Wikidata

Wikidata:Requests for comment/Use of P2389 as a qualifier

< Wikidata:Requests for comment

Issue

edit

The property organization directed by the office or position (P2389) is used as a qualifier on ~6,300 items, mostly qualifying position held (P39). It explicitly permitted the qualifier scope from six years ago until ten days ago, when this discussion was started. The argument against this use contends that position held (P39) should only take specific positions like Mayor of New York City (Q785304) as values, which would not require such qualification. However, the description in the documentation of that property has explicitly allowed generic values like mayor (Q30185) for over ten years, albeit with the problematic of (P642) as the recommended qualifier. I believe organization directed by the office or position (P2389) is the appropriate qualifier for this purpose, and should again be allowed as such. Swpb (talk) 19:49, 28 October 2024 (UTC)[reply]

Edit: Here are ~27,000 more items where organization directed by the office or position (P2389) would be the most appropriate replacement for of (P642). Swpb (talk) 09:46, 26 November 2024 (UTC)[reply]

Comments

edit

When you refer to the description, it would be great if you specify which language. On P39, I cannot see Mayor mentioned in the description in any of the languages I understand (sv, en, no, da, nl, de) (nor on P2389). Ainali (talk) 08:48, 31 October 2024 (UTC)[reply]

Sorry, the description in the documentation template on the property talk page, not the public-facing property description. It does not mention particular types of position like "mayor", it says "Qualify with:...If there is no specific item for the position: of (P642) (chamber/senate/assembly/administrative entity), but generally a specific item for the a public office should be created." (The text has varied over time and has not always had the latter bit about creating an item, but that says "generally" anyway, so it admits exceptions.) Swpb (talk) 20:35, 31 October 2024 (UTC)[reply]
Thanks for the pointer. So it has been there for less than nine years, not more than ten, but your point that it has been there a long time holds. Ainali (talk) 08:20, 1 November 2024 (UTC)[reply]
No, the instruction to qualify generic positions with an organization has been there since this edit in April 2014. Swpb (talk) 15:50, 1 November 2024 (UTC)[reply]
  • If we are being asked to vote on something, I   Support retaining use as a qualifier as described, it's a simple solution to avoid unnecessary items. ArthurPSmith (talk) 21:08, 4 November 2024 (UTC)[reply]
  • These should be solved by proper position items as these positions seems to be notable anyway and even if all 6300 cases are unique(they aren't), it's not a lot of items. It will make everything more consistent and easier to use. We should also make sure there are constraints in place to give a warning about this usage. Abbe98 (talk) 21:46, 4 November 2024 (UTC)[reply]
    • The 6300 number is not in any way exhaustive of statements where this use would apply. By "those positions seem to be notable", which positions are you referring to? We are talking about the leadership positions of *any* organization, of any size or notability, not just ones where this has already been done. If organization X is (just barely) worthy of an item, and someone who ran that organization is also (just barely) worth an item, it does not follow that head of organization X is a worthwhile item, when its only reason for existence is the avoidance of a perfectly reasonable qualifier. Swpb (talk) 22:03, 4 November 2024 (UTC)[reply]
      I would say that would be a textbook example of why we have the third notability criteria, it fulfills a structural need. Ainali (talk) 22:40, 4 November 2024 (UTC)[reply]
      "which positions are you referring to?" the ones returned by your query, bishops are one example. "We are talking about the leadership positions of *any* organization, of any size or notability" no we are not, in theory we are talking about all notable organizations or all organizations that is needed for structural reasons. Reality is that we are talking about ~6300 cases where the modeling is inconsistent with the documentation and the de-facto model making it harder for people and machines to reuse information from Wikidata. Abbe98 (talk) 12:23, 5 November 2024 (UTC)[reply]
      @Ainali, Abbe98: No, the 6300 are a fraction of the total number of cases where this applies. Here are 48,000 statements matching this use case, where of (P642) needs to be replaced. 27,000 of those positions are unique. Do you think we have a structural need for close to 30,000 new items with names like "honorary chairperson of Empresa Nacional Bazán", "rector of St. Stanislaus Novitiate", and "artistic director of Tolmec Dance Theatre", or can we avoid all that by letting a qualifier be used in the way it long has been? I have already shown this is in no way "inconsistent with the documentation and the de-facto model" - it was explicitly allowed for over a decade, so I wish you would retract that erroneous statement. It is also absolutely not a problem for some specific offices to have items and others to require a qualifier to express; that is an extremely common situation in myriad domains of data. Swpb (talk) 03:38, 8 November 2024 (UTC)[reply]
      Yes, I think that is the way to go, even if we end up having one or two orders of magnitude more items than your estimate. Ainali (talk) 21:08, 8 November 2024 (UTC)[reply]
      So 3 million items with no purpose at all outside the dubious "structural" role of making certain queries marginally simpler. I have to give you credit for consistency at least. Swpb (talk) 03:16, 9 November 2024 (UTC)[reply]
      It's not only about making queries simpler, it's also about having one way to do something so that it's easier for others not only to query but to discover and use the model. Abbe98 (talk) 17:16, 11 November 2024 (UTC)[reply]
      We may never get to three millions, as most of these marginal positions will not be held by notable people and thus never needed. It might also be the case that many of these "positions" really just are a title/work descriptions and nothing that is continuously filled, and therefore might be better under employer or occupation. Ainali (talk) 11:04, 13 November 2024 (UTC)[reply]
  •   Oppose I think this is in the same order of magnitude as of (P642), a “band-aid solution.”  – The preceding unsigned comment was added by RVA2869 (talk • contribs).
There is no basis at all for that assertion. of (P642) is problematic because it represents 200+ different relationships, with the intended one often ambiguous. The use of organization directed by the office or position (P2389) to relate a generic office to a specific organization has zero ambiguity. If you expect a closer to weigh your oppose at all, you'd better have a better reason, or in fact any reason, behind it. Swpb (talk) 03:45, 8 November 2024 (UTC)[reply]
  •   Oppose, as this is turning into a vote. Abbe98 (talk) 17:20, 11 November 2024 (UTC)[reply]
  •   Comment As someone who has been guilty of (mis)using this as a qualifier for position-held statements (both for historic mayors and historic museum directors) I suggest we create a qualifier for that purpose, because I 100% agree with the “Whoville” comment above. This would enable a parking spot for the misused cases to show up in a constraints list. I am not against creating mayor positions for each “Whoville”, but it would be nice to connect bio items to org items on the fly (these edits were mostly done while browsing items in wikidata) without messing up gov project queries. Jane023 (talk) 08:57, 26 November 2024 (UTC)[reply]
    (Just to clarify, the "Whoville" comment was part of this initial discussion.) Swpb (talk) 09:31, 26 November 2024 (UTC)[reply]
Just looking at the top, that list includes a lot of statements that shouldn't be using position held (P39) in the first place. For example, a large organization may have hundreds of people employed as a regular administrator (Q16532929) at the same time. Ainali (talk) 10:26, 26 November 2024 (UTC)[reply]
To clarify this comment that doesn't make sense here and was moved without any regards to keep the clarity: it is a clarification of the edit that was added 09:46, 26 November 2024 (UTC) in the issue section (last line). Ainali (talk) 22:23, 3 December 2024 (UTC)[reply]
  • I agree with the proposed solution, the new qualifier would provide a clearer relationship than P642. I too am guilty of using this qualifier on P39 because I am often lazy to create new position items everytime I need to insert this kind of information. Vojtěch Dostál (talk) 11:52, 12 December 2024 (UTC)[reply]
    However, shouldn't we be more prudent about the selection of statements for such conversion? See here Q717790#P39 - the first statement's qualifier is ripe for conversion to P2389, but the other positions are not leadership positions and as such their qualfiiers perhaps should not be converted to P2389, right? Vojtěch Dostál (talk) 11:58, 12 December 2024 (UTC)[reply]
Correct, we would not be converting any inappropriate statements to this property. There is an an entire Wikiproject dedicated to sorting out the uses of of (P642) into appropriate replacements. Where the qualifier links a non-leadership position to an organization, employer (P108) is often appropriate. Also to clarify, P2389 wouldn't be a new qualifier in this role, it would be a return to a years-long accepted practice. Swpb (talk) 21:21, 12 December 2024 (UTC)[reply]
Retrieved from "https://www.wikidata.org/w/index.php?title=Wikidata:Requests_for_comment/Use_of_P2389_as_a_qualifier&oldid=2286793087"