[mb-style] Granularity of works

[Brian Schweitzer]

On Wed, Dec 8, 2010 at 2:00 PM, Paul C. Bryan <email at pbryan.net> wrote:

>  I'm not sure if this has been addressed yet in other forums, but it seemed
> appropriate to get this discussed in mb-style: what should the granularity
> of works be in NGS?
>
> Works will allow ARs to other works. As I understand this, the intent is to
> allow multiple levels of granularity. Ultimately, the intent is to have a
> recording be linked to one or more works.
>
> Let's start with the following straw man (which I expect will be picked
> apart, lit ablaze, blown away, collected, compacted, fermented, packaged,
> sold, and used to fertilize someone's lawn):
>
> At the recording-AR-link level, a work should be:
>
>    - in popular music: an individual song;
>    - in jazz music: an individual jazz piece;
>    - in classical music style: an individual movement of an opus;
>    - in theatre style: a song, musical number;
>    - in opera style: a song, recitative, or musical number.
>
>
> At a higher level level, a work should be:
>
>    - in popular and jazz music: an album; (what about compilations?)
>    - in classical music: an opus;
>    - in theatre style, the title of the musical;
>    - in opera style, the title of the opera.
>
>
> Some immediately obvious problems with what's been stated above:
>
> 1. What about spoken dialogue in theatre/opera?
> Various recordings of the same opera will cut tracks at different points in
> dialog, leading to differering track titles. If we're not careful, we could
> wind-up with an arbitrary list of recording-AR works. Is this outcome
> acceptable/desirable?
>
> 2. What about more than one classical opus on the same release?
> Should a release of recordings with more than one opus be considered a
> work? I'd say no, because collections of more than one classical work in
> releases is common, somewhat arbitrary, and usually has absolutely nothing
> to do with the composition of the work. Counterarguments anyone?
>
> Paul
>

Nit picking, but I tried that simple classical split when I dealt with
originally drafting CSGv2.  Several people did their best to keep tearing
that apart.  :D

The best I could summarize what resulted would be this:

1. An artist has an oevre.  That is all works by that artist.
2. The oevre can contain 'standalone' compositions ("works") as well as
opuses.
3. An opus contains other opuses and/or standalone compositions.  An opus
can consist of just several different (single piece) compositions, but it
can also be singular (a symphony, ballet, etc - one opus of related
compositions), or it can be complex - some individual pieces *and* some
singular ones (symphonies, etc).  Or, it can be *really* complex - all of
those at the same time...  I don't recall the exact details, but someone
managed to find a case where a symphony was published, then republished in
an opus, then that opus and other opuses were collected into yet a bigger
opus - "Op. 2 No. 3, Op. 6 No. 1: Symphony in D major" as the result.
(If you followed luks' original NGS schema, think of the above as akin to
his original multiple-levels RG concept.)
4. Then you have a specific composition in an opus, be it a movement of a
symphony, a standalone single-movement work in a larger opus, or a
standalone single-movement work which isn't in an opus at all.
5. Then (to take just one of those) that symphonic movement may have
multiple parts.
6. Then, whether that movement had parts or not, there's various tempo
changes within the movement.

I think "Work" only really works as #4 above, and the list of works is 1 and
2 above.  I've been trying for a few months to figure a way to make Works
also work at the #3 level, but I think that really needs distinct entity to
work.  (Same holds true for cues from a movie score or video game, movements
of a ballet or opera, etc.)  Something along the lines of a Work version of
an RG, but with the infinite flexibility that luks' original RG concept had,
to allow for all the variation.  Otherwise, we end up fitting a lot of
square pegs into really badly fitting, or simply wrong, round holes.

Brian
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.musicbrainz.org/pipermail/musicbrainz-style/attachments/20101208/35662479/attachment.htm