Wednesday, April 6, 2011

Re: FRBR's Expression entity

Posting to Autocat

On Tue, 5 Apr 2011 16:04:30 -0400, Lin, Bin wrote:
<snip>
I have a question on one of the FRBR terms and hope to get some help here. The importance of FRBR's Expression entity, as I am struggling to understand it more fully, seems to lie in the fact that a resource seeker will get a tree-like display of a Work’s many different Expressions, that is, if that Work is realized in more than one expressions. However according to the book Introducing RDA, a guide to the basics, “Even if a work has only one expression, it is still important to identify both the work and expression entities.” I did not find further explanations on the quoted statement. Can anybody help?
</snip>

The main idea of having only a single work record and a single expression record is that in relational database structures they only need to be entered one time instead of repeated in multiple records as they are with the current unit records. These entity records can be displayed in a tree-like structure, thereby recreating 19th century book catalogs, but they can display in different ways, much as a MARC21 record can be displayed in multiple ways as we see in different catalogs today. One attempt was Fiction Finder from OCLC, but unfortunately, OCLC has shut it down. An FRBR database could be configured to display records just as they do today, although it would be absurd to change everything to provide exactly the same thing as we have today.

One of the "known unknowns" (apologies for the Donald Rumsfeld-speak http://www.youtube.com/watch?v=GiPe1OiKQuk) is exactly what you mention: with every single record, we will have to consider work and expression entities, while today we deal with works and expressions (in essence, uniform titles) only when we need to. From a strictly logical point of view, it seems difficult to imagine how the FRBR structure could possibly simplify matters and it can only result in greater complexity. I remember a discussion on the RDA-L list where people were discussing if the signers of a treaty go into work or expression or manifestation, or maybe even the item for all I remember, I got so confused! Of course, the final result was exactly the same access as today, so the patrons would not notice much, if any change, at all.

The "known unknown" is what effect this additional complexity of dealing with works and expressions for every single item will have on record creation. According to research from OCLC, less than 20% of all records in their database had more than a single manifestation, therefore I personally think switching to FRBR structures will have a great effect (additional work for over 80% of our records), but perhaps it will prove to be relatively simple.

Of course, I don't want to get into any of the "unknown unknowns" right now! :-)

No comments:

Post a Comment