Re: [RDA-L] ] The “A” in RDA

Posting to RDA-L

On 26/07/2013 22:10, JSC Chair wrote:

<snip>
Taking the bigger view is precisely what RDA will help us do – stop focusing on creating “records” and see how the resources we are describing fit into the bibliographic universe. We are living with lots of MARC limitations for now, but the data built using RDA will be especially useful when we can move beyond MARC. It is still usable in MARC just as “records” created with AAACR2 were useful in MARC, and RDA can even be used to create catalog card records, if that is your limited environment for now, but we want to look beyond the current limitations of just building a catalog to re-use of bibliographic data in the broader information community – to enable libraries to interact better in that larger realm where our users are – to connect users to the rich resources and related resources we have to offer and beyond. – Barbara Tillett
JSC Chair
</snip>

The idea that the problem is with “records” and that things will get better once they are discombobulated into various bits of data is a theory that has never been demonstrated. It also goes against reason: why should a separate bit of information such as <Paging>300</Paging> or <Title>Poems</Title> make such a big difference? On their own, these little bits and pieces of information are completely meaningless and they must be brought together again–or “recombobulated”–if anything is to make sense.

http://blog.jweinheimer.net/wp-content/uploads/2013/07/l2.jpg
 (This area apparently really exists at the airport in Milwaukee. I love it!)

The fact is: catalogs currently do not have “records” as such, because in any catalog based on an RDBMS, everything is already discombobulated into separate tables for headings, language codes, perhaps dates and all sorts of things. Internally, each catalog may separate the information in different ways. Anyway, there is nothing at all new about getting rid of the “record”–it’s been the case for decades. When a searcher of the catalog sees a record, these bits and pieces are brought together, and the human experiences the same thing as a “record”, although it can be displayed completely, partially, or it could be in many, many unique and novel ways.

I think the argument has confused database structure with data transfer. For instance, I can’t imagine anybody wanting just the <Paging> information or the <SubjectChronologicalSubdivision> without a lot of the rest of the record so that the final product will be coherent and useful. And internal database structure will continue to vary as tremendously as it does now no matter what library formats become.

In my opinion, these are side issues and the fundamental question is: if there arrives the FRBR universe that is fragmented into little bits of atoms based on works/expressions/manifestations/items, I wonder who will own what? We have already had serious issues of who owns which records, so if there are work “instances”, or as BIBFRAME seems to be leaning toward “work-expression instances”, I wonder who will own those work-expression instances? Without that information (in essence the headings but other info as well, such as language and maybe dates, etc.) the manifestation records lose the majority of their value.

Will those work-expression instances be placed into the public domain? If not, it would be like within the internal structures of your own library’s catalog, you suddenly didn’t own the information in your subject tables or the personal names in your names tables.

Or will “work-expression instances” be owned by some agency? And if they are owned, who will they be and how much will they charge?

I think that’s a pretty important issue to settle.

-175

Share