Regarding Phil's question:
> What are the methods for validating an exclusively 'oral' tradition?

That's the rub. In the English language printed broadside ballads go back to the 1500s so we can never exclude that the orally transmitted versions passed through print versions from time to time. But that's ok.

The premise would be that orality exists on a spectrum, and the more oral the transmission has been the more the song would evolve. With respect to evolutionary change of songs, if it is the ephemeral nature of oral events that is important, then perhaps the more ephemeral a written version is (e.g. broadsides more ephemeral than books or recordings) the more fluid the song remains.

Although there were broadsides in the 1500s, only 15% of the population of Great Britain was literate*, and those would be mainly in urban populations and upper classes. Generally literacy is much lower in rural populations, the poor, and disenfranchised populations (such as women and blacks in 19th c America); but these gaps have closed with time. The implication would be that songs collected from the rural poor before widespread literacy and before the dissemination of recordings and broadcasting, are the most likely to have been transmitted primarily orally.

Even with the likely involvement of written versions along the way, however, it is not at all difficult to find evidence of song evolution among versions of the same song collected as late as the 1950s from England, Ireland, Scotland, Canada, United States, and Australia, and from different regions within those countries. So "pure" sources are not required to make the point. Clearly Barbara Allen has diverged far more than, say, a Bach Oratorio, over the same time period.

Pamela


*interesting literacy data