"....ballad scholar David Buchan claimed that at the height of the tradition there were no set texts to ballads but plots and poetic commonplaces which a singer would use to re-compose at the point of performance."

And does your own collecting experience support that view, Jim? I'm aware of examples of a given singer performing a ballad differently on separate occasions (a completely different tune, for example), but does text really vary so much from one performance to another? And if Buchan's theory (as I understand it) only applied to a pre-literate era, what actual evidence did he produce to support it? I can see how it could be *inferred* from - for example - old and new world versions of ballads that tell the same story in completely different words, but direct evidence?

PS Jim, thanks for the copy of 'Songs of the Irish Travellers' - I will be writing to you more fully when I've got through my present bout of flu and Whitby folk festival next week.