SOURCE: "Chaucer and the Rhetoricians," in Proceedings of the British Academy, Oxford University Press, 1926, pp. 95-113.

[Manly was an esteemed professor of Medieval English known for his valuable contribution to Chaucer studies through his lectures and his eight-volume collection. The Text of the Canterbury Tales, Studied on the Basis of All Known Manuscripts. In the following excerpt from his published lectures, Manly describes the rhetorical styles of Chaucer's Canterbury Tales, Book of the Duchess, Parlement of Foules, and other poems. He traces Chaucer's style to the lessons given in medieval rhetorical texts, suggesting that Chaucer was following set conventions in his poetry, which he later imaginatively expanded.]

… In investigating the sources of Chaucer's notions of literature and his conceptions of style, scholars have hitherto discussed only the writings of other authors which may have served as models for imitation. The possibility of his acquaintance with formal rhetorical theory and the precepts of rhetoricians has not been considered, not-withstanding the hint that might have been derived from the allusion to Gaufred de Vinsauf and the other passages on rhetoric scattered through his works. Even a priori there would seem to be a high probability that Chaucer was familiar with the rhetorical theories of his time, that he had studied the text-books and carefully weighed the doctrines. Whatever modern scholars may have said of the errors in his references and the shallowness of his classical learning—and there are few of his critics whose errors are less numerous than his—he was a man of scholarly tastes and of considerable erudition. His works bear witness to no small reading in astronomy and astrology, in alchemy, in medicine, and in philosophy and theology, as well as in the classical authors current in his day. The ancient tradition that he was educated, in part at any rate, in the law school of the Inner Temple has recently been shown to be possible, if not highly probable. The education given by the inns of court seems to have been remarkably liberal. What more likely than that the formal study of rhetoric not only was included in his academic curriculum, as one of the Seven Arts, but also occupied much of his thought and reflection in maturer years?

What, then, was medieval rhetoric? Who were its principal authorities in Chaucer's time? And what use did Chaucer make of methods and doctrines unmistakably due to the rhetoricians?

To the first two questions satisfactory answers can be readily given. Professor Edmond Faral has recently printed the chief rhetorical texts of the thirteenth and fourteenth centuries, with illuminating biographical and bibliographical notes and excellent summaries of the doctrines. To answer the third question fully would require a volume, but a provisional view of the matter can be obtained from a rapid survey of Chaucer's best-known work.

Fortunately for our inquiry, the Middle Ages knew only one rhetorical system and drew its precepts from few and well-known sources. Moreover, there was little development of the doctrines or variety in the mode of presentation. The principal sources of the doctrines were three: the two books of Cicero entitled De Inventione, the four books entitled De Rhetorica, ad Herennium, and the Epistle of Horace to Piso. Treatises based upon these were not uncommon in the earlier Middle Ages, but after the beginning of the thirteenth century the practical spirit of the time tended in the universities to substitute instruction in letter writing and the artes dictaminis for the more theoretical and supposedly less useful study of general rhetorical principles. It is perhaps for this reason that the treatises of Matthieu de Vendôme and Gaufred de Vinsauf, written early in the thirteenth century, retained their vogue in the time of Chaucer. These treatises are the Ars Versificatoria of Matthieu, and the Documentum de Arte Versificandi and the Nova Poetria of Gaufred. The first two are prose treatises, carefully defining and discussing all processes and terms and illustrating them by examples, in part drawn from earlier writers, such as Virgil, Horace, Ovid, Statius, and Sidonius, and in part composed by the rhetorician himself, either to show his skill or to pay off a grudge. For example, Matthieu is tireless in the composition of verses attacking the red-haired rival whom he calls Rufus; Gaufred, illustrating the beauties of circumlocutio, says it is of special value when we wish to praise or [defame] a person ….

The doctrine taught by these two authorities, the common medieval doctrine, falls logically and naturally into three main divisions or heads: (1) arrangement or organization; (2) amplification and abbreviation; (3) style and its ornaments.

Of arrangement they had little to say, and that little was purely formal and of small value. They treated mainly of methods of beginning and ending, distinguishing certain forms as natural and others as artificial. Artificial beginnings consisted either of those which plunge in medias res or set forth a final situation before narrating the events that led up to and produced it, or of those in which a senteniia (that is, a generalization or a proverb) is elaborated as an introduction, or an exemplum (that is, a similar case) is briefly handled for the same purpose. It will be readily recognized that all these varieties of beginnings are in familiar use at the present day; and, curiously enough, in recent years writers for the popular magazines have shown a special fondness for beginning with an elaborately developed sententia.

We have not time to-day for a detailed examination of Chaucer's methods of beginning, but this is hardly necessary. The moment one undertakes a survey of his poetry in the light of rhetorical theory, one is struck by the elaborate artifice of its beginnings and the closeness of their agreement with rhetorical formulae. This artificiality has long been recognized but has been mistakenly ascribed to the influence of the poems upon which he drew for his materials. His French sources, however, are hardly responsible for these elaborate beginnings; they furnish only the raw materials which Chaucer puts together in accordance with the instructions of his masters in rhetoric. The apparent simplicity with which the Boke of the Duchesse begins disappears under examination: the reader is led through several long and tortuous corridors—totalling one-third of the poem—before he arrives at the real subject, which in turn is developed with amazing artificiality. The long failure of the mourning knight to make clear the nature of his loss may be regarded as an expanded form of the rhetorical figure called occupatio.

The Parlement of Foules admirably illustrates the method of beginning with a sententia:

The lyf so short, the craft so long to lerne.

This is expanded into two seven-line stanzas. Then comes, not the narrative itself, but a preliminary narrative, interspersed with various rhetorical devices, including generalizations, an apostrophe, and an outline of Cicero's Somnium Scipionis, in all 119 lines, before the story proper begins.

This method is even more elaborately developed in the Hous of Fame. In fact the poet is within twenty lines of the end of Book I before he begins to tell his story. There are sixty-five lines on dreams, sixty-five more of invocation, and more than 350 telling in outline the entirely unnecessary story of Dido and Aeneas.

Even when the narrative begins in a natural manner, as in Anelida and Arcite, the poem is given an artificial character by prefixing an invocation or by some other rhetorical device. The beginning of the Legend of Goode Women combines the methods of sententia and exemplum: our belief in the joys and pains of heaven and hell, says the poet, is based, not upon experience, but upon the acceptance of the sayings of 'these olde wise'; in like manner we must accept the testimony of books—those treasuries of wisdom—about the existence of good women, though we have never known them. A few of the separate legends begin inartificially, but it was not until late in his career that Chaucer developed the method of beginning used with such masterly skill in the tales of Miller, Reeve, Summoner, and Pardoner.

Methods of ending are treated by the rhetoricians even more summarily than beginnings, the preferred forms being the employment of a proverb or general idea, an exemplum, or a brief summary. Chaucer is fond of some sort of explicit application of his stories. In the "Reeve's Tale" this takes the form of a proverb:

And therfore this proverbe is seyd ful sooth
Him thar nat wene wel that yvele dooth:
"A gylour shal hymself bigyled be."

And the "Manciple's Tale" ends in a stream of proverbs and proverbial sayings. But the more common form of application is a generalization or an exclamatory comment. Very common also is the ending summarizing the situation at the end of the tale. On the other hand, not-withstanding Chaucer's fondness for exempla, the exem-plum-ending is very rare; perhaps the only instance, and that a doubtful one, is in the "Friar's Tale":

Herketh this word, beth war, as in this cas:
"The leoun sit in his awayt alway
To sle the innocent, if that he may"

Peculiar to Chaucer are the references to other writers for further information—as in several of the legends—and the triple demande d'amours with which the "Franklin's Tale" ends.

The technical means of passing from the beginning to the body of the work—prosecutio, as it is called—are treated with much formality by Gaufred, though he remarks with great good sense that the prime requisite is to get on with the subject: In ipsa continuatione, primum est continuare.

In Chaucer, after a rhetorical beginning, the transition to the narrative itself is usually clearly and formally indicated; so, for example, in Troilus and Criseyde:

For now wol I gon streight to my matere.

The amount of attention devoted by the rhetoricians to the second main division, that of amplification, is to the modern reader surprising, but it results quite naturally from the purely mechanical character of the art of rhetoric as conceived by them. To them the problems of composition were not problems of the creative imagination but problems of 'fine writing'—l'art de bien dire. They had no conception of psychological processes or laws. The questions they raised were not questions of methods by which the writer might most perfectly develop his conception or of the means by which he might convey it to his audience. The elaborate system of technical devices was discussed only with reference to the form and structure of each device, never with reference to its emotional or aesthetic effects. As the rhetoricians conceived the matter, if a writer had something new to say, rhetoric was unnecessary; the novelty of the material relieved him of any concern for its form. But alas! this situation seldom arose. Practically everything had already been said. All the tales had been told, all the songs had been sung, all the thoughts of the mind and feelings of the heart had been expressed. The modern writer, they held, could only tell a thrice-told tale, only echo familiar sentiments. His whole task was one of finding means and methods of making the old seem new. He might therefore well begin his task of composition by choosing some familiar but attractive text—some tale, or poem, or oration, or treatise—or by making a patchwork of pieces selected from many sources. His problem would be that of renewing the expression and especially of making it more beautiful—ornatior is the common term.

Let no one scoff at this method as incapable of producing interesting and attractive writing. It has been practised very commonly by writers in all lands and epochs. It is recommended and taught in a widely used series of French text-books. It is the method recently revealed as pursued by that most charming of stylists, Anatole France, and is perhaps the only method by which he or Laurence Sterne could have produced such effects as they achieved.

Medieval rhetoricians assume that the writer, having chosen his subject, will find his material either too great or too small for his purpose. His problem will almost necessarily be one of amplification or abbreviation. The methods of amplifying and abbreviating are derived from the technique of style. They are therefore dealt with in their proper places when style and its ornaments are under discussion, but for the sake of clearness they are also expounded elaborately with special reference to their uses and values as means of amplification and abbreviation.

The principal means of amplification are six—some writers say eight:

Description, though perhaps not the most important, may be named first, as receiving fullest attention from both Matthieu de Vendôme and Gaufred de Vinsauf. Elaborate patterns and formulas are given for describing persons, places, things, and seasons. If the description applies to externals, the features to be described are enumerated and the order in which they are to be taken up is strictly specified; if it concerns a character, the characteristics to be mentioned are listed, and those appropriate to each sex, age, social status, employment, temperament, and career are set forth in detail. Specimens are given to illustrate the doctrines. These descriptions are not, like those in Chaucer's later work, determined by the requirements of the situation in which they occur. Their use is purely conventional, for the purpose of amplifying the material, and their construction is purely mechanical. They are merely opportunities for the writer to display his rhetorical training. It is very enlightening to compare Chaucer's later descriptions—such, for example, as those of Alysoun and Absalon in the "Miller's Tale"—with the early ones; for example, with that of the Duchess Blanche, which, with the exception of one or two possibly realistic touches, is nothing more than a free paraphrase of lines 563-597 of the Nova Poetria, composed by Gaufred de Vinsauf as a model for the description of a beautiful woman. The features described in the two passages are the same, they are taken up in the same order, and the same praise is given to each. The resemblance is still further heightened by the fact that, like Chaucer, Gaufred declines to guess at the beauties hidden by the robe—a trait hitherto regarded as characteristically Chaucerian.

There seems little doubt, indeed, that Chaucer's character sketches, widely as they later depart fromthe models offered by the rhetoricians, had their origins in them. An American scholar has recently attempted to show that Chaucer derived them from the treatises on Vices and Virtues, with their descriptions of character types. The possibility of an influence from this source I will neither deny nor discuss, but the specimen sketches given by the rhetoricians seem entirely sufficient to account for Chaucer's interest in this type of description.

The next most important device was digression, of which two subdivisions were recognized: first, digression to another part of the same subject, anticipating a scene or an event which in regular course would come later; second, digression to another subject. Digression may obviously be made in many ways and may include many special rhetorical devices. Prominent among the special forms are the development of a sententia and the introduction of exempla, illustrating the matter in hand. These two devices are of (he utmost importance for Chaucer in particular and for the Middle Ages in general. The temper of the Middle Ages being distinctly practical and its literary valuations being determined, not by the criteria of art, but by those of edification, sententiae, proverbs, and exempla were used with an ardour now difficult to appreciate. The use of exempla was strongly inculcated by the rhetoricians. Matthieu de Vendôme urges the writer to provide an abundance of exempla….

But the precepts of the rhetoricians on this point had already been heeded by other writers, and in Chaucer's poems it is difficult to separate the direct influence of rhetorical theory from that of the practice of Guillaume de Machaut, whose first use of exempla was in his Dit de l'Alerion and whose later use of them gave them a vogue attested by the imitation of all hid by this astonishing fad as was any of the French imitators of Machaut. They are familiar from the series of twenty-one consecutive instances in the "Franklin's Tale" and the humorous accumulation of them in the controversy between the Cock and the Hen.

Third in importance among the devices of amplification may be placed apostrophe, with its rhetorical colours exclamatio, conduplicatio, subiectio, and dubitatic. It would be difficult to exaggerate the importance of apos trophe in medieval literature. Addresses to persons living or dead, present or absent, to personified abstractions, and even to inanimate objects are to be found in almost every composition with any pretensions to style from the eleventh century onward; and a special form, the Complainte, developed into one of the most widely cultivated types of literature. Chaucer's use of apostrophe is so frequent that no examples need be cited. Almost every tale contains from one to a dozen examples of it. Among the colours, his favourites seem to be those known as exclamatio—simply a passionate outcry addressed to some person or thing present or absent—and dubitatio, that is, a feigned hesitation what to say, a rhetorical questioning as to which of two or more expressions is appropriate to the idea and situation. Like Wordsworth's—

Fourth in order may come prosopopeia or effictio, the device which represents as speaking persons absent or dead, animals, abstractions, or inanimate objects. Widely used for purposes of amplification, this figure often furnished forth the whole of a piece of literature. Examples are numerous. A charming one contemporary with Chaucer is the débat in which Froissart represents his dog and horse as discussing their master and the journeys which he compels them to make with him. Chaucer uses it briefly many times, and elaborately in the principal scene of the Parlement of Foules.

Less important than the foregoing are the devices of periphrasis or circumlocutio, and its closely related expolitio. Circumlocutio was highly regarded as one of the best means, both of amplifying discourse and of raising commonplace or low ideas to a high stylistic level. It is too familiar to require discussion, but Master Gaufred seems not to have distinguished clearly between a statement expanded for the mere sake of amplification and one which expresses some important detail or phase of an idea. For example, he calls the opening lines of Virgil's Aeneid circumlocutio and declares, 'This is nothing else than to say, I will describe Aeneas.' And, after quoting from Boethius three lines of the metre beginning,

O qui perpetua mundum ratione gubernas,

adds,

Quod nihil aliud est quam, 'O Deus.'

These remarks and the similar ones by Matthieu de Vendôme will doubtless recall Chaucer's sly comment in the "Franklin's Tale" on his own rhetorical description of the end of the day:

Til that the brighte sonne lost his hewe,
For thorizonte hath reft the sonne his lyght,—
This is as much to seye as it was nyght.

The colour expolitio includes the repetition of the same idea in different words (one form of interpretatio) and also the elaboration of an idea by adding the reasons or authorities, pronouncing a generalization with or without reasons, discussing the contrary, introducing a similitude or an exemplum, and drawing a conclusion. Although these two figures are of minor importance, they nevertheless play a considerable part in the writings of Chaucer, as of most other medieval authors.

Other devices for amplification existed, but I will spare you even the enumeration of them.

Abbreviation is joined by the rhetoricians with amplification, but is obviously of much less practical interest. The medieval writer is, as a rule, not so much concerned to abbreviate as to amplify. Master Gaufred, however, instructs his readers that in treating a well-worn subject the best means of creating an appearance of novelty is to survey the whole subject and then run quickly over the parts that predecessors have dwelt upon and dwell upon parts they have neglected. The principal means of abbreviation recommended are certain of the figures of words: asyndeton, reduction of predication, and the like. Chaucer's favourite methods are two:

(1) The use of absolute constructions—perhaps the most striking and beautiful example of this is the opening line of the second book of the Troilus:

Out of these blake wawes for to saile,
O wind, o wind, the weder ginneth clere!

the second line furnishing an instance of the figure called epizeusis.

(2) The figure called occupatio, that is, the refusal to describe or narrate—a figure used with special frequency in "The Squire's Tale," as for example:

But for to telle yow al hir beaute
It lyth nat in my tonge, nyn my konnyng

and

I wol not tellen of hir straunge sewes

or

I wol nat taryen yow, for it is pryme

or

Who koude tellen yow the forme of daunces
So unkouthe, and so fresshe countenaunces?

…..

No man but Launcelot, and he is deed.

Into the vast and tangled jungle of the medieval treatment of Style and its Ornaments we cannot venture now. Its extent may be inferred from the fact that, notwithstanding the inclusion of very long specimens of apostrophe, prosopopeia, and description (328 lines in all) the portion of the Nova Poetria devoted to the important subjects of 'Art in General', 'Organization', and 'Amplification and Abbreviation' occupies only 674 lines, whereas that devoted to the Ornaments of Style' occupies 1125. The tangle is suggested by the fact that there are recognized, defined, and discussed thirty-five colours, or figures of words, twenty figures of thought, and ten varieties of tropes, with nine more sub-varieties. These figures fall into two very distinct classes: first, those in which human emotion and aesthetic feeling have always found utterance—metaphor, simile, exclamation, rhetorical question, and the like; and second, a vast mass of highly artificial and ingenious patterns of word and thought, such as using the same word at the end of a line as at the beginning, heaped-up rhymes, and alliteration.

Like other writers in all ages, Chaucer makes extensive use of the first class of figures; of the artificial patterns he makes only a limited use, and that solely in highly rhetorical passages, like the "Monk's Tale," certain parts of the Boke of the Duchesse, and in the apostrophes, exclamations, and sententiae of other serious compositions. The humorous tales, for which the rhetoricians forbidthe use of colores, are entirely free from special rhetorical devices, with the single and striking exception of the "Nun's Priest's Tale," a mock-heroic composition so full of rhetoric and so amusingly parodying the style of the "Monk's Tale," which immediately preceded it, as to invite the suggestion that the 'high style' and its parody were purposely juxtaposed. Is it possible that Chaucer's desire to carry out this amusing contrast explains the otherwise puzzling change of the Monk from the spectacular huntsman and hard rider of the "Prologue" to the bookish pedant of the hundred lamentable tragedies who greets our astonished ears when he is called upon for a tale?

As no one ever pays any attention to statistics and percentages, they rest the mind. This may therefore be a fitting time to introduce a few. If we list the Canterbury Tales according to the percentages of the larger rhetorical devices which they contain, they form an interesting descending series, ranging from nearly 100 per cent, to 0. Highest, as might be expected, stands the "Monk's Tale," with nearly 100 per cent, of rhetoric. Next comes the "Manciple's Tale" with 61 per cent.; then the tales of the "Nun's Priest" and the "Wife of Bath" with 50 per cent. The tales of the "Pardoner" and the "Knight" have 40 and 35 per cent, respectively; while those of the "Man of Law," the "Doctor," the "Prioress," the "Franklin," the "Second Nun," and the "Merchant" fall between 30 and 20 per cent. The half-told tale of the "Squire" stands alone with 16 per cent., and slightly below it come the tales of the "Clerk" and the "Canon's Yeoman," with 10 per cent. Quite in a class by themselves stand the tales of the "Reeve" and the "Shipman," with about 5 per cent, of rhetoric, and those of the "Miller," the "Friar," and the "Summoner," in which the rhetorical devices do not occupy more than 1 per cent, of the text.

Although some of these percentages are just what we should expect from the character of the tales and their probable dates, some are rather surprising. It is natural enough that the "Monk's Tale" should head the list, for it is professedly a collection of tragedies. But that some of Chaucer's freest and most delightful work should contain twice as much rhetoric as some of his least inspired compositions is a puzzle that demands investigation.

Let us begin by examining one of the least known and least interesting of the tales, that of the "Manciple." It is in fact so insignificant and so little read that I cannot even assume that all of you recall the plot. 'When Phebus lived here on earth, we are told, he had a fair young wife, whom he loved dearly, and a white Crow, whom he had taught to speak. But the wife was unfaithful and took a lover. This was observed by the Crow, who upon Phebus's return home told him. Phebus in sorrow and anger slew his wife, and then, repenting of his deed and disbelieving the charge brought against her, plucked the white feathers from the bird and doomed all crows to be black.'

We may note in the first place that the tale is not particularly appropriate to the Manciple or indeed to any other of the pilgrims, and that no effort is made to adapt it to him. It consists of 258 lines, of which 41 are devoted to describing Phebus, his wife, and the crow, and 50 to telling the incidents of the story. The remaining 167 lines—61 per cent, of the tale—are patches of rhetoric. Even this high percentage is perhaps too low, for the 25 lines of description devoted to Phebus are so conventional, so much in accordance with rhetorical formulas, that they might fairly be added to our estimate of the percentage of rhetoric. No effort was made by the author to conceive any of his characters as living beings or to visualize the action of the tale. The action, to be sure, seems in itself unpromising as the basis of a masterpiece of the story-teller's art, but so, if we consider them closely, are the basic narratives of the "Nun's Priest's Tale" and the tales of the "Miller," the "Reeve," and the "Friar." If Chaucer had been as well inspired when he wrote this tale as when he wrote his masterpieces, Phebus might have been as real to us as the Oxford Carpenter or the Miller of Trumpington, his wife as brilliant a bit of colour as the Carpenter's wife, and the Crow as interesting a bird as Chaunticleer or Pertelote. But he developed the tale, not imaginatively, but rhetorically. Instead of attempting to realize his characters psychologically and conceive their actions and words as elements of a dramatic situation, he padded the tale with rhetoric. Thus he thrust into it and around it 32 lines of sententiae, 36 of exempla, 18 of exclamatio, 14 of sermocinatio, 3 of technical transition, 17 of demonstratio, and 63 of applicatio—all external and mechanical additions, clever enough as mere writing, but entirely devoid of life. If the tale had been written as a school exercise, to illustrate the manner in which rhetorical padding could be introduced into a narrative framework, the process of composition could not have been more mechanical or the results more distressing.

But Chaucer was endowed with the temperament, not of the rhetorician, but of the artist; and in some way he arrived at the memorable discovery that the task of the artist is not to pad his tales with rhetoric, but to conceive all the events and characters in the forms and activities of life. For this he was well prepared by native endowment and by a habit of close observation which developed early and which redeems even his earliest poems from entire banality. Owing to the loss of so much of his prentice work and the uncertain chronology of what has been preserved, we cannot trace in detail the displacement of the older rhetorical by the new psychological methods. But certain lines in the Hous of Fame indicate that when he was writing that poem he at least had formed an idea of the new methods, even though he may long have continued in some respects under the dominance of the old. The lines in question are in the proems of the second and third books:

O thought that wroot al that I mette,
And in the tresorie it shette
Of my brayn, now shal men se
If any vertu in thee be;

and more specifically:

And if, Divyne Vertu, thou
Wilt helpe me to shewe now
That in myn hede y-marked is.

These passages, although the first is translated from Dante, seem to me to express Chaucer's growing conviction that narration and description, instead of being mere exercises in clever phrasing, depend upon the use of the visualizing imagination.

But in spite of this recognition of the true method, and in spite of his ability later in the "Nun's Priest's Tale" to parody the whole apparatus of medieval rhetoric, Chaucer did not free himself at once—and perhaps never entirely—of the idea that writing which pretended to seriousness and elevated thought was improved by the presence of apostrophes and sententiae and exempla, as he had been taught by the rhetoricians. Nor could it be expected that he should. The whole weight of the medieval conception of literature was against him—the conception, I mean, that literature, like history, is of value only in so far as it can be profitably applied to the conduct of human life, a conception which not only remained in full vigour through the Middle Ages and the period we are accustomed to call the Renaissance, but even now lies at the basis of much critical theory.

Chaucer's greatness arose from his growing recognition that for him at least the right way to amplify a story was not to expand it by rhetorical devices, but to conceive it in terms of the life which he had observed so closely, to imagine how each of the characters thought and felt, and to report how in this imaginative vision they looked and acted. And if he felt obliged, as apparently he still did, in writings of serious and lofty tone, to supply sententiae, proverbs, exempla, and other fruits of erudition, he came more and more to make only a dramatic use of these rhetorical elements, that is, to put them into the mouths of his dramatis personae and to use only such as might fittingly be uttered by them.

It is this dramatic use of rhetorical devices which we must learn to recognize in the later and more artistic poems, and which must be taken into account in our examination of the percentages of rhetoric in the separate tales of the Canterbury pilgrimage. The mere fact that the percentage in two such masterpieces of narrative art as the tales of the "Nun's Priest" and the "Wife of Bath" is nearly twice as great as in the less successful tales of the "Man of Law" and the "Doctor" would be very misleading, if taken without further investigation. But the difference in manner of introduction and use appears immediately and is of fundamental significance. In the tales of the "Doctor" and the "Man of Law" the rhetoric is prevailingly, indeed almost exclusively, used by the narrator; that is, it is not incorporated and used dramatically but stands apart from the tale. There is even a difference between the "Doctor's Tale" and that of the "Man of Law" in manner of handling. In the "Man of Law's Tale" the narrative is, for the most part, broken into comparatively brief sections and the rhetoric of the narrator is freely interspersed in the forms of apostrophe, exclamatio, collatio, sententiae, and exempla, with various digressions on astrology. In the "Doctor's Tale," on the other hand, the narrative comes in a solid block of 172 lines, preceded by 109 lines, all but 39 of which are purely rhetorical utterances of the narrator, and followed by 10 lines of rhetorical application. But both stories are, as artistic compositions, pretty crude and show no fusion of rhetorical elements. In the tales of the "Nun's Priest" and the "Wife of Bath" the situation is very different. In the "Nun's Priest's Tale," although the rhetoric is scattered through the narrative as in the "Man of Law's Tale," it is not the external comment of the narrator but the vitally dramatized utterance of speakers whose actions, and attitudes, and sentiments we accept as belonging to a world of poetic reality. In the "Wife of Bath's Tale" there are two main masses of rhetorical devices: one of them is the famous oration on 'gentilesse', poverty, and age uttered by the Fairy Wife to her humbled husband, the other is the long exemplum on woman's inability to keep a secret, uttered by the garrulous Wife of Bath herself. But in the latter instance no less than in the former the rhetoric is dramatic, is conformed to the character, and is motivated.

The tales of the "Prioress" and the "Second Nun" differ very slightly in percentage of rhetorical devices or in the placing of them. If we could isolate the tales—disconnect them from their narrators and the circumstances of their telling—we should probably agree that they show the same style of workmanship and may belong to the same period, a comparatively early one. But the difference between them in effect is very great. Why is this? Apart from the mere difference in appeal of the material of the two stories, is it not because in the one tale Chaucer has failed to visualize or to make his readers see the principal characters—Cecilia, Valerian, and Pope Urban remain to him and to us mere names—whereas both he and we have a vivid and charming picture of the little choir boy as he goes singing to his death? Is it not also because through some freak of chance the Second Nun herself is a mere name in the "Prologue" and is not mentioned at all in the pilgrimage, whereas both by the portrait in the "Prologue" and by the little episode of conversation with the Host the Prioress is endowed with lasting beauty and sympathetic appeal? Chaucer himself seems to have felt this. When the Prioress's tale is ended he tells us of its profound effect upon the whole party including himself; after the other tale he says, drily,

When toold was al the lif of Seint Cecile
Er we had ridden fully fyve mile,

we were overtaken by two men.

The tales of the "Franklin" and the "Merchant" differ only slightly in percentage of rhetorical devices from those of the "Prioress" and the "Second Nun," but in the placing and handling of these devices, as well as in other respects, they seem to belong to a much later period of Chaucer's workmanship. The dramatis personae are vividly conceived and the action is clearly visualized. Both tales show, however, the persistence of the rhetorical habit and training. In the "Merchant's Tale" most of the rhetoric is introduced dramatically as forming the speeches of January and his advisers, but there is a long undramatic passage—inappropriate either to the Merchant or to the clerical narrator for whom the tale appears to have been originally composed. In the "Franklin's Tale" a fine story finely told is nearly spoiled by one hundred lines of rhetorical exempla. The fact that they are put into the mouth of Dorigen in her complaint against Fortune indicates that Chaucer was trying to motivate them dramatically. But what reader, modern or medieval, would not have been more powerfully and sympathetically affected if Chaucer, with the psychological insight displayed in Troilus and Criseyde, had caused his distressed and desperate heroine to express the real feelings appropriate to her character and situation?

It may be noted that the tales showing a low percentage of formal rhetorical devices are, with a single exception, humorous tales and all are tales which on other grounds are regarded as of late date. The exception is the "Clerk's Tale," a pretty close translation from Petrarch. The small amount of rhetoric added by Chaucer in making this translation from Petrarch is in curious contrast to the large amount added in translating the "Man of Law's Tale" from Trivet. Can it be that his rivalry with Gower in the latter case was responsible for the rhetoric?

The absence of rhetorical devices from the humorous tales may be due in part to the specific declaration of the rhetoricians that rhetorical ornament of all sorts should be strictly excluded from such tales. But surely Chaucer's growing power of artistry, his vast observation of life, and his newly devised method of imaginative reconstruction of the scenes; characters, and events of his stories gave him such a wealth of significant detail that there was no need and no space for the older methods of amplification. Sententiae are reduced to single lines, mostly proverbs; exempla to passing allusions; apostrophes and exclamations to the briefest of utterances. For it is not only in the humorous tales that his advanced method isner's Tale" of the three roysterers who sought Death, is as vividly imagined as the tales of the Miller and the Reeve, and the long passages of rhetoric, placed between the opening twenty lines, which so wonderfully create background and atmosphere, and the narrative itself, are thoroughly explained and justified by their function as part of the Pardoner's sermon.

The survey we have made of Chaucer's work, hasty as it has necessarily been, has, I think, shown that he began his career, not merely as a disciple and imitator of a thoroughly artificial school of writing, but as a conscious exploiter of the formal rhetoric taught by the professional rhetoricians, and that it was only gradually and as the result of much thought and experiment that he replaced the conventional methods of rhetorical elaboration by those processes of imaginative construction which give his best work so high a rank in English literature. To treat his poems as if they all belonged to the same stage of artistic development and represented the same ideals of art is to repeat the error so long perpetrated by students of Shakespeare.