Aristotle's Poetics aims to give an account of poetry. Aristotle does this by attempting to explain poetry through first principles, and by classifying poetry into its different genres and component parts.
The centerpiece of Aristotle's work is his examination of tragedy. This occurs in Chapter 6 of "Poetics:" "Tragedy, then, is an imitation of an action that is serious, complete, and of a certain magnitude; in language embellished with each kind of artistic ornament, the several kinds being found in separate parts of the play; in the form of action, not of narrative; through pity and fear effecting the proper purgation of these emotions."
THE ART OF POETRY
TRANSLATED BY INGRAM BYWATER
WITH A PREFACE BY GILBERT MURRAY
OXFORD AT THE CLARENDON PRESS FIRST PUBLISHED 1920 REPRINTED 1925, 1928, 1932, 1938, 1945, 1947 1951, 1954, 1959. 1962 PRINTED IN GREAT BRITAIN
In the tenth book of the “Republic”, when Plato has completed his final burning denunciation of Poetry, the false Siren, the imitator of things which themselves are shadows, the ally of all that is low and weak in the soul against that which is high and strong, who makes us feed the things we ought to starve and serve the things we ought to rule, he ends with a touch of compunction: 'We will give her champions, not poets themselves but poet-lovers, an opportunity to make her defence in plain prose and show that she is not only sweet--as we well know--but also helpful to society and the life of man, and we will listen in a kindly spirit. For we shall be gainers, I take it, if this can be proved.' Aristotle certainly knew the passage, and it looks as if his treatise on poetry was an answer to Plato's challenge.
Few of the great works of ancient Greek literature are easy reading. They nearly all need study and comment, and at times help from a good teacher, before they yield up their secret. And the _Poetics_ cannot be accounted an exception. For one thing the treatise is fragmentary. It originally consisted of two books, one dealing with Tragedy and Epic, the other with Comedy and other subjects. We possess only the first. For another, even the book we have seems to be unrevised and unfinished. The style, though luminous, vivid, and in its broader division systematic, is not that of a book intended for publication. Like most of Aristotle's extant writing, it suggests the MS. of an experienced lecturer, full of jottings and adscripts, with occasional phrases written carefully out, but never revised as a whole for the general reader. Even to accomplished scholars the meaning is often obscure, as may be seen by a comparison of the three editions recently published in England, all the work of savants of the first eminence,  or, still more strikingly, by a study of the long series of misunderstandings and overstatements and corrections which form the history of the _Poetics_ since the Renaissance.
 Prof. Butcher, 1895 and 1898; Prof. Bywater, 1909; and Prof. Margoliouth, 1911.
But it is of another cause of misunderstanding that I wish principally to speak in this preface. The great edition from which the present translation is taken was the fruit of prolonged study by one of the greatest Aristotelians of the nineteenth century, and is itself a classic among works of scholarship. In the hands of a student who knows even a little Greek, the translation, backed by the commentary, may lead deep into the mind of Aristotle. But when the translation is used, as it doubtless will be, by readers who are quite without the clue provided by a knowledge of the general habits of the Greek language, there must arise a number of new difficulties or misconceptions.
To understand a great foreign book by means of a translation is possible enough where the two languages concerned operate with a common stock of ideas, and belong to the same period of civilization. But between ancient Greece and modern England there yawn immense gulfs of human history; the establishment and the partial failure of a common European religion, the barbarian invasions, the feudal system, the regrouping of modern Europe, the age of mechanical invention, and the industrial revolution. In an average page of French or German philosophy nearly all the nouns can be translated directly into exact equivalents in English; but in Greek that is not so. Scarcely one in ten of the nouns on the first few pages of the _Poetics_ has an exact English equivalent. Every proposition has to be reduced to its lowest terms of thought and then re-built. This is a difficulty which no translation can quite deal with; it must be left to a teacher who knows Greek. And there is a kindred difficulty which flows from it. Where words can be translated into equivalent words, the style of an original can be closely followed; but no translation which aims at being written in normal English can reproduce the style of Aristotle. I have sometimes played with the idea that a ruthlessly literal translation, helped out by bold punctuation, might be the best. For instance, premising that the words _poesis_, _poetes_ mean originally 'making' and 'maker', one might translate the first paragraph of the _Poetics_ thus:--
MAKING: kinds of making: function of each, and how the Myths ought to be put together if the Making is to go right.
Number of parts: nature of parts: rest of same inquiry.
Begin in order of nature from first principles.
Epos-making, tragedy-making (also comedy), dithyramb-making (and most fluting and harping), taken as a whole, are really not Makings but Imitations. They differ in three points; they imitate (a) different objects, (b) by different means, (c) differently (i.e. different manner).
Some artists imitate (i.e. depict) by shapes and colours. (Obs. Sometimes by art, sometimes by habit.) Some by voice. Similarly the above arts all imitate by rhythm, language, and tune, and these either
(1) separate or (2) mixed.
Rhythm and tune alone, harping, fluting, and other arts with same effect--e.g. panpipes.
Rhythm without tune: dancing. (Dancers imitate characters, emotions, and experiences by means of rhythms expressed in form.)
Language alone (whether prose or verse, and one form of verse or many): this art has no name up to the present (i.e. there is no name to cover mimes and dialogues and any similar imitation made in iambics, elegiacs, &c. Commonly people attach the 'making' to the metre and say 'elegiac-makers', 'hexameter-makers,' giving them a common class-name by their metre, as if it was not their imitation that makes them 'makers').
Such an experiment would doubtless be a little absurd, but it would give an English reader some help in understanding both Aristotle's style and his meaning.
For example, there i.e.lightenment in the literal phrase, 'how the myths ought to be put together.' The higher Greek poetry did not make up fictitious plots; its business was to express the heroic saga, the myths. Again, the literal translation of _poetes_, poet, as 'maker', helps to explain a term that otherwise seems a puzzle in the _Poetics_. If we wonder why Aristotle, and Plato before him, should lay such stress on the theory that art is imitation, it is a help to realize that common language called it 'making', and it was clearly not 'making' in the ordinary sense. The poet who was 'maker' of a Fall of Troy clearly did not make the real Fall of Troy. He made an imitation Fall of Troy. An artist who 'painted Pericles' really 'made an imitation Pericles by means of shapes and colours'. Hence we get started upon a theory of art which, whether finally satisfactory or not, is of immense importance, and are saved from the error of complaining that Aristotle did not understand the 'creative power' of art.
As a rule, no doubt, the difficulty, even though merely verbal, lies beyond the reach of so simple a tool as literal translation. To say that tragedy 'imitate.g.od men' while comedy 'imitates bad men' strikes a modern reader as almost meaningless. The truth is that neither 'good' nor 'bad' is an exact equivalent of the Greek. It would be nearer perhaps to say that, relatively speaking, you look up to the characters of tragedy, and down upon those of comedy. High or low, serious or trivial, many other pairs of words would have to be called in, in order to cover the wide range of the common Greek words. And the point is important, because we have to consider whether in Chapter VI Aristotle really lays it down that tragedy, so far from being the story of un-happiness that we think it, is properly an imitation of _eudaimonia_--a word often translated 'happiness', but meaning something more like 'high life' or 'blessedness'. 
 See Margoliouth, p. 121. By water, with most editors, emends the text.
Another difficult word which constantly recurs in the _Poetics_ is _prattein_ or _praxis_, generally translated 'to act' or 'action'. But _prattein_, like our 'do', also has an intransitive meaning 'to fare' either well or ill; and Professor Margoliouth has pointed out that it seems more true to say that tragedy shows how men 'fare' than how they 'act'. It shows thei.e.periences or fortunes rather than merely their deeds. But one must not draw the line too bluntly. I should doubt whether a classical Greek writer was ordinarily conscious of the distinction between the two meanings. Certainly it i.e.sier to regard happiness as a way of faring than as a form of action. Yet Aristotle can use the passive of _prattein_ for things 'done' or 'gone through' (e.g. 52a, 22, 29: 55a, 25).
The fact is that much misunderstanding is often caused by our modern attempts to limit too strictly the meaning of a Greek word. Greek was very much a live language, and a language still unconscious of grammar, not, like ours, dominated by definitions and trained upon dictionaries. An instance is provided by Aristotle's famous saying that the typical tragic hero is one who falls from high state or fame, not through vice or depravity, but by some great _hamartia_. _Hamartia_ means originally a 'bad shot' or 'error', but is currently used for 'offence' or 'sin'. Aristotle clearly means that the typical hero is a great man with 'something wrong' in his life or character; but I think it is a mistake of method to argue whether he means 'an intellectual error' or 'a moral flaw'. The word is not so precise.
Similarly, when Aristotle says that a deed of strife or disaster is more tragic when it occurs 'amid affections' or 'among people who love each other', no doubt the phrase, as Aristotle's own examples show, would primarily suggest to a Greek feuds between near relations. Yet some of the meaning is lost if one translates simply 'within the family'.
There is another series of obscurities or confusions in the _Poetics_ which, unless I am mistaken, arises from the fact that Aristotle was writing at a time when the great age of Greek tragedy was long past, and was using language formed in previous generations. The words and phrases remained in the tradition, but the forms of art and activity which they denoted had sometimes changed in the interval. If we date the _Poetics_ about the year 330 B.C., as seems probable, that is more than two hundred years after the first tragedy of Thespis was produced in Athens, and more than seventy after the death of the last great masters of the tragic stage. When we remember that a training in music and poetry formed a prominent part of the education of every wellborn Athenian, we cannot be surprised at finding in Aristotle, and to a less extent in Plato, considerable traces of a tradition of technical language and even of aesthetic theory.
It is doubtless one of Aristotle's great services that he conceived so clearly the truth that literature is a thing that grows and has a history. But no writer, certainly no ancient writer, is always vigilant. Sometimes Aristotle analyses his terms, but very often he takes them for granted; and in the latter case, I think, he is sometimes deceived by them. Thus there seem to be cases where he has been affected in his conceptions of fifth-century tragedy by the practice of his own day, when the only living form of drama was the New Comedy.
For example, as we have noticed above, true Tragedy had always taken its material from the sacred myths, or heroic sagas, which to the classical Greek constituted history. But the New Comedy was in the habit of inventing its plots. Consequently Aristotle falls into using the word _mythos_ practically in the sense of 'plot', and writing otherwise in a way that is unsuited to the tragedy of the fifth century. He says that tragedy adheres to 'the historical names' for an aesthetic reason, because what has happened is obviously possible and therefore convincing. The real reason was that the drama and the myth were simply two different expressions of the same religious kernel (p.
44) . Again, he says of the Chorus (p. 65) that it should be an integral part of the play, which is true; but he also says that it' should be regarded as one of the actors', which shows to what an extent the Chorus in his day was dead and its technique forgotten. He had lost the sense of what the Chorus was in the hands of the great masters, say in the Bacchae or the Eumenides. He mistakes, again, the use of that epiphany of a God which is frequent at the end of the single plays of Euripides, and which seems to have been equally so at the end of the trilogies of Aeschylus. Having lost the living tradition, he sees neither the ritual origin nor the dramatic value of these divine epiphanies. He thinks of the convenient gods and abstractions who sometimes spoke the prologues of the New Comedy, and imagines that the God appears in order to unravel the plot. As a matter of fact, in one play which he often quotes, the _Iphigenia Taurica_, the plot is actually distorted at the very end in order to give an opportunity for the epiphany.
 See my _Euripides and his Age_, pp. 221-45.
One can see the effect of the tradition also in his treatment of the terms Anagnorisis and Peripeteia, which Professor Bywater translates as 'Discovery and Peripety' and Professor Butcher as 'Recognition and Reversal of Fortune'. Aristotle assumes that these two elements are normally present in any tragedy, except those which he calls 'simple'; we may say, roughly, in any tragedy that really has a plot. This strikes a modern reader as a very arbitrary assumption. Reversals of Fortune of some sort are perhaps usual in any varied plot, but surely not Recognitions? The clue to the puzzle lies, it can scarcely be doubted, in the historical origin of tragedy. Tragedy, according to Greek tradition, is originally the ritual play of Dionysus, performed at his festival, and representing, as Herodotus tells us, the 'sufferings' or 'passion' of that God. We are never directly told what these 'sufferings' were which were so represented; but Herodotus remarks that he found in Egypt a ritual that was 'in almost all points the same'.  This was the well-known ritual of Osiris, in which the god was torn in pieces, lamented, searched for, discovered or recognized, and the mourning by a sudden Reversal turned into joy. In any tragedy which still retained the stamp of its Dionysiac origin, this Discovery and Peripety might normally be expected to occur, and to occur together. I have tried to show elsewhere how many of our extant tragedies do, as a matter of fact, show the marks of this ritual.
 Cf. Hdt. Ii. 48; cf. 42,144. The name of Dionysus must not be openly mentioned in connexion with mourning (ib. 61, 132, 86). This may help to explain the transference of the tragic shows to other heroes.
 In Miss Harrison's _Themis_, pp. 341-63.
I hope it is not rash to surmise that the much-debated word __katharsis__, 'purification' or 'purgation', may have come into Aristotle's mouth from the same source. It has all the appearance of being an old word which is accepted and re-interpreted by Aristotle rather than a word freely chosen by him to denote the exact phenomenon he wishes to describe. At any rate the Dionysus ritual itself was a _katharmos_ or _katharsis_--a purification of the community from the taints and poisons of the past year, the old contagion of sin and death. And the words of Aristotle's definition of tragedy in Chapter VI might have been used in the days of Thespis in a much cruder and less metaphorical sense. According to primitive ideas, the mimic representation on the stage of 'incidents arousing pity and fear' did act as a _katharsis_ of such 'passions' or 'sufferings' in real life. (For the word _pathemata_ means 'sufferings' as well as 'passions'.) It is worth remembering that in the year 361 B.C., during Aristotle's lifetime, Greek tragedies were introduced into Rome, not on artistic but on superstitious grounds, as a _katharmos_ against a pestilence (Livy vii. 2). One cannot but suspect that in his account of the purpose of tragedy Aristotle may be using an old traditional formula, and consciously or unconsciously investing it with a new meaning, much as he has done with the word _mythos_.
Apart from these historical causes of misunderstanding, a good teacher who uses this book with a class will hardly fail to point out numerous points on which two equally good Greek scholars may well differ in the mere interpretation of the words. What, for instance, are the 'two natural causes' in Chapter IV which have given birth to Poetry? Are they, as our translator takes them, (1) that man is imitative, and (2) that people delight in imitations? Or are they (1) that man is imitative and people delight in imitations, and (2) the instinct for rhythm, as Professor Butcher prefers? Is it a 'creature' a thousand miles long, or a 'picture' a thousand miles long which raises some trouble in Chapter VII? The word _zoon_ means equally 'picture' and 'animal'. Did the older poets make their characters speak like 'statesmen', _politikoi_, or merely like ordinary citizens, _politai_, while the moderns made theirs like 'professors of rhetoric'? (Chapter VI, p. 38; cf. Margoliouth's note and glossary).
It may seem as if the large uncertainties which we have indicated detract in a ruinous manner from the value of the _Poetics_ to us as a work of criticism. Certainly if any young writer took this book as a manual of rules by which to 'commence poet', he would find himself embarrassed. But, if the book is properly read, not as a dogmatic text-book but as a first attempt, made by a man of astounding genius, to build up in the region of creative art a rational order like that which he established in logic, rhetoric, ethics, politics, physics, psychology, and almost every department of knowledge that existed in his day, then the uncertainties become rather a help than a discouragement. The.g.ve us occasion to think and use our imagination. They make us, to the best of our powers, try really to follow and criticize closely the bold gropings of an extraordinary thinker; and it is in this process, and not in any mere collection of dogmatic results, that we shall find the true value and beauty of the _Poetics_.
The book is of permanent value as a mere intellectual achievement; as a store of information about Greek literature; and as an original or first-hand statement of what we may call the classical view of artistic criticism. It does not regard poetry as a matter of unanalysed inspiration; it makes no concession to personal whims or fashion or _ennui_. It tries by rational methods to find out what is good in art and what makes it good, accepting the belief that there is just as truly a good way, and many bad ways, in poetry as in morals or in playing billiards. This is no place to try to sum up its main conclusions. But it is characteristic of the classical view that Aristotle lays his greatest stress, first, on the need for Unity in the work of art, the need that each part should subserve the whole, while irrelevancies, however brilliant in themselves, should be cast away; and next, on the demand that great art must have for its subject the great way of living. These judgements have often been misunderstood, but the truth in them is profound and goes near to the heart of things.
Characteristic, too, is the observation that different kinds of art grow and develop, but not indefinitely; they develop until they 'attain their natural form'; also the rule that each form of art should produce 'not every sort of pleasure but its proper pleasure'; and the sober language in which Aristotle, instead of speaking about the sequence of events in a tragedy being 'inevitable', as we bombastic moderns do, merely recommends that they should be 'either necessary or probable' and 'appear to happen because of one another'.
Conceptions and attitudes of mind such as these constitute what we may call the classical faith in matters of art and poetry; a faith which is never perhaps fully accepted in any age, yet, unlike others, is never forgotten but lives by being constantly criticized, re-asserted, and rebelled against. For the fashions of the ages vary in this direction and that, but they vary for the most part from a central road which was struck out by the imagination of Greece.
ARISTOTLE ON THE ART OF POETRY
Our subject being Poetry, I propose to speak not only of the art in general but also of its species and their respective capacities; of the structure of plot required for a good poem; of the number and nature of the constituent parts of a poem; and likewise of any other matters in the same line of inquiry. Let us follow the natural order and begin with the primary facts.
Epic poetry and Tragedy, as also Comedy, Dithyrambic poetry, and most flute-playing and lyre-playing, are all, viewed as a whole, modes of imitation. But at the same time they differ from one another in three ways, either by a difference of kind in their means, or by differences in the objects, or in the manner of their imitations.
I. Just as form and colour are used as means by some, who (whether by art or constant practice) imitate and portray many things by their aid, and the voice is used by others; so also in the above-mentioned group of arts, the means with them as a whole are rhythm, language, and harmony--used, however, either singly or in certain combinations. A combination of rhythm and harmony alone is the means in flute-playing and lyre-playing, and any other arts there may be of the same description, e.g. imitative piping. Rhythm alone, without harmony, is the means in the dancer's imitations; for even he, by the rhythms of his attitudes, may represent men's characters, as well as what they do and suffer. There is further an art which imitates by language alone, without harmony, in prose or in verse, and if in verse, either in some one or in a plurality of metres. This form of imitation is to this day without a name. We have no common name for a mime of Sophron or Xenarchus and a Socratic Conversation; and we should still be without one even if the imitation in the two instances were in trimeters or elegiacs or some other kind of verse--though it is the way with people to tack on 'poet' to the name of a metre, and talk of elegiac-poets and epic-poets, thinking that they call them poets not by reason of the imitative nature of their work, but indiscriminately by reason of the metre they write in. Even if a theory of medicine or physical philosophy be put forth in a metrical form, it is usual to describe the writer in this way; Homer and Empedocles, however, have really nothing in common apart from their metre; so that, if the one is to be called a poet, the other should be termed a physicist rather than a poet. We should be in the same position also, if the imitation in these instances were in all the metres, like the _Centaur_ (a rhapsody in a medley of all metres) of Chaeremon; and Chaeremon one has to recognize as a poet. So much, then, as to these arts. There are, lastly, certain other arts, which combine all the means enumerated, rhythm, melody, and verse, e.g. Dithyrambic and Nomic poetry, Tragedy and Comedy; with this difference, however, that the three kinds of means are in some of them all employed together, and in others brought in separately, one after the other. These elements of difference in the above arts I term the means of their imitation.