olifant.html

We need to study both the context of the Song of Roland and the mis-reading of that text. We are accustomed to viewing the Chanson de Roland as French. It is not usually argued that the Song of Roland is instead a Norman twelfth-century manuscript of an eleventh-century text about an eighth-century Carolingian event, although the evidence is manifest. Like the Iliad, this is a work that had both oral and scribal states.4 We first hear of the text as sung by the minstrel Taillefer at the Battle of Hastings in 1066, William of Malmesbury telling us, "Tunc cantilena Rollandi inchoata, ut martium viri exemplum pugnatores accenderet," and Wace stating:

The Roland prevailed as well in the other Norman territories, in Sicily where it is still sung today by street story tellers and acted out in puppet theatres and where its episodes adorn painted Sicilian carts, and in the ephemeral Jerusalem Kingdom of the Franks, of the French-speaking Norman Crusaders. 9 It thus appears that the Normans adopted the tale, embroidering it, palimpsesting it, to their own needs and then scattered it about the map of Europe, in England, in Sicily, in Outremer and even back home in Scandinavia.10 Nor is it mere coincidence that the Normans, upon their Conquest of England, set to work compiling their great Doomsday Book. It could be wise to see the Doomsday Book, the Bayeux Tapestry and the Song of Roland as related cultural artifacts, as contextual mirrors, all of which are about Apocalypse and Conquest, the religious Apocalypse being appropriated by the Normans to obfuscate, excuse and "legitimate" their material Conquests, harnessing the religion of Christianity - which was pacifist - to its opposite, to the uses of power, of the military, of the national state, using thrilling poetry to recruit crusading armies.11

Fredric Jameson in "Metacommentary" noted that we should step back and observe why certain modes of literary criticism came into being at various historical periods and Thomas Kuhn has noted how new concepts engender powerfully paralysing backlashing resistance to themselves. 12 If we can do that archeologically with the Song of Roland, deconstructing its past scholarship's rigor mortis, undoing our censorship of ourselves, perhaps we can find its truer twisted hermeneutics. That the Normanness of the text has gone largely unobserved has been largely due to the cultural milieux of these critics and scholars, Francophone philologists seeing it through that language's nationalism, as if colored by the anachronistic stained glass depictions at Chartres of that tale, 13 and Anglophone scholars being humbly in their debt. Its base text manuscript was noted and edited by French scholars, first at the Revolution, then on the eve of World Wars I and II, who inevitably perceived it in a Gallic context. 14 In what follows the Song of Roland will be restored to its Norman surroundings. To truly understand it, seeing past our anachronistic overlays, we need to shift our gaze from Chartres to Jumièges, from Gothic stained glass to Romanesque manuscript illumination, from French elegance to Norman power.

Originally, the Semitic Phoenicians invented the useful technological tool of the alphabet. The invention of language was now extended through time and space by way of the invention of the means for recording that language. Even the word "alphabet" represents the un-Greek Greek form "alpha beta" of the Semitic "aleph beth."

So, too, did the Hebrew peoples first invent the sense of the Book, the history of a peoples, enshrining a politico-religious identity. In turn, European tribes, coming into contact with this Mediterranean technology of memorializing power, capable of expanding and defining both the individual mind and its society, seized hold of it, much as today's nations seek to obtain, by fair means or foul, each others' advances in computer technology. This granted to the northern European tribes access to southern Mediterranean texts, generated in Israel, Egypt, Greece and Italy, which they pirated, appropriated and adapted to their own uses. Whether the Bible or the epic, these texts - and their textual communities - conveyed and passed down codes concerning civilization. 15 At Fiesole, for instance, we see Etruscan inscriptions written in the Phoenician runic alphabet which the Germanic tribes were to adopt prior to Christianity.

Christianity - the religion of women and slaves, heir to Hebrew, Greek, Syrian and Latin alphabets - arose in the midst of imperial/colonial oppression by Rome of Israel as a proto-Gandhian non-violent revolution and was pacifist. First Constantine, then Charlemagne, then the Normans, were to wrench Christianity from pacificism to militarism and to invent the Crusade. Where one culture seizes hold of another's texts there may well be disjunction, misreadings, a wrenching clash, resulting in powerful doublenesses and absurd plagiaries. Chivalry would--by means of a paradigm shift after the Oxford Roland version--attempt to make militarism use fair play again, reviving Christianity's ethical codes of the defense of the powerless, of widows and orphans, of peasants and priests, of the unarmed non-combatant civilian, and would frequently do this both in its rituals and in its myths, its literary texts, both oral and scribal, including those concerning Charlemagne and Arthur.16 That paradigm shift paradoxically partly came about because the Christians pirated the Islamic codes which they came to admire. We recall Saladin sending two fresh horses to Richard Coeur de Lion through the lines of battle when his first was slain under him.

In the study of medieval manuscripts one becomes familiar with the palimpsest; where, to conserve valuable parchment, one text is scraped away and another written in its place, the first leaving traces of its opposing message through the second. This can serve as a metaphor for the way in which, on abstract levels, many literary and cultural texts - and especially the artifacts of the Anglo-Saxon and Norman Romanesque cultures - function, having an entwined doubleness and opposition of meanings. Let us imagine, for instance, a Christian Book of the Apocalypse superimposed uneasily upon a pagan Viking saga about blood feuds and piratical long ships, of raiders looting Christian monasteries of their Gospel books, and use that as a working image or hypothesis for the Chanson de Roland. 17 The first text is of a spiritual battle, an allegory, a psychomachia; the second of all-too-literal carnage and bloodshed, in a sense, super ego and id. In Bede's History of the Church and English People, in fact, we hear the story of abbess Hilda encouraging the cowherd Caedmon to so combine the conventions of the pagan sagas for the retelling of Hebrew texts, his translating these into the folk style of oral-formulaic epic and then her having these be made into scribal texts, weaving together the Germano-Scandinavian with the Judaeo-Christian. Possibly the "Dream of the Rood" too was a Caedmonian composition, the Ruthwell Cross consciously both Yggdrasil and Rood with its lignum vitae, upon which Odin had to hang for nine days and nights to learn the runes of life. 18 These monuments, texts which contain artifacts, such as Roland's Olifant and ivory chess pieces, and artifacts which contain texts, such as the Ruthwell Cross and the Bayeux Tapestry, are crucial to communicating across warring worlds and clashing codes.

The year 1000 had reverberated through European culture as if from the trumpet blast of Roland's Olifant. It echoed back into the past, summoning up Charlemagne from the eighth century. And it echoed into the future, to the Battle of Hastings of 1066 and other events of the Norman expansion and even into the pages of Dante's Commedia .19 The Song of Roland and the great Romanesque tympani of Doomsday with their angel trumps are as if Europe's working through of that time of anxiety, fragmenting and decomposing it across that Continent. The great Norman energy unleashed upon the map of Europe following the year 1000 attempted to legitimize and validate itself through propagandistically imposing upon itself the palimpsest of the Carolingian era, to which it had not belonged.20 Even the Bastard William's Coronation upon Christmas Day, 1066, consciously mirrored and appropriated those of Charlemagne on Christmas Day, 800, of Clovis on Christmas Day, 496. Thus the Chanson de Roland used space, Spain, France, England, Sicily, Babylon, Jerusalem, and time, combining geography and history for literature.21 So must we in our study of the poem.

Henri Focillon tells us that on Pentecost Sunday in the year 1000 the young Emperor Otto III, recently crowned in Rome, stood in the old palace chapel of Aachen gazing at the remains of Charlemagne, whose forgotten grave had been found at his command. The body of Charlemagne lay before him in an ancient sarcophagus, a cross of gold around his neck.22 The new, only recently Christianized, northern peoples were attempting to place themselves into a sense of Empire. And Charlemagne, himself from such northern origins, was the logical intermediary in this task of reconciling north and south, the Atlantic and the Mediterranean, the Germanic and the Latin. Charlemagne was a link forged between the two alien worlds of Islam and Christendom and was therefore to be exploited by the Norman propaganda machine. The Chanson de Roland would resurrect that apocalyptic Empire of the Franks. A major reason for this need was that when Islam arose it had rapidly conquered two parts of Europe, Spain and Sicily. The Normans were able to conquer back the Mediterranean island of Sicily and govern it as their kingdom, richly amalgamating Islamic architecture in Christian edifices. The Norman crusaders, likewise, initially won back Jerusalem and within that Camelot-like state, made use of the Arthurian Matter, appropriating its tales of Celts holding out against Danes for Normans holding out against Saracens. 23

Henri Focillon also tells us that the supposed Doomsday of the year 1000 itself had been spent rather quietly, that the millennarian sense was far stronger before and after that year than during it. Prior to it the pagan Norsemen had settled in France around the Seine, making that region Normandy. In the following century, the eleventh, they had become Christian and French-speaking. The eleventh-century Jumiéges Gospels contain a magnificent illumination of St. Michael combating the devil fashioned in lurid imperial purple. 24 It could well accompany the text of the Song of Roland. St. Michael was an important figure in Norman culture, appealing to the Viking military ethos, whether in Normandy, as at Mont St. Michel (mentioned, Chanson , CX, 1428, as "seint Michel del Peril"), or in Sicily. It is possible that the Norman use of St. Michael was appropriated by them from that saint's great popularity in the Byzantine Empire, a similar borrowing being that of St. George. Then, in 1066, the Normans invaded England, winning the Battle of Hastings. Anglo-Saxon peace-weaving women were commissioned to embroider the Bayeux Tapestry, probably for William's brother, Odo, Bishop of Bayeux. The cathedral of Bayeux was built by Odo to celebrate the Virgin and dedicated July 14, 1077, by Jean, Archbishop of Rouen, and Lanfranc, Archbishop of Canterbury, in the presence of Duke William. That embroidery is much like both the Old English Beowulf and the Old French Song of Roland in its Romanesque chiastic panel structuring. 25 Both the Song of Roland and the Bayeux Tapestry name a Turoldus, probably not the same person, but nevertheless an interesting coincidence. All three artifacts, whether they be textile or text, are about regal politics' struggles for power. All three place those regal political struggles within a framework of religious determinism. They do so because the presence and use of the Apocalypse conveniently validated barely legitimate claims. 26

To yoke together the concepts of Charlemagne and Apocalypse, as the Song of Roland does, was especially convenient for the Norman world. On the one hand Charlemagne provided for them a model for Christian imperialism, on the other, the Apocalypse gave them an attractively militaristic liberationist and millennarian argument for their expansionism. They subverted liberationism for the ends of imperialism. As Norman vassals to the King of France they could use the Chanson de Roland to make their feudal overlord acquiesce to their restless conquests - now guised in the sanctity of Crusade - and to justify speaking of themselves as "Franks." They were pirating a linguistic, cultural, historical identity, conquering not only space but also time. Joseph Strayer has even argued that the Crusades had to be invented by the Church to handle the energies of the newly-Christianized northern marauding and blood-feuding tribes. The Crusades' official beginning at the Council of Clermont in 1095 proclaimed by Pope Urban II certainly made conscious "legitimating" use of Charlemagne. 27 Similarly Vikings became the honored members of the Byzantine Varangian Guard - reverberating into the future with White Russians, descendants of Vikings, acculturating Byzantine politics to the Kievan state. An example of such displacement, such cultural wrenching, is a stone lion discussed by Augustus Hare in the Arsenal in Venice, upon which are carved Scandinavian runes. The lion was once in the Piraeus harbor of Athens, the runes carved upon it by Vikings conquering Byzantium boasting about that forcible appropriation of power. The stone lion was next displaced by Venetians in 1687 - looting the Greek world for the Latin one. 28 These extraordinary dislocations and transcultural manifestations are the hallmark of Viking Norman expansionism through time. Essentially piratical, they knew how to garb that energy and greed with "legitimating" borrowed/stolen myths concerning power, making slaughtering, lucrative conquest appear to be as if an allegorical Apocalypse and Crusade. 29

Similar juxtapositions occur in the sphere of textiles as they do in that of texts, where weaving and embroidery delighted in opposing symmetries, garments for the human body needing to match, with art, nature's symmetry of left and right limbs, just as much as the fabric itself conveying these designs was created from the oppositions of warp and woof. Such chiastic patterns worked equally as well for Beowulf as for the Chanson de Roland and both these poems can benefit from comparison with the oxymoronic war memorial/peace-weaving Bayeux Tapestry - such a tapestry as would have originally been displayed in pagan mead halls but which came to be hung about the walls of the cathedral of Bishop Odo of Bayeux. All three artifacts were products of Scandinavian cultures assimilating themselves to those of the Mediterranean, of pagan societies in the process of becoming Christian while at the same time resisting that foreignness. The conquered Anglo-Saxon peace-weaving women wove and embroidered the tapestry - as both protest and celebration - for and against their Norman male conquerors, in the margins of their textile which is also a history text showing forth the horrors of war wrought upon defenseless women and children in such episodes, such crimes, as rape and arson. 30 These were the propaganda/documentary mirrors of society - just as much as are Mash and Apocalypse Now our Iliads and Aeneid s, televised myths and fun house mirrors strangely reflecting ourselves. To read these texts, decoding their diplomatic, wartime ciphers, can aid us in understanding our own hermeneutics of power. These binary doublenesses gave these texts and textiles complex and twisted meanings. 31

The Bayeux Tapestry, in fact, is contemporary with the singing of the Song of Roland, though it is written with "Dick and Jane" Latin, not Old English or Old French. I would tell my students, to whom I showed slides of the embroidery, to imagine Roland and Oliver looking like its figures and to be envisaged as wearing such chain mail, such helmets with nasals, such large kite shields, and wielding such heavy swords. Similarly the panel structuring of the Bayeux Tapestry presents the symmetry of the banquet by Harold at Bosham (complete with olifant drinking horns), that of William at Hastings, the two mirror-reversed events enveloping the central actions. There is even a pun in the text (but only in English), where the soldiers "hasten to Hastings," "ETHIC:MILITES:FESTINA[ ]VERVNT:HESTINGA:" Tragically, the Anglo-Saxon forces, unhabituated to fighting on horseback with the stirrup to hold the warrior in place, lose to the superior technology of the Normans, the Tapestry carefully and chiastically again showing that difference with the horses being shown on the long ships crossing the Channel. 32

Winchester and Fleury were in constant contact. In the latter abbey was a playbook, Orléans MS 20l, that may have come from Norman England and which contains liturgical dramas and plays of saints' legends, some showing Byzantine influence. One of these is the Filius Getronis, the story of a Christian royal child, kidnapped by a wicked pagan king and forced to serve as his cup-bearer. On one side the boys' parents, the King and Queen of the Christian realm, dispense charity, bread and wine, to children, on the other side the wicked, homosexual king, with no spouse, demands that the captured child bring him wine and fruit, while above them on a pillar stands the statue of the naked, pagan idol representing Apollo which the child refuses to worship. St. Nicholas rescues the prince, named Adeodatus, carrying him from the northern left side to the southern right side. The carefully contrived, yet opposing symmetry, is typical of this period. On the one side is Christian charity, on the left, pagan cupidity. The drama, and also a tapestry of this tale, insist upon this reading of its heraldic use of space. Similarly Fern Farnham and Karl Uitti have taught us to see such chiastic panel patterns in the text of the Chanson de Roland as Romanesque, as Romanesque as are the Doomsday tympani with their oppositions of the saved and the damned, the "Chrestien" and the "Paien." 33 We could argue further that they are even more characteristically Norman, having to do with the juxtaposition, the conversion/resistance of Christian and pagan worlds.

Why do we have a Norman poem - in French - about Spain? Spain, outside the Norman sphere of influence and never to become part of it, had earlier been a Visigothic Christian state and had become overrun by Muslims, leaving only a narrow corridor to the north remaining in Christian hands, Asturias and Galicia, which struggled to reassert their dominance in a Reconquista. Christianity, under seige in Spain, had responded to Islam in several ways. Isidore, the Bishop of Seville, before the Conquest by Islam of his city, had treasured up both Christian and pagan learning in his vast encyclopedia, the Etymologiarum. Then Beatus of Liebana, under the threat of Islam, wrote a commentary to the last book of of the Bible, about Doomsday, the Book of Revelation, the Apocalypse. The Apocalypse, written by St. John on the island of Patmos, in exile, gave a political allegory of Roman imperialism and Jewish liberation, thinly disguised as the cities of Babylon versus Jerusalem and cast as a spiritual battle that would take place at the end of time in which St. Michael would combat the devil. Beatus' 776 textual commentary to the Apocalypse was to be richly illuminated in the succeeding centuries in the Christian north and west in Mozarabic style, from the eighth century, and both before and after the year 1000, two of the manuscripts being associated with Silos and Gerona.34 Next, the Apocalypse as illuminated text was to be taken up by the Anglo-Normans through the thirteenth century. The Apocalypse had already shaped Augustine's City of God, reputedly Charlemagne's bedside reading. The Apocalypse, like Achilles' shield in Iliad XVIII and Dickens' Tale of Two Cities is a tale of two cities. One is at war against the other: one is Babylon (actually Rome in this instance), the profane city of man; the other Jerusalem, "Vision of Peace," the heavenly city of God.35 The metaphorical idea within that text of the miles christi, the "knight of Christ," would become carnal and result in carnage with the Crusades.36 Behind the Christian Apocalypse lay powerful and visionary Hebrew texts, which it copied, concerning the Babylonian captivity.

At the same time the story of Saint James, Santiago Matamoros, the Moorslayer, grew up, Beatus having initially proclaimed that it was St. James who brought the Gospel to Spain, his body being conveniently found thereafter, in 813, at Compostela. We shall even have later apocryphal accounts of Charlemagne's pilgrimage to that shrine, led there by the stars of the Milky Way, the camino de Santiago - that shrine not existing in Charlemagne's day - as being the reason for his invasion of Spain. (In fact, it was Spanish Saracens who requested Charlemagne's aid in 777.) The 1140 Pseudo-Turpin text in the pilgrim's guide found in the Codex Calixtinus gives us this material. Pilgrims flocked to the wonder-working, thaumaturgical shrine. Stories were told about Santiago, that he had been beheaded in Jerusalem but his body had then been placed in a stone boat and had drifted through the Straits of Gibraltar coming to the north coast of Spain, and had then been brought inland to Compostela in a cart drawn by oxen, like the Ark of the Temple. Later he was said to have appeared at the Battle of Clavijo, mounted on a white Apocalyptic horse and that he slaughtered, while in his garb as a pilgrim with cockle shell on hat and scrip, thousands of Moors with his blood-drenched sword.37 From this he is known as Santiago Matamoros, Saint James the Moor Slayer. 38 This James was even said in the legends to be the brother of Christ. At Santo Domingo de Silos, on that pilgrim route, cloister sculptures showed scenes for the monastic liturgical dramas of the Visitatio Sepulchri on Easter Monday with the soldiers sleeping at the Tomb with huge kite shields - like those in the Bayeux Tapestry - and of the Officium Peregrinorum of Easter Monday with Christ presented anachronistically as a Compostela pilgrim with the cockle shell upon his pilgrim scrip.39 Scholars have suggested that the Chanson de Roland was partly created and propagated by Cluniac monks along the Compostela route who sang it to these pilgrims journeying to Santiago di Compostela, thereby encouraging the Reconquista.40

Pilgrims were to be peaceful, canon law forbidding them to carry arms and surrounding them with protection similar to that given in Greece to the stranger, the medieval pilgrim to be treated as if Christ in disguise, the Greek exile being under the protection of Zeus. Both the Christian pilgrim and the Greek exile could be murderers expiating their crimes, their pilgrimages being penitential, a peripatetic prison. In their distinctive garb they had the right to cross even between warring armies and were frequently for this reason the letter carriers of the Middle Ages, just as during World War II the neutral Red Cross conveyed letters between prisoners-of-war and family members. Pilgrimage existed - and co-existed - in a dialectic with Crusade, as if each were the other's anti-model. 41 The Crusades were preached, using as justification the need to protect unarmed pilgrims, and military, monastic orders were formed for this purpose, such as Jerusalem's Knights Templars and Knights Hospitallers and Spain's Knights of St. James.

The Codex Calixtinus, the pilgrim guide to St. James, spoke of the great danger to pilgrims posed by travel through the Pass of Roncesvalles, in Basque territory. These were the people who probably murdered the historical, rather than the literary, Roland in Charlemagne's rear guard for the sake of the army's baggage and loot from Spain. They also murdered pilgrims. The Pass of Roncesvalles was, for pilgrims, a Valley of the Shadow of Death, reminding them of David's Psalm. Dante is to echo it in the beginning of the Inferno, "Nel mezzo del cammin di nostra vita mi ritrovai per una selva oscura." He did so because his teacher, Brunetto Latini, before him, had written an allegorical dream vision poem, the Tesoretto, which had begun with the historical reality of his exile from Florence being told him in the Pass of Roncesvalles, so that he then, deeply sorrowing, lost his way and took another road through a different wood. In the Song of Roland we hear of the high mountains, the shadowy valleys, the dark rocks and terrifying inclines.

The exigencies of such politics required that the new Norman culture, so recently itself a pagan one, displace that sense of self upon another pagan entity - and, with conversion fervor and energy, denigrate it. It projected its own dark side, its shadow of death, upon the Saracens. (The Homeric text of the Iliad, more honestly, had spoken of that darkness as descending upon Achaian or Danaian at the time of their individual dying.)45 The Franco-Spanish mountain pass, that crossing over, liminal place, that boundary, that margin, was thus used for a magnified psycho-drama of a culture in a looking-glass war, really with itself, though seemingly with the evil empire of the other. Instead of its own former Viking paganism the Norman world thus attacked the Saracen one, justifying itself through Crusade - itself literally a crossing-over word. Similarly had the earliest English poem, the "Dream of the Rood," made use of the pagan hanging of Odin upon Yggdrasil for its runic crossing over to the Crucifixion of Christ, Odin and dreamer, Cross and Christ, each mirroring the other.

The Chanson represents the pirating of the language, the religion and the history of the Franks by the Normans. Between a paragraph and an epic, between Eginhard's brief mention of Roland, Prefect of the March of Brittany, "et Hruodlandus Britannici limitis praefectus" (Celtic Britanny adjoining Scandinavian Normandy), and his death in the Pyrenees at the hands of the Gascons or Basques and the Chanson de Roland 's epic rendition of that event - at the hands of the Saracens - much has occurred to falsify history and to engender art. We know that the historical campaign by Charlemagne in Spain was at the request of the Saracens. We know that Charlemagne's relationships with the Muslim world were not chauvinistic, Saracens having requested his aid in 777 and Haroun al-Rashid in 801 presenting to Charlemagne the elephant from whose tusks at its death were made priceless artifacts. Let us, for now, compare and contrast the two accounts, the first of horn, the second of ivory.

Eginhard wrote his Vita Karoli between 817 and 830. He described, for the year 778, the day, August 15, the following:

The Song of Roland, the Chanson de Roland, as its name implies was an oral epic, a chanted narration, sung before audiences of soldiers and pilgrims, at courts and in market places, and handed down from generation to generation of such singers before being written down in England after the Norman Conquest. Such oral epics have flourished in many cultures: Homer's Iliad and Odyssey in Greece, Beowulf in Anglo-Saxon England, epics sung today in Yugoslavia and elsewhere. Scholars, such as Albert Lord and Joseph Duggan, who have studied these works find that they are somewhat different from literary texts that are composed scribally. What characterizes them is a great deal of repetition, the repetition of phrases, and even whole stanzas or laisses as here, of speeches, of conventional clichés, of formulae. What also characterizes such oral formulaic poems is that they are not linear in the way that writing is, going from left to right, from A to B to C; instead they are symmetrical, chiastic, like woven designs, like embroidered ones, like music, like parentheses, like a cross section drawn through the ripples in a pond from a cast stone, like onion skins or like Chinese boxes, their panel structuring being A,B,C,D,C,B,A. Their poetic meter or measured line helped the poet remember how the next part of the text was to go. In the Chanson de Roland the AOI that ends many laisses or stanzas may be a refrain the jongleur both sings and strums upon his harp. It may function like punctuation in a written text, giving him time to pause and assemble in his memory the next stanza in the poem. This is poetry to be heard by the physical ear, not seen by the physical eye. Probably the best way to imagine it would be as a jaunty, jazzed up version of chant, as cantastorie - and the passage concerning the shadowy valley should recall David's psalm about the valley of the shadow of death: the Song of Roland sung in the people's language of life and death, blood and guts, in the lingua franca, in Norman French; the psalm sung in the solemn, eternal, official elite language of the Church, in Latin, in Gregorian chant. 54

The text of the Song of Roland is both a barbaric thing and one of great artistry. It is filled with lies, with military and religious propaganda. At the same time it is carefully contrived and symmetrically balanced. In it, for every action there is an equal and opposite reaction, a polarity that works like a binary computer, in itself, an artifact and tool whose original conception came from the cards for the jacquard weaving loom. The Roland comes more from oral, pre-literate culture, which, in sophisticated chiastic ways, presents a careful, self-referential design. With Norman culture is its conscious opposition to that of the Saracen, reflected, tailors' mirrors' fashion, in the poem in the oppositions of Roland and Marsile and between the world of Charlemagne and that of Baligant, their liege lords. That perception, where "Paien unt tort e Chrestien unt dreit" (LXXXIX. 1015), shapes both the Song of Roland and the illuminated Apocalypse manuscripts of Spain, England and France. Both the Iberian and the later Anglo-French manuscript illuminations stress the opposing symmetry, between Jerusalem and Babylon, the mirror-reversing quality, of their text and structure.

The text replicates the politico-religious map of the Mediterranean as its blueprint paradigm. The poem is poised between two chess game halves: "France dulce, la bele"; "clere Espaigne, la bele." A major aspect of the Apocalypse tale that lent to the Chanson de Roland an even greater immediacy than it might have otherwise had was that in the Middle Ages Babylon was thought to be the old city of Cairo in Egypt, the head of the Muslim Empire to be the Soldan of that Egyptian Babylon--as is Baligant in the Chanson. In contradistinction to Baligant of Babylon/Cairo, the Roland stresses Charlemagne as from Jerusalem-like Aix-la-Chapelle, Aachen. 55 Christian pilgrims quite deliberately recapitulated both the Exodus and the Return from the Babylonian Captivity at once by journeying to Alexandria in Egypt, then setting forth from this Egyptian "Babylon," next journeying through the Sinai Wilderness, to attain Jerusalem.56 This helps explain St. Louis' disastrous crusade strategies where he sought to conquer Jerusalem, unsuccessfully, by landing in Egypt and Tunis to do so.57 Thus these two Biblical episodes were seen geographically to be replicated by this medieval pilgrimagings, all as if a palimpsest upon each other - and in the Chanson de Roland's text. In fact, the Apocalypse distorted Ezekiel, having the Babylonian Empire of the Tigris and Euphrates rivers become instead a code name for the Roman Empire of the Tiber; then a further lie was worked into the tapestry, having Babylon become next the code name for the Islamic Empire as situated on the Nile. The map and paradigm of the text of the Chanson de Roland is that of the Mediterranean world the Viking one would seize. It presents in the text an imperialistic codification and abstraction of the Mediterranean world--and carefully plots its action upon that map as if in a Pentagon strategic session. 58 To do so, the Chanson de Roland both appropriates, projects and distorts its own Norman otherness and bipolarity between its Christian and pagan selves into a Manichean duplexity between "Chrestien" and "paien," of French Christian and Moorish Saracen as war game, chess game and psychomachia.

To explain this poetic structuring further we need to examine other aspects of Romanesque culture. That society was feudal. At a king's coronation his vassals placed their hands between his, our modern gesture of prayer, while kneeling. Then they rose and king and vassal kissed each other. The one gesture signified obedience and subservience, the other, paradoxically, love and equality. Together they represented loyalty. A substitute for the hand could be the leather or textile glove. The left hand was sinister, bastard, evil; the right just, proper, true. Therefore we have treacherous Ganelon dropping the right glove when Charlemagne hands it to him, XXV. We also have Ganelon participating in the treason of the oath upon the relics in his sword Murgleis, XLVI, which is not unlike Harold's treacherous oath in the Bayeux Tapestry. To further emphasize that act Marsile in turns swear an oath to Ganelon while seated on an ivory faldstool (" Un faldestoed i out d'un olifant," 609) upon, not so much a Koran as a book of scriptures of Mahune, Tervagan and Apollin. Then Marsile will literally lose his right hand just as Charlemagne loses his metaphorical one, that of Roland. The Amiral Baligant of Babylon will send to King Marsile of Saragossa a golden glove for his right hand - too late, 2677-8. To counter Ganelon's treachery will be the action Roland takes when dying - of profering up his right hand glove to God, " Sun destre guant en ad vers Deu tendit," 2365, 2373, 2389, repeated as with the Olifant's trumps of doom, three times for emphasis.

In this period the banner, the sign in textile, was a crucial part of the battle. It was not to be captured for that signified defeat. There were royal standards, banners which represented, like the glove, the person of the king. Pagan Vikings - and even Christian Anglo-Saxon kings - fought under a dragon standard. There was also the Crusaders' banner, mentioned in our text, the Oriflamme. The Song of Roland tells us it was St. Peter's banner given to Charlemagne by the Pope.

The Song of Roland is about Charlemagne. It is a vast epic about a battle waged against Saracens, Infidels, Muslims, by valiant Christian knights, among whom are twelve Palatinate counts, twelve peers, the douze pers, for whom Roland is as Christ, God the Son, Charlemagne, God the Father, and Ganelon, Judas. Joseph Duggan reminds me that the Roland is Charlemagne's son by incest with his sister Gisele in the Karlmagnus Saga branch I and the Oxford Roland's reference to St. Giles, lines 1095-98. Roland's death is like Christ's with an earthquake and an eclipse, like the Second Coming, like the Apocalypse, like Doomsday. The Olifant's blast is the Doomsday Trump - such as we see blown in the Autun tympanum - secularized from theology to history. A similar patterning occurs in Beowulf where its hero is abandoned by his twelve thanes, only Wiglaf staying with him during his dragon fight. The Song of Roland has a further typological patterning. Aachen was Charlemagne's capital, Aix-la-Chapelle, where he built a replica of the Church of the Holy Sepulchre, thus localizing Jerusalem on the border of France and Germany, making the capital city of his empire be as if Jerusalem.62 This city dominates the right side of the poem, its good side. This side of the poem, with the exception of Ganelon, is of the City of God in the Apocalypse and in Augustine.

The other side is of the evil empire, its diabolical parody.63 The Saracens worship three idols, in infernal trinity of Mawmet, Tervagant and Apollyon. Baligant even vows to make golden images of these three for their aid, 3490-94. We witness Marsile and Bramimonde destroy their (non-existent) idols of Apollo, Tervagan and Mahumet, CLXXXVII. Then, after Baligant's defeat, the Christians tear down the synagogues and mosques of Saragossa and smash the (non-existent) idols, 3660-65. Their cities of man are Zaragossa and Babylon, again shaped by Apocalypse and Augustine. Roland and Charlemagne are countered by the Moslem rulers, Marsile and Baligant. Even Marsile has a further decomposing nephew, Aelroth (an Anglo-Saxon sounding name that), to mirror Roland as nephew to Charlemagne, whom Roland kills. Interestingly, to stress Baligant's great age he is noted to have lived as long as Virgil and Homer, epic poets both: "le viel d'antiquitet, Tut survesquiet e Virgile e Omer" (2615-6), though this is stated to place him into the pagan era rather than to refer to the Aeneid or the Iliad./Uitti, p. 91./ Joseph Duggan has sensed that Baligant, though not historical, is essential to the text./P. 105./ A careful symmetry is worked out: Charlemagne's sword "Joyeuese" with its relic of the True Lance in its hilt is countered by Baligant's sword and lance "Precieuse" and "Maltet." The French battle cry, "Mon Joie" is echoed by the Saracen battle cry of "Precieuse." It is again as if we are witnessing a game of chess in which equal and opposite pieces of black and white or red and white are matched up against each other as indeed in the text we witness Christian knights playing at the Saracen game.

Let us return to the infernal trinity of idols. Islam, like Judaism, strictly abides by the commandment of Moses against worshiping any graven image. Medieval Christianity, likewise, said it did not worship idols--though it did permit images and statues within architectural niches. The Chanson de Roland imposing the infernal trinity of idols upon its Muslim characters projects upon them its own worst fears about itself. To speak of Muslims as worshiping idols, statues on pillars, of Mawmet, Apollyon and Termagant, is absurd. Not only did they not do so; they are rigorously unitarian: "There is no God but God and Mohammed is his prophet." But instead of reality we witness the scene where Marsile lies dying, from the loss of his arm, and his wife in her rage smashes their idols. This is preparation for her conversion to Christianity - when she will take the name of Juliana, the name of a marvelously forthright early Christian saint and martyr who told the devil in no uncertain terms to go to hell. Medieval art remembered the Roman manner of state sacrifice, to idols of naked gods or of togaed emperors upon tall pillars - though the Renaissance conveniently forgot this in its desire to erect free-standing statues such as Michelangelo's David. The Middle Ages considered such free-standing statues as against Mosaic Law: "Thou shalt not make any graven image." 64 A manuscript illumination of Augustine in Carthage shows such idolators kneeling before pillars upon which are statues of naked gods and goddesses, behind each one a fluttering demon with bats' wings.65 Saracens in this poem worshiping such idols are seen as like them, as themselves quasi-demons - in accordance with the pilgrim/crusading psalm, 113, In Exitu Israel de Aegypto, "They that make them are like unto them." This demonizing is chauvinistic propaganda. But students find the references to the Saracens' jeweled helmets splendid. And, every now and then, there are glimpses of the Iliad's good sportsmanship, where the text can pause to praise a Saracen knight.

The text begins with Charlemagne seated on his golden faldstool in an orchard beneath a pine tree, stroking his long white beard, chiastically mirror-reversed by the infidel Marsile lying on blue marble. The Saracens plot treachery under the appearance of peace, complete with olive branches. Ganelon is sent to mediate the agreement - against his will - and plots along with the Saracens against his liege lord Charlemagne out of his hatred for his stepson Roland - who is Charlemagne's nephew and right hand man. The problem here is of jealousy, like that of Cain for Abel. We know what is in store because Ganelon fumbled and dropped the right-hand glove which Charlemagne gave him to take to the Saracens. Ganelon plots the fall of Charlemagne's right hand, Roland. The messages he relays are not true ones - they are twisted distortions of Charlemagne's words designed to make Marsile mistrust the Christian Emperor. To counter Roland, Charlemagne's nephew and the twelve peers, Marsile's nephew steps forward and brings with him twelve peers, the first being Marsile's aptly-named brother, Falsaron. Later the nephew of Marsile, Aelroth, will be killed, likewise Charlemagne's nephew, Roland, and Marsile also physically loses his right hand.66These scenes are like the photographic negatives of the other's positive images. The text is strongly, heraldically, coded. It is a tissue of lies and truth, assymetrical reality made over into the symmetry of embroidery. When we come to the battle itself in the Pass of Roncevalles we have Oliver counseling Roland to blow the Olifant, the large ivory horn named "Elephant." Roland proudly - but not wisely - refuses to do so until it is too late. Now Oliver is a fictional character added to the story, although Roland is historically true. It is odd that Oliver and Olifant are so similar as to their names. Usually Oliver is thought to be associated with "olive," like those branches of olive of the initial treacherous peace embassy. There may also be an attempt in the language of the poem to create of Oliver a lying opposition to Roland's truth.67 Earlier, we had met Marsile at his second counsel sitting on an ivory faldstool. Ivory, we have noted, was considered a treacherous kind of material for it actually punned in Greek with the word for deceit which is why Homer and Virgil speak of gates of ivory, of fiction, of lies, and of horn, of fact, of truth.68 The text allegorically presents Roland as "preux," brave, proud, with utter integrity, who needs that foolhardiness to be tempered with the "sagesse," the sagacity, the cunning, the wisdom and the experience of Oliver. The two are great friends, they are each others' alter egos. The one is straightforward courage, the other the devious intelligence necessary for survival despite codes of honor. In a Homeric epic these qualities would have been decomposed into the apparitions of Ares and Athena. 69

It is too late when the Olifant summons Charlemagne to the scene of the tragedy. Roland's Olifant has been as if the Trump of Doomsday. All have already died, the twelve peers, Turpin, Oliver and Roland too, Roland having concealed his sword Durendal and smashed his Olifant to prevent their falling into Saracen hands. At his dying, Charlemagne's troops witness an almost Apocalypse, Ruskin's pathetic fallacy of God's creation responding to the tragedy.

The design of the text at this point has to broaden, to expand its chiastic structuring, its enveloping bracketting and panelling. It is no longer Roland of eastern France against Marsile of western Spain. It now becomes the battle between the Emperor Charlemagne of Aachen/Aix-la-Chapelle/Jerusalem, of northern Christendom, against the Soldan Baligant of Babylon/Cairo, of southern Islam. The poem presents a windrose of allegory palimspested upon almost all-known geography. The paradigm is Apocalyptic and Augustinian, but in it the City of God converted and subverted to the purposes of the City of Man and heresy is presented as orthodoxy. Thus the poem becomes, in fiction, the Norman Conquest of the Mediterranean world.

When the battle is over, Charlemagne, being victorious in single combat with Baligant, places the broken Olifant as both religious relic and military trophy,

Just as the Normans made use of the past, fashioning and flawing history as tool and weapon, with which to forge their "Christian" Conquest of France, England, Sicily and Jerusalem, so do we assimilate pasts not legally or necessarily our own, using these to "legitimate" or castigate Empire. To this day we have the Matters of Jerusalem (the Bible), of Athens, of Rome, of France, of Britain. The Classical world appropriated the Semitic technology of writing. Christian culture appropriated the literacy and literature of Judaism as well as that of the Classical world. Paul was to say that the liberationist Exodus tale provided us typoi hemon - shadows of ourselves.75 Republican America now - almost unconsciously - borrows the eponymous myth of British Empire - with its Departments of English in its universities teaching war-weary Aeneids and lyrics about, to us, nonsensical nightingales and cuckoos, but we scarcely dare truly teach Joseph Conrad's Heart of Darkness or its decomposition into film as Apocalypse Now. Just as, in the aftermath of conquest, we have exchanged, displaced, texts, the Bayeux Tapestry, which is Anglo-Scandinavian, in France, the Chanson de Roland, which is Anglo-Norman and about Spain, in England, so do we today, in the United States, study texts scattered in diaspora from their European origins.

Epics, unlike democratic pilgrimage poems, are texts which "legitimate" conquests. They comport themselves with bastard disjunction. In them history slips a gear or two. They need to be decoded. They are our programs and we need to tinker with them if we would write our history in an ethical mode. Homer's epics are transmitted by conquerors but are paradoxically - with fair play - the tales of the people whom they had conquered, the Ionians and Dorians, the first subjugating the second, singing of the subjugated Achaians who in turn subjugated Danaians. Arthurian epics, likewise, are tales told and transmitted through the Normans of the race whom Scandinavian-derived Anglo-Saxons/Danes drove out of their land, the Celts. In the case of the Trojan and Celtic materials there was the capacity to both praise and admire those who were subjugated. The Carolingian chansons de geste were transmitted in turn, by these Norman upstarts, exploiting the history of the people, the Franks, whose vassals they were. The Normans thus illegitimately appropriated, thus pirated, the Matters of Rome, Britain and France as well as of Jerusalem to their own "legitimating" ends. But in the case of the Matter of France, unlike those of Rome and Britain, there is an objectification, a denigration, a demonization, of the enemy - who reflected the Normans' Viking pagan northern shadow selves projected upon the southern Spanish Saracen Muslims whom they so willed to conquer and control. Edward Said has written compellingly of a similar distortion and projection by Europeans concerning Muslims in Orientalism. In our own case we find that Europeans make indigenous American peoples, our almost ancestors, heroes while we, as the new Americans, forget and silence their literature in preference to that of England.76 Thus we need to construct new theories of the epic, theories based on guilt or empathy, either piratical appropriation or silencing, decoding the entwined, ambiguous texts of the poetics of appropriated power, and preferring those where the conquerors assume the tales of the conquered, just as, among "primitive" peoples, the hunter totemically becomes the hunted, quarrying the quarry. 77 Such texts mirror and shadow hearer and reader from Troyfall to Doomsday.

1 "The Iliad, Poem of Might," The Simone Weil Reader, ed. George A. Panichas (New York: David McKay, 1977), p. 183. I thank Leslie Fiedler for this reference. I grew up near Hastings, coming home from school seeing Senlac reflecting blood-red sunsets. I wish to acknowledge SMU where I lectured on this material and the AAUW for a Founders Fellowship which enabled me to visit Gerona, Bayeux, Palermo and Monreale. The footnotes of this paper will both observe - and defy - conventions, being this paper's liminal, marginal, joco-serio area.
2 La Chanson de Roland, ed. Joseph Bédier (Paris: D'Art H. Piazza, 1931); trans. as The Song of Roland, by Dorothy L. Sayers (Harmondsworth: Penguin, 1957. Gerhart Ladner, "Homo viator: Medieval Ideas of Alienation and Order," Speculum, 42 (1967), 235, notes the medieval knowledge of "iudicium alienum," judging in others one's own faults, projecting and transforming that "inner disorder, be it moral or pathological, into mass persecution." See Edward Said, Orientalism (New York: Pantheon, 1978), discussing European projected perceptions of Muslim culture, and also the writings of Jean Paul Sartre and his translator, Hazel Barnes, on "bad faith."
3 This hopscotch, cat's cradle, cricket-like ritual game is analogous to the poem which contains it, its blueprint. Epics are encyclopedias including both codes and facts. See W.F. Jackson Knight, Cumaean Gates: A Reference of the Sixth Aeneid to the Initiation Pattern (Oxford, 1936); William H. Matthews, Mazes and Labyrinths: A General Account of their History and Developments (London: Longmans, Green, 1922); Johan Huizinga, Homo ludens: A Study of the Play Element in Culture (Boston: Beacon Press, 1962); Clifford Geertz, "Deep Play: Notes on a Balinese Cockfight," in Myth, Symbol, and Culture, ed. Clifford Geertz (New York: Norton, 1971), pp. 1-37. Chaucer's Knight's Tale similarly presents the tourney/theatre as its code/windrose.
4 Albert B. Lord, The Singer of Tales (New York: Atheneum, 1968); Joseph J. Duggan, The Song of Roland: Formulaic Style and Poetic Craft (Berkeley: University of California Press, 1973); Alan Deyermond, "The Singer of Tales and Medieval Spanish Epic," Bulletin of Hispanic Studies, 42 (1965), 1-8. Paul Zumthor speaks of epics' evolution in time as "mouvement," Parler du Moyen Age (Paris: Minuit, 1980).
5 Wace, Le Roman de Rou, ed. A.J. Holden (Paris: Picard, 1972), Société des Anciens Textes Francais, II, 183, lines 8013-8018; for William of Malmesbury, III, 157.
6 La Chanson de Roland: Réproduction phototipique du manuscrit Digby 23 de la Bodleian Library d'Oxford, ed. Alexandre de Laborde (Paris: Société des Anciens Textes Francais, 1933).
7 Laborde, p. 4.
8 Laborde, p. 11; Ettore Li Gotti, La Chanson de Roland e i Normanni (Florence, 1949); Henri Grégoire contended that the final military exploit is a propagandistic, palimpsested reference to the 1108 Norman failure at Dyrrachium, where Robert Guiscard and Behemond Hauteville, Prince of Antioch, lost to the Byzantine Emperor Alexius Comnenus, Grégoire arguing fantastically that the interpolated "Baligant" derived from "Paleologus": "Imphe, la ville d'Amphion en Terre d'Epire," Mélanges de philologie romane et de littérature médiévale offerts à Ernest Hoepffner (Paris, 1949; Genève: Slatkine, 1974), pp. 183-190.
9 For Crusading material, see Sir Steven Runciman, A History of the Crusades (Cambridge: Cambridge University Press, 1962); A History of the Crusades , ed. Kenneth M. Setton (Philadelphia: University of Pennsylvania Press, 1955). It thus appears that the Normans adopted the tale, embroidering it, palimpsesting it, to their own needs and then scattered it about the map of Europe, in England, in Sicily, in Outremer and even back home in Scandinavia.
10 Joseph Duggan, The Song of Roland: Formulaic Style, p. 41; Paul Aebischer, Rolandiana Borealia: La Saga af Runzivals Bardaga et ses dérivés scandinaves comparés à la Chanson de Roland (Lausanne: Rouge, 1954), on Karlamagnus saga compiled and translated into Old Norse for King Haakon Haakonarson, thirteenth century, etc.; Robert Davidsohn, Storia di Firenze, trans. Giovanni Battista Klein (Florence: Sansoni, 1977), V. 73, tells of the names of Roland and Turpin becoming common in Florence in 1179, while the singing by blind bards about Roland and Oliver was banned in the public squares of Bologna, 1289, because of the disorder and carnage the stories caused.
11 Duggan, p. 76, comments on Menandez Pidal's reaction to the Joshua echo in the text as "un example de mauvais gout." In fact, the Joshua material would have been typically invoked in crusading self-exegesis. The Apocalypse was a retelling of the Exodus. On exegesis, see C. Spicq, Esquisse d'une histoire de l'exégèse latine au Moyen Age (Paris: Vrin, 1944); Henri de Lubac, Exégèse mediévale: les quatre sens de l'Ecriture (Paris: Aubier, 1959-63).
A review of a book in Ashgate's series, could be of interest here:

The series Crusade Texts in Translation is proving to be one of the most significant contributions to the dissemination of crusading history of recent years. Whereas Anglophone students once used to rely almost entirely on the venerable translations of the Gesta Francorum and Fulcher of Chartres for contemporary narratives of the first crusade, plus a few Penguin Classics to take crusading into the thirteenth century, they now have the benefit, thanks to Ashgate's initiative, of a range that includes Robert the Monk and Guibert of Nogent for the First Crusade, Walter the Chancellor for the early history of Antioch, The Itinerarium Regis Ricardi for the Third Crusade, Baha ad-Din's Rare and Excellent History of Saladin, and The Templar of Tyre for the thirteenth-century kingdom of Jerusalem. The Bachrachs' translation of Ralph of Caen extends the available range by adding one of the lesser-known near-contemporary accounts of the First Crusade.
Ralph deserves to be better known and more comprehensively used than he has been. His Gesta Tancredi, although obviously covering the same ground as many of the other chroniclers of the first crusade, tells a story with its own distinct narrative and flavours. Though he became the panegyrist of the south Italian Norman Tancred, Ralph probably came from Caen, where he studied under Arnulf of Chocques at the cathedral school. He did not take part in the events of the first crusade that he describes, but he later served with Bohemond on his unsuccessful campaign in the Balkans against Alexios Komnenos in 1107- 8. Ralph had been ordained priest before 1106, and may have become Bohemond's chaplain. After the end of this campaign he joined the entourage of Tancred, who was at that time Bohemond's regent in Antioch. His connection with Arnulf of Chocques was renewed when he asked his former teacher to edit his completed Gesta Tancredi. Ralph's literary style is a combination, at times uneasy, between bluntness of narrative and full-blown rhetoric; the latter especially prominent when putting set speeches in the mouths of his protagonists. The straightforwardness of his narrative style is particularly effective in the battle scenes, for example in the description of the Turkish panic in the encounter at Tarsus. Ralph is also good at memorable epithets: of the victory at Nicaea, for example, he remarks, "Gaul assured it, Greece helped, God brought it about." He provides vivid portraits of the leaders of the crusade, especially Robert Curthose, duke of Normandy, Robert count of Flanders and Godfrey of Bouillon. Nevertheless, there is an underlying tension throughout the work between Ralph's consciousness of his role as an historian, and the value he demonstrably places on eye-witness accounts, and his use of epic verse forms in describing about a quarter of the events in the narrative. Because he was following the career of an individual leader, Ralph was more concerned with how that individual responded to the challenge posed by the crusade than with delivering a narrative of the expedition itself. This is shown most transparently in the story of Tancred and the hermit on the Mount of Olives (chs 113-14). The story functions in part to show Tancred's religious sensibility--his meditation on the city of Jerusalem bears some comparison with pilgrimage literature, and contains an echo of a pilgrimage episode in Ralph Glaber's Histories. By the time Ralph of Caen had finished writing - probably before 1118 - the main events of the crusade were in any case well known from earlier accounts. The Gesta Tancredi fills in some gaps, particularly in its account of Tancred's diversionary expedition in Cilicia, but its chief value lies in the perspectives it offers rather than in providing wholly new information. Thus, we have a more positive view of Stephen of Blois than that reported in other chronicles, emphasising his military success in skirmishes with the Turks at Antioch, and ascribing his departure to illness rather than desertion. Raymond of Toulouse, on the other hand, comes off rather poorly at Ralph's hands, being outmanoeuvred by Tancred during a quarrel over foraging at Antioch and shown up as a cynical manipulator of religious sentiment in the affair of the Holy Lance, which Ralph asserts was planted in the church of St Peter by the Provencal priest Peter Bartholomew. In his coverage of this event, and in the rather dignified quarrel between Tancred and Patricarch Arnulf after the siege of Jerusalem, Ralph, as the Bachrachs note, demonstrates considerable rhetorical and dialectical skills. The Bachrachs downplay, however, Ralph's emphasis on Tancred's Norman ancestry and its significance. It is true that Ralph speaks most often of "Gauls" in general rather than of Normans in particular when speaking of the crusaders' military success - though he is less interested than was Fulcher of Chartres in the notion of a single Christian people being forged by the common purpose of the crusade. On the other hand, there are significant moments when Ralph does show a preoccupation with the military reputation of the Normans: for example, he wants to present the vanguard at Dorylaeum as if it were entirely a Norman contingent, ignoring Flemish and Byzantine participation; likewise, the story of the hermit on the Mount of Olives is used to glorify Tancred's family association with Robert Guiscard. Conversely, Ralph dwells on the shame cast on the Normans by desertion at Antioch of the Maisnil family.

I wrote this essay ten years before the millennium, eleven years before the World Trade Center destruction. It was published in Jerusalem: Essays on Pilgrimage and Literature (New York: AMS Press, 1998), pp. 67-99, and is given here with AMS Press' consent.

The greatest violence to museums of all has been in Baghdad, Iraq, where they have been looted by war, objects, smashed or seized out of context.