On 22/04/2013 16:13, Brenndorfer, Thomas wrote:
James Weinheimer wrote:
Very often people do not want entire items, they want parts of items, so someone may need only a few paragraphs from a 300 page book but not the entire thing. But nobody had any choices before. This editor discusses this point later in the article and anyway, as he says, we don’t even know how a book is going to be read ten years from now. That is an amazing statement but I think it’s true!
That’s sound like an excellent reason to clarify entities and relationships. One would continue to expect to ask of what an excerpt is a part. The basic question behind FRBR is: can we talk about the things we’ve always used in cataloging using an entity-relationship model? The reason we want to ask this is to see if there are ways of making that data more amenable for entity-relationship databases and comparable systems that can deliver more functionality than current catalogs.The scope of RDA also includes backwards compatibility with existing displays, so there should be some demarcation between old and new elements. If anything this seems like a very sensible and basic technical requirement prior to jumping into new functionality such as Bibframe Annotations.
While the public may want bits and pieces of larger resources (that is, people do not necessarily want entire items but much smaller parts for their own purposes–and this is nothing new but has been the case for millennia) I don’t think there is any possibility of catalogers doing such in-depth analysis, which would make their work go up 40 or 50 times. I think the result would be catalogers jumping out of windows!
So, I don’t think this has much at all to do with entities and relationships. As the DUP editor says, we don’t even know how a book will be read. Even images and music have some fascinating possibilities for automatic searching today, so I think there should be primary focus on how records made by humans can best interact with automatic indexing. Automatic indexing can only get better as hardware gets faster and software improves. All this while the available stock of catalogers does not seem as if it will be “going forth and multiplying” anytime soon. So, how can our wonderful, much-superior hand-made records work with automatic indexing in the best ways possible for the public? To me, that is precisely what administrators and the public will want.
It seems to me it is time to put aside any pre-conceived ideas and start with good, old “trial and error”. While Bibframe may be good, it is but an attempt that I hope will be good, but there must be the possibility of failure. Something that has not seemed possible for certain other initiatives.