On 01/03/2013 16:24, Brian Briscoe wrote:
I have no problem with the search for more effective solutions to cross referencing, but there is currently no better way to collocate all creations by a given creator than the authority structure in use by libraries. No digital search algorithm that I have seen has come close to identifying works by creators working under different monikers. Let’s not confuse the inefficiency of traditional interfaces with the organizational and access concepts that lie behind those interfaces. Your belief that users don’t use our current catalogs does not mean that all parts of those catalogs are “broken.” Authority files and the authority concept are not broken, they just are not wrapped up in a very attractive wrapper right now.
People use our catalogs now, but primarily because they have to if they want to get into the holdings of the library–not because the catalogs work very well for them. Still, I shall repeat that it is not my belief that the catalog is broken–it’s a fact, as I demonstrated in my last podcast. I can come up with lots of other examples, but that seemed overkill and I had a pretty good test case. Even I couldn’t make the catalog work as it is designed to do. To expect a regular user of the catalog to do even 1/4 of what I did is expecting far too much. I think it is important to recognize all of this.
When an information tool breaks, the result is different from a broken water pipe or a flat tire. When most things break, it is pretty obvious. When an information tool breaks, the result is confusion and I think this is what most people who use the catalog experience.
But I am not saying that the authority files or concepts are broken. Far from it, although they do need to be completely reconsidered from scratch. The authority structures need tinkering to make them work in the real world of today. Probably a lot of tinkering.
Still, we get an idea of how it could work with “different monikers” with VIAF. There, a single person has all kinds of “monikers”. We can see all of the equally valid preferred forms of the name of Benedict XVI. http://viaf.org/viaf/102167198/#Benedict_XVI,_Pope,_1927– (Unfortunately, there are no non-roman forms, but others have those forms) When someone implements the URI to this record, all or any of this information can be displayed in any way implementors want, including displaying all forms or some forms. We see it can work. Can this concept of multiple valid forms be extended to include the cross-references (4xx)? Of course. Would things be different for the searcher? Of course.
In a physical/card/paper environment such ideas make no sense, but perhaps they can make sense in today’s environments.