The organization of information
Notice that each of the Expression records includes a Description of a Related Expression, which contains a listing of the contents of the book. The Relationship Designator “Contains (Expression)” is associated with the Related Expression.
|Related Expression\Description † \ Contains (Expression)||Organization of recorded information — Retrieval tools — Development of the organization of recorded information in Western civilization — Encoding standards — Systems and system design — Metadata — Metadata : description — Metadata : access and authority control — Subject analysis — Systems for vocabulary control — Systems for categorization — Arrangement and display — Conclusion|
RDA treats contents notes as a description of a relationship: it consists of a list of the *parts* of the *whole* resource being described. This may be somewhat obscure in RDA, but there are examples of contents notes in RDA 18.104.22.168, under “Structured Description of the Related Work”; and RDA 22.214.171.124, under “Structured Description of the Related Manifestation” – but not at RDA 126.96.36.199, under “Structured Descriptions of the Related Expression”.
It is not obvious to me whether contents notes should be recorded as Related Works, Related Expressions, or Related Manifestations – but this example forms a good basis for discussing this issue.
In this case, the contents of each edition (Expression) have been recorded, and there are differences between them. This suggests to me that contents notes should not be recorded as Related Works; the content of a work may change between successive expressions, and any single version of the contents at the work level would be misleading. Furthermore, FRBR Works are abstractions that must be realized as a particular text, image, etc. (an Expression). So I would expect that information about features of a particular text (the titles of the chapters) would be attributes of either the Expression or the Manifestation.
I chose to record the contents note as the description of a Related Expression, because the information presumably applies to all Manifestations of a given Expression. However, the contents note is created by copying the titles of the chapters from a copy of a Manifestation, so one could justify treating contents notes as descriptions of Related Manifestations.
When recording data in our current composite bibliographic records (which contain attributes of the Work and the Expression, as well as the Manifestation), we don’t need to worry about this. But the RIMMF structure forces us to make these decisions, and I think it is a useful topic to discuss.
Let us know what you think.