Wikidata talk:WikiProject Provenance

OK I started a page and created a category! Next step is to set up Nazi loot lists. Properties like inventory number (P217) per looted art register? I have no idea how many of those there are in NL, let alone other countries. Jane023 (talk) 13:35, 31 August 2020 (UTC)

Lumley collection

I’ve been meaning for ages to do some serious work on the Lumley collection. Maybe this will encourage me to dig up proper scholarly references and do the work. - PKM (talk) 05:20, 11 March 2021 (UTC)

Collections and significant events

When a painting is sold multiple times, should we capture the collections/owners plus a significant event for every sale, or is collection/owner enough? See Catherine de' Medici and her children (Q106848606) and the provenance at Yale. - PKM (talk) 22:41, 15 May 2021 (UTC)

Gah that sounds like a ton of work! I would say do whatever you feel comfortable doing. For me, I really don't like significant event for this because I want to say seller=X and buyer is X and you can't do that now (that I know of). I do use it for important dates anyway, especially if there's an auction catalog with the price. One problem with doing owned by (P127) statements is getting the owners lined up chronologically. The probem is this is only possible if you input them chronologically. For me this is pretty much impossible because I do a lot of drive-by updates with quick statements using a catalog, where I add the ownership statement with an end-time for estate sales and point-in-time for exhibitions. Jane023 (talk) 09:58, 27 May 2021 (UTC)

How to model provenance?

I just put out a call for feedback in somem proposed changes on the WD:SOAP talkpage since I only work on paintings. Link is here: Wikidata talk:WikiProject sum of all paintings/Archive/2021#How_to_model_provenance. Please respond there. Jane023 (talk) 07:40, 31 July 2021 (UTC)

Wikidata:Property proposal/beforehand-afterward owned by

Please see this discussion about new qualifiers for significant event (P793). I am currently opposed to this proposal, but I have only commented, and don't oppose it yet. This is because this is such a generic property that perhaps it could be useful for other things besides artworks, like ships and houses. I have only responded to the examples, which are all paintings. Since paintings are also my focus here, I would encourage anyone reading this with a more neutral point of view to join the discussion. Jane023 (talk) 11:52, 15 March 2022 (UTC)

Lost Artworks

Hi, I am organizing works by Polish artist Stanisław Ignacy Witkiewicz, and run into a modeling dilema. A lot of his artworks are only known from photographs (high quality negatives) as the originals have not been seen since before world war II. I was thinking about modeling it with collection (P195) -> "no_value" with qualifier has cause (P828) -> lost artwork (Q4140840). We can also use significant event (P793) -> lost artwork (Q4140840), but I feel like there is no information about any loosing "event" only that they have not been seen for 80 years. Is there a better way to model it? Jarekt (talk) 02:26, 11 September 2022 (UTC)

"Owned by", again

I am working on this painting from Christie's. So far I have Portrait of a man at half-length (Q119860603) (and that's about 8 hours of work - clearly this is not sustainable!)

If I add the owners between John Mordaunt and Lady Mary (which I can get from the owners of Drayton House on EN wiki), is there is a value for a reference based on heuristic (P887) in the format "inferred from..." that I should use to indicate that we have interpolated ownership based on the "thence by descent" comment at Christie's? (I am happy to add a new item if we agree on what to call it - maybe just "inferred from 'by descent' statement", also adding the Christie's page as a reference URL?)

And do we like owners and collections in chronological order, or reverse chronological order? (I have the Rearrange Values gadget, so I can change these). - PKM (talk) 23:55, 24 June 2023 (UTC)

Yes it's slow going! So far we still don't have "by descent" and I have in the past left bread-crumbs for myself by listing the earliest one and the latest one before leaving the family. The problem becomes those pesky families where the descent is not always father-to-son and I have been down dozens of rabbit holes only to later admire my own work which I was gladly able to copy. So to answer your question, no, there's no easy way and yes, a "by descent" option would definitely be useful, especially if you can add any text from the source under "named as". I mean notes like "by descent of the <castle location>" vs "by descent of the <blood line>", which often do not correspond. The order of statements doesn't bother me - if there's a file the start and end dates get sorted on Commons. Once I have nailed the blood line I will sometimes go back and add the intervening owners with start and/or end dates, even if they weren't big collectors themselves - it is handy to be able to get the owners at certain times because their wives (who were often not legal owners) occasionally lent works to exhibitions. Thanks for taking the time to consider this! It does get hairy but I believe it's worthwhile. I see you added unknown for unknown dates - I tend to leave those blank, because it's easier to add info than switch info. Jane023 (talk) 11:59, 25 June 2023 (UTC)

I also just added the image to Henry Macwilliam (Q79396617) and to the sitter, because the Christie's record mentions it. Weird the Ringling website has not been updated with the new info, but somehow these should be linked but I don't know how without an artist. I am linking in the files on Commons as by the same person. Jane023 (talk) 12:18, 25 June 2023 (UTC)

Okay, I have created inferred from 'by descent' statement (Q119952263) to use with based on heuristic (P887). We can tweak it as necessary.

Thanks for tracking down the MacWilliam portrait - that was somewhere vague on my to do list but I got sidetracked on Dowdeswell and Dowdeswell's Gallery (Q119860955) - the sources are very messy, and the even the BMT conflates the father and son. PKM (talk) 22:18, 25 June 2023 (UTC)

Provenance Feature / Collection Stamp

Hi! A lot of provenance research is using certain features like stamps of known collections (see especially Frits Lugt and his collection of stamps, but also in east asia stamps are very important) or other features like labels. I couldn't find any information on how to model those or if this is on the list of (many) projects to tackle? Jo.burla (talk) 12:54, 27 February 2024 (UTC)