Jim wrote:

"The broadside output runs contrary to the traditional repertoire in style and in quality - most of the published broadside collections are crammed full of unsingable songs"

This is true for a lot of them; Harry Boardman used to have me create songs out of 19th broadsides for his radio broadcasts and, if not literally unsingable, a lot of them were pretty bloody awful.

But then there's no reason that a print original should have had to be singable in the first place. Apologies for going back to 'The Wild Rover', but it's the one I know most about. The original ballad by Thomas Lanfiere, 'The Good-fellow's Resolution', is indeed wordy and moralistic - like many similar ballads of its day, composed by Lanfiere and others - although verses 1, 8 and 9 clearly belong to the song as we know it. You wouldn't look at that text and think it was the work of an unlettered toper of the lower classes. But what happened to it next - probably around 1800 - was clearly a conscious edit rather than some kind of oral processing, since in the course of cutting the song down to five verses stanzas have been deliberately cut-and-pasted, split, rejoined and boiled down. Thereafter there is a trail of 19th century broadsides each looking a bit more like the song as collected in oral tradition. So, even though the original was arguably 'unsingable', it nonetheless formed the basis for something that became highly singable.

Something similar seems to have happened with Child 243 ('The Demon Lover'), in which seven verses from the middle section of a 32 verse original 'A Warning for Married Women' - were cut out and used as the basis for a new ballad.

However, if you look at Child 286 ('The Sweet Trinity' / 'Golden Vanity' etc) The 17th century London broadside 'Sir Walter Raleigh Sailing in the Low-lands' is almost word-for-word) the same as oral versions collected in Appalachia by Cecil Sharp (apart from Sir Walter's part in the drama), and seems to have gone into oral tradition more or less unedited, then remained more or less unaltered for 200+ years.

Re Sharp's diaries:

I've spent a lot of time with his Appalachian diaries (I'm not aware that he kept one when he was collecting in England) and they don't really provide answers. Where he asked a singer about their source, the answer was usually a senior family member. He saw no printed broadsides, though he did observe one or two handwritten 'ballets'. Some of the songs he collected have texts almost identical to those in 19th century songsters, but the majority do not. The most popular ballads noted by Sharp from mountain repertoire are mostly those known to have existed in print in the 17th or early 18th century (Barbara Allen, Lord Thomas and Fair Ellender, House Carpenter etc), which fits with the notion that 18th century migrants brought them over, either on paper or in their heads. Of course the fact that most of them were in print by the 17th century does not necessarily mean that the migrants learned them directly from broadsides, but it does tell us that they were definitely around in England at the appropriate time, and suggests that they arrived with the settlers rather than being learned by later generations in America.