Rhetoric · Book III

Aristotle · a new plain-English translation from the original language

1| [3.1.1] Since there are three things that must be dealt with concerning speech — first, from what sources the means of persuasion will come; second, concerning style; and third, how one must arrange the parts of the speech — concerning the means of persuasion it has been said both from how many sources they come, that they are from three, and what these are, and why there are only so many (for all people are persuaded either because the judges themselves are affected in some way, or because they suppose the speakers to be of a certain character, or because something has been demonstrated); and it has also been said concerning the enthymemes, whence one must procure them (for some are species of enthymemes, others are topics). [3.1.2] Next in order it remains to speak concerning style; for it is not enough to have what one ought to say, but one must also of necessity say it as one ought, and this contributes much to making the speech appear to be of a certain quality. [3.1.3] Now the first inquiry pursued, according to nature, was that which is naturally first — the facts themselves, from what sources they get their power to persuade; the second was to arrange these in style; and the third, which has the greatest power of all yet has not yet been undertaken, concerns delivery. Indeed, delivery came late even into tragic performance and rhapsodic recitation;

2| for at first the poets themselves acted their own tragedies. It is clear, then, that there is such a thing in rhetoric too, just as in poetics — a matter certain others have treated, among them Glaucon of Teos. [3.1.4] This consists in the use of the voice — how one ought to use it for each emotion, for example when loud and when soft and when in a middle register, and how to use the pitches, for example acute, grave, and intermediate, and what rhythms to use for each occasion. For there are three things that people look to in this regard: volume, harmony, and rhythm. Now those who practice this art win, one might almost say, most of the prizes in the contests, and just as there actors now have more power than the poets, so it is in the political contests too, because of the corruption of the citizens.

3| [3.1.5] But no systematic art concerning these matters has yet been composed, since even the matter of style came along late; and it seems a vulgar thing, when rightly understood. But since the whole business concerning rhetoric is directed toward opinion, one must give it attention not because it is right to do so but because it is necessary, since indeed what is just is to seek nothing more in a speech than that it neither give pain nor give pleasure; for it is just to contend by means of the facts themselves alone, so that everything apart from demonstration is superfluous — and yet it has great power, as has been said, because of the corruption of the hearer. [3.1.6] Still, the matter of style does have some small necessary place in all teaching; for it makes some difference to clarity to speak this way or that, though not so much as people think — rather, all such things are a matter of appearance, directed toward the hearer; and this is why no one teaches geometry in this way. [3.1.7] That art, then, whenever it does come, will do the same work as the art of delivery; and some have attempted to say a little about it, for example Thrasymachus in his work On Pity; and being a good actor is a matter of natural gift, and belongs less to art, while the matter of style is a matter of art. This is why prizes come, in turn, to those capable of this, just as to orators who excel in delivery; for speeches that are written down have more force through their style than through their thought. [3.1.8] The poets, then, were the first to set this in motion, as was natural; for words are imitations, and of all our parts the voice is the most capable of imitation. This is why the arts of rhapsodic recitation and acting, and others as well, came into being.

4| [3.1.9] Since the poets, though saying simple things, seemed to gain their reputation through their style, poetic style was the first to arise -- the style of Gorgias, for instance -- and even now most uneducated people think that speakers of this kind converse most beautifully. But this is not so; the style of prose and the style of poetry are different. What actually happens shows this: even the writers of tragedy no longer use style in the same way, but just as they moved from tetrameters to the iambic trimeter because this meter is closest of all the meters to ordinary speech, so too they have dropped those words that lie outside everyday language, the ones with which the first poets adorned their verse, and even now the writers of hexameters have dropped them. So it is ridiculous to imitate poets who themselves no longer use that manner. [3.1.10] It is clear, then, that we should not be precise about everything that can be said concerning style, but only about as much as concerns the kind of style we are discussing here; the other kind has been discussed in the works on poetics. [3.2.1] Let those points, then, be taken as settled, and let the virtue of style be defined as clarity -- since speech is a kind of sign, and if it does not make something clear it will not do its own work -- and let it be neither low nor above the dignity of the subject, but fitting; for poetic style is perhaps not low, but it is not fitting for prose. [3.2.2] Of nouns and verbs, the standard ones make style clear, while the other kinds of words discussed in the works on poetics make it not low but adorned, since departing from the ordinary makes style seem more dignified; for just as people feel toward foreigners as they feel toward their fellow citizens, they feel the same way toward style.

5| [3.2.3] For this reason one must give one's diction a foreign air, since people admire what is far off, and what is admired is pleasant. Now in verse this is done often and suits the verse well, since the subject matter and the persons concerned there have departed further from the ordinary; but in plain prose speeches it should be done much less, since the subject matter is lesser. For even in prose, if a slave used fine language, or someone too young did, it would be unfitting, or if the matter were too trivial; but even here there is a degree of fittingness that contracts and expands. [3.2.4] Hence one must do this without being caught doing it, and must not seem to speak in a contrived way but in a natural one, for the natural is persuasive and the contrived is the opposite -- people turn against it as against someone plotting against them, just as they turn against adulterated wine. It is like how the voice of Theodorus is different from the voice of the other actors: his seems to belong to the speaker, while theirs seem to belong to someone else. [3.2.5] The trick is worked well when one selects from ordinary language in composing -- which is what Euripides does, and was the first to show how. Since speech is composed of nouns and verbs, and nouns have as many kinds as have been discussed in the works on poetry, one should use rare words, compound words, and coined words rarely and in few places -- where, and why, we will say later, since they deviate too far from what is fitting -- [3.2.6] while the standard word, the proper word, and metaphor alone are useful for the style of plain prose speeches. A sign of this is that these are the only ones everyone uses: everyone converses using metaphors, proper words, and standard words. So it is clear that if one handles this well, the result will have a foreign air and yet escape notice, and will be clear -- and this was the virtue of rhetorical speech.

6| [3.2.7] Of nouns, ambiguous terms are useful to the sophist, since it is by means of them that he does his mischief, while synonyms are useful to the poet. By standard and synonymous terms I mean, for instance, "to journey" and "to walk": both of these are standard words, and they are synonymous with each other. What each of these things is, and how many kinds of metaphor there are, and that metaphor has the greatest power both in poetry and in prose speeches, has been discussed, as we said, in the works on poetics. [3.2.8] But one must work harder over metaphors in prose speech to the degree that prose has fewer resources at its disposal than meter has; for clarity, pleasure, and a foreign air belong most of all to metaphor, and it cannot be gotten from anyone else. [3.2.9] One must also apply epithets and metaphors that are fitting, and this will come from proportion; otherwise the result will appear unfitting, because opposites show up most clearly when set side by side. One must consider what suits an old man the way a purple robe suits a young man -- [3.2.10] for the same clothing does not suit both -- and if you want to adorn something, you should draw the metaphor from the best things of the same kind, and if you want to disparage it, from the worse things.

7| I mean, for instance, since opposites belong to the same genus, to say that the beggar prays and the one who prays begs, on the ground that both are a kind of asking, is to do what has just been described -- as when Iphicrates called Callias "a begging priest of the Mother" rather than "a torch-bearer," and Callias replied that Iphicrates himself must be uninitiated, since otherwise he would not call him a begging priest but a torch-bearer: both terms concern a god, but one is honorable and the other dishonorable. And some call flatterers of Dionysus "Dionysus-flatterers," while they call themselves "artists"; both of these are metaphor, the one degrading, the other the opposite. And pirates now call themselves "providers"; hence one may say that the wrongdoer "errs" and the one who errs "has done wrong," and that the thief both "took" and "provided for himself." But the line in which Euripides' Telephus says, "ruling the oar and having disembarked into Mysia," is unfitting, because "ruling" is too grand a word for what he deserved; so the trick has not been well concealed there.

8| [3.2.11] There is also a kind of error possible in the syllables themselves, if they are not signs of a pleasant sound -- as when Dionysius the Brazen, in his elegiac verses, calls poetry "the shrieking of Calliope," on the ground that both are a kind of voice; the metaphor is poor because of the meaningless sounds. [3.2.12] Further, one must not transfer a metaphor from far off, but from things akin and of the same kind, onto things that have no name, so that once it is stated it is clear that it is akin -- as in the celebrated riddle: "I saw a man who had glued bronze onto a man with fire": the experience has no name, but both cases are a kind of attachment, and so he called the application of the cupping-glass a "gluing." And in general, one can get good metaphors from well-made riddles, for metaphors imply a riddle, so it is clear that a metaphor drawn this way has been well transferred. [3.2.13] Metaphors should also be drawn from beautiful things. The beauty of a word lies, as Licymnius says, either in its sound or in what it signifies, and its ugliness likewise. There is also a third point, which refutes the sophistical argument: it is not the case, as Bryson said, that no one speaks shamefully, on the ground that saying this instead of that signifies the same thing -- for that is false. One word is more proper to a thing than another, more like it, and more suited to it, in that it sets the matter before the eyes. Further, this word and that word do not signify a thing in the same condition, so on this ground too one must hold that one thing is more beautiful or more ugly than another: for though both words signify the beautiful or the ugly, they do not signify it as beautiful or as ugly in the same degree, or rather they do, but as more or less. Metaphors should be drawn from this source -- from things beautiful either in sound, or in force, or in visual appearance, or in some other perception. And it makes a difference to say, for instance, "rosy-fingered dawn" rather than "purple-fingered," or worse still, "red-fingered."

9| [3.2.14] And with epithets it is possible to form them from something base or shameful, as in "the mother-slayer," or from something better, as in "the avenger of his father." And Simonides, when the winner with mules offered him only a small fee, was unwilling to write the ode, as if reluctant to write about mules, but when he was given enough, he wrote: "Hail, daughters of storm-footed horses" -- and yet they were also daughters of donkeys. There is, again, the use of diminutives. [3.2.15] The diminutive is what makes both the bad and the good smaller, as Aristophanes jokes in the Babylonians, saying "little gold-piece" instead of "gold," "little cloak" instead of "cloak," "little bit of abuse" instead of "abuse," and "little ailment" instead of "ailment." One must be careful, though, and observe due measure in both cases. [3.3.1] Frigidity of style occurs in four ways: in compound words, as when Lycophron speaks of "the many-faced sky of the great-peaked earth," and "the narrow-passaged shore"; and as Gorgias named perjurers set against a man who had kept his oath "beggarly Muse-flatterers"; and as Alcidamas speaks of "the soul filling with force," and "the countenance becoming fire-colored," and thought their eagerness would prove "fulfillment-bearing," and made the persuasiveness of his speeches "fulfillment-bearing," and called "the floor of the sea" "dark-hued." All these appear poetic on account of the compounding. [3.3.2] This, then, is one cause of frigidity; another is the use of rare, foreign-sounding words, as when Lycophron calls Xerxes "a monstrous man," and Sciron "a plundering man," and Alcidamas speaks of "a plaything for poetry," and "the recklessness of nature," and of one "sharpened by the unmixed anger of his mind."

10| [3.3.3] A third cause, in the use of epithets, is using them either at excessive length, or inopportunely, or too densely packed. For in poetry it is fitting to say "white milk," but in prose some such epithets are unfitting, while others, if piled on too thickly, give the game away and make it plain that the piece is poetry -- since one certainly must use them (for an epithet varies from the ordinary and makes the diction unfamiliar), but one must aim at due measure, since overdoing it does more harm than speaking carelessly does: careless speech merely lacks excellence, but overdone speech is positively bad. This is why the writing of Alcidamas seems frigid: he does not use epithets as a seasoning but as a main dish, so densely packed are they, so exaggerated, and so conspicuous -- for instance, not "sweat" but "the moist sweat," not "to the Isthmian games" but "to the assembly of the Isthmian games," not "laws" but "the laws that are kings of cities," not "at a run" but "with the running impulse of the soul," not "a school of the Muses" but "having taken over the school of the Muses that belongs to nature," and "the gloomy anxiety of the soul," and not "a craftsman of grace" but "a craftsman of grace for all the people," and "a steward of the pleasure of the listeners," and not "with branches" but "he hid it with the branches of the forest," and not "he covered his body" but "the shame of his body," and "the desire of the soul, mimicking its opposite" (this is at once both a compound and an epithet, so that it becomes a piece of poetry), and thus "an outlying excess of depravity."

11| This is why speaking poetically through such unfittingness produces both the laughable and the frigid, and also obscurity through sheer wordiness: for when a speaker piles words onto someone who already understands, he destroys clarity by throwing shade over it. People use compound words when a thing has no name of its own and the resulting expression is easy to put together, as in "to while away time"; but if the practice is carried too far, the effect is unmistakably poetic. This is why compound diction is most useful to writers of dithyrambs (for these are full of noise), rare words to epic poets (for epic is dignified and self-assertive), and metaphor to writers of iambic verse (for these are the meters now in use, as has been said).

12| [3.3.4] And there is, further, a fourth way in which frigidity occurs, namely in metaphors: for there are also unfitting metaphors, some unfitting because they are laughable (comic poets, too, use metaphors), others because they are too solemn and tragic; and metaphors are obscure if they are drawn from too far off, as when Gorgias speaks of "affairs pale and bloodless," and says, "you sowed these things shamefully and reaped them badly" -- for this is excessively poetic. And as Alcidamas calls philosophy "a fortification against the law," and the Odyssey "a fine mirror of human life," and offers nothing of that kind as "a plaything for poetry" -- all these are unconvincing for the reasons already stated. But Gorgias's remark about the swallow, when it flew over him and let its droppings fall, is the best of the tragic kind: he said, "How shameful, Philomela!" -- for if a bird had done it, it would not be shameful, but for a maiden it is shameful. So he rebuked her well by saying what she had been, not what she is. [3.4.1] The simile, too, is a metaphor, for it differs only slightly. For when one says of Achilles, "he rushed on like a lion," it is a simile; but when one says, "the lion rushed on," it is a metaphor -- since both are brave, the speaker transferred the term and called Achilles a lion. [3.4.2] The simile is useful in prose too, but only rarely,

13| for it is poetic. Similes are to be handled the way metaphors are, since they are metaphors, differing only in how they are expressed. [3.4.3] Examples of similes are the one Androtion made about Idrieus, that he was like puppies just let off their chain: for such dogs rush at people and bite, and Idrieus too, once freed from his fetters, was dangerous. And as Theodamas compared Archidamus to Euxenus not knowing geometry, by way of an analogy -- for Euxenus, too, would be an Archidamus who knows geometry. And there is the one in Plato's Republic, that those who strip the dead are like puppies that bite the stones thrown at them without touching the one who threw them; and the one aimed at the people, that they are like a ship's captain who is strong but rather deaf; and the one aimed at the meters of the poets, that they are like people who are lovely without beauty -- for once these have faded, and those have been broken up, neither still looks the same. And there is Pericles's simile about the Samians, that they were like children who take the food offered them but keep on crying; and the one about the Boeotians, that they are like holm-oaks -- for holm-oaks are cut down by wood from their own kind, and the Boeotians fight among themselves. And there is the one Demosthenes made about the people, that they are like people who get seasick on ships.

14| And there is the one Democrates made, comparing orators to wet-nurses who swallow the food themselves and then smear the children's mouths with their own spittle. And as Antisthenes compared the thin Cephisodotus to frankincense, because it delights people precisely as it is being consumed away. All of these can be stated either as similes or as metaphors, so that whatever wins approval when stated as a metaphor will clearly also work as a simile, and similes are simply metaphors that need an added word of explanation. [3.4.4] One must always render a metaphor drawn from proportion so that it applies reciprocally, on the other side too, among things of the same kind: for instance, if the cup is the shield of Dionysus, then it is also fitting to call the shield the cup of Ares.

15| [3.5.1] Speech, then, is composed out of these elements. Now the starting point of style is speaking correct Greek, and this consists in five things. [3.5.2] First, in the connecting particles, if one renders them in the order in which they naturally come before or after one another, as some of them require -- for instance, "on the one hand" requires "on the other hand," and "I, on the one hand" requires "but he, on the other." One must supply the answering particle to the first while it is still being remembered, and must neither hang it off too far, nor supply one connective before another when it is not yet due -- for that rarely fits. For example: "I, on the one hand, since he told me (for Cleon came both asking and demanding), set out, taking them with me." In a case like this, many connectives get thrown in ahead of the one that is finally to be answered; and if the gap before "I set out" grows too long, the result is obscure. [3.5.3] This, then, is one point of excellence in the use of connectives; a second is to speak in terms proper to the thing and not in broader terms that merely cover it. [3.5.4] A third is to avoid ambiguous terms -- unless one deliberately intends the opposite, which is what people do when they have nothing to say but want to seem to be saying something,

16| for people of that sort talk this way in poetry too, Empedocles for example: the sheer bulk of matter going round and round deceives the listener, and his audience is affected the way most people are affected by seers -- for when seers speak ambiguously, their hearers nod along in agreement: "Croesus, by crossing the Halys, will destroy a great empire." And seers speak in general terms about the matter because a mistake is then less likely overall: for in a game of guessing odd or even, one is more likely to be right saying "odd" or "even" than saying exactly how many there are, and likewise more likely to be right saying that something will happen than saying when -- which is why oracle-mongers never add a further specification of when. All these devices, then, are alike, so that unless one has some such purpose in mind, they are to be avoided.

17| [3.5.5] Fourth, as Protagoras divided the genders of nouns — male, female, and inanimate: for these too must be rendered correctly. [3.5.6] 'She came, and having talked with him, went away.' Fifth is naming the many, the few, and the one correctly: 'the men who came kept striking me.' In general, what is written should be easy to read and easy to deliver — and these are the same thing. An abundance of connectives lacks this, and so does whatever is not easy to punctuate, like the writings of Heraclitus. For punctuating Heraclitus is hard work, because it is unclear which word a given phrase attaches to, the one after or the one before — as in the very opening of his treatise, where he says, 'of this account's being always, men prove uncomprehending': for it is unclear which side 'always' should be punctuated with. [3.5.7] Further, this too produces a solecism: failing to render matters correctly, if you fail to attach a word that fits both terms — for instance, with sound and color: 'having seen' is not common to both, but 'having perceived' is common to both. And your speech will be unclear if you do not put first what you mean, when you intend to insert many things in between it — for instance, say 'for I intended, after conversing with him, to do this and this and go this way,' rather than 'for I intended, after conversing, to go' — and only then 'this and this happened.' [3.6.1] The following contribute to grandeur of style: using a phrase in place of a single name — for instance, not 'circle,' but 'the plane figure equidistant from the center'; for conciseness, the opposite — a single name in place of a phrase. [3.6.2] And if a thing is shameful or unseemly: if the shamefulness lies in the phrase, use the single name; if it lies in the name, use the phrase.

18| [3.6.3] And make things clear by metaphor and by epithets, taking care to avoid what is too poetic. [3.6.4] And turn one thing into many, as the poets do: though there is one harbor, they nevertheless say 'into Achaean harbors,' and 'these many-doored folds of the tablet.' [3.6.5] And do not yoke words together, but give each its own — 'of the woman, of ours'; but if speaking concisely, the opposite — 'of our woman.' [3.6.6] And speak with a connective; but if concisely, without a connective, yet not disconnected — for instance, 'having gone and having conversed' becomes 'having gone, I conversed.' [3.6.7] And there is Antimachus's useful device — speaking of things by what they lack, which is what he does concerning Teumessus: 'there is a small windswept hill' — for in this way the description grows without limit. And this works equally for good things and bad things, whichever way is useful, by saying what a thing does not have; and this is also where the poets get names like 'stringless' and 'lyreless' melody — for they produce these out of negations. This device is well regarded when used in metaphors formed by analogy, as when one says that the trumpet sends forth a lyreless melody. [3.7.1] Style will have propriety if it is expressive of feeling and of character, and proportionate to the subject matter at hand. [3.7.2] It is proportionate if weighty matters are not spoken of carelessly, nor trivial matters spoken of solemnly, and no ornament is fitted onto a trivial name; otherwise it looks like comedy, as Cleophon does — for he said some things just as one might say 'lady fig-tree.' [3.7.3] It is expressive of feeling if, where there is outrage, the diction is that of an angry man; if the matters are impious and shameful, the diction is that of one offended and reluctant even to speak of them; if the matters are praiseworthy, the diction is that of admiration; if pitiable, the diction is that of humility; and likewise with the rest.

19| [3.7.4] Style suited to the matter also makes the matter persuasive: for the hearer's soul is misled into thinking the speaker is telling the truth, because people are disposed that way toward such things, so that they suppose — even if the facts are not as the speaker says — that the facts are so; and the hearer always shares the feeling of one who speaks with emotion, even if he is saying nothing of substance. [3.7.5] That is why many speakers overwhelm their hearers by making an uproar. [3.7.6] This demonstration from signs is also expressive of character, whenever the style that fits each class and state follows accordingly. By class I mean by age — child, man, or old man — and by sex, woman or man, and by people, Spartan or Thessalian; by states I mean those conditions in virtue of which a person's manner of life has some particular quality, since it is not in virtue of every state that ways of life have some particular quality. [3.7.7] If, then, one also uses words appropriate to the state in question, this will produce the character; for a rustic person and an educated person would not say the same things, nor say them the same way. And the audience is affected in a certain way by a device the speechwriters use to excess — 'who does not know?', 'everyone knows' — for the hearer, out of shame, agrees, so as to share in what everyone else shares in. [3.7.8] Whether one uses such devices at the right moment or not is a matter common to every kind of style. [3.7.9] A remedy for every excess is the well-worn device: one should reproach oneself for it in advance, for then it seems to be true, since the speaker plainly does not fail to notice what he is doing. [3.7.10] Further, one should not use all the corresponding devices at once — for that is how the hearer is caught unawares. I mean, for instance, if the words are harsh, one should not also match them with a harsh voice and a harsh expression; otherwise each device becomes obvious for what it is. But if one matches in one respect and not another, one achieves the same effect without being noticed. So if soft things are said harshly and harsh things softly, the result becomes persuasive.

20| [3.7.11] Double words, an abundance of epithets, and foreign words suit a speaker expressing strong feeling best of all: one forgives an angry man for calling something evil 'sky-high' or calling it 'monstrous' — especially once he already has hold of his hearers and has made them enthusiastic, whether through praise, blame, anger, or affection, as Isocrates does, for instance, at the end of the Panegyricus: 'fame and memory and those who dared.' For people utter such things when they are in a state of enthusiasm, so that their hearers evidently accept them because they are in the same state. That is also why such language suits poetry — for poetry is inspired. So one should use it either in this way, or with irony, as Gorgias did, and as happens in the Phaedrus. [3.8.1] The shape of the style should be neither metrical nor entirely without rhythm: the metrical is unpersuasive, since it seems contrived, and at the same time it distracts the hearer, for it makes him watch for the recurrence, waiting for when the same pattern will come again — just as children anticipate the herald's formula: 'Whom does the freedman choose as his guardian?' — 'Cleon!' [3.8.2] What lacks rhythm entirely is unbounded, and speech should be bounded, but not by meter, for the unbounded is unpleasant and unknowable. Everything is bounded by number, and the number that bounds the shape of style is rhythm, of which meters too are segments.

21| [3.8.3] That is why prose should have rhythm, but not meter — otherwise it will be a poem. And it should have rhythm without exactness — this will hold if it is kept up only to a point. [3.8.4] Of the rhythms, the heroic requires a solemn harmony but not one suited to speech; the iambic is the very diction of ordinary people — that is why people utter iambic feet more than any other meter when they speak — but a speech needs solemnity to arise, and the hearer needs to be carried out of himself. The trochee is more suited to the cordax; the trochaic tetrameters show this, for the tetrameter is a tripping rhythm. What is left is the paean, which speakers have used ever since Thrasymachus began it, though they could not say what it was. The paean is a third kind, and related to the ones already mentioned: for its ratio is three to two, whereas of those others one is one to one and the other two to one; and next to these ratios comes the ratio of one and a half to one — and this is the paean.

22| [3.8.5] The other rhythms, then, must be dismissed for the reasons already given, and because they are metrical; the paean must be adopted, since it alone, of the rhythms mentioned, is not a meter, so that it escapes notice most easily. [3.8.6] Now people use one and the same paean both at the beginning and at the end, but the ending ought to differ from the beginning. There are two forms of the paean, opposite to each other, one of which suits a beginning, as indeed people use it: this is the one in which a long syllable begins and three short syllables end it, as in "Delos-born, whether of Lycia" and "Golden-haired Far-darter, son of Zeus." The other is the reverse, where three short syllables begin and a long syllable ends it: "After earth and waters and ocean, night made them vanish" (Simonides, fr. 26b). This one makes an ending, for the short syllable, because it is incomplete, produces a truncated effect. [3.8.7] Rather, the sentence should be cut off by the long syllable, and the ending should be evident not through the scribe, nor through a marginal notation, but through the rhythm itself. [3.9.1] That the style must have good rhythm and not be without rhythm, and what rhythms produce good rhythm and in what condition, has been stated. Style must necessarily be either strung-together and made one by the connective, like the preludes in dithyrambs, or turned back on itself and like the antistrophes of the ancient poets. [3.9.2] The strung-together style is the old one: "This is the exposition of the inquiry of Herodotus of Thurii" (this style everyone used formerly, but now not many do). By strung-together I mean a style that has no end in its own right, unless the matter being spoken of comes to an end. It is unpleasant because of its unboundedness, for everyone wants to see the end in view; that is why runners gasp and grow faint at the turning-post, since foreseeing the finish they do not tire beforehand.

23| [3.9.3] The strung-together style, then, is this one; the turned-back style is the one in periods. By a period I mean a stretch of speech that has a beginning and an end in its own right and a magnitude easy to take in at a glance. Such a style is pleasant and easy to learn - pleasant because it is the opposite of the unbounded, and because the hearer always thinks he is grasping something and that some point has been rounded off for him, whereas foreseeing nothing and finishing nothing is unpleasant; easy to learn because it is easy to remember, and this because the style in periods has number, which of all things is easiest to remember. That is why everyone remembers verse better than prose spoken at random, for verse has a number by which it is measured. [3.9.4] The period must also be brought to completion in its thought, and not be cut off in the middle, as in these iambics of Sophocles: "This land is Calydon; of the Peloponnesian soil" (Euripides, fr. 515) - for one can take the opposite sense by the way it is divided, as in the example given, namely that Calydon is part of the Peloponnese. [3.9.5] Of periods, one kind is in clauses, the other simple. A style in clauses is one that is complete, divided, and easy to draw breath through, not in the division as also the period, but as a whole [corrupt text] (a clause being one of the two parts of it).

24| By simple I mean the one-clause kind. [3.9.6] Both the clauses and the periods must be neither stunted nor long. A too-short clause often makes the hearer stumble, for when he is still pressing forward toward the measure whose limit he holds within himself, and is suddenly checked as the sentence stops, a kind of stumbling necessarily results from the collision. Long ones, on the other hand, leave the hearer behind, like people who turn back beyond the finish-line - these too leave behind those who were walking with them. Likewise, periods that are too long become a mere speech and resemble a prelude, so that what Democritus of Chios mocked in Melanippides comes about, when he composed preludes instead of antistrophes: "The man who works evil for another works evil for himself, and the long prelude is worst for the one who made it" (Hesiod, Works and Days 265) - for this fits equally well when said against those who use long clauses. And clauses that are too short do not make a period; they hurry the hearer along headlong.

25| [3.9.7] Of style in clauses, one kind is divided, the other antithetical. Divided, for example: "I have often admired those who convened the festivals and those who established the athletic contests." Antithetical is style in which, in each of the two clauses, either a contrary is set against a contrary, or the same term is yoked to two contraries - for example: "They benefited both those who stayed behind and those who followed; for to the one group they won more than what they had at home, to the other they left enough at home" - here "staying behind" is opposed to "following," "enough" to "more." Or: "for both those who need money and those who wish to enjoy it" - here "enjoyment" is opposed to "acquisition." And again it often happens in these that the prudent come to grief and the imprudent succeed: "They were at once judged worthy of the prizes for valor, but not long after took command of the sea." "To sail across the mainland and march across the sea, by bridging the Hellespont and cutting through Athos." "And though citizens by nature, to be deprived of the city by law." "Some of them perished miserably, others were saved shamefully." "Privately to employ barbarians as household servants, but publicly to look on while many of the allies are enslaved." "Either to hold them living or to leave them behind dead." And what someone said against Peitholaus and Lycophron in the lawcourt: "These men, when at home, used to sell you, but now that they have come to you, they have bought you." All these examples produce the effect described. [3.9.8] Such a style is pleasant because contraries are most easily recognized, and are still more recognizable when set beside one another, and because it resembles a syllogism, for refutation is a bringing-together of opposites.

26| [3.9.9] Antithesis, then, is a thing of this sort; equality of clauses is when the clauses are equal in length, and similarity of endings is when each of the two clauses has similar endings. This must occur either at the beginning or at the end - at the beginning it is always nouns, at the end either the final syllables, or inflections of the same noun, or the same noun itself. Examples at the beginning: "For he received a field from him, and it was untilled"; "They could be won over by gifts, and turned aside by words" (Homer, Iliad 9.526). At the end: "You would have thought that he had not fathered a child, but had himself become a child"; "in the greatest anxieties and in the least hopes." Inflections of the same noun: "Does he deserve to be honored with a bronze statue, when he is not worthy of bronze?" The same noun: "While he was alive you spoke ill of him, and now you write ill of him." From a syllable: "What terrible thing would have happened to you, if you had seen an idle man?" It is possible to have all these features at once, so that the same expression is antithesis, and equal in clauses, and similar in endings. The beginnings of periods have been more or less fully catalogued in the Theodectea. [3.9.10] There are also false antitheses, of the kind Epicharmus used to make: "At one time I was among them, at another I was beside them." [3.10.1] Now that these matters have been distinguished, we must say from what sources clever and well-received sayings come. Producing them belongs to a naturally gifted person, or to one trained by practice, but pointing them out belongs to this method of inquiry. [3.10.2] Let us then discuss them and enumerate them; let this be our starting point. Learning easily is by nature pleasant to everyone, and words signify something, so that whatever words produce learning in us are the most pleasant. Now, unfamiliar words we do not know, and current words we already know; it is metaphor above all that produces this effect, for when the poet calls old age "stubble," he produces understanding and knowledge through the genus, since both old age and stubble belong to the class of things that have lost their bloom.

27| [3.10.3] The images used by poets do the same thing, and that is why, when well made, they seem clever. For an image, as has been said before, is a metaphor differing by an added word; that is why it is less pleasant, because it is longer, and it does not say "this is that" — so the soul does not even look for that. [3.10.4] It follows, then, that both diction and enthymemes are clever whenever they produce quick learning in us; that is why superficial enthymemes are not well thought of (by "superficial" we mean those that are clear to everyone and require no inquiry at all), nor those which, once stated, we still do not understand, but rather those where understanding occurs at the very moment they are spoken, even if it was not there before, or where the mind lags only a little behind; for something like learning occurs, but in neither of the other two cases. [3.10.5] So then, as regards the thought of what is said, enthymemes of this kind are well thought of; but as regards the diction, this is by the figure, if it is put antithetically — for example, "those who think that the peace which is common to others is war on their own private interests": war is opposed to peace. [3.10.6] And by the words, if it has a metaphor, and one that is neither foreign (for that is hard to take in together) nor superficial (for that produces no effect on us). Further, if it puts the matter before the eyes; for one must see the things being done rather than about to be done.

28| One must aim, then, at these three things: metaphor, antithesis, actuality. [3.10.7] Of metaphors, there being four kinds, those by analogy are the most esteemed, as when Pericles said that the youth lost in the war had vanished from the city as if someone had taken the spring out of the year. And Leptines, about the Spartans, said that one should not allow Greece to become one-eyed. And Cephisodotus, when Chares was eager to render his account concerning the Olynthian war, was indignant, saying that he was trying to render his account to the people while choking them by the throat. And once, urging the Athenians to sail to Euboea with provisions, he said they must go out under Miltiades' decree. And Iphicrates, when the Athenians made a truce with Epidaurus and the coastal region, was indignant, saying that they had stripped away their war supplies. And Peitholaus called the state galley "the club of the people," and Sestos "the bread-bin of the Piraeus." And Pericles urged that Aegina, "the eyesore of the Piraeus," be removed. And Moerocles said he was no more wicked than a certain respectable man he named — for that man was wicked at one-third interest, but he himself only at one-tenth. And the iambic line of Anaxandrides about daughters delaying their marriage: "My girls are overdue in their marriages." And Polyeuctus' remark about a certain paralyzed man, Speusippus, that he could not keep still, bound as he was by fortune in a disease with five holes.

29| And Cephisodotus called the triremes "dappled mills," and the Cynic called the taverns of Attica "the common messes." And Aesion said that they had poured the city out into Sicily — for this is a metaphor and puts the matter before the eyes. And "so that Greece cried out" — this too is in a way a metaphor and before the eyes. And as Cephisodotus urged, one should be careful not to let them make the gatherings of the assembly too frequent. And Isocrates likewise spoke of those who run together at the festival gatherings. And as in the Funeral Oration, that it was fitting for Greece to cut her hair at the tomb of those who died at Salamis, since her freedom was being buried along with their courage — for if he had said it was fitting to weep because their courage was being buried, that would be metaphor and before the eyes, but "freedom along with their courage" carries a certain antithesis as well. And as Iphicrates said, "for my path of words runs through the middle of what Chares has done" — a metaphor by analogy, and "through the middle" puts it before the eyes. And to say that he called on the dangers to come to the help of the dangers is before the eyes and a metaphor. And Lycoleon, on behalf of Chabrias: "not even respecting his suppliant branch" — meaning his bronze statue; for this is a metaphor in the present circumstance, though not always so, but it is before the eyes.

30| For while he is in danger, the statue is imagined to be supplicating — the lifeless thing made alive, a memorial of the city's deeds. And "practicing in every way to think small thoughts" — for practicing is a kind of increasing something. And that god kindled the intellect as a light in the soul — for both make something clear. For we do not dissolve wars but postpone them; for both are things belonging to the future, both the postponement and this kind of peace. And to call the treaty a trophy far finer than those won in wars: for the latter are for small things and a single stroke of fortune, but these are for the whole war; for both are signs of victory. And that cities render great accounts to men's blame; for the rendering of accounts is a kind of just injury.

31| [3.11.1] It has been said, then, that clever sayings come from metaphor by analogy and from putting things before the eyes; but we must now say what we mean by "before the eyes," and what produces this effect. [3.11.2] I mean that things are put before the eyes which signify things in actuality, as for example calling a good man "four-square" is a metaphor, since both are complete, but it does not signify actuality; whereas "having his prime in full bloom" is actuality, and "you, like an animal let loose and free" is actuality, and "thereupon the Greeks, darting forward on their feet" (Euripides) — "darting" is actuality and metaphor, for it conveys speed. [3.11.3] And so Homer often uses metaphor to make lifeless things alive. In all these cases it is the making of actuality that wins esteem, as in these lines: "again down onto the plain the shameless stone came rolling" (Odyssey), and "the arrow flew" (Iliad), and "eager to fly upon him" (Iliad), and "they stood fixed in the ground, longing to glut themselves on flesh" (Iliad), and "the spearpoint rushed raging through his chest" (Iliad). In all of these, because the things are made alive, they appear to be acting; for being shameless, and raging, and the rest, are all actuality. He attached these by a metaphor according to analogy: for as the stone is to Sisyphus, so the shameless thing is to the one it shames. [3.11.4] The same effect occurs in the esteemed images concerning lifeless things: "curved, crested with foam, some before, others behind" (Iliad) — for it makes everything move and live, and actuality is motion.

32| [3.11.5] One must draw metaphors, as has been said before, from things that are akin but not obvious — just as in philosophy too it takes a good eye to observe the similar even in things far apart, as when Archytas said that an arbitrator and an altar are the same thing, since to both the wronged party flees for refuge. Or as if one were to say that an anchor and a hanging pot-hook are the same thing, since both are in a way the same, but differ in being above and below. And to say that cities have been "leveled" is the same kind of thing applied to things far apart, meaning equality in reputation and in power. [3.11.6] Most clever sayings, too, come through metaphor and through an added deception: for it becomes clearer what one has learned by the outcome running contrary to expectation, and the soul seems to say, "how true, and I was wrong." And among apophthegms, the clever ones come from not saying what the words state, as in Stesichorus' saying, that the cicadas will sing to themselves from the ground. And well-made riddles are pleasant for the same reason, since there is learning and metaphor in them, and so too, as Theodorus puts it, is "saying new things." This happens when the saying is paradoxical, and not, as he puts it, measured against the previous opinion, but rather as in comic sayings, words twisted out of their proper form — which puns on a single letter can also do, since they deceive — and likewise in verse, since it does not turn out as the hearer supposed: "he walked with chilblains under his feet" — and the hearer thought he would say "sandals." This must become clear the very moment it is spoken. Puns on a letter make the hearer take, not what the phrase states, but what the word, twisted around, means, as in Theodorus' quip against Nicon the harp-singer, "it Thrax-es you" — for it pretends to say "you are a Thracian" and so deceives, since it means something else. Hence it is pleasant once understood, since if one does not grasp that it means "Thracian," it will not seem clever.

33| [3.11.7] and "do you want to destroy it." Both must be said fittingly. The same is true of witty remarks — for example, saying that for the Athenians their rule of the sea was not the beginning of their troubles, for they benefited by it; or as Isocrates said that empire was for the city the beginning of its troubles. In both cases, what one would not have expected to be said is said, and it is recognized as true: for to say that the beginning is a beginning is nothing clever, but he is not speaking in that sense, but in another; and "beginning" he denies not in the sense he first used it, but in another sense. [3.11.8] In all such cases, when the word is applied fittingly by homonymy or metaphor, then it is well done. For example, "Anaschetos is not bearable" denies by homonymy, but fittingly, if he is in fact unpleasant. And "you could not become more of a stranger than you ought" — for "a stranger, no more than you ought" is the same as saying that the stranger ought not always to be a stranger, since that too is a kind of estrangement. The same holds for the celebrated line of Anaxandrides: "It is fine to die before doing a deed worthy of death," for this is the same as saying "it is worth dying, though not being worthy to die," or "it is worth dying, not being worthy of death," or "not doing deeds worthy of death."

34| [3.11.9] The form of expression in these cases, then, is the same, but the more briefly and antithetically it is stated, the more it is admired. The reason is that learning happens more readily because of the antithesis, and more quickly because it is stated briefly. [3.11.10] What is said must always have present in it either truth relative to the person spoken of, or correctness in the saying, if what is said is to be true and not superficial. For it is possible to have these separately — for example, "one who has done no wrong ought not to die," but this is not witty; "the worthy man ought to marry the worthy woman," but this is not witty either — unless it has both at once: "it is worth dying, though not being worthy of death." The more features it has, the wittier it appears — for example, if the words are also metaphorical, and metaphorical in a particular way, and there is antithesis and balanced clauses, and it has vividness.

35| [3.11.11] Similes too, as has been said above, when they are admired, are in a sense metaphors, for they are always stated from two terms, like the metaphor by analogy — for example, we say the shield is the bowl of Ares, and the bow is a stringless lyre. Said this way the expression is not simple, but calling the bow a lyre, or the shield a bowl, is simple. [3.11.12] And people make comparisons this way too — for example, comparing a flute-player to a monkey, or a sputtering lamp to a man screwing up his eyes; for both draw the two things together. [3.11.13] It is well done when there is a metaphor: for it is possible to compare the shield to the bowl of Ares, and a ruin to the rag of a house, and to call Niceratus a Philoctetes bitten by Pratys, as Thrasymachus compared him when he saw Niceratus, beaten by Pratys in a rhapsode's contest, still unkempt and long-haired. It is in this kind of thing that poets fail most badly if they do it poorly, and win most praise if they do it well — I mean when the terms correspond, as in "his legs are as curly as celery," or "like Philammon fighting the punching-bag." All such expressions are similes. And that similes are metaphors has been said many times. [3.11.14] Proverbs, too, are metaphors from one species to another: for example, if someone brings something in expecting to benefit from it himself, and is then harmed by it, as they say the man of Carpathus was by the hare — for both have suffered what is described in the saying. The source, then, from which witty sayings come, and the reason why, has now pretty much been stated. [3.11.15] Admired hyperboles, too, are metaphors — for example, of a man with a black eye: "you would have thought him a basket of mulberries," for a bruise under the eye is something red, but this is red to an extreme degree. "Like" and "as" constructions are hyperbole differing only in wording: "like Philammon fighting the punching-bag" becomes "you would have thought him Philammon himself, fighting the bag"; "his legs as curly as celery" becomes "you would have thought he had not legs but celery, they were so curly."

36| [3.11.16] Hyperboles are juvenile, for they display vehemence. That is why people use them most when angry: "Not even if he gave me as many gifts as the sand and dust — I will not marry the daughter of Agamemnon, son of Atreus, not even if she rivaled golden Aphrodite in beauty, and Athena in her works." That is why it is unbecoming for an older man to speak this way; and it is the Attic orators who use this device most. [3.12.1] It must not escape notice that a different style is fitting to each genre; for written style and the style of oral contest are not the same, nor are the style of political oratory and that of forensic oratory. One must know both: the one is knowing how to write correct Greek, the other is not being forced into silence when one wishes to communicate something to others — which is what happens to those who do not know how to write. [3.12.2] Written style is the most precise, while the style of oral contest is the most suited to delivery (and of this there are two kinds: one expressive of character, the other of emotion). That is why actors pursue plays of this kind, and poets pursue actors of this kind. But poets meant to be read are carried about in people's hands — for example Chaeremon, who is as precise as a speechwriter, and Licymnius among the dithyrambic poets. And when compared, the speeches of writers appear cramped in actual contests, while those of orators, however well delivered, look amateurish in the hand.

37| The reason is that what suits delivery is fitting in an actual contest; so when delivery is stripped away and such speeches no longer do their proper work, they appear simple-minded — for example, asyndeton and repeating the same thing often are rightly condemned in written style, but not in the style of contest, and orators do use them, for this belongs to delivery. [3.12.3] It is necessary to vary one's manner while saying the same thing, which in a sense paves the way for delivery: "This is the man who robbed you, this is the man who deceived you, this is the man who in the end attempted to betray you," as the actor Philemon used to do in Anaxandrides' Old Men's Battle, whenever he spoke the names Rhadamanthys and Palamedes, and as he did with the word "I" in the prologue of The Pious. For if one does not deliver such passages with proper acting, the effect is like the man carrying the beam.

38| [3.12.4] The same is true of asyndeton: "I came, I met him, I begged him" — for this must be acted out, not said as a single unit in the same manner and tone. Furthermore, asyndeton has its own particular effect: in an equal amount of time, many things seem to have been said; for the connective makes the many into one, so that if it is removed, clearly the opposite will happen — the one will become many. So it has an amplifying effect: "I came, I conversed, I entreated" (it seems like many things) — "he disregarded everything I said." This is also what Homer wants to achieve in the lines "Nireus, in turn, from Syme, Nireus son of Aglaia, Nireus who was the most handsome" — for about someone spoken of at length, it follows that he is mentioned often; so if he is mentioned often, it also seems that much is said about him, so that Homer has amplified him — though he mentions him only once — through this fallacious reasoning, and has made him memorable, even though he never speaks of him again afterward. [3.12.5] Political style, then, is altogether like scene-painting: for the larger the crowd, the more distant the viewing, so that in both cases precise detail is superfluous and appears worse. Forensic style is more precise; and even more so is the style before a single judge, for it contains the least rhetorical artifice — what is relevant to the case and what is irrelevant to it are more easily surveyed together, and the element of contest is absent, so that the judgment is clear-headed.

39| That is why the same orators do not succeed in all these settings: wherever delivery matters most, there precision matters least. This is wherever the voice matters, and especially wherever a loud voice matters. Epideictic style, then, is the most suited to writing, for its proper function is to be read; [3.12.6] forensic style is second. To further divide style by saying it must be pleasant and magnificent is superfluous; for why these qualities any more than temperate and liberal, or any other excellence of character? Being pleasant will clearly result from the qualities already stated, if the excellence of style has been correctly defined. For why, after all, must style be clear, and not flat, but fitting? If it rambles, it is not clear, nor if it is too terse; clearly the mean fits. And what has been described will make style pleasant, if the familiar and the unfamiliar are well blended, together with rhythm, and the persuasiveness that comes from fittingness.

40| [3.13.1] We have now discussed style, both what is common to all kinds of speeches and what belongs to each kind separately; it remains to speak about arrangement. A speech has two parts: for it is necessary to state the subject one is speaking about, and to demonstrate it. Hence it is impossible either to state the case without demonstrating it, or to demonstrate it without first stating it: for one who is demonstrating something demonstrates some thing, and one who states something beforehand states it for the sake of demonstrating it. [3.13.2] Of these parts, the one is the statement of the case, the other is the proof, as if one were to divide it into the problem and the demonstration of it. [3.13.3] But as things stand people divide speeches ridiculously: narration, surely, belongs only to the judicial speech; how can there be narration of the kind they describe in an epideictic or deliberative speech, or arguments directed against an opponent, or an epilogue to speeches that involve demonstration? An opening, a comparison, and a recapitulation occur in deliberative speeches only when there is a dispute between opposing sides. For accusation and defense often involve these, but deliberative advice does not; and the epilogue does not even belong to every judicial speech—for instance, when the speech is short or the matter easy to remember; for it turns out that length is what is being cut away. [3.13.4] The necessary parts, then, are the statement of the case and the proof. These are the parts proper to every speech; but most speeches have an opening, a statement of the case, a proof, and an epilogue: for the arguments against the opponent belong among the proofs, and the comparison is an amplification of one's own case, so that it too is a part of the proofs (for one who does this is demonstrating something), whereas the opening is not, nor is the epilogue—rather, it reminds the hearer. [3.13.5] If, then, one divides things in this way—as the followers of Theodorus used to do—narration will become distinct from further-narration and pre-narration, and refutation distinct from further-refutation. But one must give a name only when speaking of some distinct form with a real differentia; otherwise the result is empty and nonsensical, as Licymnius does in his handbook, when he names things "impulsion," "digression," and "ramifications."

41| [3.14.1] The opening, then, is the beginning of a speech, corresponding to the prologue in poetry and the prelude in flute-playing: all of these are beginnings, and as it were a paving of the way for what follows. The flute-prelude is like the opening of epideictic speeches: for flute-players, whatever passage they happen to play well, they play as a prelude and then join it to the keynote; and in epideictic speeches one should write in the same way—saying whatever one wishes at once, then striking the keynote and joining it on, as everyone does. An example is the opening of Isocrates' Helen: there is nothing in common between the eristics and Helen; and yet even if he wanders off the point, it fits, and it is better that the whole speech not be uniform in kind. [3.14.2] The openings of epideictic speeches are said to be drawn from praise or blame (for instance Gorgias, in his Olympic speech, says: "Men of Greece, you deserve to be admired by many"—for he is praising those who organize the panhellenic festivals; while Isocrates finds fault, because they honored bodily excellences with gifts but made no prize for those of good sense), [3.14.3] and from advice (for instance, that one ought to honor good men, and so he himself praises Aristides; or those of the sort who are neither well-regarded nor base, but who, though good, go unnoticed—like Alexander the son of Priam; for this man gives advice); [3.14.4] and further, from the openings used in judicial speeches: this is drawn from what is directed at the hearer, when the speech concerns something paradoxical, or difficult, or something already talked to death by many, so as to win indulgence—for instance Choerilus: "But now, when everything has been divided up." The openings of epideictic speeches, then, are drawn from these sources: from praise, from blame, from exhortation, from dissuasion, from what is directed at the hearer; and the keynotes must be either foreign to the speech or proper to it.

42| [3.14.5] One must grasp that the openings of judicial speeches have the same function as the prologues of dramas and the proems of epic poems: for the openings of dithyrambs are like those of epideictic speeches—"because of you, and your gifts, or your spoils." [3.14.6] But in dramatic prologues and in epic proems there is a sample of the discourse, so that the audience may know beforehand what the speech is about and the mind not be left hanging; for what is indefinite leads one astray. So one who gives the beginning, as it were, puts it into the hearer's hand, so that by holding on to it he can follow the speech. That is why we have: "Sing, goddess, the wrath" (Hom. Il. 1.1); "Tell me, Muse, of the man" (Hom. Od. 1.1); "Lead me to another tale, how out of the land of Asia a great war came into Europe." And the tragedians too make clear what the drama is about, even if not immediately, as Euripides does, in the prologue—still, somewhere—as Sophocles too does: "Polybus was my father" (Soph. OT 774). And comedy does the same. The most necessary function of the opening, then, and the one proper to it, is this: to make clear what the end is for the sake of which the speech is being given (and for this reason, if the matter is already clear and small, one should not use an opening at all). [3.14.7] The other kinds people employ are remedies for defects, and are common to the other parts of the speech as well.

43| These are drawn from the speaker, from the hearer, from the matter, and from the opposing side. Those concerning oneself and one's opponent have to do with the sort of thing that either dissolves a slander or creates one (and this is not done the same way in both kinds of speech: for the defendant should put what answers the slander first, while the prosecutor should do it in the epilogue; and the reason is not obscure—for the defendant, when he is about to bring himself forward into the jury's favor, must first remove whatever stands in his way, and so he must dissolve the slander first; but the one bringing the slander must do the slandering in the epilogue, so that the jury will remember it better). What concerns the hearer is drawn from making him well-disposed and from making him angry, and sometimes from making him attentive, or the opposite; for it is not always advantageous to make him attentive, which is why many speakers try to move him to laughter instead. Everything can be brought round to promoting ease of understanding, if one wishes, and to seeming reasonable oneself; for people pay more attention to speakers of that kind. Hearers are attentive to what is important, to what concerns them personally, to what is astonishing, to what is pleasant; hence one must instill the impression that the speech concerns things of that sort. And if one wants to make them inattentive, one should suggest that the matter is trivial, that it is nothing to do with them, or that it is painful.

44| [3.14.8] One should not fail to notice that all such devices lie outside the speech proper: they are addressed to a poor sort of hearer, one who listens even to what is beside the matter; for if the hearer is not of that sort, there is no need of an opening at all, except to state the matter briefly in summary, so that the speech may have a head, as it were, the way a body does. [3.14.9] Further, making the audience attentive is a task common to all the parts, wherever it is needed; for people's attention flags everywhere more than at the beginning—hence it is absurd to place this device at the beginning, the very point where everyone listens with the greatest attention. So wherever the right moment is, one should say: "and give me your attention"; "for this concerns me no more than it does you"; and "for I shall tell you a thing as terrible, or as astonishing, as you have never yet heard." This is what Prodicus called, whenever the audience began to nod off, throwing in a bit of his fifty-drachma lecture for them. [3.14.10] That the appeal is directed at the hearer not in his role as a mere hearer of the case is clear: for everyone, in their openings, either slanders the other side or dissolves fears about themselves. "My lord, I shall not speak with any eagerness—" (Soph. Ant. 223); "Why do you preface your speech this way?" (Eur. IT 1162); and speakers who have a bad case, or seem to, do the same—for it is always better to dwell on anything rather than the matter itself, which is why slaves, when questioned, do not answer what they are asked but talk all around it, and give preliminary openings of their own. [3.14.11] Where good will is to be won from has already been discussed, and so has each of the other such points. And since it is well said, "grant that I may come dear to the Phaeacians and win their pity" (Hom. Od. 7.327), one should aim at these two things. In epideictic speeches, one should think of making the hearer share in the praise—either himself, or his family, or his pursuits, or in some other way; for what Socrates says in the Funeral Oration is true, that it is not hard to praise Athenians before an Athenian audience, but it is hard to do so before a Spartan one.

45| [3.14.12] The parts of a deliberative speech are drawn from those of a forensic speech, but by nature it has the least need of them; for the subject is one the audience already knows, and the matter needs no introduction except for the speaker's own sake or that of his opponents, or if the audience takes the matter to be of a different size than you wish — either greater or less — so that it becomes necessary either to disparage or to clear away that impression, and either to magnify or to minimize. It is for these purposes that an introduction is needed, or else merely for ornament, since a speech seems improvised if it lacks one. Such is Gorgias's encomium to the Eleans: without any preliminary sparring or limbering-up he begins straight off, "Elis, a happy city." [3.15.1] On the subject of disparagement: one topic is to argue from whatever will clear away an unfavorable impression — for it makes no difference whether someone has actually voiced it or not, so this holds generally. [3.15.2] Another topic is to meet the disputed points head-on, saying either that the thing did not happen, or that it was not harmful, or not harmful to this person, or not as serious as claimed, or not unjust, or not great, or not shameful, or not of any magnitude — for disputes turn on these very points, as Iphicrates did against Nausicrates: he admitted doing what was alleged and causing harm, but denied acting unjustly. Or one may trade off: granting the wrongdoing but claiming it was fine, granting the harm but claiming it was beneficial, granting the pain but claiming it was useful, or some other such exchange. [3.15.3] Another topic is to say the act was an error, or a misfortune, or something necessary — as Sophocles said he trembled not, as his accuser claimed, in order to seem an old man, but of necessity, since it was not by his own choice that he had reached eighty years. And one can trade off the purpose: that he did not intend to cause harm but aimed at this other thing, and did not do what he was accused of doing, but harm happened to result; and it is just to resent someone only if he acted in order to bring the harm about.

46| [3.15.4] Another topic: whether the accuser himself is implicated, now or before, either himself or someone close to him. [3.15.5] Another: whether others are implicated whom people agree are not liable to the same charge — for instance, if it is argued that so-and-so is an adulterer because he is well-groomed, then so-and-so would be one too. [3.15.6] Another: whether he has accused others, either someone else or the speaker himself, or whether people were suspected without any accusation being made, just as the speaker now is, and have turned out not to be liable. [3.15.7] Another topic proceeds from turning the accusation back on the accuser: for it is absurd that the words of a man who is himself untrustworthy should be trusted. [3.15.8] Another: whether a judgment has already been rendered, as Euripides did against Hygiaenon in the property-exchange suit, when Hygiaenon accused him of impiety for having written, urging men to perjury, "the tongue has sworn, but the mind is unsworn" (Eur. Hipp. 612); Euripides replied that his opponent was acting unjustly in bringing before the courts a case already judged in the dramatic competition, since he had already given an account of it there, or would give one there, if his opponent wished to bring the charge. [3.15.9] Another topic proceeds from accusing the disparagement itself, showing how serious a matter it is, and that it creates further judgments beyond the case, and shows that the accuser does not trust the facts of the matter itself. A topic common to both sides is to cite corroborating signs — as in the Teucer, Odysseus argues that Teucer is a relative of Priam, since Hesione was Priam's sister,

47| while Teucer replies that his father Telamon was an enemy of Priam, and that he did not betray the spies. [3.15.10] Another topic available to the one disparaging: to praise something small at length and censure something great briefly, or to set out many good points and then bring forward, briefly, one bad one that bears on the matter. Such speakers are the most skillful and the most unscrupulous, for they try to do harm by way of the good points, mixing them in with the bad. Common to both the accuser and the defendant, since the same act can be done for more than one reason, is this: the accuser must take the malicious view, interpreting it for the worse, while the defendant must take the better view — for instance, that Diomedes chose Odysseus as his companion: the one side says it was because he judged Odysseus the best man, the other that it was not for that reason but simply because Odysseus was the only one who would not compete with him for glory, as being unworthy of rivalry.

48| [3.16.1] Let this much be said about disparagement. As for narrative: in epideictic speeches it does not proceed continuously but part by part. For one must go through the actions from which the speech is composed, since the speech is made up partly of the inartistic element (for the speaker is not the cause of the actions) and partly of what comes from art — and this consists either in showing that the action occurred, if it is not credible, or in showing its quality, or its magnitude, or all of these together. [3.16.2] For this reason it is sometimes better not to narrate everything continuously, because such a demonstration is hard to keep track of: from these facts he is shown brave, from those wise or just. A speech of this kind is also simpler, while the continuous kind is elaborate and not plain. [3.16.3] One should remind the audience of facts that are already well known — which is why most speeches need no narrative at all: for instance, if you wish to praise Achilles, everyone already knows his deeds, and you need only make use of them, not narrate them. But if you wish to praise Critias, narrative is needed, since not many people know his deeds. [3.16.4] As it is, people absurdly say that the narrative must be quick. Yet it is like the answer given to the baker who was asked whether he should make the bread hard or soft: "Well," someone said, "can't you make it well?" — and the same applies here. One should not narrate at length, just as one should not make a long introduction or set out one's proofs at length; for what matters here is not doing it well by being quick or brief, but by hitting the right measure — that is, saying only as much as will make the matter clear, or as much as will lead the audience to suppose that the event occurred, or that harm was done, or that injustice was committed, or whatever degree of these you wish to establish — and, for the opposite purpose, the opposite.

49| [3.16.5] One should narrate alongside the facts whatever bears on your own good character — for example, "and I kept admonishing him, always urging what was right, not to abandon his children" — or on the other party's bad character: "but he answered me that wherever he might be, there would be other children for him," which is the answer Herodotus reports the rebelling Egyptians gave; or whatever will be pleasing to the jurors. [3.16.6] For the defendant the narrative should be shorter, since the points in dispute are either that the act did not happen, or that it was not harmful, or not unjust, or not as serious as claimed — so one should not linger over what both sides agree on, unless it bears on one of those points, as when, for example, the act is admitted to have been done, but denied to be unjust. [3.16.7] Further, one should mention as accomplished only those things whose being described as still in progress would arouse pity or a sense of outrage; an example is the account Odysseus gives to Penelope in Alcinous's hall, compressed into sixty lines, and how Phayllus handled the epic cycle, and the prologue to the Oeneus. [3.16.8] The narrative should also reveal character; and this will happen if we know what produces character. One way is by making one's choice clear: character is of a certain quality according to what the choice is, and the choice is of a certain quality according to its end. This is why mathematical discourses reveal no character, because they involve no choice either — they have no "for the sake of which" — whereas Socratic dialogues do,

50| since they deal with matters of that kind. [3.16.9] Other things are revelatory of character insofar as they accompany each type of character — for instance, that a man kept talking as he walked along, which shows boldness and boorishness of character. And one should not speak as if from calculated thought, the way people do nowadays, but as if from deliberate choice: "I wished it, and I chose it deliberately; even though I gained nothing by it, it was still better so" — for the first way of speaking belongs to a prudent man, the second to a good man, since prudence lies in pursuing what is advantageous, goodness in pursuing what is fine. And if a claim seems incredible, then one should add the reason for it, as Sophocles does — an example is the passage from the Antigone, where Antigone says she cared more for her brother than for a husband or children, since a husband or children could be had again once the ones lost were gone, "but with my mother and father in Hades, there is no brother who could ever be born again" (Soph. Ant. 911-912). But if you have no reason to give, you should still say that you are not unaware you are stating something incredible, but that you are simply that way by nature — since people find it hard to believe that anyone would willingly do anything except what is to his own advantage.

51| [3.16.10] Further, when narrating, speak from the emotional side, adding the accompanying details, even ones the audience already knows, and the features peculiar either to yourself or to the other person: "and off he went, glaring at me." And as Aeschines said about Cratylus, that he "hissed," "shaking his hands violently" -- for these are persuasive, because such details become tokens of things people do not know, drawn from things they do know. Most touches of this kind can be got from Homer: "so she spoke, and the old woman covered her face with her hands" (Homer, Odyssey 19.361) -- for people who begin to weep clutch at their eyes. And bring yourself on stage at once as a person of a certain sort, so that the audience will view you that way, and likewise your opponent -- but do it without their noticing. That this is easy, see from how messengers do it: about things we know nothing of, we still form some impression. And in many places one must narrate, and sometimes not at the beginning. [3.16.11] In deliberative speech there is least room for narrative, because no one narrates about the future; but if there is to be narrative at all, let it concern things that have happened, so that, being reminded of those, the audience may deliberate better about what comes later -- either by way of blame or of praise; but doing that at that point is not the work of the counselor. If a thing is hard to believe, you must promise to prove it and state the reason at once, and arrange the points in the order the hearers want -- as Iocaste in Carcinus's Oedipus continually keeps making promises as the man searching for his son keeps asking, and so does Sophocles' Haemon. [3.17.1] Proofs must be demonstrative; and one must demonstrate, since the dispute is over one of four points, by bringing the demonstration to bear on the very point in dispute -- for instance, if it is disputed that the act took place, then at the trial the demonstration must above all be brought to bear on this; if that it did no harm, on this; and likewise if the dispute is that it was not so great, or that it was done justly; the same holds too if the dispute concerns whether the thing happened at all.

52| [3.17.2] And it must not escape notice that only in this dispute is it necessary for one of the two parties to be wicked -- for ignorance is not the cause here, as it might be if people were disputing about what is just; so in this case one must dwell at length, but not in the others. [3.17.3] In epideictic speeches, for the most part, let amplification serve to show that the deeds are noble and beneficial; for the facts themselves must simply be believed -- rarely do speakers bring demonstrations even of these, only if they are hard to credit or if someone else is given the credit for them. [3.17.4] In deliberative speeches, one might dispute either that a thing will not happen, or, granting that it will happen if people do what is urged, that it is not just, or not beneficial, or not as important as claimed. One must also watch whether the speaker is lying about something outside the matter itself; for such lies appear as evidence that he is lying about the rest as well. [3.17.5] Examples suit deliberative oratory better, enthymemes forensic oratory better; for the one concerns the future, so that examples necessarily must be drawn from things that have happened, while the other concerns things that are or are not the case, where there is more room for demonstration and necessity -- for what has happened carries necessity with it. [3.17.6] One should not state enthymemes one after another in a row, but mix them in; otherwise they damage one another. For there is a limit even to quantity: "friend, since you have said as much as a sensible man would say" (Homer, Odyssey 4.204) -- but not more than that. [3.17.7] And do not hunt for enthymemes about everything; otherwise you will do what some who philosophize do, who draw syllogisms from premises more familiar and more credible than the conclusions they are arguing for.

53| [3.17.8] And whenever you are working on the emotions, do not also state an enthymeme (for either the enthymeme will drive out the emotion, or the enthymeme will have been stated in vain: simultaneous movements knock each other out, and either wipe one another away or weaken one another), nor, when the speech is meant to display character, should you look for an enthymeme at the same time; for demonstration carries neither character nor moral purpose. [3.17.9] Maxims should be used both in narration and in proof; for there they show character -- "and I gave it, and that though I knew one should not trust people"; and, spoken emotionally, "and I do not regret it, wronged though I was; for he has the profit, but I have justice." [3.17.10] Speaking in the assembly is harder than pleading in the courts, and reasonably so, because it concerns the future, whereas the latter concerns what has happened, which is already knowable even to seers, as Epimenides the Cretan said (for he did not prophesy about things to come, but about things that had already happened but were unclear); and in forensic speeches the law is the starting point, and once one has a starting point it is easier to find a demonstration. Nor does deliberative speaking allow for many delaying digressions -- attacking the opponent, or dwelling on oneself, or working on the emotions -- least of all any of these, unless one departs from the subject. So a speaker at a loss must do what the orators at Athens do, and Isocrates: for even while giving counsel he accuses -- as he accuses the Lacedaemonians in the Panegyricus, and Chares in the speech On the Alliance. [3.17.11] In epideictic speeches one must vary the speech with episodes of praise, as Isocrates does; for he is always bringing someone new onto the stage. And what Gorgias used to say -- that his speech never runs short -- comes to the same thing: for if he speaks of Achilles he praises Peleus, then Aeacus, then the god; and likewise he praises courage, describing the things it does or the sort of thing it is.

54| [3.17.12] So when you have demonstrations, you must speak in a way that shows character as well as in a demonstrative way; but if you have no enthymemes, speak in a way that shows character: and it suits the decent man better to appear virtuous than for his speech to be exact. [3.17.13] Among enthymemes, the refutative kind is better regarded than the demonstrative, because whatever produces a refutation makes it clearer that a real deduction has taken place: for opposites are recognized more clearly when set side by side. [3.17.14] Arguments against the opponent are not some separate kind of proof, but belong among the proofs -- the business of dissolving some points by objection and others by counter-syllogism. In both deliberative and forensic speaking, when you speak first you should state your own proofs first, and only afterward meet the opposing arguments, dissolving them and discrediting them in advance. But if the opposition is many-sided, take up the opposing points first -- as Callistratus did in the Messenian assembly: he first demolished what his opponents were going to say, and only then delivered his own case. [3.17.15] But when speaking second, one must speak first against the opposing speech, dissolving it and counter-arguing, especially if it was well received; for just as the soul will not receive a man already discredited, in the same way it will not receive a speech either, if the opponent seems to have spoken well. One must therefore make room in the hearer for the speech that is to come; and this will happen if you demolish the other side first.

55| Therefore, having fought against either all the opposing points, or the strongest, or the ones that went over well, or the ones easy to refute, one should in this way make one's own case credible. "To the goddesses I will be an ally first -- for I, Hera..." (Euripides, Trojan Women 969-971) -- in these lines the speaker seized first on the weakest point of the other side. [3.17.16] So much, then, for proofs. As for character: since to say certain things about oneself involves either envy, or long-windedness, or the risk of contradiction, and to say certain things about another person involves either abuse or boorishness, one should have someone else say it instead -- which is what Isocrates does in the Philip and in the Antidosis, and the way Archilochus handles blame: he makes the father speak about his daughter in the iambic poem, "nothing is beyond hope or beyond oath where money is concerned" (Archilochus, fragment 74), and makes Charon the carpenter speak in the iambic poem that begins "not for me the possessions of Gyges"; and just as Sophocles makes Haemon speak on Antigone's behalf to his father as though he were reporting what others say.

56| [3.17.17] One should also vary one's enthymemes and sometimes turn them into maxims — for example: "Sensible men, when they are prospering, ought to make reconciliations; for that way they would gain the most" — and put it enthymematically: "For if reconciliations ought to be made when they will be most advantageous and most profitable, then one ought to reconcile while prospering." [3.18.1] As to questioning, it is most opportune to employ it above all when the opponent has already said one of the two things needed, so that when the one further question is put to him the absurdity follows. For example, Pericles asked Lampon about the rite of initiation into the mysteries of the Savior Goddess, and when Lampon said that it was not possible for an uninitiated person to hear of it, Pericles asked whether he himself knew it; and when Lampon said yes, Pericles asked, "Then how, being uninitiated?" [3.18.2] Second, one should question when one point is evident and it is clear that the opponent will grant the other if asked: for having secured this, one should not go on to ask the premise that is evident, but should simply state the conclusion. For example Socrates, when Meletus denied that he believed in gods but said that he spoke of some divine something, asked whether spirits were not either children of gods or something divine; and when Meletus said yes, Socrates said, [3.18.3] "Is there, then, anyone who believes there are children of gods but does not believe there are gods?" Further, one should question when one is about to show that the opponent is either contradicting himself or saying something paradoxical. [3.18.4] Fourth, when it is not possible for him to escape except by answering like a sophist: for if he answers in that way — "it is so, and it is not so," or "in some respects yes, in some respects no," or "in one way yes, in one way no" — the audience grows restive, taking it that he is at a loss. Otherwise one should not attempt questioning; for if the opponent objects, one will seem to have been worsted. For it is not possible to ask many questions, because of the weakness of the hearer; that is also why one ought to compress one's enthymemes as much as possible.

57| [3.18.5] In answering, one should, in the case of ambiguous questions, draw distinctions by an extended statement and not briefly; but in the case of questions that seem to involve contradiction, one should give the solution at once, in the very answer, before the questioner asks the next question or draws his conclusion; for it is not hard to see beforehand where the argument tends. This point, and the solutions, are clear to us from the Topics. [3.18.6] And when a conclusion is being drawn, if the opponent puts the conclusion in the form of a question, one should state the reason for it. For example, Sophocles, when questioned by Peisander whether it had seemed good to him, as it had to the other special counselors, to establish the Four Hundred, said yes. "What then?" said Peisander. "Did these things not seem to you to be wicked?" "Yes," he said. "Then did you not do these wicked things?" "Yes," he said, "for there was nothing better to do." And so too the Spartan, when called to account for his ephorate and asked whether it seemed to him that the others had been justly put to death, said yes. "Then did you not vote the same measures as they did?" "I did," he said. "Would you not, then, justly be put to death yourself?" "Not at all," he said, "for they did these things after taking money, but I did not — I did them by judgment." For this reason one ought neither to ask a further question after the conclusion, nor to put the conclusion itself as a question, unless the truth is heavily on one's side. [3.18.7] As to jests, since they seem to have some use in contests, and Gorgias said one should destroy the opponents' seriousness with laughter and their laughter with seriousness, and he was right to say so — how many kinds of jests there are has been stated in the works on poetics, of which one kind befits a free man and another does not, so that each speaker may take the kind that befits himself. Irony is more suited to a free man than buffoonery is: for the ironist makes the joke for his own sake, while the buffoon makes it for another's.

58| [3.19.1] The epilogue is composed of four parts: disposing the hearer favorably toward oneself and unfavorably toward the opponent, amplifying and depreciating, bringing the hearer into the appropriate emotions, and recapitulation. For it is natural, after one has demonstrated that one is truthful oneself and that the opponent is false, that praising and blaming and putting the finishing stroke should follow in that order. One should aim at one of two things: either that one is good in relation to these particular matters, or good simply, while the opponent is bad in these particular matters, or bad simply. The topics from which one must construct this have been stated — the places from which one should establish men as good and as base. [3.19.2] The next task, once the facts have already been proved, is naturally to amplify or depreciate them: for the facts must be agreed upon, if one is going to speak of their magnitude, just as the growth of bodies proceeds from things already existing. The topics from which one should amplify and depreciate have been set out earlier. [3.19.3] After this, when it is clear both what the facts are and how great they are, one should lead the hearer into the emotions. These are pity, indignation, anger, hatred, envy, rivalry, and contentiousness. The topics for these too have been stated earlier, [3.19.4] so that what remains is to recall to mind what has already been said.

59| And this should be done not in the way people do it in their introductions, where they do it wrongly. For there, to make the case easy to follow, they urge that one repeat things often. In an introduction, then, one should state the matter, so that the point under judgment does not go unnoticed; here, in the epilogue, one should state the things through which it has been proved, summarily. [3.19.5] The starting point is that one has delivered what one promised, so that one must say both what one has proved and why. This is stated by way of comparison with the opponent's case. One may compare either what both parties said about the same point, or not point for point ("but he said this about that, while I said this, and for this reason"), or ironically ("for he said this, while I said that — and what would he have done, if he had shown this, rather than that?"), or by way of a question ("what then has been proved? or what did he prove?"). One may proceed either in this way, by comparison, or in the natural order as it was stated, taking up one's own points in that order, and then, if one wishes, separately taking up those of the opponent's speech. [3.19.6] As for the ending of the style, the asyndetic manner is fitting, so that it may be an epilogue and not a speech: "I have spoken. You have heard. You have the facts. Judge."

← All of Aristotle: The Complete Works of Aristotle