IMSLP Forums

While the tagging team have been working on the new categorization project, we've come across lots of inconsistencies in the handling of arrangements and transcriptions. Some of them appear on the page for the original work, some have their own pages under the name of the arrangers, and some appear on both of these. This is largely due to the way that IMSLP has evolved over the last five years, but it leaves us with quite a bit of duplication and potential for confusion. So on behalf of the tagging time I'd like to propose some guidelines that would address this problem:

• We should try to avoid having the same file duplicated on more than one page wherever possible
• Arrangements, transcriptions and editions should be placed on pages for the original work, unless:
  - the form of the new work is changed from the original (e.g. concert paraphrases, fantasias, cadenzas, variations), so that it doesn't represent a straightforward orchestration or reduction of a work for other forces.
  - the arrangement or transcription is part of an larger work integral (like Tchaikovsky's "Mozartiana" suite), or collections of short pieces arranged for a particular instrument that can't be split due to their layout
  - they represent additions to existing works by other composers (like Binder's overture to ''Orpheus in the Underworld'')
  - the composer of the original work can't be identified (e.g. traditionsl folk-songs)
• If the new version of a work qualifies for a different page from the original, then there should be "See also" links from the original work to the arrangement, and vice-versa.
• If the page for the original work doesn't exist yet, then it should be created, even if it only contains the "See also" link
• Every editor and arranger would have a special page called "Versions of Works by Others" in their own category, like this one for Eduard Langer or this one for Tchaikovsky, listing all their arrangements, transcriptions, editorial works, cadenzas, potpourris, variations on a theme, etc., with links to the pages for the scores in question (even if they've not yet been added). These special pages would be placed the composer category (not separate like the existing worklists) to avoid having empty categories.
• If an editor or arranger doesn't have their own composer category, then it should be created, even if it only contains the "Versions of Works by Others" page.
• We should consider giving translators and librettists their own "composer pages" as well, as knowing their dates and nationality can be helpful in checking copyright details.

The overall effect of these changes would be to create new categories for the many arrangers, editors, translators and librettists who might currently just receive a passing mention, and new lists and cross-references to help identify arrangements, transcriptions and related works in different categories.

Any comments on these proposals, or other similar ideas for improvement, will be gratefully received...

one addition:

Arrangements, transcriptions and editions should be placed on pages for the original work, unless:
- it's a part of the transcriber's or arranger's compositional oeuvre and/or is cataloged by scholar(s) as such.

Can you imagine the arguments breaking out over what constitues a "compositional oeuvre"? But if we consider, for example, Searle's scholarly catalogue of Liszt's works, then he differentiated Liszt's piano paraphrases and operatic transcriptions (S.383a to S.460) from what he called the "Partitons de piano, transcriptons, etc." (S.461 to S.577). If we were to stick rigidly to Searle's definitions, then all those from S.461 to S.577 should only appear on the pages for the original composers. However, there are some in this group, like "A la chapelle Sixtine" (S.461) and the Beethoven Symphonies (S.464) which under our proposals would remain on separate pages in the Liszt category (with "see also" references from the original work page).

I think we should stick to Searle's ordering (and other existing composer catalogues). In Liszt's section, a work that has many different transcription (versions for piano solo, two pianos, piano duet, etc.) should have one page per version, but revisions thereof should be placed on the same page. E.g. a revision of a two piano version is placed on the same page, while a piano duet version of an symphonic poem is placed on its own page.

But why should the sole criterion for an arrangement being given its own page be the existence of a separate catalogue number? Another example of a catalogue which gives separate numbers for each arrangement is David Appleby's one for Villa-Lobos. Each number doesn't necessarily represent a separate composition - Villa-Lobos arranged a number of keyboard works by Bach and Chopin for cello and piano, and each of these has its own catalogue number. But they're still arrangements, and what would make sense is for them to go on the pages for the original works just like any other arrangement. And under Davydov's proposal, there would be a page in Villa-Lobos's category entitled 'Versions of Works by Others' which would be prominently displayed (ideally) and would contain links to all of these. And I don't see why it should be any different for arrangements of the arranger's own works (except without the 'Versions of Works by Others' page, obviously) - they would go on the page for the original work. An example can be found here, although we only currently have one version. Appleby gives 7 separate catalogue numbers for this work; one for the original, 4 for arrangements for cello and piano, and 2 for arrangements for piano solo. Why should we have to find seven separate pages for this work when it would be much more convenient to list all these versions under one centralized page, since six out of the seven are just arrangements? Under your logic, a hypothetical complete listing of works by Villa-Lobos would consist of about 600 pages, when really something more like 250 would do once you factor in his numerous arrangements (and a few other things). Now, I'm not saying that we should only list Liszt's paraphrases on the pages for the original works - they are original compositions in their own right (as has been argued in many other places). But he also made a very large number of true arrangements of his and others' works, and IMO it makes no sense for each and every one to have its own page.

For the sake of being able to browse Liszt's transcription, it is necessary to have them in his section. There exists many revisions of some of his transcriptions (S.556, Die Rose (Heideröslein), S.494, Löse, Himmel, meine Seele, S.478, Russischer Galopp, S.473, Marche des Pèlerins de la sinfonie Harold en Italie, S.466, Adelaïde (Beethoven), S.461, A la chapelle Sixtine, etc.) and a multitude of editions of all of them. In the future when eventually all of them will be available on the site, the work page of the original composer will be overcrowded if they were all to be placed under that page. Moreover, and perhaps more important an issue; one wouldn't know if Liszt's transcriptions were available on IMSLP by just browsing his section. Considering that they constitute half of his compositional output, that's a major drawbackk and it will not reflect the current (and future) state of that category.

Funper wrote:In the future when eventually all of them will be available on the site, the work page of the original composer will be overcrowded if they were all to be placed under that page. Moreover, and perhaps more important an issue; one wouldn't know if Liszt's transcriptions were available on IMSLP by just browsing his section. Considering that they constitute half of his compositional output, that's a major drawbackk and it will not reflect the current (and future) state of that category.

Liszt's transcriptions will be listed in two places: (1) the existing lists of works by Searle number, and (2) the new "Versions of Works by Others (Liszt, Franz)" page. From there it will require just one click to move to the page containing the transcription. The overcrowding argument could be used both ways, as Liszt's category is currently bulging with arrangements that might be more appropriately located on the pages for the original works.

The decision that composer's arrangements of their own works should be included on the same page as the original work was taken a couple of years ago (the autumn of 2008, if I recall correctly). In the case of Liszt (most of whose self-arrangements were by then on their own pages), this still hasn't been fully implemented, which gives rise to the inconsistencies already mentioned. Hopefully that will be addressed very soon, along with the more general issue of the way that arrangements and transcriptions of other composers' works are handled...

If I understand correctly:

All works from S.461-577 (with the exception of transcriptions of his own works, but still around 100 works) are to be listed (not placed) in "Versions of Works by Others (Liszt, Franz)"?

This would have to be done manually I assume. If that's the case, there's no guarantee that the list will ever reflect the actual state of availability of Liszt's transcription, since there'll always be someone who forgets or doesn't link their submitted transcription in it. It's a shame since I personally have put my soul in that category in all respects. All in vain.

@Funper - On your first point, generally speaking that's the principle, although the list in "Versions of Works by Others" will contain links to the original work pages (if they exist), even if Liszt's transcription hasn't been added yet. Our librarians and copyright reviewers will be expected to make sure that appropriate links are included from newly-created pages, to avoid the situation you described.

I hope you won't be embarassed if I venture to say that your significant and scholarly contributions over the last few years to the Liszt category (and elsewhere) have certainly not gone unnoticed, and the changes being introduced should not for a moment be construed as critical of your enduring efforts on behalf of IMSLP (not to mention the great composer himself...).

Funper wrote: If that's the case, there's no guarantee that the list will ever reflect the actual state of availability of Liszt's transcription, since there'll always be someone who forgets or doesn't link their submitted transcription in it.

Unfortunately, no composer's work list can ever be guaranteed to be infallibly correct. IMSLP has a policy of making the upload process so easy that no one is put off by the process, but that also means that the information required to upload doesn't have to be complete or correct. Having said that, though, I think the site does a pretty good job of catching things, correcting them, and always improving them.

I understand your point about flexibility, with some composers' works being dealt with differently than others, but as one of the people who has gone through (literally) thousands of work pages in order to tag for the new search system, I have seen many inconsistencies. Most importantly, I can see where people could believe that IMSLP doesn't have something it does actually have, simply because a page is structured differently to other ones they've seen. Consistency is a preferable goal, though not always an attainable one.

My preference was to have the files on both pages, but I am in a vast minority, and I understand that there are maintenance problems with that system, so it isn't going to happen. There is, however, consistency in the guidelines that have been drawn up, so I am happy with those.

Please don't take it personally. Nothing would be lost, and the links will be there. In the end, if someone doesn't find what they want, they often ask for it in the score requests section of the forums. If we do actually, have it, it usually isn't more than a few hours before someone points them in the right direction.

This has sat here for a while without anyone raising any more questions/objections. Since we appear to have a pretty good consensus, does anyone think it's time to write this into official policy and start a general implementation of it?

OK, I've added the text to:

http://imslp.org/wiki/IMSLP:Score_submi ... scriptions

Wonderful, thanks! May I suggest that in order to facilitate browsing, the 'Versions of Other Works' page be sorted under '0' (i.e. [[Category:Tchaikovsky, Pyotr Ilyich|0]]), so it appears at the top of the list? (just thinking ahead to composers with hundreds and hundreds of work pages)

Good idea, although I've used a space instead (to avoid having the "0" subheading).

Even better I had forgotten about that.