Property talk:P195

From Wikidata
Jump to navigation Jump to search


art, museum or bibliographic collection the subject is part of
Descriptionart, museum or bibliographic collection the subject is part of
Representscollection (Q2668072)
Data typeItem
According to this template: mainly creative works, but also other items that are part of collections
According to statements in the property:
item of collections or exhibitions (Q18593264), product (Q2424752), set (Q36161), type specimen (Q51255340) and artificial physical object (Q8205328)
When possible, data should only be stored as statements
Allowed valuesitems of museums, collections, libraries, archives etc. (note: this should be moved to the property statements)
ExampleRosetta Stone (Q48584)British Museum (Q6373)
School of Athens (Q186953)Vatican Museums (Q182955)
The Starry Night (Q45585)Museum of Modern Art (Q188740)
Codex Sinaiticus (Q152962)British Library (Q23308)
Tracking: usageCategory:Pages using Wikidata property P195 (Q21037764)
See alsolocation (P276), series (P179), inventory number (P217), catalog (P972), owned by (P127)
Proposal discussionOriginally created without a formal discussion
Current uses431,717
[create] Create a translatable help page (preferably in English) for this property to be included here
Value type “collection (Q2668072), organization (Q43229), architectural structure (Q811979): This property should use items as value that contain property “instance of (P31)”. On these, the value for instance of (P31) should be an item that uses subclass of (P279) with value collection (Q2668072), organization (Q43229), architectural structure (Q811979) (or a subclass thereof). (Help)
Exceptions are possible as rare values may exist.
List of this constraint violations: Database reports/Constraint violations/P195#Value type Q2668072, Q43229, Q811979, SPARQL, SPARQL (new)
Type “item of collections or exhibitions (Q18593264), product (Q2424752), set (Q36161), type specimen (Q51255340), artificial physical object (Q8205328): element must contain property “instance of (P31)” with classes “item of collections or exhibitions (Q18593264), product (Q2424752), set (Q36161), type specimen (Q51255340), artificial physical object (Q8205328)” or their subclasses (defined using subclass of (P279)). (Help)
Exceptions are possible as rare values may exist.
List of this constraint violations: Database reports/Constraint violations/P195#Type Q18593264, Q2424752, Q36161, Q51255340, Q8205328, SPARQL, SPARQL (new)
Qualifiers “start time (P580), end time (P582), inventory number (P217), applies to part (P518): this property should be used only with the listed qualifiers. (Help)
Exceptions are possible as rare values may exist.
List of this constraint violations: Database reports/Constraint violations/P195#Allowed qualifiers, SPARQL, SPARQL (new)
if [item A] has this property (collection (P195)) linked to [item B],
then [item A] and [item B] have to coincide or coexist at some point of history. (Help)
Exceptions are possible as rare values may exist.
List of this constraint violations: Database reports/Constraint violations/P195#Contemporary, SPARQL (new)

Work in progress, figure out a common point in the tree -->

Pictogram voting comment.svg Missing inventory number
Items which have inventory number (P217) as a qualifier of collection (P195), but not as a statement (Help)
Violations query: SELECT DISTINCT ?item ?collection ?inventory WHERE { ?item p:P195 ?collectionstatement . ?collectionstatement ps:P195 ?collection . ?collectionstatement pq:P217 ?inventory . MINUS { ?item wdt:P217 [] } . } LIMIT 1000
List of this constraint violations: Database reports/Complex constraint violations/P195#Missing inventory number

Book collections too[edit]

Changed the description to include book collections too.--Micru (talk) 18:06, 22 August 2013 (UTC)

How about archival collections?[edit]

Why are archival collections not mentioned? (well, in the German translation they actually are - but is this generally accepted?) --Beat Estermann (talk) 10:20, 4 December 2014 (UTC)

this was my decision to use it also for archives and libraries. Usualy this property is to be used with inventory number (P217), so there was some need to change the definition of P217 also. --Giftzwerg 88 (talk) 12:38, 4 December 2014 (UTC)

Nesting collections?[edit]

I think there should be some guidance about nesting collections. - If an institution has collections c1, c2, and c3 - which structure should be preferred?

  • Inst has c0; c0 has c1, c2, c3
  • Inst has c1, c2, c3

In parallel, some thought may also be given to the nesting of institutions / departments within institutions. --Beat Estermann (talk) 10:25, 4 December 2014 (UTC)

It depends on the notability of c0. If c0 ist notable and has own articles and therefore an own item, you might link c1, c2, c3 to this item. Otherwise I consider it to be better to link c1, c2, c3 directly to the "top level". There is no official rule so you can choose a way of nesting, that suits best in your eyes. The British Museum has incorporated lots of collections, not all of them are notable. The nesting problem is on the level of collections. As soon as it comes to single objects of the collection like a painting or a piece of pottery, it is better to use the top level with the inventory number and give three statements of collections, one for c1, one for c0 and one for Inst. --Giftzwerg 88 (talk) 11:50, 9 December 2014 (UTC)


How about item of collections or exhibitions (Q18593264)

  • {{Constraint:Value type|class=Q18593264|relation=instance}}

instead of entity (Q35120)

  • {{Constraint:Value type|class=Q35120|relation=instance}} ? --Giftzwerg 88 (talk) 16:27, 6 January 2015 (UTC)
I added entity (Q35120) as a temporary measure to see what's possible. We have about 30.000 work (Q386724) that use collection (P195). The targets should all be a subclass of some item (preferably not entity (Q35120)). It would be logical that everything is subclassed to collection (Q2668072), but maybe that feels a bit weird. Is a museum (Q33506) a subclass of collection (Q2668072)? Maybe museum (Q33506) -> public collection (Q2982955)? I don't think item of collections or exhibitions (Q18593264) is the way forward. Multichill (talk) 17:58, 6 January 2015 (UTC)
I am working on this field for a while. We have museums, libraries, galeries, collections, buildings, churches etc. that can be targets of P195. There are two items out there for collections: collection (Q2668072) and art collection (Q7328910) they are allmost redundant, but can not be merged. I use art collection (Q7328910) in preference, because it has more langlinks and I doubt the other languages make the same distinction as nlwiki does bertween collections and collections of pieces of art. (Half of collection (Q2668072) in nlwiki is however not about collections but about art collections). Then we have collections for art and we have collections for items that are not art in its narrow sense. Museums usualy dont collect books, but some do. Unfortunately museums can collect allmost anything from samples of dust, over cut of bodyparts up to buildings in an open-air museum. Libraries are usually not museums, but some are, however many libraries are collections not only of ordinary books, but of manuscripts and incunables. I use instance of (P31)+art collection (Q7328910) with libraries only if they include such valuable items that deserve owm articles, in all other cases it is enough to give it p31+library. Latey I found two articles about concept-cars. I was lucky to be able to assign item of collections or exhibitions (Q18593264) to the items instead of assigning a whole class of cars to any form of collection. It´d be very easy to add new classes of collection items to item of collections or exhibitions (Q18593264). I have in mind that all items of collections also have inventory number (P217) and vice versa. item of collections or exhibitions (Q18593264) served very well to handle the constraints on inventory number (P217), it was just created for this purpose. Sometimes we have pieces of art beeing part of the architecture like frescos, wall paintings, mosaics, altars etc., these of course have no inventory numbers, some items have more than one inventory number just to proof the exception of the rule and to drive us crazy. Museums are not allways public collection (Q2982955), as there are some private museums out there, open to public, but still private, as well as there is no exact border to private collection (Q768717) that may be never open to the public. I also found a nice solution for all kind of weapons showed in museums: militaria (Q598227), unfortunately only the pural is used, the singular would be "militarium", because these are definitely collectors items, while most of all weapons are not. --Giftzwerg 88 (talk) 21:35, 6 January 2015 (UTC)

Great, I was reverted by Swpb. Swpb, please adjust the constraints in such a way that the false positives at Wikidata:Database reports/Constraint violations/P195 disappear. I'm working cleaning up this and these false positive reports make it hard. I'm afraid that your edits made these lists even longer. Multichill (talk) 18:18, 2 February 2016 (UTC)

Attitude unnecessary. The false positives are because private collection (Q768717) is being used (perhaps questionably) as an instance of collection (Q2668072) ("Artwork X is part of [a particular] private collection"), but it was defined only as a subclass of collection (Q2668072). I've added the "instance of" statement on private collection (Q768717); private collections in general are a subclass of collections, but a particular private collection is an instance. (If you disagree with private collection (Q768717)instance of (P31)  collection (Q2668072), then you have to accept these as true violations, not false positives.) Simply adding private collection (Q768717) to the type constraint accomplishes nothing, since it is already a subclass of collection (Q2668072); that's why the constraint template says "or a subclass thereof". Swpb (talk) 19:21, 2 February 2016 (UTC)
Sorry Swpb, that was more me fighting the constraint reports than being unhappy about your edit. I don't think adding instance of is correct. Any ideas for a more elegant solution? Multichill (talk) 20:49, 3 February 2016 (UTC)
@Multichill: All the options I can see for getting rid of the violations, in increasing order of effort:
  1. Leave the statement private collection (Q768717)instance of (P31)  collection (Q2668072)
  2. Remove all offending statements <item>collection (P195)  private collection (Q768717)
  3. Propose a new property, "collection type", which takes subclasses, rather than instances, of collection (Q2668072), and change all statements <item>collection (P195)  private collection (Q768717) to <item>"collection type"  private collection (Q768717)
  4. Modify {{Constraint:Type}} to allow instances and subclasses. {{Constraint:Type}} currently can require subclasses instead of instances, but that would not be appropriate here, since most objects of collection (P195) are instances. This change would require broad community input, since it is a very widely used template.
I'm agnostic as to which solution to go with; #1 was merely the easiest. Swpb (talk) 14:28, 4 February 2016 (UTC)

Add both as standalone statement and as qualifier to inventory number (P217)?[edit]

I'm a little bit confused about this. Is the best practice to add collection (P195) both as a separate statement and as a qualifier to inventory number (P217)? Like so? Danmichaelo (talk) 16:56, 19 February 2017 (UTC)

I think that's fine. I'd be more likely to use inventory number (P217) as a qualifier of collection (P195), rather than the other way around, but either way works. --Swpb (talk) 15:36, 24 February 2017 (UTC)
I don't think the two uses are linked or combined in the way Swpb suggests.
--- Jura 11:19, 25 February 2017 (UTC)
@Danmichaelo: collection (P195) should always be added directly and as a qualifier to inventory number (P217). That's also how the contraints on this page and at Property talk:P217 are set up. About 266.000 items use it in this way.
If you only add collection (P195) with qualifier inventory number (P217) (like Swpb suggests), it will trigger a constraint at Wikidata:Database reports/Complex constraint violations/P195#Missing inventory number. Multichill (talk) 12:33, 25 February 2017 (UTC)

@Nurni, Multichill: You both added collection (P195)exception to constraint (P2303)  allowed qualifiers constraint (Q21510851) / property (P2306)inventory number (P217): I removed it the first time and I'd like to remove it again, but: Why did you add it? There is consensus since 2013 ;-P and still (see this talk page section above) not to use the properties this way. Did I miss something? --Marsupium (talk) 15:43, 17 March 2018 (UTC)

@Marsupium: like this edit? That would cause about 17.500 constraint violations:
SELECT ?item ?itemLabel ?collection ?collectionLabel ?inv WHERE {
  ?item p:P195 ?collectionstatement .
  ?collectionstatement ps:P195 ?collection .
  ?collectionstatement pq:P217 ?inv .
  SERVICE wikibase:label { bd:serviceParam wikibase:language "en" } .
LIMIT 25000
Try it!
I don't use it in this way, but nothing wrong with using it in this way. Just let it be, definitely no consensus to just remove these. Multichill (talk) 15:56, 17 March 2018 (UTC)
@Multichill: Like that edit, yes. Wrong with it is that the co-existence of two different schemata causes disadvantages for editing and querying the information (at/from two places). Is there any gain in storing it twice outweighing the disadvantages? I worry this constraint allowing it and existing cases encourage people to use the properties this way and increase problems that now are only small. Currently (only) 460 items with inventory number (P217) as qualifier to collection (P195), but not as mainsnak:
SELECT DISTINCT ?item ?itemLabel ?collection ?collectionLabel ?inv WHERE {
  ?item p:P195 ?collectionstatement .
  ?collectionstatement ps:P195 ?collection .
  ?collectionstatement pq:P217 ?inv .
  FILTER NOT EXISTS {?item p:P217 ?invstatement}
  SERVICE wikibase:label { bd:serviceParam wikibase:language "en" } .
LIMIT 25000
Try it! --Marsupium (talk) 16:12, 17 March 2018 (UTC), 16:22, 17 March 2018 (UTC)

I do not see why inventory number (P217) is necessary with collection (P195) and as Swpb suggests, it should definetly be connected, because double edits are not state of the art. And I do not have time to add information twice. I remember that not all collections show them e.g. Gemäldegalerie (Q165631) not all, Bode Museum (Q157825), Gallerie dell'Accademia (Q338330), which causes a lot of constraint violations without solutions. Uffizi (Q51252) have parallel different inventory systems which are for all museums in Florence. We have e.g. Florentine musea Inventario 1890 ID (P1726),Florentine musea catalogue ID (P2242) and more which should be sufficient after editing it once.Oursana (talk) 00:09, 24 March 2018 (UTC)

@Oursana: I think this edit was a good idea, removing collection (P195)property constraint (P2302)  item requires statement constraint (Q21503247) / property (P2306)inventory number (P217)! --Marsupium (talk) 11:08, 24 March 2018 (UTC)
ja, thank you, the constraint violations have to be changed as well, have a nice weekend --Oursana (talk) 11:18, 24 March 2018 (UTC)

Use for creators[edit]

Can we use this property also to add in which collections a creator has work in the collection? I come across photographers, graphic designers etc that have work in a museum or archive, and it would be great if we could mention in the persons' qid a list of museums that have work of that artist without the need to add all those individual photographs and designs. Could we broaden this property or should we create a new one? --Hannolans (talk) 10:28, 12 November 2017 (UTC)

I wouldn't do that. The list would explode. If you want to know that you can do a SPARQL query. Multichill (talk) 11:51, 12 November 2017 (UTC)
@Multichill: that could work for painters as we have paintings in Wikidata, but not with photographs and graphic designers, textile designers etc, or should we start to add photopgrahs, book covers, chairs, fabrics, plates, posters etc to wikidata? Only if we have that we can sparql which museum collection has work of an artist. Or is there another trick? --Hannolans (talk) 20:23, 1 December 2017 (UTC)
If we actually need these connections I think a separate property would be better. --Marsupium (talk) 10:14, 3 December 2017 (UTC)
Agreed, and that's a big if. Some creators would have hundreds of instances of such a property, with no means of separating the trivial from the important. Swpb (talk) 16:19, 4 December 2017 (UTC)
OK, I will propose another property. Yes, I'm aware that this has limits for some wellknown top artists. --Hannolans (talk) 21:35, 18 December 2017 (UTC)

"has cause" qualifier?[edit]


I sometimes use "has cause" as a qualifier for this item, with arguments such as "gift", "lease" etc. to indicate provenance. At the moment, this raises a warning with the constraints system, which only allows "start time", "end time", "inventory number" and "applies to part" in this place. Is there a better way to describe the provenance of an item of collection, or should "has cause" be listed among the properties allowed as qualifier?

Cheers! Rama (talk) 10:06, 29 June 2018 (UTC)

How do we describe provenance in general (i.e. the entire ownership history)? - Wouldn't that information fit in there?
Cheers, Beat Estermann (talk) 10:15, 29 June 2018 (UTC)
For full provenance you generally use these properties:
Based on that I'm not sure we need the has cause (P828) qualifier, but I'm not against it, it could be a nice link to the significant event. Multichill (talk) 10:34, 29 June 2018 (UTC)
If the has cause (P828) qualifier is used, should it point to the class of the event (e.g. "purchase") or to the actual event (e.g. a specific auction)? - It should first become clearer what exactly we are trying to model. The examples given by Rama could also be interpreted in terms of ownership status (an institution may be the holder of a collection, but is not necessarily the owner of all the collection items, as some items may just be on loan). --Beat Estermann (talk) 10:48, 29 June 2018 (UTC)