This is a really interesting but difficult discussion.
In one sense one could argue that Manifestations and Items are all that exist in library collections. (In fact, I suppose you could go further and say all that exists is Items. But Manifestations are quite stable, identifiable things, so it’s not surprising how central they are to catalogs.)
The Work and Expression levels exist to show there are real relationships between Manifestations.
I think the Expression level is inherently unstable in some ways. It is something that isn’t quite a Work or quite a Manifestation, so it’s not surprising that at times, things are thrown into Expression AAP’s that we would expect to belong to those other levels.
I think I agree with John that if Arden Shakespeare helps people find things, it makes sense to include the publishers name. Also if that helps librarians keep track of volume numbers and other details.
UAB Lister Hill Library
From: Program for Cooperative Cataloging [mailto:[log in to unmask]]
On Behalf Of Stephen Hearn
Sent: Thursday, August 14, 2014 5:10 PM
To: [log in to unmask]
Subject: Re: [PCCLIST] order of language and publisher qualifiers
Bib records include a mix of work, expression, and manifestation data, and a lot of the underlying relationships are only implied. If I add a cover artist's name to a bib record for a literary work (which we do for our special collections), I'm not relating the artist directly to the manifestation. I'm saying implicitly "Literary expression is part of aggregate work; aggregate work includes cover art; Cover art has artixt X." The fact that I show only "700 $a X, $e artist" on the record for the book just elides that set of relationships, it doesn't replace it with a simpler relationship between artist and manifestation. (Cf. FRBR 5.1, third paragraph.)
The argument is that a series edition of a multipart literary work is (usually and arguably) a different work (NB: work, not expression)--an aggregate work which contains the literary work (or more precisely, an expression of the literary work) as its primary content. This all gets further elided and muddled by LC's decision that one AAP should serve for both the work and its original language expression. (Note also that for LC, the idea that "Difference in translation is clearly a difference in Expression" isn't really the case when it comes to the AAP--cf. LC-PCC PS 6.27.3, "When identifying an expression not already represented by a name authority record, do not add another characteristic to differentiate one such expression from another expression; (e.g., do not differentiate one translation of Shakespeare’s Hamlet in French from another French translation ...).")
So, if the HarperCollins Spanish series were reproduced by another publisher, the extent to which the HarperCollins series AAP could still be applied would depend on the extent to which the new publisher had added to or altered the aggregate content. If not at all, I'd be inclined to say that the HarperCollins series AAP would still be applicable, especially if the new publisher reproduced the HarperCollins series title page. On the other hand, if the new publisher also added its own series title--an important piece of the definition of a series--then an AAP for that series could also be formulated and applied.
This all reminds me of the Magic Eye graphics. You look at a flat page of dense wiggles and speckles of color, and that's all you see, until suddenly--there's a dinosaur! The FRBR perspective on resources requires this kind of stereoscopic focus on what at first seems to be fairly a flat surface but can be resolved into multidimensional layers of entities and relationships.
On Thu, Aug 14, 2014 at 3:02 PM, McDonald, Stephen <[log in to unmask]> wrote:
I agree that a Spanish translation would need a separate Expression record, with a qualifier of some sort. And a different Spanish translation (i.e. different translator) would require yet another Expression record, with a different qualifier. Difference in translation is clearly a difference in Expression.
But what if Harper Collins published a Spanish translation, and a year later MacMillan published the exact same translation, perhaps even with the exact same art. Would you give this yet another Expression record? Even if they are exactly the same except for publisher? Differences in cover art, publisher, and minor added material like introductions are considered to be part of Manifestations, not Expressions, according to FRBR. FRBR doesn’t ignore those differences, it just considers them to be in a different category.
I suggest that the differences you are talking about are differences in Manifestation, not Expression. We would not express different cover art in authority records for individual titles; that would be handled in bibliographic records, as information about the Manifestation, not the Expression. Just as a Manifestation of a single Work expresses the minor differences in cover art, introduction, and publisher, perhaps we should consider Manifestations of series Works/Expressions, which distinguish identical series differing only in these minor details. That would satisfy your desire to distinguish these things for the user. In a new RDA setting, I can imagine having Series broken into Works, Expressions, and Manifestations just as we do with individual titles.
That would be a significant change, but it is the logical extension of considering a Series as a form of aggregate Work, with Exressions.
Stephen Hearn, Metadata Strategist
Data Management & Access, University Libraries
University of Minnesota
160 Wilson Library
309 19th Avenue South
Minneapolis, MN 55455