"Re Cecil Sharp
He collected the version with St James in the words in Dewey, within walking distance of a railway station, which is how he got there himself.
You can read his diaries for the year here..."

Having spent the last four years researching Cecil Sharp's trips into the mountains, (the results are published in the current Folk Music Journal), I'm well acquainted with his diaries, the kind of communities he visited and the songs he collected. One of my principal arguments is that he and Maud Karpeles noted down a large number of songs that were not 200-year-old British ballads that had come over with the original migrants, so I do realise that the communities were not completely cut off from the outside.

"Walking distance of a railway station" actually turns out to be at least five miles away, if you read Sharp's diary entry for June 4, 1918, and do a few sums. He thought nothing of walking fifteen miles or more in pursuit of singers.

For your interest, there are several photos taken by Sharp of Laura Virginia Donald on the VWML Library website at #AC38.

However, the fact that Mrs Donald lived out in the woods five miles from a railway station isn't really relevant to the balance of probability regarding song transmission to or from Nova Scotia. The chances are tiny compared with the likelihood of the song having arrived independently from the British Isles.

Karen, although you have got a really interesting discussion going here, and although I have much sympathy for your argument about the rather reckless approach of sixties folk revivalists to evidence-based research, you seem to have got the cart before the horse. Barry's theory for Irish origins, or Lloyd's on the 'Unfortunate Rake' title or the significance of St. James Hospital, may well be based on incomplete or flimsy evidence, and deserve to be interrogated. However, you seem to be taking the position that everything they said must therefore be untrue, and that evidence that does support those theories must be dismissed - hence the repeated and rather quixotic attempts to airbrush away the pre-1920 'St James' variants of Roud 2, and to disqualify Tom Lenihan's version of the song.

On the subject of TL, you've posted the reference to American migration from his community more than once, as if this proves he got songs from there (in direct opposition to Jim Carroll's first-hand evidence), but that misses the point that his variant of Roud 2 is textually rather unique. If by some remote chance his sister had sent him the record of Armstrong singing 'St James Infirmary' it's highly unlikely he would even have recognized it as a version of the song he knew, never mind incorporating the place name. If that's not what you're suggesting I apologize for misunderstanding you, but if not, what are you trying to prove about the Lenihan version?

When I joined this discussion I was well aware of the longstanding theory that 'Unfortunate Lad' had evolved into 'St James Infirmary', 'Streets of Laredo', etc, etc, but also aware that modern scholars had questioned it. Having done a little digging myself, I've become more convinced that 'St James' is indeed part of the British Isles tradition of the song, even though no-one seems to have found anything pre-20th century as yet.