in a history of French literature, but it concerns us. We shall indeed probably find that the home-grown or home-fed Chanson de Geste did least for the novel in the wide sense—that the "Matter of Rome" chiefly gave it variety, change of atmosphere to some extent, and an invaluable connection with older literatures, but that the central division or "Matter of Britain," with the immense fringes of miscellaneous romans d'aventures—which are sometimes more or less directly connected with it, and are always moulded more or less on its patterns—gave most of all.
The Chanson de Geste.
Of these, however, what has been called the family or patriotic part was undoubtedly the earliest and for a long time the most influential. There is, fortunately, not the least need here to fight out the old battle of the cantilenae or supposed ballad-originals. I see no reason to alter the doubt with which I have always regarded their existence; but it really does not matter, to us, whether they existed or not, especially since we have not got them now. What we have got is a vast mass of narrative poetry, which latterly took actual prose form, and which—as early certainly as the eleventh century and perhaps earlier—turns the French faculty for narrative (whether it was actually or entirely fictitious narrative or not does not again matter) into channels of a very promising kind.
The novel-reader who has his wits and his memory about him may perhaps say, "Promising perhaps; but paying?" The answer must be that the promise may have taken some time to be fully liquidated, but that the immediate or short-dated payment was great. The fault of the Chansons de Geste—a fault which in some degree is to be found in French literature as a whole, and to a greater extent in all mediaeval literature—is that the class and the type are rather too prominent. The central conception of Charlemagne as a generally dignified but too frequently irascible and rather petulant monarch, surrounded by valiant and in a way faithful but exceedingly touchy or ticklish paladins, is no doubt true enough to the early stages of feudalism—in fact, to adapt the tag, there is too much human nature in it for it to be false. But it communicates a certain sameness to the chansons which stick closest to the model.
The proportions of history and fiction in them.
The exact relation of the Chansons de Geste to the subsequent history of French fiction is thus an extremely important one, and one that requires, not only a good deal of reading on which to base any opinion that shall not be worthless, but a considerable exercise of critical discretion in order to form that opinion competently. The present writer can at least plead no small acquaintance with the subject, and a full if possibly over-generous acknowledgment of his dealings with it on the part of some French authorities, living and dead, of the highest competence. But the attractions of the vast and strangely long ignored body of chanson literature are curiously various in kind, and they cannot be indiscriminately drawn upon as evidence of an early mastery of tale-telling proper on the part of the French as a nation.
There is indeed one solid fact, the importance of which can hardly be exaggerated in some ways, though it may be wrongly estimated in others. Here is not merely the largest part proportionately, but a very large bulk positively, of the very earliest part of a literature, devoted to a kind of narrative which, though some of it may be historic originally, is pretty certainly worked up into its concrete and extant state by fiction. The comparison with the two literatures which on the whole bear such comparison with French best—English and Greek—is here very striking. People say that there "must have been" many Beowulfs: it can hardly be said that we have so much as a positive assertion of the existence of even one other, though we have allusions and glances which have been amplified in the usual fashion. We have positive and not reasonably doubtful assertion of the existence of a very large body of more or less early Greek epic; but we have nothing existing except the Iliad and the Odyssey.
The part played by language, prosody, and manners.
On this fact, be it repeated, if we observe the canons of sound criticism in the process, too much stress in general cannot be laid. There must have been some more than ordinary nisus towards story-telling in a people and a language which produced, and for three or four centuries cherished, something like a hundred legends, sometimes of great length, on the single general[14] subject of the exploits, sufferings, and what not of the great half-historical, half-legendary emperor à la barbe florie, of his son, and of the more legendary than historical peers, rebels, subjects, descendants, and "those about both" generally. And though the assertion requires a little more justification and allowance, there must have been some extraordinary gifts for more or less fictitious composition when such a vast body of spirited fictitious, or even half-fictitious, narrative is turned out.
But in this justification as to the last part of the contention a good deal of care has to be observed. It will not necessarily follow, because the metal is attractive, that its attractiveness is always of the kind purely belonging to fiction; and, as a matter of fact, a large part of it is not. Much is due to the singular sonority and splendour of the language, which is much more like Spanish than modern French, and which only a few poets of exceptional power have been able to reproduce in modern French itself. Much more is imparted by the equally peculiar character of the metre—the long tirades or laisses, assonanced or mono-rhymed paragraphs in decasyllables or alexandrines, which, to those who have once caught their harmony, have an indescribable and unparalleled charm. Yet further, these attractions come from the strange unfamiliar world of life and character described and displayed; from the brilliant stock epithets and phrases that stud the style as if with a stiff but glittering embroidery; and from other sources too many to mention here.
Some drawbacks.
Yet one must draw attention to the fact that all the named sources of the attraction, and may perhaps ask the reader to take it on trust that most of the unnamed, are not essentially or exclusively attractions of fiction—that they are attractions of poetry. And, on the other hand, while the weaving of so vast a web of actual fiction remains "to credit," there are not a few things to be set on the other side of the account. The sameness of the chanson story, the almost invariable recurrence of the stock motives and frameworks—of rebellion, treason, paynim invasion, petulance of a King's son, somewhat too "coming" affection of a King's daughter, tyrannical and Lear-like impotentia of the King himself, etc.—may be exaggerated, but cannot be denied. In the greatest of all by general acknowledgment, the far-famed Roland, the economy of pure story interest is pushed to a point which in a less unsophisticated age—say the twentieth instead of the twelfth or eleventh century—might be put down to deliberate theory or crotchet. The very incidents, stirring as they are, are put as it were in skeleton argument or summary rather than amplified into full story-flesh and blood; we see such heroine as there is only to see her die; even the great moment of the horn is given as if it had been "censored" by somebody. People, I believe, have called this brevity Homeric; but that is not how I read Homer.
In fact, so jealous are some of those who well and wisely love the chansons, that I have known objections taken to ranking as pure examples, despite their undoubted age and merit, such pieces as Amis et Amiles (for passion and pathos and that just averted tragedy which is so difficult to manage, one of the finest of all) and the Voyage à Constantinoble, the single early specimen of mainly or purely comic donnée.[15] This seems to me, I confess, mere prudery or else mistaken logic, starting from the quite unjustifiable proposition that nothing that is not found in the Chanson de Roland ought to be found in any chanson. But we may admit that the "bones"—the simplest terms of the chanson-formula—hardly include varied interests, though they allow such interests to be clothed upon and added to them.
But a fair balance of actual story merit.
Despite this admission, however, and despite the further one that it is to the "romances" proper—Arthurian, classical, and adventurous—rather than to the chansons that one must look for the first satisfactory examples of such clothing and addition, it is not to be denied that the chansons themselves provide a great deal of it—whether because of adulteration with strictly "romance" matter is a question