IV. The Ballad Style

CHAPTER IV

THE BALLAD STYLE

The style of the English and Scottish ballads has often received treatment, and their appeal for the reader who is in reaction from book verse has been stressed by critics of many types. Certain conspicuous mannerisms have had attention from scholars and special students and have been utilized for special pleading. They are thought to afford ballad differentiae and to throw light upon the origin of the ballad as a lyric type. That traditional ballads constitute a distinctive species is held to be due, on the evidence of stylistic mannerisms, not to their oral or sung character or to their destination as popular poetry, but rather to their origin among the folk, especially among illiterate folk.

In the following pages no attempt will be made to repeat what has been well said by others in characterization of the ballad style. Various features of it will be examined which have been brought into the foreground of discussion because' they seemed pivotal. Usually the style of the ballads is analyzed without much reference to the pieces which exist alongside them in folk tradition. This is partly because of the tendency of many collectors to restrict their salvage to pieces of the Child type, ignoring or discarding many related types of song of equal or greater currency among the folk. In consequence of such specialization, the ballads are often endowed too distinc
tively with traits which they share with other folk-song. A study of ballads, whether mediaeval or later, which does not take into account their background, tends to foster too sharply drawn distinctions or too rigid generalizations, and to make the results arrived at less dependable.

I INCREMENTAL REPETITION AND OTHER BALLAD

MANNERISMS

"Iteration," we are told, "is the chief mark of the ballad style; and the favorite form of this effective figure is what one may call incremental repetition. The question is repeated with the answer; each increment in a series of related facts has a stanza for itself, identical save for the new fact, with the other stanzas. Babylon furnishes good instances of this progressive iteration." 1 And again, "Incremental repetition is the main mark of the old ballad structure." 2 This repetition is supposed to be bound up with derivation from the dance, as many citations will show. "It furnishes," we are told, "the connection with that source of balladry — not of mended ballads — in improvisation and communal composition, with the singing and dancing throng, so often described by mediaeval writers." References are many to "incremental repetition, obviously related to movements of the dance "; 3 or we are assured that the ballad was "meant in the first instance for singing and connected as its name implies, with the communal dance." By incremental rep

i F. J. Gummere, "Ballads" in A Library of the World's Best Literature, vol. in, p. 1308. See also The Popular Ballad, pp. 117— 134.

2" Ballads" in The Cambridge History of English Literature, vol. II, pp. 449, 459.

3 Democracy and Poetry, p. 188.

etition is meanythe ballad repetition not in the refrain way but structurally or for emphasis by which successive stanzas reveal a situation or advance the interest by successive changes of a single phrase or line. A stanza repeats a preceding one with variation but adds something to advance the story, f Lyrical repetition of this type is a marked characteristic of the Child pieces, a large proportion having this structural feature. It is upon this characteristic that many scholars rest their belief that the very structure of the ballad, the type itself, rests chiefly on the dance, the communal dance of primitive or of peasant throngs. The four examples best illustrating it, those usually cited, are Lord Randal, first heard of in the repertory of a seventeenth-century Italian singer at Verona named Camillo; * Edward, a ballad in literary Scotch, first known from the Percy manuscript; The Maid Freed from the Gallows, a ballad told with perfect symmetry by the question and answer method in a version recovered in America in the nineteenth century; and Babylon, the earliest text of which comes from Motherwell's Minstrelsy, published in 1827.

As often pointed out, the date of recovery of a ballad is no sure indication of the antiquity of a ballad, or the lack of it; but it should not be left out of account when other evidence fails. The chronology of the English and Scottish ballads lends no support to the belief that incremental repetition was a characteristic of archetypal ballads or that it points to their emergence from the dance. If the

* Countess Martinengo-Cesaresco, Essays on the Study of FolkSong (1886). "Lord Ronald in Italy," p. 214. The Delia Cruscans thought of "improving" this song. The poisoning feature of the plot is more characteristic of Italian than of English story.

repetitional type is that having greatest antiquity, repetition should appear characteristically in the earliest ballads, less often in the late — those which were composed after the dance origin so often assumed ceased to condition the structure. Yet incremental repetition does not appear in our oldest ballad text, the thirteenth-century Judas, nor is it a normal feature of those early ballad types, the outlaw and chronicle ballads.5 Unfortunately for the theorists who hold it to be fundamental, it appears most frequently in later texts, not earlier, and more often in the broadsides than in oral versions. It does not appear in the fifteenth-century Inter Diabolus et Virgo, the direct ancestor of Riddles Wisely Expounded, in the texts of which it does appear; so that Professor F. E. Bryant remarked, much puzzled, " it is a clear case of an early version not being nearly so ballad-like as a whole group of later ones." 6 Says Mr. John Robert Moore, "Unfortunately . . . the facts seem to make little provision for the theory [i. e. of incremental repetition as fundamental to the ballad structure] ; for it is the simple ballads which most often have fixed refrains, and the broadsides which exhibit the most marked use of incremental repetition. Furthermore, when oral tradition adds a refrain to an original printed broadside, it is only a simple refrain without the structural device of accretion which Professor Gummere considers so characteristic."7 Professor H. M. Belden has pointed out that the output of the nineteenth century ballad press, accessible in the British

5 There is something like it in Robin and Gandeleyn and in the learned or at least sophisticated St. Stephen and Herod. « A History of English Balladry, 1913.

7 The Influence of Transmission on the English Ballads, Modern Language Review, vol. XI (1916), p. 398.

Museum, shows this structural characteristic very markedly.8 It has been shown by Mr. Phillips Barry that iteration is a mark of the late not the earlier versions of YoUng Charlotte, whose history he has traced backward nearly a hundred years.9 Iteration can be developed, he shows, as an effect of continuous folk-singing.

Structural repetition is not a certain test of what is and what is not a ballad and it is not to be insisted upon in definition of the type, first because it is not always present in ballads, and second, because it is as characteristic of other folk lyrics as it is of ballads. Just as a ballad can be a ballad without the presence of choral repetition or a refrain, so it can be a ballad without showing incremental repetition. The. only dependable test elements in ballads are lyrical quality and a story element, and, for traditional folk-ballads, anonymity of authorship.

An excellent example of structural repetition in mediaeval song other than ballads is afforded by the following satire against women:—10

Herfor & therfor & therfor I came, And for to praysse this praty woman. Ther wer m wylly, 3 wyly ther wer,— A fox, a fryyr, and a woman. Ther wer 3 angry, 3 angry ther wer,— A wasp, a wesyll, & a woman. 8 Review of Gummere's The Popular Baltad, Journal of English and Germanic Philology, vol. vm, p. 114.

» William Garter, The Bensontown Homer, Journal of American Folk-Lore, vol. xxv, pp. 156-158.

io Bodleian MS. Eng. Poet. E. 1. f. 13. Percy Society, vol. Lxxiii, p. 4.

Ther wer 3 cheteryng, in cheteryng ther wer,— A peye, a jaye, & a woman. Ther wer 3 wold be betyn, 3 wold be betyn ther wer,— A myll, a stoke fysehe, and a woman. Or by a song on a fox and geese, which opens—11

The fals fox camme unto oure croft, And so our gese full fast he sought;     With how, fox, how, etc. The fals fox camme unto oure stye, And toke our gese there by and by. The fals fox camme into oure yerde, And there he made the gese aferd. The fals fox camme unto oure gate, And toke oure gese there where they sate. The fals fox camme to oure halle dore, And shrove oure gese there in the flore, etc. Or by this lively pastoral, which might possibly be a dance song, but which is not a ballad.12

I haue xn oxen that be fayre & brows,

& they go a grasynge down by the town;

"With hay, with howe, with hay!

Sawyste thow not myn oxen, you litell prety boyf

I haue xn oxen & they be ffayre and whight, & they go a grasyng down by the dike; 11 Cambridge University Library, M. 8. Ee, 1, 12. There are 18 stanzas.

12 MS. Balliol 354. FlUgel, Anglia, vol. xxvi, p. 197. Ed. Dyboski, E. E. T. S., Extra Series, 101 (1907), p. 104.
With hay, with howe, with hay!

Sawyste not you myn oxen, you lytyll prety boy?

I haue Xii oxen & they be fayre and blak,

And they go a grasyng down by the lak;

With hay, with howe, with hay!

Sawyste not you myn oxen, you lytell prety boy?

I haue xn oxen, and they be fayre and rede,

& they go a grasyng down by the mede;

With hay, with howe, with hay!

Sawyste not you myn oxen, you lytyll prety boyf

Or by the Song of the Incarnation of about 1400, which is quoted in full elsewhere.13

Repetition and parallelism are also characteristic of that popular type of mediaeval song the religious carol, like the well-known Cherry Tree Carol, classed by Professor Child, because of its narrative element and its currency in oral tradition, as a ballad. Other carols in ballad stave and showing very close relation to ballads — they have both structural repetition and a narrative element — are The Holy Well, telling of the childhood of Jesus, and The Bitter Withy, closely related to the preceding, perhaps as much ballad as carol. Some less ballad-like carols showing structural repetition are The Five Joys of Christmas, Bring Us Good Ale, Born is the Babe, Out of the Blossom Sprang a Thorn, This Rose is Bailed on a Byse, etc.14 The following carol is not in ballad stave but shows a type of structural repetition, or parallelism:—15

13 From the Sloane MS. 2593. See p. 175.

"Edith. Rickert, Old English Carols (1910); Jessie L. Weston, Old English Carols (1911). See especially Balliol MS. 354 and Sloane MS. 2593.

"Hill MS., ed. Dyboski, E. E. T. S. 101 (1907), p. 7.

Make we mery in hall & bowr, Thys tyme was born owr Savyowr. In this tyme God hath sent Hys own Son, to be present, To dwell with vs in verament,     God that ys owr Savyowr. In this tyme that ys be-fall, A child was born in an ox stall & after he dyed for vs all, God that ys owr Savyowr.

In this tyme an angell bryght Mete in sheperdis vpon a nyght He bade them do a-non ryght     To God that ys owr Saviowr. In thys tyme now pray we

To hym that dyed for vs on tre,

On vs all to haue pytee,

God that ys owr Saviowr.

The carol of the six rose branches, "All of a rose, a lovely rose, All of a rose I sing a song," applies the secular liking for the rose, as a poetic flower, in a poem of religious symbolism, in sequence form:—16

The fyrst branch was of gret myght, That spronge on Crystmas nyght, The streme shon over Bedlem bryght, That men myght se both brod and longe.

The nde branch was of gret honowr, That was sent from hevyn towr, Blessyd be that fayer flowr!   Breke it shall the fendis bondis. i8 From the same manuscript. There are many dramatic carols or carols in the question and response form between Mary and an angel or between Mary and her son, in the Hill manuscript.

The thyrd branch wyde spred Ther Mary lay in her bede, The bryght strem in kyngis lede To Bedlem, ther that branch thei fond.

The nnth branch sprong in to hell, The fendis bost for to fell, Ther myght no sowle ther in dwell,   Blessid be that tyme that branch gan spryng. The vth branch was fayer in fote, That sprong to hevyn tope & rote, Ther to dwell & be owr bote   & yet ys sene in priestis hondis. The vith branch by & by, Yt ys the v joyes of myld Mary, Now Cryst saue all this cumpany,  & send vs gud lyff & long. Incremental repetition and parallelism of line structure are especially characteristic of popular religious poetry, in particular of revival hymns. It is well-known that "repetition to the point of wearisomeness is a favorite form of revival hymns." 17 To cite illustration, the following song called Weeping Mary, recovered in the twentieth century among the negroes, affords an example of parallelism:

If there's any body here like Weeping Mary,

Call upon Jesus and he'll draw nigh,

He'll draw nigh.

0 glory,'glory, glory, hallelujah,

Glory be to God who rules on high.

It E. B. Miles, Some Real American Music, Harper's Magazine, vol. 109, pp. 121-122.

If there's anybody here like praying Samuel, Call upon Jesus, etc. If there's anybody here like doubting Thomas, Call upon Jesus, etc. This song is thought by Mr. H. E. Krehbiel to be an original Afro-American song, and he printed it as such.18 He limits his claim for the originality of negro songs to their religious songs, their "shouts" and "spirituals." But Weeping Mary can be traced to the singing of a white woman who had learned it at a Methodist protracted meeting somewhere between 1826 and 1830, long antedating its appearance among the negroes.19 There were many stanzas of repetitional pattern and the whole might be continued indefinitely.20 A similar history may be noted for a song included among T. P. Fenner's collection of Religious Folk Songs of the Negro as sung on plantations: 21

Wonder where is good old Daniel,  Way over in the Promise Lan', etc. Wonder where's dem Hebrew children, etc.

Wonder where is doubtin' Thomas, etc.

Wonder where is sinkin' Peter, etc.

Compare with this the old revival hymn— 22

is Afro-American, Folk-Song, 1914.

is See Modern Language Notes, vol. 33, p. 442, 1918.

20 See note 57, p. 158.

21 New Ed. 1909, p. 107.

22 See "Old Revival Hymns" in The Story of Hymns and Tunes by Theron Brown and Hezekiah Butterworth, 1896.

Where now where are the Hebrew children f They went up from the fiery furnace, etc. Where now is the good Elijah f etc.

Where now is the good old Daniel t etc.

The climax was reached with

By and by we'll go to meet him, By and by we'll go to meet him, By and by we'll go to meet him, Safely in the Promised Land. It might be mentioned also that the "chariot" frequent in negro spiritual? played a role in older revival poetry,23 But whatever their origin, negro revival hymns and plantation songs, like the folk-songs of white people, abound in instances of structural repetition and in sequences of various types. Three examples may be given: 24

Save Me, Lord, Save Me

  I called to my father;    My father hearkened to me. And the last word I heard him say     Was, Save me, Lord, save me. I called to my mother, etc.

I called to my sister, etc.

I called to my brother, etc.

23 Compare H. H. Milman's popular hymn, The Chariot of Christ, or The Last Day.

2* Marshall W. Taylor, A Collection of Revival Hymns and Plantation Melodies, 1883.
I called to my preacher, etc. I called to my leader, etc. I called to my children, etc. He Set My Soul Free

Go and call the bishops in, Go and call the bishops in, Go and call the bishops in,   And ask them what the Lord has done. Go and call the elders in, etc.

Go and call the deacons in, etc.

Go and call the leaders in, etc.

Go and call the Christians in, etc.

Resurrection of Christ

Go and tell my disciples, Go and tell my disciples, Go and tell my disciples,   Jesus is risen from the dead. Go and tell poor Mary and Martha, etc. Go and tell poor sinking Peter, etc. Go and tell the Roman Pilate, etc. Go and tell the weeping mourners, etc. Natalie Curtis Burlin's texts, though somewhat shrunken from those of the same songs in earlier collections, show the same liking for sequences:— 25

25 Negro Folk-Songs, recorded by Natalie Curtis Burlin. 1918.

O ride on Jesus,   Bide on, Jesus,   Bide on, conquerin' King. I want t'go t'Hebb'n in de mo'nin'.

Ef you see my Father,

0 yes, Jes' tell him fo' me,

0 yes,

For t' meet me t'morrow in Galilee:    Want t'go t'Hebb'n in de mo'nin'. Following verses may substitute the words "sister" and "brother" for " mother" and " father." A second song, of similar pattern, is this:—

    Good news, Chariot's comin',     Good news, Chariot's comin',     Good news, Chariot's comin', An' I don't want her leave-a me behin'. Dai^s a long white robe in de Hebb'n I know, etc.

Later verses open —

Oar's a starry crown in de Hebb'n, I know, etc.

Oar's a golden harp in de Hebb'n, I know, etc.

Par's silver slippers in de Hebb'n, I know, etc.

Repetition in iterative or sequence form is also characteristic of contemporary 'student songs, as Forty-Nine Bottles A-Hanging on the Wall, or the Song of a Tree {The Green Grass Grows All Round), or the old-time temperance songs, like The Tee-Totallers Are Coming, or The Cold-Water Pledge. And it may be a characteristic of popular laments, as The Lyke-Wake Dirge. It is found in nursery songs like One, two, buckle my shoe, or One little, two little, three little Injuns," etc., and in lullabies, like many which have been preserved from the fifteenth century. Most fundamentally, it is characteristic of orally preserved game and dance songs, which have been illustrated in another chapter; but here it is of the interweaving type, is stable and part of the fabric of the song, not iteration of the type characteristic of the ballads.

Incremental repetition appears very strikingly in American folk-songs, all of British importation, in dialogue form, which are never classified as ballads. An instance ia the familiar: —

0 where have you been, Billy Boy, Billy Boy?

0 where have you been, charming Billy?"

"I have been for a wife, she's the treasure of my life,

She's a young thing but can't leave her mother."

He is asked whether his wife can make a cherry pie, a feathei bed, a loaf of bread, a " muly cow," etc., and gives humorous responses. In The Quaker's Courtship the wooer says in repetitional stanzas that he has a ring worth a shilling, a kitchen full of servants, a stable full of horses, etc., and asks if he must join the Presbyterians; but he meets rebuff. In Soldier, Soldier, Wont You Marry Me?, the soldier answers in lyrical sequences that he has no shoes to put on, then that he has no coat, then that he has no hat. When the girl has brought these, the song ends with the question —

"How could I marry such a pretty little girl  When I have one wife to home?" In the familiar The Milkmaid, which sounds like a survival of a pastourelle, a maiden is asked in stanzas of the iterative type " O where are you going, my pretty maid?" whether she may be accompanied, what her father is, and what her fortune? She answers that she is going milking, that her father is a farmer, that her face is her fortune, etc., and when her questioner follows this last statement by the remark " Then I won't have you, my pretty maid," she responds with "Nobody asked you," etc. To cite a last example of these dialogue folk-songs showing incremental repetition, in A Paper of Pins, the wooer offers the girl a paper of pins, a little lap dog, a coach and four, a coach and six, the key of his heart, and finally a chest of gold, if she will marry him. All the offers are refused until the last. When this is accepted, he closes the sequence with —i

"Ha, ha, money is all, woman's love is nothing at all.   I'll not marry, I'll not marry, I'll not marry you." Farther, structural repetition is not a mannerism appearing in primitive poetry. There is limitless and wearisome iteration and choral response, but no telling of stories by the question and answer method of Lord Randal, Edward, The Maid Freed from the Gallows, and Babylon. As for the cowboy pieces,26 in those which their collector indicates as of communal composition, such narrative as they have is not presented by incremental repetition or by

20 An example is afforded by The Song of the "Metis" Trapper by Rolette, Lomax, Cowboy Songs, p. 320, the stanzas of which open in sequence, "Hurrah for the great white way," "Hurrah for the snow and the ice," "Hurrah for the fire and the cold," "Hurrah for the black-haired girls," but the cowboy songs as a whole do not exhibit structural repetition.

the question and answer method but in a far less skilful or lyrical way.

The truth is that repetition, structural or stanzaic, verbal, of the refrain type, or consisting of interweaving lines, may be found in all types of popular poetry, from nursery songs to revival hymns. Old French literature is that richest in mediaeval lyric poetry and in dance songs, but old French lyrics and dance songs bear no resemblance to ballads and they are plainly aristocratic. Structural iteration belongs to popular song in general, indeed it is very likely to be developed through folk-preservation when it did not belong to a song in its original form. It is not certain proof of dance origin even among primitive peoples. It characterizes not only dance lyrics but revival hymns, game and labor songs, student songs, lullabies and nursery songs, Christmas carols, laments, and songs and folk-lyrics in general. It is not a test of the ballad style, is not a ballad differentia, since it belongs to other styles also. And it is not a test of age for it is not present in some of the oldest ballads and is developed in late variants of newer ballads. Moreover it is a mannerism easily caught and of great assistance in promoting folk-participation in singing. The ballad is the only type of folk song showing structural repetition or parallelism of line in the presentation of narrative, but that is because it is the only type of folk-lyric which presents narrative. Structural repetition in ballads should not be cited as proof that the latter were composed in some manner different from other lyric verse, for it is a feature which ballads share with folk-song of many types; nor is its appearance in individual ballads proof of the antiquity in
type of such ballads.
Besides incremental repetition, other ballad mannerisms which have received emphasis are the so-called "climax of relatives" and the ballad motive of the legacy, or the giving of testamentary instructions. Both are well illustrated by The Hangman's Tree, a text of which is quoted in full in another chapter.27 In neither mannerism may certainly be seen proof of antiquity or of the communal origin of a ballad. Both appear in the later rather than in the earlier ballad texts; and the climax of relatives — better called a sequence of relatives, or better still a sequence merely, for the sequence may be of persons other than relatives, or of things28— is, as we have just seen, a characteristic of revival poetry and of general folk poetry as well. Neither mannerism appears in Judas, our oldest ballad text, nor in the ballads which go back with certainty to the fifteenth century, nor in texts from the early sixteenth century. They might for balladry, if chronology of appearance count, be termed a sign of comparative lateness. And they need not be unfailing signs of communal origin. Will and testament features played an important part in mediaeval literature,29 and by the early modern period their legacy might well appear in traditional verse. Like the sequence of relatives, the giving of testamentary instructions is a mannerism easily caught and memorable, and it is in no way remarkable that it should be found in ballads, alongside the "last goodnights," riddling, and other devices of lit

27 See p. 113.

28 See the sequence of kirks in The Gay Goshawk, or of harpstrings or of tunes in The Two Sisters.

29 E. C. Perrow, Will and Testament Literature. Publications of the Wisconsin Academy of Sciences, vol. Xvii.

erature of the past. The probability is that the legacy feature of the percentage of the English and Scottish ballads which show it is a literary heritage. The rapidity with which an easily caught mannerism may spread may be illustrated by the " come all ye" opening of the broadsides, or by the assimilation of the briar-rose motive at the end of texts of Fair Margaret and Sweet William, Lord Lovel, Earl Brand, and Barbara Allen,30 or of the stanzas beginning —

"0 who will shoe your feet, my love,  And who will glove your hand?" of The Lass of Roch Royal. In the texts of Cecil J. Sharp and Mrs. Campbell, these stanzas have spread to The Rejected Lover, and The True Lover's Farewell, and even to John Hardy, which seems to have been originally a negro song.31

A certain type of sequence of relatives is rather stock in popular song of the eighteenth and nineteenth centuries, especially in songs of the death bed or death-bed confession type. In The Cowboy's Lament (The Dying Cowboy), which derives from an eighteenth century Irish popular song,32 the speaker asks to have messages sent to his mother, his sister, his sweetheart. In Caroline E. E. Norton's Bingen on the Rhine, the sequence runs " Tell my brothers and companions," "Tell my mother," "Tell my sister," "There's another — not a sister." The Dy

so As in C. J. Sharp's text, Folk-Song from the Southern Appalachians, p. 96.

»i Folk-Song from the Southern Appalachians. See nos. 56, 61, 87.

82 Lomax, Cowboy Songs, p. 74. For its origin, see Mr. Phillips Barry's article, cited p. 207.

ing Califomian, widely known over the United States in folk song, runs in the longest of its Nebraska variants, "Tell my father when you meet him," " Tell my mother," "Tell my sister," " 'Tis my wife I speak of now," etc. In 0 Bury Me not on the Lone Prairie, as in the sea piece which was its model, is the same sequence, mother, sister, sweetheart, to whom messages are to be delivered. So in Buena Vista Battlefield,33 messages are to be sent from the dying soldier to father, mother, sweetheart; and in The Last Longhorn, a cowboy piece patterned on this type of poem —

An ancient long-horned bovine lay dying by the river; There was lack of vegetation and the cold winds made him shiver —

are found " Tell the Durhama and the Herefords," " Tell the coyotes," etc. Still another example is afforded by A Poor Lonesome Cowboy, "I ain't got no father," "I ain't got no mother," "I ain't got no sister," "I ain't got no brother," "I ain't got no sweetheart"—

I'm a poor lonesome cowboy And a long ways from home. All this illustrates how easily a familiar pattern, known through some well-known song or songs, is assimilated. None of the American pieces cited, unless the last, may fairly be said to have had communal origin.

Various other marks of style for the English and Scottish popular ballads, besides incremental repetition, the giving of testamentary instructions, and the sequence of relatives — for example, presenting narrative by question

88 Cowboy Songs, pp. 3, 34, 197.

and answer, the ballad vocabulary, the use of set epithets, alliterative formulae, and the like, have been treated in other chapters.

II DIALOGUE AND SITUATION BALLADS AND THEORIES

OF DEVELOPMENT

That situation ballads in dialogue form represent a primal type of ballad and that there is development from these to length and complexity is a view which is often brought forward. According to Professor Walter Morris Hart, "the ballad, in its simplest and most typical forms, might be called a short story in embryo. It is a song about a single situation" . . . "there is development from the simple to the complex, from the homogeneous to the heterogeneous . . . the simplest and most homogeneous ballads or groups of ballads are actually older or representative of something older, than the most complex and heterogeneous. We have already traced this development from the relatively simple ballad of Edward to the relatively complex Gest of Robyn Hode."34 An adherent of the same school formulates this theory of development as follows: "Dialogue is the primitive fact; scenario, character, and other explanatory matters come later. The older and more primitive a ballad is, generally speaking, the greater the proportion of dialogue" . . . "We can now understand what Gummere calls communal composition and can see the significance ... of such things as refrain and dialogue. They are principles of composition. They make possible the production of a fairly well-or

34 English Popular Ballads (1916), pp. 45, 49. This is the thesis of Professor Hart's Ballad and Epic, Harvard Studies and Notes m Philology and Literature (1907), IX.


dered ballad by the common activity of the whole tribe." 38 This hypothesis of development is quite unproved, and tested by the processes of living folk-song and by the songs of savage tribes, it is improbable. And an interesting feature of the assumption underlying it is its inconsistency. The writers who hold it affirm that the ballad is the earliest universal form of poetry, yet by their own theories the early simple forms only later become ballads by developing complexity and plot. The ballads are the earliest form of song, yet they develop from earlier song.

The date of recovery of ballads is not a decisive factor in determining their antiquity, yet it is to be taken into account. Judging by the date of recovery, the situation ballads, Edward and Lord Randal — from the simplicity of whose structure Professor Hart develops the epic complexity of the Robin Hood ballads — are of later rather than earlier composition. They certainly come to us in late form, as pointed out elsewhere. Edward is told as completely and with as telling use of suspense and climax as a literary ballad like Rossetti's Sister Helen. It is a somewhat doubtful evolution which passes onward from the artistic quality of these pieces into the crudeness and length of the Robin Hood narratives. But aside from the late appearance of the best ballads illustrating the "earliest" stage, it should be pointed out that in general the presence of dialogue in poetry is a sign of compara

»»G. H. Stempel, A Book of Ballads (1917), pp. xvi, xxvii. Possibly this is the view also of Professor F. M. Padelford, who speaking of the debate of holly and ivy in mediaeval literature remarks that "like other songs of winter and summer, it harks back to that communal period when dialogue was just beginning to emerge from the tribal chorus." Cambridge History of English Literature, n, p. 431.
tive lateness in composition. It does not appear in the epic poetry of early peoples. The speeches of characters in Homer, Virgil, Beowulf are long declamations. So in the older dramas, the speeches are long declamations. The breaking up of the talk of characters, in narrative and dramatic literature, into give-and-take dialogue occurs, as it were, before our eyes. In the Old English period there is very little in the poetical literature that could be called dialogue. The nearest is to be found in the works of Cynewulf and his school. It is after the Norman Conquest that it begins to enter, in lyric and narrative minstrelsy, until dialogue in one form or another, it is agreed by scholars, becomes part of the minstrel's and the song composer's stock in trade.36 In Old French literature, so largely the source of or so largely influencing Middle English literature, dialogue or semi-dialogue appears in chansons a danser of literary type between soloists and a chorus, in chansons a personnages, or chansons de mal

so An excellent example of use of the question and answer method is afforded by the early fourteenth century song of a maiden whose food was "the primrose and the violet" and whose bower was "the red rose and the lily flower," preserved in the MS. Rawlinson D 914 f 1. It is too properly a song to be termed "literary" but it is obviously for sophisticated circles and of the "conscious art" type. Middle English religious lyrics afford many examples of dialogue songs.

There is an excellent example of a question and answer lyric, between a mother and daughter in the ballad manner in Old Portuguese troubadour poetry, by King Denis (1279-1325), Das Liederbuch des Konigs Denis von Portugal, ed. H. R. Lang (1894), pp. xcv, 75; Ferdinand Wolf, Studien zur Geschichte der Spanischen und Portuguiesischen NationalUteratur (1859), p. 708. Examples may be found also in Old Italian poetry. Dialogue between mother and daughter, like other dialogue forms, seems to have been a popular troubadour mode.

mariees, in pastovrelles — all these in the form in which we have them being of minstrel origin. It appears, as so well known, in the elaborate debates, disputes, and the like, which the Middle Ages so liked. From these lyrical debates it probably entered the ballades in dialogue form 37 in which it remained popular for three centuries. Lyrical dialogue and question and answer are both characteristic enough of late mediaeval " art" song for derivation, when they appear in the ballads, from tribal improvisation or from that of peasant communes, to be unnecessary and improbable.

It is known that mediaeval minstrels often recited or gave a type of impersonation dramatically. Monologue and dialogue were rendered dramatically, though by one person. There are clear traces of this in many religious narratives and songs. Thus may have been given the early religious ballads of Judas and St. Stephen and Herod. The presence of dialogue and dramatic situation in such abundance in the ballads might well have some relation to a dramatic manner of delivery. The more song-like lyrical ballads, those with refrains, are not those preserved to us in the oldest texts but come from the Tudor period and thereafter.

Professor Hart's theory of development from the short and simple to the long and complex sounds authentic but there are many considerations which do not reinforce it. It would be easy and plausible, if we discard chronology, to build up a theory of development from the short oneact plays of the twentieth century to the five-act dramas of the Elizabethans; or from the short story of the nineteenth century to the long novel of the eighteenth; or from

37 Helen Louise Cohen, The Ballade (1915), p. 56.

the periodical essays of the eighteenth to many of the longer prose types, the sermon, the oration, the treatise, the satire, which preceded it. If there is development in literature from simplicity to complexity there is also development from length and complexity to brevity and simplicity. Fair analogy may be drawn with the development in language, as illustrated by the complex inflectional structure of Sanskrit or Greek compared with the simplified analytical structure of present English. There is a linguistic tendency to shorten and simplify forms, to drop inflections, and to analyse "sentence-words" into short elements, co-existent with the tendency, earlier recognized, to lengthen monosyllables into polysyllables by composition. In the chronology of Indo-European languages, the languages of complex structure appear early and those simplest in structure come last. And this duality of development may be paralleled from literature.

The mass of the English ballads, or lyrical narratives, certainly appear in literary history later than do the epics and chansons de geste into which they are supposed to develop. Says E. K. Chambers, "The ballad, indeed, at least on one side of it was the detritus as the lai had been the germ of romance." 38 Professor Ker points out that "... it is certain that the ballads of Christendom in the Middle Ages are related in a strange way to the older epic poetry. . . . The ballad poets think in the same manner as the epic poets, and choose by preference the same kind of plot." 39 As the epic and romantic long narratives, to be recited or sung, become outworn, new lyrical narratives to be recited or sung appear. In any

ss The Mediceval Stage, I, p. 69.

3» English Literature: Mediceval, p. 161.

case, it is not proved that the transition was from ballad to epic in mediaeval literature, from short narratives to be recited or sung to long and complex pieces to be recited or read. A case could be made out by some one caring to elaborate the thesis, for the development from mediaeval epic to mediaeval ballad. No material at all, if the facts of chronology be scrutinized, can be found to illustrate the hypothesis for ballad origins of a " traditional epic process working upon material made at a primitive stage not quite beneath our sight," while material illustrating the contrary chronological order, mediaeval epic narrative, then mediaeval ballad, exists in abundance. We are told that "even a mere comparison of early stages, in a Babylon, a Maid Freed from the Gallows, with later stages in the Robin Hood cycle, ought to place this view [of narrative development from dialogue and situation songs originating in the dance] beyond denial." 40 But the long epic narratives of Robin Hood appear early and the more song-like ones, from which the former are supposed to develop, come later. And when we watch the development of existent mediaeval dance songs, or of present-day folk-improvisations, preserved under the right conditions, we find nothing which bears out the hypothesis of development from mediaeval song to ballad, to epic. Rather is it contradicted, if we discard conjecture and stay by fact in our consideration of material.

Since both are folk-poetry and both are preserved in tradition, comparison seems especially in place between English ballads in dialogue form and game and dance songs. The chief collector of the latter, Mrs. Gomme,41

40 Gummere, The Popular Ballad, pp. 284-285.

41 Dictionary of British Folk-Lore, vol. n, p. 500.

holds that game and dance songs in dialogue form are of later origin; her opinion is based on much first-hand experience with the ways of folk-song. And it is to be noted that dialogue and situation songs appear in other lyric types beside ballads and game songs. Many carols both of the literary and of the more popular types take this form and many religious lyrics, and so do laments and dirges; and these are preserved in texts antedating those of most of the ballads.

Communal improvised folk-poetry as we can watch it among cowboys, lumbermen, negroes, European peasants, does not exhibit the ballad of situation in dialogue form telling a story. As for primitive poetry, it is rather the progenitor of modern poetry and drama in general than specifically of a dialogue (ultimately becoming an epic) ballad type42 which makes its appearance during the Middle Ages. In neither modern improvised folk poetry nor in the choral singing and response of primitive poetry is to be found the body of material needed to bear out the theory stated at the outset of this chapter. Simple as it may seem, to tell a story with completeness and cohesion by dialogue is much too difficult for folk art, whether mediaeval, modern, or primitive. The safe generalization is that the story song is not a primary but a developed type in the evolution of literature, that the story song

*2 An illustration from primitive poetry representing the nearest approach which is reached to dialogue ballads is afforded by the harvest-song dance of a Boro chief, a two-line strophe to which his wife responds, in two lines of nearly the same words, to be followed by the same two lines from a chorus. See T. Whiffen, The Northwest Amazons (1915), p. 199. But such songs of primitive peoples are not the special ancestor of that minor lyric type, the ballad, but of song of many kinds.
in which dialogue predominates is still later,43 and that both emerge from a higher origin than unlettered folkimprovisation.

Ill THE "UNIFORMITY" OP THE BALLAD STYLE

"It is a significant fact," says a well-known writer on ballads,44 " that wherever found, the ballad style and manner are essentially the same." Many make the same generalization. But this is true only in the most general sense. It presupposes too great fixity in the ballad style. The ballad is a lyric type exhibiting epic, dramatic, and choral elements; but within the type there is as great variation as within other lyric types. The ballad style is hardly more "essentially the same" than the song style in general, or the sonnet style, or the ode style. There is no single dependable stylistic test even for the English and Scottish traditional ballads; and there are wide differences between the ballads of divergent peoples, Scandinavian, German, Spanish, American. There are differences in the stanza form, in the presence and use of refrains, iteration, and choral repetition, in the preservation of archaic literary touches, in the method of narration, and the like. The similarity in style of the pieces he included was the chief guide of Professor F. J. Child

*3 It may be pointed out that when a ballad is preserved in folk tradition dialogue sometimes gains prominence as the links in the narrative drop out. When only fragments of some ballad or song are remembered, these fragments are occasionally bits of dialogue. But such a tendency is not marked. In general it is what is most striking in the individual piece, a situation, event, tragic or comic crisis, striking turn of expression, sometimes the refrain only, for dance and game songs, that lingers in the memory, when the song as a whole has been lost.

« Walter Morris Hart, English Popular Ballads (1916), p. 46.

in his selections for his collection of English and Scottish ballads; yet he encountered such variety instead of essential uniformity that he was often in doubt what to include and what to omit, and fluctuated in his decisions. He made many changes of entry between his English and Scottish Ballads, published in 1858-1859, and his final collection published in ten parts, from 1882-1898. He would not have altered his decision concerning so many pieces had the test of style been so dependable as is usually assumed.

Even the stanzaic structure of ballads is not uniform. Some of the older ballad texts are in couplet lines, while the later are usually in quatrains, and there are many variants of both forms. The ballad stanza is hardly more stable than the hymn stanza. And it varies not only in form but in movement, in the character of the expression, and in the lyrical quality. Sometimes the story is told in the third person, sometimes, as in Jamie Douglas, in the first person, as is the case in so many Danish ballads. The ballads were obviously composed to be recited, or to be sung to or by popular audiences; and, like hymns, they show brevity and simplicity of form. Otherwise there is wide fluctuation. Were the style "essentially the same" the differences in the quality of the ballads would lie only in their plots. Yet two texts of the same story often have a gulf between them. A staple example may be found in the narration of the same occurrence in the earlier and the later texts of The Hunting of the Cheviot. The earlier text contains the effective and often quoted stanza —

For Wetharryngton my harte was wo,   that euer he slayne shulde be; For when both his leggis wear hewyne in to,   yet he knylyd and fought on hys kny. The corresponding stanza in The Chevy Chase sounds like a travesty —

For Witherington needs must I wayle

  as one in dolef ull dumpes. For when his leggs were smitten of, he fought vpon his stumpes.

The same discrepancy may be noted between Percy's and Motherwell's texts of Edward.

Many critics have commented upon the relative flatness of the style of the English traditional ballads compared to the Scottish. Professor Beers 45 thinks that the superiority of the Northern balladry may have been due to the heavy settlement of Northmen in the border region. Danish literature is especially rich in ballads. It is perhaps due in part to Danish settlement in the North and to the large admixture in Northern blood and dialect that the North Countrie became par excellence the ballad land. English ballads, unlike the lowland Scotch, are often flat, garrulous, spiritless, didactic. Professor F. E. Bryant 46 thought that the ballad of the Child type was not very current in Southern England, where the institution of the printed or stall ballad came to play so large a role and established a current type of another and less poetical pattern. The discrepancy in style between Northern and Southern ballads might then be ascribed to the dominance of stall balladry in London while it played no part

*5 A History of English Romanticism in the Eighteenth Century, pp. 266, 267.

« A History of English Balladry (1913), p. 192.

in the North. Mr. T. F. Henderson 47 places emphasis upon the superiority of Scotch lyric poetry in general in the fifteenth and sixteenth centuries. Its "makers" and bards were artists of special training and descent. Their influence is dominant for generations and their legacy may be seen in Scottish song of the eighteenth century. Northern vernacular song, he points out, is more closely linked to the past than the popular minstrelsy of England. It represents more fully the national sentiments, associations, and memories. It includes many numbers that bear the hall-mark of an ancient and noble descent.

The relation is close of the Northern ballad style to that of fifteenth century Scottish poetry and to Scottish popular song as it emerges in the eighteenth century. To cite illustration, Henryson's Robyne and Makyne and The Bludy Sark are astonishingly ballad-like in stanzaic form and in expression, though they were not composed for oral currency and the themes are not heroic or border themes. The Bludy Sark opens as follows: —

This hundir yeir I hard be tald

  Thair was a worthy king; Dukes, erlis, and barounis bald He had at his bidding.

This lord was anceanne and aid,

  And sexty yearis couth ring; He had a dochter fair to fald, A lusty lady ying.

The ballad mannerism of forced accent is noticeable, and in Robyne and Makyne especially striking use is made of dialogue. If these pieces had been composed for recita

47 Scottish Vernacular Literature (1898), p. 385.

tion or singing, if they had had oral currency for some generations with consequent transformations, assimilations, and re-creation, both might possibly seem the most orthodox of traditional ballads.

To return to the subject of variation of style within the Child ballads, the precariousness of style as a test of what is properly a ballad and what is not is shown by The Nut Brown Maid. It resembles some of the traditional ballads so closely in style as to win for itself for a long time treatment as one of the latter. It was included, for example, in the first ballad collection published by Professor Child. But it has now very properly lost such classification since it is really a debate piece, a bit of special pleading, not a lyric tale.

There are some who classify the American cowboy songs as "American ballads." 48 It need hardly be said that their style is utterly different from that of the Child pieces. Conventional epithets, wrenched accent, structural repetition in narration, use of the " legacy" motive, etc., are all missing save where the songs are made over from Old World ballads. Most, however, are songs rather than ballads, and their chief collector has so termed them.49

If by the statement that ballads show uniformity of style is meant that all ballads are likely to show a certain structural mannerism, i. e., structural or lyrical repetition, so-called "incremental repetition," it should be pointed out that this is not a differentia of the ballad style, or proof of some special mode of genesis for ballads, for it is a characteristic of popular song in general. Parallel's G. H. Stempel, A Book of Ballads (1917), p. 145. "J. A. Lomax, Cowboy Songs (1910).
ism of line structure and incremental repetition are found in mediaeval songs, both religious and secular, and in folksongs of many types: carols, student songs, nursery songs and lullabies, revival hymns, etc., as well as (in a distinctive way which is not the ballad way) in game and dance songs. Lyrical repetition in presenting narrative is found only in ballads, for the ballad is the only narrative type of folk-song; but ballads can be ballads which do not show it. Its frequent presence in English ballads is a characteristic which they share with other types of folksong. It is not an essential characteristic of their structure, and it is more abundant in later than in earlier texts. There are many varieties of it; and primarily it is something to be associated not merely with the traditional ballad style but with the style of folk-song in general.

Comparison shows many points of difference as well as of resemblance in the styles of Danish, Russian, Spanish, Scandinavian, English and Scottish, and American ballads. What they have in common are the features on which we rest the definition of folk ballads as a lyric type. They are story pieces, they are singable or are easily recited, and their authors and origins have been lost to view. The real truth of the matter may be stated as follows. There is no universal ballad style essentially the "same" apart from locality or chronology, even when we limit our consideration to traditional folk-ballads. Within one community, however, through a certain duration, there is likely to be uniformity of style in the ballads preserved in folk-tradition. Popular preservation has a levelling effect on pieces which have commended themselves to the folkconsciousness and have been handed down in tradition. Pieces of all types and origins are made over to conform to the horizons of the singers. A negro song may even take on characteristics of the English and Scottish ballads when recovered from white singers in regions where Old World ballads play an important role in the folk repertory.50 Examination of a body of folk-songs may reveal wide divergence of provenance and, originally, of style. Yet, as in the cowboy pieces, the appearance of homogeneity may soon be assumed.. They seem to be the product of, and to mirror the life of, those from whom they were recovered. Pieces of all types are assimilated in folk-song; in the course of time they come to borrow elements from one another; mannerisms which are easily caught spread; until similarity of style is approximated. The ballad stanza, like the hymn stanza, has certain limitations conditioned by the powers of the singers, or by the vocal and psychological limitations of popular song in general. Yet in the long run styles change for folk poetry as they do for book poetry. British popular song of the nineteenth century is not like that of the seventeenth, nor is that of the seventeenth like that of the fifteenth. American sentimental, comic, and patriotic popular songs of the twentieth century are of other patterns from those current in the nineteenth. The song modes of John Brown, Marching through Georgia, Old Dan Tucker, Zip Goon, Lorena, have given way to those of Tipperary, Keep the Home Fires Burning, The Long Long Trail, Over There. These are songs not ballads, and some of them are of British origin; but the same general

50 Compare John Ear Ay (Campbell and Sharp, Folk-Song of the Southern Appalachians, No. 87), in which, as in several other songs in the repertory of the singers contributing, a passage has been assimilated from the Old-World ballad, The Lass of Rooh Royal.

ization is true for the style of our contemporary storysongs or ballads. The uniformity of the ballad style is a uniformity for one people, or one class of people, during one or more generations; otherwise there is only the uniformity of simplicity to be expected of popular song of all types.

IV IMPROVISATION AND FOLK-SONG

It seems clear that it is time to instil caution into our association of the primitive festal throng improvising and collaborating, and hypothetical throngs of peasants or villagers collaborating in the creation of the English and Scottish popular ballads. Primitive song and the mediaeval ballads are separate phenomena, with a tremendous gulf in time and civilization between. No doubt some of the choral improvisations of savage peoples found or find permanence, as is the case with individual improvisations, and also with songs thought out in solitude — or "dreamed" in the Indian way. But such songs — consisting of a few words, or a few lines monotonously repeated — are quite a different thing from improvisations of length, having a definite narrative element, and high artistic value as poetry. Most primitive improvisations are no tax on the memory, and hardly, in view of their brevity, on the creative power.51 A singer with a good

si In the field of primitive ritual song there are many feats of memory that are quite wonderful. Long years are required for an Indian to become a really adept Tenderer of tribal rituals. See, for examples of verbal length, in the 27th Report of the Bureau of American Ethnology, the ritual song of 39 lines on p. 42, or that of 50 lines on pp. 571-572, at the bottom very nobly poetic. Similar examples are to be found in other tribes. Also there is something remotely analogous to ballad structure in such ritual songs as voice and a turn for melody might succeed, whether he could compose words very well or not.

When it is affirmed that improvising folk-throngs created the literary type appearing in the English and Scottish ballads of the Child collection, pieces like The Hunting of the Cheviot, the Robin Hood pieces, Sir Patrick Spens, Lord Randal, etc., the affirmation is pure — and not too plausible — conjecture. We have to do with long finished narratives, obeying regular stanzaio structure, provided with rhyme, and telling a whole story — pretty completely in older versions, more reducedly in the later. To assume that ignorant uneducated people composed these, or their archetypes, having the power to do so just because they were ignorant and uneducated, finds no support in the probabilities. There is strong doubt that a "choral throng, with improvising singers, is not the chance refuge, but rather the certain origin, of the ballad as a poetic form." There is still stronger doubt of the "acknowledged aptitude of the older peasant for improvisation and spontaneous narrative song," or of a statement like this: "There can be no question, then, of the facts. Popular improvisation at the dance has been the source of certain traditional lyric narratives."52 The following position is somewhat qualified from the preceding but it, too, represents conjecture rather than what is demonstrable: "The characteristic method of ballad authorship is improvisation in the presence of a sympathetic company which may even, at times, participate in

are given on pp. 206-242 of The Hako. But these ritual songs are not improvisations; nor are they of "communal" rendering.

52 Gummere, Cambridge History of English Literature, n, p. 456; Old English Ballads, p. 312; The Popular Ballad, p. 25.

the process. Such a description is in general warranted by the evidence though it cannot be proved for any of the English and Scottish popular ballads." 53 The author "belonged to the folk, derived his material from popular sources, made his ballad under the inherited influence of the manner described, and gave it to the folk as soon as he had made it."

We should remind ourselves that m our aay attempts to solve the problems of literary history proceed from the concrete to the theoretical. The methods of the transcendentalist yield to those of the scientist, who first gathers then scrutinizes his data. Certainly this is a better method than that which generalizes from an "inner light," looking about for whatever evidence may be found by way of support. A wise thing to do before reaching

ss Kittredge, Introduction to English and Scottish Ballads, p. xvii.

This view associating the origin of the English and Scottish ballads with the gathered folk-throng and improvisation has many adherents. It is the view to be found in our best known and most accessible books treating the ballads, like Professor Gummere's The Beginnings of Poetry, The Popular Ballad, and Democracy and Poetry, and it appears in the Cambridge History of English Literature and in the Kittredge and Sargent one-volume edition of the Child ballads, and it seems to have been accepted by F. E. Bryant, A History of English Balladry, 1913. Besides the many authors holding it who have been mentioned in preceding pages, it has the support of Professor G. M. Miller, Dramatic Elements in the Popular Ballads, University of Cincinnati Studies in English (1905), and apparently of Professor C. Alphonso Smith, Ballads Surviving in the United States, The Musical Quarterly, n, 116. There are dissenters from it here and there, whose work may be found in special articles. Among them was W. W. Newell, the distinguished folklorist (see Journal of American Folk-Lore, vol. 13, p. 113). But the theory of communal origin and emergence, with its emphasis on improvisation, retains the strategic position in literary histories and in special school editions of the ballads.
conclusions concerning the processes of the past, is to make sure what is true of the present; to look for parallel contemporary material and to keep it in mind when examining the older. If the past often casts light upon the present, the present, in its turn, may often cast light upon the past.

Surely then it is advisable, in handling problems of origin, to keep an eye upon the diffusion and establishment of types, in the folk-song of our own time, holding in mind changes and parallels in conditions, especially as compared with those surrounding the folk ballads of older times. Yet this has not been a customary angle of approach in discussions of the English and Scottish popular ballads. When considering a lyric type that arose in England in the later- middle ages, critics should give it not less but rather greater weight than argument from the anthropological beginnings of poetry, which of late years has monopolized the foreground of discussion. The subjects, the authorship and composition of primitive song, and the authorship and composition of the English and Scottish popular ballads are distinct; and, for both, the affirmation of characteristic origin by communal improvisation should no longer be made.

Of late years a considerable number of pieces composed by groups of unlearned people whose community life socialized their thinking have been made available to students of folk song, namely American cowboy and lumberman songs, and negro spirituals. It is hardly likely that human ability has fallen greatly since the middle ages; yet when we see what is the best that communal composition' can achieve now, and are asked to believe what it created some centuries ago, the discrepancy becomes unbelievable. The American pieces which, according to their collectors, have been communally composed, or at least emerged from the ignorant and unlettered in isolated regions, afford ample testimony in style, structure, quality, and technique to the fact that the English and Scottish popular ballads could not have been so composed, nor their type so established. In general, real communalistic or popular poetry, as we can place the finger on it, composed in the collaborating manner emphasized by Professor Gummere and Professor Kittredge, is crude, structureless, incoherent, and lacking in striking and memorable qualities.54 Popular improvisations are too lacking in cohesion and in effective qualities, to retain identity or to achieve vitality unless in stray instances, scattered in time and place; they are too characterless to be capable of developing into a literary type like the English and Scottish ballads. There are now many collections of American folk-song, made in many States. In these collections, the pieces of memorable quality are exactly those for which folk-composition can not be claimed. The few rough improvisations which we can identify as emerging from the folk themselves — which we actually know to be the work of unlettered individuals or throngs — are those farthest from the Child ballads in their general

"For material in support of these generalizations, see the discussion of Balladry in America, especially the section entitled "The Southwestern Cowboy Songs and the English and Scottish Ballads." Compare further the improvisations of our own fast-dying-out ring games and play party songs (for references see pp. 61, 64), and of children's songs. For the Old World compare the improvisations of Faroe Island fishermen, of Russian cigarette girls, of the South German Schnadahupfln, Bohme Geschichte des Tames, p. 239, and of labor songs, Biicher, Arbeit und Rhythmus, pp. 304, 327, etc.

f characteristics and in their worth as poetry.55. Nor is there a single instance of such an improvisation developing into a good piece, or becoming, as time goes on, anything like a Child ballad, unless by direct assimilation of passages from one of the latter. Yet they emerged from throngs no less homogeneous, perhaps more homogeneous than the mediaeval peasants and villagers.

The most homogeneous groups in the world are doubtless the military groups; yet war and march songs are always appropriated, never composed by the soldiers. The examples afforded by the war for the Union are still familiar; the favorite song developed by the Cuban war 56 was adapted from a French-Creole song; and we know the origin of the songs popular among the soldiers in the European war. If the "homogeneity" theory has any value, it ought to find illustrations in army life. And do prisoners in stripes and lock step ever invent songs? Granting the "communal conditions" theory, our penitentiaries should be veritable fountains of song and balladry. As a matter of fact, the most famous of prison ballads is the masterpiece of an accomplished poet,— Wilde's "Ballad of Beading Gaol."

Another thing shown by modern collections of folksong is that the songs preserved among the folk are nearly certain not to be those composed by them. Those they make themselves are just about the first to die.57 Usu

55 It is obvious that negro songs do not tend to assume a narrative type but retrograde to a simple repetition of phrases.

58 Joseph T. Miles, "A Hot Time in the Old Town Tonight." "Hail, Hail, the Gang's All Here," popular before and during the European war, utilizes for its melody the Pirates Chorus, from Sir Arthur Sullivan's The Pirates of Penzance.

"Illustration may be drawn also from the improvisations at the ally some special impetus, some cause for persistence or popularity, is to be detected for the pieces that live. And the striking or memorable qualities, or the special mode of diffusion, necessary to bring vitality are just what the genuine "communal " folk-pieces do not and cannot have. Most improvised poetry dies with the occasion that brought it forth. This is by and large a dependable generalization. What the folk improvises is typically flat and inferior and has no such vitality as the material assimilated and preserved by the folk from other sources.

The test of subject-matter should also be taken into account, when we are considering the likelihood that some process akin to the processes of primitive choral song and dance — continued through untold centuries among villagers and peasants — produced the Child ballads. The real communal pieces, as we can identify them, deal with the life and the interests of the people who compose them. They do not occupy themselves with the stories and the lives of the class above them. The cowboy pieces deal with cattle trails, barrooms, broncho riding, not with the lives of ranch-owners and employers; and a negro piece deals with the boll weevil, not with the adventures of the owners of plantations. Songs well-attested as emerging from the laboring folk throngs of the Old-World deal

old time revival meetings, where "a good leader could keep a song going among a congregation or a happy group of vocalists, improvising a new start line after every stop until his memory or invention gave out." See The Story of Hymns and Tunes, by Theron Brown and Hezekiah Butterworth (1896), chapter vii ("Old Revival Hymns"), pp. 262-297. But these improvisations did not live or produce new hymns. The material of the revival hymns and the mannerisms of the singing, especially of improvisation and protraction, had strong influence on negro folk-song, indeed afforded the background for the negro "spirituals." See pp. 129-132.

with the interests of factory life or agricultural life, or with the adventures of those of the social class singing or composing the songs. The improvisations of folk singers are usually personal, satirical, humorous, or vituperations, are lampoons and the like, and they grow out of the immediate interests or level of life or the latest occurrence among the singers. They are not often sentimenial and are not heroic, narrative, or historical. What then must we think of the English and Scottish ballads, if the people composed them? Their themes are not at all of the character to be expected. They are not invariably on the work, or on episodes in the life of the ignorant and lowly. Would they have had so great vitality or have won such currency if they had dealt with laborers, ploughmen, spinners, peasants, common soldiers, rather than with aristocrats? The typical figures in the ballads are kings and princesses, knights and ladies,— King Estmere, Young Beichan, Young Hunting, Lord Randal, Earl Brand, Edward, Sir Patrick Spens, Edom o'Gordon, Lord Thomas and Fair Annet, Lady Maisry, Proud Lady Margaret, or leaders like the Percy and the Douglas. We learn next to nothing concerning the humbler classes from them; less than from Froissart's Chronicles, far less than from Chaucer. The life is not that of the hut or the village, but that of the bower and the hall. Nor is the language parallel to that of the cowboy and negro pieces. It has touches of professionalism, stock poetic formulae, alliteration, often metrical sophistication. It is not rough, flat, crude, in the earlier and undegenerated versions; instead there is much that is poetic, telling, beautiful. It is for its time much nearer the poetry coming from professional hands than might be expected from me
diaeval counterparts of The Old Chisholm Trail and The Boll Weevil.58 No doubt there existed analogues of these pieces, i .e., songs which were sung by and were the creation of ignorant and unlettered villagers; but we may be certain that these mediaeval analogues were not the Child ballads.

The English and Scottish ballads should no longer be inevitably related to primitive singing and dancing throngs, improvising and collaborating. We can not look upon creations of such length, structure, coherence, finish, artistic value, adequacy of expression, as emerging from the communal improvisation of simple uneducated folkthrongs. This view might serve so long as we had no clear evidence before us as to the kind of thing that the improvising folk-muse is able to create. When we see what is the best the latter can do, under no less favorable conditions, at the present time, we remain skeptical as to the power of the mediaeval rustics and villagers. The mere fact that the mediaival throngs are supposed by many scholars to have danced while they sung, whereas modern cowboys, lumbermen, ranchmen, or negroes do not, should not have endowed the mediaeval muse with such striking superiority of product.

58 See Chapter vi, iii.