Issue 649: Redrafting the scope note of R10 has member
In the 56th joint meeting of the CIDOC CRM and ISO/TC46/SC4/WG9 &49th FRBR/LRMoo SIG, the SIG resolved to start a new issue where to discuss the scope note of R10 has member.
Points that need be taken into consideration upon redrafting the scope note:
- R10 should express a generalization over all the cases that are not covered by the more specific modeling constructs available.
- The scope note of R10 should explicitly refer to its usage in documenting different databases integration, to reflect difference of opinion/practices: database A can list two distinct versions of a specific musical score/libretto/whatnot as instances of F1 Work or instances of F2 Expression.
- The scope note of R10 should explicitly state that it applies to a “superwork” construct, i.e., one that allows statements of the sort “such-and-such instances of F1 Work form distinct versions of Symphony No.something (F1)”
- Nb. The construct is easier to grasp if one’s examples are musical scores compared to adaptations of literary works into movie scripts.
- Nb’. It is often the case that whereas the distinct versions of a certain F1 Work have numbered IDs, the superwork lacks one.
- Disambiguation can come from deprecating examples that are common to both R10 and R67.
- Only keep the example about all the known different-sized versions of Rodin’s “Le penseur” –but rephrase it to avoid making a reference to “La Porte de l’Enfer”
- Consider making an example out of the known versions of Fidelio/Leonore, the creation of which has been carefully documented.
- Consider making an example out of Philokalia, a collection of texts written between the 4th and the 15th centuries AD by spiritual masters of the mystical heychast tradition of the Eastern Orthodox Church.
- Instead of making R10 a relation btw instances of F1 Work, it can apply to an instance of E89 Propositional Object and an instance of F1 Work. This way its scope is broadened, and it becomes more compatible with the notion of ‘superwork’ that we talked about.
HW: PR, TA, MZ to redefine R10 taking the points above in consideration.
Crete, May 2023
Post by Pat Riva (12 July 2023) --EVOTE
Hello all,
I am calling for an e-vote relating to LRMoo. Please vote on the list by July 26th.
Background: At the end of SIG meeting #55 in May 2023 in Heraklion, the LRMoo WG presented a sketch of a new approach to R10 after the proposal made at the meeting was considered insufficient. This was received with interest but there was not time to discuss a full proposal.
This e-vote is to approve the redefinition of R10, and the consequences on R67 (which was originally a subproperty of R10).
Please see the proposed new text in this Google doc (E-vote: R10).
Thanks,
Pat
Post by Pat Riva (30 July 2023) --evote extended
Hello all,
Not having had any votes during this vacation period, I'm extending the deadline to Sunday August 20. That is the beginning of the IFLA congress where the BCM Review Group meeting will be held and we will be bringing LRMoo for their approval.
Thanks,
Pat
Post by Martin Doerr (1 August 2023)
YES!
Martin
Post by Øyvind Eide (1 August 2023)
YES
Øyvind
Post by Pierre Choffé (1 August 2023)
Yes,
Pierre
Post by Christian-Emil Ore (1 August 2023)
Question about the direction:
R10 is member of (has member):
Domain: F1 Work
Range: E28 Conceptual Object
In all the examples the inverse is used. What is the motivation for the direction in the declaration?
Best,
Christian-Emil
Post by Pat Riva (2 August 2023) --reply to Christian-Emil (1 August 2023)
We were following the rule that a property declared in an extension must have its domain in the extension when reading in the normal direction. The range can be in CRMbase. Is it true this is the rule?
The inverse does seem more natural in the examples but maybe not in the way the property instances would be constructed.
Post by Christian-Emil Ore (2 August 2023) --reply to Pat Riva (2 August 2023)
Hi,
I thought that may be the reason. In our models the properties are bidirectional so from a purely logical point of view there is no no domain and range. All properties simply connect classes. Still it is a good and clear principle, so keep it as it is.
My vote is YES.
Best,
Christian-Emil
Post by Achille Felicetti (2 August 2023)
Dear Pat,
I confess that during the discussion at last SIG I was somewhat confused and not at all convinced of the logical consistency of the relationship between R10 and R67 and I had retained many doubts at the end of the meeting. But the way R10 has now been remodeled, stripping away any link to R67, the new scope notes and examples look completely convincing to me.
I vote YES.
Ciao,
A.
In the 57th joint meeting of the CIDOC CRM and ISO/TC46/SC4/WG9 &50th FRBR/LRMoo SIG, upon discussing Issue 360 the SIG resolved to close the issue on the grounds of there being no objection voiced at the evote that settled it.
A summary of the decisions (that have been incorporated in LRMoo since v0.9.5) can be found below:
- F1 Work. R10 is member of (has member): E28 Conceptual Object
- Redraft scope note, all new examples
- No superproperty (Outside of CIDOC CRM scope)
- R67 has part (is part of): change superproperty to P148, remove references to R10 from scope note, bring in 2 non-repetitive examples
Issue closed
Marseille, October 2023