Which takes me back to my thesis that what is academically referred to as a "source singer" is the specific traditional singer from whom a particular song collector collected a particular version of a song. This doesn't mean that this source singer is the actual source of the song. That actual source is quite possibly some medieval troubadour or minstrel, or some unknown poet/musician (professional or amateur) whose identity is lost in the mists of antiquity. That unknown minstrel is the source of the song, and its first interpreter, like the initial rivulet from a spring or from the foot of a glacier, to which other tributaries add their own interpretations.

But as McGrath points out, one can run this analogy into the ground.

Once again, I maintain that the term "source singer" is a technical term used by folklorists and ethnomusicologists, has a fairly specific meaning, and only gets screwed up when used indiscriminately and incorrectly.

Don Firth