Posting to RDA-L
Bernhard Eversberg wrote:
John Attig wrote:
> I don’t believe that FRBR deals explicitly with multiparts;
Well, the section
“126.96.36.199 Whole/Part Relationships at the Item Level”
explicitly addresses the issue. Without, admittedly, giving much guidance for dealing with it.
> in FRBR terms, the entire multivolume set would constitute one item belonging to the manifestation of the expression of the work representing the set as a whole.
And how useful ist that? Shakespeare’s “As you like it” as a part of a “Collected Plays” edition is not a manifestation of the work? Even if within this collection it is a separate volume with its own title page and perfectly citable? I believe we shouldn’t like it that way.
> Alternatively, each volume would be an item belonging to the
> manifestation of the expression of the work embodied in that volume. It
> seems to me that FRBR lets you model the situation either way — or both.
For an isolated catalog, this used to be acceptable. For cooperative cataloging, it meant lots of duplicates in the database. For the RDA vision of a Bibliographic Universe of Everything, it is not even good enough.
In my experience, the one area of bibliographic control that has the least amount of agreement is in the analytics: each bibliographic agency has its own idea of precisely what belongs to precisely what and how to describe it. Therefore, we have major problems in even getting a basic mutual understanding of series, serials, sets, and collections such as conference proceedings. I honestly do not think that we can ever hope to get anything even close to a general agreement on this, so we have to look to other solutions.
This relates back to user needs. People want the work or expression, while most more or less don’t care about the physical embodiment. I certainly agree with Bernhard that very few people know to search for “Shakespeare selections” or “Shakespeare works” to get a copy of “As you like it.” This is one of those searches that tended to work much better in a card catalog where people had no choice except to browse by author, than it does today with keyword searching.
People normally want individual articles from Time Magazine, not the whole thing. I think this can be extended to all kinds of collections, especially conference proceedings where access can be woefully inadequate. Of course, while people want individual papers they *may* also want to know about the materials related to the one they are looking at. With online resources, these considerations will probably only get more and more tricky.
I think we should rather explore ways of bringing all of these different views of works and expressions together instead of trying to mandate that everything fit to a Procrustean Bed. The power of computers is such that I have no doubt it can be done today, but the displays could be very strange. Or, it could turn out that bringing these differing views together may make the bibliographic record more understandable and useful than ever before. (Sorry for using such an obsolete term as “bibliographic record”!)
Although I am certainly no fan of FRBR, I believe the model could accommodate this.