Historical Manual of English Prosody

Now probably the whole of these Romances were more or less directly imitated from French originals, nearly all of which we actually possess; but it is extremely remarkable that they by no means always followed the metre of those originals, and that when they did they took considerable liberties with it. That metre was almost invariably Alexandrine or decasyllabic, in long batches not couplets, or octosyllabic in couplet. Of the two probably oldest of ours, Havelok and Horn, the first does attempt this octosyllabic couplet, but treats it in a very rough and independent fashion, something in the Genesis and Exodus line, while King Horn seems to favour something like what we observe in part of the English Bestiary and the whole of the French one—a split Alexandrine or six-syllabled couplet. Very soon the rime couée or Romance-six (which had not been a staple romance-metre in French) appears, and occasionally more elaborate stanzas still, such as the complicated arrangement of Sir Tristrem. Those writers who prefer couplet improve upon Horn and Havelok, but they follow Genesis and Exodus much more than The Owl and the Nightingale.

Indeed, some of them develop this couplet in a manner possessing almost infinite "future." They not merely follow the writer of Genesis and Exodus in substituting trisyllabic, if not also monosyllabic, feet for dissyllabic to the number of four, but some of them develop hints, which may be found in that composition, by extending the actual foot-length of the line to five, and sometimes repeating this in an actual "heroic" pair. Whether this was in some, or even at first in all, cases accidental, does not really matter. The decasyllable or five-foot line was already existent in great masses of French poetry, though not in single couplets; it was natural that, occasionally, more room should be wanted than the octosyllable provides; and there is the undoubted fact that, in more than one other European language, ten, or according to the structure of the particular tongue, eleven syllables were suggesting themselves as the most convenient size. The fourteener was so long as to invite breaking up quite early; the Alexandrine has never naturalised itself for continuous use in English; and the octosyllable, though its early appearance, the wealth of models for it, and its ease, fostered and sustained it, had the already mentioned drawback of lack of content. It was certain that, in a language which was showing itself so fortunately free from hide-bound qualities, the decasyllable would establish itself. It has been usual to say that, in couplet at any rate, Chaucer "took it from the French." As a matter of actual practice he may have done so; but in the order of nature and thought it was not in the least necessary for him to do it. Indeed, it would be almost literally true[64] to say that English had decasyllabic couplet before French—that it was an English invention.

For the time, however, the octosyllable was the staple for narrative, varied to no mean extent by the stanzas already described; while these stanzas, often of the most elaborate and complicated descriptions, were adopted from French (and perhaps Provençal) or extemporised by the taste and fancy of the writers. One famous collection[65] indicates the school of which our poets were scholars by alternating French poems with English. But this very collection shows amply that these same writers refused to undergo the syllabic constraints of French, and held to what were to be always the real, if frequently the unrecognised and sometimes the denied, principles of the New English in verse—that is to say, the constitution of the line by feet, not syllables—and the consequent possibility of obtaining equivalent lines by the substitution of feet, varying in syllabic constituence, but interchangeable in metrical value. Some examples of all these things will be found in the Scanned Conspectus; the student should search the books named in the notes for more, which he will find in the fullest abundance. What is important is that by this study he may and should discover the real and too commonly misunderstood relation of Chaucer to precedent English verse.

There is, however, another fact of the fourteenth century which it is not less important for him to recognise, and which also has been too often misunderstood, or at least not put in its proper place. This is the revival of alliterative-accentual verse.

As there are few things, in treating prosody, of greater weight than to keep carefully before the student the difference between controversial and uncontroversial points, it should be said at once that "revival" is not quite one of the latter. There have been some who have taken it for granted that the alliterative-accentual form never ceased out of the land. It may be so; there is even a sort of antecedent plausibility about the notion. But the important historical fact is that no such verse apparently exists of a probable date between about 1250 (the later form of Layamon itself, much further encroached upon by metre and rhyme) and about 1350. Somewhere about this latter time it does reappear; and before very long has its chief pure representative in Langland, at the same time as metre has its chief pure representative in Chaucer.

But this reappearance is conditioned and qualified by a very remarkable fact. There is, as has just been said, pure alliterative verse. It is not, indeed, an exact representation of the old A.S. line. It is somewhat longer than the shorter forms of that line, and very much shorter than the "extended" variety. In some cases, especially in the later examples, the alliteration is richer, extending to four, five, or even six syllables. Most noteworthy of all is the substitution, in the general rhythmical run, of anapæstic-iambic for trochaic basis—a fact the importance of which, in the general history of the morphology of English poetry and of the change from A.S. to M.E., cannot be exaggerated.

But it is also worthy of the most careful remark that, in a relatively large number of instances, the alliterative-accentual system is apparently unable to rely upon itself. It is tempted or driven to borrow metre, or rhyme, or both. Of the two best pieces in the alliterative division, outside Piers Plowman, Gawain and the Green Knight combines, with an unrhymed body or tirade, a rhymed "bob and wheel" in every stanza; while The Pearl, though alliterated almost to the highest possible strength, is strictly metrical and strictly rhymed throughout. Others form their stanzas of lines roughly rhythmed but fairly well rhymed.

By the last quarter of the fourteenth century, therefore, there were in England two contrasted and in a way rival, but, as has been said, overlapping, systems of versification: one a sort of atavistic revival, the other the result of a process—two centuries old to a certainty, and probably nearer four—of blending the characteristics of Low Latin and French prosody with those of Old English.

In the three chief poets of the later fourteenth century (Chaucer, Gower, and Langland) we have three object lessons as to the results of this process, which could not have been improved if the course of events had been exclusively devoted to the task of making these results, and the process itself, clear to the student. They had best be taken in reverse order.

Langland represents, in the greatest perfection that can reasonably be expected, the attempt to preserve, or revert to, verse arranged without rhyme, without metre in the strict sense, and depending for its separation from prose upon alliteration, accent, and strong middle pause. In spite of himself, and in consequence of the state of the language, actually metrical lines—decasyllables, Alexandrines, and fourteeners—do appear; but, as a rule, he avoids them either with singular skill or with remarkable luck, and on the whole achieves a consistent medium, not so much dominated as permeated by a sort of anapæstic underhum of rhythm, but otherwise maintaining its independence. Being possessed of great literary and even distinctly poetical genius, he makes it a by no means unsuitable vehicle for his tangle of apocalyptic dreams, and no ill one for the occasional passages of a more mundane description which he interlards. But it is deficient in beauty, if not in vigour; it is clearly unsuited for many of the subjects of poetry; and to any one acquainted with metre and rhyme it constantly suggests the question and complaint, "Why are we to be deprived of these already-won beauties and conveniences, and cut off with this rough makeshift?"

As Langland represents the purely accentual division or phase of English prosody at this time, so does Gower represent the almost purely syllabic. He uses, with insignificant exceptions,[66] the old octosyllabic couplet; but he comes closer than any other English writer of the Middle English period to the strict French model. He does not, like his forerunners, and like even Chaucer, allow himself the seven-syllable line as a variation; and though he does, by the admission of those who are opposed to the system of this book, occasionally admit an "extrametrical syllable," and, according to that system, much oftener a trisyllabic foot, this interferes little with the general uniformity of his verse-run. Almost the only variations that he relies upon are frequent initial trochees an occasional balanced arrangement of the halves of the line—

contrasted with less strongly marked pauses, and especially a device whereby a full stop comes at the first line of two couplets separated by another, so that a sort of In Memoriam quatrain effect, with first and last lines blank, is obtained, as thus:

To this the present writer would add distinct trisyllabic feet where others see slur, as in—

The result, especially with syncopation of these trisyllables, is what some call "pre-eminent smoothness" of metre, others dominant monotony. The metre had proved itself of old well suited for actual narrative, and, as Gower can tell a story, when he has a good one to tell, the effect, as in the passages about Nebuchadnezzar, Medea, Ceyx and Alcyone, Rosiphelè, the "Trump of Death," and other persons and things, is quite excellent. But in the didactic and conversational parts it is often terribly tedious and lamentably limp.

Thus Langland, from yet another point of view, represents the rejection of the new English prosody altogether or as far as possible, and Gower, the timid imitation of French. Chaucer, on the other hand, despite his undoubted attention to French and Italian models, is in the direct line which we have been tracing, and represents, if not completely, yet to a very large extent, at once the development and the perfecting of the processes which we have described. It has indeed been urged by some that Chaucer probably knew nothing, or very little, of English poetry before his own day. But while, on the one hand, this is quite unproven, and not a little improbable, those who urge it do not seem to see that, even if it were so, it is comparatively irrelevant. It is not in the least necessary to suppose that Chaucer must have borrowed the Vernon MS. or another like it, carried it home to the rooms above Aldgate, "stirred the fire and taken a drink" as Henryson did later with his own Troilus, and then, after discussing to himself principles of versification, have decided that this was to be followed, that to be avoided, that again to be perfected and carried further. The main and undoubted facts remain that Chaucer was an Englishman of 1340(?)-1400; that he was the greatest Englishman of letters of his time; that he spoke and wrote the English language, and that thus, by what he would himself have called "the law of kind," he entered into the inheritance of all that had been done in this English matter by Englishmen for generations beforehand. As a matter of fact, there is plenty of evidence destructive of the contention referred to. He had read the Romances, or he could not have written Sir Thopas; he knew the alliterative poems, or he could not have made the famous reference to rum ram ruf in the Prologue to the Parson's Tale, which Gascoigne caught up. It is odd if he had not heard (even if he had not read) the plays that folk like his own Absolon played "upon a scaffold high." But, as has been said, it does not matter.

For his work is there, and it is incontestably—whatever its author had read or not read—the logical and biological continuation and perfecting of all that had gone before from Godric and the Paternoster. He begins with the fluent octosyllable and the melodious and usefully stringent rhyme-royal, as well as other more or less elaborate stanzas. He communicates to the couplet[67] a greater combination of order and variety than it had ever known in English; he makes of the stanza,[68] in the case of rhyme-royal, the most perfect formal arrangement of verse that English had yet seen. Later he takes up,[69] very probably because he had written so many separate examples of it in rhyme-royal itself at the close of each stanza, the decasyllabic couplet, and makes of that something greater still—a metrical instrument or vehicle escaping at once the scanty content and slightly insignificant bearing of the octosyllable, the elaborateness and rather melancholy quality of rhyme-royal. In doing this it is inevitable that, as Spenser did in parallel case afterwards, he should lean rather towards precision than towards great laxity and luxuriance of form; for things needed order. But he sets the example of that variation of pause in rhyme-royal which was fortunately taken as a rule, and which preserved for English one of the very greatest means of metrical achievement. In the octosyllable he reproduced knowingly, and with definite apology, that "failing of a syllable" which gives acephalous or trochaic alternation, and which all the greatest masters of the metre, except (following Gower) William Morris, have imitated. And he broke up the lines very largely by conversation-fragments, by putting full stops at the end of the first line of a couplet, and by making a whole paragraph end at the same place.

But next to his provision of a perfectly finished stanza—in other words, of a complete, and pro tanto final, prosodic result—in rhyme-royal, the most important thing done by Chaucer in this department was the arranging and setting on foot of the decasyllabic couplet, which he began well in the Legend of Good Women, but carried on much better in the Canterbury Tales. Not half of his actual achievement here, and a very much smaller part of his promise and stimulus for the future, can be perceived by those who limit him to the decasyllable as such by devices of elision and syncope; still less by those who would have his varieties of line exactly to represent variations of the French decasyllable. The former proceeding is inadequate and defacing; the latter practically impossible, except as a bare and barren matter of arithmetic. You cannot imitate the prosodic effect of one language in another, even though you take the exact number of syllables and (as far as you can) divide the words, arrange the accents, etc., with the most slavish copying. The result will laugh at you prosodically; and while it is very unlikely to give you anything similar, it is nearly certain to give you something quite different.[70]

When Chaucer's verse in "heroic" or "riding rhyme" is examined, simply on its own merits and without regard to arbitrary theories of pronunciation, but with all necessary remembrance of the value of the final e, etc., it is seen to follow, in every respect, the general principles which we have seen evolving themselves in all English poetry hitherto, subject only to the general reforming or regimenting tendency which has been noticed. The normal line is beyond all question five-foot iambic, or decasyllabic with short and long syllable alternately. But there are a few instances[71] of so-called acephalous lines where the first syllable seems to have been missed—where, at any rate, there are only nine to account for, and where you consequently have to choose between a monosyllabic foot in the first place or trochaic cadence throughout. There is little doubt in the mind of the present writer that if these lines (which, after all, are very few) were deliberately written and meant to be kept, the reason of their existence was a false analogy with the octosyllable, where, as we have said, such acephalous lines, trochaic and heptasyllabic, do occur, and where they produce not only no ill, but a positively good effect. Unluckily the cutting down does not produce a good effect in the larger couplet; and if trochaic rhythm is permitted—in other words, if the missing syllable is shifted from the beginning to the end—it produces a very bad one. But they are, as has been said, in very small proportion, though there are too many of them to be simply "mended" out of existence.

Proceeding, we find, in a far larger number of instances, not a defect but an excess of syllables. As far as these syllables are found at the end of the line (in great measure caused by the final e) there is no difficulty and no dispute about them. They are allowed by everybody; and they come under that general law of almost (not quite) all prosodies which makes the final place of a line one of liberty. But it is different with those which come within the line, and with apparent extensions beyond the eleventh syllable. Many, perhaps most, prosodists would shut their eyes to the latter, regarding them as mere extra-redundances, and explain away those which occur within the line by elision before a vowel, by syncope or crasis or the like (see Glossary) when they come before a consonant.

To the present writer these devices and shifts appear unnecessary, discordant, the reverse of natural, and alike the consequence and the cause of prosodic error. With regard to hiatus (i.e. the actual contact of vowels) it has to be fully admitted that there is a strong tendency in MSS. to sink one of them and to write not merely "tharray" for "the array," but even "in thalyghte" for "in thee alyghte." The habit continued for a long time, and we find even in Wyatt and Surrey "tembrace" for "to embrace" and so forth. But it is important to observe first that this habit is not constant, as we should expect it to be if it represented a definite and reasoned wish always to reduce two such syllables to one; and further, that it will not affect the other cases of syllables, such as the last of "Heaven" (which, however, pretty certainly was monosyllabic at this time and later), "ever" the -eth of the third person singular and plural, y in "many a," i- in scores of words, and the like.

To the present writer, once more, it is certain, and even indisputable, that whether Chaucer deliberately used trisyllabic feet or not, there are trisyllabic feet by nature and poetic right in Chaucer, for any one who chooses them. And he is of opinion, though not so strongly, that Chaucer allowed himself an occasional Alexandrine or twelve-syllabled line,[72] just as preceding writers had allowed themselves occasional ten-syllabled lines in octosyllabics. What is once more certain, and almost indisputable, is that his lines can be so scanned with euphonious effect, and that similar phenomena manifest themselves all the way up to his time.

Of his rhymes nothing necessarily need be said here. He often avails himself for rhyme, as well as for rhythm, of the choice between Teutonic and Romance accent—the former always seeking the beginning of the word, the latter generally the end. This was hardly even a licence at his period.

One much-vexed point it is, however, impossible to omit, though far more, in every sense, has been made of it than it is worth. It occurred many years ago to a distinguished scholar, the late Mr. Bradshaw of Cambridge, to make a test out of the rhyme of y and ye, which, he thought (despite a famous example in Sir Thopas[73]), never occurs in the work unquestionably Chaucer's. To the present writer the occurrence of the rhyme in Sir Thopas closes the question, and he would have much to say against the establishment of the test, even if Sir Thopas were acknowledged as not Chaucerian. But from the strict point of view of this book the whole thing is really irrelevant. It does not matter to us who wrote certain pieces of English poetry, but what the characteristics of those and other pieces of English poetry are. The student of prosody may and should note that in some pieces of this period the rhyme of y and ye certainly does occur, that in others it apparently does not; but beyond this he need not, and, as a student of prosody, should not, go.

CHAPTER II
FROM CHAUCER TO SPENSER—DISORGANISATION AND RECONSTRUCTION

It might be supposed, especially in face of the unquestionable reputation which Chaucer had attained before his death—and which he maintained undisturbed, and hardly approached, for the entire period until Spenser's birth,—that his prosodic work, once done, would have been done once for all; that in points of form, though individual inferiority of poetic gift might show itself, there could be no great technical falling off. To think this, however, would be to ignore—as, in fact, men too usually do ignore, and have ignored—the necessary and intricate connection between language and prosody. Chaucer had raised the state of English versification to the highest point possible in his time; in fact, there are reasons for saying that he had screwed it up beyond the level possible to ordinary men. To mention nothing else, the exactness, and at the same time the rhythmical variety of his verse, depend on two special points—the valuing of the final e and the optional but carefully selected shift from French to English accentuation.[74] We know that, even in the mouths and on the pens of his own contemporaries, the e was breaking down, and that it "went" more and more during the fifteenth century; and we know likewise, though less certainly, that though, even at the close of the period with which we are dealing, French accentuation was still permissible to poets, an English standard was gradually establishing itself, violation of which was disapproved.[75] Moreover, the fact remains undeniable that the poetic quality of the followers of Chaucer, in Southern English of the literary kind, was low to a point unprecedented, and never yet again reached since.

The progress of prosody between Chaucer and Spenser divides itself, sharply but unequally in point of time, between a longer space (about a century and a quarter) from Chaucer to poets like Hawes, Skelton, and Barclay; a shorter (of about half a century or less) from Wyatt to Spenser. In the first division a subdivision—of matter, not time—has to be made between the literary poets in Southern English, the Scottish Chaucerians from James the First to Douglas or Lyndsay (if not even to Montgomerie, who died later than Spenser himself), and the ballad, carol, and other folk-song writers of the fifteenth century.

The history of the first division is the history of the breakdown just referred to. Except in the so-called Chauceriana—pieces such as "The Cuckoo and the Nightingale," "The Flower and the Leaf," "The Court of Love," etc., once attributed to Chaucer himself, but cast out on various kinds of evidence ranging from practically conclusive to very doubtful—and sometimes even in such poets as Lydgate and Occleve, who for no very small portion of their lives were Chaucer's own contemporaries, downwards, seem to be struck with metrical palsy or metrical blindness. Examples, given in the Scanned Conspectus above, will show the way in which they confuse different metres, vary the lengths of their lines not by intentional substitution but by sheer muddlement, violate rhythm and cadence—turn, in fact, the perfect harmony of their master into a cacophony which is not even prosaic. Sometimes, especially in Occleve, by rigid counting of syllables, they escape worse blunders, though they seldom make real music. Generally, even this resource fails them, and there is no worse chaos than in Hawes, one of the latest and not one of the least of them; while Skelton, perhaps the acutest intelligence of all, takes refuge in frank, not clumsy, and intentional doggerel.

To this spectacle of disorganisation and decay the Scottish followers of Chaucer (who, generally with acknowledgment as eager and hearty as that of their English comrades, take him for their master) present what may at first sight seem an astonishing and almost unintelligible contrast. With final e's allowed for (or in case of necessity touched in), the Kingis Quair, traditionally ascribed, and never with solid reason denied, to James the First, is a piece of rhyme-royal as soundly constructed, and as well fitted and polished, as if it were Chaucer's own. Henryson, in his following of Chaucer's Troilus, and in his other poems, never breaks down in metre, but handles every form that he touches with equal precision and charm. Even more may be said of Dunbar, whose lyrics possess the peculiar grace only given by metrical accomplishment, who can manage alliterative metre more smoothly than Langland and with not less vigour, and who, if he wrote the "Friars of Berwick," is, next to Chaucer himself, the greatest master of the early (Middle English) heroic couplet. Of the verse-chroniclers, Wyntoun, though not very poetical, uses octosyllabic couplet, with not infrequent equivalence, effectively enough, and Blind Harry writes very strict decasyllabic couplet with cæsura at the fourth syllable, after the French model. The earlier sixteenth-century writers, Douglas and Lyndsay, if not perhaps quite impeccable, appear so beside Hawes and his fellows; while the two latest strictly Scots poets, Scott and Montgomerie, manage most complicated measures—reminding us of early French and Provençal, or of those of the English fourteenth century in lyric and drama—with unerring accuracy and finished grace. Of this strange contrast the simple fact of writing in a different dialect, requiring more care in imitation, may supply some explanation; the other fact, that this dialect was rather a literary convention than a vernacular speech, some more; and the higher quality of individual genius, more still; but a margin of surprise remains.

It is difficult to say whether that margin is reduced or widened by the fact that a contrast, almost as striking, is found between the English literary poetry of the period and the "folk-song," sacred and profane. It is probable that the bulk of our older ballads date from the earliest fifteenth century or the very close of the fourteenth. The latter would seem to be true of the "Robin Hood" ballads; the former is pretty certainly true of "Chevy Chase." We have also from the fifteenth century a large body of carols, or sacred poems for singing.

Now in these, though they naturally vary much in poetic merit and in prosodic accomplishment, it is remarkable that this latter scarcely ever falls to the level of the worst literary poetry, and never falls in exactly the same way. The ballad-writers invariably, and the carol- and hymn-writers very commonly, preserve the English licence of equivalence in the fullest fashion; and this seems to relieve their motion of the staggering and fatal cramp which rests on their superiors in formal literary rank. They sing naturally: they do not aim at, and break down in, a falsetto. Although it would be impossible to have anything in a worse condition, as far as copying goes, than our oldest version of "Chevy Chase," its natural ballad motion carries it safe through all the corruptions and defacements; the sacred song of "E.I.O." is admirable metre; the Carol of the Virgin, "I sing of a maiden," is matchless in quiet metrical movement; and the famous "Nut-brown Maid," which is certainly not later than this century, deserves the same praise in more rapid melody.

These compositions, however, though they did a precious office in preserving the true principles of English prosody, could not exercise immediate influence; and the disorganising of literary versification was no doubt partly cause and partly consequence of the continuance of the alliterative revolt which did not die till after Flodden—indeed, not till after Musselburgh (Pinkie). But, indirectly, this revolt encouraged fresh developments of English metre itself. The old fourteener had taken new and lively form in such pieces as Gamelyn[76] (late fourteenth century) and Beryn (middle fifteenth), and through it and other things—the musical adaptations of songs and hymns and the like—there was arising, in dramatic literature especially, a disorderly, imperfect, but very important notion of wholly "triple-timed" or anapæstic metre. In fact, it is not excessive to regard the English fifteenth century as a period when all elements of prosody were thrown into a sort of cauldron, sack, sieve, or lucky-bag, in which, as according to the different metaphors suiting these objects, they were to be boiled down, shaken together, sifted out, and taken as fortune would have it, to supply the stock of a new venture in more orderly and polished verse-manufacture when actual speech had settled itself once more.

At what period, in what manner, and by what persons exactly, conscious discontent with this confusion and dilapidation was made manifest, is not known. That it was felt consciously about the middle of the sixteenth century we do know positively from a passage in the Mirror for Magistrates; and later still we find the precepts of Gascoigne virtually, if not always expressly, directed against it. But, as has been hinted, even Skeltonic evinces an earlier attempt to escape from it in practice as far back as the first quarter of the century; while, at an uncertain time for first efforts, during the second, and then ever increasingly during the third, till the death of Gascoigne himself, poetical practice proclaims the fact, even more emphatically than any preceptist rules of criticism could do. Indeed, there has hardly ever been any mistake, and it is difficult to think that by persons possessed of ears and eyes any could be made, about the surprising revolution manifest in the verse of Sir Thomas Wyatt, and of his younger disciple, Henry Howard, known by his courtesy title as the Earl of Surrey. Instead of the weltering and staggering discords of the poets from Lydgate onward, we come back to verse almost as clear, regular, and harmonious as Chaucer's, though with a much more modern pronunciation and accent, to which it occasionally seems to have some difficulty in reconciling itself. The final e has in most cases disappeared, though it is probably there in a few cases, and in a few others has settled itself into y.[77] The inordinate variety of syllables in the line, not explicable by any trisyllabic foot, is reformed. Indeed, the need of the reform is so strongly felt that the poets run into the opposite error—salutary for the time—of excessive syllabic uniformity.