Wikidata

Wikidata:Property proposal/object has role

< Wikidata:Property proposal

object has role edit

Originally proposed at Wikidata:Property proposal/Generic

Descriptionrole held by the object or value of a statement in the context of that statement
Representsrole (Q4897819)
Data typeItem
Domainstatements (proposed property is a qualifier only)
Allowed valuesany
ExampleLouis XI of France (Q8058)significant person (P3342)Jean Bouchard (Q3170866) w/ qualifier "object has role"=confessor (Q21500210)
lightweight class (Q26211786)mass (P2067)72.5 kg w/ qualifier "object has role"=maximum (Q21067467)
Planned useMigration of the myriad current uses of P794 (P794) and subject has role (P2868) for this purpose
See alsoP794 (P794): A poorly defined property sometimes used for this purpose, which we are attempting to deprecate
subject has role (P2868): Often used for this purpose, but more properly specifies a role of the subject item
Motivation

From this discussion about current use and misuse of P794 (P794), it has emerged that that property and subject has role (P2868) are often used to indicate a role for the object of a statement, rather than the subject, and that this distinction merits a separate property. I hereby propose that property, and propose that subject has role (P2868) be consequently restricted to statements about the subject item (which is its predominant use). @GZWDer, Jheald: --Swpb (talk) 15:37, 15 February 2017 (UTC)[reply]

Discussion

  Support should we also merge kinship to subject (P1039) and version type (P548) to this property?--GZWDer (talk) 15:40, 15 February 2017 (UTC)[reply]

Either that or make them sub-properties; between those two options, I don't have an opinion one way or the other at this point. --Swpb (talk) 15:46, 15 February 2017 (UTC)[reply]
I don't see a reason for merging. Subproperties work well enough. ChristianKl (talk) 18:35, 15 February 2017 (UTC)[reply]
It is inconsistent that relative (P1038) uses kinship to subject (P1039) but superproperty significant person (P3342) uses P794 (P794) as qualifier.--GZWDer (talk) 18:59, 15 February 2017 (UTC)[reply]
I don't think that inconsistency is a problem. It's the nature of all sub-properties that they could be replaced by their parent. --Swpb (talk) 19:12, 15 February 2017 (UTC)[reply]
Using this property does not make new confusion. The reason of we having sub-properties is that they may be used in the same condition with different meaning, otherwise we can merge them to one property. location (P276) and part of (P361) is both already merged with (at least) 3 different properties.--GZWDer (talk) 19:29, 15 February 2017 (UTC)[reply]
As noted on separate PFD, I do not favour merging "type of kinship" into this proposal. This proposal should stand on its own merits, not look to circumvent another discussion.  — billinghurst sDrewth 00:15, 20 February 2017 (UTC)[reply]
Just a note: The major use of instance of (P31) as qualifier should also be merged to this property. I have started Property talk:P31#P31 as qualifier.--GZWDer (talk) 15:53, 15 February 2017 (UTC)[reply]
Agreed. --Swpb (talk) 15:57, 15 February 2017 (UTC)[reply]

  Support I think this will help clear-up. Though I'm not sure that about that use of mass (P2067) with lightweight class (Q26211786) -- but it's the use of P2067 I'm not sure about, rather than this qualifier. Jheald (talk) 16:12, 15 February 2017 (UTC)[reply]

  Support. This would make things more clear. But probably the subject item should be something more generic than role (Q214339) (social role) - maybe role (Q4897819). Valentina.Anitnelav (talk) 20:20, 16 February 2017 (UTC)[reply]

  • Indeed. So changed. --Swpb (talk) 20:38, 16 February 2017 (UTC)[reply]
  •   Comment The use of the word object (in English) is ambiguous there are three prominent meanings and they only become obvious when there is expanded context. In common English that use of "object" (opposed to subject) is probably the third common after talking about a thing, or complaining about something. Referring to someone as an object is not usual. So here, you mention subject so it has some context, though when people type for a qualifier it isn't something that people will see your proposed context, and it may be well misunderstand. Further, I think some (many?) people are going to struggle with the use of subject versus object(ive) to separate which person is which, or which thing is which. So while the proposal is technically and grammatically right, socially I think that through proposed nomenclature it is going to fail either through no use, or further misuse.  — billinghurst sDrewth 00:37, 20 February 2017 (UTC)[reply]
    • The description as I've just amended it, "role held by the object or value of a statement in the context of that statement", will be visible to anyone using the property, and should be sufficient to resolve the ambiguity. Since "value" is the formal Wikidata lingo for the item in question, "value has role" should be an alias. I still think "object has role" should be the primary name in English, since "value" has its own ambiguities, but I could be swayed. The property will also have a "qualifier only" constraint that will make it easy to spot a good portion of the misuses that do occur. It's always a good idea to consider ambiguity in new properties, but I don't think it's likely to be a major problem here. --Swpb (talk) 14:51, 24 February 2017 (UTC)[reply]

  Support Sounds good to me. - PKM (talk) 18:46, 8 March 2017 (UTC)[reply]

Retrieved from "https://www.wikidata.org/w/index.php?title=Wikidata:Property_proposal/object_has_role&oldid=471228477"