conditions.
Now there is no doubt that medieval minstrels performed poems about the deeds or gestes of epic heroes, including Roland, and that these performances, were often sung.162 What is troubling is that the numerous references to minstrel performance at secular festivities such as chivalric feasts never suggest that this singing lasted very long.163 A case in point is the famous description of the wedding feast of the count of Archimbaud and a Flemish princess, Flamenca, in the Provençal romance that bears her name. After the second feast of the day is over and the tables have been cleared, the guests are brought fans and cushions. Now is the time for the fifteen hundred jongleurs to perform. They offer a full range of minstrelsy—singing, playing on almost every conceivable instrument, telling stories to musical accompaniment, tumbling, and juggling. They offer a wide choice of stories:
Qui volc ausir diverses comtes
de reis, de marques e de comtes,
auzir ne poc tan can si volc. (lines 617–19)
He who would hear diverse accounts of marquesses and kings and counts may hear as many as he wants.
It is possible that the jongleurs are conceived as reciting well-known passages from famous works, but equally possible that the songs they are singing would have borne only the most tangential connection to the famous poems about the same heroes that have survived. In either case, their performances do not appear to have been of any duration. Soon the king calls for the guests to join him in jousting and then dancing. The account in Flamenca is filled with names famous in French literature, including those of the heroes of Chrétien de Troyes, and Yvain is referred to by the title Chrétien gives him, the Knight of the Lion. Another work mentioned, the “lais de Cabrefoil,” might be Marie de France’s Chevrefoil.164 Yet under these conditions, the jongleurs could not have expected to deliver even one of the longer Old French lais, let alone an Arthurian romance, or a large section from a chanson de geste. The time was too short and the status of any single performer too lowly. At such a feast, no single jongleur could expect to command the attention of any but a small group of guests and even that not for long. Nor are the circumstances depicted in Flamenca such as to permit the jongleur to work through a longer piece over a series of performances. The jongleurs and their audience were assembled to add dignity to a single ritual occasion; once the feast was completed, they dispersed. To the extent that Flamenca offers a credible depiction of the performance conditions at a royal feast, the implication is that minstrels rarely had the opportunity to perform lengthy pieces to their conclusion and hence that the minstrel versions of lais, chansons de geste, or romances differed radically from those that have been preserved in manuscript.
The example of Flamenca is telling because it does seem to reflect established chivalric ceremonial, albeit filtered through literary conventions of plenitude that exaggerate the numbers of minstrels. Accounts of historical feasts, when they survive, show that the number of minstrels was considered a reflection of the dignity of the occasion. The pay record for the feast held for the knighting of the English Prince Edward in 1306 shows 119 minstrels; of these, more than 80 are explicitly classified as musicians.165 Few other feasts are as well documented, but large numbers of minstrels are recorded in many cases: 426 minstrels were paid for performing at the marriage of Princess Margaret to John of Brabant in 1290, for example.166 Nor did the more exalted minstrels who attended the great feasts necessarily stay in the area much longer than those in Flamenca. The minstrels who helped celebrate Edward’s knighting on Whitsunday 1302 had to be up before dawn the next morning in order to collect their payment before prime.
In fact, a major feast was perhaps the least suited of all these occasions for extended recitation. The audience was often disruptive, and there was too much competition; the more prestigious the occasion, the more competition there would be.167 Humbler situations might have been easier for the performer. At an isolated castle, monastery, or country house, or at a village gathering, a jongleur would have stood a better chance of holding the sustained attention of the entire audience, as he does in Gautier’s vision. Even here he would have been obliged to modify his performance to suit the occasion and retain interest. To do so, he would have needed to deliver short fragments and to modify them appropriately.
This, indeed, is a recurring accusation against jongleurs—that they distort the truth, singing whatever they think will please their listeners and offering biased and partial accounts. Peter the Chanter compares priests who change the form of the mass to increase the offerings they receive to jongleurs who switch stories to keep their audience’s interest: “Priests of this kind are like jongleurs or storytellers (fabulatori) who, when they see that the song of Landri does not please their audience, immediately begin to sing of Antioch. And if the audience is too demanding and is still not pleased by hearing of Alexander they switch to a song about Appollonius or to one about Charlemagne or about someone else.”168 Chrétien de Troyes, similarly, in the introduction to Eric et Enide accuses “those who live by telling stories” of corrupting the story:
d’Erec, li fil Lac, est li contes,
que devant rois et devant contes
depecier et corronpre suelent
cil qui de conter vivre vuelent. (lines 19–22)169
The tale is about Eric, the son of Lac [and it is one] that those who live by telling stories are accustomed to break apart and corrupt in front of kings and nobles.
The term “depecier” could be taken to mean no more than that the conteurs are accustomed to mangling a good story, but taken more literally, it accuses them of breaking the story into pieces, which is exactly what performers must often have been obliged to do as they anticipated and played upon their audience’s response. The story of Taillefer playing as he rode to his death at Hastings, however legendary it may be, provides a more plausible instance of what an actual performance might have been like than does Gautier’s vision of the dignified séance épique. Under these conditions, a jongleur might hope to sing fifty, or a hundred, or at best a few hundred lines, with little prospect of picking up where he left off. In short, the evidence suggests that sustained recitation of the kind that would have been needed to complete a poem the length of the Digby Roland was not common. The implication, and it is an alarming one, is that when minstrels recited chansons de geste these oral performances were significantly different from the written versions that have been passed down to us.
This is not to claim that the notion of sustained recitation or the séance épique is merely the product of the romantic imagination. The main source of evidence is the written chansons de geste themselves, which present themselves in a number of ways as oral performance. In Huon de Bordeaux, one of the most discussed examples, the narrator tells his listeners that it is growing dark and he is getting thirsty and will stop for the night, but that if they pay him well enough he will return on the morrow.170 Apparently they do return, but not with quite enough money, for the narrator resumes the tale but after only five hundred lines stops to make his pitch again, this time threatening to use the power of Oberon, the fairy king, to excommunicate those who do not help fill his purse when his wife brings it around.171 Here would seem to be a persuasive example of sequential performance.172 But the example is not without difficulties. It is clearly impossible for a poet to predict in advance just when a jongleur might find it appropriate to appeal to his audience, so a jongleur would be singularly ill advised to attempt to deliver Huon de Bordeaux verbatim. If we are to read the appeals as strictly functional, therefore, the only plausible explanation seems to be that the lines were copied directly from a specific live performance in which a particular jongleur improvised them.173 There is little evidence, however, to suggest that this kind of ethnographic reporting was at all common in the Middle Ages.174 It was not that the technology was lacking: although full systems of shorthand were not developed until the sixteenth century,
