UC-NRLF f\J THE BRITISH ACADEMY On the History of the Ballads 1100-1500 By W. P. Ker Fellow of the Academy [From the Proceedings of the British Academy, VoL London Published for the British Academy By Henry Frowde, Oxford University Press Amen Corner, E.G. Price One Shilling and Sixpence net ON THE HISTORY OF THE BALLADS 11001500 BY W. P. KER FELLOW OF THE ACADEMY Read December 15, 1909 AT the beginning one is met by the trouble of definition. ' Ballad ' is here taken as meaning a lyrical narrative poem (all ballads are lyrical ballads) either popular in its origin, or using the common forms of popular poetry, and fitted for oral circulation through the whole of a community. But no definition, even if it were perfect, would tell as much as a reference to the great collections of ballads made in the last century. When we speak of ballads, we mean such poems as are found in the volumes of Child and Grundtvig ; in Nigra's songs of Piedmont, and Arbaud's of Provence. In spite of Socrates and his logic we may venture to say, in answer to the question 'What is a ballad?" 'A ballad is The Milldams of Binnorie and Sir Patrick Spens and The Douglas Tragedy and Lord Randal and Child Maurice, and things of that sort.' It is not a narrative poem only ; it is a narrative poem lyrical in form, or a lyrical poem with a narrative body in it. And it is a lyrical narrative, not of the ambitious kind, like Pindar, but simple, and adapted for simple audiences and for oral tradition, from one generation to another. The ballads in the great collections, many and various though they be, are obviously not complete representatives of popular poetry or of popular narrative. There are many other ways of telling a story ; and so the ballads may be contrasted with the folk tales in prose. There are many other kinds of song ; and the ballads may be contrasted with the pure lyric, such as 'Blow, northern wind !' or 'Lenten is come with love to town 1 . It is possible now, thanks to Child and Grundtvig and their successors and compeers in their own and other lands, to survey most of the ground at leisure. Some results, some general facts, appear to be ascertained by such a survey. iv. H 1 01 A :^:C::M&fcEE$INGS OF THE BRITISH ACADEMY ; . r : : . c ; ... .* . . Thus, it has long been known how wonderfully the ballads of different countries resemble one another. The notes of Child and Grundtvig are full of parallels and correspondences, traced and verified with an industry that leaves no difficulty unattacked. Now, beyond the particular work of tracking a ballad through disguises in all the languages English, Danish, Faroese, French, Romaic, Bohemian- there is the need of grouping the different languages as far as possible ; and much of this has been done. The chief results appear to be these, v as far as the Teutonic and Latin languages are concerned, and only these are within the scope of this essay l : 1. French. In the Romance languages there is a stock of ballad poetry common to France, Languedoc, Piedmont, and Catalonia ; well represented as a whole (with a full bibliography) in Doncieux, Romancer o populaire de la France (1904), and, for particular dialects, in Arbaud, Chants populaires de la Provence (two vols. 1862, 1864), Nigra, Canti popolari del Piemonte (1888), Mila y Fontanals, Romancerillo Catalan (1882); described by Gaston Paris in a review of Count Nigra^s book, Journal des Savants, 1889. This ballad-region is bordered in the south-west by the Castilian romances, on the south- east by the purely lyrical poetry of Italy beyond the Apennines. The Castilian ballads (Lockhart's Spanish Ballads} have a different history from the French-Provencal-Piedmontese group. Italian popular poetry, except in the North, is almost wholly pure lyric. Thus the essay on the Castilian romances by our honoured Fellow, Don Marcelino Menendez y Pelayo, 2 the delightful book on Italian popular poetry by Alessandro D^Ancona, 3 have little that bears directly on the ballad poetry of the middle province, as it might be called, between Castile and Tuscany. Go out of Castile to the east, out of Tuscany to the north, and you find the Catalan and Piedmontese ballads agreeing with one another, and with France and Provence. This middle region of ballad poetry, between Spain and Italy, is well defined. It should be noted, however, that the limit seems to be much clearer and sharper in the east than in the west. The popular lyric of Italy seems to imply a distaste for ballads, south of the Apennines ; but in Spain things are different. The national romances of Castile, it is true, belong to no other country, and they have a history and a growth of their own, different from other ballads. But at the same time there are among the Spanish ballads not a few that 1 In what follows, some things have been repeated from two papers on the Danish Ballads in the Scottish Historical Review, July 1904 and July 1908. 2 Tratado de los romances viejos, 2 vols., 1903, 1906. 3 La poesia popolare italiana, 1878 ; seconda edizione accresciuta, 1906. ON THE HISTORY OF THE BALLADS, 1100-1500 3 are distinct from the Castilian national romances and related to the ballads of France. And the Portuguese ballads belong mostly to the French group. 1 So that it might be advisable to take in the whole of the Spanish peninsula, and not Catalonia merely, along with the French province ; only marking out (a large exception) the Castilian ballads on national subjects such as have descended from older heroic poetry about the Infantes de Lara or the Cid or others. Then the ballad-region of the Latin races in the West would include every land, except Italy south of the Apennines. One may look on this ballad poetry as strongest and most flourishing between Italy and Spain, the Castilian national ballads remaining as before a class by themselves. / 2. Teutonic. The ballads of the Teutonic languages fall into three groups, English, Danish^ and German. The English are all together in Child's five volumes, The English and Scottish Popular Ballads, 1882-98, and also conveniently in one volume, where every ballad but not every variant is given (ed. Helen Child Sargent and George Lyman Kittredge, 1905). The Danish ballads are in Grundt- vig, Danmarks gamle Folkeviser, 5 vols., 1853-90, continued by Axel Olrik, Danske Ridderviser, 1895-1907 (in progress). For German, there is Erk and Bohme, Deutscher Liederhort (3 vols., 1893, 1894) ; and Uhland's two volumes may be taken as representative : Deutsche Volkslieder, 1844-5. The ballads of Norway, 2 Sweden, 3 the Faroes, 4 and Iceland, 5 are not distinct from the Danish ; nor those of the Netherlands 6 from the German. All these groups are more or less closely related. Compared with the German ballads, the English and Danish form one body, owing to the larger number of common themes, and still more to likeness in poetical form. The ballad burden which is universal in Danish and very common in English is not known or not used in the same way in German. More particularly, the English and Danish forms agree as against the German in 'their use of the inset burden : with a heigh ho, the wind and the rain, and sometimes there is very close likeness in detail. A Shetland ballad, derived somehow or other from the mediaeval romance of King Orfeo, has a Norse burden of this sort, not understood by the reciter (Child, No. 19):-- 1 Hardung, Romanceiro portuguez, Leipzig, 1877. 2 Landstad, 1853 ; S. Bugge, 1858. 3 Geijer and Afzelius, 1814-16, 1880 ; Arwidsson, 1834-42. 4 Hammershaimb, 1851-5. 5 Grundtvig and Jon SigurSsson, 1854-85. 6 Hoffmann von Fallersleben, Niedertandische Volkslieder, 1856. 4 PROCEEDINGS OF THE BRITISH ACADEMY Der lived a king inta da aste Scowan ilrla griin y Der lived a lady in da wast Whar giorten han grun oarlac. In spite of the close connexion between Denmark and Germany, the [ German ballads have had comparatively little influence in Denmark ; there is a much closer relation between Denmark and France. Danish adaptations of German ballads have been detected by tricks of phras- ing, easily recognizable, e. g. the rhetorical question, of which the most striking instance is in a Dutch version of the Konigskinder, the ballad of Hero and Leander. In the plot of this the lovers are ruined by a spiteful nun, who puts up false lights in the window, and the Dutch ballad renders it thus : What stuck she up ? three candles, Three candles of twelve to the pound. (Wat stac si op? drie keersen, Drie keersen van twaelf int pont.) When a question of this sort comes in a Danish ballad it is at once detected as strange (Olrik, Udvalg, p. 147). The other favourite device is the ' said he ' or ' said she ' put between two vocatives : ' O moder ', sede se, ' moder ! ' And this, among other things, is made to prove the foreign origin of the Danish Skjon Anna} It is not true Danish. 'Moder! 1 .sagde hun, 'Fruef and 'KongenP sagde hun, 'HerreP correspond to the Dutch ' Och moeder ! ' zeide ze, ' landsvrouwe ! ' and ' Koning Alewijn ! ' zeide ze, ' heere ! ' 2 The relation of Danish and French ballads is of a different sort. Some few Danish ballads are translated from German, and without much difficulty proved to be translations. 3 On the other hand, a large number of Danish ballads correspond to French ballads more 1 Cf. Steenstrup, Vore Folkeviser (1891),, p. 103 sq. The same sort of phrase is common in the early romance of King Horn, e. g. 1. 677 C : ( Lemman, qua)? he, dere.' 2 Hoffmann, No. 11, Schim Adelheid, Grundtvig, V, p. 40. The Danish ballad was translated by Jamieson and compared (in a letter to Scott) with Lord Thomas and Fair Annie as given in the Border Minstrelsy (Popular Ballads and Songs . . . with Translations of similar pieces from the ancient Danish language, 1806, vol. ii, p. 84). 3 Cf. Steeustrup, op. cit., p. 89 sqq. ON THE HISTORY OF THE BALLADS, 1100-1500 5 or less closely, in ways hard to explain. The most notable thing in this part of the history seems to be that there is more correspondence between Danish and French than between Danish and English. Ballads are found in the Danish and French groups which are not found in English, e. g. The Dead Mothers Return, DgF. 89, translated by Jamieson (Svend Dyring) in the notes to the Lady of the Lake ; cf. Les Ourphelins, Arbaud 78, and La Mere ressuscitee, Holland, No. clxxviii. The Milk-white Doe, DgF. 58, and Olrik, Daiiske Folkeviser i Udvalg, 1899. Cf. La biche blanche, Doncieux, No. xvi. The Sister rescued from a Tyrannical Husband, DgF. 62. Cf. La manmariee vengee par ses freres, Doncieux, No. xii. The Mariners in Distress (Dei frearlause menn), Bugge, Gamle norske Folkemser (1858), No. xvii. Cf. La courte paille, Doncieux, No. xvii. The Icelandic version of La courte paitte is specially noted by M. Jeanroy in his book on old French lyric poetry x on account of the form of its refrain. It corresponds, he shows, very exactly to the old French rondet. And it seems to be generally undoubted that the Danish ballads and their Scandinavian relatives have taken up the fashion of the old French dancing songs, a fashion which began its widely extended vogue, along with many other new fashions, about the year 1100. Many of the English ballads, and all those which have a refrain, belong to the same order, though the Danish group has kept much more of the dance tradition. The old choral ballad dance and song together is still preserved in the Faroe Islands. 2 The fashion of dancing and singing caroles on the Saints' Vigils (wake-nights) is proved by many pieces of evidence ; 3 and though of course there were other places and times for dancing and singing, it seems to have been at wakes especially that the ballad was wanted. The Icelandic viJcivaJci, the Danish Vaagencetter* and many phrases in the Danish and Norse ballads have kept a record of this. Ti vokunne klaeest ho Tore R6 ut cerlege menn! den greivedottere gde A den jomfru ! ('Thora dresses for the wake, the Count's fair daughter 1 ) DgF. IV, p. 478. 1 Origines de la poesie lyrique en France, 1889, p. 415. 2 Hjalmar Thuren, Folkesangen paa Fteroerne, 1908. 3 e.g. Giraldus Cambrensis, Gemma Ecdesiastica , R. S. ii. p. 120 ; Liber Exem- plorum Dunelm. ed. A. G. Little (1908),, p. 109 ; De ludis inordinandis. 4 A. Olrik, Udvafg, Introduction, p. 6. 6 PROCEEDINGS OF THE BRITISH ACADEMY The fashion is that which is recorded in the story of the Dancers of Kolbigk, which comes from the eleventh century. They were dancing in the churchyard on Christmas night and a judgement fell on them so that they could not stop. One of the versions of the story, which is traced by Gaston Paris from England to Lorraine, 1 has a quotation, in Latin, from the song the dancers sang : Ductor furoris nostri alludens fatale carmen orditur Gerlevus : Equitabat Bovo per silvam frondosam, Ducebat sibi Mersuindem formosam ; Quid stamus ? cur non imus ? Istud ioculare inceptum iusto Dei iudicio miserabile nobis est factum. Istud enim carmen noctes et dies incessabiliter girando per continuum redintegravimus annum. Semper vero insultabat nostrae poenae cantilenae regressus : Quid stamus ? cur non imus ? qui nee restare nee circulum nostrum mutare potuimus. And Gaston Paris shows the likeness of this couplet and refrain to the old French verse : Rainaus o s'amie chevauchent par un pre, Tote nuit chevauchent jusqu'au jor clair. Je n'avrai ja mais joie de vos amer. In the Asturias there are still, or were not so long ago, ballads sung at dances, 'on pilgrimages (romerias) and like occasions.' In the old Portuguese lyric poetry there are some very pleasant examples of the true ballad style, such as those written for diversion by King Denis (1279-1325) and preserved along with his more elaborate poems of the Proven9al school : De que morredes, filha, a do corpo velido ? Madre, moiro d'amores que mi deu meu amigo : Alva e vai liero ! De que morredes, filha, a do corpo loucano ? Madre, moiro d'amores que mi deu meu amado : Alva e vai Hero!* But while the ballad custom and the form of ballad verse can be traced back thus far, the copies of the ballads themselves are compara- tively recent. With the exception of Judas (thirteenth century) there /'is nothing in Child older than the fifteenth century (Stephen and Robin - and Gandeleyri). Nor are there earlier documents in other countries. 1 Les Danseurs maudits, Itgende allemande du XI e siecle, 1900 (from the Journal des Savants). 2 F. A. Wolf, Studien, p. 708, pp. 739-40 ; Das Liederbuch des Konigs Denis von Portugal, ed. Henry R. Lang (1894), p. xcv, p. 75, v. inf., p. 27. ON THE HISTORY OF THE BALLADS, 1100-1500 7 The oldest Danish MS. collection is about _ 1550 (when the earliest Spanish Romanceros were printed), though, just as in English, there are some few earlier remnants: one ballad (a silly story) Ridderen i Hjorteham, DgF. 67, about 1450. If ballads are older than this, how is their antiquity to be proved ? The evidence is strongest in Denmark. There is a proof from language, e. g. in Dr. Axel Olrik's examination of the ballad which corresponds to our Earl Brand and our Douglas Tragedy. Dr. Olrik has shown * that the differences of rhyme in various versions of this ballad may be explained and solved by restoring old Danish forms of words that have been altered in later Danish ; the Icelandic version of the ballad is used by Dr. Olrik to help in the restoration. The evidence from diction in the ballads has to be carefully watched. Antique words and phrases do not prove, straight off, that the poem is antique which uses them, e. g. the alliterative poem of Scottish Field, on the battle of Flodden, has many old words, wye 9 freke, and others, which live in alliterative verse for a thousand years, and obviously can of themselves tell very little as to the date of the poem in which they occur. The first page of Beowulf has at least two common types of phrase, about which we can only say, when we find them later, that they show how long-enduring a fashion of this" sort may be. One is the phrase of the opening, ' We heard tell,' which is found in other words at the beginning of the Hilde- brandslied and in the Lament of Oddrun, in Muspilli, and often else- where. As it is found also in Gaelic in the Dean of Lismore^s Book it is plainly part of the nature of heroic poetry, with nothing in it to prove a date. Nor can more be made out of the second instance bleed wide sprang : ' renown sprang wide. 1 It is common in ballads and in the longer romances as well ; in Ipomedon, ' This word sprang wide withall,' and at the beginning of the Danish ballad of Ranild Jonsson Det springer nu saa vide om land at Ranild er tagen til fange and in many other places. The corruption of ' middle-earth ' into ' middlarf ' in Herd's version of Clerk Saunders has spoilt one of the most beautiful things in ballad poetry, the ghostly regret for the living world Cocks are crowing on merry middle-earth, I wot the wild fowl boding Day. 1 Riboldsvisen in Danske Studier, 1906, p. 175 sqq., an essay which starts from a review of Dr. Ernst von der Recke's Nogte Folkeviseredaktioner a book to be remembered. 8 PROCEEDINGS OF THE BRITISH ACADEMY But this does not prove antiquity, for ' middle-earth ', though a fairly ancient term and much older as a piece of mythology, is not an un- known word in later English. It is interesting to find in the ballad of Gawairfs Wedding (Child 31) a phrase That bride soe bright in bower which comes in the old Northern poetry Maer var ek meyja mdSir mik faeddi biort i buri but here again all that is proved is the tenacity and perseverance of the old poetical diction. On the other hand, there are some instances of vocabulary which have been used in proving dates. A crucial example seems to be given in a Danish ballad, Hr Bjorn paa Sonder- borg, DgF. 473 F., which has preserved the old official name Stallari^ 'Marshal, 1 Stabularius. This name and office (the editor explains) was disused by Valdemar Seir about 1200 and replaced by ' Kansler ', 'Marsk', and 'Drost' Chancellor, Marshal, and Seneschal. It is remarkable that while this old word is wanting in the sixteenth- century versions of the ballad it is found in an eighteenth-century broadsheet copy and in oral tradition in South Jutland in 1895 : Der rider han Hr Bjorn Stolden og gjsested den lidel Kirstin. There are not a few other cases where modern versions taken from recitation are older and better than the MSS. two or three centuries earlier. 1 The best evidence of age is given by the Danish ballads on histori- cal subjects. 2 This evidence as to the age of the poems is internal, and possibly to some students may seem wanting in cogency. Those who use it are bound to prove that these historical ballads or the majority of them, at any rate follow closely upon the events themselves. On the other hand, it is not easy to find in Denmark the books out of which that splendid succession of ballads could have been made. A doubtful case is pointed out by Steenstrup, 3 where the words of a ballad and those of a chronicle (Sven Aageson's) are in close 1 e. g. Karl Hittebam, DgF. 294, where precedence is given to a copy from recitation, dated 1868, Soro, Saelland. 2 Cf. Steenstrup, Vore Folkeviser (1891), p. 315 sqq. 3 Op. cit., p. 220 sq. , regarding the ballad on the murder of Knud Magnusson at Roskilde, 1157 : Udi Dannemark er kommen konger tre Der herre Knud lille hleff veyenn. ON THE HISTORY OF THE BALLADS, 1100-1500 9 agreement, and generally it is not maintained that ballads on twelfth- century personages have come down from the twelfth century. But from 1200 onward there seems to be a nearer connexion between the facts and the poetry. And at the same time written history, after Saxo, begins to wither up. The chief argument for the age and un-bookish nature of these ballads is that the subjects are often taken in different ways and treated with different poetical motives. The story of the great Marshal Stig Anderson, ' Marstig ' as he is commonly called, is passed through a number of variations. Some considerable length of time is required for all these, and on the whole the most satisfactory view is to hold that the ballads of Marstig, like the ballad of the Earl of Moray, grew out of the reality with no help from any chronicle in prose. This cycle of ballads has been carefully examined by many critics, most recently by Dr. Heusler in his Lied und Epos. They are of the greatest interest in all manner of ways not least on account of the tragic sense in them. There is room here for a short summary. First comes a ballad on the murder of King Eric Klipping on the night of St. Cecilia s Day, Nov. 22, 1286. The burden of this is All the land in danger ' (Men nu stander Landet i Vaade). It begins : There are so many in Denmark would all be Lords and Kings (Der er saa mange i Danemark som alle vil Herrer vaere), and the ballad hardly goes beyond this simple motive ; the treachery of the ambitious great men, the danger to the kingdom, the pity of the King's death. None of the conspirators is named. No one is named, except the King's man Ranild Jonson, who was a villainous person. As a matter of fact, Stig Anderson (with others) was outlawed after the murder at Finderup ; a few years later (in 1290) he took the small island of Hjelm and made it a piratical stronghold. This (the second stage) is turned into a ballad without much regard for history or for the murdered king either. Marsk Stig is the hero. First come his dreams told to his wife and interpreted, as similar dreams are in the older poetry, and others in other ballads. He rides to the Parliament (the Thing) where the inquest on the King's death is to be held. There he is insulted by the Queen, and returns the insult. She tells him scornfully that he is making himself King of Denmark ; he answers that the Seneschal her lover has taken the place of the King. Then the young King breaks out in anger and banishes Marsk Stig ; and Marsk Stig threatens that if he is outlawed he will win his bread IV. H 3 10 PROCEEDINGS OF THE BRITISH ACADEMY from Denmark. So he makes his pirate station on the island, and the Danish yeoman smarts : The goodman goes to the field abroad All for to sow his corn ; And ever he prays ' God send us help, Since Helm has gotten a horn ! ' (and the overword is * My good lord young Sir Marstig '). In the third place, another ballad has taken up the plot of Tarquin, or of King Edward and the Countess of Salisbury, as the story is told, tragically, by Jehan Le Bel. When Marsk Stig was away, the King came and dishonoured his wife. When her husband comes home she will not go out to meet him, she will not rise from her chair when he enters ; her answer, in eight lines, to his question is one of the noble things of ballad poetry : ' When you went away I was a Knight's Lady ; now I am Queen of Denmark and find it hard to praise. Never shall you sleep in my arms till you have slain King Eric who wrought this grief.' The end seems rather weak (like Jehan Le Bel's story also). Marsk Stig upbraids and shames the King, but does not kill him. The refrain is, 4 But my Lady sits in Saelland with many a sorrow and pain ' (Men Fruen hun sidder i Scelland, saa mangt der hun sorger). Last in the series comes a very remarkable poem called by Dr. Olrik ' the long ballad ' (den lange Vise\ and taken by Dr. Heusler and others as an example of what can be done when shorter ballads are stitched or otherwise combined together to make one longer compre- hensive poem. The result, as Dr. Heusler proves, is not to make an epic poem, though it runs to more than a hundred quatrains; but the poem is honourable for all that. It begins with the dreams and their interpretations. Marsk Stig is summoned to court and sent out on an expedition with the King's banner. The King goes to Marsk Stig's house, and dishonours his wife Ingeborg ; l her words on her husband's return are nearly as in the shorter ballad. The defiance follows, with the technical term undsige for casting off allegiance an important word and idea in old Danish politics. 2 Then there is a new start and a new personage introduced Rane, Ingeborg's sister's son, with whom she plots the death of the King. So Ranild Jonson, the knavish attendant of the King in the earlier ballad, is worked into this more elaborate scheme. 1 In one version, through deceit and a lying story of her hushand's death. 2 The idea is found elsewhere, of course ; very clearly in the Castilian ballads. ON THE HISTORY OF THE BALLADS, 1100-1500 11 Then follows a strange adventure. King Eric goes out hunting, and at the end of the day finds himself alone and bewildered in the wood. He comes to a little house where there is fire and light, the house of an elfin damsel, a laughing lady. This looks like the deadly en- chantress of Clerk Colvill and other ballads (la belle dame sans mercy\ but she does no wrong to King Eric, though she warns him of his death ; she does not beguile him like other fairy queens, but escapes from his embrace, and the house vanishes and the King is left alone again in the forest. Here Rane finds him and guides him to the town. In the house there he is murdered by Marsk Stig and his company. Afterwards comes the riding of Marsk Stig to the Parliament, the railing of the Queen and his answer, his outlawry as in the older ballad, and the horn of Hjelm and the yeoman's prayer against the rovers as before. It is difficult to see how all this various ballad poetry on the years 1286-90 can be understood except as coming from poetical journalism to begin with, poetical rendering of matters which were vividly felt at the time. The ballad of Nils Ebbeson (1340) is an example of a different sort. 1 In the poems about the Marshal Stig Anderson, only the general facts are preserved. At first the motive is near to reality ; the first ballad of Finderup is a lament for the King, a complaint against the ambitious nobles. In the other ballad, the outlawry of Marsk Stig and his Viking settlement in Hjelm are not greatly distorted. But the story grows, and takes up other stories, wandering romantic stuff, tragic situations not in the original reality. In earlier days and in the older heroic poetry the stories of Ermanaric, Gundahari, Theodoric grew in the same way. The different ballads of Marsk Stig may be f compared with the different renderings of the Nibelung plot, in the ' Elder Edda ' and elsewhere. But Nils Ebbeson, though it has some poetical deviations in it, is meant to keep closely to the truth, and it comes very near success. 2 It is more like the Lay of Maldon in character than any of the Nibelung poems ; it was composed immediately after the events and it is full of the spirit of that day, the national rising in Jutland against the foreigners from Holstein. 1 Cf. A. D. Jorgensen, De historiske Folkeviser og Nils Ebbesen in Historisk Tidsskrift, 1891. 8 Cf. Olrik, Udvalg, Introduction ; Steenstrup, op. cit. ; Jorgensen, op. cit. (an answer to Steenstrup) and Bidrag til Nordens Historic i Middelalderen, 1871 ; Erslev, in Hist. Tidsskrift, 1898 ; Sofus Larsen, Nils Ebbeten* Vise, Aarb. for nord. Oldk. 1903. H 4 12 PROCEEDINGS OF THE BRITISH ACADEMY At the same time, like the Maldon poem, it manages to bring in the right heroic motives, more particularly the independence of the free man, and along with that the duty of loyalty. ' Grev Gert \ Count Gerard, ' Gerardus magnificus comes Holtzatorum ', was killed at night (Apr. 1, 1340) in his lodging at Randers by Nils Ebbeson and his men, who spread confusion in the town and get away, throwing down the bridge behind them. In the ballad the motives are clearly brought out. Count Gerard wishes to insist on feudal law, which is not that of Denmark ; against him Niels Ebbeson explains that in Denmark a vassal may ' take leave ' (tage Orlov) of his lord, may renounce (undsige) his lord when he chooses so to do. For this the Count threatens him with hanging or banishment ; so Niels goes home and calls out his men to the attack. For the events of 1340 there is a good deal of historical evidence apart from the ballad. But it is impossible to refuse the ballad its place as an independent version of what happened. It has no nearer relation to the Liibeck or the Holstein Chronicles than the Maldon poem has to the English Chronicle for 991 or 993. The ballads of Marsk Stig show how readily an historical fact might be turned into tragic romance. The ballad of Niels Ebbeson proves how well the ruling motives of earlier heroic ages might be expressed in ballad form. The action is of the same sort as in many older stories ; not unlike the well-known heroic story in the English Chronicle of Cyneheard's attack, in 786, on Cynewulf of Wessex at Merton. The motive of loyalty so well represented in the old English history appears unchanged in the Danish ballad. Niels Ebbeson getting away from Randers is hindered by his brother-in-law, and calls to his brother-in-law to let him pass on account of their relationship ; he is answered almost in the very words of the old English story : * I know I am near to thy kin, but thou hast slain my lord and I may not let thee go.' * It is true that internal evidence may be deceptive ; it is certain that 1 N. loq. H6r du det Hr Ove Haas lad mig min Vej bortfare ; du vedst fuldvel, du est min Maag, du maa mig intet skade. 0. loq. Alt er det Sanden, jeg bar din Fraenke og jeg din Maag skal vaere ; du bar nu slaget min Herre ihjel, jeg maa det ikke gjore. f Maag ' and ' Herre ' are ' maeg ' and { hlaford ' in the Anglo-Saxon Chronicle : ' Ond J>a cuaedon hie J>aet him naenig maeg leofra naere Jjonne hiera hlaford, ond hie naefre his hanan folgian noldon.' ON THE HISTORY OF THE BALLADS, 1100-1500 13 ballads exist whose themes are taken from books, yet whose style and form show hardly a trace of any bookish origin. Such is the Shetland ballad of Orpheus^ the Icelandic of Tristram, the Danish of Paris and Helen. But even granting the utmost to a sceptic who would doubt the age of the historical ballads, the sceptic still has to explain the endurance and persistence of the ballad form. The ballad form in different parts of the world can take up the story of Orpheus or of Helen and turn it into its own likeness. This ballad form cannot be new when it first appears ; it must have a long history. The greater its success in transforming bookish matter to the likeness of a ballad, the stronger is the proof of its age. Ballad poetry is a long established custom before it can produce such things as these. One of the difficulties about the ballads is that while so much in them seems to be ancient or even primitive, the rhyming ballad verse is comparatively new. Some of the common ballad devices, particularly that of repetition, seem to be as old as anything in humanity, and a large number of ballad subjects are no less widely spread. But the form of verse is not old. In the Teutonic languages, the first appearance of the new rhyming measures can be roughly dated ; they can hardly be older than the eleventh century. How did the folklore themes, the ballad habits of phrasing, find expression before the rhyming stanzas and the new sort of refrains were intro- duced from France ? Where were the ballads before they were made? There is no definite answer to be given ; but it is well to recognize what is proved by the older Teutonic poetry, especially by the Anglo- Saxon, that an old civilization with an elaborate literature of its own came to an end in the eleventh century, and that there is a great division about that time between the earlier and the later Middle Ages, and great difficulty in understanding the transition. Modern poetry, including the ballads, begins about the year 1100 ; we are cut off from the time before that, and from its tastes in poetry, as we are not from any of the rhyming poetry French, Provencal, Italian, German, English from that time onward. But it should be remembered that part of the difficulty in under- standing the former age (roughly, before 1100) comes from the mere accident that so very little of its poetry has been preserved, and, in that little, so very much less of the popular unambitious sort. For the later Middle Ages (1200-1500), though there is not enough, there is a very much larger amount of popular verse in existence. But here and there in the earlier period one discovers the same sort of popular tastes as are found much more fully represented in the later. 14 PROCEEDINGS OF THE BRITISH ACADEMY There were the same comic stories ; only, whereas the later Middle Ages got them in the easy form of fabliaux, and in large numbers, the earlier time has only preserved a few by turning them experimentally and as a sort of literary game into Latin verse. 1 It seems a fair conclusion that the difference between the earlier and the later Middle Ages e.g. between 'Anglo-Saxon' and 'Middle- English' is in some respects not as great as the existing remains would make us imagine. Nothing can do away with the enormous difference, in literary ambition and aim and style, between Beowulf and Bevis of Hampton. But it is pretty certain that beneath this difference there was the same kind of folklore. The ancient Germans knew the story of Big Claus and Little Claus; they had the same jokes as the fabliaux and the Decameron, though by the literary fashions and conditions of their time they were not encouraged to put these things in writing, and only did so occasionally and acci- dentally. Later, and mainly through the influence of France and the much less pretentious narrative forms of France, it was easier for folklore to get into literature. It is a fact that many poetical subjects have been transformed from the old Teutonic verse into later rhyming forms, sometimes into pure ballads. Not to speak of the Nibelungen, or the German-Danish ballad of GrimilcFs Revenge, there is the broadsheet rhyme of the death of Ermanaric, ' de konink van Armentriken,' printed in 1560, a thousand years after the notice of the same Gothic story in Jordanes. The Norwegian ballad of Thorns Hammer agrees in sub- stance with the prymskvida. It is possible for themes of the early centuries to come through all the changes of language and poetical taste, and to accept the comparatively modern rhyming forms of the Nibelungen in one instance, of the Hammer ballad in another. So there is nothing unreasonable in thinking that other ballad-plots may have come through in the same way, though nothing happens to be extant to show them in their older form. 2 In this country, the folklore of ballads has been described by Mr. Andrew Lang 3 in esjlys to which more readers and writers are 1 Modus Florum ; Modus Liebinc ; Unibos (Miillenhoff and Scherer, Denkmaler ; Grimm and Schmeller, Lateinische Gedichte). In one passage in the ' Elder Edda ' Odin appears in the character of the Baffled Knight. 2 In this connexion one may remember the most interesting references to German, Danish, and Spanish ballads in Panzer's Hilde-Gudrun, 1901. 3 ' Ballads ' in Encyclopaedia Britannica ; ' Ballads ' in Ward's English Poets, I. (1880) ; < The Ballads, Scottish and English ' in Chambers's Cyclopaedia of English Literature, ed. David Patrick, I. p. 520 sqq. (1901). ON THE HISTORY OF THE BALLADS, 1100-1500 15 indebted than ever have been able to express their thanks properly. The ballads have modes of thought and favourite ideas such as are found over all the world, and they draw from the same sources as the folk tales. They belong to the people, it is said, and the people are in some sense or other the authors of the ballads. The term * communal authorship ' is used by some writers about the ballads, not always in the same sense. 1 Here, under the head of folklore, there are two different things to be considered ; first, the traditional subject, and next the share of 'the people ', or the community, in giving it poetical form. As to the subjects there can be no doubt that there is freedom of communication a free passage between the popular tales (Marcheii) and the ballads, with this most important condition, that nothing shall be taken up by a ballad except what is fit for the ballad form. The meaning of this' is that ballad poetry has a mind of its own, quite as definitely as Greek or French tragedy, and will not take up a subject which is too complicated or too large. Thus a number of fairy tales are unfit for ballads because there is too much matter in them, too many adventures. For the same reason, the ballad is generally different in plot from the narrative romances. Further, the ballad has no fondness for the happy ending, which is generally right in the fairy tale. The tragic motive is as common in the ballads as it is in the older heroic poetry, and the modern Greek name for a ballad, Tpayoudi, might be taken also for those of other languages. The best of them are lyrical tragedies. But this difference in spirit and purpose between the popular fairy tale and the ballad does not hinder transactions between fairy tales and ballad poetry, when the fairy tales have the right sort of com- modity to offer. And there are many folklore things which seem to belong more properly to the popular ballad than to the prose story : There are certain incidents, like that of the return of the dead mother to her oppressed children ; like the sudden recovery of a fickle bridegroom's heart by the patient affection of his first love ; like the adventure of May Colvin with a lover who has slain seven women and tries to slay her ; like the story of the bride who pretends to be dead that she may escape from a detested marriage, which are in all European countries the theme of popular song. 2 It is possible to find those themes apart from the ballad form, but it will be generally found also that the ballad form is what they want to bring out their meaning and value. 1 Cf. Francis B. Gummere, The Beginnings of Poetry (1901), especially ch. v. 2 A. Lang, in Ward's Poets, I. p. 103. 16 PROCEEDINGS OF THE BRITISH ACADEMY Apart from the complete folk tale there are incidents and machines that may be taken freely by the ballad poet and employed in his own way. Lately (by Serendipity) I came upon an example. Dr. Axel Olrik, in a paper in Danske Studier, 1 on the old Northern ball-game, has put together a number of ballads which use in one way or other the incident of the boy who plays roughly, and is told by his com- panions to go and avenge his father. In Irish, this is part of the story of Maelduin. There are four Norwegian examples, five Faroese, two or three Danish. Child Sigurd (Sigur svein) goes to the playing-field where the King's small boys are playing at the ball, and he plays them to rights; he struck one under the ear so that they laughed at him. Sigurd struck with the ball (he was stout of growth), sore hurt were the King's boys, and the blood springs out on the earth. Then out and spoke the small boys, so angry were they : ' It is fitter thou ask after thy father than hurt us so. 1 Sigurd flings away the play-ball, he has no wish to play more ; and so he went to his fair mother, and his face was wan. The following passage is given by Mr. R. A. Nicholson in his Literary History of the Arabs (p. 94) : He grew up strong in the arms, and one day he had a tussle with a youth of the Banu Zafar, who said to him : 'By God, thou wouldst do better to turn the strength of thine arms against the slayers of thy father and grandfather instead of putting it forth upon me.' 'And who are their slayers ?' ' Ask thy mother, she will tell thee.' So Qays took his sword and set its hilt on the ground and its edge between his two breasts, and said to his mother : ' Who killed my father and my grandfather ? ' In New Guinea it is much the same : Now as the child grew bigger he grew also fierce and violent. Day by day when playing with the village children he would ever strive to be chief in the games, and desired that all should bend to his will. And it came to pass on a certain day that he, as was his wont, was beating and stoning the children that pleased him not, and one bolder than the rest cried : 'Why dost thou ever beat and ill-treat us ? Is it thy vengeance for thy three kinsfolk who are dead ? ' 2 There is no need to prove the fact that there is a strong folklore element in the matter of the ballads ; things are more difficult with respect to the form. Ballads may be made out of common plots and incidents, but how do they get their poetical form, and in what sense is communal authorship a fact ? With a certain class of subject one can understand communal 1 1906 ; p. 91, Drengene paa Legevolden. 2 Annie Ker, Papuan Fairy Tales (Macmillan, 1910) ; p. 40, Kakukaku and Taureboga ; cf. also, p. 61, How the twins killed Manubada. ON THE HISTORY OF THE BALLADS, 1100-1500 17 authorship, even if one has never seen it working. The theme is given out a success to be praised, a butt to be ridiculed, a grief to be lamented and many people may take part, adding verses all on the common motive. 1 In what order their several contributions are arranged is not of very great importance. But where there is a story the case is different. Where there is a definite plot, this must surely reduce the communal share and increase the responsibility of one individual member of the community. A ballad is not the same thing as a fairy tale. The tale may be told in any language, in any style, without losing itself. But the ballad is a poem, i. e. it has a form of its own, belonging to a particular order of poetry. Much in the ballads is common folklore. But the ballads can also take up new subjects (e. g. the Battle of Otterburn, or Jamie Telfer) ; how are these to be managed by ' the people \ if ' the people ' are the authors of ballads ? At present it does not seem to be maintained by any one that ballads with a definite story or plot (as distinct from laments or lampoons or rejoicings) are really made by a crowd. But it seems to be generally taken for granted that ' the people ' exist, and that it is unnecessary to prove this dogma. Who are the People ? It is assumed too readily that there always is a 'people' or populace, distinct from the gentry, simple people in possession of folklore and the forms of thought required for ballad poetry the love of the number three, of gold and silver, of verses repeating the same thing in slightly varied words or rhymes. But different nations have different kinds of populace, and some have none at all. It is possible for a nation to be gentle all through 'the Quality 1 not a separate caste from the Quantity. Iceland is one such, andJVVales is another, with regard to literature. In W r ales the popular taste in poetry is courtly, if 'the courtly maker' is to be judged by subtilty of artifice. In Iceland the rules of the poetical game are much less exacting than in Wales, but much more than in England, and the rules are generally understood through all the country. The most popular form for many centuries in Iceland was Rimur long narrative poems using varied and difficult types of verse, under definite rules of prosody. The Rimur on the Gowrie Conspiracy were edited lately at Oxford by Mr. W. A. Craigie in an excellent book which gives the key to this sort of poetry. Mr. Craigie says in his introduction, 'the Rimur may be described as ballads.' Is this fair? 1 Oue would like to know more of the Faroese Trawlaravlsa of which Mr. Thuren speaks (op. cit. p. 35), a ballad on the English and American trawlers who have taken the place of the Algerine rovers in those seas. 18 PROCEEDINGS OF THE BRITISH ACADEMY Is it not the irony of a disillusioned lexicographer, who knows that the human race (diese verdarmnte Race) will use any word in any way it pleases, without regard to philology or any other science? The Icelandic Rimur might be called ballads, but that is not their right name. They have taken the place of ballads in Iceland ; but they are long elaborate poems, divided into several books, with a change of metre in each book, and the metres all correct and none of them too easy. They are generally paraphrases of prose documents (e. g. SJcotlands Rimur, wholly taken from the Danish translation of the Scotch official narrative of the Gowrie plot), and they are full of rhetorical ornaments. It is interesting to compare the Rimur with the Faroese ballads. The Faroe islanders are as much at one among themselves as the Icelanders in literary taste. But they have put all their hearts into ballad poetry, keeping up the old ballad customs as no other people have. At the same time they, like the Icelanders, have a liking for long stories, and some of their ballads are enormous. Also like the Icelanders they have made great use of books ; so much so that it has come to be a commonplace to begin a ballad with a reference to the book from which the story is taken. 1 There are influences here, dragging the lyrical ballad over into the other species, mere narrative. Yet in spite of all those interferences, the ballad quality is kept in the Faroese poetry, and the difference from the Icelandic Rimur is very great. In another country it can be proved how various the popular taste in poetry may be, by comparing the ballads of Piedmont with the lyrical stanzas of Tuscany or Sicily. The people of Tuscany seem to be capable of stanzas which in point of art would be dangerous company for a good deal of rather ambitious English verse. Their rispetti have more likeness to Shakespeare's sonnets than to Sir Patrick Spens or Einiwrie^ especially in the effect of their opening lines : Quando sentirai dir che saro morta, or O Sol che te ne vai, che te ne vai ! or Non ti maravigliar se tu sei bella. In Piedmont, instead, the openings are such as these : Sun tre giuvenin de scola, ch'a Tuluza volo ande (which is the Clerk's two sons of Owsenford), or Sun tre frateli, Tan ch'na sorela a marida (La Sorella vcndicata). 1 Cf. G. Vigfusson, Sturlunga, Prol. clii. ON THE HISTORY OF THE BALLADS, 1100-1500 19 These popular ballads of Piedmont, or those of the Border Minstrelsy, may be as different from literary poems as the stories of Grimm from the Sorrows of Weriher. But there is a fallacy in arguing from the more recent states of ballad tradition as shown e. g. in Kristensen's researches in Jutland, in Scott's Liddesdale raids, or even in the Percy MS. back to the times in which the ballads were flourishing. Denmark is the key of the position. There is no better account of the modern life of ballads than that given by E. T. Kristensen, one of the greatest of collectors. In Jutland in the nineteenth century the ballads were alive among the people, and along with folklore stories (JEventyr, Mdrchen) they served in all sorts of ways for entertain- ment repeated by women at their work, or to children to keep them quiet ' in huts where poor men lie ', or on the tramp over the moors. But this is not the original life of the Danish ballads. We know that in the sixteenth and seventeenth centuries they were often copied out by gentlefolk, by many ladies, and when they were first printed it was through the favour of the Queen. 1 The Danish historians are agreed that the ballads were originally, and for long, the pastime of the gentry. The Faroe islanders in their ballad-dances have preserved what was the favourite amusement in the old Danish country-houses. This came in as a fashion first among gentlefolk. The ' popular ' features here were not derived from the Danish 'populace'; or, alternatively, it may be argued that 4 populace ' here includes the whole nation. And it is possible to understand why it should be so. There were\ no other, more ambitious, literary forms to depreciate the ballads. \ Obviously, where there is a strongly prevalent literary ambition, the folklore elements in the national mind (i. e. in the common stock of ideas) will be either neglected (as in I^atin poetry for the most part) or transformed (as in the Odyssey, in Beowulf, in the Arthurian romances), or treated in a humorous observant way, as in Burns's Halloween-, Burns was an aristocrat. It is fairly easy thus to tell 1 Cf. Jamieson, Popular Ballads, ii. p. 99 : ' Of these [the Kcempe Viser] the first centenary was published in 1591 and dedicated to Sophia, Queen of Denmark, &c., by the Rev. Andrew Soffrenson [i. e. Vedel] who seems to have been a man of learning and respectability, and in habits of intimacy with his celebrated countryman Tyge Brahe, whom he calls " that worthy honourable and well-born man, my affectionate master and good old friend ", with whom it appears that the Queen happening to be storm-stayed during three days at Knutstrup, in 1586, had chatted away the time very agreeably, paa Bordet ; and to one of these conversations we are indebted for the publication of the K