I'm a piper, not a singer, but the turn which this thread has now taken is leading me to the conclusion that traditional singers may be treating the words in a manner similar to the way musicians treat a tune: keeping the basic structure, but introducing small variations which are regarded as legitimate and even desirable deviations from what was originally learnt.

I know that in the case of songs this may be due to imperfect recollection of something which was once learnt orally from an oral tradition, whereas in the instrumental music, variation of the same phrase recurring within a single performance is regarded as highly desirable.

But I have found that something similar happens on the rare occasion when I learn a tune from a book: if I go back to the book after I've been playing it for a few months, I find that it has spontaneously mutated and is no longer the same as what I initially learnt. This is one of the reasons why it is a pleasure to listen to even the most common tunes being played by a different musician, because each musician does something different with it. In trying to explain the instrumental music to people not familiar with it, I use the analogy of the Ho Chi Minh Trail - there are many broadly parallel ways of getting to the same place, rather than a single highway laid out in concrete.

This DT study has been well worth while, but I suspect that the outcome will indeed be two or three joint contenders for the title of definitive version (a bit like heavyweight boxing, really). After all, I don't think there's even agreement on how to spell "Spancil Hill", and I was also struck by the variant spelling of the author's name. "Considine" is now the standard spelling in English, but the second "o" in "Consodine" above reflects the Irish/Gaelic pronunciation of "O Consaidín" where the vowels and consonants interact.