SOURCE: Brault, Gerard J. Introduction to “The Song of Roland”: An Analytical Edition, Vol. 1: Introduction and Commentary, pp. 1-116. University Park, PA: The Pennsylvania State University Press, 1978.

[In the following excerpt, Brault discusses the historical, political, and religious background to The Song of Roland.]

1. THE HISTORICAL EVENT

The Song of Roland is an epic poem that recounts the events surrounding the death of Charlemagne's nephew Roland at Roncevaux in the Pyrenees. The Emperor and his men were journeying home after a military campaign in Spain. The disaster actually took place in the year 778, some three centuries before the poem is generally dated.¹

In 732, twenty-one years after landing on the Spanish Peninsula, the Saracens were decisively stopped at Poitiers by Charles Martel (688?-741), Charlemagne's grandfather.² Throughout this Muslim advance, however, Christians in the Asturias, the northwestern corner of Spain, had succeeded in resisting the general onslaught. By the ninth century the Christians had broadened their dominions to encompass a number of adjacent provinces, including Galicia and most of León to the south and east.

Christian reconquest of the northern tier of Spain was greatly facilitated by internal dissention among the Arabs.³ At the invitation of Suleiman, the Arab governor of Barcelona—and recognizing an opportunity to establish a buffer against the Saracen threat from south of the Pyrenees, and to make converts to Christianity—Charlemagne amassed an army and entered Spain in two main columns in 778. Alerted to a Saxon uprising to the north and forced to lift the siege of Saragossa, Charlemagne took a number of hostages including Suleiman himself, whom he suspected of treachery, and fell back on Pamplona, destroying it and forcing its inhabitants, including many Christians, to flee. As the Franks were making their way back across the Pyrenees, Gascons ambushed the rearguard, killing all its defenders. Loaded down with booty, they escaped into the night. Charlemagne was unable for the moment to avenge this stunning defeat, but he returned several years later and established a zone of Frankish influence in the northern tier of the peninsula known as the Spanish March.

This is the most faithful account of the events of the year 778, which has been pieced together by historians from scattered and often contradictory sources. The best-known narrative of these events is contained in the Vita Karoli Magni by Einhard, Charlemagne's biographer (d. 840):

Cum enim absiduo ac pene continuo cum Saxonibus bello certaretur, dispositis per congrua confiniorum loco praesidiis, Hispaniam quam maximo poterat belli apparatu adgreditur; saltuque Pyrinei superato, omnibus quae adierat oppidis atque castellis in deditionem acceptis, salvo et incolomi exercitu revertitur, praeter quod in ipso Pyrinei jugo Wasconicam perfidiam parumper in redeundo contigit experiri. Nam cum agmine longo, ut loci et angustiarum situs permittebat, porrectus iret exercitus, Wascones in summi montis vertice positis insidiis—est enim locus ex opacitate silvarum, quarum ibi maxima est copia, insidiis ponendis oportunus—extremam inpedimentorum partem et eos qui, novissimi agminis incedentes subsidio, praecedentes tuebantur desuper incursantes in subjectam vallem deiciunt consertoque cum eis proelio usque ad unum omnes interficiunt ac, direptis inpedimentis, noctis beneficio quae jam instabat protecti, summa cum celeritate in diversa disperguntur. Adjuvabat in hoc facto Wascones et levitas armorum et loci in quo res gerebatur situs; econtra Francos et armorum gravitas et loci iniquitas per omnia Wasconibus reddidit inpares. In quo proelio Eggihardus regiae mensae praepositus, Anshelmus comes palatii et Hruodlandus Brittannici limitis praefectus cum aliis conpluribus interficiuntur. Neque hoc factum ad praesens vindicari poterat, quia hostis, re perpetrata, ita dispersus est ut ne fama quidem remaneret ubinam gentium quaeri potuisset.⁴

(While the war against the Saxons is being fought energetically and almost continuously, he [Charlemagne], having stationed troops at strategic places along the borders, attacks Spain with all the forces that he can muster. He crosses the Pyrenees, accepts the surrender of all the towns and fortified places that he encounters along the way, and returns without his army having sustained any losses except that, during the withdrawal, while traversing the Pyrenees, he happened to experience Gascon treachery. While his army was marching in a long column, because of a narrow pass, some Gascons lying in ambush at the top of the mountain—for the thick woods which are very plentiful in that area afford a great opportunity for sneak attacks—swoop down on the last elements of the baggage train and on the rearguard protecting the main body of the army. They drive them back into the valley, join battle, and massacre every last one of them. Then, having looted the baggage train, they disperse very rapidly in every direction under cover of night which was falling. On this occasion, the Gascons had the advantage of light armament and control of the terrain; the Franks were greatly hindered by their heavy armament and lower position. In this battle were slain Eggihard, the royal seneschal; Anselm, count of the palace; and Roland, prefect of the Breton march, and many others. This reverse could not be avenged immediately because the enemy, having done this deed, dispersed in such a way that no one could even tell in which direction they might have been sought.)

In this account no reason for Charlemagne's invasion of Spain is provided, and the ambush is said to have taken place after he had conquered all the towns and castles in his path on the peninsula.⁵ According to Einhard, the defeat was occasioned by Gascon treachery (Wasconicam perfidiam).⁶ Duke Lupus of Gascony had earlier submitted to Charlemagne's authority but, breaking his oath, allowed a marauding band of his followers to lead a successful attack against the rearguard.⁷ Charlemagne's campaign was motivated by political and religious considerations.⁸ The Annales Mettenses priores, written shortly after 805, indicate that Charlemagne responded to appeals from the oppressed Christian community in Spain, an assertion confirmed by a letter from Pope Hadrian dated May 778 and reiterated by the Astronomer of Limoges about 840.⁹ It has been suggested that Saracens as well as Gascons participated in the ambush.¹⁰

What is most significant in Einhard's narrative is the amount of space devoted to the disaster, which not only suggests the impact of the defeat on the people of the day, but lists a number of illustrious victims, including Hruodlandus Brittannici limitis praefectus, the hero of the Song of Roland,¹¹ and mentions treachery, which will later motivate the action in the poem.¹²

2. LEGEND, POEM, AND TEXT

The myth of Charlemagne (742-814) began to grow during his lifetime, and in the two centuries after his death, numerous anecdotes about him had been circulated and his name intimately associated with the Empire.¹³ There are scattered references to the incident in the Pyrenees in medieval annals, and it is safe to assume that by the eleventh century a fairly elaborate legend had developed, perhaps giving rise to more than one work in Latin or French, none of which have survived.¹⁴ The custom of naming brothers Roland and Oliver, which dates from the beginning of the eleventh century, is an important witness to the development of the legend.¹⁵

There has been considerable speculation concerning the presumed ancestry of chansons de geste in general and of the Song of Roland in particular. If the prototype of this epic was Latin, it doubtless circulated in monastic circles in written form, like the saints' lives, which it probably resembled in form and content. On the other hand, a vernacular version, the product of oral composition, may have been part of the repertory of jongleurs, or singers of tales, who traveled about the country at the time.¹⁶ In any event, many scholars are convinced that the so-called Baligant episode—roughly a thousand lines, or about one-fourth of the poem in its present state—was not part of the original material.¹⁷

Whether inspired by legend or poem, a man of genius, referred to henceforth as Turoldus¹⁸—the name found in the Oxford manuscript—living in France about 1100,¹⁹ composed the work known today as the Song of Roland. The questions as to the precise manner and form in which Turoldus received his material and how much of the poem's style is due to his creativity and skill are much debated. However, the extant French works composed before 1100 are but pale forerunners of this remarkable epic, and the oral literature, the existence of which one can assume and even reconstruct to a certain extent, was doubtless far less complex and sophisticated.

Turoldus's masterpiece is superior to the kind of verse that singers of tales usually composed, but it has many features in common with this oral literature. Jean Rychner links the repetitions in the initial laisses of the Roland to the poet's need to have his audience clearly understand Marsile's proposal.²⁰ On the other hand, he finds only one instance of recapitulation in Turoldus's poem (vv. 2769-2787), a technique which the singer of tales often utilized to refresh his listeners' memory when a new session was beginning or to allow recent arrivals in the audience to catch up on what had already been narrated.²¹

Turoldus was obviously familiar with the formulaic diction used by the jongleurs, and he skillfully fused this procedure with the techniques of written literature taught in the schools. Other authors of his day did the same, if somewhat varying the blend, although the chansons de geste that have survived generally appear to be more the work of clerks than of jongleurs.²²

The lines between chanson de geste, chronicle, and saint's life are not always easy to discern in Old French literature. Traditional definitions remain valid for the most part; but how does one classify a poem like Ami et Amile?²³ The confrontation of, say, the Song of Roland, the Pseudo-Turpin Chronicle, and the Rolandslied (Introduction, 10), with their varying mixtures of the sacred and the profane, of edifying, historical, and legendary material, tends to negate the concept of genre. The medieval author's habit of citing well-known but often—as in the Song of Roland—nonexistent annals or chronicles for authentification complicates the matter even further.²⁴

If one is to appreciate the condition of the surviving chansons de geste, it is also important to understand the habits of medieval scribes. Copyists at times reproduced the manuscript before them quite faithfully but at other times altered their transcript, with changes ranging from occasional word substitutions to substantial abridgments and lengthy amplifications or interpolations. Transposition into a different dialect, modernization of language to reflect current usage, and, from the thirteenth century onward, switching from assonance to rhyme and from verse to prose, were common practices. Many years and even centuries sometimes elapsed between the time the original work was composed and a particular manuscript was copied, and there was often more than one intervening transcript. Turoldus's poem was subjected to all of these transformations.

Scholars have established the relationship between the extant manuscripts of the Roland²⁵ and, in general, agree that the copy preserved in the Bodleian Library at Oxford offers the oldest and the best surviving version of the poem. This manuscript, often referred to by its library reference number, Digby 23, is in a twelfth-century hand.²⁶ In Anglo-Norman dialect, it is believed to reflect rather closely Turoldus's original poem, which was composed sometime earlier.²⁷

Efforts to narrow down the date of the poem itself and of the Bodleian copy continue. There are many proposals for dating the work before or after 1100, but no convincing argument has been advanced for dissociating the poem in its present form from the other early French epics it resembles and which were composed about this time: Chanson de Guillaume, Gormont et Isembart, Pèlerinage de Charlemagne.

The copy of the Roland in Digby 23 has too many scribal errors to be an original,²⁸ and a second hand (“the revisor”) made some sixty-odd changes in the text.²⁹ The manuscript seems too small and inelegant in execution to have been a presentation copy for some important personage.³⁰ Was it, then, as some have proposed, meant to be used by a jongleur?³¹ After reviewing all the available evidence in this regard, Charles Samaran was unable to resolve the matter³² and later efforts have not removed this uncertainty.³³

A good deal is known about the circumstances under which a chanson de geste was recited,³⁴ but relatively little effort has been made to visualize how such a performance appeared to a medieval audience. New light on the oral interpretation of the Song of Roland is provided below (Introduction, 20).

To sum up, some jongleurs were also trouvères, or authors, but their compositions probably have not survived. The epics that the jongleurs performed as they traveled about the country may have been of two different types: (1) the popular form, now lost, which was heavily dependent upon the formulas, motifs, themes, and techniques about which Rychner has written, and (2) the more literate, polished, and sophisticated form, which has much in common with the popular form but which bears the hallmark of the clerical tradition. However, there is no incontrovertible proof that the Roland or any other chanson de geste was ever sung before an audience in the form preserved in medieval manuscripts.³⁵ There are limits, therefore, to any claim that can be made about the medieval audience's presumed reaction to any epic passage.

One can only speculate, too, about the manner in which jongleurs transmitted chansons de geste to the clerks who copied and doubtless altered them. Turoldus may have combined the skills of jongleur and clerk, but this does not seem likely.

One thing is certain, however: The Song of Roland is the greatest French epic and its appearance was an event of the highest importance in the history of Western literature. …

4. TUROLDUS AND HIS WORLD

In the late eleventh century Western Europe was very different from what it is today.³⁶ Its diverse peoples were just beginning to emerge from a long era marked by armed invasions and raids from far and near. They had abandoned once-flourishing cities established by the Romans³⁷ and had retreated from the countryside into isolated, self-sufficient, fortified enclaves. Royal authority was not very strong. Villages had developed in the shadow of castles and a small class of warriors—perhaps one-tenth of one percent of the population—ruled over serfs clustered around them for protection.³⁸ By now the hierarchical structure of lord and vassal had assumed its definitive form and individuals at all levels of society were bound together by vows of homage and fealty.³⁹ Whenever the term chevaler is used in the Song of Roland it refers to a fighting man, distinguished by birth and by ownership of expensive heavy armor, weapons, and a war-horse.⁴⁰ Armies of Turoldus's day also included a large number of attendants (serjanz) who assisted the warrior, but usually in a noncombatant role.⁴¹

Commerce had disappeared for the most part and virtually everyone was engaged in the serious business of subsistence agriculture. The local lord offered protection in return for ownership of all the land, a percentage of the crop and domestic animals, and various services. The baron and his knights spent a good part of their time in martial activities, warring against marauding aggressors or among themselves. Whether in a castle or in a village hut, life was rude and precarious by modern standards, alternating between times of starvation and relative plenty, between grim survival and joyous feasting. Nature and manmade violence were important factors conditioning the daily routine and seasonal cycle of work and rest. Age-old religious and folk beliefs, customs, and traditions also influenced the lives of these profoundly conservative people.

Two major activities could seriously affect an individual's life. A military expedition or a pilgrimage to a distant land might remove a person for years at a time from familiar surroundings.⁴² Powerful dukes and counts often became involved in foreign campaigns (there were numerous expeditions into Italy and Spain, for instance), and permanent resettlement at times ensued, such as that of certain Norman barons and their retinues in Sicily. The most popular pilgrimages were of course to local shrines, but they might also involve a trip to Rome, Santiago de Compostela, or even the Holy Land.⁴³ Crude itineraries—maps or narrative accounts—provided some assistance to pilgrims who were, by custom, accorded hospitality in monasteries and hostels along the way.⁴⁴ Travelers followed ancient routes, but it is misleading to think in terms of specific “pilgrimage roads,” a fiction invented by turn-of-the-century scholars.⁴⁵ Pilgrims normally traveled in groups, a custom immortalized in Chaucer's Canterbury Tales. The services of a guide (OFr. guieor, guior), that is, a person who had presumably traveled the road at least once before, helped pilgrims establish a pace and avoid obvious hazards. Whether embarking on a military adventure abroad or on a distant pilgrimage, individuals knew that travel entailed grave risks. Modern historians estimate that, in the Middle Ages, life expectancy was thirty-five years until the thirteenth century.⁴⁶

Patriotism as we conceive it today was nonexistent in Turoldus's day.⁴⁷ National and ethnic stereotypes, extending down to the regional level, were commonplace, as evidenced perhaps in v. 3796 (Icels d'Alverne i sunt li plus curteis). Foreigners and even inhabitants of neighboring lands were commonly viewed with distrust. The poet's use of the recurring expression dulce France may be in imitation of classical models.⁴⁸ On the other hand, affection and nostalgia for one's homeland and for familiar surroundings are only natural. For Turoldus, to be a Franc de France meant one enjoyed special status in Charlemagne's army.

Local political matters at times altered daily living in the eleventh century. A marriage contract negotiated between two noble families (the bride usually had little say in this matter) might bring about a change of administration and improve or worsen the serf's lot dramatically, depending on his new lord's humaneness or greed.⁴⁹ The eruption of a feud between barons could spell disaster for everyone concerned.

The sense of belonging to Christianity imparted a much stronger feeling of solidarity than any corresponding national identification, but religion then as now affected individuals' outlook on life and ethical behavior in a variety of ways. The notion of Christian unity and of a certain internationalism⁵⁰ was enhanced during the course of the eleventh century by a growing concern over the Saracen peril, a situation precipitated by the appearance of the Turks upon the scene, which sent shock waves throughout Europe and inaugurated a new phase in history.⁵¹ The many disruptions occasioned by the Turkish conquests, among many other factors, culminated in the world-shaking proclamation of the First Crusade at Clermont on 27 November 1095.⁵² The Roland's international appeal is evident from the fact that it was translated into several languages.

5. THE CHURCH AND THE ARTS

The impact of the Church on the Middle Ages is incalculable, for its influence permeated and shaped every activity known to man. In the eleventh century the Catholic hierarchy was associated with every facet of life, imparting counsel and ministering to spiritual needs in the entourage of sovereigns and lords of every degree, as well as among the lowest-ranking serfs. Regular clergy, living under vows in monasteries and various religious establishments, or as recluses, outnumbered knights by perhaps five to one; secular priests and clerks in minor orders (Introduction, 7) doubtless increased that proportion to ten to one.⁵³

A theoretical distinction had long ago⁵⁴ been drawn between the powers of Church and State, and the question of investiture was a bone of contention.⁵⁵ However, in practice the Church usually succeeded in maintaining its primacy, symbolized notably by the incident at Canossa. In January 1077, Henry IV, King of Germany and the Holy Roman Emperor, stood for three days bareheaded and barefoot in the snow in a castle courtyard until Pope Gregory VII, who was staying there, consented to grant him an audience and absolve him from excommunication. Less than a century later, in 1177, Henry II of England, having quarreled with Thomas Becket on the question of ecclesiastical authority—a dispute resulting in the latter's murder—was forced to perform a humiliating public penance at the Archbishop's shrine.

Abbots and bishops shared in the benefits of the feudal system, holding estates in fief and exercising the right to collect tolls and market dues.⁵⁶ On the other hand, sovereigns and even lesser lords commonly held churches in benefice and derived considerable income from their tithes and other revenues, a practice the Church succeeded in limiting in Turoldus's day.⁵⁷

Monks maintained schools and, although the age of the great universities was yet to come, institutions such as those at Chartres, Cluny, Laon, Orleans, Reims, Saint-Benoît-sur-Loire, and Saint-Riquier, to name but the major French establishments,⁵⁸ were already regarded as important centers of learning. Lesser schools could be found in areas boasting a few thousand inhabitants or a thriving monastery.⁵⁹ Most of the artistic activity of the period can be traced to the monastic orders, which supervised the construction and decoration of abbeys, churches, and related edifices, copied and illuminated Bibles, hymnals, pericopes, and other manuscripts, composed music and simple liturgical dramas,⁶⁰ and in general promoted esthetic values.⁶¹

Monks also produced a wide variety of didactic, exegetical, and theological writings in Latin. Every monastery corresponded regularly with its mother house, reporting on its activities and particular needs. A talent for composing homilies and hymns was highly regarded. However, even greater interest was shown in the writing and reworking of Latin chronicles and hagiographic literature. When they were in a position to do so, monasteries maintained annals and recorded important occurrences in their community life and other events deemed worthy of note. Other sources were consulted when more ambitious narratives were undertaken, the result being so-called universal chronicles. The scope and value of these compilations vary a good deal, some being a mere list of dates and occurrences, while others present biased or fragmentary accounts. To establish a simple historical fact—for instance, the incident in the Pyrenees in 778—one must often sift a bewildering mass of hearsay and contradictory evidence.

W. J. Brandt has classified medieval chronicles according to the particular mode of perception that characterizes them.⁶² He finds that chronicles present either an aristocratic or a clerical view of the world and that these perceptions of reality are essentially antagonistic: “The important discontinuity in the value system of the medieval cleric was within the profession itself, between absolutely unreconcilable views of life which were equally cherished and equally believed.”⁶³ The aristocratic ideal, he maintains, was one of worldly values and stances, great admiration being shown for profit and materialistic concerns, whereas the clerical tradition constantly judged people and events according to atemporal norms, and held innocence, purity, and the like in high esteem. The former usually adopted a narrative mode of expression, while the latter depended a good deal more on rhetorical models.⁶⁴

Without perforce disagreeing with this typology, which, like most generalizations, tends to lose some of its luster upon close examination, Paul Archambault has shown that the early chroniclers differ rather significantly from their later counterparts, and that one can also distinguish between “mirror” and “window” chroniclers, the former merely recounting events, the latter endeavoring to grasp their deep significance.⁶⁵ Archambault's categorization should be borne in mind in discussions of the meaning of Turoldus's poem.

Clerical activity in the eleventh century was not restricted to chronicling events that impressed contemporaries as being interesting or significant. The production of saints' lives was another major preoccupation. The earliest narratives of this type provide relatively few details and are documentary in style.⁶⁶ However, there arose in the Middle Ages a group of writers—known today as hagiographers—who developed a vast corpus of fiction loosely termed saints' lives. With a few exceptions, the typical Latin passio was drawn up according to a well-defined plan, but one that betrayed an astonishing lack of imagination:

Il ne faudrait pourtant point exagérer la fécondité des “trouveurs” hagiographes. Un classement méthodique des thèmes exploités par eux amène à constater que les répétitions du même trait merveilleux sont fréquentes, et que c'est surtout grâce à diverses combinaisons de lieux communs qu'il règne, dans certains groupes de légendes hagiographiques, un semblant de variété. Ce qu'il faut surtout se garder de croire, c'est qu'au point de vue de l'esthétique même, le niveau des créations merveilleuses de l'hagiographie populaire soit, en général, bien élevé. A côté de quelques trouvailles réellement heureuses et de certains motifs ingénieux et intéressants, que de banalités s'y rencontrent, que d'inventions bizarres et souvent extravagantes!

Le cadre de la narration est nettement dessiné. D'abord, une description plus ou moins détaillée de la persécution. Les chrétiens sont partout recherchés; un grand nombre tombe aux mains des soldats, et parmi eux le héros du récit; il est arrêté et jeté en prison. Mené devant le juge, il confesse sa foi et endure d'affreux supplices. Il meurt, et son tombeau devient le théâtre d'une foule de prodiges.⁶⁷

Certain hagiographic themes and narrative techniques appear in the Song of Roland. Saints' lives were read, but others—in the vernacular and in metrical form—were sung by jongleurs, constituting an important contact with the chansons de geste.⁶⁸ Turoldus's poem is neither chronicle nor saint's life; it is an epic. Each, however, was composed by clerks and it was inevitable that the first two genres would influence the other.

One cannot discuss the Church and the arts in the eleventh century without mentioning the Cluniac movement.⁶⁹ Monasticism, which dates to the dawn of Christianity, was given a decisive impetus with the foundation of the Abbey of Monte Cassino and the institution of the rule of Saint Benedict in a.d. 529. Abbeys began to proliferate throughout Europe, and, in addition to preserving for centuries the classical learning that managed to survive the Fall of the Roman Empire, they steadfastly nurtured the beliefs and exercised the moral authority that generally characterize our concept of the Church in the Middle Ages. In the year 910 the Benedictine abbey of Cluny was founded, inaugurating an unparalleled era of construction and of intellectual and spiritual vitality. About the time the Song of Roland was composed, Cluny controlled no fewer than 1,450 different religious establishments housing 10,000 monks throughout Europe.⁷⁰ Cluniac monasteries and hostels dotted the countryside and were familiar landmarks encountered by pilgrims and travelers everywhere.⁷¹ “You are the light of the world,” wrote Pope Urban II to the monks at Cluny in 1098.⁷²

Cluny's impact on monasticism was characterized by a spirit of reform that emphasized asceticism and war on secularism. This new seriousness was reflected in the contemporary strictures of monks and councils against the jongleurs.⁷³ Recriminations against scandalous personal conduct and licentious performances were part of the condemnation of public amusements and spectacles that originated with the Fathers of the Church. There is evidence, however, that some religious looked with favor upon the singing of saints' lives and epics by the jongleurs, even authorizing performances of the former in church and of the latter in the cloister.⁷⁴ This practice and the strongly clerical adaptations known as the Pseudo-Turpin Chronicle and the Rolandslied (Introduction, 10) lend considerable support to the view that the Roland and other early chansons de geste were considered to have important didactic content.

In the eleventh century the Church strove mightily to inspire Christians to cease their constant warring against one another and upon the unarmed and the helpless. It sought energetically to establish law and order by means of various associations and oaths, notably by instituting the Peace of God, designed to protect certain classes of persons and objects—such as churches, clerics, livestock, and vineyards—and the Truce of God, prohibiting violence on certain days.⁷⁵ Although enforced by spiritual penalties and eventually by temporal punishments, these movements were only intermittently successful. “The human race,” wrote the contemporary chronicler Raoul Glaber, “was like a dog that returns to its vomit. The promise was made, but it was not fulfilled.”⁷⁶ The Crusades were to afford the military class with an unprecedented outlet for their warlike energies. Historians and moralists continue to debate whether any true ethical progress was achieved during this era.

The Church transformed what was primitively a pagan dubbing ritual into the knighting ceremony, which assumed its definitive form by the eleventh century and included the blessing of the sword and other Christian observances.⁷⁷ The religious also fostered the evolution of what was originally simple class-consciousness into the code of chivalry, whose Christian coloring reached its highest expression in later times, but a number of whose essential elements can already be found in the Song of Roland.⁷⁸ The moral position of the Church had an important bearing upon such concepts as avarice, pride, right and wrong, and spoils in Turoldus's poem.

6. POPULAR TRADITION

Dancing, singing, and storytelling, like ornamental and ritual painting and sculpture, are part of the folklore of all nations, and this tradition is reflected in the Song of Roland. Demonology and belief in magic, the marvelous power of the oliphant, the naming of swords and horses, and teratology are among the most obvious forms of popular influence in Turoldus's epic. Councils, cursing, dreams, and oath-taking are frequent motifs in the tales told by people all over the world. Bruce Rosenberg believes that the image of the betrayed hero dying on a hill as the last survivor is an age-old cliché whose manifestations can be found in the Bible, the Roland, and in the myth that developed following Custer's Last Stand.⁷⁹

The authors of the chansons de geste and Old French romances were fond of inserting brief anecdotes in their narratives.⁸⁰ These miniature stories—as a rule no more than a few verses long—either concern a notable event in the character's life or the origin of an object to which he is strongly attached. At first glance they seem rather trifling, and for the most part they have been neglected by scholars or regarded as mere digressions or make-rhymes. The following examples drawn from the Song of Roland illustrate the genre:

Marsile's mules:	Que li tramist li reis de Suatilie.
Margariz's sword:	Si la tramist li amiralz de Primes.
Siglorel:	L'encanteür ki ja fut en enfer,
	Par artimal l'i cundoist Jupiter.
Valdabron:	Jerusalem prist ja par traïsun,
	Si violat le temple Salomon,
	Le patriarche ocist devant les funz.
Turpin's war-horse:	Siet el cheval qu'il tolit a Grossaille,
	Ço ert uns reis qu'il ocist en Denemarche.
Abisme's shield:	En Val Metas li dunat uns diables,
	Si li tramist li amiralz Galafes.
Durendal:	Carles esteit es vals de Moriane,
	Quant Deus del cel li mandat par sun agle
	Qu'il te dunast a un cunte cataignie:
	Dunc la me ceinst li gentilz reis, li magnes.
Charlemagne's horse:	Il le cunquist es guez desuz Marsune,
	Sin getat mort Malpalin de Nerbone.

As is evident from the preceding illustrations, anecdotes of this type utilize formulaic diction and spin fantastic yarns, often involving the supernatural. Proper names play an important and affective role in these formulas, and the objects and animals have all been acquired either as a gift or as a battle trophy. Some of these stories within a story seem plausible enough, but they are all fictions designed to place events in an appropriate epic context. Thus the medieval audience grasped full well that Charlemagne's expedition in Spain in 778 and the Muslim peril of the eleventh century were historically authentic, but they felt quite differently about the anecdotes concerning the Emperor's horse or a Saracen's shield. Nevertheless, a skillful author knew how to go about elaborating a story involving heroic exploits and extraordinary events, on the one hand, and historical facts, on the other, so as to blur the line that separates illusion and reality.

Like literary portraits in medieval literature, with which they may usefully be compared, these thumbnail sketches, which can be classified as marginal anecdotes, possibly derive from classical models,⁸¹ but it is better to study them according to their own esthetic. Alice M. Colby has defined the literary portrait in terms that also apply to the anecdote. It is, she says in part, a description that

may give many different kinds of information about an individual, which affects the listener's interpretation of the work by provoking in him an emotional reaction to the important character being described, and which stands out from the context as a semi-independent, stylistically ornate, well-organized, and completely panegyrical or censorious descriptive unit, much of the content of which is stereotyped. We have shown that this definition is partially supported by material available in the treatises on Latin rhetoric and poetic art which were widely read in the twelfth century or which, in all probability, represent twelfth-century practice, but we have also demonstrated the wisdom of tentatively basing this definition entirely on the empirical evidence to be found in the vernacular literature before seeking support from the Latin theorists. Although most of the Old French poets had in all likelihood studied formal Latin composition, it cannot automatically be assumed that, when writing in the vernacular, they applied all the rules they had learned or made no attempt to develop independent vernacular traditions.⁸²

These important principles influenced Turoldus's technique of character portrayal. Marginal anecdotes must be distinguished from what can be referred to as kernel anecdotes. The latter—for instance, the stories which Ganelon and Charlemagne tell about Roland—are of vital importance to the proper interpretation of the poem and, in certain instances, provided the nucleus around which elaborate episodes and even whole epics were later constructed.

One must not confuse rhetoric of the type originally developed by Greek and Latin authors (Introduction, 7) with forms that appeared in medieval French literature independently of this tradition, albeit often along parallel lines. The latter is rhétorique coutumière, a mode of expression that is no less formal and systematic for all its indigenous characteristics than rhétorique scolaire.⁸³ When Turoldus describes Turpin's horse, he is following quite closely a model cited by Isidore of Seville.⁸⁴ On the other hand, the brief portraits that the poet gives of certain individuals in the Song of Roland, the elaborate parallelism, and the laisses similaires technique⁸⁵ belong to a rhetorical tradition clearly distinct from that found in the works of Latin grammarians and treatise writers.

Finally, synonymic repetition deserves special mention here.⁸⁶ Classical authors—Cicero, notably—joined related expressions for the purpose of adding nuances to their thought. Old French writers paired words that were semantically identical in order to reinforce a single concept. No shading of meaning was intended. As Jones has shown, association is the key to understanding many elusive terms in the Song of Roland. The known form elucidates the doubtful expression to which it is joined.⁸⁷ On the other hand, one must also have a firm grasp of the ethical possibilities of each member before attempting to arrive at a conclusion as to the meaning intended by the author. Thus Jones is correct in pointing out that proz e vaillanz in v. 3186—the epithets concern Oliver—“is practically tautological,”⁸⁸ but he errs when he asserts that the usefulness implied in the adjective proz (< Late Latin prodis < prode est < Classical Latin prodest, a form of prodesse ‘to be useful’) alludes exclusively to the courage and physical strength of the fighting man;⁸⁹ for proz refers to all the virtues of the ideal knight, including wisdom. …

7. LEARNED TRADITION

One cannot always distinguish between popular and learned matter in the Song of Roland. For instance, knowledge of the Bible, demonology, and proverbs⁹⁰ can be both. Many, or perhaps most, of the elements cited above as popular could just as easily be relegated to the category of learned matter. It would be a gross oversimplification, too, to characterize popular influence as coarse, learned as elegant, when the opposite is often the case.

What is a clerk?⁹¹ The Old French word represents Ecclesiastical Latin clericus and is ultimately derived from clerus, the Latin term for clergy. Clerk is used in a variety of ways in the Middle Ages. It can mean a person who has been to school, a learned man, or simply a man of the church, a clergyman, although there were of course many degrees of association with the Church. In theory it is possible to distinguish between the cleric, whose duties were essentially religious, and the clerk, whose functions were purely secretarial or bureaucratic, but in practice, the clerk often served in both capacities.⁹² It is not until the fifteenth century, in fact, that the word clerk becomes specialized in its present-day meaning; that is, a person who has minor administrative duties in an office.⁹³

The situation was otherwise, however, in the eleventh century, and if Turoldus was a clerk, it may be assumed that he was in one of the several ranks of the ministry, that is, either in major or in minor orders. Being a clerk implies that Turoldus attended school and was exposed to classical learning, especially rhetoric, and received religious training involving familiarity with the Bible and other works of edification and spirituality. Clerks were often attached to royalty or to the households of noblemen or high-ranking ecclesiastics. Literary patronage at this time did not necessarily entail the explicit commissioning of a poem or other work as the clerk's sole obligation in return for payment of one kind or another. Quite frequently it resembled the sort of sponsorship or support a modern university provides professors—whose chief duty it is to teach—for creative writing or research in the humanities.⁹⁴

How did Turoldus's learning shape his poem? The most important influence was the Bible. This knowledge has been abundantly detailed by Busigny, Jenkins, Dickmann, Faral, and others.⁹⁵ To cite but the most obvious borrowings, there are certain proper names—Canelius, Dathan and Habirun,⁹⁶ for example—the sun-stopping episode, a reminiscence of the Book of Joshua; the earthquake and storm announcing Roland's death, which are modeled on the Passion of Our Lord; Ganelon as a Judas figure; the role of the angels; and a number of biblical words and expressions.⁹⁷

Fortitudo et sapientia has been accorded far too much significance for the interpretation of the Roland (Introduction, 3). This is not to say, however, that the theme does not play a role in the poem. This and the ubi sunt motif, as Charles searches for his nephew's body, are also part of the epic's religious inspiration.⁹⁸ A certain amount of material derives from the saints' lives, in particular the manner of depicting the hero's death as martyrdom in imitation of Christ, and the concept of death as a victory.⁹⁹ Some of this can be classified as the sort of knowledge with which almost anyone living in the Middle Ages would have been familiar. However, these allusions taken as a whole, together with the archetypes, stereotypes, and other clerical modes of perception discussed below, point to advanced schooling of the traditional sort.

A second type of clerical inspiration results from familiarity with classical authors. Several scholars—Tavernier, Jenkins, and Curtius, in particular—have studied the many situations, techniques, and themes in the Roland that may be said to be imitated from the classics.¹⁰⁰ One may mention, for instance, death wishes, foreboding, irony, olive branches as symbols of peace, omens, laments, nature responding to strong human emotion, nostalgia for a “sweet” homeland, and understatement. One of the most characteristic devices in the Song of Roland is that of foreshadowing, a device traditionally associated with classical writers.

No specific allusion to a classical author has been ascribed to Turoldus, but many scholars strongly sense that he was exposed to certain masterpieces of Latin literature, the Aeneid in particular.¹⁰¹ Suetonius was also widely read in the Middle Ages and influenced Einhard's Vita Karoli Magni, which may in turn have been familiar to our poet.¹⁰² There is a possible allusion in the Gautier de l'Hum episode to Suetonius's account of Augustus's cry “Vare, redde legiones.” The passage is missing in Digby 23 but may have been found in the original poem composed by Turoldus.¹⁰³ Finally, one of the major themes of the Song of Roland, the Struggle, may derive from Prudentius, whose Psychomachia exerted a lasting influence on medieval art and literature.¹⁰⁴

Virtually all the classical borrowings in Turoldus's epic were skillfully adapted to his purpose and are not immediately evident.¹⁰⁵ The poet appears to have consciously tempered his display of learning and to have artfully masked his sources.¹⁰⁶ Not infrequently, the modern scholar accurately perceives the nature of the influence but, being unable to put his finger on the exact source, concludes, as does Maurice Delbouille in his discussion of Ganelon's anecdote about Roland and the apple, that “cela sent l'école.”¹⁰⁷ Curtius, Martín de Riquer, and others have listed the rhetorical figures that abound in the Roland: amplificatio, antithesis, apostrophe, iteratio, recapitulation, repetition, and so on, characteristics never found in such profusion in popular literature.¹⁰⁸

The principal reason Latin literature was read and classical rhetoric was studied in the schools was to provide a proper foundation for interpreting the Bible and to prepare the clerk for his religious duties and for such mundane tasks as keeping records or carrying on correspondence. A by-product, doubtless considered incidental by the men of the Middle Ages but of surpassing interest to us today, was the fact that this schooling influenced clerks when they composed literature in the vernacular.

In the final analysis the difference between Turoldus and other authors of chansons de geste is not so much a question of the degree of his familiarity with the techniques of oral composition or of his mastery of clerical skills: It is rather a matter of his superior talent, that is, that spark of genius and devotion to his task which enabled him to use his knowledge and training to great effect.

8. REALISM, IDEALISM, AND THE EPIC UNIVERSE

Turoldus did not deliberately set out to transpose Charlemagne's campaign of 778 into contemporary terms or to make propaganda for the First Crusade or for any other similar endeavor. He simply wished to tell of an event that happened some three hundred years earlier. That circumstance, however, had been seriously altered by the passage of time and by the process of myth-building. The story would be further transformed now through the workings of his poetic imagination.

The chansons de geste were sung before aristocratic audiences who enjoyed being entertained in their castles after the evening meal, in their gardens, or while on long horseback journeys.¹⁰⁹ The jongleurs numbered among their many talents juggling, playing various musical instruments, singing, and tumbling, and the performance of chansons de geste was doubtless viewed in that light, that is, as entertainment. Noblemen enjoyed visualizing themselves accomplishing great deeds, striking mighty blows, and achieving the high renown associated with epic heroes.¹¹⁰

Jules Horrent points out that aristocratic audiences were sophisticated enough to appreciate esthetic effects and to recognize biblical and literary allusions.¹¹¹ This would have been even truer in the case of the clerks who happened to be present on such occasions. The moral and spiritual value of the chansons de geste was also acknowledged, for, as we have seen, authorization to have epics sung in the cloister is known to have been granted.¹¹² Finally, poems such as the Song of Roland were also heard in public squares.¹¹³ Thus while the major figures in Turoldus's poem are aristocrats involved in actions familiar, for the most part, only to members of the privileged class, the epic itself no doubt elicited an enthusiastic response from all levels of medieval society.

The Song of Roland is replete with aristocratic situations: Sovereigns and knights deal with problems in characteristic fashion, assuming stances designed to enhance their reputations and relentlessly pursuing their own advantage, profit, and revenge. There is decided realism here as well as in the description of contemporary armor, equitation, feudal relationships, military organization, and the like. However, constant exaggeration and fantasy offer important counterpoint to this reflection of contemporary manners and mores, projecting everything away from the here and now into the universe of the chansons de geste.¹¹⁴ There are a few conscious archaisms,¹¹⁵ but epic distantiation is Turoldus's main technique for situating his characters in a realm where realism plays no part.

Of even greater significance for the interpretation of the poem is the fact that Turoldus idealizes and spiritualizes these very same activities. Thus, for example, boasting and refusing help in the face of danger are familiar aristocratic stances, but in the Roland they become part of the Theme of Victory associated with the hero, a process which in the present instance is religious in essence.

The metamorphosis of worldly into spiritual achievements and attitudes, and the symbolic process by which a lion becomes the devil, and a tree the Cross, constitute the least understood phenomenon in the Song of Roland. Viewing events and utterances in isolation or with clinical detachment, as scholars studying a particular aspect of this epic frequently do, often strips them of their true significance. It is only by situating these activities in the context of the entire work and by allowing oneself to be swept along by the poem's evocative power that one can penetrate Turoldus's overall design.

The myth-making process itself also militates against realism. Given that Roland must die, Christians are in the right, pagans in the wrong, Charles is Emperor, treachery must be punished, and so on, Turoldus is constrained to make his story line conform to these inflexible parameters. On the other hand, he obviously feels free to elaborate the legend, to create new characters and situations, and to make legendary personages behave according to his poetic vision. Thus the majestic character Charlemagne, Defender of the Faith and Ruler over the Christian World, was drawn from tradition, but Turoldus imagined him as an Abraham-Job figure, knowingly yet unhesitatingly sacrificing his nephew and steadfast in his great travail. In the eleventh century royal power was at one of its low points, yet the poet instilled great authority in the Emperor, depicted him as a priest-king, and showed him in almost daily communication with the Deity.

Whether or not consciously, Turoldus utilized archetypes and topoi that are familiar to the literary critic. He makes frequent use of irony and symbols, revealing a constant search on his part for what lies beyond literalness. Individuals are often limned with a striking gesture or are involved in dramatic confrontation, thus manifesting the poet's predilection for visual impact.

9. AMBIGUITY AND LOGIC

Poetry often involves a peculiar ordering of reality and always entails exceptional use of language. In order to achieve certain effects, the author suggests meanings, employing, notably, the metaphorical mode. Allusive speech is inherently ambiguous, yet it conveys emotion and impressions with particular force and appeal. That medieval writers were acutely aware of the relevance of these matters is borne out by Saint Augustine in a famous passage: “No one doubts that things are perceived more readily through similitudes and that what is sought with difficulty is discovered with more pleasure.”¹¹⁶

In order to reconstruct the interplay of concepts and associations in the Song of Roland, one must at times put aside twentieth-century notions of what constitutes logic or common sense (this is notably true in the two oliphant scenes), for what appears bizarre or unorthodox today may not have seemed at all peculiar to the poet's contemporaries.¹¹⁷ Finally, it must not be assumed that all the workings of Turoldus's imagination were typical for his day and age.

10. SOME EXEGETICAL GUIDEPOSTS

There are indications which suggest that Turoldus was influenced by certain contemporary views regarding literary interpretation.

In medieval Christian thought the chief events of the Old Testament were held to be mere prefigurations or archetypes of those in the New Testament.¹¹⁸ This tradition dates back to the time of the Apostles. However, these typological patterns often appear forced today. For instance, Samson carrying the gates of Gaza upon his shoulders was said to prefigure Christ carrying his Cross;¹¹⁹ or associating Eve with the Blessed Virgin seemed perfectly natural.¹²⁰ Medieval madonnas, such as the one dated about 1100, preserved at Essen …, frequently represent Mary as the New Eve, holding the Infant Jesus on her lap and an apple in her right hand.¹²¹ Through Eve's fault Man was lost, through Mary's intercession he was saved.¹²²

Medieval exegetes combed the Scriptures for adumbrations of the Gospel narratives, eventually concluding that all things in Creation were, in a very real sense, metaphors of the Christian experience. By the eleventh century an impressive list of such identifications had been compiled and had found its way into glosses, homilies, and a variety of devotional and didactic literature. So far as interpretation of the Bible is concerned, many exegetes distinguished four levels of meaning: the literal, the allegorical, the tropological (or moral), and the anagogical. The latter view was crystallized in a celebrated rhyming couplet:

Littera gesta docet, quod credas allegoria,
Moralis quid agas, quo tendas anagogia.(123)

The classic illustration was that Jerusalem, in the literal sense, is a city of Palestine; in the allegorical sense, the Church; in the tropological sense, the Christian soul; and in the anagogical sense, Heaven.¹²⁴

A slight modification of this system is to be found in what Hugh of Saint Victor describes as the standard approach to all literature, whether sacred or profane, in the schools of his day. This approach involved exposing the littera, or grammatical points, the sensus, or surface meaning, then the sententia, or doctrinal content, of each text under consideration.¹²⁵ A widely attested formulation provides yet another distillation of the process and probably offers the most useful formula for approaching medieval texts in this fashion. Sacred texts were held to consist of a kernel and an outer shell. According to Honorius Augustodunensis: “Nux est Sacra Scriptura, cujus cortex est littera, nucleus vero spiritualis intelligentia.”¹²⁶

It would of course be ideal if one were always able to distinguish purely literary devices and meanings from deeper spiritual sense.¹²⁷ An effort will be made to do so here whenever possible, but more often than not the line between the two levels of significance cannot be discerned.

From the time of the Church Fathers, pagan literature was read by the exegetically-inclined for its moral precepts as well as for the Christian truths it was believed to mask. Thus Saint Jerome spoke of secular literature as a captive woman whose beauty and eloquence could lead to higher spiritual understanding,¹²⁸ while Saint Augustine compared its value to that of the gold and silver ornaments which the Israelites took with them when they fled from Egypt in order to put them to a better use.¹²⁹

It is but one step from such views of the Scriptures and of classical authors to similar interpretations of vernacular literature. D. W. Robertson, Jr., is in the forefront of scholars who have promoted the study of exegetical writings as the key to understanding medieval secular authors.¹³⁰ While the “polarities” popularly associated with the Princeton professor need not concern us here,¹³¹ many of his views are applicable to the Song of Roland.¹³²

The Song of Roland has different meanings for different readers, but the possibilities afforded by what has become known as a “medieval reading” should not be overlooked. Symbolism was part and parcel of medieval education at all levels; it had important manifestations in Romanesque art and was utilized in sermons from the pulpit. The recurrence of the ironic mode in Turoldus's poem implies that the audience will be able to distinguish surface from real meanings. Reading more than one meaning into a poem or perceiving a symbolic allusion in a particular passage is not at all the same as applying the fourfold exegetical method. The latter approach is always possible of course, but Turoldus doubtless expected a far less technical mental operation on the part of most individuals in his audience—for instance, grasping the notions referred to below in the sections relating to meaning and structure.

Modern critics use the term allegory in a variety of ways that irritate many Old French specialists who prefer to restrict its meaning to the consistent and elaborate form found, for example, in the Roman de la Rose.¹³³ Yet medieval writers conceived of allegory in a much broader sense, encompassing all figurative speech, and they uncovered symbols in a manner some scholars of today find contradictory, excessive, or incongruous.¹³⁴

The Latin Pseudo-Turpin Chronicle and the German Rolandslied are important witnesses to the process of medieval interpretation of the Song of Roland. The first of these works was strongly influenced by Turoldus's epic and the second is essentially a translation of the French poem.

A. THE PSEUDO-TURPIN CHRONICLE

Opinions vary as to the exact date of the prose work purporting to be by Archbishop Turpin, but it would appear that the earliest surviving version of the Latin chronicle was composed about 1130 by a Frenchman who was familiar with Spain.¹³⁵ The following interpretations are noteworthy, for they relate to images, motifs, and themes present in the Song of Roland and doubtless were not foreign to Turoldus's intentions.

1. THE FLOWERING LANCES.

On the eve of Charlemagne's first battle with Aigolandus near Sahagún in the province of León, the Christians plant their lances in the ground. The next morning the lances of those who are destined to die have taken root, are covered with bark, and have sprouted leaves.¹³⁶ The miracle is repeated later before the castle of Taillebourg (Charente-Maritime).¹³⁷ The image of the flowering lances ultimately stems from Psalms 92:12-13: “So the virtuous flourish like palm trees and grow as tall as the cedars of Lebanon. Planted in the house of Yahweh, they will flourish in the courts of our God.”¹³⁸ In the Song of Roland the hero is associated with a lance or spear¹³⁹ on several occasions (vv. 707-708, 1156, Roland raises or brandishes his pennant-tipped lance; v. 720, he is the spear that Ganelon shatters in Charlemagne's first dream);¹⁴⁰ this weapon is merely an extension of the right arm, a symbol of power and strength (Psalms 98:1; 138:8), and a metaphor linking Charles and Roland (vv. 597, 727, 1195).¹⁴¹ The flowers of Paradise (vv. 1856, 2197, 2898) and the blossoms stained with the blood of the Christian martyrs at Roncevaux (v. 2871) are related to this motif, as is the concept of the French being interred inside a church. …

2. DYING IN BATTLE AGAINST THE SARACENS CONSTITUTES MARTYRDOM AND IS A SOURCE OF EDIFICATION AND JOY FOR ALL CHRISTIANS.

The miracle of the flowering lances is interpreted in the following terms: “Mira res, magnumque gaudium, magnum animabus proficuum, ingensque corporibus detrimentum!… In praefata acies fas est intelligi salus certantium Christi.”¹⁴² Turoldus refers to Roland and his men as martyrs (vv. 1134, 1922) and their deeds help sustenir chrestïentet (v. 1129). The joy of martyrdom is intimately related to Charlemagne's thoughts about the slain Roland and to the terms Joyeuse and Monjoie. …¹⁴³

3. ARMING FOR BATTLE LIKENED TO GIRDING FOR SPIRITUAL COMBAT.

In what is evidently, to the anonymous chronicler's mind, a logical development, the flowering lances suggest the battle for man's soul:

Sicut enim Karoli milites pugnatori ante bellum arma sua ad debellandum praeparaverunt, sic et nos arma nostra, id est bonas virtutes, contra vicia pugnaturi praeparare debemus. Quisquis enim vel fidem contra haereticam pravitatem, vel caritatem contra odium, vel largitatem contra avaritiam, vel humilitatem contra superbiam, vel castitatem contra libidinem, vel orationem assiduam contra daemoniacam temptationem, vel paupertatem contra felicitatem, vel perseverantiam contra instabilitatem, vel silentium contra iurgia, vel obedientiam contra carnalem animum ponit, hasta eius florida et victrix in die iudicii Dei erit. O quam felix et florida erit in celesti regno victoris anima qui legitime contra vicia decertavit in terra! Non coronabitur quis nisi qui legitime certaverit.¹⁴⁴

The metaphor derives from Ephesians 6:10-17.¹⁴⁵ Arming is an epic commonplace,¹⁴⁶ but this does not mean that it is always devoid of the spiritual significance alluded to here. In fact, Christian symbolism is strongly suggested in the riverside encampment episode (Laisses 180-182), where the motif of vigilance appears together with that of weapons. Cf. Ephesians 6:13-18:

You must rely on God's armour … So stand your guard, with truth buckled round your waist, and integrity for a breastplate, wearing for shoes on your feet the eagerness to spread the gospel of peace and always carrying the shield of faith so that you can use it to put out the burning arrows of the evil one. And then you must accept salvation from God to be your helmet and receive the word of God from the Spirit to use as a sword. … Never get tired of staying awake.

The battle of the virtues and vices, overtones of which are frequently perceived throughout Turoldus's epic, notably in the single combats, was graphically portrayed in the Psychomachia, but the oppositions noted in the Pseudo-Turpin Chronicle are not those found in Prudentius's poem. The particular alignment here derives either from contemporary tracts or from Romanesque sculpture.¹⁴⁷

4. DYING IN BATTLE REPRESENTS THE OBLIGATION TO ABANDON VICES AND TO LEAD A MORAL LIFE.

In another free association the Latin chronicler links the notion of spiritual combat and martyrdom to that of “dying” to vices: “Et sicut Karoli pugnatores pro Christi fide obierunt in bello, sic et nos mori debemus viciis et vivere virtutibus sanctis in mundo, quatinus palmam de triumpho floridam habere mereamur in celesti regno.”¹⁴⁸ The comparison is a patristic commonplace¹⁴⁹ stemming from Romans 6:2: “We are dead to sin, so how can we continue to live in it?” Cf. 6:11: “You too must consider yourselves to be dead to sin but alive for God in Christ Jesus.”

5. FLEEING FROM THE ENEMY SIGNIFIES MORAL LAXNESS IN COMBATING VICES.

Following a truce, during which Charles attempts to convert Aigolandus in a theological debate, battle is resumed. Twenty, forty, then a hundred Christians slay an equal number of Saracens. However, when a hundred more Christians attack, the enemy destroys them all because they have fled out of fear.

Hii vero tipum gerunt certantium fidelium Christi. Quia, qui pro Dei fide volunt pugnare, nullo modo debent retro abire. Et sicut illi ideo occiduntur quia retro fugerunt, sic Christi fideles qui debent fortiter contra vicia pugnare, si retro reversi fuerint, in viciis turpiter moriuntur. Sed qui bene contra vicia pugnant, hi inimicos, id est daemones qui vicia administrant, leviter occidunt. Non coronabitur quis, inquit apostolus, nisi qui legitime certaverit.¹⁵⁰

The reference in the last sentence is to 2 Timothy 2:5.¹⁵¹ In the Song of Roland fleeing is an act generally associated with the Saracens, and Paien s'en fuient is a first-hemistich formula used in vv. 2162, 2164, 2460, 3625, 3634 (cf. v. 1875). Scholars have read religious meaning in the image of the Saracens running before Roland (v. 1874: Si cum li cerfs s'en vait devant les chiens). It is also worth noting that when Gautier de l'Hum makes his eleventh-hour appearance, he is said to have been forced to flee from the enemy (v. 2043: Voeillet o nun, desuz cez vals s'en fuit). …

6. CHARLEMAGNE'S VICTORY DEMONSTRATES THAT CHRISTIANITY TRANSCENDS ALL OTHER FAITHS.

There is no Baligant episode in the Pseudo-Turpin Chronicle, but the Emperor's struggle against Aigolandus, which precedes Roncevaux in the Latin work, has points in common with this lengthy incident in Turoldus's poem. When Aigolandus is decisively defeated at Pamplona, the chronicler offers this commentary: “Quapropter patet quia lex christiana omnes ritus et leges tocius mundi excellit sua bonitate. Cuncta transcendit, super angelos etiam ascendit.”¹⁵² The idea that Christianity is superior to all other religions is central to the meaning of the Song of Roland (Introduction, 11, a). Unwavering faith permeates Turoldus's entire epic. Its major themes gravitate about this ideal and its story line is constructed on the Abraham-Isaac archetype, a symbol of instant and unquestioning acquiescence to divine promptings. One of the most important metaphors in the Roland has to do with Ascent (Introduction, 15, c). The Latin chronicler thinks along the same lines, for he continues the passage in the following fashion:

O Christiane, si fidem bene tenueris corde, et operibus in quantum poteris adimpleveris, veraciter super angelos cum capite tuo Christo, cuius membrum es, sublimatus eris. Si vis ascendere, firmiter crede; quia omnia possibilia sunt credenti, dicit Dominus.¹⁵³

7. DESPOILING BATTLEFIELD VICTIMS VIEWED AS RETURNING TO SINFUL WAYS.

Following Charlemagne's great victory over Aigolandus, about a thousand Christians, coveting gold, silver, and other riches, return to the battlefield. Loaded down with spoils, they are on their way back to camp when the Almaçor of Córdoba and his cohorts ambush and slay them to the last man. According to the chronicler:

Hii vero tipum gerunt certancium Christi. Quia sicut illi, postquam inimicos suos devicerunt, ad mortuos, cupiditatis causa, redierunt, et interficiuntur ab inimicis, sic fidelis quisque, qui vicia sua devicit, et poenitenciam accepit, ad mortuos, id est ad vicia, iterum redire non debet, ne forte ab inimicis, id est a daemonibus, interficiatur malo fine.¹⁵⁴

Amplifying his thought and aiming his shafts now at monks who betray their calling, the anonymous author adds:

Et sicut illi qui ad aliena spolia revertentes praesentem vitam perdiderunt, et necem turpe acceperunt, sic religiosi quique qui seculum dimiserunt, et ad terrena negocia postea inflectuntur, vitam celestem perdunt, et mortem perpetuam amplectuntur.¹⁵⁵

There is no such episode in the Song of Roland, but readers will be struck by the parallel with the Rash Saracen incident in Laisses 178-179. As Roland, eyes closed, solemnly prepares to meet his Maker, a pagan soldier, who until this moment has been feigning death, rushes toward him, hoping to take Durendal away as a trophy. The hero, feeling his sword slip from his grasp, opens his eyes and, summoning up his last reserve of strength, strikes the Saracen a mortal blow with the oliphant. The passage lends itself to various interpretations, one of which readily suggests itself following a reading of the battlefield despoliation scene in the Pseudo-Turpin.

8. GANELON'S BETRAYAL PARALLELS THAT OF JUDAS.

At bay at Roncevaux, Roland sounds the oliphant. Charlemagne hears its call and wants to rush to his nephew's aid. Ganelon attempts to dissuade the Emperor, suggesting that his stepson blows the horn at the slightest provocation and is probably merely hunting a wild animal in the forest. However, Ganelon is well aware that Roland is dying and his counsel is the height of treachery: “O subdola controversia! O Ganaloni pravum consilium, Iudae proditoris tradicioni comparatum!”¹⁵⁶ The Judas comparison, clearly yet only indirectly alluded to in the French original (e.g., the manner in which Ganelon is introduced in v. 178, betrayal with a kiss, the greed motive including the denarii mentioned in v. 1148, and, by inference, the Christlike characterization of Roland, the individual he betrays),¹⁵⁷ is thus spelled out in the Pseudo-Turpin Chronicle. The Ganelon-Judas connection became a literary commonplace in the Middle Ages.¹⁵⁸

B. THE ROLANDSLIED

Exegetical commentary is also evident in Conrad's German adaptation of the Song of Roland. As in the case of the Pseudo-Turpin Chronicle, from which it borrowed, considerable controversy swirls around the date of the Rolandslied, which appears to have been composed about 1170.¹⁵⁹ Following are some of the observations by Conrad that relate to the matter at hand.

1. GANELON'S TREASON AGAIN LIKENED TO THAT OF JUDAS.

On the way to Saragossa Blancandrin and Ganelon reach an agreement to kill the hero. In Turoldus's poem the plot is hatched on horseback but in no particular locale (vv. 402-404: Tant chevalcherent Guenes e Blancandrins / Que l'un a l'altre la sue feit plevit / Que il querreient que Rollant fust ocis). Although the preliminary discussion in Conrad takes place in the same fashion, the final details are worked out during a halt under an olive tree (v. 1920: under einem oeleboume), mentioned only in passing at the beginning of the scene in the French original (v. 366: Guenes chevalchet suz une olive halte).¹⁶⁰ The German translator then states that Ganelon imitated “poor” Judas (v. 1925: den armen Iudas er gebildot), who betrayed Christ for thirty pieces of silver.¹⁶¹ Conrad even suggests that Ganelon's treachery was greater, for whereas Judas sold only his Master, the villain of his poem sold a large number of noble Christians to the Infidel (vv. 1936-1939). Finally, when Ganelon vainly tries to persuade Charles not to answer Roland's oliphant call, Conrad has Naimes characterize the villain as a man in Satan's power and whose behavior has been worse than that of Judas, who betrayed Our Lord (vv. 6102-6104).

2. A HANDSOME EXTERIOR MAY MASK AN EVIL HEART.

Ganelon's good looks and impressive physique are detailed by Turoldus in vv. 283-285, immediately after Roland nominates his stepfather for the mission to Saragossa, and again at Aix just as the traitor begins his formal defense (v. 3763). In the latter instance Ganelon's handsomeness is promptly followed by a crucial disclaimer: S'il fust leials, ben resemblast barun (v. 3764), a phrase linking him to the evil Saracens (vv. 899, 3164). Ganelon's attractive features have misled many scholars into believing that Turoldus stood in grudging admiration of the traitor or, for some reason or other, refrained from painting him completely black. Nothing could be further from the truth, as the German version makes abundantly clear. Conrad omits the two descriptions of Ganelon cited above, but he inserts a comparable phrase (v. 1960: Genelun was michel unde lussam) in the council scene on the way to Saragossa. He compares the handsome villain to a tree that is green on the outside but is rotten to the core. This symbol, he explains, represents the man who speaks fine words but has a false heart (vv. 1962-1975). The allusion is doubtless to Christ's warning about false prophets: “You will be able to tell them by their fruits. Can people pick grapes from thorns, or figs from thistles? In the same way, a sound tree produces good fruit but a rotten tree bad fruit.”¹⁶²

3. GANELON WAS INSPIRED BY THE DEVIL.

In Conrad the discussion between Ganelon and Marsile at Saragossa begins, as in the French version, beneath a pine tree (v. 407: Un faldestoet out suz l'umbre d'un pin; cf. v. 500: Vait s'apuier suz le pin a la tige). However, in Turoldus's poem Marsile and his advisers subsequently withdraw to a garden (v. 501: Enz el verger s'en est alez li reis) where Ganelon is later led before the Saracen king (v. 510: Enz el verger l'en meinet josqu'al rei). No garden is mentioned in the German translation, although movement away from the pine tree is clearly implied at one point (vv. 2176-2178). In the Rolandslied the council is chiefly associated with the pine tree and is, in fact, referred to in a style reminiscent of the manner of designating adventures in the romances as der Pinrat (v. 2411). Nowhere in the Song of Roland does the poet specify that Ganelon was inspired by the devil, although the villain's machinations and sacrilegious oath are clearly diabolical, and Charlemagne's angry reaction to his nomination of Roland to the rearguard (vv. 746-747: Si li ad dit: “Vos estes vifs diables, / El cors vos est entree mortel rage!”) is tantamount to such an accusation. Conrad shows no comparable reticence (v. 2365: Den tuuil gab ime den sin) and, having identified the diabolical source of Ganelon's deeds in the Pine Tree episode, likens his words to that of the Accuser (Heb. satan) in Psalms 109:6.

4. COMPANIONAGE IS A HOLY BOND.

As the French heroes prepare for battle, Conrad compares their comradely devotion to one another to the brotherly love binding priests and Levites in Psalm 133 (vv. 3455-3457). Companionage is frequently alluded to in the French original and, at times, clearly assumes the mystical quality referred to by the German poet.¹⁶³

5. THE PINABEL-THIERRY DUEL COMPARED TO THE SLAYING OF GOLIATH

Pinabel's giant stature as opposed to his adversary's slightness is plainly designed to enhance the magnitude of Thierry's victory. Such an unequal duel also, however, brings to mind the David and Goliath archetype, a connection specifically made by Conrad (vv. 8847-8850).¹⁶⁴

One must avoid the pitfall of suggesting that, because Conrad or the author of the Pseudo-Turpin Chronicle read certain meanings into characters and situations found in the Song of Roland, Turoldus necessarily shared these views. One must always distinguish, too, between the artist's work and the meanings that contemporary or later translators and adapters—not to mention modern critics—attach to it.¹⁶⁵ A good example of the gap that exists between theory and practice in the Middle Ages may be seen in the curious views found in heraldic treatises, which are often strikingly at variance with the art practiced by contemporary compilers of blazoned or painted rolls of arms.¹⁶⁶

Turoldus tends to narrate events and leave their interpretation to his audience, whereas the Latin chronicler and the German translator frequently explain them to us. Their glosses may at times seem destructive of the obscurity or variety of possible interpretations we enjoy wrestling with.¹⁶⁷

On the other hand, Turoldus had no intention of mystifying his audience. His art is often one of subtle suggestion, but he also wanted his meaning to be understood: Ki tant ne set ne l'ad prod entendut (v. 2098). Many renderings in the Pseudo-Turpin and the Rolandslied ring false and must obviously be ignored. This does not mean, however, that their novel interpretations and commentaries should all be summarily dismissed. There is, after all, a good chance that Conrad and the Latin chronicler were at times more attuned to Turoldus than modern readers can ever hope to be.

11. THE MEANING OF THE SONG OF ROLAND

In addition to the literal sense conveyed by the story line, there is another dimension in Turoldus's poem. This deeper meaning, which has to do with its moral and spiritual significance, is at times only dimly perceived, yet it is ever-present in this work and is fundamental to its correct interpretation.

A. SUPERIORITY OF THE CHRISTIAN RELIGION

The essential statement made by Turoldus is that Christianity transcends all other faiths. It is so superior, in fact, as to be in a class by itself, other forms of belief being poor excuses for religion.

B. ELECTION¹⁶⁸

In Turoldus's view God has chosen Charlemagne and the Franks for a special task, that of establishing his rule throughout the world by means of armed conquest or conversion. This election is chiefly embodied in the figure of Charlemagne. Roland fights to defend his los and the reputation of his parents, of France, and of Charlemagne. However, at the time the poem was composed, these ideals were frequently synonymous and had deep spiritual significance. … In the eleventh century Frenchmen considered themselves to be the Chosen People, the nation selected by God to accomplish his ends, and believed that the Emperor was the vicarius Christi.¹⁶⁹ This ideology is manifest in the De consecratione pontificum et regum, composed by the Norman Anonymous about 1100:

The power of the king is the power of God, in that God possesses it by his nature and the king through grace. Thus the king also is God and Christ, but through grace; and everything he does, he does not do simply as a man but as a result of having become God and Christ through grace.¹⁷⁰

This christological concept explains the bold transposition of the Majestas Domini in the Gospel Book of Otto II at Aachen …, executed about 973:

[L'empereur] est véritablement divinisé en ce sens qu'il trône comme le Sauveur carolingien de Gyulia Fehervar, comme les Sauveurs ottoniens de Darmstadt et de Heidelberg, originaires de la Reichenau, au sein d'une gloire entourée du tétramorphe. Otton II tient le globe crucigère, tandis que sur sa tête, la Main du Père dépose un diadème. Ce couronnement est un legs carolingien (cf. Sacramentaire de Corbie à Paris, B.N. lat. 1141, 2^e moitié du IX^e siècle), ainsi que l'assistance des guerriers et des clercs (comme dans le Codex aureus de Saint-Emmeran à Munich Clm. 14000, vers 870). Aussi bien, la célèbre formule “a deo coronatus” est-elle utilisée dès le règne de Charlemagne sur les actes de chancellerie. Cependant les apports de la Basse-Antiquité sont encore plus significatifs à Aix-la-Chapelle: intervention de la Terre, symbole de domination universelle, qui, accroupie, supporte le poids du tabouret impérial—mais surtout “christomimétisme” dont s'accompagne l'apothéose, comme le montre E. Kantorowicz: entrée en vigueur, sur le plan chrétien, d'un principe cher aux derniers empereurs de la Rome païenne, car Otton II est la réplique visible de la Majestas Domini, comme Probus ou Constantin le sont du “Sol Invictus” dont ils s'avèrent les sosies sur certaines de leurs monnaies. Quant au fond d'or, utilisé ici pour la première fois dans l'enluminure occidentale, il provient directement de l'art byzantin posticonoclaste.¹⁷¹

C. SAPIENTIA

Christomimeticism leads inevitably to Sapientia. The virtually inexhaustible concept of wisdom in the Middle Ages derives from popular as well as from learned sources.¹⁷² However, a distinction was always drawn between worldly wisdom and Divine Wisdom. The latter was considered to be one of the attributes of God and was identified with Christ. As a corollary, one finds the notion that all worldly wisdom was vanity, whereas the Folly of the Cross was the only true Wisdom. This idea is frequently expressed in the Epistles of Saint Paul. Saint Augustine's views, which dominated all metaphysical speculation on the subject until the end of the eleventh century, are derived from that concept.

A study of the terminology of wisdom in the Song of Roland reveals that, with one notable exception,¹⁷³ it has nothing to do with Sapientia in the biblical sense but refers instead to things of this world. Most of the time these terms have to do with military decisions—either in a council or before or during a battle—or with diplomacy. Such wisdom was held in high regard during the Middle Ages except, of course, when it pertained to an evil person whose vice transformed it into cunning or guile. Turoldus repeatedly underscores the vanity of such wisdom, represented here by the advice and counsel offered by the majority of the Franks, and by the Saracens.

Nevertheless, the type of Sapientia found in the Bible and in the writings of the Fathers of the Church is very much in evidence in Turoldus's poem. It is apparent in Turpin's revelations concerning the significance of the Battle of Roncevaux, and in Roland's arduous ascent and incarnation of the Folly of the Cross. Above all, it is evident in Charlemagne's perception of the tragedy as it unfolds and in his key role in allowing it to happen.¹⁷⁴

Wisdom in the Song of Roland implies unswerving faith in God, absolute confidence in the inevitability of Christian victory, and total commitment to the view that immediate and spontaneous compliance with divine promptings is the way to personal salvation and to the edification of mankind.

Sapientia is the supreme virtue and, like all such qualities, a gift from God. But it is not easily acquired. The Church Fathers believed that the Liberal Arts could at best provide a beginning of learning. True insight into the meaning of things was to be found in Scripture alone.¹⁷⁵ Turoldus epitomizes Wisdom in his poem by Monjoie, which often suggests Joy mixed with Suffering,¹⁷⁶ and Roncevaux, where life is viewed as a Vale of Tears, but with Joy at the End of the Journey.¹⁷⁷

D. MARTYRDOM: IMITATIO CHRISTI

Roland's passio is the central fact in Turoldus's poem: The hero's suffering and death is an imitation of Christ, and his sacrifice constitutes a new kind of martyrdom.¹⁷⁸

Anyone endeavoring to justify Roland's behavior has to contend with a number of ambivalent character traits, in particular his boasting (Introduction, 19, c) and his apparent indifference to the fate of his men at Roncevaux. … Roland may be a martyr, but his behavior does not coincide with the modern idea of the individual who voluntarily suffers death for his faith.

It is difficult for the modern mind to apprehend the medieval ideal embodied by Roland. However, an illuminating parallel is to be found in the stubborn and imprudent conduct of Thomas Becket, whom the Church has always regarded as one of its most glorious martyrs.¹⁷⁹

Turoldus created complex characters, but he surely did not mean to highlight the hero's weakness or to portray him as a scapegoat. Roland's unswerving determination to play the role assigned to him and his exemplary death are Christlike and were intended to edify, that is, to instruct and fortify Turoldus's contemporaries. “Mira res, magnumque gaudium, magnum animabus proficuum, ingensque corporibus detrimentum,” exclaims the anonymous author of the Pseudo-Turpin Chronicle at the death of the Franks, revealing indifference to what is detrimental to the body as he considers with unbounded joy the soul's enormous benefit (Introduction, 10, a, 2).¹⁸⁰

The seeming ingenuousness that characterizes the Latin chronicler's solution to the age-old problem of the premature death of just men (quia noluit ut ad propriam patriam amplius redirent, ne forte in aliquibus delictis incurrerent; et enim voluit illis pro laboribus suis coronam celestis regni per passionem impendere)¹⁸¹ has a basis in Wisdom 4:11 ([The virtuous man] has been carried off so that evil may not warp his understanding or treachery seduce his soul), an answer which, the biblical sage readily admits, leaves many looking on, uncomprehending (4:14).

Roland must die, for “unless a wheat grain falls on the ground and dies, it remains only a single grain, but if it dies, it yields a rich harvest.”¹⁸²

The hero's sacrifice gives the poem a mystical dimension that is extended by Durendal, Joyeuse, the oliphant, the landscape, and the rituals associated with the passing of the Franks. Sacredness in the Roland is further enhanced by the contrasting effect of Ganelon's sacrilegious oath and Judas-like betrayal, the pagans' demonic presence, and the intrusion of the Rash Saracen at the moment of Roland's death. The dying attitude and position on the field assumed by Roland and the way he arranges the bodies of his slain comrades are important semaphore messages for Charles and his men. … Viewed from a Romanesque perspective, Charles's triumphs over Marsile's and Baligant's forces, Bramimonde's conversion, and Thierry's victory are the fruits of Roland's martyrdom. Miraculous occurrences such as these traditionally follow the saint's passio.¹⁸³

E. THE STRICT ALTERNATIVE

The Song of Roland is not merely a paean to martyrdom; it is also, and above all, life viewed as a series of difficult choices, the correct response requiring one to follow the hard road and to enter by the narrow gate.¹⁸⁴ Roland, like Christ, provokes dissension because of the strict alternative he offers, whether in council (his recommendation with respect to Marsile's proposal) or on the battlefield (the debate with Oliver): “Do not suppose that I have come to bring peace to the earth: it is not peace I have come to bring, but a sword. For I have come to set a man against his father, a daughter against her mother, a daughter-in-law against her mother-in-law. A man's enemies will be those of his own household” (Matthew 10:34-36). At the conclusion of the poem, the prospect before Charles of yet another campaign constitutes a brilliant extension of this agonistic metaphor.¹⁸⁵ Life is a never-ending Roncevaux that must be faced with courage and with faith (Introduction, 15, a and e). …

Notes

1. On the day and month of the battle, see Jenkins, notes to vv. 1002 and 2772 (cf. also his note to v. 2628); André de Mandach, Naissance et développement de la chanson de geste en Europe, vol. 1, La Geste de Charlemagne et de Roland, Publications romanes et françaises 69 (Geneva: Droz; Paris: Minard, 1961), pp. 50-55; Jules Horrent, “La bataille des Pyrénées de 778,” Le Moyen Age 78 (1972): 197-227. On the date of the poem, see n. 27 below.

2. A History of the Crusades, gen. ed. Kenneth M. Setton, vol 1, The First Hundred Years, ed. Marshall W. Baldwin (Madison-Milwaukee-London: University of Wisconsin Press, 1969), pp. 31-34.

3. Ramón Menéndez Pidal, La Chanson de Roland et la tradition épique des Francs, trans. Irénée-Marcel Cluzel, 2d ed. (Paris: Picard, 1960), pp. 181-230, provides a detailed account of these events. Cf. Horrent, “La bataille des Pyrénées.”

4. Eginhard, Vie de Charlemagne, ed. and trans. Louis Halphen, 3d ed. (Paris: Les Belles Lettres, 1947), pp. 28, 30. This work is hereafter referred to as Halphen. English translation mine.

5. Actually, only those in the north of Spain, but not Saragossa. Saracens (Sarraceni) are mentioned by name only once in Einhard's Vita, in a passage concerning Charles Martel's victory at Poitiers (Halphen, p. 10). See, however, Charlemagne's close relationship with Harun-al-Rashid (Aaron), Caliph of Bagdad (pp. 46, 48) and his action against the Moorish pirates in the Mediterranean (pp. 52, 54).

6. Gascons, not Basques. Paul Aebischer, Préhistoire et protohistoire du Roland d'Oxford, Bibliotheca Romanica, Series prima: Manualia et commentationes (Berne: Francke, 1972), pp. 75-87. Horrent, “La bataille des Pyrénées,” pp. 202-3, suggests that the attackers were Pyrenean Gascons, rather than French Gascons (often referred to as Basques).

7. Halphen, pp. 16, 18.

8. Menéndez Pidal, p. 192.

9. Ibid., pp. 194-95; texts on p. 195, nn. 1, 2 (also pp. 528-29); Barton Sholod, Charlemagne in Spain: The Cultural Legacy of Roncesvalles (Geneva: Droz, 1966), p. 40 and n. 65.

10. Menéndez Pidal, pp. 204-9; Sholod, p. 41. Horrent, “La bataille des Pyrénées,” p. 204, expresses reservations about Saracen complicity in this respect. Aebischer, p. 88: “l'assertion de Menéndez Pidal [à propos d'une] collaboration des Arabes et des Basques à Roncevaux … ne repose que sur le fait mal interprété par lui de l'enlèvement de Sulaiman par ses fils.”

11. In the poem Roland is a Franc de France, not a Breton. Bédier, Commentaires, pp. 37-40; Horrent, La Chanson de Roland, pp. 305 and 306, n. 2. On the historicity of this personage, see Aebischer, pp. 93-145.

12. In Einhard perfidia means ‘treachery’ and is associated with the Saxons. Halphen, p. 23 and n. 4. In patristic literature perfidia often refers to the disbelief of non-Christians and is synonymous with incredulitas. Blaise, par. 392; see also Commentary, 2 (vv. 24-26) [in “The Song of Roland”: An Analytical Edition, Vol. 1. University Park, PA: The Pennsylvania State University Press, 1978].

13. Originated by Pope Leo III in a.d. 800 and styled the “Empire of the West” and the “Roman Empire,” the concept became the “Holy Empire” with Frederick I Barbarossa in 1157 and the “Holy Roman Empire” in 1254. Charlemagne was canonized at Frederick's behest on 29 December 1165.

14. André Burger, “La légende de Roncevaux avant la Chanson de Roland,” Romania 70 (1948-49): 453-73. See, however, Jean Rychner, “A propos de l'article de M. André Burger ‘La légende de Roncevaux avant la Chanson de Roland’,” Romania 72 (1951): 239-46, and Burger's reply, “Sur les relations de la Chanson de Roland avec le Récit du faux Turpin et celui du Guide du Pèlerin,” Romania 73 (1952): 242-47; Horrent, La Chanson de Roland, p. 155; Maurice Delbouille, Sur la Genèse de la Chanson de Roland (Travaux récents—Propositions nouvelles): Essai critique, Académie royale de langue et de littérature françaises de Belgique (Brussels: Palais des Académies, 1954), p. 100; Menéndez Pidal, pp. 348-49; pp. 417 ff. (speaks of a lost Cantar de Rodlane).

15. Horrent, pp. 292-7; Menéndez Pidal, pp. 355-65; Rita Lejeune, “La naissance du couple littéraire ‘Roland et Olivier’,” Mélanges Henri Grégoire, Annuaire de l'Institut de philologie et d'histoire orientales et slaves 10 (Brussels, 1950), 2:371-401; Delbouille, Genèse, pp. 98-120. …

16. Horrent, p. 302, n. 1.

17. Ibid., pp. 120-34, 138-40, 242-59; Menéndez Pidal, pp. 123-29; Joseph J. Duggan, The Song of Roland: Formulaic Style and Poetic Craft (Berkeley-Los Angeles-London: Center for Medieval and Renaissance Studies, University of California, Los Angeles, 1973), pp. 63-104; John R. Allen, “Du nouveau sur l'authenticité de l'épisode de Baligant,” Société Rencesvals pour l'étude des épopées romanes. VI^e Congrès International (Aix-en-Provence, 29 Août-4 Septembre 1973). Actes (Aix-en-Provence: Imprimerie du Centre d'Aix, 1974), pp. 147-56. For arguments in favor of the authenticity of the Baligant episode, see Paul Aebischer, “Pour la défense et illustration de l'épisode de Baligant,” Mélanges de philologie romane et de littérature médiévale offerts à Ernest Hoepffner (Paris: Les Belles Lettres, 1949), pp. 173-82; Delbouille, Genèse, pp. 32-61. For a more extensive bibliography on this question, consult Marianne Cramer Vos, “Aspects of Biblical Typology in La Chanson de Roland,” Ph.D. diss., University of Rochester, 1970, pp. 165-66, n. 26; Duggan, p. 69, n. 7; idem, “The Generation of the Episode of Baligant: Charlemagne's Dream and the Normans at Mantzikert,” Romance Philology 30 (1976): 59-82.

18. For bibliography and discussion relative to this personage, see Jenkins, pp. xlviii-lxv; Bédier, Commentaires, pp. 31-40; Horrent, pp. 326-33; Pierre Le Gentil, La Chanson de Roland, Connaissance des lettres 43 (Paris: Hatier-Boivin, 1955), pp. 32-35; Martín de Riquer, Les Chansons de geste françaises, trans. Irénée-Marcel Cluzel, 2d ed. (Paris: Nizet, 1957), pp. 105-16; Mandach, Naissance, 1: 159; Rita Lejeune, “Turold dans la tapisserie de Bayeux,” in Mélanges offerts à René Crozet à l'occasion de son soixante-dixième anniversaire, eds. Pierre Gallais and Yves-Jean Riou (Poitiers: Société d'études médiévales, 1966), pp. 419-25; Jean Dufournet, Cours sur la Chanson de Roland, Les Cours de Sorbonne (Paris: Centre de documentation universitaire, 1972), pp. 16-17. Each of these scholars concludes that Turoldus was the poet or last redactor. For the view that Turoldus was merely a copyist, see Aebischer, Préhistoire, pp. 224-28. On the manner in which medieval authors signed their works, see Ernst Robert Curtius, La Littérature européenne et le moyen âge latin, 2d ed., trans. Jean Bréjoux, (Paris: Presses Universitaires de France, 1956), pp. 624-27.

19. On the date, see Jenkins, pp. xliii-xlvi (1099-1120); Bédier, Commentaires, pp. 40-59 (c. 1100); Horrent, pp. 287-304 (first half of the eleventh century for “la première Chanson de Roland”; c. 1100 for the redaction including the Baligant episode [pp. 315-19]; reign of Henry II [1154-89] for the reworking by Turoldus [p. 330]); Michel de Bouard, “La Chanson de Roland et la Normandie,” Annales de Normandie 2 (1952): 34-38; J. C. Russell, “The Chanson de Roland: Written in Spain in 1093?” Studies in Philology 49 (1952): 17-24; Delbouille, Genèse, pp. 62-73 (c. 1100); Le Gentil, pp. 23-32 (c. 1100); M. Dominica Legge, “Archaism and the Conquest,” Modern Language Review 51 (1956): 229 (1130-50, but “1150 … may be on the late side”); Hans Erich Keller, “La conversion de Bramimonde,” Société Rencesvals. VI^e Congrès International, pp. 175-203, and Olifant 1, no. 1 (1973): 3-22 (1. “Chanson de Roncevaux,” 1086-95; 2. “version capétienne” with Baligant episode, 1147-49 at Saint-Denis; 3. “version angevine” with new version of Bramimonde, third quarter of the twelfth century); see also Keller, “The Song of Roland: A Mid-Twelfth-Century Song of Propaganda for the Capetian Kingdom,” Olifant 3, no. 4 (1976), 242-58. For Menéndez Pidal's dating, see note 22 below. On the place, see Bédier, Commentaires, pp. 37-40; Horrent, pp. 304-7; David Douglas, “The Song of Roland and the Norman Conquest of England,” French Studies 14 (1960): 99-116; Keller, “The Song of Roland.” On the significance of the late-eleventh-century Nota Emilianense, see Dámaso Alonso, “La primitiva épica francesa a la luz de una ‘Nota Emilianense’,” Rivista de Filología Española 37 (1953): 1-94; Ronald N. Walpole, “The Nota Emilianense: New Light (But How Much?) on the Origins of the Old French Epic,” Romance Philology 10 (1956/57): 1-18; Le Gentil, pp. 45-47 (including Latin text and Modern French translation); Menéndez Pidal, pp. 384-447 and pls. X, XI; Moignet, pp. 293-94 (Latin text and Le Gentil's translation); Dufournet, Cours sur Roland, p. 26.

20. Jean Rychner, La Chanson de geste: Essai sur l'art épique des jongleurs, Société de publications romanes et françaises 53 (Geneva: Droz; Lille: Giard, 1955), p. 59.

21. Ibid., pp. 49, 61-62. Cf. Duggan, Song of Roland, pp. 63-67.

22. Rychner, p. 36, suggests the exact opposite but concedes that the Roland is an exception and involved a “création poétique” of the more conventional type, presumably clerical. … Menéndez Pidal, p. 342, denies that the original Song of Roland was composed by a clerk but allows, pp. 343-76, that later reworkings (e.g., the late-tenth-century version including Oliver and Alda, the eleventh-century version mentioning the Twelve Peers) were produced by individuals with schooling. Joseph J. Duggan, “Virgilian Inspiration in the Roman d'Enéas and the Chanson de Roland,” in Medieval Epic to the “Epic Theater” of Brecht, eds. Rosario P. Armato and John M. Spalek, University of Southern California Studies in Comparative Literature 1 (Los Angeles: University of Southern California Press, 1968), pp. 9-23, refuses to see any significant clerical influence in Turoldus's poem. Duggan, Song of Roland, pp. 13-14, 16, 36-60, 193, rejects the principle of composite creation: “the Roland which we possess must be a very nearly unadulterated product of oral tradition, little changed, except for its orthography, from the form in which it was first taken down from the lips of a singer or written down by a singer who had acquired literacy” (p. 60). My own view bears similarity to that propounded by Le Gentil in the series of articles cited by Duggan, p. 4, n. 7. See also Cecil M. Bowra, Heroic Poetry (London: Macmillan, 1952), pp. 247, 250-53, 368; Maurice Delbouille, “Les chansons de geste et le livre,” La Technique littéraire des chansons de geste: Actes du Colloque de Liège (Septembre 1957), Bibliothèque de la Faculté de philosophie et lettres de l'Université de Liège 150 (Paris: Les Belles Lettres, 1959), pp. 295-428; idem, “Le chant héroïque serbo-croate et la genèse de la chanson de geste,” Boletín de la Real Academia de Buenas Letras de Barcelona 31 (1965/66): 83-98; idem, “Le mythe du jongleur-poète,” Studi in onore di Italo Siciliano (Florence: Olschki, 1966), pp. 317-27; Jean-Charles Payen, “De la tradition à l'écriture: à propos d'un livre récent,” Le Moyen Age 75 (1969): 529-39; idem, Le Moyen Age, vol. 1, Des origines à 1300, Littérature française, ed. Claude Pichois (Paris: Arthaud, 1970), p. 120: “Les plus belles chansons ont été valorisées par l'écriture, et le Roland d'Oxford lui-même est un texte trop bien composé pour procéder d'improvisations, même géniales”; Edward A. Heinemann, “La composition stylisée et la transmission écrite des textes rolandiens,” Société Rencesvals. VI^e Congrès International, pp. 253-72. Robert Scholes and Robert Kellogg, The Nature of Narrative (New York: Oxford University Press, 1966), chap. 2, “The Oral Heritage of Written Narrative,” pp. 17-56, is a useful introduction to this complex question.

23. William Calin, The Epic Quest: Studies in Four Old French Chansons de Geste (Baltimore: Johns Hopkins Press, 1966), pp. 57-117; on the problem of definition in the case of the Chanson de Guillaume, see p. 93; for the Song of Roland, Gormont et Isembart, and Girart de Roussillon, see p. 116. See also Thomas E. Vesce, “Reflections on the Epic Quality of Ami et Amile: Chanson de Geste,” Mediaeval Studies 35 (1973): 129-45, and the critique by S. N. Rosenberg in Olifant 3, no. 3 (1976): 221-25.

24. Frederick W. Locke, The Quest for the Holy Grail: A Literary Study of a Thirteenth-Century French Romance (Stanford: Stanford University Press, 1960), pp. 18-22.

25. But with important variations; see, for example, the discussion and stemmata in Jenkins, pp. xcv-xcviii; Bédier, Commentaires, pp. 83-92; Eyvind Fjeld Halvorsen, The Norse Version of the Chanson de Roland, Bibliotheca Arnamagnaeana 19 (Copenhagen: Munksgaard, 1959), pp. 272-73; and Segre, pp. ix-xviii. The medieval texts of the Roland are described in Segre, pp. xxxvii-xlvii. Mortier (see Abbreviations) is a convenient edition of the essential texts.

26. La Chanson de Roland: Reproduction phototypique du manuscrit Digby 23 de la Bodleian Library d'Oxford, ed. Comte Alexandre de Laborde, Etude historique et paléographique de M. Ch. Samaran (Paris: Société des anciens textes français, 1933), pp. 28-32. This work is hereafter referred to as Samaran.

27. On the dialect, see Bédier, Commentaires, pp. 241-62. Dating of the hand in Digby 23 varies considerably: Bédier, Commentaires, p. 66 (c. 1170); Samaran, p. 30 (1130-50); Horrent, pp. 32-42 (second half of the twelfth century); Robert Marichal, Annuaire 1969-1970 de l'Ecole pratique des hautes études, IV^e section: Sciences historiques et philologiques. Extrait des rapports sur les conférences: Paléographie latine et française (Paris, 1970), pp. 363-74 (“plus proche de 1125 que de 1150” [p. 367]); Félix Lecoy, reviewing the latter work in Romania 92 (1971): 141, accepts its findings; Ian Short, “The Oxford Manuscript of the Chanson de Roland: A Paleographical Note,” Romania 94 (1973): 221-31 (c. 1170); Charles Samaran, “Sur la date approximative du Roland d'Oxford,” Romania 94 (1973): 523-27 (1130-50); Keller, “The Song of Roland,” pp. 244-45 (“not before 1170”). On the date of the original poem, see note 19 above.

28. Samaran, in La Chanson de Roland, ed. Laborde, pp. 33-36, 38.

29. Ibid., pp. 20-22, 39-40. Many, but by no means all, of these are inept changes and give editors of the Roland a headache.

30. Ibid., pp. 36-37.

31. Ibid., p. 39, citing the catalogue of the New Paleographical Society.

32. Ibid., pp. 39-41.

33. See references in note 27 above.

34. Edmond Faral, Les Jongleurs en France au moyen âge, 2d ed., Bibliothèque de l'Ecole des hautes études 187 (1910; rpt. Paris: Champion, 1964); Bowra, Heroic Poetry, chap. 1, “The Heroic Poem,” pp. 1-90; Rychner, La Chanson de geste.

35. Scholes and Kellogg, Nature of Narrative, pp. 30-40, discusses the relationship between oral and written versions of the Roland and other epics intelligently and objectively but offers no new solutions to this problem. …

36. The best general introduction to matters discussed in this section is Marc Bloch, Feudal Society, trans. L. A. Manyon, 2 vols. (Chicago: University of Chicago Press, 1961). Another useful survey is J. W. B. Zaal, “A lei francesa” (Sainte Foy, v. 20): Etudes sur les chansons de saints gallo-romanes du XI^e siècle (Leiden: Brill, 1962), chap. 1, “La France géographique et culturelle au XI^e siècle,” pp. 27-44.

37. Some urban activity is starting up again, but major developments are yet to come. Henri Pirenne, Economic and Social History of Medieval Europe, trans. I. E. Clegg (New York: Harcourt Brace, [c. 1956]), chaps. 2 and 3.

38. Bloch, 2:288: “if the concept of nobility as a legal class remained unknown, it is quite permissible from this period [i.e., the first feudal age], by a slight simplification of terminology, to speak of a social class of nobles and especially, perhaps, of a noble way of life.” On the role of the clergy, the third element of feudal society, see Introduction, 5 and 7.

39. Ibid., 1:145-275; Ganshof, Feudalism, pp. 65-155.

40. R. C. Smail, Crusading Warfare (1097-1193) (Cambridge: Cambridge University Press, 1956), pp. 106-7. See also Oxford Text, English Translation, v. 34. The word chevaler is used indiscriminately by Turoldus for both Christians and Saracens. That a certain type of behavior was expected of the Christian knight is clearly indicated by the expression a lei de chevaler (v. 752).

41. The functions of light-armed horsemen and pedites are discussed by Smail, pp. 107-12.

42. On the significance in this regard of “collective” vs. “individual” man about 1150, see Richard William Southern, The Making of the Middle Ages (New Haven: Yale University Press, 1961), p. 222; Burgess, Vocabulaire pré-courtois, p. 7; Anthony M. Beichman, “Ganelon and Duke Naimon,” Romance Notes 13 (1971): 358-62. Such a generalization tends to break down upon close scrutiny, as does the familiar dictum that the Middle Ages is an era of participation, the modern period, one of separation.

43. See Introduction, 15, b. On medieval pilgrimages, see Jacques Le Goff, La Civilisation de l'occident médiéval (Paris: Arthaud, 1972), pp. 172-74.

44. On the significance of itineraries, see Stephen G. Nichols, Jr., “The Interaction of Life and Literature in the ‘Peregrinationes ad loca sancta’ and the ‘Chansons de geste’,” Speculum 44 (1969): 51-77.

45. See, for example, the maps in Jenkins, pp. lxxviii-lxxix, (“after Bédier”), and Menéndez Pidal, pl. IX. Cf. Le Guide du Pèlerin de Saint-Jacques de Compostelle, ed. and trans. Jeanne Vielliard, 3d ed. (Macon: Protat, 1963).

46. David Herlihy, “The Generation in Medieval History,” Viator 5 (1974): 347-64.

47. Léon Gautier observed that France occurs 170 times in this poem in the meaning ‘Charlemagne's Empire’; see Jenkins, note to v. 36.

48. Jones, p. 130 and n. 97.

49. John F. Benton, “Clio and Venus: An Historical View of Courtly Love,” in The Meaning of Courtly Love, ed. F. X. Newman (Albany: State University of New York Press, 1968), p. 20. Common sense suggests, however, that the nobleman, who depended for his livelihood on the cooperation of his serfs, realized that harsh treatment would ultimately do more harm than good.

50. Horrent, p. 307; Payen, Le Moyen Age, pp. 31-32.

51. History of the Crusades, 1:135.

52. Ibid., 1:221.

53. Statistics for the period under consideration are not available, but the percentages were probably not very different from those provided here, which are based on figures for England in Sir Maurice Powicke, The Thirteenth Century 1216-1307, 2d ed., The Oxford History of England 4 (Oxford: Clarendon Press, 1962), pp. 445-46 (clerks), and Noel Denholm-Young, History and Heraldry 1254 to 1310: A Study of the Historical Value of the Rolls of Arms (Oxford: Clarendon Press, 1965), p. 1 (knights).

54. As early as the end of the fifth century, according to Zumthor, Histoire littéraire, p. 19.

55. Bloch, Feudal Society, 2:348.

56. Ibid., 2:346-47; Ganshof, Feudalism, p. 113.

57. Ganshof, pp. 116-17.

58. Zaal, A lei francesa, pp. 42-43.

59. Bernard de Clairvaux, Commission d'histoire de l'Ordre de Cîteaux 3 (Paris: Alsatia, 1953), pp. 687-88 (“Bernard et les écolès”). Cf., however, Abbot Guibert of Nogent, writing c. 1115:

    In the time just before my birth and during my childhood there was so great a dearth of teachers that it was practically impossible to find any in the small towns, and scarcely even in the cities. And supposing that by chance they were to be found? Their learning was so meagre that it could not be compared even with that of the little wandering scholars of today.

    Text in Bloch, Feudal Society, 1:104; see also Self and Society in Medieval France: The Memoirs of Abbot Guibert of Nogent (1064?-c. 1125), trans. C. C. Swinton Bland, revised by John F. Benton (New York and Evanston: Harper & Row, 1970), p. 45. A similar statement is found in Gargantua's letter to his son. François Rabelais, Pantagruel, ed. Verdun L. Saulnier, Textes littéraires français (Paris: Droz, 1946), chap. 8, pp. 43-44.

60. Grace Frank, The Medieval French Drama (Oxford: Clarendon Press, 1954), chaps. 2-7, pp. 18-73.

61. Bernard de Clairvaux, chap. 8, “L'essor économique de Clairvaux,” pp. 95-114.

62. Brandt, Medieval History.

63. Ibid., p. 169.

64. Ibid., pp. 152-53, 171-72.

65. Paul Archambault, Seven French Chroniclers: Witnesses to History (Syracuse: Syracuse University Press, 1974), pp. 1-6, 119.

66. What follows is based upon Hippolyte Delehaye, Les Légendes hagiographiques, 4th ed., Subsidia Hagiographica 18a (1927; rpt. Brussels: Société des Bollandistes, 1955), and idem, Les Passions des martyrs et les genres littéraires, 2d ed., Subsidia Hagiographica 13 B (1921; rpt. Brussels: Sociéte des Bollandistes, 1966). Father Delehaye's views are accepted by René Aigrain, L'Hagiographie, ses sources, ses méthodes, son histoire (Paris: Bloud & Gay, 1953), the best guide available to the Old French scholar for whom Latin saints' lives constitute a terra incognita.

67. Delehaye, Légendes hagiographiques, pp. 49, 86-87.

68. Edmond Faral in Joseph Bédier and Paul Hazard, Histoire de la littérature française illustrée (Paris: Larousse, 1923), 1:7-8; Horrent, pp. 302-3; Delbouille, Genèse, pp. 137-38, 142; Riquer, Chansons de geste, p. 109; Zaal, A lei francesa, chaps. 2 and 3, pp. 45-136.

69. Bernard de Clairvaux, chap. 5, “Le monachisme à l'apparition de Bernard,” pp. 45-63. For a list of early Cistercian abbeys, see app. 3, pp. 543-47.

70. This is the traditional figure. On the difficulty of ascertaining the exact number of monks and monasteries at this time, see Bernard de Clairvaux, p. 45.

71. For Cluniac establishments in northeastern Spain, the theater of events in most of Turoldus's poem, see Sholod, Charlemagne in Spain, p. 65.

72. L. M. Smith, Cluny in the Eleventh and Twelfth Centuries (London: Allan, 1930), p. 95, Latin text (Vos estis lux mundi) quoted in n. 2. The phrase, found in a charter granted to Cluny at Abbot Hugh's request by Pope Urban II in 1098, refers to Matthew 5:14.

73. Faral, Jongleurs, chap. 2, pp. 25-43.

74. Ibid., pp. 44-47. Faral also studies the relationship between these two genres and pilgrimages. According to Bowra, Heroic Poetry, p. 29, the composer of epic poetry “wishes not to instruct but to delight his audience.” However, he concedes, p. 30, that the poet with a Christian outlook may also have a didactic purpose.

75. Bloch, Feudal Society, 2:412-20.

76. Cited by Bloch, 2:417. Cf. Proverbs 26:11; 2 Peter 2:22.

77. Ibid., 2:312-16.

78. Ibid., 2:316-18.

79. Bruce A. Rosenberg, Custer and the Epic of Defeat (University Park and London: Pennsylvania State University Press, 1974).

80. This kind of anecdote is also found in medieval French romances, e.g., Le Roman de Thèbes, ed. Guy Raynaud de Lage, Classiques français du moyen âge 96 (Paris: Champion, 1968), 2: vv. 6219-6224, 6241-6244; Les Romans de Chrétien de Troyes, vol. 1, Erec et Enide, ed. Mario Roques, Classiques français du moyen âge 80 (Paris: Champion, 1952), vv. 2355-2376, 6668-6670 (cf. the more elaborate description of Enide's dress in vv. 6674-6741). This paragraph and the next two are drawn from my article “Ganelon et Roland: Deux anecdotes du traître concernant le héros,” Romania 92 (1971): 392-94.

81. Cf. Suetonius, an important model for Einhard. Halphen, pp. x-xiii.

82. Alice M. Colby, The Portrait in Twelfth-Century French Literature: An Example of the Stylistic Originality of Chrétien de Troyes (Geneva: Droz, 1965), p. 178.

83. Paul Zumthor, “Rhétorique et langage poétique dans le moyen âge roman,” Poetyka [First International Conference of Work-in-Progress Devoted to Problems of Poetics, Warsaw, August 18-27, 1960] (Warsaw: Państwowe Wydawnictwo Naukowe; The Hague: Mouton, 1961), pp. 745-53; Zaal, A lei francesa, pp. 46, 89, 92-116. For useful distinctions between culture cléricale and culture profane, on the one hand, and culture aristocratique and culture populaire, on the other, see Payen, Le Moyen Age, pp. 33-42.

84. Jenkins, notes to vv. 1470 and 1490 ff.; Bédier, Commentaires, p. 304 (see also Foulet, Glossaire, p. 354); Edmond Faral, La Chanson de Roland: Etude et analyse, Les chefs-d'oeuvre de la littérature expliqués (Paris: Mellottée, 1934), pp. 198, 199. Duggan, Song of Roland, pp. 139-40, treats the verses in question as an elaboration framed between two formulas (v. 1649, Siet el cheval; v. 1657, Beste nen est nule). Jean Györy, in his review of Raimund Rütten, Symbol und Mythus im altfranzösischen Rolandslied, Archiv für das Studium der neureren Sprachen und Literaturen 4 (Braunschweig: Westermann, 1970), in Cahiers de civilisation médiévale 16 (1974): 345, refers to: “le cheval de Turpin, décrit en ordre inversé, de bas en haut, pour marquer la provenance chtonienne de l'animal et en même temps son élan ascensionnel.”

85. Jenkins, p. xxxviii; Menéndez Pidal, p. 375; Zaal, A lei francesa, p. 94.

86. Robert L. Politzer, “Synonymic Repetition in Late Latin and Romance,” Language 37 (1961): 484-87.

87. Jones, Ethos, p. 9.

88. Ibid., p. 22.

89. Jones's initial statements relative to this word (pp. 22-23) are judicious enough, but he soon slips into categorical assertions concerning its special meaning and, when discussing semantically related terms, repeatedly suggests that proz always refers to courage and physical strength. See my review of Burgess, Vocabulaire pré-courtois, in Speculum 46 (1971): 363-64.

90. Proverbs are usually considered to be popular in origin and transmission, but many adages are found in classical sources or in collections such as the twenty-nine medieval French compilations analyzed in Proverbes français antérieurs au XV^e siècle, ed. Joseph Morawski, Classiques français du moyen âge 47 (Paris: Champion, 1925). Maxims such as Ki tant ne set ne l'ad prod entendut (v. 2098) and Mult ad apris ki bien conuist ahan (v. 2524) have a learned flavor, whereas a phrase like Plus qu'om ne lancet une verge pelee (v. 3323) has a decidedly popular aspect.

91. Bloch, Feudal Society, 2:345-52. See note 38 above.

92. Horrent, p. 307: “Notre poète était un clerc [n. 2: Mais non un moine (voir O vv. 1880-1881)],” citing Fawtier. Cf., however, Jenkins, note to v. 1881: “Tavernier points out that this judgment necessarily implies no scorn of the monk, as such: each is useful, nay indispensable, in his own field.” On clerical self-satire, see Philippe Ménard, Le Rire et le sourire dans le roman courtois en France au moyen âge (1150-1250), Publications romanes et françaises 105 (Geneva: Droz, 1969), pp. 175-78. Bernard de Clairvaux, p. 263, n. 1:

    Nous ne confondons pas clericus et canonicus. Le clericus n'est pas nécessairement un canonicus. Mais dans le haut Moyen-Age, le terme de clericus était singulièrement ambigu. Saint Jérome avait proposé cette définition de l'étymologie du mot kleros: ‘Les clercs sont appelés de ce mot parce qu'ils sont la part du Seigneur ou bien parce que le Seigneur est leur part’ (ep. ad Nepotianum, PL. 22, 531). Définition assez imprécise, qui fut reprise par le Décret de Gratien (11^a, causa XII, qu. 1, c. 5 et 7) et qui autorise deux acceptions du mot clericus, l'une large, l'autre restreinte. Ainsi en matière favorable, c'est-à-dire quand il s'agissait de l'application des privilèges, tous les religieux, même les moniales et les frères convers étaient compris parmi les clercs. Parfois même le terme de clerici désignait des laïcs serviteurs de l'Eglise. C'est en ce sens qu'il faut comprendre une décision du concile de Tours en 567, prescrivant à l'archiprêtre de se faire accompagner d'un canonicus (clerc) ou au moins d'un clericus (serviteur laïc). En un sens plus restreint, en matière pénale, il fut reçu d'exclure du sens du mot clerc, les cardinaux, les évêques, les dignités et les chanoines des églises cathédrales. On n'était pas d'accord sur les chanoines des églises collégiales. Les clercs qui accomplissaient des fonctions déterminées dans une église (et à cause de ces fonctions percevaient une part des revenus de l'église) portaient le nom de clerici canonici, soit qu'ils vécussent selon une règle (kanón) soit plutôt parce qu'inscrits sur la table ou liste (kanón) d'une église. Un peu plus tard, vers le VII^e siècle, interviendra la notion de vie commune, qui amènera la distinction, classique au XI^e siècle, entre canonici regulares et canonici seculares, les premiers vivant dans des monastères sub abbate, les seconds dans des cathédrales ou collégiales (in domo episcopali) sub episcopo. Sur cette question voir Dict. Dr. Canon, éd. Letouzey, t. III (1942) et R. Naz. Traité de Droit canonique. Letouzey (1946), t. I. passim.

93. Oscar Bloch and Walther von Wartburg, Dictionnaire étymologique de la langue française, 4th ed. (Paris: Presses Universitaires de France, 1964), s.v. clerc.

94. Cf. Faral, Jongleurs, chap. 3, “Les jongleurs aux cours seigneuriales,” pp. 93-102; chap. 4, “Les ménestrels,” pp. 103-18; chap. 5, “Les revenus des jongleurs,” pp. 119-27.

95. Felix Busigny, Das Verhältnis der Chansons de geste zur Bible, Inaugural-Dissertation (Basel: Reinhardt, 1917); Jenkins, pp. xlvi-xlvii; Adolphe J. Dickmann, Le Rôle du surnaturel dans les chansons de geste (Paris: Champion, 1926); Faral, La Chanson de Roland, pp. 186-95; Jessie Crosland, The Old French Epic (Oxford: Blackwell, 1951), p. 74. See also Bédier, Commentaires, p. 314; Riquer, Chansons de geste, pp. 109-10; Zaal, A lei francesa, pp. 130-34; Dufournet, Cours sur Roland, p. 28. According to Marie-Madeleine Davy, Initiation à la symbolique romane (XII^e siècle) (Paris: Flammarion, 1964), p. 121: “Les moines du XII^e siècle possèdent une parfaite connaissance de la Bible. Ils savent les textes par coeur et leur propre pensée est essentiellement biblique.”

96. Jenkins, notes to vv. 3238 and 1215, respectively.

97. For discussion and bibliography, see my paper “Le Thème de la Mort,” pp. 229-30, notes 20-23; Zaal, A lei francesa, pp. 117-19. … Bédier, Commentaires, p. 314, notes the resemblance between Roland and Judas Maccabaeus. Riquer, Chansons de geste, p. 102, believes the image of the stag in vv. 1874-1875 is drawn from the Bible.

98. Faral, La Chanson de Roland, pp. 198-201; Jones, pp. 127, 182; Brault, “Le Thème de la Mort,” p. 229.

99. Brault, pp. 230-35.

100. Wilhelm Tavernier, “Beiträge zur Rolandsforschung. I. Äneide, Pharsalia und Rolandsepos,” Zeitschrift für französische Sprache und Literature 36 (1910): 71-102; Jenkins, pp. xlvii-xlviii; Curtius, pp. 111, 245, 530 (but see, especially, Curtius's article “Zur Literarästhetik des Mittelalters,” Zeitschrift für romanische Philologie 58 [1938]: 215-32). See also Bédier, Commentaires, pp. 316-17; Riquer, Chansons de geste, p. 108; Jones, pp. 130-34; Aebischer, Préhistoire, pp. 232-34; Dufournet, Cours sur Roland, pp. 27-28. Turoldus's contemporary, Saint Bernard, refers in his works to Boethius, Cicero, Horace, Juvenal, Ovid, Persius, Seneca, Statius, Tacitus, Terence, and Virgil. Bernard de Clairvaux, p. 479 and app. 4.

101. Virgil is mentioned by name in v. 2616.

102. See note 82 above.

103. See … n. 6.

104. See Introduction 15, a; 19, b. Prudentius is not, strictly speaking, a classical author.

105. Bédier, Commentaires, pp. 316-17.

106. Cf. Jones, pp. 134-35, with reference to the Waltharius.

107. Delbouille, Genèse, p. 121. …

108. For Curtius, see note 100 above; Riquer, Chansons de geste, pp. 102-5. See also Manfreid Gsteiger, “Note sur les préambules des chansons de geste,” Cahiers de civilisation médiévale 2 (1959): 213-20; Zaal, A lei francesa, pp. 84, 91 Cf. Faral, La Chanson de Roland, p. 252: “La rhétorique ne tient ici aucune place.”

109. Faral, Jongleurs, p. 59, n. 2. Payen, Le Moyen Age, p. 125, citing Legge and Duby, suggests that the chansons de geste were primarily intended for bachelers. …

110. In real life eleventh-century warriors were at times quite satisfied to win campaigns without fighting a single battle. Smail, Crusading Warfare, pp. 140-56.

111. Horrent, p. 307. Composing a chanson de geste for different audiences was not unlike preparing a sermon to be preached to the educated at the same time as to the unlettered. On the latter art, see Guibert of Nogent, Liber quo ordine sermo fieri debeat (PL, 156, cols. 11-21), English translation in Readings in Medieval Rhetoric, p. 170.

112. See Introduction, 5. See also note 74 above.

113. Rychner, p. 14. Payen, Le Moyen Age, p. 125, agrees, but cautions against going so far as to term the chansons de geste “popular” literature. See also pp. 38, 126.

114. Smail, Crusading Warfare, pp. 166-67, warns historians against using literature as a source of realistic accounts of battles. Poetry, he notes, strives to be vivid and endeavors to make clear what is essentially a confused picture. See also Stephen G. Nichols, Jr., “Historical Illusion and Poetic Reality in the ‘Chansons de geste’,” French Review 43 (1969): 23-33; Payen, Le Moyen Age, pp. 61-62. Bowra, Heroic Poetry, pp. 476-507, distinguishes three main types of heroic poetry: primitive, proletarian, and aristocratic. The latter two types are characteristic of a society that has a cultured, lettered class. More refined aristocratic poetry exists only where the ruling class shares the interests and outlook of the ruled (pp. 478-79). He considers the Song of Roland to be in the latter category (p. 478). Chap. 4, pp. 132-78, is a useful study of “The Realistic Background.”

115. Jenkins, p. xxxv, citing Baist, on the organization of Charlemagne's army and trial by combat. For possible verbal archaisms, see Delbouille, Genèse, pp. 127-29, 133-34, 149-50. … René Louis, “La grande douleur pour la mort de Roland,” Cahiers de civilisation médiévale 3 (1960): 67, n. 20, referring to Lot, dismisses the possibility of deliberate archaism in the poet's designation of the boundaries of France. Cf. Guy Raynaud de Lage, “Les romans antiques et la représentation de l'Antiquité,” Le Moyen Age 68 (1961): 247-91; Raymond J. Cormier, “The Problem of Anachronism: Recent Scholarship on the French Medieval Romances of Antiquity,” Philological Quarterly 53 (1974): 145-57. On epic distortion of reality, see Eugene Vance, Reading the Song of Roland, Landmarks in Literature (Englewood Cliffs, N.J.: Prentice-Hall, 1970), pp. 19-20. Cf. Frye, Anatomy of Criticism, p. 127: “it is of the essence of imaginative culture that it transcends the limits both of the naturally possible and of the morally acceptable.”

116. Saint Augustine: On Christian Doctrine, trans. D. W. Robertson, Jr., Library of Liberal Arts 80 (New York: Liberal Arts Press: 1958), p. 38 (2.6.8); Robertson, Preface to Chaucer, pp. 53-54.

117. Le Gentil, p. 120: “Ce serait une erreur de s'appuyer sur une logique trop positive pour contester telle ou telle de ses décisions, pour parler ici ou là d'accidents, de déficiences ou d'interpolations.”

118. For discussion and bibliography, see Réau, I, 2: chap. 4; Frye, pp. 141-50 (see also p. 359, note to p. 141, line 21); Robertson, Preface to Chaucer, pp. 286-317; Vos, “Aspects of Biblical Typology,” chap. 1, pp. 1-25. Working independently of each other, Prof. Vos and I have on occasion used a similar approach but reached different conclusions. See my “Quelques nouvelles tendances,” p. 24, n. 29.

119. Réau, II, 1:244.

120. On the Eva-Ave connection, see Pierre Jonin, Les Personnages féminins dans les romans français de Tristan au XII^e siècle: Etude des influences contemporaines, Publication des Annales de la Faculté des lettres, Aix-en-Provence, n.s. 22 (Gap: Ophrys, 1958), pp. 444-45. Jonin cites examples in Peter Damian (d. 1072) and Wace, and notes that the association may date back to the eighth century.

121. Percy Ernst Schramm, Sphaira, Globus, Reichsapfel: Wanderung und Wandlung eines Herrschaftszeichens von Caesar bis zu Elisabeth II; ein Beitrag zum ‘Nachleben’ der Antike (Stuttgart: Hiersemann, 1958), pp. 72-73; pls. 28c, 46.

122. Réau, I:197; II, 1:83; II, 2:82.

123. Ibid., I:62; Robertson, Preface to Chaucer, p. 293.

124. Réau, I:63.

125. Robertson, Preface to Chaucer, p. 315. On these terms, see D. W. Robertson, Jr., “Some Medieval Literary Terminology, with Special Reference to Chrétien de Troyes,” Studies in Philology 48 (1951): 669-92; F. Douglas Kelly, Sens and Conjointure in the Chevalier de la Charrette, Studies in French Litterature 2 (The Hague and Paris: Mouton, 1966). On the importance of exegesis in medieval education, see Payen, Le Moyen Age, pp. 44-45.

126. Réau, I:62. Honorius was born c. 1080 and was a native of Regensburg in Bavaria, not Autun, as Réau suggests. Le Goff, Civilisation, p. 602. The concept was used metaphorically by Chrétien de Troyes in the Chevalier de la Charrette and in Yvain; see Gerard J. Brault, “Chrétien de Troyes' Lancelot: The Eye and the Heart,” Bibliographical Bulletin of the International Arthurian Society 24 (1972): 145, n. 4. For Rabelais's satirical use of the notion, see François Rabelais, Gargantua, ed. M. A. Screech, Textes littéraires français 163 (Geneva: Droz; Paris: Minard, 1970), Prologue, pp. 12-13.

127. See Morton W. Bloomfield's review of Judson B. Allen, The Friar as Critic: Literary Attitudes in the Later Middle Ages (Nashville: Vanderbilt University Press, 1971), in Speculum 48 (1973): 329-30, citing Anthony Nemetz, “Literalness and the Sensus Litteralis,” Speculum 34 (1959): 76-89.

128. Curtius, p. 48 (bibliography in n. 1); Robertson, Preface to Chaucer, pp. 340-41.

129. Robertson, p. 340.

130. A. Leigh Deneef, “Robertson and His Critics,” Chaucer Review 2 (1968): 205-34.

131. The expression is Robertson's (Preface to Chaucer, p. 92, n. 67), indicating that he has modified certain views he once held.

132. Robertson's interpretation of the Song of Roland is in Preface to Chaucer, pp. 163-71. … At the Fifth International Congress of the Société Rencesvals held at Oxford in 1970, Larry S. Crist and I presented convergent views of the Song of Roland. However, my colleague declared himself to be a far more orthodox Robertsonian than I. … My paper, “Sapientia dans la Chanson de Roland,” was published in French Forum 1 (1976): 99-118. I do not know what prompts Burgess, Vocabulaire pré-courtois, p. 13, to assert that “la notion de sen était étrangère au poète de la Chanson de Roland.”

133. Curtius, pp. 248 ff.; Frye, Anatomy of Criticism, pp. 89-91; Theodore Silverstein, “Allegory and Literary Form,” PMLA 82 (1967): 28-32; Paul E. Beichner, “The Allegorical Interpretation of Medieval Literature,” PMLA 82 (1967): 33-38.

134. Robertson, pp. 297-98.

135. Historia Karoli Magni et Rotholandi ou Chronique du Pseudo-Turpin, ed. C. Meredith-Jones (1936; rpt. Geneva: Slatkine, 1972), pp. 71-75 (1130); Horrent, pp. 87-94 (1145-65); Mandach, Naissance, 1:56-58, 149 (1125-1130); Frappier, Chansons de geste, 2 (1965): 124, note (1125-30). Meredith-Jones, p. 81, does not believe the author visited Spain. Mandach's elaborate theory concerning the early evolution of the Latin text, presented in Naissance, vol. 1, is conveniently summarized in vol. 2, Chronique de Turpin: Texte anglo-normand de Willem de Briane (Arundel 220), Publications romanes et françaises 77 (Geneva: Droz, 1963), pp. 13-14. The German scholar Adalbert Hämel devoted twenty-three years to the study of the Pseudo-Turpin. His edition of that work was published posthumously: Der Pseudo-Turpin von Compostela, eds. Adalbert Hämel and André de Mandach, Bayerische Akademie der Wissenschaften, Philosophisch-Historische Klasse, Sitzungsberichte, Jahrgang 1965, Heft 1 (Munich: Beck, 1965).

136. Meredith-Jones, p. 111. On the cult of Saint Facundus (= Sahagún) and Saint Primitivus martyred near this city, see Meredith-Jones, pp. 295-96; Réau, III, 1:485; Vielliard, pp. 6-9, 83, and nn. 1, 2.

137. Meredith-Jones, p. 119 (see also note, p. 300). Cf. the miracle of the red crosses, which appear on the shoulders of the knights in Charles's army who are to die the following day (pp. 146-47; see also note, p. 300).

138. Blaise, par. 484. Cf. the flowering staff associated with Saint Christopher (Réau, III, 1:305, 309) and the attributes of Aaron (Réau, II, 1:188-89) and Saint Joseph (Réau, III, 2:757).

139. See Introduction, 19, d. Turoldus uses espiet ‘spear’ and lance ‘lance’ interchangeably, although only the former is thrown (Foulet, Glossaire, s.v. espiet); cf. hanste (< Lat. hasta ‘lance’) ‘handle [of the spear]’.

140. Brault, “Le Thème de la Mort,” p. 236. …

141. Cf. the further extension constituted by the motif of Marsile's right hand and Ganelon being torn limb from limb, Commentary, 21 (v. 1903) and 47 [in “The Song of Roland”: An Analytical Edition, Vol. 1. University Park, PA: The Pennsylvania State University Press, 1978].

142. Meredith-Jones, pp. 111, 113.

143. See also Introduction, 11, d.

144. Meredith-Jones, p. 113.

145. See also Isaiah 11:5. Blaise, par. 441; Psychomachia, v. 52 (Prudentius, ed. and trans. H. J. Thomson, The Loeb Classical Library, 2 vols. [1949; rpt. Cambridge: Harvard University Press; London: Heinemann, 1962], 1:274-343); Delehaye, Les Passions des martyrs, p. 154; Robertson, Preface to Chaucer, p. 175, n. 1, citing Alain de Lille.

146. Rychner, pp. 128, 132-33.

147. Adolf Katzenellenbogen, Allegories of the Virtues and Vices in Mediaeval Art From Early Christian Times to the Thirteenth Century, trans. Alan J. P. Crick (1939; rpt. New York: Norton, 1964), chaps. 1 and 2.

148. Meredith-Jones, p. 113.

149. Blaise, par. 462.

150. Meredith-Jones, p. 135. In devotional tracts the enemy of Fortitude is often not Fear, but Sloth (Accidia), the virtue representing not physical strength and force, but a nobler quality. Tuve, Allegorical Imagery, pp. 84, 97, 133, et passim.

151. In his Anglo-Norman translation Willem de Briane substitutes another passage from Scripture; see Mandach, Naissance, 2:39 (text on p. 64, lines 478-79).

152. Meredith-Jones, p. 143.

153. Ibid., p. 143.

154. Ibid., p. 145.

155. Ibid., p. 145. Cf. the commentary on the Christian warriors who fornicated with Saracen women, p. 185.

156. Ibid., p. 195.

157. Brault, “Le Thème de la Mort,” p. 230, n. 21. …

158. See, for example, Introduction, 10, b, 1; Robert A. Pratt, “The Old French Sources of the Nonnes Preestes Tale (Part II),” Speculum 47 (1972): 653-54, 661-62; Marianne Cramer Vos, “Ganelon's ‘Mortal Rage’,” Olifant 2, no. 1 (1974): 21. It is pointless, therefore, to deny the Ganelon-Judas connection as do, for instance, Tavernier (see Vos, p. 21, n. 20) and John A. Stranges, “The Character and the Trial of Ganelon: A New Appraisal,” Romania 96 (1975): 354-56.

159. Das Rolandslied des Pfaffen Konrad, ed. Carl Wesle, 2d ed., Altdeutsche Textbibliothek 69 (Tübingen: Niemeyer, 1967). On the date, see Lejeune and Stiennon, 1:111-19; André de Mandach, “Encore du nouveau à propos de la date et de la structure de la Chanson de Roland allemande,” Société Rencesvals. IV^e Congrès International, pp. 106-16. For a summary of the published findings concerning Conrad's use of the Pseudo-Turpin, see Mandach, pp. 108-12, 116.

160. Fig. 43 [in “The Song of Roland”: An Analytical Edition, Vol. 1. University Park, PA: The Pennsylvania State University Press, 1978]. Lejeune and Stiennon, 1:124-25; 2: fig. 96 (the illustration shows a stylized olive tree). Lejeune and Stiennon, 1:124, asserts that the scene in Conrad takes place after the plotters have arrived at Marsile's court. I find no textual basis for situating this daylong stop (v. 1982: Si wonten da allen einen tach) at Saragossa. The arrangement of the figures in this illustration is not without a certain ironic parallel with the formula of Jesus among the Doctors (Réau, II, 2:289-91). One might even consider it to be an extension of the possible parody of the Journey of the Magi in the preceding drawing (Lejeune and Stiennon, 2: fig. 95; discussion in 1:124, with no reference, however, to the Wise Men formula; on the latter, see Gabriel Millet, Recherches sur l'iconographie de l'évangile aux XIV^e et XVI^e siècles d'après les monuments de Mistra, de la Macédoine et du Mont-Athos [1916; rpt. Paris: Boccard, 1960], figs. 36, 37, 38, 39, 67, 85, 86, 87, 95, 100, 101; Gérard Cames, Byzance et la peinture romane de Germanie: Apports de l'art grec posticonoclaste à l'enluminure et à la fresque ottoniennes et romanes de Germanie dans les thèmes de majesté et les évangiles [Paris: Picard, 1966], index, p. 319, s.v. Mages (cycle des), se rendent à cheval à Bethléem; idem, Allégories et symboles dans l'Hortus deliciarum [Leiden: Brill, 1971], pl. LXXVII).

161. On the concept of “poor” Judas, see Wayland D. Hand, A Dictionary of Words and Idioms Associated with Judas Iscariot: A Compilation Based Mainly on Material Found in the Germanic Languages, University of California Publications in Modern Philology 24, no. 3 (Berkeley and Los Angeles: University of California Press, 1942), pp. 303-4.

162. Matthew 7:16-18; cf. 12:33-37. Cf. also the “whited sepulchre” metaphor in Matthew 23:27-28; and the cup, clean outside, but filthy within, in Matthew 23:25. Locke, Quest, p. 106, nn. 20, 21. …

163. On companionage, see William A. Stowell, “Personal Relationships in Medieval France,” PMLA 28 (1913): 388-416; Bloch, Feudal Society, 1:154, 155, 169, 173, 236; Jones, pp. 114, 143.

164. Similarly, in the Pseudo-Turpin, Roland fights the giant Ferracutus, who is said to be de genere Goliath (Meredith-Jones, p. 147). The David-Goliath aspect of the Pinabel-Thierry duel is noted by Jenkins, p. xxxii.

165. The lines between chronicle, epic, and saint's life are not clearly drawn at this time. See Introduction, 2.

166. Gerard J. Brault, “Heraldic Terminology and Legendary Material in the Siege of Caerlaverock (c. 1300),” in Romance Studies in Memory of Edward Billings Ham, ed. Urban T. Holmes, Jr., California State College Publications 2 (Hayward: California State College, 1967), pp. 15-16.

167. See Introduction, 9. The tendency in the Pseudo-Turpin and the Rolandslied to explain symbols and clarify mysteries found in Turoldus's poem may be characterized as Gothic. See Eleanor Roach, “Les termes ‘roman’ et ‘gothique’ dans le domaine littéraire: Essai de définition,” Les Lettres Romanes 29 (1975): 63. The Pseudo-Turpin and Conrad's adaptation are frequently viewed as distortions of the French original; see, for example, Helmut A. Hatzfeld, “Le Rolandslied allemand: Guide pour la compréhension stylistique de la Chanson de Roland,” Cultura Neolatina 21 (1961): 48-56. The Latin and German versions may well be inferior to the French epic, yet each deserves to be judged on its own terms.

168. See Karl-Heinz Bender, “La genèse de l'image littéraire de Charlemagne élu de Dieu au XI^e siècle,” Boletín de la Real Academia de Buenas Letras de Barcelona 31 (1965/66): 35-39; Dufournet, Cours sur Roland, pp. 183-85. See also Campbell, Hero, pt. 2, chap. 3, 6, “The Hero as World Redeemer,” pp. 349-54.

169. Bédier, Légendes épiques, 2d ed. (1921), 4:456; Boissonnade, Du Nouveau, pp. 265, 281-85; History of the Crusades, 1:241; and, especially, Ernst R. Kantorowicz, The King's Two Bodies: A Study in Mediaeval Political Theology (Princeton: Princeton University Press, 1957), chap. 3, “Christ-Centered Kingship,” pp. 42-86; chap. 5, par. 3, “Pro patria mori,” pp. 232-72. Cf. Menéndez Pidal, pp. 241-62; R. Foreville, “La typologie du roi dans la littérature historiographique anglo-normande aux XI^e et XII^e siècles,” Etudes de civilisation médiévale (IX^e-XII^e siècles): Mélanges offerts à Edmond-René Labande (Poitiers: Centre d'études supérieures de civilisation médiévale, 1974), pp. 275-92; Le Goff, Civilisation, p. 183.

170. Kantorowicz, p. 48, n. 11 (translation mine); Cames, Byzance, p. 40, n. 69.

171. Cames, p. 40. See also Kantorowicz, pp. 61-78.

172. For fuller details, see Eugene F. Rice, The Renaissance Idea of Wisdom, Harvard Historical Monographs 37 (Cambridge: Harvard University Press, 1958), chap. 1, “The Medieval Idea of Wisdom,” pp. 1-29; Paul Archambault, “Commynes' saigesse and the Renaissance Idea of Wisdom,” Bibliothèque d'Humanisme et Renaissance 29 (1967): 613-32; Morton W. Bloomfield, “Understanding Old English Poetry,” Annuale Medievale (Duquesne Studies) 9 (1968): 5-25. On the related notion of contemptus mundi, see Payen, Le Moyen Age, pp. 70-71.

173. E l'arcevesque, ki fut sages e proz (v. 3691).

174. Faral, La Chanson de Roland, pp. 245-46. Introduction, 19, e, f, i, k. For the figure of a Carolingian king in the traditional guise of Sapientia enthroned in the New Jerusalem (as in the conclusion of the Psychomachia), see Paul Lacroix, France in the Middle Ages: Customs, Classes and Conditions (New York: Ungar, 1963), p. 349, fig. 298.

175. Meredith-Jones, pp. 221-29; Réau, I:154-62; Curtius, pp. 47-50; Katzenellenbogen, Allegories, index, p. 96, s.v. Arts, the seven liberal. See also Introduction, 10.

176. … Cf. Bédier, Légendes épiques 3:443 (referring to Roland): “comme il convient à un martyr, sa Passion est à la fois toute souffrance et toute joie.” The notion of Joy in the Midst of Suffering is found in the Beatitudes; see Matthew 5:11-12. Cf. 1 Colossians 24. …

177. See Introduction, 15, e. For a similar view, see Vos, “Aspects of Biblical Typology,” p. 89.

178. Brault, “Le Thème de la Mort,” pp. 229-35. For Charles as a Christ symbol, see William Wistar Comfort, “The Character Types in the Old French Chansons de geste,” PMLA 21 (1906): 338; Erich Auerbach, Mimesis: The Representation of Reality in Western Literature, trans. Willard R. Trask (Princeton: Princeton University Press, 1953), p. 101. The word martyr means ‘witness’ (in the New Testament, Gr. martus ‘witness of God’; Bloch and Wartburg, Dictionnaire, s.v. martyr), that is, one who confesses his faith. According to Christian belief, this type of confession is superior to merely unburdening one's sins to a priest or to God; Blaise, par. 109.

179. Guernes de Pont-Sainte-Maxence, La Vie de Saint Thomas Becket, ed. Emmanuel Walberg, Classiques français du moyen âge 77 (Paris: Champion, 1964), introduction. Cf. Le Gentil, p. 131:

     entre Oliver et Roland il y a toute la différence qui sépare le juste du saint. L'un proportionne ses actes aux simples exigences du devoir et ne voit dans l'excès que folie et orgueil; l'autre se croit toujours en deça de ce que Dieu demande ou espère. Olivier sera sauvé. Mais au ciel, plus encore que parmi les Francs de France, il cédera la première place à Roland. Peut-il même prétendre à la seconde?

180. Meredith-Jones, p. 111. The building metaphor appears in 1 Corinthians 3:9-15. … In the Pseudo-Turpin Roland dies, his arms crossed over his breast in a very explicit imitatio Christi (Meredith-Jones, p. 205). For illustrations of the hero in this attitude, see Lejeune and Stiennon, 1: pls. L, LX; 2: figs. 288, 293, 508. … In the Latin chronicle Saint Denis appears in a vision to Charles and informs him that those who died or are about to die in Spain for the Emperor's edification (Meredith-Jones, p. 219: Illis qui tua ammonitione et exemplo tuae probitatis animati in bellis Sarracenorum in Hyspania mortui et morituri sunt) will be absolved from all sin.

181. Meredith-Jones, p. 183.

182. This passage from John 12:24 is quoted in Conrad, vv. 7885-7888. For the Pseudo-Turpin, see Introduction, 10, a, 2.

183. Delehaye, Les Passions des martyrs, pp. 213-18. Cf. also the destruction of the idols in vv. 2585-2591 and Delehaye, pp. 215-16.

184. Matthew 7:13-14. The affection and admiration with which Roland's men view their protector precludes conceiving of him as a kind of Ishmael.

185. Mickel, “Parallels in Psychomachia and Roland,” pp. 451-52. See also Introduction, 15, a. …

Abbreviations and Frequently Cited Works

Bédier: La Chanson de Roland, publiée d'après le manuscrit d'Oxford et traduite par J. Bédier. Paris: Piazza, 1921. Glossary and Index of Proper Names by Lucien Foulet (= Foulet, Glossaire) in Bédier, Commentaires. The “édition définitive” (1937) has often been reissued.

Bédier, Commentaires: La Chanson de Roland commentée par Joseph Bédier. 1927; rpt. Paris: Piazza, 1968.

Blaise: Blaise, Albert. Le Vocabulaire latin des principaux thèmes liturgiques. Turnhout: Brepols, 1966.

Châteauroux: See Mortier.

CL: Classical Latin.

Conrad: Das Rolandslied des Pfaffen Konrad. Ed. Carl Wesle. 2d ed. Altdeutsche Textbibliothek 69. Tübingen: Niemeyer, 1967. For Modern French translation, see Mortier.

Curtius: Ernst Robert Curtius. La Littérature européenne et le moyen âge latin. 2d ed. Trans. Jean Bréjoux. Paris: Presses Universitaires de France, 1956.

d.: died.

E.: English.

FEW: Wartburg, Walther von. Französisches etymologisches Wörterbuch. Basel, Bonn, Leipzig, Tübingen, 1922-. In progress (22 vols; 2d ed. of vol. 1 [1922-28] numbered vols. 24 and 25).

Foulet, Glossaire: See Bédier.

Fr.: French.

Ger.: Germanic.

Godefroy: Godefroy, Frédéric. Dictionnaire de l'ancienne langue française et de tous ses dialectes, du IX^e au XV^e siècle. 10 vols. 1881-1902; rpt. Paris: Librairie des sciences et des arts, 1937-38.

Gr.: Greek.

Greimas: Greimas, A. J. Dictionnaire de l'ancien français jusqu'au milieu du XIV^e siècle. Paris: Larousse, 1969.

Harrison: The Song of Roland. Newly translated and with an Introduction by Robert Harrison. Mentor Book. New York and Toronto: New American Library; London: New English Library, 1970.

Heb.: Hebrew.

Horrent: Horrent, Jules. La Chanson de Roland dans les littératures française et espagnole au moyen âge. Bibliothèque de la Faculté de philosophie et lettres de l'Université de Liège 120. Paris: Les Belles Lettres, 1951.

Jenkins: La Chanson de Roland: Oxford Version. Edition, Notes and Glossary by T. Atkinson Jenkins. Rev. ed. Heath's Modern Language Series. Boston - New York - Chicago - London: Heath, 1929.

Jones: Jones, George Fenwick. The Ethos of the Song of Roland. Baltimore: Johns Hopkins Press, 1963.

Lat.: Latin.

Le Gentil: Pierre Le Gentil. La Chanson de Roland. Connaissance des lettres 43. Paris: Hatier-Boivin, 1955.

Lejeune and Stiennon: Lejeune, Rita, and Jacques Stiennon. La Légende de Roland dans l'art du moyen âge. 2d ed. 2 vols. Brussels: Arcade, 1967.

Lyon: See Mortier.

ME.: Middle English.

Menéndez Pidal: Ramón Menéndez Pidal. La Chanson de Roland et la tradition épique des Francs. Trans. Irénée-Marcel Cluzel. 2d ed. Paris: Picard, 1960.

Meredith-Jones: Historia Karoli Magni et Rotholandi ou Chronique du Pseudo-Turpin. Textes revus et publiés d'après 49 manuscrits. Ed. C. Meredith-Jones. 1936; rpt. Geneva: Slatkine, 1972.

MFr.: Modern French.

Moignet: La Chanson de Roland. Texte original et traduction par Gérard Moignet. Bibliothèque Bordas. Paris: Bordas, 1969.

Mortier: Les Textes de la Chanson de Roland. Ed. Raoul Mortier. 10 vols. Paris: La Geste Francor, 1940-44.

I. La Version d'Oxford (1940).

II. La Version de Venise IV (1941).

III. La Chronique de Turpin et les Grandes Chroniques de France, Carmen de proditione Guenonis, Ronsasvals (1941).

IV. Le Manuscrit de Châteauroux (1943).

V. Le Manuscrit de Venise VII (1942).

VI. Le Texte de Paris (1942).

VII. Le Texte de Cambridge (1943).

VIII. Le Texte de Lyon (1944).

IX. Les Fragments lorrains (1943).

X. Le Texte de Conrad. Trans. Jean Graff (1944).

OE.: Old English.

OFr.: Old French.

OHG.: Old High German.

OPr.: Old Provençal.

Owen: The Song of Roland: The Oxford Text. Translated by D. D. R. Owen. Unwin Books Classics 3. London: Unwin, 1972.

Paris: See Mortier.

PL: Migne, Jacques-Paul. Patrologiae cursus completus … Series latina. 221 vols. Paris: Migne, 1884-91.

Pseudo-Turpin Chronicle: See Meredith-Jones.

Réau: Réau, Louis. Iconographie de l'art chrétien. 3 parts in 6 vols. Paris: Presses Universitaires de France, 1955-59.

Ronsasvals: See Mortier.

Rychner: Jean Rychner. La Chanson de geste: Essai sur l'art épique des jongleurs. Société de publications romanes et françaises 53. Geneva: Droz; Lille: Giard, 1955.

Samaran: La Chanson de Roland. Reproduction phototypique du Manuscrit Digby 23 de la Bodleian Library d'Oxford. Edition avec un avant-propos par le Comte Alexandre de Laborde. Etude historique et paléographique de M. Ch. Samaran. Paris: Société des anciens textes français, 1933.

Segre: La Chanson de Roland. Edizione critica a cura di Cesare Segre. Documenti di filologia 16. Milan and Naples: Ricciardi, 1971.

Sp.: Spanish.

The Jerusalem Bible: The Jerusalem Bible. Gen. ed. Alexander Jones. Garden City: Doubleday, 1966.

Tobler and Lommatzsch: Tobler, Adolf, and Erhard Lommatzsch. Altfranzösisches Wörterbuch. Berlin: Weidmann; Wiesbaden: Steiner, 1925-. In progress (10 vols.).

Venice IV: See Mortier.

Venice VII: See Mortier.

VL: Vulgar Latin

Whitehead: La Chanson de Roland. Ed. F. Whitehead. Blackwell's French Texts. 2d ed. 1946; rpt. Oxford: Blackwell, 1965. Often reissued.