care to deal great blows,
Lest a bad song be sung of us!”
But even before these lines were recovered from the Bodleian manuscript and brought into wider circulation, the image of the unified band of warriors linked in song was powerful. As we have seen, Paulmy, in one of the earliest modern evocations of the lost Song of Roland, imagined it as the marching song of the Norman soldiers. The image remains strong through the nineteenth century and beyond. It is not just Gautier who sees the Roland as “a sublime lance thrust” In numerous accounts the epic material merges with the warrior class it celebrates in the full embodiment of oral tradition.
In this vision of medieval culture, the orality of the epic is crucial. Gautier’s baron, with his “simple, vigorous, almost brutal” faith and his simple pleasures, untainted by gallantry, cannot read. The epic, as Gautier understands it, matches its audience in the purity of its primitivism:
The age that suits these works is exclusively that of primitive times, when Science and Critical Thinking do not yet exist, and an entire nation naively confuses History and Legend. Some sort of nebulous credulity permeates the atmosphere of the time, and encourages the development of this poetry that has not yet been examined by science or taken over by sophistry. The later centuries of writing are not made for these poetic narratives that circulate invisibly on the lips of a few popular singers.… One does not read these epics, one sings them.57
The grandeur of oral epic is thus part of the long history in which writing marks a fall from some lost state of primal unity. In the jongleur’s song the national and spiritual body of early France is reconstituted.58 The epic is not read, it is sung.
The Manuscript and Its Anglo-Norman Readers
The simplest means of connecting the 4,002 verses preserved in Digby 23 to this putative performance history is to claim that Digby 23 itself once belonged to a minstrel or jongleur.59 This is exactly what Léon Gautier did when he distinguished between the great illuminated manuscripts of the late Middle Ages and the earlier and simpler ones of the twelfth century, which he called “manuscrits de jongleur,” a category in which he included Digby 23. The manuscript was thus classified as a reference tool for a professional performer of some kind.
What is striking about this classification is how widely it has been accepted when even a cursory examination of the codex raises the gravest doubts. The classification of Digby 23 as a “manuscrit de jongleur” went unchallenged until 1932, when Charles Samaran, in his introduction to the facsimile edition, offered the first full codicological description of the manuscript, in the process raising grave doubts about Gautier’s identification. Samaran agreed with Gautier that Digby 23 is a cheap and somewhat worn manuscript, composed of poorly prepared parchment that was carelessly ruled, but he also pointed out that careless compilation, small size, and the wear and tear that suggests widespread circulation are not sufficient grounds on their own for associating a manuscript with a jongleur. There are numerous Latin manuscripts that are equally carelessly executed and equally battered. Samaran noted as well that the Digby copyist shows little familiarity with French epic material, frequently confusing the names of the great heroes, and that the unknown reviser shows even less, which tends to suggest that neither the copyist nor the reviser was a jongleur.60
This leaves the possibility that the book was written by a cleric for the use of a jongleur. The difficulty here is that Digby 23 is in fact a composite volume; the second half is the Roland but the first half is a glossed copy of Calcidius’s translation of the Timaeus. Since the Timaeus scarcely seems likely reading material for a jongleur, it is of some importance to establish when the two sections were brought together. Both date from roughly the same period, sometime in the twelfth century, but earlier commentators believed that the two had circulated separately until they came into the possession of Sir Kenneth Digby in the seventeenth century.61 Samaran, however, noted the presence on the last page of the Roland of what he believed to be the word “Chalcidius” in a thirteenth-century hand on folio [72] r. The same hand also adds several verses from Juvenal’s eighth satire to one of the flyleaves (folio [74]r).62 The implication is that at this point in its history, the Roland was in the hands of someone who could read and write Latin and this person also owned the Timaeus. As Samaran points out, Juvenal’s Satires and the Timaeus are hardly the reading material one would expect of a jongleur. The possibility that at some point the manuscript might have belonged to a jongleur cannot be entirely ruled out, of course. The first century or more of the manuscript’s history is unaccounted for, and it is possible that during this time it passed into the hands of a jongleur, as Maurice Delbouille has suggested.63 But there is nothing about the manuscript to encourage such a speculation. It is a cheap, portable volume; jongleurs were itinerant and were not wealthy—those circumstances, and the prevailing assumption that the poem must be associated with a jongleur are all the basis there has ever been for Gautier’s classification of Digby 23 as a jongleur’s manuscript. In comparison to many of the richly illuminated romances of later centuries, the Digby Roland is plain and humble, but considerable care has been taken to ensure that the manuscript is agreeable to the eye. It is ruled (although not always entirely consistently), and the margins are generous. The initial letter of each line is offset and there is at least one colored initial for each page; corrections are few and neat. Digby 23(2) was not just a rough draft or private copybook. It has been quite heavily used, and the ink is badly faded in places. The first and last folios are dog-eared, and half a quire is discolored at the beginning and the end, suggesting that it lay for a while unbound, but otherwise the Roland is in reasonable condition. It has not been folded, torn, stained, or scribbled in. The classification “minstrel manuscript” cannot be disproved, but it is based not on a consideration, but on an almost willful dismissal, of the codicological evidence.
A closer investigation of the manuscript can provide glimpses of the world in which the poem probably circulated and can suggest some of the ways in which it might have been enjoyed. It will also illustrate the sophistication of at least one baron and the interpenetration of clerical and chivalric culture—all points the traditional understanding of the poem tends to deny or at least to minimize.
The first approach is through the copyist, who writes a hand that has often been criticized for its awkwardness (fig. 2). Earlier readers, such as Gautier, took this awkwardness as a sign of the copy’s low social status and therefore of its association with a jongleur. But it might also be seen as a sign that the copyist was engaged in cultural negotiation, modifying a traditional script to meet new demands. This suggestion might seem highly speculative, but it is advanced by no less an authority than M. B. Parkes, who attributes the awkwardness to the scribe’s attempts to modify a large bookhand, of the kind used for Bibles and Psalters, to the demands of the smaller reading text. Parkes provides other examples of comparable hands in small single-column manuscripts, many of which seem to have emanated from the Norman or Anglo-Norman schools. On this basis, Parkes suggests the scribe may well have been “someone trained in the schools, who found service as chaplain or clerk in a bishop’s familia or a baronial household.”64 The conjunction of “worldly oriented clerics and a sophisticated urbane baronry” that was particularly marked in England and has been offered as one reason for the strength of the Anglo-Norman hagiographic tradition ensured that there were households where the Roland might have found an audience.65
Orderic Vitalis has left us a picture of one such household, that of Hugh of Avranches, earl of Chester and one of William the Conqueror’s chief supporters, who surrounded himself with “swarms of boys of both high and humble birth.”66 At many of these courts it was the custom when the knights and squires were gathered to have selected members read aloud from some suitable and improving book. At Hugh’s court this was the responsibility of the chaplain, Gerold:
To great lords, simple knights, and noble boys alike he gave salutary counsel; and he made a great collection of tales of the combats of holy knights, drawn from the Old Testament and more recent records of Christian achievements, for them to imitate. He told them vivid stories of the conflicts of Demetrius and George, of Theodore and Sebastian, of the