In 1702, just after the beginning of the new century, there appeared a book which, though it received little directly critical notice, and was spoken of with disapproval by some who did notice it, was repeatedly reprinted, and which expressed, beyond all reasonable doubt, ideas prevalent largely for a century or more before it, and almost universally for a century or more after it. This was the Art of Poetry of Edward Bysshe. The bulk of it is composed of dictionaries of rhyme, etc. But a brief Introduction puts with equal conciseness and clearness the following views on English prosody.

"The structure of our verses, whether blank or in rhyme, consists in a certain number of syllables; not in feet composed of long and short syllables." He works this out carefully—explaining that verses of double rhyme will always want one more syllable than verses of single; decasyllables becoming hendecasyllables, verses of eight syllables turning to nine, verses of seven to eight. "This must also be observed in blank verse." Then of the several sorts of verses. Our poetry, he thinks, admits, for the most part, of but three verses—those of ten, eight, or seven syllables. Those of four, six, nine, eleven, twelve, and fourteen are generally employed in masques and operas and in the stanzas of lyric and Pindaric odes. We have few entire poems composed in them; though twelve and fourteen may be inserted in other measures and even "carry a peculiar grace with them." In decasyllabic verse two things are to be considered—the seat of the accent and the pause. The pause ought to be at the fourth, fifth, or sixth syllable. The strongest accent must be on the second, fourth, and sixth. But he says nothing about accent in the last four places; indeed he is less explicit about the second half of the line throughout. And he says less about accent generally than about pause, though he is sure that "wrong placing" of it is as great a fault in English as a false quantity was in the classical languages. To make a good decasyllable you must be careful that the accent is neither on the third nor on the fifth—a curious crab-like way of approaching the subject, but bringing out in strong relief the main principle of all this legislation, "Thou shalt not." The verse of seven syllables, however, is most beautiful when the strongest accent is on the third.

More curious still is his way of approaching trisyllabic metres. As such, he will not so much as speak of them. "Verses of nine and eleven syllables," it seems, "are of two sorts." "Those accented on the last save one" are merely the redundant eights and tens already spoken of. "The other [class] is those that are accented on the last syllable, which are employed only in compositions for music, and in the lowest sort of burlesque poetry, the disagreeableness of their measure having wholly excluded them from grave and serious subjects." These are neither more nor less than anapÆstic three- and four-foot verses; though for some extraordinary reason Bysshe does not even mention the full twelve-syllable form under any head whatever. I suppose the "lowness and disagreeableness" of the thing was too much for him, and as he had disallowed feet he had, at any rate, some logical excuse in making nothing of them. He admits triplets in heroic, and repeats his admission of Alexandrines and fourteeners. "The verses of four or six syllables have nothing worth observing," though he condescends to give some from Dryden.

Under the head of "Rules conducing to the beauty of our versification," and with the exordium, "Our poetry being very much polished and refined since the days of Chaucer, Spenser, and other ancient poets," we find that you must avoid hiatus; always cut off the e of "the" before a vowel; never allow such collocations as "thy iambics" or "into a book"; never value such syllables as "amazÈd" and "lovÈd," but always contract them; avoid alliteration; never split adjective from substantive, or preposition from verb, at the end of a line. "Beauteous" is but two syllables, "victorious" but three. You must not make "riot" one syllable as Milton does.[142] You may contract "vi'let" and "di'mond," and if you do, should write them so. "Temp'rance," "diff'rent," etc., are all right; and you may use "fab'lous" and "mar'ner." But Bysshe acknowledges that "this is not so frequent." And he rejects or doubts some of the more violent and most hideous apostrophations, such as, "b'" for "by," but has no doubt about "t'amaze," "I'm," "they've," and most others. Rhyme is not very fully dealt with, but for the most part correctly enough—so far as Bysshe's principles go. Stanzas of "intermixed rhyme" (like rhyme-royal, the octave, and the Spenserian) "are now wholly laid aside," for long poems at least. Shakespeare invented blank verse to escape "the tiresome constraint of rhyme." Acrostics and anagrams "deserve not to be mentioned."

If any one has read this account carefully he will perceive at once what Bysshe's ideals and standards are. They put the strict decasyllabic couplet, with no substitution, no overrunning of lines, a fixed middle pause, and as nearly as possible an unvaried iambic cadence, into the principal place—if not quite the sole place of honour—in English poetry. They frown upon stanzas, upon varied metres of any kind, and even upon unvaried anapÆestic or "triple" measures. Strict syllabic scansion, with a consideration of accent, is the only process allowed; and even Dryden, just dead, and still regarded as the greatest of English poets, is directly though gently reproached for too great variety and laxity, as well as indirectly blamed for using "low" and "disagreeable" forms. The author seems to have been a very obscure person, of whom little or nothing is known; but any one who really knows English poetry will see that he practically expresses the mind that dominated it during almost the whole of the eighteenth century.

Either from Bysshe's starting the question; or from the same general influence which made him start it; or from the supposed tendency, not to be too hastily accepted, of a lull in creative poetry to be followed by an access of criticism—there is, from this time onward, no lack of prosodic work. John Brightland, in an English Grammar (1711), opposed Bysshe on the subject of accent; and he was also spoken of disparagingly by Charles Gildon, who produced two books, The Complete Art of Poetry (1718) and the Laws of Poetry (1721). Gildon was a pert and rather superficial writer who deservedly came under the lash of Pope; and, though neither quite ignorant nor quite stupid, he initiated a course of error which has never yet been stopped, by confusing prosody with music and arranging it by musical signs. Between Bysshe and his two critics Dr. Watts had, in the preface to his Horae Lyricae, given some prosodic remarks indicating discontent with the monotony of the couplet, an appreciation (not unmixed with criticism) of Milton, and other good things. But, before long, the question whether Accent or Quantity governs English verse—often complicated with the attempt to interpret this latter by musical notation—absorbed an altogether disproportionate amount of attention. The works of Pemberton (1738), Mainwaring, Foster, Harris, Lord Kames, Webb, and Say (1744) must be consulted by exhaustive students of the subject, and will be found duly commented upon in the larger History by the present writer. But they hardly need detailed notice here, any more than the later lucubrations of Lord Monboddo, Tucker, Nares, Fogg, and others. Their general tendency—which was indeed, as has been said, the general tendency of the century, correctly harbingered by Bysshe—was to concentrate attention on the heroic line, and indeed to regard it as strictly iambic, trisyllabic feet being wholly rejected, and even trochaic substitution either rejected likewise, as by Pemberton, or regarded as a more or less questionable licence. But the subject was also handled by persons of more literary importance, and in some cases, though not in all, of more insight and more knowledge.

The most remarkable exponent of the general prosodic ideas of the century is undoubtedly Dr. Samuel Johnson, who, though he wrote no special prosodic treatise, dealt with the subject in his Dictionary, in the Rambler (especially in connection with Milton), and in his Lives of the Poets. Except that Johnson does admit feet—or at least their names—his doctrine in the Dictionary hardly differs from Bysshe's as to the syllabic norm of lines, the strict regularity of accent constituting "harmony," and the duty of compounding superfluous syllables by elision, synaloepha, etc. He applies these doctrines in the Lives, and still more in his papers on Milton, Spenser, etc., in the Rambler. The spondees in Milton's lines—

and the trochees in his

and in Cowley's

are condemned as "inharmonious." He objects to Milton's "elisions"—that is to say, the devices necessary on his own system to avoid trisyllabic feet—and so to these feet themselves. He thinks the Spenserian stanza, Lycidas, and the end of Comus bad, because the lines and rhymes are not regularly arranged. In short, he is an unhesitating—and almost the greatest—believer in the sheer, alternately accented, middle-paused, syllabically limited decasyllable; though, with perhaps inevitable inconsistency, he does admit that, without variation of accent, the series of sounds would be not only very difficult but "tiresome and disgusting," while maintaining at the same time stoutly that this variation "always injures the harmony of the line considered by itself."

The inconveniences of this rigid system were not, however, entirely unnoticed. At an uncertain time, but probably between 1740 and his death, the poet William Shenstone urged, in a posthumously published Essay, the beauty of what he called "virtual dactyls"—that is to say, words like "watery" and "tottering,"—distinctly arguing that "it seems absurd to print them otherwise than at full length"—the "otherwise" being the established practice, based upon definite theory, of the century.

Johnson's friend the elocutionist Sheridan, in his Art of Reading (1775), calls it absurd (as it certainly is) to regard "echoing" as metrically "ech'ing." And, later, the poet Cowper, though using ambiguous and irresolute terminology on the subject, admits the "divine harmony" of Milton's "elisions"—by which, he explains in the most self-contradictory way, "the line is lengthened."

While much earlier, at the very middle of the century, John Mason, a little-known dissenting minister, who was, like Sheridan, a teacher of elocution, quoting and scanning the lines—

observes that this, "though it increases the number of syllables, sweetens the flow of the verse," "gives a sweetness that is not ordinarily found in the common iambic verse." It would be impossible to state more correctly or more definitely the case for the equivalent substitutional trisyllabic foot in English. But, as we shall see, it was to be nearly two generations before considerable poets boldly adopted (even then not always distinctly championing) the idea, and an entire century, if not more, before the principle was thoroughly accepted and understood.

Two deliberate prosodists, in two books published within a twelvemonth of each other, are memorable as (if not exactly starting) formulating, in a more elaborate way than had ever been done before, the one a mischievous and false, the other the only true method of dealing with prosody. Joshua Steele, in his Prosodia Rationalis (1775), is not always wrong; and William Mitford is not by any means invariably right—in fact, he partly shares Steele's error. But his Harmony of English Verse (1774) is even then to a great extent, and in its second edition, thirty years later, much more, occupied with a careful historical inquiry as to the actual successive forms of his subject from the earliest period. At first he had not even Tyrwhitt's invaluable Chaucer—which appeared in the year after Steele's book—to guide him: later he availed himself of the great accessions to the study of Middle and Elizabethan English which the intervening generation had seen. And so, though he believed too much in accent, and relied too much on the dangerous assistance of music, he frequently came right. He has no doubt (as it is astonishing that an historical student should have any doubt) about trisyllabic feet; he likes what he calls "aberration of accent," i.e. trochaic substitution; and he shows the possession of a fineness and cultivation of ear not as yet noticeable in any English prosodist, by observing the presence of anapÆstic rhythm in the revived alliterative verse of Langland. Except the inadequate and perfunctory, as well as of necessity merely inchoate, sketches of Webbe and Puttenham, this was the first attempt really to take English poetry into consideration when studying English prosody; and it had its reward.

On the other hand, Steele, who has been followed by many other prosodists of the same school, entirely neglected the historical contents of his subject, approaching it absolutely a priori, deciding that it is essentially a matter of music, and basing his scansions on purely musical principles. This led him to begin with an anacrusis in every case, and so to invert the whole rhythm of the line. He has been praised for his views on "time" in the abstract, and may deserve the praise; while he was certainly right in regarding pause as an important metrical constituent. But whatever merit there may be in his principles from an abstract point of view, his concrete practice is simply atrocious, and proves him to have had absolutely no ear for English verse whatever. He makes six feet or "cadences with proper rests," at least, and sometimes more, in every heroic line, so that he would scan one famous line thus—

and he arranges the opening lines of Paradise Lost for scansion thus—

It must be perfectly evident to any one who will read these examples, even to himself, but still more aloud, not merely that they entirely destroy the actual cadence and rhythm of the actual verses, but that they provide a new doggerel which is absolutely inharmonious, unrhythmical, and contrary to every principle and quality of English poetry. It would doubtless be possible to accommodate them with a tune; in fact, any one who has ever looked at a "set" song will see how they correspond to it. But then any one who has ever looked at a set song must, in a majority of cases, have been convinced at once that musical arrangement has nothing to do with prosodic.

It was inevitable that the "Romantic" movement—one of the principal causes and features of which was a demand for variety, while another was its disposition to return to older modes—should be largely concerned with prosody; but, with some notable exceptions, this concernment did not take the form of actual prosodic deliverances or discussions.

Gray, one of the chief precursors of the movement, had projected a regular history of English poetry, and has left invaluable notes under the general head of Metrum—notes in which he goes back, deliberately and directly, to Middle English, discovers therein the origin and nature of the metre of Spenser's February, etc., and has very good remarks about others. But it was not till the stir of the revolutionary period that much more was done, and even then more was done than said.

The German explorations of William Taylor of Norwich induced English writers to follow the German attempt at accentual hexameters; and another of the Norwich group, Frank Sayers, not merely wrote, but expounded and defended in prose, rhymeless metres of a choric character; both being—in part, if not mainly—revolts from the mechanical heroic couplet.

Before the end of the century, long before Coleridge published the explanatory note on Christabel metre, and not improbably before he had even thought of that note, Southey had not only used trisyllabic equivalence in his Ballads, but had formally and independently defended it as such in a letter to his friend Wynn.[143]

Wordsworth says very little about metrical detail in his famous Preface to the second edition of the Lyrical Ballads and its successors—appearing to think, and indeed in one place asserting, that "harmony of numbers" comes of itself to a person who has other poetical qualifications.

His two just-mentioned friends, however, lodged, at a slightly later period, two of the most important preceptist documents of English prosody, though they were documents differing very widely in the extent and character of their importance. These were Coleridge's note on the metre of Christabel, and Southey's Preface to the Vision of Judgment. The latter is too long to give, and is written from a mistaken point of view; but it, and the much-ridiculed poem which it accompanied, undoubtedly restarted the practice of attempting to write English hexameters, which has been continued, with some intervals and some episodes, but at times most busily, ever since. The former must be given at length, and some comment made on it:—

"The metre of the Christabel is not, properly speaking, irregular, though it may seem so from its being founded on a new principle, namely, that of counting, in each line, the accents, not the syllables. Though the latter may vary from seven to twelve, yet in each line the accents will be found to be only four. Nevertheless this occasional variation in number of syllables is not introduced wantonly, or for the mere ends of convenience, but in correspondence with some transition in the nature of the imagery or passion."

What Christabel metre really was has been expounded earlier, and its author's account of it is not a little surprising. When he called its principle "new" he must have forgotten—not exactly the Middle English writers, whom he very likely did not know, nor perhaps Gray, though the latter's remarks on Spenser's February were actually published before Christabel, but—Spenser himself and Chatterton (both of whom he certainly knew, if not Blake also), as well as the very ballad-writers whom he had himself imitated in the Ancient Mariner. His mention of "accents" and not "feet" argues an erroneous and inadequate theory which leaves much of the beauty of his own work unexplained; while it can be shown from the text itself that the variation of syllables, though metrically beautiful, often does not correspond at all with any special point of sense, passion, imagery, or anything else. But his practice more than cured any wound which his theory may have inflicted.

In comparison with Southey's and Coleridge's remarks, and still more with the practice of the latter in Christabel and the Ancient Mariner, the preceptist prosody of the extreme end of the eighteenth century, and the first third of the nineteenth becomes, except for exhaustive students of the subject, a mere curiosity, and not a very interesting one. Prosodic remarks, mostly erroneous or inadequate, found their way into popular handbooks, such as Walker's Dictionary (almost wholly wrong) and Lindley Murray's Grammar (partially right). The musical theories of Steele were taken up by others, such as Odell, Roe, and, above all, the republican lecturer Thelwall, who, escaping the consequences of his earlier extravagances, became a teacher of elocution. The new Reviews gave opportunity for occasional critical remarks on the subject—the most notable of which was the Quarterly review, by Croker, of Keats's Endymion,—usually embodying the cramped and ignorant doctrinairism of the preceding century. Southey's hexameters started a large amount of writing on that subject. In 1816 John Carey, compiler of the best-known Latin Gradus and author of many "cribs" and school editions, repeated most of the errors of Bysshe, but did grudgingly allow trisyllabic feet; and in 1827 William Crowe, a minor poet and Public Orator at Oxford, wrote a treatise of English Versification—good in method, but bad in principle—condemning the adjustment of very short to longer lines, etc. Nothing of this period comes in importance near to that second edition of Mitford (1804, with most of the historical matter added) which has been noticed.

But in 1838—after the appearance of Tennyson and Browning, but when no public attention had been paid to them—appeared the most elaborate, ambitious, and, partly at least, valuable work that had yet been written on the subject—the History of English Rhythms, by Edwin Guest, then Fellow, afterwards Master, of Gonville and Caius College, Cambridge. Guest took nearly two years between the publication of the first and second volumes of his book, and admittedly changed his opinions on some points, but his main theories are unmistakable. He goes entirely by accent, denying metrical quantity in English altogether, and imposing curious arbitrary rules (such as that two adjoining syllables cannot be accented without a pause) on accent itself. But he possessed an immense and truly admirable knowledge of English verse—Old, Middle, and Modern—up to his time; and he lavished this, in a manner useful, indeed invaluable, to the present day, on the support of general theories which, unfortunately, are quite unsound.

For Guest seems to have conducted his work under the influences of three different obsessions, no one of which he ever worked out thoroughly in all its bearings, which do not necessarily imply each other, and two of which are even rather contradictory.

The first[146] was the belief that our verse is wholly dependent upon accent, and that "the principles of accentual rhythm," whatever they are, govern it exclusively.

The second[147] was that the laws of English versification generally are somehow not only dependent on those of Old English versification, but identical with them, and always to be adjusted to them.

The third[148] was that, somewhere about the early thirteenth century, and increasingly till the end of the fourteenth, there took place a succession of alien invasions which never resulted in a coalescence or blending, but merely in the presence of two hostile elements; and that while the perfect English versifier will cling to the older and only genuine one, he must, if he does not so cling, give it up altogether, and have nothing to do with anything but "the rhythm of the foreigner."

Now what has been already and will be later given in this book seems to show that these propositions are in fact false.

In the first place, though accent plays a large part in English prosody, that prosody is as far as possible from being purely or exclusively accentual.

In the second, the oldest English poetry and its younger varieties are so utterly different that the same laws cannot, except per accidens, apply to them.

In the third, instead of two jarring elements, we find before us, from the thirteenth century, at least, onwards, a more and more distinct and harmonious blend of language, resulting, of necessity, in a more and more distinct and harmonious blend of prosody.

But there is also a fourth principle, which he adds to, rather than deduces from, the other three:—

That the collocation of accented and unaccented syllables forms sections,[149] which in turn form, and into which can be reduced, all English verse.

On these principles he went through the whole body of English verse from Caedmon to Coleridge, arranging it with infinite trouble on the "sectional" system, and classifying the verses as those of "four accents," those of "five," and so on, with suitable distinctions for stanzas, etc. Unfortunately—to mention only the crowning and fatal fault which makes mention of all others in such a book as this unnecessary—he finds himself in perpetual conflict with the practice of the greatest English poets in their most beautiful passages. Shakespeare and Milton go "contrary to every principle of accentual rhythm," and use devices which "they have no right" to use. Coleridge and Burns employ sections which "have very little to recommend them." Spenser's verse is "wanting in good taste," and Byron's versification "has never been properly censured." It may seem incredible that a writer of learning and acuteness should not have seen the absurdity of his position when he tells beautiful poetry—sometimes admitted by himself as such—that it has no business to be beautiful because it does not suit his rules. But the fact disposes of him, and of the rules themselves, without its being necessary—though it would be easy—to prove their want of intrinsic justification.