Discussion:
[mb-style] [Clean up CSG] Capitalization (and placement) of types and tempos
Brian Schweitzer
2008-01-25 15:07:08 UTC
Permalink
Part two of a trio of messages tonight... splitting these to keep the
discussions easily separated.

In CSG, we indicate worktype and tempo like this:

I. Rondo. Allegro

When it's like that, it's easy. But several things can make it complicated,
and neither CSG nor OperaStyle currently really gives clear guidance.

Several cases (the first alternate in each is my own preference):

1) Multi-word types/tempos: sentence style or English caps?
a) Tempo di minueto / Tempo di Minueto
b) Tema con variazioni / Tema con Variazioni
c) Rondo alla turca / Rondo alla Turca
d) Andante grazioso / Andante Grazioso
e) Rondeau en polonaise / Rondeau en Polonaise
etc.

2) Multiple types and or tempos for the same movement:
a) Rondo. Allegro - andante / Rondo. Allegro - Andante
b) Rondo. Allegro & Minore. Allegro di molto / Rondo. Allegro - minore.
allegro di molto / etc.
etc.

These next two are more of a mess...

3) Minuet & Trio - easy, until you mix in tempos. How do we want to
integrate these?

Minuet. Andante & Trio
Minuet. Andante & Trio. Allegro
Minuet. Andante - pi? molto & Trio
Minuet. Andante - pi? molto & Trio. Andante - allegro - primo tempo

4) Tema con variazioni

Some releases split track each variation. Some themes have identified
tempos. Some variations have different tempos, sometimes even tempo changes
within a variation. Most of these have not just one or two variations, but
up to as many as twelve...

Tema. Andante con variazioni
Tema. Andante & Variations I - IX
Tema. Andante & Variations I - X & Variation XI. Adagio cantabile &
Variation XII. Allegro
Tema. Andante & Variation I & Variation II & Variation III & Variation IV &
Variation V & Variation VI & Variation VII & Variation VIII & Variation IX &
Variation X & Variation XI. Adagio cantabile & Variation XII. Allegro
etc.

The first reads as ugly to my eye, and we miss the tempos in the last two
variations.
The second we still miss the tempos.
The third reads as a decent compromise to me.
The fourth gets obscenely long very very quickly, especially if you have
intra-variation tempo changes.

Brian
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.musicbrainz.org/pipermail/musicbrainz-style/attachments/20080125/c46bd8ef/attachment.htm
Frederic Da Vitoria
2008-01-25 15:22:52 UTC
Permalink
Post by Brian Schweitzer
Part two of a trio of messages tonight... splitting these to keep the
discussions easily separated.
I. Rondo. Allegro
When it's like that, it's easy. But several things can make it
complicated, and neither CSG nor OperaStyle currently really gives clear
guidance.
1) Multi-word types/tempos: sentence style or English caps?
a) Tempo di minueto / Tempo di Minueto
b) Tema con variazioni / Tema con Variazioni
c) Rondo alla turca / Rondo alla Turca
d) Andante grazioso / Andante Grazioso
e) Rondeau en polonaise / Rondeau en Polonaise
etc.
a) Rondo. Allegro - andante / Rondo. Allegro - Andante
b) Rondo. Allegro & Minore. Allegro di molto / Rondo. Allegro -
minore. allegro di molto / etc.
etc.
I would do like you, I assumed Italian capitalization was the rule here.
Post by Brian Schweitzer
These next two are more of a mess...
3) Minuet & Trio - easy, until you mix in tempos. How do we want to
integrate these?
Minuet. Andante & Trio
Minuet. Andante & Trio. Allegro
Minuet. Andante - pi? molto & Trio
Minuet. Andante - pi? molto & Trio. Andante - allegro - primo tempo
I don't understand your question here.
Post by Brian Schweitzer
4) Tema con variazioni
Some releases split track each variation. Some themes have identified
tempos. Some variations have different tempos, sometimes even tempo changes
within a variation. Most of these have not just one or two variations, but
up to as many as twelve...
Tema. Andante con variazioni
Tema. Andante & Variations I - IX
Tema. Andante & Variations I - X & Variation XI. Adagio cantabile &
Variation XII. Allegro
Tema. Andante & Variation I & Variation II & Variation III & Variation IV
& Variation V & Variation VI & Variation VII & Variation VIII & Variation IX
& Variation X & Variation XI. Adagio cantabile & Variation XII. Allegro
etc.
The first reads as ugly to my eye, and we miss the tempos in the last two
variations.
The second we still miss the tempos.
The third reads as a decent compromise to me.
The fourth gets obscenely long very very quickly, especially if you have
intra-variation tempo changes.
I agree here also. Not very easy to explain, but this should not happen
very often.
--
Frederic Da Vitoria
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.musicbrainz.org/pipermail/musicbrainz-style/attachments/20080125/905cc75b/attachment.htm
Brian Schweitzer
2008-01-25 15:52:21 UTC
Permalink
Post by Frederic Da Vitoria
I would do like you, I assumed Italian capitalization was the rule here.
Well, it's how I assumed it was too, but it's not actually layed out in CSG,
and there've been more than a few recent edit notes debating how such
capitalization ought to be done, lacking CSG guidance. (Just one of the
numerous "not in CSG, but people who do CSG know it's basically how it's
done" items to merge into a new CSG.)
Post by Frederic Da Vitoria
Post by Brian Schweitzer
These next two are more of a mess...
3) Minuet & Trio - easy, until you mix in tempos. How do we want to
integrate these?
Minuet. Andante & Trio
Minuet. Andante & Trio. Allegro
Minuet. Andante - pi? molto & Trio
Minuet. Andante - pi? molto & Trio. Andante - allegro - primo tempo
I don't understand your question here.
Another "CSG doesn't say how to do it, so people improvise" type of thing.
I've found a surprising number of these where people are taking "Minuet.
Andante & Trio" and using either "Minuet. Andante" or "Minuet & Trio.
Andante", either leaving out the Trio entirely, or making it seem as though
the Trio is Andante, not the Minuet.

The other question is whether we're then assuming that the Trio inherits the
Andante (which it does unless there's another tempo indicated, like "Minuet.
Andante & Trio. Allegro"), or if we need to restate it, "Minuet. Andante &
Trio. Andante"...

Personally, I suggest doing it like the examples I originally listed - I
think restatement is overkill. Plus, if we applied that idea across the
board, not just "Minuet & Trio"s it would then mean we would have to
"unwrap" Variation I - X to put in a redundant tempo on each Variation as
well. I'd really like, though, to avoid more people either leaving out
Trios, or assuming "Minuet & Trio" is a single work form - why people treat
it any differently, I can't say. (The same, also, goes for all the "&
Coda"s that get left out...)

Brian
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.musicbrainz.org/pipermail/musicbrainz-style/attachments/20080125/f8937d6e/attachment-0001.htm
Jim DeLaHunt
2008-01-27 14:35:58 UTC
Permalink
Post by Brian Schweitzer
Part two of a trio of messages tonight... splitting these to keep the
discussions easily separated.
I. Rondo. Allegro
When it's like that, it's easy. But several things can make it complicated,
and neither CSG nor OperaStyle currently really gives clear guidance.
1) Multi-word types/tempos: sentence style or English caps?
a) Tempo di minueto / Tempo di Minueto
b) Tema con variazioni / Tema con Variazioni
c) Rondo alla turca / Rondo alla Turca
d) Andante grazioso / Andante Grazioso
e) Rondeau en polonaise / Rondeau en Polonaise
etc.
Agreed. Types and tempos are capitalised according to the language in which
they are written (Italian most common, examples in German, English, etc also
exist).
Post by Brian Schweitzer
a) Rondo. Allegro - andante / Rondo. Allegro - Andante
b) Rondo. Allegro & Minore. Allegro di molto / Rondo. Allegro - minore.
allegro di molto / etc.
etc.
I'm not sure what rule you are advocating here.

I see "-" as a separator between tempo markings. I can agree with that. I'd
like us to standardise on "?" (emdash), and accept "-" (minus sign, the one
easy to type on a normal keyboard) as an alternate.

I don't follow when you use "&" and when "/" to separate multiple parts of a
work within a track. We already have a convention for "/".

2) a) Rondo. Allegro ? andante [BrianSchweitzer option before "/", because
of Italian capitalisation]
2) b) Rondo. Allegro ? minore ? allegro di molto [Treating "minore" as a
tempo mark, separating tempos with emdash or minus.]
Post by Brian Schweitzer
These next two are more of a mess...
3) Minuet & Trio - easy, until you mix in tempos. How do we want to
integrate these?
Minuet. Andante & Trio
Minuet. Andante & Trio. Allegro
Minuet. Andante - pi? molto & Trio
Minuet. Andante - pi? molto & Trio. Andante - allegro - primo tempo
Using "/" to separate parts of a work within a track, I'd suggest.

Minuet. Andante ? pi? molto / Trio. Andante ? allegro ? primo tempo
Post by Brian Schweitzer
4) Tema con variazioni
Some releases split track each variation. Some themes have identified
tempos. Some variations have different tempos, sometimes even tempo changes
within a variation. Most of these have not just one or two variations, but
up to as many as twelve...
Tema. Andante con variazioni
Tema. Andante & Variations I - IX
Tema. Andante & Variations I - X & Variation XI. Adagio cantabile &
Variation XII. Allegro
Tema. Andante & Variation I & Variation II & Variation III & Variation IV &
Variation V & Variation VI & Variation VII & Variation VIII & Variation IX &
Variation X & Variation XI. Adagio cantabile & Variation XII. Allegro
etc.
The first reads as ugly to my eye, and we miss the tempos in the last two
variations.
The second we still miss the tempos.
The third reads as a decent compromise to me.
The fourth gets obscenely long very very quickly, especially if you have
intra-variation tempo changes.
I don't quite follow your opening paragraph. Do you mean, some releases
don't split each variation out to its own track, so that one track has
multiple variations?

I agree that the third reads as a decent compromise. Again, I'm not clear
where "&" comes from, when we already have the convention of "/" as a
separator. I might suggest:

Tema. Andante / Variations I - X / Variation XI. Adagio cantabile /
Variation XII. Allegro

Stepping back a bit, part of what makes this hard is that we're trying to
stuff a tree structure of information (work - movements -
parts/movements/variations - tempo marks) into a linear TrackTitle. We're
always going to have examples that drive it to a breaking point. I'd rather
not try to solve that problem, and instead spend energy on bring a
MusicalWork data structure to MusicBrainz.

Thank you for starting this discussion on the CSG cleanup.

-----
-- http://jdlh.com/ Jim DeLaHunt , Vancouver, Canada ?
http://wiki.musicbrainz.org/JimDeLaHunt
--
View this message in context: http://www.nabble.com/-Clean-up-CSG--Capitalization-%28and-placement%29-of-types-and-tempos-tp15083823s2885p15117054.html
Sent from the Musicbrainz - Style mailing list archive at Nabble.com.
David K. Gasaway
2008-01-31 03:47:20 UTC
Permalink
Post by Brian Schweitzer
3) Minuet & Trio - easy, until you mix in tempos. How do we want to
integrate these?
4) Tema con variazioni
Here's my opinion after reviewing all the discussion. If the tempos are going to cause so much fuss, confusion, and deformity, drop them. Tempo markings on major movements is already redundant since movement numbers are included; on the other hand, they are relatively easy to standardize and add a level of _class_ to the whole affair. Not to mention that most people are going to expect the movement tempos, so there's not much point in resisting. :)

However, recording sub-movement tempos seems to me not much more than a value-deprived pedantic exercise.
--
-:-:- David K. Gasaway
-:-:- Email: ***@gasaway.org
-:-:- Web : dave.gasaway.org
Brian Schweitzer
2008-01-31 05:06:46 UTC
Permalink
Post by David K. Gasaway
Post by Brian Schweitzer
3) Minuet & Trio - easy, until you mix in tempos. How do we want to
integrate these?
4) Tema con variazioni
Here's my opinion after reviewing all the discussion. If the tempos are going to cause so much fuss, confusion, and deformity, drop them. Tempo markings on major movements is already redundant since movement numbers are included; on the other hand, they are relatively easy to standardize and add a level of _class_ to the whole affair. Not to mention that most people are going to expect the movement tempos, so there's not much point in resisting. :)
I have to admit, I seriously hope this attitude towards tempos (and it
seems you also mean form too, if movement numbers are sufficient?)
isn't the prevailing one, given that the movement number only
identifies which movement, so you lose the ability to know anything
about the work. (Minuet I - II with no trio, it's this type of
symphony, Minuet & Trio, it's this other type, no minuet, but instead
a Scherzo, it's a third type of symphony. Concerto with movement
tempos in one set of tempos, it's one type of concerto, movement
tempos in a different set of tempos, it's a different type, etc...).
Post by David K. Gasaway
However, recording sub-movement tempos seems to me not much more than a value-deprived pedantic exercise.
While it may be seen as pedantic, some people actually care about that
info... and just are looking for a good way to form it cleanly. :P

Brian
David K. Gasaway
2008-01-31 09:08:10 UTC
Permalink
Post by Brian Schweitzer
I have to admit, I seriously hope this attitude towards tempos (and it
seems you also mean form too, if movement numbers are sufficient?)
isn't the prevailing one, given that the movement number only
identifies which movement, so you lose the ability to know anything
about the work.
If I didn't make the point clear, let me state it categorically: I do
not support removing tempo markings from movement titles.
Post by Brian Schweitzer
While it may be seen as pedantic, some people actually care about that
info... and just are looking for a good way to form it cleanly. :P
If I may be blunt in turn, what's best for a few detail-oriented folks
isn't necessary what's best for MusicBrainz or CSG. You have it
entirely within your power to add whatever information you like to your
personal database, wiki master list, or tags. I do (being a person of
obsessive perfectionist slant myself), but that doesn't mean I'd foist
that cruft on CSG. :)

I've been trying (unsuccessfully?) to frame these discussions in the
perspective of the average user. My views may or may not be
representative, but I'd like to know that these folks have been given
due consideration.
--
-:-:- David K. Gasaway
-:-:- Email: ***@gasaway.org
-:-:- Web: dave.gasaway.org
-:-:- MusicBrainz: dkg
Leiv Hellebo
2008-01-31 15:11:49 UTC
Permalink
Post by David K. Gasaway
I've been trying (unsuccessfully?) to frame these discussions in the
perspective of the average user. My views may or may not be
representative, but I'd like to know that these folks have been given
due consideration.
Me too. On behalf of both me, and others in my household who're going to
look for this or that piece of Mozart... (At least I hope they will :)

And I'd like to throw in yet another consideration:

Deviating from what is printed on sleeves is not something we should do
without good reasons for it. (Yes, we do it very often already by
imposing some common formatting.)

This streamlining of the editing process seems to entail that we
shouldn't even *look* at track listings, except for finding out where in
the masterworks section we should look. This is not how things are
usually done at MB, and it will be confusing to people that are tagging
their files ("how did this information on something that looks like a
chemical element isotope enter my track title?, is this some computer
virus?"),

I hope this will all be resolved for the best eventually, but I am
concerned that this is going too fast, and undiscussed.


Leiv

(My apologies if some of this has been adressed somewhere else without
my noticing it.)
Aaron Cooper
2008-01-31 19:11:16 UTC
Permalink
Post by Leiv Hellebo
Post by David K. Gasaway
I've been trying (unsuccessfully?) to frame these discussions in
the perspective of the average user. My views may or may not be
representative, but I'd like to know that these folks have been
given due consideration.
Me too. On behalf of both me, and others in my household who're
going to look for this or that piece of Mozart... (At least I hope
they will :)
Deviating from what is printed on sleeves is not something we should
do without good reasons for it. (Yes, we do it very often already by
imposing some common formatting.)
This streamlining of the editing process seems to entail that we
shouldn't even *look* at track listings, except for finding out
where in the masterworks section we should look. This is not how
things are usually done at MB, and it will be confusing to people
that are tagging their files ("how did this information on something
that looks like a chemical element isotope enter my track title?, is
this some computer virus?"),
I don't think we need to follow the tracklists for classical because
they often leave out information. What I would like to see (to help
editors who don't know about the classical masterlists), is to put a
brief explanation and a link to the composer's masterlist in their
annotation.

-Aaron
Leiv Hellebo
2008-02-01 03:39:55 UTC
Permalink
Post by Aaron Cooper
Post by Leiv Hellebo
Post by David K. Gasaway
I've been trying (unsuccessfully?) to frame these discussions in the
perspective of the average user. My views may or may not be
representative, but I'd like to know that these folks have been given
due consideration.
Me too. On behalf of both me, and others in my household who're going
to look for this or that piece of Mozart... (At least I hope they will :)
Deviating from what is printed on sleeves is not something we should
do without good reasons for it. (Yes, we do it very often already by
imposing some common formatting.)
This streamlining of the editing process seems to entail that we
shouldn't even *look* at track listings, except for finding out where
in the masterworks section we should look. This is not how things are
usually done at MB, and it will be confusing to people that are
tagging their files ("how did this information on something that looks
like a chemical element isotope enter my track title?, is this some
computer virus?"),
I don't think we need to follow the tracklists for classical because
they often leave out information. What I would like to see (to help
editors who don't know about the classical masterlists), is to put a
brief explanation and a link to the composer's masterlist in their
annotation.
I am of course not saying we should blindly follow track lists.

But I am a bit shocked after my CD ripper told me yesterday that a
simple one and a half minute Mozart solo piano track is called:

"Modulating Prelude to Concerto for Piano in F major transitioning to C
major, K. Anh C 15.11/(K3) 626aII/I/KV deest Fr 1765a" [1]

This title behooves the study room more than the living room. Do we have
to have to all of this in the track titles?

As someone else said recently on a completely different subject: One of
the most important tags set by Picard is the one identifying the track
at MB. What if e.g. the MB track page could link to the full-blown
title. Can I then get a dumbed down version of the track title by
proposed by my ripper and subsequently on display in my living room player?


leivhe


[1] Actually, this title proved to be wrong. The track listing says
"Modulierendes Pr?ludium (F-e), KV6 deest (1776/77)", and with some help
from Brian I could confirm this by following the score at dme.mozarteum.at.
Aaron Cooper
2008-02-01 03:58:42 UTC
Permalink
Post by Leiv Hellebo
Post by Aaron Cooper
Post by Leiv Hellebo
Post by David K. Gasaway
I've been trying (unsuccessfully?) to frame these discussions in
the perspective of the average user. My views may or may not be
representative, but I'd like to know that these folks have been
given due consideration.
Me too. On behalf of both me, and others in my household who're
going to look for this or that piece of Mozart... (At least I hope
they will :)
Deviating from what is printed on sleeves is not something we
should do without good reasons for it. (Yes, we do it very often
already by imposing some common formatting.)
This streamlining of the editing process seems to entail that we
shouldn't even *look* at track listings, except for finding out
where in the masterworks section we should look. This is not how
things are usually done at MB, and it will be confusing to people
that are tagging their files ("how did this information on
something that looks like a chemical element isotope enter my
track title?, is this some computer virus?"),
I don't think we need to follow the tracklists for classical
because they often leave out information. What I would like to see
(to help editors who don't know about the classical masterlists),
is to put a brief explanation and a link to the composer's
masterlist in their annotation.
I am of course not saying we should blindly follow track lists.
But I am a bit shocked after my CD ripper told me yesterday that a
"Modulating Prelude to Concerto for Piano in F major transitioning to C
major, K. Anh C 15.11/(K3) 626aII/I/KV deest Fr 1765a" [1]
This title behooves the study room more than the living room. Do we
have to have to all of this in the track titles?
As someone else said recently on a completely different subject: One
of the most important tags set by Picard is the one identifying the
track at MB. What if e.g. the MB track page could link to the full-
blown title. Can I then get a dumbed down version of the track title
by proposed by my ripper and subsequently on display in my living
room player?
[1] Actually, this title proved to be wrong. The track listing says
"Modulierendes Pr?ludium (F-e), KV6 deest (1776/77)", and with some
help from Brian I could confirm this by following the score at
dme.mozarteum.at.
What about a simple Picard plugin that would chop out the extra
catalog info? We could trim it down to "Modulating Prelude to
Concerto for Piano in F major transitioning to C major, KV deest Fr
1765a". That's not a lot better, but I'm not sure why the transition
is in there anyways (unless the score explicitly has that in the
title. Maybe we'd be happy with "Modulating Prelude to Concerto for
Piano in F major, KV deest" - I'm sure we could rig up something that
could pull this off in Picard.

-Aaron
Jim DeLaHunt
2008-02-01 16:43:31 UTC
Permalink
Post by Leiv Hellebo
Post by David K. Gasaway
I've been trying (unsuccessfully?) to frame these discussions in the
perspective of the average user. My views may or may not be
representative, but I'd like to know that these folks have been given
due consideration.
Me too. On behalf of both me, and others in my household who're going to
look for this or that piece of Mozart... (At least I hope they will :)
This is a good point.

I think the source of the tension is that we're trying to make one data
field (the MB TrackTitle string) perform two roles: 1. give the tagger a
useful string to put in the "track title" tag in the digital music file,
which the music player displays; and 2. record a bunch of facts about the
musical work which the track represents.

What we have to do in this CSG review is to strike a good balance. So far,
I think that the CSG helps rather than hinders a non-expert contributor who
is used to classical music conventions. Most of the arcane discussion is
about edge cases.
Post by Leiv Hellebo
Deviating from what is printed on sleeves is not something we should do
without good reasons for it. (Yes, we do it very often already by imposing
some common formatting.)
For ClassicalMusic, I think we have those good reasons. I stated it at
length in another post
(http://www.nabble.com/-Clean-up-CSG--Classical-and-Release-Language-tt15083579s2885.html):
"...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."
Post by Leiv Hellebo
I hope this will all be resolved for the best eventually, but I am
concerned that this is going too fast, and undiscussed.
Well, I see the discussion taking place right here, and no-one is making
changes yet. We're just raising issues and looking for consensus. If by
"too fast, and undiscussed" you mean there are important MB authorities who
haven't chimed in yet, can we expect that they will within, say, a week or
two? Or do we run the risk that even this discussion, followed by a Request
for Comment, followed by a Request for Veto won't include the right voices?


-----
-- http://jdlh.com/ Jim DeLaHunt , Vancouver, Canada ?
http://wiki.musicbrainz.org/JimDeLaHunt
--
View this message in context: http://www.nabble.com/-Clean-up-CSG--Capitalization-%28and-placement%29-of-types-and-tempos-tp15083823s2885p15224343.html
Sent from the Musicbrainz - Style mailing list archive at Nabble.com.
Leiv Hellebo
2008-02-01 17:30:08 UTC
Permalink
Post by Jim DeLaHunt
Post by Leiv Hellebo
I hope this will all be resolved for the best eventually, but I am
concerned that this is going too fast, and undiscussed.
Well, I see the discussion taking place right here, and no-one is making
changes yet.
There are currently open Bach cantata edits done to make track titles
conform to CSGS/JSBach that seem to get voted in, despite the titles
already were according to the cantata style described in the CSG. When
these edits are accepted, they will no longer conform to the CSG.

(NOTE: This is _not_ to discredit the great work that's going on with
the master lists.)

I'll respond on other points when I have the time.


l
Frederic Da Vitoria
2008-02-02 06:08:08 UTC
Permalink
Post by Leiv Hellebo
Post by Jim DeLaHunt
Post by Leiv Hellebo
I hope this will all be resolved for the best eventually, but I am
concerned that this is going too fast, and undiscussed.
Well, I see the discussion taking place right here, and no-one is making
changes yet.
There are currently open Bach cantata edits done to make track titles
conform to CSGS/JSBach that seem to get voted in, despite the titles
already were according to the cantata style described in the CSG. When
these edits are accepted, they will no longer conform to the CSG.
(NOTE: This is _not_ to discredit the great work that's going on with
the master lists.)
I feel this is bad news. What are we doing here? Two differing rules for the
same thing? Are we going crazy? Do you have an example of such an edit?
--
Frederic Da Vitoria
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.musicbrainz.org/pipermail/musicbrainz-style/attachments/20080202/382367d6/attachment.htm
Leiv Hellebo
2008-02-04 05:56:51 UTC
Permalink
On Feb 1, 2008 12:28 PM, Leiv Hellebo
There are currently open Bach cantata edits done to make track titles
conform to CSGS/JSBach that seem to get voted in, despite the titles
already were according to the cantata style described in the CSG. When
these edits are accepted, they will no longer conform to the CSG.
(NOTE: This is _not_ to discredit the great work that's going on with
the master lists.)
I feel this is bad news. What are we doing here? Two differing rules for
the same thing? Are we going crazy? Do you have an example of such an edit?
I don't think it is very interesting, but since you ask, here's the
stuff I was thinking of:

http://musicbrainz.org/show/edit/?editid=8258236

Important note: From the little I've seen of what bgibbard has been
doing, it is excellent stuff. In this edit of course he's right "No. 21"
should go.

And translating to German might very well be correct for the BIS CD
track listing (my Suzuki-Bach passions have English in parentheses).

l
Brian Schweitzer
2008-02-01 08:19:41 UTC
Permalink
Post by Leiv Hellebo
But I am a bit shocked after my CD ripper told me yesterday that a
"Modulating Prelude to Concerto for Piano in F major transitioning to C
major, K. Anh C 15.11/(K3) 626aII/I/KV deest Fr 1765a" [1]
This title behooves the study room more than the living room. Do we have
to have to all of this in the track titles?
While the point is admittedly valid, I would draw some exception to
the particular example chosen, given that it is perhaps one of the top
2 Mozart works with the absolutely most confusing history - and of all
works that I can think of, still the one that draws the most questions
- and there, specifically *because* most liners only list "Modulating
Prelude" (on the few CDs which have this work). Would the Brilliant
Classics listing be any less "extreme", even with the non-K6 catalog
numbers stripped out, if you put the CD in and got

Prelude Transitions for Piano for Concertos for Piano in C major, K.
Anh C 15.11 Nos. a-x: (no tempo indication) - andantino / Modulating
Prelude for Piano in F major to E minor, K. deest NMA IX/27/2 No. 2:
(no tempo indication)

for a track title?

I agree, the titles can be complex. But if you put in your common
Symphonies 40 and 41 CD, you get:

1. Symphony No. 40 in G minor without clarinets, K. 550: I. Allegro molto
2. Symphony No. 40 in G minor without clarinets, K. 550: II. Andante
3. Symphony No. 40 in G minor withou clarinets, K. 550: IIIa.
Minuetto. Allegretto / IIIb. Trio & IIIa. Minuetto. Allegretto
4. Symphony No. 40 in G minor without clarinets, K. 550: IV. Allegro assai
5. Symphony No. 41 in C major, K. 551 "Jupiter": I. Allegro vivace
6. Symphony No. 41 in C major, K. 551 "Jupiter": II. Andante cantible
7. Symphony No. 41 in C major, K. 551 "Jupiter": IIIa. Minuetto.
Allegretto / IIIb. Trio & IIIa. Minuetto. Allegretto
8. Symphony No. 41 in C major, K. 551 "Jupiter": IV. Allegro molto

That's not so complicated, is it? Now, yes, I can pick out a dozen
works with rather complex histories, especially when we go completely
out of the 626 (plus letters) standard Kochel catalog works. When we
dip into the appendix works (and worse, the totally non-Kochel works),
things get complicated... hence why there has been a new Kochel
catalog under way for 8+ years. But the goal of every version has
been to keep it simple for the "standard" works - hence why the
Jupiter symphony has always been 551, the Requiem 626, La finta
giardiniera 196, etc.

(Actually, it's specifically *because* of the catalogs trying to keep
those numbers the same between versions that we did end up with the
letters and such; it would have been much easier to simply renumber
the entire list, instead of trying to insert without disrupting the
numbers for the well known works...)

But in any case, cooperaa asked me on IRC today why I used K6 / K3 /
K2 / KV instead of just K(most recent) or KV / K2 / K3 / K6.

In a perfect system, we could have all the elements in different
fields, and build your track title exactly as you wish. We don't have
that yet. In our current system, just like typography, either it's
there, or its not. If I just used K(most recent), then 1) a person
with a CD that uses original KV numbers (or K2 or K3 numbers) and not
K6 numbers would not get something that made sense to them. (You put
in a CD for "Sonata for Violin & Piano No. 29 in A major" and get K.
295d, not KV 305, etc.). 2) the same for searching - if you search
for "KV 305" (now that I inserted spaces around the /'s to work around
a lucene issue), you'll get all 295d/305's. That's not true if we
leave out the former catalog numbers. 3) If you want all the catalog
numbers in your track, they are there, and you get them.

The corallary of #3 is the answer to cooperaa's second question, why I
used that particular order. I haven't written the line of script to
do it, but I would have to believe the regex line for tagger script to
do "K. ###/(ignore until you hit end of title, a " or a :)" is much
easier than "K. (ignore until you hit the last /, then include all
after that last /". If you don't want more than K(most recent), it's
ordered such that it's in the order most simple for regex filtering.

Brian
Leiv Hellebo
2008-02-04 05:09:18 UTC
Permalink
Post by Brian Schweitzer
Post by Leiv Hellebo
But I am a bit shocked after my CD ripper told me yesterday that a
"Modulating Prelude to Concerto for Piano in F major transitioning to C
major, K. Anh C 15.11/(K3) 626aII/I/KV deest Fr 1765a" [1]
This title behooves the study room more than the living room. Do we have
to have to all of this in the track titles?
While the point is admittedly valid, I would draw some exception to
the particular example chosen,
It would be wrong to base an evaluation of the copy-from-CSGS/Mozart
methodology on this example alone. Yet, it clearly shows that the
scholarly-approach-fits-all attitude has its clear limitations: It
should not be necessary to read an instruction manual to appreciate the
hypothetical soundtrack "Piano Teacher 2".

(I am not saying that you are suggesting to use it for soundtracks and
everywhere else, but I haven't seen any discussion on scope and/or
limitations of the CSGS/M yet.)

Neither should I need to have your help to read up on K?chel editions to
be able to decipher my own track titles. (I'd much rather have
"Modulating Prelude, 1765", possibly with the keys, thank you. If I knew
what "KV deest" meant, maybe I'd like to have that there as well.)

Besides: What if the track listing on my CD uses another edition than
that of the CSGS/M, and I'm particular about my K?chels? Then what?

[snip]
Post by Brian Schweitzer
I agree, the titles can be complex. But if you put in your common
1. Symphony No. 40 in G minor without clarinets, K. 550: I. Allegro molto
2. Symphony No. 40 in G minor without clarinets, K. 550: II. Andante
3. Symphony No. 40 in G minor withou clarinets, K. 550: IIIa.
Minuetto. Allegretto / IIIb. Trio & IIIa. Minuetto. Allegretto
4. Symphony No. 40 in G minor without clarinets, K. 550: IV. Allegro assai
[Snip No. 41]
Post by Brian Schweitzer
That's not so complicated, is it?
It's better, but not quite there.

I don't really want to get bogged down in a particular discussion on
this or that work, since it to me seems obvious that we do need to
strike the balance between the needs of different user types. At least
we need to do so in the absence of more fields to enter our data and
some yet-to-be-thought-of way of handling them, and in the absence of
the dumb-my-classical-picard-plugin, which I believe will be hard to write).

In short my suggestion on track titles is that the medium packaging
should be regarded as the prime arbiter, with the CSGS/* performing
miscellaneous helping roles, much like we use Wikipedia today. (Yes, I
do think copy-pasting from CSGS/* should be encouraged, but I'm
skeptical towards entering typographically advanced stuff there.))

The rationale behind this I hope is discernible from the following
discussion on your formatting of No. 40, and various related issues:

For No. 40 I think we might disagree on at least three places:

1) Your decision to leave out the common name "Great"
2) Your decision to include the "With/Without Clarinets" part.
3) Your formatting of movement number three

1) and 2) concerns level of detail: How much should we cram into the
title? How do we measure what's relevant and what's not?

3) is not so interesting to discuss, but since I dislike it pretty much,
I'll say some about it.

----

1) Would the title you and CSGS/Mozart suggest be better or worse with
"Great" (or "The Great" or "The Great G minor") included:

Symphony No. 40 in G minor Without Clarinets, K. 550 "Great": II. Andante?

(After all you're currently using this bastard:

Symphony No. 25 in G minor, K. 173dB / 183 "Little G Minor": II. Andante

)

IMO this should be left to the track listing, and to the ad hoc
discussion among the editor and the voters:

The submitter might say "my track listing has/has not the common name
included and I'd like to follow/override it", some discussion might
follow, but in the end a good enough title for MusicBrainz will be
agreed upon. I may personally disagree and vote "no", but it's good
enough for MusicBrainz.

* If someone takes the trouble to add data for their personal music to
MB, we might do a disservice to the community if we start saying: "Thank
you for adding this, but you should note that the common name you
include here is not present on CSGS/ObscureComposer. Please step aside
while we sanitise this title now."

And do we want to continue:

"If you have any objections to how your formatting of the data on your
CD should be done, then please voice your opinion on the wiki/mailing
lists, and in due course the task force of CSGS/ObscureComposer will
start a serious discussion about in on mb-style, where such discussions
are always concluded in a sensible manner."

If we go for "the CSGS/* is always right", perhaps a common policy on
common names is in order:

"How widely used does a common name have to be to become a certified MB
common name? Does it matter if the composer coined it?"
"What if the composer disliked the common name?"

Without a common policy, the inclusion or the exclusion for each and
every one instance may be reasonably discussed and agreed upon (but
hardly always) by whoever is active at the time of discussion:

"Should we translate 'Rheinische' (by Schumann) for English track titles?"
"What about 'Path?tique' (by Beethoven), then?"
"Is 'Great' and 'Little' relevant common names at all, when we already
have the symphony numbers to distinguish them? Shortening the names is
in general a good idea, isn't it?"

I doubt we will ever be satisfied with what we agree upon, in the
unlikely event that we actually care enough to come to an agreement on
something at all. After all, who cares what common name is (NOT)
included on this particular release from 2003 which some random new
editor brings in?

And remember, the reason we started discussing this is that we feel we
need more fields for the track title.
--
2) With or Without Clarinets

Moving onto something which is slightly more musically interesting now.

(Background: Mozart's early symphonies are without clarinets, but the
late ones have them. For No. 40, Mozart wrote a second version (the NMA
score says "2. Fassung", there is no ArtistIntent for "With/WIthout C.")
of the score where he included a voice for clarinets, and altered the
oboe part a bit. The Neue Mozart-Ausgabe (henceforth NMA) has them both.
About No. 40 Wikipedia says: "Notably missing are trumpets and timpani".

Because all orchestras have clarinets these days, and because the
version with the clarinets sound better (according to some review I
found), recordings without clarinets are rare.)

The inclusion of With or Without Clarinets is IMO only relevant in
contexts where both versions of the symphony are present. This may e.g.
be a single disc release, a box set, or - the use case Jim mentioned in
another post - collection wide for the classical freak. In these cases,
this extra information is conceivably useful, much like the cluttering
of performer info into release titles is currently useful for
distinguishing between releases with otherwise identical names.

But, if track lists doesn't mention this, why should we bother?

MBVoter: "Is your version the one with clarinets, or not?"
Submitter: "Uh, dunno, how does a clarinet sound? Do you have some place
where I can upload the mp3's for this, so you guys can help me out?"

[Yes, Submitter probably has the version with the clarinets, so please
think of this as a possible example of the more general case where we
have two or more versions of something. According to the booklet of my
latest H?ndel's Messiah, there exists 10 versions of it...]

Personally, I do know how clarinets sound (at least the modern versions,
I may have trouble picking out the sound from copies of older versions,
and I guess this is more difficult if I have to pick it out from the
rest of the orchestral noise.) Nonetheless, I was previously unaware
that my three recordings of No. 40 all use clarinets. (I've probably
read it in booklets, and forgotten it right afterwards.)

And being somewhat of a classical freak myself, I do think that the
presence or not of an instrument, the use of (perhaps only some) old
instruments, the size of the orchestra, the tempo, the omission of
repeats.... etc. is fascinating.

Still, this is stuff that belongs in some wiki, or perhaps the
track/release annotations. The important thing is not whether Mozart had
the time and maybe made an extra dime on making a new orchestration of
some piece of music.

I do not want by default in the track titles that enter my living room,
because the important thing here is, to me, and for lack of better
words, the "ta-da-dam ta-da-dam ta-da-daaa, ta-da-dam ta-da-dam
ta-da-da" of the opening, and the excitement of the last movements, and ...

A final note: A remove-the-k?chels regexp is probably not so hard, but
the more diverse info you put in there, the more difficult it will be to
write a dumb-my-classical to help out those who just want the minimum.

----

3) Some words on the ugliness of

"IIIa. Minuetto. Allegretto / IIIb. Trio & IIIa. Minuetto. Allegretto"

(First I have to admit I don't even get what you're trying to convey by
the use of "/" and "&". If it's MultipleTitleStyle you want it should be
"/", what's commonly used at MB for classical subparts of movements
within the same track is, of course, "-")

Contrast your suggestion with the track list of my three No. 40:

"III. Menuetto: Allegretto & Trio"
"III. Menuetto: Allegretto - Trio"
"III. Menuetto: Allegretto"

So what if they're inconsistent, and that none mention the da capo? So
what if the last one doesn't even bother mentioning the Trio?
Why should we improve on these titles?

Perhaps even a case for ArtistIntent could be made for two tiny details:

1) Using the colon as delimiter lines up perfectly with SubtitleStyle
and what's on top here:
http://dme.mozarteum.at/DME/nma/scan.php?vsep=111&l=1&p1=161#161

2) The score use "Menuetto", which according to
http://it.wikipedia.org/wiki/Minuetto is allowed.

What I don't get is:

1) Why should CSGS/M use a bastard form that cannot be found *anywhere
else*? (Why don't scholars use it, e.g.?)
2) How are your suggestions better than my track lists?


leivhe
David K. Gasaway
2008-02-05 10:45:59 UTC
Permalink
Post by Brian Schweitzer
But in any case, cooperaa asked me on IRC today why I used K6 / K3 /
K2 / KV instead of just K(most recent) or KV / K2 / K3 / K6.
If I just used K(most recent), then 1) a person
with a CD that uses original KV numbers (or K2 or K3 numbers) and not
K6 numbers would not get something that made sense to them. (You put
in a CD for "Sonata for Violin & Piano No. 29 in A major" and get K.
295d, not KV 305, etc.).
The assumption being that "K. 295d/305" will make any more sense to the
user than "K. 295d" when they expected "KV 305". I suspect not, unless
they user already understands the ins and outs of Koechel, in which
case there's no problem with the latter. :)
Post by Brian Schweitzer
2) the same for searching - if you search
for "KV 305" (now that I inserted spaces around the /'s to work around
a lucene issue), you'll get all 295d/305's.
This is where a masterlist comes in handy. While I see the value in
what you propose, I don't think it behooves the MB database to be in
the business of recording the failings (form lack of information, of
course) of Koechel and his successors.
Post by Brian Schweitzer
3) If you want all the catalog
numbers in your track, they are there, and you get them.
And if I don't want all the catalog numbers in my track (which I
don't), then I'm left to find my own solution. I could try to write a
try regular expression to remove the excess. It might not always work.
It's not going to ship with Picard. I guess either way, one of us
loses, so which is better overall solution? IMO, it's better to
standardize on one set of catalog numbers.
--
-:-:- David K. Gasaway
-:-:- Email: ***@gasaway.org
-:-:- Web: dave.gasaway.org
-:-:- MusicBrainz: dkg
Jim DeLaHunt
2008-02-06 05:15:57 UTC
Permalink
Brian, I think I'm with David on this one. It's better to standardize on one
set of catalog numbers.
Post by David K. Gasaway
Post by Brian Schweitzer
But in any case, cooperaa asked me on IRC today why I used K6 / K3 /
K2 / KV instead of just K(most recent) or KV / K2 / K3 / K6. ...[snip]...
The assumption being that "K. 295d/305" will make any more sense to the
user than "K. 295d" when they expected "KV 305". I suspect not, unless
they user already understands the ins and outs of Koechel, in which case
there's no problem with the latter. :)
... While I see the value in what you propose, I don't think it behooves
the MB database to be in the business of recording the failings (form lack
of information, of course) of Koechel and his successors.
All these different catalogues are just different attempts to put
identifiers on musical works, right? By including multiple catalogue
numbers, you are include multiple identifiers *and* how those identifiers
correlate. By having the CSG require multiple catalogue numbers we ask every
editor to be able to cite the same set of identifiers, and ? this is what
makes it really hard ? in the same combination.

Suppose someone enters "K. 295d/305" for one release title, and I enter "K.
295d/306" in another release title? How will a third contributor know which
is correct enouch to copy? How would a fourth editor know how to correct
the mistake? If our standard is "KV 305", the picture is simpler.

In this impoverished era when we don't have a MusicalWork entity in our
structure, I think it's sufficient to have the CSG require enough
information to identify the work reliably. At some later point, when we
have a MusicalWork entity, we use this information to attach track and
release entries to the MusicalWork. Then the experts can attach the
multiple catalogue identifiers to that entity once, correctly, for everyone.

Actually, that might be an argument for a CSG that asks for a specific
preferred catalog number (e.g. "KV 305"), but accepts other catalogs instead
(e.g. "K. 295d"). The other catalog number, if accurate, is still a clear
identifier. A bot has a chance of fixing it later.
Post by David K. Gasaway
Post by Brian Schweitzer
3) If you want all the catalog
numbers in your track, they are there, and you get them.
And if I don't want all the catalog numbers in my track (which I don't),
then I'm left to find my own solution. I could try to write a try regular
expression to remove the excess. It might not always work. It's not
going to ship with Picard. I guess either way, one of us loses, so which
is better overall solution? IMO, it's better to standardize on one set of
catalog numbers.
This is another good point. Either choice will satisfy some users and
dissatisfy others. Having one catalog number at least identifies the work,
and its simpler to explain to contributors, and there's not the chance of
inconsistent combinations of identifiers. Those are positives. The loss of
the ability to search by other index numbers is a negative, but a smaller
one on the whole.


-----
-- http://jdlh.com/ Jim DeLaHunt , Vancouver, Canada ?
http://wiki.musicbrainz.org/JimDeLaHunt
--
View this message in context: http://www.nabble.com/-Clean-up-CSG--Capitalization-%28and-placement%29-of-types-and-tempos-tp15083823s2885p15300067.html
Sent from the Musicbrainz - Style mailing list archive at Nabble.com.
Brian Schweitzer
2008-02-04 11:03:30 UTC
Permalink
Post by Leiv Hellebo
Post by Brian Schweitzer
Post by Leiv Hellebo
But I am a bit shocked after my CD ripper told me yesterday that a
"Modulating Prelude to Concerto for Piano in F major transitioning to C
major, K. Anh C 15.11/(K3) 626aII/I/KV deest Fr 1765a" [1]
This title behooves the study room more than the living room. Do we have
to have to all of this in the track titles?
While the point is admittedly valid, I would draw some exception to
the particular example chosen,
It would be wrong to base an evaluation of the copy-from-CSGS/Mozart
methodology on this example alone. Yet, it clearly shows that the
scholarly-approach-fits-all attitude has its clear limitations: It
should not be necessary to read an instruction manual to appreciate the
hypothetical soundtrack "Piano Teacher 2".
(I am not saying that you are suggesting to use it for soundtracks and
everywhere else, but I haven't seen any discussion on scope and/or
limitations of the CSGS/M yet.)
Well, realistically, would we not apply CSG to a Mozart, Bach, or
Beethoven track on a soundtrack as well? If CSG only for "classical
releases"?
Post by Leiv Hellebo
Neither should I need to have your help to read up on K?chel editions to
be able to decipher my own track titles. (I'd much rather have
"Modulating Prelude, 1765", possibly with the keys, thank you. If I knew
what "KV deest" meant, maybe I'd like to have that there as well.)
I know Mozart is not the only composer for which the term deest is
used. (Its a Latin term used in music to indicate "not listed" (in
the catalog)).
Post by Leiv Hellebo
Besides: What if the track listing on my CD uses another edition than
that of the CSGS/M, and I'm particular about my K?chels? Then what?
Quite honestly, then the CD's listing is rather out of date (though
I've seen it as well)... Kochel 6 came out in 1964, and was the most
recent version - also a reason why there are a decent number of deest
works. As for being particular about your Kochels, that's partially
why I do have any/all Kochel's in there. The alternative would be our
not only having each release by box set and language, but also a
listing for each, K1, K2, K3, and K6, (same for Chopin's 4 or 5
different catalogs, etc, etc)
Post by Leiv Hellebo
Post by Brian Schweitzer
I agree, the titles can be complex. But if you put in your common
1. Symphony No. 40 in G minor without clarinets, K. 550: I. Allegro molto
2. Symphony No. 40 in G minor without clarinets, K. 550: II. Andante
3. Symphony No. 40 in G minor withou clarinets, K. 550: IIIa.
Minuetto. Allegretto / IIIb. Trio & IIIa. Minuetto. Allegretto
4. Symphony No. 40 in G minor without clarinets, K. 550: IV. Allegro assai
<snip>
Post by Leiv Hellebo
I don't really want to get bogged down in a particular discussion on
this or that work, since it to me seems obvious that we do need to
strike the balance between the needs of different user types. At least
we need to do so in the absence of more fields to enter our data and
some yet-to-be-thought-of way of handling them, and in the absence of
the dumb-my-classical-picard-plugin, which I believe will be hard to write).
In short my suggestion on track titles is that the medium packaging
should be regarded as the prime arbiter, with the CSGS/* performing
miscellaneous helping roles, much like we use Wikipedia today. (Yes, I
do think copy-pasting from CSGS/* should be encouraged, but I'm
skeptical towards entering typographically advanced stuff there.))
See, this is where, when it comes to classical, I disagree. I have
found liners to be of very little use when it comes to finding good
quality information for classical work identification - so much so,
that on the Mozart sites I frequent, nearly half the posts are along
the lines of "my liner says "this"... anyone have a clue which work
it actually is?"
Post by Leiv Hellebo
The rationale behind this I hope is discernible from the following
1) Your decision to leave out the common name "Great"
2) Your decision to include the "With/Without Clarinets" part.
3) Your formatting of movement number three
1) and 2) concerns level of detail: How much should we cram into the
title? How do we measure what's relevant and what's not?
3) is not so interesting to discuss, but since I dislike it pretty much,
I'll say some about it.
----
1) Would the title you and CSGS/Mozart suggest be better or worse with
Symphony No. 40 in G minor Without Clarinets, K. 550 "Great": II. Andante?
Symphony No. 25 in G minor, K. 173dB / 183 "Little G Minor": II. Andante
Hence why I think peer-review is a good thing; It's quite possible for
me to miss something, esp. as none of these common names is
composer-assigned. While it's happened a few times for tempos on the
Mozart list, I think the best example yet has been the JS Bach list,
where there has been more peer-review for such details. (...and /me
adds "Great" in on the local copy of the wikipage for the next update
:P)
Post by Leiv Hellebo
IMO this should be left to the track listing, and to the ad hoc
The submitter might say "my track listing has/has not the common name
included and I'd like to follow/override it", some discussion might
follow, but in the end a good enough title for MusicBrainz will be
agreed upon. I may personally disagree and vote "no", but it's good
enough for MusicBrainz.
Why? As we all know, liners rang in quality from "they ought to be
ashamed of themselves" to "rather quite good, in fact". However, all
are describing the same body of works. Why ought we to base the
inclusion or exclusion of any pertinent detail based on whether or not
a label decided the inclusion of that same detail would promote sales?
That's where I see us, as the database of releases, as different from
the labels, who have the sole goal of sales.
Post by Leiv Hellebo
* If someone takes the trouble to add data for their personal music to
MB, we might do a disservice to the community if we start saying: "Thank
you for adding this, but you should note that the common name you
include here is not present on CSGS/ObscureComposer. Please step aside
while we sanitise this title now."
Sanitize, scrutinize, or standardize? Were we not talking classical,
but pop, and someone entered a CD single with:

1. Top 40 song
2. Same top 40 song as #1
3. Same top 40 song as #1
4. Same top 40 song as #1
5. Some other random song

would we not be quick to ask for the missing version/mix information
to be added? This is the same principle, as I see it. "Adagio" we
all agre is too vague. But when the same work has multiple versions,
is it really so much to ask that the correct work be identified fully,
and named properly?
<snip to the discussion on common names - as mentioned, the exclusion
was not intentional>
Post by Leiv Hellebo
2) With or Without Clarinets
Moving onto something which is slightly more musically interesting now.
(Background: Mozart's early symphonies are without clarinets, but the
late ones have them. For No. 40, Mozart wrote a second version (the NMA
score says "2. Fassung", there is no ArtistIntent for "With/WIthout C.")
of the score where he included a voice for clarinets, and altered the
oboe part a bit. The Neue Mozart-Ausgabe (henceforth NMA) has them both.
About No. 40 Wikipedia says: "Notably missing are trumpets and timpani".
Because all orchestras have clarinets these days, and because the
version with the clarinets sound better (according to some review I
found), recordings without clarinets are rare.)
The inclusion of With or Without Clarinets is IMO only relevant in
contexts where both versions of the symphony are present. This may e.g.
be a single disc release, a box set, or - the use case Jim mentioned in
another post - collection wide for the classical freak. In these cases,
this extra information is conceivably useful, much like the cluttering
of performer info into release titles is currently useful for
distinguishing between releases with otherwise identical names.
But, if track lists doesn't mention this, why should we bother?
Not quite, nor is the history quite as you describe. In short,
however, clarinets were frequently included in symphonies then too,
but the first version of #40 did not include them. There was a
rewritten version later which did include them. I have several
recording each way.

However, your claim that only a release containing both would have
reason to include the information is puzzling. Were this not
classical, would we say the same? If I have a CD which has track 12
listed as "Foo (radio edit)", is it legitimate to leave out the (radio
edit) as there's not a non-radio edit version on the same release? If
I have a jazz CD which has track 7 as "Fubar (take 3)", can I leave
out the "take 3" because takes 1 and 2 aren't on the CD? Of course
not. Let's say I want to find all releases that do or don't have
clarinets. This information is not "clutter", nor is it "classical
freak only". It is identifying which version of a work is on the CD.
I just listened to a CD of the Levin-completion of the Requiem. It's
rather significantly different from the Sussmeyer version - yet should
we not identify which version is on a release, simply because both
versions do not appear on the same release? If so, then classical is
the only place in the database we're making this exception for
versioning information.
Post by Leiv Hellebo
MBVoter: "Is your version the one with clarinets, or not?"
Submitter: "Uh, dunno, how does a clarinet sound? Do you have some place
where I can upload the mp3's for this, so you guys can help me out?"
Rather, "is there a clarinetist included within the credited artists
on your CD?" or, perhaps, God forbid, but, "might you do a quickie 3
minute Googling to actually check on a detail about your CD?" Again,
we ask this kind of thing all the time outside of classical - why
ought that slight bit of effort really be such a grusome task when
suddenly we're dealing with "the question of what work is on your CD
anyways?"
Post by Leiv Hellebo
Personally, I do know how clarinets sound (at least the modern versions,
I may have trouble picking out the sound from copies of older versions,
and I guess this is more difficult if I have to pick it out from the
rest of the orchestral noise.) Nonetheless, I was previously unaware
that my three recordings of No. 40 all use clarinets. (I've probably
read it in booklets, and forgotten it right afterwards.)
And being somewhat of a classical freak myself, I do think that the
presence or not of an instrument, the use of (perhaps only some) old
instruments, the size of the orchestra, the tempo, the omission of
repeats.... etc. is fascinating.
I think you've overlooked the fact, however, that wiki's rather
horrible description aside (yes, I've read it before), it's not simply
"he added clarinets". There were some rather significant changes made
within the work, enough so that, as you point out, NMA found it
worthwhile to print the entire score twice, once the version with and
once the version without clarinets.
Post by Leiv Hellebo
Still, this is stuff that belongs in some wiki, or perhaps the
track/release annotations. The important thing is not whether Mozart had
the time and maybe made an extra dime on making a new orchestration of
some piece of music.
I think I'll make the same argument next time someone quibbles about
the difference between (take 1) and (take 2), or (radio edit),
(DonNMike remix) , and (original edit). Think anyone will buy this
argument there?
Post by Leiv Hellebo
A final note: A remove-the-k?chels regexp is probably not so hard, but
the more diverse info you put in there, the more difficult it will be to
write a dumb-my-classical to help out those who just want the minimum.
My sense on this, by now, is quite likely common knowledge. However,
to summarize, a "play to the lowest common factor" concept for me
makes little sense, in that it involves throwing most of the
information away - actually, rather similar to the approach of many of
the labels and liners we all complain about all the time. If someone
wants dumbed down track titles, CSG in of itself is most likely more
than they want.
Post by Leiv Hellebo
3) Some words on the ugliness of
"IIIa. Minuetto. Allegretto / IIIb. Trio & IIIa. Minuetto. Allegretto"
(First I have to admit I don't even get what you're trying to convey by
the use of "/" and "&". If it's MultipleTitleStyle you want it should be
"/", what's commonly used at MB for classical subparts of movements
within the same track is, of course, "-")
"III. Menuetto: Allegretto & Trio"
"III. Menuetto: Allegretto - Trio"
"III. Menuetto: Allegretto"
So what if they're inconsistent, and that none mention the da capo? So
what if the last one doesn't even bother mentioning the Trio?
Why should we improve on these titles?
Are you forgetting the discussion we had in edit notes where you
yourself pointed out that I was leaving out the Da capo reprises, and
suggested I add them back in? :D

Brian
Leiv Hellebo
2008-02-04 13:34:37 UTC
Permalink
Post by Brian Schweitzer
Post by Leiv Hellebo
So what if they're inconsistent, and that none mention the da capo? So
what if the last one doesn't even bother mentioning the Trio?
Why should we improve on these titles?
Are you forgetting the discussion we had in edit notes where you
yourself pointed out that I was leaving out the Da capo reprises, and
suggested I add them back in? :D
No, I was not pointing them out to have you add them back in:

http://musicbrainz.org/show/edit/?editid=8209636

I mentioned the da capo of Menuet I to stop you from considering the
Menuet I to be from a different movement than Menuet II.

Your suggestion was: II. Menuet I / III. Menuet II, and the last
movement, therefore became: "IV. Allegro".

I have never seen it done like this anywhere else.
David K. Gasaway
2008-02-05 11:21:00 UTC
Permalink
Post by Brian Schweitzer
However, your claim that only a release containing both would have
reason to include the information is puzzling. Were this not
classical, would we say the same? If I have a CD which has track 12
listed as "Foo (radio edit)", is it legitimate to leave out the (radio
edit) as there's not a non-radio edit version on the same release?
It's really hard to compare the "(version)" compromise adopted for pop
music to score editions and such for classical music. Would you accept
a pop track title that appended "(1st, 2nd, and 4th verses; 4-bar major-
fifth bridge)"? Let's not go there. Let's keep the discussion within
the classical scope.
Post by Brian Schweitzer
This information is not "clutter", nor is it "classical
freak only".
Both those points are quite arguable, sir. ;)
Post by Brian Schweitzer
If so, then classical is
the only place in the database we're making this exception for
versioning information.
As far as I'm aware, there is no provision for recording composition-
specific variation for non-classical music. The "(version)" indication
is a crude label attached when there are different instances of *the
same work* performed by the *same artist*. The remix situation is
another weird ball of wax that doesn't currently work well.
Post by Brian Schweitzer
I think you've overlooked the fact, however, that wiki's rather
horrible description aside (yes, I've read it before), it's not simply
"he added clarinets". There were some rather significant changes made
within the work, enough so that, as you point out, NMA found it
worthwhile to print the entire score twice, once the version with and
once the version without clarinets.
This is a classic example of where I would use a COMMENT tag (or maybe
VERSION). Not in my TITLE or PART, thanks.
--
-:-:- David K. Gasaway
-:-:- Email: ***@gasaway.org
-:-:- Web: dave.gasaway.org
-:-:- MusicBrainz: dkg
Brian Schweitzer
2008-02-05 14:23:12 UTC
Permalink
Post by David K. Gasaway
Post by Brian Schweitzer
But in any case, cooperaa asked me on IRC today why I used K6 / K3 /
K2 / KV instead of just K(most recent) or KV / K2 / K3 / K6.
If I just used K(most recent), then 1) a person
with a CD that uses original KV numbers (or K2 or K3 numbers) and not
K6 numbers would not get something that made sense to them. (You put
in a CD for "Sonata for Violin & Piano No. 29 in A major" and get K.
295d, not KV 305, etc.).
The assumption being that "K. 295d/305" will make any more sense to the
user than "K. 295d" when they expected "KV 305". I suspect not, unless
they user already understands the ins and outs of Koechel, in which
case there's no problem with the latter. :)
Post by Brian Schweitzer
2) the same for searching - if you search
for "KV 305" (now that I inserted spaces around the /'s to work around
a lucene issue), you'll get all 295d/305's.
This is where a masterlist comes in handy. While I see the value in
what you propose, I don't think it behooves the MB database to be in
the business of recording the failings (form lack of information, of
course) of Koechel and his successors.
David, while believe me, I do understand your point, I think each
point makes sense - but only for half the users, half the time. In
#1, you're assuming that a user always will be looking for KV 305.
Unfortunately, if I can use my own collection as a reference, only
half the time will people be looking for that KV 305. The other half
of the time, they'll be rather surprised to not see K. 295d, as that's
the K. work ID number on their CD.

In a perfect world, no CD would have anything other than a K6 number
on it. But the labels are, as we know, anything but perfect. A
master list is useful in having the numbers cross referenced, sure,
but given that labels seem to choose to use KV, K2, K3, K3a, and K6
numbers at random, using only one catalog listing would invariably be
less potentially confusing for 50% of the users, but a great deal more
confusing for the other 50%.

Brian
* Who begs that noone realize that Chopin has 4 different overlapping
catalogs all in use which are utterly unrelated to each other... *
David K. Gasaway
2008-02-05 23:15:15 UTC
Permalink
Post by Brian Schweitzer
David, while believe me, I do understand your point, I think each
point makes sense - but only for half the users, half the time. In
#1, you're assuming that a user always will be looking for KV 305.
That was your assumption, not mine.
Post by Brian Schweitzer
using only one catalog listing would invariably be
less potentially confusing for 50% of the users, but a great deal more
confusing for the other 50%.
Sigh. I don't even know how to respond to a unsupported blanket statement such as that.
--
-:-:- David K. Gasaway
-:-:- Email: ***@gasaway.org
-:-:- Web : dave.gasaway.org
Brian Schweitzer
2008-02-05 14:42:10 UTC
Permalink
Post by David K. Gasaway
Post by Brian Schweitzer
However, your claim that only a release containing both would have
reason to include the information is puzzling. Were this not
classical, would we say the same? If I have a CD which has track 12
listed as "Foo (radio edit)", is it legitimate to leave out the (radio
edit) as there's not a non-radio edit version on the same release?
It's really hard to compare the "(version)" compromise adopted for pop
music to score editions and such for classical music. Would you accept
a pop track title that appended "(1st, 2nd, and 4th verses; 4-bar major-
fifth bridge)"? Let's not go there. Let's keep the discussion within
the classical scope.
But see, there's the problem. You want to have CSG, you want the
details... yet you don't want to actually identify the work.
"Symphony No. 40" identifies *2* different works. I don't think
there's need to get all technical here; rather, I'd quote the program
notes of the Seattle Orchestra (
http://www.osscs.org/notes/mozart_k550.html ):

"Mozart's penultimate symphony is so well known, even by people who
rarely attend orchestral concerts, that little introduction is
necessary, except to answer the question, "With or without
clarinets?""

(10 second Google search to find that, by the way)
Post by David K. Gasaway
Post by Brian Schweitzer
This information is not "clutter", nor is it "classical
freak only".
Both those points are quite arguable, sir. ;)
Perhaps arguable, but I prefer, if we're going to go to all the
trouble to identify the work, that we actually identify the work. :)
Post by David K. Gasaway
Post by Brian Schweitzer
If so, then classical is
the only place in the database we're making this exception for
versioning information.
As far as I'm aware, there is no provision for recording composition-
specific variation for non-classical music. The "(version)" indication
is a crude label attached when there are different instances of *the
same work* performed by the *same artist*. The remix situation is
another weird ball of wax that doesn't currently work well.
I think the dictionary says it all here:
"Version": A particular form or variation of an earlier or original type

Isn't this the bare essence of what we're discussing? My point being,
whether we're discussing which version of a Sheryl Crow song or a
Mozart composition, the line's between classical and non-classical are
nowhere near as cut and dried as you make them seem. Adding a
qualification that correct work identification require it to also be
the same composer, I think, is disingenuous. In pop, a version by a
different artist is the same beast, under the new description of
"cover". In any case, in classical, we've already long accepted that
the composer is the artist - and thus, irregardless of the performer,
the "artist" does indeed remain the same.
Post by David K. Gasaway
Post by Brian Schweitzer
I think you've overlooked the fact, however, that wiki's rather
horrible description aside (yes, I've read it before), it's not simply
"he added clarinets". There were some rather significant changes made
within the work, enough so that, as you point out, NMA found it
worthwhile to print the entire score twice, once the version with and
once the version without clarinets.
This is a classic example of where I would use a COMMENT tag (or maybe
VERSION). Not in my TITLE or PART, thanks.
But you forget, it's not your title. It's not my title. It's a work
identification in the database we all share. But that's the point -
work identification, not work sort-of identification.

Brian
David K. Gasaway
2008-02-05 23:43:19 UTC
Permalink
Post by Brian Schweitzer
But see, there's the problem. You want to have CSG, you want the
details... yet you don't want to actually identify the work.
Yes, I want to identify the work. When did I ever say that I did not? (Hint: I didn't.) This is about what goes into the track title. Not everything needs to go in the track title.
Post by Brian Schweitzer
"Version": A particular form or variation of an earlier or original type
Isn't this the bare essence of what we're discussing? My point being,
whether we're discussing which version of a Sheryl Crow song or a
Mozart composition, the line's between classical and non-classical are
nowhere near as cut and dried as you make them seem.
You brought up the specious comparison to pop music. Never did I argue the line was cut and dried. Rather, I tried to point out that _you_ were applying a double standard to pop and classical by assuming that "with clarinets" is the same as "radio edit".
Post by Brian Schweitzer
Adding a
qualification that correct work identification require it to also be
the same composer, I think, is disingenuous.
I didn't make that argument. I said that an identification such as "(radio edit)" is only used when a composition (not necessarily by the performing artist) is performed more than once by the same artist. If one artist makes only an album edit and another makes only a radio edit, neither "(radio edit)" nor "(album edit)" would be used. Neither would the track title identify the changes (e.g., instrumentation) made to the work since original composition. There simply _is no comparison_ between "with clarinets" and the "(radio edit)" annotation used for pop tracks. Can we drop this now?
Post by Brian Schweitzer
But you forget, it's not your title. It's not my title. It's a work
identification in the database we all share. But that's the point -
work identification, not work sort-of identification.
Track title != absolute work identification. It's obvious by now that you'd like it to be, but I can't agree.
--
-:-:- David K. Gasaway
-:-:- Email: ***@gasaway.org
-:-:- Web : dave.gasaway.org
Aaron Cooper
2008-02-06 01:53:43 UTC
Permalink
Post by David K. Gasaway
Post by Brian Schweitzer
But you forget, it's not your title. It's not my title. It's a work
identification in the database we all share. But that's the point -
work identification, not work sort-of identification.
Track title != absolute work identification. It's obvious by now
that you'd like it to be, but I can't agree.
Until we have separate fields for classical releases, the "title"
field is all we have to store all the information associated with a
track.

-Aaron
David K. Gasaway
2008-02-06 02:27:52 UTC
Permalink
Until we have separate fields for classical releases, the "title" field
is all we have to store all the information associated with a track.
We have a release annotation, do we not?
--
-:-:- David K. Gasaway
-:-:- Email: ***@gasaway.org
-:-:- Web : dave.gasaway.org
Aaron Cooper
2008-02-06 03:04:08 UTC
Permalink
Post by David K. Gasaway
Post by Aaron Cooper
Until we have separate fields for classical releases, the "title"
field is all we have to store all the information associated with a
track.
We have a release annotation, do we not?
<with severe sarcasm>
Alright then, here's our solution: Only put tempo in the track title
field and then all other pieces of information that identify the track
can go in the release annotation where they will be completely
inaccessible to taggers and the webservice!
</>

Until the database schema supports splitting the classical titles up
into chunks (like work_type, catalog_number, common_name, etc.) so
every person can build their own "title", we're going to be putting up
with people whining "this is too much info" or "this isn't enough
info". Until we can satisfy all appetites by allowing custom built
titles we will never be able to find a suitable compromise.

If I were a benevolent dictator, I would say "Deal with it!" until we
have the capability to cut out the pieces of information some people
find "extraneous".

-Aaron (cooperaa)
David K. Gasaway
2008-02-06 03:45:32 UTC
Permalink
Until we can satisfy all appetites by allowing custom built titles we
will never be able to find a suitable compromise.
A compromise is possible if the involved parties are open to compromise (myself not excepted ;). But seriously, I thought the release annotation was a reasonable compromise.
If I were a benevolent dictator, I would say "Deal with it!" until we
have the capability to cut out the pieces of information some people
find "extraneous".
Again, it's not about me(*). I'd like to see a solution that doesn't inundate ordinary users and editors with esoteric trivia interesting only to a small minority - now or ever, as I'd guess that any solution to build your own title is going to have a hard time dealing with title parts like "with clarinets".

(*) In fact, I use MusicBrainz tags only indirectly (though directly enough that anything I don't like makes extra work for me).
--
-:-:- David K. Gasaway
-:-:- Email: ***@gasaway.org
-:-:- Web : dave.gasaway.org
Brian Schweitzer
2008-02-06 00:23:58 UTC
Permalink
Post by David K. Gasaway
Post by Brian Schweitzer
David, while believe me, I do understand your point, I think each
point makes sense - but only for half the users, half the time. In
#1, you're assuming that a user always will be looking for KV 305.
That was your assumption, not mine.
If I can bring back in the part that got snipped before:
------------------------------------
Post by David K. Gasaway
Post by Brian Schweitzer
If I just used K(most recent), then 1) a person
with a CD that uses original KV numbers (or K2 or K3 numbers) and not
K6 numbers would not get something that made sense to them. (You put
in a CD for "Sonata for Violin & Piano No. 29 in A major" and get K.
295d, not KV 305, etc.).
The assumption being that "K. 295d/305" will make any more sense to the
user than "K. 295d" when they expected "KV 305". I suspect not, unless
they user already understands the ins and outs of Koechel, in which
case there's no problem with the latter. :)
------------------------------------

How is this not an assumption that the user expects KV 305?
Post by David K. Gasaway
Post by Brian Schweitzer
using only one catalog listing would invariably be
less potentially confusing for 50% of the users, but a great deal more
confusing for the other 50%.
Sigh. I don't even know how to respond to a unsupported blanket
statement such as that.

How is it confusing? If a large number of CDs are using K1, and
another large number K6, plus those that perhaps list both, or use K2
or K3 numbers, if we only use any one of the numbers, where they are
not the same number, then no matter which one you pick to use,
everyone else will be confused. Just plain logic here.

Put another way, if half the people call it soda, and half the people
call it pop, and you don't tell them that pop and soda are the same
thing, if you only list "pop" in the menu, the soda people will be
confused. Much easier to list soda-pop in the menu. :)

Brian
David K. Gasaway
2008-02-06 02:45:26 UTC
Permalink
Post by Brian Schweitzer
How is this not an assumption that the user expects KV 305?
Post by David K. Gasaway
Post by Brian Schweitzer
If I just used K(most recent), then 1) a person
with a CD that uses original KV numbers (or K2 or K3 numbers) and not
K6 numbers would not get something that made sense to them. (You put
in a CD for "Sonata for Violin & Piano No. 29 in A major" and get K.
295d, not KV 305, etc.).
using only one catalog listing would invariably be
less potentially confusing for 50% of the users, but a great deal more
confusing for the other 50%.
Sigh. I don't even know how to respond to a unsupported blanket
statement such as that.
How is it confusing?
I didn't say I was confused. Care to support your assertion that users find "K. 295d/305" more confusing than "K. 295d" without resorting to weird comparisons to carbonated beverages? If you can't, then I suggest you avoid phrases such as "would invariably be less potentially confusing". We can argue these points all day, but please don't misrepresent your opinions or guesses as fact.
--
-:-:- David K. Gasaway
-:-:- Email: ***@gasaway.org
-:-:- Web : dave.gasaway.org
Brian Schweitzer
2008-02-06 00:29:11 UTC
Permalink
Post by David K. Gasaway
Post by Brian Schweitzer
But you forget, it's not your title. It's not my title. It's a work
identification in the database we all share. But that's the point -
work identification, not work sort-of identification.
Track title != absolute work identification. It's obvious by now that you'd like it to be, but I can't agree.
If (part of) the point of CSG is to identify the work better than the
liners do, and the point of the track title is, using CSG, to identify
the work, I fail to see how actually identifying the work is a bad
thing. This is like saying "Symphony in G major: I. " would be enough
for a track title - after all, he can't have written that many G major
symphonies with a first movement, right? :)

Brian
David K. Gasaway
2008-02-06 02:48:20 UTC
Permalink
Post by Brian Schweitzer
If (part of) the point of CSG is to identify the work better than the
liners do, and the point of the track title is, using CSG, to identify
the work, I fail to see how actually identifying the work is a bad
thing.
WTH? I already stated, without any qualifications, in the post you just replied to, that I _do want to identify the work_.
--
-:-:- David K. Gasaway
-:-:- Email: ***@gasaway.org
-:-:- Web : dave.gasaway.org
Jim DeLaHunt
2008-02-06 06:52:03 UTC
Permalink
Brian:

Aren't you exaggerating a bit? This isn't about perfectly identifying works
versus "identifying the work is a bad thing". It's about differing degrees
of precision in identification.
Post by Brian Schweitzer
Post by David K. Gasaway
....But that's the point -
work identification, not work sort-of identification.
Track title != absolute work identification. It's obvious by now that
you'd
like it to be, but I can't agree.
If (part of) the point of CSG is to identify the work better than the
liners do, and the point of the track title is, using CSG, to identify the
work, I fail to see how actually identifying the work is a bad thing.
This is like saying "Symphony in G major: I. " would be enough for a track
title - after all, he can't have written that many G major symphonies with
a first movement, right? :)
There are more precise degrees of work identification than what you are
proposing to record. What about identifying which edition of the printed
score the performers use? They have different typos, of course. What about
identifying which cuts the conductor took? Not all performance off the same
score play the same notes. One can go very far along this
precision-of-identification continuum.

The decisions we are talking about are trade-offs within a fuzzy continuum.
It's perfectly consistent to say that "Symphony in G major: I. " is not
precise enough for a track title, but "with clarinets" is more precise than
needed. I might agree or disagree with such a statement, but I do think
it's a valid one.


-----
-- http://jdlh.com/ Jim DeLaHunt , Vancouver, Canada ?
http://wiki.musicbrainz.org/JimDeLaHunt
--
View this message in context: http://www.nabble.com/-Clean-up-CSG--Capitalization-%28and-placement%29-of-types-and-tempos-tp15285696s2885p15300878.html
Sent from the Musicbrainz - Style mailing list archive at Nabble.com.
Brian Schweitzer
2008-02-06 03:38:33 UTC
Permalink
Post by David K. Gasaway
Post by Brian Schweitzer
If (part of) the point of CSG is to identify the work better than the
liners do, and the point of the track title is, using CSG, to identify
the work, I fail to see how actually identifying the work is a bad
thing.
WTH? I already stated, without any qualifications, in the post you
just replied to, that I _do want to identify the work_.

Actually, what you said was:

"I said that an identification such as "(radio edit)" is only used
when a composition (not necessarily by the performing artist) is
performed more than once by the same artist. If one artist makes only
an album edit and another makes only a radio edit, neither "(radio
edit)" nor "(album edit)" would be used. Neither would the track title
identify the changes (e.g., instrumentation) made to the work since
original composition. There simply _is no comparison_ between "with
clarinets" and the "(radio edit)" annotation used for pop tracks."

and in that same email, you also said:

"Track title != absolute work identification. It's obvious by now
that you'd like it to be, but I can't agree."

I think there is a direct comparison, that's the issue.

(radio edit) or (album edit) is only one of the hundreds of such
version identifications we already have in the database. We also have
such as (no vox), (with vox), (instrumental version), (vocal mix),
(drum mix), (no drums), etc. These all directly address the
instrumentation within the work.

You've suggested the annotation is where such identification belongs.
I just can't agree. For the two symphonies which WA Mozart wrote,
then later rewrote, identification such as Symphony No. 40 for
Orchestra in G minor , K. 550 "Great" or Symphony No. 35 for Orchestra
in D major, K. 385 "Haffner" just doesn't cut it. Nor would
identifying the Symphony for Orchestra in D major, K. *320 "Serenade"
as
Serenade No. 9 for Orchestra in D major, K. 320 "Posthorn" movements
1, 5, and 7, or the Sinfonia Concertante for Orchestra in G major, K.
*320 as Serenade No. 9 for Orchestra in D major, K. 320 "Posthorn"
movements 3 and 4 be a sufficient work identification.

I don't want to get bogged down specifically on Mozart here, but I
think this principal is universal within and without the classical
sphere. Would that the Kochel catalog, or the BWV catalog, or any
other catalog were a straight and perfect list, 1 to n, of every
single work, revision of a work, etc. But the real fact is, the
creators of those lists are just as human as we, and thus a work is at
the same time K. 295d and KV 305, or Op. 2 and BI 22 and C 225 and KK
6-15 at the same time, or that a single catalog entry might perhaps
refer to 2, 5 or 10 versions of the same work.

I just saw cooperaa's email. His point is essentially the same one
I'm making. Brenden's comments are in the same vein as well.
It is quite easy to go from "too much information" to "just as much
information as I want". It is quite difficult, however, to go from
"too little information" to the same.

If you want all Kochels, or you want
Chopin-Fontana/Brown/Chomi?ski-Turlo/Kobyla?ska, or you want your
Beethoven without having to deal with AnH and Hess and WoO and Opuses,
or you want your dashes correct, or, as the Seattle Orchestra put it,
"With or without
clarinets?", for the moment, our only good way to present the
information is the track title. If you're a data licensee, say the
BBC, it's quite one thing to take all that data, assuming it's well
structured, and do whatever you want with it.

You might perhaps not care whatsoever about language-correct
capitalization, dashes, commas, or quotation marks. You might want
your tracks all in Russian CSG. You might want only the Kochel number
that happens to be listed on your particular released version of a CD.
You might want it all. Right now, we just don't yet have the work
structure to support quite that degree of customization. So our
options are to either a) put it in with the rest of the kitchen sink,
or b) tuck it into an annotation where it becomes immediately far less
useful. Can I use the annotation in a tagger to add back in the data
you've decided isn't needed? Can the BBC or last.fm? Annotations are
quite useful - but not as just a good way to keep out what David
personally feels he doesn't need in a work identification.

Just as a final note, if I may quote Zaslaw in The Compleat Mozart, on
K. 385, which is the other symphony in the same situation as K. 550.

"Finally, on February 15... Mozart then proceeded to rework the score
of K. 385 sent from Salzberg by putting aside the March, deleting the
repeats in the first movement, and adding pairs of flutes and
clarinets in the first and last movements, primarily to reinforce the
tuttis and requiring no further changes in the already existing
orchestration of those movements..."

Sounds like perhaps a bit more was involved than, "hey, I got it,
let's have the clarinets double the oboes!", doesn't it?

Brian
David K. Gasaway
2008-02-06 04:03:29 UTC
Permalink
Post by David K. Gasaway
Yes, I want to identify the work. When did I ever say that I did
not? (Hint: I didn't.) This is about what goes into the track title.
Not everything needs to go in the track title.
(radio edit) or (album edit) is only one of the hundreds of such
version identifications we already have in the database. We also
have such as (no vox), (with vox), (instrumental version), (vocal
mix), (drum mix), (no drums), etc. These all directly address the
instrumentation within the work.
Ugh. What I said was that they do not indicate changes to the instrumentation in relattion to the original composition. Or a variation on the instrumentation provided by the originial composer (which is exactly the case we are considering). None of the examples you cited express those relationships.
Post by David K. Gasaway
You've suggested the annotation is where such identification belongs.
I suggested it is an alternative.
Post by David K. Gasaway
For the two symphonies which WA Mozart wrote,
then later rewrote, identification such as Symphony No. 40 for
Orchestra in G minor , K. 550 "Great" or Symphony No. 35 for
Orchestra in D major, K. 385 "Haffner" just doesn't cut it.
I'm not asking that you abandon full identification of the work. I'm questioning whether the track title is the best place to do it.
Post by David K. Gasaway
I just saw cooperaa's email. His point is essentially the same one
I'm making. Brenden's comments are in the same vein as well. It is
quite easy to go from "too much information" to "just as much
information as I want".
There again, you make an unsupported statement that is "quite easy" but I don't see that it is so.
Post by David K. Gasaway
"Finally, on February 15... Mozart then proceeded to rework the score
of K. 385 sent from Salzberg by putting aside the March, deleting
the repeats in the first movement, and adding pairs of flutes and
clarinets in the first and last movements, primarily to reinforce the
tuttis and requiring no further changes in the already existing
orchestration of those movements..."
Sounds like perhaps a bit more was involved than, "hey, I got it,
let's have the clarinets double the oboes!", doesn't it?
Who's arguing that with you? Not me. But I assume you'll want the complete text quoted above in the track title. ;)
--
-:-:- David K. Gasaway
-:-:- Email: ***@gasaway.org
-:-:- Web : dave.gasaway.org
Brian Schweitzer
2008-02-06 06:01:48 UTC
Permalink
Post by Jim DeLaHunt
Brian, I think I'm with David on this one. It's better to standardize on one
set of catalog numbers.
Post by David K. Gasaway
Post by Brian Schweitzer
But in any case, cooperaa asked me on IRC today why I used K6 / K3 /
K2 / KV instead of just K(most recent) or KV / K2 / K3 / K6. ...[snip]...
The assumption being that "K. 295d/305" will make any more sense to the
user than "K. 295d" when they expected "KV 305". I suspect not, unless
they user already understands the ins and outs of Koechel, in which case
there's no problem with the latter. :)
... While I see the value in what you propose, I don't think it behooves
the MB database to be in the business of recording the failings (form lack
of information, of course) of Koechel and his successors.
All these different catalogues are just different attempts to put
identifiers on musical works, right? By including multiple catalogue
numbers, you are include multiple identifiers *and* how those identifiers
correlate. By having the CSG require multiple catalogue numbers we ask every
editor to be able to cite the same set of identifiers, and ? this is what
makes it really hard ? in the same combination.
Suppose someone enters "K. 295d/305" for one release title, and I enter "K.
295d/306" in another release title? How will a third contributor know which
is correct enouch to copy? How would a fourth editor know how to correct
the mistake? If our standard is "KV 305", the picture is simpler.
Just to jump in a sec - there's perhaps a misperception here in what I
suggested. I suggested people always be copying from the standardized
lists, not from track titles. Thus, theoretically at least, a
misentered "K. 295d/306" ought to not occur - actually, not just with
respect to the catalog numbers, but with respect to all the data, that
(and correct caps/etc) is just the point behind the standardized lists
and having people copy directly from them... to avoid just such
incorrect data entry.
Post by Jim DeLaHunt
In this impoverished era when we don't have a MusicalWork entity in our
structure, I think it's sufficient to have the CSG require enough
information to identify the work reliably. At some later point, when we
have a MusicalWork entity, we use this information to attach track and
release entries to the MusicalWork. Then the experts can attach the
multiple catalogue identifiers to that entity once, correctly, for everyone.
Actually, that might be an argument for a CSG that asks for a specific
preferred catalog number (e.g. "KV 305"), but accepts other catalogs instead
(e.g. "K. 295d"). The other catalog number, if accurate, is still a clear
identifier. A bot has a chance of fixing it later.
Post by David K. Gasaway
Post by Brian Schweitzer
3) If you want all the catalog
numbers in your track, they are there, and you get them.
And if I don't want all the catalog numbers in my track (which I don't),
then I'm left to find my own solution. I could try to write a try regular
expression to remove the excess. It might not always work. It's not
going to ship with Picard. I guess either way, one of us loses, so which
is better overall solution? IMO, it's better to standardize on one set of
catalog numbers.
This is another good point. Either choice will satisfy some users and
dissatisfy others. Having one catalog number at least identifies the work,
and its simpler to explain to contributors, and there's not the chance of
inconsistent combinations of identifiers. Those are positives. The loss of
the ability to search by other index numbers is a negative, but a smaller
one on the whole.
Well, I still think having them and being able to chop out what you
don't want is better than not having them, and wanting them. However,
I will say I quite understand the point. I would, however, suggest
that if we limit to only a single catalog number, it not be selected
by requring only a single catalog, but rather, the best catalog number
for the job. In most cases, this would be K6, not KV - and thus,
again, given this proposition, there would not be the "KV 306" someone
might expect, as they would all be the much more up to date "K. 295d".
(I quite hope, though, that we have the "real" generic works concept
implemented as a database structure, though, before Der neue K?chel
comes out! (likely in 2010)). The problem is, while, say, a K. 581
may be quite enough to be that clear identifier", the work represented
by a KV Anh 97 would not be the same as K. 370b.

In this hypothetical "only one catalog number" proposal, I would also
strongly encourage that there also be wiggle room, such that the rare
funky case (see K. 370b and K. 371) may still be identified correctly.
For the Mozarts, the Bachs, and whoever else also uses the unspecific
deest and not the "growing actual catalog" WoO, I would also hope that
"K. deest NMA II/7/1 No. 7" wouldn't be seen as two different catalog
numbers (as dropping the K. deest, the BWV deest, etc then begs the
"why is there no Kochel/BWV number here?" question).

In essence, this all boils down to, if there's a consensus that we'd
prefer to have just one catalog number, unless two or more editors
have truely done the homework to debate it, please lets still allow
for the cases where a single catalog number just isn't enough.
(Kochel 6, remember, was last updated in 1964 - there's been quite
some large amount of scholarship since then, including the entire
volume of work by the NMA, which began work in 196*5*.)

Brian

PS: Side note on the above 370b-type confusable ones... We just ran
into one like this on one of the Mozart forums I frequent. Now that
we figured out what it is, as an example of how exotic it really can
get once you get way into the obscure works, bonus points to anyone
here who can ID the work on this CD - hint, as comprehensive as the
Mozart listing is, this one ain't on there yet... and it's not in the
NMA either http://www.jpc.de/jpcng/SESSIONID/0140e6d0266a374fd99911e9a41da690/classic/detail/-/hnum/9456336?rk=classic&rsk=novelties
David K. Gasaway
2008-02-06 11:07:31 UTC
Permalink
Post by Brian Schweitzer
I suggested people always be copying from the standardized
lists, not from track titles. Thus, theoretically at least, a
misentered "K. 295d/306" ought to not occur - actually, not just with
respect to the catalog numbers, but with respect to all the data, that
(and correct caps/etc) is just the point behind the standardized lists
and having people copy directly from them... to avoid just such
incorrect data entry.
Not to belabor the point, but it's phrases like "always be copying",
"theoretically at least", "ought to not occur" that make me
uncomfortable with this masterlist idea. Sorry, not the master lists
themselves - they're awesome! But I'm wary of expecting them to solve
our style problems.
Post by Brian Schweitzer
I would, however, suggest
that if we limit to only a single catalog number, it not be selected
by requring only a single catalog, but rather, the best catalog number
for the job.
Absolutely.
Post by Brian Schweitzer
In this hypothetical "only one catalog number" proposal, I would also
strongly encourage that there also be wiggle room, such that the rare
funky case (see K. 370b and K. 371) may still be identified correctly.
Absolutely.

Now, I've noted that you carefully avoided assenting to this proprosal.
Are you still firmly opposed to a single catalog identifier?
--
-:-:- David K. Gasaway
-:-:- Email: ***@gasaway.org
-:-:- Web: dave.gasaway.org
-:-:- MusicBrainz: dkg
Jim DeLaHunt
2008-02-06 07:16:19 UTC
Permalink
Brian, I think I'm with David on this one. It's better to standardize on one
set of catalog numbers.
Post by David K. Gasaway
Post by Brian Schweitzer
But in any case, cooperaa asked me on IRC today why I used K6 / K3 /
K2 / KV instead of just K(most recent) or KV / K2 / K3 / K6. ...[snip]...
The assumption being that "K. 295d/305" will make any more sense to the
user than "K. 295d" when they expected "KV 305". I suspect not, unless
they user already understands the ins and outs of Koechel, in which case
there's no problem with the latter. :)
... While I see the value in what you propose, I don't think it behooves
the MB database to be in the business of recording the failings (form lack
of information, of course) of Koechel and his successors.
All these different catalogues are just different attempts to put
identifiers on musical works, right? By including multiple catalogue
numbers, you are include multiple identifiers *and* how those identifiers
correlate. By having the CSG require multiple catalogue numbers we ask every
editor to be able to cite the same set of identifiers, and ? this is what
makes it really hard ? in the same combination.

Suppose someone enters "K. 295d/305" for one release title, and I enter "K.
295d/306" in another release title? How will a third contributor know which
is correct enouch to copy? How would a fourth editor know how to correct
the mistake? If our standard is "KV 305", the picture is simpler.

In this impoverished era when we don't have a MusicalWork entity in our
structure, I think it's sufficient to have the CSG require enough
information to identify the work reliably. At some later point, when we
have a MusicalWork entity, we use this information to attach track and
release entries to the MusicalWork. Then the experts can attach the
multiple catalogue identifiers to that entity once, correctly, for everyone.

Actually, that might be an argument for a CSG that asks for a specific
preferred catalog number (e.g. "KV 305"), but accepts other catalogs instead
(e.g. "K. 295d"). The other catalog number, if accurate, is still a clear
identifier. A bot has a chance of fixing it later.
Post by David K. Gasaway
Post by Brian Schweitzer
3) If you want all the catalog
numbers in your track, they are there, and you get them.
And if I don't want all the catalog numbers in my track (which I don't),
then I'm left to find my own solution. I could try to write a try regular
expression to remove the excess. It might not always work. It's not
going to ship with Picard. I guess either way, one of us loses, so which
is better overall solution? IMO, it's better to standardize on one set of
catalog numbers.
This is another good point. Either choice will satisfy some users and
dissatisfy others. Having one catalog number at least identifies the work,
and its simpler to explain to contributors, and there's not the chance of
inconsistent combinations of identifiers. Those are positives. The loss of
the ability to search by other index numbers is a negative, but a smaller
one on the whole.


-----
-- http://jdlh.com/ Jim DeLaHunt , Vancouver, Canada ?
http://wiki.musicbrainz.org/JimDeLaHunt
--
View this message in context: http://www.nabble.com/-Clean-up-CSG--Capitalization-%28and-placement%29-of-types-and-tempos-tp15083823s2885p15300067.html
Sent from the Musicbrainz - Style mailing list archive at Nabble.com.
Brian Schweitzer
2008-02-06 09:04:07 UTC
Permalink
Post by Jim DeLaHunt
Aren't you exaggerating a bit? This isn't about perfectly identifying works
versus "identifying the work is a bad thing". It's about differing degrees
of precision in identification.
Post by Brian Schweitzer
Post by David K. Gasaway
....But that's the point -
work identification, not work sort-of identification.
Track title != absolute work identification. It's obvious by now that
you'd
like it to be, but I can't agree.
If (part of) the point of CSG is to identify the work better than the
liners do, and the point of the track title is, using CSG, to identify the
work, I fail to see how actually identifying the work is a bad thing.
This is like saying "Symphony in G major: I. " would be enough for a track
title - after all, he can't have written that many G major symphonies with
a first movement, right? :)
There are more precise degrees of work identification than what you are
proposing to record. What about identifying which edition of the printed
score the performers use? They have different typos, of course. What about
identifying which cuts the conductor took? Not all performance off the same
score play the same notes. One can go very far along this
precision-of-identification continuum.
The decisions we are talking about are trade-offs within a fuzzy continuum.
It's perfectly consistent to say that "Symphony in G major: I. " is not
precise enough for a track title, but "with clarinets" is more precise than
needed. I might agree or disagree with such a statement, but I do think
it's a valid one.
I'm sorry Jim, but no, I am not willing to budge on this. If we
decide to go with single catalog numbers, fine. But when it comes to
correctly identifying a work, versus vaguely identifying a work, I
give notice right now that any RFC including such a suggestion I will
veto.

We're not talking about "the version where the conductor cut measures
3 through 5". I agree that would be overkill, and annotation
material. We are talking about correct specific work identification -
and on that, no, I won't agree that "K. 550" is enough. It may
perhaps be enough for the general public, but to be quite honest,
"Mozart's Jupiter Symphony" seems to be quite enough for the general
public to continue buying the CDs. It should not, however, be enough
for us, if we really are a database/encyclopedia of music, and if we
really do have a point at all with CSG.

I just checked google. Searching for: Mozart "K. 550" "without
clarinets" and Mozart "K. 550" "with clarinets", I turned up several
dozen recordings for each where the with/without is clearly specified.
In every decent guide to music / guide to Mozart I check, the
distinct versions of K. 550 (as well as the other 2 I mentioned) each
receive an entry. For us, then, to decide that all the
ultra-technical details of a CSG listing are required - and yet
identifying which actual work is on the CD is too much... it's beyond
my comprehension.

If I may ask, then, following this concept of vague work
identification, where would you draw the line? For K. 158a, would we
not care to identify which of the three librettos is used? I had
someone email me because they noticed their liner's libretto didn't
match what was in the listing - we researched, now all three versions
are listed. For K. 382d, would we not care if it were the "Leck mir
den Arsch fein recht sch?n sauber" libretto or the "Nichts labt mich
mehr als Wein" libretto? Or K. 382c, the distinction between the
"Leck mich im Arsch" libretto and the "La?t froh uns sein" libretto?
(German speakers can stop laughing now, lol). How about K. 495, with
an Allegro maestoso and an Allegro moderato version? That's another
someone emailed me about, as their liner didn't match the list, and
their liner turned out to be correctly identifying that it was the
alternate version.

My point is, simplification is ok, if we have to. Being vague about
the work, just because we want to avoid longer text, that to me is
purposefully destructive of data - the chances we will be able to
later to add that data back in are rather slim, once the editor is
gone and we no longer can consult the liner.

Brian
David K. Gasaway
2008-02-06 11:52:55 UTC
Permalink
Post by Brian Schweitzer
We're not talking about "the version where the conductor cut measures
3 through 5". I agree that would be overkill, and annotation
material.
Can you give us a clear defintion of where you draw the line?
Apparently, it isn't where K6 ends. For me, the catalog number seems
to be a clear-cut, unambigious, easy-to-define place to draw the line.
Post by Brian Schweitzer
We are talking about correct specific work identification
For your definition of "correct", sir. I'm willing to accept that
other people may find that "K. 550 with clarinets" is incorrect (too
imprecise - thank you Jim).
Post by Brian Schweitzer
For us, then, to decide that all the
ultra-technical details of a CSG listing are required - and yet
identifying which actual work is on the CD is too much... it's beyond
my comprehension.
How many times must I say it? I want to identify work. That is not
the same as saying that I want it all in the track title.
Post by Brian Schweitzer
For K. 158a, would we
not care to identify which of the three librettos is used?
[...]
Your notes on these issues in the wiki page are fabulous. I thank you
for your dedicated research and documentation. If anyone has questions
in the future we have somewhere to refer them for answers. That's what
makes the lists invaluable. I just don't expect track titles to be
quite so .. definitive.
--
-:-:- David K. Gasaway
-:-:- Email: ***@gasaway.org
-:-:- Web: dave.gasaway.org
-:-:- MusicBrainz: dkg
Brian Schweitzer
2008-02-06 11:35:17 UTC
Permalink
Post by David K. Gasaway
Post by Brian Schweitzer
I suggested people always be copying from the standardized
lists, not from track titles. Thus, theoretically at least, a
misentered "K. 295d/306" ought to not occur - actually, not just with
respect to the catalog numbers, but with respect to all the data, that
(and correct caps/etc) is just the point behind the standardized lists
and having people copy directly from them... to avoid just such
incorrect data entry.
Not to belabor the point, but it's phrases like "always be copying",
"theoretically at least", "ought to not occur" that make me
uncomfortable with this masterlist idea. Sorry, not the master lists
themselves - they're awesome! But I'm wary of expecting them to solve
our style problems.
In of themselves, no, I'd agree, they won't. There's simply too many
classical composers for them to quickly solve all our style problems -
see the (slowly) growing list at
http://wiki.musicbrainz.org/CSGStandard . (By the way, it could use
help if anyone's interested!). However, I think that as the number of
people working on lists grows, even if it's just people adding a
release who copy over the CSG titles they just worked out, slowly we
move to a point where the times someone needs to create a new CSG
title, rather than just copy and paste, become the rare occasion, and
not the norm. I was more referencing the typo Jim indicated, where
305 became 306 - in a copy/paste situation, such a typo is either a)
impossible, or b) an indication someone's not copy/pasting.
Post by David K. Gasaway
Post by Brian Schweitzer
I would, however, suggest
that if we limit to only a single catalog number, it not be selected
by requring only a single catalog, but rather, the best catalog number
for the job.
Absolutely.
Post by Brian Schweitzer
In this hypothetical "only one catalog number" proposal, I would also
strongly encourage that there also be wiggle room, such that the rare
funky case (see K. 370b and K. 371) may still be identified correctly.
Absolutely.
Now, I've noted that you carefully avoided assenting to this proprosal.
Are you still firmly opposed to a single catalog identifier?
I'm not dead set against it. I can quite understand how some cat
numbers might be uncomfortably long - see the near-essay I ended up
writing into the wiki on K. 370b. :) However, unlike the "vague
identification" suggestion, if there's consensus that we just want to
use a single cat #, given the flexibility noted above, I can live with
it. I do admit, however, to being uncomfortable with it - if it's not
quite obvious already, I'd always rather go with the longer title
that's info inclusive, rather than leave out info just for the sake of
the shorter title. Unlike the other situation, however, so long as we
are choosing the right cat #(s) for the job, the identification
doesn't suffer here, only the completeness of the data.

Now, since it seems I'm more on the point of view here that quite
likely won't become the consensus, anyone feel like suggesting which
would be the best Chopin catalog to try to standardize to?

Brian
David K. Gasaway
2008-02-06 12:50:28 UTC
Permalink
Post by Brian Schweitzer
In of themselves, no, I'd agree, they won't. There's simply too many
classical composers for them to quickly solve all our style problems -
Even in the cases were there are highly refined mastlist lists
available, I don't trust that they will solve our style problems.
Sorry if I didn't make that clear.
Post by Brian Schweitzer
I'm not dead set against it.
[...]
I do admit, however, to being uncomfortable with it
Hmm. Well, not quite the response I was hoping for. If I haven't
thoroughly conviced you by now, I'm not entirely sold on it myself.
Come to think of it, I'd say I'm uncomfortable with, but not dead set
against multiple catalog numbers. :)
Post by Brian Schweitzer
Now, since it seems I'm more on the point of view here that quite
likely won't become the consensus, anyone feel like suggesting which
would be the best Chopin catalog to try to standardize to?
I'm not familiar enough with the Chopin catalogs to state an opinion at
this time. Have any good resources with which to educate myself?
--
-:-:- David K. Gasaway
-:-:- Email: ***@gasaway.org
-:-:- Web: dave.gasaway.org
-:-:- MusicBrainz: dkg
Leiv Hellebo
2008-02-07 03:31:14 UTC
Permalink
David K. Gasaway wrote:

Now, I've noted that you carefully avoided assenting to this proprosal.
Post by David K. Gasaway
Are you still firmly opposed to a single catalog identifier?
Post by Brian Schweitzer
I'm not dead set against it.
[...]
I do admit, however, to being uncomfortable with it
Hmm. Well, not quite the response I was hoping for. If I haven't
thoroughly conviced you by now, I'm not entirely sold on it myself.
Come to think of it, I'd say I'm uncomfortable with, but not dead set
against multiple catalog numbers. :)
Hi guys,

if you're talking about which catalogues should go into CSGS/Mozart
here, then I think I am all for entering all of them.

This way, it will be less of a problem to let submitters just enter in
the track title what they have in their liner notes. Those interested
may look it up on CSGS/M if they need it mapped to another edition.

And in the case of fixing up existing *bad-looking* Mozart releases we
could standardise on one. (If there are any left after the batch-edits... ;)

Would that be OK?



leivhe
leivhe
2008-02-08 03:49:35 UTC
Permalink
Post by Leiv Hellebo
*bad-looking* Mozart releases
Ouch. For "bad-looking", please read "structurally incomplete" or something
(and they might be newly submitted). I was not trying to be ironic, I was
trying to be constructive and friendly.

Sorry for the spam, list
--
View this message in context: http://www.nabble.com/-Clean-up-CSG--Capitalization-%28and-placement%29-of-types-and-tempos-tp15305814s2885p15344368.html
Sent from the Musicbrainz - Style mailing list archive at Nabble.com.
Brian Schweitzer
2008-02-06 13:48:49 UTC
Permalink
Post by Brian Schweitzer
We're not talking about "the version where the conductor cut measures
3 through 5". I agree that would be overkill, and annotation
material.
Can you give us a clear definition of where you draw the line?
Apparently, it isn't where K6 ends. For me, the catalog number seems
to be a clear-cut, unambiguous, easy-to-define place to draw the line.
I think, perhaps, you have a misunderstanding of what catalogue
numbers mean. That's fine - it's quite understandable. I know little
about the other catalogues, though I'm quite sure other have similar
issues.

A Kochel listing consists of several sections. The main section is
indeed a work identification.

The important part is, however, that Kochel made the decision that,
where other works were derivatives of that work, rather than list them
separately, he simply opted to note them in a "Notes" section under
that main listing.

A Kochel catalogue number != one specific work.

A Kochel catalogue number encompasses one or several works, all
derived from the same original composition.

This is entirely like the modern concept of a track which has various
edits and remixes. Were Kochel to catalogue a modern pop star, you'd
find the original song listed with a catalogue number, but all the
variations listed only in the notes.

Thus, K. 550 != 1 single unique work. K. 550 = multiple works.

Therefore:
K. 550 with clarinets == K. 550
K. 550 without clarinets == K. 550

Also, you must keep in mind, again as I mentioned elsewhere, that the
most recent Kochel was published in 1964. It is widely recognized to
be very dated, including multiple works not by Mozart, and not
including many other works which are by Mozart - and here we're not
talking variations, but completely independent compositions. The NMA
is a quite good reference, but even there, the project began in 1965,
only a year after K6. I am aware of at least one work, now considered
100% to be a composition of WA Mozart, which was found and confirmed
as WA Mozart *after* the relevant NMA volume was published.
(Consider, the last NMA volumes were only published in 2006 - an
effort of that magnitude took 41 years to complete).

Thus why I find it rather, well, dismissive and vague to simply wave
your arms and say "It's all K. 550" without realizing that yes, *they*
all are K. 550 - but K. 550 is not one thing, but rather a *group* of
thingS.

My point is that we quite ought to identify which of those things it
actually is, if we're really concerned about identifying the work, and
not just the general group the work belongs to. As for me, I'd be
quite interested to know which recordings are of the version without
clarinets vs the version with the clarinets.

Brian
David K. Gasaway
2008-02-11 13:20:05 UTC
Permalink
Post by Brian Schweitzer
I think, perhaps, you have a misunderstanding of what catalogue
numbers mean.
I think I understand well enough. For you, the term "work" has a very
precise meaning that isn't necessarily universally shared. Yes, there
may be different works (using your definition) cataloged under a single
Koechel number. I know and accept this. But I'm inclined to let
authoritative scholars define what is or what is not considered an
object fit to be labelled as a fully identifiable and distinct unit (by
catalog number). It's not the most precise system, but it draws a very
clear and unambiguous line what to include in a track title.
Otherwise, arguments such as this may continue ad infinitum.
Post by Brian Schweitzer
My point is that we quite ought to identify which of those things it
actually is, if we're really concerned about identifying the work, and
not just the general group the work belongs to. As for me, I'd be quite
interested to know which recordings are of the version without clarinets
vs the version with the clarinets.
Have I mentioned that _I want to identify the work_? Yes, I think I
have. You still seem to misunderstand what I mean by this. For
example, I am interested to know whether a specific performance of K.
550 includes clarinets. I would record this information in my
FLAC/Vorbis tags. I would like to record this information in
MusicBrainz. However, in my opinion, the track title is not the best
place to do it.

As eager as I am to end this marathon debate, please permit me to take
one final stab at it. If we choose to record details such as "with
clarinets" in the track title, these are the possible outcomes as I see
it (assuming that the user has an intent to use the masterlists):

1) The editor knows exactly which version of the work was used for the
performance, and selects the appropriate entry in the masterlist.
2) The editor does not know which version was used, abandons the
masterlist, and builds the track title by hand.
3) The editor does not know which version was used, and randomly
selects an entry in the masterlist.
4) The editor does not know which version was used, and abandons the
edit altogether.

The negative outcomes trouble me. In (2) through (4), we end up
animosity both towards MB and towards the masterlists, which were
conceived to ease frustration, not cause it. Outcome (2) results in
track titles that are not consistent with the masterlist. Outcome (3)
results in a factually incorrect entry and invalidates the whole point
of adding this extra information to the title. In fact, it is very
difficult to differentiate this negative outcome (3) from the positive
outcome (1). The downsides to (4) are obvious.

Imagine instead a system where this information is excluded from the
track title. The masterlist page would include documentation on the
different versions of the work and ask the user, if able, to identify
the precise work in the release annotation. It could even provide the
exact text to copy-paste into the release annotation. Here are the
possible outcomes from this proposal:

1) The editor knows exactly which version of the work was used for the
performance, selects the appropriate entry in the masterlist, and
follows the annotation guidance.
2) The editor does not know which version was used, selects the
appropriate entry in the masterlist, and ignores the annotation.

In this scenario, we can easily differentiate the negative outcomes;
they are those releases that do not have a release annotation. Still,
we at least have a well-formed and factual (if imprecise) track title.
If a knowledgable editor comes across this release in the future, then
he/she will add the release annotation. Everyone seems to be mostly
satisfied (meaning, some people will be less than entirely satisfied
;).

I know that the release annotation is a bit awkward for this purpose,
but it's about the only alternative I can offer at this time.
Thankfully, if a better solution arrives in the future, it will be easy
to search track titles for the catalog number and shuffle this
information around en masse.
--
-:-:- David K. Gasaway
-:-:- Email: ***@gasaway.org
-:-:- Web: dave.gasaway.org
-:-:- MusicBrainz: dkg
Loading...