[mb-style] [Clean up CSG] Classical and Release Language

[Jim DeLaHunt]

Brian Schweitzer wrote:
> 
> In my opinion, any concept of a release language, for classical [and
> opera], is rather bogus.    Any decent non-budget release (Naxos, Philips,
> etc) tends to list either multiple languages or single listings in a mix
> of languages.  We've typically selected one, rather than enter the same
> classical release multiple times.  The non-decent releases tend to be the
> ones with a single language listing - but so minimal ("Adagio", "Allegro",
> etc) as to be rather not worth consideration.  We also have people adding
> classical in languages not even on the liner, yet we don't force those
> into the pseudo-release category, as we would for anything else.
> 

Great question, Brian.  You are right that many classical Releases have
multiple parallel languages

My big obstacle in discussing ReleaseLanguage is that I don't have a clue
why we're recording that data, what problem it's trying to solve.  My first
instinct is to turn to http://wiki.musicbrainz.org/ReleaseLanguage .  It
says there that "the language attribute (e.g. French) records the language
of the release title and track titles, (not the lyrics and not the extra
information on the disc sleeve)".  It then goes on for several paragraphs
about what notation the database uses to record the language attribute, what
scripts can be used to write various languages, etc. Not a word about which
person would use the Language attribute, and in what circumstance.  Does
someone use ReleaseLanguage to search for Releases or Tracks?  Does the
ReleaseLanguage propogate through to tags in digital music files in some
helpful way?  It doesn't seem to show up in my music player (iTunes latest
version).

There is a big debate in
http://wiki.musicbrainz.org/ClassicalReleaseLanguage about which value we
should set ReleaseLanguage to for ClassicalMusic.  I didn't see anything
there either about the function of ReleaseLanguage -- who uses it, to solve
what problem.

The most telling argument I see there is: "ReleaseLanguage has a clearly
defined function and meaning - if you need something new/different for
Classical releases, then have a new extension in the database, but don't
misuse an existing feature. Most people arguing for this I've heard about
where clueless about the use of ReleaseLanguage anyhow, which I find pretty
symptomatic." (dmppanda 2007-12-11 05:14:40)   (Though I don't find that the
wiki page ReleaseLanguage gives me "a clearly defined function". The fact
that people are "clueless about the use of ReleaseLanguage" says that maybe
the documetation isn't doing the job.)


Brian Schweitzer wrote:
> 
> Point is, "the language of the release" makes sense for most things.
> Classical, though, it just doesn't....
> 
> What I suggest then, is that, once we do have a works list for a classical
> composer, we standardize the language for that composer.  The effort
> involved in the creation of such a works list is massive,...
> 

And while it's great to have those lists as text on a wiki page, I'd hope
that those works lists could be represented as MusicalWorks database
structures.


Brian Schweitzer wrote:
> 
> ... but we already have some of us doing it.  The effort to then correctly
> identify and rename each and every track for that composer too is great. 
> But it's far better than trying to do the exact same thing, but across
> multiple languages for the same release - not only do we then have works
> by Mozart, but we have
> works by Mozart in English, French, German, Dutch, Swedish, Norwegian,
> Japanese, Chinese, etc - and for each one, then translate each track, then
> rename it.
> 

I think this leads right to the big philosophical divide in MusicBrainz
between Classical/Opera/Soundtrack on one hand and everything else
(pop/rock/folk/etc.) on the other.
* Authorship as primary vs Performance as primary
* Authorship as usually long ago vs Authorship as usually recent
* Multiple releases by multiple performers of the same Musical Work being
typical vs Single release by single performer being typical (with some
covers and re-releases, sure)
* Well developed identity for the musical work and composer independent of
performance vs musical work and composer as an appendage of the Performer
* Performance as team effort by well-known artist (composer, conductor,
orchestra, soloists, etc.) vs performance by primary artist with secondary
others.

An important use case is the person with a large collection of digital
classical music files; with metadata provided by MusicBrainz, entered by
different editors at different times. The music collector views a list of
Tracks and Releases, names provided by MusicBrainz.  Authorship is primary
in Classical/Opera, and it's common to have multiple recordings the same
musical work by different performers. The ClassicalStyleGuide (CSG)
convention of setting the ReleaseArtist/TrackArtist to composer and
standardising the track titles groups different performances of the same
musical works together. The standardisation also leads to consistency in the
list of track names as seen in the music player. By contrast, in Everything
Else genres, performance is primary, and there's no expectation of
consistent naming between albums by different performers. It's sufficient to
faithfully reproduce what's written on the release.

So in the interests of getting consistent results from different editors
entering metadata for different releases of the same musical work, we want
to unify the language used in track titles even if the language used on the
release packaging varied. We are choosing to use the TrackTitle field to
state the identity of the musical work, rather than document the text on the
release packaging.

Thus I'd argue that ReleaseLanguage a) doesn't have a clear purpose, but
that b) it is defined as describing the text on the Release packaging rather
than the well-known identity of the musical work, and c) in Classical we
have practical reasons for disregarding the Release packaging in favor of
the musical work. Thus ReleaseLanguage doesn't look very meaningful or
important to me.  And thus I'd say let's leave it as describing the text on
the Release packaging.

There's another use case at work here. A User who reads English only has
digital music files extracted from a Japanese CD of Beethoven's 9th release
in Japan with Japanese text on the Release packaging, and from a French CD
of Beethoven's 9th released in Belgium with French, Dutch, German, Italian,
and English text on the release packaging.  I'll bet what this user would
really prefer is track titles and metadata in their digital music files that
follow English language conventions, not what was on the Release. And we
could give it to them.  Once we have a database of MusicalWorks per se, we
could add a layer on top, which map from abstract musicological concepts
like "symphony", "movement", "key", "tempo mark" to textual expression of
those in various languages. For instance, a given key could be represented
as "ト長調", "Es-Dur", "Fa Majeur", and "C Major" (OK, they are different keys)
depending on which convention the user wanted to see.

I have a sense that this is where all this talk of entering a Release
multiple times in multiple languages is really heading.


Brian Schweitzer wrote:
> 
> My hope is that eventually we have real generic works lists as database
> objects (not just wikilists), and can then translate on demand to any
> release to any desired language.  But until then, just getting the data
> clean is in of itself a large job.
> 

It looks like you and I are heading in the same direction!  And maybe I'm
getting a wee bit ahead of myself.

But if we can agree on this direction, I'd sure like to write it down.
Perhaps NextGenerationSchema is the place to write it. Even if the CSG tells
editors to do something as an interim step, it will help to talk about the
end goal as a motivation for the interim steps.

-----
     -- http://jdlh.com/ Jim DeLaHunt , Vancouver, Canada  • 
http://wiki.musicbrainz.org/JimDeLaHunt

-- 
View this message in context: http://www.nabble.com/-Clean-up-CSG--Classical-and-Release-Language-tp15083579s2885p15117756.html
Sent from the Musicbrainz - Style mailing list archive at Nabble.com.