Wikidata:Property proposal/specific part

From Wikidata
Jump to navigation Jump to search

specific part[edit]

Originally proposed at Wikidata:Property proposal/Generic

   Not done
Descriptionqualifier to indicate a specific part of the object-value of a statement that the statement applies to
Representspart (Q15989253)
Data typeItem
Example 1Ripon Cathedral (Q638466)historic county (P7959)Yorkshire (Q163) → specific part = West Riding of Yorkshire (Q1934075)
Example 2Teversal Manor (Q15979549)located in the administrative territorial entity (P131)Ashfield (Q600996) → specific part = unparished part of Ashfield (Q105909200)
Example 3Holyhead (OSD 317) (Q106156311)derivative work (P4969)(O.S. Old Series sheet 78) -> specific part = northwest (Q5491373)
Planned useChange the existing qualifiers used for unparished areas (40,000) and for ridings of Yorkshire (45,000) to use the new property
See alsoapplies to part (P518), depicted part (P5961), object has role (P3831)

Synonyms[edit]

  • specifically
  • applies to part (of object-value)

Relation to other properties[edit]

Motivation[edit]

This property is intended to be analogous to applies to part (P518), but specifying a part of the object-value of a statement, rather than a part of the subject.

Use of the property is likely to be limited to specific particular kinds of situation, because in general if a statement applies to a specific part of its object, the normal tendency will be to create a item for that specific part of the object-value, and then make that more specific item the object of the statement. But in a few cases, as in the examples above, that is not appropriate or not desirable, and so this property would be useful, for instance:

  • ridings of Yorkshire. The values for historic county (P7959) are defined to be the counties that existed before Local Government Act 1888 (Q6664051). However the three ridings of Yorkshire and the Isle of Wight (Q9679) are often thought of as traditional counties in their own right, as indeed they were after 1888. For informational value, and also to dissuade people from giving them as main values of P7959, it is useful to specify them in a qualifier. location (P276) has been used so far, but this property would be a better fit.
  • unparished areas. See recent discussion on Project Chat. Use of located in the administrative territorial entity (P131) to point to a partial area defined by the absence of an administrative structure is questionable, and produces undesirable repetition for downstream interpreters of the data. statement is subject of (P805) has been used in the qualifier role, but is not really right. The qualifier now proposed would be a much better solution.
  • parts of derivative works. The derivative work as a whole is an appropriate value for derivative work (P4969). But it is useful to be able to specify which part of the derivative work it was that was based on the underlying work -- for example Holyhead (OSD 317) (Q106156311) was the basis specifically for the [File:OS old series 1 63360 78nw.jpg north-west quarter sheet] of O.S. Old Series sheet 78.
  • in the other direction, on based on (P144) statements it may be useful to be able to indicate which part of an earlier work a later work drew on.

A limited number of further use cases may also exist, that may currently be being handled by applies to part (P518). When this is the case it can be very confusing, as it can become unclear whether applies to part (P518) is referring to a part of the subject of the statement, or a part of its object. By creating this new property, specifically for parts of statement objects, that unclarity would be dispelled. P518 would then only ever refer to part of the subject of the statement, and this property would only ever refer to part of the statement's object.

Discussion[edit]

User:Zolo
Jane023 (talk) 08:50, 30 May 2013 (UTC)[reply]
User:Vincent Steenberg
User:Kippelboy
User:Shonagon
Marsupium (talk) 13:46, 18 October 2013 (UTC)[reply]
GautierPoupeau (talk) 16:55, 9 January 2014 (UTC)[reply]
Multichill (talk) 19:13, 8 July 2014 (UTC)[reply]
Susannaanas (talk) 11:32, 12 August 2014 (UTC) I want to synchronize the handling of maps with this initiative[reply]
Mushroom (talk) 00:10, 24 August 2014 (UTC)[reply]
Jheald (talk) 17:09, 9 September 2014 (UTC)[reply]
Spinster (talk) 15:16, 12 September 2014 (UTC)[reply]
PKM (talk) 21:16, 8 October 2014 (UTC)[reply]
Vladimir Alexiev (talk) 17:12, 7 January 2015‎ (UTC)[reply]
Sic19 (talk) 21:12, 19 February 2016 (UTC)[reply]
Wittylama (talk) 13:13, 22 February 2017 (UTC)[reply]
Armineaghayan (talk) 08:40, 10 March 2017 (UTC)[reply]
Musedata102 (talk) 20:27, 26 November 2019 (UTC) Hannolans (talk) 18:36, 16 April 2017 (UTC)[reply]
User:Martingggg
Zeroth (talk) 02:21, 4 June 2018 (UTC)[reply]
User:7samurais
User:mrtngrsbch
User:Buccalon
Infopetal (talk) 17:54, 9 August 2019 (UTC)[reply]
Karinanw (talk) 16:38, 24 March 2020‎ (UTC)[reply]
Ahc84 (talk) 17:38, 26 August 2020 (UTC)[reply]
User:BeatrixBelibaste
Valeriummaximum
Bitofdust (talk) 22:52, 26 March 2021 (UTC)[reply]
Mathieu Kappler
Zblace (talk) 07:22, 24 December 2021 (UTC)[reply]
Oursana (talk) 13:16, 17 May 2022 (UTC)[reply]
Ham II (talk) 08:30, 25 January 2024 (UTC)[reply]
Notified participants of WikiProject Visual arts given its potential use on paintings and drawings etc. -- Jheald (talk) 14:28, 30 March 2021 (UTC)[reply]
@Multichill: There are other reasons for this proposal than just the unparished areas -- being able to specify a specific part of the object-value is a generally useful thing.
But in respect of eg Teversal Manor (Q15979549), here are some thoughts.
  • For most places in England, other than administrative areas, one would expect located in the administrative territorial entity (P131) = <some civil parish>. When that is not the case, it is useful to have the information directly on the item to make clear that it's not just that nobody has refined the P131 down to civil parish level, it is that there is no civil parish.
  • It's useful to use a qualifier for that role, rather than an independent statement, so that if a civil parish did get created for Teversal (Q7707125) (which can happen), then at least in respect of Teversal Manor (Q15979549) the information would all be in the one statement.
  • Yes, in theory one could write queries just as well to check the chain Teversal Manor (Q15979549) : location (P276) -> Teversal (Q7707125) : location (P276) -> unparished part of Ashfield (Q105909200). But in practice it's more accident-prone. It relies on two different statements in two different places both to be right, despite whatever independent editing might have been done to one or the other.
  • Finally, it feels odd to me to treat the "unparished parts" like a real geographical area, because for a number of authorities that 'unparished part' is not a single contiguous area, but rather a collection of several areas, each one a different 'hole' in the parish map. It seems odd to me to potentially have a collection of areas as the value for a property like location (P276), rather than a single specific compact area.
Jheald (talk) 20:45, 30 March 2021 (UTC)[reply]
  •  Comment I saw a case recently where this might apply. The form was ⟨historical letter⟩ ⟨published in⟩ ⟨edition of a book⟩. It could have been ⟨historical letter⟩ ⟨published in⟩ ⟨edition of a book⟩ / ⟨specific part⟩ "Appendix 2". But that would require a string value instead of an item value. Pelagic (talk) 11:16, 18 April 2021 (UTC)[reply]
One way can try to represent this (see current Q77607750#P1365) is to write County of the City of Glasgow (Q77607750)replaces (P1365)Lanarkshire (Q530296) with qualifier applies to part (P518) = <somevalue>.
But a machine finding such a statement doesn't really know how to interpret it. Does the statement mean that part of the County of the City of Glasgow replaced all of the county of Lanarkshire? Or that part of the County of the City of Glasgow replaced part of the county of Lanarkshire? Or (?correctly) that all of the County of the City of Glasgow replaced part of the county of Lanarkshire?
(also: how would one represent that middle statement, if were the case that part of the County of the City of Glasgow replaced part of the county of Renfrewshire?)
This is why it makes sense to have two clearly distinct separate properties: one to indicate part of the subject of the statement; and a different one, if one wants to indicate part of the object of the statement.
That is what this present proposal aims to establish. Jheald (talk) 17:56, 20 April 2021 (UTC)[reply]

Thank you, that clarifies it a lot. But I think the underlying problem in the examples you give is that you don't have a value to assign applies to part (P518) to. If you had an item for the part of Lanarkshire that was moved, you could assign that as the value of the applies to part (P518) qualifier an it would be clear, without the need for this proposed property. For now, I {oppose}, but I'm glad to change my mind. Could you give an example where there is a value for applies to part (P518) but it is still ambiguous and requires this proposed property? JesseW (talk) 18:39, 20 April 2021 (UTC)[reply]

@JesseW: You can kind-of see that in the Holyhead (OSD 317) (Q106156311) example above. Does "northwest" refer to the part of the northwest of the drawing, or the northwest of the printed map? The answer is in fact the northwest of the printed map, in which the whole of drawing 317 was incorporated. (Each single map sheet combined information from multiple drawings, each of smaller areas within it). Jheald (talk) 20:08, 20 April 2021 (UTC)[reply]
OK, thanks for pointing that out (I missed it earlier). I've removed my oppose -- I'm still not sure about this, but at least I understand it now. JesseW (talk) 13:32, 21 April 2021 (UTC)[reply]
Retrieved from "https://www.wikidata.org/w/index.php?title=Wikidata:Property_proposal/specific_part&oldid=1554471433"