reproach dating from Gautier’s time and allowing empirical work on the late poems to
proceed without reference to an idealized concept of French epic.
The works involved are more numerous than is usually stated; Kibler counts twenty-
nine chansons dating from the 14th and 15th centuries, including the reworkings of older
subject matter (remaniements). The addition of epics in Franco-Italian dialect would
bring the number up to well over thirty late epic poems that we know to have been in
circulation from St. Louis’s time to about the end of the Hundred Years’ War. This
means that 20–25 percent of all medieval French texts in chanson de geste form are
“late.”
The adventures in the chansons d’aventures are stereotyped—kidnapings, treachery,
sea voyages, sieges—and usually cyclical: a character is separated from family or
beloved, travels (often incognito), and undergoes trials before the inevitable reunion. In
turn, the chansons are woven into further cyclical sequences in which the adventures of
heroes and heroines are given in succession. Several of them indeed represent prefaces or
sequels to earlier works. Along with the expansion of the Guillaume d’Orange Cycle,
largely through the creation of a sort of subcycle of adventures for the Monglane family,
and the rewriting and loosely conceived filling-in of the Crusade Cycle, there is
development of the Nantueil Cycle in the long Tristan de Nantueil (mid-14th c.; 23,361
lines) and a small cycle centered on Huon de Bordeaux. Poems with no cyclical
attachments also exist.
The late epics, as usually conceived, are immense, with a mean length of about 19,300
lines. The longest is the Chevalier au cygne et Godefroid de Bouillon at 35,180 lines;
Lion de Bourges has 34,298, the revised Ogier le Danois about 31,000, and Baudouin de
Sebourc nearly 26,000. The shortest text considered complete is Aubéron, at 2,456 lines;
Florence de Rome has 4,562 and Hugues Capet has 6,381; but some texts (Ciperis de
Vignevaux, Bâtard de Bouillon) may be incomplete, and the subtexts of the Monglane
group are nearer the 2,000–3,000-line length common in earlier works. The huge
Chevalier au cygne is in fact a 14th-century reworking of the poems of the Crusade
Cycle. In such a work, as in Tristan de Nantueil or the continuations of Huon de
Bordeaux, episode and branch are not easily distinguishable, which reminds us that the
late works were not necessarily intended or received as wholes.
This in turn raises the questions of the way these chansons were meant to be presented
and the form in which they were known to their audiences. Their texts include clear signs
of oral transmission in a traditional sense, and although it is sometimes claimed that these
must represent only archaizing window dressing, nothing indicates that oral performance
of literature died out abruptly in France with the coming of courtly romance. It is likely
that the late epics were both recited (perhaps read aloud) and read silently by individuals.
Their versification probably responds to traditional concern for a listening audience and
cannot be dismissed as merely facile any more than the earlier formulaic texts are merely
cliché-ridden. The audience itself has not been identified, but some of the narratorial
asides are addressed to a socially mixed group.
Robert Francis Cook
[See also: BAUDOUIN DE SEBOURC; CHANSON DE GESTE; CRUSADE
CYCLE; DAGOBERT, PSEUDO-CYCLE OF; HUON DE BORDEAUX; NANTEUIL
CYCLE; TRISTAN DE NANTEUIL]
Boogaard, Nico H.J.van den. “Le caractère oral de la chanson de geste tardive.” In Langue et
littérature françaises: études réunies par R.E.V. Stuip. Assen: Van Gorcum, 1978, pp. 25–38.
Medieval france: an encyclopedia 994