“Ciceronian prose is practically the prose of the human race,” says Mackail, a critic who is unusually sensitive to qualities of style. In saying this, he doubtless had in mind not only the orotund periods of the Pro Milone, the elaborately rhythmical movement of the Pro Archia, the vehement force of the first Catilinarian or the easy colloquialism of the familiar letters. It was rather the lucid and copious exposition of essays like the De Oratore in which, without revealed effort, a versatile mind found appropriate and dignified expression for all its concepts and moods. How did such prose come to be?

Cicero worked incessantly for years to acquire his command of the tools of expression. When very young he memorized the standard rules of rhetoric that emphasized the need for clarity, arrangement, conciseness, luminosity, and all the rest. Do such rules make great writers? On that point Cicero did not deceive himself. He knew that adults did not need them, but he recognized that schoolboys would save time by having their attention called to what practice would eventually reveal. Such rules might prove guide posts to intelligent beginners, but one has only to read the three books of the De Oratore[1] to discover that rhetoric was for Cicero a schoolroom crutch to outgrow and forget. Another device much recommended by the Roman teachers of his day was imitation, the study of the masters of diverse styles. It is a method that has recently been employed to good effect in the classroom for the awakening of taste and sharpening of critical acumen. Cicero did not scorn its use, but he knew too well that style is personal[2] to attempt to acquire in this way a garb that would not fit his own mental processes. When he sought out Apollonius of Rhodes as a critic it was not in order to adapt himself to that teacher’s mode of expression. He first decided what his own taste and capacity needed, and what the Roman Forum and Curia would require of him; then he sought for the teacher who could best help him by his criticism. His complete independence is shown by the fact that he traveled long, trying one after the other of the famous teachers of the east, abandoning them one by one as soon as he discovered that they did not suit his purposes.[3] Cicero did not impose the Rhodian style upon himself. He made his own curriculum to fit his temperament and sought out the tutor who could help him attain what he demanded. This procedure seems to me characteristic of the great Roman stylists. Cicero and Caesar, Sallust and Livy, Seneca and Tacitus, betray themselves in their sentence structure. The secret of their expression will never be disclosed by a search for their models nor in the rhetorical rules then current.

The aim of this chapter is limited. It cannot even attempt the important task of illustrating from a study of Cicero the valid rule that style is the man. It will attempt only to sketch the growth of Latin prose up to Cicero’s day in order to suggest how that prose became adequate to clothe the varied expression of so versatile a genius.

Roman, like English, prose developed its sonority, dignity, and rhythm in persuasive speech. As in England during the religious reformation, pulpit oratory molded speech, so in Rome, during the period of political reformation from Cato to Cicero, forensic contests in the senate house and at the tribunals transformed Roman expression. This parallel may seem obvious, but one offers it with hesitation because Roman oratorical style is generally supposed to have been shaped by the study of Greek rhetoric in the schoolroom. Quite apart, however, from the fact that true art is seldom amenable to the compulsion of precept, chronology militates against this theory. Roman prose had traveled far before it resorted to any guidance from Greece.

Like the English of Wycliffe, early Roman prose was formless. It merely followed the habits of unshaped spontaneous conversation. If anything was to be recorded with care, it employed the forms of art, that is, of verse. Naevius and Ennius wrote their chronicles in meter. Even Chaucer, who is so luminous in his verse narratives, becomes involved and at times almost incoherent in his few attempts to write prose, unless in fact he yields to the temptation of admitting rhythm into his sentences. But Chaucer is one of the last of the great writers to flounder thus. The Wyclifite Bible marked the beginning of a religious contest that continued for two centuries with more or less intensity, and finally with passionate vehemence. It was a contest that, to many, involved a question of life and death and to even more the problem of eternal salvation. The gravity of the theme called for the noblest possible expression, while the deep concern of all classes, even the most ignorant, required clarity and directness of utterance. The temptation of the learned to exaggerate rhetoric into Euphuism was immediately checked by the need of being intelligible to the congregation, while the tendency of plain persons toward colloquial formlessness was checked both by the deep respect for the sacred theme and by the high level of cultural taste among the clergy of the time. We need not deny the great influence of Ciceronian and Augustinian models upon these learned men, and in Lyly’s courtly group we know how ancient rhetoric ran pell-mell into preciosity. But that was an aberration that affected only those who had a thin message to convey. When men are intensely engaged in saving their fellows, speech will grow clear, and when these men are at the same time persons endowed with great intellects, their speech will take on dignity of structure and of sound. Before that contest English prose had babbled thus:[4]

This sentence rattles on unhaltingly through “and’s” and “that’s” for a solid page before it falls down to a stop from sheer exhaustion.

After the battle was over we have the Authorized Version with its magnificent directness:

Briefly, the parallel between early Latin and early English prose can be indicated thus. The prose of Ennius, like that of Chaucer, was very much inferior to his verse. Before Ennius died, however, the mighty struggle of statesmen had begun in the polemics of Cato, a contest which was destined to build up in time a dignified and versatile language. Cato represented the native, middle-class, agrarian population of Italy that feared the expensive and ambitious foreign entanglements which the philhellenic party of Scipio had incurred and hated the foreign culture which followed in the wake of philhellenism. Cato spoke incessantly. A hundred and fifty of his speeches were available in Cicero’s day. He attacked the Scipionic group in the senate, in public harangues, and in court. And not only he but his lieutenants—and of course his opponents—had constantly to be on their feet. This was the beginning of the party divisions that led through the Gracchan reforms and through the debating period of the civil wars to the final defeat of the Roman Republic a hundred years later. The contest of words was as bitter as in the England of Wycliffe, Tindall, Cranmer, Latimer, and Hooker. Here, too, the best intellects of the nation were exercised in the debate; here, too, the gravity of the theme and the demands of aristocratic audiences required dignified expression, while the constant necessity of winning the populace required entire clarity and lucidity of expression. The struggle was not indeed for eternal salvation, but it often involved the question of life and death, and always the future of the state. And from men like Cato, the Gracchi, Cicero and Brutus, the state claimed and won a devotion more intense than religion could. Thus there is a certain similarity between the growth of Latin prose from Ennius to Cicero and that of English from Chaucer to Hooker. And though Greek rhetorical theory and models were factors in shaping Latin prose, as Roman theory and models were factors in shaping English, it seems to me quite probable that both languages would have taken the course they did without those models, for both were determined by forensic expression, by great causes, and by intense devotion to those causes on the part of the most intelligent men of their day.

In following the evolution of Latin prose[5] we unfortunately have to deal largely with fragments quoted by later writers, and we cannot always be sure that these fragments are representative. For our purposes however they may legitimately be considered so. Before Ennius’ time very few speeches had actually been published. Cicero had at hand an old oration of Appius Claudius of about 281 B.C. and some funerary laudations, but he did not think either worth considering in a history of oratory. So far as we know, written prose documents before these were confined to laws, treaties, and meager official records. The fragments of the Twelve Tables (450 B.C.) are too scanty to afford any basis for judging style. Some of them are so wanting in lucidity, because of an ambiguous use of pronominal subjects, that a modern lawyer might readily manipulate them to prove any point. A few fourth- and third-century inscriptions from headstones and votive tablets[6] contain only blunt sentences which reveal chiefly an obvious desire to save the expense of stone-cutting. They do however show the native Latin word order and its fondness for the deferred verb. Orcevia Numeri (uxor) nationu gratia, Fortuna Diovo fileia primogenia donum dedi. This is of course a tendency in all inflected languages where the verb can be postponed till the subject and object have been visualized, since the inflectional endings indicate the direction of the verbal action. And in Latin, the hierarchy of what is important can be and was recognized by the word order. “Orcevia, Numerius’ wife, for the gift of childbirth, to Fortuna, Jove’s daughter the firstborn, this gift I give.” Strictly speaking Cicero’s best-shaped sentence is not more periodic than that colloquial tablet of a humble woman a century before any Roman scholar thought of studying style. It was not the study of Greek that determined the form of Latin prose.

The Duilian inscription of 260 B.C.—doubtless authentic in the main though found in an imperial copy[7]—is our only pre-Ennian fragment of prose that contains several complete sentences. This inscription is far more fulsome and boastful than the modest Scipionic epitaphs of two generations later, a fact probably due to Duilius’ sojourn in Sicily where he could see verbose honorary tablets at every hand. In spirit and content it is Sicilian, but its phrasing and diction are normal Latin. Its longest sentence is rambling, badly coordinated and illogically constructed despite its periodic placement of the verbs. The man who composed it had no feeling for lapidary style:

A man who composes thus is not only “hypnotized by the exuberance of his own verbosity” but unpracticed in the art of logical expression.

Our first passage of continuous prose comes from Ennius’ Euhemerus, quoted verbatim by Lactantius. A fair example is the following:[8]

For this passage I shall use Professor Rand’s translation though it introduces a modicum of style into the expression:

This Ennian passage is even more simple and devoid of stylistic qualities than is the English of Wycliffe or Chaucer. The brief plodding sentences are clear enough; in fact there is a dry legalistic explicitness in phrases like id ejus rei causa fecit uti, and deinde posterius. But the whole rattles to pieces like a mosaic set in clay. It is in the main a string of coordinate clauses loosely hung on que, atque, ibi, tum, and without any appreciation of the differences that we attempt to convey by commas, semicolons and full stops. It has not even the normal feeling for periodic structure which the epitaphs of the time reveal. It is naïve, primitive prose, and the evidence that Ennius could drivel thus is indeed illuminating to the student of literature. A nation which could be satisfied with such a medium of expression had not been very verbose.

During the next few decades, however, there was much legislation, and from the interesting Senatus Consultum de Bacchanalibus of 186 B.C. we have considerable fragments which prove that the ambiguities found in the Twelve Tables were being gradually removed and that there were enough shysters at Rome to compel legislators to evolve the intricate and all-inclusive “if-and-but style” which has ever since characterized legal expression. To this source the great prose of Rome owed very little except precision of diction. There was also not a little historical writing, chiefly however in Greek, for the use of statesmen who needed to know their precedents. But this type of prose, so far as we can judge from the fragments preserved and from Cicero’s adverse judgments,[9] made no appreciable advance upon the narrative manner of Ennius, illustrated above. Nor did such commonplace textbooks as Cato’s De Agri Cultura.

As we have said, it was public speech that moulded prose style at Rome, as in England. Among the first to make a marked impression was Cato, whose great activity on the platform begins about the year of the decree de Bacchanalibus. Nothing could be more innocent of form than Cato’s De Agri Cultura. This however, is by no means true of his speeches, several pages of which survive in the typical paragraphs quoted by later writers. Cato had not taken any course in the art of eloquence, he had not studied the Greeks to the point of appreciating stylistic qualities, and there was no literary Latin prose published for him to study, but he had, as a member of the senate, heard many elaborate arguments advanced by the foremost statesmen of his day on such weighty questions as the peace with Carthage, the proposed expedition into Macedonia in aid of the Greek democracies, the terms of peace with Philip, and the proposed war with Antiochus the Great. There can be no doubt that these debates brought out many of the characteristic qualities of Latin style. The men who argued these questions had to think soundly and to form their arguments as clearly, as definitely, as incisively, and as persuasively as they knew how. When scholastic students of style attribute Greek learning to Cato[10] because he stops to make definitions, balance arguments, and employ logical enthymemes, they astound us by their naïvete. One might as well say that Confucius, Hesiod, and Isaiah had studied Demosthenes. Indeed I doubt not the Aurignacian mother defined words for her children and that the lord of the cave had often tried to argue his wife into silence by conclusions ex contrario.

There has recently arisen another explanation for the occasional artistry of the pre-hellenistic Roman writers which has been held to apply to all of the early Latin authors including Plautus. This is that the so-called “Gorgianic figures,” used by even the earliest Romans, are of Sicilian origin, that the Romans must therefore have come into cultural contact with the Sicilians through commerce two centuries before Plautus, and that Latin prose may thus have taken on rhetorical devices in its infancy.[11] I mention these entertaining conjectures only to guard against any possible supposition that they may seem acceptable simply because they have found their way successively into recent textbooks. Cato was a man who, despite his faults, possessed a very keen and versatile mind, a visualizing, picture-making imagination, a sharp tongue, an agile as well as a retentive memory, and a penetrating power of analysis. His style, to be sure, is not malleable; the clauses cohere by logic rather than by the cement of conjunctions; he is repetitious, chiefly because he likes to hammer his nails firm; his transitions are blunt when he is impatient to be on with his argument; he does not take time to modulate his phrasing and his style has little chiaroscuro, because he is in deadly earnest all the way. His vocabulary is often of the barn and field and his imagery is apt to draw from the farmstead, as for instance when he shouts at Thermus: “You cut those ten worthy men into strips of bacon.”[12] In his Brutus, Cicero somewhat slyly likened Cato’s simple straightforward Latin to the style of Lysias.[13] Cicero, of course, knew the difference, for he later permitted Atticus to correct him on this point, but at the time he desired to recall Brutus to the logical consequences of a contemporary doctrine which somewhat naïvely overstressed the simplicity of the studied artlessness of Lysias. Cato was, of course, conscious of his effects; he drove his arguments home with intentional care, for he wrote out his speeches even though he delivered them without notes. He published them of course not as literary essays to be read by later students of oratory, but as documents designed to carry on the battles that he had begun in the court or the senate. Their art, such as it is, derives not from rhetoric but from his temperament and his fiery conviction. His philosophy of style lay in four words: rem tene, verba sequentur.

Cato’s prose was admirably suited to forensic attack. Its qualities, however, were those that spring from a practical, quick-witted, imaginative debater. Cato probably directed every word and every clause toward the precise argumentative effect that he wished to obtain. He did not pronounce them slowly in order to taste their harmony of sound or to listen to their rhythm. If they had beauty, it was by chance or by reason of the beauty inherent in the Latin of his day. He probably deleted whatever created the impression of being far-sought. Spontaneous imagery might stand if it made his meaning more clear. His antithesis, anaphora, and balance therefore belong not to the schools but spring from the instinct to strike quickly, often, and with both fists. During his fifty years of strenuous speaking he did much for Latin prose, by proving that it could be clear, pointed, and precise; that it was adapted to senatorial deliberations over world politics, as well as to legal battles in the courts and in the assembly. Cato did not have an ear for the organ qualities of the language. Nor was the time yet ripe for the elaboration of artistic effects. When Cato spoke with deepest earnestness, he could hardly escape attaining to some of the dignity that Latin speech so readily acquires, but his vocabulary was too fresh from the soil to sustain that quality for long. However, it is likely that men of taste and restraint even in his day were more concerned than he for the proprieties of diction that belonged to themes of gravity. Nobles who were learning to rule provinces the wide world over and to give commands to kings did not have to go to Greek pettifoggers to acquire dignity of address.

Toward the end of Cato’s period some nobles kept Greek teachers in their homes to teach their sons the language and the literature that prevailed in all the eastern half of the Empire. But the spirit of Rome was not then very friendly toward such teachers. The interminable wrangling of scores of Greek legations begging for favors, the disillusioning visits of Roman statesmen to Greek cities, the demoralizing influence of the country upon the soldiers stationed in Greece, the inane display of logical antinomies in the philosophical disputations, and the superficiality of a rhetorical doctrine concerned with adornments superimposed upon vacuity, these very quickly disgusted Rome. Cato’s friends succeeded in having the Greek teachers banished from Rome in 161 and again in 154.[14] It would be as great a mistake to attribute lasting cultural effects to the ambassadorial visits of Crates and Carneades to Rome as to assume that the American senate could have adopted continental rhetoric and style from the exuberant prose spoken by the French and Italian envoys, Viviani and Francesco Nitti, who were sent to Washington in 1917 to present the cause of the allied nations.

After Cato’s death more Greek teachers came, and among these the stoic Panaetius, who remained for some time and became a real cultural force in the group that gathered about the younger Scipio. Some attempt has been made to trace the Stoic rhetorical doctrine of the plain style to this contact.[15] But it is difficult to see what lessons Rome needed after Cato to illustrate the desirability of the qualities emphasized by stoic teaching: (1) pure diction, (2) clarity, (3) precision, (4) conciseness, (5) propriety. The first four of these qualities were the very spirit of Cato’s practiced though untutored Latin. The last quality concerned Cato very little in all probability, but other Roman statesmen knew the need of sloughing off barnyard diction in speaking before the august senators at Rome. Propriety of diction is after all a quality that could hardly be foreign to a people who had for centuries respected the triumphal garb, the fasces, and the august pontifical ceremonies, and it was not a quality that could be acquired from foreign teachers who did not know the tone of Latin words. We must also bear in mind that what Roman statesmen were eager to learn from men like Panaetius and Polybius and what these men desired to teach was not some clever trick of phrasing but the essence of political philosophy and of ethics. Polybius’ sixth book and Cicero’s De Republica and De Officiis are the real results of the early Stoic teaching at Rome, and Polybius’ own unwieldy sentences should warn us that contact with Stoic teaching could do little for stylistic beauty.

As the Gracchan times approached, a new division of parties became apparent at Rome. The senators were suspected of promoting expansion in the provinces for the sake of their own profit and glory, and several tribunes gained popularity by opposing the recruiting and by haling nobles into court on the charge of maladministration. Piso devised the first of the special courts, which Cicero considered of great importance for the training of orators. Then for several years there was agitation over ballot reform advocated by the populace who desired a secret ballot. Many important speeches were delivered in the senate and before the people on these measures, and if we may judge from the remarks of Cicero regarding Galba,[16] Lepidus Porcina, and Scipio Aemilianus, all this activity conduced to create a feeling for a smoother and more coherent style. Aemilianus especially, who represented the finest aristocracy in its dignity of birth and high accomplishment, spoke with that auctoritas and gravitas that were the natural concomitants of great empire.[17]

Then came the Gracchan proposals which shook the staid government to its foundations. For a dozen years the keenest minds of Rome were pitted against each other, and victory lay not in arms but in the power of persuasion. There was much discussion in the senate, but Tiberius Gracchus carried the battle directly to the popular assembly, and that is where it was fought to the end. For the words of Tiberius we have to rely chiefly upon the paraphrases of Plutarch, which are too general to permit of accurate estimate. From the speeches of Gaius Gracchus, however, we fortunately have some exact quotations.[18]

Gaius Gracchus did as much as any one man to increase the range of Latin forensic prose. Reared in the center of the dominant aristocracy where he imbibed the purest and most copious diction, trained by a mother whose urbane language delighted even Cicero, he nevertheless espoused the cause of democracy, and in the defense of this principle he acquired a lucid directness that Cicero never tires of praising. Gracchus had Greek teachers, who taught him to read and to speak Greek as well as not a little about Greek political ethics, and doubtless also the textbook rules of Greek style. Such stylistic rules, however, were not of much worth in addressing Roman voters, and they are seldom in evidence in the fragments which have survived. Cicero’s one criticism of Gracchus’ style is that he did not know how to modulate his prose so as to secure rhythmical effects. Gracchus would not have attempted to secure them had he known how. He was too concerned with the issues at stake, too fired with zeal for the cause for which he was to suffer death, to worry about the adornments of style. He published his speeches, and he doubtless prepared them beforehand, because in the revolutionary reforms that he proposed, errors of phrasing must be avoided, and the record must be kept for the sake of impressing his arguments. He certainly did not issue his speeches with a view to providing models of style.

In this period Latin prose acquired further versatility and range because Gracchus was a man of genius, believed in a cause which gave full scope to his great powers, and spoke before different audiences that required of him widely varying types of appeal. His was the task of shattering the power of the most stubborn aristocracy that the world has known, of organizing a new democratic machinery of government, of extending the suffrage throughout Italy, of saving the native stock by a vast scheme of colonization. He was stirred by an unflinching devotion to his cause, by bitterness at the murder of his brother, and by the knowledge that he too was marked for death. It is blasphemy against the informing spirit of great art to attribute his effects to rules, and not to acknowledge that genius fighting in such a cause is an independent creative force. Cato had already shaped his weapons for him. Gracchus, more richly endowed with vision, with sympathy, with intellect and wit, filed and hammered the weapons into a finished armory. There is no tool of persuasion that he did not have to employ. He used a simple, rude, staccato narrative when explaining before the rabble conditions that must be cured;[19] in elaborate argument, where the light and shade must be exact, he employed periods as well-packed, though not so musical, as those of Cicero.[20] Before senators his diction was as grave and lordly as theirs, while in the forum, though never coarse, he could be as colloquial as Cato. His vituperation carried the deep thrust of the lance rather than the rapier cut,[21] for he liked to play with lingering irony. His emotional appeal reminds one of the language of Ennius’ tragedies.[22]

And yet Cicero was not quite satisfied with his speeches as works of art. What was lacking was after all what Gracchus would have disregarded even had he lived in Cicero’s day: a more careful modulation, a studied use of rhythm, a concern for the collocation of sounds, a more elaborate sentence structure, and a more apparent contrast of light and shade. Those are qualities that do not belong to the expression of revolutionary reformers who have but a year or two in which to perform a great life-work. They must come with leisure and tranquility when men have time to try the sound and taste of phrases in patient reiteration. Meanwhile Latin prose had been fortunate in finding men like Cato and Gracchus to make it vivid, clear, versatile, and vibrant. After these two men and the scores of speakers that they drew into the arena,[23] no Roman could again write Latin with the shambling gait of Ennius’ Euhemerus without serious apology. And it is safe to say that even writers of history and autobiography, who became numerous in the Gracchan period and after, comprehended now that sentences must have clarity, unity, logic, and precision.

After the death of Gracchus there was a temporary lull in politics; the victorious aristocracy, so nearly crushed, prudently decided to compromise with the populace rather than to risk the awakening of another Gracchus by exploiting their restored power. Young men who had heard the brilliant reformer in their youth, men like Crassus, Antonius, and Catulus, grew up to be distinguished orators. They inherited the results of a great evolution of prose, they directly or indirectly received the benefits of a deeper respect for elaborate style because of a new contact with Greek teachers, and they were granted the leisure and tranquility to consider the needs of a more artistic expression.

Licinius Crassus, in whose orations Cicero found the first mature Latin prose,[24] began as a partisan of Gaius Gracchus and in his youth doubtless imitated his hero’s fiery style. He also gave some attention to the Greek rules though he held that rules did not create style but were merely a collection of deductions drawn by analysts from the practices of the eloquent.[25] He preferred observing Roman speakers to studying the precepts of the Greeks,[26] and he thought Roman oratory sounder than Greek because at Rome the pleaders were the foremost statesmen whereas in Greece only hirelings practiced the art.[27] In these views he was not far from representing orthodox opinions.[28] There were other great men who gave even less credit to scholastic practice. Antonius his rival—by many considered the more brilliant speaker of the two—claimed that rhetoric was useless in that it only formulated the obvious;[29] Scaevola pointed out that Roman statesmen who had brought Roman government to the pinnacle of glory had nothing to learn in expression from inexperienced Greek pedagogues;[30] and Cicero’s account of the style of such great orators as Sulpicius, Caesar Strabo, and Cotta reveals the fact that the oratory of these men was a home product.[31]

On the other hand there were men who tried to make up for the deficiency in practical experience by drilling at doctrine, with the usual result that their language became tangled in artifice. Men like Albucius[32] and the first Curio remind us in type and experience of the courtly Tudor wits who had little to do or say and ended in euphuism.

What was the admirable style of Crassus which Cicero now calls mature? The samples that have been saved for us by the Auctor ad Herennium unfortunately were quoted to illustrate vivid and rapid-fire argumentation, and Cicero’s longest quotation was made to indicate Crassus’ power of spontaneous reaction to a surprising situation. While these examples give proof of celerity of wit, of a forceful, picturesque, and copious diction, of the pungent and concise phrasing for which Crassus was noted, they are not normal forensic prose. They do not reveal the dignity and harmony for which this orator was praised, and they give no certain illustration of the prose rhythms that Cicero liked to find in a “mature” style. From the passages that we have we should say that Crassus spoke as a pupil of Gaius Gracchus, but with the mellowness of age and in causes of less moment.

Perhaps the real reason why Cicero found Crassus’ style mature was that the Latin language was now mature. Latin diction had now become fuller and richer. Not only had the large bulk of Accian tragedy and of hundreds of comedies enriched the language, but hundreds of speeches delivered by men who had worked hard at the task of enlarging the resources of Latin phrase and diction had now been published. The special court instituted by Piso, the frequent cases before the plebeian assembly after the Gracchan period, the new custom of attacking political opponents by means of legal prosecutions had immensely increased the scope of oratory. The factional strife introduced by the Gracchans had divided the senate into debating groups, and brought fire into electioneering oratory and into legislative discussions. Every phase of political philosophy and expediency as well as of legal and moral principles was discussed day after day. Accordingly, the Latin language matured quickly and its prose was a finished product by the time that Cicero was born, although its verse had to wait another century before attaining adequate expression.

This prose was fortunately a fairly musical thing by nature. In comparing the earliest Latin word-forms with those of the Gracchan days we find that they had improved very much in musical quality, due in part, no doubt, to the fact that the Etruscans and Sabines, who had temporarily dominated Rome, had slurred over harsh collocations of consonants till they fell away, and partly to the fact that the plebeians, who were of course less conservative in speech than the patricians, had won great positions in the fourth century. Jouxmenta of the Stele inscription had now softened to jumenta; stlis had become lis; stlocus, locus; forctis, fortis; scandsla, scala; and so on, in hundreds of words. In many positions the harsh sibilants had been eliminated: cosmis had become comis; dusmos, dumus; and intervocalic s had become r: eram was better than esam. This elimination of harsh sounds had wrought so effectively between 500 B.C. and 100 B.C. that a language that was once as rough as Gothic had acquired the mellifluous quality of Italian. Though it still contained too many sibilants for ideal speech and the final m occurred so frequently as to invite monotony, it had few sounds that could jar upon the most delicate ear. The vowels were relatively pure, and because of the abundance in inflections of the sonorous vowels a, o, and u (=oo), they gave the language an orotund quality. The Indo-European i is on the whole apt to be shrill, and the great vowel shift of sixteenth-century England which altered it to the much more musical i (=aye) undid its benefit by raising English e to the thinner sound of the old i. Latin retained the old sound, but in i-stems it frequently went over by analogy to e, and the a̅i̅, a̅ĭ diphthongs fortunately softened to the mellow æ. In all this, mechanistic forces of the speech organs were at work, but one cannot help thinking that a delicate auditory guidance helped select the desirable sounds.

Another great advantage inherent in the Latin language from the beginning was that quantities were carefully observed by it and were in fact the determining factor in its rhythm; and since time rather than stress is the guiding principle of music in human song, as in the flute and organ, Roman speech was to an unusual degree suited to modulated utterance. To be sure, in the century before Plautus, stress had threatened for a while to gain dominance in vulgar speech—enough in fact to question the rights of measured verse—, nevertheless the timely spread of the conservative, aristocratic pronunciation through political and forensic oratory, as it was heard almost daily in the open forum during the second century, gradually checked the process and standardized a precise observance of longs and shorts.

The emphatic dominance of quantity over speech went so far in controlling word-accent that about two centuries before Cicero it had drawn the accent to the penultimate vowel if that was long. Hence, in the sentence endings which so often consisted of weighty words, word-accent to a remarkable degree coincided with a natural quantitative utterance. Latin, therefore, lent itself to a rhythmical close of sentences, often combining word-stress and length of utterance in a way that Greek prose rhythm did not. Cicero had studied Greek and had observed that various writers advocated the use of iambs, dactyls, and paeans[33] for clause-endings, and he labored somewhat confusedly to justify those rhythms since Greek theory seemed to demand them, but modern analysis has proved that his ear had shaped a truer Latin rhythm than his scholarship or his logic. His favorite clausulae, though he was not fully aware of it, were cretics and trochaics, producing a rhythm that adapted itself excellently to the dominance of longs, to the penultimate law, and to a strong close. As usual, a true appreciation of the genius of the Latin language saved the art from the effect of rules that were made for another medium. Here again Latin shows its independence.

But this is not all. Cicero’s books of rhetoric emphasize periodic sentence structure with careful attention to a mobile arrangement of clauses within the period. The Greek orators had of course practiced this art, and the teachers had drawn up the rules of the game afterwards. Cicero, for instance, often patterns his clauses with care in order to reach a periodic climax. In the Orator[34] he quotes an example from a speech of his own in which he follows two pairs of balanced phrases and a pair of clauses with a tranquil dignified close.

In such studied prose as in much of our free verse, the modulation depends not only upon the measured clausulae but also upon the parallelisms of phrase.[35] It is the two-fold rhythm that we so often find in the Authorized Version, in Hooker and in Browne, before English writers knew very much about the classical theories of prose rhythm. Now the point that needs to be emphasized is that Cicero would probably have written thus had he never known rules, had he only used with his infallible ear the prose that came to him shaped by a hundred great speakers. For, in the first place, the periodic structure was native to Latin, as we have seen, from the time of the earliest inscriptions. That structure is natural in highly inflected languages where the verb can be deferred in order to make room early for the important words and concepts, while unimportant phrases can be appropriately subordinated because their inflectional forms keep them tethered in thought to their owners even though separated by space. All this invites the service of taste to provide the contrast and balance, to give light and shade, to lift and to subdue, and then to bind the whole between introductory subject and concluding verb. No speaker of taste, given leisure and rich diction, could resist the temptation of thus elaborating such a language as Latin. The sentence of the untutored Cato, quoted above, though lacking in modulation, reveals a structural form not unlike the sentence of Cicero just cited.

Cicero repeatedly calls attention to what he designates as the adornment of good prose, adornments associated in Greek learning with the name of Gorgias. These are the tropes, i.e., the figures of speech, and the schemata, i.e., the patterned expressions of sentences. But he also tells us, fortunately, that there were none of these adornments which could not be found in the works of untutored old Cato,[36] and that even unschooled rustics employed metaphor. We have already remarked how modern scholars have sought to explain their presence. Explanations are of course not necessary. Men used metaphor and simile in the caves of the Dordogne 20,000 years ago; language began in metaphor when the primitive savage first called a dog “bowwow.” Half the words of any language are still metaphorical. When a Roman tried to find some expression for thinking, whether he used puto or intelligo or concipio or cogito or arbitror or existimo or opinor or censeo, or sentio, he had to use a figure of speech. Men like Cato, Scipio, Gracchus, Cicero became powerful because they had imagination, saw visions, and put their visions into their words.

The same may be said of patterned phrases. Native Latin verse, shaped long before Greek was known at Rome, was particularly fond of balance and antithesis because it was a verse that rested on parallelism marked by the strong caesura and bound together by alliteration. Such was the form of the early prayers and proverbs of the Romans: