Bernhard Eversbergwrote:
> I mean E-R linking.
> Where MARC now has a name or subject heading (100, 110, ... 650 ...),
> there's just the text string of the heading. What belongs there to make
> it more reliable and versatile is the control number of the authority
> record.
> And so with all additional fields as they might be required by RDA
> to establish "relationships" between "entities". Any
> "entity" will
> be represented by a record (or what else?), and a record needs to have
> an address. (It can be wrapped into a URL whenever communicated in
> an HTML page, but in essence it will always be a control number.)
>
> In current MARC, there are no reliable means at all to link
> bibliographic records with each other! We are badly in need of this
> for whole-part relationships, like multipart records, and then all
> the other work - expression - manifestation relationships envisaged
> by FRBR and RDA. This has been known for a long time, but what has
> MARBI done about it?
Yes, this should have been discussed from a long time back. I suspect that one of the problems of MARBI handling it is because of the limitations of the ISO2709 format of MARC. In the MARCXML versions, this sort of linking is more or less child's play. RDF is entirely based on it.
I still believe that FRBR is primarily a look backward instead of a look forward. I see it much more of an explanation of how traditional library bibliographic records are structured instead of looking beyond. When I first read it, I couldn't understand a word and felt like a fool, so I actually read it again. The second time I understood it and realized that I knew it all already. It just restates traditional library practice in different ways and using new terminology.
I realize that FRBR is supposed to give a "generalized view of the bibliographic universe" but that is a huge task and I don't think they succeeded. I'm not faulting them--maybe when they did it, it was simply in a time when too many things were changing too quickly. I don't know.
Another thing I believe is missing--at least from my reading of it--is a sense of the unity of the catalog as an entity. As an example, you mention the whole-part relationship is not handled very well. That may be true in one sense within an individual record, but if you take the 440, 490/8xx methods for series/analyzed serial control, the catalog as a whole functions. There may not be a separate record for the serial aspect (except as an authority record) but the arrangement in the catalog, by using the series heading recreates the whole.
Jim
James Weinheimer j.weinheimer_at_aur.edu
Director of Library and Information Services
The American University of Rome
Rome, Italy
Received on Thu Nov 29 2007 - 08:54:41 EST