The Nibelungenlied, composed around 1200 ce, offers particularly interesting insights (Ehrismann 1987; Haymes and Samples 1996: pt. 1). With few exceptions, the main characters are well attested in historical sources. They lived during a time span of some 200 years, between the late fourth and the late sixth century ce. Dietrich (Theoderic, 451-526), king of the Ostrogoths and king of Italy from 493, appears in the epic as a guest at the court of the king of the Huns, Attila, who died in 453, soon after Theoderic’s birth. Theoderic’s historical opponent, Odoacer, is replaced in the epic’s tradition by Ermanaric, king of the Ostrogoths as well, who died in 375. Gunther (Gundaharius) and his brothers perished in 436, in a disastrous defeat of the Burgundians at the hands of the Huns, a battle in which Attila was not involved. Hagen, Rudiger, and even Siegfried cannot reliably be connected with known historical personalities. Finally, the two heroines, Brunhild and Kriemhild, are patterned, although in somewhat reversed roles, after two Merovingian queens, Brunichildis, a Visigothic princess, and Fredegund, whose fierce and unscrupulous rivalry in the latter half of the sixth century soon became legendary. All these were strong and colorful personalities, but historically, most of them had nothing to do with each other, and it is difficult to reconcile their epic and historical portraits. The Burgundians settled on the Rhine in the early fifth century, were defeated by the Roman general Aetius in 435 and destroyed by the Huns in 436. The epic’s concluding event, the catastrophic end of the Burgundians, thus is historical as well but totally unrelated to the ways in which the epic describes it. If we knew only the epic we would never be able to reconstruct the historical event.
Virtually nothing is known about the oral transmission of the epic material over the first few centuries. Interestingly, the epic preserves a more positive assessment of Theoderic and Attila than the (clerically influenced) written sources. Finally, the epic reflects concerns and relationships, social and political conditions of the time of its composition, when aristocratic society, exposed to new ideologies and challenges and to severe power struggles, experienced profound changes and a serious clash and crisis of values (Haymes 1998). Overall, then, the tradition underlying the epic is based on personalities and events that are mostly historical but belong to different times and contexts. Over time they were combined, reinterpreted, heroicized, and placed into a dramatic story in which we can hardly recognize the historical originals. The conflicts that are dramatized in the epic are universal and typical, encouraging the audiences to identify with them and to be challenged by them.
The Song of Roland, composed around 1100, is probably the earliest of a large number of medieval French epics, called chansons degeste, accounts of heroic deeds that were based on long oral tradition and composed in performance by singers of tales {jongleurs), although the extant versions received their final shape by written composition (Vance 1970; Brault 1978). The underlying historical event dates to 778 and is well-known fTom Einhard’s Vita Caroli Magni {28-30): upon returning from a campaign against Muslims in Spain, the rearguard and baggage train of Charlemagne’s army were ambushed and destroyed at Roncevaux in the Pyrenees. Gascon treachery was blamed, Muslim involvement is debated. The space Einhard devotes to this event reflects the impression it made on contemporaries. Roland was among the noble victims mentioned by Einhard. The other characters in the epic entered the story at later stages: their origin is largely unknown. The interpretation the event receives in the epic {the rivalry between Ganelon and Roland, the former’s treason in conspiring with the Saracenes, the fight at Roncevaux as a battle between Christians and Saracenes, the friendship between Roland and Oliver, and the trial of Ganelon) is far removed from the original: only the bare bones of the narrative are still recognizable as historical. Names are often anachronistic. So is the the epic’s entire social setting and system of ethics and values that are strongly imbued with Christian elements and the politics of the time of its composition {Haidu 1993): ‘‘The Song of Roland is an exalted stylization of the realities of the feudal world’’ at a time when these new realities still clashed with the earlier heroic ideals of the ruling aristocracy {Vance 1970: 19, 26). Again, we find old songs and themes elaborated into a complex and multidimensional epic in a period that experienced profound changes and massive challenges.
Finally, a few brief comments on Serbo-Croatian epic, studied intensively in the nineteenth and early twentieth centuries {Coote 1978, Lord 2000; see also Chapter 13, by J. Foley). Two types of epic are of special interest in our present context. One is the ‘‘Kosovo Cycle,’’ that is, various songs {never, it seems, elaborated into a big and complex epic) focusing on the battle of Kosovo in 1389, in which the expanding Turks clashed with the Serbs, apparently with an initially unclear result {first reports even celebrated it as a Christian victory), and both leaders, Prince Lazar and Sultan Murad, died. The historical record is very thin. Soon however, Kosovo came to symbolize the loss of Serbian greatness and freedom. Legends emerged, elaborated in later epic versions, to provide a fuller story {Emmert 1990; Malcolm 1998).
Prince Lazar and Milos Obilic, who assassinated the sultan, became the heroes of the legend, the one a martyr king who preferred the heavenly kingdom to an earthly kingdom, the other a perfect knight who sacrificed his life to save his honor and to display Serbian heroism. Legend also required a villain, and so made a traitor of Vuk Brankovic, even though the historical Vuk seems to haven been loyal to Lazar... Blame for the disaster was laid upon disunity among the Serbs: a quarrel arose between Lazar’s daughters, the wives of Vuk and Milos, which set the two sons-in-law at odds and caused Vuk to slander Milos as a traitor... thus provoking Milos to his daring act. {Coote 1978: 263)
The other type of epic song reflects a more limited and intimate range of ‘‘heroic activities.” It was produced in an area that was divided into many population groups distinguished by ethnic origin, nationality, religion, culture, and political orientation. This illustrates another “milieu” that fosters heroic song.
Constant conflict over the boundaries of these groups created a prolonged ‘‘Heroic Age’’ during which the deeds and qualities of heroes were essential to the survival of society. Heroism was the business of every grown male, and a necessary concomitant of heroism were the songs that celebrated it and made it immortal. {Coote 1978: 258; Djilas 1958 illustrates this well.)
As Matthias Murko writes, each song is likely to be based on a historical experience, but the depiction remains too general and vague to allow reconstruction of a specific historical event (1978: 386, 392-3; cf. 1929; Skendi 1980: chs. 8-9). The epics focus on the actions of individuals who dominate in the hostilities against members of another state or religion. Women, duels, and larger fights, caused by women or violated honor, play a big role: abduction of girls or women, weddings and processions and their interruption, ransoming of captured heroes, often for the price of a woman, gifts to heroes, often consisting of brides, revenge for violence and injury, drinking parties ending in brawls and escalating into bloody fights: these are the usual themes.
Finally, Serbo-Croatian epic also offers an opportunity to study the nature of ongoing change and transformation. Scholars tend to be impressed by the conservatism of epic song, both subjectively and objectively. The substance of a given story is often preserved over a long period of time. Singers do not crave for originality and innovation. They are convinced that in various performances they essentially present the same song. Elaboration and condensation of episodes, expansion of ornamentation, variations in the sequence of episodes, and replacement of certain episodes by others are the main forms of change that are compatible with the ‘‘truth’’ of a story (Lord 2000: 99-123). Yet, as in non-poetic oral traditions, in oral song too gradual change is inevitable, not least in order to satisfy the audience’s need for identification. Most songs studied by Murko refer to events that happened at most 150 or 200 years ago (1978: 391-92). Moreover, as Coote observes, by 1913 the withdrawal of the Turks created new conditions.
The preservation of the singing tradition in new circumstances depended upon its continuing to function as a means of edification and entertainment. Eventually, as values changed and occasions both for traditional heroism and for singing heroic songs became less frequent, singing also changed its style and subject. (1978: 258)
In other words, as is to be expected, the survival and thematic constancy of oral heroic song is tied to the continuation of the social conditions that produced it in the first place. Massive changes in the latter will necessarily cause corresponding changes in the former. It seems an important desideratum to verify this principle in a broader range of epic traditions than I was able to investigate here.
World History
