IMSLP Forums

I don't know whether the moderators can do anything about this, but there seems to be a lack of concern from users who are uploading (isolated individual) orchestral parts with regard to the appearance of the new files on the main page for the given work.

I won't name names, but there have been many recent uploads of (isolated, individual) parts without a re-edit of the work page to make the file appear in a logical place -- i.e., under the heading "Orchestra parts" and placed in orchestral score order among whatever other parts are already uploaded.

I've gone into several of these to improve the hierarchy, move files, and/or create subheadings, but it would be very helpful if all users would make sure that a work page looks organized after they've uploaded a file.

This is a skeletal version of hierarchy that I've been trying to use (entries as applicable):

===Complete score===*
====Full score====
====Piano-vocal score====
===Transcriptions===
===Excerpts other than complete score===
===Orchestral parts===

(*These need at least three "equals" signs so as to show subordinate to "1 Music Files" in the Contents box at the top of a work page.)

It gets more complicated when an opera has a suite derived from it; in that case, there would need to be two "score" headings and two "orchestral parts" headings.

We will have to set a policy on this, I suppose. Especially now since we're actually seeing a considerable number or orchestra parts being uploaded - which will only increase as time goes on. I disagree with your hierarchy, though - especially with putting the orchestra parts for a complete work in its original form beneath transcriptions and arrangements, which are frequently made long after the composer is dead. IMSLP has been somewhat weighted in favor of piano music, but the additions of items in other formats (like orchestra and band parts) is bringing some of the inherent weaknesses in the workpage format to bear.

Of course, it's what the majority of users would find convenient should ultimately be the crucial factor here. If most of our users are pianists looking for piano fantasias on operatic themes, we might wish to think using a formatting that makes such items more prominent, much as I might cringe at the idea of sticking Liszt's piano transcriptions of Beethoven Symphonies (despite their interest as piano works in their own right) on top of the orchestral originals.

Part of the problem also lies with the Wikimedia system of hierarchical headings using 2, 3 and 4 equal signs. These headings display differently in different browsers - I notice a vast difference between Firefox and Safari, for example. So, using the 3 equal heading for major divisions on a page like "Complete Opera" and "Excerpts", leaves only the 4-equal level beneath to account for permutations within the overall division. Yet, the 2-equal designation is already used for "Music Files" and it seems somehow inappropriate to use them for "Complete Opera" and "Transcriptions, Arrangements, etc."

This is one of the reasons Feldmahler has been thinking about eventually putting in a "tab" type feature which would allow each workpage to have what could be described as "sub-pages." This would also be advantageous insofar that it would allow the user interested in transcriptions and arrangements to immediately click on a tab (subpage) which listed just those items instead of scrolling down a long page of full scores, vocal scores, orchestra parts, and the like.

Handel's "Messiah" used to be our longest page, but I don't know if it still is or not with all that's been added lately. I note that Bach's Brandenburg Concerti has been split into two pages, one for the originals and another for the transcritpions. That creates problems in its own right, though, since a composer's page soon starts to take on the appearance of a telephone directory.

Carolus wrote:[...] I disagree with your hierarchy, though - especially with putting the orchestra parts for a complete work in its original form beneath transcriptions and arrangements, [...]

My logic for listing orchestral parts last is that a transcription, excerpt, arrangement, etc. is at least a full version of (a piece of) the piece. An orchestral part is one of ca. 20 or more linear layers of (any part of) a work, rather than a "whole" representation, such as a piano-vocal score, a piano reduction of a popular number, etc.

Another advantage of putting the orchestra parts at the end of the "Music Files" list would be that there is less distance during editing to move a newly uploaded instrument-part. But, as I indicated, problems arise when there are standard excerpts from a larger work, i.e., the suites from Rimsky-Korsakov's operas, or other orchestral excerpts ("Flight of the Bumblebee", etc.), for which orchestral parts would exist in addition to those from the larger work.

I guess I should mention that chamber music (i.e., 2 to 10 or so players) would be an exception for being on a main work page, rather than being segregated to a sub-page.

If we can come to an agreement about the ordering and types of scores, I can code it into the software, which would auto-sort the entries (maybe with the tabbing system). That would also answer the requests for categories with pages containing complete scores, parts, etc.

I do see your point, Lyle. There is also the probability that there will be considerably fewer orchestra librarians looking for individual parts or sets of parts, plus the musicians who wish to have one for auditions, etc. than there would be pianists, etc. who would naturally be more drawn to transcriptions.

Maybe our first level division should be:

• Complete Work (top)
  Excerpts (bottom)

Obviously, this level comes into play only for fairly large works.

Within each of those two major sections, the hierarchy could be:

• Full Score
  Vocal Score (if applicable)
  Transcriptions, Arrangements (of the complete work)
  Parts (this could include chorus scores, individual voice parts as well as the ubiquitous orchestra parts)

I agree a sort of exception to the order should be made for chamber music with 2-9(?) parts, which are frequently issued as a piano-score and parts in one package, a score and four parts for string quartets, etc.

Within each of those subsections, the hierarchy could in historical order:

• Manuscripts
  Printed editions (at least 25 years old)
  Computerized typesets

So, to put the whole thing in sort of an ordered list:

• Complete Work
  • Full Score
    • Manuscripts
      Printed editions (at least 25 years old)
      Computerized typesets
    Vocal Score (if applicable)
    • Manuscripts
      Printed editions (at least 25 years old)
      Computerized typesets
    Transcriptions, Arrangements
    • Manuscripts
      Printed editions (at least 25 years old)
      Computerized typesets
    Parts
    • Manuscripts
      Printed editions (at least 25 years old)
      Computerized typesets
  Excerpts
  • Full Score
    • Manuscripts
      Printed editions (at least 25 years old)
      Computerized typesets
    Vocal Score (if applicable)
    • Manuscripts
      Printed editions (at least 25 years old)
      Computerized typesets
    Transcriptions, Arrangements
    • Manuscripts
      Printed editions (at least 25 years old)
      Computerized typesets
    Parts
    • Manuscripts
      Printed editions (at least 25 years old)
      Computerized typesets

Carolus wrote:
So, to put the whole thing in sort of an ordered list:

• Complete Work
  • Full Score
    • Manuscripts
      Printed editions (at least 25 years old)
      Computerized typesets
    Vocal Score (if applicable)
    • Manuscripts
      Printed editions (at least 25 years old)
      Computerized typesets
    Transcriptions, Arrangements
    • Manuscripts
      Printed editions (at least 25 years old)
      Computerized typesets
    Parts
    • Manuscripts
      Printed editions (at least 25 years old)
      Computerized typesets
  Excerpts
  • Full Score
    • Manuscripts
      Printed editions (at least 25 years old)
      Computerized typesets
    Vocal Score (if applicable)
    • Manuscripts
      Printed editions (at least 25 years old)
      Computerized typesets
    Transcriptions, Arrangements
    • Manuscripts
      Printed editions (at least 25 years old)
      Computerized typesets
    Parts
    • Manuscripts
      Printed editions (at least 25 years old)
      Computerized typesets

I agree with you about this order. It seems quite logical to me. The parts (if there are many) perhaps should be in order that they appear in the score. Winds, Brass, Percussion, Strings etc. or what ever the score. Just my 2 cents

Carolus wrote: So, to put the whole thing in sort of an ordered list:

• Complete Work
  • Full Score
    • Manuscripts
      Printed editions (at least 25 years old)
      Computerized typesets
    Vocal Score (if applicable)
    • Manuscripts
      Printed editions (at least 25 years old)
      Computerized typesets
    Transcriptions, Arrangements
    • Manuscripts
      Printed editions (at least 25 years old)
      Computerized typesets
    Parts
    • Manuscripts
      Printed editions (at least 25 years old)
      Computerized typesets

So... where do the parts from a transcription/arrangement go?

Good catch, Emeraldimp. Looks like we might need another level.

So, to put the whole thing in sort of an ordered list (again):

• Complete Work
  • Original Form
    • Full Score
      • Manuscripts
        Printed editions (at least 25 years old)
        Computerized typesets
      Vocal Score (if applicable)
      • Manuscripts
        Printed editions (at least 25 years old)
        Computerized typesets
      Parts
      • Manuscripts
        Printed editions (at least 25 years old)
        Computerized typesets
    Transcriptions, Arrangements
    • Full Score
      • Manuscripts
        Printed editions (at least 25 years old)
        Computerized typesets
      Vocal Score (if applicable)
      • Manuscripts
        Printed editions (at least 25 years old)
        Computerized typesets
      Parts
      • Manuscripts
        Printed editions (at least 25 years old)
        Computerized typesets
  Excerpts
  • Original Form
    • Full Score
      • Manuscripts
        Printed editions (at least 25 years old)
        Computerized typesets
      Vocal Score (if applicable)
      • Manuscripts
        Printed editions (at least 25 years old)
        Computerized typesets
      Parts
      • Manuscripts
        Printed editions (at least 25 years old)
        Computerized typesets
    Transcriptions, Arrangements
    • Full Score
      • Manuscripts
        Printed editions (at least 25 years old)
        Computerized typesets
      Vocal Score (if applicable)
      • Manuscripts
        Printed editions (at least 25 years old)
        Computerized typesets
      Parts
      • Manuscripts
        Printed editions (at least 25 years old)
        Computerized typesets

Quick question: should vocal score with piano reduction count as an arrangement or as a separate section? From the point of view of database design (I'm thinking of the new system) it'd seem easier to classify vocal score with piano reduction as an arrangement, and vocal score without piano reduction (or if the piano reduction is basically the voices compressed into a grand staff) as parts.

If there is a good reason for vocal scores to have its own category, I'm definitely open to that.

Strictly speaking, no. It's a very commonly published format - nothing more. It is an arrangement or transcription, by definition. Vocal scores are sometimes created by the composer, sometimes another person employed to do so by the composer or other interested party (publisher, opera house).

Bizet, for example, created a vocal score for Ambroise Thomas' opera Hamlet. He did his own for Carmen and Les Pecheurs de Perles. Mussorgsky did his own vocal score for Boris. Back in Mozart's era the practice of issuing a vocal score for an opera was just beginning. I don't recall Mozart doing one for any of his own operas.

imslp wrote:[...] it'd seem easier to classify vocal score with piano reduction as an arrangement, and vocal score without piano reduction (or if the piano reduction is basically the voices compressed into a grand staff) as parts.

I think you might thinking of a "choral score," which typically has only the choral parts, with no accompaniment showing. That could be considered a "part."

As to solo voices, although something comparable to choral score doesn't seem to be common in print if at all, I somewhat recall seeing some samples of manuscripts (17th or 18th c.?) in which the solo voice is copied only with continuo or bass line, even though the piece is a regular "number" and not recitative, i.e., the other instrumental parts are omitted from the singer's manuscript part. Obviously those kinds of solo+bass parts were a rehearsal short-hand device in lieu of having printed vocal scores.

A piano reduction, in which the entire score is reduced to 2 or 4 hands, would be comparable to a piano-vocal score. IMHO, neither a piano-vocal score of a vocal work nor a piano reduction constitutes a "part," because they represent the whole structure/texture of the score rather than just one linear segment of it (flute, or trumpet, or chorus parts, etc.).

Individual vocal parts both for soloists and for the different sections in a chorus, were extracted for many of Bach's Cantatas - some of the extant manuscripts are still aroound in Bach's hand. Publishers often issued individual parts for each section of the chorus for operas well into the 19th century. Chorus scores started to become more common towards the ends of the 19th century. These will sometimes include a cue line to show what the orchestra is doing before the chorus enters. With the often steep prices encountered for complete vocal scores ($40 USD and up) - even for reprints, such chorus scores are a much more economical solution for opera companies on a tight budget.

It can be confusing to some folks, though. Look at some of the rather bitter reviews on Amazon when someone thought they would be getting a complete vocal score for 4.95 and it turned out to be a chorus score. Chorus scores are often made by extracting pertinent pages from a vocal score. Schirmer in particular had a long history of this practice.

I’d certainly agree that we need a more organized method of displaying the information on each work page. After playing around for a few hours with the system listed above, I'd like to propose two main changes.

The first of these would be to remove the top-level distinction between complete works and excerpts. There are cases where there are complete arrangements of a an orchestral work which (so far) only has a partial full score; this means that a later piano arrangement of a complete work by someone else would be listed before the composer's orchestral extracts (which seems counter-intuitive). It's also likely that the majority of people using IMSLP are used to the standard bibliographical heirarchy of 'Work > Arrangements > Extracts', and their first priority is to identify a particular version of a work, and only then to ascertain whether it's available as a complete score.

Should we also ensure is that composers' versions of their own works should always take precedence over later arrangements by others (e.g. as at present in Pictures at an Exhibition)?

In short, I'd like to suggest that the 'top-level' classification should be:

(1) Composers' versions of their own works
(2) Revisions/Transcriptions of works by others (where these differ signifcantly in form from the original)

Within these categories, the headings could be as described above:
a) Full Scores
b) Vocal Scores (where applicable)
c) Piano Reductions
d) Parts

... and within each group we could have:

i) Manuscripts
ii) First Editions
iii) Urtext Editions
iv) Other Editions (including computerized typesets)

The 'Description' fields for each file might also be standardized to show whether they cover the complete work, or an extract, which would remove any remaining ambiguities.

So, for "Swan Lake", for example, we'd have:
1. Composer's Original Version (1877)
- a) Full Score
- b) Piano Reduction [by Nikolay Kashkin]
2. Revised Version by Riccardo Drigo, etc. (1895)
- a) Full Score [nya]
- b) Piano Reduction by Riccardo Drigo
3. Arrangements
- a) Selections for solo piano by Granville Bantock

For "Carmen":
1. Composer's Original Version:
- a) Full Score [extracts only]
- b) Vocal Score
- c) Composer's Piano Reduction
2. Suites [Giraud]
- a) Full Scores [nya]
- b) Piano Reductions
3. Arrangements by Others
- a) Selections, etc.

No doubt these could be further refined, but I just wanted to throw these ideas into the hat

Perhaps Composer's versions vs. Versions by Other hands is a better first level division. It probably has more application than Complete Work vs. Excerpts. There are probably not a great many cases where the composer arranged an excerpt on his own, and it might be better to have such an except grouped with the original versions anyway.

It gets complicated in cases like Carmen and (even more so) Prince Igor. Bizet didn't live long enough to extract suites from Carmen, and the opera is most often performed with the recitatives added by Guiraud (who also extracted the suites). Prince Igor was completed by Rimsky and Glazunov, and the Polovstian Dances have been arranged by all sorts of people for all manner of combinations. I guess as we're not too strict in defining the original version we'd even be OK with those examples, and Mozart's Requiem as well.

I would have to agree with this...

Perhaps Composer's versions vs. Versions by Other hands is a better first level division. It probably has more application than Complete Work vs. Excerpts.

It also makes it easier to perhaps divide a page for some of the above mentioned popular works. I'm sure there are hundreds of arrangements of some Bizet works - everything from choir to kazoo! By divide, I mean that the arrangements would appear on a page by themselves, away from the original orchestration.

With regard to parts, is it possible to reduce the page space taken by the individual parts for a work? (Especially where there are orchestra parts). Presently the minimum seems to be two lines per file, and in most cases the second line is identical for each.