Module talk:bg-nominal

Discussion of accent and noun inflection

(moved from User talk:Benwing2)

Hi,

Are you able and interested in cleaning up the Bulgarian entries, which are full of manual transliterations. I also wonder if they should use a header similar to Russian or Ukrianian ones (with optional plural forms?). As an example, please have a look at the current роди́на (rodína), which uses a manual transliteration "ródina" (incorrect, according to this dictionary) and no stress mark on the Cyrillic form. This is very typical for Bulgarian entries. Also calling @Erutuon. If there are problematic transliterations, I'm OK to go through a list, if you make one.

BTW, we are using acute accents but some dictionaries use a grave, e.g. the dictionary above has "РОДЍНА". No strong opinion about, which one we should use. --Anatoli T. ^{(обсудить}/^вклад) 04:21, 4 March 2020 (UTC)

@Atitarev It should not be hard to do this as I already have a script to do the same in Russian. I prefer to use an acute accent, and in fact some entries already have that. I can generate a list where the Cyrillic and translit don't match, although there's not much I can do about cases like роди́на (rodína). Benwing2 (talk) 04:33, 4 March 2020 (UTC)

Yes, please, let me know if you need more info. No, we can't do anything about the incorrect information but we can strive to check against dictionaries. I heard that (not sure how true) Bulgarians are much more flexible about different stresses and there are dialectal differences, which are acceptable by other speakers. --Anatoli T. ^{(обсудить}/^вклад) 05:48, 4 March 2020 (UTC)

The Bulgarian new L2 was made by an IP in diff. I've added the manual translit later myself based on the IPA. I am going to correct it. I will just use "head=" for now. --Anatoli T. ^{(обсудить}/^вклад) 05:54, 4 March 2020 (UTC)

Apologies for many edits and stupid mistakes. --Anatoli T. ^{(обсудить}/^вклад) 05:59, 4 March 2020 (UTC)

Hi. Just bumping to check if I have progressed in the queue :) --Anatoli T. ^{(обсудить}/^вклад) 01:37, 11 March 2020 (UTC)

@Atitarev Apologies, I wrote part of the code and then got distracted. Let me see if I can finish this tonight. Benwing2 (talk) 01:39, 11 March 2020 (UTC)

No rush at all. I just wanted to know if it's still on the plan. Thank you! --Anatoli T. ^{(обсудить}/^вклад) 01:44, 11 March 2020 (UTC)

@Atitarev It's running now on CAT:Bulgarian lemmas. It should finish in an hour or so. Please let me know if you see any mistakes. Benwing2 (talk) 03:55, 13 March 2020 (UTC)

Thank you very much! --Anatoli T. ^{(обсудить}/^вклад) 04:34, 13 March 2020 (UTC)

@Atitarev CAT:Bulgarian non-lemma forms is running. After that I will see about cleaning the headword templates so they look more like the Russian ones. Benwing2 (talk) 06:10, 13 March 2020 (UTC)

BTW see User:Benwing2/bg-lemma-warnings. These are all the warnings that came out of the run on CAT:Bulgarian lemmas. There are about 100 of them. I fixed up the first few but I'm missing info on some of them; do you have a reference to a Bulgarian dictionary with stresses listed, ideally with the non-predictable inflections (e.g. cases where the stress moves onto the definite article when appended) and ideally online? Benwing2 (talk) 06:17, 13 March 2020 (UTC)

BTW in the above list, the name of the template and the parameter in question is found before the word "WARNING" or "Unable". Benwing2 (talk) 06:18, 13 March 2020 (UTC)

Cool. Did not know that “we are using acute accents” (seemed rather that we used chaos) and I used graves because this seems somewhat more common (perhaps ⅔ grave to ⅓ acute) but here it has often confused me to do one thing for East Slavic and another for Bulgarian, so good that it is now decided for the acute. Fay Freak (talk) 17:12, 13 March 2020 (UTC)

Thanks for completing this! I will help clean-up the cases.

I don't have a definite guide for Bulgarian stress patterns. There are a few unpredictable stresses in inflected forms.

1. According to this discussion, feminines ending in consonants have the definite ending stressed, e.g. есента́ (esentá, “the autumn”) but otherwise, the stress remains on the base word кни́гата (knígata, “the autumn”). есента́ (esentá) is actually pronounced есентъ́ (esentǎ́).
2. This discionary seems good, I am able to find inlfected forms with stresses and some inflections - only selected forms are stressed. E.g. есен видя. вѝдя -> вѝдиш, мин. св. видя̀х, мин. прич. видя̀л

Please check this work and this Wikipedia article. --Anatoli T. ^{(обсудить}/^вклад) 08:01, 14 March 2020 (UTC)

@Atitarev Thanks. I have cleaned up Bulgarian nouns, proper nouns, verbs, and adjectives to use {{bg-noun}}, {{bg-proper noun}}, {{bg-verb}} and {{bg-adj}}, which now take params similar to their Russian equivalents. As for stress patterns, do you know if there are adjectives or participles in Bulgarian that work like Russian accent c (ending stress on the feminine singular, stem stress on other forms) or verbs that follow Russian accent c (ending stress on the first singular present, stem stress on the other present-tense forms)? I'm not sure if https://rechnik.chitanka.info consistently indicates this. It would be great if there were a site that indicated the stress pattern on all verb forms, as I'm not too familiar with the way Bulgarian verbs work and they're quite complicated. Benwing2 (talk) 17:55, 14 March 2020 (UTC)

@Fay Freak Do you know? Benwing2 (talk) 17:56, 14 March 2020 (UTC)

@Benwing2: No, my understanding of Bulgarian is merely extended from the Serbo-Croatian studies and my Russian and I have not really read Bulgarian grammar, and I have probably never listened to Bulgarian at all. Fay Freak (talk) 18:11, 14 March 2020 (UTC)

@Benwing2: Can you light me up what is the new standard for Bulgarian? It seems that head and a= (for verbs), g= (for nouns) are no longer necessary, aren't they? Will the confusing names of inflectional templates remain? Also, as far as I get, the stress will be given as acute from now on? All modern Bulgarian dictionaries (Български етимологичен речник, Речник на българския език, rechnik.chitanka. info, etc.) and literature however use grave. It's inevitable that some new user will add a grave at some point. User:Bezimenen 19:37, 14 March 2020 (UTC)

@Bezimenen I have tried to make the templates work like the existing Russian ones. Like this: {{bg-noun|HEAD|GENDER}}, {{bg-verb|HEAD|ASPECT}}, {{bg-adj|HEAD}}. As for accents, I made them acute to follow Russian practice. If grave is the standard used elsewhere, we can do one of three things: (1) switch to use grave (in which case, what should be used for secondary stress in {{bg-IPA}}? acute is currently used but that seems very strange); (2) allow grave as an equivalent of acute, and convert to acute; (3) throw an error if grave is used. I haven't touched the inflectional templates yet; the problem is that I don't have a good reference as to what the possible grammatical variants or stress patterns are, so I don't know how to restructure them. If you have a reference, please let me know. I found something called "Граматика на съвременния български книжовен език" by Stoyan Stoyanov that may be what I'm looking for, but it's written in Bulgarian so I would need some pointers as to which volume to look in and where to look in those volumes. Benwing2 (talk) 20:03, 14 March 2020 (UTC)

We can’t throw an error if grave is used, because this is for secondary stress then. Though the only time I marked one for Bulgarian is at the descendants of German Fräsmaschine. “All Bulgarian dictionaries” is an exaggeration. As I already implied, it is used preponderatingly in native Bulgarian works, but here we do not make native Bulgarian works and it is confusing for editors to do the opposite in Bulgarian as for East Slavic (and really many contributors in Slavic languages edit all of them) and for readers, who always have to mentally switch otherwise. One must be conscious that we create something different than everybody else does. Fay Freak (talk) 20:28, 14 March 2020 (UTC)

@Fay Freak We could throw an error if grave is used without acute in the same word, which should catch most uses, or just disallow showing secondary stress anywhere but in {{bg-IPA}} (we don't normally show secondary stress in Russian headwords, inflections or links, for example). Benwing2 (talk) 20:38, 14 March 2020 (UTC)

By the way, you forgot to treat |head2= with your bot, as on левурда (levurda). Fay Freak (talk) 20:58, 14 March 2020 (UTC)

I would strongly advise against the third approach (throw an error if grave is used), because that would jeopardise the only correct spelling of ѝ (vide infra). I am also almost sure (but do not have a source to corroborate it right now) that по in combination with nouns (по̀ юнак - a more valiant fellow) is always written with a grave accent and is the second example that crossed my mind where the stress is explicitly denoted. Bogorm converſation 17:33, 16 March 2020 (UTC)

@Benwing2: The new changes are useful, indeed. Bulgarian inflectional templates also will benefit from some refreshing.

Regarding the accent: Bulgarian has stress accent, so it does not matter much which symbol is used. The current notation chosen by BAS uses grave (both for primary and secondary stress), but I don't think many people will bother if acute is used instead. In fact, old dictionaries did use acute (e.g. Gerov's dictionary). Pick whatever suits other editors. I didn't mean to advocate for either option.

PS: Volume 1 of "Граматика на съвременния български книжовен език" deals with phonology, though, I don't have access to it and cannot help much. User:Bezimenen 21:20, 14 March 2020 (UTC)

@Bezimenen: That book is on libgen. PUC 21:26, 14 March 2020 (UTC)

@PUC: Thanks. I didn't know about it until now. Безименен 21:34, 14 March 2020 (UTC)

PS I've breezed through the textbook and it discusses accent in section "Ударение" (p. 171). I don't think it will help much, though. The book does not establish accentual types as in Proto-Slavic. In "Място на ударението" (p. 178), the author give ostensive definitions to some common patterns, but there is nothing concrete. Безименен 22:15, 14 March 2020 (UTC)

I have downloaded "Maslov_Yu_S_-_Grammatika_bolgarskogo_yazyka_-_1.djvu" and "Uchebnik_bolgarskogo_yazyka.djvu" (both are easy to find), which give information about stresses, including the inflected forms. Both are in Russian. I can find individual inflected forms and what stresses are used but I'm not sure yet about what stress patterns we should use. The Russian reference mention that nouns with plural endings in -и may get a shift in stress and give some examples and there are also non-productive patterns. I've done one manually (rather awkwardly) for мъж (mǎž) (please see the inflection table). Please give me the link to the Bulgarian grammar. I seem to be able to follow written Bulgarian (more or less). --Anatoli T. ^{(обсудить}/^вклад) 23:12, 14 March 2020 (UTC)

@Atitarev I found "Граматика на съвременния български книжовен език" volume II (which deals with morphology and such) on epdf.pub. It's mentioned above that you can find volume I on libgen, although I'm not sure what that is. Benwing2 (talk) 23:18, 14 March 2020 (UTC)

@Atitarev How many types of patterns are there? If there are only two (one where the accent always stays the same and one where it moves onto all suffixes but the vocative), that should be easy to implement. Although you mention non-productive patterns, so maybe there are more. Benwing2 (talk) 23:20, 14 March 2020 (UTC)

Hiya. Sorry, I haven't ignored you guys or forgotten, the real life got in the way. So far, for nouns we may need two enhanced templates, which have stresses on the endings: for -ове and -и (but these include cases where the final consonant changes or there may be reductions) and one for irregulars, like мъж (mǎž). Shall we have a separate page dedicated to Bulgarian inflection work? We may get some native speakers involved as well. User:Bogorm is not very active but he was always willing to help when I needed advice on Bulgarian.

I'll give you more info and answer any questions when I get a bit more time. --Anatoli T. ^{(обсудить}/^вклад) 22:23, 15 March 2020 (UTC)

These masculine nouns have the stress shifted to the endings (all forms except for the vocative), e.g. връх (vrǎh) -> върхъ́т (vǎrhǎ́t) (definite subject form), върха́ (vǎrhá) (definite object form), върховé (vǎrhové) (plural).

бряг (brjag)}}, вол (vol), враг (vrag), връх (vrǎh), глас (glas), град (grad), грях (grjah), дар (dar), ден (den), дух (duh), дъжд (dǎžd), дъх (dǎh), дял (djal), ек (ek), звук (zvuk), зъб (zǎb), клас (klas), кон (kon), крак (krak), лед (led), лъх (lǎh), лъч (lǎč), мъж (mǎž), нож (nož), плат (plat), плач (plač), плет (plet), плод (plod), ред (red), рог (rog), род (rod), свят (svjat), син (sin), смях (smjah), сняг (snjag), стол (stol), страх (strah), студ (stud), труд (trud), цвят (cvjat), час (čas).

Some change the form in the inflections: e.g. бряг (brjag) -> брегъ́т (bregǎ́t), брега́ (bregá), бреговé (bregové). It is quite complicated, e.g. after numerals the form бря́га (brjága) is used (the stress is on the first syllable). --Anatoli T. ^{(обсудить}/^вклад) 23:20, 15 March 2020 (UTC)

This info came from a Russian site but I don't trust it now! Just checked against the word лъх (lǎh) in rechnik and the stress is shifting only in the definite object form лъха́ (lǎhá), not in the plural: лъ́хове (lǎ́hove). The form with numerals (not in inlfection tables): лъ́ха (lǎ́ha). --Anatoli T. ^{(обсудить}/^вклад) 23:39, 15 March 2020 (UTC)

Ек, стол and рог deserve elevated attention, since both displaced and unwavering (immovable) stress are admissible in the plural: ѐкове and ековѐ, столо̀ве and сто̀лове, and рог even has three plural forms: рога̀, рого̀ве, ро̀гове. Bogorm converſation 17:33, 16 March 2020 (UTC)

@Bogorm: Thanks, this should definitely be taken into account. I have started collected some patterns at User:Atitarev/Bulgarian_nouns (to be updated!). --Anatoli T. ^{(обсудить}/^вклад) 22:59, 16 March 2020 (UTC)

@Atitarev Thanks. It looks like all of the above words have ending stress in the definite subject and object forms, and stem stress in the count form, but varying plurals. Benwing2 (talk) 23:48, 15 March 2020 (UTC)

This site is flawed. I'll used references instead and check against the two dictionaries. --Anatoli T. ^{(обсудить}/^вклад) 23:52, 15 March 2020 (UTC)

Hi, if I may join in the conversation, here are some remarks I would like to add..

1. Regarding the accents, in the Правописен речник на българския книжовен език (1984, 10th edition), the grave accent is used. The same applies to the shortened form of ней: ѝ(dative of тя, she), which is always written with the grave accent and only occasionally, when the keyboard does not contain this symbol, as й.
2. @Atitarev Regarding the accentual flexibility: there are some cases when both the Western and Eastern variety of the stress were accepted in the written language, as обя̀д and о̀бед. In this particular case, the spelling also changes. The plural is always обѐди.
3. @Benwing2 With regard to your question: (“do you have a reference to a Bulgarian dictionary with stresses listed, ideally with the non-predictable inflections (e.g. cases where the stress moves onto the definite article when appended) and ideally online?”) I have the Правописен речник на българския книжовен език (Orthographical dictionary of the Bulgarian literary language), and with a search in Google Books after the word and the name of this dictionary in quotation marks, the relevant excerpt should be available (it was at least to me, but from an older edition).
4. @Atitarev (“есента́ ‎(esentá‎) is actually pronounced есентъ́ ‎(esentǎ́‎).”) According to the discussion some people pronounce it that way. I do not belong to this squad and do not remember having ever heard this reduction, so this assertion is at best partially correct. Bogorm converſation 17:33, 16 March 2020 (UTC)

@Bogorm: Thank you for your insight and info! Just on one pronunciation point (for now) On this video youtu.be/y61agseRVEw (very useful!) at about 25:54 the guy on the video said «и ще остана там за през нощта», which sounded like «нощ(т)ъ́» to me. I guess it doesn’t have to be handled by the module if it’s non-standard. I will try to get the dictionary you recommended. —Anatoli T. ^{(обсудить}/^вклад) 21:09, 16 March 2020 (UTC)

@Fay Freak, Bogorm, Bezimenen, @Benwing2: Hi all, maybe we should consider that all Bulgrian dictionaries use a grave accent, not an acute accent? The Russian school uses the acute accent, also in Bulgarian-Russian dictionaries. If we need to use the secondary accent in Module:bg-pronunciation, we could use the reverse, instead of {{bg-IPA|а̀втомагистра́ла}} (current), {{bg-IPA|а́втомагистра̀ла}}. Please note that secondary accent is not handled thoroughly anywhere in Russian or Bulgarian sources in terms of markings. Both Russian and Bulgarian dictionaries use just one type of accent for the main, alternative or secondary accents. For example, you won't find Russian "а̀втомагистра́ль" anywhere outside Wiki. If it's too much rework, forget it but it's also easier to Google for stressed forms if we implement the grave accent. --Anatoli T. ^{(обсудить}/^вклад) 22:59, 16 March 2020 (UTC)

 @Atitarev, Bogorm: Not sure what to consider. I have already considered it as shown above, and the argument stands that it is confusing to do one thing for Russian/Ukrainian/Belarusian and the reverse thing for Bulgarian, and it is good that it is uniform now. Or is it absolutely odd for Bulgarians to see a grave for primary stress marking? Fay Freak (talk) 23:33, 16 March 2020 (UTC)

@Fay Freak, Atitarev, Bogorm: It's not a huge amount of work at this stage to change acutes to graves. Note that originally the pronunciation module did implement primary = grave, secondary = acute, and I reversed it. If we do decide to implement primary = grave, I would not like to see secondary = acute; this is just too confusing given that Russian and English tradition is precisely the opposite. Instead, we should use some other symbol, e.g. double grave. Note that there's a tradition in English of using acute = primary, grave = secondary, which is mentioned in Wikipedia's article on secondary stress:

Another tradition in English is to assign acute and grave accents for primary and secondary stress, respectively: pronùnciátion.

It could be argued that since Wiktionary is fundamentally an English dictionary, we should follow this tradition. (On the other hand, there's also a tendency in Wiktionary to respect native traditions.) Benwing2 (talk) 00:52, 17 March 2020 (UTC)

OK, let's focus on getting this done. --Anatoli T. ^{(обсудить}/^вклад) 01:37, 17 March 2020 (UTC)

@Atitarev Sorry, getting what done? Benwing2 (talk) 02:04, 17 March 2020 (UTC)

I mean what you have started - gradual modularisation of Bulgarian inflections, if you're still interested. --Anatoli T. ^{(обсудить}/^вклад) 02:09, 17 March 2020 (UTC)

@Atitarev Yup, I'll be working on that. Benwing2 (talk) 02:39, 17 March 2020 (UTC)

@Benwing2: (“If we do decide to implement primary = grave (...) we should use some other symbol.” [for secondary]) Yes, I agree with this approach. Bogorm converſation 11:19, 17 March 2020 (UTC)

@Atitarev You asked on нощта (noštta) about adding pronunciations and stresses to inflected forms; for me it would be more helpful if you could expand User:Atitarev/Bulgarian nouns. You've indicated different special cases for plurals; can you do the same for definites, vocatives, count forms? Bulgarian nouns does a pretty good job of enumerating lots of special cases but maybe there are more. Especially helpful are pairs of nouns that are similar but differ in some respect, e.g. Wiktionary:About Bulgarian indicates that град and бик are similar in taking -ове, but град has vocative граде (гра́де?) while бик has vocative бико (би́ко?). This tells me that, unless somehow or other the final consonant predicts the vocative ending, I need a way of specifying the vocative ending. Benwing2 (talk) 04:31, 17 March 2020 (UTC)

I see. OK, I will provide the main forms of interest (the ones I can find). The stress in the vocative is quite predictable (the actual form may not). It can't fall on the last syllable and it shifts to the front. This article is good in showing how to form vocatives. The count forms should be predicatable as well (e.g. три ча́са), as opposed to the definite object form "часа́" but I don't have its full understanding yet. --Anatoli T. ^{(обсудить}/^вклад) 04:53, 17 March 2020 (UTC)

You may also find this "Maslov_Yu_S_-_Grammatika_bolgarskogo_yazyka_-_1.djvu" useful. It's not searchable but it has a decent grammar coverage. Unfortunately, I can't find online the grammar book Bogorm mentioned. --Anatoli T. ^{(обсудить}/^вклад) 04:59, 17 March 2020 (UTC)

OK, count forms are mostly predictable, it's unstressed ending -а/-я - де́сет ве́стника (ten newspapers), че́тири кра́я (four edges). Exception: три пъ́ти (three times) (път's inflection depends on senses). No vowel reduction (dropping) happens: пет теа́търа, се́дем оре́ла, suffixes are not dropped - пет бъ́лгарина - five Bulgarians. There are some exceptions: литър - два ли́тра, метър - осем ме́тра. The modern count form of ден (day) is де́на but де́ня is also used, "дни" is also used, so ден is a special case. I won't be adding count forms to my list, unless I find something unusual.--Anatoli T. ^{(обсудить}/^вклад) 06:04, 17 March 2020 (UTC)

@Atitarev: (“пет българина” — I only payed attention today to this misformation) Human male persons do not have count forms and they are always in accordance with the respective form of the numeral (in this case, петѝма), but the count form is inadmissible even without this form of the numeral. If I remember aright, in Russian, these forms (трое, четверо ... девятеро) are available for one-digit numbers and десятеро, and beyond that the main form of the numeral is used (двадцать два болгарина). In Bulgarian there is no such restriction (двадесет и четирѝма бъ̀лгари) and separate forms of the numerals for male human persons are always used (except for numerals where the last digit is 1, 5, 6, 7, 8, 9: двадесет и един българи, двадесет и седем българи, двадесет и двама българи, cf.: стотима българи, and big powers of 10: хиляда/милион... българи). Here is an informative and explanatory table: Числителни имена в българския език (PDF). In any case, nouns designating male human beings do not have count forms. Bogorm converſation 10:51, 21 March 2020 (UTC)

@Benwing2: (“бик has vocative бико (би́ко?).”) Yes, the stress falls on the first syllable. Another examples are лъ̀ко (bow), вѝко (clamour) or лу̀ко (onion), but the latter does not have plural form.

@Atitarev: The difference between два часа̀ and два ча̀са is in the meaning as well. In the first case two o'clock, in the second two hours. Bogorm converſation 11:19, 17 March 2020 (UTC)

@Bogorm: Great, thank for the explanation! Would the "o'clock" sense, the form часа́ (časá) be correctly defined as "objective form of the singular definite form of час"? --Anatoli T. ^{(обсудить}/^вклад) 12:33, 17 March 2020 (UTC)

@Atitarev, Bogorm I have tried to classify stress movement:

1. End-stressed in neither: (a)

   ек, pl е́кове, def sg е́кът/е́ка

   ези́к, pl ези́ци, def sg ези́кът/ези́ка

   теа́тър, pl теа́три, def sg теа́търът/теа́търа

   most multisyllabic words

2. End-stressed in both pl and def sg: (b)

   град, pl градове́, def sg градъ́т/града́; also враг, глас

   студ, pl студове́, def sg студъ́т/студа́

   бря́г, pl брeгове́, def sg брегъ́т/брега́, also грях

   гръб, pl гърбове́, def sg гърбъ́т/гърба́; also връх

   ред, pl редове́, def sg редъ́т/реда́

   лъч, pl лъчи́, def sg лъчъ́т/лъча́

   крак, pl крака́, def sg кракъ́т/крака́

   мъж, pl мъже́, def sg мъжъ́т/мъжа́

3. End-stressed in pl but not def sg: (c)

   вя́тър, pl ветрове́, def sg вя́търът/вя́търа

   бое́ц, pl бойци́, def sg бое́цът/бое́ца

   стол, pl столо́ве, def sg сто́лът/сто́ла; also вол, нож

   рог, pl рога́, def sg ро́гът/ро́га

   кон, pl коне́, def sg ко́нят/ко́ня

   княз pl князе́, def sg кня́зът/кня́за (archaic кня́зове)

4. End-stressed in def sg but not pl: (d)

   ця́лост, pl ця́лости, def sg целостта́

   любо́в, pl любо́ви, def sg любовта́

   зъб, pl зъ́би, def sg зъбъ́т/зъба́; also клас dial/archaic pl кла́си (normally класове́)

   лъх, pl лъ́хове, def sg лъхъ́т/лъха́; also дар (can also have no stress movement); дух, звук

   дял, pl дя́лове, def sg делъ́т/дела́

   труд, pl тру́дове, def sg трудъ́т/труда́

This suggests that we need to have a system of accent similar to Russian a/b/c/d. It appears that -о́ве vs. -ове́ is mostly predictable (-о́ве in type (c) except for вя́тър [any others?], -ове́ in type (b)), as is я -> е (always occurs when stress moves off of я?), and ръ -> ър (always occurs with any ending?). There are also weird cases like folk-poetic ве́три plural of вя́тър, with я -> е without stress shift; this also seems common in adjectives. Benwing2 (talk) 02:10, 19 March 2020 (UTC)

@Benwing2. Great analysis and plan to use stress pattern subclasses, I think. With class d, it seems that all feminine nouns ending in a consonant will fall into it любо́в - любовта́. The nouns as you may know, may fall into multiple subclasses, e.g. зъ́би or зъби́ (BTW, ъ́ with an acute is usually not visible if followed by other symbols - when typing). Category:Bulgarian noun inflection-table templates do not cater for any stress subclasses but they can be use to form classes. Some can be merged if the interchange of consonants is handled by the module.

I couldn't get the dictionary Bogorm mentioned, so we will have to rely on other sources - the two online dictionaries are good but not totally comprehensive - they don't fully show all possible forms. @Benwing2, I recommend to download the two djvu files, I find them useful but they are based on Russian. Your written Russian must be above average now, though? --Anatoli T. ^{(обсудить}/^вклад) 02:55, 19 March 2020 (UTC)

@Atitarev Yeah, Russian is easier than Bulgarian to read, for sure. I already downloaded one of the two DJVU files, I'll get the other. For me, there isn't the issue of зъ́би not showing the acute accent. Does зъ̀би show with the grave accent? Benwing2 (talk) 03:35, 19 March 2020 (UTC)

The issue is when I TYPE (edit mode), not when I post here and view. One of the reasons, if I remember correctly, for WT:BG TR to use a grave for "ъ" was the a display issue. зъ̀би is more visible but the symbol is inside "б". I have now changed WT:BG TR to match our current practice. --Anatoli T. ^{(обсудить}/^вклад) 03:46, 19 March 2020 (UTC)

@Benwing2: Ек and стол allow another plural form apart from the already listed: ековѐ and сто̀лове.

( "клас dial/archaic pl кла́си (normally класове́)" ) Кла̀си is the plural form of кла̀са (class with the meaning "social group", "admirable behaviour", "category of seats in a train/airplane"), whereas класовѐ is the plural form of клас (class with the meaning: "a class of pupils", "a taxonomical rank for organisms" and meaning from the set theory). Bogorm converſation 10:51, 21 March 2020 (UTC)

Declension of nouns

Masculine

Noun	Meaning	Pl	Def Sg	Count	Vocative	Notes
End-stressed in neither: (a)
ек<ad>	"echo, thud"	е́кове, екове́	е́кът; е́ка	е́ка	no voc?	pl екове́
кър<+ове+ища>	"field" (dated)	къ́рове, къ́рища	къ́рът; къ́ра	къ́ра	no voc?
дар<ad+ове+и[fp]/count:да́ри>	"gift"	да́рове, дарове́ (folk-poetic да́ри)	да́рът; даръ́т; да́ра/дара́)	да́ри, да́ра	no voc? да́ре? (stress?)
стол<>	"canteen"	сто́лове	сто́лът; сто́ла	сто́ла	no voc?
блок<>	"block"	бло́кове	бло́кът; бло́ка	бло́ка	no voc?
град<+и>	"grad"	гра́ди	гра́дът; гра́да	гра́да	no voc?
гост<(v)+и+е[d]>	"guest"	го́сти (dialectal го́сте)	го́стът; го́ста	го́ста	го́сте (stress?)
клон<+и+ове+ища+е[d]>	"branch"	кло́ни/кло́нове/кло́нища (dialectal кло́не)	кло́нът; кло́на	кло́на	no voc? кло́не?
храст<+и+е[collective]>	"bush"	хра́сти (collective хра́сте)	хра́стът; хра́ста	хра́ста	no voc? хра́сте?
пръч<>	"billy goat"	пръ́чове	пръ́чът; пръ́ча	пръ́ча	no voc?
грък<ad(v)(ър)+и>	"Greek man"	гъ́рци	гъ́ркът/гъркъ́т; гъ́рка/гърка́	гъ́рка	гъ́рко
край<+ища+ьове[r]>	"edge"	кра́ища (rare кра́йове)	кра́ят; кра́я	кра́я	no voc? кра́ю?	(type (ь) inferred)
змей<(v)+еве+ьове>	"dragon; kite"	зме́еве/зме́йове	зме́ят; зме́я	зме́я	зме́ю?	(type (ь) inferred)
брой<+еве+ьове[a]>	"number"	бро́еве (archaic бро́йове)	бро́ят; бро́я	бро́я	no voc???	(type (ь) inferred)
геро́й</voc:+:геро́йо>	"hero"	геро́и	геро́ят; геро́я	геро́я	геро́ю/геро́йо
чове́к</pl:хо́ра:чове́ци[a]:лю́де[a]/voc:чове́че>	"person"	хо́ра (archaic чове́ци/лю́де)	чове́кът; чове́ка	чове́ка	чове́че
ка́мък</pl:ка́мъни:ка́мъне[a][d]:ка́мъци[a][d]>	"stone"	ка́мъни (archaic/dialectal ка́мъне/ка́мъци)	ка́мъкът; ка́мъка	ка́мъка	no voc???
двига́тел<>	"engine"	двига́тели	двига́телят; двига́теля	двига́теля	no voc?	(type (ь) inferred for nouns ending in vowel + unstressed -тел; type +и inferred for multisyllabic nouns)
ле́кар<(v)>	"(male) doctor"	ле́кари	ле́карят; ле́каря ле́каря	ле́карю	(type (ь) inferred for nouns ending in vowel + unstressed -ар; type +и inferred for multisyllabic nouns)
път<(ь)+ища>	"road"	пъ́тища	пъ́тят; пъ́тя	пъ́тя	пъ́те?
път<+и>	"time/occurrence"	пъ́ти	пъ́тят; пъ́тя	пъ́ти??	пъ́ти?? no voc?
зет<(ь)/voc:зе́тко:зе́то[d]:+[d]>	"son-in-law, brother-in-law"	зе́тьове	зе́тят; зе́тя	зе́тя	зе́тко (dialectal зе́то/зе́тьо [or зе́тьо not dialectal?])
брат</pl:бра́тя:бра́те[d]:бра́кя[d]:бра́йкя[d]:бра́тия[a]/voc:+:бра́тко>	"brother"	бра́тя (dialectal бра́те/бра́кя/бра́йкя, archaic бра́тия)	бра́тът; бра́та	бра́та	бра́те/бра́тко
ези́к<>	"tongue, language"	ези́ци	ези́кът; ези́ка	ези́ка	ези́ко? no voc??
теа́тър<*>	"theater"	теа́три	теа́търът; теа́търа	теа́търа	no voc??
ко́съм<*>	"hair"	ко́сми	ко́съмът; ко́съма	ко́съма	no voc??
органи́зъм<(v)>	"organism"	органи́зми	органи́змът; органи́зма	органи́зъма	органи́зме?	(type *(d*), or type *(d)(v*) if (v) is given, inferred for nouns ending in vowel + unstressed зъм)
ста́рец<*/voc:ста́рче>	"old man"	ста́рци	ста́рецът; ста́реца	ста́реца	ста́рче
аташе́<(m)>	"attaché"	аташе́та	аташе́то	no count form?	no voc?
ори́з</n:sg>	"rice"	no pl (singulare tantum)	ори́зът; ори́за	no count form	no voc??
Пе́тър<*(v*)>	"Peter"	Пе́тровци	no def sg form (def pl Пе́тровците)	no count form??	Пе́тре
Дими́тър<*(v*)>	"Dimitri"	Дими́тровци	no def sg (def pl Дими́тровците)	Дими́търа??? no count???	Дими́тре
Ива́н<(v)>	"Ivan"	Ива́новци	no def sg (def pl Ива́новците)	Ива́на	Ива́не
ту́рчин<(ин)/pl:ту́рци/voc:~о>	"Turk"	ту́рци	ту́рчинът; ту́рчина	ту́рчина	ту́рчино
боди́л<>	"thistle"	боди́ли	боди́лът; боди́ла	боди́ла	no voc???
ме́дник</pl:ме́нци:ме́дници>	"copper pot"	ме́нци, ме́дници	ме́дникът; ме́дника	ме́дника	no voc???
наро́д<(v)+и+ища[a]>	"people, nation"	наро́ди (archaic наро́дища)	наро́дът; наро́да	наро́да	наро́де
юна́к</voc:юна́че:+>	"hero"	юна́ци	юна́кът; юна́ка	юна́ка	юна́че/юна́ко
о́рех<++и>	"walnut; walnut tree"	о́рехи	о́рехът; о́реха	о́реха	no voc??
чех<(v)++и>	"Czech man"	че́хи	че́хът; че́ха	че́ха	no voc?? (really??)
успе́х<++и>	"success"	успе́хи	успе́хът; успе́ха	успе́ха	no voc??
франк<(v)++и>	"Frank (ancient tribe)"	фра́нки	фра́нкът; фра́нка	фра́нка	фра́нко
франк<>	"franc"	фра́нкове	фра́нкът; фра́нка	фра́нка	no voc??
о́черк</pl:+:о́черки[r]>	"article, story"	о́черци (rare о́черки)	о́черкът; о́черка	о́черка	no voc??
профе́сор<(v)>	"professor"	профе́сори	профе́сорът; профе́сора	профе́сора	профе́соре
тъпа́к<(v)>	"dolt, stupid person"	тъпа́ци	тъпа́кът; тъпа́ка	тъпа́ка	тъпа́ко?
продава́ч<(v)>	"(male) vendor"	продава́чи	продава́чът; продава́ча	продава́ча	продава́чо
ни́каквец<*(v)/pl:ни́какъвци>	"despised person, nonentity (pej.)"	ни́какъвци	ни́каквецът; ни́каквеца	ни́каквеца	ни́каквецо
бъ́лгарин<(ин)(v)>	"Bulgarian"	бъ́лгари	бъ́лгаринът; бъ́лгарина	бъ́лгарина	бъ́лгарино
прия́тел<(v)>	"(male) friend, boyfriend"	прия́тели	прия́телят; прия́теля	прия́теля	прия́телю
по́длог<>	"subject"	по́длози	по́длогът; по́длога	по́длога	no voc??
((мо́лив<>,моли́в<>))	"pencil"	мо́ливи/моли́ви	мо́ливът/моли́вът; мо́лива/моли́ва	мо́лива/моли́ва	no voc???
кожу́х<>	"fur coat"	кожу́си	кожу́хът; кожу́ха	кожу́ха	no voc???
ми́тинг<>	"meeting"	ми́тинги	ми́тингът; ми́тинга	ми́тинга	no voc???
ло́зунг<>	"slogan"	ло́зунги	ло́зунгът; ло́зунга	ло́зунга	no voc???
пода́рък</pl:+:пода́рки[a]>	"gift"	пода́ръци (archaic пода́рки)	пода́ръкът; пода́ръка	пода́ръка	no voc???
пото́мък<*(v*)/pl:+:пото́мки[a][l]>	"heir, descendant"	пото́мци (archaic/literary пото́мки)	пото́мъкът; пото́мъка	пото́мъка	пото́мко
мързела́н<(v)+и+овци>	"slacker, lazy person"	мързела́ни/мързела́новци	мързела́нът; мързела́на	мързела́на	voc???
Го́спод</n:sg/voc:~и/dat:~у>	"Lord"	no pl	Го́сподът; Го́спода (???)	no count	Го́споди	dat Го́споду
ли́тър<*(c*)>	"liter"	ли́три	ли́търът; ли́търа	count ли́тра	no voc?
ме́тър<*(c*)>	"meter"	ме́три	ме́търът; ме́търа	count ме́тра	no voc?
ме́тър<*>	"type of playing card (?)"	ме́три	ме́търът; ме́търа	ме́търа	no voc?
баща́<(m)(v)>	"father"	бащи́	баща́та	count as pl?	voc??? possibly ба́що???	masculine
влади́ка<(m)(v)>	"bishop"	влади́ци	влади́ката	count as pl?	влади́ко	-ка -> -ци because masculine
съдия́<(m)(v)/pl:съди́и>	"judge"	съди́и	съдия́та	count as pl?	voc??? possibly съди́йо???	masculine; should be able to construct vocative съди́йо automatically
дя́до<(m)(v)/pl:+:деди́[p]:дя́дове[a]>	"grandfather, elder"	дя́довци (poetic деди́, archaic дя́дове)	дя́дото	count as pl?	дя́до???	masculine
чи́чо<(m)(v)>	"paternal uncle"	чи́човци	чи́чото	count as pl?	чи́чо???	masculine
ву́йчо<(m)/pl:+:ву́йчевци[d]:ву́йчови[d]:ву́йчеви[d]:ву́йчи[d]:ву́йци[d]/voc:+:ву́йче[d][endearing]>	"maternal uncle"	ву́йчовци (dialectal ву́йчевци/ву́йчови/ву́йчеви/ву́йчи/ву́йци)	ву́йчото	count as pl?	vocative (dialectal/endearing) ву́йче; normal vocative ву́йчо???	masculine
End-stressed in both pl and def sg: (b)
град<b(v)+ове+ища[r]>	"city"	градове́ (rarely гради́ща)	градъ́т; града́	гра́да	гра́де?	(type b with plural ове automatically infers +ове́; type c with plural ове automatically infers +о́ве; types b and c with plural ища automatically infer +и́ща)
враг<b+ове+и[p]/voc:вра́же>	"enemy"	врагове́ (poetic врази́)	врагъ́т; врага́	вра́га	вра́же
глас<b(v)>	"voice"	гласове́	гласъ́т; гласа́	гла́са	гла́се???
клас<b,d+и[a][d]>	"ear of corn"	класове́ (archaic/dialectal кла́си)	класъ́т; класа́	кла́са	no voc???
клас	"grade; class, category"	класове́	класъ́т; класа́	кла́са	no voc???
плат	"cloth, fabric"	платове́	платъ́т; плата́	пла́та	no voc???
плач	"cry, crying"	плачове́	плачъ́т; плача́	пла́ча	no voc???
страх	"fear"	страхове́	страхъ́т; страха́	стра́ха	no voc???
час	"hour"	часове́	часъ́т; часа́	ча́са	no voc???
град	"hail"	no pl?	градъ́т; града́	no count?	no voc? гра́де?	singulare tantum
студ	"(the) cold"	студове́	студъ́т; студа́	сту́да? no count?	сту́де? no voc?
дъжд<b(v)>	"rain"	дъждове́	дъждъ́т; дъжда́	дъ́жда	дъ́жде?? no voc?
син	"son"	синове́	синъ́т; сина́	си́на	си́не/си́нко
бря́г	"coast, shore"	брeгове́	брегъ́т; брега́	бря́га	no voc?	(type (я) i.e. я -> е when unstressed, is automatically inferred)
also грях "sin", свят "world", смях "laughter", сняг "snow", цвят "flower"
гръб<b(ър)>	"back, rear"	гърбове́	гърбъ́т; гърба́	гъ́рба	no voc?
връх<b(ър)+ове+ища[r]>	"top"	върхове́ (rarely върхи́ща)	върхъ́т; върха́	въ́рха	no voc?
ред	"row"	редове́	редъ́т; реда́	ре́да	no voc?
лед	"ice"	ледове́	ледъ́т; леда́	ле́да	no voc?
бой<c(v)+еве́+ове[a][d]>	"fight"	боеве́ (archaic/dialectal бойо́ве)	бо́ят; бо́я	бо́я	бо́ю???
плет<b+ове+ища>	"fence"	плетове́/плети́ща	плетъ́т; плета́	пле́та	no voc?
плод<b(v)>	"fruit"	плодове́	плодъ́т; плода́	пло́да	пло́де?
лъч<b+и>	"ray, beam"	лъчи́	лъчъ́т; лъча́	лъ́ча	no voc?
ден<b(ь)(v)/pl:дни:де́ньове[a][d]:дене́[a][d]/count:де́на:дни:де́ня[a][d]>	"day"	дни (also archaic/dialectal де́ньове, дене́)	деня́т; деня́	count де́на/дни (also archaic/dialectal де́ня)	де́не?? stress??
крак<b+а>	"leg, foot"	крака́	кракъ́т; крака́	кра́ка	no voc?
мъж<b+е/voc:мъ́жо>	"man, husband"	мъже́	мъжъ́т; мъжа́	мъ́жа	мъ́жо
съд<b+илища>	"court"	съди́лища	съдъ́т; съда́	съ́да	no voc??
граде́ц</pl:градовце́>	"small city"	градовце́	граде́цът; граде́ца ?? stress??	граде́ца	no voc???
листе́ц</pl:листца́:листица́:листовце́[a]>	"small leaf"	листца́/листица́ (archaic листовце́)	листе́цът; листе́ца ?? stress??	листца́ (???)	no voc???
гласе́ц</pl:гласовце́[d]>	"little voice, weak voice"	гласовце́ (dialectal)	грасе́цът; грасе́ца ?? stress??	гласе́ца (dialectal? nonexistent?)	no voc???
End-stressed in pl but not def sg: (c)
вя́тър<c*/pl:ветрове́:ве́трища[r]:ве́три[fp]/voc:ве́тре>	"wind"	ветрове́ (rare ве́трища, folk-poetic ве́три)	вя́търът; вя́търа	вя́търа	ве́тре?? (?? stress)
оте́ц<*/voc:о́тче>	"father"	отци́	оте́цът; оте́ца	оте́ца	о́тче	gen/acc (obsolescent) отца́
хубаве́ц<*(v)>	"good-looking man"	хубавци́	хубаве́цът; хубаве́ца	хубаве́ца	хубаве́цо
чуждене́ц<*(v)>	"(male) foreigner, stranger"	чужденци́	чуждене́цът; чуждене́ца	чуждене́цо
бое́ц<c*(v)>	"soldier; fighter"	бойци́	бое́цът; бое́ца	бое́ца	бое́цо
оре́л<*/voc:о́рльо>	"eagle"	орли́	оре́лът; оре́ла	оре́ла	о́рльо
стол<c>	"chair"	столо́ве	сто́лът; сто́ла	сто́ла	no voc??
вол<c(v)>	"ox"	воло́ве	во́лът; во́ла	во́ла	во́ле???
нож<c(v)+ове+еве[d]>	"knife"	ножо́ве (dialectal ноже́ве)	но́жът; но́жа	но́жа	но́же???
бог<c(v)+ове́/pl:богове́:бо́зи[a]/dat:бо́гу>	"god"	богове́ (archaic бо́зи)	бо́гът; бо́га	бо́га	бо́же	dative бо́гу
о́гън<c*(ь)+ьове:+еве[d]/count:о́гъня:о́гня[d]/voc:о́гньо>	"fire, flame"	огньо́ве (dialectal огне́ве)	о́гънят; о́гъня	о́гъня (dialectal о́гня)	о́гньо (correct?? stress??)
жрец<c(v)+и>	"pagan priest"	жреци́	жре́цът; жре́ца	жре́ца	жре́цо
рог<c/pl:рога́:ро́гове[a]:роги́[d]/count:ро́га:ро́гове[a]>	"horn"	рога́ (archaic ро́гове, dialectal роги́)	ро́гът; ро́га	count ро́га (archaic ро́гове)	no voc?
лист<c/pl:листа́:ли́сти[p]:ли́сте[d]:листя́[d]:ли́стове[a][r]>	"leaf"	листа́ (poetic ли́сти, dialectal ли́сте/листя́, archaic/rare ли́стове)	ли́стът; ли́ста	ли́ста	ли́сте?? no voc??
но́мер<c+а>	"number, size"	номера́	но́мерът; но́мера	но́мера	no voc?
кон<c(ь)(v)+е,ac(ь)+и[d],ac(ь)+ьове[d]>	"horse"	коне́ (also dialectal ко́ни́/ко́ньо́ве)	ко́нят; ко́ня	ко́ня	ко́ню (stress??)
цар<c(ь)+е/voc:царю́>	"czar, king"	царе́	ца́рят; ца́ря	ца́ря	царю́
крал<c(ь)+е/voc:кра́лю:кралю́>	"king"	крале́	кра́лят; кра́ля	кра́ля	кра́лю/кралю́
княз<c(v)/pl:князе́:кня́зове[a]>	"prince"	князе́ (archaic кня́зове)	кня́зът; кня́за	кня́за	кня́же
господи́н<c/pl:господа́:господи́новци[pej]/voc:~е>	"sir, gentleman"	господа́ (pejor. господи́новци)	господи́нът; господи́на	господи́на	господи́не
боди́л<c/pl:бодли́:боди́ли[a][d]>	"thorn, spine"	бодли́ (archaic/dialectal боди́ли)	боди́лът; боди́ла	боди́ла	no voc???
мо́мък<c*(v*)>	"youth; bachelor"	момци́	мо́мъкът; мо́мъка	мо́мъка	мо́мко
мъдре́ц<c(v)/pl:+:мъдърци́[r]>	"wise man; wisdom tooth"	мъдреци́ (archaic мъдърци́)	мъдре́цът; мъдре́ца	мъдре́ца	мъдре́цо
храбре́ц<c(v)>	"brave man"	храбреци́	храбре́цът; храбре́ца	храбре́ца	храбре́цо
подле́ц<c(v)>	"scoundrel"	подлеци́	подле́цът; подле́ца	подле́ца	подле́цо
бегле́ц<c(v)/pl:бегълци́:+[r]>	"(male) fugitive"	бегълци́ (rare беглеци́)	бегле́цът; бегле́ца	бегле́ца	бегле́цо
мъртве́ц<c(v)/pl:мъртъвци́:+[a][d]>	"dead man"	мъртъвци́ (archaic/dialectal мъртвеци́)	мъртве́цът; мъртве́ца	мъртве́ца	мъртве́цо
End-stressed in def sg but not pl: (d)
зъб<d(v)+и/pl:+:зъби́(c)>	"tooth"	зъ́би (colloquial зъби́)	зъбъ́т; зъба́	зъ́ба	зъ́бе??
звук<d+и+ове>	"sound"	зву́ци (also зву́кове)	звукъ́т; звука́	зву́ка	no voc??
лъх<d>	"breeze" (poetic)	лъ́хове	лъхъ́т; лъха́	лъ́ха	no voc??
дъх<d>	"breath"	дъ́хове	дъхъ́т; дъха́	дъ́ха	no voc??
съд<d>	"vessel, dish"	съ́дове	съдъ́т; съда́	съ́да	no voc??
кръг<d>	"circle"	кръ́гове	кръгъ́т; кръга́	кръ́га	no voc?
дух<db(v)>	"spirit, ghost"	ду́хове (also духове́???)	духъ́т; духа́	ду́ха	ду́хо???
дял<d>	"part, share"	дя́лове	делъ́т; дела́	дя́ла	no voc???	type (я) automatically inferred
труд<d>	"effort, work"	тру́дове	трудъ́т; труда́	тру́да	no voc???
род<db(v)>	"tribe; family; kind; gender"	ро́дове (also родове́)	родъ́т; рода́	ро́да	ро́де???
сън<d/n:sg>	"sleep"	no pl	съня́т; съня́	съ́ня	no voc???
сън<d+ища>	"dream"	съ́нища	съня́т; съня́	съ́ня	no voc???
плъх</voc:плъ́хо>	"rat"	плъ́хове	плъ́хът; плъ́ха	плъ́ха	плъ́хо
чорбаджи́я<(m)/voc:~~:~~йо>	"(hist. Ottoman Turkish) master, rich man"	чорбаджи́и	чорбаджи́ята	count as pl?	чорбаджи́, чорбаджи́йо	masculine, чорбаджи́ - non-productive vocative, the same pattern applies to other nouns suffixed with -джия or -чия; note: ~~ indicates the stem чорбаджи́
пете́л<*/voc:пе́тльо>	"rooster"	петли́	пете́лът; пете́ла	пете́ла	пе́тльо
януа́ри(m)??	"January"	януа́ри	януа́ри	януа́ри	no voc	masculine, indeclinable as all masculines ending in -и; no suffixes in the definite form for month names (only?)
Ду́нав<(v)>	"Danube"	-	Ду́навът; Ду́нава	no count	Ду́наве	singulare tantum
друг</pl:дру́ги:дру́зи/voc:дру́же>	"friend (rare or dated)"	дру́ги, дру́зи	дру́гът; дру́га	дру́га	дру́же
кът</pl:ове+ища>	"corner"	къ́тове, къ́тища	къ́тът; къ́та	къ́та	no voc	къ́тища - might be also a plurale tantum
кита́ец<*(v)>	"Chinese person (male)"	кита́йци	кита́ецът; кита́еца	кита́еца	кита́ецо

Feminine

Noun	Meaning	Pl	Def Sg	Count	Vocative	Notes
кни́га<>	"book"	кни́ги	кни́гата	count as pl	no voc??
зи́ма<>	"winter"	зи́ми	зи́мата	count as pl	no voc?? or зи́мо?
Кристи́на<(v)>	"Christina"	pl???	no def sg	count as pl	Кристи́но
ба́ба<(v)/dat:ба́би>	"grandmother, old woman"	ба́би	ба́бата	count as pl	ба́бо	dative ба́би
вя́ра<(v)/pl:ве́ри[r]>	"faith"	(rare) ве́ри	вя́рата	count as pl	вя́ро
ма́йка</voc:+:ма́йке[d]/dat:майки[d]:ма́йци[d]:ма́йце[d]>	"mother"	ма́йки	ма́йката	count as pl	ма́йко (dialectal ма́йке)	dative (dialectal) ма́йки/ма́йци/ма́йце
ка́ка<(v)/dat:ка́ки[r]>	"older sister"	ка́ки	ка́ката	count as pl	ка́ко, dative (rare) ка́ки
друга́рка</voc:~~о:+[lc]>	"female friend"	друга́рки	друга́рката	count as pl	друга́рко (low-colloquial друга́рке)	per Bulgarian nouns, -е is the normal vocative for nouns in -рка
ве́щица<(v)>	"witch"	ве́щици?	ве́щицата	count as pl	ве́щице
тъпа́чка<(v)>	"stupid woman (colloq.)"	тъпа́чки	тъпа́чката	count as pl	тъпа́чке
дире́кторка<(v)>	"female director"	дире́кторки	дире́кторката	count as pl	дире́кторке
Сте́фка<(v)>	"Stefka"	pl??	no def sg (def pl??)	count as pl	Сте́фке
река́<>	"river"	реки́	река́та	count as pl	no voc?? or ре́ко??
жена́<(v)>	"woman, wife"	жени́	жена́та	count as pl	же́но
сестра́<(v)>	"sister"	сестри́	сестра́та	count as pl	се́стро
ръка́<+е>	"hand, arm"	ръце́	ръка́та	count as pl	no voc??
овца́<(v)+e+и>	"sheep, ewe"	овце́/овци́	овца́та	count as pl	о́вце??
нога́</pl:нозе́:ноги́:нози́>	"foot, leg" (archaic/dialectal)	нозе́/ноги́/нози́	нога́та	count as pl	no voc??
госпожа́<a/voc:+:~:госпо́же[a]:госпо́жа[a][d]>	"Mrs., ma'am"	госпо́жи	госпосжа́та	count as pl	госпо́жо/госпожа́ (archaic госпо́же, archaic/dialectal госпо́жа)
сълза́<ab>	"tear (of crying)"	съ́лзи/сълзи́	сълза́та	count as pl	no voc??
ба́ня<>	"bath, bathroom"	ба́ни	ба́нята	count as pl	no voc??
ле́ля<(v)/pl:+:ле́лини[d]/dat:ле́ли>	"aunt"	ле́ли (dialectal ле́лини)	ле́лята	count as pl	ле́льо	dat ле́ли
ста́я<>	"room"	ста́и	ста́ята	count as pl	no voc??
Бълга́рия<(v)/n:sg>	"Bulgaria"	no pl	no def?	no count	voc Бълга́рийо
земя́<(v)>	"earth, land"	земи́	земя́та	count as pl	зе́мьо
змия́<(v)>	"snake"	змии́	змия́та	count as pl	зми́йо
свиня́<(v)+и+е>	"sow"	свини́/свине́	свиня́та	count as pl	сви́ньо??
вест<(f)>	"message (literary), news (plural)"	ве́сти	вестта́	count as pl	no voc??	(fem)
цел<(f)>	"purpose, goal; target"	це́ли	целта́	no count	no voc	type (d)+и automatically inferred for feminine nouns ending in a consonant
нощ<(f)>	"night"	но́щи	нощта́	no count	no voc	type (d)+и automatically inferred for feminine nouns ending in a consonant
сто́йност<(f)>	"value, cost"	сто́йности	стойността́	count as pl	no voc??	(fem)
ця́лост<(f)>	"whole, unity"	ця́лости	целостта́	no count	no voc	type (d)+и automatically inferred for feminine nouns ending in a consonant
любо́в<(f)>	"love"	любо́ви	любовта́	no count	no voc	type (d)+и automatically inferred for feminine nouns ending in a consonant
кръв<(f)(ър)+и:+ища[c]:+ове[d]>	"blood"	къ́рви (colloquial къ́рвища, dialectal къ́рвове)	кръвта́	count as pl	no voc??	(fem)
гръд<(f)/n:sg>	"women's breasts (collective)"	no pl (sg tantum)	гръдта́	count as pl	no voc??	(fem)
пе́сен<(f)*>	"song"	пе́сни	песента́	count as pl	no voc??	(fem)
ми́съл<(f)*>	"thought"	ми́сли	мисълта́	count as pl	no voc??	(fem)
за́хар<(f)>	"sugar"	за́хари	захарта́	count as pl	no voc??	(fem)
дъщеря́</voc:дъ́ще:дъ́щи[dated]>	"daughter"	дъщери́	дъщеря́та	count as pl	дъ́ще (dated/archaic дъ́щи)
мома́</voc:мо́ме:мо́мне[d]>	"young girl"	моми́	мома́та	count as pl	мо́ме (dialectal мо́мне)	expected vocative is мо́мо
Мари́йка<(v)/n:sg>	"Mariyka (f. name)"	no pl	no def?	no count	voc Мари́йке	voc "-ке" when "-ка" is diminutive
ни́ва<(v)/pl:+:нивя́[p]:ни́ве[d]	"field"	ни́ви (poetic, collective нивя́, dialectal ни́ве)	ни́вата	count as pl	ни́во	нивя́ - also considered a plurale tantum
((долина́<>, доли́на<>))	"valley, dale"	долини́/доли́ни	долина́та; доли́ната	count as pl	no voc???

Neuter

Noun	Meaning	Pl	Def Sg	Count	Vocative	Notes
дете́</pl:деца́>	"kid, child"	деца́	дете́то	count as pl	no voc
гове́до<>	"cattle"	гове́да	гове́дото	count as pl	no voc
се́ло<c>	"village"	села́	се́лото	count as pl	no voc
у́тро<c>	"dawn"	утра́	у́трото	count as pl	no voc
ста́до<c>	"flock"	стада́	ста́дото	count as pl	no voc
пиа́но<c>	"piano"	пиана́	пиа́ното	count as pl	no voc
огледа́ло<c>	"mirror"	огледала́	огледа́лото	count as pl	no voc
тя́ло<c+а+еса[a]>	"body"	тела́ (archaic телеса́)	тя́лото	count as pl	no voc
дърво́</pl:дърве́та:дърва́:дървя́[d]:дъ́рве[d]:дървеса́[p]>	"tree"	дърве́та/дърва́ (dialectal дървя́/дъ́рве, poetic дървеса́)	дърво́то	count as pl	no voc
писмо́<>	"letter"	писма́	писмо́то	count as pl	no voc
листо́</pl:листа́:ли́сте[d]>	"leaf"	листа́ (dialectal ли́сте)	листо́то	count as pl	no voc
око́</pl:очи́>	"eye"	очи́	око́то	count as pl	no voc
ухо́</pl:уши́>	"ear"	уши́	ухо́то	count as pl	no voc
колело́<+а+ета>	"wheel"	колела́/колеле́та	колело́то	count as pl	no voc
петънце́<>	"spot"	петънца́	петънце́то	count as pl	no voc
лице́<>	"face, person"	лица́	лице́то	count as pl	no voc
учи́лище<>	"school"	учи́лища	учи́лището	count as pl	no voc
носле́<>	"nose"	носле́та	носле́то	count as pl	no voc
а́гне<>	"lamb"	а́гнета	а́гнето	count as pl	no voc
море́<+ета+я[a][p]>	"sea"	море́та (archaic/poetic моря́)	море́то	count as pl	no voc
поле́<+ета+а[a][p]>	"field"	поле́та (archaic/poetic поля́)	поле́то	count as pl	no voc
въже́<+ета+а[a]>	"rope"	въже́та (archaic въжа́)	въже́то	count as pl	no voc
момче́</pl:+:момчети́и[colloquial pejorative]>	"boy"	момче́та (colloquial pejorative момчети́и)	момче́то	count as pl	no voc
бра́тче<>	"little brother"	бра́тчета	бра́тчето	count as pl	no voc
цве́те<c+я>	"flower"	цветя́	цве́тето	count as pl	no voc
ло́зе<c+я>	"vineyard"	лозя́	ло́зето	count as pl	no voc
изключе́ние<>	"exception"	изключе́ния	изключе́нието	count as pl	no voc
допълне́ние<>	"supplement, addition"	допълне́ния	допълне́нието	count as pl	no voc
и́ме<c+ена>	"name"	имена́	и́мето	count as pl	no voc
вре́ме<c+ена>	"time"	времена́	вре́мето	count as pl	no voc
здра́ве</n:sg>	"health"	no pl	здра́вето	count as pl	no voc	(sg tantum)
чу́до<+еса>	"miracle, wonder"	чу́деса	чу́дото	count as pl	no voc
небе́<+еса[p]+ета[d]/gen:небесе́[a]/dat:небеси́[a]>	"sky"	(usually poetic) небеса́ (dialectal небе́та)	небе́то	count as pl	no voc	gen (archaic) небесе́, dat (archaic) небеси́
живо́тно<+и>	"animal"	живо́тни	живо́тното	count as pl	no voc
насеко́мо<+и>	"insect"	насеко́ми	насеко́мото	count as pl	no voc
бижу́<>	"jewel"	бижу́та	бижу́то	count as pl	no voc	(neut)
меню́<>	"menu"	меню́та	меню́то	count as pl	no voc	(neut)
такси́<>	"taxi"	такси́та	такси́то	count as pl	no voc	(neut)

Plurale tantum

Noun	Meaning	Pl	Def Pl	Count	Vocative	Notes
три́ци</n:pl>	"bran"	already pl	три́ците	count as pl	no voc	(pl tantum)
очила́</n:pl>	"glasses"	already pl	очила́та	count as pl	no voc	(pl tantum)
дърва́</n:pl>	"wood, firewood"	already pl	дърва́та	count as pl	no voc	(pl tantum); regular plural of дърво́ (tree) is дърве́та or дърве́са
обу́ща</n:pl>	"shoes"	already pl	обу́щата	count as pl	no voc	(pl tantum)
га́щи</n:pl>	"pants, trousers"	already pl	га́щите	count as pl	no voc	(pl tantum)
хо́ра</n:pl>	"people; plural of чове́к"	already pl	хо́рата	count as pl	no voc	(pl tantum), plural of чове́к
фина́нси</n:pl>	"finance; finances"	already pl	фина́нсите	count as pl	no voc	(pl tantum)
А́лпи</n:pl>	"Alps"	already pl	А́лпите	count as pl	no voc	(pl tantum)
везни́</n:pl>	"scales (for weighing)"	already pl	везни́те	count as pl	no voc	(pl tantum); sg везна́ is rare
лю́де</n:pl>	"people; plural of чове́к"	already pl	лю́дете	count as pl	no voc	(pl tantum), plural of чове́к
турча́</n:pl>	"Turks (collective); plural of ту́рчин"	already pl	турча́та	count as pl	no voc	(pl tantum), plural of ту́рчин
нивя́</n:pl>	"field(s) (collective, poetic); plural of ни́ва"	already pl	нивя́та	count as pl	no voc	(pl tantum), plural of ни́ва
къ́тища</n:pl>	"corner(s) (collective); plural of кът"	already pl	къ́тищата	count as pl	no voc	(pl tantum), plural of кът, also къ́тове
((масло́<>, ма́сло<>))	"butter, oil"	масла́/ма́сла???	масла́та; ма́слата???	count as pl	no voc	"ма́сло" according [1]. Is plural also "ма́сла"?

Explanation

The general format is as follows: LEMMA<ACCENT(INDIC)(INDIC)+PLURAL1+PLURAL2/PROPERTY1:VALUE1:VALUE2/PROPERTY2:VALUE3:VALUE4>. More specifically:

1. There can be zero or more accent codes at the beginning. The allowed accent codes are

   • a = stem-stressed in both def sg and plural
   • b = ending-stressed in both def sg and plural
   • c = stem-stressed in def sg, ending-stressed in plural
   • d = ending-stressed in def sg, stem-stressed in plural

   Note that the interpretation of these is a bit different from the Russian equivalents.

   The defaults are as follows:

   1. Masculine nouns with reducible plural where the stress is on the reducible vowel (e.g. чуждене́ц, def sg. чуждене́цът, pl. чужденци́) default to c; others default to a.
   2. Feminine nouns ending in stressed -а́ or -я́ (e.g. жена́, def sg. жена́та pl. жени́) default to c; feminine nouns ending in a consonant (e.g. нощ def sg. нощта́, pl. но́щи) default to d; others default to a.
   3. Neuter nouns ending in stressed -о́ or -е́ default to c; others default to a.
2. Following the accent codes are zero or more indicators, which are mostly inside of parens:

   • * (indicates reducible plural)
   • (d*) (indicates reducible def sg)
   • (v*) (indicates reducible vocative)
   • (c*) (indicates reducible count form)
   • (v) (indicates that the noun has a vocative)
   • (ь) (indicates that the def sg is in -ят/-я and the count is in -я, instead of def sg -ът/-а and count in -а; automatically inferred for nouns ending in -й, vowel + -тел or vowel + -ар)
   • (-ь) (prevents (ь) from being inferred when it normally would be)
   • (ър) (indicates ръ -> ър change when a vowel-initial ending is added)
   • (ин) (indicates that -ин is lost in the plural)
   • (m) (indicates unpredictably masculine, e.g. when ending in a vowel)
   • (f) (indicates unpredictably feminine, e.g. when ending in consonant)
   • (n) (indicates unpredictably neuter)

3. Following this are zero or more plural specs, e.g. +и, +ове, +ища. +и normally causes palatalization of word-final к, г, х (except for word-final нг) -> ци, зи, си; +е similarly causes palatalization of word-final ка, га, ха -> це, зе, се. To disable this, use ++и or ++е.

   The defaults are as follows:

   1. All feminine nouns default to +и.
   2. All multisyllabic masculine nouns default to +и.
   3. Monosyllabic masculine nouns normally default to +ове, or +еве if ending in й, or +ьове if the (ь) indicator is present (either explicitly, or inferred for nouns in -тел or -ар).
   4. Neuter nouns in -о, -ще or -це default to +а; neuter nouns in -ие default to +я; other neuter nouns in -е default to +ета; remaining neuter nouns (loanwords, e.g. бижу́, меню́, такси́) default to +ета (note that for these nouns, the stem is the entire word, whereas for neuter nouns in -о and -е, the stem does not include these endings).
4. Following this is zero or more property specs. Currently recognized properties are:

   • pl (specify the indefinite plural form)
   • def (specify the definite singular subjective form)
   • count (specify the count form)
   • voc (specify the vocative form)
   • n (restrict the noun to singular-only or plural-only).

   All the properties specifying particular forms should be followed by one or more colon-separated forms, where the special value + means "use the default form" and ~ and ~~ are substituted by the whole word and the stem, respectively.

   Each form can optionally be followed by a footnote in brackets, e.g. [collective] or [dated]. Certain abbreviations are recognized for the footnotes, e.g.

   • [a] = [archaic]
   • [c] = [colloquial]
   • [lc] = [low colloquial]
   • [p] = [poetic]
   • [fp] = [folk-poetic]
   • [d] = [dialectal]
   • [r] = [rare]

   Note that specifying a plural form using a property spec overrides any plurals specified using +и or similar, so only one of the two mechanisms should be used.

5. Finally, more than one of the entire specification (including the initial specs, plurals and properties) can be given, comma-separated.

Comments

@Atitarev, Bogorm I created the above table to help in creating the new {{bg-ndecl}} template. The stuff in the first column is what would be specified in the argument to {{bg-ndecl}} (except for the few cases without <> after the noun, which need to be fixed up); I will create a guide that explains it shortly. For now I need help with the remaining columns, particularly those cases which I marked with one or more ?'s. Note that almost all the nouns in the table are masculine; I will be adding more masculine nouns, as well as feminine and neuter ones, shortly. Benwing2 (talk) 03:19, 20 March 2020 (UTC)

@Benwing2, Bogorm: I have mostly the same questions. Unfortunately, dictionaries don't give extensive info on vocatives. I noticed that not all our templates provide this either. Apparently, not every word may have a vocative or it could be theoretical or in case of personification.

градче́та (correct stress) is a plural of градче́, not граде́ц. Please see [2] and [3]

I have also downloaded "Bolgarsko-russky_slovar_Bernshtein.pdf". It also provides some grammar, when it's irregular and has inflection guide at the end. --Anatoli T. ^{(обсудить}/^вклад) 04:54, 20 March 2020 (UTC)

@Benwing2: град does have a count form - e.g. "два гра́да" - two cities (only in the plural column).

Please note how the stress is shifted in multisyllabic vocatives - госпожа́ - госпо́жо. This is in the Russian-Bulgarian grammar.

I'll give you a tip. I sometimes succeed by searching for stressed forms in quotes using a grave accent. --Anatoli T. ^{(обсудить}/^вклад) 05:03, 20 March 2020 (UTC)

@Benwing2 I think it's the best to start a Bulgarian noun declension module and add straighforward (easy to understand) classes/subclasses (basic form/stress pattern) and build on it. For example, feminines are simpler, both ending in vowels and consonants. {{bg-noun-f-a1}} is the simplest pattern but need to take care of the vowel shift in the vocatives - госпожа́ (gospožá) - госпо́жо (gospóžo). Even the ones with mutating vowels seem not so hard. E.g. вя́ра (vjára) uses template {{bg-noun-f-yat}}. It only uses two parameters - sg and pl stems (вяр and вер). --Anatoli T. ^{(обсудить}/^вклад) 06:49, 20 March 2020 (UTC)

{{bg-noun-f-ya1}} and {{bg-noun-f-ya2}} belong to the same simple class. They just have different vocative endings (1. змия - змийо, 2. зeмя - земьо) . Count forms coinside with the plural forms. --Anatoli T. ^{(обсудить}/^вклад) 06:55, 20 March 2020 (UTC)

@Benwing2: Irregular noun (no plural): го́спод (góspod), it has a historical dative го́споду (góspodu), just like бог (bog) has бо́гу (bógu) in set expressions. It has an irregular vocative го́споди (góspodi) (coincides with the vocative of Russian госпо́дь (gospódʹ)). Not sure how to fit into your table. --Anatoli T. ^{(обсудить}/^вклад) 07:50, 20 March 2020 (UTC)

@Atitarev I updated the table (now 4 tables) with several feminine and neuter nouns and more masculine nouns. I think I have enough nouns now. I have added the datives and genitives in the last column. Benwing2 (talk) 01:38, 21 March 2020 (UTC)

@Benwing2: Thanks for the efforts. I was away bushwalking the whole weekend. I am back. Trying to understand the codes. What is "прия́тел<(v)>" for example? --Anatoli T. ^{(обсудить}/^вклад) 01:01, 23 March 2020 (UTC)

@Atitarev I cleaned up the explanation and added more info. <(v)> means this noun has a vocative; nouns that don't specify this won't have a vocative unless it's explicitly supplied using /voc:.... Also, I think your changes for стол "canteen" and град "grad" may not be correct; according to http://ibl.bas.bg/rbe/lang/bg/, these nouns are declined differently from стол "chair" and град "city; hail". Benwing2 (talk) 02:24, 23 March 2020 (UTC)

Also, I have written most of the code; in another day or two it should be ready to test. Benwing2 (talk) 02:28, 23 March 2020 (UTC)

@Atitarev I made a few changes. The vocative code (v) doesn't need to be given if you explicitly list the vocatives. I'm following Bulgarian nouns for what counts as the "regular" vocative of a given noun, hence мома́ would have regular vocative мо́мо (where the stress shift is regular, as you've noted), and the actual vocative мо́ме (and dialectal мо́мне) need to be listed explicitly. Also I've currently made the footnote code [d] mean "dialectal", and I've been translating "остар." as "archaic" with code [a], but I have no problem translating it as "dated" and using [d] to mean "dated", in which case maybe [di] = "dialectal". Benwing2 (talk) 14:51, 24 March 2020 (UTC)

Vocatives of nouns in -ка

@Bogorm, Bezimenen, Atitarev Bulgarian nouns says that feminines in -рка and -чка as well as all names in -ка have vocatives in -е but implies that other feminines in -ка have vocatives in -о. Is this true? I see that хубаве́лка "pretty woman" probably has a vocative in -е. Should the rule be modified? Benwing2 (talk) 00:02, 30 March 2020 (UTC)

Similarly for кали́нка. Benwing2 (talk) 00:15, 30 March 2020 (UTC)

Also ше́фка, коле́жка. Benwing2 (talk) 01:04, 30 March 2020 (UTC)

@Benwing2: I don't know the rule.

1. -е - хубаве́лке, коле́жке, ше́фке, Мари́йке.
2. -о - бъ́лгарко/бъ́лгарке(? - I found in songs), америка́нко, учи́телко, друга́рко

We need to check words (mixed gender) with -ца as well - це or цо - пия́ница, уби́йца (is it feminine in modern Bulgarian). Native speakers are silent and there is not enough material on this. --Anatoli T. ^{(обсудить}/^вклад) 01:19, 30 March 2020 (UTC)

@Benwing2: The historically correct form is -кo. In modern times, though, this form is considered serious and even rude. For example, if you refer to a girl called Мария (Marija) as Марийо, this either implies that you're going to scold her, or you are mocking her. The forms with -e are considered more caressing and are often found with diminutive nicknames as is the case with Bulgarian Марийка (Marijka). (Remark: The respective caressing vocative of Мария (Marija) is Мари, so not all feminines use -e as a sign of cuteness.) Notice that -к- is not palatalized by -е, so -ке is clearly a recent innovation. Безименен

@Atitarev: The vocative of Bulgarian учителка (učitelka), другарка (drugarka), лекарка (lekarka), etc is always with -ко (unless you want to show disrespect to the holder of the title). българка (bǎlgarka) can be both - e.g. in this song Велико дульбер бугарко it's with -ко (technically the song is in Aegean Macedonian, but I couldn't find a song in eastern dialects), while in Раде, бяла българке it's with -ке.

Obscure terms for inhabitants

@Bezimenen, Bogorm, Atitarev Need native speaker help here. I converted all nouns using {{bg-noun-f-ya1}} to {{bg-ndecl}}, but there are several that were missing accents and are not in any dictionaries: бургазлия (burgazlija), хасковлия (haskovlija), одринлия (odrinlija), русенлия (rusenlija), търновлия (tǎrnovlija), странджалия (strandžalija). I assigned the stress position based on the stress of the proper noun the term is derived from, which may be wrong. Also note ва́рналия (várnalija), кърджа́лия (kǎrdžálija), добри́члия (dobríčlija), which contained accents previously but may be wrong. Benwing2 (talk) 04:01, 31 March 2020 (UTC)

@Benwing2, Ted Masters: I've got answers to some questions in http://ibl.bas.bg/rbe/lang/bg:

1. http://ibl.bas.bg/rbe/lang/bg/бургазлия/ - БУРГАЗЛЍЯ
2. http://ibl.bas.bg/rbe/lang/bg/одринлия/ - О̀ДРИНЛИЯ
3. http://ibl.bas.bg/rbe/lang/bg/варналия/ - ВА̀РНАЛИЯ
4. http://ibl.bas.bg/rbe/lang/bg/кърджалия/ - КЪ̀РДЖАЛИЯ: EDIT: @Benwing2: only the 2nd sense (outlaw) --Anatoli T. ^{(обсудить}/^вклад) 05:38, 31 March 2020 (UTC)

I am inviting User:Ted Masters from the Bulgarian Wiktionary to our discussions. --Anatoli T. ^{(обсудить}/^вклад) 04:54, 31 March 2020 (UTC)

@Benwing2: I'm not sure about the stress, but the above terms use Bulgarian -лия (-lija). That's based on the adjective suffix -lV from Turkish (apparently interpreted as a *-i stem noun, e.g. as in the native съдия (sǎdija), мълния (mǎlnija)). As a borrowing, the place of the stress is going to be hard to generalize.

PS ру́сенлия (rúsenlija) has stress on -у-. Безименен 14:16, 31 March 2020 (UTC)

A couple of auto-detection issues -ин and -ар - exception or another rule?

@Benwing2: Hi, a couple of issues, you already know about делфин/пингвин/магазин:

1. делфи́н - vocative form - делфи́но, should be делфи́не, also пингвин
2. ку́фар - all inflected forms are incorrect - ку́фарят, ку́фарю, should be be a hard declension. --Anatoli T. ^{(обсудить}/^вклад) 00:08, 1 April 2020 (UTC)

I just made a quick post describing the issue but not sure if a solution is easy. I think, generally auto-detecting the inflection for native -ин and -ар nouns is a good idea but it would be good to have an easy way to override it and tell the module what type it should be? --Anatoli T. ^{(обсудить}/^вклад) 00:16, 1 April 2020 (UTC)

@Atitarev For ку́фар (kúfar) there's already an indicator (-ь) for exactly this circumstance; it says to not assume the (ь) indicator, which would cause the forms ку́фарят, ку́фарю. For делфи́но, пингви́но, I would recommend for now using /voc:~о to explicitly give the vocative; perhaps in the future I'll come up with something better. Please note, I've greatly expanded Template:bg-ndecl/documentation, which should make your life easier. Benwing2 (talk) 04:03, 1 April 2020 (UTC)

@Benwing2: Thanks, mate! Happy with your solutions. Is User:Benwing2/test-bg-ndecl getting out of date? It's been good for me too look up cases. As with Russian or Arabic, I often use terms with a similar paradigm first before digging into the documentation. Categorisation would also help but first things first.

(Did you check Bezimenen's responses to your queries and his edits? Do they answer most of outstanding items?) --Anatoli T. ^{(обсудить}/^вклад) 04:14, 1 April 2020 (UTC)

@Benwing2: лиси́ца (lisíca), пти́ца (ptíca), probably many more require manual vocative override as well: {{bg-ndecl|лиси́ца</voc:лиси́цо>}}. --Anatoli T. ^{(обсудить}/^вклад) 05:58, 1 April 2020 (UTC)

@Atitarev, Bezimenen, Bogorm I have updated User:Benwing2/test-bg-ndecl so it contains the latest syntax; not much has actually changed. As for лиси́ца (lisíca), пти́ца (ptíca), etc., is there a rule concerning nouns in -ица that says the vocative should be in -о? Bulgarian nouns says "all nouns ending in -ца" take -е; maybe it's wrong? (Note, you can use the syntax /voc:~~о to save some keystrokes when adding a manual vocative; ~~ stands for the stem.) Also I added support for |adj= to {{bg-noun}} to display the "related adjective" of a noun, e.g. водоро́ден (vodoróden) for водоро́д (vodoród, “hydrogen”). Benwing2 (talk) 05:20, 2 April 2020 (UTC)

@Benwing2: Thanks. These https://rechnik.chitanka.info/w/птицо, https://rechnik.chitanka.info/w/лисицо suggest that these vocatives are correct. We can test more -ца words to make sure. --Anatoli T. ^{(обсудить}/^вклад) 05:45, 2 April 2020 (UTC)

@Atitarev I went through all the nouns I could find in -ица. ку́кувица "cuckoo" is listed on chitanka with vocative кукувице, similarly ве́щица "witch", госпо́жица "miss", у́личница "whore", развра́тница "slut". Meanwhile, проти́вница "female opponent", съпе́рница "female opponent", со́бственица "female owner" and maybe кърти́ца "mole", па́тица "female duck", гъсе́ница should have a vocatives but none are listed. The only pattern I see is that пти́ца and лиси́ца have stressed -и́ца while the others have unstressed -ица. Benwing2 (talk) 06:03, 2 April 2020 (UTC)

@Atitarev Also, p. 148 of Maslov "грамматика болгарского языка" says that "many feminine nouns" in -ка, when it is a diminutive suffix, form vocatives in -е, but mentions as exceptions учи́телка, друга́рка with vocatives in -о, "поскольку здесь другой суффикс -/к/-", presumably a suffix -ка that forms feminine equivalents of masculine nouns rather than diminutives. Benwing2 (talk) 06:09, 2 April 2020 (UTC)

This same page mentions that nouns in -ица typically form vocatives in -е. Benwing2 (talk) 06:10, 2 April 2020 (UTC)

@Benwing2: Thanks, I find that page a bit confusing, even if it's in Russian. We still have inconsistencies, don't we? These links show vocatives:

"гъсеницо" https://writecraft.io/posts/5bcc61857002abe9622cd3a0

"гъсенице" https://otkrovenia.com/ru/prose/edinstvenoto-priklyuchenie-na-malkata-gysenica-haj

"къртицо" https://hotarena.net/kyrticooo-borisa-se-izdade-tq-shte-e-noviqt-sabotyor-vij-kak-podskaza

"къртице" https://chitanka.info/text/2065/15

Do we have to specifically search for each vocative, which may also give a wrong result? LOL--Anatoli T. ^{(обсудить}/^вклад) 07:06, 2 April 2020 (UTC)

@Atitarev Hmmm. Sounds like this is an unsettled area in the language. Maybe native speakers can give their intuition: @Bezimenen, Bogorm Benwing2 (talk) 07:12, 2 April 2020 (UTC)

@Atitarev, Benwing2 Both vocatives are possible - the one with -цо is formal and sounds strict/serious, while the other with -це is more informal and caressing. In general, you use -е when you refer to a youngling in a gentle manner. You use -o when you have to be formal (e.g. you should not say учителке /even though technically acceptable/, but учителко) or when you exert authority - in commands, orders, scolding. Безименен 10:29, 2 April 2020 (UTC)

@Atitarev, Benwing2: If the definite article of the said word is -ът (subj.) / -а (obj.), the vocative ends in -е; while those with definite article -ят (subj.) / -я (obj.) end in -ю. The definite article of куфар (kufar) is куфар(ът)/(а), so the vocative form here should be куфаре, not куфарю; and the vocative of цар (car) is expected to be царю because its definite article is цар(ят)/(я). Note: that's my personal observation on the -ар/-ер -ending words. I'm not sure if such a rule exists in Bulgarian language. --Ted Masters (talk) 16:49, 2 April 2020 (UTC)

@Ted Masters, Benwing2: Thank you, @Ted Masters. we have now sorted this problem. цар (car) is a soft declension "(ь)" and ку́фар (kúfar) is a hard declension "(-ь)". --Anatoli T. ^{(обсудить}/^вклад) 22:57, 2 April 2020 (UTC)

@Ted Masters, Bezimenen, Bogorm: We still have questions regarding vocative cases with -о / -е. Your input as native speakers would be highly appreciated. --Anatoli T. ^{(обсудить}/^вклад) 01:36, 3 April 2020 (UTC)

I do not have much to add to what Безименен already outlined. Regarding птица, I remember the vocative form from the famous poëm Обесването на Васил Левски of Ivan Vazov: Гарване, и ти, птицо проклета. Bogorm converſation 11:56, 5 April 2020 (UTC)

Comparatives of certain adjectives, stress position of certain adjectives

I have some questions about whether various adjectives have comparatives:

1. друг (drug, “other”)
2. diminutives of adjectives: ба́вничък (bávničǎk), бе́дничък (bédničǎk), бле́дничък (blédničǎk), бли́зичък (blízičǎk), бо́сичък (bósičǎk), бъ́рзичък (bǎ́rzičǎk)
3. абсолю́тен (absoljúten, “absolute”)
4. аванга́рден (avangárden, “avant-garde (relational); first, front, leading”) (esp. in its figurative sense)
5. възбя́л (vǎzbjál, “off-white”), възголя́м (vǎzgoljám, “largish”), and similar adjectives in въз-
6. глухоня́м (gluhonjám, “deaf-mute”)
7. млечнобя́л (mlečnobjál, “milk-white”)
8. невря́л (nevrjál, “unboiled”), неголя́м (negoljám, “not big, not large”), недовря́л (nedovrjál, “undercooked, underboiled”), and other negative adjectives
9. второразре́ден (vtororazréden, “second-rate”), първоразре́ден (pǎrvorazréden, “first-rate”), треторазре́ден (tretorazréden, “third-rate”)
10. се́рен (séren, “sulphurous”)
11. безавари́ен (bezavaríen, “flawless, trouble-free”), безбро́ен (bezbróen, “countless, numberless”), безде́ен (bezdéen, “inactive, lethargic”), безиде́ен (bezidéen, “lacking in ideas, apolitical”), безкра́ен (bezkráen, “endless, ceaseless”), and other adjectives in без-
12. комерсиа́лен (komersiálen, “commercial”)
13. прекра́сен (prekrásen, “beautiful, magnificent”) and other adjectives in пре-
14. отли́чен (otlíčen, “excellent”), отличе́н (otličén, “distinguished, honored”)
15. естети́чен (estetíčen, “aesthetic”)
16. кра́лски (králski, “royal, regal”)
17. анони́мен (anonímen, “anonymous”)
18. забрани́телен (zabranítelen, “inhibitory, proscribing, which forbids”)
19. бре́менен (brémenen, “pregnant”)
20. медици́нски (medicínski, “medical”)
21. вто́ри (vtóri, “second”), тре́ти (tréti, “third”), and other ordinals

Also some other questions:

1. Where is the stress in възбя́л (vǎzbjál, “off-white”), възголя́м (vǎzgoljám), etc.? Can въз- take a primary or secondary stress?
2. Where is the stress in неголя́м (negoljám)? talkoven says не́голям.
3. What is the feminine of безве́рен (bezvéren, “unbelieving, infidel; treacherous, deceitful”)? безве́рна, безвя́рна, or both? chitanka.info appears to list both.
4. Where is the primary stress in свръ́хестествен (svrǎ́hestestven)? RBE says on есте́ствен, PONS and talkoven say on свръ́х-.

@Atitarev, Bezimenen, Bogorm, Nauka Thanks for your help. Benwing2 (talk) 06:59, 6 April 2020 (UTC)

@Benwing2: - my reply based on some checking and my strong assumption.

1. No comparatives for all on the list - safer that way, if some of them do, they can added later
2. възбя́л (vǎzbjál), възголя́м (vǎzgoljám) No secondary stress
3. Both безве́рна and безвя́рна are valid and are included in chitanka dictionary
4. It should be свръ̀хесте́ствен (using the Russian notation for stresses). Secondary stresses are all flawed in all Slavic dictionaries. Also calling @Ted Masters. BTW, we can update the Bulgarian group ping for the duration of this project.--Anatoli T. ^{(обсудить}/^вклад) 07:12, 6 April 2020 (UTC)

@Benwing2:: PONS says "свръ̀хестѐствен", which is a usual way in Slavic dictionaries to mark both primary and secondary stresses, as if there are two accents, only Russian dictionaries use acutes and Bulgarian dictionaries use graves for both primary and secondary accents. Making "свръ̀х-" the primary would be unnatural, IMO but this is my strong hunch only. --Anatoli T. ^{(обсудить}/^вклад) 07:20, 6 April 2020 (UTC)

@Atitarev, Benwing2: Adjectives that do have comparative forms (по- and най-):

1. друг (drug, “other”): по-друг; най-друг (in some rare occasions); Note: there's also по друг (no hyphen between the words), which is not a comparative form; it is another way of saying различен (različen, “different”)
2. ба́вничък (bávničǎk), бе́дничък (bédničǎk), бле́дничък (blédničǎk), бли́зичък (blízičǎk), бо́сичък (bósičǎk), бъ́рзичък (bǎ́rzičǎk)
3. аванга́рден (avangárden, “avant-garde (relational); first, front, leading”)
4. комерсиа́лен (komersiálen, “commercial”)
5. прекра́сен (prekrásen, “beautiful, magnificent”); Not completely sure, but I think that the words with пре- always tend to have comparative forms (have to check for possible exceptions, but for now can't think of any).
6. отли́чен (otlíčen, “excellent”), отличе́н (otličén, “distinguished, honored”)
7. естети́чен (estetíčen, “aesthetic”)

Special cases (the ones with без-):

1. безде́ен (bezdéen, “inactive, lethargic”), безиде́ен (bezidéen, “lacking in ideas, apolitical”) do have comparative forms
2. Those that end in -less when translated to English (like безбро́ен (bezbróen, “countless, numberless”)), lack such forms

The rest do not have comparative forms. And for the ones with въз — they are already in comaparative form. Въз would be like saying that something is more or less than what is supposed to be. Е.g. възбял is another way of saying that something is по-малко бял(o) (“less white”, “less of a white”); възголям is like saying that something is по-голям(o) (“larger”). Sometimes it means more, sometimes it means less — depends on the word that is being used.

And now about the stresses (the ones that Atitarev is unsure of):

1. нéголям — https://ibl.bas.bg/rbe/lang/bg/неголям/
2. свръхе́стествен — https://ibl.bas.bg/rbe/lang/bg/свръхестествен/

Also an advise: it's probably a better idea to check RBE first (way more reliable; also more frequently updated) and then PONS and talkoven (if RBE is missing the information needed). --Ted Masters (talk) 12:23, 6 April 2020 (UTC)

@Ted Masters, Benwing2:: Thanks, Ted Masters! https://ibl.bas.bg/rbe/lang/bg/свръхестествен/ shows "СВРЪХЕСТЀСТВЕН" (not свръхе́стествен), which is the main stress. Is there a secondary stress? Just to clarify, my point was, if there is a secondary accent, it would be on the prefix, not on the main word. And to clarify further, in the stress notation we use at the English Wiktionary "свръ̀хесте́ствен" means [ˌsvrɤxɛˈstɛstvɛn], where the main stress is on the есте́ствен (estéstven) word.

BTW, https://ibl.bas.bg/rbe/lang/bg/България/ shows БЪЛГАРЍЯ, which can't be right. --Anatoli T. ^{(обсудить}/^вклад) 12:50, 6 April 2020 (UTC)

@Atitarev: Yeah, it is on the second е (свръхесте́ствен). My mistake, sorry about that. And yes, according to slovored.com, it does have a secondary stress (on the ъ̀ in the свръх prefix). Not quite sure about that, however. The word БЪЛГАРЍЯ is an obsolete form of the word българи (bǎlgari, “Bulgarians, Bulgarian people”), and that's why the stress is on the И — to imply the difference with the word БЪЛГА́РИЯ (the one that means the country Bulgaria). And for the record: I find it kind of funny that they've added the obsolete form and not the one that is being used in everyday speech. :-) --Ted Masters (talk) 13:59, 6 April 2020 (UTC)

@Ted Masters: No problem at all. I will address the other sense of България (Bǎlgarija) later. Thank you again for your help! --Anatoli T. ^{(обсудить}/^вклад) 22:16, 6 April 2020 (UTC)

Template:bg-adecl

I have documented {{bg-adecl}}. Some of the indicators are similar to those for nouns, but the overall structure is much simpler. Currently there is no way to override individual forms, but I may implement that if there's a need for it. There's probably a need to specify footnotes, e.g. to indicate that comparatives/superlatives are attested but rare. The module also can't yet support a few types of adjectives, e.g. късметли́я (kǎsmetlíja, “lucky”) (singular all genders късметли́я, definite късметли́ята, plural късметли́и, plural definite късметли́ите). There are also adjectives like сербе́з (serbéz, “bold”), which are more or less indeclinable but sometimes may take endings -а and -ът (under circumstances I'm not quite sure of). Benwing2 (talk) 08:11, 6 April 2020 (UTC)

@Atitarev, Bezimenen, Bogorm, Nauka Benwing2 (talk) 08:12, 6 April 2020 (UTC)

@Benwing2:: Thanks! Don't forget the "curly" cases like кой (koj), ни́кой (níkoj), чий (čij, “whose”), numerals(?) like еди́н (edín), два (dva), pronouns мой (moj), etc. Not sure if they have to be added to adjective or noun modules. Some of these use cases, e.g. кой (koj). Using {{subst:wgping|bg}} for the first time: (Notifying Benwing2, Bogorm, Bezimenen, Nauka, Ted Masters): . --Anatoli T. ^{(обсудить}/^вклад) 08:22, 6 April 2020 (UTC)

@Benwing2: All nominals in -лия (-lija) are treated as nouns (likely from earlier substantivized adjectives) in Modern Bulgarian. In particular, късметлия (kǎsmetlija) means a lucky person, not lucky (that would be късметлийски (kǎsmetlijski)). There is no need for bg-adecl to handle them. Безименен 09:42, 6 April 2020 (UTC)

@Bezimenen, Benwing2: Thank you. I've made късметли́я (kǎsmetlíja). --Anatoli T. ^{(обсудить}/^вклад) 22:16, 6 April 2020 (UTC)

@Benwing2: Hi. I am going through remaining adjectives, which haven't been converted by your bot. I assume your issue was missing stresses and whether or not comparatives are required. I will complete them over time - there were too many but there are less than a hundred remaining now, so that, if you want, you could focus on the modularisation of nominals - there pronouns, numerals, etc. --Anatoli T. ^{(обсудить}/^вклад) 02:15, 8 April 2020 (UTC)

@Atitarev I had already gone through about 55-60 of the remaining adjectives after my previous run in a text file so I went ahead and merged those changes with yours. I'll leave the remainder to you and work on the module. The first thing I'm working on is nouns with adjectival declension. At some point after that I'll get to pronouns and numerals. Benwing2 (talk) 05:17, 8 April 2020 (UTC)

@Benwing2: Thanks! Your bot has modified some of my changes as in diff and I have already mostly answered your question "does this have a comparative?" in my edits. --Anatoli T. ^{(обсудить}/^вклад) 05:27, 8 April 2020 (UTC)

(Notifying Benwing2, Bogorm, Bezimenen, Nauka, Ted Masters): Hi. All adjectives have now been converted to use {{bg-adecl}}. One adjective всежегом (vsežegom) is obscure, I don't know its stress pattern or declension type. --Anatoli T. ^{(обсудить}/^вклад) 05:29, 9 April 2020 (UTC)

Gender of place names

(Notifying Benwing2, Bogorm, Bezimenen, Nauka, Ted Masters): Hi. What is the gender of cities names like Ско́пие (Skópie) or Сара́ево (Saráevo)? --Anatoli T. ^{(обсудить}/^вклад) 06:25, 7 April 2020 (UTC)

Hello! The gender of the both names is neutrum. The checking method is an Noun phrase with attributive adjective as dependand, in this case the adjective accepts the gender of the argument of the Noun phrase: старо (n.) Скопие (n.), прескрасното (n.) Сараево (n.), but стара (f.) София (f.) or старият (m.) Пловдив (м.) --Nauka (talk) 08:34, 7 April 2020 (UTC)

@Nauka: Thank you, I have used this method myself but I forgot about it, LOL. —Anatoli T. ^{(обсудить}/^вклад) 09:10, 7 April 2020 (UTC)

Adjectivally-declined nouns

(Notifying Atitarev, Bogorm, Bezimenen, Nauka, Ted Masters): How do adjectivally-declined nouns like ру́ски (rúski, “Russian (language)”), ма́лък (málǎk, “child, youngster”), вое́нен (voénen, “soldier”), feminine ма́лка (málka), вое́нна (voénna) work? In particular, the vocative forms of nouns referring to people. The current declension of ма́лък (málǎk), which I copied from what was there before, says the vocative of ма́лък (málǎk) is either ма́лък or ма́лкия, both of which seem strange. Shouldn't it be ма́лки? Also, is the plural of ма́лък (málǎk) ма́лки or ма́лци? Finally, under which circumstances is the "extended (vocative) form" in -ий such as ма́лкий, но́вий, пра́вий, вое́нний, ру́ский used? Thanks for your answers. Benwing2 (talk) 00:57, 9 April 2020 (UTC)

@Benwing2: We need native speakers on this. Most likely, there are no additional vocatives, vocative = base form, малък=малък. I'll do some research. Also calling two more active Bulgarian editors - @Kiril kovachev, Tigerbaron. --Anatoli T. ^{(обсудить}/^вклад) 05:28, 9 April 2020 (UTC)

@Benwing2: Not sure if you saw it but on page 169 of Maslov's "Грамматика болгарского языка" the following vocative forms of adjectives are mentioned: (драг) "дра́ги прия́телю" - "dear friend", (беден) "бе́дни" - "poor". Only used for some masculine sg adjectives. Adjectives, like "бълга́рски", "балка́нски" are unchanged.

There are also some indeclinable adjectives of Turkic origin: ина́т (inát, “stubborn”), кески́н (keskín, “strong”) (кески́н раки́я - strong vodka), сербе́з (serbéz, “brave”), коджа́ (kodžá, “grown up”) or borrowings from European languages: е́кстра (ékstra, “selected, extra”). Some of these may get definite forms, only if substantivated, e.g. Иван Сербе́за (nickname). --Anatoli T. ^{(обсудить}/^вклад) 06:59, 9 April 2020 (UTC)

@Benwing2, Atitarev, as far as I know, the subject of substantivized adjectives is not fixed by the Standard language. More or less, the rule is to use your gut. For example, with ма́лък (málǎk) you can use the aforementioned ма́лък or ма́лкия, but you can also switch to ма́лчо, мъни́че. Technically, these are vocatives of different words - ма́лчо n (málčo), мъни́че n (mǎníče) - but they bear the same meaning. The plural of the adjective малък (malǎk) is малки. I don't think the noun has a plural - it is mainly used as a nickname, so one would rarely use it in plural form.

Names of languages in -ски are uncountable, declined as the masc. sg. form of the adjective - you just drop the had noun език - e.g. руски, руският /full article/, руския /short article/. No vocative or plural. военен (voenen, “one who serves in the army”), любим m (ljubim, “dear one”) /f. любима/, близък m (blizǎk, “close one, relative”) /f. близка/ have both vocative and plurals:

военен, военният, военния, военний /sg/
военни, военните, военни /pl/

The extended forms in -ий have mostly poetic function. I'm not sure what is their current status. Until 1945, they were acceptable, but probably not nowadays. You can find extended forms of adjectives in the works of earlier poets: e.g. новий, новая, новое, but these are no longer productive in Modern Bulgarian. It uses -ят, -та, -то instead. Dialectally, one can also use -ос, -са, -со (from Proto-Slavic *sь), -он, -на, -но (from Proto-Slavic *onъ), etc. One noticeable difference between the two classes of definite articles is that the innovated series follow head-second order, while the original *jь-based are usually head-first:

зелената гора /the green forest/, новото време /the new age/
vs отпор последний /the last stand/, борба свещеная /the holy struggle, fight/

There are exceptions, of course - родний край /native land, place of origin/, милое чедо /dear child/, and others. In these cases, though, -й, -я, -е are used for emphasize, not so much as a definite article. Безименен 09:49, 9 April 2020 (UTC)

@Bezimenen Thank you for the detailed explanation! Could you clarify a few things regarding vocatives of substantivized adjectives? You gave военний as the vocative of военен, which looks like an extended form in -ий. Can the vocative also be военни? Is the vocative of любим similarly любимий, and близък similarly близкий? What about любима, близка? Do they have vocatives любима, близка or любимо, близко (or близке)? Benwing2 (talk) 15:00, 9 April 2020 (UTC)

@Benwing2: I think the correct vocative is the -ий extension, but I'm not 100% sure. If an official source claims that it's -и, I'm willing to agree with it. I should clarify that historically военни, любими, близки. etc. are just contractions of the -и(й) form (from earlier pSl *-ъjь > OCS -ъɪй > Bulg. -и(й)). The difference is only orthographic, not phonetic. Both variants are pronounced /more or less/ the same. On this account, whatever academics from BAS say, that would be the official form. This is another example of "democratization" of the standard language, done by the communist regime, and I don't know which variant was accepted as the standard. These vocatives are mostly used in colloquial speech, so we didn't cover them in school.

Любима = Eng. honey (when referring to your girlfriend/spouse) and близка = Eng. female relative, close female friend exist. I'm not so sure about любимо, близко, though. My gut feeling tells me they do NOT exist, but I'm not absolutely sure. Probably, if you refer to a child, you may use them as nouns, but it would be odd. Безименен 01:57, 10 April 2020 (UTC)

PS мило n (milo, “dear one”) comes to my mind as a neuter substantivized noun. It can be used both for men and women.

Support for adjectival nouns, overrides, short forms, мой/твой/вой/не́ин/etc.

(Notifying Atitarev, Bogorm, Bezimenen, Nauka, Ted Masters): I added support for adjectival nouns (ру́ски (rúski, “Russian (language)”), ма́лък (málǎk, “youngster”)) and made various other changes to support possessive pronouns such as твой (tvoj, “your”) and не́ин (néin, “her”). See User:Benwing2/test-bg-ndecl and User:Benwing2/test-bg-adecl (at the bottom in both cases). It should now be possible to support all the possessive pronouns and probably also еди́н (edín). A little more work is needed for два (dva) but it should be possible to create tables for the remaining numbers (три, че́тири, пет, ...) using the /def_pl:... override. Benwing2 (talk) 06:17, 9 April 2020 (UTC)

@Atitarev I added наш, негов and един to User:Benwing2/test-bg-adecl. Benwing2 (talk) 15:29, 9 April 2020 (UTC)

@Atitarev I added support for два as well as кой-type and чий/какъ́в-type pronouns. See User:Benwing2/test-bg-adecl. Benwing2 (talk) 02:45, 10 April 2020 (UTC)

@Benwing2: Great, thank you! I have used this in entries. --Anatoli T. ^{(обсудить}/^вклад) 15:09, 10 April 2020 (UTC)

@Benwing2:: Hi. Could you please suppress the stress marks from the headers in all inflection tables, e.g "inflection of "", forms of ""? Hopefully it's an easy change:) --Anatoli T. ^{(обсудить}/^вклад) 23:20, 10 April 2020 (UTC)

@Atitarev Sorry, can you explain further what you mean? Benwing2 (talk) 00:04, 11 April 2020 (UTC)

@Benwing2: I want tables displaying "Positive forms of ве́сел", "Inflection of че́до", "Forms of ни́кой" without the stress mark on the inflection table titles themselves. --Anatoli T. ^{(обсудить}/^вклад) 00:19, 11 April 2020 (UTC)

@Atitarev I see. Yes that can be done. Just wondering why you want the stress marks removed? The Russian tables have stress marks in the headers. Benwing2 (talk) 00:57, 11 April 2020 (UTC)

@Benwing2: Maybe I had a fresh look and I didn’t like it but it’s OK, forget it, it may be useful for terms with multiple etymologies and stresses. —Anatoli T. ^{(обсудить}/^вклад) 01:07, 11 April 2020 (UTC)

@Atitarev OK just let me know if you want it removed in the future. I'm already removing monosyllabic stress marks so it's not hard to remove all of them. Benwing2 (talk) 01:14, 11 April 2020 (UTC)

lack of count forms for human males

(Notifying Atitarev, Bogorm, Bezimenen, Nauka, Ted Masters): It's been mentioned a couple of times that human males don't have count forms. I added the (h) indicator as a shortcut for "human". This is equivalent to writing (v)/-count (i.e. vocative but no count) but is easier and shorter to type. I'm gonna go through the nouns we have and convert the vocative indicators for male humans to the (h) indicator. Benwing2 (talk) 01:50, 11 April 2020 (UTC)

@Benwing2: Most human males have a countable form. човек m (čovek), ловец m (lovec) for example have човеци (pl. coll.), човека (pl. count.); ловци (pl. coll.), ловеца (pl. count.). The countable form мъжа of мъж (mǎž, “man”) is somewhat archaic - e.g. most people would say петима мъже (5 men) instead of петима мъжа (it depends on the dialect). This however is an exception, not the rule^{Edit: Apparently, not anymore.}. Безименен 09:54, 11 April 2020 (UTC)

(Notifying Atitarev, Bogorm, Bezimenen, Nauka, Ted Masters): OK, I am confused then. User:Bogorm specifically said

"Human male persons do not have count forms and they are always in accordance with the respective form of the numeral (in this case, петѝма), but the count form is inadmissible even without this form of the numeral."

Bulgarian language says:

"However, a recently developed language norm requires that count forms should only be used with masculine nouns that do not denote persons. Thus, двама/трима ученици ('two/three students') is perceived as more correct than двама/трима ученика, while the distinction is retained in cases such as два/три молива ('two/three pencils') versus тези моливи ('these pencils')."

Under бройна форма (brojna forma) it says:

"In modern Bulgarian the use of the count form for nouns, which represent people, is frowned upon and is considered incorrect by some."

It sounds like the idea that count forms should not be used for human males is recent and not universally accepted. I'm not sure what we should do. I already made the change to add (h) for human males to disable the count form. If we decide this is not a good idea, I can either undo this change or change the handling of (h), e.g. to indicate that count forms for male humans exist but are obsolescent, or often considered incorrect, or something. Benwing2 (talk) 19:34, 11 April 2020 (UTC)

(Notifying Atitarev, Bogorm, Bezimenen, Nauka, Ted Masters): Does this rule concerning the count form not existing any more apply to mythological beings and such, e.g. а́нгел (ángel), бог (bog), дя́вол (djávol), стопа́нин (stopánin, “home spirit (dialectal)”)? How are they counted, using два or два́ма? Benwing2 (talk) 19:38, 11 April 2020 (UTC)

@Benwing2: I didn't know about that. I knew of an even earlier norm which was revoked, but I didn't know the rule got even more simplified. Before 1945, the norm was indeed confusing. It stated that one should use:

* countable form when the referred objects are enumerable and in orderly state: e.g. двама играча = 2 players (in determined order);

* collective form when they were disordered, even if they are enumerable: e.g. двама играчи = 2 players (in undetermined order); няколко играчи = several players;

Anyway, we should stick to the newest rules made by BAS. Ignore my previous comment.

Regarding the other question about "два"/"двама" - the translation of Genesis 19:1 uses двама ангела... On the other hand, there are several jokes centered around 2 devils, who discuss their devilish deeds. Most of the jokes use the expression два дявола. To add to the confusion, if they don't communicate, but argue, the form двама дяволи is used. So, yeah... I'll leave other native contributors share their observations. I'm in a pickle on this topic. Безименен 22:49, 11 April 2020 (UTC)

@Benwing2, Bezimenen: Thank you for this enhancement (h). For me it's a bit confusing but it seems that modern Bulgarian grammarians frown upon the usage like "три мъ́жа" vs "три́ма мъ́жа". We probably want numeral tables include the "-ма" forms as in {{Bulgarian cardinal numerals}}. There is still a count form for male humans, even if it's with "-ма" numerals, isn't there? Maybe the other usage "три мъ́жа" should be marked as dispreferred, obsolete or rare? --Anatoli T. ^{(обсудить}/^вклад) 01:57, 12 April 2020 (UTC)

@Bezimenen From looking through Google Books, it appears that both "два ангела/два дявола" and "двама ангели/двама дяволи" are possible. I imagine usage is somewhat unsettled. Benwing2 (talk) 03:04, 12 April 2020 (UTC)

(Notifying Atitarev, Bogorm, Bezimenen, Nauka, Ted Masters): I found a Bulgarian grammar in English ("A Concise Bulgarian Grammar" by John Leafgren, written in 2011 or later) that has the following to say about "special" forms like два́ма, три́ма, etc.:

The use of these special forms with nouns denoting male humans, however, is not obligatory.

They occur more often with smaller quantities, with the use of the primary cardinal numbers increasingly frequent with larger numbers. Note also that the nouns used with these special masculine human cardinal numbers must be in their regular plural forms (not the numerical forms), while when such a masculine human noun is, in fact, used with a primary cardinal number, both the regular plural and the numerical form of the noun are possible, with the former increasingly likely as quantity increases. Here are some examples:

‘the 2 students (male or mixed gender10)’ два́мата студе́нти/два́та студе́нти/два́та студе́нта

‘3 pupils (male or mixed gender)’ три́ма учени́ци/три́ учени́ци/три́ учени́ка

’25 teachers (male or mixed gender)’ два́йсет и пети́ма учи́тели/ два́йсет и пе́т учи́тели/ два́йсет и пе́т учи́теля

Does this resonate with the native speakers? If so it suggests the count forms are not ungrammatical. Benwing2 (talk) 04:02, 12 April 2020 (UTC)

Here's what Maslov says:

Надо учесть, что лично-мужские формы широко применяются только при указании на небольшие количества лиц, причем и в этих случаях не являются строго обязательными: рядом с "пети́ма войни́ци" "пять солдат" всегда возможно "пет ду́ши войни́ци", а изредка встречается и "пе́т войни́ка"; но если речь идет о более чем десяти лицах, явно предпочитается неспециализированная форма числительного в сочетании с обычной формой множественного числа: "петдесе́т и пе́т войни́ци".

which translates as "It should be noted that personal-male forms are widely used only when referring to small numbers of people, and in these cases are not strictly mandatory: next to "пети́ма войни́ци" “five soldiers” is always possible "пет ду́ши войни́ци", and occasionally "пе́т войни́ка"; but when it comes to more than ten persons, an unspecialized form of the numeral is clearly preferred in combination with the usual plural form: "петдесе́т и пе́т войни́ци"." Benwing2 (talk) 05:01, 12 April 2020 (UTC)

@Benwing2: Oh, yes, I forgot to mention that large number of dynamic entities are treated as uncountable, exactly as Maslov says. The distinction is not so much between countable vs uncountable rather than orderly vs disorderly. If you can't properly imagine the group arranged as a tuple, you use the normal plural form.

I don't know if you have read that somewhere, but diachronically the countable form in Bulgarian continues the historical "dual", while the collective form is the historical normal "plural". All numerals with -ма are also based on the dual numeral in Proto-Slavic. Originally, the dual was used when entities come in pair, which apparently was generalized to any countable, organized set /triple, quadruple, etc/. Entities which are disorganized (even if enumerable) are not marked as countable. This is the case with large number of animate objects.

The best solution may be to give the countable form of animate male nouns as archaic/rare, but the topic is really confusing, so independent opinion from other native speakers is needed. Безименен 08:49, 12 April 2020 (UTC)

@Bezimenen, Benwing2: Let's do it this way, rather than removing the count forms altogether. We can come up with a label or description.

My question above was probably missed. Do you think it's a good idea to add the human masculine forms to numeral declension tables as in {{Bulgarian cardinal numerals}}? --Anatoli T. ^{(обсудить}/^вклад) 11:59, 12 April 2020 (UTC)

@Atitarev IMO два́ма is a separate word from два, but we should definitely have a usage note on два saying something about два́ма, and similarly три/три́ма, etc. Benwing2 (talk) 15:23, 12 April 2020 (UTC)

@Bezimenen:, regarding the two angels, in the translation of the Liber Genesis into Bulgarian according to the edition approved by the Bulgarian Orthodox Church they are mentioned as двамата Ангели (capitalised and with the regular plural form). The random website containing the purported translation is apparently repræsentative of another confession. Bogorm converſation 20:43, 12 April 2020 (UTC)

@Atitarev: (There is still a count form for male humans, even if it's with "-ма" numerals, isn't there? Maybe the other usage "три мъ́жа" should be marked as dispreferred, obsolete or rare?) They are ungrammatical, both три *мъжа and трима *мъжа. Трима мъже is the correct form. But if you insist on allowing those forms, then I would opt for proscribed as a tag and perhaps a usage note discouraging them. Bogorm converſation 20:43, 12 April 2020 (UTC)

@Benwing2, Bogorm, Bezimenen: Thank you for responses. I've made a stub for два́ма (dváma). It needs an inflection table to include the definite form два́мата (dvámata) (it's also beneficial since you include the accelerated form creation, which is great). @Benwing2, not if any of existing patterns will fit. Please check otherwise. @Bogorm, No, I won't insist on including count forms for males. --Anatoli T. ^{(обсудить}/^вклад) 23:58, 12 April 2020 (UTC)

@Atitarev You can just use {{bg-ndecl}} here. Benwing2 (talk) 01:45, 13 April 2020 (UTC)

Nouns in -ане

(Notifying Atitarev, Bogorm, Bezimenen, Nauka, Ted Masters): It's not mentioned in Bulgarian nouns, but it looks like most nouns in -ане with a plural use the plural -ания, e.g. схва́щане (shváštane), пъту́ване (pǎtúvane), присти́гане (pristígane), ди́шане (díšane). The current module has them using the plural -анета, which I'll fix. What about nouns in -ене? I see тъ́рсене (tǎ́rsene, “search, demand”) pl. тъ́рсения, but хо́дене (hódene, “walking, going”) pl. хо́денета or хо́дения, я́дене (jádene, “meal”) pl. я́денета, си́рене (sírene, “cheese”) pl. си́ренета. Which plural is more common, -ия or -ета, or does it "just depend on the word"? Benwing2 (talk) 20:23, 11 April 2020 (UTC)

@Benwing2: It's subjective. In principle, all verbal nouns ending in -не < *-nьje should be declined with -ния. These with -нета are reinterpreted as nt-stems, which has happened irregularly. A similar phenomenon occurred with the outcome of *moře, *lože, *ǫže (originally neuter yo-stems). The reason for the mixing was because proto-Slavic *-е, *-ьje, and *-ę got all mixed in Modern Bulgarian, so the declension patterns mixed as well. Безименен 00:12, 12 April 2020 (UTC)

Changing ==Inflection== to ==Declension==

(Notifying Atitarev, Bogorm, Bezimenen, Nauka, Ted Masters): If no one objects, I'm gonna change the header of noun and adjective declensions from ==Inflection== to ==Declension== for consistency with Russian and other languages. Benwing2 (talk) 01:49, 14 April 2020 (UTC)

@Benwing2: Fine by me. --Anatoli T. ^{(обсудить}/^вклад) 02:25, 14 April 2020 (UTC)

@Benwing2: Why would you præfer Declension over Inflection for Bulgarian? Is this the prævalent terminology in English-language grammar books? Vestiges of cases have survived only in the personal, relative and interrogative pronouns (some archaic forms for demonstrative pronouns too) in modern Bulgarian. Bogorm converſation 20:58, 15 April 2020 (UTC)

@Bogorm, Benwing2: Hi. It's important to have a lot of consistency in a multilingual dictionary, so that editors, user and bots can work with multiple languages. I have checked multiple languages with inflections and all nouns use ==Declension==, e.g. using the word for "fork", as I just created a Belarusian entry відэ́лец (vidéljec), شَوْكَة (šawka), Gabel, widelec, vidlička, काँटा (kā̃ṭā). I think it's a good rationale, unless there is something really wrong for a particular language. --Anatoli T. ^{(обсудить}/^вклад) 00:06, 16 April 2020 (UTC)

@Benwing2: @Atitarev: Unlike all those languages, Bulgarian does not have cases except for some pronouns. In this regard it can be compared to Danish or Norwegian. I just took a look at two accidental Danish words like tid#Danish and højde and to may dismay one of them had inflection and the other declension as title, so there is clearly a lack of consistency in the titles, but the template already had declension. The Danish forms have eight distinct forms, the nominative and genitive for definite and indefinite singular and plural. The Bulgarian noun has four distinct forms for the nominative (like a Danish noun in singular) and only some Bulgarian nouns have a singular vocative form which is distinct from the singular nominative, but the plural form always coincides with the plural nominative, so there would be even less distinct entries in the table (at most five). If this could be considered enough for a whole table, I would not object, but I am still curious about the prævalent term in English-language grammar books of the Bulgarian language, since I am unfamiliar with them. Bogorm converſation 16:15, 16 April 2020 (UTC)

@Atitarev: I know User:Rua much prefers ===Inflection=== to ===Declension=== and has used that header for the languages she's worked on, so there definitely isn't any cross-website consistency! PUC – 16:43, 20 April 2020 (UTC)

@PUC: Thanks, you probably want to notify User:Bogorm about this. I don't have the answer about the English usage. It seems it's up to dictionary makers. Also, Bulgarian has at least remainders of cases, North Germanic languages don't. --Anatoli T. ^{(обсудить}/^вклад) 22:32, 20 April 2020 (UTC)

Support for multiword expressions

(Notifying Atitarev, Bogorm, Bezimenen, Nauka, Ted Masters, Kiril kovachev): Now present. See documentation on {{bg-ndecl}}. Benwing2 (talk) 00:05, 23 April 2020 (UTC)

@Benwing2: Thank you! --Anatoli T. ^{(обсудить}/^вклад) 00:18, 23 April 2020 (UTC)

Nouns in -ан

@Atitarev I changed things so nouns in -а́н have -и as default plural. There are actually a lot of multisyllabic nouns in -ан:

аероплан
балкан
банан
бутан
великан
волан
вулкан
гарван
диван
дуриан
дъждобран
езан
екран
калкан
капитан
кардан
катран
колан
лантан
манган
мегдан
мързелан
наркоман
океан
оман
омбудсман
паткан
патладжан
продан
ресторан
стопан
таван
титан
тъкан
ураган
уран
фазан
хулиган
чобан
шаран
юрган

Some of them are not stressed on the last syllable, and some don't have a plural, but as far as I can tell all of them with a plural have -и not -овци; only мързела́н (+и or +овци) and готова́н (+овци) use -овци. I went through and removed an explicit +и where it existed. Benwing2 (talk) 16:23, 3 May 2020 (UTC)

@Benwing2: Thanks for addressing this. --Anatoli T. ^{(обсудить}/^вклад) 22:52, 3 May 2020 (UTC)

Template name?

@Benwing2 Will the name remain as it now: bg-ndecl? I think bg-decl-noun (which already exists) or at least bg-noun-decl is better. Безименен 11:33, 29 March 2020 (UTC)

@Atitarev, Bogorm I like 'bg-ndecl' because it's short and relatively easy to type, but if others prefer something else e.g. 'bg-decl-noun' or 'bg-noun-decl', I can change it. 'bg-decl-noun' could potentially be made to behave the new way if the first param has <> signs in it, and the old way otherwise. Benwing2 (talk) 06:18, 30 March 2020 (UTC)

@Benwing, Bezimenen, Bogorm: What Безименен is asking makes sense for consistency and will make it easy to find by its parts but I'm not insisting on the change. The template's parameters are not easy to follow because they are very different from any others but I guess it just takes some getting used to. Good work in a short time, especially considering how relatively poor online resources are! Still plenty to do, though. (BTW, Ben, please share your offline resources with me as well, if there are any I haven't got).--Anatoli T. ^{(обсудить}/^вклад) 06:26, 30 March 2020 (UTC)

@Bezimenen, Bogorm, Atitarev I modified {{bg-decl-noun}} so you can use the new syntax as well as the old one. I am in the process of documenting {{bg-ndecl}} properly, which should make it easier to understand. Benwing2 (talk) 01:54, 31 March 2020 (UTC)

Accent used to Indicate Stress

Shouldn't the accent used be a grave accent rather than an acute one since that is the standard within Bulgarian? Terinuva 17:58, 25 April 2020 (UTC)

@Terinuva:: Hi. We are using acute accents at Wiktionary. WT:BG TR --Anatoli T. ^{(обсудить}/^вклад) 23:23, 25 April 2020 (UTC)

-овци

@Benwing2: Hi. In this revision I got the plural form "бана́новци" but I was expecting "бана́ни". What's the difference between e.g. журна́л (žurnál) and бана́н (banán)? Is -овци ending common for this type of common nouns? --Anatoli T. ^{(обсудить}/^вклад) 08:05, 28 April 2020 (UTC)

@Atitarev Per Maslov, nouns in -а́н tend to have a plural of -овци (e.g. мързела́н (mǎrzelán)). Loanwords are an exception; I'm not sure how many loanwords in -а́н there are. Benwing2 (talk) 15:05, 28 April 2020 (UTC)

@Benwing2: Thank you. I have added three more loanwords I could think of: екра́н (ekrán), капита́н (kapitán) and вола́н (volán). --Anatoli T. ^{(обсудить}/^вклад) 00:17, 29 April 2020 (UTC)

Vocatives for masculine country names ending in a consonant

@Bezimenen, Kiril kovachev, Bogorm, Benwing2: Hi. User:Agnese (mostly active on the Polish Wiktionary) asked me regarding vocative forms, such as Азербайджане, Иране. I based my edits on Yu. S. Maslov's "Грамматика болгарского языка" and some research. Could you please confirm that such vocative forms for masculine are valid? It's easier to find feminine forms like -ийо for obvious reasons. --Anatoli T. ^{(обсудить}/^вклад) 00:14, 19 January 2021 (UTC)

@Atitarev: On a theoretical level, these forms do exist, but I can't think of any example where one would use them. Except for България (Bǎlgarija) (e.g. in the poem Българийо, за тебе те умряха) and Македония (Makedonija) (e.g. Гордей се майко Македонийо), I don't think any other country has usable vocative. Безименен (talk) 10:30, 19 January 2021 (UTC)

@Bezimenen: Thank you! I was able to confirm vocatives for feminine countries but not the masculine. I will remove them. --Anatoli T. ^{(обсудить}/^вклад) 10:40, 19 January 2021 (UTC)

Way to override definite singular and plural in multi-word terms?

Hello, I have been editing lately, and found an instance where I couldn't generate a sensible declension table for контактна леща. If I place basic <> on both terms, it generates forms like „контактната лещата,“ whereas the „-та“ should only be on „контактната“. If I meanwhile try to override the definite form to „леща,“ the module raises this error: "Lua error in Module:bg-nominal at line 1269: Unrecognized ending for definite subjective singular: 'конта́ктната ле́ща'". Presumably because it ought to end in „та“ or similar. I couldn't find a way to generate the correct declension. Is there any way to do this currently? Thanks. @Benwing2 Kiril kovachev (talk) 12:41, 11 July 2021 (UTC)

@Kiril kovachev See Template:bg-ndecl#Multiword expressions. Benwing2 (talk) 15:14, 11 July 2021 (UTC)

Thanks for your help :) Kiril kovachev (talk) 08:56, 12 July 2021 (UTC)

Automatically detect feminine for nouns ending in -ост

@Benwing2 Hello, what would you say to the idea of detecting nouns that end in -oст as being feminine, precluding the need to write out the (f) indicator for them? This is a fairly common suffix, shared by most abstract nouns, so it might be a shout. I also remember forgetting the (f) once, admittedly out of carelessness, but I think it's a much more likely default to have than the opposite (I don't know of any words with this ending that aren't feminine). Kiril kovachev (talk・contribs) 20:02, 9 August 2023 (UTC)

@Kiril kovachev Yup, the Ukrainian and maybe Belarusian modules already do this. (I don't think the Russian module does this but it should, and for Czech it's required in all cases to specify the gender because there are so many irregularities in Czech declension.) The Ukrainian module (see Module:uk-noun#L-1706) detects -ець/-єць as masculine, -тель as masculine, and -ість/-їсть as feminine. I don't know if the other suffixes need to be handled specially for Bulgarian, maybe they already default to masculine. Benwing2 (talk) 21:34, 9 August 2023 (UTC)

@Benwing2 I believe consonant-ending words already default to masculine, so it should just be this if I'm not mistaken. Kiril kovachev (talk・contribs) 21:36, 9 August 2023 (UTC)

OK, sounds good. I guess the difference is the palatalization marked by the soft sign has gone away at the end of words in Bulgarian. Benwing2 (talk) 21:37, 9 August 2023 (UTC)

The only thing is, I don't know for certain that there's no word ending in -ост where (f) doesn't apply: I think I'll try to create a tracking category so we can see if there are any before changing it. Kiril kovachev (talk・contribs) 21:37, 9 August 2023 (UTC)

@Kiril kovachev Sounds good. The only words I can imagine being problematic are those like кост and гост, and we can check for there being a syllable preceding the -ост. Benwing2 (talk) 21:39, 9 August 2023 (UTC)

@Benwing2 Good point, I'm hoping that these will be so few though that they can just be manually itemised; I'm not confident I could write a safe check for preceding syllables to run at runtime in the module, I don't think... maybe just counting vowels? Kiril kovachev (talk・contribs) 21:45, 9 August 2023 (UTC)

@Kiril kovachev Just check for a preceding vowel in the Lua pattern, no need to count vowels. Almost certainly the number of exceptions is very small. Benwing2 (talk) 21:51, 9 August 2023 (UTC)

@Benwing2 It turns out the only exceptions (praying my tracking code was correct) were лост, тост, гост, and мост. I also fished out some that should've well been marked (f), but weren't: благодарност, несерио́зност, сладост, and смелост. Anyway, I now implemented what you suggested: if it ends in -ост and contains a preceding vowel, it assumes a feminine gender, otherwise behaves as it did before. Feel free to critique any changes (diff). What do we now do with all these marked as (f) unnecessarily? Just leave them? Thanks for your help, Kiril kovachev (talk・contribs) 23:11, 9 August 2023 (UTC)

@Kiril kovachev In terms of the pattern, I'd just do this:

rfind(lemma, com.vowel_c .. "[^ ]*о́?ст$")

This uses the list of vowels already in Module:bg-common, although probably we should add ѐ and ѝ to that list (or alternatively, call mw.ustring.nfd beforehand to convert the string into decomposed format). Benwing2 (talk) 23:36, 9 August 2023 (UTC)

@Kiril kovachev BTW I'll do a bot run to remove (f) from nouns in -ост. Benwing2 (talk) 23:43, 9 August 2023 (UTC)

Also, I moved the talk page stuff on Template talk:bg-ndecl to this page, where the other related discussions are. Benwing2 (talk) 23:43, 9 August 2023 (UTC)

@Benwing2 Ah okay, so the vowels were there after all. I couldn't find them in this module, like we had them in bg-pronunciation, so I guessed that we didn't have it—do we re-write it in bg-pronunciation to save on the module import perhaps? Else I'd have considered re-using the vowels from there. Also duh, I don't know why I overcomplicated that match... got it in my head once that I had to strictly consider the left side and so starting doing some substring gymnastics. I'll port that pattern instead. What do you mean by ѐ, BTW? I'm not sure that letter is used in Bulgarian, although ѝ definitely is. It's also difficult because in some places you can be certain there won't be an ѝ, e.g. in the pattern I used the set of vowels really didn't need to include it, because it only ever occurs by itself. I guess it's still a vowel, and it shouldn't cause trouble to include it in there, but since we use it in patterns we should just be careful that doesn't somehow affect those that use vowels_c. And also thanks for the bot run. Kiril kovachev (talk・contribs) 00:12, 10 August 2023 (UTC)

@Kiril kovachev We should probably have Module:bg-pronunciation use Module:bg-common; the only reason not to would be to possibly save a bit of memory, and it's unlikely to matter for Bulgarian pages since they aren't memory hogs (which happens normally for Latin-script pages with just 2-4 letters). I'm including ѐ because it might show up e.g. in respelling with secondary stress indicated (ѐ is a single Unicode char). Benwing2 (talk) 01:06, 10 August 2023 (UTC)

@Benwing2 Right, that makes sense-do you have a ballpark estimate as to what the memory difference even would be? At any rate I agree it would be good to depend on bg-common, if just to synchronise all the like data. Kiril kovachev (talk・contribs) 09:45, 10 August 2023 (UTC)

@Kiril kovachev I'm not really sure; I could guess on the order of a few hundred kilobytes but that would be just a guess. It might vary a lot. User:Theknightwho can you speak to this? Benwing2 (talk) 18:52, 10 August 2023 (UTC)

Hmm, as long as it's around there I take it it would be fine; the module itself isn't that heavy datawise, so I hope it wouldn't weigh down any pages that use it; and moreover like you said the memory limit is likely way above what we tend to use on most pages... I would guess, but I don't know either... Kiril kovachev (talk・contribs) 18:57, 10 August 2023 (UTC)

@Benwing2 I can't speak to memory issues any more than you can, unfortunately, as the design of Scribunto means that memory use can fluctuate unpredictably and disproportionately. I would say that there's nothing to worry about if only 2 or 3 pages are pushed into CAT:E, as that's likely down to random statistical variance affecting pages already very close to the limit. Theknightwho (talk) 19:33, 10 August 2023 (UTC)

Replacing subjective/objective with subject/object

Hi @Benwing2 and @Kiril kovachev,

Per the discussion regarding Module:bg-verb, I'd like to request that the inflection tags sbjv and objv be replaced with sbj amd obj. The former two refer to grammatical cases which Bulgarian does not have, while the latter two refer to the subject/object of a verb, which tracks closely the distinction between the two forms of the definite article.

I made the change to Module:bg-verb, but I don't have permissions to edit this module. Could one of you please make the necessary change? We'll also need to run a job to fix noun/adj/verb non-lemma form entries that were created with "subjective"/"objective".

Thanks,

Chernorizets (talk) 21:36, 31 August 2023 (UTC)

@Chernorizets Done. I hope this diff was what you meant? Where are we supposed to check what effect this has had? Kiril kovachev (talk・contribs) 00:05, 1 September 2023 (UTC)

@Kiril kovachev yes, that's the change I had in mind. It wouldn't affect any already-created non-lemma forms for adjectives or nouns, but you could go to any e.g. Bulgarian noun and, if you try to create a definite masculine noun form (using the accelerated links, if you have that enabled), you'll notice that it says "subject/object" instead of "subjective/objective". Chernorizets (talk) 00:32, 1 September 2023 (UTC)

@Chernorizets Ah, okay, nice one. I'm not yet acquainted with the whole accelerated system—do you use it, and do you find it useful?

Also, re:fixing existing generated entries: this seems to be a trivial bot task, and I could have it ready as soon as tomorrow. We just need to change "sbjv" and "objv" to "sbj" and "obj", correct? In this case, I can make a temporary tracking category for entries with these former attributes, and just change them to their new values one by one. In fact, I'll see if I can put the tracking category in right now. Kiril kovachev (talk・contribs) 00:58, 1 September 2023 (UTC)

Never mind that. It would seem I don't have the permissions, as if I'm not mistaken the change would have to go all the way up to {{inflection of}}. Fortunately, there's an information dump from kaikki.org, which I've already got downloaded, which should have a list of entries with this template as part of the data. Kiril kovachev (talk・contribs) 01:03, 1 September 2023 (UTC)

@Kiril kovachev I do use it. You can enable it from Preferences > Gadgets > Add accelerated creation links for common inflections of some words. What you'll see then when you expand a declension/conjugation table is green links that, when you click them, auto-populate a new page with the non-lemma form, passing all the correct information to the {{inflection of}} template. Lately I've decided that I want to add plurals for BG nouns, and this does (most of) the job for me. I'm also thinking of using it to add verb participles (разбран, разбрал, разбиращ, разбирайки), which may not be too easy to guess from the root verb form. Chernorizets (talk) 01:04, 1 September 2023 (UTC)

@Chernorizets Thanks for the detailed instructions. Indeed that does sound helpful; so far I've not been in the business of adding inflected forms, but that might be a good goal for the future (I feel like we don't have that many right now). Does it already have support for those verb forms right now? Kiril kovachev (talk・contribs) 01:22, 1 September 2023 (UTC)

@Kiril kovachev Yes. Check out Category:Bulgarian non-lemma forms and drill down by whatever seems interesting, e.g. noun forms, verb forms, etc. A few concrete examples: абдикиращият, абитуриентът, магазина, авангардният. Chernorizets (talk) 01:52, 1 September 2023 (UTC)

@Chernorizets Thanks again! I just tried it out, and it works very nicely. I feel a luddite for having never thought to use it before, lol. Kiril kovachev (talk・contribs) 18:11, 1 September 2023 (UTC)

@Chernorizets I now have the script principally ready to convert the forms: please see GitHub if you want to look at it. I'll test it some more, but whenever we're ready, please let me know and I can set it off. Kiril kovachev (talk・contribs) 18:12, 1 September 2023 (UTC)

@Kiril kovachev There are a few things to keep in mind:

• {{infl of}} is an alias of {{inflection of}}, so if there are any non-lemma forms created with the alias, the code won't touch them. I don't know OTOH if that's really a thing today.
• with words like часа (časa), this could either be the object definite article of час (čas), or the countable form as in "два часа по-късно". Because the stress is different in each case, you have two separate subsections within the Bulgarian section. I'm not sure how the code handles that.
• {{inflection of}} can be used to specify multiple inflections at the same time, which is what would happen with a word like "моряк", where "моряка" is both the object definite form, and the countable form (no change of stress). Again, I'm not sure how the code would handle that. You can see an example of multiple inflections in кука (kuka), under the verb section.

Cheers,

Chernorizets (talk) 06:40, 2 September 2023 (UTC)

@Chernorizets Thanks for the feedback, I hadn't considered that there would be an alias template, so that will definitely be of use. Bear in mind that the list I had generated to iterate through, though, fortunately, is not based on the presence of those raw templates, but rather the presence of the "objective" or "subjective" data tags within the pre-processed data set, so in as far as knowing what pages to edit, we are still set.

As for the effect on multiple cases: the program does indeed apply the "fixing" to every inflection template it sees, but this should be a no-op for any {{inflection of}} that doesn't have a sbjv or objv parameter: see the code here, which is the only place the page contents get modified:

def fix_inflection(template: mwparserfromhell.wikicode.Template):
    for param in template.params:
        param: mwparserfromhell.nodes.extras.Parameter
        if param.value == "sbjv":
            param.value = "sbj"
        elif param.value == "objv":
            param.value = "obj"

I.e. for any other {{inflection of}} attribute, it won't do anything, since neither if arm is executed.

What the code does do is find every single {{inflection of}} template under the ==Bulgarian== section, but as mentioned above, it's my belief that this won't cause any problems at all, because it doesn't affect cases that it shouldn't.

Does this sound good? Kiril kovachev (talk・contribs) 12:51, 2 September 2023 (UTC)

@Kiril kovachev sounds good, as long as it iterates through every occurrence of the template (which I thought it did) and replaces just those parameters, then we should be good to go. Chernorizets (talk) 21:36, 2 September 2023 (UTC)

@Chernorizets Yep, then we should be golden. I actually noticed a truly fatal bug prior to updating it today, which was that the regex was not escaped, and so it was actually iterating all templates whatsoever! Fortunately, no adverse effects so far — I've checked all the existing test attempts — because no other template ever uses "sbjv" or "objv" as a parameter. However, the problem is fixed now anyway, so it should be fine. I don't know if I should wait any more; it doesn't seem like anyone is objecting... Kiril kovachev (talk・contribs) 22:32, 2 September 2023 (UTC)

For the record, I am going to set it going now. I figure that we're in agreement. If not, undoing this change will be just as easy as doing it, so no harm done. Kiril kovachev (talk・contribs) 22:59, 3 September 2023 (UTC)

Suppress vocative forms of adjectives by default

Hi @Benwing2, Kiril kovachev,

I'd like to bring parity between Bulgarian nouns and adjectives in suppressing vocative forms by default. The vocative is used idiomatically in Bulgarian, and doesn't make sense on most nouns and adjectives, even if it's grammatically possible. E.g. you could in theory say бавний компютъре (bavnij kompjutǎre, “o, slow computer”), but it sounds like a line from a joke or a skit. The change I am requesting is to suppress vocative adjective forms by default, and be able to toggle them on explicitly with (v) as for nouns.

For backward compatibility, I would preserve the -voc option, making it essentially a no-op, since many adjectives use it today to explicitly suppress the vocative (including all the ones I've created). In essence, this change would remove the need for me and others to almost always have to specify -voc. If we do make the change, there could be a bot job later to just nix that option on the pages where it's used, allowing us to fully deprecate it.

Thoughts? I don't have the permissions to edit this module, so if we want to keep it that way, I'd be looking to one of you to make this change.

Thanks,

Chernorizets (talk) 03:55, 4 November 2023 (UTC)

@Chernorizets I'm a bit confused why you want this, since any adjective (notionally at least) can be attached to any noun, meaning that animacy (which is more or less a requirement for the vocative) isn't inherent in an adjective, unlike for nouns. I imagine a word like бавен, for example, can be used to describe people. It seems like the decision regarding whether a given adjective merits a vocative form in its declension table would be hard to make, esp. for less-common adjectives, and would lead to a lot of inconsistencies. Benwing2 (talk) 05:03, 4 November 2023 (UTC)

@Benwing2 I had to go back and consult some grammars before replying here. @Bezimenen keep me honest, but my understanding (backed by what I've read and my own experience as a speaker) is that the use of the vocative for adjectives in standard contemporary Bulgarian is even more restricted than that for nouns. In fact, I can only think of a handful of examples - emotionally-charged adjectives like мил (mil), скъп (skǎp) and любим (ljubim) in fixed-ish expressions like мили татко (mili tatko, “dear dad”) and скъпи приятелю (skǎpi prijatelju, “dearest friend”). At least one grammar went on to claim that case distinctions are neutralized for adjectives. Ultimately, vocative forms are most commonly used with given names, and secondarily with some nouns denoting people (literally or metaphorically). Even there, a whole class of feminine vocatives is now considered rude and to be avoided. Chernorizets (talk) 07:00, 4 November 2023 (UTC)

@Chernorizets OK. So presumably, since бавен can mean "retarded", how would you say something like "hey, retarded dude, get away from me!"? Obviously something like that would be quite rude, but since we're a descriptive dictionary rather than a prescriptive one, if people actually use бавни in such a case, we should include it (possibly with a note about its perceived rudeness); same for those rude feminine vocatives. But if the actual set of adjectives that can be used in the vocative is limited, it does seem to make sense to make the vocative setting be opt-in rather than opt-out. Benwing2 (talk) 07:51, 4 November 2023 (UTC)

@Benwing2 I would use бавен (baven), as in ей бавен/тъп, разкарай се (ej baven/tǎp, razkaraj se). If I used бавни (bavni) it would imply I was addressing multiple people, not one person. I'd like to get Bezimenen's take on the overall proposal before making changes, but I feel confident. Chernorizets (talk) 08:09, 4 November 2023 (UTC)

We should also update Wiktionary:About Bulgarian with this, since I've seen non-speaker editors blithely create various vocative forms because they show up in inflection tables. Chernorizets (talk) 08:11, 4 November 2023 (UTC)

@Chernorizets OK. It looks like you nominalized the adjective; can you construct a sentence where the vocative expression is an adjective-noun phrase involving бавен or тъп? Does it still appear in the nominative/unmarked form? Benwing2 (talk) 08:37, 4 November 2023 (UTC)

@Benwing2 I would still say something like ей, идиот тъп (ej, idiot tǎp, “hey you dumb idiot”), as opposed to ей, идиоте тъпи (ej, idiote tǎpi), which sounds like it would be a line in a 19th-century period piece of theater. Chernorizets (talk) 08:50, 4 November 2023 (UTC)

@Chernorizets OK, thanks. Benwing2 (talk) 08:55, 4 November 2023 (UTC)

@Chernorizets One final thing: Are these vocatives commonly found in older writing that readers might encounter? If so, it might make sense to include them and mark them as archaic. We do currently include some dated/dialectal/archaic/etc. plurals because they are found in dictionaries. Benwing2 (talk) 08:57, 4 November 2023 (UTC)

@Chernorizets Hi, I'm in agreement with everything in the thread so far. I'm rather ambivalent whether we choose to remove them by default, but I can understand this from the perspective that the vast majority of adjectival vocatives I have never once seen in the wild, and it's basically only the few examples you listed above that I hear semi-regularly today; even in those, they are to me more like fossilized set expressions rather than a productive category. Regardless, if they technically exist for every adjective, providing the hypothetical form is potentially still desirable. I was just thinking today about vocatives of nouns as well, and although indeed most adjectives and nouns have seldom seen their vocatives used, wouldn't it be best for readers to at least know (a) that such a thing as "vocative" even exists, and (b) what form to use for each word? A marking, like the kind that Ben proposed above, might just be a good solution to this. In this variant of the solution, we would bring nouns to parity by including a "archaic"/"fanciful"-labelled vocative for any noun that could have one; but I'm not particularly advocating for that necessarily, just bouncing ideas. What do you think? I will support whatever we choose anyway. Kiril kovachev (talk・contribs) 23:46, 4 November 2023 (UTC)

FWIW we do include vocatives for all Ukrainian and Czech nouns, although I don't know if that's because of a difference in usage between vocatives in Bulgarian vs. Ukrainian and Czech or simply because Ukrainian and Czech dictionaries are better at listing the vocative for all nouns. User:Atitarev might be able to comment more. Benwing2 (talk) 00:01, 5 November 2023 (UTC)

@Benwing2: Apparently, vocatives are much less common in Bulgarian than most Slavic languages, except for Macedonian or Russian. Bulgarian dictionaries use much less vocatives, so it's OK to reduce and make optional where appropriate. Anatoli T. ^{(обсудить}/^вклад) 02:25, 5 November 2023 (UTC)

@Benwing2 one of the defining characteristics of Bulgarian, which sets it apart from all other Slavic languages except for Macedonian, is the transition from being synthetic to being analytic. That transition was gradual, and by the modern Bulgarian period (19th c - today) there were only relics of the nominal case system left. The transition continues to the present day, where some of the relics that were still around in 19th century prose (e.g. dative and genitive forms of personal names) have become obsolete.

This is not too dissimilar from the situation with English possessives - one could say that English has nominative and genitive for nouns, but AFAIK some modern analyses simply refer to possessive/genitive forms, just like some modern Bulgarian grammars refer to "vocative forms". That's why it's also likely misleading to try to compare Bulgarian vocative forms with vocatives in other Slavic languages, whose nominal case systems are alive and well and not restricted to narrow subsets of the vocabulary.

RE: marking vocative adjective forms as e.g. "archaic", for the sake of accounting for older texts - personally I don't see that as a huge benefit to the user. As I mentioned, the further back in time you go, the more there are relics of older case forms, beyond just the vocative, especially for given names (e.g. Иван (Ivan) → accusative Ивана (Ivana), → dative Ивану (Ivanu)). A few other case relics survive in fixed expressions, like своего рода (svoego roda) and на полза роду (na polza rodu). These are things that I myself had to ask about as a schoolchild because they struck me as foreign and strange. I'd assume that someone deep enough in Bulgarian that they are reading historical fiction would just ask similar clarifying questions. I'd rather we have a clean, trim representation of the modern language in inflection tables, and only introduce case variants when they are still in common circulation amongst speakers. The worst that can happen is that we forget to do that here and there, but I don't see how that detracts in a meaningful way from the quality of the information we provide in those tables. Chernorizets (talk) 01:15, 5 November 2023 (UTC)

Also, what's true even in the cases where we'd want to include vocative adjective forms is that the forms ending in -ий are definitely archaic. The forms with -и are what people would use today. Chernorizets (talk) 01:22, 5 November 2023 (UTC)

@Benwing2 @Kiril kovachev sorry I dropped the ball on this. To close this out, I'm requesting that:

• we suppress vocative masculine forms for Bulgarian adjectives by default, while keeping the -voc option around for compatibility with all adjective entries which use it today
• we introduce an explicit option to show vocative forms, which would be used sparingly. It can either be +voc or the one used in {{bg-ndecl}}.
• we mark masc. vocatives in -ий as archaic, since even in the rare cases when vocative adjectival forms are used today, they don't use this ending.

This is one of the modules I don't have permissions to edit, so I'd be looking to one of you to make the changes. I can help identify what in the code needs to change to what and where, if needed.

Thanks,

Chernorizets (talk) 05:31, 23 November 2023 (UTC)

@Chernorizets You should be able to edit these modules. You have autopatroller permissions and I set the permissions of all Bulgarian modules to be autopatroller and up, except for Module:bg-pronunciation/testcases which is set to autoconfirmed, since that is a less visible module. Benwing2 (talk) 07:33, 23 November 2023 (UTC)

@Chernorizets Sorry, I thought this was already done, so I didn't realize we still have to change things. Plan sounds good still ^^ Are you indeed able to edit the module after all? Kiril kovachev (talk・contribs) 09:24, 23 November 2023 (UTC)

@Kiril kovachev @Benwing2 I have now made the following changes:

• masculine singular vocative forms for adjectives are suppressed by default. They can be explicitly toggled on with the new option +voc. The existing option -voc is now a no-op, and we could write a bot job to remove it from existing adjective entries, before we can safely remove it from the code.
• vocative forms in -ий are marked as "Archaic" via a footnote.

I have the following action items:

• add +voc to the handful of adjectives that still regularly have vocative forms, like мил (mil) and драг (drag).
• update the {{bg-adecl}} documentation to reflect the new option.

I'm not sure what we ought to do with already-created non-lemma forms like абдоминалний (abdominalnij), which is a nonsensical vocative of "abdominal". We only have 326 total adjective non-lemma forms, so either by myself or with help, I could go thru the list and mark the ones we don't want to keep for deletion.

Let me know if you have any comments or concerns.

Thanks,

Chernorizets (talk) 08:13, 10 December 2023 (UTC)

@Chernorizets All fine with me. IMO we should just delete the bad non-lemma forms; if you give me a list, I can delete them, or if they are identifiable by ending (e.g. -ий?) I can make a list for you to review. Removing the existing -voc should be easy by bot. Benwing2 (talk) 20:28, 10 December 2023 (UTC)

@Chernorizets This is the list of adjective forms in -ий:

абатский
абдоминалний
абисинский
абитуриентский
аборигенский

Benwing2 (talk) 20:30, 10 December 2023 (UTC)

@Benwing2 yep, these are the ones I could find as well. For all other adjective non-lemma entries, I manually removed the non-archaic vocative inflections, which coincide with the plural in -и. I only kept them in three cases where I could find modern sources using the vocative - малък (malǎk), храбър (hrabǎr) and августейши (avgustejši). Feel free to delete the 5 entries in the above list. Chernorizets (talk) 22:55, 10 December 2023 (UTC)

@Chernorizets: FWIW, here are all the current uses of -voc (34 terms) [also, I deleted the 5 bad vocatives above]:

• Page 309 виолетов: Found match for regex: {{bg-adecl|виоле́тов</-voc>}}
• Page 355 вълчи: Found match for regex: {{bg-adecl|въ́лчи<!(ч)/-voc>}}
• Page 401 готварски: Found match for regex: {{bg-adecl|готва́рски<!/-voc>}}
• Page 402 готин: Found match for regex: {{bg-adecl|го́тин</-voc>}}
• Page 405 градински: Found match for regex: {{bg-adecl|гради́нски<!/-voc>}}
• Page 410 графичен: Found match for regex: {{bg-adecl|графи́чен<*!/-voc>}}
• Page 410 графичен: Found match for regex: {{bg-adecl|гра́фичен<*!/-voc>}}
• Page 420 гръклянов: Found match for regex: {{bg-adecl|гръкля́нов<!/-voc>}}
• Page 448 дементен: Found match for regex: {{bg-adecl|деме́нтен<*/-voc>}}
• Page 482 друг: Found match for regex: {{bg-adecl|друг<!/-voc>}}
• Page 506 едновременен: Found match for regex: {{bg-adecl|едновре́менен<!*/-voc>}}
• Page 507 едногодишен: Found match for regex: {{bg-adecl|едногоди́шен<*!/-voc>}}
• Page 518 електричен: Found match for regex: {{bg-adecl|електри́чен<*!/-voc>}}
• Page 519 електрически: Found match for regex: {{bg-adecl|електри́чески<!/-voc>}}
• Page 527 епичен: Found match for regex: {{bg-adecl|епи́чен<*!/-voc>}}
• Page 532 етичен: Found match for regex: {{bg-adecl|ети́чен<*/-voc>}}
• Page 590 зимен: Found match for regex: {{bg-adecl|зи́мен<*!/-voc>}}
• Page 673 кварков: Found match for regex: {{bg-adecl|ква́рков<!/-voc>}}
• Page 674 керамичен: Found match for regex: {{bg-adecl|кера́мичен<!*/-voc>}}
• Page 739 кухненски: Found match for regex: {{bg-adecl|ку́хненски<!/-voc>}}
• Page 841 минал: Found match for regex: {{bg-adecl|ми́нал</-voc>}}
• Page 853 множествен: Found match for regex: {{bg-adecl|мно́жествен<!/-voc>}}
• Page 894 наг: Found match for regex: {{bg-adecl|наг<!/-voc>}}
• Page 980 неутронен: Found match for regex: {{bg-adecl|неутро́нен<*!/-voc>}}
• Page 1062 памучен: Found match for regex: {{bg-adecl|паму́чен<*/-voc>}}
• Page 1109 поливен: Found match for regex: {{bg-adecl|поли́вен<*!/-voc>}}
• Page 1151 преносен: Found match for regex: {{bg-adecl|прено́сен<*!/-voc>}}
• Page 1189 пълнозърнест: Found match for regex: {{bg-adecl|пълнозъ́рнест<!/-voc>}}
• Page 1284 свързан: Found match for regex: {{bg-adecl|свъ́рзан<!/-voc>}}
• Page 1316 сказуемен: Found match for regex: {{bg-adecl|сказу́емен<!*/-voc>}}
• Page 1354 смокинов: Found match for regex: {{bg-adecl|смоки́нов<!/-voc>}}
• Page 1429 същ: Found match for regex: {{bg-adecl|същ<!/-voc>}}
• Page 1453 топен: Found match for regex: {{bg-adecl|топе́н<!/-voc>}}
• Page 1571 чеснов: Found match for regex: {{bg-adecl|чесно́в<!/-voc>}}

Benwing2 (talk) 23:42, 10 December 2023 (UTC)

@Benwing2 thanks! I can probably fix them manually. Appreciate the quick work on deleting the 5 entries. Chernorizets (talk) 01:05, 11 December 2023 (UTC)

@Chernorizets I just fixed them semi-manually. Benwing2 (talk) 01:08, 11 December 2023 (UTC)

@Benwing2 I guess the option is used on several pronouns as well. I'm fixing that manually by watching what pops up on CAT:E. Chernorizets (talk) 04:39, 11 December 2023 (UTC)

@Chernorizets Cool, sounds good. Benwing2 (talk) 04:41, 11 December 2023 (UTC)