"How did they mediate what they chose to sing and what they actually performed and recalled and passed on to the collectors. Part of this will be their families, communities, and audiences if any."

Good question. Or maybe questions, for there is a lot in it.

Somebody told me to read Wilgus and somewhere he makes a point about collectors and researchers in the early days not being trained, which I took as implying that in the US where 'folklore' has been more of an academic subject, such training has taken place.

So maybe here we have a problem in terms of evidence, quality and quantity? Sorry if I sound like a broken record, but it is an honest point of view.

Jack's point about stuff not being 'real' outside particular contexts is a good one.

I think there may also be issues about informants actually creating stuff because somebody wants to collect it. There will be those who know more than I do, but is it not suspected that Mrs Anna Gordon may have played a part in 'creating' some of the ballads that she reported and which were published by Child?