Posting to NGC4LIB
On 10/8/2013 5:19 PM, Joseph Montibello wrote:
I don’t think the point of moving to linked data is to make the catalog more coherent, or relevant, or useful. In my own opinion:
- The goal of getting library data into a linked data format is to contribute library data to the information environment.
- The benefit of such contribution is not that libraries’ tools get used more but that libraries themselves get used more, directly or indirectly.
- Moving data from our catalogs into linked data format(s) may cause the catalog to disappear as a distinct search silo at the same time that it brings library data to more people.
Just as “the child lives for our sakes not for his own”, the library’s data does not live for its own sake, nor for the sake of the libraries themselves, but for the users – and they’re increasingly out there in the information environment instead of inside our buildings.
You very may well be right: that there are no specific purposes to libraries going into the linked data universe. It may just be a jump into the dark, or as I prefer to think of it: a leap of faith.
But the reason I think that just putting our records into linked data won’t work is that any coherence our records have now is because they are related to other records in the collection. In the linked data universe, even that will be gone. What do I mean by this?
Imagine a record with a heading like “World War, 1939-1945–Campaigns–Eastern Front–Personal narratives, Hungarian”. I can guarantee that nobody–absolutely nobody will ever come up with a query like that. The heading, based on its own merits, is completely and totally incoherent and 100% unfindable. That is, except for a cataloger whose mind has been warped, somebody like me. Unfortunately, I also think in terms of “Africa, Northern” and “Russia (Federation)” but as I mentioned, my mind is warped.
And yet, even though the string we see is impossible for someone to find, I maintain that people still want the materials that use that string. In other words, people want the materials within the set defined by the library catalog “World War, 1939-1945–Campaigns–Eastern Front–Personal narratives, Hungarian” because that set represents materials selected by experts, and organized according to specific professional criteria that are completely removed from any personal or corporate advantages. The Googles cannot do that, and they will not, although they would be happy to ingest anything we do and make money on it.
Therefore, the problem is not with what catalogers make, that is: creating and maintaining these sets according to specific professional criteria, but the problem is that the public cannot find those sets today. As a consequence, the sets we make are irrelevant to the public. Back in the days of the card catalog (when these searching methods were introduced), such a heading made a certain amount of sense but now that
everything is searched by keyword, that same heading becomes crazy and incoherent. This is why I have said that the problem is not so much the way we catalog and our methods, but that the dictionary catalog no longer works for our society today.
So, until we figure out how to help someone who wants materials on memoirs of people from Budapest about WWII to get to the heading “World War, 1939-1945–Campaigns–Eastern Front–Personal narratives, Hungarian” then anything we do will remain incoherent to the public. In the past, there was the reference librarian who understood how things worked and could help people find such a heading, but those days are long gone.
Our catalogs just don’t work for people anymore. RDA and FRBR are irrelevant to this. Bibframe is irrelevant. Linked data is irrelevant. What difference does it make if it is the text string “World War, 1939-1945–Campaigns–Eastern Front–Personal narratives, Hungarian” or http://id.loc.gov/authorities/subjects/sh2010007818.html? In either case, nobody’s going to find it. How do we make the materials within that set findable?
That is the issue facing catalogs and libraries, and until we address it there will be little advance. We can put our records into linked data now (no need for RDA or Bibframe, but now) cross our fingers and hope
somebody else solves the problem for us. That’s what I think is the purpose of the jump into the linked data universe.