Jim Carroll wrote:
'Yes, they [the source singers] most certainly did [distinguish between 'traditional' and 'commercial' songs in their repertoire] - this has been argued ad nauseum elsewhere and apparently remains one of the great myths that haunt this subject. I would be interested to learn on what evidence people who make this claim base it on.'

Even if this were the case - i.e. that recreational singers in rural areas divided their songs into 'the old/orally transmitted ones' and those of more obviously recent (theatre/music hall, then gramophone/radio) provenance - clearly the majority ended up preferring the latter, or the oral tradition would not have died out.

It is interesting that the purist wing that emerged from the 60s revival and saw itself as reacting against the supposed middle-class insensitivity of Cecil Sharp - I'm thinking of people like Reg Hall and everyone involved with the Musical Traditions magazine/label - continue to reproduce his romantic notion of a 'tradition' untouched by interactions with a commercial entertainment industry, even though such a thing had begun to come into existence at least as early as the 17th century. Dave Harker, whatever one might think of his critical Marxist angle, did some persuasive research on this subject.

As Steve Gardham points out a couple of posts above, far from all being improvised by farm labourers in their spare hours, many of what are now considered 'traditional' songs had origins as printed broadsides. In fact, what strikes me when listening to CDs in the Hall-curated Voice of the People series is how few of the songs recorded correspond to the Child/Sharp/Vaughan Williams ideal of folk songs. Looking at the texts, there's a lot of doggerel that smacks more of the street ballad than the neater or slightly more 'poetic' lyrics (supposedly polished by the unconscious literary instinct of the 'folk') favoured by the collectors and the latter-day folk musicians (M. Carthy et al) who worked from the Penguin Book.