Comedy sets out to show men worse than they are in actual life. Tragedy, better.

And as for the manner: the poet may narrate — either taking on another personality, as Homer does, or speaking in his own — or he may present all his people to us alive and moving.

So Sophocles imitates the same kind of object as Homer, since both take the higher type of character. And from the other angle, Sophocles is doing what Aristophanes does, since both put people in front of us acting.

Let me stop here to correct something, because it goes wrong constantly. People attach the word poet to the name of a metre and speak of elegiac poets and hexameter poets — as though it were the verse and not the imitation that made the man a poet. If a treatise on medicine is brought out in verse, they call the author a poet. But Homer and Empedocles have nothing whatever in common except the metre. One is a poet. The other should be called a natural scientist.

It is the imitation that makes the poet. The verse is incidental.

Why There Is Poetry at All

Poetry seems to have come into existence from two causes, and both of them lie deep in our nature.

The first is that imitation is instinctive in us from childhood. This is one of the differences between a human being and the other animals: we are the most imitative creatures alive, and we get our earliest lessons by imitating.

The second is that everyone takes pleasure in things imitated. And I want to dwell on this, because the evidence for it is stranger than people notice.

Objects we cannot bear to look at in themselves — I mean the forms of the most ignoble animals, and dead bodies — we take delight in contemplating when they have been reproduced with real accuracy.

Why should that be?

The reason is that learning gives the liveliest pleasure, not only to philosophers but to everybody, though most men's capacity for it is limited. When you look at a likeness, what you are enjoying is the act of inference. You are saying to yourself: ah, that is he.

And you can test this. If you happen never to have seen the original, the pleasure you take will not come from the imitation at all. It will come from the execution, or the coloring, or something of that kind.

So imitation is one instinct. Alongside it runs the instinct for tune and for rhythm — metres being obviously divisions of rhythm. Men who began with those natural gifts improved on them little by little, until their rough improvisations gave birth to poetry.

The Plot Is the Soul

Let me now give you the definition of tragedy, gathered from everything I have said.

Tragedy is an imitation of an action that is serious, complete, and of a certain magnitude; in language made pleasurable, with each kind of ornament used separately in the different parts; performed by people acting and not by narration; and, through pity and fear, accomplishing the purgation of those emotions.

From that definition six parts follow, and they are the only six.

1. Plot
2. Character
3. Diction
4. Thought
5. Spectacle
6. Song

And of these, by far the most important is the structure of the incidents — which is what I mean by plot.

Here is why, and this is the point on which I part company with almost everyone.

Tragedy is not an imitation of men. It is an imitation of an action, and of life. And life consists in action; its end is a way of acting, not a quality. Men have their qualities by their character, but it is by their actions that they turn out happy or the reverse.

So the poet does not construct an action in order to display a character. The character comes in as an accessory to the action.

Consider two proofs.

• First: you cannot have a tragedy without action, and you can have one without character. Most of our contemporary poets fail entirely at character, and they still produce tragedies. It is the same in painting — Polygnotus renders character well and Zeuxis does not, and Zeuxis is not therefore no painter.

• Second: string together a set of speeches expressing character, finished beautifully in their diction and their thought, and you will not produce the tragic effect nearly so well as a play that is deficient in all of that but has a plot with its incidents artfully arranged.

And notice that the most powerful sources of emotion in tragedy — the reversal, and the recognition — are parts of the plot. Not of anything else.

Beginners, incidentally, get to precision of diction and portraiture long before they can build a plot. That was true of nearly all the early poets.

The plot, then, is the first principle and, so to speak, the soul of a tragedy. Character is second.

Note

The same holds in painting. The most beautiful colors, laid on at random, will not give you as much pleasure as a portrait drawn in outline with chalk.

And I will put the spectacle last of all. It has an emotional attraction of its own, but it is the least artistic of the parts and the least connected with the poet's art — because the power of a tragedy is felt without any performance and without any actors at all. Besides, producing spectacular effects depends more on the stage machinist than on the poet.

A Whole — and One Man Is Not One Action

I said the action must be complete and whole, and of a certain magnitude, so let me define those.

A whole is what has a beginning, a middle, and an end. A beginning is what does not itself necessarily follow from anything, but after which something naturally is or comes about. An end is the reverse — what naturally follows something else, by necessity or as a rule, but has nothing following it. A middle follows one thing and is followed by another.

So a well-made plot must not begin or end just anywhere.

As for magnitude: beauty depends on size and order both. A very small creature cannot be beautiful, because you take it in during an almost imperceptible instant and the view is confused. Nor can an enormous one — if it were a thousand miles long, the eye could not take it in at once and the sense of the whole would be lost.

So with a plot. It needs a length the memory can hold.

And here is the practical rule. The proper magnitude is whatever allows the sequence of events, following probability or necessity, to admit a change from bad fortune to good, or from good to bad. Now, a warning about unity, because this is where poets most often go wrong.

Unity of plot does not consist in unity of the hero.

An infinite number of things happen to one man, and they cannot be reduced to one thing. This is the mistake of everyone who has ever composed a Heracleid or a Theseid. They assume that because Heracles was one man, the story of Heracles must be one story.

Homer, superior in everything else, saw this too — whether by art or by instinct. In composing the Odyssey, he did not put in everything that happened to Odysseus. He left out the wound on Parnassus and the feigned madness at the mustering, because there was no necessary or probable connection between them.

Here is the test, and it is severe. The parts must be so joined that if any one of them is shifted or taken out, the whole is dislocated and disturbed. Anything whose presence or absence makes no visible difference is not a part of the whole at all.

Poetry Is Truer Than History

Which brings me to what I most want you to understand.

It is not the poet's job to relate what has happened. It is his job to relate what may happen — what is possible according to probability or necessity.

The poet and the historian do not differ by writing in verse or in prose. You could put the work of Herodotus into verse and it would still be history, with metre exactly as much as without it.

The real difference is that one relates what has happened, and the other what may happen.

And for that reason poetry is a more philosophical thing than history, and a higher thing, because poetry tends to speak of the universal and history of the particular.

By the universal I mean: how a person of a certain type will, on a given occasion, speak or act — by probability or necessity. By the particular I mean what Alcibiades did, or what was done to him.

It follows that the poet must be a maker of plots rather than of verses, since he is a poet by virtue of imitating, and what he imitates are actions. And if he happens to take a subject out of real history, he is no less a poet for that — there is no reason why some things that actually happened should not also conform to probability, and it is in respect of that quality that he is their maker.

The worst plots of all are the episodic — by which I mean plots where the episodes follow one another with no probable or necessary sequence. Bad poets write them through incompetence. Good poets write them to please the actors: writing showpieces for competition, they stretch a plot past its capacity and are forced to break the natural continuity.

Reversal, Recognition, and the Statue That Fell

Plots are either simple or complex, because actions in life divide the same way. A simple action is one where the change of fortune happens without reversal and without recognition. A complex one has one or both.

And both must arise out of the internal structure of the plot, so that what follows is the necessary or the probable result of what preceded. It makes all the difference in the world whether an event happens because of the previous events or merely after them.

• Reversal is a change by which the action swings round into its opposite. Consider the Oedipus: the messenger arrives to cheer Oedipus and free him from his fear about his mother — and by revealing who Oedipus is, produces precisely the opposite effect.

• Recognition, as the word says, is a change from ignorance to knowledge, producing either love or hatred between people marked out for good or bad fortune. The finest form of recognition coincides with a reversal, as it does in the Oedipus.

Now — tragedy imitates events that inspire fear and pity, and the effect is best when the events arrive unexpectedly and follow one another as cause and effect. The wonder is greater then than if they had happened by themselves or by accident. Because even coincidences are most striking when they have an air of design.

Take the statue of Mitys at Argos, which fell on the man who had murdered Mitys, while he was standing there watching a festival, and killed him. Events of that kind do not feel like chance.

The Man Who Is Not Eminently Good

So what should the poet aim at, and what should he avoid?

Work through the possibilities and they eliminate themselves.

• Do not show a thoroughly good man passing from prosperity into misery. That moves neither pity nor fear. It merely disgusts us.

• Do not show a bad man passing from misery into prosperity. Nothing could be further from the spirit of tragedy. It has no tragic quality whatever — it does not satisfy our moral sense and it produces neither of the two emotions.

• Do not show the downfall of an utter villain. That would satisfy the moral sense, but it will still inspire neither pity nor fear — because pity is aroused by undeserved misfortune, and fear by the misfortune of a man like ourselves.

What remains is the character between those extremes. A man not pre-eminently virtuous and just, whose misfortune is brought on him not by vice or depravity but by some error — and who is highly renowned and prosperous. A man like Oedipus, or Thyestes.

And the change should be single in its issue, not double. It should run from good fortune to bad, and not the reverse.

I know some people place first the other kind — the kind with a double thread, like the Odyssey, where the good and the bad meet opposite ends. It is ranked first because of the weakness of audiences, and because poets follow what audiences want. The pleasure it gives is not the tragic pleasure. It belongs to comedy, where the deadliest enemies in the story walk off the stage as friends, and nobody kills anybody.

This is also why the people who criticize Euripides for ending unhappily are wrong. It is the right ending. And the proof is on the stage: in competition, those plays, when well worked out, have the most tragic effect. Euripides may be faulty in the general management of his material. He is still felt to be the most tragic of the poets.

You Should Be Able to Shudder With Your Eyes Shut

Fear and pity can be produced by spectacle. They can also be produced by the inner structure of the piece — and that is the better way, and the mark of the better poet.

Here is the standard I hold to. The plot should be so put together that a man who only hears the story told, without seeing anything at all, will shudder with horror and be melted to pity by what happens. That is what you get from hearing the story of the Oedipus.

To get the same effect out of spectacle is a lesser method and depends on external help. And those who use spectacle to produce not the terrible but merely the monstrous have nothing to do with tragedy at all. We must not demand every kind of pleasure from tragedy. Only the one proper to it.

As for where to look for your material: the deed must fall between people who are close to one another. If an enemy kills an enemy, there is nothing to pity, in the act or in the intention, beyond the suffering itself. The same with people indifferent to each other.

But when it happens between those who are near and dear — a brother killing, or about to kill, a brother; a son his father; a mother her son — those are the situations the poet should hunt for.

And of the ways of handling it, the worst is this: to be about to act, knowing who the person is, and then not to act. That is shocking without being tragic, because no disaster follows. Better that the deed be done. Better still that it be done in ignorance and the discovery come afterwards — then there is nothing to revolt us, and the discovery comes as a blow.

Prefer the Probable Impossibility

Two rules on the making, and I have nearly finished.

When you are constructing the plot and working out the words, put the scene in front of your own eyes. See it with the greatest possible vividness, as if you were standing there watching, and you will find what fits and be least likely to miss a contradiction.

And work it through with the gestures. A man who actually feels the emotion is the most convincing, because he shares the nature of what he is representing: the man who is genuinely agitated storms, and the angry man rages, with the most lifelike reality. Which is why poetry demands either a happy gift of nature or a strain of madness — in the first case a man can take the mould of any character, and in the second he is lifted out of his own.

The other rule concerns truth, and it is the one that sounds like a paradox until you have tried to write anything.

The wonderful is pleasing. You can see that from the fact that everybody telling a story adds something of his own, because he knows his hearers like it. And it is Homer above all who taught the other poets the art of lying skilfully.

The trick works on a fallacy. If one thing being true means a second thing is true, men assume that if the second is true the first must be as well. It is a false inference — but the mind, knowing the second, quietly infers the first.

From which follows the rule:

the poet should prefer a probable impossibility to an improbable possibility

And keep the irrational out of the action. If you cannot get rid of it altogether, put it outside the play — as the ignorance of Oedipus about the manner of Laius's death lies outside the play. Do not untie your plot with a god lowered from the machine. Reserve the god for what lies beyond human knowledge, before or after the action, since it is to the gods we attribute seeing everything.

Inside the action there must be nothing irrational at all.

Which Is the Higher Art

Some people hold that epic is the superior form. Let me finish by saying why I think the reverse.

Tragedy has everything epic has — it may even use the epic metre — and in addition it has music and spectacle as powerful accessories, which produce the most vivid of pleasures. It is vivid in the reading as well as in performance. And it reaches its end within narrower limits, and a concentrated effect gives more pleasure than one diluted over a long stretch of time.

What, for instance, would become of the Oedipus of Sophocles if it were stretched to the length of the Iliad?

Epic also has less unity, and the proof of that is that any epic will furnish subjects for several tragedies.

So if tragedy is superior in all these respects, and fulfils its own function better — and every art ought to produce not any chance pleasure but the pleasure proper to it — then plainly tragedy is the higher art, since it attains its end more perfectly.

Of the poetry that imitates in hexameter verse, and of comedy, I shall speak hereafter.