As these types may be varied by resolutions in the same way as the primary types, there arise various additional formulas such as the following:

Other variations may be effected by disyllabic or even polysyllabic theses in the beginning (‘anacruses’) or in the middle of the verse instead of monosyllabic theses.

Apart from these another frequently occurring variation of type C must be mentioned which corresponds to the formula (×)×̀×–́×–́×, and may be designated (with Professors Paul and Luick) as type C^a, because the position of its accented syllables points to type C, while on the other hand it bears a certain resemblance to type A.

The following examples, many of which have been quoted before by Luick, may serve to illustrate these types of short lines or rather hemistichs and their combination in couplets or long lines, in which a normal hemistich is often followed by a lengthened one and vice versa:

Types with resolutions:

The first hemistich of the last line offers a specimen of a variation of the ordinary types with feminine endings (chiefly of A, C, and C^a), designated by Prof. Luick as A₁, C₁, C^a₁, and showing the peculiarity that instead of the ending –́× somewhat fuller forms occur, consisting either of two separate words or of a compound word, and thus corresponding either to the formula –́×̀, or, if there are three syllables, to the formula –́××̀, or in case of a resolution (as in the above example) to the formula –́×⏑́×. We differ from Prof. Luick, however, in admitting also endings corresponding to the formula ⏑́×̀×.

As a rule, if not always, such forms of verse are occasioned by the requirements of rhyme. This is not the case, it is true, in the following purely alliterative line:

but in other verses it is so, e.g.:

and similarly (not corresponding to –́××̀, as Prof. Luick thinks):

but the formula –́××̀is represented by the following verses:

In the last but one of these examples this accentuation is corroborated in the Jesus College MS. by the written accent on the word gódne, whereby not only the rhyme -lyng: king is shown to be an unaccented one, but at the same time the two-beat rhythm of the hemistich is proved as well as that of the preceding hemistich. Moreover, the alliteration in all these examples is a further proof of the two-beat character of their rhythm.

§ 48. It was owing to the use of these two more strongly accented syllables in each verse which predominate over the other syllables, whether with secondary accents or unaccented, that the poets, who wrote in this metre, found it possible to regard the different kinds of verse they employed as rhythmically equivalent. These were as follows: (1) purely alliterative lines with hemistichs of two stresses, (2) extended lines of this kind with secondary accents in the middle of the hemistich, (3) rhyming-alliterative or merely rhyming lines with a feminine ending and a secondary accent in the middle of the verse, or with a masculine ending and two secondary accents, one on the last syllable, as is also the case with the corresponding verses mentioned under the second heading. These two last-mentioned verse-forms are very similar to two popular metres formed on the model of Romanic metres. The former of them—the hemistich with three stresses (one of which is secondary) and feminine ending, together with the much rarer variety that has a masculine ending—resembles the sections of the Alexandrine; and the hemistich with a masculine ending (more rarely a feminine) and four stresses (two of which have secondary accents only) is similar to the short four-beat couplet, and also to the first section of the Septenary line (the second section being similar to the former three-beat group). It is, therefore, not to be wondered at that this metre of Layamon in its different forms (that of the purely alliterative line included) is in several Middle English poems, chiefly in The Bestiary, employed concurrently (both in separate passages and in the same passage) with the above-mentioned foreign metres formed on Romanic or mediaeval-Latin models. By this fact the influence of the Romanic versification on the origin and development of this form of the native verse gains increased probability.[102]

The limits of our space do not permit of further discussion of this peculiar metre, which, as presented in the extant examples, appears rather as in process of development than as a finished product, and of which a complete understanding can be attained only by elaborate statistical investigation.

C. The progressive form of the alliterative line, rhymed throughout. ‘King Horn.’

§ 49. The further development of the Layamon-verse is very simple and such as might naturally be expected from its previous history.

The use of final rhyme becomes constant, and consequently alliteration, although remnants of it still are noticeable even in short lines connected together, becomes more and more scarce.[103]

The unaccented syllables are interposed between the accented ones with greater regularity; and among the unaccented syllables the one (or, in some sub-species of the verse, more than one) which is relatively stronger than the rest receives full metrical stress, or at least nearly approaches the fully-stressed syllables in rhythmical value.

This form of the metre is represented by a short poem[104] consisting of only twelve lines, belonging to the first half of the thirteenth century, and by the well-known poem King Horn[105] (1530 lines) which belongs to the middle of the same century.

The prevailing rhythmical form of this poem is exemplified by the following verses, which for the sake of convenience we print here, not in the form of couplets (as the editors, quite justifiably, have done), but in that of long lines as they are written in the Harleian MS.:

This form occurs in more than 1300 out of the 1530 short lines of which the poem consists. It is evident that the rhythm of these lines is nearly the same as in the following taken from earlier poems:

If those syllables which have the strongest accent in the unaccented parts of these verses are uttered a little more loudly than was usual in the alliterative line the rhythm becomes exactly the same as in the corresponding verses of King Horn, where the three-beat rhythm already has become the rule.

This rule, however, is by no means without exceptions, and even the old two-beat rhythm (which may have been the original rhythm) is, in the oldest form of the poem, sometimes clearly perceptible, rarely, it is true, in both hemistichs, as e.g. in the following line:

but somewhat oftener in one of them, as in the following:

Of this type of verse a great many examples are of course to be met with in the earlier alliterative poems:

The third type (three beats with masculine ending), which is of rarer occurrence, is represented by the following lines:

As corresponding lines of earlier poems may be quoted:

The fourth type (four beats with masculine ending), which occurs somewhat oftener, has the following form:

The corresponding rhythm of the earlier poems occurs in verses like:

The fifth type (four beats with feminine endings) is represented by the following verses:

As corresponding verses of earlier poems we quote:

The circumstance that these different types of verse occur in different poems promiscuously makes it evident that they must all have been developed from one original rhythmical form. It is clear that this fundamental type can only be found in the old two-beat alliterative hemistich, the more so as this kind of verse is the very metre in which the earlier poems Byrhtnoth and Be Dōmes Dæge for the greatest part are written, and which is exemplified in about a third part of the poetical piece of the Saxon Chronicle of 1036 and a fifth part of the later-piece of 1087, and again very frequently in Alfred’s Proverbs and in Layamon’s Brut, and which still can be traced as the original rhythm of King Horn

§ 50. The evidence of the metre of this poem, showing its affinity to the alliterative line and its historical origin from it, is so cogent that it is unnecessary to discuss the theories of Prof. Trautmann and the late Dr. Wissmann, both of whom, although from different points of view, agree in ascribing a four-beat rhythm to this metre.[106]

The frequent use again in this poem of the types of line occurring in Layamon’s Brut, as pointed out by Prof. Luick (l. c.), puts the close connexion of the metre of King Horn with that form of the alliterative line beyond doubt. We cannot, however, in conformity with the view we have taken of Layamon’s verse, agree with Prof. Luick in assigning a secondary accent to the last syllable of the feminine ending of the ordinary three-beat verse, in which the greater part of King Horn is written. Prof. Luick himself does not insist upon that particular point so strongly for this poem as he did for the earlier poems written in a similar metre.

The following examples serve to show that the same extended types of line which were found to be the commonest in Layamon’s Brut (cp. p. 77) recur as the most usual types also in this poem:

A + C: Álle bèon he blíþe | þat tò my sóng lýþe! 1–2.

A + A: A sáng ihc schàl ȝou sínge | of Múrrȳ̀ þè kínge. 2–3.

A + A: He fónd bì þe strónde, | aríued òn his lónde, 35–6.

B + C: Àll þe dáy and àl þe nī́ȝt, | tìl hit spráng dái lìȝt. 123–4.

B + B: Fàirer nis nón þàne he wás, | hè was bríȝt sò þe glás. 13–14.

C + C: Bì þe sé síde, | ase hè was, wóned (⏑́×) ríde. 33–4.

C + A: Of þìne méstére, | of wúde and òf rivére. 229–30.

D + A: Schípes fíftène | with sárazìn[e]s kéne. 37–8.

C + A: Þe chìld him ánswérde, | sóne so hè hit hérde. 199–200.

B + E: Hè was whít sò þe flúr, | róse-rèd was hìs colúr. 15–16.

In most cases we see that identical or similar types of verse are connected here so as to form a couplet (printed by us as one long line). Even where this is not so, however, the two chief accents in each short line serve to make all the different forms and types of verse occurring in this poem sound homogeneous. This admits of a ready explanation, as the poem, in which no stanzaic arrangement can be detected, although styled a ‘song’ (line 2), was certainly never meant to be sung to a regular tune. On the contrary, it was undoubtedly recited like the ‘Song’ of Beowulf—probably not without a proper musical accompaniment—by the minstrels.

At all events the treatment of the words with regard to their rhythmic use in this poem does not deviate from that of Layamon.

§ 51. The two poems are of the same period, and in both the etymological and syntactical accentuation of natural speech forms the basis of the rhythmic accentuation. Monosyllabic words and the accented syllables of polysyllabic words having a strong syntactical accent are placed in the arsis; unaccented inflectional syllables as a rule form the theses of a verse; second parts of compounds and fully sounding derivative syllables are commonly used for theses with a somewhat stronger accent, and may, if placed in the arsis, even bear the alliteration, or, if they are less strongly accented, the rhyme:

CHAPTER IV
THE ALLITERATIVE LINE IN ITS CONSERVATIVE
FORM DURING THE FOURTEENTH AND FIFTEENTH CENTURIES

A. The alliterative verse without rhyme.

§ 52. The progressive or free form of the alliterative line came to an end as early as the middle of the thirteenth century, when it broke up into short rhyming couplets. The stricter form was for nearly three centuries longer a very popular metre in English poetry, especially in the North-Western and Northern districts of England and in the adjacent lowlands of Scotland. The first traces, however, of its existence after the Norman Conquest are to be found in the South of England, where some poetical homilies and lives of saints were written at the end of the twelfth and in the beginning of the thirteenth century which are of the same character, both as to their subjects and to their metre, as the poetical paraphrases and homilies written by Ælfric. These poems are Hali Meidenhad (a poetical homily), the legends of St. Marharete, St. Juliana, and St. Katherine. These poems have been edited for the Early English Text Society, Nos. 18, 13, 51, 80; the first three by Cockayne as prose-texts, the last by Dr. Einenkel, who printed it in short couplets regarded by him as having the same four-beat rhythm (Otfrid’s metre) which he and his teacher, Prof. Trautmann, suppose to exist in Layamon and King Horn.[107] The Homilies have no rhymes.

The form of these homilies and legends occasionally exhibits real alliterative lines, but for the most part is nothing but rhythmical prose, altogether too irregular to call for an investigation here. Some remarks on passages written in a form more or less resembling alliterative verse may be found in our Englische Metrik, vol. i, § 94.

It is quite out of the question to suppose these Southern works, with their very irregular use of alliteration and metre, to have had any influence on the metrical form of the very numerous alliterative poems written in the fourteenth and fifteenth centuries in the Midland and Northern districts of England. It is, however, not at all likely that alliterative poetry should have sprung up there without any medium of tradition, and that it should have returned to the strict forms of the Old English models. Nor can we assume that it was handed down by means of oral tradition only on the part of the minstrels from Old English times down to the fourteenth century. The channel of tradition of the genuine alliterative line must be sought for in documents which for the most part have been lost.

A few small remnants, however, have been preserved, viz. a charm in a MS. of the twelfth century (cf. Zupitza, Zeitschrift für deutsches Altertum, xxxi. 49), a short poem, entitled ‘Ten Abuses’, belonging to the same period (E. E. T. S. 49, p. 184), a prophecy of five lines contained in the chronicle of Benedict of Peterborough (Rerum Britannicarum Scriptores, 49, ii. 139), finally a prophecy ascribed to Thomas of Erceldoune (E.E. T. S., vol. 61, xviii, Thom. of Erc., ed. by A. Brandl, p. 26). But these pieces, treated by Prof. Luick in Paul’s Grundriss, ed. 2, II. ii, p. 160, are either too short or are too uncertain in text to admit of our making definite conclusions from them.

But from the middle of the fourteenth century onward we have a large number of poems composed in regular alliterative verse, e.g. King Alisaunder (Als.) and William of Palerne (W.), both in E. E. T. S., Extra-Ser. No. 1; Joseph of Arimathie (J.A.), E. E.T. S. 44; Sir Gawain and the Green Knight (Gr.), E.E. T. S. 4; Piers Plowman (P. P.), by W. Langland, E. E. T. S., Nos. 17, 28, 30, 38, 54; Pierce the Plowman’s Crede (P. P. Cr.), E. E. T. S. 30; Richard the Redeles (R. R.), E. E. T. S. 54; The Crowned King (Cr. K.), ibid.; The Destruction of Troy, E. E.T. S. 39, 56; Morte Arthure, E. E. T. S. 8; Cleanness and Patience, E. E. T. S. 1; The Chevalere Assigne, E. E.T. S., Extra-Ser. 6; and others of the end of the fifteenth and the beginning of the sixteenth centuries: see Prof. W. W. Skeat’s list in ‘Bishop Percy’s Folio MS.’, London, 1867 (ed. Furnivall and Hales), vol. iii, p. xi, and many recent publications of the Early English Text Society.

On the structure of this metre the opinions of scholars differ a good deal less than on that of the progressive or free form of the alliterative line. Yet there are a few adherents of the four-beat theory who apply it to the alliterative line of this epoch, amongst others Rosenthal (‘Die alliterierende englische Langzeile im 14. Jahrhundert,’ Anglia, i. 414 ff.). The two-beat theory, on the other hand, has been upheld also for this form of the alliterative line by Prof. W. W. Skeat, Essay on Alliterative Poetry, Percy Folio MS. 1867 (ed. Furnivall and Hales), by the present writer in Englische Metrik, i, pp. 195–212, and by Prof. Luick, Anglia, xi, pp. 392–443 and 553–618, and subsequently in Paul’s Grundriss, ed. 2, II. ii, pp. 161–3.

§ 53. The use and treatment of the words in the verse is on the whole the same as in the Old English period. The chief divergence is, that in this period of the language the difference between long and short syllables was lost, in consequence of the lengthening of short vowels in open syllables which had taken place in the interval, and that consequently the substitution of a short accented syllable and an unaccented one for a long accented syllable (the so-called resolution) was no longer admissible. Otherwise syllables with a primary accent, syllables with a secondary accent, and unaccented syllables are treated just as in the Old English poetry. Accented syllables are as a rule placed in the arsis, as are also second parts of compounds. Other syllables with secondary accent (derivative and inflectional syllables) are only in exceptional cases placed in the arsis of a verse.

It is of special interest, however, to notice that words of Romanic origin which in the course of time had been introduced into the language are in many cases accented according to Germanic usage. Words of which the last syllable was accented in French have in their Middle-English form the chief accent thrown on a preceding, frequently on the first, syllable, and in consequence of this the originally fully accented syllable in trisyllabic words receives the secondary accent and is treated in the rhythm of the verse in the same way as syllables with a secondary accent in English words. The laws, too, which in Old English affect the subordination and position of the parts of speech in their relationship to the rhythm of the verse and to the alliteration, remain, generally speaking, in force. It is remarkable that ‘if an attributive adjective is joined to a substantive, and a verb to a prepositional adverb, the first part of these groups of words still has the chief accent’ (Luick). The relationship, on the other hand, of verse and sentence is changed. While in Old English poetry run-on-lines were very popular and new sentences therefore frequently began in the middle of a line, after the caesura, we find that in Middle English, as a rule, the end of the sentence coincides with the end of the line. Hence every line forms a unity by itself, and the chief pause falls at the end, not, as was frequently the case in Old English times, after the caesura.

§ 54. Alliteration. On the whole, the same laws regarding the position of the alliterative sounds are still in force as before; it is indeed remarkable that they are sometimes even more strictly observed. In the Destruction of Troy, e.g. triple alliteration according to the formula a a a x is employed throughout.

In such cases it may sometimes be noticed that a line which has no internal alliteration is linked by alliteration with a preceding or with a following line, in the same way as was to be observed already in the last century of the Old English period (cf. p. 50):

Again an excess of alliteration is found, which happens in different ways, either by admitting four alliterative sounds in one line (a a a a) as was sometimes done even in Old English:

or by retaining the same alliterative sound in several consecutive lines, e.g. :

or, finally, by allowing the somewhat more strongly accented syllables of the theses to participate in the alliteration:

By the increasing use of this kind of alliteration it ultimately degenerated so much that the real nature of it was completely forgotten. This is evident from the general advice which King James VI gives in his Revlis and Cavtelis to be observit and eschewit in Scottis Poesie (Arber’s Reprint, p. 63):

He then gives a description of this kind of verse which makes it evident that he looked upon ‘tumbling verse’ as a rhythm of two beats in each hemistich or four beats in the full line, for he says:

King James VI was a contemporary of the last poets who wrote in alliterative lines in the North and therefore undoubtedly had heard such poems read by reciters who had kept up the true tradition of their scansion. We have here then the very best proof we can desire not only of the four-beat rhythm of the line, but also of the fact that unaccented words, although they may alliterate intentionally, as they do often in poems of the fifteenth century, or unintentionally, as earlier, do not get a full accent in consequence of the alliteration, as some scholars have thought, but remain unaccented.[109] As to the quality of the alliteration the same laws on the whole still prevail as in Old English poetry, but are less strictly observed. Thus frequently voiced and unvoiced sounds alliterate together, and the aspiration is neglected; f alliterates with v, v with w, w with wh, s with sh or with combinations of s and other consonants, g with k, h with ch:

On the other hand, sometimes (as e.g. in the Alisaunder fragments) greater strictness may be noticed in regard to alliteration of vowels, as only the same vowels[110] are allowed to alliterate:

Later on, in the fifteenth century, vocalic alliteration in general falls into disuse more and more.

§ 55. Comparison of Middle English and Old English alliterative verse. With regard to the rhythmic structure of the verse the Middle English alliterative line is not very different from the corresponding Old English metre. Two beats in each hemistich are, of course, the rule, and it has been shown by Dr. K. Luick, in a very valuable paper on the English alliterative line in the fourteenth, fifteenth, and sixteenth centuries,[111] that all the different types which Prof. Sievers has discovered for the two sections of the Old English alliterative line occur here again, but with certain modifications.

The modifications which the five chief types have undergone originated in the tendency to simplify their many varieties exactly in the same way as the Old English inflexional forms of the language were simplified and generalized in the Middle English period.

Only three of the five old types, viz. those with an even number of members (A, B, C), are preserved in the second section of the verse, and those not in their original forms. They show further a certain tendency to assimilate to each other.

In types B and C the variations with disyllabic anacrusis occurred most frequently, as was also the case in type A, and verses of this kind now become predominant. Furthermore, in the Old English alliterative line, endings consisting of an accented and an unaccented syllable (feminine endings) prevailed; and type B was the only one of the symmetrical types ending with an accented syllable. In Middle English the use of feminine endings goes so far that the original type B has disappeared altogether and given place to a new type with an unaccented last syllable corresponding to the form ××–́×–́×.

Prof. Luick very properly calls this type BC, holding that it originated from the variations ××–́×⏑́͜× and ××⏑́͜×–́× of the old types B and C in consequence of the lengthening of the originally short accented syllable. Verse-ends with two unaccented syllables, which might have arisen in the same way from –́× = ⏑́͜××, did not become popular; and verse-ends with one unaccented syllable predominated. Lastly, an important feature of the later verse-technique deserves notice, that a monosyllabic anacrusis (an initial unaccented syllable) is generally allowed in types where it was not permitted in the Old English alliterative line. The consequence of these changes is that the rhythm of the verse which was in Old English a descending rhythm, becomes in Middle English ascending, and is brought into line with the rhythm of the contemporary even-beat metres.

This is the state of development presented by the Middle English alliterative line in one of the earliest poems of this group, viz. in the fragments of King Alisaunder, the versification of which, as a rule, is very correct.

Here the three types only which we have mentioned occur in the second hemistich.

Type A is most common, corresponding to the formula (×)–́××–́×:

lórdes and óoþer 1, déedes of ármes 5, kíd in his tíme 11, térme of his lífe 16,

or with anacrusis:

or stérne was hólden 10, and sóne beráfter 25.

More than two unaccented syllables may occur after the first accented syllable. These two peculiarities seldom occur together in one and the same second hemistich (though frequently in the first hemistich); but there are some examples:

is túrned too him álse 165, and príkeden abóute 382, hee fáred òn in háste 79;

in this last example with a secondary accent on the word òn as also in the verse: þe méssengères þei cámme 1126.

Type C, (×)××–́–́×:

was þe mán hóten 13, þat his kíth ásketh 65, as a kíng shólde 17, withoute míscháunce 1179.

Type BC, (×)××–́×–́×:

or it týme wére 30, in his fáders life 46, of þis méry tále 45, þat þei no cómme þáre 507.

The same types occur in the first hemistich; but type C disappears almost entirely, and in the other two the last syllable not unfrequently is accented, especially if a considerable number of unaccented syllables occur in the middle of the hemistich; such verses may be looked upon as remnants of types B and E:

þo was cróuned kíng 28, hee made a uéry uów 281, and wédded þat wíght 225, þe bérn couth þerbý 632, &c.

Type D also seems to occur sometimes:

móuth méete þertò 184, what déath drý[e] thou shàlt 1067.

Besides these types the first hemistich has, as in Old English times, some forms of its own. The succession of syllables –́××–́× (type A) is extended either by several unaccented syllables before the first accented one (polysyllabic anacrusis) or by the insertion of a secondary accent between the two main accented syllables, or after the second accented syllable, with a considerable number of medial unaccented syllables.