The text of Ferdowsi's 'Shahnama' and the burden of the past.

• Author: Davidson, Olga M.
• Position: Persian poet Abu'l-Qasem Ferdowsi

The thesis of this note is that the textual tradition of Ferdowsi's Shahnama, by way of its variant readings, reflects an oral tradition of formulaic composition. In my book on classical Persian poetry, I had tested an earlier version of this thesis on a randomly selected passage from the Shahnama, finding that "every word in this passage can be generated on the basis of parallel phraseology expressing parallel themes," and that "the degree of regularity and economy in the arrangement of phraseology is clearly suggestive of formulaic language."(1) More important for the moment, I found that "the variations between variant lines in different manuscripts correspond to those between variant lines in different passages" (p. 175). In other words, the variant readings of different manuscripts of Ferdowsi's Shahnama result from a system of formulaic variation typical of oral poetics.

The comparative method offers criteria for establishing what is or is not typical of oral poetics. There are two levels of comparison. One involves the evidence of living traditions of oral poetics as observed and described in fieldwork. The primary case in point is Parry's and Lord's research in South Slavic heroic song.(2) Another level involves the evidence of texts revealing patterns of formulaic composition that are demonstrably analogous to what is observable in living oral traditions. In my book on classical Persian poetry, the main point of comparison was classical Arabic poetry (Poet and Hero, 64), as analyzed by Michael Zwettler in The Oral Tradition of Classical Arabic Poetry: Its Character and Implications (Columbus, 1978). For Zwettler, the high degree of variation in the variant readings of any given sample textual tradition in Arabic poetry reveals the underlying system of an oral tradition: "nowhere does the inherent instability or, better, fluidity of the early Arabic poem - its essential multiformity - emerge with greater clarity than through consideration of the body of those lectiones variae that the textual tradition has preserved" (p. 206). Equally important, Zwettler shows that scribal mistakes "do not constitute a major source of variation" (ibid.).

In oral poetics, as the empirical fieldwork of Parry and Lord had shown, the burden of the poetic past becomes in varying degrees re-created with every recomposition-in-performance: variation is a primary mark of recomposition-in-performance, and Zwettler applies this discovery of Parry and Lord to classical Arabic poetry, describing it as "a poetry that lives through variants" (p. 189). In other words, the variae lectiones in the Arabic textual evidence can be seen as a reflex of the formulaic variants in what was formerly a living oral tradition.

In this way, classical Arabic textual traditions may be usefully compared with those of classical Persian - as may, to a lesser degree, medieval Irish and French textual traditions, and those of ancient Greece.

There are those who object to such comparisons, however, arguing that Ferdowsi's poetics are too different or even superior to be compared to the South Slavic traditions studied by Parry and Lord - and the objection extends to the poetics of the oral Irish, French, and Greek traditions.(3) The premise of such an argument is that, "unlike Homer and the Serbo-Croatian bards of Parry's and Lord's acquaintance, who practiced their art in the context of an 'oral culture,' Ferdowsi composed self-consciously, and formally, in the context of a highly literate society."(4) After all, "Khurasan of Ferdowsi's time was a hub of cultural and intellectual activity, not some backwater haven of oral tradition," and the city of Tus in Khorasan province "was an urban center, and no matter where in the environs of the city Ferdowsi lived, he was a sophisticated urbanite, not a provincial bard."(5)

It is important, however, to avoid unfounded assumptions about oral poetry as something by nature unsophisticated and necessarily non-urban. Let us take as a point of comparison such documented cases of "sophisticated urbanites" as we find in medieval French poetic traditions. In this context I turn to Rupert Pickens' edition of the medieval troubadour songs of Jaufre Rudel (floruit twelfth century A.D.).(6) Pickens' editorial work is pertinent to my main thesis concerning textual variation as a symptom of compositional variation. Pickens proves that most of the textual variations in the songs of Jaufre Rudel, that is, most of the different readings transmitted by the different manuscripts, are part of a compositional system that goes beyond any individual composer. Some of the variations in the manuscript tradition of Jaufre Rudel may be due to Jaufre himself, while others are due to the transmitters of his songs in the song culture of the troubadours. But the point is, both kinds of variations are part of one system (Pickens p. 35):

The difference is that the modern editor and critic might consider Jaufre's changes to be "authentic," because authorized by him, and the transmitters' to be intrusive and destructive, despite the latters' positive motivation. Certainly, as we know by observing the manuscript tradition, authenticity was not a concern of the transmitters; moreover, Jaufre himself affirms the principle of change as esthetically proper to his genre, so that it might be said that mouvance is an aspect of the intention of his songs.

The concept of mouvance, as Pickens is using it here, was formulated by the medievalist Paul Zumthor. According to this formulation, medieval texts that derive from oral traditions are not a finished product, un achevement, but a text in progress, un texte en train de se faire.(7) No matter how many times a text derived from oral traditions is written down, it will change or move: hence the term mouvance. Following both Zumthor and Pickens, Gregory Nagy has applied the concept of mouvance to the history of the Greek Homeric text: both the papyrus fragments (from the Hellenistic and Roman periods) and the medieval manuscripts of the Homeric Iliad and Odyssey preserve a number of variant readings that are demonstrably authentic from the standpoint of the formulaic system that generates Homeric diction.(8) In any given case, where two or more authenticated variant readings are attested, Nagy argues that the editor's task is to establish which variant was used at which historical point in...