Collections

[Carolus]

Thanks to Feldmahler, we now have a new category - Collections - which functions similarly to the Arrangers and Editors categories. There are two ways to include a work page in Collections. The first, used on those collections consisting of the works of a single composer, is to add the following to the General Information section: |Page Type=Collection . The second, for collections which include more than one composer, adds the page to the Collections sections of the category pages of composers whose works are present: {{ColCat|Firstname|Lastname}} .

Sibley has been uploading lots of collections lately, so now we have a method of dealing with these. It's also good for handling items such as the volumes of the Handel works edited by Chrysander, where keyboard works (for example) were gathered together in a couple of large volumes.

[Eric]

In the former case- collections of works by a single composer -do we have a good standard way of defining "Collections" in this context (defining what one is and is not) ?

(I may be overusing or misusing (probably not underusing) this. Anycase, if not definition- in the sense of greater definition, not dictionary definition- any help is appreciated! )
Eric

[Carolus]

The usual definition of a 'collection' is any volume which contains more than one work. Single-composer collections can be assembled by composers themselves (frequent), by others after a composer's death (extremely common), by publishers and others during the composer's lifetime (common) - with or without the composer's blessing. There are of course plenty of multi-composer collections as well. Multi-composer collections are typically assigned to the 'Various' category, with the {{ColCat|Firstname|Lastname}} added for every composer represented. Single-composer collections use the |Page Type=Collection added field in the General Information section.

The great advantage of having the single-composer collection option available is that it will allow for the upload of items such as the complete volumes of the BGA, with all of the critical commentary and the like. Collection pages like this should ideally be confined to just the collection itself - no arrangements or other items appearing there. These pages don't normally appear in the composer's regular category, but in the "Collections with..." subcategory.

[Eric]

I think I understand completely but not entirely. Couple of questions still-

1) what distinguishes suite-like pieces- especially those some of whose items first appeared separately - from collections generally?

2) I'm actually a little reluctant, despite (my) current behavior, to overuse the collections feature, since the features (tags, etc.) of any item put under page type collections do not get counted in the Category Walker. (Last I checked, anyway... !)
So if a sonata only appears in a sonata collection or some similar case, it can't, I think, be narrowed-in-on and found using the Category Walker - and this it seems to me, is a bug.

[Carolus]

Interesting point. I'll check with Feldmahler about this.

[Davydov]

So if a sonata only appears in a sonata collection or some similar case, it can't, I think, be narrowed-in-on and found using the Category Walker - and this it seems to me, is a bug.

Wouldn't the page for the sonata collection be tagged as "Sonatas", and thereby picked up by the category walker ?

(Ideally there would links to the collection from every work page involved in the collection, but that's a lot of work to ask of our librarians.)

[Eric]

that's what I meant in the first place- there was nothing else I could have meant, since the hypothetical sonatas in this one-composer sonata collection didn't have their own individual pages. (Well, it could have been something like "Complete Wind Quintets (Danzi, Franz)" in which case both the quintets collection and the individual quintets pages are tagged, but I meant that a page like Dew Drops or something similar - or a similar page which is a set of rondinos etc. under another composer - does not seem to show up under the category walker at all. Maybe this has been fixed (or maybe not - even Dew Drops doesn't even show up in the category walker under "Pieces" of all things... ! ... ) - who knows.

Also, not every collection is a page that links to various work pages. Why- should they be? If so, then I really do misunderstand the idea from point 1. The dew drops page is a collection in part because its various items have different opus nos., but they don't have separate work pages. (If anything, I removed some items' separate work pages in a similar case, once I was told about this functionality, and redirected them -to- a collections page for a similar reason, since the only reason I gave them separate workpages was because they had different op.nos.)

[Davydov]

Hi Eric. I'm not aware of any bug affecting the categorization of collections pages, but if you're still finding issues then please let me know.

As far as definitions are concerned, I'd say:

• Work = A composition intended by its composer to be performed independently from any other composition.
• Collection = Two or more works (as defined above) published together in a single edition.

Thus a suite from a ballet or opera meets the definition of a "work" because it's independent from the original ballet or opera.

I don't think anyone's suggesting that the basic principle of one page per work should change, but where it's not possible or desirable to split collections into their constituent works, we can include link from the individual work pages to the pages for the collections, and vice-versa.

In an ideal world, if a work in a collection doesn't have its own work page, then one should be created for it, even if for the moment it only includes a link back to the collection. But our librarianship resources are spread very thinly, so we can't do everything all at once.

[Eric]

Even this definition, as noted, bows to tradition with enough counterexamples to make a mathematician's head explode (Chopin's preludes op.28 were published as one but were apparently not intended to be performed as a group, but are not generally regarded as a collection, similarly Medtner's opp. 38 to 40 sets individually, and other examples) - but will attempt the general idea. Thanks

[steltz]

The problem as I see it is that a collection of keyboard works might include sonatas and toccatas, as well as many other taggable titles. With the "collections" tag, those works won't show up on the list of sonatas, etc.

And, of course, if links aren't made between the individual work pages and the collections page, then a user would miss one edition. This will be tricky because someone uploading a collection may not find work pages for everything in the collection, but a year from now, if someone creates a work page for a "new" sonata without realizing that we already have one edition, the link won't be made.

Ideally, the collections should be checked periodically to see if new links need to be made.

[Notenschreiber]

As far as I see, compositions with the same opus number are treated as one work here at IMSLP. But this contradicts clearly the definition of
a work given by Davydov above. The meaning of the notion "work" is not so clear after all... But my head is not exploided, in spite of being a
mathematician

[Carolus]

As with so many things, the dividing line between a "work" and a "collection" can be rather amorphous at times. The Chopin Op.28 cited is a good example. Since it falls under a single opus, the normal thing to do would be to treat it as a single work - even though it is really a collection which was assembled and published by the composer himself and falls under the succinct definition given by davydov. Interestingly, Brown divides the 24 preludes into odd groups organized by year of composition, while Chominski assigns a number to each one.

From what I can tell, there is no actual bug. The genre and instrumentation tags remain just as they are when something is transferred to the 'Collections' subcategory. The big advantage of this new feature is that we can now add the numerous collections out there as collections, which means that the individual items can be split out and added to the proper work pages at a later, more convenient time. The collection page (of arrangements, for example) remains even after the individual works are split out, and can be linked to in the individual items added to work pages.

[steltz]

Here is a good example:

http://imslp.org/wiki/Violin_Sonatas_(Aubert,_Jacques)

which has been put into the Collections category. I've tagged it now as sonatas, so it will show up in the lists, but I don't think it belongs in the Collections category.

Firstly, it isn't part of the collected works of Aubert in general (which I don't think exists), but if this is a Collection, then any set of sonatas published as a unit is a collection, and then the uniqueness of that category is watered down heavily.

I don't think a set is necessarily a collection -- it should be something bigger.

Any other opinions on whether this page should be a collection in and of itself?

[Davydov]

The definition of a collection is that it's an edition containing more than one work. So far so good

It's really the definition of a "work" that's the key. For that I usually check Grove, or the Library of Congress authorities. They indicate there were 5 sets of violin sonatas by Aubert, each of which constitutes a "work" with its individual opus number (1, 2, 3, 4 and 25).

The Auber Violin Sonatas page contains only book 3, so I've moved them to "10 Violin Sonatas, Op.3 (Auber, Jacques)".

[Notenschreiber]

But why under the collection tab?