BY

CAMBRIDGE
HARVARD UNIVERSITY PRESS
1924

LONDON: HUMPHREY MILFORD
OXFORD UNIVERSITY PRESS

COPYRIGHT, 1922
BY HARVARD UNIVERSITY PRESS

Third printing

PRINTED AT THE HARVARD UNIVERSITY PRESS
CAMBRIDGE, MASS., U. S. A.

TO
C. H. N. B.

[vii]

Most of the older discussions of English versification labored under two difficulties: an undue adherence to the traditions of Greek and Latin prosody more or less perfectly understood, and an exaggerated formalism. But recently the interest and excitement (now happily abated) over free-verse have reopened the old questions and let in upon them not a little light. Even today, however, a great deal of metrical analysis has wrecked itself on the visible rocks of a false accuracy, and it is therefore not only out of caution but also out of mere common sense that we should eschew the arbitrary, even at the risk of vagueness and an 'unscientific' admission of uncertainty. For the only great and annihilating danger of writing on versification is dogmatism. Our theorists, both old and new, are first tempted and then possessed with their theories—all else becoming wrong and intolerable. In the following pages I have perhaps erred in a too frequent insistence on doubts and perplexities; perhaps also, on occasion, in a too plain statement of opinion where judgments are bound to differ—sic se res habent.

Now it is plain that rhythm is one of the ultimate facts of nature and one of the universal principles of [viii]art; and thus versification, which is the study of the rhythms of verse, is both a science and an art. But it differs from the other sciences in that its phenomena are not 'regular' and reducible to law, but varying and subject to the dictates, even the whims, of genius; inasmuch as every poem involves a fresh fiat of creation. Of course, no poet when he is composing, either in the traditional "fine frenzy" or in the more sober process of revision, thinks of prosody as a science, or perhaps thinks of it at all. If he did he would go mad, and produce nothing. But the phenomena remain, nevertheless, and the analysis of them becomes for us a science.

This analysis has what Bacon would call two inconveniences. The first is complexity. The various ways in which the formal rhythms of verse combine with the infinitely modulated rhythms of natural prose produce a resultant which is complicated to the last degree and which almost precludes orderly exposition. No system has been devised to express it. The simpler ones fail through omission of important difficulties, the more elaborate totter under their own weight. And thus the Gentle Reader is either beguiled by false prophets—looks up and is not fed—or loses heart and saves himself by flight. There is, to be sure, an arcanum of prosodic theory which is the province of specialists. It has its place in the scheme of things; but it is no more necessary for the genuine enjoyment of Milton (or the 'moderns') than a knowledge of the formulae for calculating the parallax of Alpha Leonis is necessary for enjoying the [ix]pillared firmament. We must then compromise with a system which reveals the existence of all the phenomena and tries to suggest their interrelated workings.

The other inconvenience is that of seeming to deny the real poetry by our preoccupation with its metrical expression. "Under pretence that we want to study it more in detail, we pulverize the statue." This is an old charge, and our answer is easy. For, however it may be with the statue, a poem is never pulverized; it is still there on the page! No amount of analyzing can injure the poem. If we think it has injured us, even then we err, and need only recall our natural aversion to hard labor. In nearly every instance it was the work and not the analysis that bothered us.

This is a small book and therefore not exhaustive. And since it is as elementary, especially in the treatment of the principles of rhythm, as is consistent with a measure of thoroughness, the apparatus of mere learning has been suppressed, even where it might perhaps seem needed, as in footnote references to the scientific investigations on which part of the text is based. I have consulted and used, of course, all the books and articles I could find that had anything of value to offer; but I have rarely cited them, not because I wish to conceal my indebtedness, but because there is no room for elaborate documentation in such a book as this. On the other hand, I owe a very great deal, both directly and indirectly, to Professor Bliss Perry—although my manuscript was finished before I saw his Study of Poetry; and this debt I wish to acknowledge most fully and gratefully.

[x]

In lieu of a formal bibliography, I think it sufficient (in addition to the footnotes that occur in their proper place) to refer the reader to the larger works of Schipper and Saintsbury, to the smaller volumes of Professor Perry and Professor R. M. Alden, and particularly to Mr. T. S. Omond's English Metrists, 1921.

P. F. B.

[xi]

[1]

[3]

Rhythm, in its simplest sense, is measured motion; but by various natural extensions of meaning the word has come to be used almost as a synonym of regularity of variation. Whatever changes or alternates according to a recognizable system is said to be rhythmic, to possess rhythm. In this sense, rhythm is one of the universal principles of nature. We find it in the stripes of the zebra, the indentation of leaves, the series of teeth or of crystals, the curves of the horizon; in the tides, the phases of the moon, the rising and setting of the sun, the recurrence of seasons, the revolutions of planets; in the vibrations of color, sound, and heat; in breathing, the throbbing of the pulse, the stride of walking. All action and reaction whatever is rhythmic, both in nature and in man. "Rhythm is the rule with Nature," said Tyndall; "she abhors uniformity more than she does a vacuum." So deep-rooted, in truth, is this principle, that we imagine it and feel it where it does not exist, as in the clicking of a typewriter. Thus there is both an objective rhythm, which actually exists as rhythm, and a subjective rhythm, which is only the feeling of regularity resulting from a natural tendency of the mind to 'organize' any irregularity that we meet.

[4]

There are two fundamental forms of rhythm, though these are not altogether mutually exclusive, (1) spatial, and (2) temporal.

Spatial Rhythms. The simplest spatial rhythm is a series of equidistant points—

..........

More complex forms are the succession of repeated designs in mouldings and wainscotings (for example, the alternation of egg and dart), the series of windows in a wall, or of the columns of a Greek temple, or of the black and white keys of a piano. Still more complex is the balanced arrangement of straight lines and curves in a geometrical design, as in certain Oriental rugs or the Gothic rose windows. And probably the most complex spatial rhythms are those of the facades of great buildings like the Gothic cathedrals or St. Mark's of Venice, where only the trained eye perceives the subtleties of alternation and balance.

Temporal Rhythms. Temporal rhythms, apart from those of planetary motion, the alternation of seasons, and the like (which are called rhythmic by a metaphorical extension of the term), manifest themselves to us as phenomena of sound; hence the two concepts time-rhythm and sound-rhythm are commonly thought of as one and the same.

The simplest form is the tick-tick-tick of a watch or metronome. But such mechanical regularity is comparatively rare, and in general the temporal rhythms are all highly complex composites of sounds and silences. Their highest manifestations are music and [5]language. The rhythm of language, and a fortiori that of verse, is therefore primarily a temporal or sound rhythm, and as such is the particular subject of the following pages.

Combinations. Language, however, when addressed to the eye rather than to the ear, that is, when written or printed rather than spoken, is partly a spatial phenomenon; and, as will appear presently, the arrangement of words and sentences on the formal page is a real factor in the rhythm of verse. Moreover, most of the rhythms of motion, such as walking, the ebb and flow of tides, the breaking of waves on the beach, are composites of temporal and spatial.[1]

Sound Rhythm. These elementary generalizations must be narrowed now to the special phenomena of sound, and then still more particularly to the sounds of language.

All musical tones, including the phonetic sounds of words, have four characteristics: pitch, loudness or intensity, quality or tone-color, and duration. The last, of course, needs no definition.

Pitch is dependent on the number of vibrations per second. The greater the number of vibrations, the higher the pitch and the more 'acute' the tone. The [6]lowest pitch recognizable as a tone (as distinguished from noise) is 8 vibrations a second; the highest pitch the ear can hear is between 20,000 and 30,000 a second. In normal English speech among adults the voice ranges from about 100 to 300 vibrations, but in animated speaking this range is greatly increased.

Loudness is a comparative term for the strength of the sensation of sound in the ear. It is determined by the energy or intensity of the vibrations and varies (technically speaking) as the product of the square of the frequency and the square of the amplitude (I=n²A²). But for ordinary purposes it is sufficient to regard loudness and intensity as the same. The distinction, however, is clear in common practice; for whether one says "father" loudly or quietly, there is a relatively greater intensity of sound in the first syllable than in the second. In speech this intensity is called accent or stress.

The third characteristic, variously called quality, timbre, tone-quality, tone-color, is that which distinguishes sounds of the same loudness and pitch produced by different instruments or voices. It is the result of the combination of the partial tones of a sound, that is, of the fundamental and its overtones. In music, tone-quality is of the utmost importance, but as an element of speech rhythm it is practically non-existent, and may be wholly neglected, though it plays, of course, a prominent part in the oral reading of different persons.[2]

[7]

What is the relation of these physical attributes of sound to sound rhythm? The answer lies in a closer examination of the nature of rhythm, especially as it concerns the rhythm of speech.

Rhythm means measured flow or succession. Now first, in order that any succession may be measured, there must be something recognizable which distinguishes one unit from the next. In spatial rhythms the point of division is almost always easily perceived; hence the greater difficulty of analyzing the simplest time-rhythms as compared with the most complex space-rhythms. Moreover, the basis of measurement, that by which the 'distance' between any point of division and that which follows it is determined, must, by definition, be duration of time. Suppose, however, that the time-distance between successive points of emphasis or division is equal, is the rhythm therefore necessarily regular? No, because the points of emphasis themselves may vary in force or energy. Thus if in the following scheme (´ = point of emphasis; - = equal time-distance):

´-´-´-´-´-´- etc.

every ´ is not of the same value, the result might be (´´ = twice as much emphasis as ´; ´´´ = three times as much):

´´-´-´´´-´´-´´´- etc.

[8]

and this could not be called regular. A simple illustration of this is the difference in music between 3/4 time, where we count 1´ 2 3, 1´ 2 3, 1´ 2 3 and 6/4 or 6/8 time, where we count 1´´ 2 3 4´ 5 6, 1´´ 2 3 4´ 5 6. Furthermore, apart from any question of force or energy applied in the production of a sound, it is clear that high notes seem to possess a greater strength than low notes, and must therefore be recognized as an element in rhythmic emphasis. For example, if the following series of notes were sounded on a piano, and each struck with equal force—

etc.

a certain 'accent' would probably be felt on the e which was not felt on the a. And it is well known that shrill sounds and high-pitched voices carry farther and seem louder than others.

In the simplest kind of temporal rhythm, therefore, where the beats are, say, drum-taps of equal force, the primary element is time. But if there is the added complication of drum-taps of unequal force, the element of comparative stress must be reckoned with. And if, finally, the drum-taps are not in the same key (say, on kettledrums differently tuned), then the further element of comparative pitch must be considered as a possible point of emphasis. In a word, pitch may sometimes be substituted for stress.

In music rhythmic units may be marked by differences in tone-quality as well, and thus the potential complexity is greatly increased; but in spoken language, [9]as has been said, this element of rhythm is negligible. In speech-rhythm, however, the three conditions of time, stress, and pitch are always present, and therefore no consideration of either prose rhythm or verse can hope to be complete or adequate which neglects any one of them or the possibilities of their permutations and combinations. And it is precisely here that many treatments of the rhythm of language have revealed their weakness: they have excluded pitch usually, and often either stress or time. They have tried to build up a whole system of prosody sometimes on a foundation of stress alone, sometimes of time alone. The reason for this failure is simple, and it is also a warning. Any attempt to reckon with these three forces, each of which is extremely variable, not only among different individuals but in the same person at different times—any attempt to analyze these elements and observe, as well, their mutual influences and combined effects, is bound to result in a complication of details that almost defies expression or comprehension. The danger is as great as the difficulty. But nothing can ever be gained by the sort of simplification which disregards existent and relevant facts. It is to be confessed at once, however, that one cannot hope to answer in any really adequate way all or even most of the questions that arise. The best that can be expected is a thorough recognition of the complexity, together with some recognition of the component difficulties.

Moreover, only a part of the problem has been stated thus far. Not only is all spoken language the [10]resultant of the subjectively variable forces of time, stress, and pitch, but these three forces are themselves subject to and intimately affected by the thought and emotion which they express. Though educated persons probably receive the phenomena of language more frequently through the eye than through the ear, it is true that words are, in the first instance, sounds, of which the printed or written marks are but conventional symbols. And these symbols and the sounds which they represent have other values also, logical or intellectual and emotional values. Language is therefore a compound instrument of both sound and meaning, and speech-rhythm, in its fullest sense, is the composite resultant of the attributes of sound (duration, intensity, and pitch) modified by the logical and emotional content of the words and phrases which they represent.

For example, utter the words: "A house is my fire," and observe the comparative duration of time in the pronunciation of each word, the comparative stress, and the relative pitch (e. g. of a and fire). Now rearrange these nearly meaningless syllables: "My house is afire." Observe the differences, some slight and some well marked, in time, stress, and pitch. Then consider the different emotional coloring this sentence might have and the different results on time, stress, and pitch in utterance, if, say, the house contains all that you hold most precious and there is no chance of rescue; or if, on the other hand, the house is worthless and you are glad to see it destroyed. And even here the matter is comparatively simple; for in reading the [11]following sentence from Walter Pater, note the manifold variations in your own utterance of it at different times and imagine how it would be read by a person of dull sensibilities, by one of keen poetic feeling, and finally by one who recalled its context and on that account could enjoy its fullest richness: "It is the landscape, not of dreams or of fancy, but of places far withdrawn, and hours selected from a thousand with a miracle of finesse."[3]

The last step of the complication, which can only be indicated here, and will be developed in a later chapter, comes with the mutual adjustment of the natural prose rhythm and the metrical pattern of the verse. Such a sentence as the following has its own peculiar rhythms: "And, as imagination bodies forth the forms of things unknown, the poet's pen turns them to shapes, and gives to airy nothing a local habitation and a name." Now read it as verse, and the rhythms are different; both the meaning and the music are enhanced.

[12]

These then are the problems and the difficulties. The solutions can be only partial and tentative, but they are the best we are able to obtain with our present knowledge and our present capabilities of analysis. As science today has advanced in accuracy of knowledge and understanding of the facts of nature far beyond the powers of our ancestors to imagine, so in the future psychologists may, and let us hope will, enable us to comprehend the subtleties of metrical rhythm beyond our present power. Yet there will always remain, since the ever-inexplicable element of genius is a necessary part of all art, a portion which no science can describe or analyze.

The Psychology of Rhythm. That nearly all persons have a definite sense of rhythm, though sometimes latent because of defective education, is a familiar fact. The origin and source of this sense is a matter of uncertainty and dispute. The regular beating of the heart, the regular alternation of inhaling and exhaling, the regular motions of walking, all these unconscious or semi-conscious activities of the body have been suggested; and they doubtless have a concomitant if not a direct influence on the rhythmic sense. Certainly there is an intimate relation between the heart action and breath rate and the external stimulus of certain rhythmic forces, as is shown by the tendency of the [13]pulse and breath to adapt their tempo to the beat of fast or slow music. But this can hardly be the whole explanation. More important, from the psychological point of view, is doubtless the alternation of effort and fatigue which characterizes our mental as well as physical actions. A period of concentrated attention is at once followed by a period of indifference; the attention flags, wearies, and must be recuperated by a pause, just as the muscular effort of hand or arm. In truth, the muscles of the eye play a real part in the alternations of effort and rest in reading. The immediate application of this psychological fact to the temporal rhythms has been clearly phrased by the French metrist, M. Verrier:

Thus we have always at hand both a more or less efficient bodily metronome in the pulse and in respiration, and also a "cerebral metronome" capable not only of easy adjustment to different rates of speed but also of that subtlest of modulations which psychologists call the 'elastic unit,' and which musicians, though not so definitely or surely, recognize as tempo rubato.

The sense of rhythm, as has been said, differs remarkably in different individuals—just as the sense of touch, of smell, of hearing.[5] To some, rhythm appears chiefly as a series of points of emphasis or stresses alternating with points of less emphasis or of none at all; such are called, in scientific jargon, 'stressers.' To others the principal characteristic of rhythm is the time intervals; such are called 'timers.' But this is a practical, not a philosophical distinction. For it is the [15]succession of points of emphasis which even the most aggressive stresser feels as rhythmic; and succession implies and involves a temporal element. The stresser's only difficulty is to feel the approximate equality of the interval. The essential thing, however, is to understand that, while time is the foundation of speech-rhythm, stress is its universal adjunct and concomitant.[6]

The explanation of this duality is simple. A series of identical tones

etc.

contains a simple objective rhythm. The pronounced timer will feel it clearly; the extreme stresser will not. Change the series to

etc.,

or

etc.,

and both will feel it; for in the last example both time and stress are obvious, and in the other the longer notes of the series produce the effect of stress.[7] Most persons, therefore, with a greater or less degree of consciousness, allow their physical or cerebral metronome to affect the simple

etc.,

[16]

so that they hear or feel either

etc.,

or

etc.,

It is thus that the clock says tick-tock, tick-tock, the locomotive chu-chu, chu-chu. Timers are in the minority.

A converse phenomenon of the subjective introduction of stress into a series of identical tones at equal intervals is the subjective 'organization' of a series of irregular beats. Some do this more easily and naturally than others, but the tendency is present in all who are not absolutely rhythm-deaf. The "minute drops from off the eaves" beat out a tune, the typewriter develops a monotonous song, the public speaker 'gets his stride' and continues in a sing-song.

Thus, when there are equal intervals but stress is absent, we more or less unconsciously supply it; when there are distinct stresses at irregular intervals we organize them into approximately regular intervals. We have in us by instinct and by development both the ability and also the need to draw forth rhythm wherever it is latent. Rhythm becomes one of our physical and mental pleasures, manifest in primitive dancing and balladry, sailors' chanteys, and the simple heave-ho's of concerted labor. It induces economy of effort, and so makes work lighter; and it has, though perhaps not always, a certain æsthetic value, in making labor more interesting as well as easier. It is one of the attributes of the god we worship under the name of System.

[17]

Coördination, Syncopation, Substitution. The processes of the subjective organization of rhythm may best be explained under the heads of coördination, syncopation, and substitution. Their application to the particular problems of verse will be apparent at once, and will, in fact, constitute the bulk of the following pages.

Coördination has two aspects, according as it is thought of simply as an existing fact or as a process. In the former sense it is the agreement or coincidence (or the perception of agreement or coincidence) between the simple normal recurrence of beats and the actual or predetermined pattern. Thus in the lines

the 'natural' beat of the words uttered in the most natural and reasonable manner coincides with the 'artificial' beat of the metrical line.

On the other hand, coordination is the process which results in one's reduction of irregular beats to an approximately regular series. When we hear a haphazard succession of drum-taps or the irregular click-click of the typewriter, most of us soon begin to feel a certain orderly arrangement, a rhythmical swing in the repeated sounds, a grouping according to a sort of unit which recurs with nearly equal intervals. The units are not absolutely equal, but are elastic, allowing of some contraction and expansion; yet they are so nearly equal, or we feel them so, that the series seems regular.

[18]

Now this process of coördination involves two activities, syncopation and substitution. The workings of both are highly complex and somewhat uncertain; they differ greatly in different individuals, and when analyzed scientifically seem to produce more difficulties than they explain. But fortunately the outstanding ideas are beyond dispute, and detailed examination can properly be left to the scientists.

Syncopation is the union, or the perception of the union, of two or more rhythmic patterns.[8] A familiar example is perhaps the 'three against two' in music, where one hand follows a tum-te-te, tum-te-te rhythm, the other a tum-te, tum-te. This complexity, which strikes us as sophisticated subtlety and is not always easy to reproduce, is in fact both simple and familiar to the untutored savage. We must remember that the evolution of language and of music has been for the more part in the direction of greater simplicity of structure. Primitive music, as we find it in the undeveloped Indians and Australasians, is often too complex to be expressed by our regular notation. Another familiar example of syncopation is the negro dance, in which the "dancer taps with his feet just half-way between [19]the hand-claps of those who are accompanying his performance."[9] And of course the commonest example is the strongly marked syncopation of ragtime.[10]

In prose, this syncopation is evident in the apparent recognition, and even reproduction in reading aloud, of a regularity of rhythm where none really exists; as when protracted reading or listening develops or seems to develop a monotonous sing-song. But this phenomenon cannot be explained briefly, and the details must be omitted here.[11] In verse also syncopation frequently occurs, though it is seldom recognized except as an 'irregularity.' In the following lines of Paradise Lost the first two coincide pretty closely with the normal beats of the measure; while in the third line the series is an entirely different one.

Here to stress distinctly but,-tent, could utterly ruins both the meaning and the music of the line: to utter the words as if they were ordinary prose would preserve the meaning, but destroy the verse-movement. In Milton's ear, however, and in ours if we do not resist, there is a subtle syncopation of four beats against five. (Of course syncopation alone does not explain the [20]rhythm of this line.) A most startling syncopation is ventured by Milton in Samson Agonistes (1071-72):

Substitution is simpler. It merely means recognizing the equivalence, and therefore the possibility of interchange, of a long interval with two or more shorter intervals whose sum equals the one long. That is, in music two quarter-notes are equal to a half-note, and they may be anywhere substituted one for the other; or a dotted half-note equal three quarter-notes, etc. In verse it means that three syllables (or one, or even four) may be substituted for the normal two syllables of a foot if the three (or one or four) are uttered in approximately the same period of time.

The term substitution, however, may be used in a larger sense. Thus far only the purely temporal element of the rhythm has been thought of. When the two others, stress and pitch, are recalled, it becomes clear that another sort of substitution is both possible and usual, namely, that of either pitch or stress for duration. In other words, the groups that make up a rhythmic series may be determined or marked off by emphasis of pitch or emphasis of stress as well as by duration of time. In fact, it is from this habitual interplay of the three elements that most of the complexity of metre arises; as it is the failure to recognize this substitution which has given the older prosodies much of their false simplicity and their mechanical barrenness.

Summary. The fundamental problems of versification are all involved in the principles of rhythm, especially [21]the temporal rhythm of language. The rhythm of both prose and verse is a resultant of the three attributes of sound: stress, duration, and pitch (the first two being usually the determining elements, the third an accessory element) modified by the thought and emotion of the words. The feeling for this rhythm, or perception of it, has both physical and psychological explanations, and varies considerably among individuals, some being 'timers,' others 'stressers,' apparently by natural endowment. The processes of our perception of rhythm are those of coordination, or partly subjective reduction of actual 'irregularities' to a standard of 'regularity'; this reduction being accomplished mainly by syncopation and substitution.

[22]

It is clear now that all language is more or less definitely rhythmical; and that the two fundamental and determining elements of speech-rhythm are time and stress. It is clear also that the essential thing in our perception of rhythm is the experience or recognition of groups, these groups being themselves distinguished and set off by stress and time. When there is an easily felt regularity of the groups, when the alternation of stress and unstress and the approximate equality of the time intervals are fairly apparent, then the rhythm is simple. When the regularity is not obvious, the rhythm is complex, but none the less existent and pleasing.[12] In other words, the character of language rhythm is determined by the relative proportion of coincidence and syncopation. In verse, coincidence preponderates; in prose, syncopation (and substitution). Between absolute coincidence, moreover, and the freest possible syncopation and substitution, infinite gradations are possible; and many passages indeed lie so close to the boundary between recognizable preponderance of the one or of the other that it is difficult to say this is verse, that is prose. [23]Various standards and conventions enter into the decision.

For practical convenience three main sorts of rhythmic prose may be distinguished: (1) characteristic prose, or that in which no regularity (coincidence) is easily appreciable; (2) cadenced prose, or that in which the regularity is perceptible, but unobtrusive, and (3) metrical prose, or that in which the regularity is so noticeable as to be unpleasing. No very clear lines can be drawn; nor should one try to classify more than brief passages in one group or another. And, obviously, longer selections will combine two or more sorts in succession. A few examples will serve to show what is meant.

Characteristic Prose. No prose, as has been said above, is without rhythmic curves; but the best prose, that which always keeps in view the best ideals of prose, carefully avoids consecutive repetitions of the same rhythmic patterns. It is the distinction of verse to follow a chosen pattern, with due regard to the artistic principles of variety and uniformity; it is the distinction of prose to accomplish its object, whether artistic or utilitarian, without encroaching on the boundaries of its neighbor. Prose may be as 'poetic,' as charged with powerful emotion, as possible, but it remains true prose only when it refuses to borrow aids from the characteristic excellences of verse.

To be sure, it is not always easy to avoid regular patterns in writing the most ordinary prose. They come uncalled; they seem to be inherent in the language. Here is, chosen casually, the first sentence of a current [24]news item, written surely without artistic elaboration, and subjected, moreover, to the uncertainties of cable transmission. It was no doubt farthest from the correspondent's intention to write 'numerous' prose; but notice how the sentence may be divided into a series of rhythmic groups of two stresses each, with a fairly regular number of accompanying unstressed syllables:

No one would read the sentence with a very clear feeling of this definite movement; in fact, to do so rather obscures the meaning. But the potential rhythm is there, and the reader with a keen rhythmic sense will be to some extent aware of it.

Again, there is in the following sentence from Disraeli's Endymion a latent rhythm which actually affects the purely logical manner of reading it:

Here one almost inevitably pauses after dreams (or prolongs the word beyond its natural length), though there is no logical reason for doing so. Why? Partly, at least, because persisted in her dreams and of riding upon el-have the same 'swing,' and the parallelism of mere sound seems to require the pause.

For these reasons, then, among others, the most 'natural' spontaneous and straightforward prose is not always the best. Study and careful revision are [25]necessary in order to avoid an awkward and unpleasant monotony of rhythmic repetition, and at the same time obtain a flow of sound which will form a just musical accompaniment to the ideas expressed. Only the great prose masters have done this with complete success. Of the three following examples the first is from Bacon; the second is from Milton, who as a poet might have been expected to fall into metre while writing emotional prose; the third is from Walter Pater—the famous translation into words of the Mona Lisa painted by Leonardo da Vinci. The first is elaborate but unaffected, the second is probably spontaneous, the third highly studied.

Here no continuous patterns are recognizable, yet the whole is felt to be musically and appropriately rhythmic. In the next excerpt, however (from John Donne), and in many passages in the Authorized Version of the Psalms, of Job, of the Prophets, there is a visible balance of phrases and of clauses, a long undulating swing which one perceives at once, though only half consciously, and which approaches, if it does not actually possess, the intentional coincidence of cadenced prose.

Cadenced Prose. Cadenced prose is in English chiefly an historical phenomenon of the seventeenth century. It is part of the late Renaissance literary movement, when prose, after vaguely classic models, was held worth cultivating on its own account; and is in some degree a tempered afterglow of the crude brilliance of euphuistic balance and alliteration. It made no effort to conceal its definite rhythmic movements—rather, it gloried in them; but was always careful that they should not become monotonous or too palpable.

In the following examples the rhythmic units are for the sake of clearness indicated by separate lines, after the fashion of 'free-verse.' The passages should be read first with the line-division uppermost in the attention; then as continuous prose. The result of the second reading will be perhaps a fuller appreciation of the rhythmic richness of the sentences, both as to variety and uniformity. Sing-song and 'pounding' are by all means to be deprecated.

(a) Simple two-and three-beat rhythms—

(b) Simple three-and four-beat rhythms—

(c) Mainly two-beat rhythms—

(d) Mainly three-beat rhythms—

Metrical Prose. The above passages are daring, but greatly daring. So great is the subtlety, the variety, the art, that they never fail of their intended effect. They are justifiable because they justify themselves—partly by their lofty and dignified content, partly of course by their sheer artistry. But when the same thing is attempted by unskilful hands it fails ingloriously. We say it has "a palpable design upon us," [30]and balk. Gibbon and Burke, as inheritors of the seventeenth-century tradition, sometimes fell into the error; Ruskin, with his 'poetical' style, was sometimes guilty; but the worst and most conspicuous offenders were Dickens and Blackmore. Examples are abundant. Not all are equally unpleasant; the individual taste of some readers will approve passages which others will reject. With Dickens and Blackmore, however, the phenomenon approaches downright deliberate trickiness.

A peculiar instance of metrical prose, avowedly an experiment and fortunately (as most will think) not repeated, is the passage near the end of Kingsley's Westward Ho! Kingsley called it 'prose shaped into song.' The objection is simply that in such a situation song is out of place. Let prose do the legitimate work of prose; and when the intensity of feeling justifies song, let there be song. No hybrids, no cross-breeding—unless, as here, for purposes of experiment. Here is a part of the passage:

Elements of Prose Rhythm. Thus far the discussion of language rhythm has been confined to a general perception of rhythmic movement. When an attempt is made to carry the investigation into greater detail, more difficult and from a prosodic point of view really crucial [32]problems present themselves. The essential thing in any perception of rhythm is the experience of groups; but what are the nature and determining qualities of these groups? In music there are bars—the primary rhythmic group, comprising a single rhythmic wave, that is, covering the time-distance from one point of division to another—phrases, cadences, etc. The dual nature of language, however, its union of sound elements and thought elements, gives the question another aspect. Corresponding to the musical bar there is the metrical foot; to the musical phrase, the logical phrase; to the musical cadence, a similar melodious flow of word-sounds. But there are also in prose what are called breath-groups and attention-groups, series of words bound together by the physiological requirements of utterance and the mental requirements of perception and understanding.[13] The first step towards clearness will be a closer distinction between prose and metrical rhythms.

Syllable. The simplest and smallest unit of speech-sound is the syllable; then follow, in increasing magnitude, the word, the phrase (that is, words held together by their meaning or by their sound), the clause, the sentence, the paragraph. These units exist in verse as well as in prose, but while verse has other units (which are arbitrary and artificial), prose rhythm has only these. The rhythm of a paragraph is determined by the length, structure, content, and arrangement of the sentences; that of a sentence by the length, structure, [33]content, and arrangement of the phrases; that of the phrase by the length, structure, content, and arrangement of the words; that of a word by the character of the syllables. Now syllables, as has been explained above, have the sound attributes of duration, intensity (or lack of intensity), and pitch—called, however, in the terminology of phonetics, length or quantity, accent (or no accent), and pitch. These must be studied individually before their combined effects can be understood.

Length. Length is of course comparative. Some vowels require a longer time to enunciate than others: the e in penal than the i in pin, the o in coat than the o in cot, etc. Again, some consonants are shorter by nature than others: the explosives, p, t, k, etc., than the continuants s, z, th, f, m, n, l, etc. When vowels and consonants are combined into syllables the comparative length is still more apparent: thus form is longer than god, stole than poke, curl than cut, etc. Moreover, it is not alone the natural quantity of vowels and consonants that affects or determines their length, but also their position in a word and in a sentence. Thus, for example, the same sounds are uttered more rapidly when closely followed by one or more syllables than when alone: as bit, bitter, bitterly; hard, hardy, hardily. This elasticity of syllabic quantity is clearly shown in Verrier's examples:[14]

[34]

These indications, moreover, cover normal utterance only; in emotional language or elocutionary delivery there are deliberate and arbitrary lengthenings and shortenings.[15]

Accent and Stress. The term accent may best be reserved for grammatical or dictionary accent—the greater emphasis placed according to standard usage upon one syllable of a word as compared with the others. Thus portion has an accent on the first syllable, material on the second, apprehension on the third, deliberation on the fourth. The other syllables are either unaccented, as the first of material and the second of portion, or have a secondary accent, as the second of deliberation.

Accent should be distinguished from stress, which is the rhythmical emphasis in a series of sounds. In prose the rhythmical stress is determined almost wholly by accent; in verse the two sometimes coincide and sometimes differ markedly.

In certain words whose accent is somewhat evenly divided between two syllables, and in certain combinations [35]of monosyllables, there is a tendency to subject even grammatical accent to rhythmical stress. Hence the common pronunciations Newfoundland, Hawthornden; the alternation of stress in poor old man, sad hurt heart; and the shift of accent in "In a Chinese restaurant the waiters are Chinese."

Pitch. Pitch is a very uncertain and variable phenomenon. For the most part it is an ornament or aid to simple language rhythms, but under some conditions it plays an important rôle which cannot be neglected. Because of the physical structure of the vocal organs pitch is constantly changing in spoken discourse, though often the changes are not readily perceptible. Usually it coincides with accent.[16] It is also a frequent but by no means regular means of intensifying accent: compare "That was done simply" (normal utterance) with "That was simply wonderful" (intensive utterance). On the other hand pitch and accent sometimes clash: compare "The idea is good" (normal utterance) with "The idea!" (exclamatory). Other examples of [36]pitch as a significant factor in prose are: "One should not say 'good' but 'goodly,' not 'brave' but 'bravely'"; "Not praise but praising gives him delight."[17]

Another aspect of pitch is that which in the rhetorics is usually called inflection. A question is uttered with rising inflection, that is, with a higher pitch at the end. Declarative sentences usually have a falling inflection just before the final period, that is, a lower pitch. Exclamations often have a circumflex inflection, as "Really!" spoken in a sarcastic tone; that is, the pitch rises and falls.

Experimental attempts to indicate variations of pitch by our common musical notation are given by Verrier. A single example will suffice here.[18]

Perhaps the most important aspect of pitch from the point of view of rhythm is its actual influence upon accent. We say naturally: "He was fifteen years old"; but place the numeral for emphasis at the end of the sentence and it receives a kind of pitch accent: "His [37]age was fifteen." Compare also Chinese and Chinese in the example above.

Observe carefully the elements of duration, stress, accent, and pitch in the following sentences:

Word and Phrase Rhythm. The next larger unit after the syllable is the word; after the word, the phrase. Something has already been said in the previous paragraphs on word and phrase rhythm: it remains to examine them more closely.

Words vary in length from one to eight or even ten syllables; and the accents (main and secondary) may fall on any of these syllables according to the origin and historical development of the word—thus words of two syllables: ápple, alóne; of three syllables: béautiful, accéssion, appercéive; of four syllables: ápoplexy, matérial, evolútion, interreláte. But generally in polysyllables the tendency to rhythmic alternation of stress produces one or more secondary accents more or less distinctly felt; thus on the first syllable of apperceive and on the third of apoplexy there is an obvious secondary accent; on the third syllable of beautiful and the fourth of material there are potential accents, not regularly felt as such but capable, under certain circumstances, of rhythmic stress. For example, in the phrase 'beautiful clothes' there is no accent and no stress on -ful; but in 'beautiful attire' the syllable -ful receives a very slight accent (properly not recognized [38]by the dictionaries) which can well serve as a weak rhythmic stress. Long words illustrate the same principle: antitranssubstantionalistic, pseudomonocotyledonous, perfectibiliarianism. This potential stress is of the utmost importance in verse—as when Milton out of three words, two of which have no recognized secondary accent, makes a 5-stress line:

The result of this tendency to alternation, or in other words of the difficulty of pronouncing more than three consecutive syllables without introducing a secondary accent or stress, is that English phrases fall naturally into four rhythmic patterns or movements (and their combinations): 1. accent + no-accent (a. one syllable, b. two syllables); 2. no-accent (a. one syllable, b. two syllables) + accent. Examples: 1a beauty, 1b beautiful, 2a relate, 2b intercede. These four movements are variously named: the first two are called falling, the second two rising; 1a and 2a are called duple or dissyllabic, 1b and 2b triple or trisyllabic; 1a is called trochaic, 1b dactylic, 2a iambic, 2b anapestic (after the names of the metrical feet in classical prosody). Beauty, by this usage, is a trochee, beautiful a dactyl, relate an iamb, intercede an anapest. But these patterns alone are by no means sufficient to explain or register all the phrasal movements of English prose—as a single sentence will show.

[39]Here the first phrase is in falling rhythm, the second (probably) in rising rhythm, the third is—rising or falling? To some readers it will appear of one sort, to others of another. The fourth phrase is probably rising, the fifth doubtful, the sixth falling, the seventh probably rising. To say that the first phrase is made up of a dactyl and two trochees means very little. The primary fact to be recognized and understood is that these four patterns exist in English speech not as absolute entities but as tendencies. In prose they are discontinuous, irregularly alternating, often hardly perceptible; but they are there as potential forces whose latent effects are brought out by regular metre.

Another problem at once obvious is to determine the limits of a phrase. Some readers will feel "to fortune" in the above sentence as a separate phrase, others will join it to the three words that precede. No rules can be laid down. Two tentative but useful criteria are possible, however. A phrase may be regarded as purely musical, a group of sounds that either by their own nature or by their possibility of utterance in a single expulsion of breath seem to belong together. But this is an uncertain criterion, since we separate the sounds of words with great difficulty from their meaning, and the periods of breathing are subject to arbitrary control. And some phrases are uttered in much less than the time required in normal breathing. The other criterion, sometimes supporting sometimes contradicting the former, is the logical content of words. But this also is uncertain, since logical content ought to hold subject and verb together, whereas in the example [40]above it clearly does not. And neither breath grouping nor logical grouping will enable us to determine whether "either of virtue or of mischief" is two phrases or one.

The limits of the sentence, with its clauses, are, largely through the modern conventions of printing, more distinctly felt and observed. But its rhythm is none the less complex. For it is not only the sum of the smaller rhythmic movements of word and phrase and clause, but forms a new entity of itself, created by the union of the lesser elements—just as a building is more than its component bricks, stones, and timbers.

Composite Speech Rhythm. Such, briefly described, are the rhythmic elements of spoken English prose. When only small sections are analyzed singly, it is possible to understand something, at least, of the intricate pattern of forces which are interwoven in the rhythms of ordinary language. When one undertakes to analyze and express the combined rhythms—musical, logical, emotional—of connected sentences and paragraphs, one finds no system of notation adequate; the melodies and harmonies disappear in the process of being explained. Those who wish to enjoy to the fullest the rhythmic beauties of English prose must patiently scrutinize the smallest details, then study the details in larger and still larger combinations—the balance and contrast of phrases, the alternation of dependent and independent clauses, the varieties of long and short sentences, of simple, compound, periodic sentences—and finally endeavor to rejoin the parts into a complete [41]whole. To pursue the subject further would be to encroach upon the domain of formal rhetoric and would be out of place here. The best counsel is the old counsel: try to understand and feel the great passages of the great prose masters. A few examples have been given on pages 25 ff., above; they should be studied diligently.

Prose and Verse Rhythm. It is but a short step from the occasional regularity of rhythm in the passages on pages 27-29 to the deliberately continuous regularity of verse. A tendency to rhythmic flow, it has already been shown, is inherent in ordinary language. When the words are made to convey heightened emotion this tendency is increased, and "the deeper the feeling, the more characteristic and decided the rhythm" (John Stuart Mill). Then, as Coleridge says, "the wheels take fire from the mere rapidity of their motion," and finally we have

Intensified, regularized rhythm is reciprocally both a result of impassioned feeling and a cause of it: hence its double function in poetry. It springs, on the one hand, from "the high spiritual instinct of the human being impelling us to seek unity by harmonious adjustment and thus establishing the principle that all the parts of an organized whole must be assimilated to the more important and essential parts." On the other hand, it "resembles (if the aptness of the simile may excuse its meanness) yeast, worthless or disagreeable [42]by itself, but giving vivacity and spirit to the liquor with which it is proportionally combined."[19]

The question is as old as Aristotle, whether metre, that is, regularized rhythm, is an inalienable and necessary concomitant of poetry. The answer rests on a precise understanding of terms; for the right antithesis, so far as there is one, is not between prose and poetry, but between prose and verse. High and passionate thoughts, true poetical feeling and expression may and do exist in prose, but their most natural and characteristic expression is in verse. The old question has been lately reopened, however, by the anomalous form called 'free-verse.' Only the name is new; the thing itself is, at its best, but a carefully rhythmed prose printed in a new shape: an effort to combine in an effective union some of the characteristics of spatial rhythm with the established temporal rhythms of language. Free-verse will be discussed more fully on a later page; it is mentioned here because it is a natural transition between prose and verse, claiming as it does the freedom of the one and the powers of the other.

[43]

Another means of recognizing the close relations of verse and prose is to try to determine which of several passages of similarly heightened emotion, printed in the same form, was originally verse and which prose.

On the other hand, it is worth observing what effect metrical arrangement has upon the emotional quality and power of words and phrases. Hardly anyone would, perhaps, find the following passages strikingly melodious:

But turn these words back to their original metrical order, and it is almost a miracle performed. One recalls Coleridge's definition of poetry as the best words in the best places.

It should now be clear that prose and verse are not so antithetical as is often supposed; that they are only different forms of the same substance, language; two branches from the same root. At certain points they overlap and are practically one; at other points the divergence is obvious but not great; and even in their extreme differences the common basis of the rhythms is the same. In both prose and verse are the same relations of time, stress, and pitch, except that in verse the arrangement and order of them are according to a perceptible pattern. Verse is but prose fitted over a framework of metre. Herein lies the whole art of versification, the whole psychology of poetic rhythm, the whole problem of metrical study and investigation.

[46]

We must always remember that "a line of verse is a portion of speech-material with all its phonetic features (corresponding to its ethos as well as its logos) adjusted, without violence, to a fixed and definite metrical scheme. The two entities, metrical scheme and portion of speech-material adjusted thereto, are distinct and the chief study of the metricist is the manner of adjustment of the latter to the former, the way in which a suitable portion of phonetic liquid is chosen and poured into metrical bottles."[21] Only after having grasped what can be grasped of the subtleties of prose rhythm, and having learned the common forms and patterns of metre, can we put the two together, recognize their new unity, perceive the new rhythmic beauties, harmonies, modulations that spring from their mutual adjustment.

A word may be added here, though the subject is one rather of æsthetics than of prosody, on the function of metre in emphasizing and reinforcing the beauties of thought, emotion, and expression that poetry offers. Two practical illustrations have just been given above. Every writer on poetics, from Aristotle down, has had something to contribute, but the substance of it all may be found in the eighteenth chapter of Coleridge's Biographia Literaria, from which a few sentences have already been quoted.[22] It is not merely that verse by its external appearance notifies the reader, or by its [47]perceptible regularity notifies the listener, that the writer is putting forth his highest efforts, that language is being driven to its highest possibilities; it is not that the use of verse signalizes greater aims and intentions than the use of prose; but rather that the higher efforts, the greater aims, turn by a natural, spontaneous, but partly mysterious instinct to metrical forms for adequate or fit expression. The poets themselves have proved this. No one, barring a few notable exceptions, who felt the creative powers of poetry within him has dared neglect or refuse the added difficulties and the potential beauties of metre. Not the sense of obstacles overcome, but of possibilities realized prompts to formal rhythms. Music, in Dryden's phrase, is inarticulate poetry; but poetry, while it remains articulate and endeavors to accomplish its own destinies, will always approach as close as its own conditions permit to the powers of music. Some poets are inclined more powerfully to music than others. Burns composed with definite melodies in mind; Shelley often began with a little tune which he gradually crystallized into words; Schiller tells us that inspiration often came to him first in the form of music. Tennyson, Swinburne, and others, have chanted rather than read their poetry aloud. And even Browning, who sometimes appears to prefer discord to music, is found to have studied not only the science of music, but also the musical effectiveness of words.

While it is unquestionably going too far to insist as Hegel does that "metre is the first and only condition absolutely demanded by poetry, yea even more necessary [48]than a figurative picturesque diction"; or even to say that the finest poetry is always metrical; still it remains a simple fundamental truth that metre is the natural form of poetic language. The great exceptions to this—the poetic prose of a Sir Thomas Browne, a Pater, a Carlyle, or the free-verse of Whitman—do but prove its soundness; for we always feel them to be something exceptional, something not quite natural though not quite amiss, something wonderful, like tours de force. We would not wish them otherwise, perhaps; but we should doubt them if we did not actually have them before us.

[49]

Elements of verse rhythm. The simplest metrical unit is the syllable; the next higher unit is the foot, a group of syllables; the next higher unit the line, a group of feet; then the stanza or strophe.

In some prosodies—as the French and Italian, for example—the standard unit of verse is the syllable. The first essential of a line is that it have a certain number of syllables; the accents or stresses may, theoretically at least, fall anywhere in the line. In English verse also the syllable has sometimes been regarded as the unit, but for the most part only by a few poets and prosodists of the late sixteenth, the seventeenth, and eighteenth centuries.

The foot corresponds in English verse to what has been described in Chapter I as the rhythmic unit of all rhythms, namely that which recurs in regular sequence. It comprises, therefore, a point of emphasis and all that occupies the time-distance between that point of emphasis and the following one. In other words, a foot is a section of speech-rhythm containing a stressed element and an unstressed element, usually one or two unaccented syllables. So much is clear and undisputed in theory. But there are few single topics on which writers on English prosody are so much at variance as [50]on the further, more accurate definition of the foot. One of the main sources of difficulty, however, is easily removed. The metrical foot is not a natural division of language, like the word or the phrase, but an arbitrary division, like the bar in music, an abstraction having no existence independent of the larger rhythm of which it is a part. The analogy between the metrical foot and the musical bar is very close: they are both artificial sections of rhythm which either in whole or in part may be grouped into such phrases as the ideas or melodies may require.[23] They may be isolated and treated by themselves only for the purposes of analysis, for they are merely theoretical entities, like the chemical elements. There is no reason, therefore, that the foot should correspond with word divisions, no objection to the falling of different syllables of one word into different feet. Thus in Gray's line

both curfew and parting are divided.[24] Further, the division between clauses may fall in the middle of a foot, as in Wordsworth's lines

[51]

But another difficulty remains, which is apparent in the second line just quoted from Wordsworth. The general rhythm of the whole sonnet of which these two lines are the beginning is plainly duple rising, or iambic. The first line and the latter part of the second are easily divisible into iambs; but how shall Getting and spend- be divided? Clearly and spend- is an iamb, but Getting is not. Can trochees and iambs occur together in the same line without either obscuring or actually destroying the rhythm? The simpler solution would be to keep the whole line in rising rhythm by regarding -ing and spend- as the second foot and ‸ Gett- as the first. (The sign ‸ indicates a missing syllable or musical rest. See below, page 63.)

The most common feet are the iamb, the trochee, the anapest, and the dactyl (see above, page 38), to which may be added the spondee. The names are borrowed, not quite felicitously, from classical prosody. Various symbols are in use:

Classical prosody distinguished several other feet, some of which are occasionally mentioned in treatises on English verse: amphibrach ◡_◡, tribrach ◡◡◡, pyrrhic ◡◡, paeon _◡◡◡, choriamb _◡◡_.

The objection to the use of these classical terms is not so serious as is frequently supposed. Since Greek and [52]Latin prosody was primarily quantitative, that is, based upon syllabic length, and every long syllable was theoretically equal to two short syllables, an iamb or ◡- had the musical value of ♪♩, a trochee of ♩♪, a dactyl of ♩♪♪, etc. And since no such definite musical valuation can be given to English feet, a Greek iamb and an English iamb are obviously different. But after all there was inevitably an element of stress in the classical feet, and there is a very positive element of time in the English, so that the difference is not so great, and no confusion need result once the facts are recognized. Another set of terms, however, borrowed from the Greek and Latin is open to more grave objection, for no real equivalence exists between the classical and the modern phenomena. The iambic trimeter in Greek consists of three dipodies or six iambs; as used by English prosodists it consists of three iambs. The Greek trochaic tetrameter, similarly, contains eight trochees, the English 'trochaic tetrameter' but four. The common term iambic pentameter is not so objectionable, but is to be rejected because of its similarity to the others, which are actually confusing.

The next larger metrical unit after the foot is the line or verse. It is distinguished (1) mechanically by the custom of printing, (2) phonetically by the pause usual at the end, and (3) structurally by its use as a unit in forming the stanza. Lines are of one, two, three, or more feet, according to the metrical form used by the poet (see Chapter IV). In rimed verse the end of the line is so emphasized that the line itself stands out as a very perceptible rhythmic unit; in unrimed verse, [53]however, the line is frequently not felt as a unit at all, but is so interwoven with the natural prose rhythm of the words as to be almost indistinguishable to the ear, though of course visible to the eye on the printed page. This fact is easily apparent in reading the second, fifth, and sixth illustrative selections on pages 43, 44.

The stanza or strophe is a combination of two or more lines of the same or varying lengths, according to a regular pattern chosen by the poet. 'Irregular' stanzas sometimes occur, in which the thought rhythm is said to control and determine the stanzaic rhythm; that is, the length of line and position of rimes are regulated by the logical and emotional content of the words. On the various kinds of stanzaic structure, see pages 88 ff., below.

Metrical Patterns. It must be fully understood that these metrical patterns of line and stanza are purely formal. They are the bottles into which the poet pours his liquid meaning, or better, the sketched-in squares over which the painter, copying from an old masterpiece, draws and paints his figures. They have no literal or concrete existence. They are no more the music of verse than

is the music of a waltz. They are absolutely fixed and predetermined (though the poet may invent new patterns if he chooses). But he uses them only as forms on which he arranges his words and phrases. For the rhythm of language is extremely soft and malleable: [54]by skilful handling it can be moulded into an infinite variety of shapes. Perhaps the comparison of a stanza by John Donne with a stanza by W.B. Yeats, both based on the same metrical scheme, will help to make this clear. The formal scheme is

Even more striking is the difference of rhythmical effect observable in reading, one after the other, a page of Pope's heroic couplets in the Essay on Man, of Keats's same couplets in Endymion, and Browning's same couplets in My Last Duchess.

While the formal pattern remains fixed and inflexible, over its surface may be embroidered variations of almost illimitable subtlety and change; but always the formal pattern must be visible, audible. The poet's skill lies largely in preserving a balance of the artistic principles of variety in uniformity and uniformity in variety. Once he lets go the design, he loses his metrical rhythm and writes mere prose. Once we cease to hear and feel the faint regular beating of the metronome we fail to get the enjoyment of sound that it is the proper function of metre to give. On the other [55]hand, if the mechanical design stands out too plainly, if the beat of the metronome becomes for an instant more prominent than the music of the words, then also the artistic pleasure is gone, for too much uniformity is as deadly to art as too much variety.

These verses are regular (as is appropriate for the theme), and vary comparatively little from the formal metrical pattern. The coincidence of prose rhythm and metrical rhythm is almost complete. Yet by means of small subtleties of variation in pause, word order, long and short syllables, Gray always saves the poem from monotony. How far the variations may be carried, how much the ear may be depended upon for rhythmic substitution and syncopation, is determined by many things. Certain lines are unmistakably metrical to all ears and in all positions—such as these verses of Gray's Elegy. Certain lines are generally felt to contain daring variations and yet be successful and effective—such as

Other lines stretch our metrical sense to the breaking point, and according to individual taste we judge them bold or too bold—such as Tennyson's

[56]

or Milton's

In all of these examples the metrical pattern is the same: five consecutive iambs. The modifications illustrate plainly the extraordinary flexibility of language.

Time and Stress. Probably the most disputed point in all prosodic theory is the relative importance of time (duration, syllabic length) and stress (accent) in English verse. Some writers have attempted to explain all the phenomena entirely by stress; others entirely by time. Neither side, of course, has been very successful.[25] The difficulty is partly one of theory and partly one of correct analysis of the facts. Thanks, now, to the attention paid in recent decades by the experimental psychologists to rhythm and metre, we are in a position to reach at least approximate clearness on this vexed point. Since the older theorists have mostly started either from the traditional conceptions of classical prosody or from examination of but a part of the phenomena, their work may be left out of account here. Certainly no great blame attaches to them; they are the Bacons and Harveys and Newtons of metrical science. A more nearly correct analysis of the facts is possible now because with the minutely accurate [57]instruments of the scientists to aid us we need no longer trust to the uncertainties of perception and statement of separate individuals. Of course no one today holds the extreme belief that science explains everything; and of course the scientific experiments on the nature and effect of rhythm must have a starting point in the personal equations of those who have submitted themselves to the scientific tests. With all its patience and thoroughness of investigation, experimental psychology is only now establishing itself. But it does offer, on this one mooted point of versification, invaluable help.

The theory presented in the previous pages states that sound rhythm consists of a succession of points of emphasis separated by equal time divisions. This is the ideal rhythm. When subjected to the conditions of metrical language it suffers two alterations. In the first place, our notions of time are extremely untrustworthy. Days vanish in a moment and they drag like years. Very few of us can estimate correctly the passage of five minutes: syllables are uttered in a few hundredths of a second. We are satisfied with the accuracy shown by an orchestra in keeping time; but if we took a metronome to the concert we should find the orchestra very deficient in its sense of time. The fact is that the orchestra knows better than the metronome, that perfectly accurate time intervals become unpleasantly monotonous, that we rebel at 'mechanical' music. Thus the time divisions of pleasurable rhythm are not mathematically equal, nor even necessarily approximately equal, but are such as are felt to be equal. [58]The second alteration of ideal rhythm is that which results from the conformity of fluid language to its metrical mould. This metrical scheme, based theoretically on equal time units marked by equal stresses, becomes a compromise of uneven stresses and apparently equal time divisions.

Almost every line of verse is a proof of this: both the fact and the explanation are clear when approached from the right angle, and may be tested by carefully prepared statistics. In the following examples the figures beneath each syllable give the time of utterance in tenths and one-hundredths of a second; the figures in parentheses represent pauses.[26] The first, from Paradise Lost, II, 604-614, is in blank verse, with five iambic feet to a line; the second, from Shelley's The Cloud, is apparently irregular, but the basis is clearly anapestic. The ideal rhythm or metrical pattern of the first is

regularly repeated. The ideal rhythm of the second is

six times repeated.[27]

[59]

[60]

Two facts emerge from these statistics at once: (1) that in about 90 per cent of the feet the ◡ or unstressed element is shorter than the _̷ or stressed element, or, in other words, stress and syllabic length nearly always coincide; and (2) that while there is very great variation in the absolute lengths of short syllables and long syllables, the proportion of average lengths is about 2:4.[28] One need not suppose that the conscious mind always hears or thinks it hears the syllables pronounced with these quantitative proportions. Though we deceive ourselves very readily in the matter of time, it is not true that we have no sense of duration whatever. Quite the contrary. Our cerebral metronome is set when we read verse for about .6 seconds for a foot (.2 seconds for the unstressed element;.4 seconds for the stressed element). If we read faster or more slowly the proportions remain the same. When, however, in Paradise Lost, II, 607,

[61]

the normal proportions are so patently departed from that the theoretically unstressed syllable small is actually longer than the theoretically stressed syllable drop, and the foot small drop takes 1. second, or 2/5 longer than the average foot beside it (with one, .61 seconds)—when divergences so great as this are both possible and pleasurable, the conclusion should be, not that the ear makes no recognition of the time, but that it is capable, by syncopation and substitution, of adjusting itself to a very great possibility of variation without losing hold of the rhythmic pattern. Looked at from one point of view, the extreme variations would appear to be irregularities and warrant the judgment that no element of duration exists as a principle of English verse; but from the right point of view these variations mean only that the metrical time unit is extraordinarily elastic while still remaining a unit; that the ear is willing and able to pay very high for the variety in uniformity which it requires.

Pause. The time element of English verse is affected also by different kinds of pauses. Three kinds may be distinguished, two of which belong properly to prose rhythm as well. (1) The logical pause is that cessation of sound which separates the logical components of speech. It helps hold together the members of a unit and separates the units from each other, and never occurs unless a break in the meaning is possible. It is usually indicated in printed language by punctuation. (2) The rhythmical pause separates the breath groups of a sentence and therefore concerns language chiefly as a [62]series of sounds independent for the most part of logical content or symbolism. Though its origin is primarily physiological, it soon induces a psychological state and results in an overuse or overdevelopment of the cerebral metronome. Both readers and writers get into a certain 'swing' which turns to monotony and sing-song in reading and to excessive uniformity of sentence length and structure in writing—what is called a jog-trot style. This pause as it affects the reading of verse is only slightly dependent upon the logical content of words, for it takes its pace, especially in rimed verse, from the normal line length, and tends to make every line sound like every other, regardless of the meaning. (3) Metrical pause is primarily independent of the other two, but most frequently falls in with them. It belongs to the formal metrical pattern, and serves usually to mark off the line units. There is thus theoretically a pause at the end of every line, and a greater pause at the end of every stanza. When verses are 'run on,' i. e., when there is no logical pause at the end, many readers omit the metrical pause or reduce it to a minimum. Others, whose rhythmic sense is very keen, preserve it, making it very slight but still perceptible. The metrical pause is greatly emphasized by rime.

There are two other time elements in English verse, related in different ways to each of these three pauses, one which is nearly equivalent to the musical rest; the other which is nearly equivalent to the musical hold. The latter is common to both verse and prose, and is emotional or elocutionary in origin; "If....," "Well——?" "'These roses?' she drawled." In verse [63]it often coincides with and supports a metrical pause, especially on rime words. Many readers in fact combine the hold and the metrical pause or use them interchangeably. The former, the rest, is a pause used to take the place of an unstressed element. As such, however, it does not altogether compensate the break in the normal time-space, but fills in the omission sufficiently to preserve the rhythm of the verse.

These various pauses are all well illustrated in Tennyson's lyric, Break, Break, Break.

Logical pauses occur at the end of ll. 2, 4, 6, 8; and probably after stones in l. 2. After stones there would be also a rhythmic pause, but it is reinforced and practically replaced by the logical pause. Another rhythmic pause might occur after tongue in l. 3, but it is absorbed partly by the length of tongue and partly by the necessity of preserving the line rhythm through utter. It will be felt, however, if the lines are read thus:

[64]

The metrical pause appears clearly after utter in l. 3. The pauses after boy (l. 5) and lad (l. 7) are both metrical and logical. The hold is illustrated by O in l. 5 and l. 7. [29] The rest appears distinctly in l. 1. From reading the whole poem we know that the movement is anapestic. The pattern rhythm for the first line would be

The number of syllables is three, whereas the other lines have from seven to nine syllables each. That is, before each break two light syllables, or their time equivalent, are lacking, their place being supplied by the rest-pause (which is also logical and emotional).[30]

The reader may analyze the comparative lengths of foot, line, pause, and rest in the following record:[31]

[65]

Pitch. Pitch appears to be sometimes a determining element in rhythm, as has been shown above; but since its chief function in verse is that of supporting the recognized determinants and adding grace-notes to the music, it is omitted here and discussed in Chapter V, below.

Balance of Forces. It is not to be inferred from the foregoing sections that the basis of English metre is time. For the basis of English metre is dual: time and stress are inextricable. Beneath all metrical language runs the invisible current of time, but the surface is marked by stress. The warp of the metrical fabric is time; stress is the woof. And from the surface, of course, only the woof is visible. Moreover, the poet's point of view in composing and generally the reader's [66]point of view in reading has always been that of the 'stresser.' No poet ever wrote to a metronome accompaniment; extremely few readers are fully conscious—few can be, from the nature of our human sense of time—of the temporal rhythm that underlies verse. Thus it has come about, historically, that modern English verse is written and regarded as a matter of stress only, because to the superficial view stress is predominant.[32] Probably the truth is that most poets compose verse with the ideal metrical scheme definitely in mind and trust (as they well may) to their rhythmical instinct for the rest. Whatever device they employ for keeping the pattern always before them, they do keep it distinctly before them—except perhaps in the simpler measures which run easily in the ear—and build from it as from a scaffolding. They may not know and may not need to know that this metrical scheme does itself involve equal time units as well as equal stresses. They vary and modulate both time and stress according to the thought and feeling the words are asked to express. And though it is a point on which no one can have a dogmatic opinion, [67]one inclines to the belief that usually the finest adaptations of ideas and words to metre are spontaneous and intuitive. Skill is the result of habit and training, and metrical skill like any other; but there is also the faculty divine. One is suspicious of the

for when we can see how the trick is done we lose the true thrill.

It would be absurd to imagine a prosody which was independent of its own materials. It would be absurd therefore not to find in all language the elements out of which verse is made. Indeed, M. Jourdain, having recovered from his first shock on learning that he had actually been talking prose, must prepare for a second: that he has actually been talking potential verse. The three acoustic properties of speech—duration, intensity, pitch—modified by the logical and emotional content of which the sounds are symbolic, combine to produce an incredibly subtle and elastic medium which the poet moulds to his metrical form. In this process of moulding and adjustment, each element, under the poet's deft handling, yields somewhat to the other, the natural rhythm of language and the formal rhythm of metre; and the result is a delicate, exquisite compromise. When we attempt to analyze it, its finer secrets defy us, but the chief fundamental principles we can discover, and their more significant manifestations we can isolate and learn to know. In all the arts there is a point at which technique merges with idea and conceals the heart of its mystery. The greatest [68]poetry is not always clearly dependent upon metrical power, but it is rarely divorced from it. No one would venture to say how much the metre has to do with the beauty of the

[69]

Line Length. A line of English verse may contain from one to eight feet. Theoretically, of course, more than eight feet would be possible; but just as there are sounds which the human ear cannot hear and colors which the eye cannot see, so there appears to be a limit beyond which we do not recognize the line as a unit. The most frequently used lines are of four and five feet, most conveniently called, respectively, 4-stress and 5-stress lines;[33] those of one, two, and three feet tend to become jerky, those of more than five to break up into smaller units.

Line Movement. The movement of a line is determined primarily by the foot of which it is composed. It is iambic, trochaic, anapestic, dactylic, according as the metrical pattern is made up of iambs, trochees, etc. Thus

is plainly iambic.

is plainly trochaic.

is plainly anapestic.

is plainly dactylic.

But very few poems conform exactly to the metrical pattern. For example, Blake's

seems clearly to be trochaic; yet the last trochee of each line lacks its unstressed element, and the fourth line has an extra-metrical syllable, Could. By itself the fourth line would be called iambic: in this context it is [71]called trochaic with 'anacrusis,' i. e., with one or more extra-metrical syllables at the beginning.[34] Or again in Clough's stanza,

the movement is clearly iambic, yet the first and third lines have an extra-metrical syllable at the end. This is called 'feminine ending.'

Moreover, sometimes the word or phrase rhythm clashes with the metrical rhythm and makes the resultant seem doubtful. Thus

are unmistakably iambic, and Wordsworth's

is unmistakably trochaic; but in Tennyson's

there are metrically five iambs in each line, but also in each four words that are trochaic. The result is a conflict [72]of rhythms, a kind of syncopation, which produces a very pleasing variant of the formal rhythm.

Furthermore, in a passage like the following, which everyone recognizes as exquisitely musical, it is not obvious whether the rhythm is iambic or anapestic or trochaic.

If the first two syllables be regarded as anacrusis, the first line would be trochaic, with a dactyl substituted for a trochee in the second foot. The third line is apparently trochaic. But only three lines of the eight have a feminine or trochaic ending, and all except the third have iambic or rising rhythm in the first foot; so that it is more simple and natural to consider the last syllable of the first, third, and seventh lines as extra-metrical, and call the rhythm iambic-anapestic, or rising. Since the ◡_̷ and ◡◡_̷ are both rising rhythm they may be readily substituted one for the other—the appearance of equal time values being preserved—without disturbing the musical flow of sounds. Thus of the thirty-two feet in the eight lines, seventeen are iambs and eleven anapests, two are weak iambs (-orous, -ylus), one a spondee (bright night-), and one monosyllabic with a rest (‸ Fills). Tennyson's Vastness may [73]also be studied for its combinations of trochees, dactyls, and spondees. Here is one stanza:

Similar combinations, still freer, with frequent anacrusis as well, are characteristic of Swinburne's Hesperia; e. g.—

The first line of a poem is not always a good criterion of the metre of the whole poem—though Poe declared that it should be. For Tennyson's The Higher Pantheism is chiefly in triple falling rhythm, but it begins

The first stanza of Campbell's famous Battle of the Baltic runs:

[74]

Here the first line might be 3-stress or 2-stress; the second, third, fourth, sixth, seventh, eighth might have three stresses or four; the fifth five or six; the ninth two or one. It is not, in fact, until we reach the

of the fourth stanza that we are sure how the poem ought to be read. But Campbell was not a faultless artist. There is the same metrical ambiguity, however, in Tennyson's

until the second line shows us we should read it with three stresses, not four. There is a curious verse in Gay's Beggar's Opera which well illustrates the necessity of consulting the context to determine the pattern, for it can, taken by itself, be scanned in three different ways:

viz., ◡_̷◡◡_̷◡_̷◡_̷◡ or ◡_̷◡◡◡_̷◡_̷◡ or ◡_̷◡◡_̷◡◡◡_̷◡.

But sometimes it is difficult, if not impossible, to say whether a line or series of lines is in rising or falling rhythm, or what sort of foot is predominant—in other words, what is the formal metrical pattern. This difficulty is, of course, no fault of the poet's: it lies in the complexity of the phenomena, and is after all a weakness of our power of analysis. In the spectrum blue merges into green, red into yellow, and though we invent names for various tints, others still escape classification. And just as some verses combine iambic and [75]anapestic (rising), or dactylic and trochaic (falling) movements, so others combine rising and falling rhythms. For example,

This may be trochaic with anacrusis or iambic with feminine endings, but neither quite adequately describes it. Is Shelley's To Night prevailingly iambic or trochaic? All of the twenty-five long lines end with an iamb, but only eleven begin with rising rhythm (thirteen begin with falling or trochaic rhythm, and one is ambiguous). Two of the short lines are definitely iambic, the other eight are doubtful, but apparently trochaic. If it is read as iambic, eleven of the hundred feet in the long lines will be 'irregular'; if it is read as trochaic, eleven likewise will be 'irregular.' Milton's L'Allegro and Il Penseroso contain lines that are purely iambic, as

some that are purely trochaic, as

and others which are a combination, as

Again, how shall the following stanza from F. W. H. Myers's Saint Paul be classified?

[76]

The metrical scheme appears to be

that is, 5-stress trochaic, with dactylic substitution in the first foot and truncation or catalexis of the last foot in the second and fourth lines; or perhaps iambic, with anapestic substitution in the second foot and a feminine ending in the first and third lines. But when many of these stanzas are read in succession, the movement is found to be

that is, 4-stress falling rhythm, with intermixed duple, triple, and quadruple time.

This introduces a new question, whether English verse admits of a foot resembling the Greek paeon, _̷◡◡◡. The answer seems to be that theoretically it does not, but practically it does.[35] It would, doubtless, [77]be more accurate to describe the foot as _̷◡◡`[A]◡, for some stress, however slight, is regularly felt on the third syllable. But the poets have had their way, and written what certainly try to be paeonic feet. Thus Macaulay's The Battle of Naseby begins:

And Mr. Kipling's The Last Chantey:

And Mr. E. A. Robinson's The Valley of the Shadow is in this same rhythm, the first four lines being almost perfectly regular:

Some have read Browning's A Toccata of Galuppi's to the same tune, but at grave risk of destroying the music.

[78]

Rightly described, this movement is a discontinuous syncopation of fours and twos; the prevailing formal unit is _̷◡◡◡, but it is varied now by _̷◡_̷◡, and now by simply _̷◡, with the usual substitution of _̷◡◡ for _̷◡. It is an excellent exercise to analyze Jean Ingelow's Like a Laverock in the Lift and observe the pauses, holds, and substitutions. The most notable are _̷_̷◡ for _̷◡◡◡ (we too, it's), and _̷◡_̷ (lass, my love, l. 5; thou art mine, l. 6; missed the mark, l. 7, etc.). The third line may be read

or

The former seems preferable.[37]

[79]

How musical and effective this rhythm is, judgments will differ. It is clearly capable of great variety, but the large proportion of light syllables forces heavier stress on some of the accents, and the number of naturally heavy syllables which do not coincide with the metrical stress is excessive; and the almost inevitable result is a thumping which only the deftest manipulation can avoid.[38]

Probably the most striking and successful use of the 4-beat movement is that of Meredith's Love in a Valley. So marked is the time element, with the compensatory lengthenings and pauses, that the poem almost demands to be chanted rather than read; but when well chanted it is peculiarly musical, and when ill read it is horribly ragged and choppy. The whole poem will repay study for the metrical subtleties, but the first stanza is sufficient to illustrate the rhythm (there are normally four _̷◡◡◡ in each line).[39]

[80]

Examples. There occur examples of 1-, 2-, 3-, 4-, 5-, 6-, 7-, 8-stress iambic, trochaic, anapestic, and dactylic lines, sometimes used continuously and sometimes used in combinations with other lengths. But many of these are unusual, and may be found only by diligent search.[40] Some have already been illustrated in the previous section, others occur here and there throughout this volume, especially in the paragraphs on the stanza; some of the more important, however, are given below. But, of course, the line rhythm is significant mainly as a unit of the longer composition, and brief selections cannot well represent the rhythmic movement of a whole poem. Whenever possible the poem should be read complete.

Attempts have been made to characterize the different feet as slow or rapid, solemn or light, and so on, but they are generally unsuccessful. For though certain measures seem to be inherently unsuitable for dignified themes, or for humorous subjects, there are always contrary instances to be adduced, and it is dangerous to be dogmatic. Anapests are said to be characteristically rapid, hurried, because they crowd more syllables than iambs do into a line; but anapests [81]are often slow-moving, because there is frequent iambic substitution and because many important words—monosyllables, for the most part—have to do duty for light syllables metrically. Perfect anapests, like perfect dactyls, are comparatively few in English.

Two-stress and 6-stress anapestic:

Three-stress anapestic:

[82]

Four-stress anapestic:

Five-stress anapestic. This is a peculiar metre, usually felt to be choppy and harsh. It has been said that no one can read Browning's Saul and follow both metre and meaning at the same time:

Eight-stress anapestic. This is on the whole the longest line possible in English.[42] It is really a tour de force.

The trochaic line is generally stiff and thumping. It does not admit of frequent substitutions, for many substitutions [83]destroy the trochaic effect. It usually comes to an abrupt close because feminine endings are not easy or natural in English. Moreover, there are in the language so many dissyllabic words of trochaic movement that the resulting frequent coincidence of word and foot tends to produce monotony. Tennyson once said that when he wanted to write a poem that would be popular he wrote in trochaics. Certainly the stresses are more prominent in trochaic verse than in iambic or even anapestic; and the untrained ear likes its rhythms well marked.[43] The Locksley Hall poems are good examples:

Notable is Tennyson's skill in this 8-stress line in avoiding the natural break into 4 + 4. This break occurs regularly and is enforced by the rime in Poe's The Raven. One of the most successful metrically of purely trochaic poems is Browning's One Word More, a few lines of which are quoted on page 70.

Four-stress trochaic.

Five-stress trochaic.

(Note here the substitutions for special imitative effect.)

Shelley's To a Skylark is in trochaic metre of 3-stress and 6-stress lines.

Dactylic lines are not common except in the imitations of the classical hexameter. Hood's familiar Bridge of Sighs in 2-stress lines, and Tennyson's still more familiar Charge of the Light Brigade (which is, however, only partly dactylic) are good illustrations.

Iambic lines are by very far the most frequent in English verse. No special examples need therefore be [85]given except of the less usual 6-stress and 7-stress lines. On blank verse see pages 133 ff.

The 6-stress line is called the alexandrine (probably from the name of an Old French poem in this metre). It is still the standard line in classical French verse; but the French alexandrine differs from the English, principally in having four stresses instead of six. In English it is usually awkward when used for long stretches, and tends to split into 3 + 3. Lowell called it "the droning old alexandrine." It was employed for several long poems in Middle English; and certain of the Elizabethans tried it: Surrey, Sidney, and Drayton—Drayton's Polyolbion (1613) contains about 15,000 alexandrines. It has not commended itself to modern poets, with one exception, for sustained work. Browning wrote his Fifine at the Fair (1872) in this measure; and while he succeeded in relieving it of some of its monotony, he only demonstrated again its unfitness, in English, for continuous use. A peculiar musical effect is obtained from it, however, by Mr. Siegfried Sassoon in his Picture-Show:

On the other hand, as the last line of the Spenserian and similar stanzas the alexandrine has proved very melodious and effective, largely by contrast with the shorter lines. A few isolated examples will illustrate [86]some of its powers, but of course the whole stanza should be read together.

Alexandrines were occasionally in the eighteenth century (and more frequently in the late seventeenth) inserted among heroic couplets for variety and special effect, as in Pope's

[87]

But Pope himself condemned the 'needless alexandrine'

One of the oldest lines of modern English verse is the so-called septenary (septenarius), having had a nearly continuous tradition from the twelfth-century Poema Morale down (in its divided form) to the present. It began as a single line of seven stresses or fourteen syllables, and continued to be used as such through the Elizabethan period, and sporadically even later.[44] But on account of its customary pause after the fourth foot, it very early broke into two short lines of four and three stresses each, and thus the septenary couplet became the ballad stanza. For example,

is essentially the same metre, though printed differently, as

The more notable long poems in septenaries are Warner's Albion's England (1586), Golding's translation of Ovid's Metamorphoses (1565, 1567), and Chapman's translation of the Iliad (1598-1611).

[88]

Couplet. The line unit is used sometimes singly and continuously, as in blank verse, and sometimes in groups usually held together by rime. These groups are called stanzas or strophes. The simplest stanza is, therefore, the couplet rimed aa.[45] Couplets are either unequal or equal in length.

The only much-used unequal couplet is the combination, now old-fashioned, of an alexandrine and a septenary, and called, from the number of syllables, Poulter's Measure, because, says Gascoigne (1575), "it gives xii. for one dozen and xiii. for another." Wyatt and Surrey and Sidney wrote in it; the older drama employed it occasionally; Arthur Brooke's Romeus and Juliet (1562) on which Shakespeare's play was based, is in this measure. The following example is by Nicholas Grimald (1519-62).

[89]

It survives chiefly in the S.M. (short measure) of the hymn books and such stanzas as that used by Macaulay in his Horatius:

Other unequal couplets are found in Herrick's A Thanskgiving to God for his House (a⁴a²) and Browning's Love among the Ruins (a⁶a²).

The equal couplet is used both continuously and, more rarely except with long lines, as a single stanza. Sometimes two or three couplets are combined into a larger stanza. The usual forms of the couplet used continuously are the 4-stress or short couplet ("octosyllabic") and the 5-stress or heroic couplet ("decasyllabic").

Short Couplet. The short couplet in duple iambic-trochaic movement has proved its worth by its long history and the variety of its uses. The English borrowed it from the French octosyllabic verse, and employed it chiefly for long narrative poems. Chaucer used it in his earlier work, the Book of the Duchess, and the House of Fame; Butler in the serio-comic [90]Hudibras; Scott, Byron, Wordsworth, and Morris in their Romantic narrative verse. For lyric purposes it was used by Shakespeare and other dramatists, by Milton in L'Allegro and Il Penseroso, and since then by most of the greater and lesser poets. But its effect, especially in long poems, is often monotonous because of the rapid recurrence of the rimes, and its powers are somewhat limited. Except under expert handling it is likely to turn into a dog-trot, and it seems sometimes to lack dignity where dignity is required. On the whole it is better for swift movement, for the obvious reason that the line is short: the frequent repetition of the unit, both line and couplet, produces the effect of hurry.

Never has the short couplet revealed its flexibility to better advantage than in Milton's L'Allegro and Il Penseroso and in Coleridge's Christabel. In Christabel Coleridge believed he was inventing a new prosodic principle, that of counting the stresses rather than the syllables;[46] and though he erred with respect to the originality of his principle, he succeeded in getting a freer movement than the couplet had had since Chaucer. Some of the roughness of Chaucer's short couplets is probably due to the imperfections of our texts, and some also to the haste with which he wrote—it is in this metre that the fatal facility of certain poets has proved the worst bane—but the Chaucerian couplet stands as a prototype (though not literally a model) of the freer flow of Byron's[47] and Morris's [91]couplets, in contrast to those of Scott and Wordsworth, which resemble the stricter, syllable-counting couplets of Chaucer's friend Gower.

The chief drawbacks of the short couplet, besides monotony, are the tendency to diffuseness of language and looseness of grammatical structure (as in Chaucer and Scott, for instance), and rime-padding, i. e., the insertion of phrases and sometimes even irrelevant ideas, for the sake of the rime.

The chief sources of variety are substitution, pause, run-on lines, and division. The first is very apparent in the much-quoted passage in Christabel:

The pause offers more difficulties for the poet, and more opportunities; since the line is so short, and the rimes reinforce the regular metrical pause at the end of the line, important grammatical pauses cannot well occur in the middle of the line without danger of breaking the rhythm. The logical pause must, therefore, usually coincide with the metrical and thus emphasize unduly the line unit. Moreover, the quick return of the rime sound causes the couplet itself to be felt as a unit and produces what are called 'closed couplets,' in [92]which the two lines contain an independent idea. To avoid irksome uniformity in this regard three devices are customary: to 'run-on' the meaning from one line to the next, thus momentarily obscuring the metrical pause, to 'run-on' the couplets themselves, and to divide the couplet so that the second verse belongs to a new sentence or independent clause.

In this passage, which is on the whole conservative and stiff in movement, observe (1) how the pause in the middle of ll. 4, 13, and 17 helps to vary the measure; (2) how many of the verses end with a logical as well as metrical pause; (3) how in ll. 3, 5,16, and 17 the meaning runs over without pause into the next lines; (4) how the first two couplets and the last two are run together, whereas the third and fourth are both closed and independent; [93]and (5) how at ll. 9 and 10 the couplet is divided. This last device is not very frequent in the practice of any poet except Chaucer; it is well illustrated, however, in these lines from Shelley's With a Guitar to Jane:

Two other means of varying the swing of the short couplet are to change the order of the rimes (as in the example above from Christabel) or introduce a third riming line (that is, to use triplets with the couplets), and to intermingle shorter lines, as Coleridge does occasionally in Christabel, and Byron at the beginning of The Prisoner of Chillon:

Heroic Couplet. The 5-stress line, both rimed and unrimed, is the most flexible and best adapted to all kinds of subjects that English versification possesses. Its powers range through the tragedy and comedy of Shakespeare, the dignity of the sonnet, and the grandeur of Milton, to the satire of Pope and the informal conversational verse of Mr. Robert Frost. The 4-stress line is too short, the 6-stress is too long (when it does not split into two equal parts); the 5-stress seems to hit the golden average. It is less inclined to 'go' by [94]itself, and therefore is suitable for slow movements; on the other hand, it is easily divided by pauses and hence is easily relieved of monotony and adjustable to almost all tempos.[48]

The earliest form, historically, of the 5-stress line in English was in rimed couplets; the first poet to use the rimed couplet continuously (as distinguished from occasional use in a stanza) was Chaucer.[49] Blank verse is a modification of the couplet by the simple omission of the rimes at the end.

The history of the heroic couplet may be divided into two periods, that of Chaucer and his followers, Gavin Douglas and Spenser, and that beginning with Marlowe, Chapman, and other Elizabethans and continuing down to the present. This division is peculiar, for it represents a double curve of development, the one comparatively short, the other long. Chaucer's couplet has all the marks of ease and freedom of a fully matured medium: great variety in the pauses, run-on lines and couplets, and divided couplets. (All the means of securing variety for the short couplet, explained above, apply a fortiori to the heroic line.) Douglas, in large part, and Spenser pretty fully, adopted and preserved this unfettered movement, though the former anticipates here and there the neat [95]balance of the Popian couplet. Then the measure seems to have begun all over again, partly on account of an attack of syllable-counting, with close formal recognition of the line unit and the couplet unit, and gradually worked its way back to its original flexibility.[50]

The following characteristic examples illustrate the chief varieties of the couplet. (Again, they should be supplemented by the reading of longer passages. Pope's couplet, in particular, with its perfection of form according to a few well-marked formulas, reveals its great weakness, monotony, only in the consecutive reading of several pages.)

[97]

Though Chapman sometimes uses the pause and run-on lines freely, the regularity of the foot makes for a certain stiffness and inflexibility.

Donne's metres were notoriously careless—or deliberately irregular. They therefore stand somewhat out of place in the general trend of development.

Waller, though his lifetime (1605-87) embraces that of Milton, is the natural precursor of the eighteenth century. His couplets are almost all characteristic of eighteenth-century couplets, which seem to seek perfection within themselves. The aim of Waller, Dryden, Pope, and Johnson was primarily to exalt the couplet and extract from it all its potentialities, not to obscure [98]it by varied pauses and run-on lines. Waller was praised by the best critics of his own and the following generation for the great 'sweetness' and smoothness of his verse.

It is interesting, from a metrical point of view, to compare Chaucer's couplets with Dryden's where he is translating Chaucer, e. g., in the Knight's Tale and Palamon and Arcite.

Between 1664 and 1678 it became the fashion, partly as a reaction against the liberties of the late Elizabethan blank verse, and partly under French influence, to write drama in heroic couplets. But the undertaking soon proved abortive.

[99]

Pope's couplets represent the acme of polish and metrical dexterity—a perfect instrument for wit and satire.[51] Thus in the mock-heroic Rape of the Lock these well-modeled couplets prove their mettle, but in the translation of Homer their fatal limitations are easily apparent.

The departure from the petrified couplet was gradual and natural, and influenced greatly by the simpler language and content of the verses. These two specimens show Goldsmith writing in two manners, only a few lines apart. Still freer are Cowper's couplets in his On the Receipt of My Mother's Picture. Byron in English Bards and Scotch Reviewers (1809) and Crabbe in his earlier work, still practised the eighteenth-century couplet (in the Tales of the Hall, 1819, Crabbe varied it to a considerable degree), but the new spirit of the Romantic Movement leavened all the metrical forms, as it did the themes, of poetry. Compare the following examples.

Stanzas of three lines riming aaa (called tercets or triplets) are not very common. Familiar, however, is Herrick's Upon Julia's Clothes:

Other examples are: Threnos (in The Phœnix and the Turtle), Herbert's Trinity Sunday, Quarles' Shortness [103]of Life, Browning's A Toccata of Galuppi's, Tennyson's The Two Voices, Swinburne's After a Reading, and Clear the Way; and (with a simple refrain) Cowper's To Mary:

Crashaw's Wishes to his Supposed Mistress rimes a²a³a⁴.

Tennyson's 'O Swallow, Swallow' in The Princess is in unrimed triplets.

On the terza rima see below, page 164.

The most important quatrains are the ballad stanza, riming a⁴b³c⁴b³ or a⁴b³a⁴b³ (the Common Measure of the hymnals), with the related Long Measure riming abab⁴ or abcb⁴; the In Memoriam stanza abba⁴; and the elegiac quatrain abab⁵. These are often combined into 8-and 12-line stanzas, as abab bcbc⁵ (called the Monk's Tale stanza), abab cdcd, etc., sometimes with alternating long and short lines. And these, as well as longer stanzas, are frequently varied by the use of repetitions and refrains.[52]

The ballad stanza, with its frequent variations of internal rime and additional verses is excellently illustrated by Coleridge's Ancient Mariner. Similar is Tennyson's Sir Galahad, a 12-line stanza of three [104]quatrains, a⁴b³a⁴b³cdc⁴d³efgf⁴. Another common variation is that of Hood's The Dream of Eugene Aram, Wilde's Ballad of Reading Gaol, and Rossetti's Blessed Damozel, a⁴b³c⁴b³d⁴b³. The musical roughness of the old ballads should be contrasted with the regularized modern imitations, such as Longfellow's Wreck of the Hesperus. Better imitations are Rossetti's Stratton Water and The King's Tragedy, Robert Buchanan's Judas Iscariot, and W.B. Yeats's Father Gilligan. Sometimes a shorter quatrain is printed as a long couplet and combined into larger stanzas, as in Mr. Alfred Noyes's The Highwayman (which has an additional variation in the inserted fourth and fifth lines):

The variations in Tennyson's The Revenge should be carefully studied.

The ballad stanza is closely similar to the abab⁴ and abcb⁴ quatrains, and (as in the Sir Galahad mentioned just above) the two are sometimes united. All three were much used by Wordsworth and many minor poets for lyrics as well as narratives; the result is often an undignified tinkle that takes the popular ear and "makes the judicious grieve." The stanzaic unit is so easily carried in one's mind and so rapidly repeats itself, that there is little opportunity for the necessary pleasing surprises. But that the measure is capable of a simple expressive music is evident from such examples [105]as Wordsworth's 'Lucy' poems. These stanzas, both alone and doubled (as in To Mary in Heaven), were favorites with Burns.

A striking musical effect was obtained by Swinburne in Dolores by shortening the last line of a double quatrain:

Similar interesting variations are Coleridge's Love, aba⁴b³ and Wordsworth's The Solitary Reaper.

The In Memoriam stanza (abba⁴) is named after Tennyson's poem (though that was by no means its first use), because Tennyson gave it a peculiar melody, and, partly for this reason and partly from the length and subject of the poem, almost preëmpted it for elegiac purposes.[53] Characteristic stanzas metrically are these:

One of the peculiarities of the stanza is the increased emphasis which the rime of the third verse receives from its proximity to that of the second; and this is noticeable both when there is a logical pause after the third verse and when there is none:

Run-on stanzas are very frequent; especially remarkable is the periodic movement of the four stanzas of LXXXVI, leading up to the last line—

"By the rhyme-scheme of the quatrain," says Corson, "the terminal rhyme-emphasis of the stanza is reduced, the second and third verses being the most closely braced by the rhyme. The stanza is thus admirably adapted to the sweet continuity of flow, free from abrupt checks, demanded by the spiritualized sorrow which it bears along. Alternate rhyme would have wrought an entire change in the tone of the poem. To be assured of this, one should read, aloud, of course, all the stanzas whose first and second, or third and fourth, [107]verses admit of being transposed without affecting the sense. By such transposition, the rhymes are made alternate, and the concluding rhymes more emphatic. There are as many as ninety-one such stanzas.... The poem could not have laid hold of so many hearts as it has, had the rhymes been alternate, even if the thought-element had been the same."[54] Examples for this experiment are:

Compare the slightly different effect of the same stanza printed as two lines, in Wilde's The Sphinx:

The name 'elegiac stanza' for the abab⁵ quatrain comes apparently from its appropriate use by Gray in the Elegy Written in a Country Churchyard, but it is not altogether fitting; for it is simply the quatrain movement of the English sonnet, where no lament is intended, and it was employed effectively by Dryden in his Annus Mirabilis, and has been often employed [108]since, without elegiac feeling. For examples see the stanza from Gray, page 55, and the sonnets on pages 129 f. An especially interesting modification is that of Tennyson's Palace of Art, a⁵b⁴a⁵b³.

Five-line stanzas are formed in various ways, e. g., aaaba, aabba, aabab, abbba, ababa, ababb, etc., in lines of three, four, five, etc., stresses.

Six-line stanzas are formed by similar combinations; the most frequent is the quatrain + couplet, called, from Shakespeare's poem, the Venus and Adonis stanza, ababcc⁵ (compare the end of the English sonnet and the ottava rima).[55] Familiar examples are Wordsworth's To a Skylark and his fine Laodamia.

The same rimes with 4-stress verses are also common,[56] for example, Wordsworth's

Another important 6-line stanza is the tail-rime or rime couée, a stanza much used in the Middle English romances and chosen by Chaucer for his parody, Sir Thopas. Harry Bailey, mine host of the Canterbury pilgrims, called it 'doggerel rime.' The simple and probably normal form is aa⁴b³cc⁴b³ or aa⁴b³aa⁴b³, which to save space in the manuscripts was written thus:

Variations are extremely common: the aaa⁴b²ccc⁴b² of Wordsworth's To the Daisy, aaaa⁴b²ccc⁴b³ of Tennyson's Lady of Shalott, aa³b²ccc³b² of S. F. Smith's America, aaa³b²ccc³b² of Drayton's Agincourt, and the so-called Burns stanza, in which Burns wrote some fifty poems, aaa⁴b²a⁴b², e. g., To a Mouse and Address to the Deil.

The most important 7-line stanza is the rime royale or Chaucer (or Troilus) stanza, ababbcc⁵. In the Parlement of Foules, the Man of Law's Tale, and Troilus and Criseyde, Chaucer made it a splendid vehicle both for narrative and for reflective analysis, for humor, satire, description, and all the gamut of emotions; in the fifteenth and sixteenth centuries James I, Lydgate and Hoccleve, Henryson and Dunbar, and Skelton, Hawes and Barclay employed it, [110]largely in imitation of Chaucer; Wyatt used it in his Vixi Puellis Nuper Idoneus; and Shakespeare in The Rape of Lucrece. Since then it has not proved attractive to the poets—though no reason for its disuse is obvious—except Wordsworth (in his translations of Chaucer) and Morris, Chaucer's latest disciple.

In comparison with the formality of Shakespeare's and the evenness of Morris's, the ease and smoothness of Chaucer's stanza are striking. Wyatt's stanzas are musical in their way.

[111]

Eight-line stanzas are variously formed—chiefly by the doubling of quatrains, sometimes with different rimes, as ababcdcd, sometimes preserving one or another or both rimes, as ababbcbc, abcbdbeb, ababacac, abababab, etc. Other varieties are abcdabcd (Rossetti) and aaabcccb (tail-rime), and aabbccdd.

One of the commonest 8-line stanzas is that imported from Italy and called ottava rima, abababcc. It has been charged with tediousness, and tedious it may become if not sedulously varied. It was introduced, along with so much else from Italy, by Wyatt, and was then employed for different purposes by Sidney, Spenser, Daniel, and others.[57] At the close of the eighteenth century it enjoyed a rebirth. "It had already been used by Harrington, Drayton, Fairfax (in his translation of Tasso's Jerusalem Delivered), and ... in later times by Gay; and it had even been used by Frere's contemporary, William Tennant; but to Frere belongs the honour of giving it the special characteristics which Byron afterwards popularized in Beppo and Don Juan.... Byron, taking up the stanza with equal skill and greater genius, filled it with the vigour of his personality, and made it a measure of his own, which it has ever since been hazardous for inferior poets to attempt."[58] Byron had first adopted the stanza in his translation of Pulci's Morgante Maggiore, which is itself in ottava rime. Beppo was written [112]in 1817, and Don Juan begun in the next year. In 1819 the first four cantos of Don Juan were published; in 1820 Keats published his Isabella, and Shelley wrote his Witch of Atlas, both in the same metre.

By far the most important of 9-line stanzas, and one of the finest of all stanzas in English poetry, is the ababbcbc⁵c⁶ invented by Spenser—a double quatrain of [113]5-stress lines plus an alexandrine. This particular octave had been used by Chaucer in the Canterbury Tales, and is sometimes referred to as the Monk's Tale stanza: the stroke of metrical genius lay in adding the 'supplementary harmony' of the alexandrine, by which the whole stanza climbs to a majestic close or ebbs in a delightful decrescendo as the poet wills.[59] The long swing of nine verses on three rimes, with the combined effect of the interwoven rimes (abab and bcbc) united by the couplet in the middle, culminating in the unequal couplet at the close, the extraordinary opportunity of balancing and contrasting the rime sounds, and of almost infinitely varying the pauses—all these render the Spenserian stanza incomparable for nearly every sort of poetic expression.

After the Faerie Queene, the chief poems in this metre are: Shenstone's The Schoolmistress (1742), Thomson's The Castle of Indolence (1748), Burns's [114]The Cotter's Saturday Night (1786), Scott's Don Roderick (1811), Byron's Childe Harold's Pilgrimage (1818 et seq.), Shelley's Laon and Cythna (The Revolt of Islam) (1817, 1818), and Adonais (1821), Keats's Eve of St. Agnes (1820), and the opening of Tennyson's Lotos Eaters (1833).

From the following examples only a limited conception can be gained of the stanza's varied capabilities. Long passages should be read together—and read, for this purpose, with more attention to the sound than to the meaning—in order that the peculiarities of handling of the different poets may be felt.

(Childe Harold begins with many deliberate imitations of Spenser's language and style, but soon neglects them. Here perhaps more than in any other metre the tone and subject of the poem determine the movement of the stanza. The above is but one example of Byron's great variety.)

During the earlier half of the seventeenth century a small group of poets, imitating Spenser both in substance and in external manner, introduced a number of stanzas, some of them not to be admired, whose chief characteristic is the alexandrine for a last line—e. g., abababcc⁵c⁶, ababcc⁵c⁶, ababbcc⁵c⁶, and ababbc⁵c⁶ (which last is that of Milton's On the Death of a Fair Infant, The Passion, and the introduction to On the Morning of Christ's Nativity). Another modification is that of Milton's Ode itself, aa³b⁵cc³b⁵d⁴d⁶. Matthew Prior attempted to improve the Spenserian stanza in his Ode on the Battle of Ramillies by a rime scheme (suggested perhaps by the English sonnet) ababcdcde⁵e⁶—of which Dr. Johnson says: "He has altered the stanza of Spenser, as a house is altered by building another house in its place of a different form." Still farther from the Spenserian original, but probably a development from it, is Shelley's To a Skylark abab³b⁶ (mainly in falling rhythm); and an extension of this last is Swinburne's Hertha (see above, page 81) abab²b⁶ in triple rising rhythm.

[118]

The only English stanza that can be said to rival the Spenserian in artistic merit is the sonnet: but the two are for very different purposes, the one being nearly always used in long, clearly connected series, generally narrative, the other nearly always as an independent poem. Even when sonnets are written in 'sequences,' the relation of the individual sonnets to each other is rarely very close; the unity of the whole sequence (as in Rossetti's House of Life, for example, or Mrs. Browning's Sonnets from the Portuguese) is one merely of general tone and subject. Some of Shakespeare's sonnets are bound together by an intimate unity like stanzas of one poem; others are completely detached. Occasionally a poem is composed of three or four sonnet-stanzas, as Leigh Hunt's The Fish, the Man, and the Spirit, but even then each sonnet remains an independent whole.

The word 'sonnet,' borrowed with the metrical form from Italy in the late sixteenth century,[61] was at first used loosely for almost any short poem on love not obviously a 'song'; but soon the term became restricted to a poem of fourteen 5-stress iambic lines arranged according to one of two definite rime schemes or their modifications. These two rime schemes are [120]the original Italian abba abba cde cde and the English abab cdcd efef gg.

Italian Sonnet. The organization of the subject matter of an Italian sonnet is (at least theoretically) as fixed as that of the rimes. The whole should aim to convey without irrelevant detail a single thought or feeling. The first quatrain, abba, should introduce the subject; the second, abba, should develop it to a certain point, at which a pause occurs; such is the octave. The sestet continues in the first tercet, cde, the thought or feeling in a new direction or from a new point of view, and in the second, cde, brings it to a full conclusion.[62] The rime sounds of the octave and those of the sestet should be harmonious but not closely similar.

It stands to reason that very few poets have enslaved themselves to such an imperious master without assuming certain liberties. Very few sonnets of any poetic value can be found conforming strictly to all these requirements. But the general purport of the formal division may be seen in Christina Rossetti's poignant "Remember"—

The first quatrain says: Remember me when I am gone and we can no longer meet and part as in life. The second quatrain adds: when we can no longer enjoy the companionship of mind, planning what might have been. The sestet continues: Nevertheless, do not let the memory of me become a burden, especially if you ever learn what was in my living thoughts.

Most sonnet writers, while regarding the form as in the abstract something almost sacred, have felt free to mould it in some measure to the immediate demands of their subject—not all, however, with the same success.[63] For the sonnet demands perfection, a single flaw almost cripples it; and few have the absolute command of language necessary to forge a single idea without irrelevance and without omission according to so strict a pattern. Those who are too subservient to the form weaken their poetic thought; those who, like Wordsworth often, are inobedient to the form, produce a poem which is imperfect because it is neither a sonnet [122]nor not a sonnet. Few have come as near the true balance as Milton at his best. "A hundred Poets," says Sir William Watson,

And yet Milton, while preserving the rime scheme, generally disregards the thought divisions, and in half of his sonnets has the pause, not after the eighth line but within the ninth. Commenting on this division Wordsworth says: "Now it has struck me, that this is not done merely to gratify the ear by variety and freedom of sound, but also to aid in giving that pervading sense of intense unity in which the excellence of the sonnet has always seemed to me mainly to consist. Instead of looking at this composition as a piece of architecture, making a whole out of three parts, I have been much in the habit of preferring the image of an orbicular body—a sphere or dew-drop."

Such a close unity can easily be obtained from the Italian sonnet, as hundreds of examples prove,—Milton's On his Blindness is a striking case, with no full stop until the end of the fourteenth line,—but even better for this object is the rime scheme invented by Spenser and used in a hundred and twenty-one sonnets: ababbcbccdcdee. The Spenserian sonnet, however, has found no favor with later poets.

Certain variations in the Italian form are regularly admitted as legitimate. The quatrains must always rime abba, but the sestet may rime cdecde or cdcdcd or [123]cdedce or cdedec, or almost any arrangement of two or three rimes which does not end in a couplet. And even this last caveat is sometimes disregarded by careful sonneteers. A greater liberty is to vary the rimes of the octave to abbaacca. The division of the sestet into two distinct tercets is very rarely maintained; and that of the octave into quatrains is frequently neglected with impunity. Thus the poet adjusts his theme to the strict rules of the sonnet much as he adjusts the natural rhythm of language to the strict forms of metre; the one inescapable requisite being that in neither may he lose hold of the fundamental pattern. But there is this difference, that the sonnet form is extraordinarily firm, and breaks if forced very far from normal. How far one may go can be determined only in special cases, for "the mighty masters are a law unto themselves, and the validity of their legislation will be attested and held against all comers by the splendour of an unchallengeable success" (Pattison).

The early Italian sonnets in English, those of Wyatt, Surrey, and Sidney, are very irregular: Sidney's nearly always end in a couplet and rime the octave abbaabba or abababab or ababbaba. Sometimes he uses such a scheme as ababbababccbcc. Wyatt has one rimed abbaaccacddcee, and Surrey one ababababababaa.

Donne's Holy Sonnets (written about 1617, though not printed till 1633, 1635) were regular in form, and were practically the first English sonnets not concerned with love. Milton followed this tradition, and expanded it to further themes—his only successful poems in lighter mood are sonnets—occasional and political subjects—

[124]

On the formal side Milton handled the sonnet, as has been said, with the freedom of a master.

From the time of Milton's (1642-58) very few sonnets were written in England till towards the end of the eighteenth century. Then the form was revived, under the original impulse of the Wartons in the mid-century, by Bowles, and given a new life by Wordsworth and Keats. In 1850 Mrs. Browning published her Sonnets from the Portuguese, and in 1870 and 1881 Rossetti his sonnet-sequence, The House of Life. The latter contains on the whole the truest representatives of the Italian model.

Here the rime scheme is peculiarly irregular, and the result is hardly a sonnet at all. Shelley's manuscript shows that the poem cost him a great deal of trouble.

English Sonnet. Out of the 'irregularities' and experiments of the early English sonneteers there rapidly developed a new form based on an entirely different principle of division, a series of three quatrains abab, cdcd, efef, followed by a couplet gg. This looser structure, simpler in music and in arrangement of subject matter, soon became a favorite, was used by Surrey and by Sidney, and was adopted by Shakespeare for his hundred and fifty-four sonnets[64]—hence it is [128]sometimes called the Shakespearian sonnet. "With this key," said Wordsworth,

But a sonnet in the stricter sense this 14-line stanza of course is not; for it does not aim to possess the balance, contrast, and functional organization of the Italian stanza. It has qualities of its own, however, which give it its own distinction; and, moreover, it is frankly what many sonnets of the stricter form, without the justification of a difficult and definitely organic structure, are: simply a poem of fourteen lines. For many of Wordsworth's and most of Mrs. Browning's sonnets, though they have the rime-scheme of the Italian, have the simple thought arrangement of the English sonnet.

Not many examples are necessary. Some, like the first two below, preserve the metrical division of the quatrains, with the couplet for an epigrammatic summary; others more or less obscure the division.

Combinations of the two sonnet forms not infrequently occur (as in the last example below), but they are not approved by the critics or the theorists, and generally they miss the excellences of both forms, however successful they may be in other respects.

Besides the stanzas described above, which are but the most familiar or most important of the great variety of regular English stanzas, there are others which, because they are peculiarly constructed or not regularly repeated, may be called Complex. Such are, for example, the 'trailing vine' stanzas of Spenser's Prothalamion (abba⁵a³bcbc⁵dded⁵ee³ff⁵) and Epithalamion (ababc⁵c³dcde⁵e³fggf⁵f⁴hh⁵), and also the simpler ababcde⁵c³de⁵ of Keats' Ode to a Nightingale.

Many of these complex stanzaic forms, moreover, belong in the tradition of the so-called Pindaric ode, imitated freely from the Greek choric odes of Pindar. The closer imitations are in fixed though complex stanzas regularly repeated, and are called Regular Pindarics. These have first a strophe of undetermined length, then an antistrophe identical in structure with the strophe, and then an epode, different in structure from the strophe and antistrophe. The second strophe and second antistrophe are identical metrically with the first, the second epode with the first epode; and so on. The best examples in English are Ben Jonson's On [132]the Death of Sir H. Morrison, and Gray's Progress of Poesy and The Bard.[66]

About the middle of the seventeenth century, Cowley, misunderstanding the structure of Pindar's verse, invented another sort of Pindaric ode, which is called Irregular because, as he himself explained, "the numbers are various and irregular," and there was no formal stanzaic repetition. The lines were long or short according as the thought-rhythm demanded (or seemed to demand), and in respect to arrangement were not bound to any formal pattern. This freedom, under skilful control, may well produce felicitous results, but when not managed by poets of a strong and sure rhythmic sense—as it was not by the many Cowleyan imitators—it results merely in metrical license and amorphousness. "That for which I think this inequality of number is chiefly to be preferred," said Dr. Sprat, the first historian of the Royal Society, intending no sarcasm, "is its affinity with prose." But this argument, which is in part also that of the modern free-versifiers, is simply a confusion of two functions, the verse function and the prose function.

But before very long Cowley's invention found a true master in Dryden, whose To the Pious Memory of ... Mrs. Anne Killigrew (1686), Song for St. Cecelia's Day (1687), and Alexander's Feast (1697) are justly praised for their 'concerted music.' The example had in fact already been set by a still greater master; for Milton [133]with his early experiments in unequal rimed lines (On Time and At a Solemn Music), his incomparable success with the irregular placing of rimes in Lycidas, and his choral effects both with and without rime in Samson Agonistes, had shown what English could do under proper guidance. Then, after Dryden, the regular Pindarics of Gray and certain of Collins' Odes helped to carry on the tradition down to Coleridge's Dejection, Monody on the Death of Chatterton, and Ode on the Departing Year, and its culmination in Wordsworth's Intimations of Immortality ode (1807). After that, both in time and in interest, come Shelley's Mont Blanc (1816) (which he himself described as "an undisciplined overflowing of the soul") and Tennyson's On the Death of the Duke of Wellington (1852) (which has at least Tennyson's almost unfailing technical dexterity). The work of Coventry Patmore in this kind of verse has not been generally approved. This is partly because of the subjects on which he wrote and partly because of his inability to compose lines of haunting melody—perhaps his deliberate avoidance of them. But in certain poems like The Azalea and The Toys the very intensity of the feeling both creates and sustains and in the end justifies the 'irregular' metre.

Perhaps three-fourths of the greatest English poetry is in the unrimed 5-stress line called blank verse—nearly all the Elizabethan drama, Paradise Lost, some of the best of Keats and Shelley, Wordsworth's Michael, The Prelude, The Excursion (the good with [134]the bad!), Tennyson's Princess and Idylls (notable poems of their age, though not to be ranked with 'the greatest'), and Browning's The Ring and the Book, together with most of the dramatic monologues. No other metrical form has such an interesting history; no other form has manifested so great a variety and adaptability for every kind of poetic thought and feeling. These two facts alone—its bulk and its variety—would justify a much fuller treatment than is possible here. But it will perhaps be sufficient to follow rapidly in outline the development of blank verse, with illustrations of the most significant stages, and then, in the following chapter, to devote more attention to blank verse than to rimed stanzas in the exposition of metrical harmonies and modulations.

The idea of writing unrimed verse was no doubt the most valuable result to English poetry of the academic attempts, towards the end of the sixteenth century, to write classical verse in English. It could be pointed out triumphantly that all the splendid poetry of classical antiquity—Homer and Lucretius and Virgil, Sappho and Catullus and Horace and Ovid—had been independent of rime; and whatever might be the disagreement on quantitative feet in English, it was impossible to deny that English could successfully copy this element of the great classical verse and recover, as Milton said, the ancient liberty "from the troublesome and modern bondage of riming."

The movement had already begun in Italy with Trissino's Sophonisbe, written in 1515, the first modern tragedy. It reached England in the middle of the [135]century with the influence of the Italian Renaissance brought chiefly by Wyatt and Surrey. Surrey translated two books of the Æneid (II and IV) into blank verse (published in 1557); Sackville and Norton adopted it for the first English tragedy, Gorboduc (1565); and then Gascoigne used it in his Steele Glas (1576) for general didactic and satiric purposes. Thus the beginning was made, and it remained only for the new form to justify itself by its children. Experiments continued, with the first great achievement in Marlowe's Tamburlaine the Great.

The early examples show plainly both the influence of the parent couplet—for, as was said above, blank verse was written first as the old couplet without rime—and the syllable-counting principle: the line unit is prominent, there are comparatively few run-on lines or couplets, and some of Surrey's verse, for example, though it has the ten syllables then regarded as necessary, refuses to 'scan' according to more recent practice because the stresses are wholly irregular. On the other hand, there is often so great a regularity in coincidence of natural rhythm and metrical pattern, reinforced by some awkward wrenches of the conventional order of word and phrase, that the result is unpleasantly stiff and formal.

This is not so much monotonously regular as intolerably rough and unsteady.

Note here the monotonous pauses, indicated by the original punctuation.

Marlowe, inheriting the defects of his predecessors, [137]succeeded, by virtue of his "plastic energy and power of harmonious modulation" in recreating the measure. He found it "monotonous, monosyllabic, and divided into five feet of tolerably regular alternate short and long [i. e., unstressed and stressed]. He left it various in form and structure, sometimes redundant by a syllable, sometimes deficient, enriched with unexpected emphases and changes in the beat. He found no sequence or attempt at periods; one line succeeded another with insipid regularity, and all were made after the same model. He grouped his verse according to the sense, obeying an internal law of melody, and allowing the thought contained in his words to dominate their form. He did not force his metre to preserve a fixed and unalterable type, but suffered it to assume most variable modulations, the whole beauty of which depended upon their perfect adaptation to the current of his ideal."[67] No metre responds so readily and so completely to a poet's endowment of genius as blank verse, and hence the secret of Marlowe's improvements over his predecessors is his superior poetic gift. He seems to have felt and thought and written with an enormous imaginative power; by making his verse an organic expression of this power he achieved an almost new medium, ranging in variety from the simplicity and pathos of—

[138]

to the "swelling bombast of bragging blank verse" (Thomas Nash's hostile phrase) in Tamburlaine—

But even in Marlowe the 'mighty line' is still felt as the unit. All his volubility, his extravagance, his passion, his occasional tenderness did but develop the line to its fullest possibilities; the larger unit of the long harmonious period or 'blank verse paragraph' is rare and exceptional with him, though credit is due him for foreshadowing this also:

Shakespeare's blank verse is the supreme manifestation of the measure for dramatic purposes. In his plays it modulates and adapts itself to the changing emotions [139]of every speaker, "from merely colloquial dialogue to strains of impassioned soliloquy, from comic repartee to tragic eloquence, from terse epigrams to elaborate descriptions." It is customary to distinguish three 'periods' in Shakespeare's blank verse, corresponding closely to his whole artistic development: first, the more formal, 'single-moulded' line of the early plays; second, the perfect freedom and mastery of the great tragedies; and, third, the daring liberties, verging on license, of the later plays. These distinctions have, of course, no more absolute value than all similar classifications of impalpable modifications, but they at least suggest the underlying truth that Shakespeare began as a beginner, and then, having mastered the difficulties and subtleties of the form, treated it with the easy familiarity of a master. To illustrate these developments adequately would require pages of quotation; but one may compare the restricted movement of such a passage as this from Two Gentlemen of Verona (III, i)—

with the fine modulations, fitting exactly the nuances of meaning in this from Hamlet (III, iii)—

[140]

or this from King Lear (II, iv)—

and also with the flowing, slightly 'irregular' lines of this from The Tempest (II, i)—

The greater freedom of syncopation and substitution, of extra syllables and unusual pauses, which characterizes Shakespeare's later blank verse, became almost a norm with Beaumont and Fletcher, Shirley, Ford, and the Jacobean dramatists. They often carried [141]freedom to the extreme limit, where an inch further would change verse into prose. They were capable, to be sure, of more careful regular verse, and wrote it when the occasion seemed to call for it; but partly from choice, and partly no doubt from haste or indifference or both, they made a very free blank verse their staple. Shakespeare had alternated prose and verse as the subject or tone required; the later dramatists seemed to seek a verse that might be, in a sense, midway between prose and verse. Thus they avoided a necessity of frequent change, except a loosening or tightening of the reins. To call this verse decadent is somewhat unjust. It is in truth a special form which is certainly well justified for certain subjects and occasions.

[142]

When Milton composed Comus in 1634 it was natural for him to model his blank verse on the best of Shakespeare's and Ben Jonson's, rather than on that of the contemporary playwrights; for his finer taste, his more delicate ear, and his classical training and tendencies would at once lead him to reject the metrical laxities of Ford, Shirley, Davenant,and the other writers of 'broken down' blank verse. And though his language shows great familiarity with the later plays of Shakespeare, especially The Tempest, he admitted comparatively few of their metrical licenses and followed in the main the versification of the Midsummer Night's Dream and the earlier tragedies. There is generally a tendency to make the line the unit—but the verse paragraph or stanza effect is also present in nearly fully developed form, as witness the opening lines of the poem—weak or feminine endings are not frequent, alexandrines very few. The 'short fit of rhyming' (ll. 495 ff.), disapproved by Dr. Johnson, would be explained partly by the tradition of the masque and partly by the model of Shakespeare's comedies.

But the great Miltonic blank verse of Paradise Lost is not a copy of any master; it is a development and a consummation of two influences, the slow maturity of Milton's mind, deepened and broadened by the Commonwealth controversies "not without dust and heat," and the exalted sublimity of the yet unattempted theme of justifying God's management of human and divine affairs. His maturity brought him his great familiarity both in matter and in style with nearly all that was best in European literature, and his peculiar [143]subject, with only gods and angels (Adam and Eve are scarcely human, even after the fall) for characters and selected portions of eternity and infinity for time and place, gave him the tendency to artificiality and strain to the outmost verges of sublimity, and to extraordinary involution of phrase and idea—for all of which he must have a suitable prosody. He chose blank verse when the poetical fashion was for rime and described it, in words not altogether clear, as consisting "only in apt numbers, fit quantity of syllables, and the sense variously drawn out from one verse to another."[68] Apt numbers, that is, appropriate rhythms, Milton's verse certainly has; but it is the last item, the great variety of movements subordinating the line-unit, and running-on of verses into longer periods, for which his blank verse is famous. Every page of Paradise Lost contains examples; some of the finest occur in the rhetorical display of the Pandemonic Council in Book II. Note the position of the pauses in the following passage, and then compare the specimens of early blank verse given above.

One may surely say that Milton extracted from blank verse all its possibilities of variety and movement so far as his subject matter permitted. He is lyrical, dramatic, didactic, and of course epic, in turn. He even showed that it is possible to imitate hollowly his own "planetary wheelings"—as though the instruments kept on playing and the music ceased.[70]

Since Paradise Regained and Samson Agonistes, though various poets have adapted it to their own uses, blank verse has shown only one significant development, the conversational, or so-called 'talking,' style. In the eighteenth century Milton's mannerisms [146]dominated nearly all blank verse, both for good and for evil. What freedom Thomson allowed himself he got from Milton; most of Cowper's thin grandiosity he took from Milton; and much also of Wordsworth's false and empty elaboration which make the Prelude and Excursion so dull in places—the whole tribe of verses of which

is the pilloried example—came from the Miltonic tradition. Keats fell partially into the error, but was wise enough to recognize it. Shelley, with much of Milton's intensity and somewhat too of his sublimity, could successfully follow the great stride and at the same time preserve his own idiom. Tennyson, keeping both the freedom and as much of the "continuous planetary movement" as was consistent with his themes, softened the metre—weakened it, some will say—by his decorative tendency and indulgence in only half-concealed virtuosity.[71] And the famous Oxus ending of Arnold's Sohrab and Rustum is a studied reproduction of the Miltonic music in a lower key. But it was Landor who, taking a hint perhaps from Milton's unadorned didacticism of Paradise Regained and also from the straightforward verse used on occasion by the Elizabethan dramatists, showed the way to what has often been called a strictly contemporary development of blank verse, the talking style. Since this is less [147]familiar than most of the phenomena of blank verse, it will require fuller illustration.

The uneven line which separates blank verse and prose is easily apparent in such a passage as the following from Much Ado about Nothing (V, i)—

In the first part of this passage the language is the simple natural expression of prose, yet so devised that it also fits the metrical pattern. It is either prose or verse according to the way one reads it. But in Leonardo's long speech (after the first line, which is 'irregular') the verse pattern becomes more and more prominent, until in the last three lines it predominates over the natural utterance of the words and produces a certain stiffness. Here the two different manners [148]stand side by side: a natural simplicity so great that the metrical quality is almost obscured, beside a formality so obvious that the feeling of natural expression is partly lost. Now Milton, and after him Dryden and the eighteenth century, regarding poetry generally as a thing apart, followed the latter sort; but when the Romantic Revival brought poetry back to ordinary human life there reappeared, tentatively, of course, a simpler blank verse in Thomson, Crabbe, Cowper, and Wordsworth. A clear example is the opening of Landor's Iphigeneia and Agamemnon—

Again, compare Mrs. Browning's Aurora Leigh—

These are from the metrical point of view nearly identical with Mr. Robert Frost's talking verse, so often called a 'contribution' to verse technique—

The obvious difficulty is to maintain dignity along with relaxation—a feat which Mr. Frost and Mr. E. A. Robinson have occasionally accomplished. And from this it is but a step to the extreme simplicity of Miss Lowell's To Two Unknown Ladies—

Thus the wheel has come nearly full circle, but with a longer radius. For just as blank verse developed from the early Elizabethan—and pre-Elizabethan—strict formality to the laxity of the Jacobean dramatists and found a true balance of freedom and restraint in Milton, so from the monotonous eighteenth-century couplet (and it should be recalled that in the beginning blank verse sprang from the couplet) it has gradually enlarged its freedom into the extreme license from a metrical point of view of its adopted cousin free-verse. Already, moreover, there have been signs of a reaction against the extreme, and the wheel is coming to an artistic balance again.

Free-verse (or, as Miss Lowell prefers, 'unrhymed cadence') is a hydra-headed phenomenon. It can never be adequately discussed; for when one head is disposed of, two others appear in its place. Its origins are involved in obscurity and—what is worse—ignorance; and its practitioners and staunchest defenders are as variable in their points of view as it itself is in its rhythmic impulses.[73]

[151]

Behind all the utterances of friend and foe seems to lie the ultimate belief that the 'voluntary thraldom' of formal metrical patterns is a monstrous error which can only be removed by unrestricted appreciation and application of the natural rhythms of idea and of language. There is in every thought, however simple or subtle, in every feeling, however evanescent or profound, an inherent rhythm which is as a material body to the thought's or emotion's soul. This native, inevitable rhythm—one might call it the rhythme juste, the exact rhythm—is the only fit expression for an intellectual or emotional idea; all others are foreign to it, tyrannous usurpations, in a word, impossible substitutions for it. To attempt, therefore, to twist these natural and exact rhythms to the formal predetermined patterns of traditional versification is a suicidal impertinence, foredoomed to failure.

Such a position has in theory much justice. It means briefly that the basis of poetical form should not be the metrical pattern freely varied yet always perceptible, but the natural organic rhythm of the ideas expressed; that is, there should be no harmonized difference between what have been explained above as thought rhythms, sound rhythms, and metrical rhythms, but all three should be one original and indivisible unit. This would make a combined thought-and-sound unit (breath group and logical-emotional [152]group) the foundation of verse, whereas this is really the characteristic of prose as distinguished from verse. These exact organic rhythms "differ from ordinary prose rhythms," says Miss Lowell, "in being more curved, and containing more stress"; which, though not very perspicuous, seems to mean that free-verse is more carefully cadenced, or, in other words, more nearly metrical, than ordinary prose. Perhaps it would be no injustice to the upholders of free-verse in its best manifestations to say that, while metre requires that beneath all variations the regular beat should never be missed, free-verse requires as much rhythm (i. e., regularity) as is possible without its becoming perceptible.

If this is true, or as near the manifold truth as one can get, then the free-verse movement in English is mainly a return to the cadenced prose of the seventeenth century with the additional trait of the appearance of verse. This is an important addition, however. It involves a careful recognition of what psychology calls the 'prose attitude' and the 'verse attitude,' and also (as has been suggested above) the peculiar union of prose with the spatial rhythm of verse. We read with ear and eye together, though with varying proportions of emphasis on the one or the other; for some 'vocalize' whatever they read, others read almost entirely with the eye. Since it is the eye that takes the earlier and quicker perception of printed language, we tend to judge by the appearance of a page whether it contains prose or verse. Columns of irregular but approximately equal line lengths, regular blocks of printing regularly [153]spaced and separated as stanzas, indentation of every second or every third line—these at once announce that the page contains verse. And they at the same time constitute an obvious spatial rhythm to the eye, and prepare the attention of eye and ear and mind for the approximate regularity of verse. Then, when so prepared, we unconsciously organize as fully as possible any irregularities that appear in the language and transform into actual verse the verse potentialities which pervade our speech.

Some kinds of free-verse, however, do not, so far as one can see, aim to be more than ordinary prose printed in segments more or less closely corresponding with the phrase rhythm or normal sound rhythms of language. It is then prose in actuality and verse in appearance—no more.

On the justification of this peculiar amalgam there is little agreement. No doubt for certain swift effects free-verse is the natural and most serviceable medium. Many short poems in this irregular form are like snapshots or like rapid sketches as compared with finished paintings. But the ultimate æsthetic judgment must be precisely that of the snapshot as compared with finished painting. Nature is always wrong, says the paradox; art depends upon a deliberate selection of details and structure. It balances freedom and restraint, variety and uniformity, one against the other; and even when it appears spontaneous it is but the result of an unconscious choice which is itself born of long training or of the mysterious faculty divine. In very little of what at present is called free-verse does art have a [154]real place. It is all freedom and variety, with almost no restraint and uniformity: all stimulation and no repose. There is sometimes a rapid alternation of verse rhythm and prose rhythm, which, in Bacon's phrase, may cleave but not incorporate; they succeed each other but do not melt into each other. Now and again, to be sure, this uncertainty, this very irregularity, powerfully represents the thought and emotion of the poem; but nevertheless there can be little doubt that except in the limited field of instantaneous flashes the most adequate and pleasing medium is the skilfully varied regularity of formal verse.[74]

The many kinds of free-verse are recognizable chiefly by the greater or less feeling of metrical form lying behind them. For convenience they may be distinguished, according as verse or prose predominates, as (1) irregular unrimed metre, (2) very free blank verse, (3) unusual mingling of metre and prose, a kind of recitative, and (4) mere prose printed as verse, or what may be called free-verse par excellence. A few illustrations will help to make clear the distinctions.

Of the first sort are the unrimed choruses in Milton's Samson Agonistes, the metre of Southey's once-admired [155]Thalaba and the Curse of Kehama, and parts of Shelley's Queen Mab. Here the lines are irregular in length (as in the 'irregular' Pindaric odes), but they are usually felt as truly metrical, though they do not repeat a single pattern.

Very free blank verse, when taken in small excerpts, often seems devoid of metrical regularity. The reason for this is that in long poems much greater freedom is possible because the ear and the attention, accustomed for longer periods to the formal pattern, hold it more easily where it becomes faint. Examples of this approximation to prose have been given above, pages 43, 44. The famous first lines of Paradise Lost, if printed after the contemporary fashion of free-verse, would by very few be recognized as blank verse; and the same is true of many passages throughout the poem, and indeed throughout all long poems in blank verse.

Among the finest free-verse in English are the Evening Voluntaries of Henley.[75] In these poems clearly metrical lines (sometimes only parts of lines) alternate with simple prose. The line length is now based on phrasal rhythm, and at other times on no discoverable principle except that of beginning a new line with some emphatic word.

[158]

Modern free-verse, or free-verse par excellence, which is mere prose with the spatial rhythm of verse, has been skilfully written by various contemporaries. Let a single example suffice. Such a bare but moving situation as that of Miss Lowell's Fool's Money Bags could no doubt be adequately presented in traditional metre, but perhaps not so directly as in her 'curved' prose—

A good example of combined metre and confessed prose (not to be confused with the mingling of verse and prose illustrated on the previous page) with easy transitions from one form to the other may be seen in a poem called Spring by Mr. Clement Wood. The rapid change from verse to prose is, of course, familiar in Shakespeare and his fellow dramatists, sometimes even in a single speech.

[159]

As wide as are the possibilities of variety in native English verse, the poets have endeavored to extend its boundaries by the annexation of foreign prosodies from ancient Greece and Rome and from mediaeval France. In absolute contrast to free-verse, which is the denial of metrical formalism, this is the apotheosis of it. They admittedly place form above content and are satisfied (for the most part) with the mere exhilaration of dancing gracefully in chains.

A group of Elizabethan experimenters, among whom were Sidney and Spenser, sought diligently to compose in the quantitative metres of the classics; Puttenham, the author of one of the first English treatises on the Art of Poetry (1589), declared that by "leisurable travail" one might "easily and commodiously lead all those feet of the ancients into our vulgar language"; but while they may have satisfied themselves (Spenser certainly did not) these experimenters produced nothing of genuine significance. The result was candidly anticipated by Ascham, who said in the Schoolmaster (1570) that "carmen exametrum doth rather trot and hobble than run smoothly in our English tongue." Thomas Nash confirms this opinion in his criticism of Stanyhurst's attempt to translate Virgil into hexameters: "The hexameter verse I grant to be a gentleman of an ancient house (so is many an English beggar); yet this clime of ours he cannot thrive in. Our speech is too craggy for him to set his plow in. He goes twitching and hopping in our language like a man running upon quagmires, up the hill in one syllable and [160]down the dale in another, retaining no part of that stately smooth gait which he vaunts himself with amongst the Greeks and Latins" (Four Letters Confuted). Coleridge's judgment was the same:

Thereafter, apart from isolated attempts, efforts were abandoned until the nineteenth century, when Southey, following William Taylor, who in turn had been induced by Goethe's Hermann und Dorothea to try a new principle of frankly substituting sentence stress or accent for length of syllable, wrote his Vision of Judgment (1821). Out of this revised experimenting came ultimately Longfellow's Evangeline (1847) and the Courtship of Miles Standish (1858) and Clough's Bothie of Tober-na-Vuolich (1848). These alone, not to mention the lesser imitations, were enough to discredit the movement metrically. Meanwhile Tennyson and Kingsley, followed later by William Watson, and still enthusiastically by the present Poet Laureate, undertook to harmonize syllabic length and stress by more or less occult processes. As a matter of learned experiment and debate these problems have a certain academic interest, but only the staunchest and (one may say) blindest adherents find in them any practical importance.

The storm centre of all classical adaptations has been the dactylic hexameter, the standard measure of Greek and Latin narrative poetry. The most nearly successful English hexameters are probably those of Kingsley's Andromeda (1858), which occupy a middle [161]ground between the purely accentual and the purely (so-called) quantitative experiments. An example of this and one of Mr. Bridges' quantitative hexameters must suffice. Though both have good qualities, neither approaches the melodic variety and dignity of Homer and Virgil, or even Ovid.[77]

After the hexameter the most frequently imitated metre is the Sapphic strophe. Swinburne's Sapphics in Poems and Ballads are the best known; but though they are finely musical they do not pretend to give more than an echo of the Greek music.

Both Tennyson and Swinburne tried the Catullan hendecasyllabics. Tennyson's Milton, in alcaics, is famous, and has a well-marked Miltonic sound, but little of the sound of Horace's alcaics. Admirable also are the elegiac distichs of Watson's Hymn to the Sea—

Of the French lyrical metres that have been imitated in English, mainly for lighter themes, the ballade and the rondeau are the most important. These and the villanelle, triolet, and pantoum are not, like imitations of classical forms, semi-learned attempts to do in English what is foreign to the nature of the language, but games of skill in phrasing and riming, wholly legitimate once their artificiality is granted. For the impassioned overflowing of a sincere spirit they are unfitted, but for [163]grace, point, and delicate charm nothing could be better devised; and when occasionally they are used for the expression of genuine feeling, the unexpected union of lightness and seriousness has a peculiarly poignant effect.

The ballade in its commonest form consists of three 8-line stanzas riming ababbcbc and a 4-line stanza called 'envoy,' bcbc; the last line of each stanza being repeated as a refrain, and the a, b, and c rimes throughout the poem being the same. The lines contain usually either four or five stresses. The envoy is a sort of dedication, addressed traditionally to a "Prince." Variations of all kinds occur, encouraged by the difficulty of satisfying all the demands of the form. Examples may be found (with an excellent introduction) in Gleeson White's collection of Ballades and Rondeaus (Canterbury Poets), and Andrew Lang's Ballades of Blue China.

Rondeaus and rondels (two forms of the same word) are written with greater freedom of variation. Their organic principle is the use of the first phrase or first line, twice repeated, as a refrain (R). The commoner model in English is: aabba, aabR, aabbaR, in which the first half of the first line constitutes the refrain. Another type rimes ABba, abAB, abbaAB (the capital letters indicating the lines repeated). For examples see the reference above. Austin Dobson, Henley, and Swinburne have written successfully in this form.

The triolet is a sort of abbreviation of the second variety of rondeau. Its lines are usually short and rime ABaAabAB.

[164]

The villanelle, in its normal form, consists of five 3-line stanzas (aba) and a concluding 4-line stanza, all with but two rimes, the first line, moreover, being repeated as the sixth, twelfth, and eighteenth, the third line as the ninth, fifteenth, and nineteenth.

The pantoum is of Eastern origin, but it came into English through the French. It is extremely rare. It consists of a series of quatrains abab, with the second and fourth lines of each stanza repeated chainwise as the first and third of the next stanza. The closing stanza completes the chain by taking as its second and fourth lines the first and third of the first stanza.

From Italy have come, besides the ottava rima and the sonnet, two other metrical forms, the sestina and the terza rima. The sestina is composed of six 6-line stanzas and a final 3-line stanza. Instead of rimes the end words of the lines of the first stanza are repeated in this order 1.2.3.4.5.6.—6.1.5.2.4.3.—3.6.4.1.2.5.—5.3.2.6.1.4.—4.5.1.3.6.2.—2.4.6.5.3.1.—and the last stanza 5.3.1. with 2.4.6. in the middle of the lines. Gosse, Swinburne, and Kipling have written sestinas; Swinburne one with the additional embellishment of rime.

The terza rima is the metre of Dante's Divine Comedy. The rimes are aba, bcb, cdc, etc.... yzy, zz. It has not been very successfully used in English, except in the stanzaic arrangement of Shelley's Ode to the West Wind,—aba, bcb, cdc, ded, ee. Other examples besides translations of Dante are short poems by Wyatt and Sidney, Browning's The Statue and the Bust, and Shelley's unfinished The Triumph of Life.

[165]

The terms melody, harmony, and modulation, being borrowed from music, are not to be applied too literally to the art of versification. They represent metaphorically, however, certain important qualities of verse which, with the exception of rime, cannot from their very impalpability be formally explained, but can only be suggested and partially described. They are not the determining and fundamental characteristics of verse—those have already been discussed—but rather its sources of incremental beauty, of richness and, subtle power. To draw an illustration from another art, they add light and shadow, fullness, roundness, depth of perspective, vividness, to what would else be simple line-drawing.

The language of ordinary prose has its own melody and harmony, its own sonorous rhythms, and its own delicate adjustments between sound and meaning. All these natural beauties verse inherits from prose and then adds the further beauties that result from the union of prose rhythms and the formal patterns of verse. Some of these qualities which are the peculiar enhancements of verse will now be examined.

The simplest and most tangible of these is rime in its various forms. Rime is, in its most general signification, the repetition, usually at regulated intervals, [166]of identical or closely similar sounds. According to the circumstances of the identical or similar sounds, four varieties are distinguishable: (1) alliteration, or initial rime, when the sounds at the beginning of accented syllables agree, as tale, attune; (2) consonance, when the vowel sounds differ and the final consonantal sounds agree, as tale, pull; (3) assonance, when the vowel sounds agree and the consonants differ, as tale, pain; and (4) rime proper, when both the vowels and the final consonants agree, as tale, pale.

Alliteration is a natural and obvious method of emphasis in English—and often difficult to avoid rather than to obtain. Popular sayings—wind and weather, time and tide, kith and kin, ever and aye, to have and to hold—are fond of it for its own sake. The early English, German, and Scandinavian prosodies made it a determining principle; and in the north of England it survived well into the fifteenth century; but since then it has been considered a too 'easy' kind of metrical ornament, one to be used sparingly and only for very special effects. "Apt alliteration's artful aid" is very well when it is apt and artful; but when some poets in their simplicity have gone so far as to "hunt the letter to the death," one cannot but condemn it, in John Burroughs' ironic phrase, as a "leprosy of alliteration." Most of the poets, however, have made skilful use of it, notably Tennyson and Swinburne, though the latter frequently overdid it, as in—

[167]

Very remarkable is the combination of rime and frequent alliteration in Browning's Abt Vogler.

Analogous to alliteration and perhaps to be classed as a by-form of it is the subtle use of the same sound in unstressed parts of neighboring words, as in—

Consonance is very similar to this latter form of alliteration. Its use is irregular and usually hidden. Note the alliteration and consonance in Milton's line, both the s's and the n's—

Assonance, like alliteration and consonance, occurs in modern verse sporadically, almost accidentally, but with great frequency in all languages. As a regular principle of verse (in place of rime) it is characteristic of Spanish and of Old French; in English its deliberate use is very rare—the best example is perhaps the song "Bright, O bright Fedalma" in George Eliot's The Spanish Gypsy.

Minute analysis is tedious and unsatisfactory, often indeed misleading, but a single example will perhaps suggest some of the ways in which alliteration, consonance, and assonance are interwoven for harmonic effects that, not being altogether obvious, are felt rather than directly perceived. Similar experiments may be made by the reader with other passages. The opening stanza of Gray's Elegy, quoted on page 55, above, is remarkable for its smooth and quiet flow, [168]symbolic of the atmosphere described by the words. How is this 'atmosphere' produced? or rather, what is there that produces in us this sense of appropriate atmosphere? In the first place, the lines are 5-stress and have the "long iambic roll," and the rimes are simple abab. Furthermore, the coincidence of prose and verse rhythms is noticeable; there are only three variations: wind in the second line, which is too important to occupy the metrically unstressed position, and o'er in the second line and the second and in the fourth, which are not quite strong enough to stand in the stressed position. By a sort of substitution or 'occult balance' the weakness of o'er is compensated by the slight overweight of wind. And the weakness of and is strengthened by the rhetorical pause after darkness. A rough approximation in semi-musical notation would give for the second line

◡—◡———◡◡◡—

There is a syncopation by which — — and ◡ ◡ combine (the natural syllabic length of o'er helping considerably) without destroying the fundamental rhythm. In the fourth line, instead of

◡—◡—◡—

we have

the pause being supported by the meaning as well as by the structure of the verse. Alliteration is appropriately inconspicuous; it is limited to plowman ... plods and the conventional weary way. The consonance is [169]significant. The most frequently repeated consonantal sounds are: l 10, d 9, r[78] 8, th 6, n 6, and w 5; that is, of the seventy consonantal sounds (counting th as one, p and l as two sounds) in the stanza, thirty-five, or one-half, are the comparatively soft sounds l, r, th, n, w. From the point of view of the line, a tabulation shows two or more occurrences in each line of—

That is, there is a kind of RTLD motif throughout the stanza. The assonance is even more striking. The stressed vowel sounds (which are of course the most important[79]) line by line are as follows:[80]

Here the five ō-sounds and four ī-sounds and three ŭ^R-sounds are noticeable.

Now while no one would dream of saying that such a mechanical examination unlocks the mystery of this quatrain's music, it cannot be denied that the predominance of some sounds (especially those that are peculiarly suggestive) over others is significant. And certainly such a tabulation reveals parts of the mystery which are not plain even to the trained eye and ear.

[170]

The origin of rime is much disputed, but it occurs, at least sporadically, in the poetry of nearly all peoples, and is likely to have been a spontaneous growth arising from a natural human pleasure in similar sounds. "It lies deep in our human nature and satisfies an universal need." It is an established phenomenon in Sanskrit and Persian prosody, in Arabic, in Chinese, in Celtic, in Icelandic. Greek prosody, and Latin, which was based upon Greek, rejected it, partly perhaps because it was too simple an ornament for the highly cultivated Greek taste, especially on account of the great frequency of similar inflectional endings, and perhaps because it was not entirely consistent with the quantitative principle.[81] In the popular Latin verse, however, which was accentual, rime is found; and when, before the fall of the later Empire, quantity was gradually abandoned, rime returned as a regular feature of Latin verse. From thence it passed into the Romance languages—Provençal, Italian, French—where it was for a time rivalled by assonance; and finally, under French influence after the Conquest, it made its way into England. But it had not been unknown in earliest English verse, though it occurred only here and there, as in Greek and Latin.[82] And from the fact that rimes appear with greater frequency in the later than in the earlier Anglo-Saxon verse, as the native poets became more [171]familiar with the rimed Latin hymns, one may feel sure that it would have developed into a staple of English verse independently of French influence. From the twelfth century until the introduction of blank verse by the Elizabethans, practically all English verse, except that which belongs to the Alliterative Revival (mainly in the north of England) of the second half of the fourteenth century, was rimed.

From the æsthetic point of view rime has been severely attacked and faithfully defended. A lively controversy was waged at the end of the sixteenth century between the Renaissance classicists, who of course condemned it, and the native rimers, but was brought to a peaceful conclusion by Samuel Daniels' A Defence of Rhyme in 1603. In a prefatory note to the second edition of Paradise Lost, Milton delivered an arrogant but ineffectual counterblast. Rime, he said, was "no necessary adjunct or true ornament of poem or good verse, in longer works especially, but the invention of a barbarous age, to set off wretched matter and lame metre; graced indeed since by the use of some famous modern poets, carried away by custom, but much to their own vexation, hindrance, and constraint to express many things otherwise, and for the most part worse, than else they would have expressed them."

The chief arguments against rime are those mentioned by Milton, its tendency to conceal "wretched matter and lame metre," and the necessity it often forces upon poets of either twisting unpleasantly what they have to say or of adding irrelevant matter. Besides [172]these there is also what Cowper called "clock-work tintinnabulum"—mere empty jingle. But all the arguments are double-edged. For although many inferior poets have imposed for a while on readers and critics by the superficial melody of rime alone, "wretched matter and lame metre" were never long successfully concealed by it. And although, as Hobbes wrote, rime "forces a man sometimes for the stopping of a chink to say something he did never think," it is a fact nevertheless that the second thought, induced by rime-necessity, "the rack of truest wits,"[83] is sometimes if not better than the first, at least a worthy and handsome brother to it. Whether rime be a hindrance, vexation, and constraint to the poet depends almost wholly on his mastery of the technique of verse. It is not always easier to write in unrimed measures, for, as Milton proudly implied, good blank verse is the most difficult of all metres. And although the jingle of like sounds may become tedious and mechanical if unskilfully handled—"to all judicious ears trivial and of no true musical delight," says Milton again—it has also proved a source of richness and beauty of sound; and it should never be forgotten that in the true æsthetic judgment of poetry sound plays a very important part.[84]

The satisfaction which the ear receives from rime at the end of a verse has been aptly compared to the pleasure we feel when a long arch of melody returns to the dominant and then the tonic. More elaborate is [173]Oscar Wilde's praise of rime—"that exquisite echo which in the music's hollow hill creates and answers its own voice; rhyme, which in the hands of a real artist becomes not merely a material element of metrical beauty, but a spiritual element of thought and passion also, waking a new mood, it may be, or stirring a fresh train of ideas, or opening by mere sweetness and suggestion of sound some golden door at which the Imagination itself had knocked in vain; rhyme, which can turn man's utterance into the speech of the gods; rhyme, the one chord we have added to the Greek lyre."

The real problem in the arguments on rime is its fitness or unfitness in particular kinds of poetry. No rules or laws can be formulated; men have judged differently at different times; but it has been generally felt that shorter poems, inasmuch as they are in a way the concentrated essence of poetry, and must make their full impression almost instantaneously, require all the advantages of the poetic art. Tennyson's unrimed lyrics and Collins' Ode to Evening are unusual, though successful, experiments. For long poems, however, there is not this necessity of immediate effect. Here rime is sometimes a vexation, sometimes not. Justification lies in special circumstances. The classical French drama found it indispensable; English poetic drama gave it a trial in the seventeenth century and rejected it. Narrative poems which contain a large lyrical element, like the Faerie Queene and the Eve of St. Agnes, are, all agree, enhanced by the rime. But no one would now wish to have Paradise Lost in [174]rimed verse, though it is clear from the publisher's note in 1668 that many readers at the time were 'stumbled' because it was not. On the other hand, we feel that Chapman's and Pope's Homer and Dryden's Virgil might have been better without rimes. Once more, it lies with the poet—and with the poem—to justify his use of rime or his refusal of it; if he is a good poet and his judgment is not warped by local or temporary conditions there will rarely be any doubt.

Rimes are called masculine when they consist of one syllable, as cries: arise; feminine when they consist of two or more syllables, as heedless: needless, beautiful: dutiful. When both vowel and following consonant agree the rime is called perfect, as might: right, solemn: column. When the preceding consonant as well as the vowel and following consonant agree the rime is called identical or echo rime, as reed: read, perfection: infection, ours: hours. When there is a difference either in the vowel sound or in the following consonantal sound, that is, when assonance or consonance is substituted for rime, the rime is usually said to be approximate or imperfect, as worth: forth, was: pass, gusht: dust (Coleridge). When the rime words look alike but are pronounced differently, they are called eye rimes, as war: car, brow: glow. Sometimes false rimes occur which have no similarity of sound or appearance, but are more or less sanctioned by earlier pronunciation or by custom, as high: humanity. Sometimes also unaccented syllables are rimed with accented syllables, as burning: sing.

Imperfect rimes of all sorts are used for various reasons. Compared with some languages, English is not [175]very rich in rime words; and for many words which poets are prone to use, such as love, God, heaven, etc., few available rimes exist. When good rimes are few, older pronunciations are often resorted to, as the familiar love: move, blood: stood, north: forth. In reading the older poets we find many rimes which are now imperfect but were once entirely correct, as the eighteenth century fault: thought, join: shine, tea: way. On the other hand, the poet's carelessness or indifference is sometimes to blame for approximate rimes, as Gray's beech: stretch in the Elegy, and his relies: requires, Blake's lamb: name and tomb: come, Coleridge's forced: burst, Whittier's notorious pen: been, etc. But to dogmatize on a point like this is obviously very dangerous. Certain poets, especially among the moderns, may be said to choose imperfect rimes deliberately, both as a fresh means of securing variety and avoiding the monotony of hackneyed rimes, and also as a means of subtly suggesting the imperfection and futility of life. A few famous examples, defensible and indefensible, are: Wordsworth's robin: sobbing, sullen: pulling; Tennyson's with her: together, valleys: lilies; Keats's youths: soothe, pulse: culls; Swinburne's lose him: bosom: blossom. Keats and Rossetti are noted for their free use of approximate rimes. The humorous rimes of Byron and Browning, among others, are of course in a different category.

Feminine rimes have been frequently rejected as undignified. They are, said Coleridge, "a lower species of wit"; and he instanced, not very justly, the couplet of Smart:

[176]

But again the right justification is successful use, and no one will deny that Swinburne's double and triple rimes have greatly enriched his verse and revealed to others unused possibilities of metre. Such rimes as grey leaf: bay-leaf were practically a new thing in 1865.[86]

Too evasive for explanatory analysis, almost too delicate and impalpable even for descriptive comment, are many of the best musical effects of fine poetry. The poet's ear and his sixth prosodic sense enable him to make his verse a perfect vehicle of his meaning and emotion. He chooses an appropriate stanza for his poem, discovers an unguessed power in some common measure, makes the words hurry or deliberately holds them back, varying the tempo with the spirit of the words, gives the pattern an unusual twist when the idea is unusual, startles or soothes by the sound as well as by the intellectual content of his lines—and accomplishes all these metrical nuances, not with the whip-snapping of the ring-master, but with the consummate art that conceals art. When his prosodic effects are obvious they lose their power; we can see how the trick is done and we do not marvel. But when we feel vaguely the haunting quality of a melodious line or the perfect metrical rightness of a phrase without knowing why the melody haunts us or the phrase just fits, then we both marvel and applaud; then the poet's [177]gift, his divine authorization, is patent, and we recognize his superiority with awe.

Some of these effects have already been mentioned in the preceding paragraphs; but besides the 'tone-color' of assonance and consonance and rime proper there are also effects of pitch and of tempo and of repetition, and imitative effects, more or less concrete and explainable. It is true that many trained readers find subtleties of sound and suggestiveness where others find none, and also that many find rich beauties that the poet himself was not aware of and did not intend. This latter case may be accounted for in two ways: sometimes a reader is supersubtle and imagines embellishments that do not exist; and sometimes the poet builds better than he knows. His intuition, or inspiration, or whatever one chooses to call it, endows him with powers of whose complete functioning he is not at the time conscious. As readers must steer carefully between these two dangers, so also the poet has to avoid on the one hand repelling us by the appearance of a metrical device and on the other losing an effect which he intends but which may be too delicate to be seen or felt. No one probably ever missed the simple melody of Poe's

or the imitative effectiveness of Swinburne's

and though these beauties are obvious they are for most tastes not too obtrusive. But Tennyson's

[178]

is not so obvious, and there is danger of its escaping notice. One hears the line with increased pleasure after the imitation of sound is pointed out; but only the trained ear catches it at first.

This correspondence of sound and sense is called onomatopoeia. It may appear in a single word, as buzz, whack, crackle, roar, etc.; or a combination of imitative words, as Tennyson's

or a suggestive echo rather than direct imitation, as Shelley's

or a suggestion of motion rather than of sound, as Milton's sea-fish

and the

or an attempt to imitate the motion described, as Tennyson's picture of Excalibur when Sir Bedivere hurls it into the lake—

and Swinburne's more simple

[179]

or even the correspondence of a harsh line and a harsh thought, as Browning's famous

Sometimes there is obtained an effect of altered tempo; of which the best illustration, though hackneyed, is still Pope's clever couplets in the Essay on Criticism—

Examples of similar metrical skill may be found everywhere, especially among the more conscious literary artists, such as Shelley, Tennyson, Rossetti, Swinburne, and Browning, too. A few worth study follow:

Something has already been said above on the nature and effects of pitch in spoken rhythm (pages 35 ff.). It is a constant factor of language, but its usual function is special emphasis or intensification. By itself it rarely dominates or determines the rhythm. And since the regular determinants of spoken rhythm are time and stress, it follows of course that pitch serves usually to reinforce these determinants.[89] But not always; for not only does pitch sometimes clash with rhythmic stress, but also it is sometimes a substitute for it. All three of these functions—strengthening, opposing, and replacing stress—are operative in verse.

In Shelley's line

a great deal of the effect is due to the combination of word accent and emphatic pitch in the syllable-ting-, so that not merely the one word but the one syllable dominates the whole verse. In such frequent conflicts of stress as "on the blue surface," where the prose [182]rhythm is ◡◡ ̷ ̷◡ while the verse pattern has ◡_̷◡_̷◡, the so-called hovering accent (as it is usually described, with the theory that somehow the normal quantity of stress is divided between the and blue) is properly a circumflex accent, which in other words means pitch. Similarly in "If I were a dead leaf," the peculiar rhythm is to be explained as a balance of pitch against stress. And in that metrically notorious line of Tennyson's—

the chief irregularity or dissonance is the clash of pitch against stress in "own time." If the line read—

there would be an unusual arrangement of stresses and unstressed syllables, a peculiar syncopation, but no great difficulty.[90] Much simpler and clearer is the conflict of stress and pitch in such passages as

and Keats's

and Marvel's

The most interesting, and the rarest, effect of pitch in verse is its use as a substitute for stress. In the much-discussed first line of Paradise Lost—

there is a metrical stress on dis-of "disobedience." This is not so much, however, an intensification of an already existent secondary accent, as in, for example, Shelley's

as the substitution of pitch for stress.[93] The adaptability of language to metre appears very clearly in such a line as Paradise Lost, III, 130—

in which the first compound shows a conflict of pitch and stress ('self' having a pitch-accent, but occurring in an unstressed part of the line), while the second shows pitch taking the place of stress. The whole line, [184]and indeed the whole passage, though not of high poetic value, is an admirable illustration of the Miltonic freedom of substitution and syncopation—pitch playing a very important rôle. One should read the lines first as prose, with full emphasis on the expressive contrasts; then merely as verse, beating out the metre regardless of the meaning; finally, with mutual sacrifice and compromise between the two readings, producing that exquisite adjustment which is the characteristic of good verse. There is a similar example of pitch and stress in the familiar

Repetition is a rhetorical not a metrical device, though it is employed with great effectiveness in verse as well as in prose: