Wikidata talk:WikiProject Names/Archive/3

This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Deletion

Hi,

Notified participants of WikiProject Names

Y'all may want to take a look at Wikidata:Requests_for_deletions#Q23005924 (it seems pretty forward but maybe there is an exception or a special case here).

Cheers, VIGNERON (talk) 20:03, 9 February 2021 (UTC)

Brazilian multiple surnames

Here is another question about Brazilians' surnames. It's very common to register a person with both mother and father's family names, and many Brazilians sign their intellectual production using all their surnames (that maybe be two, three or four names, for example). Any suggestion will be very appreciated to find the best way to use P734 for a person's multiple surnames. I've posted similar question on the discussion page, I got good answer from @billinghurst:, but I was instructed to come ask here at the project. Thanks. AmarilisMGC (talk) 18:40, 14 February 2021 (UTC)

Amarílis I don't have an answer, but I think we need more guidance/examples. Example: Oscar Niemeyer (Oscar Ribeiro de Almeida de Niemeyer Soares Filho), should this use P734 + serial ordinal P1545 for his apparently four-part compound surname? And, for this example should his P9139 first family name in Portuguese name be "Ribeiro"? Thanks, Paul / Prburley (talk) 16:15, 17 February 2021 (UTC)
- Hi, @Prburley:. Thank you. Yes, Ribeiro is Oscar's first family name in Portuguese (Filho signs that his name is the same as his father's. For this type of relationship we have in Portuguese Junior, Sobrinho, Neto depending on the relative). If I may add another question: Besides P734, would it be a good practice also use P9139, and P1559 and P1477?

Other examples of compound surnames are: Neves Gonzaga (Q465902), Cavalcanti de Albuquerque e Melo (Q1793637), Machado de Assis (Q311145), Piccolotto Siqueira Bueno (Q93793869), Faggin Pereira Leite, and Refinetti Rodrigues Martins.AmarilisMGC (talk) 18:25, 18 February 2021 (UTC)

@Prburley:, I'm sorry. I have to rectify my message above. Just trying to make it clear, these examples I've given you aren't compound surnames (as we see at https://en.wikipedia.org/wiki/Category:Compound_surnames), maybe we could call them multiple or "complex" (as made of different parts), because they are formed by the combination of surnames from both mother and father's, and in many cases from spouses'.AmarilisMGC (talk) 18:25, 18 February 2021 (UTC)

Maybe @DaddyCell: who proposed first family name in Portuguese name (P9139) wants to comment. --- Jura 07:36, 2 March 2021 (UTC)

Hello! I will try to help.

Portuguese and Brazilian surnames are a bit complex. Traditionally, the first surname comes from the mother and the second, from the father. Those surnames, indeed, are the second ones of each parent, that is, are the ones of the respective grandfathers. So, the second surname is usually used as the main one, and people can present themselves mentioning only the second surname.

Until here it may be simple. But, traditionally as well, people can receive both surnames from the mother in first place, and both surnames of the father in the second place, having a kind of compound surname. Even though, it is flexible, as those "compound" surnames may not be given together to their children.

It can be a bit more complex. Traditionally some women have changed one or both of their surnames when they got married, adopting one or both surnames of their husbands. There is flexibility in the position of those new surnames, and even, they can be added intead of changed. Nowadays, it is also permited to the husband to do the same, and change (or add) his surname(s) by the one(s) of his wife.

Finally, nowadays the fist surname can come from the father and the second from the mother. So with all this, it may be very messy when we try to put order to all this information.

So my suggestion is to use all the posibilities for surnames in Wikidata: family name (P734) as the second (and main) surname(s), first family name in Portuguese name (P9139) as the first surname(s), and using Q1376230 and Q30232378 when clarification is needed. In cases of complex surnames, it is posible to add 2 or more values as well. So if a person has 4 surnames, I would add the second 2 in family name (P734) and the first 2 in first family name in Portuguese name (P9139).

I don't know if it helped. Regards! --DaddyCell (talk) 10:59, 4 March 2021 (UTC)

St Leger vs. St. Leger

Notified participants of WikiProject Names

Should the family names "St. Leger" and "St Leger" be separate items? I believe the distinction is US vs. UK spelling convention. Thanks! - PKM (talk) 22:10, 21 February 2021 (UTC)

Nope. Solved. Klaas `Z4␟` V: 08:11, 22 February 2021 (UTC)

Why? --- Jura 09:52, 22 February 2021 (UTC)

@Jura1: Though I don't agree with the way Klaas modified data of St. Leger (Q94144228) while a discussion is in progress, I tend to agree that "St" without dot is plain misspelling. Currently, only 2 linked items use that spelling: John St Leger (Q6258753) and Anthony St. Leger (Q89487309). The first one differs from a reliable source and its enwp article also uses the dot; the second item is contradicted by its ORCID and by the first paper I opened to check. There are a few more items with no linked family names (William St Leger (Q18811424), St Leger St Leger, 1st Viscount Doneraile (Q16830036), Anthony St Leger (Q4121000)...) that might be worth checking, though. Of course if one person DOES use spelling with no dot, a distinct family name item with an undotted native label (P1705) must be created, IMHO. Lαδδo chat ;) 14:42, 22 February 2021 (UTC)

See your own talkPage, colleague Jura1 and the one of Q94144228 (St[.] Leger). Revert without proper explanation is unnice, sorry. Klaas `Z4␟` V: 11:33, 6 March 2021 (UTC)

Aren't both just abbreviations of some full name like "Sergeant"? --Infovarius (talk) 21:38, 8 March 2021 (UTC)

I believe it's an abbreviation of "Saint". St. John, St. Louis, etc. It certainly isn't "Street Leger". I believe both should be merged (as should St. John (Q7593634) and St John (Q65953328), etc.), as the 'differences' appear to be be catalog mistakes and/or regional variants, of which both forms can be found depending on whichever style some database or encyclopedia fancies. We don't separate Mister (Q177053) due to regional variations in punctuation. -Animalparty (talk) 21:59, 10 March 2021 (UTC)

Well, names are all about regional variations. However, it's unclear if this applies to Leger. If we follow @Laddo:'s reasoning, we can do with one item for now. --- Jura 10:59, 16 March 2021 (UTC)

surname from a specific origin (Q33283501)

Hello everyone,

I don't know if it's been discussed before. I was trying to translate some element labels that were surname from a specific origin (Q33283501). I noticed that some of them were distinguished by language of work or name (P407), some by facet of (P1269), some by country of origin (P495), some by country (P17).

I find it very messy and I think some elements are simply duplicates.

Just as an example:

Do you think there should be a rule? Which is the correct use of surname from a specific origin (Q33283501)?

This is the query I started with:

Items used: surname from a specific origin (Q33283501)

Properties used: subclass of (P279)   , language of work or name (P407)   , facet of (P1269)   , country of origin (P495)   , country (P17)   , instance of (P31)

SELECT ?item ?itemLabel ?lang ?langLabel ?partOf ?partOfLabel ?origin ?originLabel ?country ?countryLabel
WHERE {
  SERVICE wikibase:label { 
    bd:serviceParam wikibase:language "[AUTO_LANGUAGE],en". 
  }
  { 
    ?item wdt:P279 wd:Q33283501.
    OPTIONAL { ?item wdt:P407 ?lang. }
    OPTIONAL { ?item wdt:P1269 ?partOf. }
    OPTIONAL { ?item wdt:P495 ?origin. }
    OPTIONAL { ?item wdt:P17 ?country. }
  } UNION {
    ?item wdt:P31 wd:Q33283501.
    OPTIONAL { ?item wdt:P407 ?lang. }
    OPTIONAL { ?item wdt:P1269 ?partOf. }
    OPTIONAL { ?item wdt:P495 ?origin. }
    OPTIONAL { ?item wdt:P17 ?country. }
  }
}

Try it!

Many thanks, --Luca.favorido (talk) 07:00, 20 March 2021 (UTC)

Delete all of them? We should really not be making this entire class of items when a statement of e.g. country of origin (P495) on the items of family names themselves could suffice? I hope I'm not missing something obvious here. The obvious thing I'm missing here is that there are Wikipedia articles about e.g. "polish surnames"? And we have been creating this hierarchy to put these items in? I do hope it suffices to use country of origin (P495) on family name items instead of making them instances of a surname from a specific origin (Q33283501).

+1 for making the labels consistent. --Azertus (talk) 09:32, 28 May 2021 (UTC)

Same spelling, different origin

What to do in this case? Example: Caucasian Eldar (e.g. Azerbaijani has Latin spelling) vs. Scandinavian Eldar. Current system presumes they should be in the same item, but it loses info about origin of the names. --Infovarius (talk) 19:58, 24 May 2021 (UTC)

One aspect to consider: Having only separate items for separate origins would make it necessary to know for every named person which origin their name has. Maybe the etymologies could be stored in Lexemes? Julian Jarosch (digicademy) (talk) 13:29, 26 May 2021 (UTC)

@Infovarius: I'm not sure how we could deal with 2 items in practice. I had the same thought for Berger (in French it means "shepherd" but in German and other close languages it a totally unrelated name meaning "mountainer") but in the vast majority of case, there is no way to know if it's the French name or the Germanic one. That said, the current situation is suboptimal and should be improved (Q1260304#P31 is a data-nightmare, clearly not something that should be in P31, P642 is a bad qualifier, why two values as qualifier and the statement is only true for the French name, in bonus: it's redundant with P138 which is closed to the Lexemes idea proposed by @Julian Jarosch (digicademy):). Cheers, VIGNERON (talk) 16:42, 10 June 2021 (UTC)

Perhaps a short summary of the approach in Digital Dictionary of Surnames in Germany (Q61889795) might be somewhat helpful? Let’s take the name Elben (Elben (Q37299927)) as an example entry: Under »Etymologie« (etymology) separate hypotheses or sources of the name are listed. (»Hauptbedeutung« is the main, ascertained interpretation; »In Einzelfällen« means these etymologies may be applicable in a few cases.) Each etymology may have a different derived from lexeme (P5191), a different class of name (occupational surname (Q829026), locational surname (Q1714800) …), possibly different languages of origin, different area of origin, different sources … All these bits of information would correspond to statements in Wikidata, I’d say.

I don’t think the statements could be properly grouped by etymology unless each etymology has a separate Wikidata element? (Lexemes seem a good fit to me, but who knows!)

Julian Jarosch (digicademy) (talk) 17:12, 10 June 2021 (UTC)

Name class

I see a recommendation in main page of the project: "use only basic classes but not Belarusian name (Q35002801) or similar". But why? Why are we afraid of using more exact classes and try to move such values to some other property like has characteristic (P1552)? @1Veertje, Jura: For me, P31 is better to query than to seek which other property to use. --Infovarius (talk) 21:13, 9 June 2021 (UTC)

We are not afraid, it's just bad practice. instance of (P31) defines the nature of an item, it does not defines every characteristics of it. Values used in P31 should stay the most general possible, and other properties should be used to differentiate the items. P31 is not a garbage collector if you don't know how to properly write a query using several properties. If you start putting every characteristics of an item in P31, you will end up with humans instances of items like "French doctor born in 1950". It would be unmaintainable and a nightmare to query.

In the case of names, there are plenty reasons not to use items like Belarusian name (Q35002801) in P31:

First, why do you want to differentiate names in P31 by their origin? Why wouldn't you differentiate them by their writing system (P282) or other characteristics? You can thus imagine several combinations. It's a slippery slope and it would become unmaintainable.
Second, the origin of names is something fuzzy. You will find sources for the same name stating several different origins. Borders (political, ethnic, or others) evolve all the time in history, with complex definitions (should Belarusian be about the current country, an ethnic group, or anything else?). We even don't have sources for most of them. I think values in P31 should be unique and stable.
Third, it's easier to query instance of (P31) = family name (Q101352) and use another property to narrow your search, rather than to try to use the class structure (like ?name wdt:P31/wdt:P279* wd:Q101352), which is hardly maintained in Wikidata and source of many errors. See also KISS principle: in general, I think that a characteristic of an item should be matched to a specific property; in other words, a property should have a limited scope, but would be easy to maintain.

Cheers, — Envlh (talk) 16:17, 10 June 2021 (UTC)

@Envlh: I understand you. But "Values used in P31 should stay the most general possible" in general is not true. Otherwise we wouldn't have all these "films, books, albums" when we have a general work (Q386724). --Infovarius (talk) 12:26, 11 June 2021 (UTC)

There is a specialized property for this: language of work or name (P407)=Belarusian (Q9091) --- Jura 13:38, 3 July 2021 (UTC)

pap: prome nomber maskulino vs. di prome nomber maskulino

@Edoderoo: In male given name (Q12308941) under description for Papiamento it says “prome nomber masculino” (entered by yourself), but your bot writes “di prome nomber masculino”. What's right now? --HarryNº2 (talk) 06:20, 18 August 2021 (UTC)

Good question. I will have to ask a native speaker, which of the two is preferable, and change it from there. Edoderoo (talk) 08:29, 18 August 2021 (UTC)

It is being worked on... on getting an answer which one will be more suitable. Edoderoo (talk) 11:50, 18 August 2021 (UTC)

Should be "nòmber dilanti maskulino". In Pap-CW an individual's first name is "nòmber dilanti" and her/his last name would be "fam". Both "promé nòmber" and "di promé nòmber" would gloss as "first name" in a sequence of listed names rather than as the first name of an individual. ObaTango (talk) 12:25, 18 August 2021 (UTC)

Osypenko (Q108524859) vs Osipenko (Q12798037)

Please have a look: Osypenko (Q108524859) vs Osipenko (Q12798037). Would someone with knowledge of Cyrillic please separate the two data records properly and fill them in. Thx, --HarryNº2 (talk) 08:16, 16 September 2021 (UTC)

Here is the same problem: Harmash (Q108550378) vs Garmash (Q4133810). --HarryNº2 (talk) 08:38, 16 September 2021 (UTC)

In general, we would have just one item per script (generally Latin, but here Cyrillic). Are there uses that would require separate items?

@Бучач-Львів: FYI.

It's about different pronunciation and transliteration. See similar in Latin. --Infovarius (talk) 16:03, 17 September 2021 (UTC)

See Santos (Q20813204) vs. Santos (Q20813183) and José (Q29043257) vs. José (Q29043256). --Infovarius (talk) 16:05, 17 September 2021 (UTC)

@HarryNº2:. Hi! I don't see any problem. On the contrary, this is exactly how it should be, because these surnames, in addition, are also write differently in Polish - ukr. Osypenko and rus. Osipienko, Harmasz - Garmasz.
@Infovarius:. Your ping didn't work ... --Бучач-Львів (talk) 13:55, 16 November 2021 (UTC)

Notified participants of WikiProject Names @Бучач-Львів: Now there are also two items for "Бондаренко", as the pronounciation/Polish transcription differs: Bondarenko (Q31033336)/Bondarenko (Q102273568).

Should the basic principles be changed to have an own item for each name string/pronounciation combination? E.g. should Paul (Q12668666) be split up in "Paul in English language", "Paul in German language", "Paul in French language", "Paul in Dutch"? - Valentina.Anitnelav (talk) 13:28, 16 November 2021 (UTC)

I'm personally against creating more Q-items only because of different pronunciation, because this may differ even within dialects e.g. the name Ger(h)ard in Dutch, German and Flemish. Klaas `Z4␟` V: 10:12, 17 November 2021 (UTC)

@Klaas van Buiten:. Here we are talking primarily about different writings in different languages. Besides, if things are somehow different from one another, then they are different. Whereas we can artificially unite something? --Бучач-Львів (talk) 12:49, 18 November 2021 (UTC)

I'm a mergist so when as in the example Osypenko/Osipenko there are two different transcriptions for Осипенко we can use the field 'also known as'
Kind regards, Klaas `Z4␟` V: 18:30, 19 November 2021 (UTC)

Wingfield: given or family name for "Penrhyn Wingfield Coussens"?

For Penrhyn Wingfield Coussens (Q109349440), is "Wingfield" a given name (which doesn't yet exist), or is it a family name (Wingfield (Q37436734)? His mother's name was Maria Wingfield, so it could be a double family name (Q29042997)? His books were usually under "Penrhyn W. Coussens". Inductiveload (talk) 16:41, 1 November 2021 (UTC)

This happens a lot in English naming. In general, if both names are passed on to children consistently, I'd consider it a double surname. Otherwise it's a given name based on a family name (which occurs commonly also as a "first" name in both the UK and US). The form "Penrhyn W. Coussens" pretty much supports that it's a given name. - PKM (talk) 21:58, 16 November 2021 (UTC)

actually male given name that is exceptionally also used for women

In the German-language Wikidata forum, an IP in relation to the data object Jodie Foster (Q41351) noted that the middle name given there is not entirely correct or confusing. Because Jodie Foster's second given name is Christian and according to the data object Christian (Q18001597) it is a male given name. Therefore, actually, this data object cannot be used with Jodie Foster's data object. Is there a basic procedure in such a special case? --Gymnicus (talk) 08:07, 19 January 2022 (UTC)

You can use Christian (Q18001597), e.g., (from the enwiki page) Christian Maclagan (Q5109875), Christian Pitre (Q16734573), Christian Ramsay (Q18670910), Christian Serratos (Q235719). The claim "male given name" doesn't mean it's never used by females. Ghouston (talk) 09:31, 20 January 2022 (UTC)

The question is if the claim is even correct. en:Christian (given name) claims that the name was used as a female name in the 17th and 18th in Scotland and is being used as a female name in the present. Emu (talk) 10:36, 20 January 2022 (UTC)

Names can often change usage in this way. I don't think claims like "male given name" are a good thing, as discussed previously. Ghouston (talk) 10:44, 23 January 2022 (UTC)

Qualifiers for given names and surnames - establish a guideline

Notified participants of WikiProject Names Hi! As of now, Wikidata:WikiProject Names doesn't make clear which properties should be used as qualifiers for given name (P735) and family name (P734), so many different uses have been tried. As of now, both properties have a long list of property constraint (P2302)allowed qualifiers constraint (Q21510851) and for most of them I think the use is relatively clear; however, I would like to reach a standard for some cases:

Property	Qualifier-value	Qualifiers used	Query	Most used qualifiers
given name (P735)	usual forename (Q3409033)	six	https://w.wiki/4tkT	in nearly all cases object has role (P3831)
given name (P735)	name at birth (Q2507958)	six	https://w.wiki/4tkR	mainly object has role (P3831), applies to part (P518)
given name (P735)	religious name (Q1417657)	six	https://w.wiki/4tkP	mainly object has role (P3831), applies to part (P518), subject has role (P2868)
family name (P734)	maiden name (Q1376230)	eleven	https://w.wiki/4tkM	criterion used (P1013) in 3/4 cases and object has role (P3831) in 1/4 cases
family name (P734)	married name (Q30232378)	ten	https://w.wiki/4tkU	criterion used (P1013) in 3/4 cases and object has role (P3831) in 1/4 cases

In my opinion we should probably reach one qualifier-property for each of these values (and others too, surely). I would propose simply to adopt object has role (P3831) in all these cases, and to bot-apply the change. Opinions? --Epìdosis 10:16, 28 February 2022 (UTC)

Support for using object has role (P3831) consistently for this. ArthurPSmith (talk) 18:17, 28 February 2022 (UTC)
Support for using object has role (P3831) consistently for this, but pinging @Daniel Baránek: who used criterion used (P1013) for his mass imports of invaluable data and probably had a good reason for that. --Emu (talk) 20:15, 28 February 2022 (UTC)
@Emu: Thanks for pinging. See my vote for explanation. I only do not understand why You are writing about "invaluable data". You probably ment "invalid data" = not in accordance with existing modelling? --Daniel Baránek (talk) 08:29, 1 March 2022 (UTC)
@Daniel Baránek: No, I did mean invaluable! I think your imports are among the best additions to Wikidata in the last time! --Emu (talk) 11:17, 1 March 2022 (UTC)
@Emu: Oh, thanks :-) Daniel Baránek (talk) 11:39, 1 March 2022 (UTC)
Support The only reason for using criterion used (P1013) was that I wasn't familiar with Wikidata:WikiProject_Names/Properties#Qualifiers. Proposed unification is surely desirable. --Daniel Baránek (talk) 08:29, 1 March 2022 (UTC)
Comment given name (P735) with religious name (Q1417657) should be rather moved to religious name (P1635) than modifying the qualifier. --Daniel Baránek (talk) 08:31, 1 March 2022 (UTC)
- that's a different datatype. --- Jura 09:22, 1 March 2022 (UTC)
Comment I'm not entirely convinced the methodology applied here: object has role (P3831) implies some qualification of the value, whereas criterion used (P1013) only qualifies the relation. The counts provided above don't really show anything about this difference.

P735 can apply to parts of such names and then the relation isn't object has role (P3831). Bulking changing qualifiers to P3831 would be an incorrect use of P3831. --- Jura 09:22, 1 March 2022 (UTC)

@Jura1 Could you give us an example? You are probably raising an important point, but I think I need an example to fully grasp your comment. Thanks! --Emu (talk) 11:20, 1 March 2022 (UTC)

Probably most with name at birth (Q2507958) for P735 should have criterion used (P1013) or applies to part (P518) (a person's name at birth generally includes one or several given names and a family name e.g. "Marya Salomea Skłodowska".

Even if you consider it being limited to the given name(s) only, it may not work if a person has several given names. Q7186#Q7186$959EBDBB-C0C5-41AA-970F-62769DD4538E lists just one of them.

For usual forename (Q3409033) as value, it may not matter.

Not sure if subject has role (P2868) could ever apply (on items for human (Q5)).

The qualifier object has role (P3831) seems to have come in as a replacement of instance of (P31). Maybe we should have stayed with that or picked something else. We used to have the qualifier "as" that avoided such problems. --- Jura 12:03, 1 March 2022 (UTC)

@Jura I’m not quite sure if I get it. So your problem is that object has role (P3831) implies some sort of uniqueness and can’t therefore be used if there are several given names at time of birth? --Emu (talk) 19:22, 13 March 2022 (UTC)

Should we start uniforming to qualifier object has role (P3831) at least for the value usual forename (Q3409033), for which there is no objection so far? --Epìdosis 08:49, 6 April 2022 (UTC)

Yes please. I still don’t understand what @Jura is trying to say … --Emu (talk) 10:09, 6 April 2022 (UTC)

usual forename (Q3409033) now uniformed manually to object has role (P3831); 50 items changed. The next ones could be name at birth (Q2507958) and religious name (Q1417657), I think, if the consensus is sufficient. --Epìdosis 12:15, 6 April 2022 (UTC)

uniformed to qualifier object has role (P3831) of given name (P735) the following: usual forename (Q3409033), name at birth (Q2507958), religious name (Q1417657), legal name (Q666791), stage name (Q1055303), hypocorism (Q1130279). I wait some more weeks for family name (P734). --Epìdosis 18:05, 30 April 2022 (UTC)
uniformed to qualifier object has role (P3831) also: pen name (Q127843), Christian name (Q2396897), nickname (Q49614), given name (Q202444), middle name (Q245025). --Epìdosis 14:44, 1 May 2022 (UTC)
If no other objection arises, tomorrow I would go with object has role (P3831) also for family name (P734) as initially proposed nearly three months ago. --Epìdosis 15:19, 19 May 2022 (UTC)
{{Autofix}} was ineffective and in fact it doesn't work for qualifiers, so I have just opened a bot request. --Epìdosis 07:49, 17 July 2022 (UTC)

Help with a Chilean name

Could someone help me out with Millarca Valenzuela (Q29610805) (Edith Millarca Valenzuela Picón)? I'm not sure where to apply series ordinal (P1545) qualifiers, or if they should be used at all. Thanks! Aluxosm (talk) 12:32, 27 May 2022 (UTC)

Notified participants of WikiProject Names: If anyone fancies a challenge, the top of this list has some more tricky ones that I've been struggling with. Any help would be much appreciated, cheers! Aluxosm (talk) 14:44, 27 May 2022 (UTC)

@Aluxosm: For Millarca Valenzuela (Q29610805), I think it's ok now. --Epìdosis 14:47, 27 May 2022 (UTC)

Diacritics in Vietnamese names

Vietnamese uses many diacritics. Consequently, Dung, Dũng, and Dụng are 3 different Vietnamese given names (there are other similar cases). When I have a look at the items linked to Dung, there are actually two situtations. The given name of some of them is actually Dung in Vietnamese (for example Q7833324). In other cases (for example Q99596693), the source where the name appears omits the diacritics, so we actually do not know if the original name is Dung, Dũng, or Dụng. Would it be desirable to create two distinct items for Dung: one for the authentic Vietnamese name without diacritics and one for the cases where we do not know which diacritics are used in Vietnamese? (In my opinion, this second item could also be used when we do not know if it is a Vietnamese name or a name of another origin.) BrightRaven (talk) 07:50, 21 June 2022 (UTC)

Merge of family names starting with lowercase "van" and uppercase "Van"

I saw recently some family name items where merged in which one was the family name with a lowercase infix (P7377) and the other of a uppercase infix (P7377). Examples are Q2177230 (van der Meer / Van der Meer, change) and Q1645815 (van den Heuvel / Van den Heuvel, change).

To give some context: in the Netherlands the lowercase version is almost always used. It can turn into uppercase in some cases, like when used without a given name, so it's when at the start of the name. This is similar as how normal lowercase words can turn into uppercase. From itself it's lowercase. In Belgium different variants exist alongside each other, in both case and concatenation. So there can be a van der Meer, a Van der Meer and a Vander Meer, while in the Netherlands there would only be a van der Meer.

On these merged items there are now multiple versions of infix (P7377), which isn't allowed according to the property description. The labels now also don't show up correctly anymore in the family name (P734) on items about humans.

Where these merged correct?

If not, how can they be untangled, including all the links and how can these merges be prevented in the future?

If yes, how should the duplicate infix (P7377) and incorrect labels be handled? Koen WikiTree (talk) 20:09, 28 June 2022 (UTC)

I think they should be merged, since the names are the same, there's just a couple of different formatting conventions. Ghouston (talk) 10:52, 29 June 2022 (UTC)

That means, in Belgium one person’s name could be “van der Meer” and another person’s name “Van der Meer”? My opinion: let’s not merge, then, otherwise we can’t distinguish these names on Wikidata. Julian Jarosch (digicademy) (talk) 12:03, 30 June 2022 (UTC)

Possibly. Here's an article on a Flemish site that uses "Mathieu van der Poel" for a Dutch cyclist. They also use Wout van Aert for a Belgian cyclist, so perhaps they really are distinguishing on a case-by-case basis. I'm not sure if somebody migrated between Belgium and the Netherlands, that the capitalization would be preserved for generations on end. Ghouston (talk) 07:56, 1 July 2022 (UTC)

People actually discuss such things elsewhere too. It's possible Van Aert has Dutch citizenship, despite living in Belgium. Ghouston (talk) 08:05, 1 July 2022 (UTC)

In Belgium there can indeed be a person named "van der Meer" and another named "Van der Meer" (and a "Van Der Meer", Vander Meer, etc.). So in that sense they can be considered different surnames. The article about Mathieu van der Poel only uses "Van der Poel" when the first name is ommitted. In that case the "v" is capitalized because it's at the start of the name, not because it has a capital V from itself. This is similar as when you call a house Blue House. In that case blue starts with a capital B, while the word blue hasn't got a capital from itself. Koen WikiTree (talk) 16:16, 2 July 2022 (UTC)

What Koen WikiTree says is correct. Unfortunately, Wikipedia automatically writes all lemmas in capital letters at the beginning, which causes additional confusion. Van, van, Van der, van der, etc. are still to be treated as different name suffixes and may not be merged. --HarryNº2 (talk) 19:45, 2 July 2022 (UTC)

I'd say even if we want to have Dutch and Belgian versions of the same surname, that the Dutch label should still be capitalized, since that's how it's written when the name is used on its own. Ghouston (talk) 10:17, 3 July 2022 (UTC)

These are two different words or suffixes. It's the correct spelling that matters, not how it's spelled as a standalone word. Each word is capitalized at the beginning. Only the use of the word in context is of interest here, not how it is spelled as a stand-alone word. --HarryNº2 (talk) 15:08, 3 July 2022 (UTC)

Note that these are not the Belgian and Dutch versions of the same name. These are different versions of a name, for which one version is used in the Netherlands and several versions are used in Belgium.

In the Netherlands the 'v' is only capitalized when used in a sentence at the start of the name (so without a given name). In that case it gets a capital letter, because of the position, not because it has a capital v from itself. You can see this when a surname starting with 'van' is used on its own outside a sentence, for example in a passport. There an lowercase v is used. Koen WikiTree (talk) 15:43, 7 July 2022 (UTC)

P1705 language for given names and family names

Notified participants of WikiProject Names Hi all! Until a few days ago Wikidata:WikiProject Names and Wikidata:WikiProject Names/Help didn't give advices about which language was to be selected for native label (P1705) in given names' and family names' items; I have recently specified that this language should be "mul", mainly for three motives: P1705 is already used in the majority of cases with "mul" language, while the language(s) using the name could be specified using language of work or name (P407); having all the names with P1705 using "mul" would help with merges (cfr. Wikidata:Database reports/Constraint violations/P1705 and also Wikidata:WikiProject Names/Possible surname duplicates, resulting from Wikidata:Project_chat/Archive/2022/06#Please_advise_on_surnames); I have found a few scattered opinions about this in old discussions (2016, 2017, 2019). However, this is not the only possible approach, let's list some:

native label (P1705) always "mul" (with language of work or name (P407) stating all languages using the surname) - in the guidelines after my intervention
native label (P1705) "mul" only if the surname is used in more than one language, otherwise the only language using the surname (with language of work or name (P407) stating all languages using the surname)
one native label (P1705) for each language using the surname (with language of work or name (P407) stating all languages using the surname)

Please comment if you agree with my intervention (= option 1) or if you prefer one of the other options. On the basis of the consensus it will also be possible request a bot intervention to try to adequate existent items to the chosen standard, insofar as possible. Thanks, --Epìdosis 12:31, 2 July 2022 (UTC)

A specific language should only be specified if a source is given for it. Setting native label (P1705) to "mul" across the board would not be correct. The original language should already be specified in native label (P1705) if it can be verified by a reputable source, in this case Johannes Gutenberg University Mainz (Q161982).

Here are two examples of what I mean:

Müller (Q8157228): native label (P1705) is "de", language of work or name (P407) is German (Q188), see DFD "Müller"
Miller (Q1605060): native label (P1705) is "mul", language of work or name (P407) is German (Q188) and English (Q1860), see DFD "Miller"

Using "mul" for e.g. Cyrillic script (Q8209) ("ru", "uk" etc.), Japanese (Q5287) ("ja") or Korean (Q9176) ("ko" or "ko-kp") would also not be correct. --HarryNº2 (talk) 13:55, 2 July 2022 (UTC)

Diminutives in Russian

What are current view about diminutive forms of Russian names? I know that e.g. for English names there are separate items with P31=hypocorism (Q1130279) are created. I would explain this because there are person known primarily by such diminutives. But there are (almost?) none for Russian though. So my opinion is the following: in English (in Latin script? in some other languages?) probably Ann and Anya and Annushka are different names - they deserve separate items; in Russian Аня = Анна = Аннушка so we have single Anna (Q22713652) with a bunch of aliases. Any thoughts? @Jura1:? Infovarius (talk) 13:01, 13 July 2022 (UTC)

About “Ataúlfo” and other similar cases

In Spanish there is the given name “Ataúlfo”. Currently Wikidata has the item Ataulfo (Q17621889) which appears to be in Italian. To use on people like Ataúlfo Morales Gordillo (Q112161389) or Ataúlfo Sánchez (Q5710499), do I have to create the corresponding item? In the case of King Ataulf (Q159716), what given name is used? I exemplify it with “Ataúlfo”, but the same happens with other similar names. Thanks. --Fantastoria (talk) 09:22, 17 July 2022 (UTC)

@Fantastoria: A separate data record must be created for each different spelling, see e.g. Müller (Q8157228), Muller (Q28800754), Mueller (Q15042884), Miller (Q1605060) etc., or Andre (Q16274896)/André (Q6298851), Maria (Q325872)/Maria (Q56449214)/Mária (Q6949677)/Mariá (Q81914052). --HarryNº2 (talk) 16:11, 1 August 2022 (UTC)

Proposal to re-configure name model

The more I notice of it, the more I am convinced that the current model is unsustainable and inconsistent with the rest of the data, and that we are letting the potential to create more useful data escape us while other areas have addressed similar concerns with better suited properties for lexicographic data. I will start by identifying some major problems with the current model, and then propose some solutions.

Problems

The notion of names as script-specific requires the exclusion of many people's names right away. For one, illiterate people are not nameless, yet no provision exists for how to link names to people who do not or did not read and write at all. This includes most people historically, and a non-trivial number of people presently.
Further, it is not required that a person have a single standard spelling for a name, and if there are spelling conventions there is no way to state which
The idea of "name in native language" as being an essential component of a name item is flawed given that a person can have a name that can only be understood as one originating from a language which they do not speak, or that the language of a person's name may be known but the language they speak natively is not publicly known. Both of these situations are very common.
In multilingual countries and regions with multiple writing systems, it is typical for a given person to write their name regularly in one script. Due to the adoption of the parameters outlined by WikiProject Names being largely adopted by users of the Latin script in practice, what this means is that nearly everybody who speaks a native language which is not commonly written with a Latin script is linked to an "English" Latin script name that no other scripts are allowed on. It is unclear if these people were supposed to have multiple items for each of their names linked to represent each script, or if the items have just never been corrected to have multiple scripts, but in any case the resulting current situation is one which is not consistent with the stated purpose of this project, and just not factually accurate for most people who have names linked to them. The majority of contemporary Indian people have names that are attested in at least three scripts - "English" Latin script, "Hindi" Devanagari script, and one or more of the several writing systems used for regional languages in India. Yet I have never once encountered a Wikidata item for an Indian person that actually reflects this fact, an especially baffling situation when we consider that India is one of the most populous countries in the world.
We can look at specific examples which I think demonstrate the failure of the current WikiProject Names project to produce any usable information. Mahatma Gandhi (Q1001) is indisputably one of the most well known of recently living people all of the world, and naturally his item bears one of the lowest Q numbers for a person on Wikidata, and was once touted as an exemplary model for a human item. Yet today, the name properties on Mahatma Gandhi (Q1001) link to items which are informationless at best and misleading at worst. Gandhi is a native Gujarati speaker with the given name Mohandas, which has the item Mohatmas (Q20658450). A look at the link backs makes it clear that only two other people and a fictional character are also linked to this name. The two other people look to be Bengali speakers. However, the Mohandas item declares the language of the name as being Hindi, the writing system as being Devanagari, and lacks a Gujarati label entirely despite the most notable bearer of the name being a native Gujarati speaker. Further contradicting itself, the name in the native label property is ਮੋਹਨਦਾਸ, with the monolingual text code for Hindi, yet written in the Gurmukhi script commonly used for Punjabi. The Hindi termbox label has a Punjabi label in it, the actual Punjabi termbox labels are empty, and a Devanagari Hindi label is nowhere to be found. As for the question of what the "true" native language of this name, it is actually a Sanskrit name. This is a historical language not currently spoken yet Sanskrit names are used across India and they may be understood as the same name. The current model has no way of accounting for this, but does seem to have caused more confusion given that the linked "identical family name" is an item which only uses Latin script, which is also linked exclusively as names of individuals who likely do not write solely with Latin script.
Let us take another example of a famous person, Imran Khan (Q155164), celebrity cricket player turned populist Prime Minister of Pakistan. Though I have left the Gandhi item alone for now, I could not help but correct this one. Khan's native language was listed as Urdu, which was also listed as the native language of his name. This was likely done simply because Urdu is an official language of Pakistan, but a claim like this should have a substantive reference given that only a small minority of Pakistani people actually speak Urdu as a mother tongue. This is somewhat more prevalent among those who had an affluent upbringing like Khan, but given that he was raised in an area that is primarily Punjabi speaking by parents who spoke Saraiki and possibly Urmuri, a rare language spoken by only a few thousand people, it is unclear where he would have picked up Urdu as a mother tongue. The fact that nobody has been able to verify what his mother tongue is notable enough to have been remarked on in Pakistani newspapers. The best we can say is "unlikely to be Pashto and would be unsurprising if Punjabi," and the best we can do for Wikidata is "unknown value" for native language and "undetermined" for native language label. The problem still stands though, that despite the fact that this is one of the most notable contemporary politicians on the global stage, we apparently should have no way to model his name if we follow the WikiProject Names rules because we do not know his mother tongue. This seems pointless, as we should not actually need to know his mother tongue to retrieve information about his name, and we can say that he has the Arabic-origin given name Imran and the Turkic/Mongolic family name Khan.

Proposed solution

I would like to propose that names be given single items per common origin rather than by script or spelling difference - splitting names by the latter as recommended is in practice very inconsistently implied and apparently only ever done for Latin and possibly Cyrllic script forms. The vast majority of non-European-origin names are not modelled correctly when considered under any terms. This would mean merging the several items for Mohammed, the most common given name in the world, which only differ in that they are spelled slightly differently in the Latin script, despite being spelled identically in many other scripts and in all cases linked mostly to people who do not primarily write with a Latin-based script. It would also mean splitting the Scottish name Harris and the Arabic name Harris (Q20631613), which currently share an item despite only being coincidentally related. Further, each name would be labeled in a way that is readable to speakers of the language to that label. The Arabic name Harris may be spelled هاريس in Arabic, and is also used by Punjabi speakers, who often write with an Arabic-based script. However, the same name may be written as حارث in Punjabi, and the letter used at the front of the Arabic spelling is not typically supposed to be used in the same position in Punjabi words. Instead of making the labels look like spelling mistakes in other languages which use Arabic-derived scripts, the name should be spelled appropriately for each given language so that it is actually recognizeable for what it is. After all, when the work of a writer is published in multiple languages, this is what is done with their name.
The relatively recent personal pronoun (P6553) has been excellently modeled to use lexicographic data rather than item links to represent pronouns. This way, a number of nuanced linguistic notions of pronouns may be determined from the lexicographic entity links that would not be possible with an item lacking language specific or grammatical information. Names may be modeled using lexicographic data as well, without even needing a new property - we can simply use the qualifier object form (P5548) to link specific representations of a name to a statement which links the general single item for a name.
A demonstration: let us take Mohammed Rafi (Q156815), a native speaker of Punjabi, bearing the common name Mohammed. Both India and Pakistan use English in an official capacity, and naturally people named Mohammed in Punjab often have their name represented in Latin script within this context. However, the given name Mohammed does not have a standard Latin script spelling in this region and the same person may even spell their named variably as Mohammed or Muhammad without worrying about the difference. Mohammed Rafi is no exception in this regard, and his full given name can be found spelled multiple different ways in the Latin script. Note that importantly, one form of his name in the Latin script does have a standard spelling: the abbreviation Mohd. used in Punjab is always spelled the same way. Muhammad Rafi is a perfectly valid way to spell the same person's name, but Muhd. is never correct. Further complicating things, despite Mohammed Rafi spelling his name as محمد in Punjabi Shahmukhi (Perso-Arabic based script), Mohammed has always has a Punjabi specific pronunciation differing from the Arabic محمد. The sound the middle character ح represents within Arabic orthography is non-existant in Punjabi, and the correct pronunciation of Mohammed Rafi's name does not involve this sound. If we were to create the multiple additional Mohammed items apparently needed to model this one person's name under the WikiProject Names guidelines, it is not clear how this data would be usable to anybody. Most would justifiably see such an endeavor as a waste of time.
Instead, we can link object form (P5548) as a qualifer to given name (P735) = Mohammed (Q5837762). This is an arbitrarily selected Mohammed item, any or all would work here until they are merged. The object form we will enter is invalid ID (L686069#F1). In doing so we have clarified two valid Punjabi spellings of his given name and implicitly the specific Punjabi pronunciation of his name without obscuring the fact that he shares his given name Mohammed with every other Mohammed/Muhammed/Mohammad/etc. in the world. Note that this lexeme currently has to have several Mohammed items linked to it to represent all the duplicates that would simply be spelled the same way in Punjabi and represent the same name - if/when merged, only one item for each sense of given and family name would be necessary, and any number of valid transliterations of this name could be derived from the merged items rather than this information being dispersed or undocumented.
From here, we can learn interesting things about this given name that were not possible using the current model. We can trace the property derived from lexeme (P5191) to the original Arabic given name مُحَمَّد (L52356), and further to the original Arabic word form from which we may learn that this name actually started as a particular passive participle form of a verb invalid ID (L686791#F2). If implemented at scale, this would open up the possibility of all kinds of queries that were not possible for. What if you wanted to know the most common names derived from verbs? Or what if there are more people with the name Ivan have a Cyrllic form linked or a Latin form linked? Which typically male given names are derived from grammatically feminine nouns? These could become quite creative given some effort to this end.
As an example of representing Mohd. specifically, we may add property short name (P1813) pointing to value "Mohd.", say that the object has role as given name, and then link an object form invalid ID (L450083#F8) containing the statement variety of lexeme, form or sense (P7481) pointing to South Asian English (Q7566032). The lexeme Muhammad (L450083) should be linked to the same single Mohammed item. This way, we can indicate this particularly spelled, locale specific version of the name as a valid short form for this person's name with out over-complicating things.

Naturally there are some hairy details surrounding names which I have not touched on here, but there is too much to cover in a single post. I do think we can account for most necessary nuance with this as a starting point though. Switching models would shift the data to a transitional state temporarily, but I would say that this is preferable to the current situation in which the data does not seem to be usable for anything at all. --Middle river exports (talk) 23:49, 14 August 2022 (UTC)

I understand that it’s impossible to select one specific spelling for Indian (or even Medieval European) people, but the exact spelling does count in case of most present-day European/North American people. For example, the Hungarian names Mária (Q6949677), Marika (Q3293081), Marietta (Q17578110) and Mariann (Q1165172) (and more) all originate (or likely originate) from the name of Virgin Mary (Q345), but calling a Marietta Mária is not considered acceptable. Maybe multi-lingual/multi-spelling names should be allowed in case a single native language or a canonical spelling is unknown or doesn’t exist, but items focusing on a single spelling should also be allowed when the spelling is definitely known. —Tacsipacsi (talk) 09:51, 16 August 2022 (UTC)

I would say that these are all different names with different etymologies even though they have the same origin - what I mean by this is that even though they are similar, they were each derived as separate names. They are not just alternate spellings, they are not pronounced the name either. So I do not think there is any reason to merge these.

The problem I am talking about is for situations where the choice of how to spell the name is entirely arbitrary. "Saad Ullah" and "Sadullah" are both choices for spelling the same name that might even lead someone unfamiliar with naming conventions in Pakistan to think that this is two names for example. Having one item that indicates that this is one name with multiple spellings would not only reflect how this name is used properly, but would help clarify to people unfamiliar with it that these spellings are in fact interchangeable. (Unlike Marika and Marietta.)

Different spellings are also not even obvious a lot of the time. Can we really expect there to be different items for every Arabic-origin name which has ة in it to have a counterpart version with ۃ in it? These are different characters, the second one being an Urdu version of the first one which you can only tell is different based on the cursive attachment to another letter (which does not always appear). Middle river exports (talk) 04:00, 14 September 2022 (UTC)

Thank you for raising fascinating and valuable points! I’m quite unfamiliar with the languages and writing systems you bring up, so I can hardly offer much of a solution. Quite broadly, and possibly unhelpfully, speaking, I don’t think the solution is to use one model for all cultures and scripts. I’m with Tacsipacsi that it would not be so good for current European and North American cultures to remove spelling distinctions, because I think many of these people generally tie their identity to a precise spelling. But I don’t mean to say that I don’t want a solution that’s also adequate for the cases you describe. I just tend to think, right now, that there will be duplications and parallel structures to account for different naming practices. I.e. in some way parallel items for distinct spellings in Latin script and generalised items (lexemes?) for name-types covering various spellings, maybe mostly not for Latin script.

A lateral point I’d like to bring up are migration and codified transliterations. I think it would be important for e.g. Germans to maintain the distinction of someone in Germany officially registered as “Mohammed” and someone registered as “Muhammad”. Essentially, by moving into a different cultural mindset, the distinction is ascribed significance again.

Regarding illiterate people: Yes, that’s a complication. A part of a solution can be to look at if they were in a literate context (e.g. recently) – did the family chose a spelling for their name? Did they have official documents? Maybe this is somewhat comparable to phonetic names of/for Deaf people.

Julian Jarosch (digicademy) (talk) 17:44, 24 August 2023 (UTC)

Citation needed for said to be the same as (P460)

Notified participants of WikiProject Names

said to be the same as (P460) is to be provided with a source in future at Silver hr's request, see Property talk:P460#Citation needed constraint. said to be the same as (P460) is used practically in every name article, e.g. Andre (Q16274896)/André (Q6298851), Maria (Q325872)/Maria (Q56449214)/Mária (Q6949677)/Mariá (Q81914052) and hundreds of thousands more, which you cannot or do not have to prove because it is simply logical. --HarryNº2 (talk) 04:27, 16 August 2022 (UTC)

To provide a quick background/explanation:

The proposal for said to be the same as (P460) states that the intent for its use is to document claims about historical people by some scholars who say that different historical people are in fact the same person (while at the same time other scholars may claim differently), and it was approved with that in mind.
There was already a wider discussion about this in Wikidata Project Chat. My summary is here. Briefly: Wikidata has 3 ways for editors to claim that items are about the same concept: merging, permanent duplicated item (P2959) with potential qualifiers such as sourcing circumstances (P1480)presumably (Q18122778) or nature of statement (P5102)hypothesis (Q41719), and equivalent class (P1709) and its superproperty exact match (P2888). said to be the same as (P460) is for when someone (not a WD editor) is making the claim in some source document.
Relevant to WikiProject Names: some names are connected in some way. For example, the connection between Miller (Q1605060) and Möller (Q1184002) is that they are both instance of (P31)occupational surname (Q829026)of (P642)miller (Q694116). The connection between Andre (Q16274896) and André (Q6298851) is Andre (Q16274896)has characteristic (P1552)name without diacritical marks (Q89410507)of (P642)André (Q6298851). The connection between Andre (Q16274896) and Drejc (Q20523907) could be Drejc (Q20523907)has characteristic (P1552)version (Q105062328)of (P642)Andre (Q16274896); I can't say for sure as names aren't my domain of expertise. What I can say for sure is that none of these names are exactly the same concept, which is currently claimed by said to be the same as (P460) statements.

Silver hr (talk) 16:57, 16 August 2022 (UTC)

said to be the same as (P460) and its description ("this item is said to be the same as that item, but it's uncertain or disputed") expresses that it is controversial. Therefore, the indication of a source is unnecessary. --HarryNº2 (talk) 14:08, 25 August 2022 (UTC)

I have already addressed this twice: here and here. Let me repeat it yet again: the "uncertain or disputed" bit in the description means that it's to be used for cases when it's not completely obvious that two concepts are the same; instead someone is making that claim, but not everyone necessarily agrees with it. This is why it's important to state who is making the claim.

Please start addressing the substance of my arguments instead of merely repeating yourself. Merely repeating oneself does not make for a constructive discussion. Silver hr (talk) 15:21, 25 August 2022 (UTC)

To merge or not to merge?

Hi,

I let people knowing the subject better answer on MediaWiki_talk:Gadget-Merge.js#Merging_Multiple_Pages (I feel it should not be merged, but I'm a bit uncertain).

Cheers, VIGNERON (talk) 17:44, 12 September 2022 (UTC)

Issues with name items

I get the feeling that there is a ton of crossover with several types of nicknames, surnames etc and that these there has been several cases of merges with items like these after-name (Q4116295), nickname (Q15635173), that don't always end up making sense. I've been trying to move around and clean up a lot here but I can't read all these languages so its hard for me to really get if the links that are left are really about the same things. StarTrekker (talk) 18:56, 17 August 2023 (UTC)

P6597 (Digital Dictionary of Surnames in Germany ID) constraint violations seem to have gone up recently

Just a short note, not sure if anyone wants to get into it further:

I suspect this batch in particular (@JotaCartas) introduced a number of unique value constraint violations. (Not the entire batch is flawed, though.)

I’ve used up all the time I currently have available to sort through some of the constraint violations, but won’t personally be able to get to the bottom of it.

Julian Jarosch (digicademy) (talk) 17:56, 24 August 2023 (UTC)

I can check it out.StarTrekker (talk) 08:48, 29 August 2023 (UTC)

Wikidata talk:WikiProject Names/Archive/3

Contents

Deletion

Brazilian multiple surnames

St Leger vs. St. Leger

surname from a specific origin (Q33283501)

Same spelling, different origin

Name class

pap: prome nomber maskulino vs. di prome nomber maskulino

Osypenko (Q108524859) vs Osipenko (Q12798037)

Wingfield: given or family name for "Penrhyn Wingfield Coussens"?

actually male given name that is exceptionally also used for women

Qualifiers for given names and surnames - establish a guideline

Help with a Chilean name

Diacritics in Vietnamese names

Merge of family names starting with lowercase "van" and uppercase "Van"

P1705 language for given names and family names

Diminutives in Russian

About “Ataúlfo” and other similar cases

Proposal to re-configure name model

Problems

Proposed solution

Citation needed for said to be the same as (P460)

To merge or not to merge?

Issues with name items

P6597 (Digital Dictionary of Surnames in Germany ID) constraint violations seem to have gone up recently

Navigation menu

Wikidata talk:WikiProject Names/Archive/3

Deletion

Brazilian multiple surnames

St Leger vs. St. Leger

surname from a specific origin (Q33283501)

Same spelling, different origin

Name class

pap: prome nomber maskulino vs. di prome nomber maskulino

Osypenko (Q108524859) vs Osipenko (Q12798037)

Wingfield: given or family name for "Penrhyn Wingfield Coussens"?

actually male given name that is exceptionally also used for women

Qualifiers for given names and surnames - establish a guideline

Help with a Chilean name

Diacritics in Vietnamese names

Merge of family names starting with lowercase "van" and uppercase "Van"

P1705 language for given names and family names

Diminutives in Russian

About “Ataúlfo” and other similar cases

Proposal to re-configure name model

Problems

Proposed solution

Citation needed for said to be the same as (P460)

To merge or not to merge?

Issues with name items

P6597 (Digital Dictionary of Surnames in Germany ID) constraint violations seem to have gone up recently

Navigation menu

Search