One thought re “main entry”… maybe the concept as such is no longer relevant in an online environment, but then some online environments may have a need for something similar but perhaps even more detailed.
Take search results lists. If your entry (work, manifestation or whatever) has multiple individuals/entites with same level of credit for creation of the work in your metadata scheme (whether that’s Dublin Core creator/contributor or something else, perhaps based on roles or whatever), but you have a limit on the number of characters in your “brief” results for “creator”, and they won’t all fit, designating one as “primary” (perhaps followed by “…” or “+” to indicate there are more in the fuller display), allows your system to make a pick based on something other than “first one encountered in the metadata or database”. In other words, gives you some control.
Citations is another… where somehow the system generating them has got to make a pick of which name comes first.
Actually you might want to be able to duplicate the order in which responsible entities (agents) are found on the title page, package or whatever, in order to generate citations following some external rules that call for that.
Technology and Metadata Librarian
Library Core Systems, Library and Information Technology Services
Room 208 Woodruff Library, Emory University, Atlanta, Ga. 30322
(404) 727-6888 / [log in to unmask]
Maybe we should belabor this point – if not now, when?
And adding to your point, in RDA RDF there are properties "has other person, family, or corporate body associated with a work” (rdaw) and "has other agent associated with a resource" (rdau) – a clear legacy of main entry, which hasn’t had real utility since we stopped using card catalogs. At what point do we let go of such outdated practices and constraints and apply practices that reflect our current environment?
Nancy J. Fallgren
Metadata Specialist Librarian
Cataloging and Metadata Management Section
Technical Services Division
National Library of Medicine
I missed "bf:authorizedAccessPoint" - which does exist. I should have looked online rather than in my downloaded files. Still, it is unclear what the meaning and function of access points are in bibliographic
data in RDF. Access points were literally "access points". But they also had a collocation function. In fact, access was through collocation/alphabetical listing. The "work" access point often combines work and expression
I can see how some access points could be considered identifiers, but that hasn't been their function -- their function really has been as access points. And if you need to access Hamlet in French in a machine-readable world it should be possible without a [Hamlet. French] access point.
It's not worth belaboring this particular point, but it's another example of how we need to do more than just blindly copy over practices from MARC to any future format.
On 4/20/15 4:54 PM, Ehlert, Mark K. wrote:
Thanks, Stephen. The access point info is particularly interesting because neither RDA in RDF, nor BIBFRAME, nor FRBR in RDF have *any* properties for authorized access points. All three do have something that is essentially "title of the work." With your reply it is now entirely unclear to me what information should be encoded as "title of the work" as well as whether there is any place for access points in these vocabularies.
* * * * *
I can't speak for BIBFRAME, but in FRBR and RDA, an access point is not defined as an entity. In fact, it didn't come out as an entity (with attributes) until FRAD came along--"controlled access point" it's called there. An access point as viewed through RDA's prism is an artificial construct that acts as a substitute identifier, though not defined as an identifier per se. Still, if I read the RDA Registry correctly, access points for works would latch onto the "has identifier for the work" property:
Mark K. Ehlert
University of St. Thomas