One of the great characteristics of multiple entity models like RDF is that an identifier can be assigned to the proper entity/RDF resource. When we don’t have to jam everything into a single MARC bib record we will be able to assign identifiers explicitly to the correct RDF resource. From that resource patrons will discover associated resources.  

So for your kit example… when someone searches on the ISBN for one of the items they will find that specific item AND also find out that it’s part of a kit. If there are two items in the kit with different ISBN’s they can be treated as separate RDF resources with different ISBN’s. The kit would be a separate RDF resource and could get its own identifier/s. 

As to how to handle multiple ISBN’s for different bindings, there will always be a role for PCC (and others) to generate best practices. 


Yes, we will be working with MARC for a while longer. I, too, hope in the meantime that we make strides in sanctioned additions and work arounds to MARC that position us for better days, e.g. more $0 URI’s that are dereferenceable.



-- 
Steven Folsom

Discovery Metadata Librarian
Cornell University Library






On 4/4/15, 3:08 PM, "J. McRee Elrod" <[log in to unmask]> wrote:

>Posted to Autocat, Bibframe, and MARC.
>
>MARC needs development, and Bibframe needs provisions, to cope with
>the complexities of the bibliographic universe.
>
>As just one example, consider the ISBN.
>
>Contrary to the assumption of some MARC duplication identification
>programs, and some Bibframe advocates, each ISBN does not represent a
>different manifestation.
>
>Some manifestations have a different ISBN for each binding, be it
>hardback, paperback, library, or delux.  Many sets have an ISBN for
>the set, and for each individual volume.  Kits may have ISBNs for
>items in the kit, with or without one for the kit itself.  Serials may
>have an ISSN for the serial, but ISBNs for individual volumes,
>particularly yearbooks.  Contrary to expectations, some producers of
>nonbook materials such as DVDs are assigning ISBNs; an ISBN no longer
>means a book.
>
>Utlas (unlike MARC) had an 021 field for analytical ISBNs.  The only
>way we have found to prevent unwanted merging by ISBN is to code
>analytical ISBNs as 020$z. which is clearly wrong (MARCEdit does not
>like 020 in a serial record), but this does allow collections folk and
>patrons to find by ISBN an individual volume or item in a larger
>resource.  We need a MARC field or subfield code for analytical ISBNs,
>as one example of needed MARC development, as would Bibframe need a
>label (which are getting quite lengthy to make distinctions aren't
>they?).
>
>While working on a possible replacement, please let us continue to
>improve what we have.  We will be using present systems for some time.
>
>
>
>   __       __   J. McRee (Mac) Elrod ([log in to unmask])
>  {__  |   /     Special Libraries Cataloguing   HTTP://www.slc.bc.ca/
>  ___} |__ \__________________________________________________________