Wikidata:Property proposal/Study for
Study for[edit]
Originally proposed at Wikidata:Property proposal/Creative work
Description | MISSING |
---|---|
Data type | Item |
Domain | work (Q386724) |
Allowed values | work (Q386724) |
Example | Supper at Emmaus (Q29211543) - study for - Disciples at Emmaus (Q29211544) |
Planned use | potentially Commons:Template:Study for |
- Motivation
Quite a few items are about artworks that are sketches or studies for another one.
Alternative solutions:
- Link from the finished painting to the study using based on (P144) but that sounds vague, and a bit odd, does not mean we don't need another property to show the reciprocal relation. Zolo (talk) 12:12, 9 April 2017 (UTC)
- Use P31 with qualifiers, like in Supper at Emmaus (Q29211543)instance of (P31)study (Q2647254)
of (P642)Disciples at Emmaus (Q29211544). But qualifiers make things more complicated
Notified participants of WikiProject Visual arts
--Zolo (talk) 12:12, 9 April 2017 (UTC)
- Discussion
- Support So for one way we can use based on (P144), but we have nothing the other way around probably because we're afraid of the 1:n problem (a lot of works are based on just one work). This seems to be an excellent solution to limit the scope, but still able to make the link. Multichill (talk) 13:48, 9 April 2017 (UTC)
- Support Jane023 (talk) 18:19, 9 April 2017 (UTC)
- Support Support as very useful. - PKM (talk) 18:33, 9 April 2017 (UTC)
Support Great!Neutral --Marsupium (talk) 20:03, 9 April 2017 (UTC), 12:20, 9 May 2017 (UTC)- Support Seems useful! – Susanna Ånäs (Susannaanas) (talk) 06:10, 10 April 2017 (UTC)
- Oppose Hi folks, we should study this a bit more (pun not intended) and not be hasty about it.
- CCO pp22 "Relationship Type and Reciprocity" says "CCO recommends that relationships between entities be reciprocal" and then gives a great table
- CONA In Depth pp46 explains the various associative relations
- CONA Editorial Guidelines is most detailed and gives 23 symmetric relations and 19 inverse (reciprocal) relation pairs (some curious ones are cartoon, pastiche, counterproof).
I think we should decide an approach and then model it comprehensively. We should copy the CONA relations to a new page "Visual arts/Relations", and argue/decide and document the representation for each of those relations.
Now my personal opinion: we shouldn't add 23+19+19=61 props for these relations (some of these exist, eg "different from"). I think we should group them into semantically coherent groups (eg "predecessor-successor", but I'm not claiming this is a good name for it) and use one relation per group, qualifying with an item to elaborate the detail. I think the qualifier should be on the relation, not on "instance of", i.e. something like this. The benefit is that we can extend the CONA nomenclature of relations as needed, eg see the last bullet (which is not in CONA):
- Supper at Emmaus - successor - Disciples at Emmaus / role Study
- Disciples at Emmaus - predecessor - Supper at Emmaus / role Study
- Rodin's Gates of Hell in Paris - same genesis (same production) - Rodin's Gates of Hell in Zurich / role Cast
- Cup - same context (used together) - saucer / role Conjuncted with
- Roman marble statue - predecessor - Greek bronze original / role Copy
- Greek bronze original - successor - Roman marble statue / role Copy
- Andy Warhol's Colored Mona Lisa - predecessor - Leonardo's Mona Lisa / role Parody (I think it's this piece)
Hmmm, those roles don't sound entirely good when made symmetric... What do you think, maybe we should use the original detailed relation names --Vladimir Alexiev (talk) 13:42, 12 April 2017 (UTC)
- @Vladimir Alexiev: Interesting resourcse, but if we try to solve the whole issue, I feel like will end up in one of those endless and sometimes ultimately fruitless discussions. Still, a few thoughts:
- About symmetric properties
- We should probably avoid that 1 property in 1 way correspond to two properties in the other way. Like use "map of" and "study for" in one direction. And use "based on" as a symmetric for both.
- I am not convinced we really need to have the create all symmetric properties. I think we need "study for", but not sure for its symmetric "study is". More generally the properties that tend to be N:1 seem more usable than 1:N (at least it makes items easier to read).
- Subproperties vs qualfiers:
- Using subproperties documented with subproperty of (P1647) is less annoying to type, and also more human-readable than more general "predecessor" / "successor" + qualifiers~. I don't think we should shy away from creating new properties when needed. Actually there are not so many of what CC0 calls "works that are related as steps in the creation proces". And all those in CCO are also in the CONA.
Label | Symmetric | Property | Symmetric property | CONA | CCO | Note |
---|---|---|---|---|---|---|
preparatory for | based on | - | - | Y | Y | based on (P144) has a wide scope, not sure it includes this "based on" |
study for | study is | Y | Y | |||
prototype for | prototype is | Y | Y | |||
cartoon for | cartoon is | Y | ||||
model for | model is | Y | Y | |||
plan for | plan is | Y | Y | |||
original print | counterproof from | Y | Y | |||
printing plate for | printed from plate | Y | Y | |||
negative for | printed from negative | Y | ||||
- I would be fine with creatign a property for each --Zolo (talk) 12:01, 13 April 2017 (UTC)
@Zolo: "avoid that 1 property in 1 way correspond to two properties in the other way": please elaborate, I don't understand. "no symmetric": ok, agreed, but can you always tell 1:N or N:1? "create CCO props but not CONA props": not sure this is the good decision, but ok we can start with this. "based on": yes, that is a bad name in CONA. --Vladimir Alexiev (talk) 14:47, 13 April 2017 (UTC)
- @Vladimir Alexiev:
- "avoid that 1 property in 1 way correspond to two properties in the other way" I mean if we want a symmetric property to "study for", it should be "study in". It should not be, say, a property that is symmetric to both "study for" and "prototype for". It imply that we should rescope based on (P144).
- "no symmetric": we can't always tell 1:N or N:1, that why I wrote "tend". Generally speaking, the preparatory works are the N and the final work the 1. Said differently, I think we need properties linking from a preparatory work to its final work, but not so much the symmetric linking from the final work to the preparatory work.
- "create CCO props but not CONA props": I don't think I suggested that. That said, I feel we can safely create properties supported by both systems.--Zolo (talk) 15:33, 13 April 2017 (UTC)
- @Zolo: Sorry, that's how I interpreted the limited table above. CONA has plenty more --Vladimir Alexiev (talk) 05:29, 18 April 2017 (UTC)
- @Vladimir Alexiev: CONA has many more "associative relatonships", but I think the table shows all those linking a preparatory work to a finished work. --Zolo (talk) 15:13, 18 April 2017 (UTC)
- @Zolo: Sorry, that's how I interpreted the limited table above. CONA has plenty more --Vladimir Alexiev (talk) 05:29, 18 April 2017 (UTC)
- This property currently lacks a description. ChristianKl (talk) 12:59, 15 April 2017 (UTC)
- What prior art exists for this relationship? Are there existing ontologies that model it out there? ChristianKl (talk) 13:02, 15 April 2017 (UTC)
- @ChristianKl: The prior art is CONA and CCO, see links above --Vladimir Alexiev (talk) 05:29, 18 April 2017 (UTC)
- Oppose - the property does not have any description currently. The other properties mentioned should have their own proposal. − Pintoch (talk) 16:31, 13 December 2017 (UTC)
- Marking as Not done as this proposal looks abandoned. Please create one proposal for each property mentioned if you want to go forward with the more elaborate schema, or reopen this proposal if you think one property is enough. In the latter case, please provide a description in English at least. − Pintoch (talk) 15:53, 19 February 2018 (UTC)