Romances. Romance has long since lost its original signification in every country except Spain, where it is still occasionally used in speaking of the vernacular, as it was in the middle ages when Latin was the language of the lettered classes and of documents and writings of all kinds. But even there its commoner application is, as elsewhere, not to a language, but to a form of composition. In English it has been almost invariably applied to a certain sort of prose fiction, and, in a secondary sense, to the style and tone prevailing therein. By 'the romances,' using the term specifically, we generally mean the prose fictions which, as reading became a more common accomplishment, took the place of the lays and Chansons de geste (q.v.) of the minstrels and trouvères, and were in their turn replaced by the novel. Of these the most important in every way are the so-called romances of chivalry, which may be considered the legitimate descendants of the chansons de geste. The chivalry romances divide naturally into three families or groups: the British (which, perhaps, would be more scientifically described as the American or the Anglo-Norman), the French, and the Spanish; the first having for its centre the legend of Arthur and the Round Table; the second formed round the legend of Charlemagne and the Twelve Peers; and the third consisting mainly of Amadis of Gaul followed by a long series of sequels and imitations of one kind or another. In strict chronological order the Charlemagne cycle should stand first, for the Charlemagne legend was apparently of an earlier formation than the Arthurian; but on the other hand the materials out of which the Arthur legend shaped itself must of course have been the older, and the prose romances which either grew out of it or were grafted upon it are for the most part of an earlier date than those belonging to the Charlemagne story.
The first appearance of Arthur is in the history of Nennius, where he is presented in a quasi-historical shape, simply as the chosen leader of the Britons in twelve successful battles fought with the Saxons; but it is in the Historia Regum Britanniae of Geoffrey of Monmouth (1140) that he first appears as the hero of a connected story. Geoffrey, in fact, may be fairly claimed as the founder of the Arthurian legend. Whatever his materials may have been or whatever the source from which he obtained them, he contrived to give them 'un caractère chevaleresque et courtois,' to use the words of M. Gaston Paris, which was altogether foreign to them when they came to his hands, and thus succeeded in presenting a picture of Arthur and his court which at once proved acceptable to the age in which he lived. It is this character, impressed upon the Arthur legend by Geoffrey, that led Cervantes to regard it as the fountain-head of chivalry and chivalry romance, as he does in Don Quixote (part I, chap. xiii.). The story, however, as Geoffrey left it, is little more than the foundation of the structure raised by his successors a century later. Whether we accept in its entirety or in part only his account of the 'very ancient book' from Brittany which he professed to have translated, or hold that his authorities were simply Nennius, Welsh traditions, and Breton lays and tales, it is clear that his sources of information conveyed no hint of the Round Table or of the Grail, to say nothing of Lancelot and other personages who have come down to us as part and parcel of the Arthurian story. The first reference to the Round Table is in the Brut of Wace (1155), which is in fact an amplified metrical version of Geoffrey's history, and from the words used—'Fist Artus la roonde table, dont Breton dient mainte fable'—we are left to suppose that it was through Breton tradition that it found its way into the story. By some it has been conjectured that in the Round Table we have only an imitation of the Peers of the Charlemagne legend, but in truth the two institutions represented two totally distinct ideas. The peers were simply a fraternity, 'xii. compaignuns,' as the Chanson de Roland calls them, bound together by mutual affection alone, with no ulterior aim or object, and entirely uninfluenced by the sovereign. The Round Table, on the other hand, was a knightly fellowship in which the bond of union was the pursuit of chivalrous adventures and 'deeds of worship,' of which the king was the head, and by which he was 'upborne' and the quiet and rest of his realm insured. The distinction deserves notice, for it is characteristic of the difference between the two legends and the romances that represent them. The Arthurian stories were knightly and courtly, their authors were courtiers, sometimes knights—if we may trust the statements of early editors, they were written to order at the instance of magnates, among whom Henry II. and Henry III. of England are named, and at any rate were obviously addressed to what would now be called the aristocratic section of society. With the Carlovingian it was very different; the chansons de geste from which they were derived were made for and sung to no one class in particular, and it is manifest that the selection for translation into prose was always governed by considerations of popular interest. Hence the phenomenon noticed by more than one observer, that the Arthurian stories have never become in the strict sense of the word popular in any age or country, while the Carlovingian have enjoyed a wide-spread popularity, and in some instances continued to hold their own as popular stories down to the present day. Mr J. A. Symonds observes that in Italy the Arthurian stories, though relished by the cultured classes, never took the fancy of the people at large in the same way as the Carlovingian; and in Spain the romances and ballads that treat of Arthur are few and meagre, while the Charlemagne literature is extensive and rich, and the History of Charlemagne and the Twelve Peers is still a current chap-book in high request. A more obscure question is how the Holy Grail came to be linked to the Arthurian story. There can be no doubt that Celtic tradition and mythology present sufficient analogies to justify a theory that the idea of the Grail is a purely Celtic one which may be traced back to pagan times. But none of these analogues, not Fionn's healing cup or the mystic basin which figures in Percy, can be in any true sense called a Grail. The essence of the Arthurian Grail lies in its character of a Christian relic, and the very name suggests that the conception as it is there presented to us was an Anglo-Norman one. It is very possible, no doubt, that Celtic tradition may have had a share in shaping the conception, but that is all that can be safely said. Some little light, perhaps, is thrown on the question by the curious coincidence between the book presented in a vision in the year 717, which Robert de Borron (circa 1190) sets up as the prime authority for his Sainct Greal, and the vision in the same year in which the Grail itself was seen by a British hermit, as recorded by Helinand in 1204. The return of the first Crusaders stimulated that enthusiasm for relics of the Passion of which we have a proof in the Sacro Catino at Genoa and its rivals. A very natural consequence would be an eagerness to discover the hiding-place of the true catino, and this, when the Round Table idea had been once imported into the Arthurian story, would furnish the 'deed of worship' par excellence necessary to its constitution, while an equally natural consequence would be that the poets in working out the idea would avail themselves of any floating traditions of mystic vessels endowed with miraculous properties which could be pressed into their service. Arthur himself has, no doubt, been treated in the same fashion. Hero-worship is almost always accompanied by annexation. The Charlemagne legend is largely made up of fragments that properly belong to Charles Martel, Pepin, and Charles the Bald. Even in the comparatively modern case of the Cid, one of the most famous exploits, the unseating of the
French ambassador, is in reality the property of the 15th-century Conde de Cifuentes. It would be strange if so remote a figure as Arthur's did not show signs of some such process; but even if we find there, as Professor Rhys holds, traces of the culture hero, or of the solar myth, the question of his personality cannot be said to be thereby affected. It would be almost as unreasonable to treat him as a purely mythical being on such grounds, as to deny Sheridan's existence because jokes attributed to him are to be found in early editions of Joe Miller. There is very little certainty connected with the construction of the Arthurian story. It seems plain that the History of the Grail, which properly should precede the Quest, was in reality a later composition; and the respective shares of Chretien de Troyes and Robert de Borron in the Grail, Perceval, and Lancelot are pretty clearly defined. But in most other respects the Arthurian cycle deserves the title M. Gaston Paris applies to it of 'dédale inextricable.' In no case, as Mr Alfred Nutt says, do we possess a primary form; all the versions that have come down to us presuppose something earlier; all is uncertainty, the order in which the component parts were produced, the sources from which they were derived, the authors to whom they are attributable, the relationships of the various versions and forms one to another; and research seems ever to reveal new nebulae and discover fresh clusters of difficulties. Even on the question as to whether the primary form was in verse, as analogy would lead us to expect, we are for the most part left to conjecture. That Breton popular poetry may have contained the germs of Tristram, Perceval, and Lancelot is no doubt a probability; but of one thing at least we may be certain, that veritable creations like the Lancelot of the Arthur story could have had no place in the simple naïve lais of which we have examples in the translated specimens of Marie de France. The stones may have come from a Celtic quarry, but the building was Anglo-Norman.
It was inevitable that the Arthur stories proper should be followed by romances claiming a supplementary or an introductory character, such as Meliadus, Guiron le Courtois, Artus de Bretagne, and Perceforest, but it would be an injustice to treat these, as Dunlop has done, as though they were legitimate members of the Arthurian cycle, nor have they been admitted into it by the compilers or arrangers who have now and then attempted to present it in a consecutive shape. Facile princeps of these is our own Sir Thomas Malory, whose work is, as Dr Sommer says in his masterly edition, 'by far the best guide to the Arthur romances in their entirety.' Malory's judgment may not be, perhaps, invariably impeccable. He has not always chosen the best or most poetical form, and he has left unculted many beauties of the old MSS. But this may not have been so much his fault as that of the materials with which he had to content himself. Of his general good taste and literary skill there can be as little question as of his English which has made his book one of the classics of his language. Malory, furthermore, as the exhaustive researches of Dr Sommer show, is the sole authority for portions of the series, in particular the story of Gareth in the seventh book. See ARTHUR, GRAIL.
In the romances of the Charlemagne cycle we stand on much firmer ground. It is true that we know even less of the authors than in the case of the Arthur stories, but on the other hand the whole process of production lies plain to view. The starting-point of the legend is undoubtedly the disaster of Roncesvalles, and the Song of Roland—not, of course, the Chanson de Roland that has come down to us, but some older form, the existence and nature of which are matters of inference—may be taken as the foundation of the whole Charlemagne cycle of romance (see ROLAND). Of this, apparently, we have a prose version at the end of the Latin history of Charlemagne, which pretends to be the work of his contemporary the Archbishop Turpin. Nothing was farther from the intention of the writers than to produce a romance; but among the romances, or at the head of them, their work must be placed. About its intention there can be no mistake. By Charlemagne's example it points out the advantages here and hereafter of serving the church liberally and zealously, endowing holy shrines, encouraging pilgrimages, converting the heathen or exterminating them when unconvertible. It records a military pilgrimage to Compostella made by Charles at the call of St James, and is plainly the work of different hands. M. Gaston Paris believes the first five chapters to have been written by a monk of Compostella about 1050; but it is not very obvious why a Spaniard who had his own national legend of Compostella should have gone out of his way to make a patron of a foreigner and an invader. The remainder, he thinks, was written by a monk of Vienne between 1109 and 1119. The book was soon translated into French, and became the chief source of the story of Roland and Roncesvalles, for which it was believed to be the prime authority until the discovery of the chanson proved the existence of a common ancestor. The influence of the latter was mainly through the chansons de geste of which it was in most cases the model. Of these the number is large. M. Leon Gautier's list enumerates above a hundred belonging to the Charlemagne cycle, and this of course only represents survivors. Only a few, however, gave birth to prose romances. The Roland had been forestalled by the Turpin history, and of the others the majority were in interest too local, not sufficiently popular, or for other reasons unsuitable for prose. The story of Ogier le Danois (who possibly had nothing to do with Denmark, but was merely warden of the Ardenne-mark) was too famous to be left in the verse of Adenes le Roy; the traditions of the struggles between the sovereign and his vassals in Aquitaine, not so much in Charlemagne's time as in Pepin's, lent an interest to Renaud de Montauban, the Rinaldo of Italian poetry, but best known as the hero of the Four Sons of Aymon (q.v.), a romance that has probably never been out of print since the introduction of printing; and similar reasons, more or less strong, influenced the selection of Doon de Mayence, Maugist d'Aygremon, Guerin de Montglave, Mille et Amys, Jourdan de Blaves, Galien Rhétoré, and divers others. One of the most notable, independently of its connection with Don Quixote, is Fierabras. In the 15th century it was translated into prose by one Jean Baignon of Lausanne, who prefixed to it the early account of Charlemagne by Vincent de Beauvais, and added the concluding chapters of Turpin with the Roncesvalles story, the whole forming a kind of consecutive Charlemagne romance resembling the Arthur compilations. In this shape, and under the title of La Conqueste du grant roy Charlemaigne des Espaignes, it achieved extraordinary popularity, became a regular chap-book, was translated into Spanish by Nicolas de Piamonte, whose version supplied the balsam of which Don Quixote made trial, and from Spanish into Portuguese about the middle of the 18th century; when it was supplemented by an entirely new Charlemagne romance by the translator, a curious proof of the vitality of the legend.
From the lays of the minstrels of the same period there came also many independent prose romances not necessarily connected with any particular cycle: Valentine and Orson, which, however, is sometimes linked with the Charlemagne cycle; Cleomades, or Clamades, where Cervantes found the magic wooden horse, which by a lapse of memory he assigns to Pierre of Provence and Magalona, another romance of the same kind; Partenopeus of Blois; Melusina; The Knight of the Swan, in some respects the most interesting of all, and curious as an illustration of the growth of a romance. Originally a folklore legend of Brabant, the source of Lohengrin, the story was turned into a poem and incorporated in the series on Godfrey de Bouillon, who was made a descendant of the Knight of the Swan; then it was annexed by Vincent de Beauvais for his Speculum Historiale, from which it passed into the shape of a romance, and was translated into English at the instigation of Edward, Duke of Buckingham, who claimed to be one of the knight's descendants.
Cervantes correctly claims Amadis de Gaula as the founder of Spanish chivalry romance, though he may have been in error as to its being the first work of the kind printed in Spain; the Valencian Tirant lo Blanch must have preceded it. It was long held to be of Portuguese origin on the bare statement of Gomez de Azurara that it was entirely the work of Vasco de Lobeira; but there is ample proof that an Amadis was extant in Spain at least as early as the middle of the 14th century, very probably as early as 1300, but at any rate before Lobeira was born. Southey, in whose time the evidence was not forthcoming, may be excused for asserting the Portuguese origin of the romance; but it is strange to find M. Gaston Paris still describing it as 'portugais puis espagnol aux XVe et XVIe siècles.' Whether this Amadis was in verse or in prose is uncertain; we only know from one witness that it was in three books, and Garci de Montalvo, who is responsible for the existing Amadis, merely claims to have corrected three books, which previous editors and scribes had left in a corrupt state, and to have added a fourth. Nor is it a certainty that it was of purely Spanish origin. The influence of the Arthurian romances is manifest, but what is far more suspicious is the absence of Spanish colour and indications of Spanish parentage; the names are almost all akin to those of the Arthur stories, the faye Urganda is a distinctly Celtic creation, and the scene throughout is laid on Arthurian ground, Wales, England, Brittany, or Normandy, a choice not easily explained in a romancer whose business was to interest Spanish hearers or readers. But whether or not the original may have been some northern French story, it certainly was not, as has been sometimes suggested, Amadas et Ydoine in which there is no more resemblance to Amadis than there is in Aucassin and Nicolette.
The earliest known edition of the Amadis (q.v.) is of 1508, but this cannot be the first; it is too near the date of other romances obviously inspired by it and born of its success, and it is evident that it was finished shortly after the fall of Granada in 1492. The date is significant in its bearing on the curious phenomenon of the sudden outburst of a chivalry romance literature in Spain, just as the middle ages were drawing to an end and other nations were beginning to put away chivalry among the bric-à-brac of bygone days. But in Spain it marked the close of a campaign of seven centuries and the end of a national life of sustained excitement. Under the new order of things, the triple despotism of crown, church, and Inquisition, the nobles and minor nobility were left with a superabundance of leisure on their hands, a condition, as every seaside librarian knows, always favourable to the circulation of fiction, so that Montalvo could not have chosen a better time for his venture. But it would be unjust in the extreme to deny to the merits of the Amadis their share in the creation of Spanish chivalry romance. In almost every respect, story, incidents, characters, and human interest, it will bear comparison with the best of its predecessors, and as a romance of chivalry, pure and simple, it has no equal. In this lay the secret of its success. For Spain chivalry romance had a reality unknown elsewhere. Amadis came to a generation which had seen round Ferdinand and Isabella knights who could match any of Arthur's or Charlemagne's exploits. Coming at such a time it is no wonder that Amadis was followed by a cry for more, and that more was promptly supplied. But Esplandian, Florisando, Lisuarte, Amadis of Greece were of a very different vintage. It was by Feliciano de Silva, the object of Cervantes' special detestation, that the work of continuation was chiefly carried on. He was a clever man, with a facile pen, and if not imagination, at least invention in abundance, but his greatest gift was his intuitive perception of the tastes of his readers. He perceived that it was not so much recreation as excitement they wanted, and that so far from objecting to rant, bombast, and extravagance, the more they got the better they were pleased. He seems to have been the first author who reduced writing nonsense to a system, and also the first who made a handsome fortune by his writings. The professed continuations formed, however, only a small portion of the romances, more or less in imitation of the Amadis, and infected by the style of Feliciano de Silva, the Felixmartes, Belianises, Olivartes, which continued to flow from the press until the long line ended with Poliscine de Boecia, two years before Don Quixote was sent to the press.
With Don Quixote, fittingly, the history of romances as a branch of fiction comes to a close. There are, indeed, two other groups that claim the title, the Pastorals, and those sometimes called the Heroic, an epithet better deserved by the readers who were bold enough to face entertainment in such a formidable shape. But to these quite as much space as their merits entitle them to has been already given (see NOVELS).
See Paulin Paris, Les Romans du Table Ronde (1868-77); Gaston Paris, La Littérature Française au Moyen Âge (2d ed. 1890), Histoire poétique de Charlemagne (1865), De Psuedo Turpino: Hist. Caroli Magni (1865); Oskar Sommer, Morte Darthur (3 vols. 1889); A. Nutt, Studies on the Legend of the Holy Grail (1888); Professor Rhys, The Arthurian Legend (1891); G. Paris and J. Ulrich, Merlin, Roman en Prose d'après le MS. appl. à M. Huth (Société des Anciens Textes Français, 1886); W. F. Skene, The Four Ancient Books of Wales (1868); J. S. Stuart Glennie, Arthurian Localities (1868); Birch-Hirschfeld, Die Sage vom Gral (1877); Herz, Sage vom Parzival und dem Gral (1882); E. Martin, Zur Gral Sage (1880); H. Zimmer (on the Breton sources of the Arthur Legend—Göttingische Gelehrte Anzeigen, Oct. 1890); L. Gautier, Les Épopées Françaises (1878-82); Melzi, Bibliografia dei Romani Italiani (1865); Gayangos, Libros de Caballerías (Bib de Autores Españoles, vol. xi.); Mily a Fontanals, Poesia heroico-popular Castellana (1874); Turpini Historia Caroli Magni, Texte Revue par F. Castets (1880); Ward, Catal. of Romances in the Dept. of MSS., British Museum (1883); Quaritch, Catal. of Romances of Chivalry (1882); Early English Text Society's publications; Romania; many papers by Gaston Paris; the section on literature in the article SPAIN: George Saintsbury, The Flourishing of Romance (1897); W. P. Ker, Epic and Romance (1897).