Not unfrequently, also, this licence is caused by the rhyme, as in the following examples:

The same manner of treatment may be found applied to words which end in -lyng, -esse, -nesse, and similar syllables, and which have a secondary accent on the last syllable and the chief accent on the preceding root-syllable.

In Modern English verse the absence of a thesis between two accented syllables sometimes arises from phonetic conditions, i.e. from the pause which naturally takes place between two words which it is difficult to pronounce successively. This pause supplies the place of the missing thesis, as e.g. in the following lines:

In other instances the emphasis laid upon a particular word compensates for the absence of the unaccented syllable, especially, if the accented syllable is long: e.g.

This licence is of frequent occurrence in even-beat measures.

§ 87. Another metrical peculiarity caused by the influence of the rhythm is the lengthening of a word by the introduction of an unaccented extra syllable, commonly an e, to supply a thesis lacking between two accented syllables.

§ 88. Another deviation from the regular iambic line is the inversion of the rhythm; i.e. the substitution of a trochee for an iambus at the beginning of a line or after the caesura. The rhythmical effect of this licence has some resemblance to that of the suppression of anacrusis. In both cases the rhythmic accent has to yield to the word-accent. But while in the latter case the whole verse becomes trochaic in consequence of the omission of the first syllable, in the former the trochaic cadence affects one foot only (generally the first), the rest of the verse being of a regular iambic rhythm. Hence the number of syllables in each line is the same as that in all the other regular lines (including those with level stress), whereas verses with suppressed anacrusis may easily be distinguished from the former by their smaller number of syllables. On the other hand, the number of syllables (being the same in both cases) affords no help in distinguishing between change of word-accent and inversion of rhythm. Which of these two kinds of licence is to be recognized in any particular case can be determined only by the position which the abnormal foot occupies in the line. Inversion of rhythm (i.e. the substitution of a trochee for an iambus) occurs, as a rule, only at the beginning of a line or hemistich, where the flow of the rhythm has not begun, so that the introduction of a trochee does not disturb it. If, therefore, the discord between normal word-stress and iambic rhythm occurs in any other position in the line, it must be regarded as a case of level stress.

The following examples will serve to illustrate the difference between these three species of metrical licence:

Omission of anacrusis:

Level stress:

Inversion of rhythm:

Inversion of rhythm may be caused in the interior of a rhythmical series only when a particularly strong emphasis is laid upon a word, e.g. to express an antithesis or for similar reasons:

We may distinguish between two kinds of inversion of rhythm, viz. (i) natural inversion, and (2) rhetorical inversion. The former is caused by word-accent, the latter by the rhetorical accent, as illustrated by the last examples. The second kind differs very clearly from level stress, as the word in question or the first syllable of it (see the second line of the following quotation) is to be uttered with an unusually strong emphasis, e.g.:

In the second example inversion of rhythm occurs (as it often does) twice over, viz. at the beginning of the verse and after the caesura.

Not unfrequently also two inversions of rhythm follow immediately upon one another, e.g.:

Such verses, however, may also be looked upon as instances of the omission of anacrusis combined with epic caesura.

This would be the only admissible explanation in verses the first accented word of which is a word which usually does not bear an accent or is not accented rhetorically, e.g.:

But in a line with an emphasized first word inversion of rhythm is the more probable explanation: e.g.

§ 89. Disyllabic or polysyllabic thesis. Another important deviation from the regular iambic rhythm, which is clearly to be distinguished from the double thesis caused by inversion of rhythm, consists in the use of two or sometimes even more unaccented syllables instead of one to form a regular thesis of a verse. This irregularity, which is almost as common in Modern English as it is in Early English poetry, may occur in any part of the verse. If it occurs in the first foot, it may be called disyllabic or polysyllabic anacrusis, as in the following examples:

This metrical licence may occur also immediately after the caesura, e.g.:

It most frequently occurs, however, in the interior of the rhythmical sections, and there it is found in any of the feet, except the last, as will be seen by the following examples:

§ 90. Unaccented extra syllables are found also before a caesura or at the end of the line. In the former case they constitute what is known as epic caesura, in the latter they form feminine or double endings (if there is only one extra syllable) or tumbling endings (if there are two extra syllables). In both cases this irregularity is softened or excused, so to say, by the pause, except where the accented or masculine ending of the hemistich is required by the very nature of the metre, viz. in the first acatalectic half of the Septenary line. It does, however, not unfrequently occur in some Early Middle English poems written in Septenary metre, e.g. in the Moral Ode and several others, but this may be only owing to want of skill or carelessness on the part of the authors of these poems. The following example taken from the Moral Ode may serve to illustrate this:

§ 91. Double or feminine endings are more frequent than epic caesuras, especially in Middle English poetry. They become rarer, however, in the course of time in Modern English in consequence of the gradual disappearance of the inflexional endings, e.g.:

The two last quotations are noteworthy because the number of extra syllables after the last accented one is two, three, or even four, a peculiarity which is one of the characteristics of Fletcher’s versification. Other poets, e.g. Shakespeare, preferred feminine endings in some periods of their literary career, so that it is possible to use the proportion of masculine and feminine endings occurring in a play, compared with others of the same poet, as a means of ascertaining the date of its origin.

It is also to be observed that in certain epochs or kinds of poetry feminine endings are more in favour than in others. In the eighteenth century they are very scarce, whereas they become more frequent again in the nineteenth century. Byron and Moore especially use them copiously in their satirical and humorous poems to produce burlesque effects.

§ 92. Another metrical licence also connected with the end of the line is what is known as the enjambement or run-on line—that is to say, the carrying over of the end of a sentence into the following line.

or a verb from its subject or object, formed by a monosyllabic word:

But if, on the other hand, two closely connected parts of a sentence are each of them long enough to fill up two measures, they may be separated by enjambement:

The admissibility or inadmissibility, however, of run-on lines depends on many different and complicated considerations, for which the reader may be referred to ten Brink, Chaucer’s Sprache und Verskunst, §§ 317–20, and to our own larger work, vol. ii, pp. 59–62.

In Shakespeare’s versification, and probably also in that of other poets, the more or less frequent use of run-on lines is characteristic of certain periods of their literary career, and is therefore looked upon as a valuable help in determining the date of the different plays (cf. § 91). The largest percentage of run-on lines probably occurs in Milton’s epics.

§ 93. The judicious use of run-on lines is often resorted to for the purpose of avoiding monotony. Another metrical licence connected with the line-end, which is adopted for the same purpose, is rhyme-breaking. This occurs chiefly in rhyming couplets, and consists in ending the sentence with the first line of the couplet, instead of continuing it (as is usually done) till the end of the second line. Thus the close connexion of the two lines of the couplet effected by the rhyme is broken up by the logical or syntactic pause occurring at the end of the first line. This is used rarely, and so to say unconsciously, by the earlier Middle English poets, but is frequently applied, and undoubtedly with artistic intention, by Chaucer and his successors. The following passage contains examples both of rhyme-breaking and of the more normal usage:

Rhyme-breaking may, of course, also take place in other metres, as e.g. in four-foot iambic verses:

Chapman, in his translation of Homer, often uses it in Septenary verses as well as in five-foot iambic verses. In certain stanzas rhyme-breaking at particular places is a strict rule, as e.g. in the Rhyme-Royal stanza (a b a b . b c c), in the ballade-stanza of eight lines (a b a b . b c b c), and also between the two quatrains of the regular Italian sonnet.

On the other hand this licence is rare in the works of the poets of the eighteenth century who wrote under French influence, and in modern times (especially at the present day) it seems to be rather avoided than intentionally admitted.

§ 94. Another peculiarity of frequent but irregular occurrence in even-beat verse is alliteration, a feature which is derived from the old native metre, and is still (consciously or unconsciously) employed by many poets as an ornament of their verse.

In the thirteenth and fourteenth centuries it is mostly used merely to give a stronger emphasis to those words of the verse which bear the logical and rhythmical accent,[131] but even as early as this we can observe a decided predilection for accumulated alliteration. Sometimes the same alliterative sound is retained through several successive lines. In other instances a fourth alliterating word is admitted in the line (as in the example referred to above). In the fifteenth and sixteenth centuries this striving after accumulation of alliteration was carried to such a length that it became a rule that as many words in the line as possible, whether accented or not, should begin with the same letter. This accounts for King James VI’s metrical rule quoted above (p. 89), that in ‘Tumbling verse’ the line is to be ‘literal’. Even Chaucer, in spite of his well-known hostile attitude to regular alliterative poetry,[132] allowed his diction to be influenced strongly by it, e.g.:

This accumulation of alliterative sounds occurs in the works of many Modern English poets, some of whom, as Peele and Shakespeare, have themselves ridiculed it, but were unable, or were not careful, to avoid it altogether in their own practice.

For particulars see Neuengl. Metrik, pp. 68–76, and the following treatises:

CHAPTER VII
THE METRICAL TREATMENT OF SYLLABLES

§ 95. As the root-syllables of words (leaving out of account the words of Romanic origin) almost universally retain their full syllabic value, whether occurring in arsis or in thesis, they require no notice in this chapter. We therefore confine our remarks to the formative and inflexional syllables, which, though as a rule found only in thesis, admit of being treated metrically in three different ways. (1) A syllable of this kind may retain its full value, so as to form by itself the entire thesis of a foot. (2) It may be slurred, so that it combines with another unaccented syllable to form a thesis. (3) It may lose its syllabic value altogether, its vowel being elided and its consonantal part (if it has any) being attracted to the root-syllable. By the last-mentioned process, as is well known, the number of inflexional syllables has been greatly reduced in Modern as compared with Middle and Old English.

-e in a certain number of inflexional forms of the verb (as e.g. in the inf. and in the past part. of strong verbs, where n is dropped), and of the substantive and adjective, and as an ending of Romanic words, &c.

Of all these endings only the comparative and superlative suffixes -er, -est are preserved in an unreduced state in Modern English. The final -e has disappeared in pronunciation (with some exceptions occurring in Early Modern English). The important suffixes -en, -es, -ed, -est (2nd pers. sing.), -eth (for which -s, the northern ending, instead of -es, is commonly substituted) have been contracted through syncope so as to form one syllable with the root, except where the nature of the final consonant of the stem prevents syncope, e.g. in -es and -est after sibilants, in -ed after dentals, in -en after v, s, t, d, k (as in houses, ended, risen, written, hidden, broken, driven). As, however, these are always full syllables they may here be disregarded. The ending -edest has been shortened into -edst.

It is to be observed that the syncopation of the vowel (e) of the inflexional endings was not so nearly universal in Early Modern English as it is at present; and further, that it is still much less prevalent in poetry than in prose, because the poets for metrical reasons often preserve the fuller endings when in ordinary speech they are no longer used.[133] In examining the metrical treatment of the Early English inflexional endings, we shall therefore have occasion to consider the usage of the present day, notwithstanding the fact that some of these endings are obsolete in modern prose.

The chief difference between Early and Modern English with regard to the treatment of the inflexions is that in Early English poetry the full pronunciation is the rule—in accordance with the practice in ordinary speech--and the syncopation of the vowel (e, rarely i or u) is the exception; while in Modern English it is the shortened pronunciation that is normal, the full syllabic form being used only exceptionally as a poetic licence.

§ 96. The first point that requires notice is the treatment of the unaccented e of words of three and four syllables in Middle English. The following observations are founded on those of ten Brink, Chaucer’s Sprache und Verskunst, § 256.

1. If each of the two last syllables of a trisyllabic word has an unaccented e, one of them is generally elided or slurred over under the influence of the rhythmical accent. Thus the past tense singular of the weak verbs clepede, werede, makede, lovede may be scanned either clepte, werde, made, lovde, or cleped, wered, maked, loved. Just in the same way the plural forms clepeden, makeden, &c., may be read either clepten, maden, &c., or cleped, maked, &c.; likewise the plural endings of nouns faderes, hevenes may be pronounced fadres, hevnes or faders, hevens. In Early Middle English, however, and also in the language of Chaucer, exceptions to this rule are found, trisyllabic scansion occurring chiefly in the plur. pret., e.g.:

The e following upon an unaccented syllable which is capable of receiving the accent, whether in a word of Teutonic or Romanic origin, is commonly mute. E.g. banere, manere, lovere, ladyes, housbondes, thousandes are generally to be pronounced in verse (as, indeed, they were probably pronounced in prose) as, baner, maner, lover, ladys, housbonds, thousands. But this e, on the other hand, not unfrequently remains syllabic, especially in the Ormulum, where it is dropped only before a vowel or h. E.g. cneolénn meoklík(e) annd lútenn 11392, meocnéss(e) is þrínne kíness 10699, Forr án godnéss(e) uss háveþþ dón 185. Before a consonant or at the end of a line, however, it is always sounded: Ennglísshe ménn to láre 279, God wórd and gód tiþénnde 158, forrþí birrþ áll Cristéne fóllc 303. Goddspélless hállȝhe láre 14, 42, 54, þa Góddspelléss neh álle 30. Other examples are: And þó þet wéren gítserés Moral Ode, MS. D. l. 269; For thóusandés his hóndes máden dýe Chauc. Troil. v, 1816; enlúminéd id. A B C 73.

In words of four syllables a final e which follows upon an unaccented syllable with a secondary accent may at pleasure either become mute or be fully pronounced. So words like óutrydère, sóudanèsse, émperòures, árgumèntes may be read either as three or four syllables. Examples of e sounded: Bifórr þe Rómanísshe kíng Orm. 6902; Annd síkerrlíke trówwenn ib. 11412; þurrh hállȝhe góddspellwríhhtess ib. 160; Till híse lérninngcníhhtess ib. 235; Annd þúrrh þin góddcunndnésse ib. 11358; An Gódd all únntodǽledd ib. 11518; I glúternésse fállenn ib. 11636; þurrh flǽshes únntrummnésse ib. 11938; in stránge ráketéȝe Moral Ode, 281; a thíng(e) unstédeféste ib. 319; bifóre héovenkínge ib. 352, &c. Examples of e mute: And þá, þe úntreownéss(e) dide þán Moral Ode, 267; þéosternéss(e) and éie ib. 279. Orm has it only before vowels or h: Forr són se glúternéss(e) iss dǽd 11663, &c.

§ 97. Special remarks on individual inflexional endings.

The syncopation of the -e in the adverbial -es is indicated, as is well known, by the spelling, in certain cases: e.g. in else, hence, thence, whence (instead of the Middle English forms elles, hennes, &c.); but even in words where it is preserved in writing, as e.g. in whiles, unawares, it has become mute and has, as a rule, no metrical value in Modern English poetry. The archaic certes, however, is still always treated as a disyllabic, e.g.

§ 98. The ending -en (plur. nom. of nouns; prepositions; infinitive; strong past part.; plur. pres. and pret. of verbs) is in Middle English (a) commonly treated as a full syllable during the first period, and later on mostly, although not always, to avoid hiatus, before vowels and h, e.g. His éyen stépe Chauc. Prol. 201; Bifórenn Críst allmáhhtig Gódd Orm. 175; Befóren ánd behýnde Alexius, ii. 393; abóven álle nációuns Chauc. Prol. 53; þú schalt béren hím þis ríng Floris and Blanch. 547; Fór to délen with no swích poráille Chauc. Prol. 247; bifrórenn Orm. 13856; forlórenn ib. 1395; Sche wás arísen ánd al rédy díght Chauc. Kn. T. 183; Hir hósen wéren óf fyn scárlet reed id. Prol. 456; For thís ye knówen álso wél as I ib. 730; Swa þátt teȝȝ shúlenn wúrrþen þǽr Orm. 11867; þatt háffdenn cwémmd himm í þiss líf ib. 210; Ál þet wé misdíden hére Moral Ode, 99; (b) syncopated or slurred, especially in later times, after the n has been dropped already in prepositions and verbal inflexions, e.g. His póre féren he delde Alexius, ii. 210; Hálles and bóures, óxen and plóugh ib. 12; Bifórr þe Rómanísshe kíng (instead of biforenn) Orm. 6902; Hastów had fléen al nýght Chauc. Manc. Prol. 17; She bóthe hir yónge chíldren untó hir cálleþ id. Cl. T. 1081; is bórn: þat wenten hím bifórn id. Man of Lawes T. 995–7; withínne a lítel whýle id. Sq. T. 590; And únderfóngen his kínedóm Flor. and Blanch. 1264; þei máde sówen in þát cité Alexius, i. 577; Bíddeþ his mén cómen him nére ib. 134; Hórn: i-bórn King Horn, 137–8; forlóren: Hórn ib. 479–80; Was rísen and rómede Chauc. Kn. T. 207; my líef is fáren on lónde id. N. Pr. T. 59; And fórth we ríden a lítel móre than páas id. Prol. 825; þei drýven him ófte tó skornínge Alexius, i. 308; þei rísen alle úp with blíþe chére ib. 367; þei cásten upón his cróun ib. 312; And wíssheden þat hé were déd Alexius, ii. 335, &c.

In Modern English this ending is much more rare, and is hardly ever used as a full syllable of the verse. The plural ending -en of the substantive occurs now and then in Wyatt’s and Surrey’s verse, as e.g. in éyen instead of éyes, both in rhyme, e.g. éyen: míne Sur. 14, and in the interior of the line, ib. 126, 128; Wyatt 8, 17, &c.

Prepositions ending in -en are scarcely ever used now; sometimes the archaic withóuten is to be met with in some Early Modern English poets, and then, of course, as a trisyllable: withóuten dréad Sur. 95; withóuten énd Spenser, F. Q. II. ix. 58. The obsolete infinitives in -en may also be found sometimes in the writings of the same and other early Modern English poets: in váyn: sáyen Sur. 31; his flócke to víewën wíde Spenser, F. Q. I. i. 23; to kíllën bád Shak. Pericles, II. Prol. 20. Likewise certain antiquated plural forms of the verb in -en: dischárgën cléan Sur. 30; fen: lífedën Spenser, F. Q. II. x. 7; and wáxën ín their mírth Shak. M. N. Dr. II. i. 56.

§ 99. The comparative and superlative endings -er, -est are, as a rule, syllabic. Hórn is fáirer þáne beo hé King Horn, 330; No lénger dwélle hý ne mýghte Alexius, ii. 85; But ráther wólde he yéven Chauc. Prol. 487.

These endings are treated, moreover, as full syllables in the unaccented rhymes Hǽngest: fǽirest Layamon, 13889–90; Hǽngest: héndest ib. 13934–5. If an inflexional -e is added to such words, so as to make them trisyllables, it is commonly elided or apocopated, e.g. Fór he ís the fáireste mán Horn, 787; hire grétteste óoth Chauc. Prol. 120; The férreste in his párisshe ib. 494. Slurring or syncopation takes place in the following examples, Sche móst wiþ hím no lénger abíde Sir Orfeo, line 328; No lénger to héle óf he bráke Alexius, ii. 127; more rarely in the superlative, Annd állre láttst he wúndedd wáss Orm. 11779, 11797; Was thóu not fárist of ángels álle? Towneley Myst. p. 4.

In Modern English these endings are treated similarly. The comparative-ending -er is mostly syllabic on account of the phonetic nature of the final r, and even if slurred, it does not entirely lose its syllabic character, e.g.: