On Literary Composition

1. This gift too, dear son, I give to you—just as Helen in Homer, entertaining Telemachus, says she does—on the day you keep this first birthday since you came to a man's estate, to me the sweetest and most honoured of festivals. Only it is not a piece of my own handiwork that I send you, as she says she gave the youth the robe, nor a thing suited only to the season of marriage and a fitting favour for a wife; rather it is the offspring and creation of my own learning and soul, a possession that is at the same time an instrument useful for all the needs of life that arise through speech—the most indispensable of all possessions, if I for my part happen to be thinking rightly. It is useful alike to all who practise civic oratory, whatever their age or condition, but most of all to young men like yourselves who are just laying hold of the subject—you, Rufus Metilius, son of a good father and to me the most honoured of friends.

For since there are two kinds of training applying to virtually all discourse, one concerned with thoughts and one concerned with words—of which the former would seem to touch more on the matter, the latter on the expression—and since all who aim at speaking well are equally diligent about both these studies of discourse, the knowledge that leads us to the subject matter and to good judgement about it is slow and hard for the young, or rather impossible to fall to the lot of beardless boys and youths; it is more proper to an understanding already in its prime and seasoned by grey hairs, since it grows together with much experience of words and deeds, and much trial and suffering of misfortunes both one's own and others'. But the love of beauty that concerns words naturally blossoms together with youthful years. For the whole soul of a young man is in a flutter about the embellishment of expression, taking up irrational and, as it were, inspired impulses toward it. Such impulses need a great deal of intelligent guidance and direction at the outset, if the young are not to say "every word that comes upon an untimely tongue," nor to put together at random whatever falls in their way, but to use a selection of words at once pure and noble, and to adorn them with an arrangement that has the sweet mingled with the dignified. For this part, then, which the young must practise first, I contribute to you a song to win your love: the treatise on the arrangement of words—a subject that has occurred to few, only to those of the ancients who composed handbooks of rhetoric or dialectic, but which up to the present has been worked out by no one accurately or sufficiently, as I am persuaded. And if I find leisure, I shall publish for you another work on the selection of words as well, so that you may have the whole department of expression completely worked out. That treatise, then, look for again next year at this same season, if the gods keep us unharmed and free from sickness—if indeed it has been firmly fated for us ever to attain it. For now receive the treatise which the divine power has brought to my mind.

The chief points of it, which I propose to set forth, are these: what the nature of arrangement is and what power it has; what objects it aims at and how it attains them; what its most general classes are and what is the character of each, and which of them I am persuaded is the best; and, besides these, what exactly that poetic, eloquent, and honey-sweet quality is that falls upon the ear—the quality that naturally accompanies the arrangement of prose expression when artful construction imitates the unstudied utterance of ordinary speech and succeeds excellently in the imitation—wherein its power lies, and by what practice each of these qualities might be produced. Such, roughly outlined, are the matters I intend to discuss. And here the treatise begins.

2. Arrangement is, as the very name shows, a certain placing side by side of the parts of speech, which some also call the elements of expression. These Theodectes and Aristotle and the philosophers of that age carried up to three, making nouns, verbs, and conjunctions the primary parts of expression. Those who came after them, and especially the heads of the Stoic school, advanced them to four, separating the articles from the conjunctions. Then those of the next generation, distinguishing the appellatives from the proper nouns, declared the primary parts to be five. Others, detaching the pronouns from the nouns, made this a sixth element. And others, separating the adverbs from the verbs, the prepositions from the conjunctions, and the participles from the appellatives—and still others introducing yet further divisions—made the primary parts of expression many; concerning which the discussion would be no small one. However that may be, the interweaving and juxtaposition of these primary parts—whether they be three or four or however many—produces what are called clauses; then the fitting together of these completes the so-called periods; and these periods bring the whole discourse to perfection. Now the functions of arrangement are these: to set the words suitably side by side, to assign to the clauses their proper harmony, and to divide the discourse well into periods.

Though it is the second division in order among the studies belonging to the department of expression—for the selection of words leads the way and naturally precedes it—arrangement nonetheless has in discourse a charm, a persuasiveness, and a power not a little greater than the other. And let no one think it strange that, while there are many great studies concerning selection, about which philosophers and statesmen alike have discoursed at length, arrangement, which holds second place in order and has by no means won an equal share of discussion, should have so great a force and capacity as to surpass and master all the works of selection. Let him reflect that in the other arts too—all that take various materials and from them produce a composite end, such as building, carpentry, embroidery, and the like—the powers of composition are second in order to the powers of selection but first in efficacy; so that if the same has happened in the case of discourse, it should not be thought absurd. And nothing prevents me from offering proofs of my proposition, lest I seem to be taking for granted matters that are open to dispute.

3. Now all expression by which we signify our thoughts is either in metre or without metre; and each of these, if it meets with a fine harmony, is able to make both the verse and the prose fine, but if it is flung out carelessly and at random, it ruins along with itself even the usefulness of the thought. Many poets, then, and prose writers, philosophers, and orators have selected with care words altogether fine and fitting to their subjects, yet, by assigning to them a harmony that was haphazard and unmusical, have reaped no good from that labour; while others, taking words contemptible and humble, but arranging them sweetly and admirably, have invested their discourse with much charm. And one might almost say that arrangement stands to selection in a relation analogous to that which words bear to thoughts. For just as there is no use in a good thought unless one gives it the adornment of fine wording, so here too it is of no profit to find pure and well-worded expression unless one also clothes it with the harmony proper to it.

But that I may not seem to be making an assertion without proof, I shall try to show by actual demonstration the grounds on which I was persuaded that arrangement is a finer and more perfect exercise than selection, bringing forward a few first-fruits both of verse and of prose. Let Homer be taken among the poets, and Herodotus among the prose writers; for from these it is enough to form a judgement about the rest. There is in Homer the passage where Odysseus, lodging with the swineherd, is about to take his early-morning meal, as was the custom among the ancients; and then Telemachus appears to them, returning from his journey to the Peloponnese—matters of everyday life, expressed supremely well. But where the excellence of the expression lies, the verses themselves, when set before you, will show:

The two within the hut, Odysseus and the noble swineherd, were making ready their morning meal at dawn, having kindled fire, and had sent out the herdsmen with the gathered swine. About Telemachus the loud-baying dogs fawned and did not bark as he drew near. Noble Odysseus marked the dogs fawning, and the sound of footsteps came to him; and at once he spoke to Eumaeus, who was close beside him: "Eumaeus, surely some companion of yours is coming here, or some other acquaintance, since the dogs are not barking but are fawning all about; and I hear the tread of feet." The word was not yet fully spoken when his own dear son stood in the doorway. The swineherd sprang up in amazement, and from his hands fell the vessels at which he was busied, mixing the gleaming wine. He ran to meet his master, and kissed his head and both his lovely eyes and both his hands; and a warm tear fell from him.

That this passage attracts and enchants the ears, and holds a place inferior to none even of the very sweetest poems, all, I am sure, would testify. But where does their persuasive charm reside, and why are they such as they are—is it because of the selection of words or because of the arrangement? No one would say it is because of the selection, as I am persuaded; for the whole passage is woven of the cheapest and humblest words, such as a farmer, a sailor, a craftsman, and anyone who takes no care to speak well might pick up ready to hand and use. Dissolve the metre, at any rate, and these same words will appear poor and unenviable; for there are in them no noble metaphors, no interchanges of meaning, no catachreses, no other figurative diction whatever, nor indeed any archaic words or any foreign or coined ones. What then remains but to credit the arrangement with the beauty of the expression? There are countless such passages in the poet, as I am sure all know; but for me, speaking only by way of reminder, it is enough to have cited these alone.

Come now, let us pass on to prose as well, and consider whether it too has experienced this same effect, so that from small and trivial matters and words well arranged great pleasures and graces arise. There is in Herodotus a certain king of the Lydians, whom he calls Candaules but says was called Myrsilus by the Greeks, who, being in love with his own wife, then thought fit to have one of his companions look upon the woman naked; and the man fought hard not to be compelled, but, since he could not persuade the king, submitted and gazed at her—a thing not only far from dignified or fit to be told beautifully, but childish and dangerous, and nearer the shameful than the noble; yet it has been told with great skill, and has come out better to hear when narrated than to see when done. But that no one may suppose it is the dialect that is the cause of the pleasure in the passage, I shall change its character into the Attic tongue, meddling with nothing else, and so present the dialogue.

"Gyges, since I do not think you believe me when I speak of the woman's beauty—for men's ears happen to be less trustworthy than their eyes—contrive that you may behold her naked." But he cried out and said: "Master, what unsound proposal is this you make, bidding me look upon my own mistress naked? When a woman puts off her tunic, she puts off her modesty along with it. Long ago men discovered what is honourable, and from these discoveries we must learn; and among them is this, that each should look upon what is his own. I am persuaded that she is the fairest of all women, and I beg you not to demand of me what is unlawful." So saying, he resisted . . . , but the king answered him thus: "Take courage, Gyges, and fear neither me—as though I spoke this word to test you—nor my wife, lest any harm to you come from her. For I will so contrive the matter from the beginning that she shall not even know she has been seen by you. I will set you behind the open door of the chamber where we sleep; and after I have gone in, my wife too will come to bed. Near the entrance stands a chair; on this she will lay her garments one by one as she takes them off, and you will be able to gaze at her quite at your ease. And when she goes from the chair to the bed and you are at her back, then let it be your care thereafter that she does not see you going out through the door." So he, since he could not escape, was ready.

Here too one could not say that the dignity and gravity of the words has made the diction comely; for they are unstudied and unselected, just such symbols as nature has set upon things; and indeed it would perhaps not have been fitting to use other and grander ones. For necessarily, when a thing is uttered in the most proper and most immediate words, the thoughts can be no grander than the words are. And that there is nothing dignified or extraordinary in them, anyone who wishes will discover by changing nothing but the harmony. There are many such passages also in this writer, from which one might infer that the persuasive charm of the expression lay not in the beauty of the words but in their conjoining. And on these matters this is enough.

4. But that one may perceive far more clearly how great a strength the power of composition has, both in verses and in prose, I shall take certain passages reputed to be excellent, and by changing their harmony I shall make both the metres and the prose appear quite different. Let these lines first be taken from Homer:

But she held them even, as a true working-woman holds the scales, who, taking the weight and the wool on either side, draws them up balancing them, that she may earn a sorry wage for her children.

This metre is the heroic, a perfect hexameter, marching by the dactylic foot. Now by altering the arrangement of these words I shall make the same lines, instead of hexameters, into tetrameters, and instead of heroic, into prosodiacs in this fashion:

But she held the scales even, as a true working-woman holds, who, taking up the wool on either side and the weight, draws, balancing, that she may earn for her children a sorry wage.

Such are the so-called Priapean lines, these:

Not profane, O ye initiates of the new Dionysus . . . And I too am come, made one of the mystic revel through his benefaction.

Again, taking other lines of Homer, and neither adding to them nor subtracting anything, but only changing the arrangement, I shall give them another kind, the one called the Ionic tetrameter:

So he lay stretched out in front of the horses and the chariot, bellowing, clutching at the bloody dust. So he lay stretched out in front of the horses and the chariot, clutching at the bloody dust, bellowing.

Such are these Sotadean verses:

There upon the topmost pyres the corpses lay, in a strange land, having left behind their orphaned walls, the walls of sacred Hellas and the inner hearth of their fathers, and lovely youth and the fair face of the sun.

I could go on to exhibit many other kinds of metre, and various ones, that fall into the heroic line, and to show that the same thing has happened to almost all the other metres and rhythms as well; so that, the selection of the words remaining the same and the arrangement alone being shifted, the metres are remoulded and there change along with them the figures, the colours, the characters, the passions, the whole dignity of the verses. But I should be forced to touch upon a greater number of theorems, some of which are familiar to very few. On many occasions, and not least on such as these, it would be fitting to bring in these words of Euripides:

Touch not, O my soul, on subtle reasonings; why think beyond measure? unless you mean to give yourself airs among your equals.

These matters, then, I think I shall let pass for the present. But that prose too can undergo the same experience as verse—the words remaining the same and the arrangement being changed—is open to anyone who wishes to observe. I shall take from the writing of Herodotus the opening of his history, since it too is familiar to most, changing only the character of the dialect. "Croesus was a Lydian by birth, son of Alyattes, and lord of the nations within the river Halys, which, flowing from the south between the Syrians and the Paphlagonians, issues toward the north wind into the sea called the Euxine." I transpose the harmony of this passage, and it will become for me no longer a leisurely, narrative composition, but rather upright and combative: "Croesus was the son of Alyattes, a Lydian by birth, lord of the nations within the river Halys, which, flowing from the south between the Syrians and the Paphlagonians, discharges into the sea called the Euxine toward the north wind." This character would seem not far removed from this passage of Thucydides: "Epidamnus is a city on the right as one sails into the Ionian gulf; bordering on it are barbarians, the Taulantians, an Illyrian people." Again, altering the same passage, I shall give it yet another form in this fashion: "Of Alyattes the son was Croesus, a Lydian by birth, of the nations within the river Halys the lord; which, flowing from the south between the Syrians and Paphlagonians toward the north, issues into the sea called the Euxine." This figure of arrangement is Hegesianic—paltry, ignoble, soft; for the priest of these trifles, that fellow, writes such things as: "Out of a good festival we keep another good one." "From Magnesia am I, the great Sipylene." "For Dionysus spat no small water into the drink of the Thebans; for it is sweet, but it makes men mad."

Enough of examples. For I think I have made sufficiently clear what I set out to show: that arrangement has greater strength than selection. And it seems to me that one would not err in likening it to Homer's Athena. For she made Odysseus, who was the same man, appear different at different times—now small and wrinkled and unsightly,

like to a wretched beggar and an old man,

and again, touching him once more with the same wand,

made him taller to look upon and stouter to behold, and down from his head she sent curling locks, like the hyacinth flower.

So arrangement too, taking the same words, makes the thoughts appear at one time shapeless, humble, and beggarly, and at another lofty, rich, and beautiful. And this was very nearly the chief point in which poet differs from poet, and orator from orator: the dexterous arrangement of words. For almost all the ancients made great study of it, and it is owing to this that their metres, their songs, and their prose are beautiful; the later writers, however, no longer did so, save a few; and in time it was wholly neglected, and no one thought it necessary or believed it contributed anything to the beauty of discourse. And so they left behind such compositions as no one endures to read through to the very end—I mean Phylarchus and Duris and Polybius and Psaon and Demetrius of Callatis and Hieronymus and Antigonus and Heraclides and Hegesianax and countless others; if I should care to recite the names of them all, the day's time would fail me. And why wonder at these, when even those who profess philosophy and publish handbooks of dialectic are so wretched at the arrangement of words that one is ashamed even to speak of it? It is enough to use Chrysippus the Stoic as evidence of my statement—for further than him they could not go—since no one has worked out the art of dialectic more precisely than he, nor produced discourses put together with worse harmony than his, at any rate among those thought worthy of name and reputation. And yet some of them pretended to take this branch seriously too, as being necessary to discourse, and even wrote certain treatises on the arrangement of the parts of speech; but they all wandered very far from the truth, and did not so much as dream what it is that makes arrangement sweet and beautiful. When I myself, then, resolved to compose this treatise, I sought whether anything had been said about it by my predecessors, and especially by the philosophers of the Stoa, knowing that those men gave no small thought to the department of expression; for one must bear true witness to them. But seeing that nothing whatever, great or small, had anywhere been said by any of those at least who are thought worthy of name, bearing on the inquiry I have chosen, and that the two treatises Chrysippus left bearing the title "On the Arrangement of the Parts of Speech" contain not a rhetorical but a dialectical study—as those know who have read the books—being concerned with the arrangement of propositions true and false, possible and impossible, admissible and changeable, ambiguous, and others of the same kind, and contributing neither benefit nor use to civic discourse, at any rate toward the pleasure and beauty of expression which arrangement must aim at—from that treatise I withdrew, and inquired on my own account whether I might find some natural starting point, since this seems to be the best beginning of every matter and every investigation. And after laying hold of certain theorems and supposing that the subject was proceeding methodically for me, when I learned that this road was leading me elsewhere than where I was going and where it was necessary to arrive, I withdrew. Yet perhaps nothing prevents me from touching on that study too and stating the reasons for which I gave it up, so that no one may think I passed it by through ignorance rather than by choice.

5. It seemed to me, then, that, following nature above all, we ought so to fit the parts of speech together as she wishes. For instance, I thought it right to set nouns before verbs—for the former denote substance, the latter accident, and substance is by nature prior to accidents—as these Homeric lines have it:

The man tell me, O Muse, of many turns, and

The wrath sing, goddess, and

The sun leapt up, leaving . . .

and the like; for in these the nouns lead the way and the verbs follow. The argument was plausible, but it did not seem to me true. For one could bring forward other examples in the same poet arranged in the opposite way to that in which these are arranged, yet no less beautiful and plausible. Such are these:

Hear me, child of aegis-bearing Zeus, Unwearied One, and

Tell me now, ye Muses who hold Olympian dwellings . . . Remember your father, Achilles, peer of the gods.

For in these the verbs lead the way and the nouns are subordinated; and no one would find fault with this construction as unpleasing. Furthermore, I thought it better to place verbs before adverbs, since by nature that which acts or is acted on is prior to its attendant circumstances—I mean manner, place, time, and the like, which we call adverbs—using these examples:

He smote about him on every side, and a hideous groaning arose . . . He fell backward, and gasped out his life . . . He sank down on the other side, and the cup fell from his hand.

For in all these the adverbs are set after the verbs. And this too is plausible, like the first, but not true, any more than that was. For these passages in the same poet are expressed in the opposite way:

In clusters they fly upon the spring flowers . . . This day shall a man into the light Eileithyia, goddess of childbirth, bring forth.

Have the verses, then, become any the worse because the verbs are subordinated to the adverbs? No one would say so. Again, I thought one ought not carelessly to observe this rule too: that things prior in time and in order be taken first. Such are these:

They first drew back the victims' heads, and slew them, and flayed them, and

The bow twanged, the string sang loud, and the arrow leapt, and

The ball then the queen threw to her attendant; she missed the attendant and cast it into the deep eddy.

"By Zeus," one might say, "yes—were there not also many other verses, not so arranged, yet no less beautiful than these."

He raised himself and struck with an oaken billet, which he had left when splitting wood—

for surely the raising of the arm is prior to the striking. And again,

He drew near and stood, and the axe cut through the tendons of the neck—

for surely it behoved him who was about to drive the axe into the bull's tendons first to stand near it. Furthermore, I thought it right to place nouns before epithets, appellatives before nouns, pronouns before appellatives, and in verbs to observe that the active forms lead the way before the passive, and the finite before the infinitive, and many other such rules. But all these experience shook and proved to be worth nothing. For at one time the arrangement turned out sweet and beautiful from these and the like rules, and at another time from rules that were not such but their opposites. For these reasons, then, I gave up that kind of study. And I have mentioned these rules even now not as worthy of serious attention; and I have cited the handbooks of dialectic not as necessary, but in order that no one, supposing them to hold something useful for the present inquiry, may think much of knowing them, lured by the titles of the works—titles which bear a certain resemblance—and by the reputation of those who composed them.

I return, then, to the position I started from before I digressed to these matters: that the ancients took great forethought for this department—poets, prose writers, philosophers, and orators alike—and thought it right neither to join word to word, nor clause to clause, nor period to period at random; but there was among them a certain art and there were theorems which they used to compose well. What these theorems were, I shall try to teach, so far as I am able, as far as it was in my power to discover them along with the rest—not stating all of them, but only the most necessary.

6. It seems to me that the science of composition has three tasks: one, to see what naturally fitted with what will take a fine and pleasing conjunction; another, to discern how each of the things to be fitted together with one another, if shaped in such and such a way, might make the harmony appear better; and the third, if any of the things taken needs alteration—I mean subtraction, addition, or change—to discern this too and to work it out suitably for the use intended. What each of these can effect I shall state more clearly by using illustrations from certain of the manual arts which all know—I mean building, shipbuilding, and the like. For the builder, when he has procured the material out of which he is to construct the house—stones and timbers and tiles and all the rest—then composes the work out of these, busying himself with these three things: what kind of stone, timber, or brick he must fit to what kind of stone, timber, or brick; next, how each of the things fitted together is to be set, and on what face; and third, if anything is ill-set, to knock it off and trim it and make this very thing well-set. And the shipbuilder busies himself with these same things. In like fashion, I say, those who are to arrange the parts of speech well must first consider what kind of noun or verb or other part, set together with what kind, will be suitably placed, and how this will be well or better placed—for surely not all words, set together with all, naturally affect the hearing alike. Next, they must discern how, if a noun or verb or any other part whatsoever be given such and such a shape, it will be more gracefully set and more becoming to the subject; I mean, in the case of nouns, whether taken in the singular or plural they will take a better conjunction, and whether brought out in the nominative case or in one of the oblique ones; and if certain nouns can naturally become feminine from masculine, or masculine from feminine, or neuter from these, how they might be better shaped, and all such matters. In the case of verbs, whether they will be better taken active or passive, and according to what moods, expressed (which some call verbal cases), they will get the best position, and what differences of tense they should display, and whatever else naturally attends upon verbs—and these same things must be observed in the other parts of speech also, that I may not speak of each one by one. Besides these, they must discern, of the things taken, whether any noun or verb needs alteration, how it might become more harmonious and better set. This element is more abundant in poetry, rarer in prose; yet it occurs in prose too, so far as is possible. For the speaker who says "into this contest [tutoni ton agona]" has added a letter to the pronoun with a view to the arrangement, since it would have been sufficient to say "into this contest"; and again, the one who says "perceiving Neoptolemus the actor [katidon]" has lengthened the word by a prefix, for "perceiving [idon]" would have done; and the writer who says "having come for the sake of no private enmity" has, by elisions, reduced the parts of speech and struck off some of the letters; and the one who says "he made [epoiese]" instead of "he made [epoiesen]" without the final n, and "I shall take away [aphairesomai]" instead of "I shall be deprived [aphairethesomai]," and all such things; and he who says "loved his place [echorophilese]" for "was fond of his place [ephilochorese]," and "shall have been loosed [lelysetai]" for "shall be loosed [lythesetai]," and the like, alters the expressions so that they may become for him fairer to fit together and more suitable.

7. This, then, is one study of the science of composition—that concerning the primary parts and elements of expression themselves. The other, as I said at the outset, concerns the so-called clauses, and requires a more varied and a greater treatment, about which I shall now try to say what I think. For these too must be fitted together so as to appear akin and friendly, and shaped as well as possible, and prepared beforehand, where needed, by reduction and amplification, and by whatever other alteration the clauses admit. Each of these things experience itself teaches. For often this clause, set before that one or after it, displays a certain euphony and dignity, but, taking another conjunction, appears graceless and undignified. What I mean will be clearer if one sees it in an example. There is in Thucydides a passage in the speech of the Plataeans, very gracefully composed and full of pathos, this: "And you, O Lacedaemonians, our only hope—we fear lest you should not stand firm." Come, then, let someone dissolve this conjunction and rearrange the clauses thus: "And you, O Lacedaemonians, we fear lest you should not stand firm, our only hope." Does the same grace, or the same pathos, still remain when the clauses are fitted together in this way? No one would say so. And what of this expression of Demosthenes: "He writes that to take the things granted, while admitting it to be lawful, is illegal"—if someone, dissolving and transposing the clauses, should bring it out in this fashion: "Admitting, then, that to take the things granted is lawful, he writes that to return thanks for them is illegal"—will it be equally forensic and rounded? I think not.

8. Such, then, is the study concerning the fitting together of clauses. But what is its character as regards their shaping? There is not one mode of bringing out all our thoughts; rather, we utter some as making assertions, some as inquiring, some as praying, some as commanding, some as in doubt, some as supposing, and others under some other figure; and in accordance with these we try to shape the expression too. There are doubtless many figures of expression, just as there are of thought, which it is not possible to comprehend under headings, and which are perhaps infinite—matters about which there is much to say and the study runs deep. The same clause, then, cannot produce the same effect when shaped this way or that. I shall speak by means of an example. If Demosthenes had brought out this expression in the following way—"Having said these things I drafted the decree; having drafted it I went on the embassy; having gone on the embassy I persuaded the Thebans"—would it then be composed as gracefully as it is now composed?—"I did not say these things and then fail to draft the decree; nor draft it and then fail to go on the embassy; nor go on the embassy and then fail to persuade the Thebans." It would be a long discourse for me if I should wish to speak of all the shapings that clauses admit. But for the sake of an introduction it is enough to have said so much.

9. Moreover, that some clauses also admit of alterations—at one time receiving additions not necessary as regards the sense, at another suffering subtractions that leave the thought incomplete, which poets and prose writers make for no other reason than that of harmony, that it may become sweet and beautiful—of this I think very little argument is needed. For who would not agree that this expression which Demosthenes has uttered is enlarged by an addition not necessary, for the sake of the harmony?—"For the man who does and contrives those things by which I might be caught, this man wars against me, even though he does not yet strike or shoot at me." For here "shoot at me" is added not for the sake of necessity, but in order that the last clause, "even though he does not yet strike," being harsher than is needful and not pleasant to hear, may by this addition become more graceful. And again, who would not say that that Platonic period which the man writes in the Funeral Oration has been pieced out with a filling not necessary to the expression?—"For when deeds have been nobly done, by a discourse nobly spoken there comes to the doers remembrance and honour from those who have heard it." For here "from those who have heard it" is said with no necessity, but in order that the last clause, "to the doers," may be made equal and matched with those before it. And what of that famous saying in Aeschines—"You summon him against yourself, you summon him against the laws, you summon him against the democracy"—a tricolon much admired: does it not partake of the same character? For what might have been comprehended in a single clause in this way—"you summon him against yourself and the laws and the democracy"—has been divided into three, the same expression having been set down repeatedly, not for the sake of necessity but to make the harmony sweeter. Such, then, is the manner of the addition that is made to clauses. And what of the subtraction? It occurs when something of what is necessary, if said, would offend and trouble the hearing, but, removed, makes the harmony more graceful. Such, in verse, are these lines of Sophocles:

I close my eyes and open them and rise, watching others rather than watched myself.

For here the second line is composed of two clauses, neither of them complete; for the expression would have been complete if brought out thus: "watching others more than being watched by others"; but the metre would have been wronged and would not have had the grace it now has. And in prose such things as this: "But that, while accusing certain men, to take away from all the exemption is the part of unjust men, I shall pass over." For here too each of the first two clauses has been reduced; they would have been complete if one had brought them out thus: "But that, while accusing certain men of not being fit to have the exemption, to take it away from all—even from those who justly obtained it—is the part of unjust men, I shall pass over." But Demosthenes did not think it right to take greater forethought for the precision of the clauses than for their good rhythm.

Let the same be said by me about the so-called periods also; for these too, both those that lead and those that follow, must be suitably fitted together, when it is fitting to bring out the discourse in periods. For surely the periodic style is not useful everywhere. And this very point is proper to the science of composition: when one ought to use periods, and to what extent, and when not.

10. These matters being settled by me, it would follow to state what the things are at which one who wishes to compose expression well must aim, and through what theorems one might attain what one wishes. It seems to me that there are two most general things at which those who compose verse and prose must aim: pleasure and beauty. For the hearing seeks both these, experiencing something like sight; for sight too, when it looks upon moulded works and paintings and carvings and all the products of human hands, if it finds in them both the pleasant and the beautiful, is satisfied and desires nothing further. And let no one think it strange if I make two ends and separate the beautiful from the pleasant, nor consider it absurd if I hold that some expression is composed pleasantly but not beautifully, or beautifully but not pleasantly; for truth allows of such a thing, and I claim nothing novel. The expression of Thucydides, at any rate, and of Antiphon of Rhamnus is composed beautifully, by Zeus, if any other is, and one could not blame it on that score; yet not altogether pleasantly. And that of Ctesias of Cnidus the historian, and of Xenophon the disciple of Socrates, is pleasant in the highest degree, yet not beautiful so far as it should be. I speak in the more general sense, not absolutely, since in those writers too some things are fitted together pleasantly, and in these beautifully. But the composition of Herodotus has both these qualities, for it is at once pleasant and beautiful.

11. The things from which, in my opinion, pleasant and beautiful expression will come are these four, the most important and the most powerful: melody, rhythm, variety, and the propriety that accompanies these three. Under pleasure I rank charm, grace, euphony, sweetness, persuasiveness, and all such things; under beauty, grandeur, weight, solemnity of language, dignity, mellowness, and the like. For these seem to me the most important and, as it were, the chief heads of the rest. These, then, are the things at which all who write metre, song, or what is called prose seriously aim, and I do not know whether there is anything else beyond these; and those who have taken the first place in some one of them, and in both, are many good men. To bring forward examples of each of them at present is not possible, lest I wear out the discourse on this; and at the same time, if anything should fitly be said about any of them and there should be need of citations, there will be another more suitable occasion for them, when I sketch the characters of the harmonies. For now, this much said about them is enough. I return again to the distinctions I made between pleasant and beautiful composition, that the discourse may, as they say, proceed for me along its road.

I said, then, that the hearing takes pleasure first in melodies, then in rhythms, third in variations, and in all these in propriety. And that I speak the truth, I shall offer experience itself as witness, which one cannot impeach, since it is agreed with by our common feelings. For who is there who is not led and bewitched by this melody, while by some other he feels nothing of the kind; and who is not made at home by these rhythms, but troubled by those? Indeed I myself, even in the most crowded theatres, which a motley and unmusical throng fills, have thought I perceived how there is in all of us a certain natural kinship with melodiousness and good rhythm; for I have seen a good lute-player, highly esteemed, hooted by the crowd because he struck a single discordant string and spoiled the melody, and a flute-player of the highest skill in handling his instruments suffer this very thing because, blowing a hollow note or not pressing his lips, he piped a squeak or what is called a false note. And yet, if one should bid the layman take the instruments and himself do any of the things which he charged the artists with doing amiss, he could not. Why so? Because the one is a matter of knowledge, of which we have not all partaken, while the other is a feeling which nature has given to all. The same thing I have observed happening with rhythms also—all the people at once vexed and displeased whenever anyone made a stroke or movement or figure in disproportionate times and destroyed the rhythms. And it is not that things melodious and well-rhythmed lead us to pleasure, and we are all charmed by them, while variations and propriety have not the same charm and grace, nor are equally felt by all; on the contrary, these too greatly enchant us all when rightly done, and lead to great vexation when mismanaged. I infer this from the fact that, even when the music of instruments, of song, and of dance succeeds in all in point of grace, yet, if it does not make timely variations, or strays from propriety, the satiety is grievous, and what does not fit the subject appears unpleasing. Nor have I used an illustration foreign to the matter; for the science of civic discourse too was a kind of music, differing from the music of song and instruments in quantity, not in quality. For in it too the expressions have melody, rhythm, variation, and propriety; so that in this too the hearing is delighted by the melodies, led by the rhythms, welcomes the variations, and longs in all for what is proper—the difference being one of degree, of more and less.

Now the melody of speech is measured by a single interval, that which is called approximately the fifth; and it is neither raised more than three tones and a semitone toward the high, nor lowered more than this space toward the low. Yet not every expression set under a single part of speech is uttered at the same pitch, but one on the high pitch, another on the low, and another on both. Of those that have both pitches, some have, on a single syllable, the low blended with the high—these we call circumflexed—while others have each separately, in one syllable and another, each preserving its own nature by itself. And in disyllables there is no intervening space between low and high; but in polysyllables, however large they may be, the one that has the high pitch is single among the many other low ones. The music of instruments and of song, however, uses more intervals, not the fifth alone, but, beginning from the octave, it sings both the fifth and the fourth and the third and the tone and the semitone, and, as some think, the quarter-tone too perceptibly; and it claims to subordinate the words to the melodies, and not the melodies to the words, as is plain from many examples and especially from the songs of Euripides, which he has composed for Electra to speak in the Orestes, addressing the chorus:

Hush, hush! lightly set down the white tread of your sandal; make no noise; step away from there, away from the bed.

For in these the words "hush, hush, white" are sung upon a single note, although each of the three words has both low and high pitches. And "of the sandal [arbyles]" has its third syllable on the same pitch as the middle, although it is impossible for one word to take two high pitches. And of "set down [tithete]" the first syllable becomes lower, but the two after it are high-pitched and of the same tone. And of "make no noise [ktypeite]" the circumflex is done away with; for the two syllables are spoken on a single pitch. And "step away [apopro bate]" does not take the high accent on the middle syllable, but the pitch of the third has shifted to the fourth syllable. The same thing happens also with the rhythms. For prose does no violence to the times of any noun or verb, nor transposes them, but keeps the syllables, long and short, as it has received them from nature; whereas music and rhythmic art alter them, shortening and lengthening, so that they often pass over into their opposites; for they do not adjust the times to the syllables, but the syllables to the times.

Now that the difference by which music differs from speech has been shown, it remains to say this too: that the melody of the voice—I mean not the singing voice but the bare speaking voice—if it disposes the hearing pleasantly, may be called "tuneful," but not "in tune"; and the proportion among the times of the parts, preserving the melodic form, is "well-rhythmed," but not "in rhythm"; and how these differ from one another I shall say at the fitting time. For now I shall try to render what follows: how civic expression might, by the arrangement itself, sweeten the hearing—in respect of the melodies of the sounds, the proportions of the rhythms, the varieties of the variations, and the propriety to the subjects—since these are the heads I have laid down.

12. Not all the parts of expression naturally affect the hearing alike, just as not all visible things affect the sense of sight alike, nor all tasteable things the sense of taste, nor the other senses the things that move each. Rather, some sounds sweeten the hearing and some embitter it, some roughen it and some smooth it, and they produce many other effects upon it. The cause is both the nature of the letters out of which the voice is composed, which has many and various powers, and the interweaving of the syllables, shaped in all manner of ways. Since, then, the parts of expression have such a power, and since it is not possible to change the nature of each, it remains by mixture, blending, and juxtaposition to conceal the awkwardness that attends some of them—mingling smooth with rough, soft with hard, euphonious with ill-sounding, easily pronounced with hard to pronounce, long with short, and in like manner setting the rest together at the right moment; not taking many monosyllables in a row (for the hearing is fatigued by this), nor more polysyllables than enough, nor things of like tone beside things of like tone, nor things of like duration beside things of like duration. One must also quickly change the cases of nouns (for, prolonged beyond measure, they greatly offend the ears), and continually break up the similarity of many nouns set in a row, and of verbs and of the other parts, guarding against satiety; and not always remain in the same figures, but change frequently; and not bring in the same tropes, but vary them; nor begin often from the same words nor end in the same, exceeding the right measure of either.

And let no one suppose that I am laying these down absolutely, as though they would always be causes of pleasure, or their opposites of annoyance; I am not so foolish. For I know that pleasure often arises from both, at one time from things of like kind, at another from things of unlike kind; but in all cases I think one must hunt for the right moment, for this is the surest measure of pleasure and displeasure. And of the right moment neither orator nor philosopher has ever yet defined an art; nor did Gorgias of Leontini, who first attempted to write about it, write anything worth mention; nor has the matter a nature that can fall under any general and systematic comprehension; nor is the right moment in any way to be hunted by knowledge, but by judgement. And this some, having exercised it on many occasions and often, find better than others, while those who leave it unexercised hit upon it more rarely and, as it were, by chance.

But, to speak of the rest as well, these are the things I think one must observe in composition who is to dispose the hearing pleasantly: either to fit together with one another the tuneful, well-rhythmed, and euphonious words, by which the sense is sweetened and softened and on the whole made at home; or, the words not having such a nature, to interweave and weave them in with those that can charm it, so that by the grace of the latter the unpleasantness of the former may be overshadowed—as prudent generals do in marshalling their armies; for they too conceal the weak troops with the strong, and none of their force is rendered useless. And I say one must relieve sameness by introducing timely variations; for variation is a sweet thing in every work. And last—what is indeed the most important of all—one must assign to the subjects a harmony proper and becoming to them. As for being ashamed of any noun or verb, however worn, on the ground that it will be spoken without disgrace—I think there is no need; for I declare that there will be no part of speech so humble, sordid, or otherwise distasteful, by which some body or thing is signified, that will not have a suitable place in discourse. I exhort men, trusting to the arrangement, to bring such words out boldly and confidently, using Homer as my example—in whom even the cheapest of words are found—and Demosthenes and Herodotus and the rest, whom I shall mention a little later as fits each one. Let this be my account of pleasant composition—few things out of many theorems, but enough to serve as heads.

13. Well then. If anyone should ask me how a fine harmony might arise, and from what theorems, I should say, by Zeus, in no other way and from no other things than those from which the pleasant arises; for the same things produce both: a noble melody, a dignified rhythm, a magnificent variation, and the propriety that accompanies all these. For just as there is a pleasant expression, so there is a noble one; and as there is a polished rhythm, so there is another that is dignified; and as variation has its grace, so it has its loftiness too; and as for propriety, if it does not have the greatest share of beauty, it will scarcely have a share of anything. From all these, then, I say, the beautiful in the harmony of expression must be cultivated, the same from which the pleasant comes too. And here too the cause is the nature of the letters and the power of the syllables, out of which words are woven; about which it would now be the right time to speak, as I promised.

14. The beginnings, then, of the human articulate voice, which no longer admit of division, are what we call elements and letters: letters, because they are signified by certain marks; elements, because every voice takes its origin from these as its first things and is dissolved into them as its last. Of the elements and letters the nature is not one and the same for all; their first difference is, as Aristoxenus the musician declares, that according to which some produce vocal sounds and others mere noises—vocal sounds being those called vowels, noises being all the rest. The second is that according to which, of the non-vowels, some by themselves naturally produce noises of some sort, indicative of a hiss or a whistle or a murmur or some other such sounds, while others are wholly devoid of voice and noise and cannot be sounded by themselves; for this reason some have called the latter mutes and the former semivowels. But those who divided the primary and elementary powers of the voice into three classes called vowels those which are voiced both by themselves and with others and are self-sufficient; semivowels those which are uttered better with vowels than by themselves, and by themselves worse and not self-sufficiently; and mutes those which have neither the complete nor the half-complete sounds by themselves, but are sounded out with others.

As to what their number is, it is not easy to say with precision, since the matter caused much perplexity to our predecessors too. For some thought that the elements of the voice were thirteen in all, and the rest constructed out of these; others thought there were more even than the twenty-four we now use. The study of these things, however, belongs more properly to grammar and metric, or, if one wishes, to philosophy; and it is enough for us, supposing the first principles of the voice to be neither fewer nor more than these, to state the properties belonging to them, beginning with the vowels.

These, then, are in number seven: two short, epsilon and omicron; two long, eta and omega; and three of doubtful quantity, alpha, iota, and upsilon, for these are both lengthened and shortened; and these some call common, as I said, while others call them variable. All these are sounded by the windpipe resounding with the breath, the mouth simply shaped, and the tongue doing no work but resting still. Only the long ones, and those of the doubtful quantity that are spoken long, take a sustained and continuous channel of breath, while the short ones, or those spoken short, are brought out by a cutting-off and a single stroke of breath, the windpipe being moved only for an instant. Of these the best, and those that produce the most pleasant sound, are the long ones and those of the doubtful quantity that are prolonged in utterance, because they are sounded for a long time and do not cut short the strain of the breath; the worse are the short ones or those spoken short, because they are small-voiced and scant of sound. Among the long vowels themselves, again, the most euphonious is alpha when it is prolonged; for it is uttered with the mouth opened to the fullest and the breath carried upward toward the palate. The second is eta, because it forces the sound downward about the base of the tongue, not upward, and the mouth is moderately opened. The third is omega; for in it the mouth is rounded and the lips are drawn together, and the breath makes its stroke about the edge of the mouth. Less than this is upsilon; for, a considerable contraction occurring about the lips themselves, the sound is choked and comes out narrow. Last of all is iota; for the breath strikes about the teeth, the mouth being slightly opened and the lips not brightening the sound. Of the short vowels, neither is comely, but epsilon is less ill-shaped than omicron; for it parts the mouth better than the other and takes its stroke more about the windpipe.

Such, then, is the nature of the vowel letters. The nature of the semivowels is as follows. They are eight in number: five simple—lambda, mu, nu, rho, sigma—and three double—zeta, xi, psi. They are called double either because they are composite—zeta of sigma and delta, xi of kappa and sigma, psi of pi and sigma, blended with one another and taking on their own sound—or because each, when taken into the syllables, occupies the place of two letters. Of these the doubles are better than the simples, since they are larger than the others and seem to approach nearer the complete sounds; the simples are inferior because the sound is gathered into shorter spaces. Each of them is sounded in some such way as follows. Lambda, with the tongue rising to the palate and the windpipe resounding. Mu, with the mouth closed by the lips and the breath divided through the nostrils. Nu, with the tongue shutting off the passage of the breath and transferring the sound to the nostrils. Rho, with the tip of the tongue fanning the breath and rising to the palate near the teeth. Sigma, with the tongue brought wholly up toward the palate, the breath carried through the middle of them and pushing out about the teeth a thin and narrow whistling. The three remaining semivowels take a mixed noise: from one of the semivowels, sigma, and from three mutes—delta, kappa, and pi.

These are the formations of the semivowel letters. They cannot all move the hearing alike; for lambda sweetens it and is the sweetest of the semivowels, while rho roughens it and is the most noble of its kind; and the two that resound through the nostrils, mu and nu, dispose it in an intermediate way, producing sounds of a hornlike character. But sigma is graceless and unpleasant, and, used to excess, gives much offence; for the hissing seems to belong rather to a brutish and irrational voice than to a rational one. At any rate, some of the ancients used it sparingly and guardedly, and there are those who composed whole songs without a sigma; and this Pindar too makes plain where he says:

Before, the song of the dithyramb crept on, drawn out at length, and the san counterfeit upon the lips of men.

Of the other three letters, those called double, zeta sweetens the hearing more than the others; for xi gives out its hiss through kappa and sigma, and psi through pi and sigma, both of these being unaspirated, whereas zeta is softly aspirated by the breath and is the most noble of its kind. So much, then, concerning the semivowels.

Of the so-called mutes, nine in number, three are smooth, three rough, and three intermediate between these: smooth, kappa, pi, tau; rough, chi, phi, theta; common to both, gamma, beta, delta. Each of them is sounded in this manner. Three from the edge of the lips, when, the mouth being closed, the breath then thrust out from the windpipe looses its bond. The smooth of these is pi, the rough phi, the intermediate beta; for it is smoother than the rough and rougher than the smooth. This is one set of three mute letters, spoken with a like formation but differing in smoothness and roughness. Three others are spoken with the tip of the tongue pressed against the front of the mouth at the upper teeth, then fanned by the breath, which gives it a passage downward about the teeth; these differ in roughness and smoothness; for the smooth of them is tau, the rough theta, the intermediate and common delta. This is the second set of three mute letters. The three remaining mutes are spoken with the tongue rising toward the palate near the throat and the windpipe sounding under the breath; these too differ in no way from one another in formation, except that one is spoken smooth, one rough, and one moderately and between the two. Of these the best are those spoken with much breath, the second those spoken moderately, the worst those spoken smooth; for these have only their own power, whereas the rough ones have in addition the breath, so that they are somewhere near more complete than those.

15. Out of the letters, then, which are so many and have such powers, the so-called syllables arise. Of these the long are those composed of long vowels or of the doubtful-quantity vowels when brought out long, and those that end in a long letter, or one spoken long, or in one of the semivowels and mutes; the short are those composed of a short vowel, or one taken short, and those that end in these. But of length and shortness of syllables the nature is not single; rather, some are longer than the long, and some shorter than the short. This will become clear from examples.

It is agreed that a syllable is short which the short vowel omicron makes, as it is spoken in hodos (road). Let someone add to it a single one of the semivowel letters, rho, and let it become Rhodos; the syllable still remains short, only not in like manner, but it will have some slight difference from the former. Let him further add to it one of the mute letters, tau, and let it become tropos; this will be larger than the former syllables and yet remains short. Let yet a third letter be added to the same syllable, sigma, and let it become strophos; by three additions perceptible to the ear it will become longer than the shortest, while still remaining short. So these are four differences of the short syllable, the irrational sense having the measure of the difference. The same account holds of the long also. For the syllable produced from omega, being long by nature, when increased by the addition of four letters—three set before and one after, in the form in which splen (spleen) is spoken—would surely be said to be larger than the former one of a single letter; and when reduced again by each of the added letters one by one, it would have perceptible differences toward the lesser. Now what the cause is that the long ones do not, when lengthened to the limit of the letters, go beyond their own nature, nor the short ones, when contracted from many letters to one, fall away from their shortness, but that the former are still regarded as double the short ones, and the latter as half the long ones, it is not necessary to consider for the present. For it is enough, so far as fitted the present argument, to have said that short syllable differs from short syllable and long from long, and that not every short and not every long has the same power, whether in bare prose or in poems and songs constructed by metre or rhythm.

This, then, is the first theorem concerning the affections of syllables. Another is this. Since the letters have many differences not only in length and shortness but also in their sounds, of which I spoke a little before, it is wholly necessary that the syllables composed of these, or woven through them, should preserve at once the individual power of each and the common power of all, which arises through their blending and juxtaposition; from which there come soft sounds and hard, smooth and rough, sweetening the hearing and embittering it, astringent and diffusive, and producing every other natural disposition; and these are countless in number.

Having noted these things, the most graceful of the poets and prose writers partly construct the words themselves, suitably interweaving them with one another, and partly contrive the letters and syllables that are proper to the feelings they wish to present, in various and ingenious ways—as Homer often does. Thus, in the case of windswept shores, wishing by the prolongation of the syllables to express the ceaseless roar:

The beaches resound as the brine belches forth.

And in the case of the blinded Cyclops, wishing to express both the greatness of the pain and the slow groping of his hands for the cave's door:

The Cyclops, groaning and travailing in his pangs, groping with his hands.

And somewhere else, wishing to display much earnest entreaty:

not even though Apollo the far-worker should suffer very much, grovelling on and on before father Zeus the aegis-bearer.

One may find countless such passages in him, indicating length of time or greatness of body or excess of suffering or stillness of rest or some such thing by nothing else so much as the construction of the syllables; and others wrought in the opposite manner for shortness and speed and haste and the like, as these:

Suddenly wailing she spoke among her handmaids, and

The charioteers were astounded when they saw the unwearied fire.

For in the one is shown the catching of the breath and the disorder of the voice; in the others, the distraction of the mind and the suddenness of the terror; and both of these the diminution of the syllables and letters produces.

16. The poets and prose writers, then, construct of themselves, looking to the thing, words proper to and indicative of the subjects, as I said; but they also take much from their predecessors as those constructed it, namely such words as are imitative of the things, as these are:

For the great wave roared against the dry land. And he, screaming, flew on the breaths of the wind. On the beach it roars greatly, and the sea thunders. He watched for the whir of arrows and the thud of javelins.

The great source and teacher of these is nature, who makes us imitative in our words, by which the things are signified according to certain reasonable likenesses that move the mind. By her we were taught to speak of the lowings of bulls, the neighings of horses, the snortings of goats, the roar of fire, the crash of winds, the creaking of rigging, and countless other things like these—some of them indications of sound, some of shape, some of action, some of feeling, some of motion, some of rest, some of any other thing whatever; about which much has been said by our predecessors, the best of it by Plato the disciple of Socrates, who first introduced the discourse on etymology, in many places and especially in the Cratylus.

What, then, is the chief point of this argument of mine? That the power of the syllables becomes various according to the interweavings of the letters, the nature of the words manifold according to the composition of the syllables, and discourse multiform according to the harmonies of the words; so that it is altogether necessary that an expression should be beautiful in which the words are beautiful, that beautiful syllables and letters should be the causes of the beauty of words, that pleasant diction should arise from words, syllables, and letters that please the hearing in like manner, and that the particular differences among these—according to which characters, passions, dispositions, deeds of the persons, and their attendant circumstances are revealed—should arise such as they are from the primary construction of the letters.

I shall use a few examples for the sake of clearness in this argument; for the many others you will find by collecting them for yourself. Homer, the most full-voiced of all poets, whenever he wishes to display beauty of a comely sight and a charm that draws on pleasure, uses the best of the vowels and the softest of the semivowels, and does not crowd the syllables with mutes, nor cut short the sounds by setting hard-pronounced letters beside one another, but makes the harmony of the letters gentle and flowing painlessly through the hearing, as these are:

And she came forth from her chamber, wise Penelope, like to Artemis or to golden Aphrodite. At Delos once, beside Apollo's altar, such a thing I marked, a young shoot of a palm springing up. And Chloris I saw, most lovely, whom once Neleus wedded for her beauty, when he gave countless gifts.

But when he brings in a piteous or fearful or haughty sight, he sets down not the best of the vowels but the most ill-sounding, and takes the hardest to pronounce of the noisy letters or the mutes, and crowds the syllables with these, as these are:

Terrible he appeared to them, fouled with brine. And on it the Gorgon, grim-eyed, was set as a crown, glaring dreadfully, and round about were Terror and Fear.

And when he wishes to imitate in his expression the confluence of rivers into one place and the crash of waters mingling, he does not work out smooth syllables but strong and resistant ones:

As when winter torrents, rivers flowing down from the mountains, into a meeting-place of glens hurl their mighty water.

And bringing in a man forcing his way against the opposing stream of a river with his arms, partly holding out and partly swept along, he produces checkings of syllables and delays of times and resistances of letters:

Terribly about Achilles the troubled wave stood up, and the stream, falling on his shield, pushed him, nor could he keep his footing.

And displaying the noise and piteous death of men dashed against rocks, he lingers over the most unpleasant and ill-sounding letters, nowhere smoothing the construction or sweetening it:

Seizing two together, like puppies, against the ground he dashed them; and the brains ran out upon the earth and wetted the ground.

It would be a great task to speak, if I should wish to bring forward examples of all the things one might demand under this head; so, content with what has been said, I shall pass on to what follows. I say, then, that he who wishes to work out beautiful expression, in setting the sounds together, should gather into one place such words as comprise fine sound, magnificence, or dignity. Some things on this matter have been said by Theophrastus the philosopher, in a more general way, in his work On Style, where he defines what words are by nature beautiful, by the composition of which he thinks the diction will become beautiful and magnificent, and again other words small and humble, from which he says neither a good poem nor a good prose discourse can come. And, by Zeus, the man has not said this wide of the mark. If, then, it were possible for all the parts of the expression by which the matter is to be revealed to be euphonious and well-worded, it would be madness to seek the inferior; but if this be impossible, as is the case with many things, one must try by interweaving, mixture, and juxtaposition to do away with the nature of the inferior—which is what Homer is wont to do in many places. For if anyone should ask any poet or orator whatever what dignity or fine wording these names have that belong to the Boeotian cities—Hyria and Mycalessus and Graia and Eteonus and Scolus and Thisbe and Onchestus and Eutresis, and the rest in order that the poet mentions—no one could tell of any whatever; yet he has so finely woven them together and divided them with euphonious fillings that they appear the most magnificent of all names:

The Boeotians Peneleos and Leitus led, and Arcesilaus and Prothoenor and Clonius, who dwelt in Hyria and rocky Aulis, and Schoenus and Scolus and ridgy Eteonus, Thespeia and Graia and broad Mycalessus, and those who dwelt about Harma and Eilesion and Erythrae, and those who held Eleon and Hyle and Peteon, Ocalea and Medeon, the well-built city.

Speaking to those who know, I do not think more examples are needed. For his whole Catalogue is of this sort, and many other passages, in which, compelled to take names not beautiful by nature, he adorns them with other beautiful ones and dissolves their harshness by the comeliness of the latter. And about these matters, enough.

17. Since I said that the rhythms too have no small share in dignified and magnificent composition, that no one may think I speak carelessly when I bring rhythms and metres, which belong to the study of music, into a diction that is neither rhythmic nor metrical, I shall render the account of these too. It is as follows.

Every noun and verb and other part of expression that is not monosyllabic is spoken in some rhythm; and I call the same thing foot and rhythm. Of disyllabic expression, then, there are three differences. For it will be either of two short syllables, or of two long, or of one short and one long. Of this third rhythm the manner is twofold: one beginning from a short syllable and ending in a long, the other from a long and ending in a short. The short-syllabled one is called the hegemon and the pyrrhic, and is neither magnificent nor dignified; its form is such as this:

Speak, then, by the foot the newly poured strains.

That which has both its syllables long is called the spondee, and has great dignity and much gravity; an example of it is this:

What course, then, shall I take—this or that, that or this?

That composed of a short and a long, if it takes the leading syllable short, is called iamb, and is not ignoble; but if it begins from the long, trochee, and is softer than the other and more ignoble. An example of the former is such as this:

At leisure you are present, son of Menoetius.

And of the other:

O heart, my heart, troubled with helpless cares.

So many, then, are the differences, rhythms, and forms of disyllabic parts of expression. Of trisyllabic ones there are others, more than those mentioned, and admitting a more varied study. That which consists of all short syllables, called by some the choreus, of which an example is such as this:

Bromius, spear-bearer, war-god, raiser of battle-din,

is humble, undignified, and ignoble, and nothing noble could come from it. That which consists of all long syllables—the metrists call it the molossus—is lofty and dignified and far-striding; an example of it is such as this:

O Zeus' and Leda's fairest saviours.

That made of a long and two shorts, taking the long in the middle, is named amphibrach, and is not very much among the well-formed rhythms, but is broken-up and has much that is effeminate and unpleasing, such as these:

Iacchus, triumph, you the leader of these dances.

That which takes the two shorts first is called anapaest, and has much dignity; and where one must invest the subject with some grandeur or feeling, it is suitable to be employed; its form is such as this:

A heavy crown it is for my head to bear.

That which begins from the long and ends in the shorts is called dactyl, and is altogether dignified and most notable for the beauty of expression; and the heroic metre is for the most part adorned by it; an example of it is this:

Bearing me from Ilium, a wind brought me near the Cicones.

The rhythmists, however, say that the long syllable of this foot is shorter than the perfect long, and, not being able to say by how much, they call it irrational. There is another that has a rhythm corresponding to this, which, beginning from the shorts, ends in the irrational; this one, separating it from the anapaests, they call cyclic, bringing forward an example of it such as this:

The high-gated city is laid low upon the earth.

About these there might be another discussion; only both rhythms are among the very beautiful ones. One kind of trisyllabic rhythm still remains, which consists of two longs and a short, and makes three forms. For, the short becoming the middle and the longs the extremes, it is called cretic, and is not ignoble. An example of it is such:

And they hastened on swift sea-cars beaked with bronze.

But if the two longs hold the beginning and the short the end, such as these:

To you, Phoebus, the Muses too share an altar,

the form is very manly and suited to dignified language. The same will result if the short is set before the longs; for this rhythm too has dignity and grandeur; an example of it is this:

What shore, what woodland shall I run to? whither shall I go?

To both these feet names have been given by the metrists: bacchius to the former, hypobacchius to the other. These are the twelve rhythms and feet, the primary ones that measure out all expression, metrical and unmetrical, out of which arise lines and clauses; for all the other feet and rhythms are composite of these. A simple rhythm or foot will be neither less than two syllables nor more than three. And about these I do not know that one need say more.

18. The reasons for which I was now led to say these things by way of preface—for it was certainly not my purpose otherwise to touch on metrical and rhythmic theorems, but for the sake of necessity—are these: that through noble, dignified, and grand rhythms, composition becomes dignified, noble, and magnificent, while through ignoble and humble ones it becomes without grandeur and undignified, whether each of these rhythms be taken by itself or they be interwoven with one another in their combinations. If, then, it shall be possible to compose the expression out of all the best rhythms, it would be for us according to our prayer; but if it should be necessary to mix the worse with the better, as happens in many cases—for the words are attached to the things just as they chance—one must manage them artfully and steal away the necessity by the grace of the arrangement, having besides much license; for no rhythm is banished from unmetrical expression, as it is from metrical.

It remains to set out evidence of what I have said, that my discourse may gain credence too. It will be little about much. Come, then: who would not agree that this expression of Thucydides in the Funeral Oration is composed dignifiedly and magnificently?—"Most of those who have already spoken here praise the man who added this speech to the law, as a noble thing that it should be delivered over those who are buried after the wars." What, then, is it that has made this composition magnificent? That the clauses are composed of such rhythms. For the first three feet preceding the first clause are spondees, the fourth an anapaest, and the one after this again a spondee, then a cretic, all dignified. And the first clause is solemn for these reasons. The next one—"praise the man who added this speech to the law"—has two hypobacchii as its first feet, a cretic as the third, then again two hypobacchii and a syllable by which the clause is completed; so that it too is rightly solemn, composed of the noblest and finest rhythms. And the third clause—"as a noble thing that it should be delivered over those who are buried after the wars"—begins from the cretic foot, takes the anapaest second, and the spondee third, and the anapaest again fourth, then the two next dactyls, and the two last spondees, then a final cadence. This too has become noble through its feet. Most passages in Thucydides are of this kind—or rather, few are not so—so that he is rightly thought lofty and of fine diction, as bringing in noble rhythms.

And this Platonic passage—by what else than its being composed of the finest and most notable rhythms could one say it is so dignified and beautiful? For it is among the very famous and renowned, that which the man used at the beginning of the Funeral Oration: "In deed these men have for us what is due to them; and having attained it they go their appointed journey." In these there are two clauses that complete the period, and the rhythms that divide them are these: bacchius the first—for I should not think fit to scan this clause as iambic, considering that not nimble and swift but deliberate and slow times should be assigned to those who mourn—spondee the second; the next a dactyl, the elision being divided; then after this a spondee; the next rather a cretic than an anapaest; then, in my opinion, a spondee; the last hypobacchius, or, if one wishes, an anapaest; then a cadence. Of these rhythms none is humble or ignoble. Of the next clause—"and having attained it they go their appointed journey"—the first two feet are cretics, the two after these spondees; after which a cretic again, then last a hypobacchius. Necessarily, then, the discourse composed of all fine rhythms is fine. Countless such things can be found in Plato too. For the man was most divinely gifted at discerning melodiousness and good rhythm; and if he had been as clever at selecting words as he was extraordinary at arranging them, he would either have outstripped Demosthenes in beauty of expression or made the contest doubtful. As it is, in selection he sometimes errs, and especially where he pursues the lofty, extraordinary, and elaborate phrase, about which I have shown more clearly elsewhere. But he arranges words both pleasantly and beautifully, by Zeus, and one could not blame him in this respect.

I shall set out the expression of one more man, to whom I assign the prize for power in discourse. For Demosthenes is, as it were, a standard of selection of words and beauty of composition. In the speech On the Crown, then, there are three clauses that complete the first period, and the rhythms that measure these are the following: "First of all, men of Athens, I pray to all the gods and goddesses." The clause begins with the bacchius rhythm, then a spondee follows, then an anapaest, and after this another spondee, then in order three cretics, and last a spondee. Of the second clause, this—"as much goodwill as I continue to feel toward both the city and all of you"—the first foot is a hypobacchius, then a bacchius, or, if one wishes, a dactyl; then a cretic; after which are two compound feet, the so-called paeans; on these follows a molossus or a bacchius, for he may be divided either way; and last a spondee. Of the third clause, this—"that as much may be granted me by you toward this contest"—two hypobacchii begin, then a cretic follows, with which a spondee is joined; then again a bacchius or a cretic, and last again a cretic, then a cadence. What, then, prevented this from being a fine harmony of expression, in which there is neither a pyrrhic foot nor an iambic, nor an amphibrach, nor any of the choreus or trochee? And I do not say this, that each of those men never used the more ignoble rhythms too—for he did—but that they have well concealed them and woven them in, distributing the worse among the better.

But those who took no forethought for this department brought out their writings, some humble, some broken-up, and some having some other disgrace and shapelessness. Of these the first and middle and last is Hegesias of Magnesia; about whom, by Zeus and all the other gods, I do not know what to say—whether there was in him such insensibility and dullness that he could not see which rhythms are ignoble and which noble, or such infatuation and corruption of mind that, knowing the better, he then chose the worse, which I rather believe; for ignorance succeeds even by accident in many cases, but deliberate choice never. At any rate, in all the writings the man has left, one could not find a single page composed felicitously. He seems to have supposed these things better than those, and to make them with zeal, into which a man of sense, falling by necessity in an extempore discourse, would set them down as a disgrace. I shall set down an expression of his from his history, that it may become clear to you by the contrast how much dignity the noble has in rhythms, and how much disgrace the ignoble. The matter the sophist takes up is this: Alexander, besieging Gaza, a very strong place in Syria, is wounded in the assault and takes the place in time; carried away by anger, he slaughters all who are caught, ordering the Macedonians to kill whomever they meet, and, taking captive their commander, a man of high rank in fortune and in beauty, he orders him bound alive from a chariot and the horses driven at full speed, and destroys him in the sight of all. One could not name sufferings more terrible than these, or more fearful to behold. How, then, the sophist has expressed these, it is worth seeing—whether dignifiedly and loftily, or humbly and laughably.

"And the king, with his troop, led the way. And somehow it had been planned by the best of the enemy to meet him as he advanced; for it had been resolved that, having mastered one man, they would also drive out the multitude. So this hope ran together into daring, so that Alexander never before was in such danger. For a man of the enemy, bent down to his knees, seemed to Alexander to have done this for the sake of supplication. Letting him come near a little, he dodges as the man drove his sword at the lappets of the breastplate, so that the blow was not most deadly. But him he himself destroyed, striking him on the head with his blade, while the rest fresh anger inflamed. So, then, the recklessness of the daring deed drove out the pity of each—of those who saw, of those who heard—so that six thousand of the barbarians were cut down at that trumpet-call. Baetis himself, however, Leonatus and Philotas brought up alive. Seeing him fleshy and large and grim-eyed—for he was black in colour too—Alexander, hating both his designs and his appearance, ordered them to thread a bronze ring through his feet and drag him round naked. Crushed by his ills against many rough places, he screamed. And it was this very thing, as I say, that gathered men round. For the toil intensified, and he cried out barbarously, beseeching his master; and the solecism made men laugh. The fat and the bulk of his flesh revealed another Babylonian animal, fat. So the crowd mocked, hurling soldierly insolence, ugly and crooked in manner, against the foe."

Now are these like those Homeric lines in which Achilles is outraging Hector after his death? And yet that suffering is less; for the outrage is on an insensible body. Still it is worth seeing how far the poet surpasses the sophist.

He spoke, and devised foul deeds against noble Hector. He pierced the tendons of both feet behind, from heel to ankle, and fastened thongs of oxhide, and bound them to the chariot, but let the head trail. Mounting the chariot and lifting up the famous armour, he lashed the horses to go; not unwilling the two flew on. A cloud of dust rose from the dragged corpse; and about it his dark hair was strewn, and the whole head lay in the dust— the head once fair; but then Zeus to his enemies gave it to be outraged in his own native land. So all his head was fouled with dust; and now his mother tore her hair, and flung far the gleaming veil, and wailed aloud most loudly when she beheld her son. Piteously groaned his dear father, and round about the people were held by wailing and lamentation throughout the town. It was most like this, as if all beetling Ilium were smouldering with fire from the heights.

So a noble body and dreadful sufferings ought to be spoken of by men who have sense and understanding. But as Hegesias of Magnesia has spoken, they would be spoken by women or broken men—and not even by these in earnest, but in jest and derision. What, then, was the cause of the nobility of those verses, and of the humbleness of these trifles? The difference of the rhythms more than anything, even if not it alone. For in those there is no undignified line, while here there is no period that will not give pain.

Having now spoken about the rhythms too, how great a power they have, I shall pass on to what remains.

19. My third theorem of the things that make fine harmony was variation. I mean not the change from the better to the worse—for this is quite silly—nor again from the worse to the better, but the variety within things of the same kind. For all fine things have their satiety, just as pleasant things do, when they lie in sameness; but varied by variations they always remain fresh. Now to those who write metres and songs it is not possible to vary everything, or not all of them, nor to the extent they wish. Thus the epic poets may not vary their metre, but all the lines must be hexameters; nor their rhythm, but they will use the feet beginning from a long syllable, and not even all of these. To those who write songs it is not possible to alter the melody of the strophes and antistrophes, but, whether they assume enharmonic, chromatic, or diatonic melodies, in all the strophes and antistrophes the same modulations must be preserved; nor the rhythms that comprise whole strophes and antistrophes, but these too must remain the same; but about the so-called epodes it is permitted to vary both these, the melody and the rhythm. The clauses out of which the period is composed it is granted, with much liberty, to divide variously, setting upon them now these, now those magnitudes and forms, until they complete the strophe; then again they must make the same metres and clauses. The ancient lyric poets, I mean Alcaeus and Sappho, made their strophes short, so that in their few clauses they introduced not many variations, and used very few epodes; but Stesichorus and Pindar, working out larger periods, divided them into many metres and clauses for no other reason than love of variation. The dithyrambic poets even changed the modes, making them now Dorian, now Phrygian, now Lydian in the same song, and varied the melodies, making them now enharmonic, now chromatic, now diatonic, and continually exercised their full license in the rhythms—I mean those of the school of Philoxenus, Timotheus, and Telestes, since among the ancients at least the dithyramb too was orderly.

But prose has full liberty to vary the composition by variations as it wishes. And the best expression of all is that which has the most pauses and enharmonic variations—when one part is spoken within a period and another outside a period; and this period is woven of more clauses, that one of fewer; and of the clauses themselves one is shorter, another longer, one more rough-hewn, another more finished; and there are now these, now those rhythms, and figures of all kinds, and the different pitches of the voice, the so-called accents, stealing away satiety by their variety. There is also a certain grace, in such things, in that which is so composed as not to seem composed. And I do not think much discourse is needed on this point; for that variation is the sweetest and most beautiful thing in discourse, I am persuaded all know. As an example of it I take all the expression of Herodotus, all that of Plato, and all that of Demosthenes; for it is impossible to find others who have used more numerous episodes, more timely varieties, and more multiform figures—I mean the one as in the form of history, the other as in the grace of dialogue, the third as in the use of forensic contests. But the school of Isocrates and his disciples was not like these; for, though these men composed much pleasantly and magnificently, in respect of variations and variety they are not very successful. With them there is one cycle of the period, a uniform order of figures, the same guarding against the clashing of vowels, and many other such things that fatigue the hearing. I do not, then, approve that school in this respect. In Isocrates himself, perhaps, many other graces bloomed that concealed this shapelessness; but among his successors, since their other successes are fewer, this fault becomes more conspicuous.

20. There still remains for me the discourse on propriety. For propriety must be present in all other things too, and if any work falls short in this department, it falls short, if not of everything, at least of the best. About this whole subject, then, this is not the time to inquire; for its study is deep and requires much discussion. But of all that bears on this part with which I happen to be dealing, if not all nor even most, yet so much as is possible, let it be said.

It being agreed among all that the proper is what is suited to the underlying persons and matters, then, just as there would be a selection of words, one proper to the subjects and another improper, so surely also of composition. Of this one must take the truth as one's example. What I mean is this: we do not use a like composition when angry and when rejoicing, nor when lamenting and when fearing, nor when we are in any other passion or evil, as we do when we reflect that nothing at all is disturbing or grieving us. There are countless other things one might mention, wishing to reckon up all the kinds of propriety; but the one I have readiest and most generally to say about it, this I shall state. The same men, being in the same state of soul, when they report matters at which they happen to have been present, do not use a like composition about all, but become imitative of the things reported in arranging the words too, doing nothing deliberately but led to this naturally. Observing these things, the good poet and orator must be imitative of the matters about which he brings out his discourse, not only in the selection of words but also in the composition. This the most divinely gifted Homer is wont to do, although he has but one metre and, as it were, few rhythms; yet still he is ever creating something new in them and exercising his art, so that it makes no difference to us whether we see the things happening or hear them spoken. I shall say a few things, by which one might be able to use a few examples for many. Reporting to the Phaeacians his own wandering and his descent into Hades, Odysseus, having told the sights of the evils there, renders them. Among these he relates also the sufferings of Sisyphus, for whom they say the gods below have set a limit to his release from torments when he shall have rolled a stone up over some hill; but this is impossible, the stone falling down whenever it comes to the top. How, then, he will reveal these things imitatively by the very arrangement of the words, it is worth seeing:

And Sisyphus too I saw, having mighty pains, heaving a monstrous stone with both his hands; indeed, propping himself with hands and feet, he kept thrusting the stone up toward a crest.

Here the arrangement is what reveals each of the things that happen: the weight of the stone, the laborious movement from the ground, the man bracing himself with his limbs, climbing toward the hill, the rock scarcely thrust up; no one could say otherwise. And by what means has each of these come about? Not, by Zeus, at random or of itself. First, in the two lines in which he rolls the stone up, except for two verbs the remaining parts of the expression are all either disyllabic or monosyllabic; next, the long syllables are by half more numerous than the short in each of the lines; then all the harmonies of the words have made wide strides, and stand apart very perceptibly, either through the clashing of the vowel letters or through the joining of the semivowels and mutes; and the whole is composed of dactylic and spondaic rhythms, the longest and those with the widest stride. What, then, can each of these effect? The monosyllabic and disyllabic words, leaving many intervening times between one another, imitated the protracted duration of the labour; the long syllables, having certain supports and seatings, the resistance and the heaviness and the difficulty; the chill space between the words and the juxtaposition of the roughening letters, the intervals of the action, the checks, and the greatness of the toil; the rhythms, regarded in their length, the stretching of the limbs, the dragging-out of the man who rolls, and the bracing of the stone. And that these are not the works of nature acting at random but of art trying to imitate the things happening, what is said next reveals. For the stone turning back again from the summit and rolling down he has expressed not in the same manner, but quickening and contracting the composition; for, having said in the same figure—

But when he was about to heave it over the top—

he sets down this:

Then the mighty force would turn it back again; and then down to the plain the shameless stone rolled.

Does not the arrangement of the words roll down together with the weight of the stone—or rather, has not the speed of the report outstripped the rush of the rock? It seems so to me. And what again is the cause here? For this too was worth seeing. The line that reveals the descent of the stone has no monosyllable, but only two disyllabic words. First, then, this does not separate the times but quickens them; next, the line having seventeen syllables, ten are short syllables and seven long, and not even these complete; necessarily, then, the phrase is drawn down and contracted, dragged along by the shortness of the syllables. Besides this, no word has taken a notable separation from the next; for there is nothing set side by side—vowel beside vowel, or semivowel beside semivowel or mute—of the kind that naturally roughens and parts the harmonies. No perceptible separation arises, the words not being severed, but they slip together and are carried down together and become, as it were, one out of all, through the precision of the harmonies. And what is most worthy of wonder above the rest, no rhythm of the long ones that naturally fall into heroic metre—neither spondee nor bacchius—is mixed into the line, except at the end; all the others are dactyls, and these too have irrational syllables mixed in, so that some do not differ much from trochees. There is nothing, then, to prevent the phrase from being nimble and rounded and flowing down, struck out of such rhythms.

One could show many such things spoken in Homer; but to me these too seem enough, that I may have room to speak of the rest. The things, then, at which those must aim who are to make pleasant and beautiful composition, both in poetry and in unmetrical discourse, these, in my opinion, are at least the most important and the most powerful. As for those it was not possible to treat—being lesser than these and dimmer and, through their multitude, hard to comprehend in a single writing—these I shall add in the daily exercises and shall use the testimonies of many good poets, prose writers, and orators. For now, having added to my discourse the remaining things I promised, no less necessary than the rest, I shall stop: what the differences of composition are, and what the character of each is for the most part, and to mention those who took first place in them and to furnish samples of each; and when these have reached their end for me, then to clear up also those matters perplexing to most—what it is that makes prose expression appear like a poem while remaining in the form of discourse, and poetic phrase resemble prose while preserving its poetic dignity; for almost all who have spoken or composed best have these excellences in their expression. I must try, then, to speak about these too, what I think. I shall begin from the first.

21. I hold that the specific differences of composition are very many indeed, and able to come neither under a single view nor under exact reckoning; and I think that each of us has his own peculiar character of composition of words, just as of countenance—using no mean example, that of painting; for as in painting all who paint living creatures, taking the same pigments, make mixtures in no way like one another, so in poetic and in all other diction, all using the same words, do not compose them alike. The generic differences of composition, however, I am persuaded are only these three, to which anyone who wishes may give appropriate names, when he has heard their characters and their differences. For myself, not being able to address them by proper names, since they are nameless, I call them by metaphorical names: the one austere, the other polished, and the third well-blended. How this third ever comes to be, I am at a loss, and "my mind is divided to speak the truth"—whether by privation of each of the two extremes or by mixture; for it is not easy to conjecture the clear truth. Perhaps, then, it is better to say that by the relaxation and the intensification of the extreme limits the intermediate ones, which are very many, come to be. For it is not as in music, where the middle string is equidistant from the highest and the lowest, that in like manner in discourse the middle character stands equidistant from each of the extremes; rather it is among the things contemplated in breadth, like a herd and a heap and many other things. But this is not the time fitting for that study; I must speak, as I proposed, about the characters too—not all that I might say, for I should need very long discourses, but the most manifest things themselves.

22. The character of the austere harmony, then, is as follows. It wishes the words to be set firmly and to take strong positions, so that each word may be seen from every side; it wishes the parts to stand apart from one another at notable distances, divided by perceptible times. It does not at all mind using rough and resistant junctures in many places, such as occur, among stones laid together at random in buildings, in the bases that are not well-cornered and not smoothed flat but unworked and casual. It loves to draw out at length with large words spread broadly; for to be contracted into short syllables is hostile to it, except where necessity sometimes compels.

These things, then, it tries to pursue in the words and longs for. In the clauses it likewise cultivates the same things, and the dignified and magnificent rhythms; and it does not wish the clauses to be equal to one another, or alike, or slaves to necessity, but consequent and noble and brilliant and free; it wishes them to resemble nature rather than art, and to be spoken according to passion rather than according to character. Periods that round off the sense within themselves, for the most part it does not even wish to compose; but if ever it should be carried to this of itself, it wishes to display the unstudied and simple, using neither certain additions of words, that the cycle may be filled out, that in no way help the sense, nor showing any zeal that the cadences may be theatrical or polished, nor being very carefully measured so as to suffice for the speaker's breath, nor having any other such study whatever. These too are proper to such a harmony: it is poised in the matter of cases, varied in its figures, sparing of conjunctions, without articles, in many places disdainful of consequence, least of all florid, high-minded, blunt, unadorned, having archaism and mellowness for its beauty.

Of this harmony there were many devotees in poetry, history, and civic discourse, surpassing the rest: in epic poetry Antimachus of Colophon and Empedocles the natural philosopher; in lyric poetry Pindar; in tragedy Aeschylus; in history Thucydides; in civic discourse Antiphon. Here the subject demanded that I furnish many examples of each of those mentioned, and perhaps the discourse would have become not unpleasant, embroidered with many examples as with spring flowers; but the treatise would have appeared beyond measure, and scholastic rather than instructive. Yet neither was it fitting to leave the things said untested, as if they were manifest and needing no testimony; somehow one had to take the mean between both, and neither exceed the right measure nor fall short of conviction. This I shall try to do, taking a few samples from the most eminent men. Of poets, then, Pindar taken alone will suffice, and of prose writers Thucydides; for these are the best in the austere harmony. Let Pindar begin, and of him a dithyramb of which this is the beginning:

Come into the chorus, ye Olympians, and send upon it the famous grace, ye gods, who haunt the much-trodden, fragrant navel of the city in sacred Athens, and the all-cunning, glorious market-place, to win the violet-bound garlands and the spring-plucked songs; and look upon me, sent from Zeus with splendour, coming a second time with songs to the ivy-crowned god, whom we mortals call Bromius and the Loud-shouter, to sing him, the offspring of the highest fathers and of Cadmean women. The clear signs of fulfilment do not escape him, whenever, the chamber of the crimson-robed Hours being opened, spring brings on the nectar-breathing plants; then, then upon the immortal earth the lovely tresses of violets are flung, and roses are mingled in the hair, and the voices of songs ring out with the flutes, and the choirs ring out for Semele of the circling diadem.

That these are strong, compact, and dignified, and have much austerity, and roughen the ears without giving pain and embitter them in moderation, and are protracted in their times and stride widely in their harmonies, and display not this theatrical and polished beauty but that archaic and austere one—all, I am sure, would testify who have a moderate sense for discourse. By what study they were so wrought—for it was not without art and reason, but using spontaneity and chance, that they took this character—I shall try to show.

His first clause is composed of four parts of speech: a verb, a conjunction, and two appellatives. The verb and the conjunction, blended by elision, have made the harmony not unpleasing; but the appellative set together with the conjunction has notably roughened the joining; for "into the chorus [en choron]" is resistant and not smooth-flowing, the conjunction ending in the semivowel n and the appellative taking its beginning from one of the mutes, ch. These elements are by nature unmixable and unglueable; for n does not naturally precede ch within a single syllable, so that, not becoming boundaries of syllables, they do not join the sound, but necessarily a certain silence arises between the two, distinguishing the powers of each of the letters. The first clause, then, is thus roughened by the arrangement. (Allow me to call clauses not those by which Aristophanes or some other of the metrists arranged the songs, but those by which nature thinks fit to divide the discourse, and by which the children of the orators divide their periods.)

The clause that lies next to this—"and send upon it the famous grace, ye gods"—has made a notable stride from the former, and has comprised within itself many resistant harmonies. For it begins with one vowel letter, epsilon, and is set beside another vowel, iota; for into this the preceding clause ended. These too are not blended together, nor does the iota precede the epsilon within a single syllable; rather a certain silence arises between the two, propping each of the parts and giving them a sure footing. And in the detailed composition of the clause, to the conjunctions "and upon [epi te]," from which the clause begins—whether one ought to call the leading one a preposition—the appellative part lying upon it, "famous [klytan]," has made the composition resistant and rough. In what respect? Because the first syllable of "klytan" wishes to be short, but is longer than the short, being composed of a mute, a semivowel, and a vowel. Its not being purely short, and at the same time the difficulty of pronouncing the blend of the letters, produces a delay and a check of the harmony. If, at any rate, one were to remove the kappa from the syllable and make it "epi te lytan," both the slowness and the roughness of the harmony will be dissolved. Again, to the appellative "klytan" the verbal "send [pempete]" lying upon it has not a sound concordant or well-blended; rather the m must be supported and, the mouth being sufficiently pressed, the p then becomes audible; for the p is not subordinate to the m. The cause of this is the shaping of the mouth, which brings out each of the letters neither in the same place nor in the same manner; for the sound of the m is made about the palate, the tongue rising at the tips of the teeth and the breath being divided through the nostrils, while that of the p is made with the mouth closed and the tongue doing no work, the breath taking its noise all at once at the opening of the lips, as I said before; and in the mouth's passing from one shaping to another, neither kindred nor alike, a certain time is comprehended, with which the smoothness and good flow of the harmony are parted. And at the same time, neither does the preceding syllable of "pempete" have a soft sound, but it roughens the hearing somewhat, beginning from a mute and ending in a semivowel. And to "grace [charin]" the "gods [theoi]" lying beside it checks the sound and makes a notable propping of the parts, the one ending in a semivowel, n, and the other having a mute, th, as its leader; and the semivowels do not naturally precede any of the mutes.

Upon these is brought a third clause, this: "who haunt the much-trodden, fragrant navel of the city in sacred Athens." Here, to "navel [omphalon]," ending in the n, the "fragrant [thyoenta]" lying beside it, beginning from the th, renders a resistance like the former; and "thyoenta," ending in a vowel, a, joined to "in sacred [en tais hierais]," which takes its beginning from a vowel, e, has torn apart the sound by the time between, which is not small. Upon these follow those words, "the all-cunning, glorious market-place [pandaidalon t' euklea agoran]." Here too the conjunction is rough and resistant; for a mute is joined to a semivowel, t to the n, and the time has made a notable stride between the appellative "pandaidalon" and the elision joined to it; for both syllables are long, but the syllable that elides the two is by no little larger than the moderate, being composed of a mute and two vowels. If, at any rate, one should remove the t from it and make it "pandaidalon euklea agoran," coming to the just measure it will make the harmony most smooth-flowing.

Like these too are those words, "to win the violet-bound [iodeton lachein]." For two semivowels are set beside one another, the t and the l, having no natural conjunction, since they are brought out neither in the same places nor with like shapings of the mouth. And the words said upon these are drawn out in their syllables and stand apart in their harmonies for a long way: "garlands and the spring-plucked [stephanon tan t' earidropon]"; for here too long syllables clash, exceeding the just measure—both the final syllable of "stephanon," comprising within two semivowels a vowel long by nature, and the one joined to it, drawn out by three letters, a mute, a vowel spoken long, and a semivowel; there has arisen, then, both a propping by the lengths of the syllables and a resistance by the juxtaposition of the letters, the n having no sound concordant with the t, which I said before too. There is set also beside "songs [aoidan]," ending in the n, "from Zeus [Diothen te]," beginning from the mute d; and beside "with splendour [syn aglaiai]," ending in iota, "look upon me, journeying with songs [idete poreuthent' aoidan]," beginning from iota. One could find many such things examining the whole song.

But that I may have room to speak of the rest, let Pindar suffice, and let there be taken of Thucydides this expression from the preface:

"Thucydides the Athenian wrote the war of the Peloponnesians and the Athenians, how they warred against one another, beginning as soon as it broke out and expecting that it would be great and the most notable of those that had gone before, inferring this because both sides went into it at the height of their full preparation, and because he saw the rest of the Greek world taking sides with one party or the other, some at once, some intending to. For this was the greatest movement that ever befell the Greeks and a part of the barbarians, and, one might say, the most of mankind. As for the events before these, and those still more ancient, it was impossible to find them out clearly through length of time; but from such evidence as, examining as far back as possible, I am led to trust, I do not think they were great, either in their wars or in other respects. For the land now called Hellas, it appears, was not anciently settled with security, but there were migrations in former times, and each people readily left its own land, forced by others ever more numerous. For, there being no commerce, and no mingling fearlessly with one another either by land or by sea, and each cultivating their own only so much as to live, having no surplus of wealth and not planting the ground—since it was uncertain when someone, coming upon them, would take it away, the more so as they were unwalled—and thinking they would everywhere command the daily necessary food, they removed without difficulty."

This expression—that it does not have smooth or accurately smoothed harmonies, nor is fluent and soft and slipping imperceptibly through the hearing, but displays much that is resistant, rough, and harsh; and that it does not touch even slightly upon festive or theatrical grace, but exhibits a certain archaic and self-willed beauty—as to all the well-educated, who know it equally with me, I need not say, the more so as the writer himself has agreed that his writing is less delightful to hear, and that it is composed to be a possession for ever rather than a prize-essay to be heard for the moment. What the theorems are which this man used to make the harmony rugged and austere, I shall indicate in few words; for it will be easy for small things to be samples of great, for those who do not pass with difficulty to the study of the like and consequent.

First, at the beginning, to the appellative "Athenian" the verb "wrote [xynegrapse]" being fitted parts the harmony notably; for the s is not set before the x by a co-utterance arising within a single syllable; rather, the s must be caught up in silence and then the x become audible. This affection produces roughness and resistance. Next, the syncopations of sounds that arise after this—the x, the s, the n, the s, and the g lying beside one another four times in a row—well grate the hearing and notably shake up the harmonies, when he says "the war of the Peloponnesians and the Athenians"; for of these parts of the expression there is none that must not be caught up in silence and pressed first by the mouth about the last letter, that the one joined to it may take its own power clear and pure. Besides this, the juxtaposition of vowels that occurs at the end of this clause in "and the Athenians" has broken the continuity of the harmony and parted it, taking a very perceptible time between; for the sounds of the a and the e are unmixed and cut off the sound; whereas it is continuous and smoothed sounds that make good flow.

And again, in the second period, the leading clause, this—"beginning as soon as it broke out"—the man, having fitted it moderately so that it appears most euphonious and soft, the part after this he again roughens and tears apart by the gapings of the harmonies: "and expecting that it would be great and the most notable of those that had gone before"; for three times in a row, at no great distance, vowels are set beside one another, producing clashings and checks and not allowing the hearing to take the impression of one continuous clause; and the period, ending for him in "of those that had gone before," has not a well-traced and rounded cadence, but appears, as it were, headless and unrounded, as if it were a part of the second rather than an ending.

The same thing the third period has suffered too; for of it the cadence is uncircumscribed and unseated, its last part being "some intending to," it too comprising at once many resistances of vowels against vowels and of semivowels against semivowels and mutes, which produce the roughnesses of things not concordant by nature. To put it in brief, there being some twelve periods that I have set out, if one should divide them in due measure for the breath, and there being clauses comprised in them not fewer than thirty, one could not find six or seven clauses in all composed with good flow and smoothed in their harmonies; but clashings of vowels in the twelve periods one would find nearly thirty, and juxtapositions of semivowels and mutes that are resistant, harsh, and hard to pronounce so many in number that there is something of the kind for nearly every part of the expression. Great too is the disproportion of the clauses to one another, and the irregularity of the periods, and the novelty of the figures, and the disdain of consequence, and all the other things I reckoned to be characteristic of the unadorned and austere harmony. For to go through all of them again on the examples and to waste time on this I do not think necessary.

23. The polished composition, which I placed second in order, has a character such as this. It does not seek that each word singly be seen from every side, nor that all stand upon a broad and secure seat, nor that the times between them be long; nor in general is this slow and steady quality dear to it; rather it wishes the wording to be in motion, and some words to be carried over others, and the chain of connection to be borne along, taking its footing like things that flow and never rest still; and it requires that the parts be blended together by elision and woven together so as to produce, as far as possible, the appearance of a single word. This the precisions of the harmonies effect, comprising no perceptible time between the words; and in detail it resembles finely woven fabrics, or paintings in which the lights are blended with the shadows. It wishes all the words to be euphonious, smooth, soft, and maidenly, and is somewhat hostile to rough and resistant syllables; and everything bold and hazardous it holds in wary avoidance.

And it wishes not only the words to be suitably fitted and smoothed together with the words, but also the clauses to be well woven with the clauses and all to end in a period, defining the length of a clause—which shall be neither shorter nor greater than the mean—and the measure of a period, which a man's full breath shall command; an unperiodic expression, or a period without clauses, or a disproportionate clause, it could not endure to make. It uses rhythms too, not the greatest but the middle and shorter ones; and it wishes the cadences of the periods to be well-rhythmed and set, as it were, by the rule, doing the opposite in the fitting of these to what it does in the fitting of the words; for those it blends together, but these it parts and wishes them to be conspicuous, as from a watch-tower. As for figures, it loves to use, for the most part, not the more old-fashioned ones, nor those to which some dignity, weight, or tone belongs, but the dainty and flattering ones, in which there is much that is deceptive and theatrical. And, to speak more generally, it has the opposite form to the former in the greatest and most important respects, about which I need not speak again.

It would follow to enumerate those who took first place in this style too. Of epic poets, then, Hesiod seems to me to have worked out this character most beautifully; of lyric poets Sappho, and after her Anacreon and Simonides; of tragic poets Euripides alone; of prose writers, accurately no one, but rather than the rest Ephorus and Theopompus; of orators Isocrates. I shall set down examples of this harmony too, choosing of poets Sappho, of orators Isocrates. And I shall begin from the lyric poet:

Many-throned, immortal Aphrodite, child of Zeus, weaver of wiles, I beseech you: do not with surfeits and with griefs subdue, O lady, my heart; but come hither, if ever at another time you heard my voice from afar and hearkened, and, leaving your father's house, came in your golden chariot, having yoked it. Fair swift sparrows drew you about the dark earth, beating their thick wings, from heaven down through the mid-air; and at once they arrived. And you, O blessed one, smiling with your immortal face, asked what I had suffered this time, and why this time I called, and what I most wished to happen for me in my maddened heart: "Whom this time shall I persuade to lead you back into her love? Who, O Sappho, wrongs you? For if she flees, soon she shall pursue; and if she takes not gifts, yet she shall give; and if she loves not, soon she shall love, even against her will." Come to me now too, and loose me from hard cares, and all that my heart longs to accomplish, accomplish; and you yourself be my ally.

The good flow and grace of this expression have arisen from the continuity and smoothness of the harmonies. For the words lie beside one another and are woven together according to certain natural affinities and conjunctions of the letters; for the vowels are joined to the mutes and semivowels almost throughout the whole song—all those that naturally precede and follow one another, brought out together within a single syllable; while collisions of semivowels with semivowels or mutes, and of mutes and vowels with one another, which shake the sounds, are very few in it. I, at any rate, examining the whole song, find in so many words, verbs, and other parts perhaps five or six interweavings of semivowel and mute letters that are not naturally apt to be blended with one another, and even these do not greatly roughen the good flow; and of juxtapositions of vowels, those that occur within the clauses themselves are still fewer than these, and those that join the clauses to one another a little more numerous. Reasonably, then, the expression has become fluent and soft, the harmony of the words in no way breaking the sound into waves.

I would have stated the rest of the peculiarities of this composition too, and shown by the examples that they are such as I say, were the discourse not going to become long and to give some appearance of repetition. For it will be possible for you and for anyone else, taking up one by one each of the things I have enumerated in the foregoing description of the character, to consider them upon the examples at much leisure and ease; but for me it is not possible to do this. It is enough only to point out sufficiently what I wish to those who are able to follow.

I shall set down the expression of one more man fashioned into the same character, Isocrates the orator, whom I think to have refined this harmony more accurately than all who have used prose. The expression is this, from the Areopagiticus:

"Many of you, I think, wonder what mind I had in coming forward about safety, as if the city were in dangers or its affairs were tottering, and not rather possessing more than two hundred triremes, keeping peace about the land, ruling the sea, and having besides many allies—some ready to help us if there be any need, and far more who pay their contributions and do what is enjoined. With these things ours, one might say it is reasonable for us to be confident, as being far from dangers, and that it befits our enemies to be afraid and to take counsel for their safety. You, then, I know, using this reasoning, both despise my coming forward and hope to hold all Hellas under this power; but I, for these very reasons, happen to be afraid. For I see those of the cities that think they fare best taking the worst counsel, and those that are most confident falling into the greatest danger. The cause of this is that none of the goods or the evils comes by itself to men, but with wealth and dominion are ranged and follow folly, and with folly licence; and with want and lowliness, temperance and much moderation. So that it is hard to discern which of these two portions one would pray to leave to one's own children; for one might see, from the one that seems the meaner, affairs for the most part advancing to the better, and from the one that appears the stronger, accustomed to change for the worse."

That these are blended together and shaded together, and that not each word singly stands upon a conspicuous and broad seat, nor is divided by long intervening times and stands apart from the others, but they appear to be in motion and in continuous flux and flow, and the harmonies that join the expression are gentle and soft and headlong—to this the irrational feeling of the hearing bears witness. And that the causes of these things are none other than those I stated before about this conduct of discourse, it is easy to see. For one would not find a single clashing of vowels, at least in the passages I have set out, and, I think, not in the whole speech, unless something has escaped me; and of semivowels and mutes, few and not very conspicuous nor continuous. These, then, are the causes of the good flow in the expression, together with the proportion of the clauses to one another, and the cycle of the periods, having something rounded and well-traced and dispensed to a nicety in its proportions. And above all these, the figures, having much that is fresh; for they are antitheses and assonances and balanced clauses and the like, out of which festive diction is wrought. I do not think it necessary to be prolix in going through the rest; for enough has been said about this composition too, so far as fitted.

24. The third and middle of the two harmonies mentioned, which I call well-blended for want of a proper and better name, has no peculiar form of its own, but is blended moderately, as it were, out of those two, and is a kind of selection of the best in each. This seems to me fitted to bear away the first prize, since it is a kind of mean—and the mean is virtue, in lives and in deeds, as Aristotle thinks and the others who philosophize after that school—but it is seen, as I said before too, not by exact definition but in breadth, and has many specific differences; and those who have used it have not all pursued the same things nor in like manner, but some these more, some those, and different men intensified and relaxed the same things differently, and all became worthy of note in all the kinds of discourse. The summit, then, and the goal of all,

from whom all rivers and the whole sea and all springs flow—

would rightly be said to be Homer. For with him every passage, wherever one lays hold of it, is embroidered to the utmost with both the austere and the polished harmonies. Of the others who pursued the same mean, those who are later than Homer would appear far behind him when examined beside him, but, if one should regard them by themselves, they are worthy to be looked at: of lyric poets Stesichorus and Alcaeus, of tragic poets Sophocles, of prose writers Herodotus, of orators Demosthenes, of philosophers, in my opinion, Democritus, Plato, and Aristotle; for it is impossible to find others who have blended their discourse better than these. And about the characters this is enough. For examples of these I do not think it necessary to bring, since they are quite manifest and need no discussion.

But if it seems to anyone that these things are worthy of much toil and great labour, he is quite right to think so, in the words of Demosthenes; yet if he reckons up the praises that follow upon them when rightly done, and how sweet is the fruit from them, he will count the toils a pleasure. The chorus of the Epicureans, however, who care for none of these things, I beg off; for the writing not being laborious—as Epicurus himself says—to those who do not aim at the standard, which shifts constantly, it was an antidote against much idleness and clumsiness.

25. These things being at an end for me, I think you still long to hear this: how unmetrical expression becomes like a fine poem or song, and how a poem or song becomes like fine prose. I shall begin first from bare prose, choosing one of the men whom I think to have copied poetic phrase most of all—wishing to choose more, but not having time enough for all. Come, then, who would not agree that the discourses of Demosthenes resemble the best poems and songs, and especially the harangues against Philip and the public forensic contests? Of which it will suffice to take the proem of one, this:

"Let none of you, men of Athens, suppose that I have come to accuse Aristocrates here, neither for the sake of any private enmity, nor that, seeing some small and slight fault, I so readily bring myself into hostility on this account; but if I reckon and consider rightly, all my zeal is about this: that you hold the Chersonese safely and be not again, deceived, deprived of it."

I must try to say what I think about these things too. They are now like mysteries, and cannot be brought out among the many, so that I should not be vulgar if I should invite those for whom it is lawful to come to the rites of the discourse, and should bid the profane shut the doors of their ears. For some take the most serious things for laughter through inexperience, and perhaps they suffer nothing strange. What I wish to say is this.

All expression composed without metre cannot take on poetic inspiration or lyric grace, at least as regards the composition itself—since the selection of words too has great power, and there is a kind of poetic naming, of glossematic, foreign, figurative, and coined words, in which poetry delights, when these are mingled to satiety into the unmetrical expression, which many others do, and not least Plato. But I do not speak of the selection; let the inquiry about these things be set aside for the present. Let the study be about the composition itself, which displays poetic graces in common and worn and least poetic words. As I said, then, bare expression cannot become like the metrical and tuneful, unless it comprise certain metres and rhythms set within it imperceptibly. It is not, however, fitting that it should seem to be metrical or rhythmical—for thus it will be a poem and a song, and will simply depart from its own character—but it suffices only that it appear well-rhythmed and well-metred; for thus it would be poetic, yet not a poem, and tuneful, yet not a song.

What the difference of these is, is very easy to see. The expression that comprises like metres, preserves ordered rhythms, and is carried through, line by line, period by period, or strophe by strophe, by the same figures, and then again uses the same rhythms and metres in the following lines, periods, or strophes, and does this for a long time, is rhythmical and metrical, and to such expression the names metre and song belong; but the expression that comprises wandering metres and disordered rhythms and displays neither consequence among them nor combination nor responsion, is well-rhythmed, since it is embroidered with certain rhythms, but not rhythmical, since not with the same ones nor in the same way. Such, I say, is all unmetrical expression that displays the poetic and the lyric; and such I say Demosthenes too used. And that this is true and that I am introducing nothing new, one might gain conviction even from the testimony of Aristotle; for the philosopher has said, among other things about civic expression in the third book of the Rhetorical Art, what its nature ought to be, and especially about good rhythm, from what kind such a thing would arise; in which he names the most suitable rhythms and shows in what respect each of them is useful, and sets down certain expressions by which he tries to confirm his account. But apart from the testimony of Aristotle, that it is necessary for certain rhythms to be comprised in prose, if the poetic beauty is to bloom upon it, one will know from experience itself.

For instance, the speech against Aristocrates, which I mentioned a little before, begins from a comic tetrameter line composed of anapaestic rhythms, but falls short by a foot of the complete one, owing to which it escapes notice: "Let none of you, men of Athens, suppose that I." For if this should take on the missing foot, either at the beginning or in the middle or at the end, it will be a complete anapaestic tetrameter, which some call Aristophanean:

Let none of you, men of Athens, suppose that I am present,

equal to:

I shall tell, then, of the ancient education, how it stood.

Perhaps someone will say to this, that this came about not by design but of itself; for nature improvises many metres. Let this be true. But the clause joined to this also—if one should dissolve its second elision, which has made it indistinct by linking it to the third clause—will be a completed elegiac pentameter, this:

For the sake of any private enmity whatever,

like these:

Maidens, lightly lifting the traces of their feet.

And this too let us suppose to have come about, according to the same spontaneity, without design. But, one clause in between being composed in the prose manner—"to have come to accuse Aristocrates here"—the clause again woven with this is composed of two metres: "nor that, seeing some small and slight fault, so readily on this account." For if one should take that Sapphic epithalamium, this—

For there was no other maiden, O bridegroom, such as this—

and take the last three feet and the cadence of the comic tetrameter, the so-called Aristophanean, this—

When I, speaking justice, flourished, and temperance was held in honour—

and join them in this fashion:

For there was no other maiden, O bridegroom, such as this and temperance was held in honour—

it will differ in nothing from "nor that, seeing some small and slight fault, so readily on this account." And what follows is equal to an iambic trimeter with its last foot removed: "to bring myself into hostility"; for it will be complete on taking a foot and becoming such:

to bring myself into some hostility.

Shall we pass over these too as having come about not by design but by spontaneity? What, then, does the clause close upon this wish again? For this too is an upright iambic trimeter:

but if I reckon and consider rightly,

the conjunction taking its first syllable long. And what of "and consider [kai skopo]," falling in between, by which the metre is overshadowed and obscured? The clause taken up after this is composed of anapaestic rhythms and proceeds to eight feet, preserving the same figure:

that you hold the Chersonese safely and not deceived—

like this in Euripides:

King of the land of many clods, Cisseus, the plain gleams with fire.

And the part of the same clause set after this—"be again deprived of it"—is an iambic trimeter falling short by a foot and a half; it would have been complete thus:

be again deprived of it in turn.

Shall we still say that these are improvised and undesigned, being so varied and so many? I do not think fit; for one can find the things following upon these too of the same kind, full of many metres and rhythms of all sorts.

But that no one may suppose this speech alone to have been so constructed by him, I shall lay hold of another again, of the one thought to be expressed quite divinely, the one On behalf of Ctesiphon, which I declare to be the best of all speeches. I see in this too, after the address to the Athenians, immediately the cretic rhythm—or, if one wishes to call it a paean, it will make no difference—composed of five times, not, by Zeus, improvised but woven as deliberately as possible through the whole of this clause:

To all the gods I pray, and to all the goddesses.

Is not that rhythm too of such a kind?

In Cretan rhythms let us sing the child.

To me, at least, it seems so; for, apart from the last foot, the rest are altogether equal. Let this too, if one wishes, be improvised; but the clause joined to this also is an upright iambic, falling short of the complete by a syllable, so that here too the metre may be indistinct, since, one syllable being added, it will be complete:

as much goodwill as I continue to feel.

And then that paean or cretic, the five-timed rhythm, will come in what follows these: "toward the city and all of you, that as much may be granted me by you toward this contest." For this seems, except that it has two broken feet at the beginning, in all the rest like that in Bacchylides:

It is no work of sitting still nor of delay, but to the gold-shielded Itonian we must go to the well-wrought temple and show some delicate thing.

I suspect a certain onslaught against these things by men inexperienced in liberal education, who pursue the vulgar part of rhetoric apart from method and art; against whom it is necessary to make a defence, lest we seem to have left the contest unfought. They will say this: "Was Demosthenes, then, so wretched that, when he wrote his speeches, setting metres and rhythms beside him like modellers, he tried to fit the clauses into these moulds, turning the words up and down and watching over the lengths and times and cases of the nouns and the moods of the verbs and busying himself with all the accidents of the parts of speech? Surely such a man would be foolish to give himself over to such elaborate trifling." These things, then, and the like, deriding and mocking, one might without difficulty repel by saying this: First, that there was nothing strange if the man thought worthy of such reputation as none of those named before him for power of speech, composing works for eternity and giving himself accountable to all-testing time, wished to take up nothing at random, neither matter nor word, but to have much forethought for both these—the management of the thoughts and the comeliness of the words—especially as the men of that time brought out their discourses resembling not painted but carved and chased works, I mean Isocrates and Plato the sophists. For the one composed the Panegyricus, as those who write the shortest time declare, in ten years; and Plato did not cease combing and curling his dialogues and weaving them in every way until, having reached eighty years, he died; for the stories told of the man's industriousness are surely familiar to all lovers of letters—among others, the one about the tablet which they say was found, after his death, with the beginning of the Republic arranged in various ways, having this opening: "I went down yesterday to the Piraeus with Glaucon the son of Ariston." What, then, was strange if Demosthenes too had a care for euphony and melodiousness, and for setting down neither word nor thought at random and untested? For it seems to me to befit far more a man constructing civic discourses as eternal monuments of his own power not to neglect the least of things, than for the sons of painters and engravers, in perishable matter, displaying the skill and toil of their hands, to wear out the precision of the art upon little veins and feathers and down and such minutiae. Using these arguments, then, one might, I think, claim nothing beyond what is reasonable, and might add this too: that, while he was a youth and just laying hold of the subject, it was not unreasonable for him to look round at everything that could fall within human practice; but when long exercise, gaining much strength, had wrought certain moulds and impressions in his mind of every thing practised, he then did these things from the readiest source and from habit.

Just such a thing happens in the other arts also, whose end is some activity or production. Thus those who know how to play the lute, to pluck the strings, and to play the flute to perfection, when they hear an unfamiliar piece, without much trouble reproduce it at once upon their instruments, simultaneously with the thought; but in learning they take up the powers of the notes by much time and toil, and their hands were not at once in the habit of doing what was enjoined, but only at last, when much exercise made the practice into the strength of nature, then they became successful in their works. And why need one speak of the others? What we all know is enough to cut through their whole nonsense. What is this? When we are taught our letters, first we learn thoroughly their names, then their shapes and powers, then in this way the syllables and their affections, and after this the words and their accidents—I mean lengthenings and shortenings and accents and the like; and when we have gained the knowledge of these, then we begin to write and to read, syllable by syllable and slowly at first; but when considerable time, coming on, has wrought strong impressions of them in our souls, then we do them from the readiest source, and read through whatever book one may give us without stumbling, with habit and speed. Just so one must suppose it to be also with the composition of words and the good flow of clauses among the athletes of the work. But for those inexperienced or unpractised in this to wonder and disbelieve if something is achieved with mastery by another through art, is not unreasonable. Against those, then, accustomed to scoff at the precepts of the arts, let this be said.

26. About tuneful and metrical composition, which has much likeness to prose expression, I have some such things as these to say: that the first cause here too, in the same manner as in unmetrical poetry, is the fitting of the words themselves; the second, the composition of the clauses; the third, the proportion of the periods. He who wishes in this department to succeed must turn and fit together the parts of expression in manifold ways, and make the clauses at intervals proportionately, not rounding them off with the lines but cutting the metre across, and make them unequal and unlike, and often gather them into commata shorter than clauses, and make the periods, at least those lying next to one another, neither equal in size nor alike in form; for that which wanders about the rhythms and metres appears closest to prose. Now those who construct epic and iambic verse and the other metres of like kind cannot divide their poems with many metres or rhythms, but must always remain in the same figure; whereas lyric poets may throw many metres and rhythms into a single period. So that those who compose in a single metre, when they break up their lines, dividing them by clauses now this way, now that, dissolve and do away with the precision of the metre, and, when they make the periods varied in size and figure, they cast us into forgetfulness of the metre; while the lyric poets, working out their strophes in many metres, and making the divisions of the clauses—each time again unequal and unlike one another—unequal and unlike, and through both these not allowing us to take hold of a uniform rhythm, wrought much likeness to prose in their songs; and, with figurative, foreign, glossematic, and the other poetic words remaining in the poems, it is none the less possible for them to appear like prose.

Let no one suppose me to be ignorant that the so-called prosaic quality is held to be a fault in a poem, nor condemn me for this stupidity, as though I were ranking a fault among the virtues of poems or discourses; but let him hear and learn how I think one ought to distinguish, in these things too, the excellent from the worthless. Knowing that of discourses one is that of the ordinary man—I mean the babbler and trifler—and the other the civic, in which there is much that is constructed and artful, whatever of poems I find resembling the trifling and babbling discourse I count worthy of laughter, but whatever resembles the constructed and artful I think fit to receive emulation and zeal. If, then, each of the discourses happened to have a different name, it would follow to call by different names each of the poems that resemble them; but since both the excellent and the worthless are alike called discourse, one would not err in counting the poems that resemble fine discourse fine, and those that resemble bad discourse bad, in no way disturbed by the common name of discourse. For the likeness of the naming, applied to different things, will in no way prevent one from seeing the nature of each.

Having now spoken about these matters too, I shall close my discourse, setting down here a few examples of what has been said. From epic poetry, then, these will suffice:

But he from the harbour went up by the rugged path—

this, then, is one clause. Another:

over a wooded place—

shorter than the former and cutting the line in two. A third, this:

through the heights—

a small comma, less than a clause. A fourth:

to where Athene showed him the noble swineherd—

composed of two half-lines and in no way like the former. Then the last:

who, of his substance, most of all cared for the household which noble Odysseus had won—

making the third line complete, but the fourth deprived of its precision by the addition. Then again:

But him in the forecourt he found seated—

this too not running out with the line.

There he had a court high-built—

this too unequal to the former. And then the following sense is unperiodic, spoken in clauses and commata. For having set down—

in a conspicuous place, he will again add:

fair and large—

a comma shorter than a clause; then:

running round about—

a word having by itself some sense. Then in order he will construct the rest in the same manner; for why need one be prolix?

From iambic poetry, these of Euripides:

O fatherland, which Pelops bounds, hail—

this is the first clause, up to this point.

you who tread the wintry rock of the Arcadians—

the second, up to this.

thence I claim my birth—

this is the third. The former were larger than a line, this one less.

For Auge, daughter of Aleus, bore me secretly to the Tirynthian Heracles—

after this:

Mount Parthenion knows it too—

neither of these measured by a line. Then again another, both less than a line and greater than a line:

there my mother, released from her travail, Eileithyia loosed—

and the things following these are similar.

From lyric poetry, these of Simonides. They are written according to divisions not those which Aristophanes or anyone else constructed in clauses, but those which prose demands. Attend, then, to the song and read it according to the divisions, and be assured that the rhythm of the ode will escape you, and you will not be able to make out either strophe or antistrophe or epode, but it will appear to you one continuous discourse. It is Danae, borne over the sea, lamenting her own fortunes:

When in the carven ark the wind blowing bore her, and the heaving sea, she shivered with fear, her cheeks not unwetted, and threw her loving arm about Perseus and said: "O my child, what trouble I have; but you sleep, and in your milk-fed temper you slumber in this joyless brass-nailed timber, sent forth in the night unlit and in the dark-blue gloom. The deep brine over your hair, as the wave passes above, you heed not, nor the wind's voice; lying in your crimson cloak, your fair face against my breast. But if to you the dreadful were dreadful, you would have lent your delicate ear to my words. I bid you sleep, my babe; let the sea sleep, let the measureless evil sleep. And may some change of purpose appear, Father Zeus, from you; but if any bold word I pray, beyond justice, forgive me."

Such are the metres and songs that resemble fine discourses, coming about for the causes I told you before.

This you shall have as my gift, Rufus, worth many others, if you should be willing to keep it continually in your hands, like anything else of the very useful kind, and to train yourself in it by the daily exercises. For the precepts of the arts are not of themselves sufficient to make able combatants of those who wish to do without practice and exercise; rather it rests with those who are willing to toil and endure hardship whether the precepts shall prove serious and worth discourse, or paltry and useless.