Wikidata:Property proposal/Taxon holotype location

taxon holotype location

Originally proposed at Wikidata:Property proposal/Natural science

   Not done

Description	location where the holotype specimen of a taxon is stored
Represents	holotype (Q1061403)
Data type	Item
Domain	taxon (Q16521)
Example 1	Pholcus phnombak (Q106573712) → Senckenberg Museum (Q706441)
Example 2	Nimiokoala greystanesi (Q42602605) → Queensland Museum (Q1537684)
Example 3	Giant koala (Q2340546) → Queensland Museum (Q1537684)
Expected completeness	always incomplete (Q21873886)
See also	taxonomic type (P427): name-bearing genus or species of this nominal family or genus (or subfamily or subgenus)

Motivation

Capture information on the location of holotype specimens of taxa. Please be gentle, my first attempt at writing a property proposal and the requirements are far from clear. CanadianCodhead (talk) 15:14, 21 April 2021 (UTC)

Discussion

WikiProject Taxonomy has more than 50 participants and couldn't be pinged. Please post on the WikiProject's talk page instead.

•  Support, an important property for biology.--Arbnos (talk) 15:33, 27 April 2021 (UTC)
•  Question Why is this property restricted to holotypes? --Succu (talk) 16:11, 27 April 2021 (UTC)

  @Succu:My intention assuming this was approved was to then create proposals for paratype etc, but as this was my first attempt at creating a proposal, I wanted to ensure I understood the template etc first rather than making the same mistakes over and over in multiple proposals. Perhaps it is worth considering a single property, something liker 'Specimen location' where the type (holotype, paratype etc) is separated by a qualifier and perhaps even the specimen number and link to the relevant webpage should it exist be captured?CanadianCodhead (talk) 14:21, 28 April 2021 (UTC)

  @Succu: what else would you include? --- Jura 17:24, 13 May 2021 (UTC)

  In zoology there is an endless number of types. Types can have a description, a locality, a date, a collector, a collecting number, a herbarium code (here called location), a specimen identifier and so on. I don't think this information should be added as qualifieres to the value of proposed property. A type is not allways a specimen. It could be an illustration too. And we have taxonomic type (P427). I don't think the proposed property is the right way to make progress. --Succu (talk) 15:42, 14 May 2021 (UTC)

  • Ok. Thanks for your input. I guess if only minimal information is available, one could start out with qualifying an unknown value of P427. --- Jura 16:48, 14 May 2021 (UTC)

•  Question Another modelling option that does not require extra properties is to have items for the holotypes themselves. They could have instance of (P31) holotype (Q1061403) (or paratype or any other types) and some relation to the taxon. It could be also instance of (P31) Giant koala (Q2340546) (considering taxons as classes) or use the qualifier. Or it could use the (rather broad) qualifier "of", as it is done here. There are 329 "types" on Wikidata, and the modelling varies: https://w.wiki/3Hqz. Anyways, great to see progress towards standardization! TiagoLubiana (talk) 04:28, 7 May 2021 (UTC)
  • Oh. and in that case, one can use collection (P195) to indicate the collection (example) TiagoLubiana (talk) 04:32, 7 May 2021 (UTC)

@TiagoLubiana:I agree standardization is a great goal, it is kind of where I was going about the more general property. I do think there could be a couple of issues with making items themselves for specimens. It works well when you know the collection ID number. For example if you known the holotype of species X is Royal Ontario Museum specimen JX-77454-P2, it doesnt work so well in the likely many cases where all you know is the specimen is at the Royal Ontario Museum. You could of course create a generic 'species X holotype' item, but then if you want to do that for things like paratypes where there can be multiple, it gets messy in terms of naming etc.CanadianCodhead (talk) 13:51, 11 May 2021 (UTC)

@CanadianCodhead: I'd be okay with "species X paratype" in the name and a description of where it is in the "description". The greatest benefit is to be able to add extra metadata, like the ID number, but also images and, who collected it, and the like. If there is an article describing an holotype, than one can link the qid of the speciment directly. TiagoLubiana (talk) 20:39, 11 May 2021 (UTC)

@TiagoLubiana: just to restart the discussion, how would you propose capturing this data. Here are 3 real world examples: - Spiclypeus shipporum is a dinosaur known from a single specmen. It is held in the Canadian Museum of Nature as specimen CMN 57081 - Ectatosticta yukuni is a newly described spider species. I will be creating the item for it in a moment. It has a holotype specimen held in the Institute of Zoology at the Chinese Academy of Sciences in Beijing under the catalog # Ar42454 as well as 3 paratype specimens held at the same place using catalog #'s Ar42455, Ar42456 and Ar42457 - Augacephalus ezendami is another spider which has 1 holotype and 1 paratype held at the British Museum of Natural History. So far that is all I know, where they are held, no collection numbers. They dont appear to be available or digitized in the museum's online portal. So how could examples like the above be entered in a consistent manner? CanadianCodhead (talk) 13:48, 19 May 2021 (UTC)

@CanadianCodhead:Do you have the source articles/URLs for that information? I can create an item for each notable specimen and set up links to the institutions and the species. TiagoLubiana (talk) 16:38, 19 May 2021 (UTC)

@TiagoLubiana: The 2 spiders are here and here. The search portal on the Canadian Museum of Nature is offline for some reason as of when I write this. The original paper describing the species which also lists the specimen number is here CanadianCodhead (talk) 16:44, 19 May 2021 (UTC)\

@CanadianCodhead: That is one way to do it: holotype of Ectatosticta yukuni (Q106906682) ; holotype of Augacephalus ezendami (Q106906769). I've also used taxonomic type (P427) to link the species to the types (which is one of the uses on its examples). I believe the same can be done for the paratypes and so on. TiagoLubiana (talk) 17:10, 19 May 2021 (UTC)

@TiagoLubiana: not sure P427 is the right place for the link back on the taxon Q item, that's supposed to be for the type species/genus, not specimen according to its documentation. So for Ectatosticta, that should be Ectatosticta davidi in there. There needs to be a different place for the link back on the species item. CanadianCodhead

Not sure I get the examples in the P427 discussion page. The description for the item in there states ' This is used to indicate the type genus of this family (or subfamily, etc) or the type species of this genus (or subgenus, etc). It is not intended to indicate a type specimen, etc.'CanadianCodhead (talk) 17:31, 19 May 2021 (UTC)

@CanadianCodhead: I agree with you, the usage is not clear. That said, the descriptions here on Wikidata are often imprecise, and things that are semantically close sometimes are mashed up together. Take a look at the properties used for types, it is really a mix: https://w.wiki/3MK2. It is reasonable to use P427 (close enough) and in the future the different uses (taxon-taxon and taxon-specimen) can be distilled into specific properties. TiagoLubiana (talk) 18:01, 19 May 2021 (UTC)

@CanadianCodhead: I mean, you can always propose a new property, but usually there is some overhead (as you can see :) ) TiagoLubiana (talk) 18:03, 19 May 2021 (UTC)

@TiagoLubiana: That's kind of the initial problem statement that prompted the proposal. There is no clear or defined way to document specimens. I really am indifferent if we have them done as separate Q items and a single link on the taxon page back to that, or a property with multiple qualifiers. I just see the need for something to do it that is as standardized as much as possible. I'm really not a fan of using properties for purposes that are explicitly stated in their documentation it is not for. That just creates a mess of data.CanadianCodhead (talk) 18:35, 19 May 2021 (UTC)

@CanadianCodhead: You have a good point, I agree with you in theory, but community processes are harder. Anyways, I'd definitely support a new property that takes this burden out of P427 and into a taxon-to-specimen property. Also, maybe you can start a data model page (i.e. the models on https://www.wikidata.org/wiki/Wikidata:WikiProject_COVID-19/Data_models) where work on sistematization can go beyond just one property. TiagoLubiana (talk) 18:53, 19 May 2021 (UTC)

Remark to holotype of Augacephalus ezendami (Q106906769): According to On some Southern African Harpactirinae, with notes on the eumenophorines Pelinobius muticus Karsch, 1885 and Monocentropella Strand, 1907 (Araneae, Theraphosidae) (Q90007362) the holotype of Ceratogyrus ezendami (Q106907504) is OUMNH-2009-043 from Oxford University Museum of Natural History (Q1760005). --Succu (talk) 19:27, 19 May 2021 (UTC)

Remark to „taxon-taxon and taxon-specimen“: Saying Aaa bbb is the type species (Q252730) of a genus Aaa refers to type specimen (Q51255340) of Aaa bbb. taxonomic type (P427) works fine for me for both types. --Succu (talk) 21:22, 19 May 2021 (UTC)

@TiagoLubiana: I'm not sure if you are proposing trying to create a data model for specimens or taxonomy on the whole. If it is taxonomy, I don't think that is a realistic goal. People have been arguing about how to model taxonomic data since probably the first week the site was launched. Seemingly simple issues like if the label for a Q item should be the scientific name or a common name or if 'Common Loon', 'Common loon' and 'common loon' should be 1 or 3 entries under the taxon common name property lead to vicious arguments.

I'm happy to close out this specific property proposal if the consensus it is better to add specimens as separate Q items and then do a linkback on the taxon Q item is a better approach. Frankly it likely is. But I really oppose using the P427 for the linkback. It's one thing to use a property for something it is somewhat related to, it is a separate really bad precedent and approach however in my mind to use it for something it specifically says it is not for. To my mind it is less an issue about entering data as it is about the mess and uncertainty it creates for querying and retrieving data. It just makes it very difficult to understand what data means in a query etc if a property is being used for multiple completely separate concepts. CanadianCodhead (talk) 15:05, 21 May 2021 (UTC)

 Oppose I prefer to have the location in the item for for the particular specimen. ChristianKl ❪✉❫ 11:36, 2 December 2022 (UTC)

•  Not done, no consensus of proposed property at this time based on the above discussion. Regards, ZI Jony ^(Talk) 18:20, 1 June 2023 (UTC)