¹ There appears no reason to doubt the priority of the Chanson de Roland to the Fierebras and Otuel as claimed by Comfort, P. M. L. A. XXI, 345 ff.

² I have in mind only what I have heard in conversation.

³ Into the well known story of the earlier meanings of the term romance it is perhaps hardly necessary to enter. Nevertheless the story is one of such inherent interest and one that helps so far to explain the present meaning of the term that it will not be amiss to review it briefly. Originally romance bore a meaning quite distinct from that which it bears today. It signified something pertaining to Rome and was applied (1) to a language derived from popular Latin and (2) to a literary composition written in such a language. A good example of this secondary meaning of the term is afforded by the title Roman de la Rose, where the term roman has no reference whatsoever to the “romantic” character of the contents of the poem, which is, furthermore, only in part romantic anyway. It bore the title of course, because of the language in which it was written, the term romance (O. F. romanz, etc.) being used, as it is still today, to distinguish the vernacular from the Latin or learned language. In England, however, romance was used to distinguish the Anglo-French from the native language or literature (cf. Voelker, P., Zs. fuer rom. Philogie X, 489). But from the mere accident that those who used romance languages, particularly the French, chanced to excel in the composition of tales of an incredible character, the term came at length to be applied to a tale of this sort in whatever language written. Having once come to clear itself of all association with language, the term could be extended backward to include whatever of a fanciful character had been written before the time of the French by the Hindoos, Arabs, Persians, Greeks, or Romans, as well as forward to embrace those productions of a like character which, largely as a result of French influence came to be written by the Celts, Teutons, Slavs, and other peoples of late medieval and modern Europe. Thus the term now signifies exhibitions of fancy wherever found—whether standing alone in romances so-called, as in the Panchatantra, the Arabian Nights' Entertainment, the Milesian Tales, the Dolopathus, the Baalam and Josophat the Greek fables regarding the Aethiopians and Hyperborians, the Decameron of Boccaccio, the Heptameron of Margaret of Navarre, the Chaucerian tale of Cambuscan and the wondrous horse of brass, and the story of Oberon as recounted in Huon de Bordeaux or by Wieland, or incidentally incorporated in epic or drama, as in the Homeric episodes of Ulysses's adventures with the Circe, the Cyclops, and the lotus-eaters, in the fabulous episodes in the Beowulf or Niebelungenlied, and in the Shakespearean episodes of “the three caskets” and “the pound of flesh,” of Oberon and Titania, and of Othello's recital of “moving accidents by flood and field,” and of “men whose heads do grow beneath their shoulders.”

⁴ Since the book from which Geoffrey claims to have extracted his History of the Kings of Britain was an “old British book” and since Geoffrey himself appears to have been a Briton, the objection may be raised that we have romance produced by those to whom the story of the British kings was native and that such a circumstance violates our conception of romance as the story of one folk refashioned by another. But the remoulding of a story of one folk by another, though a usual, is by no means an invariable prerequisite to the production of romance. The essential characteristic of romance is that the story be incredible. But a story may become incredible through changes subsequently wrought in the character of the people who first produced it. The record of the British kings found in the book from which Geoffrey purloined the materials of his history must have been epic to the people who produced and read it because there can be no doubt that they believed in it. To Geoffrey, however, who even if a Briton, was a highly educated scholar and wrote in Latin with an Anglo-French audience in mind, this same chronicle of events could not have seemed credible and must therefore, even though in no wise altered, have been romance.

⁵ Not because she is mourning the loss of her husband, for whose decease it would have been ludicrously out of character for her to have expressed regret, but because his own mistress Maria d'Aquino, whom, as he tells us in his Proemio, Criseida represents, had, when he first beheld her at the church of St. Lorenzo at Naples, been so attired because it was Holy Saturday, as previously explained in the Filocolo, ed. Moutier, Firenze, 1827-34 VII, 4ff.

⁶ On the Date of Chaucer's Troilus, etc., Chaucer Soc., Second Ser., No. 42.

⁷ Les Légendes Épiques, 4 vols., Paris, 1910.