I think the nature of the problem you identify in your examples is not really with 'manifestation' as an entity, but with the lack of fleshing out of how to model "aggregations" in FRBR, a somewhat tricky problem.
A 3 DVD set is really an aggregation of 3 manifestations (each of a certain expression). A re-published edition with 'added bits' (new introduction, etc) is really, in my opinion: A manifestation establishing a new expression representing the aggregation as a whole including the new bits, which includes: a manifestation of an existing expression (the 'main body' whose text is unchanged), as well as manifestations of new expressions (the 'added bits'), these new 'added bits' expresisons may or may not be 'important' enough to spend much time describing/recording, but they're there. [That is, in my analysis it's simplest to say that any given manifestation (belonging to an expression set, which belongs to a work-set) can be an aggregation of other manifestations (belonging to their own expression sets, and possibly different work sets)] So, in my understanding/analysis, the problems here are less with any ambiguity or overlap in the 'manifestation' entity, and more in the trickiness and lack of established ways to model 'aggregations', whether a boxed set, an anthology, or an existing expression republished in 'aggregation' with 'new bits'. The solution lies in some better instructions for modelling aggregations in FRBR; I know there is/was a FRBR working group on that issue, but I'm not sure what it's status is. Jonathan ________________________________________ From: Resource Description and Access / Resource Description and Access [[email protected]] On Behalf Of Karen Coyle [[email protected]] Sent: Friday, March 05, 2010 3:34 PM To: [email protected] Subject: Re: [RDA-L] Question about RDA relationships (App. J) Quoting Stephen Hearn <[email protected]>: > The web statements would presumably be derived from a large set of > records, not from an individual record. The bib record for Sturges' > "Magnificent 7" if constructed the same way as the Kurosawa record > would inferentially provide the data needed to create the statement > establishing his connection. The logic would break down, though, when > the bib record describes multiple objects. A MARC bib record describing > a 3-DVD set of the three "King Kong" movies would have a hard time > associating individual directors with their movies in a way that > machines could process. As Adam observed, an authority record focused > on a particular work or expression in a way that many bib records are > not would provide a better basis for establishing these kinds of > relationships. The more I think about FRBR, the more I think that the Manifestation is the fly in the ointment. Manifestations do more than manifest an expression. They often have added bits that don't follow from the work/expression, such as bibliographies supplied by the author or publisher but not part of the main text, indexes, illustrations that are added in to a text previously published without, etc. etc. etc. So you example here of trying to make sense of a record for a 3-DVD set illustrates that well. I'm beginning to think of the manifestation as having some overlap with W/E, but not entirely --- can't come up with the word, something like "contiguous" -- with it. I suspect that's how catalogers see it, it just doesn't seem to be expressed that way in FRBR. I think a better way of saying it would be that the manifestation "manifests and often enhances" the expression. I have a terrible feeling I'm the last one to realize this. kc -- Karen Coyle [email protected] http://kcoyle.net ph: 1-510-540-7596 m: 1-510-435-8234 skype: kcoylenet
