Posting to NGC4LIB
On 09/08/2011 20:54, Jonathan Rochkind wrote:
I’m not sure if that’s supposed to be a parody of FRBR but in fact it accurately represents our _current_ non-FRBR catalogs, right?
What do you get if you to our legacy catalogs, in a large library, and enter “Huckleberry Finn”. You get ALL that, 200 hits, videos, audios, multiple publishers. The user has to deal with all the manifestations, and figure out the differences on their own.
In fact, it’s FRBR behind the scenes in our data that would allow our systems to provide an interface that would let them ignore FRBR.
But of course FRBR can’t tell the user what the actual content differences are between the 1962 and the 1972 editions. Only expensive human time noting and recording that stuff can do that, and we can’t afford it. Maybe we can convince users to enter it recreationally.
But it’s only FRBR behind the scenes that would let the system figure “Okay, they want a copy of huckleberry finn, they probably want a print copy, let’s just pick one for them.” With the data in our systems right now, even that’s pretty infeasible for the system to do.
Both. I am trying to demonstrate that for the average user, they don’t care one bit whether something was published in 1982 or 1993, published by Harpers or Verso or MacMillan or Random, and the number of pages means nothing. They are interested in “expressions”, e.g. Huck Finn in English as a text. From that point on, “selecting” manifestations, most don’t care about, or to put it more specifically: a *very few* people need publishers, page numbers, and so on, and even for those who do, it is for a *relatively small percentage* of their own information needs. So, someone may be researching the various printings and states of the text of Huck Finn for an article. This person needs the manifestation information (but usually, they need more than we give them). Yet, if you are a professor researching the text of Huck Finn, it still represents a lesser part of your *total information needs* which go far beyond your research on this article or book. And this is for the very few people who need this kind of information in the first place.
The average user is focused on Twain’s text and *perhaps* would be interested in knowing there are differences in the text from one manifestation to another. But that does *not* mean they are interested in the manifestations themselves, or in other words, “that information has no meaning for me.”
Who does it have meaning for? The librarian/library catalog who absolutely needs this information to maintain the collection. So, this is why I believe that the FRBR user tasks are not for the user so much as for the librarian to manage the collection. While I certainly agree that the public wants expressions, less so “works”, real “user selection” is based on factors outside of the catalog record, and occurs only when they examine the item. Of course, once the catalog went online and everyone began to use keyword, the traditional organization of the cards fell apart, so even the expressions were essentially lost from the user’s perspective.
This is why I say that FRBR will not be a solution, and this is demonstrated in the “conversation”. We are pretending that people want to select manifestations when we provide them with information that is meaningless for them, and this will still be the case if we change to FRBR structures.
What I am trying to demonstrate with the conversation is that there is a fundamental disconnect from what the users want and expect, and what the catalog record furnishes them. It has led to many misunderstandings and bewilderment among the public, and the cataloging community has deluded itself into believing that this is what the users want. If you listen to my entire podcast at http://catalogingmatters.blogspot.com/2011/08/cataloging-matters-podcast-12.html I discuss this a bit more.
As you point out, in our current online card catalogs, the search for Huck Finn breaks down, but in systems with other indexing, e.g. Worldcat http://www.worldcat.org/search?q=au%3A%22Twain%2C+Mark%22+ti%3A%22Adventures+of+Huckleberry+Finn%22&qt=results_page people can limit in all kinds of ways using information that would have been buried otherwise: format, related names, year, language, etc. I maintain that these capabilities fulfill the “FRBR user tasks” right now, and even overfulfills them. I’m sure the current options for limits and sorts could be improved still further.
If we accepted this, we could “declare victory” in the implementation of FRBR, put this tedious discussion to rest for good, and move on to trying to figure out how to provide the public with the real information that they want by trying to provide “added value” to the records in all kinds of new ways using the resources of the entire library community: e.g. doing more work with the authority files to make them useful, working with Wikipedia, figuring out what “selection” means in an open, networked environment, making reference work more vital today, ______ (please fill in the blank!) and centering it all within a new form of catalog, or portal, or service, or nexus, or something.
I think that would be exciting!