Why do we add composer to a classical release with multiple titled works only if composer is different

I dont understand the reasoning behind this

e.g
From Style / Classical / Release Title - MusicBrainz

An example of an album with multiple works both composed by Johann Sebastian Bach

Concertos pour 1 & 2 violons / Concerto pour violon & hautbois

And then an example of an album with multiple works composed by different composers

Vaughan Williams: Fantasia on a Theme by Thomas Tallis / Britten: Variations on a Theme of Frank Bridge / Elgar: Introduction and Allegro

I dont understand why we only add the composers for the second album because it would be just as helpful to be able to see the composer on the first album.

e,g
Bach: Concertos pour 1 & 2 violons / Concerto pour violon & hautbois

And in both cases the composers are shown on the cover but seperately to the title. And in both cases if you actually go into the release and look at the tracks it is equally obvious who the composer is for any particular track.

So what is the reasoning behind the difference ?

Actually Im wondering if the examples in the guidelines have misunderstood the guidelines because for works by different composers it says

For every work, add the composer(s) as credited. Separate the composer(s) from the work with a colon.

the word credited refers to adding a composer

whereas for multiple works by composer it says

Use the works as credited and separate them with a slash.

so although it does’t mention composer it again says credited, so does the credited implicity mean add the composer as written on cover ?

The reasoning is that for a release like “Piano Concerto no. 1 / Symphony no. 2”, if the only composer is Beethoven, then there’s no question about what works it contains. But if the composers are Beethoven and Schumann, it’s useful to know at a glance if this is the Schumann concerto and the Beethoven symphony, or the other way around.

No. It just means “if the cover says just Symphony 1 don’t expand it to Symphony no. 1 in D major even if you’d like that more” and it has nothing to do with the composer, which should not be part of the title, as per “The Release title should not contain composers except for titled releases containing their names as above, or for multi-composer releases as outlined below.”

2 Likes

Thanks! Im not an expert in Classical but it seems if you’re only looking at the release title and not the tracklist, there’s no clear way to tell that Beethoven composed both works (or even any of them). It’s not reasonable to assume someone would just know that, any more than they’d know one piece was by Beethoven and another by Schumann.

Now, if you’re looking at the release with the tracklist, sure, you’d see the composers. But that applies to any release, not just this one.

To me, it doesn’t really make sense to have a field that sometimes includes the composer and sometimes doesn’t, for me these are are some issues with it

  • It’s not intuitive – The rule isn’t obvious, so anyone editing needs to read the Classical Style Guide (CSG) very carefully to get it right.
  • It makes using MusicBrainz harder – For example, on the Release Duplicates page, similar titles get flagged when adding a new release. If composer names aren’t in the title, we end up checking lots of releases that clearly aren’t duplicates—just because there’s no composer info to help filter them.
  • It leads to inconsistency in the database – Classical album covers usually show both the works and the composers. If we only include the composer from the cover sometimes, it feels a bit random.
  • It causes tagging inconsistencies – A lot of tagging apps rely on MusicBrainz. I think users would appreciate having composer names included (or excluded) consistently in titles—it would make tagging much smoother if the data followed a clear and consistent rule.

Paul

But the release title is supposed to work in tandem with the release artist, and the two combined have all the needed info :slight_smile:

So would it be correct to say the history of this is that the composer was originally never added to release title. But then at some point it was realized that combination of Release Title and Release Artist did not help identify the composer of work when there were multiple composers and works on the release. So then an amendment was made to add the composers to the title in the case of multiple works and multiple composers to try and address this?

Honestly, the guidelines have been like that for so many years that I cannot tell you if that was the order in which they were developed or not :slight_smile:

Fair enough.

But I think the implementation as it stands does cause the four issues I listed.

Now Release Duplicates page could be resolved by simply addig Release Artist column, but the other three all have the underlying issue is the unituitive and seemingly inconsistent approach to the title field. it would seem this would be resolved by either always adding composers to title or never adding them to title, this should apply to single work releases as well.

By seemingly inconsistent I mean there is a rule but it would not be easy to work it out by just looking at a list of classical releases, and to my mind that means it is probably a bad rule.

If I was designing from scratch it seems to be the the best approach would be to always add the composer because that would be most useful, and the composer seems like part of the title for Classical and remove the composer from album artist unless they were actually involved directly in this release, e.g. a contemporary composer making a recording of their music.

Now of course we are not going to do this but would be nice to get a discussion going to think of a way forward.

The main place for the composer should be always the artist field.
You shouldn’t regard the release artist a performing artist. Any role can be a release artist, if that’s the choice of the publisher of the release.

For instance field recorders of folklore & traditional music are in many case musicological audio documents, where the field recorder is considered the “author” of the release. They didn’t compose anything, they didn’t perform anything. They made recordings, made photographs, and/or wrote liner notes, but are in many cases the release artist.

Adding the composer to the title is just a method to clarify which work is by which release artist.
Other methods could be possible, but each method has it’s own advantages & disadvantages. Usually it’s no good to change the rules in the middle of the game, however.

The composer on MB is always before the ;
Adding the composer in the title to a single composer release would result in double information & longer titles (Where the full release title should always be regarded the Artist(s) + Title(s))
Once you get accustomed to the meaning of the ; joiner, things start to look way better.

Classical is a very hard genre to mould in a database format, compared to pop/rock/etc.
The focus is mostly equal on the composer & recording, and often there isn’t a “release title” but just one or more “work titles”. To make it even harder, the work titles are sometimes very generic in nature.
(Even worse is many players/music library interfaces not being able to handle composer tags in a proper way, or multiple artist releases in general.)

1 Like

Hi, thanks for responding, I dont think I did I said the release artist should be the recording artist, I said

unless they were actually involved directly in this release

and in your example the field recorder clearly were involved directly in this release.

Whereas the composers probably were not, compare it to Pop Music where often the artist did compose the music, and the composer/writer just has a composer/writer credit and is not the release artist. So I see little real difference here in release artist, just added complexity in the case of classical releases.

Not if composer was added to title but not album artist. I mean the release title is the release title, would you say the full release title of Pop/Rock is album artist and title ?

I dont really agree, I feel there is a tendancy to say Classical Music is so complicated and so different to other music that we need to introduce all these hacks. In reality the only difference is they are just a bit more complex in the number of people involved - releases tend to be collabaration between diffferent independent entities (e.g Conductor. Orchestra, Soloist). And in addition to the release - track relationship there is often the work - movement relationship, which is handled nicely by works, but that happens in non -classical as well just much more rarely.

To make it even harder, the work titles are sometimes very generic in nature.

Well yes, that is why adding the composer to the title would be of benefit.

Possibly not, but it would be worth properly considering if the current way is the right way or not rather than accept it as the right way. And if it is not the right way then some mitigation should be considered.

Just a couple more thoughts

Of course the opposite is true in the case of multi composer releases the CSG means we do have double information with composers being added to both title and release artist

If you asked a Classical Afficiando for their favourite album I reckon they would say something like

Vivaldi’s Four Seasons performed by Nigel Kennedy’s with the English Chamber Orchestra

not

Four Seasons by Antonio Vivaldi

or

Four Seasons by Vivaldi, Nigel Kennedy’s and the English Chamber Orchestra

i.e in the response I think they would make they use the composer name as part of the title not as the artist.

It took me a long time to write the following comments, but I wanted to be sure I wrote them as clearly as I could.

I don’t remember in what order the rules were set, but there were indeed hacks added when it was realized that the first simple rules didn’t work properly.

Three considerations:

  • Classical music lovers tend to think of composer before performer(s). This is also how classical albums are sorted in physical stores.
  • As you know, in MB, as a general rule, the release title should be exactly what is printed.
  • MB shows release lists with release title and release artist.

This works well when there is only one composer. But as you can see in your Williams / Britten example, it could be difficult to know which composer composed what just by looking at the release artist and the release title. So this hack was set. I agree it is a deviation from the “title as printed” rule. But classical music lovers, like me, prefer this deviation to having a list of release titles which could be tricky to understand. Actually, we are doing what we would have liked the music industry editors to do.

Also, it should be noted that adding the composer to all titles, as you suggested, would be a deviation in all classical titles from the general rule “titles as printed”. So which is better: having all classical titles inconsistent with all other titles, or having only classical titles with 2 or more composers inconsistent?

P.S. My comments don’t mean that I agree or disagree with how MB does things. I have always felt that release artist and release title weren’t really important. What really matters when I am browsing or exploring my library is the work (which implies the composer in classical music), the performer(s), and sometimes the recording (which implies the year and the original recording company). IMO shoehorning so many relevant data into two fields cannot be done elegantly. Even more so when adding rules like “title as printed”. ARs are a much better way to represent data.

P.P.S. The physical or cultural (or commercial) reality is not really structured. It simply is, many things happen without anybody thinking of how those things could fit it in a general structure. We are trying, because this is how we think, to put our representations into a structure. But there are cases where reality will not fit in a structure, so that ugly hacks to the structure’s rules will have to be set. I don’t know of cases like this with ARs, but there probably are some. The Release title + Release artist issue is a good example of it.

2 Likes

I looked at the spine of some of my classical releases and the overriding format is something like

ComposerName: Work(s) - Conductor, Orchestra

So my point is yes the composer is very important, there is no denying that. But to me it seems the composer is part of the title, if it was considered the artist then the spine would be


Work(s) - Composer, Conductor, Orchestra

And when you look at the front cover the composer comes before the obvious title, it is not listed with the performers

So that is my point really the composer as written on both front cover and spine seems to be part of the title it is certainly not part of the release artists, am I the only one seeing it this way ?

And look at this example, a current composer conducting their own music


He has his own name twice, once before the work and once after, and on the spine the first name is grouped with the work distinct from the second reference to name and orchestra name.

Firstly I think the composer is part of the title. But assuming it is not we have two set of style guidelines, the standard ones and the CSG, and all releases should use one or the other. So within those guidelines it would be preferable for everything using one guideline to work in a similar way therefore it would better for all classical titles to be inconsistent with all other titles rather than just some.

ijabz Jaikoz and SongKong developer
June 6 |

  • | - |

davitof:

Three considerations:

  • Classical music lovers tend to think of composer before performer(s). This is also how classical albums are sorted in physical stores.
  • As you know, in MB, as a general rule, the release title should be exactly what is printed.
  • MB shows release lists with release title and release artist.

This works well when there is only one composer. But as you can see in your Williams / Britten example, it could be difficult to know which composer composed what just by looking at the release artist and the release title. So this hack was set. I agree it is a deviation from the “title as printed” rule. But classical music lovers, like me, prefer this deviation to having a list of release titles which could be tricky to understand. Actually, we are doing what we would have liked the music industry editors to do.

I looked at the spine of some of my classical releases and the overriding format is something like

ComposerName: Work(s) - Conductor, Orchestra

So my point is yes the composer is very important, there is no denying that. But to me it seems the composer is part of the title, if it was considered the artist then the spine would be


Work(s) - Composer, Conductor, Orchestra

And when you look at the front cover the composer comes before the obvious title, it is not listed with the performers

So that is my point really the composer as written on both front cover and spine seems to be part of the title it is certainly not part of the release artists, am I the only one seeing it this way ?

And look at this example, a current composer conducting their own music

He has his own name twice, once before the work and once after, and on the spine the first name is grouped with the work distinct from the second reference to name and orchestra name.

You made me realize that there is a little flaw in our discussion: I think most classical releases don’t have a real title, not in the sense “title” has in most other genres. I believe it is more like a designation, a practical text which has been added by the publisher in order to make the release easier to store and retrieve. Re-releases often have small variations in the “title” like adding or removing a catalog number, or a performer, or even a work. This is of course not true of more recent releases like https://musicbrainz.org/release-group/b81e3f12-64fb-4560-adf6-4eea6ecc39df , which has a real title. BTW, I am not sure how this kind of “really-titled” releases fit inside CSG…

davitof:

Also, it should be noted that adding the composer to all titles, as you suggested, would be a deviation in all classical titles from the general rule “titles as printed”. So which is better: having all classical titles inconsistent with all other titles, or having only classical titles with 2 or more composers inconsistent?

Firstly I think the composer is part of the title. But assuming it is not we have two set of style guidelines, the standard ones and the CSG, and all releases should use one or the other. So within those guidelines it would be preferable for everything using one guideline to work in a similar way therefore it would better for all classical titles inconsistent with all other titles rather than just some.

OK, the composer could be considered as a part of the title. As you suggested, we would have to remove the composers from release artists, taking care of keeping them when they are also performers, as in your example. Maybe it could be done automatically. There would be another modification to do: the search would have to be modified to allow to search by composer easily. I believe that, for the search to give results quickly, there should be a direct relation between composers and releases. Currently, performers and main performers are added as release artists, but I don’t think that the database is able to distinguish between performers and composers at the release level. If I am correct, the database structure would probably have to be modified in order to be able to distinguish between these two categories of artist-release relationships. Plus any consequences I haven’t seen because I don’t really know the MB database structure. So now I am wondering, would fixing this “ugly hack” be worth all this work?

I am one :wink:
Having to sort out a few thousands of records.

I can say: composer & performers are both to be regarded equally important as release artist by about any classical aficionado.

In most cases I’d first care about the composer as artist, then about the performer. I’d only put the performer first in case of recitals.

But obviously the performer does matter a lot. But I don’t really think in terms of “favourite”, rather in terms of “interesting interpretations”.

On some albums the solo performer is listed before the work title, together with the composer. The position of the artist doesn’t tell us anything about whether to consider them release artist or not.

What I do know, is I want to have the composer as a release artist equally as the performer.
(In case it is a release artist, as in recitals of various composers, the performer(s) could be the only release artist)

Composers performing their own work should be listed as Composer; Performer (if presented that way) so it’s immediately clear it’s a recording involving a performance by the composer itself. (On a record called “Lecuona Plays Lecuona” where Lecuona performs solo on piano, I think it’s fine to just list Lecuona as artist however)

You really can’t draw any conclusion out of the order.

Any artist that is presented as release artist is a release artist. It has nothing to do with involvement (otherwise we would need to remove any artist appearing on a post-mortem compilation as well.)
The composer is definitely not part of the title.
Well, not more or less as any artist is part of the title of the albums they release. All artists are part of the title of a release. You don’t split them apart, unless within a context where only titles of a single artist are listed (but even there, the artist can’t be disconnected from the title.)

Thankyou for being open to this, the question is should it be part of the title. I take your point that for Classical releases sometimes the title is more of a psuedo title created by the record company rather the the performers, but it does seem clear to be that this pseudo title usually includes the composer and is of the form
composer: work

I was trying to focus on what is right in terms of the data rather than concernign ourselves with implementation of such a change. but since you ask with the size of the MusicBrainz database this change would certainly not be trivial, although I think it could mostly be handled in an automatic fashion.

Im not clear of the problem here. If searching for a release then a search that include composer would work beause they would now be part of title. If searching for a composer then in artist search that would work because composer are artists. So I cant see the issue

So i think the only change is modification od the metadata. But the point of is not to simply fix the “ugly hack”, the “ugly hack” is simply a symptom of what seems to be a fundamental issue with the composer being stored in release artist field rather as part of the title

So Im certainly not saying the composer is less important, just that they should be regarded as part of the title not the release artist. Just to clarify are you saying you would say

Four Seasons by Vivaldi, Nigel Kennedy’s and the English Chamber Orchestra

rather than

Vivaldi’s Four Seasons performed by Nigel Kennedy’s with the English Chamber Orchestra

Do you have such an example, that would be interesting to see

Yes, that is my point the Composer is not presented as a release artist, that is why on the spine it typically says

Composer: Work - conductor, Orchestra

it is seperated from the release artists and preneding the work title

It has nothing to do with involvement (otherwise we would need to remove any artist appearing on a post-mortem compilation as well.)
But the conductor, orchestra, soloist are involved in making the recording that is the difference.

That makes no sense, they are split into title and release artist fields.

Nirvana: Nevermind
→ Artist before title, so not a release artist?

It really doesn’t matter where the artist is placed compared to the title.
Some releases even only show the composer as release artist. These kind of conclusions are a dead end and lead to nowhere.

Things is: in any library or record catalog, the composer is presented as an artist, not as a title.

If you have a collection of a few thousands of classical records, you want to go to the artist profile of a composer, and see all the releases by that composer.
If you would only enter the composer in the title, the database will be completely broken and disfunctional.

1 Like

You just use the title :slight_smile:

With respect that is a spurious argument because

  1. We are talking about classical release not Pop/Rock
  2. There are just two words on the cover and spine Nirvana and Nevermind there are no other artists, so this separation between composer and others artist does not apply
  3. There is no colon separating artist from title, they are two distinct things, whereas for classical on the spine it is usually composer: title , i,.e it is in essence a sentence

You can go to that composer and select the Works tab and see a list of all their works, that seems more useful to me, if you wish to see all releases that a work is available on then select the work, and then the linked recording to see a list of releases for that recording.

I am not the only one who thinks this, if you look at Roon, which I think has the best audio player available currently they put composer in title rather than album artist,

See