Page:Old English ballads by Francis Barton Gummere (1894).djvu/103

INTRODUCTION. xcvii with memory, is a necessity; and we have thus, by steps legitimate in every way, taken our narrative ballad back to a communal origin, and removed it from the conditions of individual authorship. Modern critics would teach us that the ballad, made as any other poem is made, gets its impersonal and differencing quality by oral transmission alone; a glance at the relations which melody, dance, refrain and improvisation bear to the later narrative ballad shows us that its earliest form could never have been that of a poem such as individual authors compose, and it is these four elements, moreover, dwindled and uncertain as they are, which give us our best notion of primitive poetry in its habit as it lived.^ The ballad, then, is a survival from this vanished world of poetry. The particular ballads of the present collec- tion cannot be referred directly to communal authorship; but their differencing qualities, the impersonality, the hint of something which we cannot define but as little can deny, are due to this older connection, and are not to be explained by mere oral transmission.^ Sincere, 1 The relations of ballad and epos» intricate enough in detail, offer no difficulties from this point of view. Germ or aftergrowth of the epos, the ballad has its independent place as a poetical fact. If, however, the reader cherishes a Shandean love of comparisons, let him make what he can of this (from Ferdinand Wolf, in Wiener Jahrbiichery CXVII, 87): "The pure and original epos is the evening-dream after sunset, sinking peacefully into vague memories of the past; the lyric-epic ballad (volkslied) is the shadow of the forward-hastening star of day." Wolf was one of the first to protest against " nebulous " definitions. 2 Nigra compares the changes of a ballad with the changes of a dialect; new words or phrases do not destroy the general character of a dialect, and modern phrases or turns of thought make nothing against the traditional ballad. Yet another parallel between the making of ballads and the making of language may be drawn from the reference of Schleicher {Compendium,* p. 641, note) to that hesitation felt by all Aryan languages to put forth stems for the Digitized by LjOOQIC