The Cambridge History of English and American Literature in 18 Volumes (1907–21).
Volume II. The End of the Middle Ages.
§ 2. The Canute Song
Yet another line of demarcation must be drawn. English and Scottish ballads as a distinct species of poetry, and as abody, can be followed back through the fifteenth century, occur sporadically, or find chance mention, for a century or so before and then altogether cease. Owing to the deplorably loose way in which the word “ballad” is applied, not only the references of early historians, like William of Malmesbury, to the “popular songs,” the cantilenae, the carmina vulgaria, from which they draw for occasional narrative, but also the passages of older epic that tell a particular deed or celebrate a popular hero, are, alike, assumed to indicate a body of ballads, similar to those of the collections, extending back to the Norman conquest, back even to the Germanic conquest of Britain, but lost for modern readers by the chances of time and the lack of written record. Such a body of ballads may, indeed, be conjectured; but conjecture should not pass into inference. Not a single specimen is preserved. It is, to be sure, unlikely that the primary instinct of song, the tendency to celebrate heroes and events in immediate verse, and the habit of epic tradition, main constituents of balladry, should cease as we cross the marches of the Transition period and pass from the modern speech and modern metres, in which our ballads are composed, into that more inflected language, that wholly different form of rhythm, which prevailed in Old English and, with some modifications, in all Germanic verse. To claim for this older period, however, ballads of the kind common since the fifteenth century in England, Scandinavia and Germany, is an assertion impossible to prove. The Old English folk must have had popular ballads of some sort; but it cannot be said what they were. Singing, to be sure, implies a poem in stanzas; and that is precisely what one cannot find in recorded Old English verse—the one exception, Deor’s song, being very remote from balladry. It is true that the subject of a popular ballad can often be traced far back; Scandinavian ballads still sing the epic heroes of “Old Norse.” Community of theme, however, does not imply a common poetical form; and it is the structure, the style, the metrical arrangement, the general spirit of English and Scottish ballads, which must set them apart in our literature and give them their title as an independent species. We find a relative plenty of “popular” verse in the thirteenth, fourteenth and fifteenth centuries—songs by a political minstrel of some sort, which had their immediate vogue, were recorded here and there, and soon forgotten—but this sort of thing should not be confused with songs made among the people, passed down by oral tradition and marked with those peculiarities of structure and style which are inseparable from the genuine ballad of the collections. In the absence of texts, conjecture is useless. The earliest recorded piece of English verse which agrees with balladry in all these important characteristics is the famous song of Canute, preserved in the chronicles of Ely. The king’s actual part in the case is doubtful, and unimportant. Coming by boat, it is said, with his queen and sundry great nobles to Ely, Canute stood up, bade his men row slowly, “called all who were with him in the boats to make a circle about him … and to join him in song; and composed in English a ballad (cantilenam) which begins as follows:
The verses are familiar; but their significance is not always noted. The chronicler turns them into Latin, and, with clear reference to popular tradition, adds—“and so the rest [of the song] as it is sung in these days by the people in their dances, and handed down as proverbial.…” That is, the song was traditional a century and a half after the supposed fact, and it seemed natural to the chronicler that such a cantilena should be improvised to the singing of a chorus. Perhaps songs of this kind were in Malmesbury’s mind when he apologised for using as material for his history cantilenae “worn by the friction of time”; but the political verse of minstrels like the later Laurence Minot is a more likely assumption; and, whatever the likelihood, the verse itself has vanished. In Canute’s case there is a fragment of actual song, of the highest value; for it is not only one of the earliest recorded pieces of English poetry to break away from the uniform stichic order of old English metres, but it is in the rhythm which belongs to the best English and Scandinavian ballads of tradition. Grundtving thinks that the quoted lines are the burden or chorus of the piece, which was doubtless narrative in its further course, and told, one may conjecture, of Canute’s own deeds. This desire of the warrior to sing the battles he has fought did not pass away with the lost songs. A passage in bishop Leslie’s History of Scotland, used in part by Andrew Lang for the solution of the problem of ballad origins, declares that “our bordir men,” as Dalrymple translates, delight in their own music and in the songs that they themselves make about their deeds and about the deeds of their forbears. The bishop’s Latin is unequivocal: cantiones quas de majorum gestis, aut ingeniosis praedandi precandive stratagematis, ipsi confingunt. Gaston Paris, on good evidence, has made a similar assertion about the early Germanic and English warriors, who, before the days when the minstrel existed in a professional class, sang their own deeds and furnished the prime material of later epics. Even in Beowulf a warrior is described improvising a song on the defeat of Grendel. There is, thus, a presumption that border ballads, like Cheviot and Otterburn, owed their earliest form to the improvisation of fighting men who could sing their own deeds; and thus, too, one draws a faint line, mainly touching theme and conditions of origin, from the “old song of Percy and the Douglas” back to those lost lays that inspired the poet of Beowulf.
But this is all. Of the actual structure and form of those old lays nothing is known; and it must be remembered that even Cheviot and Otterburn, while of the undoubted general type of balladry, are not, in more exact analysis, of the typical construction which one finds in ballads recovered from genuine oral tradition. All that can be said of material gathered from older chronicles, or suspected in older poems, is that it lends itself to conjecture, not to proof. The one exception is this song of Canute, which may pass as a genuine ballad fragment.