lcnetdev / bibframe-ontology

Repository for versions of BIBFRAME ontology.
http://www.loc.gov/bibframe/
50 stars 7 forks source link

Proposal: Introduce CollectionArrangement, etc; Redefine Arrangement, etc #77

Closed kefo closed 3 years ago

kefo commented 3 years ago

The existing properties and class - bf:arrangement, bf:organization, and bf:Arrangement - present some unfortunate naming collisions.

At this time – May 2021 - the class – bf:Arrangement - refers to the arrangement of a collection of resources. The property - https://id.loc.gov/ontologies/bibframe-2-0-1.html#p_arrangement – associates a Work or Instance with a bf:Arrangement. These map to MARC Bib 351 – Organization and Arrangement of Materials.

The collision here is with the more common association of “arrangement” with music. Musical Work X is an arrangement of Musical Work Y. Musical Work Y is an Arrangement. This rather useful and good relationship has been shoved in bflc for a while now, but really needs to be elevated to the main ontology because it is a readily derivable relationship and a good one. The problem here is that we risk having very similarly named properties in the vocabulary with the stronger property– bf:arrangement – referring to a rather uncommon concept (when compared to that of a musical arrangement).

The property bf:organization - https://id.loc.gov/ontologies/bibframe-2-0-1.html#p_organization – is the organization of resources within a bf:Arrangement, but it has the unfortunate collision with the class bf:Organization, which means something entirely different. bf:organization also breaks a wider pattern in the Ontology

And so on….

Which is to say that this pattern does not hold (and wouldn’t make sense):

Yet, on the surface, you couldn’t blame anyone for thinking that that was a viable pattern just by reviewing the names of BF properties and classes. That person is in for a surprise.

The existing properties and the class are not used anywhere at this time. LC has not incorporated them into any profiles nor has Sinopia (per email communication with Nancy Lorimer). At LC, a scan of the MARC stats reveals that there are 3,595 MARC bib records using the 351 field. Such a small number comes nowhere close to even qualifying as a rounding error out of 22M bib records.

Expected/recommended changes:

kefo commented 3 years ago

https://id.loc.gov/ontologies/bibframe.html#c_CollectionArrangement https://id.loc.gov/ontologies/bibframe.html#p_collectionArrangement https://id.loc.gov/ontologies/bibframe.html#p_collectionArrangementOf https://id.loc.gov/ontologies/bibframe.html#p_collectionOrganization https://id.loc.gov/ontologies/bibframe.html#p_pattern https://id.loc.gov/ontologies/bibframe.html#p_hierarchicalLevel

https://id.loc.gov/ontologies/bibframe.html#c_Arrangement https://id.loc.gov/ontologies/bibframe.html#p_arrangement https://id.loc.gov/ontologies/bibframe.html#p_arrangementOf