Wikidata:WikiProject Visual arts/Questions



Item structure








Examples of tricky items … Please use hard coded labels instead of {{Q}} and {{P}} since these templates break cross-language linking to section headings, thanks!

Mona Lisa (Q12418)Edit

  • !: depicts a woman idenfited by some as X and by some as Y--Marsupium (talk) 01:49, 23 September 2013 (UTC)   Done--Oursana (talk) 13:56, 15 October 2014 (UTC)

Belly Amphora by the Andokides Painter (Q540716)Edit

  • !: creator

Duenos Inscription (Q657989)Edit

  • !: the inscription is hard to decypher with parentheses reflecting uncertainties

Napoleon I on his Imperial Throne (Q1282840)Edit

  • !: two inventory numbers for the Louvre, one for the Musée de l'armée
    This shouldn't be a problem. We may simply state both of them. Additionally why should probably use qualifiers for the institutions like collection (P195) or owned by (P127). (Perhaps even the period of validity should be stated.) As well for The Night Watch (Q219831). --Marsupium (talk) 01:49, 23 September 2013 (UTC)

The Night Watch (Q219831)Edit

  • !: two inventory numbers for the Amsterdam Museum (owner), one for the Rijksmuseum (on permanent loan)

Pillar of the Boatmen (Q1394918)Edit

One object in four pieces, with a four different inventory numbers

I suppose that it may be the best solution to state the four inventory numbers with different qualifiers. However, there will be other cases like Parthenon Frieze (Q3087894) that will force us to create more items to describe one object consisting of more objects completely. I think this case will always occur if there is a greater number of statements that differ between the particular parts like the object history of the Parthenon Frieze. --Marsupium (talk) 01:49, 23 September 2013 (UTC)

Massacre of the Innocents by Rubens (Q1247103)Edit

One item for two artworks: At least the linked en, es, fr and la articles (I have not checked the others) cover two paintings. How shall we deal with this? I think the only way will be to create two separate items for the statements and to find a way to show their relationship to Massacre of the Innocents by Rubens (Q1247103). What do you think? --Marsupium (talk) 00:16, 8 November 2013 (UTC)

There are 2 Versions of this motif by Rubens, in Munic and in Ontario, + a 3rd Version of Rubens workshop in Brussels. I checked all wikilinks: az covers only the Munic version, th covers this motif generally, so these have to be dislinked, th forms one super cat. All the other interwikilinks refer to both paintings, so this combi-item has to be kept. I will create a matching commons cat. I changed the linked Property:P18 image from the Rubens workshop Brussels version to the Ontario version and will link my new upload of the Munic version. (User:Jean-Pol GRANDMONT changed this image and destroyed all 8 wikilinks which I reverted and repaired.)
These commons:Category:Motifs by painters (work in progress) generally or by painter are regularly used on commons, e.g.: Q1787460. The question remains how to deal with the diverging painting's data: Technique, Date, Institutions, property etc. There are also mixed motif cats on commons, which do not differ among the different paintings, e.g.:File:Categorization of Eleanor of Toledo with her son Giovanni de' Medici by Angelo Bronzino.jpg, here categorization has been changed meanwhile. This is going to be changed on wikicommons and mixed informations are not suitable on wikidata. I wonder, if we have to create separate items for e.g. the musea information or if we only link information we have items for. As the wikilinks to Massacre of the Innocents by Rubens (Q1247103) are about a combination of two paintings, in any case we must keep this combi-item of the two paintings on wikidata too. But in this example we must perhaps be prepared, that perhaps the wikilinks will also expand and include Version 3 of Rubens workshop in Brussels.--Oursana (talk) 13:36, 28 November 2013 (UTC)
I am pretty sure that we should have an item for each object. --Marsupium (talk) 08:11, 5 December 2013 (UTC)
What are the benefits from this? It is no need to do so. If you keep them together you can link to all wp language versions -az,th, and my new created commons cat, which has still to be filled completely. The link to the wp articles may not get lost. And what about the 3rd workshop copy version in Brussels? No third item I propose. And what will be the way to combine with Massacre of the Innocents by Rubens (Q1247103) ? This has to be solved before creating two items. I'd like to see more statements than from the two of us.--Oursana (talk) 00:10, 8 December 2013 (UTC)
An item per object seems necessary, because many things can differ between the two version (the museum, the date, the conservation history etc.) and it would be complicated to document them in the same item. What I do not know is how they should be linked. Perhaps keep the item groping both versions essentially empty except for Wikipedia links and use part of (P361) in items about individual versions. Or use the earliest version for Wikipedia links and use "influenced by" or a new property in items about posterior versions ? --Zolo (talk) 08:44, 8 December 2013 (UTC)
Thank you Zolo, I agree with your first proposal: keeping Massacre of the Innocents by Rubens (Q1247103) for Wikipedia and commons cat links, two new items for Munic and Ontario versions and a 3rd one for the workshop Brussels copy (!?) and use part of (P361) in items about individual versions. I do not like 1st (Ontario) version and all Wikipedia articles dealing with two versions in one item, and I am not happy with "influenced by" for the two original versions by Rubens, which differ completely. The 2nd (Munic version) and the az article originally within this item can be linked. The workshop copy in Brussels is almost a copy of the 2nd Munic version, so I would not use influenced by (P737) /P738 (P738), but something like copy? The th wikipedia link, meanwhile separated from this item Massacre of the Innocents by Rubens (Q1247103), must get a new item as icon class code subject of numerous sculptures and paintings and the other items above are part of (P361). --Oursana (talk) 03:08, 9 December 2013 (UTC)
  • However, I'd prefer another solution: I'd create an item for each artwork and select one to shift all interwikis which refer to more than one artwork to one of the one-artwork items since I think as many sitelinks as possible should go to one item for practical causes: Even different versions should be interlinked through Wikidata without using local interwikis for all "orphaned" painting versions. (With this I oppose my own first thought above.) I do not think that it will be possible or useful to create items for all possible Wikipedia article topics of ≥287 Wikipedia language versions. Wikidata should be object-oriented rather than Wikipedia article oriented (as it is in most parts ATM). I can explain this opinion more precisely, but have not much time now. The sitelinks should be considered with a semantic like the property that is used for example by DBpedia (Q465), cf. [1]. If we do not think about that it causes problems like this one: Property talk:P31#Inconsistency? I think that in most cases this will be the most simplest solution.
  • But if we will create items just for the Wikipedia articles about more than one artwork let us then think about making the item a class (perhaps collecting the properties the artworks have in common) with the one-artwork-items instance of (P31) of it rather than using part of (P361) (implicating that the grouping item is an instance, too.) And problems like this are not equal enough to use influenced by (P737) for all of them, the property is not equal enough. Thus I'd prefer the use of instance of (P31) relationships. Regards, --Marsupium (talk) 14:31, 9 December 2013 (UTC)
This is an icon class code that has been the subject of numerous sculptures and paintings, many of which will probably have an article added in some wiki in the future. I asked for icon class codes to be added to the artwork template, because this would be a great way to discover categories on Commons and link them to wikipedia articles. I think in this case both paintings should have their own data items and be linked to a Massacre of the Innocents article. To properly differentiate, icon class, year painted, and artist should be enough to separate them out. For cases when the attribution is uncertain, the current or last-known owner can be used to differentiate. Jane023 (talk) 23:54, 7 December 2013 (UTC)
Sorry Jane, I am not sure if you read the whole section. Massacre of the Innocents by Rubens (Q1247103) is consequently only about the two versions of this motif by Rubens (Alte Pinkothek Munic and Ontario). 8 articles about this two versions of Rubens are linked. I separated the th versions, which is about the motif/iconclass generally. If we make a cut within Massacre of the Innocents by Rubens (Q1247103) we destroy everything. What do you mean with should have their own data items and be linked to a Massacre of the Innocents article? We already have 8 articles, each dealing about the two Rubens versions together, these you do not want to link? Linking the articles about two items to two commons files (two items) and the articles are a third item?? Could you please explain exactly how many items which for you want to create and to link, this would be very helpful to communicate.
We also have to deal with az only about the Munic version of the Massacre by Rubens, and hasnot the quality of an article. As I stated above there is also a Rubens workshop copy in Brussels.
And very important we have to consider Notability-policy, which does not allow an item per commons cat/file. Notability is per article, wp-cat or commons gallery only (see RfD, I suppose you do not need the link?) And here the articles are about 2 versions, so the item must combine the two versions too. The notability forms the organization of the items with regard to existing articles etc., a theoretical structure not regarding articles has no notability on wikidata.
That I prefer to keep this extra item for the two Rubens versions, does not oppose a general icon class item as subject of numerous sculptures and paintings, perhaps with regard to the th article, e.g. including Giotto, Tintoretto etc..This is another general question not exactly dealing with the 2 Rubens versions, which should be instance/ part of this Icon class. We must deal with the two Rubens versions as separate item (instance/part of), because the Icon class Massacre of the Innocents is very big. And for this instance the two Rubens versions are enough and the rationale of their combination is that they form the artists versions of this motif. Otherwise I am afraid we will lose all the articles value and create a big mass with no benefit at all. --Oursana (talk) 01:34, 8 December 2013 (UTC)
  • I agree with Jane, an item for each artwork which is linked to a separate item for a motive linked to a code of Iconclass (Q1502787) sounds nice! But we cannot use the Iconclass item for the sitelinks for this example of course, that's even an other problem I have answered to above.
  • We would need a new property for the Iconclass codes as you have yet demanded here, Jane023, with the use Zolo has proposed there and similar to the way we are yet handling the object classification with Art & Architecture Thesaurus ID (P1014). I have already studied the Iconclass model a bit, but that may require a more precise look into the intricacies of Iconclass' system. I do not think it was useful to create an item e.g. for 11HH(GENEVIEVE)81(+2) – post-mortem occurrences ~ St. Genevieve - finding, transportation of relics of female saint (+ Mary), but we can use a code like that for a <has Iconclass code>-property of a single artwork item. So I think we will need a combination of the two possibilities Zolo has pointed out.
  • Oursana, notability shouldn't be a problem as I pointed out here: Even painting versions should be notable by points 2 and 3 of WD:N.
Regards, --Marsupium (talk) 14:31, 9 December 2013 (UTC)
Yes, actually, I agree that that it should be P31 rather than p361. I would propose that items about individual paintings belonging to such groups should be handled exactly like other paintings except that we should add "p31: the group". I mean, we should state 'depicts: massacre of the Innocents' on the item about the Alte Pinakothek version, not just on the item grouping both versions, so that if we make reckless generalization in the generic item, it does not affect items about individial paintings, and also because it sounds easier to read and to query. --Zolo (talk) 08:07, 10 December 2013 (UTC)
Yes, we should make the statements for all single artwork items leaving the grouping items (if we want to use them) without detailed information. Only a statement like subclass of (P279) <class of things covered together by a Wikipedia article> might be useful to keep track of those constructions. --Marsupium (talk) 15:44, 10 December 2013 (UTC)
  Done--Oursana (talk) 18:04, 30 December 2015 (UTC)

Christ Washing the Disciples' Feet (Q75846)Edit

And one has to consider, if this is not a general problem also with other paintings, where different versions of a painter's motif are described in one article, see Christ Washing the Disciples' Feet (Q75846) which is different.

en:deals with 4: (S Marcuola/Prado version√, S Marcuola/Shipley version√ from the 6 mentioned original versions (two are only mentioned) + two copies (S Marcuola/Shipley copy or Original version and San/Marcuola Toronto Copy, this motifs had all companion paintings of the last supper which also must be linked to each other; es, eus= Prado vers. So here en/es,eu must become two items, all have Prado Vers+ copies in common, but we cannot link to only one section. The commons cat is meant for all versions, but has until now only version (S Trovaso/ NG) and the sub cat for the S Marcuola/Prado version. It is a disputed if the Shipley or the Prado version is originally from S Marcuola, the latter is general opinion. So commons cat goes with en, commons:Category:Jesus Christ washing the feet of the apostles by Jacopo Tintoretto (San Marcuola) sub cat goes with es,eus. Will we take Massacre of the Innocents by Rubens (Q1247103) for en or es,eus, the rest I can do.

What do you think, shall we create a subpage for these kind of items? LG--Oursana (talk) 18:55, 5 December 2013 (UTC)

Napoleon Crossing the Alps (Q59201)Edit

same problem, 5 versions, actual 4, we have to separate.--Oursana (talk) 10:43, 8 January 2014 (UTC)


How do we say that a painting depicts a sitting girl with a blue dress and a bearded man, in such a way that we unambiguously know that the it is the man who is bearded and the woman who is sitting ? A qualifier ? Using qualifiers ? I would be fine with that, but it will make queries more complicated because every time we need to find a painting with a bearded man, we have to query qualifiers as well. And would we call such a qualifier ? The Joconde database wisely dodges the issue in its 'représenté' field. --Zolo (talk) 08:07, 10 December 2013 (UTC)

Qualifiers seem to be the best solution to me ATM. I have even another dream for tagging things in visual artworks connected to your question: I think it would be great if we found a way to annotate shapes in a file, if possible not only rectangles and if we connected the depicts (P180) statements to the shapes. Cf. my thoughts at Wikidata talk:WikiProject Visual arts/Item structure#Additional items for images? A happy New Year rolling out to all of you around the world now, cheers, --Marsupium (talk) 07:19, 31 December 2013 (UTC)

We have to find a way to describe several representations of the same thing, especially a person, in one image room(?) (in German „Bildraum“) (at different times). It does not come a better example to my mind at the moment but The Procession to Calvary (Q686107) which depicts Jesus and his cross two times. Unfortunately, the resolution of File:Pieter Bruegel d. Ä. 007.jpg makes it hard to recognize the second crucifixion scene in the background on the right. How shall we describe several depictions like there? --Marsupium (talk) 13:46, 19 January 2014 (UTC)

@Marsupium: Easy: mention the person twice with two different qualifiers --Vladimir Alexiev (talk) 03:29, 8 January 2015 (UTC)

Items for single object parts!?Edit

To denote the measures of this object: File:Klapphocker Daensen Abschlusskappe.jpg we should create a new item with a statement part of (P361) <Daensen folding chair (Q1744373)>, right? --Marsupium (talk) 07:06, 31 December 2013 (UTC)

That seems like the clearest solution, though p518 could arguably be used as well. --Zolo (talk) 10:55, 8 January 2014 (UTC)
But there must be a link from Daensen folding chair (Q1744373) to the detail. I would prefer Zolos solution with applies to part (P518) not to loose information.

several creator (P170) e.g. Brancacci Chapel (Q1996075)Edit

I would like to continue the question of several painters see Wikidata:Property proposal/Place#painter (en), Maler (de), peintre (fr). Here we even have the problems with 3 painters and one combination of two thereof. I added creator (P170) an do not know how to add applies to part (P518), Zolos suggestion, but here not fresco but part of the frescos, the parts have own categories. The solution cannot be too create several items brancacci/artist. I will be happy for help and suggestions.--Oursana (talk) 13:46, 13 January 2014 (UTC)

In this case, I think creating an item for the frescos, or even an item by fresco ok. But I guess we could also use p518 with something like fresco painting (Q134194), left (Q14565198)
Thank you Zolo. As I mentioned above I am afraid separating is no good solution. We have 12 frescoes on 2 walls and two rows telling a story painted by three painters +fourth painter combination. I started to add creator (P170) + Masolino and do not know /cannot add applies to part (P518). Could you go on to show me. Regards--Oursana (talk) 23:52, 13 January 2014 (UTC)

How to link Felice Brancacci (Q3742116) as commissioner of Brancacci Chapel (Q1996075)?Edit

How can we connect this commissioner of the frescos to Brancacci Chapel (Q1996075)--Oursana (talk) 13:51, 13 January 2014 (UTC)

You mean what should we add in , I do not think we should add anything, just like for creator (P170) (we dont add all paintings by Picasso to the item about Picasso). It can be retrieved through . On Wiki, I am not sure about what we can do. It will be retrievable through query, but all individual queries will have to be hardcoded if we want to use them in Wikidata, so that is not so convenient. --Zolo (talk) 18:56, 13 January 2014 (UTC)
I doubt because with Picasso we state his work, at least the important work. But Felice Brancacci is only famous for having commissioned the frescos of Brancacci Chapel (Q1996075). If we cannot add this, we do not need any further information about him.--Oursana (talk) 23:52, 13 January 2014 (UTC)
I think having several items is useful here, otherwise, it is a bit difficult to describe what is depicted on the frescoes, the dimension of each of of them etc. I have added a p518, but in this case that does not work very well, because there are rather many different frescoes. If we create items for each different fresco, it works better (creator: Masaccio: qualifier:p518: The Expulsion from the Garden of Eden etc.)--Zolo (talk) 18:38, 14 January 2014 (UTC)
Thank you Zolo. I cannot put (creator: Masolino: qualifier:p518: Temptation of Adam and Eve. Do I have to create an item for Temptation of Adam and Eve by Masolino first or do you want an extra item Temptation of Adam and Eve by Masolino? Regards--Oursana (talk) 20:41, 17 January 2014 (UTC)
Sorry, I do not understand your question. Yes, you need a new item, but what is the difference between "an item for Temptation of Adam and Eve by Masolino" and "an extra item Temptation of Adam and Eve by Masolino" ? --Zolo (talk) 08:12, 18 January 2014 (UTC)
1st I must tell you that you created some double items, which I deleted. Do we really need right (Q14565199), left and right (Q542952), right (Q15272410) and left (Q15272420)? The last two do not belong to relative direction (Q14565197).
Sorry for my troubling question, by creating the item I gradually understood. Perhaps I will slowly try to continue. Please keep an eye on it. Shall I link Brancacci Chapel (Q1996075) to every fresco? --Oursana (talk) 01:28, 22 January 2014 (UTC)

Scope of this pageEdit

I would prefer if we used this page only to talk about the application of our rules to special cases (items). Let us use new sections at Wikidata talk:WikiProject Visual arts/Item structure for all problems which arise in general and in several cases to keep some more clearity (especially for non-project-members)! Most of the current sections of this page belong there. We might even consider to move them!? Regards, --Marsupium (talk) 13:52, 19 January 2014 (UTC)

Castelfranco Madonna (Q1720631), significant event (P793), theft (Q2727213)Edit

How can we give the date?--Oursana (talk) 01:38, 22 January 2014 (UTC)

With a qualifier : click on modify, click on add a qualifier and choose the property date GautierPoupeau (talk) 07:46, 23 January 2014 (UTC)
Thank you. --Oursana (talk) 02:06, 26 February 2014 (UTC)

cycle of paintingsEdit

should we expand painting (Q3305213) also for cycle of paintings or should we create a new item to express cycle of paintings e.g. for Legend of Saint Ursula (Q605725), where only painting (Q3305213) is used incorrectly.--Oursana (talk) 02:06, 26 February 2014 (UTC)

It is certainly quite a bad idea to claim and similar things. Probably, ShonagonBot claimed this because of Category:Paintings by Carpaccio (Q9135007) in en:Legend of Saint Ursula.
Ideally single items should be created for the paintings with claims like . Claims like
⟨ Legend of Saint Ursula (Q605725)      ⟩ instance of (P31)   ⟨ cycle of paintings ⟩
bring the disadvantage with them that they would force us to maintain a separate item tree with items like <cycle of artworks>, <cycle of paintings>, <cycle of panel paintings> etc. However, I prefer this compared to the current solution in Legend of Saint Ursula (Q605725). Please go on and change it! ;) LG, --Marsupium (talk) 16:43, 8 March 2014 (UTC)

I created cycle of paintings (Q16905563) and cycle of frescos (Q16905550), see [2]--Oursana (talk) 01:34, 19 May 2014 (UTC)

The Farewell of Telemachus and Eucharis (Q3230385)Edit

The depicted episode is inspired by Fenelon's Telemachus, but much of it is David's invention [3], p46-50. Can that be expressed through based on (P144) ? --Zolo (talk) 09:14, 13 March 2014 (UTC)

IMHO you better take inspired by (P941), see example Property talk:P144. We should add this to Wikidata:WikiProject Visual arts/Item structure.--Oursana (talk) 11:06, 13 March 2014 (UTC)
Yes, that's already better, but any idea how we can make it really clear that it is a modified version of Fénelon's story ? --Zolo (talk) 14:42, 13 March 2014 (UTC)
you should in any case add your reference url described at URL (P973), or is this different used?--Oursana (talk) 19:42, 3 April 2014 (UTC)
If you mean adding the URL in the source, that should be reference URL (P854).
described at URL (P973) means: webpages about the item (sort of = external links). In this case, I do not think we should use it, because the resource is a book, and can have its own item (accessing it through the Google Books URL is just one solution among many). There is an item about an edition of the book (The Farewell of Telemachus and Eucharis, 1st edition (Q15933427)) and it already says: main subject (P921): The Farewell of Telemachus and Eucharis (Q3230385). Arguably, we could also have the reverse item-type property, something like The Farewell of Telemachus and Eucharis (Q3230385): "P:Works about the item": (The Farewell of Telemachus and Eucharis, 1st edition (Q15933427) (not quite sure it would be a good idea though, because for some items the list would be impossibly long). -Zolo (talk) 17:06, 4 April 2014 (UTC)

Angelica After Ingres (Q15975385) (modelling uncertainty)Edit

Mentionning it here, although it is a much broader issue relating to the fact that there does not appear to be any good way to express alternatives in Wikidata. It is uncertaine whether Angelica After Ingres (Q15975385) is drawn after Ruggiero Rescuing Angelica (Q3439636) or from another drawing. How do we say that ? --Zolo (talk) 06:18, 23 March 2014 (UTC)

Properties for motif items like Massacre of the Innocents (Q15676570) or Adoration of the Magi (Q488841): subject heading (P921), depicts (P180), Iconclass notation (P1256)/depicts Iconclass notation (P1257)Edit

still busy with Massacre of the Innocents I created this item, linking only to wp-th, dealing with this motif massacre of the Innocents (Q643474) in art.

The problem is, I cannot use main subject (P921), depicts (P180), based on (P144) linking to massacre of the Innocents (Q643474) without constraint violation when I deal with this item as motif and not a work. We typically have articles about different versions, see above. I was trying to link with aforesaid properties, but now I found out about causing constraint violations.
One solution is wether we can change this for motifs in art.
On the other hand I would have a problem stating for articles dealing with several artworks instance of (P31) painting (Q3305213) (similar problems with cycles see above). This will be confusing the motif and the certain artworks.
So how do we treat items for certain motifs in art to which we link the precise art works. Can the icon class properties help here? But we also have do deal with a qualifier for instance of (P31).
Solution will be instance of (P31) motif (Q1229071) or even better with Iconclass notation (P1256), depicts Iconclass notation (P1257) and changing/permitting constraint for using motif/icon class theme. Different from Rijksmuseum, who only have to classify artwork,in wp we are dealing mostly with items/articles dealing with icongraphic themes (/by painter). Two more examples:
Adoration of the Magi (Q488841): (de:Anbetung der Könige: bekanntes Motiv in der bildenden Kunst); (en:Adoration of the Magi: is the name traditionally given to the subject in the Nativity of Jesus in art in which the three Magi); (fr:Adoration des mages: thème iconographique).
The Bible text is linked via part of (P361), which makes no sense when the article is not about the narrative but the motif in art. based on (P144) would be better but one has to make changes to avoid constraint violation.
Virgin of the rocks (Q269342), which has to be changed to show the Louvre and National Gallery versions like in the linked articles --Oursana (talk) 13:37, 23 April 2014 (UTC)

You are surely right that stating Massacre of the Innocents (Q15676570) instance of (P31) painting (Q3305213) would be a quite bad idea and lead to awful mess. I fear that the question is not that easy.
  • A possibility might be to use classes, in this case a class item "works depicting the Massacre of the Innocents". The class items can then get depicts (P180)- and based on (P144)- and depicts Iconclass notation (P1257)-statements (perhaps the property constraint rules have to be changed slightly to facilitate this, but that should not be a problem). We could simply make "works depicting foo" class items out of the current "foo" motif items. I would prefer this solution if we deal with only a few dozen items.
Wikidata Query currently gives 40 items for a search for CLAIM[31:2160811] OR CLAIM[31:1229071] and not all of them are really motifs, but there a probably several motif items which do not yet have a instance of (P31) artistic theme (Q2160811)- or instance of (P31) motif (Q1229071)-statement.
  • But if there are more such motif items we should perhaps allow them as a domain for some of the properties you mentioned.
What do you think of those two approaches? --Marsupium (talk) 11:34, 20 July 2014 (UTC)
Hi Marsupium, I do not get right your second proposal: domain (?). --Oursana (talk) 08:57, 7 September 2014 (UTC)
The domain is the class of items the property is allowed on (cf. LG, --Marsupium (talk) 15:34, 11 October 2014 (UTC)

group of works of same icon classEdit

I think Virgin of the rocks (Q269342) should be treated as a group of works like Zwei Reiter am Strand (Q15132660) rather than a motif. --Marsupium (talk) 11:34, 20 July 2014 (UTC) (Yesterday I changed the instance of (P31)-claim to Zwei Reiter am Strand (Q15132660) instance of (P31) cycle of paintings (Q16905563), however it is not really good. We should better use a "group of works" item rather than cycle of paintings (Q16905563). You might change this if you agree.)

I will start with this property, will you please support. I think we can handle also the above questioned motif items by this way.--Oursana (talk) 09:15, 7 September 2014 (UTC)

For something like a class of icons highlighting a particular religious story, wouldn't main subject (P921) be more appropriate and versatile? Among other things, this wouldn't limit it to just a group of paintings, but could include all kinds of works.--Pharos (talk) 20:07, 5 January 2017 (UTC)

Francesco Ubertini /Bacchiacca Francesco Bacchiacca (Q7770), catalog code (P528)Edit

Could you please have a look here. I do not agree with the use of catalog code (P528) together with Web Gallery of Art Web Gallery of Art (Q1464063). It is nor a number nor a catalogue and is often used by this user.--Oursana (talk) 12:33, 13 September 2014 (UTC)

Actually it is a catalog, Oursana, and the code is the name used by the Web Gallery of Art (right or wrong, including some doubles). What is your objection specifically to this catalog? Jane023 (talk) 10:36, 14 October 2014 (UTC)
Can Web Gallery of Art ID be added as another property like Joconde or RKD? - PKM (talk) 21:15, 8 October 2014 (UTC)
I suppose it could be, but I think it is better to include the WGA object title and WGA url. So for example for a painting, just include the catalog code in the same way as it has been done for the painter, but use the title for the catalog name and use property reference URL (P854) to include the WGA url as a reference. Jane023 (talk) 10:36, 14 October 2014 (UTC)
Sorry, for me it is still strange to add the name as a code--Oursana (talk) 01:19, 25 November 2014 (UTC)

Sorry for the delay in response. You can create an alternate name for the property - what would you like to call it? "name according to the catalog" or "catalog listing" ? Jane023 (talk) 11:46, 8 December 2014 (UTC)

Yes, that helps, perhaps catalog listing. I still hesitate to use this like catalogue raisonnée--Oursana (talk) 12:59, 10 December 2014 (UTC)

Armada Portrait (Q4792503)Edit

There are three well-known versions of this portrait - how do we reflect this? Also, the National Portrait Gallery is now rejecting the attribution to George Gower. How do we indicate a former and "current" attribution with dates and sources? - PKM (talk) 21:13, 8 October 2014 (UTC)

I have the same question. Jane023 (talk) 10:38, 14 October 2014 (UTC)
Would the new manifestation of (P1557) property be appropriate, with a somewhat abstract item like Virgin of the rocks (Q269342) for the class to which the pictures all belong?
As for the attribution, can the standard "preferred" and "deprecated" attributes for values attached to a property handle it ? Jheald (talk) 14:05, 14 October 2014 (UTC)
Yes I suppose preferred and deprecated would work for me. Very often what one museum says does not match with what a cataloger writes in a catalog raisonné, so having dates for attribution is not as useful as having attributions at time of sale, museum database, or per catalog per cataloger. As for the manifestation property, I am still not so sure. Jane023 (talk) 09:09, 15 October 2014 (UTC)
I am not sure the Armada Portrait is a manifestation of either the Spanish Armada or of Elizabeth I, though it might be a manifestation of the portraiture of Elizabeth I - but that doesn't let us group the a Armada portraits separately from the others. Would cycle of paintings work here? It's not a sequential cycle, but rather an original and its copies and variants. - PKM (talk) 17:52, 30 November 2014 (UTC)
I just created group of paintings (Q18573970), see Massacre of the Innocents by Rubens (Q1247103)--Oursana (talk) 01:13, 1 December 2014 (UTC)
That should work, but depicts: Massacre of the Innocents should be removed from the generic item, right? - PKM (talk) 21:58, 7 December 2014 (UTC)
Done Jane023 (talk) 12:30, 10 December 2014 (UTC)
I edited group of paintings (Q18573970) to be a subclass of group of works (Q17489659) rather than of just 'group'. - PKM (talk) 17:53, 8 December 2014 (UTC)

──────────────────────────────────────────────────────────────────────────────────────────────────── Along these same lines, I have created tapestry series (Q18609875) and group of tapestries (Q18609861). - PKM (talk) 18:19, 8 December 2014 (UTC)

You should move Wikidata:WikiProject Visual arts/Item structure#Series or Cycle? here--Oursana (talk) 11:17, 10 December 2014 (UTC)

Series or Cycle?Edit

(Copied from Wikidata:WikiProject Visual arts/Item structure#Series or Cycle?) I just stumbled across painting series (Q15727816) which seems to mean the same thing as cycle of paintings (Q16905563). There are more than 100 items tagged as painting series (Q15727816). How shall we handle this? - PKM (talk) 01:40, 9 December 2014 (UTC)

I'd think "cycle" is a series that has some coherent order like in "narrative cycle". So that would be a subclass of series. Marie de' Medici cycle (Q2558141) would a cycle because it tells a story with a beginning and an end. Rouen Cathedral Series (Q637414) does not have such a clear order, and even though it is a series, it is not a cycle. --Zolo (talk) 21:32, 9 December 2014 (UTC)
Can we move this discussion to site Questions as at least similar questions are discussed there--Oursana (talk) 11:10, 10 December 2014 (UTC)

Cranach's Martyrdom of Saint Catherine in BudapestEdit

I am trying to set up a Wikidata item for the institution we have identified as c:Institution:Collection of the Reformed Church, Budapest, noted as location "Ráday Library". A VIAF search is giving me three options, and no amount of googling seems to tie this specific painting to one of these specific institutions. This seems to be another instance of WGA vague specifications of the sources. Does anyone have more info on the ownership of this painting? - PKM (talk) 21:54, 7 December 2014 (UTC) (my antivirus avast blocks the english version), Cranach Digital Archive is a reliable source: Ráday Library of the Reformed Church, Budapest, [4], Raday Collections of the Danubian District of the Hungarian Reformed Church, Budapest, hu:Ráday utca… ;Ráday Collection at Budapest including bible Museum; p93,; The Ráday Collections of the Danubian District of The Hungarian Reformed Church contains three departments: the Library, Archive and Museum--Oursana (talk) 00:28, 8 December 2014 (UTC)
Thank you, @Oursana:. That's very helpful. This is the VIAF. I'll set that up now. - PKM (talk) 02:48, 8 December 2014 (UTC)

  Done - PKM (talk) 01:22, 9 December 2014 (UTC)

Cenacolo di Fuligno Q2944202Edit

Cenacolo di Fuligno (Q2944202) has been tagged as an instance of a religious painting depicting Jesus Christ, but more correctly it is a museum in a former refectory which houses Perugino's Last Supper. I think the painting and the museum are conflated here. Does anyone know if the phrase "Cenacolo di Fuligno" is correctly used to describe the painting? The linked articles in FR and IT Wikipedias clearly mean the museum. - PKM (talk) 03:10, 13 December 2014 (UTC)

The fresco is Last Supper (Q3208044). I tried to fix it. So you can see the different names. Happy New Year for you and everybody in this project.--Oursana (talk) 04:45, 8 January 2015 (UTC)
Looks good! Thank you. - PKM (talk) 17:45, 8 January 2015 (UTC)

  Done ---Oursana (talk) 21:58, 8 January 2015 (UTC)

Museums at Chateau de BloisEdit

Looking at this VIAF record, it seems to that C:Institution:Musée du Château, Blois and C:Institution:Musée des beaux-arts de Blois and their associated categories should be combined and connected to a single Wikidata item that is 'part of' Chateau de Blois, which has a different VIAF id. Is anyone familiar with this museum? If we agree here I'll start a conversation on Commons Village Pump before proceeding. - PKM (talk) 03:20, 20 December 2014 (UTC)

Mapping creators to their biographiesEdit

I am looking for pointers regarding mapping creators of any country to their biographies on Wikisource in any language. As an example of what I mean, please see how I linked the Wikisource entry for Vasari's biography of Properzia de' Rossi (Q293106) (Properzia de' Rossi) to her item. This is the definitive first biography of her published, but of course not the only one. Any ideas how to make use of original-language biographies now by dictionary-of-biography author on Wikisources (multiple languages, not just Italian)? Thx Jane023 (talk) 08:51, 8 April 2015 (UTC)

Hi @Jane023: (also pinging @Charles Matthews:) -- cf Wikidata_talk:WikiProject_UK_and_Ireland#DNB00: please don't sitelink a biographical article on Wikisource to an item about a person. Instead, the advice from Wikisource is to create a specific Q-item for the article, and then use eg instance of (P31) -> biographical article (Q19389637) and something like main subject (P921) to link that to the biographical subject. That way we can handle multiple biographies for the same person; and also record properties about those biographies (such as part of (P361), author (P50) etc) in a structured way. Jheald (talk) 10:30, 8 April 2015 (UTC)
That's right. A recent "protocol" is based on the "described by source" property to link from the person's item to a dedicated item for the source: which may be an item that links to the actual text on a Wikisource. As James says, the "main subject" property should be used to link back. This seems good librarianship to me, in that the page for the source can collect up metadata on that source. For "article-like" sources, I believe in using the "part of" property to link to the item on the main work. Where the version on Wikisource is divided up as one acticle per page, this gives what I consider an ideal result. I've done Reynolds, Joshua (DNB00) (Q19097329) to demonstrate. Charles Matthews (talk) 10:44, 8 April 2015 (UTC)
Thanks for the quick replies! The Reynolds solution does look like what I meant. Now I am looking at it though, I think we need something in between. I was thinking because in the case of creators (like Properzia and Reynolds) they should probably have their own "author" pages on Wikisource where all Wikisource articles about them can be aggregated. Otherwise, you will get tons of these flooding into Wikisource for all languages that eg. Vasari has been translated into, no? Jane023 (talk) 10:57, 8 April 2015 (UTC)
@Micru: What's the opinion of the Wikisourcers on this? Jane023 (talk) 12:55, 8 April 2015 (UTC)
@Jane023: Sometimes the author pages have a section called "Works about [author]", but that is of course depends on the case. For most creators I wouldn't create an author page just to say that there is something written about them, too much overhead, but for some of them might be worth it. It is up to you to decide, I guess.--Micru (talk) 15:47, 8 April 2015 (UTC)

Maybe we need something more than an author page - something that aggregates the material locked in the language-silo versions of Wikisource. I want to be able to easily link to original-language versions of important biographies. I guess a painter is an author, just not of the written word. That's why they have creator pages on Commons. Maybe it should be something on Commons? Jane023 (talk) 15:58, 8 April 2015 (UTC)

@Jane023: If you look at the Reasonator page for Joshua Reynolds [5], you can see that scrolling down to the "From related items" part, (ie not the described by source (P1343) property, but further down) this section has automatically gathered in that he is the main subject of Reynolds, Joshua (DNB00) (Q19097329); mousing over that property produces a pop-up with more information about the biography, with a link to wikisource. Is this then not doing exactly what you are looking for?
For Commons, one could create a gadget that automatically reproduced this behaviour for any creator page -- ie saw what items were linked to the subject of the creator page by a main subject (P921), and listed them, with more info available on mouseover. In principle, the code would be very similar to the gadget TheDJ made for me, that on any Commons category page looks up to see if the page is the target of a Commons category (P373) property, and if so adds a box linking to the relevant Reasonator page -- something I've been finding very useful. Jheald (talk) 18:07, 8 April 2015 (UTC)

Sculptures by CanovaEdit

Referring to this thread, I'd want ask you two questions:

  1. Is instance of (P31) sculptural group (Q2334148) a correct standard?
  2. How should I indicate the period during which a sculpture was sculpted?

Jane023 (talk) 08:50, 30 May 2013 (UTC)
User:Vincent Steenberg
Marsupium (talk) 13:46, 18 October 2013 (UTC)
GautierPoupeau (talk) 16:55, 9 January 2014 (UTC)
Multichill (talk) 19:13, 8 July 2014 (UTC)
Susannaanas (talk) 11:32, 12 August 2014 (UTC) I want to synchronize the handling of maps with this initiative
Mushroom (talk) 00:10, 24 August 2014 (UTC)
Jheald (talk) 17:09, 9 September 2014 (UTC)
Spinster (talk) 15:16, 12 September 2014 (UTC)
PKM (talk) 21:16, 8 October 2014 (UTC)
Vladimir Alexiev (talk) 17:12, 7 January 2015‎ (UTC)
Sic19 (talk) 21:12, 19 February 2016 (UTC)
Wittylama (talk) 13:13, 22 February 2017 (UTC)
Armineaghayan (talk) 08:40, 10 March 2017 (UTC)
Musedata102 (talk) 20:27, 26 November 2019 (UTC) Hannolans (talk) 18:36, 16 April 2017 (UTC)
Zeroth (talk) 02:21, 4 June 2018 (UTC)
Infopetal (talk) 17:54, 9 August 2019 (UTC)
S.ann.adams (talk) 21:05, 13 December 2019 (UTC)
Karinanw (talk) 16:38, 24 March 2020‎ (UTC)
Ahc84 (talk) 17:38, 26 August 2020 (UTC)
  Notified participants of WikiProject Visual arts Thank you, --Epìdosis 12:31, 15 October 2016 (UTC)

@Epìdosis: I'm sorry that nobody has responded here yet … better late than never …
  1. >5k uses does not look too bad :-)
  2. I don't know of anything better than something with significant event (P793) like significant event (P793) creation (Q11398090) / start time (P580) start time; / end time (P582) end time .
Cheers, --Marsupium (talk) 16:48, 30 December 2017 (UTC)

Sorting artists named John BradleyEdit

There are three artists named John Bradley active in the 1830s and 1840s and even ArtUK has two of them confused. This fascinating discussion (login required) indicates that ArtUK has misidentified their works by John Bradley as John Bradley of Pall Mall (1786-1843) when they should be John Bradley (of Keighley) (1787-1844) [6]. Neither of these is the John Bradley who was an itinerant American painter in the 1830s and 1840s.

WD has:

  • John Bradley (Q21289252) (1786-1843), the artist and printmaker resident in Pall Mall with VIAF and ULAN records
  • John Bradley (Q21460529) "British painter (1786-1843)" with the ArtUK artist ID. This is apparently JB of Keighley, but with incorrect dates that ArtUK will eventually change.
  • John Bradley (Q21543222) "English artist" with RKD identifier only; we should figure out which JB is meant and merge with one of the others.
  • John Bradley (Q16197540) The American painter; this record seems to be fine

I'm not sure how to sort this out until such time as ArtUK changes their dates. I suppose we could add the other dates from Oxford Reference to John Bradley (Q21460529), and add "from Keighley, Yorkshire" in the label instead of dates? Thoughts? - PKM (talk) 00:16, 9 February 2017 (UTC)

Fascinating thread.
I've gone ahead and added the places and dates of birth and death from that thread, keeping but deprecating the dates Art UK have for JB of Keighley (it's useful to indicate that there is an error out there in the wild).
Now to try to also work out which of the other identifiers apply to which JB. The print of Henry VII in the NPG is presumably JB of Pall Mall; and the print of Mrs Dickons in the NLI catalogue says as much (the first time I've ever followed a link in that cat and found it to be useful!). The ULAN ID is definitely for the printmaker. Which leaves the RKD one -- which I think needs them to be contacted, to decide who they want their entry to be for. Pinging @RKDdata, Multichill:. -- Jheald (talk) 10:20, 9 February 2017 (UTC)
Thank you, @Jheald:. I'll go back to working with them on Commons shortly, and I should have enough references to make at least a stub on JB of Keighley for EN wiki. - PKM (talk) 01:29, 10 February 2017 (UTC)

work with multiple executions (Q28886448)Edit

How do we deal with prints, sculptures?, e.g.The Great Wave of Kangawa (Q252485)

is a print and therefore has many versions. I would appreciate in this case not to retrieve to the commons file the wikidata acc number automatically, as it is wrong by version. So we better do not put acc nr to Wikidata. For the important versions and the reproductions we should create separate items and link it to the general one to which are linke the wps. Some prints like this have hundreds of versions.

So we have do deal like we do with painting series.

The main, wp linked item may have only claims valid for all executions of the artwork, so not different museums, especially not different accession number per Wikidata file with view that commons files get the accession number retrieved from wikidata this is a mass. Instead the artwork executions have to be linked with part of and opposite.

By the way we did not decide about retrieving the Wikidata accession number, as in these cases it causes problems.

I found The Age of Bronze (Q526178), The Age of Bronze (Q21647332)

We have now work with multiple executions (Q28886448), group of casts (Q28890616), Q28912972, single execution of an artwork (Q28912688)

One problem remains, I would like to use for title tl:label, this is not possible for the separated items. Perhaps we must create special technical solutions in this cases --Oursana (talk) 00:49, 8 March 2017 (UTC) --Oursana (talk) 17:30, 7 March 2017 (UTC)

The Great Wave off KanagawaEdit

(copy from my c disc)
Why have you created d:Q28912605 while d:Q252485 exists? --jdx Re: 19:09, 7 March 2017 (UTC)

we have to and always do with several items of an artwork. We will have to do so for other prints and the reproduction of the Great Wave. Before there was British and Metropolitan museum and the Inventary/accession number of the Met altogether in one item The Great Wave of Kangawa (Q252485). And there are many more impressions. As long as the inventory number of the Metropolitan was written in The Great Wave of Kangawa (Q252485) I it was retrieved from Wikidata to all files of the Great wave, even 1930 reproductions. So I had to separate the items but linked it through instance of (P31) and has part (P527) to each other. The same we do with series of paintings and casts of a sculpture. I will copy this discussion to Wikidata:WikiProject Visual arts/Questions, as I do not understand why you discuss it on my commons disc. There you can see similar questions and we can have a broader response.--Oursana (talk) 19:23, 7 March 2017 (UTC)


General question for the project team - in the case of a common motif like "cinquefoil", WD has cinquefoil d:Q3414584 (the heraldic symbol in the hierarchy of heraldic terms). Should we have a separate Q item for the figure itself under motifs > geometric motifs? My gut instinct is that there should be two separate items, with the heraldic charge both an <instance of> heraldic fsymbol and <instance of> cinquefoil (= geometric motif).

This comes up frequently when I am mapping AAT terms in Mix'n'Match. So far I have skipped these since I am unsure how to handle. AAT doesn't have a hierarchy for heraldic terminology, but the vocabulary is richly represented in Wikipedias. - PKM (talk) 20:02, 4 April 2017 (UTC)

Wikidata talk:WikiProject Heraldry could be a place for that … I wanted to write when I noticed that you did the only edit since 2015. But to your question: I'd also think two separate items are useful for the heraldic figure and the geometric motif. --Marsupium (talk) 19:17, 22 May 2017 (UTC)
@Marsupium: Sorry, it seems I never responded on this! Thanks for your comments. I didn't ask at the Heraldry project because I thought the project was dormant - the last question I asked there was never answered. But you're right, I should post the question there before I do a lot of work on motifs. I separated pattern and its subclasses from pattern and its subclasses, and then got distracted by something else. But I should get back to this soon. - PKM (talk) 02:46, 17 July 2017 (UTC)

Works after Leonardo da Vinci's Last SupperEdit

Jane023 (talk) 08:50, 30 May 2013 (UTC)
User:Vincent Steenberg
Marsupium (talk) 13:46, 18 October 2013 (UTC)
GautierPoupeau (talk) 16:55, 9 January 2014 (UTC)
Multichill (talk) 19:13, 8 July 2014 (UTC)
Susannaanas (talk) 11:32, 12 August 2014 (UTC) I want to synchronize the handling of maps with this initiative
Mushroom (talk) 00:10, 24 August 2014 (UTC)
Jheald (talk) 17:09, 9 September 2014 (UTC)
Spinster (talk) 15:16, 12 September 2014 (UTC)
PKM (talk) 21:16, 8 October 2014 (UTC)
Vladimir Alexiev (talk) 17:12, 7 January 2015‎ (UTC)
Sic19 (talk) 21:12, 19 February 2016 (UTC)
Wittylama (talk) 13:13, 22 February 2017 (UTC)
Armineaghayan (talk) 08:40, 10 March 2017 (UTC)
Musedata102 (talk) 20:27, 26 November 2019 (UTC) Hannolans (talk) 18:36, 16 April 2017 (UTC)
Zeroth (talk) 02:21, 4 June 2018 (UTC)
Infopetal (talk) 17:54, 9 August 2019 (UTC)
S.ann.adams (talk) 21:05, 13 December 2019 (UTC)
Karinanw (talk) 16:38, 24 March 2020‎ (UTC)
Ahc84 (talk) 17:38, 26 August 2020 (UTC)
  Notified participants of WikiProject Visual arts

This item needs cleanup. I would think it should be <subclass of> derivative work and all the existing <subclass of> statements should be changed to <has parts of the class> - but is there a better way to handle this item? - PKM (talk) 02:24, 17 July 2017 (UTC)

Hmm I don't think it should be a subclass of anything. It is just an article that discusses a series of paintings based on another painting. It is not a "class" of paintings at all and should just be listed as a painting series. Whether or not the painters were aware they were building a series is unimportant for the purpose of this specific item. They all have a religious genre and they are all based on the same painting. They should each have their own item and be listed from this item with "has part". All of those "instance of" statements need to be trimmed. Jane023 (talk) 11:51, 17 July 2017 (UTC)
Later I thought maybe it should be a Wikimedia list article, but I see Shonagon has made it into a group of works. That is also a good solution. Best Jane023 (talk) 08:13, 18 July 2017 (UTC)
Yes, I like that too. - PKM (talk) 06:54, 20 July 2017 (UTC)

Cleaning up life datesEdit

I am having a tour in Commons:Category:Creator templates with Wikidata link: mismatching birthdate, which is a good occasion for cleaning up and sourcing statements in Wikidata. Any idea what Valenti, 'Paragone' (1972) is referring to in [7] ?--Zolo (talk) 09:29, 4 August 2017 (UTC)

Probably some modern interpretation of Da Vinci's en:Paragone? Jane023 (talk) 16:31, 4 August 2017 (UTC)

This ResearchSpace/Metaphactory platform shows only 10 in/out statements so is impossible to search. The full info about this person is at but unfortunately I don't see how to relate the two IDs: this query at doesn't show anything useful

 SELECT * {?sub ?pred "129249"}

@Zolo, Jane023: The record has these citations of "Paragone":

  • (Valenti, 'Paragone', 265 (1972), p. 92, n. 1),
  • S.P. Valenti, 'Un pittore fiorentino a Roma e i suoi committenti', Paragone 265 (1972), pp. 80ff.;

I think Paragone is an art journal or other periodic publication. Worldcat returns many matches, and from correlation of volume and year I'd say this is part of the same series. Would be very curious whether you (or Wikidata Libraries) has access to it --Vladimir Alexiev (talk) 15:24, 19 September 2017 (UTC)

Unkown artists associated with "schools"Edit

Jane023 (talk) 08:50, 30 May 2013 (UTC)
User:Vincent Steenberg
Marsupium (talk) 13:46, 18 October 2013 (UTC)
GautierPoupeau (talk) 16:55, 9 January 2014 (UTC)
Multichill (talk) 19:13, 8 July 2014 (UTC)
Susannaanas (talk) 11:32, 12 August 2014 (UTC) I want to synchronize the handling of maps with this initiative
Mushroom (talk) 00:10, 24 August 2014 (UTC)
Jheald (talk) 17:09, 9 September 2014 (UTC)
Spinster (talk) 15:16, 12 September 2014 (UTC)
PKM (talk) 21:16, 8 October 2014 (UTC)
Vladimir Alexiev (talk) 17:12, 7 January 2015‎ (UTC)
Sic19 (talk) 21:12, 19 February 2016 (UTC)
Wittylama (talk) 13:13, 22 February 2017 (UTC)
Armineaghayan (talk) 08:40, 10 March 2017 (UTC)
Musedata102 (talk) 20:27, 26 November 2019 (UTC) Hannolans (talk) 18:36, 16 April 2017 (UTC)
Zeroth (talk) 02:21, 4 June 2018 (UTC)
Infopetal (talk) 17:54, 9 August 2019 (UTC)
S.ann.adams (talk) 21:05, 13 December 2019 (UTC)
Karinanw (talk) 16:38, 24 March 2020‎ (UTC)
Ahc84 (talk) 17:38, 26 August 2020 (UTC)
  Notified participants of WikiProject Visual arts

In Wikidata we have very few items of the type "British School 17th century" (see Q23906819). Currently this is instance of "Art Movement", as specified at Wikidata:WikiProject Visual arts/Item_structure.

Many museums and auction houses use structures like "Netherlandish School 16th century" as the "artist" for works. How should we model this in Wikidata so that it can be correctly captured in Commons and doesn't cause constraint violations in Wikidata?

My thought is that our model should be <creator> = "anonymous Q4233718", modified with a new property for "associated art movement" 'art movement or style with which an anonymous artist is identified' (values to include any Q968159). Commons would then pick this up as "Anonymous, British School 18th century" or the like.

Does this make sense? Does anyone have a better idea how to model these? - PKM (talk) 19:32, 17 September 2017 (UTC)

I assume you already know about Wikidata:WikiProject_Visual_arts/Item_structure#Use_of_creator_.28P170.29_in_uncertain_cases? Maybe something with school of (P1780)? Multichill (talk) 19:36, 17 September 2017 (UTC)
I do, but honestly I hadn't thought of using "school of" with an art movement. That should work! - PKM (talk) 19:44, 17 September 2017 (UTC)
I'd propose to use movement (P135) and skip creator (P170) and reserve school of (P1780) like creator (P170) for individuals and to delete British School 17th century (Q23906819) with the arbitrary century and put that information in inception (P571) where it fits better. --Marsupium (talk) 21:16, 17 September 2017 (UTC)
I agree the dates belong in a date field (I'd go for "point in time" = 1600s). But I think as used in art historical circles, "British School 17th century" should be the label. We need some structure that allows Commons and 3rd party sites using our data to generate the correct attribution "Anonymous, XXX School XXth Century" or "Unknown Artist, XXX School XXth Century". - PKM (talk) 17:47, 18 September 2017 (UTC)

Hmm I think this is a mistake and "British School 17th century" should be culture, not movement. The appropriate movement if it exists at all would be Dutch-style realism I think, for British paintings of the 17th-century. Jane023 (talk) 07:30, 18 September 2017 (UTC)

I support changing "XXX School XXth Century" from "art movement" to "culture". I think that's a better characterization. That would still work with qualifier "school of". - PKM (talk) 17:47, 18 September 2017 (UTC)
The more I think about this change, the less I am sure about it. A school defines a set of professional approaches and practices. I'm not sure it's either a movement or a culture exactly. - PKM (talk) 18:41, 19 September 2017 (UTC)

I think "creator" should not be recorded (because it's unknown) and instead style/culture should be recorded. "British School 17th century" needs to be dealt with carefully: if "British School" is a well-defined school that spans say 15-18 century then indeed the century designation should be removed. But many "cultural periods" go together with a time designation (even though not strictly defined). Regarding "school" vs "culture": this distinction is not always clear-cut, especially for ancient/indigenous/ethnographical cultures ("people"): see Wikidata_talk:WikiProject_Visual_arts/Item_structure/Art_movements#Ethnic_groups. --Vladimir Alexiev (talk) 14:58, 19 September 2017 (UTC)

Here we should always use whatever is in the referenced source material. If a museum catalog identifies a "culture" we should add one. If a museum identifies an unknown artist working within a (more or less) recognized stylistic "school" we should document that. From the Commons perspective, "unknown artist" is a standard usage. I suppose from a strict Wikidata perspective that could either be recorded as "creator" = unknown value, or "creator " = Q4233718. This project's current best practice seems to be the latter, but Commons should probably recognize either one as valid.
@jarekt, jheald:, do you have any insights here? - PKM (talk) 18:38, 19 September 2017 (UTC)
It just happen that I proposed creation of strip down version of commons creator template that would be applicable to schools, manufacturers, photo studios, photo agencies and other types of group creators which can not be associated with a single individual. I was thinking to place such templates in "Creator" namespace and use regular Commons Creator page (P1472). So this is the plan to deal with such creators from the Commons perspective. --Jarekt (talk) 14:30, 20 September 2017 (UTC)
@Jarekt: Thanks for the link to your proposal. - PKM (talk) 22:27, 26 September 2017 (UTC)

Use of instance of (P31): anonymous master (Q474968)Edit

Copied over from User talk:Jarekt#Use of instance of (P31): anonymous master (Q474968).

Hi Jarakt and AndrewNJ, you have changed Zolo's use of human (Q5) for instance of (P31) on Pseudo Granacci (Q30581143). Were there any public considerations for the use of anonymous master (Q474968)? It does not seem to be widely used. I don't have (yet) a preference here, I just guess we should follow one line with this. --Marsupium (talk) 17:49, 24 October 2017 (UTC)

I added Q5 to Pseudo Granacci (Q30581143). My edit was to fix P1472 constraint violation and have instance of (P31) which is some sub-class of people and anonymous master (Q474968) made sense at the moment. --Jarekt (talk) 18:03, 24 October 2017 (UTC)
My edit, was actually automated by mix'n'match. Probably not the best place to discuss it. My fuzzy feeling is that a special P31 for this kind of case might be a good idea, but that we should settle on which one. I see that notname (Q1747829) is also used. It is more general (which I like) but it is also a subclass of name (Q82799) which sounds like bad semantics. --Zolo (talk) 19:43, 24 October 2017 (UTC)
I see, I agree that anonymous master (Q474968) makes sense in that way, indeed there are some important use cases where it's an advantage to filter those (out), e.g. for inter-language copying of labels. I just think we should find a consensus and record the result somewhere in the guidelines.
Pseudo Granacci (Q30581143) was already sub-subclass of human (Q5) via anonymous (Q4233718) that is subclass of (P279) of human (Q5) (instance of (P31) was shortly discussed at Talk:Q4233718#Instance of human; P31 or P279:Q5 together already been removed 7 times in the item's history if I've counted correctly).
Multichill has also made anonymous master (Q474968) a subclass of (P279) of notname (Q1747829) two years ago.
I'd suggest to keep the subclass of (P279):human (Q5) at anonymous (Q4233718), but no instance of (P31):human (Q5) following Multichill and against Infovarius (I've mixed up P31 and P279). I'm sceptical about anonymous master (Q474968) at notname (Q1747829). And I think anonymous master (Q474968) is a good choice for instance of (P31) on our anonymous person items, not notname (Q1747829) because their sobject isn't really the name, but the person that appears to have such a name – the name has no "floruit" and isn't a creator of works. --Marsupium (talk) 09:44, 26 October 2017 (UTC)

Proposal to unite our discussions on one pageEdit

I just posted the section above here since it looked like this page has the most traffic recently compared to Wikidata talk:WikiProject Visual arts/Item structure and Wikidata talk:WikiProject Visual arts and Wikidata talk:WikiProject sum of all paintings. Thematically there isn't a difference between the pages (any more) as far as I see. I'd like to propose to unite our discussions on one page or perhaps one per WikiProject and freeze the others (or the other two on this WikiProject). If there will be more than one page we should somehow state the relation between the remaining pages. I'm afraid the current fragmentation is especially complicating the collaboration for new or infrequent users. What do you think? --Marsupium (talk) 13:19, 26 October 2017 (UTC)

  Support I have difficulty finding discussions I'm looking for and i am not a new user. - PKM (talk) 19:56, 6 November 2017 (UTC)


I stumbled across this picture Photograph of Flag Raising on Iwo Jima (NAID 520748) (Q117693) with a coordinate location (P625) that seems erroneous. But anyway, should it be the place where the picture was taken, or the place where the picture is now? Thanks! Syced (talk) 08:07, 23 April 2018 (UTC)

coordinate location (P625) is where the work is now. For the place where is was taken, you can use coordinates of the point of view (P1259). -Zolo (talk) 11:05, 23 April 2018 (UTC)

Creator issuesEdit

Hi, Please see Wikidata talk:WikiProject Visual arts#Creator issues. Regards, Yann (talk) 11:24, 9 July 2018 (UTC)

collection (P195)Edit

Hi! I'd ike to add collection (P195) in Sanint Philip Neri and Saint Aloysius Gonzaga in Glory (Q22949398); it is placed in San Filippo Neri baptistery (Q19828270), but it is property of the parish of Basilica di San Giovanni Battista (Q2887005), which is the curator of this painting. So, the value for location (P276) is San Filippo Neri baptistery (Q19828270), but which is the value for collection (P195)? San Filippo Neri baptistery (Q19828270) or Basilica di San Giovanni Battista (Q2887005)? --Yiyi .... (talk!) 08:21, 26 July 2018 (UTC)

It's your choice - personally I prefer the institution that manages the art, but this is often not very clear from sources or even from people in the institutions themselves! Jane023 (talk) 10:25, 26 July 2018 (UTC)
Thanks! It's Parrocchia di San Giovanni Battista (Q55758382). --Yiyi .... (talk!) 18:02, 26 July 2018 (UTC)

Scientific illustrationsEdit

I have been working with scientific illustrations, botanical plates to be specific. Lithographs, even. My knowledge of and language for the art part

The plates are paintings and have annotations, printer, and "et. lith.isms". I tried to note these using "media legend" and that was seriously wrong.

The prints have a historical significance to the science. Some of them are part of the first description of a species which gets cited and cited by scientists every time the species is written about.

Do you have a way to handle artworks such as this?

Here is one: Q56878923--RaboKarbakian (talk) 18:06, 10 October 2018 (UTC)

Hello RaboKarbakian,
those illustrations are definitely in the scope of this WikiProject. It's documentation is mainly on Wikidata:WikiProject Visual arts/Item structure.
I hope that helps. If you have more questions, go ahead to ask them, the more precise the better! If anyone else has more remarks, please add them! --Marsupium (talk) 19:22, 10 October 2018 (UTC)
That is just great! I made a new entry for it as the one I posted is for an article about it Trineuron spathulatum (Q57209102).
I am fairly certain that I can navigate the scientific parts, and the publication parts -- however, any help with the art parts that you can think of as my knowledge is practically only of the digital arts.--RaboKarbakian (talk) 21:46, 10 October 2018 (UTC)
It looks like I can make an ensemble of works of art (Q28870066) of all of the plates from the "voyage"? --RaboKarbakian (talk) 21:55, 10 October 2018 (UTC)
artwork series (Q15709879) with property has parts of the class (P2670) print (Q11060274) might be appropriate.
One thing that one should try to be clear on, at least for oneself, is whether the item for the print like Trineuron spathulatum (Q57209102) is intended to be for the print; or an edition/state of the print (different printings may have differed slightly, and could have different items); or a single copy of the print. Probably on Wikidata we want items at the level of the edition/state of a print; but we might want others as well.
Looking at Trineuron spathulatum (Q57209102), I would possibly not make the statement collection (P195) Flora Antarctica Plates (Q57241635) - P195 tends to be reserved for physical collections of actual objects. part of the series (P179) might be more appropriate here, or part of (P361). Similarly, inventory number (P217) is definitely for physical objects. series ordinal (P1545) would be more appropriate to indicate a position is a series, as a qualifier on a P179 statement.
Finally, a depicts (P180) statement is more useful to indicate the subject of the print, than qualifier of (P642) on a instance of (P31) statement -- searchers are more likely to look for the P180. Jheald (talk) 09:37, 12 October 2018 (UTC)

The scans show a stamp, maybe an embossing? Where the paper was warped around a metal press, especial for the voyage.... Is there a WORD for that?--RaboKarbakian (talk) 22:00, 10 October 2018 (UTC)

watermark (Q43065). - PKM (talk) 20:36, 11 October 2018 (UTC)
@RaboKarbakian, PKM: The word you are looking for may be platemark [8]. As far as I can see, we don't have an item for this yet; but we probably should, eg in case we want to specify the size of an image with qualifier applies to part (P518) "platemark". Jheald (talk) 08:46, 12 October 2018 (UTC)
Alternatively (I'm not 100% sure what you were describing), we have an item for chop mark (Q22907215) in the context of coins; but such marks also get applied by printers [9]. We should also probably have an item for "accession stamp", or more generally "stamp mark", but I'm not sure whether those exist as yet. And perhaps also an item for "embossed mark", the result of embossing (Q1335051) - Jheald (talk) 09:51, 12 October 2018 (UTC)
Blind mark from the vocabulary seems to be what it is. Platemark is similar but found in the margin as a border. The last time I got a copy of my birth certificate, they "did that" to the document, over the signature. The device looks like a stapler but leaves not a staple but a three dimensional impression. It might be on my passport also, but mine (unused since 2007) seems to have disappeared, so I am unable to verify this.
And, probably they should be editions or versions. For books, the same book might have been scanned twice. I have no idea how these plates were first issued or how many were made. There could be scans of a different set of the plates. The scan of the document they illustrate that was transcribed at wikisource only references them. --RaboKarbakian (talk) 14:04, 12 October 2018 (UTC)
@Jheald, RaboKarbakian: I have been working on various sorts of mark (Q57248625) since yesterday, using AAT as a starting point. I can research and add blind mark, plate mark, and “embossed mark” (same as stamp?). I’m sorting Commons as I go, so I’m slow with this work. - PKM (talk) 19:43, 12 October 2018 (UTC)
@Jheald, RaboKarbakian: [10] has "blind stamp' AKA "chop mark". I think this is AAT's chop (printer's mark). Does that sound right to you? - PKM (talk) 20:19, 12 October 2018 (UTC)
And also stamps " Marks that are printed, impressed, perforated, or embossed, made by a relatively hard object, especially a die, block, or other tool, on a softer or absorbant material such as paper or wax." - PKM (talk) 20:27, 12 October 2018 (UTC)
"Stamp" was the word I used when first trying to express what I thought this was via "officious language".--RaboKarbakian (talk) 03:08, 13 October 2018 (UTC)
@Jheald, RaboKarbakian: We already had blind stamp (Q884372) from EN and DE wiki, and I have assumed this the same as AAT "chop (printer's mark)" based on the definitions at and AAT. I have made it a subclass of new item stamp (Q57305415) as well as of mark (Q57248625) following AAT, which is slightly redundant. - PKM (talk) 22:14, 13 October 2018 (UTC)
I always thought that a watermark was part of the paper making process. Stamp is what I will use.--RaboKarbakian (talk) 15:26, 14 October 2018 (UTC)
Sounds good. You are correct about "watermark"; I misunderstood what you were looking for originally. - PKM (talk) 19:13, 14 October 2018 (UTC)
I approximated the size(s) based on scan information and well-known fractions (rounding up or down to the nearest quarter of an inch). I feel like I should have verified this with the available paper sizes from the mid-1800s, but there are maybe as many paper sizes as there are "little yellow flowers" (a highly technical botany classification which is akin to "little brown bird")....--RaboKarbakian (talk) 14:04, 12 October 2018 (UTC)


I'd like to add to our best practices for Creators that appellations such "the Elder", "the Younger" should be translated in labels. I have a spreadsheet which can be posted as a table to help with this work. Does anyone object? - PKM (talk) 22:00, 24 May 2020 (UTC)