English, Department of


Date of this Version



PMLA, vol. 35, no. 2; New Series vol. 28, no. 2, (1920) pp. 161-188


Many origins have been suggested for the type of narrative song appearing in the English and Scottish traditional ballads: minstrel genesis, origin in the dance, improvisations of media3val peasant communes, or descent from the dance songs of primitive peoples. The hypothesis of minstrel origin was that first to be advanced and it has always retained supporters. There remains a possibility not yet brought forward which deserves to be presented for what it is worth, since the problem, though it may be insoluble, has its attraction for critic and student. We have but meager knowledge of the ballad melodies of pre-Elizabethan days, and we can get but little farther with the study of the ballads by way of research into medieval music. Moreover the earliest texts remaining to us seem to have been meant for recital rather than for singing. In general, the melodies of ballads are more shifting, less dependable, than are the texts, in the sense of the plots and the characters which the texts present. This is true of contemporary folk-songs and it was probably true earlier. One text may be sung to a variety of airs or one air may serve for many texts. Nor can we get much farther with the study of ballads by way of the minstrels. They have had much attention already; and nothing has ever been brought out really barring them from major responsibility for ballad creation and diffusion in the ear- lier periods. Again, we can get but little farther by studying the medieval dance, or folk-improvisations, or the dance songs of primitive peoples, all of which have been associated with the Child ballads to an exaggerated degree. It is time to try a new angle of approach-the last remaining-although the hypothesis which it suggests is far re- moved from the theory of genesis enjoying the greatest acceptance at the present time, and although it-like its predecessors--may not take us very far. It has been customary among theorizers completely to discard the chronological order of the ballad texts remaining to us, and to argue toward origin and development from a type of ballad like Lord Randal and Edward, of comparatively late appearance, when such reversal of chronology best suited the theory to be advanced. The contrary procedure, theorizing from the facts of chronology, is the logical one. If the ballad texts which are oldest are given attention and emphasis, actual fact adhered to and conjecture omitted, can anything distinctive be reached? This method of approach is one to which the ballads have never been subjected in more than a cursory way. If it is tried, in what direction does it lead?