Posting to RDA-L
J. McRee Elrod wrote:
I’ve had not one suggestion on or off list with any provision in RDA which makes it easier to catalogue electronic resources than using AACR2, which might have been added to AACR2.
That is very interesting and it certainly mirrors my experience. Cataloging electronic resources that you do not control, e.g. a digital copy of a book on the web, is not more difficult than cataloging a regular book. You just handle it differently, and decide if it is, in FRBR terms, a new item or a new manifestation. Keeping a valid URL is the major difficulty.
Working with the “real” web materials is completely different, but this does not mean that they are any more difficult to catalog. In my experience, the problems fundamentally stem from the nature of the materials: 1) it’s difficult to examine the item. With a book or map or recording, etc. it is pretty easy to examine the entire item before you begin cataloging. With web sites, you don’t know where it starts or finishes, when you have left the site or not, etc.; 2) the information on the item changes constantly and unpredictably, so you can never be sure whether the record you made last week–or even five minutes ago–has anything to do with what the resource looks like now.
But none of this is really new, since these are essentially the same problems we face when cataloging serials and looseleaf publications. That’s why I think that the real problem is: 3) web resources change with no notice. With physical resources, the mail arrives, is sorted out, and the relevant materials eventually get sent to someone who updates the record. With web materials, this no longer applies since the web master can change the title of the resource, the site can even be hijacked, or whatever and the record does not change because we have not been notified. The note we provide that gives the date the title was viewed on: “Title from home page (viewed April 22, 2002)” is just pathetic and is similar to: “Don’t blame me, I’m only the cataloger.”
None of this has anything to do with cataloging *rules* and much more to do with procedures and using technology to deal with a different kind of material. I still believe my ideas from an article I wrote in Vine Magazine from 1999 could point toward a solution. The article was much too long (my normal problem, I know) but in essence suggested that using embedded metadata within the resource could be checked by a web spider periodically and if certain information were updated, the catalog record would be updated as well.
I saw the workflow as: a selector decides a site is worthwhile and provides some instructions to the cataloger (e.g. analyse certain subsections of the site). This goes to a cataloger who creates the record(s). The web master is notified that the site has been selected as especially worthwhile and given a copy of the metadata record(s) to be placed on specific page(s). Then, if the web master changed the title of the site or other information such as the dates or basic description, he or she would be required to change the <dc:title> or <dc:date> in the embedded metadata record. (This is simple for the web master) A spider would check it periodically, and if any changes are found, they are added in the catalog record automatically, and messages sent both to catalogers and the web master to notify everyone of the changes. I saw it as an interactive CIP and apportioning the labor where it best belongs.
But I have never seen how changing cataloging *rules* have much to do with the matter.