So I am quite interested in folk music, and traditional music, and getting it documented as best as possible in the database. Roud Folk Index is quite important as it covers a lot of folk songs from my home country, England - as well as other parts of the “commonwealth”.
The best resource for checking whether a folk song is part of the Roud index is to use the VWML website (https://www.vwml.org) and typing in the first line of the song, then finding a match and it will give you the index.
Most of the time, the index will show small variations on ultimately the same composition (maybe a different spelling, changing pronouns etc.) but index entry 269 displays wildly differing variants: https://www.vwml.org/roudnumber/269
I am aware of it, and have linked some songs to it, but it is a long and scary list so I would say experience is very very low… I am just a folk \ punk cross-over person
But it is lists like this is why I love the interconnectedness of MB’s database.
I do admit to sometimes being confused as to when a recording is to link to a Work, or a new arrangement. It is a bit of a confused grey area to me.
Do you know that these are different works?
If I lookup other RN entries there are also seemingly unrelated titles but that’s probably not unusual with traditionals.
To be honest I think I’m going to take up @aerozol 's suggestion and contact someone from the library to try and guage why this has happened, and what measurement they use to match variations of the same folk song
As someone who has added folk/trad stuff to the database pretty much since I started with MB more than 15 years ago… err on the side of adding duplicates.
Due to how many of these larger collections of traditional Works function, oftentimes groupings are thematical. E.g., all the R269’s seem to follow a common story of a servant girl meeting a sailor/soldier and becoming pregnant (note also how several of them are also denoted as Laws K43). Some of the “first lines” in some of the VWML versions seem to be non-first lines in other versions, so likely just verses in different order or lyrics getting cut up (or e.g. a chorus’ first line instead of first line of first verse). (There’s a reason that this is a whole field of studies. )
In the end, I’d say it’s up to editor discretion for whether something should be a new MB:Work or not, and I would absolutely not find fault with associating multiple MB:Works with the same Roud or other index number, if applicable.
I was re-reading [STYLE-385] Add "Roud Folk Song Index" - MetaBrainz JIRA and it really seems to me like it’d be sensible to add a work attribute for Roud Folk Song Index, so that the numbers can be stored there (possibly in addition to the series). Does anyone think this is a bad idea?
So long as there is some kind of X-reference (as from what i gather Work Attribs are static values that don’t dynamically update a series entity) to the existing series then yes I’m all for it.
My initial gut reaction to this is that it’s probably not a good idea? It feels like this just adds data duplication with all the pitfalls of the data getting out of sync etc. Is there a reason why it being an attribute would be preferable? I guess being more easily queryable…
Well, series numbers were mostly meant to be a 1:1 thing, so each series should not have a ton of parts with the same number - although it’s not like it actively breaks stuff I guess. It’s also huge so the series will eventually be a huge mess to edit - probably something can be done about it, just not sure what
In general I guess it seems to me more like a work attribute (such as the key or a PRO ID) than a work catalogue, but I might be wrong
)