arm:Volume

[sfolsom]

In the bibliographic citations model what purpose does the :Volume class serve? Why wouldn't the :page link to a bf:Instance or bf:Item that is typed/classified as volume?

[rjyounes]

Would each volume in a set of books, such as an encyclopedia, be a distinct instance (presumably they would be separate items)? If so, then the Volume class does seem redundant. If not, then I think that was the intended purpose, so that the citation could be located precisely. @timathom would remember best.

[melanieWacker]

This is interesting. A bit of a semantic dilemma. In BIBFRAME a multi-volume set as a whole would be a bf:Work, a bf:Instance and apparently also a bf:Item. I just ran one of LC's encyclopedia description through the comparison viewer and bf:Item is the item-level description of the set, not one per volume. In current cataloging we may have an authority record (work) for the set, a bib. record (manifestation) for the set -- but then we would have in our local system two item records, one per volume.

[melanieWacker]

This is the bib ID I used for the comparison viewer: 3588686

[rjyounes]

@melanieWacker I'd like to look at the results in the comparison viewer. Where is it and how do I use it (if not self-evident)? Thanks!

Hard to see how you would not have an item-level description for each volume, since a library could hold some but not all of the volumes - and they would all have their own custodial history, conceivably (e.g., one volume gets stolen). At work- and instance-levels it seems to me that semantically they'd be one resource, unless you could publish one volume separately from others, in which case they would have to be multiple instances.

[melanieWacker]

@rjyounes the comparison viewer is located here: http://id.loc.gov/tools/bibframe/compare-id/full-ttl you can just plug any bib. ID from the LC catalog into the search box, I used this one: 3588686 which is for a two-volume set.

[melanieWacker]

I assume that the comparison viewer only converts the bib. information and not the item records in the system. So it might not "know" that there are two items. Still, the definition of bf:Item is "Single example of an Instance." So if the Instance represents two volumes, would that be true for the item?

[timathom]

Apologies, I just noticed this thread--it's a good question. The BF Item proposal from 2015 (which doesn't seem to be available on the LC website anymore) introduced a distinction between compound and simple Items, but didn't go into detail about multivolume monographs or serials.

However, in the context of the citation recommendation, an arm:Volume would typically correspond to a bf:Instance (since the use case was being able to cite entries in published bibliographies). We haven't really discussed compound Instances much, as far as I can recall. But I do think it makes sense, as Steven suggests, to link the Page to an Instance that is also typed as arm:Volume. Except we would have to be clear about how we were modeling things vis-à-vis multivolume Works/Instances.

[rjyounes]

Are we saying here that we would define arm:Volume as a subtype of bf:Instance? I think the issue was that a volume could also be an Item, and therefore Volume doesn't subclass either Instance or Item, given that we are avoiding the complexities of union subclass axioms.

[timathom]

Seems like a broader discussion needs to be had about multivolume resources. Could we type something as both a bf:Instance and an arm:Volume as a way to indicate that it is part of a multivolume Instance?