This issue brings up one of my pet peeves with the Digital Tradition database. The database started its life as a couple people's own personal collections of songs that they liked and scribbled down the words of as best they knew them. This meant that words were frequently misheard or mistranscribed, songs weren't attributed to anyone or to the wrong author, etc - but that's typical for personal song books.

Now the DT has gotten huge, and is a major reference for people all over the world. Unfortunately, being an all volunteer effort, those same mistakes and misattributions are becoming more and more etched in stone. We could just brush it off with a "Oh, that's just the folk process...", but where the author of a piece is known and his or her actual words are obtainable, it's disrespectful to leave the mistakes in place.

The wonderful people who have been doing "Lyric Add:"s recently have been doing a much better job of documenting them, with identification of sources, stating if it's from a written source or a transcription from a recording (and from WHICH recording or source), etc. Maybe we could find a grad student who wants to make editing and analysis of the DT his or her thesis project for a degree in Ethnomusicology or Library Science?