N[ikola Vujnovich]: Did you decorate everything in the song just the way it is in the songbook?
A[vdo Mejedovich]: No, I did more of that.
N: More, you say?
A: Yes, more, by at least twice.
N: Aha! At least twice as much?
A: That’s how I do it. Do you want me to lie to you, or tell you the truth?
N: The truth, just tell me the truth!
A: Aye!
N: Yes, we’ve got to get at the truth.
A: Well, then, mate, as I’ve told you before, when some other good singer takes five hours to sing a tale, I need ten.

( Serbo-Croatian Heroic Songs, vol. 3, eds. A. B. Lord, D. E. Bynum, Cambridge MA, 1974, p. 74)

I regret that N. has taken offense at my review of his book Poetry as Performance in the belief that I have misrepresented his views by claiming that he rejects Albert Lord’s theory of the dictated text (I attribute the same views to him en passant in a footnote in my chapter “Homer and Writing” in the recent E. J. Brill volume A New Companion to Homer, which I co-edited with Ian Morris).

We must remember that the Parry-Lord theory of oral composition is based on analogy. As Avdo Mejedovich sang the Wedding of Smailagich Meho, so Homer sang the Odyssey; as Nikola Vujnovich took down Avdo’s song, so someone took down Homer’s. Criticism of Parry-Lord has always come from those who reject in principle such arguments by analogy, unjustly in my opinion, and N. would appear to reject them too.

Lord’s theory of the dictated text is after all a theory of text-fixation. Oral songs on similar themes sung by different singers are very different, and even by the same singer are different every time they are sung. Dictation to a scribe, however, fixes the text. The dictated text continues to exist so long as it is copied and preserved, a basis for memorized reperformance by literati, liable to distortion, interpolation, and other mishaps, but no longer governed by the laws of oral composition.

For Lord’s lucid theory, based on analogy, of the unique dictated text, N. has in Poetry in Performance and other publications substituted an “evolutionary theory of text-fixation through crystallization.” What N. means by this is never clear to me, but he appears to believe that the Iliad and the Odyssey were somehow recorded (by dictation?) many times, over and over, each time incorporating small changes, the origin of the minor variations in our textual tradition. According to Lord’s model, once Homer died, say in the eighth century, later versions of songs he had sung, fashioned by later poets, would no longer be the poems of Homer; yet according to N. (if I understand him rightly), they somehow are still the poems of Homer, even if recorded, as he thinks, in the 6th century. But there is no need to reject Lord’s theory of the unique dictated text (and the refinement of that theory by others, including myself), which solves so many problems. Why replace it with a complex description without analogies in field experience, which fails to take satisfactory account of whole bodies of linguistic, historical and archaeological evidence?