Re: [ACAT] New Bibliographic Framework: Update with Eric Miller

Posting to Autocat

On 09/10/2012 23:26, Humpal, Nathan wrote:

I think he made an interesting point that RDA, and FRBR have sort of helped librarians conceptualize the work that they do within a larger linked data framework. I feel that, despite a great prevailing frustration with RDA, and a general head-scratching towards FRBR, both have dragged catalogers, and librarians in general, away from the linear, one book to one record concepts of the past to an understanding of a web of concepts. Even if the theoretical ideal of FRBR is never realized, even if Work, Expression, Manifestation and Item are never fully integrated into libraries, a germ of it will probably always be there, which may facilitate a more nuanced understanding of how the catalog fits within the broader framework of the web.

Of course, I’m saying this only having come out of library school three years ago. I’ll fully admit to not understanding the entire history of the data structure of catalogs, nor the prevailing attitude of catalogers. But, at the same time, things like FRBR, this new bibliographic framework, and, to a lesser extent, RDA are bright eyed beacons for me as a young idealist.

An interesting point, and perhaps it explains (for me) some of the attitudes toward RDA among the younger people. Has RDA and FRBR “dragged catalogers, and librarians in general, away from the linear, one book to one record concepts of the past to an understanding of a web of concepts”? I think not. Why? Because FRBR was never really discussed by the library community; it was simply given to them as a fait accompli, and the only task for librarians was to understand it–not to question it. As a result, FRBR was never questioned in a serious way until much later, after it had already been accepted as dogma. I believe that many now are finally questioning the FRBR user tasks, seriously wondering if they really are the primary tasks that users want, realizing it has never been proven, and they can simply examine their own searching patterns and needs. This is why I think the purpose has subtly shifted away from “implementing the FRBR user tasks”, and toward “entering the Linked Data Universe”. This shows the importance of Eric Miller’s talk, I believe. He says that to implement linked data, you don’t need FRBR; you don’t need RDA. He thinks it all should be done in as simple a way as possible because the more complicated you make it, the less willing others will be to work with your data. I agree.

Since FRBR is based on an ideal vision that has little to do with reality, now with RDA, which is involved in the nitty-gritty world, we are discovering that it is not such an easy thing to create these different entities of work/expression/manifestation/item. If it were possible to start all over from scratch, with a blank database, I shall grant for the moment that FRBR structures may be OK. But such entirely new starts rarely happen in the world. Everything has a past: the handsome prince or the beautiful princess entrancing our minds turn out to actually have a past and a reality, much of which may be unpleasant. Everyone has their own limitations and their weaknesses, and they will make those limitations known sooner or later. We learn to move beyond those ideals we have in our minds, to accept people more or less as they are so that everyone can move forward.

Bibliographic information is much the same. It has a past. There is some information that simply does not exist in the older records, and ignoring that fact harms the whole (as I tried to show with my recent podcast In the card catalog, which unfortunately can still not be ignored when considering how our data functions, the WEMI were not “entities” but how the cards were arranged. This was all completely lost in the OPACs, especially when keyword searches appeared.

Translating these arrangements of the cards into “entities” with different “attributes” may not have seemed that big of a deal when trying to wrestle intellectually with the concepts of FRBR, but in its realization with RDA, it turns out to introduce complexity rather than taking any away. Plus, any utility to the searcher is non-existent because not only are the FRBR user tasks subtly accepted as not being the tasks users want, but especially when we compare what can be done today with the new faceted catalogs such as Worldcat, where a search for a uniform title immediately allows the searcher to navigate through the WEMI in any number of ways. Today, right now. Here is the search in Worldcat for Dicken’s “Oliver Twist”, and the searcher can choose the facets to get German, or Roman Polanski, or by date, or whatever. The user interface, along with everything else, could be improved of course. How could an actual WEMI improve on that? I have yet to hear anything and it seems that faceted catalogs are more embarrassing than anything else since nobody wants to talk about them.

I find that very telling.