Shortcuts: WD:PP/GEN, WD:PP/Generic

# Wikidata:Property proposal/Generic

 Before proposing a property Check if the property already exists by looking at Wikidata:List of properties (manual list) and Special:ListProperties. Check if the property was previously proposed or is on the pending list. Check if you can give a similar label and definition as an existing Wikipedia infobox parameter, or if it can be matched to an infobox, to or from which data can be transferred automatically. Select the right datatype for the property. Start writing the documentation based on the preload form below and add it in the appropriate section. Creating the property Change status=ready on template to attract the attention of a property creator. Creation can be done after 1 week by a property creator or an administrator. See steps when creating properties.
 On this page, old discussions are archived. An overview of all archives can be found at this page's archive index. The current archive is located at 2016/10.

## Generic properties

### stub of

Under discussion
Description This property is concern in giving a stub subject of the items. There are many articles that have size less than (5000 bytes), so we put a template in the bottom of the article, that said: "This is an article about Computer Science (e.g.). You can help wikipedia by expanding it.". This template is useful for many purposes, like: stats issues, drew the attention of the users, and trace the articles with same subject. I think we should have a property like this, and the bots can read the property and put it in the bottom of the article -If the article is a stub-, instead of putting a public stub template in the article. Item Q8554261, Q6683323, ...etc For example, there is an article in Arabic wikipedia called "ar:أزرق سماوي", which is about blue color, and it's a stub, but it contains a public stub template (no subject specified), so if we have a property (stub of); we can add it to the item, with a value of "Q7643731", and then, we let the bots read the property from wikidata, and then put it in the article. For the previous article, the bots supposed to add "template:color-stub" instead of "template:Stub". The bots will read the property (stub of) from wikidata, and instead of adding generic stub template, we can add a specified stub template to the article; so we can benefit from it with stats, classifications.
Motivation

The stub system is applied in many wikipedias, and I think the stubs should be centric in wikidata. ASammour (talk) 12:37, 15 October 2016 (UTC)

Discussion
• Oppose -- quality indicators are too-page specific. We have previously rejected this; please review the archives. --Izno (talk) 13:08, 15 October 2016 (UTC)
• @Izno: I can't find the proposal in the archive. Please refer the link to the discussion. And about your reply about quality indicators; I don't think this is a problem, it's true that the metrics differs from wikipedia to other languages, but if any wikipedia need to benefit from the property, they can use it depend on their metrics, and run the bots upon this.--ASammour (talk) 14:53, 15 October 2016 (UTC)
• Any property solely reliant on bots based on user-contributions on over 700 other wikis and some 40 million pages just in those wikis' content spaces? No, that's a problem. Anyway, the wikis will have their own programmatic basis for quality assessment coming soon. --Izno (talk) 15:32, 15 October 2016 (UTC)
•  Oppose "Stub" is a quality indicator on a linked wikipedia-article. Properties should be about the item subject, not on single links among them. Perhaps it could be done by creating a "stub"-badge, but that is beyond discussion here, and that would of course not yield stub-categories. Besides, stub-categories may be very different among wikipedias. Another solution would be the rejected proposal with stub-categories: see Wikidata:Property proposal/Archive/52#Category for stubs. Lymantria (talk) 15:59, 15 October 2016 (UTC)
•  Oppose Wikidata properties aren't applied to Wikipedia articles but to Wikidata items. ChristianKl (talk) 12:20, 16 October 2016 (UTC)
•  Oppose. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 13:26, 16 October 2016 (UTC)
•  Oppose what do you do if an article is a stub on some wiki's and a full-length article on others? This isn't appropriate for wikidata. ArthurPSmith (talk) 18:19, 17 October 2016 (UTC)

### to

Under discussion
Description For use as a qualifier of statements with relegated (P2882) or promoted (P2881). Item . Add it to many pages that are using relegated (P2882) or promoted (P2881)
Motivation

I need a qualifier to show where a team promoted or relegated, to show where the team played the next season due to promotion or relegation of the previous season. Xaris333 (talk) 23:11, 17 August 2016 (UTC)

Discussion
•  Comment In English at least, I don't think that "at" is the right thing for your use case. I would expect "at" to have a value of a location, an event or a time of day. "to" is what I'd expect, or possibly something more specific if "to" is too generic for others. Thryduulf (talk) 00:35, 18 August 2016 (UTC)
Thryduulf Ok, rename to "to". Xaris333 (talk) 12:19, 18 August 2016 (UTC)
•  Oppose This should be computable, once the old league is marked with its next highest and lowest league. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 11:36, 18 August 2016 (UTC)
Andy Mabbett An example:

Is different from:

Xaris333 (talk) 12:19, 18 August 2016 (UTC)

• (edit conflict) @Pigsonthewing: It is only computable when both leagues have exactly one league above and exactly one league below, this is not always true. For example lower leagues very often have different geographical scope. For example a team relegated from National League (Q58916) may play the next season in either National League North (Q59041) or National League South (Q58915) depending on geography not only of the relegated team but also that of other teams relegated to, promoted from and promoted to the level below National League (Q58916). A team in the Midlands may also move between leagues without promotion or relegation to balance the leagues (e.g. if only southern teams are relegated from the national level). The representation method on Wikidata must work for all levels, not just national to national. Thryduulf (talk) 12:28, 18 August 2016 (UTC)
It seems to me computable from the fact that the team is in session A in league X and in season B in league Y. If I additional know that the team was relegated it seems to me like all information is available. ChristianKl (talk) 20:41, 1 September 2016 (UTC)
•  Support as "to". Thryduulf (talk) 12:29, 18 August 2016 (UTC)
•  Oppose as "to". There should be a better (maybe generalized term) for promoted or relegated --Succu (talk) 19:38, 15 September 2016 (UTC)
Succu Like what? Suggest one... Xaris333 (talk) 21:55, 15 September 2016 (UTC)
Nope, Xaris333, it's your turn to clarify your propsal (at/to). --Succu (talk) 21:58, 15 September 2016 (UTC)
But you are saying There should be a better (maybe generalized term).... How can you be sure about that? What are we going to do? We need the qualifier. Xaris333 (talk) 13:45, 16 September 2016 (UTC)

ChristianKl, why? Can you suggest something else; I am trying to find something. Xaris333 (talk) 19:17, 7 October 2016 (UTC)

As I said above (and without getting a reply) the information is already possible to be calculated. Furthermore a short word like "to" as a high likelihood of being abused in for different purposes besides the one's defined here. ChristianKl (talk) 19:19, 7 October 2016 (UTC)

### orthogonal of

Under discussion
Description Nonoverlapping dimesion of the same class orthogonality (Q215067) Item longitudinal axis (Q380410) → lateral axis (Q2050624) longitudinal axis (Q380410) → yaw axis (Q9288712) backness (vowel) (Q24287581) → height (vowel) (Q24287574)
Motivation

It seems like there should be a property to model the relationship between the vertical axis and the lateral axis. I think orthogonal of is a good name because it's a quite general term in mathematics. I also think it makes senes to define it wide enough to model the relationship between the different properties of vowels.

– The preceding unsigned comment was added by ChristianKl (talk • contribs) at 9 June 2016‎ (UTC).

Discussion
•  Comment added to Wikidata:Property proposal/Generic today only.
--- Jura 08:03, 26 August 2016 (UTC)
•  Weak oppose A proposal affecting both very abstract and very real subjects!? Does that work cross language barriers? In Swedish we for example do not use the same word to describe a relation between two abstract vectors in Mathematics (ortogonal) and the relation between two Geometric entities (vinkelrät). I would therefor prefer if this proposal would be split into Mathematics and Geometry. -- Innocent bystander (talk) 13:54, 26 August 2016 (UTC)
Are you sure? The Swedish Wikipedia seems to [say](https://sv.wikipedia.org/wiki/Vinkelr%C3%A4t) that ortogonal is another word for vinkelrät. I would guess that 'ortogonal' is a hypernym of 'vinkelrät'.
The point of math isn't to talk mainly about abstract examples but also to use math to model real world entities. Mathematical vectors are a concept that's useful for modelling many real world problems.ChristianKl (talk) 09:58, 29 August 2016 (UTC)
"Vinkelrät" and "Ortogonal" are exchangeable in a simple Cartesian 2D or 3D vector space. But as soon as it becomes more complicated, it looks very strange to talk about "vinkelrät". "Vinkelrät" almost demands that you can use a machinist square (Q1778014), but that is hard when the vector looks like: ${\displaystyle A\sin x+B\cos x}$ -- Innocent bystander (talk) 16:01, 29 August 2016 (UTC)
What you said doesn't conflict with my claim that "Ortogonal" is a hypernym of "Vinkelrät". You can simply translate my proposed property into Swedish by using the word "Ortogonal". It doesn't create any conflicts. I don't see a reason to have an extra property that's focused on geometry when the mathematical term of orthogonalty can also be used for geometry. ChristianKl (talk) 17:15, 29 August 2016 (UTC)
"Ortogonal" and "Vinkelrät" are both useful in a small range with 2D/3D-Cartesian vectors. In pure Geometry and pure Abstract Math both cannot be used. -- Innocent bystander (talk) 06:08, 30 August 2016 (UTC)
What are cases where "Vinkelrät" can be used but "Ortogonal" can't? ChristianKl (talk) 08:23, 30 August 2016 (UTC)
The example above about "position of the tongue" is such an example. There is a narrow range where Mathematics and Geometry overlaps where both can be used, but hardly otherwise. -- Innocent bystander (talk) 08:50, 30 August 2016 (UTC)
In my understanding we have a three dimensional vector to describe a vowel that that has a height, a backness and a rounded dimension. Those three dimensions are independent of each other and thus orthogonal. In this particular case height and backness are also geometrically orthogonal but height and roundedness isn't.
Currently the only way to link roundedness (Q1425556), backness (vowel) (Q24287581) and height (vowel) (Q24287574) together would be different from (P1889). I think it would be worthwhile to have a more specific property for it and I think the mathematical term of orthogonality fits it. Do you have another idea of how a property for this purpose could be called? ChristianKl (talk) 09:58, 30 August 2016 (UTC)
As I said above, split the proposal into Geometry and Mathematics. Or skip the Mathematics for now. All examples above are within the range of Geometry from my point of view. Also note that my oppose is a weak oppose. We already have properties that works poorly in some languages, it does not corrupt Wikidata completely if we introduce this problem in my main language. -- Innocent bystander (talk) 10:39, 30 August 2016 (UTC)
Support This seems well-defined and useful and makes sense to me. ArthurPSmith (talk) 18:43, 30 August 2016 (UTC)

### legal status (medicine)

Under discussion
Description legal status for pharmaceutical drugs, e.g. general sales list for paracetamol in the UK legal status (medicine) (Q26715005) Item pharmaceutical drug (Q12140) paracetamol (Q57055) → General sales list (UK) (Q26715239)
Motivation

Match property from Wikipedia Infobox. Domswaine (talk) 17:27, 29 August 2016 (UTC)

Discussion
•  Question do you anticipate one instance of this property for each jurisdiction? Would values like "prescription only" or "over-the-counter" qualified with applies to territorial jurisdiction (P1001) be better? I support the idea of a property for the legal status of a drug (probably not just a medicine though, as what is classed as a medicine and what as a (recreational/harmful) drug varies between jurisdictions and within over time), but I don't want it restricted to just one or a handful of jurisdictions. Thryduulf (talk) 00:31, 30 August 2016 (UTC)
I would also prefer a property that's qualified with applies to territorial jurisdiction (P1001). I don't see why the word medicine should be used. Controlled Drug or controlled substance seem to be better names for me.ChristianKl (talk) 10:06, 30 August 2016 (UTC)
Agree with jurisdiction qualifier and expanding to non-medicinal drugs. Still think proprietary item values for each jurisdiction would be required as drugs policy far from internationally uniform with some nations having several levels, e.g. general sales, pharmacy only or prescription only for medicinal drugs in the UK. These could be, however, a subclass of more wide reaching values, such as "prescription only" or "over-the-counter". Domswaine (talk) 11:38, 30 August 2016 (UTC)
Would there be perhaps benefit in a property for a broad category (e.g. "over-the-counter", etc) for easy international comparison and a second for the detailed restrictions and requirements (or lack of) (maybe an item for "legal status of $drug in$country" that contained details?)? My thinking is something comparable to the manner of death (P1196) and cause of death (P509) pair. Thryduulf (talk) 14:34, 30 August 2016 (UTC)
I would advice against using a broad "over-the-counter" in a way that isn't subclassed. If you take the phrase literally many countries allow the selling of those medicine by putting them on shelves. Other don't and require that the pharmacist who sells the drug speaks to the patient. Whether or not businesses that aren't pharmacies can sell the product also depends. ChristianKl (talk) 10:20, 31 August 2016 (UTC)
OK, the actual category names will need agreeing but I think the general categories are something like this "completely illegal", "prescription only, special license/authority needed to prescribe" (, "prescription only, all/most doctors can prescribe", "available from pharmacist (person) only", "available in pharmacy (type of shop) only", "available without restriction", "not regulated". Do those seem OK to you? Are more needed? Thryduulf (talk) 10:49, 31 August 2016 (UTC)
Is there an argument that having multiple, potentially overlapping, generic values would end up being more complicated and less representative than proprietary, country specific, values.Domswaine (talk) 21:38, 31 August 2016 (UTC)
The values shouldn't be overlapping (which is why it is needed to discuss what the values are) but are sort or hierarchical, e.g. I have in the past been prescribed by a doctor throat sweets that are available on supermarket shelves. The idea is to have a generic value per jurisdiction for easy and high level comparison between jurisdictions AND a separate property with the specifically detailed value per jurisdiction that is required for detailed study of one jurisdiction but is difficult to compare across jurisdictions. Thryduulf (talk) 17:27, 1 September 2016 (UTC)
General sales, pharmacy medicines, pharmacist only medicines, prescription medications and controlled drugs seem a good place to start for the generic labels. For the propriety labels for each jurisdiction, general sale list (GSL), pharmacy medicines (P) and prescription only medicines (POM) are the classifications for what are judged to be medicinal drugs in the UK with controlled drugs categorised into one of class A, class B and class C. Domswaine (talk) 18:22, 05 September 2016 (GMT)
The example provided, Paracetamol, illustrates some of the challenges here. Where you can buy this drug in Sweden, partly depends on the package. 1000 mg demands a prescription from a physician. To get 500 mg you have to visit a drug store. 125 mg can be found in any approved shop. Suppository versions of this drug can have other rules. -- Innocent bystander (talk) 08:00, 3 October 2016 (UTC)
•  Question Would a quantity qualifier be sufficient here? Domswaine (talk) 18:39, 09 October 2016 (UTC)
• I don't think so, as for paracetamol I would expect quantity to refer to the number of tablets. For example in the UK "the largest pack size of paracetamol that shops without a pharmacist working in them can sell is 16 tablets, but pharmacies can sell packs of 32 tablets." [1], which I'd represent with a quantity qualifier using tablet (Q206077) as the unit. For the situation Innocent bystander raises (which also applies to ibuprofen in the UK according to [2]) I'd expect a qualifier called "strength", "dosage", "active ingredient amount" or something like that, which afaict doesn't currently exist. Thryduulf (talk) 10:56, 10 October 2016 (UTC)
• Yes, and it may also apply to "usage". I know one drug against mycosis (Q464067) that can be bought without prescription for treatment for your feet. But you need a prescription anywhere else. Some drugs can be prescribed by a dentist or a vet. And some requires a prescriptions from a specialist in some specified medical discipline. -- Innocent bystander (talk) 11:34, 10 October 2016 (UTC)

@ChristianKl: Thanks for the ping. Some other uses which are related are food additives, agriculture chemicals, dietary supplements, and alternative medicine. There are some places which group each of these with medical regulation and some places which treat these separately. Food additives might include colors, preservatives, and wax coatings; agriculture might include pesticides, herbicides, and veterinary drugs like livestock growth hormones and antibiotics; dietary supplements can mean vitamins or physiologically active non-allopathic treatments; alternative medicine can mean a dietary supplement or any sort of drug. Any of these might have a mention in their own schedule. I would want the available labels to be inviting for sorting any of these. Blue Rasberry (talk) 00:59, 24 October 2016 (UTC)

### male form of label

Under discussion
Description male form of label Monolingual text occupation (Q12737077) and position (Q4164871) hétaïre (Q466118) → no value acteur (Q33999) → fr"acteur" nephew or niece (Q780) => en"nephew" frato / fratino (Q31184) => eo"frato" ondernemer (Q131524) → nl"zakenman" politikisto (Q82955) → io"politikistulo" dramatan (Q33999) → vo"hidramatan" ûndernimmer (Q131524) → fy"sakeman" I plan to add a masculine form of label to names of title and function that are feminine because at the moment some masculine terms have feminine labels but the reverse is not true, and one cannot know what is the masculine label of actress (in French) it would be great to have one automatized sparql request listing feminine labels of title and functions which do not have a pending masculine label, and to enter the masculine form when it exist as well as notifying for the others that there is no feminine word associated. female form of label (P2521)
Motivation

Il y a une propriété forme féminine du libellé pour les noms d'occupation, et le pendant n'existe pas pour les formes masculines du libellé, ce qui serait intéressant à avoir pour les langues genrées. Nattes à chat (talk) 21:05, 7 September 2016 (UTC)

Hi there,
Being in the process (in French) of adding female form of label (P2521) to male denominations of names of occupation and professions, I have stumbled upon other difficulties:
• some feminine words like hétaïre (Q466118) appear in the search of elements referring to an occupation or profession performed by a woman that does not have a female label. This is a strange fact for me, as these are feminine words in the first place, and not masculine. For e.g., troubadour did not have a female label, so it appeared in the search which is logical because it is a masculine word in french. Hetaire is a word used solely for women in the antiquity (ses wikipedia article on the subject: https://fr.wikipedia.org/wiki/Hétaïre)
Therefore could we envisage to clarify this by:
• either separating the two labels (for exemple acteur (Q33999) and actrice Q21169216 are separate) and include in every element a female and a masculine label. For exemple acteur (Q33999) has a female label but actrice Q21169216 does not because the property dose not exist.
• either putting the male and female denomination together "acteur ou actrice" in the label definition to match for e.g. the english version where there is no feminine or masculine attached to a profession or occupation. If we consider the occupation, it is the same performed by both genders so should not be in two different elements. Recently on the french wiki, a tendency to separate articles describing female and male occupation (in sports) that were similar were reverted on the grounds that the activities were the same, and the separation, because of the existing gender gap, would lead to further invisibilisation of woman's articles, the male articles being far more viewed that the female.
In both cases the property male form of label should in my view be created, so we could start there and discuss the rest at length later. --Nattes à chat (talk) 11:27, 8 September 2016 (UTC)
Discussion
• This has been rejected before. What is different now? --Izno (talk) 12:43, 8 September 2016 (UTC)
There was no use case in the previous discussion, and now we already have 4 languages (Dutch, French, Frisian and Ido) that require this property. Robin van der Vliet (talk) (contributions) 16:34, 13 September 2016 (UTC)
• last discussion: Wikidata:Property proposal/male form of label --Pasleim (talk) 17:44, 8 September 2016 (UTC)
• Is there any place in Wikipedia where having the label would be important for infoboxes? Currently I don't see such a usecase. * For the sake of Wikidata I think it makes sense to wait for the Wikidictionary integration. Given the recently discussed WMDE plan for having Wikidictionary and Commons integration as a priority for Wikidata development in 2017, I think it's worth the waiting. In Wikidata there's also no issue with adding additional subclasses that are gender specific.
It is currently important for the Frisian Wikipedia, take a look at this infobox, it shows both gender forms because we can't retrieve the male form. It works fine though on articles about females. Robin van der Vliet (talk) (contributions) 16:34, 13 September 2016 (UTC)
As far as hetaera (Q466118) being only applicable to woman, that's worthwhile information to store but that information isn't about the label that used. Using "gender:female" might work better.
As far as actor (Q33999), goes I don't think it's a item that's supposed to say anything about the gender of the person. actress (Q21169216) is instance_of Wikipedia page. It's also not a occupation that a female actor can have. I don't see the point of actress (Q21169216) as it stands currently.
There could be subclasses of actor (Q33999) for male/female actors where the items contain information about the gender. ChristianKl (talk) 12:20, 9 September 2016 (UTC)
actor (Q33999) is actually gendered: it is a masculine word, and it is never used to design a female actor, who would be designed by {{Q|21169216}. It would be useful for someone to know that the male label of {{Q|21169216} is actor (Q33999). For the moment when this type of information appears, it is only appearing to give the female label. I think it should be balanced. --Nattes à chat (talk) 07:55, 10 September 2016 (UTC)
Honestly, dealing with linguistic issues in Wikidata makes me unconfortable. I'd prefer to use subclass items, as items have labels, for this and dealing with this conceptually : one item for the male gender people with this occupation, and one for the female form. All of those would be subclasses of a generic item for any people who have this occupation. This would have no But IF WE REALLY can't adopt such a model, I  Support this property - as the current solution is actually an unfortunate gender bias - we implicitely assume the label of the item is a male form, actually. author  talk page 08:54, 10 September 2016 (UTC)
if it was possible to have a generic term for an occupation and two subclasses at the same hierarchical level that would be great, but then how would we name this in French if there is no "neutral" term? which is why, before adressing this question I suggest to create this male form of label to adress the gender bias.--Nattes à chat (talk) 07:57, 11 September 2016 (UTC)
@Nattes à chat: Well, it does not really matter we don't have a french neutral - first because terms do not really matters on Wikidata so we totally can use the same label for the parent item and the male one if we want to, no problem in this project because what's important is the definitions. Second because we totally can use label such as "Bucheron/bucheronne" or "Bucheron-nne". No problem at all. author  talk page 18:33, 11 September 2016 (UTC)
•  Strong support Sometimes we use a neutral form for the main label, sometimes the male form and sometimes even both forms (separated by "/" or "or"). This will result in that Wikipedias using female form of label (P2521) will always show the female label for females, but will sometimes have to show a neutral or even both forms for males, because it is not possible to request a specific male form. This problem will be fully symmetrically solved by this proposal. Robin van der Vliet (talk) (contributions) 00:55, 11 September 2016 (UTC)
Comment We need this property on the Frisian Wikipedia. Most Dutch and Frisian words are neutral by default and some of them have female versions, but there are also words that don't have a neutral form and only have a male and female form. For example, the Frisian word for carpenter is "timmerman" for males and "timmerfrou" for females. Because of the unavailability of a specific male label property it looks like this in our new Wikidata template. It should look like "timmerman" for males, "timmerfrou" for females and as a neutral compromise it should show "timmerman / timmerfrou" for the few people who don't have a gender assigned. Robin van der Vliet (talk) (contributions) 18:01, 12 September 2016 (UTC)
• One thing we have to find out is what aspect of a word we here imply? The Swedish word "Sjuksköterska" is feminine as a word. But it is used both about male and female nurses. The male form of this word: "Sjukskötare" is male as a word, but it is also used both about males and females. These tow words are very different as occupations. The former demands an academic degree while the latter only can be found within psychiatry (Q7867). -- Innocent bystander (talk) 07:37, 11 September 2016 (UTC)
@Innocent bystanderI dont think this situation exist in French, a feminine word designing a male occupation. I agree that every use of a word has to be considered carefully, but then the label would be appied to specific words at a time. We would be speaking of words which design the same activity (same profession and same diploma), with the possibility of linking to the female and male expression of that same activity when it is expressed differently in a binary structured language such as French. --Nattes à chat (talk) 07:50, 11 September 2016 (UTC)
@Nattes à chat: In Swedish it is today common that a gender specific word is used for both males and females. Then it has become considered the "gender neutral form". "Sjuksköterska/Sjukskötare" is in that aspect little of an exception, since it has different meanings in different forms. "Lärarinna" when is is used about the 19th century also have a different meaning than "lärare" since teachers of the 19th century had different kinds of diploma depending of the gender. On svwiki the use of gender neutral forms is a common reason to editwars. The use of gender could also depend on if you are from Sweden or Finland. Gender neutral form is the normal form in Finnish and that influences also the Swedishspeaking people there. -- Innocent bystander (talk) 12:17, 11 September 2016 (UTC)
•  Oppose per the discussion closed last month. --Sebari (Srittau) (talk) 01:19, 12 September 2016 (UTC)
You proposed previously to use aliases, but that makes it impossible to retrieve the male form systematically, which is now needed for the Frisian Wikipedia. Take a look at this infobox, it shows both gender forms because we can't retrieve the male form. It works fine though on articles about females. Robin van der Vliet (talk) (contributions) 16:45, 13 September 2016 (UTC)
•  Comment what would be the domain for this property? Currently P2521 is mainly used for occupation (Q12737077) and position (Q4164871). The most recent proposal we had tried to expand this to the domain of taxon common name (P1843). If this is meant to be the equivalent of P2521, I think we should explicitly define the scope.
--- Jura 10:45, 12 September 2016 (UTC)
This property would be the counterpart of the female property and both properties are to be used symmetrically. The scope of both should be at least occupation (Q12737077) and position (Q4164871). Robin van der Vliet (talk) (contributions) 15:39, 16 September 2016 (UTC)
Ok. I added that above. I think we should avoid the animal farm we had in the last proposal.
--- Jura 10:55, 26 September 2016 (UTC)
Fine with me, let's only add them for human occupations and positions. Robin van der Vliet (talk) (contributions) 12:01, 27 September 2016 (UTC)
•  Oppose per previous. --Yair rand (talk) 07:03, 13 September 2016 (UTC)
Your previous argument was that this is in the domain of Wiktionary. In that case, would you support deletion of female form of label (P2521), which is also in the domain of Wiktionary? Both properties are equally needed for the Frisian Wikipedia. Robin van der Vliet (talk) (contributions) 16:34, 13 September 2016 (UTC)
•  Strong support This label is needed for the Ido Wikipedia. In Ido, there's always different formes for the two genders. Aktoro = actor (neutral – either or), aktorulo = actor (male), aktorino = actress. We want to show female labels for the women and male labels for the men, but this is currently impossible. The women get the female label, while the men have to get the neutral one. See io:Module:Wikidata (line 509). – Algentem (talk) 15:29, 13 September 2016 (UTC)
•  Support after hearing the Dutch infobox use case. Jane023 (talk) 17:12, 15 September 2016 (UTC)
•  Strong support Terms previously used to indicate males exclusively are being used more and more to include women as well, the use cases support that. When you search for people in a certain profession, you want to see both genders, and therefore you need a separate label to indicate profession and gender MarjonW (talk) 19:32, 15 September 2016 (UTC)
This property has nothing to do with searching. The fact that it appears to have something to do with it suggests that it's misleading and no good property. ChristianKl (talk) 20:40, 17 September 2016 (UTC)
No good property? It is currently needed for already 2 Wikipedias, just as much as female form of label (P2521). Or is that also a bad property? Robin van der Vliet (talk) (contributions) 00:25, 18 September 2016 (UTC)
It feels like you didn't read the sentence towards which you are replying and the argument that it makes. ChristianKl (talk) 12:30, 20 September 2016 (UTC)
•  Oppose given the lack of defined scope for this property (my question above remained unanswered), this proposal suffers the same problems the previous ones.
--- Jura 15:11, 16 September 2016 (UTC)
Scope is defined now. Robin van der Vliet (talk) (contributions) 12:01, 27 September 2016 (UTC)
•  Support It would be useful in many languages. - Scailyna (talk) 06:55, 26 September 2016 (UTC)
•  Strong support after hearing the case including the examples in different languages. Ecritures (talk) 08:41, 26 September 2016 (UTC)
•  Support We need this property in several languages, including French or Spanish for instance. 77.207.177.82 15:52, 26 September 2016 (UTC)
•  Support Useful, in line with female form of label (P2521). Lymantria (talk) 16:32, 21 October 2016 (UTC)

### Craft

Under discussion
Description Craft (or vessel/vehicle) used in a journey, expedition or mission. craft (Q1294755) Item Expedition, Journey, Mission would like to use this property to connect various historic expeditions to the main ship used. vessel (P1876) should be a subclass of this class
Motivation

There's currently a vessel (P1876) property for use within the domain of spaceflight missions, but no generic property for connecting an expedition / journey / mission to the main craft or vehicle used. I'd like to use this particularly within the realm of historic voyages, e.g. Discovery Expedition (Q1031706) and first voyage of James Cook (Q3178839) Frankieroberto (talk) 22:08, 12 September 2016 (UTC)

Discussion
• A good use-case, but we should re-purpose vessel (P1876). I therefore formally  Oppose. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 17:14, 13 September 2016 (UTC)
• Happy to re-use vessel (P1876), but is there a mechanism to rename it 'craft' to be more widely applicable? --Frankieroberto (talk) 18:01, 13 September 2016 (UTC)
• Hmm, so it looks like I can edit the [English] label and description, but I don't want to do that without agreement (and also a bit of help to update the labels in other languages). --Frankieroberto (talk) 18:44, 13 September 2016 (UTC)

### inferred from statements on

Under discussion
Description (use as reference) statement added based on related statement found on the following item Item MISSING
Motivation

Currently we have a bots that automatically add inverse statements. They sometimes use "stated in". I think it would be worthwhile to have a more specific property for the use case. ChristianKl (talk) 10:07, 27 September 2016 (UTC)

Discussion
•  Comment: it can be useful, but such a source is easily lost. --AVRS (talk) 12:16, 27 September 2016 (UTC)
Especially in cases where the source it's lost, having a "inferred from" property could help with recognizing that the statement isn't justified anymore. Those are cases where data provenance is important. ChristianKl (talk) 14:35, 27 September 2016 (UTC)
•  Comment - this doesn't sound like a good idea to me. Can you give a specific example or two to illustrate though? ArthurPSmith (talk) 18:48, 27 September 2016 (UTC)
Let's say we have a property Alice (Qa) and a property Bob (Qa). We have the statement Qa child (P40) Qb. Now a Bot like User:LandesfilmsammlungBot comes along and wants to add the statement Qb father (P22) Qa. The current behavior of that bot is to give as a reference stated in (P248) Qa for such a claim (it also adds retrieved (P813). I think that's using stated in (P248) in a way it isn't designed to work and it would be good to have an alternative that make the data provenance more clear. ChristianKl (talk) 21:24, 27 September 2016 (UTC)
•  Support we had various approaches suggested for this and each time someone adds statements based on this some other group wants yet another solution. I think a more explicit label might even work better, e.g. "statement added based on related statement found on the following item". I don't think it should be used for properties other than symmetric or inverse ones. This new reference-qualifier could replace corresponding Wikidata item (Q20651139).
--- Jura 06:56, 5 October 2016 (UTC)
@Jura1: I think your sentence is a bit too long for a property name. I modified the name and added it as description. ChristianKl (talk) 09:50, 5 October 2016 (UTC)
•  Support. Thryduulf (talk) 12:15, 6 October 2016 (UTC)
•  Oppose. Automatic inverse statements should just copy the actual source, rather than referencing other Wikidata items. If there is no source, than perhaps the inverse shouldn't be automatically created in the first place. --Yair rand (talk) 18:11, 6 October 2016 (UTC)
@Yair rand:Simply copying statement means that there's no provenance that allows a data user to know that a bot created the entry based on reasoning about other Wikidata items. That especially an issue when the subject is more complex that a simple inverse statement. If you infer that Alice is the mother of Bob based on the fact that Alice is female and Bob is her child, that might be an error in cases of transgender individuals (e.g. the source might say that Alice is the father of Bob). ChristianKl (talk) 19:58, 6 October 2016 (UTC)
• Hesitant. It would be better if we could link directly to a real statement, or all this will look like a dirty hack. This will easily get ambiguous on which claim was used, so we actually would need to duplicate the claim main snak at least in some source of ours or something ... author  talk page 19:48, 12 October 2016 (UTC)
• Could we use uses property (P3176) as a qualifier to record which statements were used to infer from? Thryduulf (talk) 16:53, 13 October 2016 (UTC)
• This would not be enough, we would have to store also the value as there could be several statements with the same property. But this still could be several statements with the same property/main value pair (main snak) ... so maybe in some cases we would also have to store some qualifier snaks ... in the end this amout to copy the whole statement which is not really practicable and difficult to implement. Of course maybe a partial solution could be implemented right now but a more close examination of the needs of something like that should occur - and this is in the scope of WikiProject Reasoning so maybe the discussion should happen there. author  talk page 17:12, 13 October 2016 (UTC)
• The information might not be completely comprehensive but it would still provide more information about the provenance than the status quo. We could have an additional property like "Automatic reasoning schema" that a bot can use if we want to provide more detailed information. ChristianKl (talk) 12:02, 22 October 2016 (UTC)
• @ChristianKl: A usecase where this would actually be useful would help. Do you have something in mind ? If the only usecase is "complete information for a symmetric property", it's obvious that the item in case is the suject of the claim. Knowing the rule is enough to know that the item has been inferred either from statement in the item of a parent or from a brother/sister. Thinking about this, it would probably be more useful to have an item for the rule of inference where the rule is described and a property to link to this item, like we have a "based on heuristic (P887) " - that could be used in this case, by the way.
(also, rereading the proposal, it should be more appropriate to say to the bot owners who does that to just stop putting items in "stated in") author  talk page 13:33, 22 October 2016 (UTC)

### is a set of

Under discussion
Description set of items of the class set (Q36161) Item Sets/Groups Set of Intercostal nerves (Q1515961) → Intercostal nerve (Q27058076)
Motivation

In Anatomy there are a lot of cases where there are items for groups and it's worthwhile to be able to have a more specific property than "has part". ChristianKl (talk) 18:29, 30 September 2016 (UTC)

Discussion
Question why ? author  talk page 09:06, 1 October 2016 (UTC)
Being able to specify the identity of items is valuable. When the identity is specified it's less likely that people with different languages use the item differently. ChristianKl (talk) 12:04, 1 October 2016 (UTC)
Wat's the identity of an item ? author  talk page 12:46, 1 October 2016 (UTC)
In the case of the FMA defined concepts that are defined as sets, I would call them being a set part of their identity. ChristianKl (talk) 13:33, 1 October 2016 (UTC)
What's FMA and where are those definitions ? author  talk page 15:55, 1 October 2016 (UTC)
FMA is the Foundational Model of Anatomy, which happens to be a controlled vocabulary for human anatomy. ChristianKl (talk) 16:21, 1 October 2016 (UTC)
Why is that "Set of Intercostal nerves" a set and not a class "Intercostal nerve(s)"? --AVRS (talk) 16:03, 1 October 2016 (UTC)
FMA has a concept for the set (FMA:75467) and also a class that the individual intercostal nerves can subclass (FMA:6028). Those two concepts have different FMA IDs. ChristianKl (talk) 16:21, 1 October 2016 (UTC)
That's exactly what has part (P527) is supposed to represent. To say this is a set in the FMA sense, I'd instead propose to tag the set-class with a metaclass :
< Set of Intercoastal nerves > instance of (P31) < FMA-set class >
. author  talk page 16:49, 1 October 2016 (UTC)
No. The nature of a set (S) of X is that every member of S is an X. That implies that if A isn't an X, it also isn't part of S. The "has part" relationship on the only hand only indicates that there is a B that is part of S. ChristianKl (talk) 17:22, 1 October 2016 (UTC)
That should be a part of the property description then. What I'd propose to model this is then something like a "unique value" constraint on the "has part" property for each class-instance of the FMA class metaclass. author  talk page 17:34, 1 October 2016 (UTC)
I don't think 'FMA class' is something special. UMBEL has a general set concept. In general we don't use "X class" when we add terms from controlled vocabularies and link them to the concepts we use. We don't let muscle organ development (Q14863685) subclass "GO_class".
It's not good to have important information that's just mentioned in the description. Descriptions don't auto-translate into other languages. It's a bad idea to have information that's only available in description in some languages when it would be possible to capture the information via statements. The statements also allow computers to understand what a concept is about. Being able to express information with statements is the core of what structured data is about. ChristianKl (talk) 18:29, 1 October 2016 (UTC)
@ChristianKl:
•  The Umbel concept is about mathematical sets, actually sets of mathical objects like numbers or other sets, described by set theory (Q12482)  . The relations on elements on a math set are "set membership" and "set inclusion". So it's very different from the concept of the FMA ontology which is a relation between classes of organs, if I understand well. Although the class concept in ontologies is also based on set theory, ontologies like Cyc that are cited on your UMBEL link does actually separate quite clearly mathematical sets and classes of concrete objects to not mix them up. Although the "class" concept that is the base of our ontologies has an analog in mathematical set theory.
• We need also to talk to humans :) and then you should provide the statements that would describe your proposal, they are not presented as well. author  talk page 19:13, 1 October 2016 (UTC)
• The point of math is that you can model the world with it. Math tells me that 2+2 is 4, so I know that if I add two apples to two apples I get four apples. In the same sense the mathematical concepts of set is useful in this instance. In particular if I have a set of apples with cardinality two and merge that with non-overlapping set of apples of cardinality two, I will get a set of apples with cardinality four.
To me the notion what a set happens to be, is a primitive that I formally learned math classes. In this case I think Set of Intercostal nerves (Q1515961) has part (P527) Set of Intercostal nerves (Q1515961), Set of Intercostal nerves (Q1515961) has part (P527) Second intercostal nerve (Q27058089), etc. {{Q|1515961} subclass of (P279) Intercostal nerve (Q27058076). That leaves the relationship of Set of Intercostal nerves (Q1515961) with Intercostal nerve (Q27058076). The straightforward term from math is that Set of Intercostal nerves (Q1515961) is a set of Intercostal nerve (Q27058076).
In set theory the terms set and class both exists and have clearly distinct meanings.
en.wiki has a definition for a mathematical set "A set is a gathering together into a whole of definite, distinct objects of our perception [Anschauung] or of our thought—which are called elements of the set." ChristianKl (talk) 19:39, 1 October 2016 (UTC)
@ChristianKl: You don't seem to get my point. In UMBEL and in Cyc of course classes and set - sets are a special kind of classes in set theory - to model the world. But the "set" class in Cyc is not used to model the concrete world, but to model maths themselves. This is different than sets defined here who are used in a different sense "an organ composed of (smaller) organs of that type". author  talk page 19:49, 1 October 2016 (UTC)
No. UMBEL says "An instance of Set_Mathematical can be any arbitrary set of Things". It doesn't say that the Things have to be mathematical in nature. It also has the subclass set of colors. Colors are on that level are not mathematical in nature. ChristianKl (talk) 20:00, 1 October 2016 (UTC)
@ChristianKl: OK, from what I understand then we don't need this. A type has a natural extension and if we want to say that we know the number of its instances it's enough to use quantity (P1114) who was intended to denote the number of instances of a class - then we know the extension of the class is fully known. Second, the "is a set of" relation can be expressed by subclass of (P279)  : It links a "set", mathematically a special kind of classes, and any instance of the subclass is an instance of the parent class, so this relationship express that the subclass is a set of object of the parent class in a sense : the set of all its instances is a subset of the subset of all the parent classes. Actually it contradicts the fact that anything can be a member of the UMBEL set as it's a set of something they all have something in commons : the fact to be instances of that class, and it associated definition. More philosophically, I don't like that any item who do not denotes a mathematical object is denoted as a mathematical set, this blurries the line between the properties of the model and the properties of the objects of the items. And last, your "group" notion is not well defined and does really not seem to fit the definition of "set" of UMBEL. A "set of intercoastal nerve" is essentially a composition relationship : any person has this group of nerve, they form a network in your body. It's more like it defines a composition relationship - this set defines a bigger object who are parts of it.
Alternatively we can define a class of nerve from "the intercoastal nerves in your body" and "the intercoastal nerves in my body" - those two classes would have actually instances IRL. Not very interesting actually as we would have to create such a class for everyone. A more interesting approach would have to create a "metaclass" : "class of intercoastal nerves of a human". The class of "the intercoastal nerves in your body" would be an instance of it as it is such a class. This would be very close to the definition of a mathematical set made of sets.
But do we really need this ? We totally can define an item intercoastal nerve system. It's part are intercoatal nerves - the "has part" relationship is definitely the right one to link objects - or system of objects - who are interlinked to form a bigger object. A model like
< intercoastal nerve system > has part (P527) < intercoastal nerve >
quantity (P1114) < x+-10 >
is enough then, isn't it ?
Actually I think we're dealing with an ambiguous meaning of "set" in english who can be understood as "system" in some expressions like "TV set". author  talk page 09:31, 2 October 2016 (UTC)
I filled subject item to remove possible ambiguity of "set" in English. As I said above the "has part" relationship doesn't imply that all elements are listed.
The word system implies interaction between parts.
I understand that both set theory and system theory aren't necessarily taught to everyone in school but it's still fundamental mathematics that's useful for describing the world. People without an mathematical education might use terms differently but when it comes to creating well-defined structured data it's very useful to use the formal mathematical notions of concepts.
Your claim that a member of a set is a subclass of the set is wrong. That's not how it works.
As far as math goes every group is a set but not every set is a group.
I'm instance of (P31) human (Q5). My heart is instance of (P31) heart (Q1072). We call neither of those items human class or heart class because it's implied that they are classes. Every abstract anatomical concept talks about classes.
Would you consider it helpful to rename this into "is a set of subclasses of"? ChristianKl (talk) 11:31, 2 October 2016 (UTC)
@ChristianKl: This has definitely nothing to do with the "group" notion in maths as there is no involved operation at all at sake here. :)
Your claim that a member of a set is a subclass of the set is wrong. That's not what I mean. Actually thinking about it all of this seems to be related to the disjoint union of (P2738) who involves completeness. Say that human intercoastal nerve is the class of all nerve. Say that human intercoastal nerve 1-2 / human intercoastal nerve 2-3 / ... (fantasy numbers) is the class of all intercoastal nerves beetween coast 1 and 2, 2 and 3 ... (resp) , whose instances are mine, yours and so on. Then the claim
< human intercoastal nerve > disjoint union of (P2738)
of (P642) < human intercoastal nerve 1-2 >
of (P642) < human intercoastal nerve 2-3 >
of (P642) < ... >
is enough to model the relationship (I think you) you intend to model. The completeness is obtained because a single statement is supposed to be complete. The subclasses partitions the parent class, and it is implied that any of the types intercoastal nerves beetween coast 1 and 2, 2 and 3 are types of intercoastal nerves because of the subclass relationship. author  talk page 12:33, 2 October 2016 (UTC)
Defacto Wikidata isn't complete. There's also no documented assumption that it is complete or a way to mark a certain item has listed all things that could be listed as part_of.
Given the broadness of part_of completeness is also really hard. You might trivially think that if you list all the states of Germany, you have all parts but you don't. Bundesrat (Q146138) has country (Germany). Country is a subsubproperty of part_of so if you want completeness you would have to say Germany has part Bundesrat (Q146138).
Furthermore there are cases where I want to speak about the "set of all human muscles" without knowing every muscle. Earlier this year we discovered a new muscle, so it's quite possible that we still discover new muscles in the future. That makes it worthwhile to have a way to specify a relationship between "set of all human muscles" and "human muscle". That's even more important when talking about subjects where we know less. ChristianKl (talk) 13:18, 2 October 2016 (UTC)
Unconvinced. We already have is a list of (P360) , and your new property just seem to add confusion to me, and the precise definition is still very blurry to me.
That's defined as Wikidata property giving Wikimedia list. I don't think this is a Wikimedia list even if it could be called a list in the more general sense. ChristianKl (talk) 14:04, 2 October 2016 (UTC)
Wikidata property giving Wikimedia list is kind of misleading, as actually it is used to give the type of the elements. But after some digging on the links on external identifiers - see my vote comment - what would actually model you example seem definitey to be has part claim. Your objection to this was As I said above the "has part" relationship doesn't imply that all elements are listed. which is why I was trying to find a solution for the completeness. Now you claim after all you don't want completeness, so either you get your thoughts straight, or I'll be totally sure you're making a fool out of me. author  talk page 15:35, 2 October 2016 (UTC)
I want completeness of items but I want to be able to specify the set in an intensional way without constructively describing it. Even muscles that aren't yet discovered are part of the set of all muscles. The set of all muscles is by definition complete even if I don't know all items in it.
In general the word list is used in English to specify a set that has an order. If I pick two elements of a list I can say which one comes first. Those Wikimedia lists might not have that property but they still imply that Wikimedia projects list the items in some order whereby every item of the list is explicitly listed.
I also generally considered "has_part" to be quite overloaded with means that when making statements about completeness it's useful to make more specific claims. ChristianKl (talk) 16:10, 2 October 2016 (UTC)

Also, a comment about an earlier stuff of yours We don't let muscle organ development (Q14863685) subclass "GO_class". => we definitely should not. They should be linked with instance of (P31), definitely not subclass of (P279). author  talk page 14:01, 2 October 2016 (UTC)

@ChristianKl: „Bitte übersetze das in Deutsch“. Danke --Succu (talk) 22:00, 1 October 2016 (UTC)

@Succu:Ich habe eine deutsche Übersetzung hinzugefügt. ChristianKl (talk) 22:44, 1 October 2016 (UTC)

### mirrors data from

Under discussion
Description The website automatically crawls or mirrors another website as it's data source Item websites Yahoo! Japan Talent Database (Q27048656) → Wikipedia (Q52)
Motivation

There are a bunch of websites that automatically mirror Wikipedia data. To prevent citogenesis it would be good to document when this is the case. Especially tools like the Primary Sources tool could benefit from having this data. ChristianKl (talk) 17:04, 2 October 2016 (UTC)

Discussion

Under discussion
Description institution where the subject teaches or taught Item employer (Q3053337) Q template Richard Dawkins (Q44461) → professor (Q121594) → University of Oxford (Q34433) Add it to scholars pages.
Motivation

The university or academic institution where a scholar taught is not just an "employer". Suzanne Beulemans 02:45, 3 October 2016 (UTC)

Discussion
• so is the intent to make this a sub-property of employer (P108)? Or is the issue that some people teach at an institution without actually being employed there (?) I'm not sure what this really adds. ArthurPSmith (talk) 19:55, 3 October 2016 (UTC)
• no, the intent is to make this an alternative to employer (P108) for scholars and researchers, as lecturing and conducting research at an academic institution isn't quite the same as just being "employed" there. employer (P108) is rendered as "worked for", which I think doesn't accurately describe advancing higher learning and engaging in academia. I would hope an "academic institution" item would yield something like, "taught at". Thank you for your time. Suzanne Beulemans 04:27, 4 October 2016 (UTC)
• well, at least in the case of your example, employer (P108) has been used to indicate Oxford is where Richard Dawkins works, and his position properties indicate what kind of work he does. I don't think we would want distinct properties for every different kind of work - for example people in charitable organizations may consider their service to be of a quite different character to bankers etc, but it is still where they work (unless they are volunteers, I'm not sure we have a property to capture volunteer service). What about school teachers - are they at an "academic institution"? Yet they teach, and do something different from what the custodial or administrative staff do. If you could give perhaps a clearer example of why this might be necessary we could perhaps understand better? ArthurPSmith (talk) 16:10, 4 October 2016 (UTC)
• Professors don't just "work" at academic institutions: upon being hired they must engage in "progress toward tenure" where said tenure is supposed to grant them academic freedom of publication of their research. Charitable organizations or banks don't grant tenure, that I know of. Likewise, upon retirement, charitable organizations or banks don't grant professors emeritus status, which makes them lifetime members of the institution. In that respect, a clearer example would be that professors are more like persons in the military who "serve" a country and are granted (or not) "honorable discharge" at the end of their service, which entitles them to many benefits. Thank you. Suzanne Beulemans 16:36, 4 October 2016 (UTC)
• well, lots of jobs have probationary periods of one sort or another. But I think your military analogy is a good one. We do have military branch (P241) and military rank (P410) which I think would be good examples to follow - so I'm thinking what might actually be helpful here is to have a detailed "academic affiliation" property that designates the specific department or institute (rather than the general top-level institution) and an "academic rank" indicating what type of academic position they hold (this seems to be a little different from professorship (P803) which is associated with particular named chairs or professorships). So for Richard Dawkins it would be perhaps <<academic affiliation>> "New College, Oxford University", <<academic rank>> "emeritus professor"? ArthurPSmith (talk) 19:51, 4 October 2016 (UTC)
• The fact that someone doesn't just work doesn't mean that they aren't employed. Given the current wording holding one guest lecture at a university would qualify a person to receive this property.
This property seems ill-defined to me so I  Oppose.
The purpose of a property also isn't to define words that resonator uses to describe it. ChristianKl (talk) 21:39, 4 October 2016 (UTC)
• Awesome, ArthurPSmith! The "academic affiliation" property designation perfectly defines what I had in mind. This would work well with an "academic rank" property indicating the academic position held. I don't think we have to worry about guest lecturers as they don't have an academic position, and they could fall within the "employer" property of that university, like librarians, etc. Thank you so much for your help. Suzanne Beulemans 17:08, 5 October 2016 (UTC)
position held (P39) is already perfectly capable of indicating the academic position that a professor holds. ChristianKl (talk) 11:55, 6 October 2016 (UTC)
Except it doesn't display properly in the Infoboxes. Either the Professor property shows, with the start and end date, who the person replaced, but you can't display the academic affiliation. Or the "Employer" property will display the University but not their title. Suzanne Beulemans 00:13, 8 October 2016 (UTC)
If it doesn't display in the infoboxes it's possible to modify the code of the infobox. I'm not sure why it can't be modified to show the data in this case.
Whether someone has the title of being a professor is independent from the person being employed or affiliated to an university. There are cases where a person keeps their title even when they leave an university.
Apart from that students also have an academic affiliation to their alma mater. ChristianKl (talk) 19:37, 12 October 2016 (UTC)
No, you can't display that "affiliation" for instance, with your name when attending a conference or when signing an article, like an emeritus can. Suzanne Beulemans 02:43, 13 October 2016 (UTC)
Oppose I've thought about this some more, and it seems to me everything here can be handled just fine by existing properties - we also have affiliation (P1416) that is useable in case the relationship is not one of employment. ArthurPSmith (talk) 19:33, 20 October 2016 (UTC)

### HQ photo

Under discussion
Description photo of the headquarter of a company Commons media file business enterprise (Q4830453) Wikia (Q17459) → File:Wikia and Wired Building location-9387.jpg import from Wikipedia wdfist image (P18), logo image (P154)
Motivation

The property image (P18) is used extremely arbitrary at companies. Sometimes a product is shown, sometimes it's duplicate to logo image (P154). The primary visual identification mark for companies is the logo and not the HQ, which is currently shown by the visual editor or the Wikipedia mobile app, when it occupies P18. This should be solved by this proposal.

Generally is the question whether entities should have set an image. The reuse of P18 is most easy if all P18 statemtn only show that object. Related images pollute the property somehow.--Kopiersperre (talk) 15:47, 6 October 2016 (UTC)

Discussion
• I would prefer if the property doesn't abbreviate HQ. ChristianKl (talk) 18:03, 6 October 2016 (UTC)
• Why not use P18, with a applies to part (P518) qualifier? Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 18:28, 6 October 2016 (UTC)
• I'd prefer an item for the headquarter, and property headquarter. TThere also should exist a property for the logo. If there is not, it shoudl be created. author  talk page 18:39, 6 October 2016 (UTC) Also, this is not a generic property but more like an organisation related one. author  talk page 18:44, 6 October 2016 (UTC)
•  Oppose. Use a qualifier. Thierry Caro (talk) 00:38, 7 October 2016 (UTC)

### Quora topic ID

Description identifier for a topic, on Quora External identifier https://www.quora.com/ https://www.quora.com/topic/\$1
Motivation

Quora (Q51711) is a popular Q&A site. Users can choose to subscribe to individual topics. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 18:26, 6 October 2016 (UTC)

Discussion
• By that notion, should we have a property for every StackExchange website as well? ChristianKl (talk) 18:31, 7 October 2016 (UTC)
•  Oppose Till me concern is answered. ChristianKl (talk) 12:07, 15 October 2016 (UTC)
• That is a question about a different property, not this one. I have no view on such a property, at this point in time, but would be happy to review a proposal if you would like to write one. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 11:25, 16 October 2016 (UTC)
• My concern is that allowing this property to be created would mean that there are many equivalent questions and answers sites and creating a property for every one creates a lot of data. ChristianKl (talk) 18:54, 16 October 2016 (UTC)
• Each property is debated on its own merits. Creating one does not "allow" another, nor does not doing so block it. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 21:01, 16 October 2016 (UTC)
•  Support. This one is OK. Thierry Caro (talk) 12:10, 8 October 2016 (UTC)

### 3D model

Under discussion
Description url of file with 3D model representing the item URL Common Objects, Open Source Hardware Objects Url to 3D file repository or 3D file directly like .step .stl .fcstd LM8UU Linear ball bearing (Q27148424) → [3] https://en.wikipedia.org/wiki/3D_modeling To add 3D model to Items, like common objects. http://opensourceecology.org/gvcs/
Motivation

As a part of wikiobject project meta:WikiObject Qupro (talk) 15:30, 8 October 2016 (UTC)

Discussion
•  Comment @Qupro: I fixed the template. Don't hesitate to ask a property creator for help with the template.
--- Jura 14:09, 9 October 2016 (UTC)
• @Jura example fixed too. @Andy Mabbett meta:WikiObject is in early stage. I am working on more examples and make it known. Thanks both.
• Sooner or later commons will host 3D files. Given that this is supposed to link to an external 3D file, it would have to be named appropriately. ChristianKl (talk) 17:40, 9 October 2016 (UTC)
• Do we know about such plans or is it just a general thought?
--- Jura 11:19, 11 October 2016 (UTC)
We do know about such plans: https://phabricator.wikimedia.org/T3790 ChristianKl (talk) 14:33, 12 October 2016 (UTC)
Interesting read. Looks like it isn't going to happen anytime soon.
--- Jura 12:12, 19 October 2016 (UTC)

Description the channel, network, website or service that broadcast this item on tv, radio or online. broadcaster (Q15265344) Item instance or subclass of broadcasting program (Q19968906), event (Q1656682) instance or subclass of broadcaster (Q15265344), website (Q35127), film distributor (Q368290), video hosting service (Q559856), service on internet (Q1668024), Online service provider (Q1641122) original network (P449)
Motivation

The need for this was identified in two independent but concurrent discussions at project chat: WD:PC#Top goalscorer and tv network (permalink) and WD:PC#original network (P449) (permalink).

It is intended as a superproperty of original network (P449) (which seems to be specifically important for drama series) as has a narrower domain and a single value restriction. This property will have many values in some cases (e.g. Olympic games) and those values should (but not must) be qualified with valid in place (P3005) and can also take other qualifiers such as applies to part (P518) live television (Q431102) when live coverage and highlights are carried by different broadcasters/channels. The allowed values for this property are intended to be broad to cope with the different structure of particularly TV broadcasting around the world (brought up as an issue in the second linked discussion). The hierarchy of internet broadcasters/broadcasting networks on Wikidata does not appear to be well structured at present hence the several entries here and I still expect to see violations of that until it's cleaned up. Thryduulf (talk) 16:25, 11 October 2016 (UTC)

Discussion
•  Support --Edgars2007 (talk) 18:12, 11 October 2016 (UTC)
•  Support Sjoerd de Bruin (talk) 18:43, 11 October 2016 (UTC)
•  Support Runner1928 (talk) 19:19, 11 October 2016 (UTC)
•  Support I'd suggest this should be used with start/end date qualifiers. Possibly also a location qualifier of some sort as well. Danrok (talk) 01:41, 12 October 2016 (UTC)
• I've used valid in place (P3005) as a location qualifier in the examples, and suggest that including that (or another location qualifier if there are better ones) is added as a constraint. Start time and end time qualifiers should be used for drama series but I don't want to add them as a requirement as they wont add anything for live sporting events for example. Thryduulf (talk) 02:03, 12 October 2016 (UTC)
•  Oppose Use original network (P449), relabelled if necessary, with qualifiers. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 09:24, 12 October 2016 (UTC)
• @Pigsonthewing: As noted in the linked project chat discussion and directly above, original network (P449) is not suitable for this use as it has a narrower domain and single value restriction that do not match this proposed usage. As also noted above original network (P449) is widely used for its intended use and that is apparently an important element of data for drama series (and possibly other uses). Further the proposed constraint requiring a geographical qualifier is not necessary for the intended and actual use of original network (P449). Together this means that if we did convert it into a "broadcast by" property we would be required to:
• Remove the existing constraints
• Change the label and description in every language
• Determine a suitable qualifier to indicate that the value is for the original network
• Add that qualifier to each of the current uses (just shy of 28,000)
• Rewrite any and all templates and queries that use the property in it's current form to use the qualifier-based form to extract the same information
• Write and monitor a complex constraint to ensure that it has a single value qualified as the original network, regardless of the total number of values
• Write and monitor a complex constraint to ensure that values not qualified as the original network are qualified with valid in place (P3005) or another geographical qualifier.
Or we could avoid all of that by not trying to get one property to fulfil multiple roles. Thryduulf (talk) 12:27, 12 October 2016 (UTC)

──────────────────────────────────────────────────────────────────────────────────────────────────── Yes, I read the linked discussion; I don't agree with the comment you cite; nor that a single value qualifier is correct even for the current limited usage; nor that it is necessary to "add [a] qualifier to each of the current uses"; but even if it is, that's a trivial task for a bot or for QuickStatements. We shouldn't use new properties as sticking-plasters for problems with existing properties or their constraints. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 17:00, 12 October 2016 (UTC)

• I'm not going to oppose (yet), but why does "original network" exist? Why this proposal? My suggestion is that these are both duplicate properties to "publisher". Can someone suggest otherwise? --Izno (talk) 15:44, 12 October 2016 (UTC)
• Good point. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 17:00, 12 October 2016 (UTC)
• It is currently being used with a single value restriction to represent the single network that originally broadcast the item. Templates, etc that use the property will be expecting a single value and will need some way to determine which is the single original network if the property changes to accepting multiple values. I don't know why it is important, but given that it is in very widespread use for it's current role the evidence strongly points to it being important, and consequently some alternative way of representing the same data will be needed. These are questions that should be asked of the people who are currently using the property rather than changing it under them (yet none of you have even attempted to let them know change is being proposed). The people proposing the change are the ones who need to demonstrate that no functionality will be lost. This new superproperty is being proposed because original network (P449) is too specific and not suitable for broadening. Thryduulf (talk) 17:02, 13 October 2016 (UTC)
A date qualifier (mandatory) seems sufficient to me--and where two networks broadcast it simultaneously (for which "original network" completely fails to account), what then? Those wikis are still out in the dark with a "single value" constraint. --Izno (talk) 19:47, 14 October 2016 (UTC)
• I have left a note at Property talk:P449. There are 5 Russian projects noted as using that property who need to be informed, but a Russian speaker needs to do that. Thryduulf (talk) 17:06, 13 October 2016 (UTC)
•  Oppose. Relabel original network (P449). Thierry Caro (talk) 06:37, 13 October 2016 (UTC)
•  Support I agree with Thryduulf, P449 should not be meddled with (and it is used in more Wikipedias than what's listed on the talk page). However, the single value constraint is quite meaningless on P449 as well, as a series may have several original broadcasters over its course, but they definitely need to be dealt with separately from simulcast, encore and foreign market broadcasters. (We've been having a similar discussion about title (P1476)). – Máté (talk) 07:35, 14 October 2016 (UTC)
•  Support original network (P449) may need some tweaking but it seems it is naturally only a sub-property of this proposal, and stretching it to include all cases represents a loss of information. ArthurPSmith (talk) 19:07, 14 October 2016 (UTC)
• We need to change this practice of "notify wikis first, then decide". It completely calcifies our property change process. I will likely be starting an RFC or some such. --Izno (talk) 19:47, 14 October 2016 (UTC)
• People who are using a property need to have a voice in discussions about whether and how to change that property, and they cannot be expected to follow every discussion on Wikidata on the off-chance that it might be relevant to their work. Thryduulf (talk) 22:38, 14 October 2016 (UTC)

### Storm category

Under discussion
Description – (Please translate this into English.) Saffir–Simpson hurricane wind scale (Q205801) Item Q27058656 (Hurricane Matthew) > 4 (start date = *, end date = *) National Hurricane Center Adding hurricane category to hurricanes. Not currently.
Motivation

Hurricane strength category on Saffir-Simpson Scale

Discussion
•  Oppose This is a classification property, use instance of (P31) or something like that. author  talk page 19:42, 12 October 2016 (UTC)
• But you shouldn't use start and end dates with instance of (P31). ArthurPSmith (talk) 14:09, 13 October 2016 (UTC)
Yes, it does not seem like a problem to me. author  talk page 16:33, 13 October 2016 (UTC)
•  Comment I don't think this is restricted to hurricanes, is it? Or is a different scale used for tropical cyclones in other ocean basins? ArthurPSmith (talk) 14:11, 13 October 2016 (UTC)

### formal citation

Under discussion
Description a formal citation string, if possible in the way the cited source prefers to be cited Monolingual text References Thomas Hanna (Q4495505) place of birth (P19) Waco (Q128244) Reference: → en:"Texas Birth Certificates, 1903-1935," database with images, FamilySearch (https://familysearch.org/ark:/61903/1:1:VH82-Q2T : 4 December 2014), Thomas Louis Hanna, 21 Nov 1928; citing Waco, Mclennan, Texas, United States, certificate 95667, Bureau of Vital Statistics. State Registrar Office, Austin; FHL microfilm 2,240,748."
Motivation

There are source that specify how they want to be cited. In the case of the FamilySearch records the information of the citation isn't easily expressible with existing Wikidata properties and it's useful to have the data as a text-string even if the content of the text string isn't structured data.

In the spirit of making it easy to reference statements I also think this property is useful for people who want to cite books but don't want to go through the effort of creating an item for the book.

I'm not sure about the name of this property. If you have a suggestion for a better name, feel free to suggest a better name. ChristianKl (talk) 19:15, 12 October 2016 (UTC)

Discussion

Many citations have special typography, such as italics for book titles or boldface for volume numbers. I don't think we can reproduce this in a property value. In unusual cases the loss of special typography might even make it difficult to identify the source. Jc3s5h (talk) 19:51, 12 October 2016 (UTC)

Oppose There is no good reason to encourage laziness of importing unstructured datas into a structured project, this would defeat the whole purpose. The citation some projects provide are there to be complete about the information, but in no world someone can impose a very specific way to cite him. Usually publications have there own formats for citation its customary to follow, not the way of the cited - this would make bibliographies totally inconsistent. author  talk page 20:31, 12 October 2016 (UTC)
@TomT0m: The current form of laziness is that people often don't contribute any sources. When Wikipedia's want sourced data's I think most of them would be happy if having the text instead of having nothing. Given that providing citation-backed data seems one of the main constraints from getting Wikipedia to integrate Wikidata I think it's worthwhile to make contributing sources easier. It's also possible to extract data from citations that are written in the common form of academic citations via a bot. ChristianKl (talk) 20:38, 12 October 2016 (UTC)
It's the worst possible solution. There is citoid and a currently WMF funded project to make this easier, I'd hope they will solve it properly. There is billions of possible format for citation. Consider your example : why on earth would we replicate the datas present on the linked page in an unstructured way ??? It's probably generated from their datas ! we would almost have statement from each sourced by the raw URL ... this definitely would be absurd and to much effort to copy paste this string. This is for paper citations. author  talk page 20:45, 12 October 2016 (UTC)
Because the current solution is to not record the data at all. ChristianKl (talk) 12:26, 13 October 2016 (UTC)

### ASCII code

Under discussion
Description number of the character in the ASCII character set ASCII (Q8815) Number role (Q1707847) integers in [0;255] none Wikidata property  <  97 >
Motivation

common character set, a subset of unicode code points

(Ajoutez ici vos motivations pour la création de cette propriété) author  talk page 20:10, 12 October 2016 (UTC)

Discussion

### has code

Under discussion
Description code for Number Q184759 none has code search < 0xE9 >  ; has code search < 0xE9 > codes for search Ascii code (Par exemple: vérifier les autres propriétés afin d'être cohérent, collecter des données, automatiser un lien externe, etc.) -->

code for

encodes
Description qualifier to link a code to the item of the character it encodes, if relevant. Number Q184759 none see above example
Motivation

Compared to the other attempt, this property is restricted in domain and is in reverse sense : from character set encoding to the caracter, and not from character to code. This mean a letter can has as many code we want without being polluted by as many statements as there is character sets. There is also a qualifier to link to the item of the character if relevant. author  talk page 20:26, 12 October 2016 (UTC)

Discussion

Comment I imagine that the character set properties would get quite polluted themselves, most especially for a set like GB 2312 (Q1421973), Big5 (Q858372), Shift_JIS (Q286345) or KS X 1001 (Q489423). As for pollution of the items for the characters themselves--if anything, numerous codepages may end up sharing a codepoint for a single character, so the main risk is just a lot of qualifiers on a single code. When Wikidata-Wiktionary integration goes live, it may become much more beneficial to add character codes to items (especially Chinese characters) rather than to character sets. Mahir256 (talk) 01:57, 15 October 2016 (UTC)

Good point. Maybe another relationship should be used to minimize redundancy like "share/duplicates codes with/of", qualified by intervals of shared codes. Many ISO european standards charater sets shares a lot with ascii, that would avoid a lot of claims if we avoid duplicating all the shared codes with ascii. Do you know if there is a similar situation on oriental character sets ? Arabian and Cyrillic characters sets should be in a similar situation as there is about the same number of letters in arab alphabet and in latin one. Japan and China shares a lot of characters ... author  talk page 08:28, 15 October 2016 (UTC)

### Category for people deaths by this disease

Under discussion
Description link to Wikimedia category for people deaths by this disease. disease (Q12136) Item Wikimedia category (Q4167836) from User:Mr. Ibrahem/disease category for people who died here (P1465)
Motivation

I want to add the categories to many items so I can use it for autocategorization Mr. Ibrahem (talk) 13:21, 13 October 2016 (UTC)

Discussion
• @Nikki: Well, what about a new property with name "Related Categories" just like website account on (P553), We can use it to do any type of autocategorise with specific item in the qualifier.
for example I will create new item with name "Category for people deaths by this disease" and I will use it in the qualifiers with property: applies to territorial jurisdiction (P1001). and this way can work on award holders, deaths places, births places, and may things.?--Mr. Ibrahem (talk) 20:13, 15 October 2016 (UTC)
I support this latter suggestion, except that the property used as a qualifier should not be applies to territorial jurisdiction (P1001). Thierry Caro (talk) 13:36, 19 October 2016 (UTC)
•  Comment Maybe a new property for properties would work ("stroke" (Q12202) > "Property for autocategorization" > "Cause of death" (P509)). A qualifier on that could link to the applicable category ("Category:Deaths from stroke (Q6509490)"). So if an infobox an article relates to an item with P509:Q12202, it could add the article into the category found on Q6509490. This is similar to what some wikis do based on date/place of birth/death.
--- Jura 13:45, 19 October 2016 (UTC)
Comment Actually, the other way round categorization might be easier. A statement linking to the category and possibly a qualifier to indicate the property.
--- Jura 16:49, 19 October 2016 (UTC)
• The other reason "auto-categorization" confuses me (besides my above oppose) is that we should be relying on humans with their actual causes of death filled here. --Izno (talk) 17:37, 19 October 2016 (UTC)
• This actually makes categorization rely on that. Which can be seen as a step forward compared to suggestions to query Wikidata and than add "manually" categories.
--- Jura 17:41, 19 October 2016 (UTC)
• Then it's not necessary per my original oppose. --Izno (talk) 18:06, 19 October 2016 (UTC)
• Can you provide us with a proof of concept? You can use any article on any wiki (with P509 set) to illustrate how your approach can work (either in LUA or with #property).
--- Jura 18:15, 19 October 2016 (UTC)
• That seems clearly not to be the user's intent. Even if it were, I would disagree with that as a useful scope for Wikidata. Propping up the shitty category system is not our goal at Wikidata. --Izno (talk) 18:32, 19 October 2016 (UTC)
• I think from the proposal alone it's not entirely clear what they were looking for. I assume that you agree they couldn't use P971 to categorize articles at Wikipedia and a new property (in one way or the other) would be needed. Personally, I think it works well for cawiki and a few others.
--- Jura 18:36, 19 October 2016 (UTC)

### "Since" and "Until"

Under discussion
Description It defines the item since which the proposition is true Item start time (P580)
Under discussion
Description It defines the item until which the proposition is true Item end time (P582)
Motivation

Some statements cannot be defined temporally with start time (P580) and end time (P582). That's why I think it will be useful to have two generic properties to own the same meaning of start time (P580) and end time (P582) but the data type is not a time value but an item. See the example to be more clear ★ → Airon 90 15:49, 14 October 2016 (UTC)

Discussion
•  Support ChristianKl (talk) 17:47, 14 October 2016 (UTC)
•  Support. I can forsee usage of this for historical figures for when we know things about them relative to events but don't know the date of that event. Thryduulf (talk) 16:35, 15 October 2016 (UTC)

### is a historic designation for

Under discussion
Description historic, previous, or outdated designation for an object Item besonders geographische Objekte, aber auch andere Cöln (Q18019200) → Cologne (Q365) Cöln ist eine historische Bezeichnung für Köln
Motivation

Die bisher verwendete Eigenschaft follows (P155), mit welcher zum Beispiel Constantinople (Q16869) und Istanbul (Q406) verknüpft sind, ist definitionsgemäß nur zulässig, wenn keine Identität zwischen geographischem Vorgänger- und Nachfolgeobjekt besteht, was in diesem Fall jedoch gegeben ist. Einem Konzept von Vorgänger und Nachfolger entspricht auch nicht, dass gerade in historischen Zeiten verschiedene Bezeichnungen nebeneinander existierten. Die inverse Eigenschaft „ist eine rezente/gegenwärtige/aktuelle Bezeichnung für“ wäre dann ebenfalls anzulegen. Volmar (talk) 11:19, 21 October 2016 (UTC)

Discussion
Gibt es überhaupt so viele solcher Seiten? Ich weiß nicht, ob diese Eigenschaft eine gute Lösung wäre, weil "Cöln ist eine historische Bezeichnung für Köln" nur für Deutsch stimmt. Vielleicht wäre es besser, official name (P1448) oder native label (P1705) zu benutzen, mit end time (P582) und subject of (P805) als Qualifikatoren (d.h. der Name war "Cöln" bis Enddatum und Cöln (Q18019200) beschreibt das). Es sei denn, ich habe die Benutzung von P805 komplett falsch verstanden. :) - Nikki (talk) 19:40, 21 October 2016 (UTC)
Oppose In general Wikidata items are about concepts and not to names of concepts. You can use official name (P1448)/native label (P1705) with start time (P580) and end time (P582). You can use said to be the same as (P460) to link the two items together. ChristianKl (talk) 12:25, 22 October 2016 (UTC)