Epic and Romance: Essays on Medieval Literature

by W. P. Ker

Roland is a succession of separate scenes, with no gradation or transition between them. It still bears traces of the lyrical origins of epic. But the narrative, though broken, is neither stinted nor laboured; it does not, like Beowulf, give the impression that it has been expanded beyond the convenient limits, and that the author is scant of breath. And none of the later chansons de geste are so restricted and reserved in their design as Roland; most of them are diffuse and long. The French and the Teutonic epics are at opposite extremes of style.

The French epics are addressed to the largest conceivable audience.[69] They are plain and simple, as different as possible from the allusive brevity of the Northern poems. Even the plainest of the old English poems, even Maldon, has to employ the poetical diction, the unprosaic terms and figures of the Teutonic School. The alliterative poetry down to its last days has a vocabulary different from that of prose, and much richer. The French epic language is not distinguished and made difficult in this way; it is "not prismatic but diaphanous." Those who could understand anything could understand it, and the chansons de geste easily found currency in the market-place, when they were driven by the new romances from their old place of honour in "bower and hall." The Teutonic poetry, even at its simplest, must have required more attention in its hearers than the French, through the strangeness and the greater variety of its vocabulary. It is less familiar, less popular. Whatever dignity may be acquired by the French epic is not due to any special or elaborate convention of phrase. Where it is weak, its poverty is not disguised, as in the weaker portions of Teutonic poetry, by the ornaments and synonyms of the Gradus. The commonplaces of French epic are not imposing.[70] With this difference between the French and the Teutonic conventions, there is all the more interest in a comparison of the two kinds, where they come into comparison through any resemblance of their subjects or their thought, as in Byrhtnoth and Roland.

The French epics have generally a larger political field, more numerous armies, and more magnificent kings, than the Teutonic. In the same degree, their heroism is different from that of the earlier heroic age. The general motives of patriotism and religion, France and Christendom, prevent the free use of the simpler and older motives of individual heroism. The hero of the older sort is still there, but his game is hindered by the larger and more complex political conditions of France; or if these are evaded, still the mere size of the country and numbers of the fighting-men tell against his importance; he is dwarfed by his surroundings. The limitation of the scenes in the poems of Beowulf, Ermanaric, and Attila throws out the figures in strong relief. The mere extent of the stage and the number of the supernumeraries required for the action of most of the French stories appear to have told against the definiteness of their characters; as, on the other hand, the personages in Beowulf, without much individual character of their own, seem to gain in precision and strength from the smallness of the scene in which they act. There is less strict economy in the chansons de geste.

Apart from this, there is real and essential vagueness in their characters; their drama is rudimentary. The simplicity of the French epic style, which is addressed to a large audience and easily intelligible, is not capable of much dramatic subtlety. It can be made to express a variety of actions and a variety of moods, but these are generally rendered by means of common formulas, without much dramatic insight or intention. While the fragments of Teutonic epic seem to give evidence of a growing dramatic imagination, and the Northern poems, especially, of a series of experiments in character, the French epic imagination appears to have remained content with its established and abstract formulas for different modes of sentiment and passion. It would not be easy to find anything in French epic that gives the same impression of discovery and innovation, of the search for dramatic form, of the absorption of the poet's mind in the pursuit of an imaginary character, as is given, again and again, by the Northern poems of the Volsung cycle. Yet the chansons de geste are often true and effective in their outlines of character, and include a quantity of "humours and observation," though their authors seem to have been unable to give solidity to their sketches.

The weakness of the drama in the French epics, even more than their compliance with foreign romance in the choice of incidents or machinery, is against their claim to be reckoned in the higher order of heroic narrative. They are romantic by the comparative levity of their imagination; the story, with them, is too much for the personages. But it is still the problem of heroic character that engages them, however feebly or conventionally they may deal with it. They rely, like the Teutonic epic and the Sagas, on situations that test the force of character, and they find those situations in the common conditions of an heroic age, subject of course to the modifications of the comparatively late period and late form of society to which they belong. Roland is a variation on the one perpetual heroic theme; it has a grander setting, a grander accompaniment, than Byrhtnoth or Waldere, but it is essentially the old story of the heroic age,—no knight-errantry, but the last resistance of a man driven into a corner.

The greatness of the poem of Roland is that of an author who knows his own mind, who has a certain mood of the heroic imagination to express, and is at no loss for his instrument or for the lines of his work.

The poem, as has been already noted, has a general likeness in its plan to the story of Finnesburh as told in Beowulf, and to the poems of the death of Attila. The plot falls into two parts, the second part being the vengeance and expiation.

Although the story is thus not absolutely simple, like the adventures of Beowulf, no epic has a more magnificent simplicity of effect. The other personages, Charlemagne, Ganelon, Oliver, King Marsile, have to Roland nothing like the importance of Agamemnon, Ajax, Diomede, or Hector, as compared with Achilles in the Iliad. The poem is almost wholly devoted to the praise and glorification of a single hero; it retains very much of the old manners of the earlier stages of epic poetry, before it ceased to be lyric. It is a poem in honour of a chieftain.

At the same time, this lyrical tone in Roland and this pathetic concentration of the interest on one personage do not interfere with the epic plan of the narrative, or disturb the lines of the composition. The central part of the poem is on the Homeric scale; the fighting, the separate combats, are rendered in an Homeric way. Byrhtnoth and Roland are the works that have given the best medieval counterpart to the battles of Homer. There is more of a crisis and a climax in Roland than in the several battles of the Iliad, and a different sort of climax from that of Byrhtnoth. Everything leads to the agony and heroic death of Roland, and to his glory as the unyielding champion of France and Christendom. It is not as in the Iliad, where different heroes have their day, or as at Maldon, where the fall of the captain leads to the more desperate defence and the more exalted heroism of his companions. Roland is the absolute master of the Song of Roland. No other heroic poetry conveys the same effect of pre-eminent simplicity and grandeur. There is hardly anything in the poem except the single mood; its simplicity is overpowering, a type of heroic resistance for all the later poets of Europe. This impressive effect is aided, it is true, by an infusion of the lyrical tone and by playing on the pathetic emotions. Roland is ideal and universal, and the story of his defeat, of the blast of his horn, and the last stroke of Durendal, is a kind of funeral march or "heroic symphony" into which a meaning may be read for every new hero, to the end of the world; for any one in any age whose Mood is the more as the Might lessens. Yet although Roland has this universal or symbolical or musical meaning—unlike the more individual personages in the Sagas, who would resent being made into allegories—the total effect is mainly due to legitimate epic means. There is no stinting of the epic proportions or suppression of the epic devices. The Song of Roland is narrative poetry, a model of narrative design, with the proper epic spaces well proportioned, well considered, and filled with action. It may be contrasted with the Death-Song of Ragnar Lodbrok, which is an attempt to get the same sort of moral effect by a process of lyrical distillation from heroic poetry; putting all the strongest heroic motives into the most intense and emphatic form. There is something lyrical in Roland, but the poem is not governed by lyrical principles; it requires the deliberation and the freedom of epic; it must have room to move in before it can come up to the height of its argument. The abruptness of its periods is not really an interruption of its even flight; it is an abruptness of detail, like a broken sea with a larger wave moving under it; it does not impair or disguise the grandeur of the movement as a whole.