The Zombie Stack Exchanges That Just Won't Die
My cataloging class (back in the day) focused mainly on describing individual items rather than the relationships among them. I'm curious about the state of the art when it comes to related items. Let me give you a scenario:
Bib #24601 describes the hardcover edition of Harry Potter and the Order of Fries by J. K. Elemenopee.
Bib #31337 describes the abridged audiobook on CD.
Bib #54321 describes the widescreen DVD version of Harry Potter and the Order of Fries, the film based on the book.
It's clear to me that these bibs describe things that are related to each other in an important way. But is there a good way of describing that relationship in MARC? By "a good way" I mean a practice that
would also result in appropriate behavior from the ILS, such as
Ben Ostrowsky
I think you might be talking about the 500 field which is for "General Notes". You'll often find DVDs that are based off books have "Based on the novel by..." in this field.
http://www.oclc.org/bibformats/es/5xx/default.shtm
As for the records to recognize each other and allowing a patron to select which formats to put on hold based off one entry in the OPAC, I have never heard of this, though it would be nice.
BUT
RDA is changing a lot of this. How it will work in the OPAC setting to create a relationship that you describe? It might make linking the records to each other a bit easier. BYU has RDA implemented in their catalog. I can't directly link to an entry but for Harry Potter and the Deathly Hallows, there are links on each name or title that guide you further along. It is worth taking a look at.
An RDA example of "Relationship" fields:
700 1_ ǂi motion picture adaptation of (work) ǂa Austen, Jane, ǂd 1775-1817. ǂt Emma.
700 fields: Encoding the ǂe Relationship designator term is cataloger’s judgment. It is not a core element in RDA, so you may not always see it in every record. RDA Appendix I Relationship Designators: Relationships Between A Resource And Persons, Families, And Corporate Bodies Associated With The Resource. Examples: ǂe screenwriter, ǂe actor, ǂe film director RDA Appendix J Relationship Designators: Relationships Betw, Expressions, Manifestations, And Items.
Examples: ǂi motion picture adaptation of (work)
http://tinyurl.com/7rosbmv (A powerpoint presentation given at the ULA Conference 2011 based off BYU's testing of RDA.)
LoC published some documentation about how MARC records could be used to identify holdings that are the same work or a related work (in the context of FRBR). In the relevant section about identifying matching or related works, it has this to say:
...records for the "same work" are considered to include two groups:
records that match on the name subfields in the main entry (100/110/111) and title (240/243/245 - in that order of precedence
- subfields \$a, \$n, and \$p). In cases where main entry is under title (130, or 245 - where there is no 1XX field), records that match on the title subfields are considered to be records for the same work.
records that contain an analytical added entry (700/710/711/730/740, with a second indicator value of 2) that matches the main entry/title elements in records from group one (1) above.
Added entries of an unspecified type (700/710/711/730/740, where the second indicator value is blank) are considered, by default, to be either secondary added entries or related work added entries...
I'd expect your hypothetical hardcover and audiobook would (or should) qualify as a "group 1" match already without doing anything differently.
That same document has some theoretical examples of a hierarchical display for different expressions of the same work, along with related works, and the example of The English Patient is nearly exactly what you're talking about. (I think).
That being said, there's an awful lot of "could" floating around in there. As an end-user, I've yet to use an OPAC that seems to actually try to do this.
The answer to your question is a matter of opinion; with that disclaimer, I think the answer should be: "No, there is no good way of encoding the relationship of the adaptation to the original item". MARC can only manage a general note, as Katie R. mentions, which isn't usable by an OPAC, and oft overlooked by the human searchers. John Flatness has discovered combinations which could work, as he says, but again the relationship isn't explicit, and the OPAC can make no use of this without coding which MARC currently doesn't embody.
What we really need is linked data that can explicitly describe the nature of the relationship. Ian Davis and Richard Newman's Expression of Core FRBR Concepts in RDF may not be the most current thinking on the subject, but the list of terms gives a clear idea of the relationships that could be embodied with linked data.
Linking entries in MARC already exist: the 76x-78x fields. They've been used for serials (continuing resources) for years.
http://www.loc.gov/marc/bibliographic/bd76x78x.html
The 776 field is specifically for other formats. So linking between the print and the online versions, or the print and the audio versions, for example. This is limited to a purely horizontal relationship however.
While these specific relationships are currently described in MARC via the linking fields, many of our systems are not able to fully utilize them in the way you describe above. Links can be provided, but each record still remains separate (they aren't grouped together for requesting functions).
With RDA, the use of relationship designators allows us to expand beyond the specific linking fields and make other MARC fields linking fields in a way by adding the designator subfield and code. Again, we will likely be limited by our current systems in terms of what we can actually do with those relationships besides describe them.