I agree with Chew Chiat Naun that conventional title data for collections may eventually be more usefully found in genre/form data elements, but that will depend on whether collections are described as works.The FRBR Review Group's "Final report of the Working Group on Aggregates" (
http://www.ifla.org/files/assets/cataloguing/frbrrg/AggregatesFinalReport.pdf ) recommends that collections of expressions (one of the three identified types of aggregates) can often be sufficiently described at the FRBR Manifestation level--"The aggregating work may, or may not, be deemed important enough to be recorded." (Report, p. 7)
The problem is that FRBR itself states that subject relationships, which presumably will someday include or be parallel to genre/form, are recorded only in relation to works (FRBR 3.1.3). If an aggregating work is not deemed important enough to be recorded, then where do subjects fit in? Should FRBR be revised to permit subjects to be assigned to manifestations? Can subjects assigned to aggregating works described only at the manifestation level be considered to belong to an otherwise unidentified and unrecorded FRBR work? Will relating an aggregate to subjects and genre/form be a case that requires that the aggregating work be recorded as a work? As the Working Group on Aggregates report makes its way toward incorporation into FRBR, these issues will need attention.
Another case on which more guidance could be helpful would be that of collections known by a component work's title. "The Tennis Court Oath" is the title of both a John Ashbery poem and a collection of John Ashbery poems, established for a bilingual AACR2 expression as "Ashbery... Tennis court oath. $l Spanish & English." Assuming this preferred title decison survives the transition to RDA, we could still want to distinguish between it and the single poem with our authorized access points. Does the unqualified title signify the collection, as now, and would a (Poem) qualifier be added to the authorized access point for the single poem? Should the single poem's AAP be unqualified and the collection qualified by (Collection)? Guidance on how to handle such cases consistently would be welcome.
Stephen