Medieval rhetoric and poetic

CONTENTS

ABBREVIATIONS RECURRING IN THE NOTES

CHAPTER I THE SOPHISTIC TREND IN ANCIENT RHETORIC — References and Abbreviations — A. The Two Historic Conceptions of Rhetoric — B. The Second Sophistic [10] — 1. Philostratus, The Lives of the Sophists [11] — 2. The Character of Sophistic — FOOTNOTES:

Plato’s distrust of rhetoric is a permanent reminder. It is so significantly typical that it recurs throughout the history of education, and must recur. Again and again educational practise has found that it cannot do without rhetoric; again and again educational theory has grudgingly inquired what to do with it. For distrust of rhetoric may be more than the impatience of the philosopher with the orator, of speculation with ordered presentation, of the quest for truth with persuasion. This is involved too; for philosophers have often been impatient of presentation. They have wished to think aloud, or to question and answer, or merely to analyze for themselves, without being held to consecutive explanation. Plato himself falls short of discerning the importance of making truth available and effective for the mass of men incapable of scientific analysis; and this is the ground of Cicero’s rejoinder.[1] But Plato’s distrust is more deeply of rhetoric as he heard it taught. Even in his day Greek rhetoric was largely sophistic, the rhetoric of personal display and triumph. In the Gorgias[2] Socrates admits the function of a nobler rhetoric, but cannot find it in use. In the Protagoras[3] he asks the vital question, “About what does the sophist make a man more eloquent?” In the Phædrus, which discusses rhetoric more specifically, his satire is most evidently of sophistic. To this the ultimate objection is moral. A man should train himself “not with a view to speaking and acting before the world, but for the sake of making himself able ... to please the gods.”[4] Plato challenges not merely the method of the sophists, but their ideal. Since rhetoric has almost always had some part in education, and since it always ultimately involves morality, Plato raises a leading question.

The ultimate, the only final answer to Plato’s challenge is the Rhetoric of Aristotle. This proceeds from a conception not only larger than the sophistic of Gorgias and Protagoras, but also significantly divergent in aim. The true theory of rhetoric as the energizing of knowledge, the bringing of truth to bear upon men,[5] is there established for all time. Aristotle amply vindicated rhetoric by defining its place among studies, its necessary correlation with inquiry and with policy, its permanent function. He settled the question of rhetoric philosophically. He established its theory. But this theory was oftener accepted than followed. The sophists had, indeed, been put in their place more surely by Aristotle than by Plato; but they continued to thrive, until ancient rhetoric became more and more sophistic.

The conception animating the practise and the teaching of sophistic, far from being limited to antiquity, is medieval as well, and modern. Apparently it is permanent. Rhetoric is conceived by Aristotle as the art of giving effectiveness to truth; it is conceived alike by the earlier and the later sophists and by their successors as the art of giving effectiveness to the speaker. The conceptions are not contradictory. The second may be theoretically included within the first; and actually Demosthenes may learn something from Isocrates. But to embody them in educational procedure, to carry out either as the controlling idea of a course of study, is to discover that sooner or later they become practically incompatible. Ingenuous youth will be devoted either to energizing truth or to exploiting itself. There will come a parting of the ways; for the two conceptions are divergent. What Aristotle discerned as differentiating is differentiating still. The flaw in sophistic is moral. It may not impair technical training; but by deviating motive it tends to impair education.

For Aristotle’s theory is a touchstone. To recall rhetoric to the true function discerned by him has repeatedly been the object of reform in teaching. What has intervened to deviate rhetoric and frustrate its best use has again and again been the preoccupation with giving effectiveness not to the message, but to the speaker. Ancient sophistic is thus typical. It is not merely historical; it is historic. The false conception divined by Plato, and exposed finally by Aristotle’s demonstration of the true conception, led ancient rhetoric through empty personal triumphs into an elaborate art of display, devoid, at its worst, of other motive. As sophistic spread, as its idea of rhetoric became dominant, ancient education was narrowed;[6] and ancient oratory eddied in shallows until it found a new course with the new motive of Christian preaching.

In exorcising the false conception Aristotle removed the false sophistic emphasis from style. He does not despise, nor even slight, technic. He finds analysis of sentence rhythms necessary. But his goal in this, as in his analysis of figures, is beyond the technical means of securing particular effects. He does not classify figures for reference; he seeks in both phrase and cadence the function; and he discusses neither until he has spent some two-thirds of his treatise on the function of rhetoric as a whole course of study. This he finds philosophically necessary. Otherwise rhetoric cannot be justified; otherwise, he clearly implies, it is narrowed and degraded. For him rhetoric is so inextricably moral that it should never be divorced from subject matter of real significance.

But what subject matter of real significance has oratory when it is barred from discussion of present policy? Here appears a strong external cause of the spread of sophistic. The sophistical trend, already marked, was furthered by the narrowing of public discussion. Of the three fields[7] of oratory distinguished by Aristotle, deliberative, forensic, and occasional, the first was restricted by political changes. It faded with democracy. So later it faded at Rome, and still later in other realms. Deliberative oratory presupposes free discussion and audiences that vote. The steady increase of government from above administered by an appointed official class hastened also the tendency of the second kind of oratory, forensic, to become technical, the special art of legal pleading. Thus the only field left free was the third, occasional oratory, encomium, or panegyric, the commemoration of persons and days, the address of welcome, the public lecture. A favorite field even in Plato’s time, it is in any time the freest field for imaginative and emotional appeal and for personal triumphs. Thus it was early and assiduously cultivated by the sophists. Though it opens, on the other hand, the highest reaches of eloquence, though Isocrates is more than a sophist and Lincoln’s Gettysburg address is as far from sophistic as possible, still its becoming the main field of Greek oratory gave the lead to sophistic. In such conditions sophistic could control education; and its control of education reacted upon the conditions to make a vicious circle. Oratory and the training for it became preponderantly an art of display; and the rhetoric finally bequeathed by the ancient schools was sophistic.

In sum, the sophistic tendency, which may be found in any highly developed literature, was confirmed in Greek by causes both intrinsic and extrinsic. Becoming a habit, it became a scheme of education. Against this Plato represents Socrates in fundamental opposition. Aristotle does more than oppose it; he establishes constructively a rhetoric whose persuasion shall be more than personal appeal and personal triumph. But the rhetoric nobly and philosophically conceived by him did not succeed in supplanting the tendency seen at its best in Isocrates. The conception of Isocrates in Philostratus, though inadequate, is not wrong essentially.

The Isocratean ideal of eloquence, influential even upon its critic Dionysius of Halicarnassus, even upon so great an orator as Cicero,[8] became more sophistic in the practise of the schools. In the first century of our era sophistic had won its place, by the second century an eminence undisputed till Christian preaching returned to the sound ancient tradition.

For sophistic is the historic demonstration of what oratory becomes when it is removed from urgency of subject matter. Seeking some inspiration for public occasions, it revives over and over again a dead past. Thus becoming conventionalized in method, it turns from cogency of movement to the cultivation of style. Cogency presupposes a message. It is intellectual ordering for persuasion, the means toward making men believe and act. Style, no longer controlled by such urgencies of subject, tends toward decoration and virtuosity. A necessary study in any rhetoric, it had been highly cultivated during Greek democracy; but under monarchy and Empire it became a preoccupation, almost a monopoly. Sophistic practically reduces rhetoric to style. The old lore of investigation (inventio), paralyzed by the compression of its trunk nerve, has little scope beyond ingenuity. Organized movement[9] (dispositio), similarly impaired at the source, tends to be reduced to salience and variety, or to be supplanted by pattern. Memory becomes verbal. But style and delivery, becoming the main reliance, are elaborated into a systematic technic to a degree almost incredible to-day. In sheer virtuosity the second sophistic has hardly a parallel in earlier or later centuries. It is more like the art of Paderewski or Bernhardt than like that of Demosthenes.

The Lives of the Sophists by Philostratus derives the “new” sophistic of his day from the old. New, says he, it is not. We may call it a second sophistic; but it keeps an old tradition. Far from apologizing for sophistic, old or new, Philostratus is proud of it. He sets out to celebrate it as a great tradition. Gorgias is not defended from Socratic exposure; he is claimed as a distinguished ancestor.[12] That rhetoric is not what Aristotle urged, that it is after all sophistic, Philostratus assumes. Nor is the assumption merely provincial vanity; it was widespread and secure. There was no need to vindicate what was generally accepted. Moreover the facts justify not only the assumption of Philostratus, but his history. Though he is not historical in method, his assertion of continuity from Gorgias down to the platform artists of his day has been approved by studies really historical, and affirmed for the fourth century as well. The Lives of the Sophists, therefore, exhibit the second sophistic as the fixing of an old tendency in habitual practise and teaching.

During the second, third, and fourth centuries, and throughout the Roman world, rhetoric meant a sophistic generally constant. The leanings of particular schools, such as Stoic Pergamum, were not wide enough to bring new departures. They were merely shifts of emphasis within a common doctrine and practise. The cult of “Atticism” was too artificial to check the general tendency to “Asianism.”[13] What was learned at Athens could be practised at Rome; and neither Athens nor Rome dimmed the glory of Smyrna and Antioch.[14] The schools of Bordeaux were to become essentially like those of Gaza and Carthage. The same “Gorgian figures” were learned by St. Augustine in Latin Africa, by St. Gregory Nazianzen in the Greek East, and by the pagan Libanius. Greco-Roman rhetoric was as pervasive as Roman law and almost as constant.

a. VIRTUOSITY

In estimating this rhetoric, then, little allowance need be made for place or date.[15] Its main characteristics were so constant as to stand out clearly. The most obvious arise from the general aim of virtuosity. This is the constant assumption of Philostratus. Individual triumphs were not so much triumphs of individuality as outstanding exhibitions of skill in working out a pattern. In method, in composition, there was little difference between a teacher’s assignments to his amateur pupils and his own professional orations.[16] Sophistic is largely an oratory of themes.

(1) Declamatio[17] (μελέτη)

Fictitious Themes

Evidently the themes were generally the same as those of the declamationes celebrated by Seneca.[20] Some of them were identical. Such subjects give the oratory of the imperial centuries, both Greek and Latin, the air of athletics, and make its teaching seem largely gymnastic. Gregory Nazianzen, indeed, calls the sophists “oratorical acrobats.”[21] But instead of dismissing sophistic with so obvious a sarcasm, we may learn something from its delight in verbal artistry.

Since the oratory of display is still with us, the second sophistic should be taken to heart as a complete historic demonstration of what must become of rhetoric without the urgencies of matter and motive.

Philostratus has no qualms. For him declamatio (μελέτη), far from being merely a school exercise, is a form of public speaking on a par with any other. It is even the form of his sophists. He pays little attention to any other except encomium, which is also a school exercise. Reading the past with the eyes of the present, he finds it in Gorgias,[23] who elaborated “encomia of the Medic trophies.” “Medics” (Μηδικά), dilations on the old glories of the Persian wars, were the favorite subjects of declamatio. This is evident both from their frequent recurrence and from Lucian’s satire.[24] Scopelian, Philostratus thinks, was best

Ptolemy of Naucratis, however, was nicknamed Marathon.[25] The wonder is that the nickname was sufficiently distinctive.[26]

(2) Improvisation and Memory

The vogue of such subjects does much to explain the otherwise incredible accounts of improvisation. “Propose a theme,” the sophist’s challenge which Philostratus traces back to Gorgias,[27] becomes less startling when we find that the theme, as well as the treatment, might come from stock. Even so the readiness and fluency seem phenomenal and were the great boast. See Mark of Byzantium recognized by pupils in the school of Polemo.

Aristides was exceptional in declining to speak thus; and Philostratus thinks the less of him.[28] Generally improvisation was expected as a mark of virtuosity.[29] The locus classicus, perhaps, of improvisation, the most daring and phenomenal virtuosity, is ascribed by Eunapius to Prohæresius.

This performance is extraordinary only in degree. That of Isæus, as reported by Pliny,[30] seems the same in kind; and so, apparently, are many other recorded triumphs. Taken together, they reveal strict limits. The improvisation was mainly of style. It consisted in fluency of rehandling, of variations upon themes, and in patterns, so common as to constitute a stock in trade. It permitted the use over and over again not only of stock examples and illustrations, but of successful phrases, modulated periods, even whole descriptions. It was the art of a technician, not of a composer.[31] Memory, too, thus trained, was no longer the orator’s command of his material;[32] it was the actor’s command of words. Though a sophist might, indeed, be a thinker, he hardly needed to be for the purposes of his oratory. His fluency was typically not in seizing and carrying forward ideas and images, but in readiness to draw upon a store.

(3) Delivery

The character of this oratory is further expressed in the records of its delivery. Even more than modulation Philostratus exhibits sonority and force. Polemo’s delivery was thrilling as an Olympian trumpet.[33] Scopelian imitated the volume of Nicetes and had the sonority of Gorgias.[34] Favorinus fascinated even those who did not understand Greek.[35] The carrying voice spoke in marked rhythms. Gesture, pushed sometimes to the extent of acting, was habitually demonstrative. Sitting at first, the orator might then leap to his feet, smite his thigh, walk, stamp, sway as a Bacchante. If such theatrical delivery seems to moderns of the West more violent than it seemed to its own audiences, it has never been extinct; and any one familiar with the oratory of display in any time will recognize the sophist’s heavy frown, his mien of deep thought, his air of authority.[36] Chaucer’s Pardoner speaks for the whole sophist line:

Such weight and vehemence of delivery, sometimes conceding a benignant smile, oftener relying on arrogance,[37] was the bodily expression of the impressiveness (δεινότης) cultivated no less assiduously in style. The sophist was over-expressive lest for a moment he should cease to be impressive. The audience need not be held to any course of thought; it must not be held too long by any one device of style; but it must unflaggingly admire. It must be spellbound. The constant implication of Philostratus probably echoes the ideal of orator and audience alike: behold a great speaker!

b. DILATION

Such oratory must be dilated, even inflated. That it was so in fact any one may satisfy himself who has the patience. The amplification[38] practised by Cicero and taught by Quintilian, though in print it may seem over-anxious, is an oratorical necessity. It is not merely Greek expansiveness; for it moves also the more stinted Latin. In any language it must almost always be practised as a means to oral clearness.[39] But sophistic amplification has no such warrant. It is often purely decorative. Instead of marking a stage of progress, it often merely dwells on a picture, or elaborates a truism, or acts out a mood. It is there for itself, expecting its own applause. Many of the figures of speech are devices of dilation; for sophistic is an art not only of elaboration, but of elaborateness.

(1) Ecphrasis

Without enumerating devices that constitute a large part of sophistic, we may see the characteristic dilation at the full in the single form known as ἔκφρασις.[40] An ecphrasis is a separable decorative description, usually of a stock subject. “I will draw this for you in words,” says Himerius,[41] using the formula of introduction, “and will make your ears serve for eyes.” The natural beauty of a prospect or of the human body is detailed for admiration, even oftener the artistic beauty of statue or temple. The orator turns on, as it were, a storm, a feast, the prospect of a city. The essentially artificial character of the ecphrasis is obvious in the favorite exercise of word-painting a peacock.[42] Apparently a boy could carry this peacock from school to the platform and continue to use it with merely verbal variations.

Of course the ecphrasis might rise to a higher level. So it did often. An accomplished orator might make it splendid, even really moving. Oratory cannot afford to neglect the appeal of oral description. None the less the ecphrasis had two essential vices. First it was extraneous, separable, detachable, a clear sign that sequence did not count. Secondly, instead of following the Aristotelian counsels of specific concrete imagery, it habitually generalized and rapidly became conventional.

Ecphrasis is no less significant for poetic. A form of Alexandrianism[43] avoided by Vergil and adopted with enthusiasm by Ovid, it perverts description because it frustrates narrative movement. The habit of decorative dilation in oratory confirmed a decadent habit of literature.[44] That the habit is decadent even when indulged with more taste is suggested by certain passages in De Quincey, in Pater, most clearly perhaps in that English sophist Laurence Sterne. Among the ecphrases of the Sentimental Journey is one that he executed upon a theme taken from the most expert mocker of the sophists, Lucian,[45] and has made quite typical of the soothing rhythms and the elegant dilation of sophistic eloquence.

c. PATTERN

For the composition of the whole speech sophistic generally had little care. That planned sequence, that leading on of the mind from point to point, which is the habit of great orators and the chief means of cogency, presupposes urgency toward a goal. Sophistic often had no goal. The audience need be won only to admiration, not to decision. Easily, therefore, rhetoric came to pay no more attention to logical movement than poetic to movement in narrative. Like Alexandrian narrative, sophistic oratory cares little for onwardness; and its lore is reduced to prescription for detail.

If Philostratus seems occasionally aware of the value of planned movement, scrutiny will reveal that he is thinking not of the order of the whole, but only of sentences. For instance, the passage praising Isocrates for his “brilliant composition”[46] specifies his handling of rhythms; “for thought after thought concludes upon a balanced period.”

As if to mark the lack of individual planning for cogency, sophistic is commonly composed upon set patterns. No other body of oratory has so uniformly resigned itself to forms. The orator could devote his whole attention to each separate development because its place was predetermined in a traditional series of topics. The encomium[48] of a country was expected to deal with its situation, climate, products, its race, founders, government, its advancement in learning and literature, its festivals and its buildings, unless indeed the whole encomium were based on one of these topics. Similar topics controlled the praise of a city, a harbor, a bay, an acropolis. The classification of these as separate forms goes on to enumerate the speech at an embarkation, a marriage, a birthday, a festival, etc., as in a complete letter-writer.[49] Similarly prescribed was the encomium of a person. So pervasive were its topics that they invaded even written biography.[50] Philostratus follows them in his account of Herodes Atticus.[51] Basil, on the other hand, explicitly rejects them as inept for encomia of Christian martyrs; and his protest shows at once their prevalence and their typical vice.

Such composing upon a pattern is legitimately a school exercise. Its use in elementary education is not confined to sophistic. What is sophistic is its extent and its prescriptiveness, still more its extension from school into adult and professional practise. How far the oratory of the imperial centuries was controlled by fixed topics becomes startingly evident in its conformity to the rules set forth by the manuals of elementary exercises.[53] Theon’s of uncertain date may have been superseded[54] by that of Aphthonius in the fourth century, which at any rate had a long life.[55] But the pattern is most concisely shown in the second century by Hermogenes,[56] whose work is typical of them all. There some of the most characteristic habits of form in sophistic oratory are seen as prolongations of school exercises.

Arid, impersonal as arithmetic, pedantically over-classified, sometimes inconsistent, these rules[64] are nevertheless illuminating. They expose sophistic oratory. The patterns set forth for boys are recognizably the patterns of the public oratory of men. Such higher attainment as might come with experience was not in composition. In composition adult oratory too, as well as these elementary exercises, was feeble at the source. For lack of animating conception and advancing urgency of thought, it eddied in forms. It is the historic demonstration of the doom of an oratory of themes. The resounding reputations so expertly cultivated for themselves time has reduced to absurdity. Hippodromus, Mark, Polemo, Scopelian—which of the beadroll of Philostratus is even the echo of an echo?

d. ELABORATION OF STYLE

The long reign of sophistic reduced rhetoric to style.[65] That this was the preoccupation even of the earlier sophistic we may guess from the derision of the Phædrus and from other references.[66]

The wider scope demanded by Aristotle’s different conception is recognized in the traditional fivefold division[67] found in Cicero and Quintilian. This division, which has such validity as to be essential for securing the educational values proposed by the ancients, seems to have been inactive from Quintilian on to the fall of Rome.[68] That the limitation to style impoverishes rhetoric and impairs even the study of style itself is evident in the sophistic period and is confirmed in the medieval.

Incidentally, the focus on style contributed to the confusion of rhetoric with poetic.[69] Neither being conceived often in its larger aspects of movement, both being studied habitually for words and sentences, the distinction between the two was the more easily blurred. Here poetry had the more to lose. The use of poetic diction to decorate oratory must have confirmed the tendency to conceive poetic itself as an art of decoration.

But the main results of giving to style a monopoly are the cultivation of literary flavor, with conformity to past usage, and the forcing of figure and rhythm. The style inevitably acquired by those who seek style is decorative and elaborate. In order to sound literary, the orator is impelled both to depart from common speech and to force his note. Devices valuable in revision, to clarify and impress a message, become artificial in practise and unduly elaborated in theory by being pursued for themselves.

(1) Literary Allusion and Archaism

The preoccupation that seized any opportunity for “Medics”[70] led to frequent literary allusions. Allusion is a legitimate, sometimes an important, means of heightening eloquence. Reviving old associations by familiar words and rhythms, it helps to suggest a mood or intensify an appeal.

But the sophists used this form of suggestion so incontinently, and often so conventionally,[71] as to betray an anxiety to sound literary.

The same anxiety led to their frequent use of obsolescent words. Archaism became an habitual form of decoration. They borrowed the language of Demosthenes to welcome a proconsul and win from him some otiose appointment, or played upon an audience to capture its applause for literary tone. “Atticism” was often little more than pride in a highly sanctioned diction. The aim of the sophists was not to model their composition on Demosthenes, still less on the restrained habit of Lysias, but to borrow from them words enough to give antique flavor. Presuming to be apostles of Hellenism, they were anxious to sound traditional. That they often thereby became stilted is early evidence that this conception of elegance is false.[72]

(2) Decorative Imagery

Metaphor, which is a reliance of all popular oratory, seldom has in sophistic the suggestiveness of fresh observation. Rather the sophist relied on far-fetching or on the abundance of his literary stock.[73] “Living tombs,” said of vultures, is ascribed to Gorgias by the treatise On the Sublime,[74] and passed on through the schools. If Athens was “the eye of Greece,” another city might be “the eye of Asia.” Sophistic metaphor generally lacks vitality. That it should achieve so little imaginative suggestion is a clear sign of artificiality. No less conventional and decorative are the frequent similes. They are more sophistic only in being more elaborate. In both cases it is not the imagery that is sophistic; it is the straining, or the conventional decoration, or the dilation. For all its store of tropes, for all its lavish use of them, sophistic is poor in active imagery. Quintilian’s eighth book analyzes the heightening of diction which comes from concreteness[75] (iii), amplification (iv), epigram (sententia, v), and tropes (vi). Of these the second and third were the reliance of the sophists. The first they neglected; the fourth they had conventionalized.

(3) Balance

Imagery, what the ancients called trope, covers all that is usually meant by the term figure in modern use. Ancient manuals and their medieval derivatives generally use figures to mean typical adaptations of sentence movement. These are minutely classified even in the older rhetoric. The treatise Ad Herennium,[76] which does not distinguish them from tropes, enumerates sixty-one and groups them by the traditional twofold division followed by Quintilian: figuræ sententiarum (σχήματα διανοίας), and figuræ verborum (σχήματα λέξεως). Quintilian (IX), distinguishing them from tropes, both reduces the number of figures and by grouping simplifies the analysis. Of figuræ sententiarum he enumerates twelve.[77] Figuræ verborum he groups as: (1) variations of syntax, (2) modes of iteration, (3) word-play, (4) balance and antithesis. The sophists especially cultivated these figures, most of all the last, those forms of balance which were traditionally called the figures of Gorgias (Γοργίεια σχήματα).[78]

Balance, as an obvious way of marking a comparison or a contrast, is so familiar in every language and in every period as hardly to be thought of as a figure. It becomes a figure by becoming a preoccupation; and the preoccupation, evident in certain modern literary periods, has never been stronger than in sophistic. The sophists pursued balance with such zeal as to display its typical faults of padding and superficiality. A habit of balance tends to slip in here and there a makeweight of mere words, or to force the sense into the form. Over-balancing, sophistic shows abundantly, invites false balance. It is the way not to precision, but to epigram.

Description of the several forms of balance distinguished by sophistic cannot go far without examples. The only sufficient examples must be sought in Greek and Latin; for the sophistic refinements often depend upon the recurrence of inflections or upon transpositions possible only in a language that is highly inflected. Modern languages depend so much less on inflection that they chime less readily and can transpose for symmetry sometimes only by more conspicuous violation of normal sentence order. Nevertheless some of the sophistic forms of balance, with other figures of words, can be exhibited accurately, and the character and effect of them all can be generally suggested, by English examples. Both the charm and the danger of the ancient figures are exemplified by De Quincey in what he called “impassioned prose.”

De Quincey’s encomium Joan of Arc, insistent in apostrophe, has one hyperbole that might have been uttered by Polemo or Scopelian. “The graves that had closed sixty years ago seemed to fly open in sympathy with a sorrow that echoed their own.” His alliteration, too, often suggests the same anxiety to enhance. In the ecphrasis on the forest and fountain of Domrémy (paragraph 12), and again in the corresponding one toward the close, he is more delicate. On the other hand he uses with sophistic fondness the device of a carrying iteration. The encomium opens:

An even more marked example is the paragraph next to the last. Its opening and its close are as follows:

There are few more striking examples of a value in iteration much sought by the sophists, its carrying on to a climax. Refrain carrying to climax is used at greater length, and with finer balances and allusions, in the twenty-eighth paragraph of The English Mail-Coach.

Whether De Quincey’s reading of Greek may have dwelt too long on Isocrates is less important than that his devices of style spring from similar preoccupations. The balances of Sir Thomas Browne, whose style he tells us that he studied, have none of this sophistic chiming and oral dilation. De Quincey reminds us of the sophists because he is a sophist. Sophistic was not extinguished with the Roman Empire; and De Quincey’s style has marked family traits. Thus it is easy to detach many suggestive examples of the Gorgian figures, balances used not for clearness, but generally for emotional emphasis and sometimes for emotional expansiveness.

These are simple balances. “Which was heaven’s vicegerent, and which the creature of hell” marks the antithesis by reverse balance (chiasmus). The following are enhanced by alliteration:

Nor does De Quincey’s refinement stop there. The following balances have antithesis, alliteration, chiasmus, hyperbaton. The first varies its contrasting rhythms; the second leads up to the climax quoted above (“For all, except this comfort”).

The sophistic marking of balance by rime (homœoteleuton), easy in Greek or Latin through the recurrence of inflectional endings, is so forced in English as to be very rare and in very bad taste. Word-play, on the other hand, has always been one of the commonest devices for enhancing balance into epigram. “Figures do not lie. The trouble with statistics is not that figures lie, but that liars figure.” Paronomasia is not sophistic; but, like other jingles, it attracted the sophists too much, as to-day it attracts Mr. Chesterton and Mr. Bernard Shaw.

(4) Clausula

Balance is only one mode of rhythm. The sophists were so preoccupied with it as often to risk monotony. For though they boasted of variety, they were too fond of certain rhythms, and too anxious to mark them, to achieve much flexibility. Their idea of aptness as conformity to an assumed character or occasion led them rather to cast a whole passage in one stylistic pattern.

Next to the perfecting of balances, they studied most attentively sentence cadences.[80] Of clausulæ, as of other effects of style, they had a classified store for selection. Though we find it hard to follow them here, and impossible to translate their clausulæ in terms of English stress rhythms, we are not warranted in dismissing their studies of rhythm as idle. True, they often overdid rhythm as they overdid technic in general; but English prose has rarely been in danger of this excess, and in particular it has been surest with those who have controlled cadence. Sentence emphasis is the clue to mastery of sentence movement. Its greater masters, modern as well as ancient, have grasped this not only as logic, but as cadence. The flaw in sophistic rhythms is their emptiness, the pursuit of them for themselves. The difference between the sounding clausula of dilation and the solving clausula of mounting emotion can be heard in the same English sophist. All the following sentences conclude well for the ear; but whereas the first two are prolonged by decorative additions, the last is an ascending period.

(5) Vehemence

“Scopelian, when one of Polemo’s pupils said that his instrument was the drum, picked up the sneer with ‘the drum, indeed; but it is the shield of Ajax.’”[81] The passage is characteristic not only in allusion, figure, and ingenuity, but in grandiloquence. It is itself a drum-beat; and the sophistic harmony was fond of drums. The stylistic effects most sought are those most marked.[82] Scott’s deprecatory “The big bow-wow strain I can do myself”[83] has neither this aim nor this attitude. He is generously wishing that he could control the quiet sureness of Jane Austen. A sophist was complacent in his own style. He was anxious only that his bow-wow should always be big, or, to return to Scopelian’s more precise figure, that the audience should always hear the drum. It is the drum that marks sophistic. Few of the devices of style so carefully cultivated are sophistic in themselves. What is sophistic is the use of them all, as from a classified store, in excess and with insistent emphasis. The sophistic style cannot be escaped. It is always saying, Here is style.

Such rhetoric is not worthless. Some of its technical skill is available for better ends. But as other arts, to survive and progress, must be more than technics, so especially the art of words cannot go far without being animated by power of conception.[84] Technic is promotive and educative only as it gives free course to motive and vision. As a system of education, therefore, sophistic was hollow. This is the issue raised by Plato; and he is justified by history. Sophistic could use its many devices only to exhibit skill, not to guide either the state or the individual. The only force that could revive rhetoric with the lore older than this spent tradition was a new motive.

CHAPTER II [1] ST. AUGUSTINE ON PREACHING ( DE DOCTRINA CHRISTIANA , IV) — TABULAR VIEW OF ST. AUGUSTINE’S DE DOCTRINA CHRISTIANA IV — FOOTNOTES:

With this elaborate pedagogical tradition a clean break is made by St. Augustine. The fourth book of his De doctrina christiana[2] has historical significance in the early years of the fifth century out of all proportion to its size; for it begins rhetoric anew. It not only ignores sophistic; it goes back over centuries of the lore of personal triumph to the ancient idea of moving men to truth; and it gives to the vital counsels of Cicero a new emphasis for the urgent tasks of preaching the word of God.

Abstractly and in retrospect the very character of Christian preaching seems necessarily to reject sophistic. But at the time this seemed anything but inevitable. Sophistic was almost the only lore of public speaking then active. It dominated criticism and education. The Greek fathers Gregory of Nyssa and Gregory Nazianzen might expose its falsity of conception; but they could not escape it. It had brought them up. Its stylistic habits were ingrained in their expression. Augustine too had been brought up on sophistic. Nor could he escape it. Again and again his style rings with its tradition.[3] Not only had he learned it for good; he had taught it. He had been himself, in Plutarch’s sense and Strabo’s, a sophist. We must hasten to add that the great Christians of the fourth century, if they could not escape sophistic, at least redeemed it by curbing its extravagance and turning it to nobler uses. But Augustine did much more. He set about recovering for the new generation of Christian orators the true ancient rhetoric. He saw that for Christian preaching sophistic must not only be curbed; it must be supplanted. Against the background of his day his quiet, simple book, renouncing the balances and figures of his other works without renouncing their fervor, is seen to be a startling innovation.

Not the least striking trait of the innovation is its reserve. Augustine does not attack sophistic as the Gregorys do; he ignores it. In Chapter xxxi of Book II he had, indeed, mentioned it. Discussing there not style, but matter, he had contrasted the necessary training in argument with sophistic quibbling, and had then added, forecasting Book IV, that superfluous stylistic ornament also is sophistic.

But an uninformed modern reader of Book IV would hardly be aware that sophistic existed. No denunciation could be more scathing than this silence. In Augustine’s view of Christian preaching sophistic simply has no place. A good debater, instead of parrying he counters. He spends his time on his own case. A good teacher, he tells his neophytes not what to avoid, but what to do. He has so far renounced sophistic that he has no concern to triumph. He wishes simply to teach sound rhetorical doctrine. He achieves an extraordinary conciseness not so much by compression as by undeviating straightforwardness.

A reader familiar with the times, however, will be reminded of sophistic by many allusions. Single phrases or sentences some of them, a few more extended, they all serve to illuminate by contrast the true rhetoric.

Display, inflation, thirst for applause—every reader of Augustine’s time would recognize in these allusions a repudiation of sophistic.

For Augustine thinks that Christian preaching is to be learned best from Christian preachers. As if in reply to Julian’s scornful “Let them elucidate their Matthew and Luke,”[6] he recommends not only for doctrine, but for rhetoric, the Epistles, the Prophets, and the Fathers, and proceeds to analyze their style. The analysis, though based on the current Latin version, is generally transferable to the Greek, since it is much simpler than the classification set forth by sophistic. It exhibits sentence movement simply in climax, period, balance—those devices which are most easily appropriated and most useful. The general ancient counsels of aptness and variety are applied specifically to preaching. As to cadence (clausula), Augustine dispenses with all subdivisions, and even makes bold to assert that it must sometimes be sacrificed. Similarly omitting all classification of figures, he manages to suggest in a few words what figures are for. In a word, he shows how to learn from the Canon and the Fathers the rhetoric that is vital to homiletic.

This rhetoric, not only simpler than sophistic, but quite different in emphasis, is set forth in the terms of Cicero. Augustine has gone back four and a half centuries to the days before declamatio. The instruction that he draws from his analysis of Christian literature is planned upon the “instruct, win, move” (docere, delectare, movere) of De oratore and upon the corresponding three typical styles (genus tenue—medium—grande) of Orator.[7] Evidently Augustine had the greater Cicero, not the lesser that sufficed for the Middle Age. He neither quotes nor cites any other rhetorician; and though his doctrine of aptness and of variety is common throughout the older rhetoric, for this too he had no need to go beyond the master’s two great works. Nor have any others been more persuasive as to imitation,[8] which is Augustine’s controlling idea. This first Ciceronianism, too immediately aware of the perverted imitation of style taught by sophists to fall into the archaism and redundancy of later worship of Cicero, is a penetrative recovery of Cicero’s larger meaning. Augustine’s application of the three typical styles is more just and more practically distinct than Cicero’s own. Would that all Ciceronians had been equally discerning!

The fourth book of the De doctrina Christiana is specifically linked by its proem to the preceding three as setting forth presentation (modus proferendi). Books I-III have dealt with study of the subject matter (inventio); Book IV is to deal with expression. Augustine thus makes the traditional fivefold division twofold. Inventio, which under sophistic had lapsed, he restores to its rightful place and gives it a new application to the exegesis of Scripture. Of the remaining four left to his second heading he discusses only style (elocutio). Delivery and memory are mentioned incidentally; plan is omitted. The omission is not negligent. The first chapter warns us not to expect a manual of rhetoric. Nevertheless a modern student cannot help wishing that so suggestive a treatise had both applied to preaching the ancient counsels as to plan and exhibited the New Testament in this aspect. Thus to analyze for imitation not only the style of the Pauline epistles, but their cogency of order, would doubtless have made the work unduly extensive. One hopes that seminarians of the fifth century were stimulated, and that seminarians of the twentieth century will be stimulated, by the example of the treatise itself to study Romans not only for appeal, but for cogency. Meantime Augustine’s fourth book remains one of the most fruitful of all discussions of style in preaching.

But to learn such skill from rules, he goes on, is the way rather for boys than for men who have immediately before them the urgent tasks of preaching.

They must, indeed, know the principles of adaptation (iv), and develop their expression as far as they can; but they will do so best by imitation.

Expressed in modern terms, Augustine’s position is that rhetoric as a classified body of doctrine is properly an undergraduate study. It is not the best approach for seminarians because its method is analytical. The young preacher, needing rather promotion than revision, will advance more rapidly by imitation.

Starting from this principle, that the more fruitful study for learning to preach is imitation of Christian eloquence, Augustine proceeds to show (vi-viii) how distinctive is the eminence of such models and how repaying to analysis. His vindication should be pondered by those who still permit themselves to disparage without distinction the literary value of the New Testament, and by those who, granting poetic to Ambrose, remain unaware of his rhetoric.

The vindication of an eloquence distinctly Christian has the more weight because its doctrine of form and substance echoes from Cicero the best ancient tradition. The older tradition had in Augustine’s time been so overlaid that he could do no better service to rhetoric than to recall it. In fact, Christian eloquence redeemed public speaking by reviving the true persuasion.

The insistence on the Ciceronian doctrine that style is not separable has a bearing more than historical. Not only for Augustine’s time, but for any time, the truism must be reasserted. His iteration is more than preoccupation with Cicero, more than repudiation of sophistic. It springs from the cardinal importance of the truism for homiletic. In the pulpit the sophistic heresy of art for art’s sake becomes intolerable.

Augustine’s next step (vii) is to support his general claims for Christian eloquence, and to show how it may be studied, by analyzing briefly three typical passages. In the first, Romans v. 3-5, he analyzes prose rhythm under the familiar heads of classical sentence movement (compositio): phrases and subordinate clauses (cæsa), coordinate clauses (membra), period (circuitus), climax (gradatio), adding the equivalent Greek terms.[14]

RHYTHMICAL ANALYSIS OF ROMANS V. 3, 4, 5

The passage is short enough, and the sentence movement simple enough, to be grasped readily. Its balance is striking without being monotonous, and is reinforced by a linking iteration that leads to a climax.[15] He is a wise teacher who begins with an instance so memorable. It must have seized even more quickly a generation familiar with both the terms and the method.

The next example, 2 Corinthians xi. 16-31, shows the same sentence devices carried through a much longer reach, and is therefore used both to reinforce the first and to add the importance of rhythmical variety. The counsel of variety, though a commonplace of the older rhetoric, had especial point by contrast with the sophistic fondness for trimming and prolonging balances. Incidental to the exhibition of variety is a reminder of aptness; and the analysis concludes:

The analysis of the third example, Amos vi. 1-6, leads the study to longer and more sustained rhythmical reaches. Lest it seem the more difficult in the more figurative version of the Septuagint, Augustine quotes it “as translated from the Hebrew into Latin style through the interpretation of the priest Jerome, expert in both languages.”

ANALYSIS OF AMOS VI. 1-6

Much more urgent, leaping to attack, rising, prolonging, varying, subsiding to a pregnant close, the prophecy widens the conception of rhythmical range. Marking the rhythms briefly, Augustine uses it also to show the oratorical force of figures.[16] Thus a few pages of analysis are made to yield wide and definite suggestion. This, perhaps, is their outstanding merit; while they show the student what to look for, they invite him to go on for himself. But the pedagogical achievement does not stop there. The professor of rhetoric has seen that rhetorical analysis must be simplified, and that it must be made progressive. Where else shall we find so much drawn from three analyses? The first reduces the complicated lore of rhythm to its essentials. The second, reinforcing and extending these, dwells upon aptness as a corrective of rhetorical zeal, and as a constructive principle. The third, quoting rhythms still more urgent with emotion, passes to the emotional value of concrete words. To bring the over-classified lore of sophistic back to the simplicity of Aristotle was a service not only to homiletic, but to all rhetoric. A greater service was to substitute for the static and formalized pedagogy of the day a vital order. Augustine had been doubtless a popular professor; Christianity made him a great teacher.

Pedagogically, therefore, even his incidental definitions are worth noticing. That the function of grammar is traditionally to impart correctness of speech (iii) is used to support the contention that even this elementary skill comes best in fact from imitation. The period (vii) is defined so as to throw the emphasis on delivery. Its “clauses are suspended by the speaker’s voice until it is concluded at the end.” Therefore it “cannot have fewer than two clauses.” So he points out in the passage above from Amos that the rhythm is available for delivery (in potestate pronuntiantis) either as a series of six or as three pairs, and that the latter is more beautiful. So he suggests limiting analysis to give room for oral interpretation.

The next and longest section (ix-xix) is based on Cicero’s “inform, please, move” (docere, delectare, movere). Distinguishing each of these tasks clearly, Augustine is at the same time careful to unite them, by progressively iterative transitions, in the single and constant task of persuasion. In exposition (docere) clearness may demand the use of popular expressions. What avails correctness in a diction that is not understood?

The correctness (integritas) of diction boasted by the sophists, and carried by them even to the pedantry of archaism, is here faced squarely. The assertion that it must sometimes be sacrificed, the making of clearness absolutely paramount, is the bolder at a time when Christian preaching was not yet recognized as having secure command of elegance. Unmistakable clearness, Augustine goes on, is so much more important in preaching than in discussions permitting question and answer that the speaker must be quick to help unspoken difficulties.

No warrant here, he adds (xi), for dilation beyond the demands of clearness, but good warrant for making instruction pleasant and appealing in order to hold attention. Passing thus to the two other tasks of oratory, he quotes (xii) Cicero’s “to instruct is of necessity, to please is for interest, to move is for victory.”[18] The three are then both carefully distinguished and shown to be a sort of geometrical progression. The first is first of necessity. It must be mastered; but it is rarely sufficient. To supply the lack, the second demands more rhetoric by demanding further adaptation to the audience; but it too must remain insufficient. So the third task, to move, is not merely the third item in a classification; it is the final stage in a progress. That progress is increasingly emotional. The last stage demands not only all the rhetoric of the preceding, but also the art of vivid imagery[19] and of urgent application. So Augustine arrives at one of those linking summaries which constitute almost a refrain.

The next chapter (xiv) warns against resting in the second stage.[20] To make the pleasing of the audience an end in itself is the typical vice of sophistic. If preaching tolerates it, “the time will come when they will not endure sound doctrine; but after their own lusts shall they heap to themselves teachers, having itching ears.” Augustine quotes, not these words of St. Paul, but Jeremiah, and rises to denunciation of mere pleasing. “Far from us be that madness.” One of Cyprian’s rare descriptive passages is adduced to show how “the wholesomeness of Christian preaching has recalled his diction from [sophistic] redundancy and held it to a graver eloquence of less display.” As the ultimate objection to the sophistic ideal is moral, so is the preacher’s ultimate resource. Since his strength is derived from a source deeper than human skill, his best preparation is prayer. Augustine is not above enforcing this reminder by playing upon the words orare, orator, oratio. Nevertheless human skill is to be cultivated. Prayer itself proves the folly (xvi) of making no other preparation. He who abjures human lore of preaching because God gives us our messages might equally well abjure prayer because God knows us and our needs. The Pauline counsels specify how Timothy should preach. As God heals through doctors and medicines, so he gives the gospel to men by men and through man.

The transition (xvii) to the final task of moving men to action is another full and explicit iteration of all three, and at the same time a preparation for the next section on the corresponding three typical styles. Since the subject matter of preaching is always great, at least in implication (xviii), does it not always demand a great style? No; for a great matter (xix) may at the time rather demand exposition; and this in turn demands a restrained style. Again, a great matter may at the time rather demand praise or blame; and here enters the second task of so adapting the style as to win sympathy.

Augustine has passed (xvii-xix) from Cicero’s three tasks of oratory to his three typical styles by applying to the preacher Cicero’s definition of the orator: “He, then, shall be called eloquent who can speak small things quietly, larger things proportionally, great things greatly.”[22] Thus the three styles are genus submissum (or tenue), genus temperatum (or medium), and genus grande. As in Cicero, these correspond to docere, delectare, movere, and the second is connected with panegyric.

Augustine now proceeds to exemplify the first style (xx) from Galatians as calling for skill in reasoning and for a memory trained to bring in objections and difficulties where they can best be met. This debater’s memory is precisely the ancient memoria, the fifth of the traditional parts of rhetoric. It seems to have fallen into abeyance under sophistic. What the sophists boasted was verbal memory, which Augustine merely mentions in his appendix as something quite different.[23]

The same chapter (xx) exemplifies the second, or median style from Timothy and Romans as having the charm of aptness. Here Augustine confronts squarely the sophistic habit of making rhythmical beauty paramount and the pagan disparagement of Christian style. Some one may find the cadence of Romans xiii. 14 defective. Certainly it would soothe the ear more rhythmically if the verb came last.

The third, or great style, whether it be elegant or not, has for its distinguishing quality the force of emotional appeal. The instances are from 2 Corinthians vi and Romans viii. Romans is a long epistle, not a sermon. Though it was read aloud, of course, it is essentially a treatise, a philosophy of history. It is largely expository and argumentative. Since it is addressed primarily to reflection and reason, its main artistic reliance is on cogency of order. But even here presentation does not remain purely logical. For persuasion it must rise also emotionally. As we read in Acts xvii the outline of the apostle’s Areopagus speech, we discern beyond the logical chain of propositions an expanding conception of the Life-giver. Who can doubt that the style too, as in Romans, rose to grande? The traditional doctrine of the peroration, easily as it may be abused, is only the expression in rhetoric of the audience’s final demand and the speaker’s final answer. That demand and that answer are emotional.

Adding Galatians iv, Augustine says of it:

After quoting without further comment examples from Cyprian and Ambrose, Augustine shows (xxii, xxiii) the need of variety. More even than other forms of oratory, preaching seems to suffer from a stylistic level. No one of the three styles, least of all the third, can effectively be prolonged; the change from style to style gives relief; and subordination of what might be heightened may enhance the emotion of what must be. What must be heightened is what is to rouse the audience to action. So the test of achievement in the third style is not applause, but tears and change of life (xxiv). So also the end of all eloquence, in whatever style, is persuasion (xxv).

Thus Augustine is more explicit than Cicero in showing that the three typical styles are but three ways (xxvi) of achieving a single end, even as the three corresponding tasks, though one of them absorbs attention at a time, are but three aspects of the single task. Nor can persuasion dispense with a means beyond art, the appeal of the speaker’s life[25] (xxvii). Though the Church speaks not merely through a man, but through his office, persuasion needs for full effect his whole influence. Because his life is without shame, the preacher speaks not shamelessly (xxviii), not only with restraint and charm, but with power, to win obedience to the truth.

The historical significance of the De doctrina christiana, important as it is, should not obscure its value as a contribution to homiletic. The first homiletic, though one of the briefest, remains one of the most suggestive. It omits no essential; while it reminds us of the general principles of rhetoric, it emphasizes those applications to preaching which are distinctive; and it proceeds pedagogically. Though the doctrina of the title refers strictly to exposition, and this is amplified and iterated as a constant necessity, Augustine includes specifically and from the start both charm and appeal, and concludes by showing emotional appeal to be the final stage of the comprehensive task of persuasion. Homiletic is an application of rhetoric long established as permanent, consistent, and in both materials and conditions fairly constant. That it is also comprehensive, demanding all three typical styles, including argument in its exposition, winning sympathy in order to urge action, varying its art[26] while holding to its single aim, is most suggestively established here in its first great monument.

Not only does Augustine forbid the arid and the tedious, not only does he insist on emotional appeal; he also vindicates for Christian eloquence the importance of charm. This was the more delicate because charm was both abused by contemporary sophists and still suspected by contemporary preachers. Augustine presents it at once frankly and with just discrimination. To make it an end in itself, he is careful to show, is indeed sophistic; but to ignore it is to forget that preaching is a form of the oratory of occasion.[27] The Areopagus speech of St. Paul,[28] though it is only summarized in Acts xvii, is evidently occasional, and has clear indications of that adaptation to win sympathy which is Augustine’s interpretation of Cicero’s delectare. The speech on occasion, favorite form of oratory in Augustine’s time, had been conventionalized to the point of recipe. The recipes, though he knew them all, Augustine simply ignores; the field he redeems. He shows Christian preaching how to cultivate it for real harvest. History has shown no other direction of rhetoric to be so peculiarly homiletic.

Already Christian eloquence had reached conspicuous achievement in panegyric and more widely in the field of occasional oratory. The pagan sophist must look to his laurels. But these very triumphs had brought the danger of lapsing into too familiar conventions. What in pagan oratory might be no worse than pretty or merely exciting, in Christian oratory would be meretricious. To hold his difficult course, the preacher, as Augustine reminds him again and again, must at every moment steer for his message. He must never deviate. Though sophistic lost its dominance centuries ago, it has never been quite dead, and it always besets preaching. Therefore a constant concern of homiletic is to exorcise it by a valid rhetoric; and no book has ever revealed this more succinctly, more practically, or more suggestively than the De doctrina christiana.

CHAPTER III THE LAST ROMAN SCHOOLS AND THE COMPENDS (FIFTH TO SEVENTH CENTURIES) — References and Abbreviations — A. The Schools of Gaul — 1. Ausonius [2] — 2. Sidonius Apollinaris — 3. Textbooks — B. The Trivium in Compends of the Seven Liberal Arts — 1. Martianus Capella — 2. Cassiodorus — 3. Isidore — FOOTNOTES:

The need of the Church and of the Roman world for such forward counsels, the tenacity of sophistic among reactionaries and conformists, are amply exhibited by the last days of Roman Gaul. In the ancient province a cultivated leisure class, living in the twilight of a great past outworn and doomed, cherished the sophistic conception of elegance and the sophistic habit of education by rhetoric[1] as symbols of their Romanism. Provincials, they were sometimes more Catholic than the Pope; their writers and teachers would not risk being thought less literary than the capital of the world. But in fact Gaul of the fifth and sixth centuries was more Roman than Rome. It was the last territory of the ancient world.

“The poetical fame of Ausonius,” says Gibbon in a contemptuous postscript to a footnote,[3] “condemns the taste of his age.” Whether or not Gibbon took too seriously the fourth-century habit in compliments, Ausonius at least reflects the taste of his age. Fading, therefore, long since to the shadow of a shade, hardly any longer an author, he is nevertheless an important document. Such fame as he may have had once is hardly even considered. Like the rhetors celebrated by Philostratus,[4] like those at Bordeaux whom he himself commemorates, he has ceased to be even a name. He is a collection of trivial fourth-century verses recalled only because they incidentally record contemporary preoccupations. Expert in the metric and the diction that he taught as grammaticus, he could turn a stanza on anything—on a city, a Cæsar, or a sage. Only he preferred topics that came handily in series: the order of the daily round or of noble cities, twelve Cæsars, seven sages, a roster of Trojan heroes. These are topics for Latin verses in school. Many of his poems are evidently, and others probably, school exercises. This in itself, the acceptance of themes as literature, is eloquent of the literary habit of sophistic.

For the work of Ausonius may be summed up as declamatio.[5] Its being mainly in verse hardly modifies its character beyond emphasizing sententiæ. The conciseness imposed upon some of his work by his predilection for epigrams is the balance and word-play of rhetoric, not the focus of poetry. That poetic was rhetoric applied to verse Ausonius was not the man to doubt. His praise of the Bordeaux rhetors for both alike[6] increases the probability that the two were in fact alike; and this is confirmed by his own verse. It is confirmed also by the appearance of several local celebrities now as grammatici, now as rhetores. The function of the former was traditionally both to teach elegant correctness and to expound the poets; of the latter, to train directly for oratory. But the two functions thus distinguished by Quintilian seem in fourth-century Bordeaux to have been combined, or at least to have been exercised successively by the same person.[7]

Evidently the sophistic conception of style as dilation by decoration and literary allusion was still prevalent. To be literary was to dilate. Of this Ausonius sometimes shows a humorous awareness.

But the humor of his address to his stenographer does not make the description of composing the less true to the habit of his time.

The typical declamator appears in Exuperius, “majestic in gait and in great words.”[10] Literary tags, especially in panegyric, come thick.

The speeches and poems of these men, and of Ausonius himself,[12] are typically occasional. Their ideal of aptness is sought by following the recipes for encomium. Literature, long fixed, was to be attained by expert conformity.

Evidently too these encomiasts still had their reward.[13] As official spokesmen they were appointed to public office. Ausonius himself rose with the fortunes of Gratian. From being his tutor he mounted by the steps of comes and quæstor to be in 378 præfectus Galliarum. Such a position dispenses a man from writing for posterity, and effectually prevents him from questioning the permanence of an order of life and thought already spent. Until the Roman world fell apart, it was satisfyingly enclosed in the schools of Bordeaux.

Cherishing of the past, rhetorical education, obsession of style—all this is embodied in the Roman prefect and Christian bishop Sidonius Apollinaris.[14] His letters and poems, current through the middle age as models of style, are consistently in the modes of sophistic. Commemorations, addresses of welcome or congratulation, above all panegyrics, they follow the tradition of declamatio;[15] and their allusions abundantly exemplify both the school practise of his day and the literary preoccupations.

Sending to Perpetuus his discourse at Bourges, Sidonius apologizes for its defects of style.

A long letter to Domitius, describing villa life in detail most interesting to the historian, contains a lively passage of description.

So far Sidonius sounds as if he had a respite from style. The picturesqueness seems to spring from observation. But habit is too strong to let him either stop there or dispense with literary allusions. He goes on:

Balance has never been pursued more anxiously. The teaching of Probus is praised for handing down:

The letter of welcome to Constans has a similar series of balanced contrasts.

Such contrasts in series are sometimes made even more artificial by word-play.

This is the “pomp of Roman speech.”[21]

Sidonius feels himself the representative of a great tradition.[23] That tradition was generally ceremonious regard for usage and a certain anxiety to exhibit culture by literary allusions and by command of the technic of style. Generally, that is, the literary tradition of Gaul was sophistic. Specifically it was declamatio.[24] The school tradition celebrated by the elder Seneca, and made by him and by the declamatores after him the exemplar of oratory, is seen in the pages of Sidonius equally to monopolize the schools and the platforms of Gaul. This is rhetoric, and there is no other. It is fully displayed in the long letters to Lupus[25] and to Claudianus Mamertius,[26] not only in conclusive evidence of detail, but as a conception that is pervasive because it is exclusive. As in the letter above to Probus, there is frequent use of the words declamatio, declamare, controversia, etc.

The most extended reference is the letter to Remigius about the collection which had originally, perhaps, been his desk-book at Rheims, and which had been revised and copied for imitation.

Interesting as is the glimpse of the transmission and traffic of books, it is quite overshadowed by the significance of such a collection. Evidently it was regarded not only as a storehouse for imitation in school, but as a work of literature. Similarly were collected early in the sixth century the declamationes of Ennodius, Bishop of Pavia.[31] The letter of Sidonius specifies, moreover, those literary excellences which were sought alike in teaching and in professional practise. What he means, for instance, by firmness of structura is not cogency of composition, but smoothness of style; and revision to this end involves meticulous adjustment to tongue and ear.

Other references show equally that the exercises of the schools were carried, as in the earlier imperial centuries, into public speaking and literature. The orator’s achievement is not the persuasion of his fellow men; it is his own virtuosity.

The climax of the praise is his disciplina. Declamatio teaches boys to develop an outline at length.

But this is no less the achievement of the finished orator.

The habit of dilation is carried into poetry. Is it not sanctioned by Horace?

Horace is so misapplied because declamatio tends to merge poetic in rhetoric.[36] The poems of Sidonius differ from his prose in little but verse. Three of the longer ones are panegyrics; all are occasional; all show the same habits of style. The literary tags used to sum up the teaching of Probus[37] assign to poetry a kind of appropriateness which belongs to rhetoric. Sapaudus owes his literary reputation to his training in declamatio under Pragmatius, who used to “break the rhetoric benches” with a peroration.[38]

In short, the rhetoric and the poetic of fifth-century Gaul are seen in Sidonius to be following faithfully the sophistic tradition of declamatio. He knows all its recipes. From correctness conceived as archaism and elegance conceived as ceremony, through dilation by the Gorgian figures and by literary allusion, he is constant to the sophistic ideal of expert impressiveness.[39] Augustine had been the pioneer of the Christian future of rhetoric; Sidonius was a complacent reactionary of its decadent Roman past.

a. GRAMMATICA AND DIALECTICA

In this period were written the Latin grammars authoritative throughout the middle age, those of Donatus and Priscian.[40] The former, used generally in two parts as an introductory manual, was so current, indeed, as to become common property and to reduce its author’s name to a common noun.[41] Priscian came to be used as a second book.[42]

The ancient tradition including in the scope of grammatica not only meters and some of the figures of speech, but also the study of poetry through prælectiones,[43] is recognized in several of the preliminary definitions.

St. Augustine refers to this induction into poetry as habitual.

For such teaching the favorite author was Vergil; the favorite authority, the “Ars Poetica” of Horace.

From this period come also the standard medieval textbooks of logic. The logic of Aristotle was mediated to the whole middle age by the translations and commentaries of Boethius. “He translated the Εἰσαγωγή of Porphyry and the whole of Aristotle’s Organon. He wrote a double commentary on the Εἰσαγωγή and commentaries on the Categories and the De interpretatione of Aristotle, and on the Topica of Cicero. He also composed original treatises on the categorical and hypothetical syllogism, on Division and on Topical Differences.”[46] The medieval order of studies was: Porphyry’s Introduction, the Categories, Interpretation, Syllogisms, Topics.

b. RHETORICA

Quintilian was known to both Ausonius and Sidonius, and doubtless generally to the rhetors of Gaul.[47] But his work is addressed rather to teachers than to pupils. More available for the schools was Cicero’s youthful compend De inventione. The early and continued use of this is widely attested. Its contents are as follows:

To this was added in general medieval use the Rhetorica ad Herennium, probably by Cornificius, but thought throughout the middle age to be Cicero’s.[48] When this began its medieval vogue is difficult to determine. Since it is not mentioned by either Cassiodorus or Isidore, it may not have been generally current before the Carolingian revival. Meantime the fourth century added, besides the commentary of Victorinus on the De inventione, the compendious catechism of Fortunatianus;[49] and the fifth century, the longer work of Julius Victor.[50]

The relations of rhetoric to grammar on the one hand and to logic on the other must be considered in determining its function at any period of its history. The ancient grammatica, for instance, included at its best not only metric and some of the figures of speech, but a certain induction into poetry. Its field of composition was pretty definitely marked out in traditional elementary exercises.[51] Rhetorica as conceived by Aristotle or by Cicero concerns itself of necessity with logic. With Quintilian not only proof and refutation, but that estimate of the whole line of argument for which he provides systematic analysis under the traditional head of status, are largely logical. Logic may be used for analysis without presentation. This, indeed, is abstractly its proper function, and indicates its relation to philosophy; but in actual practise, or in a given system of teaching, its relation may be rather to rhetoric, and conversely rhetoric, by yielding its field of inventio to logic, may be reduced to the study of style.

So the estimate of the rhetoric or the poetic of a given period must consider the contemporary view of the whole Trivium. This obligation is obvious where the Trivium is conceived as a unified group of studies; but it is no less important where the three studies are pursued without explicit relations. Indeed, one of the measures of effective functioning in any one of them is the fostering or the ignoring of their relations. A survey of the history of the Trivium in this aspect distinguishes the lingering of ancient educational traditions in the fourth and fifth centuries and their lapse in the sixth and seventh, then that increasing range of cathedral and especially monastic schools which received its historic impulse from Charlemagne. The great monastic schools come to their prime in the eleventh and twelfth centuries; in the thirteenth they yield to the universities. The dominant member of the Trivium in the earliest of these periods is the rhetorica of decadent antiquity; from the seventh into the tenth it is grammatica; from the eleventh on, it is dialectica.

The division of studies into seven liberal arts came to the middle age from Varro’s Disciplina largely through Martianus Capella.[52] His Marriage of Philology and Mercury was widely current for centuries. The allegory implied by his title is carried out in ornate verse and prose through two books. In the other seven books allegory is reduced to the conventional description introducing each of the arts in turn with appropriate costume, symbols, and speech. Thus the work divides sharply into a grandiose allegorical prelude of two books, with a similar prelude to each following book, and a sober, concise, pedestrian compend of grammatica (III), dialectica (IV), rhetorica (V), geometria (VI), arithmetica (VII), astronomia (VIII), and harmonia (IX).

Grammatica, the first of the language studies, claims its ancient territory, including the exposition of poetry. The subsequent treatment is conventional and incomplete.[53]

The Lady Dialectica is more assertive.

The fifth norma annexes that part of grammatica which was most remote from logic; the sixth trenches on rhetorica. The actual extension of dialectica into the ancient domain of rhetoric during the high middle age[55] may have found here some warrant. But Martianus himself does not include these items either in the prospectus immediately following or in the subsequent compend.[56]

Rhetorica enters Book V with such pomp and noise as to frighten some of the minor symbols.

After this fanfare Rhetorica settles down to drill. Though the predilection for declamatio transpires now and then even in this, the compend covers all five parts of the ancient program, and gives to inventio and dispositio more than twice the space granted to elocutio. There is room not only for character and the feelings (ἤθη and πάθη), but for status and analysis of types of argument. Most of the examples are from Cicero. The rhetoric that Martianus preaches is quite distinct from that which he practises. Its proportions are those of the better ancient tradition. It does not even dilate on colores.[58] For all the rodomontade with which it is introduced, the ancient program is still comprehended.

The survey included by the monk Cassiodorus[59] in his Institutiones is hardly more than an enumeration. The seven arts are not considered in their larger aspects; and the Trivium is merely sketched. Rhetorica, after being summarily distinguished from dialectica and divided into its typical considerations and objects, is hastened through the parts of a speech into the emotions. Then defining it anew and dividing it into its traditional five parts, Cassiodorus enumerates status, reverts to the parts of a speech, cites Quintilian and Fortunatianus, and classifies arguments. The meager summary is not even clear. Nevertheless the influence of Cassiodorus in carrying forward the idea of the seven arts is attested by frequent reference.[60]

In the seventh century, and for centuries afterward, a chief purveyor of the lore of the seven arts was the Etymologiæ, or Origines, of the Spanish bishop Isidore of Seville.[61] The work is an aggregation of summaries, not only of the seven arts, which occupy only three books of the twenty, but of medicine, law, holy writ, history, the hierarchy celestial and terrestrial, zoölogy, geography, metallurgy, agriculture, and crafts. Though it has no single guiding scheme, its brief chapters are grouped under headings fairly convenient for reference. It is a guide-book, rendering its first service when books were few and hard to get, and long continuing in vogue. In the thirteenth century Vincent of Beauvais,[62] undertaking more systematically a compend hardly less comprehensive, his vast Speculum, transferred to it whole passages from Isidore.

Isidore appealed to the earlier middle age as a mediator not only of manifold lore, but especially of ancient tradition.

Isidore’s grammatica has the usual contents, including schemata, tropi, and metric.[64] Fabula and historia reflect the ancient elementary exercises.[65] His second book contains both rhetorica and dialectica.

A chapter (xxiv) inserted here groups all the seven arts under philosophia, which is threefold:[66] (1) naturalis (physica), (2) moralis (ethica), (3) rationalis (logica), “in which is discussed how in the causes of things or in the conduct of life truth itself may be sought.” Isidore does not, however, pursue this larger scope of logica. Though he cites additionally Plato’s use of the term to include both dialectica and rhetorica, he goes on to enumerate other divisions of philosophy, and then (xxv) to the usual items of dialectica as presented by Boethius: Porphyry’s Introduction, Aristotle’s Categories, interpretation, syllogisms, division and definition, topics.

The program for the Trivium keeps in the main the ancient proportions. Some usurpation of composition by dialectica, some leaning of rhetorica toward dilation and decoration, may be read into it; but each of the two studies clearly has its own function, and work enough for serious occupation. As in other parts of his aggregation, Isidore makes his summary of the Trivium a list of studies, not a group. He attempts no unified plan. Therefore his putting of rhetorica before instead of after dialectica should hardly be pressed for significance.

CHAPTER IV POETIC, OLD AND NEW (FIFTH TO SEVENTH CENTURIES) — References and Abbreviations — A. Claudian and Boethius — B. Prudentius, Sedulius, Fortunatus — C. The Earlier Latin Hymns [22] — 1. Iambic — 2. Trochaic — 3. Other Measures — 4. Poetic Conceptions — FOOTNOTES:

Lack of literary vitality must show itself conspicuously in poetry. Claudian[1] at the end of the fourth century is typical of the fashionable versified rhetoric. Poet of those who taught and quoted Vergil, but liked Lucan and Statius, he is imitative fluently, prettily, and, except for a certain monotony, expertly. His Raptus Proserpinæ had the longer life, doubtless, because the abundant allusions and decorative descriptions give literary atmosphere without taxing either memory or imagination, and because, though diffuse, it is briefer than much verse narrative current in its time. Most of Claudian’s other poems were occasional, and faded with the occasion.

But the old modes, for all the trivial use of them, were not spent. Boethius[2] showed himself here too the last of the Romans. The man who interpreted to century after century the logic of Aristotle was a poet. Not only is the Consolatio philosophiæ poetic in conception; its lyric interludes are so far above the facile expertness of the time as to suggest comparison with the best Latin achievement. Nothing less was the inspiration of their firm and various technic. Though he seems at first like his contemporaries in looking backward to what was going or gone, he knew better than they what to look for in the Roman tradition. He is never either archaist or sophist. Images of classical reminiscence are used with fresh realization of their source in the sounds and colors of sea and fields and clouds. They are not tags, nor merely allusions, nor decoration. So used, they have much the same freshness as of old, or as the poet’s own immediate observation.

is at once old and original; and the last line of this poem focuses the whole in a pregnant phrase:

Conceived by him habitually in cosmical aspects, nature none the less struck his senses in distinct detail. The philosopher was a poet.

Boethius thus revived the old poetic terseness, as in

not merely because he was too original and too serious to be deviated by the literary habit of dilation, but because he grasped constructively the economy set forth in the De sublimitate[5] as a poetic principle.

Having thus suggested the mood of oppression through images of the blind frustration of man by the heavens and the earth, the lyric contrasts the first, joyous contemplation of nature “under the open sky,” advances to the inevitable later pondering upon the causes of movements so stupendous, and returns to the unyielding inertia of environment, still sharply imaged, but now carried from mood to conviction. The essentially poetic composition of his lyrics saves them from what would otherwise seem too intellectual solutions. Stolidam cernere terram is, indeed, as some other closes of his, a logical conclusion; but it is not reached by a logical process, nor is it mere epigram. It is the final satisfying image of a composition intensely and progressively imaginative. The immense medieval vogue of the Consolatio must often have reminded apt spirits of the true method of poetry.

In detail also Boethius set a chastening example. The spiritual elevation of this very lyric may have suggested for hymns[7] a measure that otherwise would hardly have seemed available. Analysis reveals a fondness for cæsural effects of syncope, for rime, and for subtler recurrences. The next lines are:

But the italics exaggerate recurrences which to the ear are not exaggerated. Boethius does not remind us of literary devices. He never descends to word-play. His rime[8] is neither insistent nor inclined to the later art of stanzas. It is merged in the other suggestions of a various harmony.

The third measure of the third book, echoed now and then in hymns, slows its pace not only by the cæsural pause, but by the predominance of spondees at the onset.

Such reflective lyrics were so readily assimilated to medieval thought that their grave and restrained form must have been instructive in the centuries of poetic transition.

Prudentius[9] devoted much capable and dignified verse in many meters and in several distinct literary forms entirely to the service of religion. The Psychomachia, in some nine hundred hexameters, is an allegory of the soul’s warfare. Such description as that which introduces Avarice[10] handed on from Roman antiquity to the middle age the poetic habit of presenting personified abstractions by appropriate costume, gesture, and speech. The habit is allegory in its most obvious form; the method is the descriptive ecphrasis[11] transferring to poetry the rhetorical doctrine of appropriateness. It is the method of Martianus Capella in the same age and of Alain de Lille in the thirteenth century.[12]

Reviewing his life in the preface to the Days (Cathemerinon), Prudentius seeks a poetry proper to his old age in hymns.

The twelve poems that follow, cast in nine different measures, are lyric reflections on the daily recurrences of cockcrow, food or fast, lamplighting, sleep; or on the festal recurrences of Christmas or Epiphany. They are poems of some length, not hymns in the specific sense to which the word was soon limited. But stanzas selected from them were combined to make the Breviary hymns for Lauds on Tuesday, Wednesday, and Thursday, and for the feasts of the Holy Innocents, the Epiphany, and the Transfiguration. Almost equally familiar in the middle age was the Corde natus[13] hymn taken from the ninth poem.

The Crowns (Peristephanon) is a group of fourteen poems commemorating martyrs. Some are even longer than those of the preceding group; and the narrative of St. Romanus extends to eleven hundred iambic trimeters in five-line stanzas. Other lyric stanzas are pressed into the service of narrative, as well as the better adapted hexameter and elegiac. There is little narrative movement. What Prudentius sought was detailed description, the making of martyrdom vivid. This, rather than the vogue of declamatio,[14] explains a certain diffuseness and the violence of many physical horrors of torture and execution. His descriptive habit is utterly different from Claudian’s. That he was not thinking of rhetoric seems sufficiently evident from his ignoring the recipes of encomium.[15]

The remaining verses of Prudentius, nearly four thousand hexameters, more than one-third of his work, present theology by exposition and argument.[16]

The main work of Sedulius[17] is a hexameter paraphrase of the Bible; but he is better known for his Carmen paschale because this gave to the Breviary two hymns.[18] Its rimes suggest increasing inclination toward regularity, forecasting the time when what the ancient prosody felt as a device of style was to become a device of composition, a recognized method of emphasizing words and of making stanzas.

But rime was recognized slowly as a composing principle. Fortunatus[19] in the next century used it often enough, indeed, to show intention, but still incidentally. His famous hymn Vexilla regis, ignoring rime in its first quatrain, ends every line of the second with the same sound, and three of them with double rime.[20] Contemporary interest in meter appears in the poem replying to Gregory’s request for Sapphics.[21] Most of his many other graceful occasional poems are in elegiacs. One other hymn achieved a fame second only to that of the Vexilla regis, the Pange, lingua, gloriosi. These two great Passion hymns may be taken as typical of the transitional poetic of the time. They are at once old and new. Each is a veritable hymn, not adapted as were the selections from Prudentius, but composed in a popular measure for community singing. Turning back to them from their successors of the great medieval period, one almost inevitably renders them with the same strong stresses. So read, they seem not to lose, but to gain in emphasis and swing. Who shall prove that they were not so rendered in the poet’s own time? But who shall prove that they were? Against such rendering is clear evidence that Fortunatus controlled expertly the ancient quantitative prosody. For it are, first, the measures themselves, which come from popular verse accentual even in ancient Rome, and, second, a shift of speaking habit spreading slowly through the new Roman world. Both these must now be examined.

Here is new poetry, and the beginning of a new poetic. Not graver than Boethius in rejecting decoration, not terser in rejecting dilation, it is more direct, more simply responsive. The communal expression, the imaginative answer of people united, is as distinct poetically from individual reflection as is the communal hope from pensive resignation. The hope that Boethius had, but did not express in his Consolatio, appears here as above all a communal inspiration. The hymns are popular essentially in being the poetry of a society, the kingdom of heaven. Often intensely lyric, they express typically in these earlier centuries the emotions of a community. Their poets, many of them soon forgotten, if ever known, without thought of individual fame sought to give voice to what all felt together. The wide and continued vogue of the early hymns testifies to the validity of their popular poetic.

Their popularity is conspicuous in their verse. The measure above may be found, indeed, among learned poets; but it is originally and usually the verse of common people. Latin popular verse, even in classical times, was probably accentual. Though still called by the grammarians dimeter, this particular measure is certainly accentual as it is used in hymns of later centuries, and as it is rendered later even in hymns of this early period. Is it accentual as rendered even here at the beginning of Latin hymnody? The answer, though still disputed, has been much advanced in the last fifty years. To begin with clear terms, accentual means controlled by stresses; quantitative, controlled by time. All verse beyond mere mechanical exercise and doggerel has both elements; but every verse has one or the other for its control, its rhythm. All verse is something like dance, something like song; but every verse is dominated by the one rhythm or the other. In this sense English verse is accentual. Every expert English poet regards time also, as he is aware of alliteration or subtler recurrences, or as he uses rime; but he sets and holds his rhythm by stresses. In this sense modern French verse is quantitative. Though it regards other elements, including stress, it makes time, as English verse does not, essential in its pattern. The same difference distinguishes the verse of Bernard of Morlaix from the verse of Vergil, and generally medieval Latin verse from ancient. Medieval Latin verse has a different rhythm. Probably the new rhythmical habit began early.

But when we try to determine dates, we should remember that old verse habits give way to new gradually. Poetic does not progress by revolutions. The decorative habit of the Roman de la Rose, though it is now a curious piece of antiquity, survived long after Chaucer had outgrown it and had even exploded it in satire. In verse, too, the three centuries including Chaucer help us to understand the centuries from Sedulius to Bede. With other court poets, Chaucer was bilingual. He not only understood and spoke French; he had French verse in his subconscious mind. True, the French verse of his time shows, more than that of to-day, awareness of stress. The two rhythms were so much less distinct that Chaucer could more easily turn French to the profit of his own development in metric. None the less his rhythm is English. In spite of his ready tolerance of a shift of stress in foreign words, and of the enhancing of his harmonies by long vowels, his rhythm, the pattern or control of his verse, is consistently accentual. The first great English poet, as he used French, indeed, but turned for his poetry to English, kept no less confidently English rhythm.

In the fourth century Latin verse showed distinctly side by side two rhythms. Quantitative verse, long confirmed by Greek example and often directly imitative of Greek models, held the field of culture. The verse of Horace, Vergil, Ovid, it imposed itself upon all educated poets. Its influence controlled the schools through the archaistic teaching of style; and its quantitative prosody continued for centuries to be taught as part of grammatica. But all this while another rhythm was heard from the mouths of soldiers in songs of marching beat. What is somewhat indefinitely known as Saturnian verse moved beneath and behind literature, sometimes broke in half-conscious echoes through learned poetry, then gained the ground lost by literary standards, and finally won recognition as valid poetry in the hymns.[24]

Eventually the manuals of metric, which in any age proverbially lag behind current habit, distinguished the new verse by a new application of an old name. They called it rhythmus. The ancient quantitative verse is generally referred to in the middle age as metrum (metra, metricus, etc.); the new Latin verse, as rhythmus (rithmus, rithmicus, etc.). Bede’s De arte metrica, written early in the eighth century and widely used as a textbook, distinguishes as follows:

Bede is apparently feeling his way, and evidently trying to find warrant for a new poetic in ancient authority. Victorinus, whom he is quoting,[26] may intend by rhythmus and numerus nothing more than their older, more general sense. Bede’s variations and additions make specific and unmistakable application to the verse of the hymns as distinct from metra. The testimony, first to the new habit, and secondly to the recognition of it as a valid and beautiful poetic, is testimony to fact. In a schoolbook of the early eighth century Ambrosians are exhibited as rhythmi; and rhythmi are recognized as a distinct, self-sufficient method of verse.

Though the change of verse habit was slow, though metra were composed long after rhythmi had won the field, and though on the other hand even early hymns that are metrically correct may have been rendered as rhythmi, the progressive prevalence of accentual composition can hardly be doubted; for it answers a shift in the habit of speech itself. By the seventh century Latin was no longer spoken even by the learned, much less by the average monk in England or Spain, as it had been spoken by Cicero. The change of speech tune, doubtless more rapid among men not born to the language, certainly unequal and gradual, seems to be a fact in the history of the language.[27]

Some of the earliest verse generally recognized as rhythmus comes from Ireland.[28] The earliest known hymn manuscript is the so-called Antiphonary of [Irish] Bangor,[29] which is not an antiphonary, but a collection of hymns, prayers, and canticles. In poetic art these range all the way from measures of noble beauty, some of which appear to have been composed as metra, to mnemonic jingle. The following hymn is probably of the fifth or sixth century:

On its face this is a metrum, the familiar iambic dimeter. Rendered as a rhythmus, it would freely disregard quantities, stress the final syllable of each line, and elsewhere generally stress the word-accent: ígnis creâtor ígneús. Though either measure is satisfying, the safer assumption is that the composer intended a metrum.

Probably even older, on the other hand, is St. Sechnall’s (Secundinus, fifth century) fervent but rude praise of St. Patrick in twenty-three stanzas beginning with the successive letters of the alphabet.

Here the ignoring of quantities is such as to preclude a metrum.[30] But accentual interpretation also is difficult. Read as a rhythmus based on the Corde natus measure,[31] which was frequently thus used later, it moves tolerably in certain stanzas (e.g., 4 and 7), but in others, including the two quoted, intolerably violates word-accent. Read by word-accent as a rhythmus of three stresses, it makes extravagant use of two intervening unstressed syllables.[32] Indeed, without some clearer clue than has yet been offered, we can hardly be sure what measure was intended.[33] Much less should we regard it as representative of the early poetic capacity of rhythmus. That capacity is amply vindicated in other hymns of this very manuscript.

Examination of the early hymns measure by measure begins with St. Ambrose, Bishop of Milan at the end of the fourth century. Both his own fame and the recognition of a typical stanza appear in the general use of the adjective Ambrosian to describe many hymns of unknown authorship and sometimes of uncertain date. The canon of his own hymns has, indeed, been determined;[34] but it is less important than his achievement of a type. His answer to a common need established a common form. He had the discernment, first, to select a popular measure often used accentually, and then so to use it as to obey both the popular stress habit and the learned poetic of quantity. That his hymns are valid either as metra or as rhythmi means, though his regard for word-accent shows that he foresaw the latter rendering, that they must have been composed in the former, and probably that time rhythm was still heard in speech. The increase of the stress habit gradually worked a transformation. The regular dactylic close hardly satisfying so short a line, the final syllable, which even in the ancient metric might be long sometimes, came to be stressed always. As a metrum the iambic dimeter ran Ætērnĕ rērūm cōndĭtŏr; as a rhythmus it became Ætérne rérum cónditór. The effect, though different, is not inferior. For the distinctive verse values of rhythmi, heard at their best in later hymns freely so composed, can be discerned even here in the earliest centuries by rendering as rhythmi—as in fact they came to be rendered—two of the most familiar hymns taken from the dimeters of Prudentius.

The accentual habit spread slowly and intermittently as a change of control from one element of verse to another. Both elements continued in the better hymns to vary and enhance what must otherwise become monotonous or bald; but the control, the rhythm, gradually changed from time to stress. Variety through shifting the places of stress[35] appears in several hymns of the sixth century. The Ambrosian quoted by Bede as an example of rhythmus[36] even begins with a stress;[37] and the probable following of word-accent often gives a dactylic opening.

Similar variations appear in the “Versus Flavii ad Mandatum.”

Though possible lingering or revival of a sense of quantities, and the shifting pronunciation of proper names, leave uncertainties (as above, for example in Adam and vitæ), there is little doubt that such variations were not only accepted, but even sought. They often relieve dubious measures; and conversely they are sometimes used in ruder rhythmi so excessively as to blur the verse pattern. One of the Bangor hymns seems to run best as follows:

If so, the syncope in line 2 serves to emphasize vox. But the verse of this hymn is inferior.[38] The better hymns use the variations with better art.

More immediately suggestive of accentual rendering was another soldiers’ measure used by Prudentius in the first poem of his Crowns (Peristephanon), a commemoration of two soldier martyrs. It was better known through a hymn taken from the ninth poem of his Days.

Rendered as a rhythmus, as in the hymn quoted by Bede,[40] the measure, besides ending with a stress, was often divided at the cæsura, to make stanzas of six four-stress lines; for the original line is long to handle singly; it tends to break. None of the other hymns taken from this poem of Prudentius is so familiar and so stirring as this of Fortunatus:

Possibly Fortunatus composed this as a rhythmus; more probably he thought of the soldiers’ marching stresses as reinforcing his metra. The long Bangor hymn ascribed to St. Hilary of Poitiers, and probably very old,[42] is quoted with admiration by Bede[43] as a metrum.

It has the deeper significance of exhibiting an early form of medieval symbolism.

The very different trochaic measure of the sixth-century Irish “breastplate” hymn, with abrupt pauses and rime, seems clearly a rhythmus, and deserves attention least of all for its art.

Less popular, more literary measures, though some of them were used early in the hymnaries, seem to have been interpreted as rhythmi more slowly. This, indeed, is what one would expect; but the evidence is not decisive. The Sapphic stanza, the beautiful meter of the second poem of Boethius,[44] even iambic trimeter, were in time transformed. But in hymns they were comparatively infrequent; and for other uses they were revived with what seems clearly quantitative intention and achievement. Either rendering is beautiful for one of the few Sapphic poems that are essentially hymns, a hymn for Sunday Lauds.

The most striking verse in the Bangor manuscript is that of the familiar Communion hymn.

The spondaic opening and the marked cæsura may be reminiscent of the Quamvis fluente dives auri gurgite of Boethius;[46] and the composition is far superior to the Bangor habit in sense of time values. Not only Irish and popular, but probably much older than this manuscript,[47] the hymn may have been composed as a metrum. By either rendering it is unusually, and at the same time expertly, free. Though the spread of stress rhythm carried some verse that is rude and mechanical, as in this manuscript, it did not of itself forfeit time values, and it opened in the old language new effects of verse.

But the newness of the verse is at best less significant than the newness of the poetry. A jaded world has been refreshed by new imaginative expression. Bede’s exhibition of Christian poetry is not merely pious; nor is it either timid or complacent. He is convinced of a new Latin poetry. As Augustine redeemed rhetoric, so Ambrose transformed poetic, by new motives. The tender image of the Holy Innocents playing with their palms and crowns[48] is neither old nor new; it is the timeless language of individual lyric. But the habitual conceptions of Prudentius and Ambrose, of Hilary and Gregory, are not individual; they are communal. The lyric of the hymns exalts the common emotions of common observance. It is the poetry of aspirations shared not only with all Christians everywhere, but with immediate companions turning work into worship. It expresses the visions of a fellowship.

No exception is found in the early hymns commemorating martyrs. The triumph is not personal. The individual heroism passes into the common hope of released energy and of the triumph of the kingdom of God. As early churches were built over the tombs of martyrs, and later additions made the nave look both down to the original confessione and up to the high altar, so the martyr hymns express both a common gratitude and a common devotion. So a hundred images of light, suggestions of dawn, noon, stars, the ordinary lamp, the candle in church, lead not to individual emotion, but to the poetry of theology.

In every light is the light of the world. The night-light (vigil lucerna) leads up to the giver of light, the creative fire (Ignis creator igneus, Lumen donator luminis).[49] Poetry discerns a new earth because of a new heaven, and finds both one. The poets of the hymns do not, as the Stoics, look down on the material world; they look through it. They neither belittle physical reality nor bow to it; they go on from it. Life in the hymns is not an urgent present and a visioned future; it is all one. The frame of the Christian year opens in the hymns on eternity. Eternal life begins now, and is not survival, but progressive release of human energy by God. Thus the Incarnation is revealed by Prudentius not in versified theology, but in poetic truth. It becomes a cosmic vision.

Though this conception was widespread through the Prudentian hymn Corde natus,[50] and though Prudentius was quite as much philosopher as poet, literary influences are insufficient to account for the consistent continuity of the hymns as thought. Even Fortunatus, who certainly was no philosopher, was uplifted by common visions. The single line Hymnum dicat turba fratrum[51] might be taken as a formula for the poetic of the early hymns. Hilary discerned also the communal appeal of the symbolism of the Old Testament, as Ambrose had discerned a communal verse; but though the hymns owe much to individuals, the character of their lyric grows from unifying communal conceptions of life.

CHAPTER V THE CAROLINGIANS AND THE TENTH CENTURY — References and Abbreviations — A. The Trivium in the Greater Monasteries — B. Grammatica — 1. POETIC — 2. LATIN HYMNS — 3. NARRATIVE HEXAMETERS AND ELEGIACS — C. Dialectica — D. Rhetorica — E. The Poetic of Germanic Epic — FOOTNOTES:

Hibernia, Northumbria, Francia were successively the seats of learning in the period of readjustment after the invasions. Men born to Celtic speech, to English, to Frankish learned the Latin culture and transmitted it to the middle age. In the circumstances the writers of the eighth, ninth, and tenth centuries were primarily teachers; and their chief field was grammatica. The language of learning was no longer for any of them common speech; it had to be acquired even by Italians and Gauls as a second language and a superior. In compensation it was universal, halted by no frontiers. The whole western world of culture was a quartier Latin. Scoti, as the Irish Celts were commonly called, wrote and taught in it by Humber, Rhine, or Loire; Bede wrote in it his Church History of the English People; Alcuin was brought from the school of York to apply it to the education of the Frankish empire, and his companion and successor was the Bishop of Orléans, Theodulf, a Goth from Spain.

Grammatica became thus more important than ever. It opened not only learning in general, not only literature, but especially the interpretation of the liturgy, the offices, the creeds, and the Scriptures. Charlemagne’s care was to secure a clergy that should be first educated and then educating. The mission of the Church to teach through the universal language of western Christianity was exercised partly through the cathedrals, mainly through the monasteries. In these centuries rose such great monastic schools as Fulda, St. Gall, and Tours. As the physical preservation and circulation of texts depended on the scriptoria, so on the masters of the monasteries depended not only specific training for the religious life, but much of the more general training in divinity and most of the seven liberal arts.

The monastic slant suggests a narrowing of culture. But the restriction of the seven arts must have been due quite as much to those other conditions which gave the preponderance to grammatica. The achievement of the age was preparatory. That the age of preparation was also an age of revival is witnessed best by the outstanding teachers. The Scoti who fled from the Danes to the Continent were remarkable no more for their preservation of the last vestiges of Greek than for their intellectual eagerness. They were a stirring leaven. The greater schools founded during this period, such as that at Wearmouth to which English Benedict had brought store of books from Rome and Vienne, show an impressive succession of teachers: Bede, Alcuin, Rabanus, Loup, Remi, Gerbert.[1] The former three are sufficient assurance that culture was safe; the latter, that it was advancing.

A work on the Elements of Philosophy, ascribed to Bede, closes with the following list:

Toward the close of the tenth century the same order of studies, except for the transposition of geometria and musica, appears in the school at Speier. Though the details of the reminiscences prefixed by Walter of Speier to his Passion of St. Christopher[3] are obscured by figurative language, allusions, and other devices of style, he shows unmistakably, after his first lessons in psalmody, a full course of grammatica, including much metric. His references seem also to indicate both the elementary exercises beginning with fabula, and prælectiones on Vergil, Ovid, Horace’s Ars poetica, and the metra of Boethius. Dialectica, which he entered by the door of Porphyry, he recalls less distinctly. Rhetorica, though remembered in her usual garb of flowers, evidently included declamatio. By the end of the tenth century, then, a typical monastic school, though still spending most time on grammatica, seems to have offered an ample trivium.

Donatus and Priscian, with the other grammarians of the declining Empire,[4] kept their authority. They were successively adapted to changing needs in manuals by Bede, Boniface, Paulus Diaconus, Alcuin, Loup, Remi, Gerbert, Abbo, Ælfric. That grammatica thus engaged the best teachers of the time is evidence of its cardinal importance. At Chartres, by the tenth century, grammaticus was the usual name for headmaster.[5]

The study of figures, both those usually included in grammatica and those assigned to rhetorica, was applied to the interpretation of holy writ. Augustine[6] had pointed out that the Scriptures not only use figures, but explicitly mention allegory and parable. Bede’s brief summary De schematibus et tropis sacræ scripturæ[7] is thus typical both of elementary teaching and of medieval habit of reading.

That neither Bede nor Alcuin specifically defines grammatica in the traditional terms as including the study of Latin poetry may mean no more than that neither wrote comprehensively on the whole subject. The prælectio can hardly have been neglected by the Scoti, or by Bede himself. The definition of Rabanus in the ninth century[8] not only resumes the whole ancient scope, but puts the interpretation of the poets first. That Boethius was added to the list of classics[9] is significant of the influence of his metra even on the hymns.

When Bede tells his boys to look at all the first syllables[10] of a manuscript page of hexameters, because these syllables must be long, he is not precluding either nicer points of metric or wiser consideration of poetry; he is very practically teaching Latin quantities. His book offers much more; and though its subject is only metric, it takes pains to distinguish rhythmic,[11] and closes with that classification of poetry by Diomedes which was to be often repeated.

Far as this is not only from Aristotle, but from Vergil, the shift of emphasis from composition to style remained for centuries characteristic of medieval Latin poetic, and opened the way for the confusion of poetic with rhetoric.

(1) Iambic

The best known hymn of this period is by Rabanus (ninth century).

This keeps generally the quantities of the metrical dimeter.[13] Rime is insistent in ruder hymns. The two following, Irish of the eighth century, though substantially correct as metra in some stanzas, seem to be rhythmi. The first is alphabetical; the second, a lorica. As rhythmi they are relieved by shift of stress in the places indicated, and probably intend it elsewhere; i.e., they are most satisfactorily read by word-accent.

(2) Trochaic

Rhythmical use of the Corde natus measure[14] is suggested by frequent disregard of the distinction in the original meter between trochee and spondee within the line, and of the dactyl at the end.[15] The popular swing of this rhythmus is felt in a well known hymn of the eighth century:

Freer use, with both end-rime and occasional internal rime,[17] as in lines 3 and 5 above, appears in an Irish hymn ascribed to St. Cuchuimne.

A Septuagesima hymn in a tenth-century manuscript is so united by the iteration of Alleluia that the poet felt no need of rime.

A much simpler trochaic measure is heard in one of the most popular of medieval hymns. Found in a manuscript of the ninth century, it may well be earlier.

(3) Other Measures

Sapphics, which of course exercised the skill of the learned,[18] are occasionally convincing in a hymn.

One Sapphic, doubtfully ascribed to Paulus Diaconus, had currency enough to furnish later a memory stanza for the notes of the scale.

The striking measure of Sancti venite[19] is not forgotten.

The second measure of Boethius, exhibited by Bede in two poems,[20] is echoed in the Tanquam præcipitans turbo regentes of Sedulius Scotus[21] and appears in a fine Assumption hymn of the ninth century.

In a few hymns of this period the measure seems to be derived from one used twice by Prudentius.[22]

A rhythmus of uncertain date is most plausibly assigned to tenth-century Verona.[23]

To read this as a rhythmical senarius disregards not only many quantities, as might be expected, but also many word-accents. The measure of the hymn just above is suggested by the doubly dactylic close and the generally long and stressed opening. But so to render it is again to violate many word-accents, including almost all those of the characteristic second foot. The word-accent is generally kept by rendering:

It is hard not to think that at least the final rhythmical dactyls were in the composer’s mind. If so, we have rhythmical dactyls not only as occasional substitutions, but as constituent; and the easiest rendering makes them constitute the whole measure.

Carolingian hymn-writers ranged in art all the way from the expert and fluent metrist Sedulius Scotus to the undisciplined Gottschalk,[25] who would write twelve stanzas on a single rime such as Christe, mearum | Lux tenebrarum,[26] or Spes mea, Christe, | Rex benedicte.[27] Thus he rimes even Sapphics. With much diffuseness and jingle he has sometimes a lyric appeal that forecasts the more sentimental hymns of later centuries, and is as far from the gravity of the elder habit.

Hymnody is typically communal and popular. Such poetic opportunities inspired and authorized in the Carolingian period some verse more valid as devotion than as poetry. Here and there manuscripts have preserved local commemorations which make little pretense beyond grateful mnemonic. But hymns of higher achievement show that the new Latin verse given by the Church to the last days of the Empire was appreciated and carried forward as poetry; and the rhythmical adaptations of measures not used before evince an active, and often an expert, poetic.

But the characteristic verse of the period is hexameter or elegiac. Thus Theodulf composed even the familiar Palm Sunday hymn.

Alcuin devotes fifty-five hexameters to a conflictus between spring and winter, and celebrates York in over sixteen hundred.[28] The elegiacs may seek Ovidian recurrences.[29]

Of the many narrative poems employing these measures the commonest were the passiones, or saints’ legends, usually with at least a rhythmic bent, and by the tenth century habitually rhythmi with internal rime. The tenth-century Vita et passio sancti Christophori of Walter of Speier, though it abundantly exemplifies both the metrical training and the study of Latin poets that he mentions in his introduction, shows also the trend of the time.

Some, at least, of these longer poems were cumulative school exercises. A promising theme in the imitative verse that was commonly part of the study of the Latin poets would be commented by the grammaticus, revised according to his criticism, and kept by him for later rehandling or extension.[31] Thus the verse, even with the rhythmical habit established, attended to Latin quantities. Thus also classical reminiscences, especially Vergilian, are frequent; and Walter shows the continued vogue of Prudentius. For the hexameters most typical of the period are literary exercises.

Logic followed the Boethian tradition handed down by Isidore.[32] Alcuin, though his manual is meager, repeats in his tract on the Trinity St. Augustine’s view of the importance of this study[33] for the defense of the faith. Rabanus makes it theoretically central.

But the turn of dialectica to dominate the Trivium was not yet.

Alcuin’s adaptation of the De inventione of Cicero, Walafrid Strabo’s enumeration of the five ancient parts of rhetoric, do not prove the use of the whole ancient program. Even the ancient texts would not of themselves carry on the ancient method. The “quæstiones civiles” often quoted from Cicero’s opening definition could hardly carry their ancient content either in a society disturbed by the invasions or in a society reorganizing under feudalism. Moreover the teaching of rhetorica, even when it kept touch with Roman method, was likely to lean on the declamatio handed down by the schools of Gaul.[34] For all these reasons the ancient lore naturally most sought and most used was elocutio, the counsels of style. The function of rhetorica is usually described by some such verb as ornare.

Little beyond this is suggested by the summary of Rabanus. Repeating once more that the field of rhetoric is quæstiones civiles, he adds:

The passage is reminiscent of St. Augustine’s De doctrina christiana;[35] and eight later chapters (xxix-xxxvi) follow this closely, sometimes continuously and word for word. But Rabanus is at once less specific and narrower as to style. As to composition in the large he says hardly anything; and he seems to miss the cogency of Augustine’s own order.

The larger and more vital conception of rhetoric, which was at least before the eyes of Rabanus, seems more to beckon Loup de Ferrières. Man of letters in his intellectual eagerness[36] as well as in his style, and teacher as well, he makes requisitions on the libraries of his friends. The Quintilian that he needs is not the volume of selections, but all twelve books;[37] the Cicero, not only the common De inventione, but also the book whose recovery by Poggio in the fifteenth century was one of the literary events of the Renaissance, the De oratore.[38] If these two cardinal works of the better ancient tradition were not much sought, at least they were available.

The “three styles” seem already to have been transferred in school from rhetoric to poetic, and exemplified from Vergil.[39] Mature practise was already attentive to prose rhythm. Abbo of Fleury, much preoccupied with this, was also fond of alliteration, and sometimes marked his balances with rime.

Though the extremes of this passage go beyond Abbo’s normal practise, they appear also in tenth-century ceremonious letters.[41]

This is the period also of Germanic epic: the Anglo-Saxon Beowulf (probably eighth century), Waldere, Finnsburgh, and Maldon, the Hildebrand (early ninth century) of the continental Germans, the Scandinavian “Elder Edda.”[42] Though little connection is apparent between these verse narratives and the Latin poetic with which they are contemporary, there may have been some.[43] Anglo-Saxon epic is of the time of Bede. The Walter legends are known largely through the Latin hexameters of Ekkehard;[44] and other learned clerks found native epic worth while not only as history, but as literature.

What has been preserved shows the primary epic appeal of legend not exotic and imported, but handed down in folklore still orally active. This is not at all to say that they are history as opposed to fiction. Their historical value, however great, is accidental. Their facts, already centuries old, have been shaped by tradition. Their Sigurd or Hildebrand is seen through a magnifying mist. Epic is never, in our modern sense, history. It is the glorification in song of a hero; and primary epic has its own authentic appeal from singing a hero that still belongs to the poet and to his hearers and still beckons their communal dreams.

Thus Germanic epic, taking us farther back through legend into myth, gives a more immediate sense of oral tradition. There is even an eery likeness, as of the most ancient poetic repeating itself, between the minstrel in the Beowulf and the minstrel in the Odyssey.

So Hrothgar’s minstrel is represented as singing songs of former heroes to awaken joy in hall along the mead-bench. Among those thus inserted in the Beowulf is the lay of King Finn, which has come down also in another form. As the Greek minstrel turns old songs to the praise of the hero present before him, so the warriors celebrating in hall Beowulf’s killing of Grendel turn the legend of Sigmund.

In such passages we seem to be near the roots of verse narrative.

The verse narratives of the Germanic peoples during this period are poetically homogeneous. Hildebrand, indeed, is more stinted than Beowulf, and the north inclines more than the west toward lyric; but they all have essentially the same poetic.[45] Their epic conception is typically not of a progressive story, but of a situation. The hero is imagined in a crisis. Sometimes abrupt or stinted, they nevertheless prevail by unity. This mainspring of their poetic is their habitual means toward tragic intensity.[46] Even more constant is their movement in detail. The verse consists of two staves separated by a marked cæsura, but corresponding by alliteration. The alliteration is not, as in Latin verse, an added suggestion; it is constituent; it makes the verse.

No less essential is the two-stave movement, so strong in Germanic habit that it may well have been influential in handling even Latin hexameters with cæsura reinforced by rime.[47] In Anglo-Saxon the staves show distinct recurrent types; and the verses generally tend, as above, to “run on,” whereas the Old Norse are oftener composed in the fashion of the “closed couplet.” But these differences are unimportant beside the constant binary movement. The verses are not equal in number of syllables; their stress rhythm is patterned in alliterated pairs.

All primary epic is thus concrete. It speaks habitually in the immediate terms of the five senses. But the habit of images crystallized among the Germanic poets in a conventional epic diction. Their style is deliberately removed from common speech. Its most obvious traits are designation by descriptive compounds and accompaniment by descriptive epithets. A lord is “land-ruler,” as above, or “prize-giver,” or “hoard-ward.” His warrior is “hall-counselor,” “earl’s hope,” “rugged-in-war.” The ominous raven is “sallow-brown, swarthy.” Ships especially command a whole store of such phrases as those of the seventh and eighth lines above. Germanic epic has a distinct poetic language.

CHAPTER VI RHETORIC AND LOGIC IN THE TWELFTH AND THIRTEENTH CENTURIES — References and Abbreviations — A. The Trivium at Chartres, Eleventh and Twelfth Centuries — B. The Trivium in Hugh of St. Victor — C. The Metalogicus of John of Salisbury — DIGEST OF METALOGICUS — D. Thirteenth Century Surveys — 1. The Anticlaudianus of Alain de Lille — 2. The Speculum Doctrinale of Vincent of Beauvais — 3. St. Bonaventure de Reductione Artium ad Theologiam — 4. The Trésor of Brunetto Latini — FOOTNOTES:

At the fall of Rome the Trivium was dominated by rhetorica; in the Carolingian period, by grammatica; in the high middle age, by dialectica. The shift of emphasis to logic probably began in the eleventh century. Even Chartres, renowned for its teaching of grammatica, shows hints of this under Fulbert.[1] In the next century the theory of logic was fortified by commanding in Latin translation those parts of Aristotle’s Organon which had not been available; and its practise became more urgent through the historic debates as to universals. By offering thus the most active training in composition, logic confirmed the restriction of rhetoric to style. John of Salisbury, after giving full scope to grammatica, focuses his great book on dialectica. Rhetorica he merely mentions; it claims none of his thought. Nor does any other leader of the high middle age treat rhetoric as active in the intellectual processes of composing. Rhetoric has no educational vitality. The vital study that taxes and develops men’s minds is logic.

In detail, grammatica at Chartres[2] during the great century of the school (about 1050-1150), shows a full development of prælectio and distinct cultivation of rhythmic.[3] Rhetorica, except in dictamen and in some application of the larger ancient precepts of composition to preaching, is at a standstill.[4] The more significance therefore attaches to a short poem of Fulbert summing up the differences between rhetoric and logic: the one concrete, current, reasoning in enthymemes, aiming at persuasion; the other abstract, syllogistic, aiming at conviction.[5] This giving of the full ancient scope to rhetoric in theory may be a reminder, may be even a protest. At any rate the practise had no such scope. The Chartres manuscript[6] containing these verses is an eleventh-century collection of traditional materials for the study of logic. The study was advanced by Gilbert de la Porrée.[7] Thierry’s collection[8] of the traditional writers on the seven arts, Bibliotheca septem artium, or Heptateuchon, gives 190 leaves to grammatica,[9] 88 to rhetorica,[10] 154 to dialectica.[11] The remaining 160 leaves of the two large volumes are devoted to the Quadrivium. Thierry’s prologue,[12] summarizing the functions of the several arts viâ Martianus Capella, distributes those of the Trivium as follows: through grammatica, elegance; through dialectica, logical coherence; through rhetorica, ornament. Oral composition, as distinct from revision for style, seems to have no pedagogy except through logic.

Hugh of St. Victor’s[13] Lore of teaching (Eruditio didascalica, or Didascalicon) is neither a compend nor a program; it is a concise philosophical survey of education. Though his primary concern is with what is to be studied, and to what ends, he sometimes gives also acute hints of method. Book I (De studio legendi), a preliminary survey of typical directions of study, ends (xii) with logica, Hugh’s general term for the Trivium as a group, i.e., for all language studies. Considering language as expression (logica sermocinalis), these are indeed the usual three; but considering language as thought (logica rationalis), they are only two, dialectica and rhetorica. Book II (De discretione artium) expands the exposition to determine the places of all seven arts in a scheme of philosophy. First the Quadrivium, then the manual arts, are classified under the traditional threefold division:[14] theory (1) of thought, (2) of morals, or conduct, (3) of technical skill. Outside this division remains the Trivium, Hugh’s logica, as a fourth and final group. Though logica may be divided simply, as in Book I, Hugh now offers a more philosophical division by function in composing. By the latter we have (a) grammatica by itself as having no such function, and (b) rhetorica and dialectica together as both involving investigation (inventio) and the processes of arrangement, and revision (iudicium). These two composition studies are thus seen to belong, as probable proof (probabilis), between absolute, or abstract proof (demonstratio) and plausible, or illegitimate proof (sophistica).

PHILOSOPHIA

From these classifications Hugh proceeds in Book III to practical considerations: of studies as training (disciplina), of the interrelations of the seven arts, of a scheme of reading (ordo legendi), of meditation, etc. The three remaining books deal with sacred studies.

Hugh’s term logica expresses a conception of the Trivium as an integrated group. Less obvious, but hardly less significant, is the importance given implicitly to logic (dialectica). Though rhetoric is recognized as having theoretically a function in composing, it receives otherwise but little more attention than in the cardinal treatise of medieval pedagogy, the Metalogicus.

The most extensive reasoned medieval survey of the Trivium is the Metalogicus[15] of John of Salisbury. Though this devotes most space to logic and to the logical aspects of other language study, it is a unified and carefully coherent presentation of all teaching that deals with words. Statesman as well as scholar, more widely known, perhaps, than any other man of his time, and more widely conversant with its movements in church and state, secretary to two archbishops of Canterbury, rounding out his life as Bishop of Chartres, he devoted his best thought to the Metalogicus. The classification underlying its first two books is that of Hugh of St. Victor.[16]

By the conception of the Trivium as twofold, rhetorica is theoretically subordinated; and in John’s working out of the scheme it is ignored. Barely mentioned,[17] it appears to have no distinctive composing function. Part of its ancient function seems to be implied now and then under grammatica; more is certainly transferred to dialectica, with which John connects most of what he discerns of composition as a study. His scheme can be comprehended only as a whole and in sequence. The survey of logica in Books I and II is not primarily an analysis. Rather it develops the functions of language studies in progressive stages. This procedure, too, as well as some of its important details, may have been suggested by Quintilian.[18] John is concerned less with division than with order. His own order is so significant and so carefully marked[19] that it should be followed step by step.

Book I. Grammatica

i-vi. The opening takes occasion from certain opponents of the Trivium.[20] “When logica was derided, and its envious opponent provoked me, in spite of my indignation and protests, by almost daily disputes, at length I accepted trial, and have studied to reply to his calumnies” (824 A) ... “Since I have undertaken the defense of logica, the book is entitled Metalogicon.”

vii-ix. Eloquence is natural, not in the perverted sense that the full exercise of speech is instinctive, but only to the extent that speech is the peculiar opportunity of mankind.

x (837 B). “Logica, then, to show the widest meaning of the word, is the theory of speaking or of discoursing. Sometimes it is contracted to the extent of limiting the force of the word to theories of discourse. Whether, therefore, it teaches the ways of reasoning or offers a rule for all speech, they are evidently unwise who call it useless; for either [the narrower or the wider scope] is taught by most famous theory as necessary. The twofold meaning of the word comes from its origin in Greek; for there λόγος means now speech (sermo), now theory (ratio). But that its meaning may be extended most widely, let us assign to it at present the control of all speech, so that nowhere it may be proved useless, and so that in its more general sense it may appear as a whole very useful and necessary.”

xi. The idea of any art is to further nature by theory.

xii (839 C). “But since artes are of many kinds, the first of all for a mind bent on wisdom are the artes liberales. All these are included in the theory either of the Trivium or of the Quadrivium; and so great efficacy they are said to have achieved among the ancients, who taught them assiduously, that they opened all reading, roused the mind to everything, and sufficed to resolve the difficulties of all questions which can be settled. They to whom the theory of the Trivium expounded the secrets of all speech, the law of the Quadrivium the secrets of all nature, needed no teachers to explain their books or resolve their questions.”

xiii (840 A). “Of all these the first is logica, in that part of it which deals with the first teaching of speech.... This is grammatica, the lore of speaking and writing correctly, the origin of all liberal disciplines ... the cradle of all philosophy ... the first nurse of all literary study.”

xiv-xvi. Grammatica imitates nature by keeping congruity of thought.[21] For instance, it does not tolerate adjectives of secondary application with nouns of primary application.

xvii (847 A). “In other things, too, grammatica imitates nature; for the precepts of poetica set forth the habits of nature and exact of the craftsman in this art that he follow nature—to that degree, indeed, that the poet shall not depart from the footprints of nature, but apply himself to stick to them in manner, gesture, even word. Moreover the theory is to be kept not only in feet or tenses, but in ages, places, seasons,[22] and other details beyond our present purpose; for all these come from the workshop of nature. Indeed, poetica stays so close to the things of nature that many have refused to include it in grammatica, asserting that it is an art in itself, pertaining no more to grammatica than to rhetorica,[23] though so far related to both as to have precepts in common. Let them snarl about this who will. I will not keep up the dispute; but under favor of them all I think that poetica is to be assigned to grammatica as to the mother and nurse of its study.... Either grammatica will hold on to poetica, or poetica will be turned out from the number of the liberal disciplines.”

xviii-xx. Grammatica deals both with precision and with imagery, both with denotation and with connotation. It includes letters, syllables, phrase, sentence-form, punctuation, figures, metric—everything that can be taught verbally.

xxi-xxiii. It has occupied persons no less eminent than Cæsar and Cicero. It is a practical guide to utterance and to learning. The objection derived from Seneca is insufficient. For the practise of philosophy and of virtue the important approaches are reading, teaching, meditation. Of these the root and foundation is grammatica.

xxiv. Actually the prælectio is vindicated by such a grammaticus as Bernard of Chartres.[24]

(853 C) “He, then, who aspires to philosophy, let him lay hold of reading, teaching, and meditation, with the practise of good works, lest God be angry and what he seemeth to have be taken from him. But since lectio is equivocal, applicable both to the practise of teacher and learner and to the absorption of one studying writings for himself, let the one, the interchange of teacher and learner, be called, to use Quintilian’s word, prælectio, the other, applied to the scrutiny of meditation, simply lectio. On the authority, then, of Quintilian, the grammaticus in his prælectio ought to attend to such details as to ask to have the verse analyzed into the parts of speech and the appropriate feet, which ought to be known in poems. He should take exception to barbarisms, improprieties, or other transgressions of the law of speaking—not, however, that he should find fault with poets for metrical necessities which, though faults in prose, are called virtues in verse, since force of necessity commonly wins the praise of virtue for what cannot be denied without sacrifice. Metaplasm, sentence variation, figures of speech and such various iterations as may be present, the theory underlying this way of speaking or that—all these the prælectio should point out and impress upon the hearer’s memory by frequent warnings.

(854 A) “The prælectio should make authors yield, without holding them up to ridicule, the feathers with which, crow-fashion, they have decked their works from various disciplines, to make the style more becoming. The more disciplines the teacher is imbued with, and the more abundantly, the more fully he will discern the elegance of authors, the more clearly he will bring it out in teaching. For they by the diacrisis which we may call illustration[25] or visualization, when they had undertaken in bare outline story, plot, fable, or whatever else it might be, would develop it with such abundance of disciplines and such charm of sentence and style that the work when completed seemed the image of all the arts.

(854 B) “Grammatica and poetica, indeed, are entirely fused and control the whole surface of what is expounded. Campologica, as it is called, contributing descriptive amplification of proof,[26] looses its theory in a blaze of gold; and rhetorica with store of persuasions and brilliance of style rivals the brightness of silver. Mathematics is borne on the wheels of its Quadrivium and, hard on the heels of the others, has woven its own figures and charms in manifold variety. Science, having searched the counsels of nature, brings from its own storehouse manifold charm of figures. Moreover that which rises above the other parts of philosophy—I mean ethics—without which not even the name of philosopher abides, surpasses all the others in the gift of ornament that it brings. Sound Vergil or Lucan, and there, whatever philosophy you profess, you will find its making. In proportion, therefore, to the capacity of the pupil, or to the industry and diligence of the teacher, the fruit of the prælectio auctorum is constant.

(854 C) “This used to be the habit of Bernard of Chartres, in our modern times the most overflowing spring of literature in Gaul. In his reading he would show first what was simple and regular. Grammatical forms, rhetorical figures, quibbles of sophistry, relations of the passage to other disciplines, he used to bring out clearly—not, however, by teaching everything at every point, but by adjusting to the capacity of his pupils and to the time of the instruction. Since appeal of discourse is either in precision, that is in the nice adjustment of adjective or verb to noun, or in imagery, that is in passing by comparison from one sense to another, he used to inculcate these in the minds of his hearers whenever he found occasion. Since memory is strengthened and talent is sharpened by practise, he would spur some by exhortation, others by punishments, to imitate what they had heard. Each of them was required to account on the following day for what he had heard on the preceding, some more, some less. For with them the preceding day always taught the following.[27]

(855 A) “The evening exercise, which was called declinatio,[28] carried such abundance of grammar that any one keeping at it for a whole year, provided he were not too stupid, would control the principles of speaking and writing and could not remain ignorant of the meaning of expressions in common use. But since no school, nor any day, should be without religion, such a subject was proposed as would upbuild faith and morals and animate the group, as by common discussion, toward good. The final item, moreover, of this declinatio, or rather of this philosophical discussion, exhibited the footsteps of pious remembrance. The souls of the departed, by devout offering of the sixth penitential psalm [De profundis] and the Lord’s Prayer, were commended to their Redeemer.

(855 B) “For those whose assignments were elementary exercises in imitating prose or poetry he set poets or orators and prescribed close imitation after showing the art of connection and of sentence close.[29] If a boy had brightened his work by sewing on a piece from some one else, he would show that the theft was detected, but very often would inflict no punishment. But if the borrowing was misplaced, with modest kindliness he bade the boy come down to express his author’s likeness; and his own practise was such that in imitating his predecessors he became a model for his successors. He also taught among the elements and fixed in mind the force of composition,[30] the achievement of thought and of phrase, the character of the style, whether thinness or plausible abundance, extravagance or just measure.

(855 C) “Stories and poems, he used to say, were to be read carefully, not on the run; and of each pupil he required as a daily task something memorized with careful attention. But superfluous reading, he would add, should be shunned, famous authors are enough. To follow what every one, however unimportant, has ever said is to regard oneself either too meanly or too boastfully. It holds back and obstructs minds which would otherwise make better use of their leisure; and what displaces something better is so far unavailing that it cannot even be called good. To explore all papers and ponder all writings, even those not worth reading, is no more to the purpose than to attend to old wives’ tales. For, as Augustine says in his De ordine: ‘Who shall call that man uncultured who has not heard of the flight of Dædalus, or a liar for asserting it, or impudent for questioning it? I always feel deep pity for those of our friends who are accused of ignorance if they have not answered what was the name of the mother of Euryalus, and who dare not call the people who ask such questions shallow, impertinent, and curious.’ So says Augustine both neatly and truly. Therefore it was rightly reckoned by the ancients among the virtues of a grammaticus that there should be some things which he did not know.

(856 A) “Since in all the preliminary exercises nothing is more useful than to accustom oneself to what ought to be done expertly, Bernard’s students would daily write prose and verse and practise themselves by exchange of criticism. Nothing is more useful than this exercise for expression, nothing more promotive of learning; and its greatest contribution is to the conduct of life, provided this insistence be controlled by charity, and progress in literature contribute to humility.”

The last chapter (xxv) quotes at length the laus grammaticæ of Quintilian I. iv. 5-6.[31]

Book II. Demonstrativa and Dialectica

Proem. “The course of the former book has sufficiently, I think, disengaged the truth that grammatica is not useless, and that without it not only eloquence falls short, but the way toward the other expressions of philosophy is not open.”

i. Logica, being the theory of discourse, embraces both investigation and judgment.[32]

ii. Knowledge of truth being for them the highest good, the Peripatetics divided philosophy into two parts: natural, or physical, and moral, or ethical.[33] But the difficulties arising from insufficient control of discourse “demonstrated the need of determining and publishing a lore which should distinguish words and concepts[34] and dissipate the mists of fallacies. Here, indeed, as Boethius asserts in his second commentary on Porphyry, is the origin of the logica disciplina. For there had to be a lore which should distinguish the true from the false and teach which reasoning holds the path of [absolute] truth, which of probable, which of assumed,[35] and which should be distrusted. Otherwise truth cannot be found by reasoning (858 C).... The rules of the art were seized and handed down finally by Aristotle.”

iii. (859 C) “Later in time than the other disciplines of philosophy, this is first in place. For beginners in philosophy it is a prerequisite, as the interpreter of words and concepts, without which no item of philosophy comes precisely to light. He who thinks that philosophy is taught without logica, i.e., by [direct] cultivation of wisdom, may as well do away with theory in everything, since this is the domain of logica.... The very name comes from its being an aid and a test of theory. Plato divided it into dialectica and rhetorica; but those who estimate its efficacy higher give it more, i.e., demonstrativa, probabilis, sophistica. Demonstrativa begins in the first stages of training, and passes on into the next. It is satisfied only by necessity; provided a thing ought to be so, it pays little attention to whether or not it appears to be so. This becomes the philosophical majesty of those who are teaching precisely, a majesty grounded, quite apart from the assent of an audience, on its own will. Probabilis, on the other hand, is occupied with what appears to be so to all, or to many, or to intelligent observers, with what is best known and most probable to them, or with what follows therefrom. This includes dialectica and rhetorica, since logician and orator alike striving to persuade, the one an opponent, the other a judge, think the [abstract] truth of their arguments makes little difference, provided they keep what seems to be true. But sophistica, which is apparent and not serviceable wisdom, assumes the likeness either of probability or of necessity, little caring what this or that may be, while it involves whatever is discussed in fanciful images and deceptive shadows. Dialectica, that member of the Trivium which all approach from this side and from that, but few, in my judgment, really pursue, neither aspires to dogma, nor is drowned in the waves of politics, but analyzes truth by prompt and reasonable probability.”

iv. Dialectica, moreover, is the art of effective debate.

v-viii. Logica has for its distinctive function to serve as effective instrument. It is not an end in itself. So perverted, it becomes the absurd and deplorable occupation of senility.

ix-xi. (866 C) “Dialectica, which among the servants of eloquence is most alert and prompt, avails each man according to the measure of his knowledge.... Deprived of the strength of the other disciplines, it is maimed and almost useless; thriving with their vigor, it is strong to overthrow all falsehood, and always suffices at least to reach a probable conclusion.” From my own teachers, to whom I returned years after they had schooled me, I conclude that “as dialectica advances other disciplines, so if it remain alone, it lies bloodless and sterile, and does not engender the fruit of philosophy in a soul not impregnated from other sources.” Of itself it can only despatch issues, not rise to others.

xii. Dialectica operates in all disciplines wherever the issue is abstract.[36] It leaves to rhetorica whatever is hypothesis, i.e., whatever involves concrete circumstances: who, what, when, why, how. It makes no address to the public, expects no legal decision.

xiii-xv. Though each division of philosophy has its own field of inquiry and its own principles, yet logica supplies methods common to all, as it were theory in a nutshell. A problem in dialectica considers choice and avoidance, truth and knowledge, whether for itself or as aiding inquiry where opinion divides.

xvi-xvii. Review of the value and place of Aristotle, of the right use of Porphyry’s Introduction, and of other typical cases in teaching. Bernard (875 D) of Chartres and his followers took great pains to heal the breach between Plato and Aristotle; “but in my opinion they came late and labored in vain to reconcile in death those who differed as long as they lived.”

xviii-xix. Certain errors of those who profess Aristotle can hardly be overlooked: the burdening of tender shoulders, the making of Porphyry cover the whole ground, the misinterpretation that simplifies Aristotle by substituting Plato or something remote from both.

xx. The last chapter, much the longest, presents Aristotle on genus and species.

Book III. Topica

i-iv. A survey of the teaching of categoriæ, prædicamenta, and interpretatio as preliminary begins with general advice (890 D). “The exposition of every book should be such as to furnish most readily the knowledge of what is written. No occasion should be sought of introducing difficulty; everywhere the way should be opened. That was the practise, I remember, of Abelard ... (891 A). Thus Porphyry should be read so that the significance of the expressions in question may be retained and the sense of the words got from the surface. He will be sufficiently introductory so, and conspicuous for being quickly intelligible ... (891 D). For the text is to be searched mannerly, not bitterly racked, as if it were a prisoner, until it gives up what it has not taken.” This preliminary closes with a reminiscence (900 C). “Bernard of Chartres used to say that we, like dwarfs on the shoulders of giants, can see more and farther not because we are keener and taller, but because of the greatness by which we are carried and exalted.”

v-x. Forecasting the rest of the program, John wonders why Aristotle’s Topica should have been so long neglected (903 A). “Single words of it, in both rules and examples, are valuable not only for dialectica, but for almost all the disciplines. It comprises eight books, each more potent than the last.” The following digest, book by book, iterates (910 C) the general value. “The precepts of all eloquence seem to be derived originally from it as from the primary source. For it is indubitably true, as Cicero and Quintilian say, that rhetors and rhetoricians have found in it not only an aid, but a source.”

Book IV. Analytica

In contrast to the ten long chapters of III, IV is divided into forty-two short ones: i-v, analytica in general; vi-viii, demonstrativa; ix-xx, the progress of knowledge: sensus, imaginatio, prudentia, ratio, intellectus; xxi-xxiii, hypothetica, sophistica; xxiv-xlii, critical review: objectors to Aristotle the place of logica in teaching, typical conceptions of ratio, and of truth and error, the relation of ratio to veritas.

This last book iterates the importance of correlation (xxviii. 982 B). “But though logica is useful generally, he who is ignorant of other arts[37] is not so much helped by it toward philosophy as he is hindered by a habit of verbosity and presumption. For logica is almost useless if it be alone. It stands out when it shines by the power of correlated studies.”

John’s slighting of rhetoric cannot be explained merely by his preoccupation with logic. Why was he thus preoccupied in a consistent and progressive scheme of the whole Trivium? He begins with a logica embracing all studies of words; he devotes a whole book to grammatica; in his last pages he is speaking of an organon that shall be a minister to eloquentia. Yet rhetorica he merely mentions when he must. That he was aware of its ancient importance in such a scheme as his is evident from his large use of Quintilian’s Teaching of Rhetoric. No other medieval writer gave this work more attention. The much-quoted chapter (I. xxiv) on prælectio uses not only Quintilian’s ideas, but his very words;[38] and other correspondences are no less significant. The following list, though not complete, is typical.

The correspondences above, verbatim for considerable stretches, involve here and there transpositions or other variations. The following are quotations or adaptations.

Even if these correspondences were confined to Quintilian’s first and second books, there would still be no sufficient ground for the inference that John, consulting him primarily for grammatica and further for his general ideas on education, did not think of him as a rhetorician. For Quintilian not only presents rhetoric from the beginning; he frequently in these first books cites and quotes Aristotle as a rhetorician. But the matter seems to be put beyond doubt by the use of Book X. It is only fair to assume of so careful a scholar reading the first books and one of the last, and occupied with Quintilian’s idea of educational sequence, that he read the whole work.[39]

Having read and admired one of the chief ancient works on rhetoric, why did he leave rhetoric out of his own scheme? The answer is probably in the contemporary conditions to which the Metalogicus is adjusted, and especially in the contemporary state of rhetoric. It seems not to have been in the twelfth century worth more than mention from a John of Salisbury seeking a vital sequence of studies. It lacked what he sought above all, vital relations. If he had known Aristotle’s Rhetoric, he might conceivably have sought to recall the ancient study to its better ancient aim. What he found vital in Quintilian’s rhetoric he transferred to grammatica or to dialectica, partly, no doubt, because the transfer was actually going on, partly, one may think, because he saw in these other studies the real opportunities of his time for composition. The current lore of ornament which passed for rhetoric could hardly detain his consideration.

What he discerns in grammatica, and had found in the teaching of Bernard, is of course training in precision. But though he makes much of this, he does not slight the value of concreteness for presentation; he sees the importance of studying style by imitation; and he adopts from Quintilian a word not common in medieval treatises, œconomia. All these point to composition; and composition seems to have been one of the essential applications of the master’s analysis. The prælectio as John describes it is tinged with Quintilian because it realizes fully the ancient function of grammatica with the poets.

In the rapidly expanding teaching of dialectica also he sees opportunity for composition. To this end, very likely, he urges repeatedly that the debates of the schools keep touch with reality by insisting on subject matter through correlation with other studies. Above all he desires that language studies be unified and progressive, that they call for expanding correlation of inventio and iudicium. The study that in his time actually demanded and exercised these was not rhetorica, which was by way of ignoring both; it was dialectica.

The allegorical survey entitled by Alain[40] Anticlaudianus has but superficial likeness to the De nuptiis philologiæ et Mercurii[41] of Martianus Capella. Each presents the seven liberal arts in allegory; each is in nine books; but there is no real resemblance, nor any indication of Alain’s having used Martianus except as a stock source. The allegory of Martianus is confined to his first two books, and is purely decorative; the allegory of Alain pervades his whole work as a controlling idea. That idea is the function of education in the redemption of mankind. The sub-title “de officio viri boni et perfecti” is akin to Hugh’s[42] “animæ perfectio.” The consistent, elaborate theological symbolism shows some force of conception and, in spite of occasional excursions into style, some ardor and elevation. The four thousand hexameters are usually above the medieval fatal fluency in this verse. The survey, though it has not Hugh’s originality and does not attempt John’s unification, is equally serious.

Alain incidentally defines the character of each of the seven arts, and summarizes its scope. Each has its function—except rhetorica. The other members of the Trivium[43] provide the car of Prudentia with essential pole and axle; the Quadrivium supplies essential wheels; but all that Rhetorica has to offer is quite unessential adornment. Though Alain rehearses the traditional parts of her ancient lore,[44] he sums up her actual occupation in two lines of cardinal significance.

In other words, rhetoric is not operative as composition, but only as style after the fact. Her gifts to the soul are only colores, decor, clausula.

Alain’s own use of rhetoric is consistent with this point of view. Each of his allegorical figures is introduced with the conventional descriptive ecphrasis; and his diffuseness arises from the idea that poetica involves decorative dilation by those colores of which Cicero and Vergil are equally patterns.

The vast Speculum of Vincent of Beauvais[45] is a compend of all knowledge. Its second part, Speculum doctrinale, Mirror of Teaching, devotes two books to the Trivium. Grammatica,[46] including metric, follows Isidore. The following book[47] devotes ninety-eight chapters to logic (logica), ten to rhetoric, twenty-three to poetic. The proportion is significant; and poetica, taken from under the ægis of grammatica, appears as a separate, coördinate section. For rhetoric Vincent, still following Isidore,[48] repeats the classical definition and division. The following chapters (101-108) deal briefly with the elements of a speech, the ideals of oratory, the types of cases, status, syllogisms, loci rhetorici, style.

Vincent then goes on to poetica,

Except meter, which belongs under grammatica, and the two forms of drama, which are not thought of as forms of composition, there is nothing to distinguish this poetic from rhetoric.

St. Bonaventure[52] distinguishes four “lights”: (1) the exterior light of a manual art, illuminating with regard to artistic form; (2) the inferior light of knowledge through the senses, illuminating with regard to the patterns of nature; (3) the interior light of philosophical knowledge, illuminating with regard to truth comprehended intellectually; (4) the superior light of grace and revelation, illuminating with regard to truth as a means of salvation.[53] For (3), the “interior light,” his division is stated first in its commonest form. Philosophia is: (a) rationalis (i.e., sermonum), (b) naturalis (i.e., rerum), (c) moralis (i.e., morum). This division[54] he then repeats in another form and order:

• (a) physica, for knowing the causes of being;
• (b) logica, for knowing the theory of understanding;
• (c) practica, for knowing the conduct of living.

Finally he considers it in a third aspect. The domain of moralis is motive; of physica is self-limited and self-sufficient; of sermocinalis is interpretation. The last item he subdivides into:

Bonaventure does not use the word dialectica. His word for this, not for the whole Trivium,[56] is logica. Though his movendum and motivam suggest association of rhetoric with morals, and remind one of Aristotle’s conception, and Cicero’s, and Quintilian’s, he is content to give rhetoric the narrow and barren field of ornatus, assigning docendum and verum to logic.

The Old French Trésor of Brunetto Latini[57] devotes a far larger proportion of space than Vincent’s Speculum to rhetoric. Book I, surveying the seven arts,[58] includes history, geography, and the zoölogy of the bestiaries. Book II adds to Aristotelian ethics a collection of moral apothegms.[59] Book III opens politics with rhetoric.

The following chapters present the ancient division into five parts and the lore of status and quæstio.[61] But at this point the distinctive ancient function of rhetoric begins to fade. After saying that the main division of all expression is into prose and rime, he adds that “the teachings of rhetoric are common to both,” save for the restrictions of meter.[62] The section on dispositio, the ordering of a speech, takes from Martianus Capella not his sober survey, but a single passage which Brunetto perverts by subdivision and by transfer to narrative.

Here Martianus is apparently extending to the whole plan of a speech the commonplace of ancient rhetoric that the narratio (statement of facts) might be either continuous as a single, distinct part, or distributed for the sake of giving salience to its separate items. Brunetto’s further extension is not his own; it is common stock of the contemporary artes poeticæ.[64] These manuals make much of “natural” and “artificial” order, subdivide each, and apply both to narrative. Brunetto is typical.

The mechanical division reflects a wide abeyance in theory, not only of poetic shaping and movement, but even of the dispositio of ancient rhetoric; for its real concern is not with composition at all, whether in prose or in verse, but with the phrasing of certain patterns.

Brunetto’s preoccupation with style, and with style mainly as decorative dilation, soon appears in “how to dilate one’s tale in eight ways ... which are called colors of rhetoric.”[65] Among these colores, under demonstrance (demonstratio), is his celebrated, but entirely conventional, ecphrasis on Yseult.

The traditional parts of a speech (exordium, etc.) are used[66] as an approach to dictamen, of which Brunetto appears to be thinking through much of what follows. The concluding chapters on statecraft are far from a discussion of politics. Rather they set forth the conduct of a seigneur. Thus they fail to carry out the relations of rhetoric to government which in his opening he borrows from Cicero. Book III, then, of the Trésor lapses further and further from the ancient conception of rhetoric with which it begins. The rhetoric which it actually presents, and which it applies to poetry as well as to dictamen,[67] is a meager, though pleasant, review of style.[68]

What rhetoric appears in these surveys to lack most is distinct function. Writers as different as John of Salisbury and Brunetto Latini seem to think of it as polishing, decorating, especially dilating, what has been already expressed. It comes in after the real job is done; it has lost its ancient function of composing. The ancient lore of inventio kept rhetoric in contact with subject matter and with actual presentation. This had so much less scope in feudal society that the lore easily lapsed, or was perverted. The only large field for its exercise was preaching. Education, therefore, naturally threw its weight on grammatica for boys, on dialectica for men. Between the two rhetoric was crowded into narrow room. Whether it would still have vindicated itself if it had been the rhetoric of Aristotle, or oftener the rhetoric of Quintilian, can be only conjectured. Actually it was the rhetoric of De inventione and ad Herennium, and inculcated the sophistic of Sidonius. That may explain why there was no medieval rhetorician who really advanced the study.

[1] Fulbert, c. 960-1028. His distinguished successor Ives, who ruled Chartres 1090-1115, had been, with Anselm, a pupil of Lanfranc at Bec, and had himself taught there (Clerval, 146).

That Chartres had the whole Organon even before Gilbert de la Porrée, Richard l’Evêque, and John of Salisbury is shown by Clerval, 244 seq.

Haskins notes in the twelfth century the primacy of the cathedral schools (Laon, Tours, Chartres, Orléans, Paris) over the monastic.

(The Normans in European History, 177; The Renaissance of the Twelfth Century, 49.)

[2] The traditional manuals and authors continue: Donatus, Priscian, Martianus Capella (not much used), Bede, De arte metrica; Livy, Valerius Maximus, Orosius, Gregory of Tours; Vergil, Ovid, Horace, Terence, Statius; Fortunatus, Sedulius, Arator, Prudentius, Boethius. Prælectio appears at its best in John of Salisbury’s account of the teaching of Bernard of Chartres (below pages 160-164). See the admirable summaries, period by period, in Clerval.

The library at Bec in the twelfth century had Priscian, Isidore, Martianus Capella (with Remi’s commentary), Anselm De grammatica, and Claudian (PL 150: 775-782).

[3] Clerval, 111-113, suggests an influence on the earlier forms of liturgical sequences.

[4] Colson (ed. of Book I, Cambridge, 1924) finds no clear indication of Quintilian in the eleventh century; and the large use of him by John of Salisbury in the twelfth (below, page 169) is not specifically related to rhetoric. Cicero is limited to De inventione and the commentary of Victorinus (Clerval, 115). The Rhetorica ad Herennium is the source of the Colores of Onulf, who taught at Speier c. 1050. The Bec library had in the twelfth century “utraque rhetorica” (i.e., De inv. and Ad Herenn.), Cicero, De partit. orat., Suetonius, and the letters of Sidonius. At Chartres Thierry cites Quintilian and, according to Clerval, 233, Cicero, De oratore. The latter, with Quintilian, appears in A list of textbooks from the close of the twelfth century, Haskins in Harvard Stud. Class. Phil. 20:92.

[5] Clerval, 115-116.

[6] *100 as described in Clerval, 117. Besides Fulbert’s verses, it has (*10) two short treatises: De rhetoricæ cognatione, and Locorum rhetoricorum distinctio. The Topica, of course, are common ground between rhetoric and logic.

[7] Gilbertus Porretanus, c. 1075-1154, Chancellor at Chartres 1126, teacher at Paris 1141, Bishop of Poitiers 1142; Liber sex principiorum, Comment. IV libr. Boeth., Liber de causis. See Clerval, 163-168, and Poole. Clerval suggests that Gilbert, as well as Thierry and Bernard Silvester, had relations with the Toulouse translators of Arabic books into Latin.

[8] Terricus Carnotensis, Breton, scolarum magister at Chartres 1121, taught rhetoric at Paris c. 1140, died c. 1150; Comment. De invent. and Rhet. ad Herenn.; Heptateuchon c. 1141 (Chartres MSS. *497, 498) described in Clerval, 220 seq., with a synoptic table of contents.

[9] Donatus and Priscian.

[10] Cic. De invent. and Partit. Orat., Rhet. ad Herenn., the summary of Severianus, and Martianus Capella.

[11] The usual Boethian items.

[12] Translated in Clerval, 221.

[13] Hugh of St. Victor, Saxon, c. 1096-1141, entered the Abbey of St. Victor at Paris c. 1116 and taught there; works PL 175-177. See especially Mignon, whose second chapter is a good summary of studies before Hugh. A critical study of the MSS. by Hauréau (B.), Les œuvres de H. de St. V., Paris, 1886 (a revision of a study published in 1859), shows that Eruditio didascalica has only six parts, the seventh added in PL being a separate work. See also Fourier Bonnard, Histoire de l’abbaye royale et de l’ordre des chanoines réguliers de St. Victor de Paris, Paris (n. d.); and the index to Manacorda. John of Salisbury refers to Hugh, and uses part of his classification.

[14] The division, which is common in the middle age (e.g., John of Salisbury, Metalogicus, II. ii), is in Quintilian, Inst. Or. XII. 2. 10.

[15] About 1159. For John see Poole, the index to Clerval, and the antiquated, but still suggestive study of Schaarschmidt, J. S. nach Leben und Studien ..., Leipzig, 1862. Webb (C. C. I.) has edited the Policraticus, Oxford, 1909, with an introduction especially valuable for John’s reading, and announced an edition of the Metalogicus, for which meantime the only available text is in PL 199, referred to in this section by column. The letters, also in PL 199, were printed with Gerbert’s and Stephen of Tournai’s in 1611 (Paris, Ruette). The Historia pontificalis, printed in MGH as anonymous, has been edited by Poole, Oxford University Press, 1927. Policraticus IV, V, VI, with selections from VII and VIII, are translated, with an introduction, by Dickinson (J.) as The statesman’s yearbook of J. of S., New York, 1927.

[16] See the preceding section. John mentions Hugh at 833 A, 924 B. The word disserere Hugh and John may have taken, as John took other things at the opening of Book II, from Isidore (see above, page 97); but the ultimate source of the phrase ratio disserendi is probably Cicero’s Topica: “omnis ratio diligens disserendi duas habet partes, unum inveniendi alteram iudicandi.” Top. 2. Cf. Fin. 1. 7. 22; Fat. 1. 1. Disserere seems limited to dialectica in De orat. I. 9 (see Wilkins’s note); but using the same word in II. 157, Cicero points out that the proper function of dialectica is analysis (iudicium, not inventio). In Orator 113, discussing the common ground of rhetoric and logic, he says “utrumque in disserendo est,” and goes on to distinguish logic as ratio disputandi from rhetoric as ratio dicendi. Dicere is generally his word for rhetoric. Quintilian, whom John uses largely, probably has in mind the same distinction in X. 1. 81, though in other places he uses disserere more generally.

[17] I. xvii (847 C), xxiv (854 B), II. x (868 B).

[18] For John’s use of Quintilian, see below, page 169.

[19] In this respect Metalogicus is conspicuously different from Policraticus.

[20] These “Cornificians” are mentioned 825 C, 827 A, 852 B, 857 A. See Clerval 182, 211, 227.

[21] Quadam proportione rationis, 841 B.

[22] This poetic is none the less rhetoric for being confirmed by quotations from Horace (Ars poetica, 102-105, 108-111; see ARP 245). The general doctrine of appropriateness was the basis of the specific recipes for encomium (above, page 31).

[23] Cf. Vincent of Beauvais below, section D. 2.

[24] The following translation renders entire the most specific extant account of prælectio in the middle age. Long selections from it in French translation will be found in Clerval, 225-227; a shorter selection, with the corresponding Latin text, in Faral, Les arts poétiques du xiie et du xiiie siècle, Paris, 1924, 99-101. Haskins, 135, has a short selection in English translation.

For prælectio, see the index to ARP and to this volume.

[25] Illustratio is Quintilian’s word (and Cicero’s, he says) in VI. ii. 32, where he is exhibiting the same sort of development as is here indicated by John. For historia, argumentum, fabula see the index. That these were elementary school exercises suggests that the subject of this sentence (Illi) refers not to authors, but to schoolboys engaged in developing a materia; but the reference can hardly be determined in PL, in which the whole sentence seems to me dubious.

[26] Probandi colores. The word colores in this sense, unusual at this time, is characteristic of Seneca Rhetor, to whom John refers in II. viii. But the solution of a passage apparently corrupt may well await a better text.

[27] Cf. III. vi (904 C). Et sicut juxta ethicum: discipulus prioris est posterior dies.

[28] Declinatio is the eighth item, under Priscian, in the list from Thierry’s Heptateuchon cited above, page 153.

[29] For clausula see the indexes to ARP and to this volume; for John’s own cadences, below, Chapter VIII.

[30] Œconomia.

[31] See below, page 170.

[32] For inventio and iudicium see Hugh above, page 154, and note 16. Quintilian III. iii. 5 objects to the application of this to rhetoric.

[33] See Hugh’s division above, page 155.

[34] Vocum et intellectuum.

[35] Ficta, hypothetical? Sophistica seems to be intended in the following clause.

[36] Thesis.

[37] Reading aliarum.

[38] This is pointed out by Colson in his edition of Quintilian’s first book, Cambridge, 1924, page 1, note 2. It is the stranger that the quotations should not have been noticed before since John himself calls attention to them: “ut verbo utamur Quintiliani,” “ab auctoritate ejusdem Quintiliani” (I. xxiv. 853 D).

[39] Illustratio in I. xxiv (854 A/B) is probably a reminiscence of Quintilian VI. ii. 32, to judge not only from the word, but from the context in both passages. Cf. 844 A, 851 C, 860 C, 910 D.

[40] Alanus de Insulis (about 1128-1202), Cistercian at Citeaux, sometimes called “doctor universalis,” wrote mainly on theology. Anticlaudianus was edited by Wright in the second volume of his Anglo-Latin Satirical Poets and Epigrammatists of the Twelfth Century, London, (Rolls Series), which contains also De planctu naturæ. The latter has been translated by D. M. Moffat in Yale Studies in English, New York, 1908.

The citations in this section are by column from PL 210. For the Summa de arte predicatoria see below, Chapter IX. C.

[41] See above, Chapter III. B. 1.

[42] For Hugh, see above, page 153.

[43] Alain’s order is the usual (1) grammatica, (2) dialectica, (3) rhetorica. His logica is dialectica, not the inclusive term of Hugh and John. Its auctores (III. ii. 510 A) are Porphyry, Aristotle, Zeno, Boethius.

[44] III. ii. 512-513.

[45] About 1190-1264. The Speculum was printed at Strasburg in 1473, and by the Benedictines at Douai in 1624.

[46] Book III (II in the Benedictine edition).

[47] Book IV (III in the Benedictine edition).

[48] Vincent refers in Chapter 100 to Quintilian. “Tullius in libro de oratore” (100 and 127) is probably not a reference to De oratore. In 101 “Tullius in rhetorica prima,” and in 127 “idem in rhetorica secunda,” are the usual references to De inventione and ad Herennium. For the references to Quintilian see Bassi, Giornale storico, XXIII, 186, and Colson’s introduction to his edition of Quintilian’s first book, page lii.

[49] Farabi. For translations from Arabic see above, page 153, and Haskins, Chapter IX.

[50] Compare Alain above in the quotation on page 174.

[51] For the three school exercises that conclude this list see the indexes to ARP and to this volume; for the rest of the classification, Bede, above, Chapter V. B. 1.

[52] 1221-1274. The Opusculum de reductione artium ad theologiam is number 4 (pages 317-325) in volume 5 of the Works edited by the College of St. Bonaventure, Quaracchi (Florence), 1891.

[53]

• (1) Lumen exterius artis mechanicæ illuminat respectu figuræ artificialis;
• (2) lumen inferius cognitionis sensitivæ illuminat respectu formæ naturalis;
• (3) lumen interius cognitionis philosophicæ illuminat respectu veritatis intellectualis;
• (4) lumen superius gratiæ et sacræ Scripturæ illuminat respectu veritatis salutaris.

[54] For the classifications of Hugh of St. Victor and John of Salisbury see above, sections B and C.

[55]

• (a) moralis regit motivam;
• (b) naturalis regit se ipsam;
• (c) sermocinalis regit interpretativam:
• (1) grammatica ad exprimendum respicit rationem ut apprehensivam ... congruum;
• (2) logica ad docendum respicit rationem ut iudicativam ... verum;
• (3) rhetorica ad movendum respicit rationem ut motivam ... ornatum.

[56] That he does not intend logica in the larger sense of Hugh and John will be clear from comparative study of the three forms of his division, and also from Collationes in hexaëmeron (in the same volume), IV. 18-25. This briefly sums up rhetorica according to its ancient topics, including the three fields and the five parts. Here again dialectica is not used, and logica merely supersedes it.

[57] 1230-1294. The Trésor is edited, with an introduction, by Chabaille in the Collection des documents inédits sur l’histoire de France, Paris, 1863. References in this section are to the pages of this edition. The Italian version, also widely current, is edited by L. Gaiter, Bologna, 1878. See also F. Maggini, La rettorica italiana di B. L. (Pub. del R. Istituto di studi superiori), Florence, 1913.

[58] The proem, beginning with theology as the highest lore of theorica, makes the usual enumeration under mathematica, divides practica, “la seconde science de philosophie,” into “éthique . économique . politique,” and so arrives at rhetoric, but with some confusion as to logic.

[59] See Chabaille, xv.

[60] Villani may be merely repeating this when he says: “Egli fu cominciatore e maestro in digrossare i Fiorentini, e farli scorti in bene parlare et in sapere guidare e reggere la nostra repubblica secondo la politica.” VIII. x.

[61] Brunetto’s main source is Cicero De invent.

[62] La grans partisons de touz parleors ... en prose ... en rime; mais li enseignement de rectorique sont commun andui, sauf ce que la voie de prose est large et pleniere, si comme est ore la commune parleure des gens; mais li sentiers de rime est plus estroiz et plus fors. 481.

[63] His igitur ad fidem faciendam prudenter inuentis ordo rerum est sociandus, quæ pars dispositio uocitatur ... duplex igitur huius partis est ratio; aut enim naturalis est ordo aut oratoris artificio comparatur: naturalis, cum post principium narratio, partitio, propositio, argumentatio, conclusio epilogusque consequitur; artificio oratoris, cum per membra orationis quæ dicenda sunt digerimus. V. 506 (Teubner, ed. Dick, 248).

[64] See below, Chapter VII. B.

[65] “Comment l’om puet acroistre son conte en viij manieres ... qui sont apelees color de rectorique.” The first of these, aornement, is defined so bluntly that the humorous implication may be intentional: “tout ce que l’om porroit en iij moz ou en iiij, ou a mult po de paroles dire, il les acroist par autres paroles plus longues et plus avenans qui dient ce meisme.” 486.

[66] 490.

[67] For dictamen, see below, Chapter VIII.

[68] Jean d’Antioche in his introduction to the translation of De inventione and Rhetorica ad Herennium (end of XIIIth century) has two divisions of Philosophye: (A) into ethique morale, rationele, and naturele; (B) as follows:—

Delisle in Notices et Extraits, 36:216.

Neither Le mariage des sept arts nor Henri d’Andeli’s Bataille des sept arts contains any information about teaching.

CHAPTER VII LATIN POETIC IN THE TWELFTH AND THIRTEENTH CENTURIES — References and Abbreviations — A. Poetica included in Grammatica — B. Poetria — 1. Matthieu de Vendôme, Ars Versificatoria (before 1175) — 2. Geoffroi de Vinsauf, Poetria Nova (c. 1210) — 3. Évrard, Laborintus (c. 1213) — 4. Johannes de Garlandia, Poetria (probably before mid-Thirteenth Century). — 5. Common Traits — C. Hymns — 1. Progress of Rimed Stress Verse — 2. Variations in Trochaic Stanza — 3. Symbolism — FOOTNOTES:

The grammatica of this period continued the traditional inclusion of metric and of certain figures of speech; and the master’s prælectio on the Latin poets involved at Chartres, as two centuries before at Speier,[1] imitative writing of Latin verse. About 1200 appeared two hexameter summaries: the Doctrinale[2] of the Norman Alexandre de Villedieu, and the Græcismus[3] of the Flemish Évrard de Béthune. The former had so long and wide a vogue that it may be called the standard medieval mnemonic of grammatica. Reviewing successively inflections, syntax, metric,[4] accents, figures, it includes under the last the lists of Greek terms that show at once a preoccupation of the time and the shifting boundary between grammatica and rhetorica.[5]

Distinctive of this period is the separate ars poetica, or poetria. Often itself in verse, this sort of manual differed from Bede’s, first in being less a reference book for the study of meters than an exercise book for the actual writing of Latin verse, and secondly in giving less space to prosody than to poetic diction. The four most conspicuous[6] may be assigned approximately to the half-century divided by 1200 (c. 1175-1225).

Matthew’s prose manual, though it omits prosody, is otherwise connected even more obviously than the others with the teaching of grammatica. Not only is he known to have been grammaticus at Orléans; his book is inclined throughout in the direction of such teaching,[7] and it contains specimen school exercises. The grammatical slant is most obvious in those on adjectives in -alis, -osus, -atus, -ivus, -aris.[8] The longer examples of descriptive verse may well be such successively revised themes as were seen earlier at Speier.[9] The use of Horace’s “Ars poetica” is so extensive, even for the time, as to suggest that Matthew’s book may have begun in his prælectiones on that poem. Whatever degree of probability may be attached to these suggestions, there is no doubt of Matthew’s intention and preoccupation. His book seeks to further the writing of Latin descriptive verse. The idea behind it is that poetry is mainly description, which in turn proceeds mainly by dilation. Style, which is his only concern, is conceived as decoration. Though his lists of figures for this purpose (III) generally agree with those of the Doctrinale,[10] rhetoric is evident not only in the phrase colores rhetorici,[11] but as a constant preoccupation. That poetica as style is identical with rhetorica he assumes; that it is distinct as composition can hardly have entered his head, but composition in either field is beyond his scope. His sections on beginning (I. 3-16) refer not to introducing the subject, but to phrasing the first sentences. The faults then enumerated (I. 30-37) are of style. Description is expounded (I. 38-113) as appropriateness of phrase to condition, age, place, etc., and as the seeking of “attributes” in a person’s physical and mental habit, his deeds, his speech,[12] or in the cause, quality, and time of an event. Reference to subject, thought, or composition goes no further; the rest of the book is purely verbal.

The extraordinary vogue of Geoffrey’s two thousand hexameters in itself suggests that his too is an exercise book. To suppose that it was cherished for its literary achievement is to impute to several centuries a larger and more general appetite for bombast than other evidence warrants. As school mnemonics his verses are more tolerable. As exercises in synonyms they have excuse for their redundancy. As suggestions and examples in the pursuit of figures they have more warrant for exaggeration as a means of distinctness. Caricature, which can never have been the intention of a man devoid of humor, has before Geoffrey’s time, and since, resulted from sheer overemphasis. His own incessant word-play is so anxious as to verge sometimes on reductio ad absurdum. Or were some of these tirades made to order by his students? May they be such progressively revised composite themes as the earlier schools assigned in similar hexameter tasks?[13] At any rate, a probable explanation of the portentous style, as well as a charitable one, is that the Poetria nova was a museum for boys.

After bowing to the ancient inventio and dispositio, neither of which is in point and neither handled as a process of composing, Geoffrey devotes most of his book to the rhetorical means of dilation. This is the aim of the colores, not only of verbal expansions in general, but in particular of two sorts of deliberate interpolation: apostrophe and description.

Chaucer,[15] whose ironical homage has made Geoffrey a laughing-stock, was exploding the use of these figures in verse narrative. In oratory also of a certain sort he exhibits their deviation in the specimen preaching of the Pardoner because they are perennial in sophistic. Geoffrey’s description is precisely the ecphrasis cultivated throughout the Empire in declamatio.[16] In a word, Geoffrey’s poetic, as Alain’s, Vincent’s, Brunetto Latini’s,[17] is mainly the rhetoric of dilation. The sophistic of the ancient encomium, walking the schools once more, is now called Poetria.

The Laborintus[18] is at once briefer and more inclusive. Opening with rueful mock-heroic on the lot of a schoolmaster, it glances through the seven arts in a series of allusions, and expands upon grammatica. Thus poetry is reached by the traditional approach.[19] Most space, however (269-598), is given to exemplifying rhetorical ornament, especially figures. Though the list, as usual, is long, the manufactured or borrowed examples are short. Évrard’s plan of tucking away each within a closed hexameter-pentameter couplet is carried out with some ingenuity. Though he has to take more room for demonstratio (573-594), he generally abjures Geoffrey’s dilating upon dilation. His examples are rather mnemonics than exhibitions. The fourth section (599-686) is a list of authors for school reading.[20] The order, though not obviously progressive, suggests: (1) certain brief moralizing works for elementary study, i.e., a “Cathonet” (the so-called “Distichs of Cato”), a “Théodolet,” or “Theudlet” (the allegorical verse dialogue entitled Theodulus),[21] an “Ysopet,” or collection of fables;[22] (2) the classics and their imitators, with Vergil applied as the exemplar of all “three styles”;[23] (3) the Latin summaries of the Trojan war, Dares and the Ilias latina, (4) Sidonius and the earlier Christian poets, Alain’s Anticlaudianus, and Matthew’s Tobias; (5) a group of works on style, i.e., Geoffrey’s Poetria; the Doctrinale and Græcismus, Matthew’s Ars versificatoria, Martianus Capella, and Bernard Silvester.[24]

The section on metric (687-834) exemplifies the leonine verses repudiated by Matthew,[25] the handling of phrases and clauses—even to such ingenuities as reversible lines, several patterns of internal rime in the hexameter,[26] and typical faults. After lamenting a schoolmaster’s hardships (835-990), Évrard concludes (991-end) with classified specimens of rhythmi. There is a noticeable preponderance of trochaic measures.[27] The final group of quatrains in a measure common both in hymns and in Goliardic use is clearly a school exercise in framing a stanza to end upon a familiar quotation.

John’s work differs from the other three in specific application to dictamen.[28] The inclusion of poetria and dictamen in one treatise, though ill managed, is a practical adjustment to the teaching of the time. As taught, both were rhetorica, and both were confined within the single department anciently called elocutio. John begins, indeed, as Geoffrey does, with inventio; but his treatment of it[29] shows how faint in his time were even the echoes of its ancient function. Invoking for it simultaneously the “Ars poetica” and the Rhetorica ad Herennium, he first misapplies it to adaptation of style to person, occasion, etc., as in a letter—and as in the “three styles” of which Vergil is again made the exemplar! Then he perverts it to the search for appropriate proverbs, of which he provides a classified list for use in dictamen. His further applications show that inventio in his practise is purely verbal and leads, as fatally as all other approaches, to the lists of figures. “Nor should it be forgotten,” he adds (897),[30] “that any theme (materia) can be expressed in six ways according to the six cases of the noun.” As if uneasy at the ancient application of the term, he appends a final section De arte inveniendi materiam, as a separate device for “boys wishing to amplify and vary a theme.” For example, in writing about a book they might find occasion for praise or blame in the efficient cause, i.e., in the writer; in the material cause, i.e., in parchment, ink, etc. This can hardly be the mere dotage of John the Englishman. It is a sharp reminder of the educational level of these manuals; and it shows that the old inventio had departed from rhetorica.[31]

Otherwise he could hardly go on (Chapter II): “After inventio ... follows electio. Tully after inventio puts dispositio, then the art of memorizing, and finally delivery; but, for writers of poetry or dictamen, after inventio the useful art is that of choosing”! The choosing that he means is of the right style, “brief for affairs of the Curia (i.e., dictamen), diffuse for poetry” (897). Again he exemplifies by a specimen letter. Memoria is considered merely as mnemonic; and the cardinal mnemonic is a diagram of the “three styles” (900), each with its proper furniture of persons and things occupying a segment of a circle, the Rota Virgili.[32] As in the other manuals, the “art of beginning” (905), though divided into several modes, has little to do with composition. A letter (907) may begin with a proverb, an example, a comparison, with si, or cum, or dum, or an ablative absolute. The “six parts of a discourse” (911) are summarily defined as by the ancients, but thereupon exemplified in seventy-eight elegiacs. As in the other manuals also, the art of concision (913) is mentioned; but the art of dilation (914) by figures is dilated.

Such clumsy handling makes obvious the misapplications of rhetoric to poetic that are current among John’s contemporaries. Another perversion equally general is that of the narratio[33] of a speech to narrative. After exhibiting quite properly the statement of facts in a letter as an application of narratio, John deviates as follows:

Brunetto Latini shows not only the same deviation,[35] but also that “natural order and artificial order” which John (905) and the other pedagogues had perverted from Martianus Capella’s narratio to narrative, and, behind both, the general misconception of the ancient dispositio.[36] Terms traditional in ancient rhetoric for the processes of composition are deviated at once to poetic and to style because the consideration never extends beyond figures, feet, or clauses.

Thus John is able to add (928) to the “three poetic styles ... four other styles in modern use: (1) the Gregorian, (2) the Tullian, (3) the Hilarian, (4) the Isidorian. By the first he means the Roman style in dictamen.”[37] The second he distinguishes not by rhythm, but by colores, and as “used both by poets when they write prose and by teachers in school exercises.” The third, defined metrically, is exemplified both by an ancient hymn[38] and (929) by a letter. The fourth, that of Augustine’s Soliloquies, marks the balance of clauses not by equality of length, but by chiming cadences. To the usual list of figures are added the ten commonplaces (939) for the description of a person;[39] to a lust-and-blood plot (tragedia), further specimens of dictamen. Rhetoric and poetic are merged in one scheme of style. The scheme is confused; but the intention is single.

The final section, ars ritmica, is significant, as Évrard’s is, by its very presence in a schoolbook. It shows (56)[40] the same school assignment of a stanza framed to end on a familiar quotation. Its classification of examples, though not illuminating, is much clearer than the subdivisions of the chapters preceding. Rhythmus, the verse of the hymns, distinguished from metrum though discussed in some of the ancient metrical terms, is presented for study and practise. For the schoolboys who used it, as for the modern explorer, this section must have offered the relief of an active poetic after the drill of deviated rhetoric.[41]

The pedantic subdivision of these manuals shows that their aim was not to organize the study of poetic, but to cover its elements by as many exercises as possible. Imitative writing of Latin verse, long part of the study of grammatica, has been combined with the theory of rhetorica through exercises in figures, and with its practise through exercises in dictamen. Doubtless the resulting aggregation was called poetria both because the exercises were still connected with the traditional prælectio and were oftenest in verse, and because, whether in verse or in dictamen, they were focused on that heightening by ornament and by dilation which was conventionally regarded as poetic. Poetria, then, meant generally the study of style, and specifically the study of stylistic decoration. The lore for this was rhetoric,[42] partly indeed by misapplication, partly from the vagueness of the boundary in Latin tradition. The “colors of rhetoric,” not always clearly distinguished, sometimes strangely spelled, were faithfully recited as a sort of Greek ritual of poetic. The confusion went to its bitter end in that stock perversion by which Vergil’s poetic was broken on the wheel into three pieces of rhetoric.

The vital difference between rhetoric and poetic in composition, probably beyond the ken of these writers, was certainly beyond the intention of these manuals. Composition for them goes no further than the adjustment of a sentence. The ancient inventio and dispositio, sometimes dragged in by misapplication, are generally ignored. The distinction of “natural” from “artificial” order provides a pattern, not to promote composition, but to obviate its necessity. So the “methods of beginning” are presented as verbal devices. The scope of these manuals suggests that rhetoric, whether in its own name or as poetria, did not teach composition. What had once been part of rhetoric was now left to logic and the debates of the schools. As for poetic composition, the active progress of vernacular verse narrative would hardly be represented in a schoolbook. What is represented, what appears alike in school Latin and in professional vernacular,[43] is surviving conventional pattern, the passive voice of poetic, not its active. Marie had found another poetic. Chrétien, though he had accepted some of the same conventions of style, had learned otherwise what he knew of narrative movement. Even Latin narrative had been otherwise studied by Walter Map. To poetic in this larger composing activity the poetria of the schools offers no clue. At the turn of the next century Dante, who knew all its poetic conventions, ignored them in a supreme composition; and within that century Chaucer, who knew them too, laughed them away.

The ancient quantitative metric learned in school was often practised in such occasional verse as Baudry de Bourgeuil’s.[44] Both his elegiacs and his partially rimed verses are literary exercises. More significant is the tentative use of rime in the same eleventh century by Fulbert.

Heribert, Bishop of Eichstätt in 1021, rimed more confidently, but still without insistence.

The danger of insistence in rimed stress verse was already evident in many accentual hexameters riming the end of the verse with the middle. How easily the combination lapses into doggerel appears in a common seven-stress verse with marked cæsura (4 + 3).

Such mechanical versifying merely makes obvious that stress rhythm was the established habit.[46] By the twelfth century anything else was merely literary exercise, quite out of the literary current; and the danger of obtruding the stress pattern of failing to fuse it with the other means of suggestion, was more and more expertly avoided.

How far even a very marked pattern of stress and rime could be carried was demonstrated in the famous De contemptu mundi of Bernard of Cluny.[47] Though its arraignment of his age soon faded, its detail of doom and redemption, its realization of eternal life, showed poetic vitality. Bernard’s failure in organizing the movement of the whole is thus forgotten in the striking success of these parts.[48]

The heavenly Jerusalem has never been contrasted with the cities of the perverted present more eloquently. For Bernard’s achievement of style is rather in the ample realizations of eloquence than in poetic compression. The poetic distinction is in the verse. The sheer technical mastery of insistently rimed stress rhythm shows the possibilities of the verse habit of his time.

Thus the poem begins; and the cento made from it is often entitled[49] Hora novissima. Accentual hexameters rimed within—the Leonine system carried a step further—are rimed together as couplets. They are entirely dactylic except in the last foot; the typical ancient variation by spondee within the line is never used. Finally they are divided into three staves of two beats each, thus foregoing another ancient variation, the shift of cæsura. The first two staves of each line rime together; the last rimes with the last stave of the next line. The pattern could hardly be more marked. Yet the drumming dactyls, the insistent rime, are kept in movement; they do not stall even when the weaving of the line becomes sheer virtuosity.

Bernard demonstrated that even insistently rimed stress rhythm could be kept from jar and jingle by attending constantly to movement. For this he exhibited further the capacity of accentual dactyls. More generally he displayed with great technical skill the range and flexibility of rhythm always chiming with word-accent, that is always answering the habit of speech.

The currency of rimed stress verse is obvious in student songs and other jocular and satirical poems known generically as Goliardic.[50] A little satire on masters and bachelors of arts, assigned to the eleventh century, is typical of the ease with which such verse could be turned.

Of the numerous twelfth- and thirteenth-century poems of this sort, many of which have been repeated ever since, none is more famous than the drinking song once attributed to Walter Map.

Rime no longer incidental, but integral and composing, opened to medieval poetic wide artistic possibilities of stanza. The austere requiem sequence of Thomas of Celano, Dies iræ, is cast in three-line trochaic stanzas of a single rime.

That forty of the forty-five hymns assigned by Misset-Aubry to Adam of St. Victor are trochaic shows the strong preference of the time. The commonest of these trochaic stanzas, the favorite hymn measure of the twelfth and thirteenth centuries, expands Corde natus into a six-line stanza by doubling the first stave in a riming couplet twice and using the second stave between the two couplets and at the riming end.

Using this scheme oftener than any other, Adam rarely holds to it throughout a hymn. His way is rather to vary it in one or two stanzas; and in some of his hymns he even interpolates a stanza of a different measure. For example, the ninth stanza of this Christmas hymn departs entirely from the pattern except in the middle and end lines.

This is the only variation; all the other stanzas are regular. The Lauda Sion sequence of St. Thomas Aquinas varies the measure but slightly; and the Stabat Mater keeps it strictly—even reinforces it with additional rimes.

Adam’s harmonizing of rimed rhythmic, conspicuous above, has great range and variety. The range appears in the following contrast:

Within a hymn the variations are delicate adjustments. The hymn on the Cross strikes the familiar measure, swerves from it, returns to it, varies it.

Quite as widely Adam realized the poetic possibilities of symbolism. Imagination in the middle age was stirred habitually by types. As these spoke in sculpture and glass they spoke in the hymns. Medieval symbolism sought to induce mood, to stir emotion, not by individualizing concrete details, but by familiar typical associations: lamb, vine, star of the sea. Such symbols, long ago drawn from Messianic prophecy,[51] had become both numerous and familiar. They differ essentially from the figures of the poetriæ in being not decoration, not epithets or periphrases used instead of proper names, but immediate lyrical approaches. Light is used, not instead of the sacred name, or of some such title as Redeemer or Savior, but to focus attention on the Light of the World. So Cornerstone,[52] or Lamb, or Bread, suggests redemption immediately in one aspect. So the Redeemer is seen to be foreshadowed in Isaac,[53] Joseph, or David; for medieval art sees history as the progress of the redemption of mankind.

In this aspect the symbols of the Virgin Mother are lyric not merely in warmth of emotion, but in visions of human progress as divine. In turn Bush burning but unburnt, Flower, Fleece bedewed, Star immemorially guiding sailors, she embodies personally hope after hope. This habitual symbolism of stone and glass and hymn is less sentimental than intellectual. While it appeals to childhood memories, it opens vistas. The surcharging of the Corpus Christi hymns does not cloud their scholastic precision.

Keeping much of the tradition of its earliest centuries, hymnody has nevertheless widened its range; keeping communal devotion, it has risen in contemplation. This is the character of the poetry written for the new feast by the Angelic Doctor.[54] The enthusiasm of the popular processions is answered, but it is also brought to its goal. The greatest medieval hymns obliterate the crude distinction between “reason” and “feeling,” between “thought” and “emotion.” They remind us of that ancient saying about the sublime, that it springs from intellectual vigor of conception.[55] That is why, of all medieval poetry, they are the best approach to the Divina Commedia.

CHAPTER VIII DICTAMEN — References and Abbreviations — A. The Rhetoric of Dictamen — B. Digest of Candelabrum I-V [18] — Book I. Choice of Words, Rhythmical Composition of Sentences — Book II. Figures — Book III. Salutatio — Book IV. Exordium, Narratio, Petitio, Conclusio — Book V. Summary Review of the Preceding Books — C. Cursus — FOOTNOTES:

The art of letter-writing, especially the composition of official and other ceremonious letters, was of cardinal importance in the middle age. Even to-day, when letters of affairs are facilitated and multiplied by devices for dictation and despatch, it has a field smaller than the medieval in proportion to the other means of communication. Since these conditions lasted into modern times, letter-writing kept its importance throughout the Renaissance; but its art, developed in Latin antiquity and keeping for all affairs of moment the Latin language, became in the middle age a necessary ally alike to law and to diplomacy. Dictamen was a recognized profession and an habitual means of education. The model for the official correspondence of the western world was the Papal Chancery. The chief center of teaching was Bologna.[1] Though there was rivalry at Orléans, there was a recognized primacy of stilus Romanus.

No medieval form of writing has come down to us more abundantly in manuscripts; and none is more abundantly available in print. In addition to the collections of letters by famous men, or included with them, are many form-letters preserved as models.[2] Medieval epistolary habits are thus amply exemplified as the most widespread applications of the study of style. As a fine art medieval diplomatic correspondence is seen at its finest in the letters of John of Salisbury. The expertness that illuminated even routine affairs and determined the direction of great ones was recognized in his own day. The Archbishop of Rheims, writing to the Pope concerning the exile of the Archbishop of Canterbury, leaves the form of his letter to John.[3] At an earlier crisis of that exile, when Becket cried out against Henry’s influence at Rome, the indignation of the great prelate was submitted to the great secretary; and he was too faithful a friend to forget his art.

Having shown a list of grievances to be ineffective as composition, he adds immediately that accusations are dangerous as style. But not stopping with criticism, he took the great hazard upon himself in the following letter to the same Cardinal.

Medieval letters, long studied for history, should be better known as literature.

Less available, though for our purposes more directly important, are the manuals. Called generically ars dictaminis (or dictandi), or summa dictaminis, they are sometimes brief introductions to the formularies, sometimes substantial and detailed separate works. The latter have much less often than the formularies been printed. The ampler manuals apply to letter-writing a review of both grammar and rhetoric. This is instructive in Conrad’s Summa de arte prosandi.[5] The Boncompagnus,[6] so widely used that Boncompagno’s name, like Donatus, became a common noun, is amplified rather by diffuseness and by abundance of specimens. The Rhetorica novissima[7] of the same author is at once over-divided and ill-digested. Equally diffuse and equally relying on specimens, it appeals to students of law by deliberate and often inept variations from the traditional lore of rhetoric. The vogue of Thomas of Capua is hardly explained by the meager printed text of his Dictator.[8] None of these is at once so ample and so specific as a work by a Florentine of the Bolognese school, perhaps Bene of Florence, entitled Candelabrum.[9]

The international affairs of the Roman Curia demanded and developed professional notarii.[10] Their first and abiding concern was precision. Legal correctness of language, exactitude, systematic verification and record, precaution against tampering and forgery, all demanded an elaborate technical skill. This was ars notaria in the stricter sense, an important branch of the practise of law, especially of canon law.[11] Beyond legal correctness and dignity it developed style befitting Rome. Privileges,[12] decrees, mandates, dispensations, commissions, and other forms observed exact appropriateness and rhythms that were at once marks of authenticity and models. The same care extended to diplomatic correspondence. The documents in both fields, a mine for students of medieval history, amply attest the importance of dictamen as a profession.[13]

So wide a demand would of itself have maintained schools of professional technic. But since the technic demanded preliminary general training, it became a development of rhetorica. Not only so; it divided with preaching the whole field of daily prose composition. In current application oral composition was preaching, written composition was dictamen. The two were the typical medieval fields for the ancient lore of persuasion. The manuals of dictamen often begin by defining it in the general sense of writing.[14] Dividing all writing into (1) metricum, the ancient quantitative verse still taught as a branch of grammatica, (2) rithmicum, the accentual rimed verse of the hymns, and (3) prosaicum, they confine themselves to the third, and make it equivalent to letter-writing (prosaicum vel epistolare). To this they apply the current ancient authorities, De inventione and Ad Herennium, adding for its maxims of aptness Ars poetica. The ampler manuals include the sacred list of figures obligatory in the poetriæ.[15] Otherwise the application of the ancient rhetoric is practical and pointed.

For in fact the ancient lore was immediately and practically applicable. It did not, as in the poetriæ, have to be perverted. Of the traditional five parts of ancient rhetoric, inventio, dispositio, and elocutio, though not pronuntiatio and memoria, bear directly on letters, whereas the first two have nothing properly to do with verse-writing. Elocutio is applied practically by being focused on the cardinal ancient virtue of appropriateness; artistically, by elaborating compositio as prose rhythm. Immediately adaptable were the five parts of a speech. The exordium is always cardinal in a letter as benevolentiæ captatio. Narratio applies exactly in its proper sense of statement of the facts. Petitio, though it has less scope, is quite pertinent. Conclusio, though varying most from its ancient function, has some general correspondence. In a word, the classical doctrine for the parts of a speech applied to a letter by mere reduction of scale.

Dictamen was equally practical in actual teaching. Besides giving exercises in correctness, it compelled attention to elegance. Its study of appropriateness was readily extended by such imaginary adaptations as were inculcated in the ancient prosopopœia, and thus provided in writing the kind of practise sought orally in the ancient controversiæ.[16] John of Garlandia’s confused combination of verse and prose in a single manual[17] is insufficient to prove that the two were often taught concurrently; and the higher tone of the artes dictandi suggests that they were addressed to older pupils.

Dictamen is defined with the usual inclusiveness as apt and elegant writing, inseparable from subject matter, depending on native ability, teaching, and practise, using to some extent all five traditional parts of rhetoric, but mainly style (elocutio).

All the rest of Book I is devoted to compositio as rhythm.

Considering the positions of nominatives and of oblique cases, of relative clauses, of infinitives, of locutions fixed in certain places by usage, the author quotes Geoffroi de Vinsauf: “A noble gravity comes from order itself when what is joined by syntax is separated by order.”[21] A word ending in a consonant should be followed by one beginning with a vowel, and vice versa. The cursus should not be continuously swift or slow but varied:

So the dictator will add, subtract, or transpose.

He will consider the rhythms of terminations:[22]

The rule for cadences (de finibus distinctionum) is that they must satisfy.

These considerations open the large principle, fundamental in ancient rhetoric and clearly discerned by this writer, that sentence skill consists in composing rhythmical units in a total movement, i.e., in composing by heard clauses.

In punctuation the author prefers the simple and sparing Roman use, and deprecates applying to dictamen the rules followed in pointing the religious offices.[25]

The final consideration of this section is of sentence length. Book I closes with general advice.

The second book, on stylistic ornament (ornatus, dignitas) is devoted to the traditional figures. These are the same list from the Rhetorica ad Herennium as serves for the poetriæ, and are rehearsed in the same order.

The three kinds of dictamen are: (1) prose, i.e., free composition (sermo communis, or solutus); (2) metrical; (3) rhythmical (genus rithmicum), which observes syllabic equality and rime.[26] “Since we have nothing, however, to do with the two others, ... let us now proceed to the prose dictamen, the dictamen of letters; for that is demanded of all and recognized as of great utility. It increases eloquence, promotes favor, enlarges honors, and often enriches the needy.” The typical parts of a letter—there may be more, or fewer—are salutatio, exordium, petitio, conclusio. This whole book is devoted to prescriptions for the first. The salutatio must always be in the third person. Its order is determined by the relation of the rank or dignity of the sender to that of the recipient, though this in certain cases is waived. It is careful of titles and of their appropriate modifiers.

The book closes with a protest. Merely to consult a formulary is a poor substitute for studying the art of salutatio. It is like swimming with corks. “The ideal (forma) of salutatio which we have here worked out, well grasped and held, will save writers both from borrowed plumage and from deficiency.”

The exordium is such a prelude to the statement of the facts as will make the hearer[28] open-minded, well disposed, and attentive.

Narratio, statement of the facts,[32] should be concise, transparently clear, and plausible (brevis, dilucida, verisimilis). The application of the ancient maxim is to the connection of narratio with exordium by proper conjunctions and in several possible cases. Similarly are set forth the ways of connecting it with the following petitio, which is summarily defined. Conclusio in a letter is not, as in a speech, the logical result of proof. Rather it is the satisfaction of whatever expectations have been aroused. It may be affirmative, negative, or conditional, so long as it is a satisfying close. Modes of connecting it with the rest of the letter again have most space.

The fifth book, reviewing all these items in the same order, constitutes a summary manual, a summa dictaminis.

This book by itself provides about the same quantity of precept as suffices in other manuals to introduce collections of specimens; but it is exceptionally systematic.[33]

In the twelfth and thirteenth centuries the practise of cursus became more technical, and the use of the term more special. The Chancellor appointed by Pope Urban II was John of Gaetà (Giovanni Gaetano), who had learned dictamen from Alberic of Monte Cassino,[34] and who was specifically commissioned to “reform the style of ancient grace and elegance in the Apostolic See” and to “restore the Leonine rhythm with its lucid rapidity.”[35] The dominant chancery style, thus reformed, embodied the principle of rhythmical close in three types of prose cadence:

The first sentence of a mandate of Innocent III shows all three.

The indignant closing paragraph of Dante’s letter to the Florentines begins:

The fixing of these three cadences in Roman use would of itself have given them a vogue beyond letter-writing. Moreover they were chosen not at random, but as typical fulfilments of rhythmical expectation. For they appear also, though not exclusively, in the collects of the liturgy and in the daily offices.[39] The collect for the fourth Sunday after the Epiphany has all three.

The longer measures of the collect composed in 1264 by St. Thomas Aquinas for Corpus Christi fall upon velox.

The English translation, following this cadence in the first measure and the third, departs in the second.

Hearing these cadences over and over, preachers of course used them often in Latin sermons. Thus the cadences that were obligatory in ceremonious letters, and habitual in other use, confirmed the conception and practise of Latin prose as rhythmical.

The rhythmical conception of prose, as the middle age knew well, had been dominant in classical antiquity. But medieval Latin prose, though its tradition came through the schools of Gaul, had had its own development. As in verse,[42] so in prose, the shift of rhythmical control to stress opened, not a breach, but a new artistic life. Medieval Latin ran as a living language with the movement of living speech. This was recognized even by the grammarians.[43] So the thirteenth-century manuals of dictamen kept the terminology of ancient metric to describe accentual rhythms[44] because Latin to them was present as well as past, and seemed to reach indefinitely into the future. The future, by the time of the Renaissance, was seen to be with the vernacular for poetry. More slowly it was seen to be there also for prose. Meantime Latin poetic and rhetoric came to be studied by revival of antiquity. Revival translates Renaissance exactly. Italian scholars, and after them French and English, sought to turn Latin prose back to Cicero. Though they thus accomplished some most worthy ends, their theory and their practise interrupted Latin prose composition. Their contemptuous rejection of medieval rhythms hastened the processes by which Latin became a “dead” language. In the middle age it was living. A conspicuous evidence of its vitality, not merely a legal technic, but the formulation of a rhythmical habit, was the cursus.

CHAPTER IX PREACHING — References and Abbreviations — A. Vernacular to the People, Latin to the Clergy — B. Collections — C. Manuals — D. Symbolism — E. Composition — 1. Imaginative Method — 2. Logical Method — F. Style — 1. Rhythm — 2. Balance and Rime — 3. Refrain — FOOTNOTES:

The historian of twelfth-century preaching finds the “prodigious success”[1] of St. Anselm’s sermon on the Assumption not even faintly echoed in the flat record. The same disappointment awaits any one who shall turn from the fame of John Bunyan’s preaching to his printed sermons. In both cases the reputation is more convincing than the record. The testimony leaves no doubt that the preacher was eloquent; but his eloquence has not been preserved. The seventeenth-century record and the medieval, though in different directions, are alike insufficient.[2] Even when we have on the page before us the very sentences—and often we have something less, or even something different, we lack something of the eloquence. For the difficulty of transmission is so fundamental that it thwarts even the records of our own day. Oratory is typically the energizing of a message by a speaker for a specific audience. Its style depends on all three. In varying degrees all three enter into its composition. The occasion becomes part of the message. Between speaker and audience there is mutual response. Therefore of all the arts oratory is the most perishable.

But the history of oratory, though often baffling, is not hopeless. So perishable that it can never be quite recaptured, oratory may be so charged that even through imperfect record much often transpires. For any given period its records not only reveal much of the habit of the time, but conversely can often be so interpreted by that habit as to become really significant. Medieval sermons can thus be related not only to the medieval pedagogy of rhetoric, but to actual habits of composition and of style. In a form of oratory continuous for nearly two thousand years they will show at least what in preparation, composition, utterance, record is typically medieval.

For the middle age preaching is the characteristic form of oratory. Political oratory being in abeyance, legal oratory having little scope, preaching practically monopolizes the third field distinguished by Aristotle, occasional oratory, the oratory of here and now. Teaching, clearly of course a sermon function, sometimes becomes the main object; but even so it is not incompatible with other oratorical use of the occasion. Indeed, a sermon hardly succeeds as teaching except through the typical means of occasional oratory. For a sermon is different from a lesson, and even more different from an essay. As a form of oral composition it has opportunities and methods distinct from those of the bar or the senate. Occasional oratory, always beset by temptations toward sophistic,[3] has always opened on the other hand the highest ranges. In preaching, the safeguard against sophistic is in the distinctive use of the occasion to move men to action. Always emotional, occasional oratory becomes in preaching a distinct form of persuasion. Relying less than political oratory on argument, reasoning less and pleading more, it is even more urgent toward a goal. This character, achievement alike and misuse, is vivid in the medieval preaching of the friars.

Medieval preaching is occasional oratory not only in the ancient application to special solemnity, but more significantly in using the familiar recurrences of the Church year. The whole western world kept not only Christmas, but Candlemas, Michaelmas, and the rest. Nothing is more characteristic than the focus of popular emotion upon Corpus Christi. The calendar of feasts and fasts was the actual calendar. A medieval auditory had a great common fund of conscious and subconscious associations. The surrounding symbolism of stone and glass was familiar from land to land. The pulpit was beside the altar. The regular sermon on a Sunday or a feast was at Mass after the gospel, which furnished its text.[4] The atmosphere of prayer was intensified by specific petition preceding and following. “Let us therefore ask our Lord to give me good words for you” ended the exordium, or “Pray that we may be illuminated,” or “Pray the Lord, therefore, that by the power of spiritual teaching to-day your hearts may be uplifted”;[5] and the congregation united in Pater and Ave. So the sermon ended with prayer, and was immediately followed by common intercessions. Though some of these conditions are constant throughout the history of Christendom, never before or since have they combined at once so amply and so widely to constitute the conditions of preaching. In this special sense medieval preaching was the oratory of occasion. At no other time has oratory had at command so large and constant a fund of common emotional associations.

The appeal of medieval preaching to this common emotional fund, though it cannot often be measured, can be analyzed often as to style and composition, oftener as to habit of thought, in abundant documents. Even in print hardly any medieval material is more abundant—and there are further stores in manuscript—than collections of sermons. In order to interpret these, we must constantly remember how they were made and why. First, medieval sermons are habitually gauged to one of the two typical audiences. Either they are for the lay folk in parish church or cathedral (sermones ad populum), and were preached in the vernacular; or they are for the clergy (sermones ad clerum), i.e., before synods, councils, schools, oftenest of all in monastery chapels, and were preached in Latin.[6] In either case, until the late middle age, the record was always in Latin; in either case the preparation also, the notes and outline and more or less of the further composition, was usually in Latin. In the former case of a sermon preached to a lay congregation the record gives at most only a translation.[7] In the latter case of a sermon preached to a monastic community the record is at least nearer to the spoken word. We can press further, therefore, the analysis of sermons preserved by such communities as the Cistercians and the Victorines.[8]

Though popular preaching is no less worthy of study, less of it has been transferred to the written page.[9] But translation, though it could convey little of the spoken vernacular style, must not be thought of as involving the difficulties of to-day. In the middle age every cleric was bilingual; and every preacher of distinction enough to be recorded not only read and wrote Latin habitually, but spoke it fluently.[10] His schooling having been in Latin, both written and oral, his daily offices, his reading, and whatever writing he did being still in Latin, he used Latin easily and naturally in thinking out and ordering a sermon or in giving another’s sermon, or his own afterward, the permanence of writing. Since he would not be embarrassed, as young preachers sometimes are to-day, by trying to recall phrases and sentences, preaching in the vernacular from Latin notes might even leave oral composition the more free.

The other process, the final rendering in Latin, was by no means always of the sermon as preached. It might be merely a digest, or more expansive, or the whole sermon. All three degrees seem to be exemplified among the printed sermons of Fulbert.[11] The compilers of Odo of Morimond say: “he was very eloquent; but in writing his sermons we have ignored the form for the substance.”[12] Collections often show distinctly less concern for record, for commemoration of achievement, than for guidance. Some even of the more important collections have evidently this practical aim. They propose not only inspiration, but practical suggestion and direction for future sermons on recurrent themes. They are less anthologies than repertories. Maurice de Sully, Bishop of Paris, thus designed his widespread collection primarily for the pastors of his diocese.[13] The study of preaching, not merely the preservation of notable sermons preached in his time (sermones reportati de auditu), seems to have been the object of the collections of Pierre de Limoges. Clearly the best known of these, his Distinctiones,[14] arranging not only whole sermons, but outlines and suggestive passages alphabetically by subjects, is meant for reference. That the collections commonly had in view the practical use of successful sermons as models is suggested again and again, and sometimes indicated.[15]

The great thesaurus of Jacques de Vitry[16] is systematically practical, a collection not primarily of sermons, but for sermons. Models, outlines, suggestions, intended for adaptation in the vernacular[17] are arranged according to the Church calendar: four books for the seasons, a fifth for the saints, with three expositions for each day, the first of the introit, the second of the epistle, the third of the gospel. The sixth book is classified for adaptations to typical social groups. What has naturally most attracted modern historians in this storehouse is the abundance of illustrative descriptions and stories, the exempla. A resource prized in all times for popular address, the exemplum was so cultivated in medieval preaching as to call forth many collections.[18] It looms unduly large to modern readers because the exempla have now an extraneous interest in reflecting medieval life, and because they were taken not only, as now, from contemporary life, from history, from legend, but also from the bestiaries.

The mythical zoölogy is now often too diverting to keep its effectiveness as illustration. But to recover this, and to realize that the bestiaries do not convict the middle age of credulity, we need not go back of St. Francis de Sales. Exempla, after all, are a commonplace of preaching because they are a necessity.

The general thesaurus and the special[19] alike confirm the impression of wide use of a common fund of topics and illustrations. But sermon helps are not peculiar to the middle age. Every age has offered to its preachers something equivalent to the medieval handy manuals.[20] Not only preaching, but all occasional oratory, has always been beset by the temptation to rely on pattern and stock.[21] That the middle age had many conventional sermons is less significant than that it had St. Bernard and the early Dominicans.

Medieval manuals of preaching are no more specific than modern manuals as to their rhetoric. That is typically generalized, or even taken for granted. For instance, Guibert de Nogent’s Liber quo ordine sermo fieri debeat,[22] concerned entirely with interpretation, has no rhetoric at all. The anonymous Chartres work on popular preaching quoted in part by Clerval[23] gives, besides the usual procedure and advice, a practical counsel for preparation and one for delivery. The first is to fix meditation on the specific object of that sermon and on the specific means most likely to bring it home, not on the words. The words will come if the matter is surely ordered. The second is to increase urgency from quotations of texts through exempla uttered with loud appeal until the congregation shows emotional response, then to moderate the tone so as to lead quietly to the final benediction.

Another anonymous manual[24] sets forth eight methods of expansion (De dilatatione sermonum): (1) definition and exposition of terms, especially of their “moral” significance; (2) division, not to be so minute in a sermon as in a treatise; (3) proof, by contraries, enthymemes, examples, again to be guarded from subtlety; (4) citation of confirmatory texts; (5) degrees, i.e., positive-comparative-superlative, as in exhibiting one of the virtues; (6) metaphors, not to be multiplied nor mixed; (7) symbolism, i.e., allegory, tropology, anagogy; (8) cause and effect. Most of these are commonplaces of rhetoric. Some of them, and their use for dilation, suggest the same preoccupations as the poetriæ.[25]

The Dominicans, devoted to preaching, maintaining a long and severe discipline of studies, provided systematically for oral composition. “Legendo, studendo, disputando”[26]—the last tallies not only with Dominican fame, but with the focus, and the increasing emphasis, of oral training on logic.[27] Otherwise the manual of the fifth Dominican General, the De eruditione prædicatorum of Humbert de Romans, says little of rhetoric. Rather it is practical psychology, as are the second book of Aristotle’s Rhetoric and the sixth book of Jacques de Vitry.

Practical by another method is the manual of Alain de Lille, Summa magistri Alani doctoris universalis de arte prædicatoria.[28] “First we must see,” he says at the end of his preface, “what preaching is, its distinguishing qualities of expression and of thought, and its typical species; secondly, who should preach; thirdly, to whom; fourthly, why; fifthly, where.” His first chapter, after brief consideration of preaching as public speaking, sets forth the special sanction for sermons of the gospels, the Psalms, the Pauline epistles, and the “books of Solomon” as offering instruction in morals. Winning his audience by his own humility and by the practical import of his message,[29] not by appeals to applause, the preacher proceeds to the twofold exposition of the text. “Let his words be sometimes emotional (commotiva), to soften the minds of his hearers and bring tears. But let the sermon be concise, lest prolixity bring instead boredom.... Finally he should use exempla.”

Chapter ii proposes for a theme the scorning of the world (De mundi contemptu), and for its text, “Vanity of vanities; all is vanity.” Distinguishing the vanity of this world as thus typically threefold, Alain quotes Romans and the Psalms and adds Persius. “Thus the preacher should confirm each division of his text. His next step is to show where is vanity, where vanity of vanities, where all is vanity.... The scorning of the world, thus established by the way of the world itself, should be established also by consideration of man’s end (a causa finali).” The peroration apostrophizes human blindness (O homo, si mundana, etc.).

In a word, Alain furnishes a specimen order and procedure. Succeeding chapters thus present other usual topics: scorning of oneself, gluttony, the other deadly sins, the heavenly hope, spiritual grief and joy, patience and each of the other virtues, love of God and of one’s neighbor, etc. The remaining headings of his prefatory division are taken up at chapter xxxviii; and his consideration of different types in congregations has the humor to include the sleepy. The title Summa, then, implies a preacher’s book of reference, with advice and direction subordinated to classified material. The distinction of this Summa is its systematic conciseness.

Much more distinctive than any provision of collection or manual for ordinary use of a common fund is the prevalence of symbolism. The “moral” exposition that followed the literal commonly had this direction. For medieval symbolism was not merely a habit of exposition; it was seen as well as heard because it was a habit of conception. The cathedrals still exhibit in sculpture and glass what came in words from their pulpits. The Psalms, especially those recurring in the daily offices, thus conveyed from generation to generation the eternal word of God. For every psalm was the lyric cry of David, or of some other ancient Hebrew, only in the first instance (litteralis expositio). It always meant more; and its further meaning (moralis expositio), as cumulating progressive experience of God, was more important. It always meant not only David, but every other soul thus answering God, the “poor man” of every time and of every clime, every individual then and there focusing and enlarging by those words his own experience.

So far the “moral” interpretation treats the psalm as an extreme case of literary inspiration, appropriates it as having the universal validity of every lyric that has survived the centuries. A religious classic, more widely and intensely spiritual than any other body of lyrics, the psalter is still only a classic. But medieval interpretation always took a third step. Every psalm meant not only David, and further not only every “poor man” making it his own, but also “the man,” the Lord incarnate to share all humanity and to give men “power to become the sons of God.” Thus the Psalms spanned the centuries, connected the New Testament with the Old indivisibly, and were the voice of man in every age answering God in the Church. The habitual use of the Old Testament, and even of secular history, as prefiguring is here focused in its simplest form.[30] Simplest and commonest instance of symbolism, the interpretation of the Psalms is characteristic. Though not, indeed, peculiar to the middle age, it opens a widespread medieval habit.

The habitual use of the Song of Songs as symbolic of the Christ and his Church reminds us that symbolism is essentially poetic. It was the usual poetic of medieval preaching in interpreting not only the Scriptures, but all human experience. As behind the word lay spiritual meaning, so behind the veil of the senses lay the realities of life. The so-called “otherworldliness” of medieval preaching is no ignoring of physical facts; it is a sustained and consistent call to see through them. What medieval philosophy defined as seeing “in the aspect of eternity” (sub specie æternitatis) medieval preaching inculcated as a habit of vision. The habit might, indeed, degenerate, as in that popular manual of the fourteenth century, the Speculum humanæ salvationis,[31] into extravagance or minute formalism; but even so it kept some view of history as a providential progress. It might rise, on the other hand, to express the immediate apprehension of the mystic, the vision of spiritual genius. But the distinction of the medieval mystic is rather in degree than in kind. The whole age is characteristically habituated to mysticism. No other poetic vision is equal to Dante’s; but no other form of poetry in his age, whether in verse or in sermon, is commoner than the vision.

Not only to the rapt was the visible world eloquent of the unseen eternal.

In the following sermon of this series St. Bernard refers seriatim to the accepted prefigurations of the Blessed Virgin. She is the perennial antagonist of the serpent (Genesis iii. 15), the mulier fortis of Solomon (Proverbs xxxi. 10), the burning bush (Exodus iii. 2), Aaron’s budding rod (Numbers xvii. 8), Isaiah’s rod from the root of Jesse (xi. 1), the rod that smote the rock (Exodus xvii. 6) and divided the sea (Exodus xiv. 16), Gideon’s fleece (Judges vi. 37-40), the woman who is the Lord’s new creation (Jeremiah xxxi. 22). The prefigurations of this list had an obvious appeal in sermons through being familiar in hymns, in sculpture, in glass. Evidently St. Bernard loved them for their poetry; but he is also convinced of their sanction. He even formulates the theory of symbolic prefiguration.

His second sermon for Advent carries out the prefiguration of the rod, or stem (virga).[32]

The symbolism of his sermon in the Epiphany octave on the marriage at Cana is specifically moralized.

Such preaching shows the same preoccupation as the symbolic windows of the cathedrals, their carved capitals, above all the thronged but harmonized groups of their great porches. It is not merely conventional illustration; it is a constant and consistent reminder of the history of mankind as a scheme of redemption.

The most conventional form of symbolism is allegory. This may be no more than what Lowell calls “personification by capital letters,” as in the Vice and Virtue, the Reason, Feeling, or Nature, of the eighteenth century. Even so it has more point when the convention is as instantly recognizable as the seven deadly sins of medieval capitals and poems, or the four cardinal and the three theological virtues, or even the seven liberal arts, each with appropriate costume, attitude, gesture, or action. How pleasant allegory may be in verse is evident in the wide popularity of the Roman de la Rose.

But no less evidently such pictorial aggregation offers little to preaching, since it has no vigor of composition. To be moving, allegory as well as higher symbolism must be narrative. The classic demonstration of this, the Pilgrim’s Progress, was written by a preacher. The art has never been more widely or more expertly pursued than in the middle age. The fourteenth-century English Piers Plowman suggests even more vividly, with its strongly oral manner, how allegory might move in preaching. A sermon might be not only enlivened by passages of easily recognizable description; it might further be vivified by making the figures individuals and by putting them into speech and action. A complete demonstration is furnished in the interlude, prologue, and tale of Chaucer’s Pardoner. That accomplished rascal is made to display the whole art of oral narrative, from its vulgar drama to its high ranges of symbolic suggestion. The case is extreme; but the method has been a resource of popular preaching from the days of Nathan’s rebuke to David down to the present. Oral narrative, often with dramatic dialogue, has been heard again and again, both in such base uses as the Pardoner’s and in such urgency of message as Bunyan’s, because it has been found surely persuasive.

Here is the high art of the exemplum. Often the medieval record puts us off with a summary; sometimes the manuals give a mere note, “Tell the story of so-and-so”; but sometimes there are clear clues to such narrative art as makes the Pardoner’s tale breathless. An exemplum of Walter Map’s lets us divine how dramatic the tale might be in the pulpit.

Composition as a progress of thought, logical sequence, ordering by paragraphs, shows on the other hand little that is distinctively medieval. As in other times, differences here are rather individual. Bourgain does not establish his assertion of a general weakness of order;[33] and his addition that even St. Bernard proceeds by leaps[34] is extravagant. Lecoy de la Marche has more warrant for asserting that medieval sermons generally lack such sustained progress[35] as Bossuet’s. For often the course of a medieval sermon is not only shorter;[36] it is less sustained by a progress of paragraphs than forwarded by stages of emotion. The latter method may be for its audience no less valid. Indeed, preaching must learn poetic as well as rhetoric. Its composition, ideally embracing both, will be bent toward the one or the other by the particular preacher and audience. The medieval abeyance of sustained logical progress, if in fact it was general, may have been due to preoccupation with those poetic methods which were effective with some of the greatest preachers.

The school lore of rhetoric, however, was weak in the inculcation of logical progress. Busy with style, confusing in this study rhetoric with poetic, it offered too little practical guidance for preaching as composition. Doubtless the lack was filled in some cases by the teacher;[37] but the shift of pedagogy to logic implies, among other things, dissatisfaction with rhetoric.[38] What medieval rhetoric lacked, medieval logic, for all its triumphs in other directions, could not quite supply for preaching. The disputations of the schools, excellent for general training in oral fluency, in precision, in detecting and meeting error, could not reach the specific skill of continuously instructing, winning, and moving a silent congregation. They might even lead a vain or unwary preacher astray. Medieval preaching has been accused of habitual over-division, of tedious minuteness in headings and subheadings. Unfortunately such sermon divisions are not peculiar to any period. They seem no less common, for instance, in printed English sermons of the seventeenth century. But the medieval record, at least, does not always reproduce the actual preaching method.[39] Sometimes it gives little beyond the outline, which thus seems the more barren. Sometimes it uses a sermon as the occasion for further development, rather providing for future than following actual exposition of the text. Where the record seems to follow, there is evidence enough to make probable that the dominance of logic did deviate many late medieval sermons into over-analysis.

Analysis, the very function of logic, can never suffice, as Aristotle makes clear, for presentation. In preaching, its value is not for organizing, but for preliminary study. It is rather for the preacher than for his audience. It belongs rather in his notes than in his sermon. But though doctors analyzing daily in the schools might rely unduly on their habitual method when they preached, they preached oftenest to the schools themselves, that is to special audiences habituated to logical method. The popular preaching of the Franciscans can hardly have had that bent. The Dominicans, severely trained in logic, were also trained specifically to preach. St. Bernard, preaching both to monastic communities and to the people, will not be accused of over-division. Neither St. Anthony of Padua nor St. Bernardine of Siena can have set Italy on fire with elaborate analysis. These considerations, though largely a priori, are at least as weighty as contrary inferences from the sermon record. They permit the belief that medieval preachers realized as generally as preachers of other times the apothegm of St. Ambrose: Non in dialectica complacuit Deo salvum facere populum suum.

Medieval sermons, for all the defects of transmission, offer ample and various demonstration of the force and beauty of medieval Latin. In contrast to the archaistic composition of the Renaissance, medieval sermon Latin moves at its best with ready variety and even in ordinary use with easy fluency.

Pierre de Celles,[40] fairly typical of ordinary expertness, is clearly rhythmical. St. Bernard’s ardor and winsomeness, the oral immediacy of his expression, appear in no resource of his style more strikingly than in his very pace.[41]

Balance, beloved of occasional orators from Gorgias down,[42] is useful in sharpening the iterations and contrasts of preaching.

Such balance often leads, through mere inflectional correspondence, to rime. That rime of clause with clause was avoided as a vice even in classical Latin is an exaggeration of Renaissance scrupulosity.[44] The middle age, following rather Augustine and the tradition of the schools of Gaul,[45] found rime desirable, and often sought it to mark parallel or contrast. Simple, ordinary use of it is heard again and again in a sermon by Hilduin.[46]

Elinand uses it to point an epigram.[47]

Richard of St. Victor’s rime often marks insistently exact balances.

It may mark a progressive iteration. Repeating florida and transeamus from what he has just said, he goes on:

Even such insistence remains within the limits of rimed prose; it is only extreme use of a widespread habit. Exceptional, on the other hand, since it repeatedly verges toward verse, is St. Anselm’s Lament of the Magdalen at the tomb.[48]

The device was so easily abused as to call forth more than one warning;[49] but it was so obvious a means of emphasis as to be widely prevalent.

St. Bernard, though he does not avoid rime, shows habitually no need of it to strengthen his iterations. With him iteration often advances from rhetorical cumulation to poetical refrain.

His fifth sermon for the vigil of Christmas uses incremental iteration to contrast to-day and to-morrow.

Refrain reinforces other incremental iteration in an ardent sermon on the Magnificat.[50] The very insistence exhibits strikingly the value of cumulative progress for charging exposition with emotion.

CHAPTER X POETIC ACHIEVEMENT IN VERNACULAR — A. Lyric and Epic — B. Experiment and Convention in Romance — 1. Romance in Latin and in Vernacular — 2. Walter Map and Marie — 3. Chrétien de Troyes — 4. Conventional Composition — C. The Poetic Composition of the Divina Commedia — D. The History of Medieval Verse Narrative in Chaucer — 1. Poetic Conventions in the Earlier Poems — 2. Poetic Innovation in Troilus and Criseyde — 3. Criticism of the Poetriæ — 4. The Poetic of the Canterbury Tales — FOOTNOTES:

Lyric, earliest medieval poetic achievement in Latin, shows full development in the vernaculars. On this the widest Latin influence was of course the hymns. The most distinct and conspicuous vernacular poetic, that of the Provençal troubadours, is mainly a minute and elaborate stanza technic. The fame of troubadour virtuosity gave vogue in other vernaculars to metrical skill. This is its historical significance in the development of medieval lyric. Lyric in any time, however, has a short history. As in Greece, so again and again, having perfected its diction and its metric, it comes to its own and abides. For more than any other form of poetic it is timeless; not ancient, nor medieval, nor modern, but perennial.

Germanic epic,[1] surviving as a strong influence in medieval English romance, persisted as a distinct poetic among the Scandinavians. In remote Iceland they developed an elaborate technic of verse epic, and later a distinct art of prose epic. Not even the troubadour poetic in the south is more detailed and elaborate than the epic doctrine of the Skáldskaparmál.[2] No medieval stories have more force of narrative directness than the prose sagas; and few have as much. In narrative stripped and stark the chief medieval author, perhaps, is Snorri Sturlasson. But in spite of Viking travels the Old Norse poetic of verse and of prose remained apart from medieval habit. Bounded by its own civilization, it was so little appropriated in the general medieval development of narrative that its recovery in modern times was as startling as that of the Etruscan sculpture at Perugia.

French epic, flowering later than Germanic, is already in the eleventh century tinged with romance. The French chansons de geste reflect a society distinctly feudal; their communal sense is of a larger community; and they constitute a first chapter in European vernacular poetic. Epic still in characterization and in the youthful zest of their combats against odds, they are swayed by new literary currents. The physical environment is less sharply detailed than in Germanic epic, perhaps because a kindlier nature was more taken for granted. A sharper difference is in style. They have none of that Germanic conventional elaboration which made epic diction a language apart. Much simpler,[3] they are also more diffuse, in both respects nearer to common speech. The movement in detail is of ten-syllable verses gathered into irregular stanzas (laisses) by assonance. A more significant distinction is the habit of narrative. Roland presents substantially the same situation as Maldon. In Anglo-Saxon it is focused by concentration; in French it is approached as the culmination of a series. Habitually French epic is more extended and fluent, expansive not descriptively but historically, running through more events in a single composition. Whereas Anglo-Saxon epic typically selects and intensifies a crisis, French epic tells the whole story.

Medieval story-tellers in turning to the vernacular found wider opportunity. The Latin poetic that they all studied in school, practised in historia, fabula, argumentum, and at greater length in saint’s legend, was too much absorbed in descriptive elaboration to teach them much of narrative. Outside the cloister, at the regular stations of the pilgrimage routes, waited audiences of increasing size and diversity; and the development of feudalism established the castle hall as a social center. The development of vernacular narrative answered with the more social art of romance. The narrative future was increasingly for the vernacular. Meantime Latin narrative continued not only for its more special audience, but in various interplay. Exempla, sketched or summed in Latin, were preached in vernacular. An ugly old legend of folklore, moralized in Latin for an exemplum, Chaucer gave back to the wider audience in the exquisite art of his Prioress. Romances and legends of saints interpenetrated. The complicated development of the Grail legend must be followed in both Latin and vernacular. Nevertheless poetic was beckoned into new narrative paths mainly by the wider appeal through the vernacular.

Something, of course, was carried over from the poetriæ,[4] oftenest the conventional lore of style. By contrast we are made the more sharply aware of the escape and adventure of individual bent and native inspiration. Aucassin et Nicolete has another appeal by another method. Its style, in spite of occasional conformity, is not in the tradition of the poetriæ. Chrétien, though conforming more, draws his psychology rather from observation of social habits and even of individual character than from the classified lore of appropriateness to type. Dante not only commits the most serious enterprise of medieval poetic to the vernacular; he utterly ignores the Latin school lore. At the end of the middle age, the Renaissance already tinging his thought, Chaucer passes through a whole course of Latin conventions, adapting critically as he goes, and ranges beyond.

In the twelfth century Walter Map’s Latin notebook De nugis Curialium[5] shows both the persistence of old forms and the stirring of new in rendering new material. Some of his stories are clearly oriental. Historical anecdotes and exempla, often given in mere lucid summary, occasionally indicate narrative composition.[6] He tries a story in the way of Ovid, or again in the Petronian way of the “Matron of Ephesus.” Two longer tales show not only a firmer conciseness, a sharper vividness, but what is more significant, grasp of narrative movement. The Friendship of Sadius and Galo (III. ii)[7] and Sceva and Ollo (IV. xvi) are enlivened in scene and furthered in plot by dialogue. Both are carried through progressively to an issue. The former skilfully combines the Orestes-Pylades motive of friendship with the Joseph-Zuleika motive of the scorned queen. The latter turns the cynicism of its oriental source to brisk social satire. Here in Latin are narrative experiments full of promise. Looking over Walter’s shoulder, we can divine a new literary stir.

More suggestive of response to a new interest of his audience is the abundance of Celtic folklore. There are banshees, wandering dead, a nicker, a pact with a demon. King Herla (I. xi), like Rip Van Winkle, stays too long in fairyland. The fairy mistress appears four times. Once (II. xi) the old tale is hardly more than notes. In Edric the Wild (II. xii) the moonlight dance (chorea feminarum) keeps its ancient spell. Meridiana (IV. xi) associates a different and longer version quite disconcertingly with the great Gerbert. Though none of these brings the tale to the sequence of Sadius and Galo, we feel an artist at work. Some of these shorter tales may be Latin summaries, like the exempla of the collections, for telling in French. Whether he worked out his narrative art in the vernacular we do not know. Though some critics have inferred for him a considerable part in the development of the Grail legend, the only surviving work that is certainly his is this artist’s sketch-book.

There are fewer suggestions of narrative experiment in the contemporary French lais of Marie.[8] Her Yonec is satisfied with the rapid summary of Latin tradition. Her charming Honeysuckle is an episode of the Tristram story. Iseult, finding a peeled wand by her way through the wood, knew that Tristram was near. As much lyric as narrative, such episodic poems suggest what the lai may have been in its earliest use of Celtic fairy adventure. Lanval and Eliduc handle with more narrative progress what we now call a situation. Another rendering of the former situation, the anonymous Graalent, shows by contrast Marie’s discernment either in choosing a version narratively superior or in herself reshaping. Generally her art is less of composition than of style. Her habit with a situation is not the fluent onwardness of Aucassin et Nicolete,[9] nor the intensive and progressive sequence of the later Chastelaine de Vergi.[10] The technic of the latter, though distinctly realized in the middle age, remained exceptional. It receives little attention even in the Decameron; and it stands out among the Canterbury Tales.

A romance of five or six thousand lines, but still selective, was developed by Chrétien de Troyes.[11] His Erec tells the tale which Lady Charlotte Guest translated in her Mabinogion as Geraint the Son of Erbin from a later Welsh version. Enid is stricken with shame that Geraint’s love of her should run to ignoble fondness. Her lament, overheard and mistaken by Geraint, rouses his jealous pride to prove her long and cruelly. This moment the Welsh writer sees in its setting.

At such a point the Welsh writer’s abundance of description is not merely pretty; it is fitting. What he thus dwells upon until we must feel it because we can see it, is an important moment in the story. But in general he blurs his story by spending equal elaboration on what is quite subordinate, or even irrelevant. His incidents claim attention equally in succession; Chrétien’s are lengthened or shortened with a sense of their narrative values. For Chrétien, relying less on the traditional picturesque description, has the better narrative. Using descriptive detail less for itself, he uses it more to bring out character or give to important moments salience. Instead of merely accumulating adventures and telling each for what it is worth by itself, as is the habit of the longer romances, he has selective art enough to bring out those that will keep attention on the conquest of the proud, selfish devotion of the husband by the nobler devotion of the wife. And at the end he impresses the significance of the whole course of adventures, the meaning of the story as a whole. He discerns, though he does not always achieve, the narrative force of unified progress.

How far the art of such romances was studied in the middle age appears again on comparison of the fourteenth-century Ywain and Gawain with its original, Chrétien’s Chevalier au Lion. The Englishman, without losing anything of the plot, reduces the length one-third by compressing, modifying, or even omitting Chrétien’s detail. And since Chrétien’s detail is not merely added for richness, but spent to bring out character or mood, a change here is a change in the total effect, a shifting of interest from the persons to the events. Thus to Chrétien the central situation is this. A widow forced to marry again, as medieval widows were if they had property, accepts the slayer of her husband. How would she feel? Might she not, from making a virtue of necessity, come to love her second husband if he were young and brave? If, tiring of her riches and ease, he would be off to his wars again, would she forgive him for breaking his promise to return on a day? And might he not then, learning from the loss of her to value her truly, devote himself to winning her back by proving his better manhood? The situation is almost the reverse of that in Erec. Such questions of character and feeling lead Chrétien to dwell upon the scenes between Ywain and Alundyne, and even to comment satirically now and then on their mental attitudes. Most of this the English translator omits.

But that the omissions were deliberate appears in his throwing emphasis upon what directly furthers the movement of the story. What makes a story quick and strong he understands so well that he even quickens and varies Chrétien’s pace by turning some of the indirect discourse into direct dialogue. Nor was he insensible to Chrétien’s suggestions of character or mood. At the final revelation, where Chrétien says simply that Alundyne started (la dame tresaut), he renders:

This is exactly in Chrétien’s habitual manner. Rendering with discernment and spirit, then, keeping the plan and transitions that hold the tale together, he bends the story in the direction of his own interest and skill. He too knows something of narrative art. His simpler, more onward version throws into relief Chrétien’s superiority not only in delicacy of verse and style, but in those ampler and finer suggestions of character which bring out of a situation its deeper narrative values.

The increasing audience for romance invited much mere retelling, many versions inartistic or even unintelligent. But the scientific study of sources, while it defines the diffusion of current tales and carries their history back to folklore and myth, should not blur the history of medieval narrative art. In the midst of almost impersonal transmission were both artistic experiment and artistic achievement. Chrétien de Troyes and a few others have survived by name. Nameless, but no less convincing, is the narrative art of Aucassin et Nicolete, of the Chastelaine de Vergi, of Gawain and the Green Knight.

Generally the artist is less clear, the impersonal processes of transmission and the conventions of telling clearer, in the long romances. Adventure, when fairyland was no longer new, was sometimes supplied by mere aggregation. Love, presented mainly as wooing, crystallized into the code of amour courtois. Chivalry, a motive even more clearly ideal, could be no less presented as a code. Adventure, love, chivalry, established as the three habitual motives of romance, were readily combined in stock forms. Better artists vitalized them by characterization; but since romance can be composed acceptably without characterization, its journeywork, medieval or modern, tends to content itself with types. The hero is “a very perfect gentle knight”; and his action consists in having many adventures. The heroine is a beautiful young lady, and needs no action at all. Both may be described at length without being individualized. Even on these terms, much as they have always amused satirists, romance may have charm of style and some of the perennial appeal of youth. But it has no properly narrative vigor. Its composition tends so easily to repeat that the stock adventures can be assigned as well to one knight as to another.

In fact, the middle age saw adventures transferred from one hero to another who had meantime become more popular. Thus Gawain, coming straight out of fairyland, gradually lost more and more of his glory; and the process is one of the ways in which there grew up cycles of romance. The longer Lancelot stories are aggregations of many separate adventures. One of the best remembered feats in Malory’s compilation is the three-days’ tournament. Not only is this told of other knights in other romances; it is found also by itself, as in the tale of Ipomedon. Some medieval rewriter added it to his Lancelot; and there it stayed. The Lancelot story was further swelled by adventures formerly ascribed to Gawain, who had already lost some of his feats to Percival. Even the winning of the Grail quest, which had early been transferred from Gawain to Percival, went to Lancelot’s son Galahad. Finally the Lancelot aggregation was attached to the cycle of Arthur.

Whether or not it was attached to one of the cycles, a conventional long romance could thus aggregate. Bevis of Hampton or Guy of Warwick might be longer or shorter without the slightest narrative difference. It is long because it is interminable. Even in better hands the medieval long romance prevails part by part, as it was read. It was not composed as a single narrative. Such singleness as the middle age cultivated in romance must be sought in the parts considered as separate stories, and will be found oftener in the shorter romances that remained by themselves. For lack of it the most conventional long romances become series of typical descriptions. The typical hero, typically equipped without and within, has one typical encounter after another.

The difference between such aggregative transmission and narrative progress can be discerned among the many versions of the Grail. A magic talisman from folklore had been transmuted by the popular emotion focused at Corpus Christi;[13] but it became a narrative goal only for those with art enough to conceive the great quest as something more than a series of adventures. Meeting an earlier form of it in his Balin and Balan, Malory brought it into no distinct relation to that tragedy. Its recurrence in his later books is similarly unharmonized and inconclusive. Wolfram von Eschenbach, also using more than one version, had focused his Parzival sufficiently to give a long romance some movement onward.

The solitary eminence of Dante is a perpetual reminder of the limits of any lore of poetic. As for medieval lore, its approach to the Divina Commedia is hopelessly short. The ultimate reason, of course, for its inadequacy is that the greatest poetic achievement of the middle age is far more than medieval. It is for all time even more clearly than the Œdipus Rex or the Æneid or Othello; and like them it derives its greatness from something beyond poetic.

Nevertheless it is also medieval and also a great achievement of poetic composition. So to consider it is to discern both more of its greatness and more of medieval habit. In the current medieval lore the main lack is seen by contrast to be in the larger movement of composition. Medieval poetic carries us so short a distance toward the Divina Commedia because it is preoccupied with style. Vergil is currently cited and quoted, but usually as an exemplar of the three styles.[14] Medieval romances carried nationalism back to Troy for centuries without discerning the larger art of the Æneid in composition. For Dante Vergil was guide not only in a deeper sense, but also in poetic.

This makes clearer why Dante turned his back on current lore. He knew that lore, not only the elaborate technic of the troubadours, but the Latin poetria. His ignoring of it is tantamount to an arraignment of its cardinal weakness. It had too little technic for his main poetic concern, the movement of the whole. It had too much schoolmasterly fiddling with words to carry orchestration beyond a few conventional modes. No one can study the history of poetic without finding this deficiency exceptional only in degree. In kind it is historic. Here, again and again, appears the gap between pedagogically formulated poetic and poetry.

In Dante’s style the medieval poetria is not merely outdistanced; it is repudiated. Matthieu de Vendôme, Geoffroi de Vinsauf, John the Englishman, might have tolerated figures of wrestlers and dogs in hell, but hardly broth, cooks, oxen, and swine. They would have challenged bellows in purgatory; they would have been shocked at rooks and a fish-pool in heaven. The decorative descriptions of a rhetoricated poetic seem nowhere else quite so futile as beside Dante’s figures of precise geography.

As for Paolo and Francesca, Ugolino, and a hundred other passages, with those single lines that enrich memory,[16] no better praise can be compassed than Matthew Arnold’s word “touchstones.” They tell us, better than any definition, what poetry is. But the mere historical significance of Dante’s luminous precision is its vindication of that true theory of poetic diction which was formulated in the De sublimitate.[17] The essential character of poetic style is not dilation, which belongs to rhetoric; it is sublimation. Dante’s extraordinary conciseness, austere, ascetic, is never bare; it is surcharged. For style too, as well as for composition, he must have discerned that ancient critic’s fine distinction: rhetoric and poetic must never be confused; but at high temperatures they can be welded together.

Poetry conveyed vision oftener, perhaps, in the middle age than in any other period. Piers Plowman and Pearl carry on in the next century a persistent medieval preoccupation. The Divina Commedia, fulfilling this aspiration, reveals man’s need, his quest, his attainment, of vision. “Blessed are the pure in heart; for they shall see God.” Purity remains negative to those who have no desire to see God. But Dante answers those for whom it is positive and constructive.

The poetry of this “concreate and perpetual thirst” (III. ii. 19) animates for all time a whole cosmogony antiquated merely as science, and communicates the great philosophy of the middle age to the heart. The Divina Commedia achieved the high quest of the Grail romances. It is the vision poem.

The heaven of Dante’s vision transcends time and space.

But it is not reserved for rapt contemplatives.

For the Divina Commedia ranges beyond the lyric exaltation of the individual poet. It is suffused with social sense, as constantly aware as the Piers Plowman of the “fair field full of folk,” of the world of striving men and women. Its vision begins with society gone bad and frustrating itself; it arrives at society perfectly realized as immediate intercommunion and interaction. So the Church, the heavenly society, recurs constantly in communal echoes of hymns and canticles: Ave, Agnus Dei, Benedictus qui venit, Salve Regina. In the same spirit Cynewulf had composed his Christ upon the seven great antiphons of Advent.[18] Familiar gospels are recalled. Most frequent echoes, as of symbols most suggestive of common human experience, are of the Psalms.[19] The artistic harmony is based upon the medieval conception of social history as the redemption of mankind.

The word allegory has come to suggest vagueness, or conventionalization, or fancy. None of these has any place in the Divina Commedia. Allegorical, indeed, it may be called in a sense large enough to include Piers Plowman, the Faerie Queene, and the Pilgrim’s Progress; but no other allegory is so large in scope, so consistent and complete in composition. For such details as the sacred chariot, the eagle, wolf, and dragon (II. xxxii) we search the carved capitals and the windows; but including all these and ranging beyond them is the constant interpretation of the whole sensible, transitory world as typical of the eternal. Ephemeral matters, politics, even quarrels and grudges (I. xvi. 73), are interpreted permanently. The intensely specific realism is made constantly to serve idealization. Human life is revealed sub specie æternitatis.

For life beyond is seen as the prolongation of the line that we give to life here and now. Its motive power throughout is love radiating from God and leading men back to him. The disturbance of life is sin. The horror of the Inferno is of sin as perversion. The warpings of life there tormenting Italians or ancients are typical to the fearsome degree of revealing our own. Thus the course of the Divina Commedia is of ideas and principles of action so embodied as to lead emotion on. What is familiar in lyric is here carried through a whole conception of life. Theology is translated through vision into emotion and will, as if lyric were carried through into drama.

Thus the persons are typical, not merely as wearing recognizable costumes and uttering appropriate sentiments, but more intensely as acting in our drama. The thieves transformed into serpents, which in turn absorb them (I. xxv), are thus akin artistically to the sculpture of medieval capitals. They are not pictorial imitation of the other arts, as are the conventional descriptive pauses of the Roman de la Rose and the poetriæ; for they are narrated. There is no ecphrasis.[20] The method is never static. The speaking flames move.

Such narrative progress in detail is integral with the progress of the total symbolism toward culmination. The famous ascent in Paradise moves both in and by itself and with the movement of the whole.

In method, as in degree, this is the individual achievement of a great poet; in kind the symbolism is characteristically medieval. For in detail and in total conception the Divina Commedia starts not with the individual event or person, not with Beatrice Portinari,[21] but with the idea. Whether Dante recalls a youthful love is a question so subordinate as to be artistically immaterial. He has not sublimated earthly passion into heavenly. His conception is heavenly from the beginning, and progressively throughout. He carries a single, controlling idea forward imaginatively, stage by stage. Each stage is vivid with intensely specific realism; but this realism, as that of the carved capitals, is neither the object nor the occasion; it is only the imaginative means to impress the idea. The idea is constant; it is beginning and end. The final vision is not transformation; it is conclusion.

Such grasp of poetic movement doubtless owed much to Vergil. Dante’s homage to the “courteous soul of Mantua” (I. ii. 58) is more than conventional tribute.

The larger narrative movement of the Æneid, which quite escaped the average medieval romancer, was not lost on so great a composer. Of all the ancient art available in his time this alone could give him that instruction. But if the Æneid inspired him with its poetic scope and reach, and gave him a sense of the poetic energy of onwardness, it did not prescribe his method. The Divina Commedia is not an imitation of the Æneid. Its movement is both different and more compelling.

For the movement of the Divina Commedia is at once logical and imaginative, an extraordinary fusion of rhetoric and poetic. The first canto closes with a forecast of the whole progress.

Throughout, the progress is so reasoned that it can be mapped and briefed. None the less the Divina Commedia is a great exemplar of poetic movement. It arrives not at a demonstration, but at a catharsis. Its conception, at once constant and widening, is carried forward imaginatively. We move not from proposition to proposition, but from scene to scene. The Faerie Queene, no less imaginative in detail, has none of this force of imaginative composition. The Christ in the earliest middle age, the Pearl in the latest, have far less scope. The Divina Commedia reveals the whole capacity of medieval symbolism by sustained poetic movement.

So sustained a movement has no time for the conventional descriptive pause. In spite of the abundance of its illustrations, the Divina Commedia is never merely pictorial.[23] Nor is the steady movement monotonous. The transitions are beautifully various. Purgatorio does not repeat the plan of Inferno; and it has more interaction. As the action advances through Purgatorio into Paradiso, there is increasing exposition of ideas. There is even the Gothic variety that enlivens the cathedrals. The grotesque devils of the twenty-second canto of Inferno are brothers to those on the capitals of Vézelay. At the other pole of sentiment, the eleventh canto of Purgatorio opens with a poetic amplification of the Paternoster. The fifth canto of Paradiso is Justinian’s summary of the history of Rome. But the variety is never merely picturesque. Never merely aggregative, the variations of the Divina Commedia are more like those of Bourges than like those of Chartres. They are still more like the variations in a symphony.

The logical order fused with this poetic movement is: Inferno, the punishment of self-enslavement; Purgatorio, the progress of self-mastery; Paradiso, the rapture of self-expression. From perversion and frustration we pass through discipline to the liberation of personality.

It is through this liberation of personality that the human drama becomes divine. “I am come that they might have life, and that they might have it more abundantly.” For the Divina Commedia is far more than a vindication of free will; it is a vision of the progressive freeing of the will. After the terrible revelation of hell as self-enslavement, free will is asserted from purgatory as essential to divine justice.

But this liberty is not innate merely as are the physical impulses. God’s gift of free will is a progressive energizing of man’s struggle to will higher and higher till he attains perfect freedom.

So developed by the mutual response of God and man, free will becomes the greatest gift (“lo maggior don,” III. v. 19). Heaven is the final release and achievement of personality.

The culmination is quite beyond what is usually meant by self-satisfaction. As personality can be progressively released and achieved only through giving, so the self-expressive joy of Paradiso contributes to the common joy. It is simultaneously giving and receiving, utterance and response. No longer, as the poet of Pearl discerned again in the parable of the laborers, is giving and receiving frustrated by competition. Immediately every joy is the joy of all, every expression of personality is a gift and a response, in the creative activity of love.

The amazing figures in which this triumph is symbolized sum up also Dante’s poetic: vividness of charged simplicity in expression carried forward in a composition of progressive movement.

No other single work of medieval art conveys so much of the middle age; for no other concedes so little to medieval convention. The centuries of St. Bernard, Adam of St. Victor, and St. Thomas Aquinas, of Chrétien and the Roman de la Rose, of Vézelay and Bourges, like other centuries, had their artistic conventions. These, as too often in other centuries, seemed to the makers of manuals to comprise the theory of art. Such theory needs both correction and expansion from great artists.

Chaucer, more clearly than most poets, shows the whole artistic progress from expert verse translation through convention accepted and convention modified to creation. From the Roman de la Rose through his study of Boccaccio to the Canterbury Tales his work comprehends in itself much of the history of medieval poetic. He was ahead of his time; that is, he was artist enough to feel new currents of thought and new ventures in expression, to bend toward these the received modes, and finally to enlarge the scope and perfect the technic of medieval narrative. The Renaissance movement, which he was quick to feel in Italy, he could not communicate to England. It had to be brought again, and in different ways. But the art of narrative, most popular of medieval arts, he led from accepted poetic habits in new directions and to new achievements. Consummate metrist, he knew what to do in English with the richer couplet, the more fluent stanza, of Boccaccio. Composition in the larger sense he learned more slowly. The road is long from the leisurely conventions of the Book of the Duchesse to the intensity of the Pardoner and the sustained poetic progress of the Troilus and Criseyde. All the more clearly he illuminates the significance of previous experiments by revealing what was vital in their technic. For since he was a studious artist, and even a critic, as well as a genius, his career epitomizes the progress of medieval verse narrative. Medieval poetic may fairly be said to end with his death in 1400.

The first part of the Roman de la Rose, Chaucer’s literary point of departure, represents the conventions and the mood of the fashionable fiction of courtly love. Its allegory is simple personification carried out by costume and attitude, less by speech and action, appropriate to type. It is less narrative than descriptive. Relying mainly on style, it has so little vigor of composition that in spite of its great medieval vogue it has been long dead. Chaucer’s translation shows no small command of diction and of verse. Though the French octosyllabic and the English line of four stresses differed less then than later, translation had much to teach a bilingual poet not only reading, but constantly hearing and speaking two rhythms. It made him early aware both of the difference between the two and of the directions of development for the English of his choice. He has already found variations to evade the tendency of this short couplet toward jingling monotony; and still more verse control is evident in his catching one of the chief charms of the original, its easy fluency.

His own Book of the Duchesse uses the same magazine of courtly love: rehearsal of a tale of Ovid, May-morning dream-vision, “love’s servant” in “complaint,” praise of the lady, tapestry description, couplet of facile fluency. How much poetry could still be conveyed through these symbols he showed later in the pageant prologue to the Legend of Good Women. Kittredge[25] finds the conventional dream manipulated in the Book of the Duchesse to give some sense of actual dreaming. Every attentive reader has been relieved by the realism of the half-grown, unbroken dog,

The touch of individualizing stands out, indeed, because most of the description is conventionally generalized and decorative; but it may be significant as forecasting Chaucer’s later use of gesture, and as a means of transition. There is an attempt at variety in the punctuation of the long praise of the lady by dialogue; and the closing dialogue is really rapid. An expert metrist in conventional modes, with originality enough for pleasing variations—we can hardly read more of Chaucer into the Book of the Duchesse. The Ovidian thin lucidity of narrative, the usual aggregation, the involution, above all the stalling for description, show that he was not yet giving his mind to narrative composition.

He was spurred to explore narrative by the timely stimulus of Italy. First of English artists to learn there, he was among the most responsive. The Parlement of Foules shows immediate artistic development and promises more. The verse is enriched. The five-stress line, so near the beginning of its long English history, is realized not only for variety, but for flexibility to the mood of speech. Most of its typical variations, shift of cæsura, syncope, doubling of the unstress as in anapest or dactyl, close on stress or on unstress, are here. Though there is not yet that narrative fluency which makes the verse constantly further the story, there is narrative stanza. Stanza, of course, was no innovation. Such set lyric forms as balade and rondel, which were to hold their popularity into the sixteenth century, were exploring the technic of refrain. Linking refrain, used simply in verse narrative,[27] was carried in Pearl to intricate harmonies. In French and in English, stanza was an assured technic. But its narrative development owes much to Chaucer. His adaptation of the Italian stanza in the Parlement of Foules opened the way toward that technical mastery which is one of the characteristic achievements of his narrative.

The Parlement of Foules shows development thus in detail. The total composition, the movement of the whole, is still to seek. Here again are conventional allegory, this time mainly from Alain de Lille,[28] rehearsal from the medieval treasury, dream-vision, decorative description, without any narrative progress. But the allegory is vivified as in Piers Plowman. The classes are differentiated by the lively speech of their representatives. The interaction, though elementary, is more than mere débat. There is even approach to individual character in distinguishing the second tercel from the first. In these modifications there is promise[29] of a still more characteristic achievement, the interludes of the Canterbury Tales.

The narrative stanza of Troilus and Criseyde is a technical triumph. Even the rich harmonies of The Faerie Queene hardly dispute its eminence; for they are adjusted rather to description. Chaucer’s stanza is narrative in sure and fluent onwardness. It so furthers the movement of the story as rarely to invite separate attention, so deftly merges with the other means of narrative suggestion that its values transpire not from quoting this stanza or that, but from reading on and on. No other verse narrative is more satisfying to read aloud. The subtle harmonies of Pearl are adapted to lyric reflection. The easy movement of Childe Harold pauses again and again on picturesqueness. A fairer comparison is Byron’s triumph of fluency, Don Juan. Those stanzas have the same achievement of onwardness where Byron gives himself to the story. Chaucer’s story, deeper, more consistent, more progressive, is always his main concern. His distinction is in making his stanza constantly serve this.

For Troilus and Criseyde carries verse narrative beyond Boccaccio’s scope and beyond his own habit.[30] Perhaps the challenge of the Filostrato was the more provocative because he already suspected the ways of romance. Here was the perennial situation of romance with a new emphasis. Passion, so little realized in French, was presented convincingly; and the lover’s complaint was not the conventional tribute of devotion, but the anguish of disillusionment. Passion, then, left dust and ashes. The French literary lover had been forever wooing a literary goddess; the Italian lover, smitten with the beauty of flesh and blood, won it easily, held it in ecstasy, and could not recover from its loss. As if divining here something truer, which yet was not the whole truth, Chaucer planned one of the few great love-stories. What the romances generally left out of Tristram and Iseult, what only the best of them saw in Lancelot and Guenevere as motive, was the effect of illicit love on character. Passion, stronger than amour courtois, was it stronger than the whole social code? And what then? Chaucer lived to present love in other aspects and in other ways; but first Boccaccio’s story moved him to carry the passion of the noble and beautiful through to its bitter end. He composed not at all a French romance, not the lyric Italian story, but a verse narrative at once so realistic and so dramatic that we naturally call it a novel.

Amour courtois, both conventional and feeble as plot, he relegated to the setting as one of the fictions of high society. Troilus languishes appropriately and writes a complaint. Cressid keeps a proper distance. Pandarus makes allowance for moonshine. What Chrétien had found surely appealing in social romance, the habit of society itself, is appropriated as amply as in Gawain and the Green Knight. Medieval rendering of all stories in contemporary terms is here carried into realism. Pandarus, visiting Criseyde,

The dialogue is more than realistic; it gives leads for the story.

Description, too, is made to serve the narrative economy. The garden setting (II. 813) combines with the song of Antigone to turn Criseyde’s mood, which is then revealed in their brief dialogue. Setting in Gawain and the Green Knight, equally distinct as social background, oftener picturesque as scenery, is less forwarding, less woven into the texture of the story. The clue to this advance in poetic is in Chaucer’s accompaniment of significant speech by significant gesture. He often indicates the stage “business.”

Finally dialogue is used to bring a scene to an issue. Stage business, scene, issue, the terms are not too dramatic for the method of the passage between Pandarus and Criseyde over the first love-letter.

The scene becomes a situation. The eleven lines are dramatic not only in presenting the situation as spoken and acted, but in developing by interaction both character and plot. Chaucer has found how to tell the whole story of illicit love progressively by revealing in speech and action not merely type, not merely situation, which had rarely been so used to its full significance, but character moving to its issue. In this larger movement, sustained and advanced by progressive characterization, Troilus and Criseyde is the great medieval love story.

The achievement of creation becomes the more conspicuous on review of what the story had been.[31] Mere unrelated episodes of the older sources, mere scattered incidents in Benoît’s huge Roman de Troie, had kindled Boccaccio’s Filostrato. Of its 5512 lines Chaucer used 2730, or about half, but carried his own story to 8239. His addition, most of the first half of his story, is devoted to the gradual yielding of Cressid. Boccaccio has no such approach because he has no such person. His story is of Troilus; and his Cressid, remaining the typical fair inconstant that he found in Benoît, falls as readily into the arms of her first lover as into those of her second. Chaucer’s Cressid is the first great character of English fiction. The characterization, delicate enough to keep her graciousness even in ruin, and to give the unashamed materialism of Pandarus engaging frankness and humor, is dynamic. It gives the story motive. Cressid is always dominant, not in having more space, but in bringing about before our eyes alike the heartbreak of Troilus and her own degradation. This is Chaucer’s achievement of narrative progress. Cressid, we say to ourselves, would yield to Diomed. But though dominated by Cressid, the movement is of all three characters interacting, of characterization advancing by stages.[32] Thus Chaucer’s refocusing is neither modification of Boccaccio’s story nor such variation as had been already achieved; it is progressive motivation toward a significant issue.

The dramatic interaction effective in Troilus and Criseyde is less continuous, though no less striking, in the Canterbury interludes. It is not drama, it remains subordinate to narrative, partly because the time for drama was not ripe, more because Chaucer knew verse narrative for his own art. The bent of his genius he followed in both experiment and study. One of his first technical achievements, the weaving of description into the action, appears not only in progressive mastery, but also as distinct theory. Here evidently he discerned as a critic the deviation and the insufficiency of the poetic that he had learned at school.[33]

Let me tell an old British tale in my own plain way; for I am unversed in ornaments of style. This is all that the Franklin’s prologue means on its surface. The connotation beneath is inviting. Are “aventures,” “layes,” “rymeyed,” “instruments” intended precisely? How much grasp they suggest of early medieval poetic is perhaps beyond our determination.[34] But the term “colours of rethoryk” occurs also in the Hous of Fame. The interlude before the Clerk’s Tale has a sarcasm of the Host against these same “colours” in the same connection. The Squire’s disclaimer has the same significant terms as the Franklin’s, and the same point as the Clerk’s reply to the Host. The satire in the Tale of the Nun’s Priest on Geoffroi de Vinsauf confirms the suspicion that in all these passages Chaucer implies specific criticism of the poetriæ.

For the language of the Franklin’s prologue, in spite of his disclaimer, is literary. Chaucer knew as well as Shakspere that he who announces “a plain, unvarnished tale” may command a better art than the rhetoric that he disclaims. It is worth while to explore, therefore, the mention of rhetoric in connection with story-telling, the conjunction of Cicero and Parnassus. As Sir Thopas parodies not only the conventional motives of romance, but also particular faults in its conventional technic, so Chaucer’s references to “colours of rethoryk,” instead of being taken as general disparagement of grandiloquence, may well be sounded for their particular significance. In any age, indeed, the man of letters contemplating the rules of his art laid down by pedagogues is moved to sarcasm; but Chaucer’s sarcasms may suggest specifically wherein the pedagogues that he knew went wide of the narrative art that he came to comprehend as artist and as critic. His reference in the Hous of Fame merely glances at “prolixitee.” The passages in the four Canterbury tales, ampler and more specific, together suggest that the application of “colours of rethoryk” to narrative is a perversion, that Cicero is out of place on Parnassus.

The notion that the citing of “Pernaso” and “Cithero” in the same breath is meant to exhibit the Franklin as “a burel man” is dispelled by literary history. Cicero as a master of style had long been invoked to teach poetry. Poetria, as conceived by the medieval manuals, is essentially elaboration of style.[35] That it is a distinct mode of composition is never even hinted. Focused on diction and devoted to elaboration, it draws upon the ancient colores until poetic is indistinguishable from rhetoric. Either, then, Chaucer is merely accepting this merger, without particular intention in his “Pernaso” and “Cithero,” or he is hinting that the very conception hinders straightforward narrative.

In the fourteenth and fifteenth centuries the English word rhetoric denoted style generally, whether in prose or in verse. So rhetor or rhetorien meant master of style, and was freely applied to poets. The familiar reference to Petrarch, therefore, is usually taken as general praise.

Nor should we pause over the conjunction of “rethoryke” and “poetrye,” were it not identical with that of “Cithero” and “Pernaso.” Even if this pairing, like that, merely reflects medieval habit without reflecting on it, the Clerk’s prologue as a whole is sufficiently definite. Repeating the Host’s term “heigh style,” it goes on to consider Petrarch in this aspect. The Host had been quite precise.

Whether or not “heigh style” is a misappropriation of a phrase of Petrarch’s, its implication here is both definite and significant. “As whan that men to kinges write” makes “endyte” refer unmistakably to dictamen.[36] The Host deprecates the perversion of this to the telling of a tale. His sarcasm is directed not vaguely at grandiloquence, but specifically at the application of dictamen to story-telling. To measure its significance, one must read the doctrine pervading poetria and dictamen alike. That doctrine is rhetoric; it is nothing more and nothing else.

The Clerk, taking up “heigh style,” admits that it is a hindrance in Petrarch’s descriptive opening.

What the Clerk is made to challenge here in the application of “heigh style” to a tale is the separable description. No device for dilation was more magnified in the poetriæ. More than apostrophe, contrast, and other “colours,” it was inculcated to give poetry what Chaucer’s eagle calls

Chaucer’s bête noire of “prolixitee” among these pedagogues was Geoffroi de Vinsauf; and that brings us to the familiar sarcasms of the Nun’s Priest.[38] They pursue not merely an ass, but still more a perverted poetic. Chaucer pillories Geoffrey not merely because the Nova poetria is an Ovidian nightmare, but because its constant object, alike in the rules and in the manufactured examples, is to inculcate the stalling of composition by “colours.”[39]

The separable description challenged by the Clerk is challenged also by the Squire, and in terms that confirm the significance of the Franklin’s.

The Squire’s later sarcasm is more open and more constructive. It may well sum up Chaucer’s criticism of the deviation of poetic by rhetoric.

A rhetoricated poetic, though in other mouths than Geoffrey’s it has often had a fairer sound, is always a perversion. The confusion of poetic with rhetoric has always tended to obscure the imaginative value to narrative of onward movement. The medieval pedagogues who reduced the greatness of Vergil to mastery of all “three styles” of rhetoric would doubtless have recommended embellishing the eloquences of Aucassin et Nicolete at the expense of the story, dilating the Tombeor de Notre Dame, and making the Chastelaine de Vergi static. Their conception of narrative has no room for such composition as makes the Pardoner’s Tale a fatal sequence.

Even Chaucer is too responsive to the taste of his time to abandon the accepted means of dilation. Some of his own apostrophes differ from Geoffrey’s to Friday[40] only in eloquence, not in method. But as early as the House of Fame he was critically aware of the “prolixitee” inherent in “colours of rethoryk.” The interpolated description, which he had elaborated to make the Knight’s Tale magnificent,[41] he parodied in Sir Thopas and challenged through the Clerk and the Squire. In these passages he reminds us that his achievement of narrative composition had taught him to distrust rhetoric as a means of enhancing when the tale, Pardoner’s, Squire’s, or Franklin’s, was really the thing. Further he implies general denunciation of the staleness and ineptitude of medieval rhetoric as poetic method.

Within the Canterbury frame narrative composition is widely various. The early tale of the Man of Law follows the pattern of a saint’s legend; the Clerk’s is rather iterative than progressive. In such cases Chaucer has not recomposed; he has limited his art to style. What distinction he could achieve even so is most conspicuous in the exquisite tale of the Prioress. Without crisis, without even salience of moment, with no more reshaping than the shift of focus from the horror of ancient superstition to the beauty of childlike devotion, he wrought by adjustment of tone and cadence a marvel of simplicity. The Merchant infuses into his fableau realistic suggestions of senile sexuality; but the fableaux generally, Miller’s, Reeve’s, etc., are told in their typical form for their typical values. Beast epic in the mouth of the Nun’s Priest, never livelier and seldom so rich in suggestion, is little reshaped as a story. Even the Wife of Bath tells her fairy mistress tale without originality of composition.

The artistic interest of the narrative couplet in the Knight’s Tale has been heightened by Dryden. Beneath the superficial differences which mark each verse with its time is a difference of conception. Chaucer’s time gave him easier variations; but he himself bent them narratively. The larger movement of the Knight’s Tale conveys the magnificence of princely chivalry less in action than in pageantry. Romance as story was not for Chaucer, one might say but for the tale of the Franklin. Here he shapes a plot complementary to that of Troilus and Criseyde in briefer compass, but with equal onwardness. The issue again is convincing because it is reached by distinct stages. The fourth, for instance, culminates upon the squire’s desperate triumph.

And the next begins:

The crisis upon Dorigen’s taking all to her husband, thus fulfilling her character while she flies in the face of the code of secrecy, is rendered in dialogue of sharp interaction.

No less eloquent is her reply to the squire at their meeting.

Such passages, as in Troilus and Criseyde, are narrative leads. The last leads directly to the squire’s fine renunciation; for the tale of generosity has been embodied in individuals working out its issue. Even the illusion of the rocks is lifted above magic because they were first an obsession of the anxious wife. An exemplum and a tale of marvel have been reconceived and recomposed in a moving story.

No triumph of Chaucer’s is more evident than the art of narrative swiftness in the tale of the Pardoner. Verse narrative has nothing more seizing, more breathless, more fatal. Yet even this is no greater than the art that reveals the Pardoner himself. The tale is no more triumphant than the interlude and the prologue. Rather they belong together. So the ordinary romance told by the Wife of Bath is heard amid the echoes of her brutal realism. The familiar story of the cock and the fox takes both color and shape from the Nun’s Priest. Readers opening their Canterbury Tales for the second time, or the third or the tenth, are as likely to turn to an interlude as to a tale. Even an adequate account of the poetic of every tale would fall short of the total artistic value. For the interludes play a part in a scheme more ambitious than any other medieval “framework”. Gower’s plan in the Confessio Amantis is no more than a classified series. The preface to the Decameron proposes “a hundred tales ... told in ten days by a noble company of seven ladies and three youths in the time of the late pestilence ... in the which tales appear pleasant or rude chances of love and other incidents of fortune happening as well in modern times as in ancient.” After describing the plague in Florence, the “noble company”, and the fair country house to which they withdrew for safety, Boccaccio makes each of his ten persons tell a tale each day on the same general theme. Thus he arranges ten groups of ten tales each, with charming interludes of conversation, song, and description. But only the charm of style saves the connective from monotony. The narrators are merely mouthpieces; the interludes are not used, as by Chaucer, to bring about contrast and interchange; the setting, though more attractive than the allegorical fiction of the Confessio Amantis, is merely repeated with variations. The ancient plan of The Seven Sages remains inflexible through all the versions. The king, stayed by the tale of the first sage, finds the queen in tears, and is won back by her counter-tale to reaffirm his sentence on the prince. With mere variation of the dialogue this scene is repeated six times. Instead of being always different, as in the Canterbury Tales, the interlude is always substantially the same. Not only are the tales generally alike in form, being all by the necessity of the plan exempla, but the plan itself is little more than a vehicle.

The scheme of the Canterbury Tales is at once more flexible and larger in scope. The fiction of a traveling company offers more opportunity than that of a confession, a trial, or even a house-party. But further the interludes, instead of being pauses, whether pleasant as Boccaccio’s or tedious as Gower’s, act upon the tales. They add an individualized teller in action and interaction. The parody romance of Sir Thopas gains in point by the rude interruption of the host. Revenge for this may be meant as excuse for the dulness of the following prose morality. At any rate, the host’s rueful comparison between the wife of Melibeus and his own, and the domestic comedy of Chantecler and Pertelote, suggest cues for the marvelous prologue of the Wife of Bath; and her tale opens the way for other maistrye of women in marriage.[42] Though the grouping remains conjectural because the scheme remained incomplete and the manuscripts do not agree in the order, there is no doubt that Chaucer projected a larger technic of dramatic setting. The general prologue describes each teller by summary indications of make-up, costume, and personal style, not in order to review medieval social classes, but to prepare for the various interaction of the interludes. Far from exemplifying Chaucer’s descriptive habit, it is specifically adjusted to a list of dramatis personæ.

Medieval verse narrative was recited or read aloud. Chaucer had learned early that setting need not interrupt its oral course. Making description run instead in that course, he had learned further that inanimate background is both less tractable orally and less significant narratively than the environment of men and women. Dialogue, the most oral means of liveliness, he pointed by gesture, and from mood and emotion advanced it to interaction. Finally he staged his tales by characterizing the tellers and suggesting their interplay with an audience. Though this is not often carried out so dramatically as with the Pardoner, it is evident as a narrative scheme. For the “framework” of the Canterbury Tales is human.