Ah, you're right. Current practices make it difficult or impossible to tell (in an automated fashion) whether a uniform title authority record represents a Work, an Expression, or a Manifestation, and exactly what 'headings' correspond to which as well.
Jim, does this help you at all see why _some_ change of our cataloging practices is required? What we do is just too ambiguous for the machine world.
This question concerns me because I don't believe I am against changes in our cataloging--I think it's quite to the contrary. It's just that I don't think that RDA is a productive way of going about it. This colors the project I have initiated with the Cooperative Cataloging Rules, so I feel I must explain.
What do I think needs to be changed? Let's look at the same examples I gave before:
100 1_ |a Shakespeare, William, |d 1564-1616. |t Sonnets. |l German & English. |k Selections
The suggestions I have seen, and the project of subject headings at http://id.loc.gov/ does the same, is that this heading must be turned into a URI to be useful. I disagree since I don't think that even if this headings were turned into a URI, it would be very useful at all. What would be extemely useful however, is to link what can be linked, and there is an awful lot here: separate links to Shakespeare's authority record (an extremely rich source); to the Name/title record of the Sonnets; to the different German/English versions, and to different Selections. As a result, this single heading is extremely rich in linked data. So, this text-based string should be dismantled and linked wherever possible, and it is very possible here.
Therefore, the above heading would be (in very poor XML):
<name>[link to Shakespeare's heading]</name>
<title>[link to record for the Sonnets]</title>
<language>[link from list of languages]</language>
<version>[link to record for selections]</version>
Therefore, the above should all be separate links when it becomes very useful.
Let's examine the other heading I gave before:
130 _0 |a Bible. |p O.T. |p Genesis. |l Catalan. |s Clascar. |d 1914
Again, when viewed *not as a text string* but as an entire number of separate linked data, this is extremely rich. I can imagine in both cases, that the OPAC would present the user with different possibilites depending on where you clicked, using various onmouseovers to display the options. Someone would see the incredible NAF heading for the Bible, and would run the mouse over Genesis, and would get all kinds of links about the book of Genesis, going into the authority file, but it could be enriched with links into Wikipedia, into versions, into online webpages, and so on. With subjects, you could run your mouse over the main heading and get all the cross-references and scope notes immediately.
Now, from the point of view of *cataloging* i.e. what the actual work that the cataloger does, exactly what needs to change here? Are the rules for constructing the name heading for Shakespeare going to change? No, because they don't need to. How about for determining the title "Sonnets"? No. Or Selections? No. Any changes for the Bible heading? No. For individual books of the Bible? No. And this is what AACR2 and RDA deal with, not with coding.
Now, let's consider the |l in the Shakespeare heading. Here I could see a change in that |l could be made repeatable, thereby simplifying both the cataloger's and the user's tasks (no need to learn the order). From these simple examples, I think it's clear that our formats must change substantially, but the standards for cataloging do not, or not all that much. Still, I submit that deciding to "link what can be linked" would constitute a huge advance in the catalog and that our users would love it.
Do we need WEMI and RDA for this? I don't see why. First, I don't think that WEMI fits the world of information very well and to me, such an abstruse theoretical/philosophical argument is far less important at a moment when we should be concentrating on *linking wherever we can*. There are lots of places for links in all of our records, and so I see it as primarily a systems problem that can be implemented *right now.* Certainly the models that FRBR and WEMI present are at the least far too dubious for the library world to focus its diminishing resources on at this very difficult moment, as the "Study of the North American MARC Records Marketplace" makes very clear. (I'm still reading and considering it, by the way)
Ultimately, we may discover that changes in cataloging procedures are necessary or useful in some cases, e.g. the multiple languages in the uniform title I showed, but these types of changes will make themselves clear as experience is gained.
I hope this makes my own ideas clearer: I am not saying that the current situation should not change, I am saying that FRBR, WEMI and their RDA version do not provide the changes that either we or our users need. At the same time, there's a lot that can be done right now, if people are willing to open up their data for experimentation by the general information community.