Module talk:be-noun

@Atitarev This module is live now. Could you help fix the resulting module errors? They are caused by missing stresses, multi-stressed forms or missing forms. Benwing2 (talk) 05:16, 10 July 2020 (UTC)Reply

@Benwing2: OK, thank you. --Anatoli T. ^{(обсудить}/^вклад) 05:19, 10 July 2020 (UTC)Reply

@Benwing2: I have fixed all Belarusian entries (generating errors only) but I don't know what to do with {{be-noun-inan-f-а-1}}. --Anatoli T. ^{(обсудить}/^вклад) 07:49, 10 July 2020 (UTC)Reply

@Atitarev You mention genitive plurals of feminines (also neuters) in -аў/-яў. I am working on automating Belarusian nouns and I'm planning on adding a flag to indicate this type of genitive plural, probably (ў) attached to the accent pattern, similarly to the reducible indicator. So ды́ня (dýnja) might be indicated as {{be-ndecl|ды́ня<a,a(ў)>}} or similar. Benwing2 (talk) 04:14, 19 July 2020 (UTC)Reply

@Benwing2: Thanks for that. Whenever I find these cases every now and again, I add them to the declension. Such cases may also be included in slounik.org as well. Another case is пусты́ня. --Anatoli T. ^{(обсудить}/^вклад) 06:53, 19 July 2020 (UTC)Reply

@Benwing2: Pls note that this also happens with some neuter nouns, e.g. крэ́сла->крэ́слаў (also in slounik). --Anatoli T. ^{(обсудить}/^вклад) 07:31, 19 July 2020 (UTC)Reply

@Atitarev I looked into this and the majority of neuter nouns actually have genitive plural in -аў or -яў, so this will be the default. Benwing2 (talk) 01:14, 21 July 2020 (UTC)Reply

@Benwing2: Thanks, as long as you cover cases like дрэў/дрэваў.

I sometimes can't tell with 100% certainty if a form is standard, even if it seems attestable or common. E.g. I can find examples of "сініцаў" (сініца) and there will be many other alt fem. gen. pl forms, so I am not sure, if they all should be included. Slounik mostly provides the latest standardised forms. More research is required. --Anatoli T. ^{(обсудить}/^вклад) 01:28, 21 July 2020 (UTC)Reply

@Atitarev Yup, those cases will be handled. As for feminines in -аў, I read somewhere (I think on Wikipedia) that feminines in -аў are more characteristic of Taraškievica and those with a null ending are more characteristic of the current Russianized norm. Benwing2 (talk) 03:35, 21 July 2020 (UTC)Reply

@Benwing2: Yes, that's about right, which makes it difficult to determine, which parts of Taraškievica inflections can be included in the declension tables. For example, I have manually collected inflected forms of ме́ста (mjésta). Gen. pl: "мест, място́ў" but loc. sg. "ме́сьце" would definitely apply only to Taraškievica, with modern form being "ме́сце". So I only included "ме́сце". --Anatoli T. ^{(обсудить}/^вклад) 04:20, 21 July 2020 (UTC)Reply

@Benwing2 I think it would be great to embrace and provide for Taraškievica as well eventually. I just searched for "надзеяў" (надзея) and found a few links going https://www.svaboda.org/, which is a good read now that all Belarusian media is controlled by Lukashenka and all "standard" Belarusian texts ar efull of propaganda. I think it would be safe to add -аў/-яў as alt feminine forms to all feminine nouns ending in -а/-я for now. Belarusians seem to mix old and new grammars. It won't hurt to have both надзе́й and надзе́яў. --Anatoli T. ^{(обсудить}/^вклад) 05:52, 21 July 2020 (UTC)Reply

@Atitarev Hmmm. If I add those forms to all feminine nouns in -а/-я, maybe I should footnote them as Taraškievica norm unless the template call specifically requests these forms using (ў). Otherwise you'll get a random mixture of different norms, which isn't so good. What do you think? Benwing2 (talk) 05:59, 21 July 2020 (UTC)Reply

@Benwing2: There is not always a clear border between standard and Taraškievica. In case of на́зва (názva), gen. pl "на́зваў" is the only way, perhaps because of the consonant cluster, you'll find many other similar cases. Dative and locative sg. "на́зьве" is a more clear Taraškievica case. Different sources in slounik may give one or the other or both inflections. --Anatoli T. ^{(обсудить}/^вклад) 06:10, 21 July 2020 (UTC)Reply

@Atitarev What I mean is, there will be a way in the module of adding -аў as a feminine ending (probably the (ў)), either by itself or as an alternative. We can do this when Slounik, esp. Граматычны назоўніка 2013, lists -аў as an ending, either as the only possibility or as an alternative. If not given in Граматычны назоўніка 2013, the module can still add it, but footnoted as Taraškievica, to make it clear it's less likely to be used by the "standard" norm. Benwing2 (talk) 06:16, 21 July 2020 (UTC)Reply

@Benwing2: That would be fine, thanks. --Anatoli T. ^{(обсудить}/^вклад) 06:19, 21 July 2020 (UTC)Reply

(moved from Talk:востраў) @Benwing2 It's the second time I noticed that the count form is different from the usual. Typically, nouns with numbers 2, 3, 4 (22, 23, 24, etc.) are similar to Russian or Ukrainian (but "тры дні" (three days) = uk: "три дні" but <> ru" "три дня").

In this case, however, во́стравы (vóstravy) doesn't even appear in the declension table. Slounik sometimes shows this grammatical feature. I added usage notes to another BE entry as well but I forgot what it was :(. Not sure what to do about this yet, just FYI. --Anatoli T. ^{(обсудить}/^вклад) 03:00, 22 July 2020 (UTC)Reply

@Atitarev Yes, I've noticed this. Ukrainian also has count forms. The rule for Ukrainian seems to be that you take the nominative plural and stress it according to the stress of the genitive singular. Sometimes this produces forms that aren't found in the declension tables. In Belarusian it's more complex because of vowel alternations and such, but the same idea seems to apply. Here it seems like you take the entire genitive singular, chop off the ending, and add the ending of the nominative plural. Maybe the same thing actually happens in Ukrainian, too; you could determine this by taking a word like друг (which has a different singular vs. plural stem) and looking it up in mova.info, which lists count forms. Unfortunately the site is currently down so I can't verify this. Also, I'm not sure if the count forms apply only to masculines (like in Bulgarian) or to all nouns -- do you know? My plan is to include the count forms in the declension tables once I finish automating Belarusian noun declensions, since they can be generated in a systematic fashion. Benwing2 (talk) 03:21, 22 July 2020 (UTC)Reply

@Atitarev It's not just masculines. Slounik has an example of a count form for пятля́ (feminine). Benwing2 (talk) 04:37, 22 July 2020 (UTC)Reply

Also for чало́ (neuter). Benwing2 (talk) 04:48, 22 July 2020 (UTC)Reply

@Atitarev I added support for count forms using |count=. See во́страў for an example. Benwing2 (talk) 05:04, 22 July 2020 (UTC)Reply

@Benwing2: Thanks. --Anatoli T. ^{(обсудить}/^вклад) 05:32, 22 July 2020 (UTC)Reply

@Benwing2: Adding "з ліч. 2, 3, 4" in the search box in slounik brings back a lot of results. --Anatoli T. ^{(обсудить}/^вклад) 05:36, 22 July 2020 (UTC)Reply

@Benwing2: Thanks for adding the count form into the table. I think it would be appropriate to add a footnote that these forms are to be used with numbers 2, 3, 4 (22, 23, 24, etc.). It's not like the Bulgarian count forms. --Anatoli T. ^{(обсудить}/^вклад) 23:53, 22 July 2020 (UTC)Reply

@Benwing2, @Atitarev: With numbers 2, 3, 4, the count form should be just nominative plural without any extra fanciness (дзень - дні - два дні, дом - дамы́ - два дамы́, след - сляды́ - два сляды́, год - гады́ - два гады́). Except for a small number of words that preserved the old dual grammatical number form and, as you noticed, have this extra form listed in the dictionaries (во́ка - во́чы - два во́кі, сястра́ - сёстры - дзьве сястры́, во́страў - астравы́ - два во́стравы). A brief explanation in Belarusian can be found here: https://www.svaboda.org/a/24852780.html --Ssvb (talk) 03:43, 6 April 2025 (UTC)Reply

@Ssvb Is there anything needing to be done to the module? It's not clear from what you've posted here or the link you gave. Benwing2 (talk) 03:48, 6 April 2025 (UTC)Reply

@Benwing2: The currently used "take the entire genitive singular, chop off the ending, and add the ending of the nominative plural" approach produces incorrect count forms for many words. It should be "take the nominative plural" by default. For example, the declension table of мост (most) suggests that the count form should be "мосты". But in reality it's "масты" (same as the nominative plural form) and it's easy to find quotations on Wikisource to demonstrate this: "два масты", "тры масты". Or on Belarusian news sites: "site:svaboda.org "два масты"", "site:nashaniva.com "два масты"".

That said, the "take the entire genitive singular, chop off the ending, and add the ending of the nominative plural" algorithm seems to work for at least some of those rare old "dual grammatical number form" cases, such as сястра (sjastra), востраў (vóstraŭ), etc. So it's not entirely useless, but shouldn't be used by default. --Ssvb (talk) 05:35, 6 April 2025 (UTC)Reply

@Ssvb How do you propose I fix this? It looks like hard masculine and neuter nouns with a plural default to adding a count form using the algorithm you describe above, while other nouns default to no count form. Should I keep the default count form and make it identical to the nominative plural, or just not include the count form by default (i.e. only including it when it's overridden)? Benwing2 (talk) 05:26, 7 April 2025 (UTC)Reply

@Benwing2: It's probably better to always show the count form even if it's identical to the nominative plural, because this differs from Russian (incorrectly assuming the Russian rules happens to be a common source of mistakes). And it will be necessary to review all the Belarusian noun entries for the correctness of the count form, but this can be easily automated using the data from https://github.com/Belarus/GrammarDB --Ssvb (talk) 07:06, 7 April 2025 (UTC)Reply

@Atitarev Note the locative in -у with a noun referring to a person. We've seen this with nouns ending in -ц, maybe it applies to all the always-hard endings (ш ж ч as well)? 06:26, 20 July 2020 (UTC)Reply

@Benwing2: Yes, "тхор"'s (ferret) locative is "тхары́" but is animate otherwise.

слухач ->слухачу́ (listener, person), вуж -> вужы́ (grass snake, animal), нож->нажы́ (knife, object). (your signature was stuffed)--Anatoli T. ^{(обсудить}/^вклад) 06:49, 20 July 2020 (UTC)Reply

@Benwing2: тру́цень (trúcjenʹ) is another example of person vs animal difference. "drone" sense loc. sg. "тру́тні", "sluggard" sense loc. sg. "тру́тню". --Anatoli T. ^{(обсудить}/^вклад) 07:46, 22 July 2020 (UTC)Reply

@Atitarev The same thing happens with nouns in -й, e.g. яўрэ́й "Jew" (loc sg. яўрэ́ю) vs. папуга́й "parrot" (loc sg. папуга́і) vs. гай "grove" (loc sg. га́і). Benwing2 (talk) 08:33, 22 July 2020 (UTC)Reply

@Atitarev I've decided to require 'pr' vs. 'anml' in {{be-ndecl}}, as it's needed in many masculine nouns; by requiring it to be specified, we reduce the chance of error vs. requiring that people specify an override of the locative singular for people or animals (depending on which one is chosen as the default). Do you think we should introduce this distinction in {{be-noun}} headwords as well, or should we just have 'an' as we do currently? Benwing2 (talk) 03:42, 23 July 2020 (UTC)Reply

@Benwing2: It's OK to model everything on the Ukrainian nouns for animacy, IMO.

IMO, it would be great if Polish labels converted to categories as well, ideally consistent in all three languages. --Anatoli T. ^{(обсудить}/^вклад) 04:10, 23 July 2020 (UTC)Reply

@Atitarev I assume almost all remaining proper nouns without declensions have standard declensions (stress pattern a, no special indicator codes required). Is that true? Can you look over the nouns in Category:Requests for inflections in Belarusian noun entries and let me know if any of them require special handling? For the ones that don't, I can fairly quickly add declensions to all of them using a script. Benwing2 (talk) 01:10, 29 July 2020 (UTC)Reply

@Benwing2: I am glad you asked. These are time-consuming but quite repetitive tasks. Эўрапе́йскі Зьвяз (see "звяз") is "genu" but otherwise, they all (proper nouns) seem straightforward. Perhaps, need to be cautious with terms requiring "sg" (Вікіпэ́дыя, Ўэльс, Пенсільванія, Руская зямля, Эўрапейскі Зьвяз), etc but you know this. --Anatoli T. ^{(обсудить}/^вклад) 01:22, 29 July 2020 (UTC)Reply

@Benwing2: Don't forget about new cases, pls - Ілья́ (per Talk:Ілья) and Хрысто́с (I have made it manual - irreg), which will help handling Ісу́с Хрысто́с (voc: Ісу́се Хры́сце). --Anatoli T. ^{(обсудить}/^вклад) 01:26, 29 July 2020 (UTC)Reply

I can see Хрысто́с and Ісу́с Хрысто́с have already been taken care of :) --Anatoli T. ^{(обсудить}/^вклад) 01:28, 29 July 2020 (UTC)Reply

@Atitarev Hi. I'm working on this now but I realize there are some cases where accents are needed. Could you fix them up if possible? See Category:Requests for accents in Belarusian proper noun entries. Thanks! Benwing2 (talk) 02:13, 29 July 2020 (UTC)Reply

@Benwing2: Only accents will do? Are you able to add {{be-IPA}} on accented terms as well, at least single-word? --Anatoli T. ^{(обсудить}/^вклад) 02:18, 29 July 2020 (UTC)Reply

@Atitarev Yes, just the accents are enough. I can add {{be-IPA}} after adding the declensions. Benwing2 (talk) 02:26, 29 July 2020 (UTC)Reply

@Benwing2: Thanks. I've done Эўрапейскі Зьвяз, pls exclude from the bot run. --Anatoli T. ^{(обсудить}/^вклад) 02:34, 29 July 2020 (UTC)Reply

@Atitarev Should be all done. Benwing2 (talk) 03:24, 29 July 2020 (UTC)Reply

Still need to add {{be-IPA}}; will do that soon. Benwing2 (talk) 03:24, 29 July 2020 (UTC)Reply

@PUC, Jarash, Atitarev Under what circumstances is the count form irregular as след -> сле́да (not сле́ды) or мур -> муры́ (not му́ры)? If there is a rule for it, I'd rather put it in the module instead of adding overrides everywhere. Also please add overrides using the indicator counta (or count:сле́да) to add an override in -а, or countý (or count:муры́) to add an override in -ы́. Benwing2 (talk) 23:45, 28 August 2022 (UTC)Reply

@Benwing2, PUC, Jarash: Sorry, I am not familiar with these exceptions and couldn't find any evidence either. "тры следы" has a slightly more hits than "тры следа", no Taraškievica hits with "сьл-". Slounik.org has an example "баранава́ньне ў два сьляды́" (Taraškievica) but this is different from both "сле́да" and "сле́ды". Can @Jarash perhaps provide some examples from literature or references? --Anatoli T. ^{(обсудить}/^вклад) 00:09, 29 August 2022 (UTC)Reply

@Benwing2, @PUC, @Jarash, @Atitarev: The word след (sljed) is just regular and should use the nominative plural form: "тры сляды́". I believe that both "тры сле́ды" and "тры сле́да" are incorrect, but this doesn't stop less grammar-savvy people and such examples can be surely found on the Internet. I have added a few quotations from the published books to the след entry.

One more minor nitpick: overriding the count form makes it lose the "Used with the numbers 2, 3, 4 and higher numbers after 20 ending in 2, 3, and 4" footnote in the table. --Ssvb (talk) 04:15, 6 April 2025 (UTC)Reply

Thanks, I'll fix the count override issue. Benwing2 (talk) 04:26, 6 April 2025 (UTC)Reply

Hello! I couldn't get the magic of this template to work with «бор» (in the first meaning). I think it has a stress pattern not covered by Zaliznyak’s system? (Or maybe I don't know Zaliznyak's system well enough)

The closest I could get was able to get is {{be-ndecl|бор<c.genu>}}, but it produces incorrect locative (бо́ры instead of бары́) and incorrect count form (бо́ры instead of бары́).

The expected declension is as follows: sg. N бор, G бо́ру D бо́ру, A бор, I бо́рам, L бары́; pl. N бары́, G баро́ў, D бара́м, A бары́, I бара́мі, L бара́х, CF бары́.

If someone could tell me the correct magic code to get the declension table right, I would be grateful. Хтосьці (talk) 10:50, 7 April 2025 (UTC)Reply

I think I've understood how this is done, бор<c.genu.locý.countý> seems to produce the correct result. Хтосьці (talk) 09:19, 9 April 2025 (UTC)Reply