Module talk:ru-pron/Archive 2

Russian pronunciation

Moved from Module_talk:links#Korean_transliteration ... I don't know Russian well but do you think it'd be possible to make a pronunciation module for Russian? Using accent marks, and some extra tricks for irregularities. Wyang (talk) 06:26, 28 May 2014 (UTC)

... Re: a pronunciation module for Russian. Yes, please! It is predictable in 95-99% of cases, there are some variants and exceptions can use phonetic respelling, e.g. сегодня as сево́дня (sevódnja). I can teach you some Russian too, if interested. --Anatoli ^{(обсудить}/^вклад) 06:32, 28 May 2014 (UTC)

Great, thanks. Now that I have the confirmation... If no one wants to take the lead, I might do so, in which case I will have to bombard you with questions. :) Wyang (talk) 06:48, 28 May 2014 (UTC)

Well, that's expected. If I can help. One place to start is w:Russian phonology, also you can familariase yourself with the IPA for Russian. Things to pay attention to at first are perhaps: devoicing/voicing of consonants, palatalisation, vowel reduction (akanye, ikanye) in non-stressed positions. --Anatoli ^{(обсудить}/^вклад) 07:01, 28 May 2014 (UTC)

It helps that there are a lot of entries in Category:Russian terms with audio links. If you have a good ear for foreign languages, you might try those recordings. E.g. разгово́р (razgovór). --Anatoli ^{(обсудить}/^вклад) 07:06, 28 May 2014 (UTC)

['(.[ое])го'] = '%1во'

@Kc kennylau and @Wyang regarding this diff - we need this change but it has to be parameterised, e.g. "adj=y". It's applicable primarily to adjectives (and participles), numerals and pronouns in the genitive and accusative case (animate) for masculine and neuter, also some derivatives, like сегодня, сегодняшний but otherwise, the string "ого" and "его" can be read as expected, with [g], not [v], e.g. много, лего, Егор. Not that although it's true that "ого" and "его" as adjective ending with [v] is usually at the end of a word, there are some words where it appears not at the end, also the final "ого" and "его" is NOT always pronounced with [v] but also with [g]. --Anatoli ^{(обсудить}/^вклад) 11:21, 7 June 2014 (UTC)

Maybe we should document exceptional case where ого is still /ogə/. By the way, I added a dollar sign at the end of the expression. --kc_kennylau (talk) 11:33, 7 June 2014 (UTC)

Pronunciation [g] is expected (the rule) and [v] is a common exception but it's an exception, not a rule, so it has to be parameterised, IMO. Yes, I saw the dollar sign but "мно́го" is [ˈmnogə] and "ле́го" [ˈlʲegə] but "сего́дня" is [sʲɪˈvod⁽ʲ⁾nʲə] and "сего́дняшний" is [sʲɪˈvod⁽ʲ⁾nʲɪʂnʲɪj] (from сего́ дня́). --Anatoli ^{(обсудить}/^вклад) 11:43, 7 June 2014 (UTC)

Corrections

• карьере sounds like [kərʲˈjerʲɪ], not [kərʲˈjerʲə] (*карьеря).--Cinemantique (talk) 20:20, 3 July 2015 (UTC)
• Архангельск should be [ɐrˈxanɡʲɪlʲsk], not [ɐrˈxanɡɨlʲsk].--Cinemantique (talk) 17:01, 4 July 2015 (UTC)
• нецелесообразный should be [nʲɪt͡sɨlʲɪsɐɐbˈraznɨj], not [nʲɪt͡sɨlʲɪsɐːbˈraznɨj].--Cinemantique (talk) 22:07, 5 July 2015 (UTC)
• I think АНБ should be [ɐɛnˈbɛ], not [ɐɨnˈbɛ].--Cinemantique (talk) 09:41, 8 July 2015 (UTC)

For the final unstressed "е" (also in поле, здоровье) we need to find a more definite rule or establish patterns and ways to override the behaviour, e.g. you can use {{ru-IPA|карье́рʲɪ}} -> IPA^(key): [kɐˈrʲjerʲɪ] as one of the ways to pronounce the form. I've added just two cases for now - Архангельск and нецелесообразный to Module:ru-pron/testcases. We should get rid of long vowels in the Russian pronunciation and just have duplications. @Wyang, are you able to help, please? --Anatoli T. ^{(обсудить}/^вклад) 11:00, 8 July 2015 (UTC)

I think АНБ is fine. Spelling pronunciation is also acceptable, of course. You can use the usual templates IPA^(key): [ɐɛnˈbɛ] as an alternative pronunciation. --Anatoli T. ^{(обсудить}/^вклад) 11:05, 8 July 2015 (UTC)

This pronunciation (по́ле like ['polʲə] = по́ля) is obsolescent. Please read paragraph 5.7.2 of the book Russian Literary Pronunciation by Ruben Avanesov (he uses ъ for ə and ə for ɪ) and listen to Forvo's recordings.--Cinemantique (talk) 12:31, 8 July 2015 (UTC)

I see. I pronounce поле more like [ˈpolʲe]. Not sure if the module will be able to handle these rather complicated rules for the pronunciation of the final е. We could establish a parameter, choose a default pronunciation and use parameters when the pronunciation is different. --Anatoli T. ^{(обсудить}/^вклад) 13:19, 8 July 2015 (UTC)

I think so, too. But maybe it's only in my mind.--Cinemantique (talk) 13:37, 8 July 2015 (UTC)

• в ко́же should be ['fkoʐɨ]. Avanesov denotes this sound by the symbol ы^э, so maybe it is ɪ̈. It occurs after ʐ, ʂ, t͡s at the end of the words.--Cinemantique (talk) 16:41, 8 July 2015 (UTC)

I do notice your edits, just don't know what to do about them right now. Are you able to summarise or just write out in English the rules, as in your link? I'll contact Wyang and/or try to get some help with other Lua editors. --Anatoli T. ^{(обсудить}/^вклад) 01:25, 9 July 2015 (UTC)

4. should сск be converted to ск?

@Benwing2 Yes, but маркси́стский (marksístskij) could be both [mɐrkˈsʲist͡skʲɪj] and [mɐrkˈsʲisːkʲɪj]. --Anatoli T. ^{(обсудить}/^вклад) 05:55, 19 October 2015 (UTC)

OK, I will mark this as [s(t)skʲɪj], will that work?

Also, what was the outcome of discussions on н[ндт]ск? I think we decided to write [n(t)sk], right? But what about ннск? I forgot whether the gemination of н is preserved, i.e. should it write [nsk] everywhere, or [nːsk] after the stress? Benwing2 (talk) 06:40, 19 October 2015 (UTC)

No, the gemination of the н is not preserved there. The general rule is fricatives cannot be geminated when followed by a plosive (counting /n/ as a fricative). [mɐrkˈsʲisːkʲɪj] is an exception because it is a much newer gemination and the /t/ phoneme is still present. --WikiTiki89 15:08, 19 October 2015 (UTC)

where to put primary stresses?

@Atitarev, Cinemantique, Wikitiki89 I'm assuming that single-syllable words that aren't prepositions should be pronounced as stressed. I'm trying to implement that but running into issues with the primary and secondary stress markers. Some possibilities:

1. All single-syllable words that aren't prepositions get primary stress by default, and we put the primary stress mark by all of them.
2. Only put primary stress on words that either directly follow a preposition or particle or words explicitly marked with primary stress, leaving the remaining single-syllable words without a stress marker, even though their pronunciation is stressed.
3. Put primary stress as in #2, but put secondary stress on remaining single-syllable words pronounced as stressed.
4. Other possibilities?

What do you think? Benwing2 (talk) 06:08, 19 October 2015 (UTC)

#2 is OK but hyphenated words should get a stress mark as well. --Anatoli T. ^{(обсудить}/^вклад) 06:38, 19 October 2015 (UTC)

OK, there's a test case with ка-гэ-бє́ which is displayed with primary stress only on the third (stressed) word. Do you think the other hyphenated words in this expression should get a stress mark, and if so, primary or secondary? Benwing2 (talk) 06:42, 19 October 2015 (UTC)

I meant words like что-либо. If the secondary stress is marked, it should be displayed. The primary stress should be shown in any case. --Anatoli T. ^{(обсудить}/^вклад) 06:51, 19 October 2015 (UTC)

I'm still not quite sure what you mean. I fixed this test case so that the primary stress on что́ is displayed properly, but are you suggesting the что should have a primary stress even without the stress mark? If so, what's the algorithm that determines that что should get a primary stress but ка in ка-гэ-бє́ presumably should not? Or should ка also get a primary stress even without an accent mark on it? Benwing2 (talk) 08:03, 19 October 2015 (UTC)

No, I am not suggesting that. I am agreeing with #2.--Anatoli T. ^{(обсудить}/^вклад) 08:56, 19 October 2015 (UTC)

OK, I'll implement that. Benwing2 (talk) 09:17, 19 October 2015 (UTC)

It should be "words that aren't prepositions or other short particles". Also, что́-либо has no secondary stress, the -либо/-нибудь/-то suffixes are completely unstressed. --WikiTiki89 15:11, 19 October 2015 (UTC)

Right, that's what I meant, thanks for the correction. I'm not sure why что-либо is written with secondary stress in the testcases. Benwing2 (talk) 15:32, 19 October 2015 (UTC)

Suffixes -либо and -нибудь may have a secondary accent, as an alternative pronunciation. --Anatoli T. ^{(обсудить}/^вклад) 15:45, 19 October 2015 (UTC)

implemented ɐɐ sequences, do they apply across words?

@Atitarev, Cinemantique, Wikitiki89 I have implemented the generation of ɐɐ sequences from hiatuses involving ɐ and ǝ. One question ... should this apply across word boundaries? Benwing2 (talk) 01:49, 23 October 2015 (UTC)

Like Южная Америка? I don't think so. Let's not do it. --Anatoli T. ^{(обсудить}/^вклад) 02:06, 23 October 2015 (UTC)

@Atitarev Thanks. BTW can you give me an example of words with ндск and ннск sequences in them? Benwing2 (talk) 02:38, 23 October 2015 (UTC)

Never mind, I found англо-норманнский and исландский. Benwing2 (talk) 02:41, 23 October 2015 (UTC)

I think it does apply across word boundaries; I just tried pronouncing Южная Америка and the -jə definitely shifted to -jɐ. By the way, I was just reminded of a joke involving the homophony between голландские and голанские. --WikiTiki89 15:01, 23 October 2015 (UTC)

Optional (j) at beginning of ельча́нин or not?

@Atitarev, Wikitiki89, Cinemantique Is there an optional (j) at the beginning of ельча́нин or is it required? Thanks. Benwing2 (talk) 06:21, 23 October 2015 (UTC)

Optional, same with unstressed "я".-Anatoli T. ^{(обсудить}/^вклад) 06:47, 23 October 2015 (UTC)

Thanks, that's what I thought. Benwing2 (talk) 08:24, 23 October 2015 (UTC)

@Atitarev, Wikitiki89, Cinemantique Another question .... In домохозя́ин (domoxozjáin), the pronunciation currently appears as [ˌdoməxɐˈzʲæɪn]; should it be [ˌdoməxɐˈzʲæ(j)ɪn]? Benwing2 (talk) 08:24, 23 October 2015 (UTC)

No, that would sound like Ukrainian accent. Russian tends to lose "j" between vowels before cognates of "i", Ukrainian tends to acquire it. Cf. Ukrainian хазя́їн (xazjájin) [xɐˈzʲɑjin] = Russian хозя́ин (xozjáin) [xɐˈzʲæɪn]. "и" is never pronounced with "j", only after "ь", or "й". After hard prepositions ending in consonants it even becomes "ɨ", e.g. "в И́ндии" - [ˈv‿ɨn⁽ʲ⁾dʲɪɪ]. --Anatoli T. ^{(обсудить}/^вклад) 09:26, 23 October 2015 (UTC)

OK, thanks, just wanted to make sure. Benwing2 (talk) 09:28, 23 October 2015 (UTC)

The optional "j" happens with iotated letters "е" and "я" when unstressed in absolute initial positions and between vowels. --Anatoli T. ^{(обсудить}/^вклад) 09:30, 23 October 2015 (UTC)

I agree that post-vocalic и just always be preceded by an optional [(j)], but Anatoli was against it last time I brought it up. --WikiTiki89 15:04, 23 October 2015 (UTC)

Incidentally, I was skimming through a 19th century Russian grammar book and it mentioned a distinction between имя [ˈi-] and ими/их/им [ˈji-] (see Ukrainian ім'я and їх/їм), and implied that this was the norm among Russian speakers at the time. --WikiTiki89 15:15, 23 October 2015 (UTC)

Words with сск

@Atitarev, Cinemantique, Wikitiki89 What about words like дама́сский? Currently we reduce сск to ск. Should we keep сск as such? Then what about words like русский? Should we handle дамасский with gem=y or русский with gem=n? Benwing2 (talk) 20:35, 23 October 2015 (UTC)

Normally "сск" and "зск" are reduced to "ск", even in post-stress positions. Yes, we should make "gem=y" work for cases where [s] should be geminated but in most cases with "[сз]ск" it's not, so maybe leave the default ungeminated. --Anatoli T. ^{(обсудить}/^вклад) 21:31, 23 October 2015 (UTC)

@Cinemantique Would you agree to that? The module won't be able to tell why "русский" and "дамасский" are pronounced differently. Besides, the test case module doesn't take "gem" parameter, so maybe you should remove "дамасский". --Anatoli T. ^{(обсудить}/^вклад) 22:38, 23 October 2015 (UTC)

• Please read §13.4.3 (page 177) in Avanesov's book. He mentions only three ungeminated words - русский, белорусский, and французский. I think it's easy to remove one letter.--Cinemantique (talk) 11:47, 24 October 2015 (UTC)

Alright, we can make three words with gem=n.--Anatoli T. ^{(обсудить}/^вклад) 22:23, 24 October 2015 (UTC)

OK, I made this change. It was a lot more than three words because of various inflected forms, terms like русский язык and Французская Гвиана etc. but manageable. Benwing2 (talk) 22:01, 25 October 2015 (UTC)

Thanks for that. --Anatoli T. ^{(обсудить}/^вклад) 22:16, 25 October 2015 (UTC)

розеттский камень and words with double consonants before other consonants

@Atitarev, Cinemantique, Wikitiki89 Cinemantique says this word does not have geminate /t/ in it. Is there a general rule regarding geminates before other consonants, or should we handle this with a special rule, or with gem=n? Benwing2 (talk) 22:46, 24 October 2015 (UTC)

Generally, no gemination before other consonants. --Anatoli T. ^{(обсудить}/^вклад) 00:47, 25 October 2015 (UTC)

Purpose of pal=y?

@Atitarev, Wikitiki89, Cinemantique Can you explain the purpose of pal=y? Benwing2 (talk) 02:20, 25 October 2015 (UTC)

Must be for forcing palatalisation, what's the example? --Anatoli T. ^{(обсудить}/^вклад) 02:27, 25 October 2015 (UTC)

I don't know, I'm trying to understand the code concerning this. Benwing2 (talk) 02:41, 25 October 2015 (UTC)

Maybe (without having looked at the code) it has to do with the assimilation in consonant clusters? --WikiTiki89 03:07, 25 October 2015 (UTC)

That's what I thought but that doesn't seem to be the case. It seems like it forces palatalization of consonants before non-palatal vowels, including consonants like шцж that normally can't be palatalized. But I have no idea under what circumstances this would be required. Benwing2 (talk) 03:50, 25 October 2015 (UTC)

отсчитываться

[ɐt͡sˈɕːitɨvət͡sə] - @Atitarev, is this transcription correct?--Cinemantique (talk) 14:27, 25 October 2015 (UTC)

No, it should be [ɐtˈɕːitɨvət͡sə], or perhaps [ɐt͡ɕˈɕitɨvət͡sə]. --WikiTiki89 18:37, 25 October 2015 (UTC)

We should correct it somehow...--Cinemantique (talk) 19:09, 25 October 2015 (UTC)

The module won't be able to determine this. Use "(phon=)отщи́тываться" in {{ru-IPA}}, just like отсчёт uses "отщёт". --Anatoli T. ^{(обсудить}/^вклад) 19:28, 25 October 2015 (UTC)

['tsč'] = 'čɕ'?--Cinemantique (talk) 19:38, 25 October 2015 (UTC)

Yes, perhaps but I would use 'tɕ', if it won't create any other problems. --Anatoli T. ^{(обсудить}/^вклад) 19:43, 25 October 2015 (UTC)

If тсч is always pronounced тщ, this would be easy enough to fix... Benwing2 (talk) 21:24, 25 October 2015 (UTC)

The string seems to always happen in positions "prefix+stem", also with "дсч," so it should be [тд]сч. Examples: отсчёт, подсчёт. Perhaps, as Wiki suggested, we should assimilate [тд]сч to чщ but it can really go either way. чщ is for the fast speech. --Anatoli T. ^{(обсудить}/^вклад) 21:38, 25 October 2015 (UTC)

The reason I thought it would be hard it because I thought "[тд]с" would be analysed first, before seeing that there's a "ч" after that. --Anatoli T. ^{(обсудить}/^вклад) 22:18, 25 October 2015 (UTC)

подтрибунный

[pətrʲɪˈbunːɨj] - Can we fix it without gem?--Cinemantique (talk) 13:56, 26 October 2015 (UTC)

Do you mean this should be geminated? Benwing2 (talk) 18:54, 26 October 2015 (UTC)

If so is there a general rule concerning дт or дтр? Benwing2 (talk) 18:55, 26 October 2015 (UTC)

We can't fix it without "gem=y". It's unpredictable. Here, it's a morpheme boundary - prefix + stem. In Кронштадт and it's inflections, there is no gemination. --Anatoli T. ^{(обсудить}/^вклад) 19:30, 26 October 2015 (UTC)

'^podt'? We have many such words.--Cinemantique (talk) 20:55, 26 October 2015 (UTC)

It's still driven by the grammar, not letter combinations. If your approach is taken, a list of commonly geminated sequences should be made with prefixes под-, от-, над-, с-, ис-, etc. and the following consonants. It won't be possible to make it fully automatic. --Anatoli T. ^{(обсудить}/^вклад) 21:06, 26 October 2015 (UTC)

It is possible to do something like this. We already have a special case for '^o[dt]s', and we already have a list of unstressed prepositions and particles. Benwing2 (talk) 21:10, 26 October 2015 (UTC)

In fact there's already a hack that's supposed to handle exactly this case:

local geminate_pref = {
	--'^abː',
	'^adː', '^bezː', '^braomː', '^vː', '^voszː', '^izː', '^inː', '^kontrː', '^nadː', '^niszː',
	'^o[cdmtč]ː', '^podː', '^predː', '^paszː', '^pozː', '^sː', '^sverxː', '^subː', '^tröxː', '^čeresː', '^četyröxː', '^črezː',
}

Not sure why it doesn't work. Benwing2 (talk) 21:13, 26 October 2015 (UTC)

@Atitarev, Cinemantique, Wikitiki89 Can you collectively give me a fairly complete list of verbal/nominal prefixes that end with a consonant? I want to clean up the geminate_pref handling. The problem with the above list is that it doesn't handle things like '^potː', but to prevent it overgenerating to words like Поттер I will have it check the original word to make sure it has the expected под- prefix. However, I need an accurate list of such prefixes. The list above looks odd -- why '^niszː' for example? And what's '^braomː' for? And presumably '^isː' should be present as well and '^ra[zs]ː', but '^pozː' may be wrong. Help? Benwing2 (talk) 08:58, 28 October 2015 (UTC)

I assume '^pozː' is the combination of по- and с-. These are all the native Slavic prefixes that end in consonants:

• в- (v-)
• с- (s-)
• из- (iz-)/ис- (is-)
• об- (ob-)
• от- (ot-)
• без- (bez-)/бес- (bes-)
• над- (nad-)
• под- (pod-)
• раз- (raz-)/рас- (ras-)/роз- (roz-)/рос- (ros-)
• пред- (pred-)

Keep in mind that many of them can be preceded by another prefix, especially by по- (po-). --WikiTiki89 14:44, 28 October 2015 (UTC)

• воз- (voz-)/вос- (vos-) (воззвание, воссиять)
• меж- (mež-) (only two words: межжаберный, межжелудочковый)
• через- (čerez-)/черес- (čeres-)/чрес- (čres-) (череззерница, чересстрочный, чресседельник)

Also, they can be preceded by prefix не- (ne-): невосстановимость.--Cinemantique (talk) 17:14, 28 October 2015 (UTC)

• '^braomː' - I don't know.
• '^abː'. аббревиату́ра (abbreviatúra), аббревиа́ция (abbreviácija) - not geminated. абба́т (abbát), аббати́са (abbatísa), абба́тство (abbátstvo) - two variants, with or without gemination.
• '^adː'. адде́нда (addénda), адде́ндум (addéndum), аддикти́вный (addiktívnyj), адди́кция (addíkcija), аддисо́нов (addisónov), аддити́вный (additívnyj), адду́ктор (addúktor), адду́кция (addúkcija). My orthoepic dictionary haven't these words. I think they have no gemination.--Cinemantique (talk) 17:30, 28 October 2015 (UTC)
• '^inː'. Few words: иннаби́ (innabí) - don't know; иннерва́ция (innervácija) - with or without (+ иннерви́ровать (innervírovatʹ), иннерви́рованный (innervírovannyj)); иннова́ция (innovácija), инновацио́нный (innovaciónnyj) - geminated.--Cinemantique (talk) 17:51, 28 October 2015 (UTC)
• '^kontrː'. Words beginning with контрр... have syllable-boundary between two 'р', for example ко̀нтррефо́рма (kòntrrefórma) - [ˌkontr-rʲɪ-ˈfor-mə]. My dictionary says that.--Cinemantique (talk) 17:51, 28 October 2015 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ @Cinemantique Thanks! I have changes that will fix this issue and avoid the need to specify gem=n or gem=y in a number of words (e.g. суббота); just waiting currently for a tracking category to fill up with all the changed words to make sure nothing bad will happen. Benwing2 (talk) 22:47, 28 October 2015 (UTC)

I can think of two more prefixes, that would cause a gemination.

• низ-/нис-
• супер-

The word расска́з (rasskáz) and derivatives is an exception, no gemination, although it has the prefix рас-

Benwing2, Not related but I think the use of symbols for various sounds in the module should be fixed (I saw you comments too). --Anatoli T. ^{(обсудить}/^вклад) 23:25, 28 October 2015 (UTC)

Also, расспроси́ть (rassprosítʹ) and derivatives, расстега́й (rasstegáj) have no gemination. Words with two variants: расста́ться (rasstátʹsja)+, расстегну́ть (rasstegnútʹ)+, расстоя́ние (rasstojánije), расстра́ивать (rasstráivatʹ)+, расстреля́ть (rasstreljátʹ)+, расстро́ить (rasstróitʹ)+.--Cinemantique (talk) 02:22, 29 October 2015 (UTC)

I've pushed the code. I edited the word рассказ to add gem=n; does this also apply to the verbs рассказать and рассказывать and related words like рассказец? Benwing2 (talk) 04:58, 29 October 2015 (UTC)

Yes, as I said "and derivatives" (they are derivatives of рассказ). --Anatoli T. ^{(обсудить}/^вклад) 05:16, 29 October 2015 (UTC)

OK, i fixed those words. Can you check the various other words beginning with расс- (e.g. рассол)? There are many: рассадник, рассада, рассвет, рассердить, рассеянный and related words, рассеиваться, рассудок, рассуждать, рассудительный, etc. Benwing2 (talk) 05:51, 29 October 2015 (UTC)

• @Benwing2, please add обезз- and обесс- (more than 50 words).--Cinemantique (talk) 06:54, 29 October 2015 (UTC)

Done. Benwing2 (talk) 07:05, 29 October 2015 (UTC)

Thanks!--Cinemantique (talk) 07:11, 29 October 2015 (UTC)

@Wikitiki89 I didn't add рос- or роз-; the only words I see with рос- are variants of Россия and I assume those don't have gemination. Benwing2 (talk) 07:07, 29 October 2015 (UTC)

They are just stressed variants of раз- (raz-)/рас- (ras-). All the words we and ru.wikt currently have that have this prefix with a geminate are ро́ссыпь (róssypʹ), ро́сстань (rósstanʹ), ро́ссыпный (róssypnyj), and ро́ссказни (rósskazni). --WikiTiki89 14:45, 29 October 2015 (UTC)

• Few words with two prefixes (via Ruwiki):
  • безнаддувный
  • безрассветный, безрассудно, безрассудность, безрассудный, безрассудство
  • врассыпку, врассыпную
  • всеподданнейше, всеподданнейший
  • гиперссылка, гиперссылочный
  • дорассветный
  • зассать
  • квазирасслоение, квазирасстояние, квазисуперрешение
  • макроисследование
  • микроисследование
  • наподдававший, наподдавать, наподдавший, наподдать
  • нарассказать, нарасследовать
  • обоссанный, обоссать, обоссаться
  • перерасследовавший
  • перессоривший, перессорить, перессориться
  • предрассветный, предрассудок, предрассудочек, предрассудочность, предрассудочный, предрассуждение
  • суперподделка

--Cinemantique (talk) 07:11, 29 October 2015 (UTC)

OK, I added various double prefixes with ra[sz] but took out checking for double prefixes with po- and nepo- because these words seem few and far between. Benwing2 (talk) 07:43, 29 October 2015 (UTC)

Optional palatalization

@Cinemantique, Atitarev When exactly does optional palatalization occur before /j/? Always? Benwing2 (talk) 17:27, 29 October 2015 (UTC)

• I think it's in [бвмп]ь[еёиюя]. All words having it.--Cinemantique (talk) 18:21, 29 October 2015 (UTC)
• I'm trying to analyse (yes = optional palatalization, no = only palatal)
  • бье: no (бьеннале, безрыбье)
  • бьё: yes (жеребьёвка, бьёт)
  • бьи: yes (воробьиный)
  • бью: usually no (бьюик, дистрибьютор), yes in бью, бьющий.
  • бья: воробья - yes, исподлобья - no.--Cinemantique (talk) 18:52, 29 October 2015 (UTC)
• I think we're getting bogged down in details out of scope for a dictionary. --WikiTiki89 18:56, 29 October 2015 (UTC)

As always, if there's a consistent rule, I can implement it, otherwise we shouldn't bother. Do the books have anything to say about it? Benwing2 (talk) 19:08, 29 October 2015 (UTC)

I get this information from another dictionary. Is Wiktionary worse than it?--Cinemantique (talk) 19:12, 29 October 2015 (UTC)

There are many rules that are much more consistent and yet we don't bother implementing them. For example, the value of stressed /ɨ/ is different after a labial compared to after any other consonant. --WikiTiki89 19:15, 29 October 2015 (UTC)

Yes, we won't show everything but orthoepic dictionaries have info about palatalization.--Cinemantique (talk) 19:25, 29 October 2015 (UTC)

• I have to say I don't like parameters in this template because word combinations can have different cases. I prefer to change word a little, for example {{ru-IPA|он смо́трит исподло́бья на вороб(ь)я́}}.--Cinemantique (talk) 19:12, 29 October 2015 (UTC)

  I agree. --WikiTiki89 19:15, 29 October 2015 (UTC)

  We also need a way to specify that {{ru-IPA|ба́шням}}, {{ru-IPA|ба́шнями}}, {{ru-IPA|го́нят}}, {{ru-IPA|мо́рем}}, and {{ru-IPA|си́нем}} should have [-ʲə-] rather than [-ʲɪ-] in their second syllables. --WikiTiki89 19:39, 29 October 2015 (UTC)

The module seems to support using ə directly, along with pal=; or at least that's what I think. pal= isn't used anywhere so I don't really know its purpose. Alternatively we could use some special symbol.

There's also the more general issue of the pronunciation of final -е; Anatoli said that he and Cinemantique would specify the rules for how this should be pronounced. Benwing2 (talk) 20:44, 29 October 2015 (UTC)

I suppose an alternative would be to write ьə; that could be supported. Benwing2 (talk) 20:46, 29 October 2015 (UTC)

Or ə̈. Benwing2 (talk) 20:47, 29 October 2015 (UTC)

The thing is, -е at the end of a word is easy to override by replacing it with -я or -и as necessary (and since it can only be predicted grammatically, I don't think we should bother automating it). But the -я-/-е- in the examples I mentioned above cannot be overridden because -я- also generates [ʲɪ] by default. I don't like the idea of using ьə inside the template because it is not an actual Russian letter, perhaps we should use ьа (and йа after a vowel). --WikiTiki89 20:59, 29 October 2015 (UTC)

Do ьа or йа ever occur in actual Russian text? I know that ьо does. Benwing2 (talk) 00:43, 30 October 2015 (UTC)

I don't like using ьа or йа (I don't think they occur in Russian, or extremely rarely, maybe in some unusual borrowings). There should be another way to override the default behaviour like in Tiki's examples above (e.g. ба́шням) or a parameter (reduc=n) to suppress the vowel reduction or replace with the vowel that should be pronounced (си́нем actually should be [-ʲɪ-], not [-ʲə-]). We already use ӂ and ɣ symbols for [ʑ] and [ɣ].

As for the final -е, Cinemantique was the one who mentioned it several times. I'm not entirely sure if we can establish more or less reliable rules, at least for some combinations but I don't mind using -я or -и to change the default. If he starts writing out the rules from Avanesov, we can analyse, what's feasible. --Anatoli T. ^{(обсудить}/^вклад) 00:56, 30 October 2015 (UTC)

I pronounce си́нем with [-ʲə-] (in case it wasn't clear, I'm referring to the prepositional case of си́ний and other soft adjectives). --WikiTiki89 01:27, 30 October 2015 (UTC)

In a normal speech "си́нем" and "си́ним" (cases of синий) are pronounced identically by native Russians in Russia - [ˈsʲinʲɪm]. In a slow speech, or in order to highlight the differences, they say [ˈsʲinʲem] (the spellings often effect how people speak, so vowel reductions are not always observed or not all types of reductions). --Anatoli T. ^{(обсудить}/^вклад) 01:49, 30 October 2015 (UTC)

"Native Russians in Russia" don't all pronounce everything the same way. --WikiTiki89 03:22, 30 October 2015 (UTC)

I refer to typical, most common and standard, what sounds authentic, which I think I can judge in most cases. --Anatoli T. ^{(обсудить}/^вклад) 04:40, 30 October 2015 (UTC)

-е

Now the module transcribes '-е' (at the end of words) like [ə]. I think we should make it only after 'ц'. According to Avanesov (§5.8, page 102), 'е' after 'ж, ц, ш' in post-stressed syllable is pronounced like [ə], except prepositional case, dative feminine, and comparative. What we have at the end of words:

• -же: 13 words (же, даже, понеже, иже, также, боже, негоже, ложе, цветоложе, тоже, ничтоже, вчуже, дюже). Does 'даже' sound like 'дажа'? I'm not sure. Maybe 'ложе, цветоложе' only...
• -це: many words (солнце, сердце...). 'е' should be manually changed to 'и' or 'ы' in prepositional case.
• -ше: 0 words.

'по́ле' should be [ˈpolʲɪ] (compare with 'По́ля'), 'мо́ре' is [morʲɪ] (not like 'мо́ря'), etc. That's all.--Cinemantique (talk) 11:11, 30 October 2015 (UTC)

Are you saying you pronounce по́ле the same way in the nominative and prepositional cases? I pronounce по́ле (nom/acc) exactly the same as по́ля; and same goes for мо́ре. But I think people from St. Petersburg tend to pronounce по́ле (nom/acc) differently from both по́ля and по́ле (prep). --WikiTiki89 15:50, 30 October 2015 (UTC)

нт

ма́нтия gives [ˈman⁽ʲ⁾tʲɪjə]

антило́па gives [ɐn⁽ʲ⁾tʲɪˈlopə]

Why is palatalization optional? It shouldn't. How does the rule work?--Cinemantique (talk) 16:24, 2 November 2015 (UTC)

It comes from this code:

local cons_assim_palatal = {
	compulsory = ut.list_to_set({'stʲ', 'zdʲ', 'nč', 'nɕ'}),
	optional = ut.list_to_set({'slʲ', 'zlʲ', 'snʲ', 'znʲ', 'tnʲ', 'dnʲ',
		'nsʲ', 'nzʲ', 'ntʲ', 'ndʲ'})
}

I'm not sure where the list of compulsory and optional palatal assimilations comes from. Benwing2 (talk) 21:34, 2 November 2015 (UTC)

I have no objection if н[дт] sequence is removed from the optional list. --Anatoli T. ^{(обсудить}/^вклад) 00:46, 3 November 2015 (UTC)

Do you mean it should become a compulsory assimilation or not palatalized at all? Benwing2 (talk) 01:02, 3 November 2015 (UTC)

Compulsory. --Anatoli T. ^{(обсудить}/^вклад) 01:15, 3 November 2015 (UTC)

OK, I made it compulsory but that broke a bunch of testcases. Could you verify that they should be fixed and fix them? Thanks! Benwing2 (talk) 02:29, 3 November 2015 (UTC)

Thank you very much!--Cinemantique (talk) 07:34, 3 November 2015 (UTC)

расчертить

With this new case, I think we need an override parameter for "сч". It's mostly identical to "щ", including prefix+stem boundaries. --Anatoli T. ^{(обсудить}/^вклад) 09:37, 2 November 2015 (UTC)

• What about a hypothetical combination "исчислять счёт"? Would we use parameters twice?--Cinemantique (talk) 09:57, 2 November 2015 (UTC)

The best way is to use sch1, sch2, unless the separate reading is more common than "щ" (I don't think it is), otherwise, reverse the rule. "gem" could also use gem1, gem2, maybe gem3? --Anatoli T. ^{(обсудить}/^вклад) 10:59, 2 November 2015 (UTC)

I think if you rewrite the word to have -ɕч- in it, it will end up how you want it. This might not be an elegant solution, though. Benwing2 (talk) 11:21, 2 November 2015 (UTC)

It would be easy if I write -щч- (we don't have such words).--Cinemantique (talk) 11:46, 2 November 2015 (UTC)

It's closer but "щ" is always geminated. If we make -щч- to read as [-ɕt͡ɕ-], then it's fine with me. (|phon= shouldn't be used, "щч" is not a phonetic respelling.). Alternatively use the hard-coded IPA template {{IPA|lang=ru}}, which is always a choice. --Anatoli T. ^{(обсудить}/^вклад) 12:03, 2 November 2015 (UTC)

• We can add such a rule: ^(бе|и|ра)сч(not! ёт) = щч. I mean all the words with prefixes 'бес-', 'ис-', and 'рас-', except words having 'счёт' (list). 'исчезать' and derivatives have two variants - with [ɕː] or [ɕt͡ɕ]. We can add the first variant like this: ищеза́ть.--Cinemantique (talk) 14:25, 2 November 2015 (UTC)

@Cinemantique I've got a little confused with the explanation, sorry. @Benwing2 Does it make sense to you? --Anatoli T. ^{(обсудить}/^вклад) 00:47, 3 November 2015 (UTC)

I think Cinemantique wants щч to represent [-ɕt͡ɕ-] and also have a rule that determines when to convert сч to щч instead of щ by looking at certain prefix boundaries as he described above. Benwing2 (talk) 00:59, 3 November 2015 (UTC)

It's the rule I don't quite get. I admit I missed that "сч" is not always equal to "щ". We can also remove the existing assimilation rule, make [-ɕt͡ɕ-] the default pronunciation, replace most occurrences of "сч" with "щ" in the template calls (I can imagine it's not too hard), "phon=" could be used.

The sequence "щч' to represent [-ɕt͡ɕ-] is not an ideal solution either but could also work. E.g. in the sequence "-вств-", the first "в" is silent in many occasions (e.g. чувство) but far from all (e.g. девственность). In cases where it's silent I used |phon=-ств-. --Anatoli T. ^{(обсудить}/^вклад) 01:09, 3 November 2015 (UTC)

Well, I added some tracking to see the situations where сч, зч, шч, and жч occur. The tracking takes awhile to cycle through all the pages so the lists are so far incomplete; you'll have to wait awhile. Look here for сч and зч: [1] and here for жч and шч: [2] Benwing2 (talk) 01:25, 3 November 2015 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ @Cinemantique What do you think? Take a look at the tracking categories above. Benwing2 (talk) 08:03, 3 November 2015 (UTC)

Well, [-ɕt͡ɕ-] occurs only between prefix and root, generally prefixes бес(ч)-, ис(ч)-, and рас(ч)-. чересчур has [ɕt͡ɕ], too. Words having -счёт- is always pronounced with [ɕː].--Cinemantique (talk) 08:18, 3 November 2015 (UTC)

What about зч and жч? Are these always pronounced щ? Benwing2 (talk) 12:45, 3 November 2015 (UTC)

зч = [ɕː]

жч - few words

with [ɕː]: мужчина, перебежчик, перебежчица.

with [ʂt͡ɕ] after меж-: межчелюстной, межчеченский.

шч - few words: веснушчатый has two variants [ɕː]/[ɕt͡ɕ]; черешчатый, сиводушчатый too, probably.--Cinemantique (talk) 13:07, 3 November 2015 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ OK, I implemented the following rules, which should be a compromise based on the above:

• шч, зч, жч always become щ.
• сч becomes щ word-initially, in счёт, and in sequences like ссч; otherwise it becomes [ɕt͡ɕ].
• You can force [ɕt͡ɕ] by writing щч.

Benwing2 (talk) 02:15, 5 November 2015 (UTC)

BTW please check the tracking categories I listed above for all the words with сч, шч, зч, жч in them and verify the pronunciations (NOTE, it may take awhile for the new pronunciations to get propagated; you can speed things up by making a null save). Benwing2 (talk) 02:17, 5 November 2015 (UTC)

@Benwing2 Thanks for your efforts again! I've just made веснушчатый with two pronunciations. My reason for not liking "щч" is because it's not how it would be pronounced by Russians and it may get abused or misunderstood. Can I suggest a parameter "shch=y" instead, with the same effect as using "щч"? --Anatoli T. ^{(обсудить}/^вклад) 02:27, 5 November 2015 (UTC)

@Atitarev If you don't like "щч", then how about ɕч? This is what щч maps to in order to implement its pronunciation. I'd like to avoid another parameter as it's complicated to implement those parameters properly, esp. if you want sch1 and sch2 etc. to deal with the problem that Cinemantique mentioned, where you might have two occurrences of сч with different pronunciations. Benwing2 (talk) 02:34, 5 November 2015 (UTC)

"ɕч" is more palatable, although harder to input. ("щ" is geminated as [ɕː] in most cases.) --Anatoli T. ^{(обсудить}/^вклад) 02:41, 5 November 2015 (UTC)

@Benwing2 What about чересчур, which can go both ways? It defaults to [ɕˈt͡ɕ], which is more common but [ɕː] is also quite common. --Anatoli T. ^{(обсудить}/^вклад) 02:59, 5 November 2015 (UTC)

Not quite sure what the issue is ... you can rewrite with щ to force the [ɕː] pronunciation. Benwing2 (talk) 03:03, 5 November 2015 (UTC)

дн, тн

• @Benwing2, palatalization is optional only after [r] and [l]: куро́ртник, сболтну́ть сболтнёшь. In the other cases [d/t] must be palatal: па́мятник, спу́тник, отнёс.--Cinemantique (talk) 00:35, 6 November 2015 (UTC)

Is this from Avanesov? I reckon, родня, дни can be pronounced without palatalisation as well, also with the examples you gave - just my opinion. (сболтну́ть is a bad example, no palatalisation here on [n]). --Anatoli T. ^{(обсудить}/^вклад) 00:41, 6 November 2015 (UTC)

I would say that when the [t] or [d] is released nasally, as is most often the case, then Cinemantique's rule applies. When, as in careful speech, the [t] or [d] is released normally (i.e. orally), it would usually be unpalatalized. --WikiTiki89 01:23, 6 November 2015 (UTC)

It's from Avanesov's orthoepic dictionary (§93).--Cinemantique (talk) 01:57, 6 November 2015 (UTC)

Avanesov mentions отнёс, поднёс may be pronounced with unpalatalised [t] and [d] but the palatalised is preferred/more correct. Since palatalised is considered the standard pronunciation, I don't mind making this mandatory, thanks for the reference. --Anatoli T. ^{(обсудить}/^вклад) 02:13, 6 November 2015 (UTC)

@Atitarev, Cinemantique, Wikitiki89 This is now implemented. However, it also applies to words beginning with дн, such as днище. Is this correct? Benwing2 (talk) 04:45, 8 November 2015 (UTC)

Yes, it is standard.--Cinemantique (talk) 11:05, 8 November 2015 (UTC)

нн and optional gemination (moved from Talk:девственная плева)

@Benwing2 The pronunciation of де́вственная плева́ (dévstvennaja plevá) should be [ˈdʲefstvʲɪn(ː)əjə plʲɪˈva] - with optional gemination in positions further from the stress, like at девственный. --Anatoli T. ^{(обсудить}/^вклад) 05:07, 6 November 2015 (UTC)

The code seems to check specially for нный and makes that optional. Should it also check for нная and нное or should it make ALL occurrences of нн other than directly following the stress optionally geminated? Benwing2 (talk) 05:33, 6 November 2015 (UTC)

I hope it would be correct to make the gemination optional for ALL occurrences of "нн" (surrounded by vowels) other than directly following the stress. @Cinemantique, Wikitiki89, any objections? --Anatoli T. ^{(обсудить}/^вклад) 05:58, 6 November 2015 (UTC)

Yes, we have two variants if -нн- is in suffix (промы́шленность, безжи́зненный, проду́манный). In other positions there could be the one and only variant (анне́ксия, аннули́ровать) or two variants (аннигиля́ция). Maybe your rule should work in post-stress (but not directly) syllables. I'm not sure now. I'd like to write {{ru-IPA|ан(н)игиля́ция}} or something like this.--Cinemantique (talk) 07:31, 6 November 2015 (UTC)

You must mean pre-tonal rules. For post-tonal we can use the usual "gem=y", "gem=n". --Anatoli T. ^{(обсудить}/^вклад) 10:06, 6 November 2015 (UTC)

I think Cinemantique says that the optional gemination of нн should occur in all syllables after the stress except directly after the stress, and that the gemination should be preserved pre-stress, but it should be possible to write н(н) to indicate optional gemination. I think this is doable. Benwing2 (talk) 11:32, 6 November 2015 (UTC)

@Atitarev, Cinemantique, Wikitiki89 Both of these things have been implemented. Benwing2 (talk) 07:21, 10 November 2015 (UTC)

optional gemination using |gem=opt

You can now use |gem=opt to force optional gemination. It doesn't affect щ or ӂ, which are always displayed with gemination (except in the sequence щч), but otherwise should cause optional gemination, displayed as (ː). I've used it in сеттер and глиссер, and I'll use it in группа providing I get confirmation that gemination is indeed optional, and I imagine there are many more such places. Benwing2 (talk) 14:24, 9 November 2015 (UTC)

• @Benwing2, thank you so much! Can we do the same thing for the optional palatalization? I mean скамья, пьянка, etc.--Cinemantique (talk) 19:33, 11 November 2015 (UTC)

вѐтслу́жба

@Cinemantique Can you explain the general rule behind this? Does it apply to all тс sequences? Benwing2 (talk) 15:20, 9 November 2015 (UTC)

In general тс ≠ ц (the -т(ь)ся ending is an exception). There is usually a syllable break between the т and the с, and in careful phonetic analysis the /t/ non-phonemically affricates to [t͡s]. --WikiTiki89 15:45, 9 November 2015 (UTC)

Yes,

[t] before [s] → [t͡s]: вѐтслу́жба, отсыла́ть

[tʲ] before [sʲ] → [tʲ͡sʲ]: пять селёдок, отсюда

[d] before [z] → [d͡z]: надзор

[dʲ] before [zʲ] → [dʲ͡zʲ]: пядь земли

[t] or [d] before [ʂ] → [t͡ʂ]: отшибить, младшая

[t] or [d] before [ɕ] → [t͡ɕ]: тщетно, дщерь

[t] or [d] before [ʐ] → [d͡ʐ]: таджик, отжить.--Cinemantique (talk) 16:32, 9 November 2015 (UTC)

Thanks! Can you create test cases for these situations? Benwing2 (talk) 16:51, 9 November 2015 (UTC)

OK, later.--Cinemantique (talk) 21:48, 9 November 2015 (UTC)

I disagree with the fourth (after edit: fourth and fifth) rule, they should produce [t͡ʂ] and [t͡ɕ], respectively. More precisely, the first part of the affricate should be [ʈ] (and for the fifth (after edit: sixth) rule [ɖ]). Also, the first rule should produce [tʲ͡sʲ] (and the third's Unicode representation should actually be [dʲ͡zʲ], i.e. the combining inverted breve comes after the superscript j). Also the palatalization of the original /t/ or /d/ is irrelevant: in отсюда, /t/ → [tʲ͡sʲ]. --WikiTiki89 22:12, 9 November 2015 (UTC)

• @Atitarev, I've added many tests.--Cinemantique (talk) 19:56, 10 November 2015 (UTC)

Thanks. I agree with these new cases. (the assimilation is light and is not always heard, it disappears, if there is even a slight pause, e.g. between morphemes or words)--Anatoli T. ^{(обсудить}/^вклад) 22:07, 10 November 2015 (UTC)

Help understanding existing тс rules (and related)

@Atitarev, Cinemantique, Wikitiki89 Can you help me understand some existing parts of the pronunciation module that involve тс? I'm trying to apply the above assimilations but there is a bunch of complicated stuff I don't completely understand and I suspect isn't quite right. It looks like this (these operate in the order given):

1. {'n[dt]sk', 'n(t)sk'},
   • This affects нтск and ндск. I think this is OK.
2. {'s[dt]sk', 'sck'},
   • This affects стск. I think this is OK.

3.     -- the following is for отсчи́тываться and подсчёт and is
       -- ordered before changes that convert ts to c 
       {'[cdt][sš]č', 'tšč'},

   • This affects the sequences [дтц]сч, [дтц]щ, and converts them to тщ. I presume this should instead convert to чщ?
     • I think you're right. BTW I cannot find words with цсч or цщ.--Cinemantique (talk) 09:47, 13 November 2015 (UTC)
4. {'^o[dt]s', 'ocs'},
   • This affects words beginning specifically with unstressed отс- or одс- and converts them to оцс. Is this bogus and should be removed (subsumed by more general тс code)?
     • Yes, it's a general rule. And I don't see words beginning with одс-.--Cinemantique (talk) 09:50, 13 November 2015 (UTC)
       • Can it be subsumed into a more general rule that affects тс? Benwing2 (talk) 10:07, 13 November 2015 (UTC)
         • See in ветслужба above.--Cinemantique (talk) 10:46, 13 November 2015 (UTC)
           • So I gather that all тс becomes цс? That would mean that this отс rule is unnecessary, and the complex rule below (#7) can be greatly simplified. Benwing2 (talk) 11:11, 13 November 2015 (UTC)
5. {'́tʹ?sja', '́cca'},
6. {'([^́])tʹ?sja', '%1ca'},
   • These two convert -тся- to -цца- if directly after a stressed vowel, and to -ца- otherwise. Note that this also applies in the middle of a word, e.g. a made-up word питсял, which I presume is a bug, and it should apply only at the end of a word; but other than this, is this correct?
     • That's right, however we have no words with тся in the middle.--Cinemantique (talk) 10:46, 13 November 2015 (UTC)

7.        {'[dt](ʹ?)s(.?)(.?)', function(a, b, c)
           if not (b == 'j' and c == 'a') then
               if rsub(b, vowels_c, '') ~= '' or b == 'j' and rsub(c, vowels_c, '') ~= '' then
                   return 'c' .. a .. b .. c
               else
                   return 'c' .. a .. 's' .. b .. c
               end
           end end },

   • This code is a bit complex and is buggy, but its basic effect is that [дт]с before a vowel becomes цс, and otherwise [дт]с becomes ц. This is why вѐтслу́жба is currently rendered as if written вѐцлу́жба. Should this always convert [дт]с to цс?
8. {'([^rn])[dt]c', '%1cc'},
9. {'[td]č', 'čč'},
10. {'r[dt]c', 'rc'},
11. {'r[dt]č', 'rč'},
12. {'n[dt]c', 'nc'},
    • These five are buggy as written, but the intent appears as follows: [тд]ц becomes цц, and [тд]ч becomes чч, except that [рн][тд][цч] becomes [рн][цч], i.e. the middle [тд] drops out. This seems OK, is this right?

Thanks for your help. Benwing2 (talk) 08:11, 13 November 2015 (UTC)

• Is[ t͡sː] a right symbol for отцепить, собираться? Avanesov (p. 205) and Russian Grammar say there's 'долгий затвор' (articulation delay before plosion) and they use a symbol [^тц].--Cinemantique (talk) 10:46, 13 November 2015 (UTC)
  • I'm not sure how to write that superscript т in IPA notation. Benwing2 (talk) 11:11, 13 November 2015 (UTC)

Цюрих

Does this word actually have palatal [tʲ͡sʲ] in it? I thought ц was always hard. Benwing2 (talk) 11:12, 13 November 2015 (UTC)

There is a small number of exceptions, only with ю, я and ё - Цюрих, хуацяо and фрицёнок (dimin. of фриц). It would be palatalised with ь as well but it only occurs when imitating Ukrainian or Belarusian (перець, пяць). Before и and е ц is not palatalised. --Anatoli T. ^{(обсудить}/^вклад) 12:40, 13 November 2015 (UTC)

I think the pronunciation [ˈt͡surʲɪx] exists as well. --WikiTiki89 14:51, 13 November 2015 (UTC)

Please review affricate assimilations

@Atitarev, Cinemantique, Wikitiki89 I implemented the affricate assimilations described above. However, it caused some increased test failures and all sorts of changed pronunciations, and I'm not sure they're all correct. The new code is currently only in my private module; please check out the corresponding unit tests in Module:User:Benwing2/ru-pron/testcases. I added a bunch of cases I'm not sure how should be pronounced; please check all the failures and fix the ones that are incorrect, and for the others, let me know how to restrict the affricate assimilations appropriately. Thanks! Benwing2 (talk) 03:00, 14 November 2015 (UTC)

I've changed some. With "дж" after consonants it seems hard to pronounce it the same way as the first ""дж"" in Джордж, you have to make a pause. I think [ʐ] becomes optional in such cases - [ˌlos ˈand͡ʐ(ʐ)ɨlʲɪs], [ɐlʲd͡ʐ(ʐ)ɐˈzʲirə], [nd͡ʐ(ʐ)ɐˈmʲenə]. --Anatoli T. ^{(обсудить}/^вклад) 03:32, 14 November 2015 (UTC)

Some failed cases like Пи́тсбург, всле́дствие, Петрозаво́дск should just have [d͡z] and [t͡s]. --Anatoli T. ^{(обсудить}/^вклад) 03:45, 14 November 2015 (UTC)

Hmm, that suggests that [тд]с should be just [d͡z] and [t͡s] before stop consonants (e.g. бтк), probably fricatives as well (вфх). Does the same thing apply to [дт]ш and [дт]ж (are there any instances of these sounds occurring before other consonants)? Benwing2 (talk) 09:21, 14 November 2015 (UTC)

I think all fricatives are a short version [d͡z], [t͡s], [d͡ʐ], [t͡ʂ] front of consonants (even when spelled with clusters) and long version дз - [d͡zz], тс, дс, цс - [t͡ss], дж - [d͡ʐʐ], тш - [t͡ʂʂ] in front of vowels, no? --Anatoli T. ^{(обсудить}/^вклад) 09:36, 14 November 2015 (UTC)

• I've found another rule here (§165 in Knyazev/Pozharitskaya's textbook): тс and дс is just [t͡s] if the next sound is consonant.--Cinemantique (talk) 09:40, 14 November 2015 (UTC)
  • Then what about вѐтслу́жба? This was the first word where you claimed the "long version" occurred, and here тс is before a consonant. Is there an exception for л and maybe р?
  • BTW Anatoli, if we have the short versions before consonants and the long versions before vowels, what happens end of word e.g. джордж? You indicated it as the long version, is this always true? Benwing2 (talk) 09:50, 14 November 2015 (UTC)

Good question. I think ветслужба is a case of morpheme boundaries ("vet(erinary) service") but making an exception for л and р may not be a bad idea, e.g. in отсро́чить (otsróčitʹ) there will also be [[t͡ss] but "от-" is a prefix. --Anatoli T. ^{(обсудить}/^вклад) 10:01, 14 November 2015 (UTC)

As for Джордж, I think [d͡ʐʐort͡ʂ] and [d͡ʐʐort͡ʂʂ] are both OK. --Anatoli T. ^{(обсудить}/^вклад) 10:05, 14 November 2015 (UTC)

How about words ending in дж or тш after a vowel? Do they exist and if so how are they pronounced? Benwing2 (talk) 10:09, 14 November 2015 (UTC)

@Atitarev, Cinemantique, Wikitiki89 Also, why does джорджтаун go one way but мундштук the other way? Benwing2 (talk) 10:38, 14 November 2015 (UTC)

I have fixed "Джо́рджтаун", it's [ˈd͡ʐʐort͡ʂʂtəʊn]. --Anatoli T. ^{(обсудить}/^вклад) 11:19, 14 November 2015 (UTC)

@Atitarev, Cinemantique, Wikitiki89 BTW if you (any of you) could try and spell out explicitly all the rules for when the short and long forms should be used, I'd be very grateful. If it depends on morpheme boundaries, then try to automate this as much as possible, and to fix things further I'll add a character e.g. _ or : that can be used to indicate a morpheme boundary. Benwing2 (talk) 10:52, 14 November 2015 (UTC)

• Unfortunately, different authors give different information. @Atitarev, can you download this book and read §81 (page 210). Matusevich says (p. 213) that 'т' doesn't assimilate with 'с' between words (от села).--Cinemantique (talk) 10:54, 14 November 2015 (UTC)
  • Well then, do your best ... whatever rules you can state, I can implement. Benwing2 (talk) 10:59, 14 November 2015 (UTC)
• @Atitarev, Benwing2, OK, let's avoid it at the end of word and in front of consonant but use between words and parts of word. We can use slash sign: вет/служба.--Cinemantique (talk) 11:10, 15 November 2015 (UTC)

OK, I think you're saying that [тд]с, [тд]з should be t͡ss and d͡zz before vowels, and t͡s and d͡z elsewhere (before consonants and at the end of a word), and [тд]/с, [тд]/з with an explicitly written slash should be t͡ss and d͡zz, and so should [тд] с, [тд] з across a word boundary. Right? What about [тд]ш and [тд]ж? Same rules? If so, what about Джо́рджтаун, Джордж, мундштук? Benwing2 (talk) 07:20, 16 November 2015 (UTC)

I think, yes. Actually, мундштук has unpronounceable д, that is, it's pronounced like мунштук: [mʊnˈʂtuk].--Cinemantique (talk) 10:30, 16 November 2015 (UTC)

@Cinemantique What about Джо́рджтаун, Джордж, котте́дж? Anatoli thinks that d͡ʐʐ and t͡ʂʂ occur even before consonants or at the end of a word. Benwing2 (talk) 09:45, 17 November 2015 (UTC)

Well, okay. Let's do it everywhere.--Cinemantique (talk) 12:18, 17 November 2015 (UTC)

OK, sounds good. I've got the code set up in Module:User:Benwing2/ru-pron, I'll push it live soon. Benwing2 (talk) 12:24, 17 November 2015 (UTC)

Geminated [l]

Actually, [l] can be geminated but rarely (гулливый, галломан, вилла), usually as additional variant (аллитерация, балловый), very rarely at the end of a word (вилл because of geminated [l] in other forms).--Cinemantique (talk) 05:17, 15 November 2015 (UTC)

I can make лл become ungeminated even after a stressed syllable unless gem=y or gem=opt, if you think this is reasonable. Benwing2 (talk) 07:14, 16 November 2015 (UTC)

@Cinemantique What do you think? Benwing2 (talk) 10:17, 17 November 2015 (UTC)

OK, let's do it.--Cinemantique (talk) 10:35, 17 November 2015 (UTC)

@Cinemantique Can you insert some testcases for words with лл directly after the stress that aren't geminated? Benwing2 (talk) 11:06, 17 November 2015 (UTC)

проце́нтщик - [prɐˈt͡sɛnʲt͡ɕɕːɪk]

@Benwing2 Re: your question. If we establish that assimilation does happen, then yes, your test case is correct. The assimilation is not strong and even a tiny micropause will make it [prɐˈt͡sɛntɕːɪk]. --Anatoli T. ^{(обсудить}/^вклад) 01:59, 16 November 2015 (UTC)

брюзжа́ть

Yes, зж, жж; сч, зч (=щ) are always geminated, in any position. --Anatoli T. ^{(обсудить}/^вклад) 02:03, 16 November 2015 (UTC)

бв

What's the difference between e.g. любви and обвинительный? --Anatoli T. ^{(обсудить}/^вклад) 05:47, 16 November 2015 (UTC)

See Cinemantique's comments in Talk:помпезный. Benwing2 (talk) 06:07, 16 November 2015 (UTC)

Yes, Avanesov said that б in любви should be palatalized. It was in 1984. The modern edition of his orthoepic dictionary says that palatalization is optional.--Cinemantique (talk) 06:18, 16 November 2015 (UTC)

проце́нтщик, мундштук, is deletion of [дт] here automatic?

@Atitarev, Cinemantique, Wikitiki89 Is deletion of [дт] in sequences like нтщ, ндш (between н and [шщ]) automatic? Benwing2 (talk) 10:17, 17 November 2015 (UTC)

Yes, I've done.--Cinemantique (talk) 10:34, 17 November 2015 (UTC)

Thanks! I fixed it so it also handles н[дт]ш, but this causes a failure in ре́гентша. Can you fix the test case appropriately? Benwing2 (talk) 11:04, 17 November 2015 (UTC)

@Cinemantique So I take it deletion of [дт] in н[дт]ш isn't automatic, if it remains in регентша but is deleted in мундштук? Benwing2 (talk) 11:27, 17 November 2015 (UTC)

It's right for ландшафт and мундштук, but wrong for эндшпиль. I'm not sure about андшпуг, гандшпуг, гёрлфрендша, ландштурм. т in -нтш- isn't silent (комендантша, коммерсантша, музыкантша, регентша).--Cinemantique (talk) 11:33, 17 November 2015 (UTC)

OK, that means we need to use phon= when deletion occurs. Is it automatic though in нтщ and ндщ? Benwing2 (talk) 11:35, 17 November 2015 (UTC)

All such words I've found (I say 'yes' if the orthoepic dictionary says т/д is silent): алиментщик (yes), алиментщица (probably yes), бриллиантщик (I think, yes, like алиментщик), брильянтщик (like previous), диссидентщина (like the next?), интеллигентщина (yes), комплиментщик (yes), комплиментщица (probably yes), монументщик (like алиментщик?), позументщик (yes), процентщик (yes), процентщица (probably yes), фундаментщик (like алиментщик?); not sure: хай-эндщик, шаландщик.--Cinemantique (talk) 11:59, 17 November 2015 (UTC)

OK, it sounds like in нтщ, the т is consistently silent. Benwing2 (talk) 12:06, 18 November 2015 (UTC)

Should we have a special case for отс-?

@Atitarev, Cinemantique, Wikitiki89 Lots of words begin отс- + consonant, esp. отст- and отср-. These will all be treated as having the "short" variant (same as ц). But maybe we should have a special rule that treats отс as оцс, since presumably there's a morpheme boundary after от-? Benwing2 (talk) 12:49, 17 November 2015 (UTC)

I suspect we may need a special rule for подс- too. Benwing2 (talk) 12:50, 17 November 2015 (UTC)

Yes, there is always a morpheme boundary after a prefix. --WikiTiki89 15:33, 17 November 2015 (UTC)

Yes, all words with ^отс[вклмнрт].--Cinemantique (talk) 16:00, 17 November 2015 (UTC)

I just did it with all ^отс, ^отз, ^подс, ^подз. Benwing2 (talk) 11:15, 18 November 2015 (UTC)

тс/дз code pushed; please review new testcases, esp. those involving sequences like цз, чж (esp. various Chinese names and words)

@Atitarev, Cinemantique, Wikitiki89 Please review the various testcases I indicated in the commit messages to Module:ru-pron/testcases; also check the remaining failing test cases and see which if any need to be corrected; probably дух бодр, плоть же не́мощна, де́сять за́поведей, Дза̀уджика́у. Thanks! Benwing2 (talk) 11:14, 18 November 2015 (UTC)

Checked. The actual "ˈdʲesʲɪdʲ͡zʲ ˈzapəvʲɪdʲɪj" looks better. Not sure if it makes sense to check the suffix -ин in isolation. @Cinemantique, could you check "собра́ние" in your references, please? I need some confirmation that [j] is optional and the final [ə]. --Anatoli T. ^{(обсудить}/^вклад) 11:29, 18 November 2015 (UTC)

@Atitarev, Cinemantique, Wikitiki89 What's the rule regarding optional (j) in this case? Is it specifically in final -ɪjə? Also -ijə? Can it occur with other vowels preceding the j, or other than word-final? Benwing2 (talk) 11:43, 18 November 2015 (UTC)

It occurred to me when I saw it in ru-wikt and pronouncing "-ие" with or without [j] both sound natural to me, unlike "-ия". I don't know the rule. --Anatoli T. ^{(обсудить}/^вклад) 11:48, 18 November 2015 (UTC)

@Atitarev, see §5.7.2 б). Avanesov uses 'ə' for [ɪ]; 'ɪ̯' means [ɪ̯], but we don't use this sign, only [j].--Cinemantique (talk) 14:58, 18 November 2015 (UTC)

@Cinemantique, Benwing2 If I read it correctly, зна́ние (znánije) is [ˈznanʲɪjɪ] and собра́ние (sobránije) should then be [sɐbˈranʲɪjɪ]? It makes "собра́нии" (prep. sg) sound almost the same as "собра́ние" (nom/acc sg). Although, it's not exactly "собра́ния" (gen sg, nom/acc pl), which is definitely [sɐbˈranʲɪjə], making the nominative and accusative sg forms [sɐbˈranʲɪjɪ]]. It doesn't sound quite right. Do you have another source? Did I read Avanesov's notation correctly? Please confirm. @Benwing, my test case needs to be reviewed in any case. I can't confirm the presence of an optional [j]. --Anatoli T. ^{(обсудить}/^вклад) 04:26, 20 November 2015 (UTC)

Anatoli, can you fix up the test cases in Module:ru-pron/testcases that are wrong? Benwing2 (talk) 11:45, 18 November 2015 (UTC)

Only one spotted. Cinemantique suggested that geminated [lː] or [lʲː] is rare. I don't if we can default them to ungeminated and maybe use "gem=y" for words like "вилла". --Anatoli T. ^{(обсудить}/^вклад) 11:53, 18 November 2015 (UTC)

Indeed. I have implemented this in my sandbox module. What about Гуйчжо́у and Чжэцзя́н? Is assimilation of the чж cluster to дж correct? Also, surely Дза̀уджика́у should be pronounced ˌd͡zzaʊd͡ʐʐɨˈkaʊ, as it's actually rendered, not the "expected" ˌd͡zzaʊdʐɨˈkaʊ? Benwing2 (talk) 12:05, 18 November 2015 (UTC)

Oh shit, you know Russian better than me now, you've put me to shame! :) You really have an eye for details and very thorough! I've tried again. --Anatoli T. ^{(обсудить}/^вклад) 12:13, 18 November 2015 (UTC)

Syllable boundaries with C‿C, V‿C, C‿V

@Atitarev, Cinemantique, Wikitiki89 I did a bunch of new work on my sandbox module Module:User:Benwing2/ru-pron. Please check out the testcases in Module:User:Benwing2/ru-pron/testcases. One fix is in cases like [pɐt͡s‿ˈstuləm], which the module in Module:ru-pron renders incorrectly as [pɐˈt͡s‿stuləm]. In the process, however, what used to be [nʲɪ‿ˈfsʲɵ to ˈzolətə] becomes [nʲɪ‿fˈsʲɵ to ˈzolətə] -- it's almost ignoring the ‿ connector. Is this correct? What about от а́бы, which both the new and old modules render as [ɐˈt‿abɨ], which seems correct to me. If you think that in cases of ‿C or ‿CC the stress should go before the C and after the ‿ connector, but in cases of C‿V the stress should go before the C (and before the connector), I can implement that; or whatever else you think is the right way can be implemented. Benwing2 (talk) 11:41, 18 November 2015 (UTC)

I find it a bit hard to read when the stress is before ‿ but I am not sure what's the right way. These connectors are a good thing (I like them) but are they standard? --Anatoli T. ^{(обсудить}/^вклад) 11:58, 18 November 2015 (UTC)

The connector should imply that the previous word attaches to the following word phonologically as a single word. Benwing2 (talk) 12:02, 18 November 2015 (UTC)

I personally feel that phonological words should be spelled without a space rather than with the ‿ symbol. --WikiTiki89 15:39, 18 November 2015 (UTC)

будь что будет

@Atitarev, Cinemantique, Wikitiki89 What should happen with final дь,ть, etc. before initial ш or ж? The module currently generates [but͡ʂʲ ʂto ˈbudʲɪt] but my new sandbox module generates [but͡ʂ ʂto ˈbudʲɪt] without the palatalization. Is there palatalization? Is there assimilation to the [ʂ]? Benwing2 (talk) 06:33, 20 November 2015 (UTC)

Assimilated but still palatalised, waiting for more input, complex case. --Anatoli T. ^{(обсудить}/^вклад) 09:11, 20 November 2015 (UTC)

I think, it sounds like буч што будет, isn't it? So it has [t͡ɕ], not [t͡ʂʲ].--Cinemantique (talk) 10:28, 20 November 2015 (UTC)

Yes, [t͡ɕ] represents the palatal [t͡ʂ]. --Anatoli T. ^{(обсудить}/^вклад) 10:54, 20 November 2015 (UTC)

@Atitarev, Cinemantique, Wikitiki89 OK, I will make this change. A couple of questions:

1. What about words ending in сь or зь and beginning with ш or ж? Does сь assimilate to ɕ, for example?
2. What about words with чж in them, like Тайчжун, чжуинь, чжуанский, Сучжоу, цзяньчжи, etc.? Do we get [d͡ʑʐ] or [d͡ʐʐ]? Benwing2 (talk) 08:55, 21 November 2015 (UTC)

1. I think, yes. гусь жа́ренный [guʑ ˈʐarʲɪn(ː)ɨj], за́пись жи́зни [ˈzapʲɪʑ ˈʐɨzʲnʲɪ], за́пись Шу́берта [ˈzapʲɪɕ ˈʂubʲɪrtə], гусь швеи́ [guɕ ʂvʲɪˈi]. 2. [d͡ʐʐ] would be good.--Cinemantique (talk) 10:07, 21 November 2015 (UTC)

@Atitarev, Cinemantique, Wikitiki89 Sounds good.

1. How about words with с щ and с ч across a word boundary, does it become [ɕ ɕː] [ɕ t͡ɕ]?
2. How about words with сь щ and сь ч across a word boundary, same?
3. How about words with т щ and т ч across a word boundary, does it become [t͡ɕ ɕː] and [t t͡ɕ]?
4. How about words with ть щ and ть ч across a word boundary, what happens then? Benwing2 (talk) 12:19, 21 November 2015 (UTC)

1-3. I think, that's right. 4. [t͡ɕ ɕː] and [tʲ t͡ɕ] (ждать щуку, прождать час).--Cinemantique (talk) 16:55, 21 November 2015 (UTC)

Thanks! Here's another one: бред си́вой кобы́лы. Currently pronunciation is given as [brʲet͡s ˈsʲivəj kɐˈbɨlɨ]. Should the t͡s become palatalized by the sʲ in the next word? In general, does palatalization assimilation of this sort work across word boundaries? Benwing2 (talk) 07:28, 22 November 2015 (UTC)

No, I think the current pronunciation is fine. Assimilation yes, palatalisation - no. That's how I pronounce it without any pause. --Anatoli T. ^{(обсудить}/^вклад) 07:47, 22 November 2015 (UTC)

Thanks. Benwing2 (talk) 15:26, 22 November 2015 (UTC)

ха́ос, Агринион, Черкесск, вепсский, о̀ргкомите́т, Борне́о, Альбире́о, адажио

@Atitarev, Cinemantique, Wikitiki89

How are these three words pronounced?

1. Does ха́ос end in [-os] or [-əs]? Similarly, does Агри́нион end in [-on] or [-ən]? Are there any other words ending in -VоC (V=vowel, C=consonant), and how are they pronounced?
2. Does Черкесск have [sːk] or [sk]?
3. Does вепсский have [sːk] or [sk]?
4. Is о̀ргкомите́т pronounced [ˌorkːəmʲɪˈtʲet] or [ˌorkəmʲɪˈtʲet]?
5. Do Борне́о (phonetic respelling Борнэ́о) and Альбире́о (phonetic respelling Альбирэ́о) end in [-o] or [-ə]? What about адажио?

Thanks!

Benwing2 (talk) 07:46, 20 November 2015 (UTC)

1. o
2. geminated
3. geminated (optional?)
4. geminated (morpheme boundary!). Alt stress is a good signal for that
5. o, defined by the rule - vowel + final о --Anatoli T. ^{(обсудить}/^вклад) 09:09, 20 November 2015 (UTC)

Гуйчжо́у vs. отжи́ть, where's the syllable boundary?

@Atitarev, Cinemantique, Wikitiki89 These two words disagree on whether the syllable boundary should be ˈd͡ʐʐ or d͡ʐˈʐ. Which one is it? Benwing2 (talk) 12:28, 21 November 2015 (UTC)

Ideally, they should differ. The former is a separate syllable, the latter - prefix plus a stem letter. --Anatoli T. ^{(обсудить}/^вклад) 02:21, 22 November 2015 (UTC)

I second what Anatoli said, as I believe I have already said before. --WikiTiki89 14:38, 23 November 2015 (UTC)

pronunciation of final -е, let's finally figure this out

@Atitarev, Cinemantique, Wikitiki89 The module currently makes платьице pronounced like платьицы (final [ɨ]) but оконце pronounced like оконца (final [ə]). The difference is whether a vowel precedes the ц. Same thing happens with дороже,даже,княже (final [ɨ]) vs. глубже,также,позже,же (final [ə]) and превыше,краше (final [ɨ]) vs. лучше,дальше (final [ə]). This seems strange to me; is there any logic to it? Note, for example, it will make же when attached phonologically to a preceding consonant-final word pronounced [ʐə], but pronounced [ʐɨ] when attached to a vowel-final word. I know that final -е is sometimes [ɨ/ɪ] and sometimes [ə], and it depends on what grammar role the word fills, e.g. neuter singular in -е is [ə], right?

What should be done here?

Benwing2 (talk) 04:53, 20 November 2015 (UTC)

I'll wait for Cinemantique re -це words but all other words, comparative forms you listed and даже, княже, также, же (always unstressed) have [ɨ]. --Anatoli T. ^{(обсудить}/^вклад) 04:57, 20 November 2015 (UTC)

I see no difference between оконце and платьице, there should be final [ə]. After ж or ш [ɨ] should be. After soft consonant [ɪ] should be. [polʲə] (по́ле) is the Old-Moscow pronunciation, as Avanesov said.--Cinemantique (talk) 10:17, 20 November 2015 (UTC)

I think these should just be manually overridden as {{ru-IPA|око́нца}} and {{ru-IPA|пла́тьица}}. --WikiTiki89 15:03, 20 November 2015 (UTC)

@Atitarev, Cinemantique, Wikitiki89 Cinemantique, is your comment about по́ле indicating that all final -е should be pronounced like final -и except after ц? I can make this page but I suspect it will affect a whole lot of words. What about words like моро́женое? Isn't this pronounced with final [ə]? I calculate that about 2000-2500 entries will change if this is done. The following, for example, is a list of all the words beginning with д and ending with е, and many would be changed (I can't say how many because I'm not sure which ones have overridden the pronunciation already):

• давайте
• давление
• даже
• дайте мне
• далее
• дальше
• данные
• дарование
• датирование
• двадцать четыре
• две
• движение
• двое
• двоемыслие
• двоеперстие
• двоеточие
• двуперстие
• де-юре
• девяносто четыре
• деепричастие
• действие
• декольте
• делать предложение
• делегирование
• деление
• деревенение
• деревце
• десантирование
• десятилетие
• детище
• дефиле
• деформирование
• дешевле
• деяние
• дистанционное обучение
• дистанционное управление
• днище
• добавление
• добродушие
• доведение
• доверие
• довольствие
• дозволение
• дознание
• доколе
• долготерпение
• дольше
• домашнее задание
• домашние
• доминирование
• домище
• донесение
• дополнение
• допущение
• дороже
• доселе
• достижение
• достояние
• досье
• драже
• древние
• древоизмерение
• древонасаждение
• дреколье
• дробление
• друг познаётся в беде
• дружище
• дружнее
• дуновение
• дурнее
• дуче
• Душанбе
• дыхание
• дюже

Benwing2 (talk) 08:50, 21 November 2015 (UTC)

@Atitarev, Cinemantique, Wikitiki89 Anatoli, Cinemantique, Wikitiki, do you have any comments about the above?

How are the following words pronounced (in particular the ending -е)?

1. Neuter nouns and adjectives ending in -ое, e.g. моро́женое, былое, луковое, насекомое, отглагольное, существительное, животное

   [jə] (Avanesov, §5.7.1 ж), page 99).--Cinemantique (talk) 17:55, 22 November 2015 (UTC)

2. Neuter nouns ending in -ие, e.g. дуновение, дыхание, добродушие, двуперстие, двоемыслие

   [jɪ]; [jə] is Old-Moscow accent (Avanesov, §5.7.2 б), page 100).--Cinemantique (talk) 17:55, 22 November 2015 (UTC)

3. Neuter nouns ending in -ье, e.g. дреколье, веселье, плоскогорье, запястье, воскресенье, платье, здоровье, аканье, счастье

   [jɪ]; [jə] is Old-Moscow accent (Avanesov, §5.7.2 б), page 100).--Cinemantique (talk) 17:55, 22 November 2015 (UTC)

4. Neuter nouns ending in soft paired consonant + -е, e.g. по́ле, море, горе

   [ɪ]; [ə] is Old-Moscow accent (Avanesov, §5.7.2 б), page 100).--Cinemantique (talk) 17:55, 22 November 2015 (UTC)

5. Neuter nouns ending in -це, e.g. деревце, солнце, сердце, платьице, полотенце, блюдце, зеркальце, оконце, щупальце, болотце, кружевце

   [ə] (Avanesov, §5.8, page 102).--Cinemantique (talk) 18:23, 22 November 2015 (UTC)

6. Neuter nouns ending in -ше (maybe there aren't any)

   [ə]. No example. The rule is that 'е' after 'ж, ш, ц' in post-stress syllable sounds like [ə], not only at the end of word, e. g. плю́шем [ˈplʲʉʂəm] (Avanesov, §5.8, page 102).--Cinemantique (talk) 18:23, 22 November 2015 (UTC)

7. Neuter nouns ending in -же (could only find ложе)

   [ə] (Avanesov, §5.8, page 102).--Cinemantique (talk) 18:23, 22 November 2015 (UTC)

8. Neuter nouns ending in -ще, e.g. училище, жилище, кладбище, туловище, влагалище, чудовище, чудище, сборище

   [ɪ]; [ə] is Old-Moscow accent (Avanesov, §5.7.2 б), page 100). Listen there.--Cinemantique (talk) 19:01, 22 November 2015 (UTC)

9. Neuter nouns ending in -че (could only find вече)

   Like previous.--Cinemantique (talk) 19:01, 22 November 2015 (UTC)

10. The adjective plural endings -ие, -ые, e.g. данные

    [ɪɪ], [ɨɪ] (Avanesov, §16.4, page 198).--Cinemantique (talk) 17:35, 22 November 2015 (UTC)

11. Comparatives in -ее, e.g. дурнее, дружнее

    [jɪ]; [jə] is Old-Moscow accent (Avanesov, §5.7.2 в), page 100).--Cinemantique (talk) 17:55, 22 November 2015 (UTC)

12. Comparatives in -ше, же e.g. больше, лучше, выше, дальше, краше; дороже, глубже

    [ɨ] (Avanesov, §5.8, page 102).--Cinemantique (talk) 18:27, 22 November 2015 (UTC)

13. Comparatives in -че, ще, e.g. легче, богаче, короче, крепче; гуще

    [ɪ].--Cinemantique (talk) 18:27, 22 November 2015 (UTC)

14. Adverbs ending in vowel + -е (any?)
15. Adverbs ending in soft paired consonant + -е, e.g. досе́ле

    [ɪ].--Cinemantique (talk) 18:27, 22 November 2015 (UTC)

16. Adverbs ending in ше, же, e.g. также, даже, дюже, неклюже

    I think, [ɨ].--Cinemantique (talk) 18:27, 22 November 2015 (UTC)

17. Adverbs ending in че, ще, e.g. иначе; блестяще, зловеще, потрясающе

    [ɪ].--Cinemantique (talk) 18:27, 22 November 2015 (UTC)

18. Adverbs ending in це (any?)

    I've found only one - вкра́тце. I think, with [ɨ].--Cinemantique (talk) 18:36, 22 November 2015 (UTC)

19. Cardinals ending in -е, e.g. двое, четыре

    I think, двое and трое - [jɪ] or [jə] (Avanesov-2, §55), четыре - [ɪ].--Cinemantique (talk) 18:31, 22 November 2015 (UTC)

20. The prepositional singular ending -е after a soft consonant

    [ɪ] (Avanesov, §5.7.3 г), д), page 101).--Cinemantique (talk) 18:01, 22 November 2015 (UTC)

21. The prepositional singular ending -е after ше, же

    [ɨ] (Avanesov, §5.8, page 102).--Cinemantique (talk) 18:23, 22 November 2015 (UTC)

22. The prepositional singular ending -е after че, ще

    [ɪ].--Cinemantique (talk) 18:01, 22 November 2015 (UTC)

23. The prepositional singular ending -е after це: австралийце

    [ɨ] (Avanesov, §5.8, page 102).--Cinemantique (talk) 18:23, 22 November 2015 (UTC)

24. Verbal endings in -е, e.g. давайте

    [ɪ] (Avanesov, §5.7.3 и), page 101).--Cinemantique (talk) 18:01, 22 November 2015 (UTC)

25. Neuter nouns and adjectives in -ее, e.g. си́нее
26. Neuter adjectives in -ье, e.g. ли́сье

    [jə] (Avanesov, §5.7.1 ж), page 100).--Cinemantique (talk) 15:47, 24 November 2015 (UTC)

27. Vocatives in -е, e.g. кня́же, ста́рче, сы́не
28. The prepositional singular after vowels, e.g. бо́е, кра́е, бу́е, музе́е, Тайбэ́е, и́нее

    [(j)ɪ] (Avanesov, §5.7.3 г), page 101).--Cinemantique (talk) 15:47, 24 November 2015 (UTC)

29. Dative and prepositional singular of feminine nouns in -а, –я

    [ɪ] (Avanesov, §5.7.3 д), page 101).--Cinemantique (talk) 15:47, 24 November 2015 (UTC)

Thanks very much for your help! I am trying to determine what should be done with final -е. Benwing2 (talk) 08:43, 22 November 2015 (UTC) BTW Anatoli, Cinemantique, I know you've commented previously on some of these words but I'd like to get everything in one place. Benwing2 (talk) 08:45, 22 November 2015 (UTC)

@Cinemantique: What symbol does Avanesov actually use for the final [-jɪ] and [-ɪ]? I think he is wrong about [-jə] and [-ə] being the "Old" Moscow accent, I think it is still the current Moscow accent. --WikiTiki89 18:54, 22 November 2015 (UTC)

He uses 'ə' for [ɪ], 'ъ' for [ə] (unfortunately). Listen here. Is this pronunciation normal or marginal?--Cinemantique (talk) 19:05, 22 November 2015 (UTC)

One more time in his orthoepic dictionary, §55 ('ь' for [ɪ], 'ъ' for [ə]). Yes, we have two variants. Which of them should we use here?--Cinemantique (talk) 19:26, 22 November 2015 (UTC)

Ok, so I think you've been misusing the symbol [ɪ]. If a word ends in [-ɪ], then it would sound the same as one that ends with the letter и. What I hear in those recordings is something in between [-ɪ] and [-ə], which we can call [-e]. --WikiTiki89 19:23, 22 November 2015 (UTC)

[e] appears in stressed syllables (unstressed in some foreign words). I hear it there. I think, it's not natural.--Cinemantique (talk) 19:41, 22 November 2015 (UTC)

I didn't mean that it is [e], just that we can call it [e]. If you want to add some IPA diacritics to that to make it more accurate, go ahead. But I think we can just say that the quality of [e] is different in stressed and unstressed syllables and that in the recording you just linked to, the [e] is actually under secondary stress. --WikiTiki89 19:46, 22 November 2015 (UTC)

Hmmm, if we take what Cinemantique wrote above at face value then it looks like we want to treat -ое and -це as ending in [-ə] and all the rest as ending in [-ɪ]. But I also hear a sound something like [-ɛ] in the recording of жилище labeled as Russian although it's so short it's hard to tell (the one labeled Ukrainian sounds more like [-ɪ]). In the recording of училище живописи, the two words sound like they have clearly different endings. Wikipedia in w:Russian phonology#Unstressed vowels claims that жи́тели, жи́теле and жи́теля all sound different, respectively [ˈʐɨtʲɪlʲɪ], [ˈʐɨtʲɪlʲɛ] and [ˈʐɨtʲɪlʲə], and they cite Avanesov. Benwing2 (talk) 02:27, 23 November 2015 (UTC)

@Benwing2 I am more inclined to take this approach but not for all cases (I couldn't find these examples in Wikipedia, though). I am sometimes confused by Avanesov's rendering of the final "-е" but if I offer my vision, it could be considered an original research. I can suggest my voice recordings for your analysis. I could try to make them as natural as possible but this couldn't be used as a source. In my opinion, there's a trend to pronounce the final "-е" more or less clearly in Russia. It could be an effect of the spelling on how people speak or to avoid any misunderstandings, like жи́тели, жи́теле and жи́теля, which are all different forms. --Anatoli T. ^{(обсудить}/^вклад) 03:54, 23 November 2015 (UTC)

@Atitarev The Wikipedia examples are under the "Vowel mergers" section on the page I linked to, second bullet point under the line that says "There are a number of exceptions to the above vowel-reduction rules". I'm OK with making final unstressed -е be rendered [-e] or [-ɛ], as Wikipedia has it, and using a final spelling with -a/-я or -и to get a pronunciation with [-ə] or [-ɪ]. Possibly we could make exceptions for cases like -ое, and -це, if there are no examples where the [-e] or [-ɛ] sound is needed. If there are definite part-of-speech rules regarding when a spelling with -a/-я or -и is required, I could probably have a bot do the fixing-up. But I'd want to make sure Cinemantique is OK with this. Benwing2 (talk) 04:05, 23 November 2015 (UTC)

• I absolutely disagree with the transcription [(o) ˈʐɨtʲɪlʲɛ]: [ɛ] appears after hard consonants. It's really complicated. Knyazev thinks, there's 4 variants of pronunciation 'на во́ле' (§172). We can stay with the Old-Moscow accent for the cases 2, 3, 8, and 11. We have few words for the case 4: поле, биополе, горе, море, (живе́те? мысле́те?), кофе, вече, (дуче?). That's all. We can add manually: поля, горя, моря, etc.--Cinemantique (talk) 09:50, 23 November 2015 (UTC)

[ˈʐɨtʲɪlʲɛ] (жи́теле) must really be a misreading of Avanesov's transcription. (He should have used IPA to avoid the misreading). [ˈʐɨtʲɪlʲe] would be more accurate rendering and IMO, the real modern pronunciation of the form. --Anatoli T. ^{(обсудить}/^вклад) 10:58, 23 November 2015 (UTC)

Let's not be nitpicky about the choice of symbols [ˈʐɨtʲɪlʲɛ] and [ˈʐɨtʲɪlʲe] mean the same thing as far as I'm concerned. I think the latter is a better fit, but neither is "wrong" or "right". --WikiTiki89 14:43, 23 November 2015 (UTC)

@Cinemantique Do you object to [e] as much as [ɛ]?

I'm thinking of doing the following:

1. final unstressed [-е] after ц and о is [ɘ]
2. final unstressed [-е] after ш and ж is [ɨ]
3. final unstressed [-е] anywhere else is [e]
4. To force [ə], write the letter a (after ц ш ж) or я (elsewhere).
5. To force [ɨ], write the letter ы.
6. To force [ɪ], write the letter и.
7. To force [e], write the letter е̂ (or е̏).

This will require us to do the following:

1. Use a bot to rewrite all instances of final -и in the phonology section that correspond to -е in the lemma to be -е.
2. Rewrite the few instances where the above rules don't apply to use a different letter in the phonology section (e.g. for ложе apparently we want {{ru-IPA|ло́жа}}).

An alternative is to add a pos= argument with values such as pos=neut (neuter noun), pos=adj (adjective, autodetected as singular or plural based on the ending), pos=comp (comparative), pos=verb (verb second-person plural forms), pos=adv (adverbs), pos=pre (prepositional singular of nouns), pos=voc (vocative singular, as in кня́же). This would take more bot work but has the advantage that we can change the representation of final -е without massive disruption. Benwing2 (talk) 08:19, 24 November 2015 (UTC)

BTW I added five new categories of -е, #'s 25-29, above. Benwing2 (talk) 08:21, 24 November 2015 (UTC)

• Why can't we use [ɪ]? Listen how it sounds there and there. It looks like [e] for me.--Cinemantique (talk) 15:47, 24 November 2015 (UTC)

  The question is, does жили́ще (by Avanesov's research and by your own pronunciation) end in the same vowel as ве́щи? If the answer is yes, then we can use [ɪ], if the answer is no, then we cannot use [ɪ], because we already use [ɪ] for ве́щи. I strongly believe that the answer is no, but then again, I use the Moscow pronunciation of [ə], which makes it difficult for me to judge. --WikiTiki89 16:04, 24 November 2015 (UTC)

  Avanesov said that this mysterious sound is something between [i] and [e] but closer to [i], sometimes more close to [i] or just [i]. I've downloaded this file and made it slow, that's a result. Do you hear something like [pʲeeerʲeee...]? We use [ɪ] for it. Also, I've made the same with this file (мири́ть) - what we hear? I hear the same sound.--Cinemantique (talk) 18:22, 24 November 2015 (UTC)

  I was asking only about the end of a word, not the middle. --WikiTiki89 19:04, 24 November 2015 (UTC)

  Alright, this is the end of word: сиськи (by Yarka) > slowly. I hear something like "...кеее". In another tests I hear something like [i] or uncertain.--Cinemantique (talk) 21:29, 24 November 2015 (UTC)

  It's better to compare with the same consonant and in a similar stress and palatalization environment. That's why I chose жили́ще and ве́щи as a good example. --WikiTiki89 21:32, 24 November 2015 (UTC)

  @Atitarev, Cinemantique, Wikitiki89 I agree with Wikitiki here, what's most important is whether there's a difference between final -е and final -и, less exactly what they sound like in isolation. In the sound file provided by Cinemantique, I hear a clear difference between жили́ще and ве́щи. The Wikipedia page that claims the difference between жи́тели and жи́теле gives as references Avanesov 1975 pages 121-125 ("Фонетика современного русского литературного языка") and Avanesov 1985 page 666 ("Сведения о произношении и ударении", in Borunova, C.N.; Vorontsova, V.L.; Yes'kova, N.A., Орфоэпический словарь русского языка. Произношение. Ударение. Грамматические формы). What exactly do these references say? Benwing2 (talk) 02:13, 26 November 2015 (UTC)

  @Benwing2 Please tell me what I can do to help. There are things I agree and disagree with Avanesov but I don't have a source for points I disagree on. --Anatoli T. ^{(обсудить}/^вклад) 04:23, 26 November 2015 (UTC)

  @Atitarev If you made recordings of normal pronunciation of pairs of words that are as close as possible but differ in final -е vs. -и, it might help, also quoting the above pages from the above sources if you have them. Benwing2 (talk) 04:57, 26 November 2015 (UTC)

  @Benwing2 I am considering this but I can't promise. --Anatoli T. ^{(обсудить}/^вклад) 23:05, 26 November 2015 (UTC)

  OK, thanks. Benwing2 (talk) 02:14, 27 November 2015 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ @Atitarev, Cinemantique, Wikitiki89 I implemented a first pass of final -е work. You can see the results in Module:User:Benwing2/ru-pron/testcases. I added a pos= parameter that determines how final -е is rendered, according to the descriptions above, where I have taken Cinemantique's [ɪ] to mean [e], in accordance with what Wikitiki, Anatoli and I seem to have agreed upon. I assume Cinemantique doesn't like this; part of the benefit of the pos= param is that it's easy to change the way all or some final -е's look (e.g. to make some [e] become [ɪ]) just by editing some tables in the module.

Anatoli, if you could record some of the words with final -е that I added in Module:User:Benwing2/ru-pron/testcases and upload them, it would be very helpful. Even better would be if you could compare each of them to the same word ending in -и if it exists, and if not to another word that rhymes or otherwise preserves as much of the last few letters as possible. Benwing2 (talk) 10:39, 30 November 2015 (UTC)

Sorry, I couldn't do it again. We are busy with the renovation again. :) I'll try in the next few days. --Anatoli T. ^{(обсудить}/^вклад) 22:36, 30 November 2015 (UTC)

@Cinemantique I've noticed you've added some test cases with final -е represented as [ɪ]. Have you listened to Anatoli's recordings in User talk:Atitarev/recording? What do you think? Also, can you quote what it says in the following references, if you have them available?

• Avanesov 1975 pages 121-125 ("Фонетика современного русского литературного языка")
• Avanesov 1985 page 666 ("Сведения о произношении и ударении", in Borunova, C.N.; Vorontsova, V.L.; Yes'kova, N.A., Орфоэпический словарь русского языка. Произношение. Ударение. Грамматические формы)

These are what Wikipedia cites to justify writing final -е as [ɛ]. Thanks! Benwing2 (talk) 13:19, 10 December 2015 (UTC)

• We can read the first on-line. Avanesov uses three symbols, [и] (in unstressed syllables, too), [и^е], and [ь]. Some quotes (can someone translate them? @‎Atitarev, Stephen G. Brown, Wikitiki89):--Cinemantique (talk) 17:58, 10 December 2015 (UTC)
  • page 121: “Однако сложного заударного вокализма после мягких согласных определяется чрезвычайной широтой колебаний в произношении гласного в пределах одной и той же морфемы, неразграниченностью [и] и [ь], весьма слабой разграниченностью [ь] и [ъ].”
  • page 123: “... безударное [и] с высшей степенью редукции и вялости артикуляции, с пониженным подъёмом языка и есть тот звук, который в фонетической транскрипции обозначается буквой [ь].”
  • page 124: Avanesov contrasts жители with жителе but for the first case he uses the same symbol like in stressed syllables, so it's [i].
  • page 125: “... все четыре описанные слабые фонемы заударного конечного открытого слога различаются лишь в отчётливом, чеканном произношении. В беглой речи [и] и [и^е]... могут не различаться: вместо них может произноситься открытого типа [и]...”
• Page from the second one. Also, read §5.--Cinemantique (talk) 18:14, 10 December 2015 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ @Atitarev Can you help translate the above quotes, and perhaps some of the text in the pages that Cinemantique linked to? Benwing2 (talk) 01:30, 18 December 2015 (UTC)

@Benwing2 Sorry, I didn't get around to do it. Can you remind me what IPA symbols Avanesov's symbols refer to, if you can? Cinemantique gave the explanations earlier but I can't find them. It would make more sense to me. I'll try to translate tonight. --Anatoli T. ^{(обсудить}/^вклад) 02:33, 18 December 2015 (UTC)

@Atitarev Unfortunately I don't know what the symbols stand for exactly. Cinemantique said originally that [ь] stands for [ɪ] and [ъ] stands for [ə], but perhaps they mean something else. There's also [и] and [и^е]. I suggest you leave the symbols as-is; it should hopefully become clearer what they mean in context. Benwing2 (talk) 03:05, 18 December 2015 (UTC)

• Russian Wikipedia uses [ɪ] both for [и^е] and [ь] (see section "/i/ и /ɨ/ (/и/ и /ы/)" and table "Основные аллофоны гласных").--Cinemantique (talk) 10:05, 18 December 2015 (UTC)
• @Benwing2: Here is a quick, unpolished translation:
  • page 121: “Yet a complex posttonic vocalism after soft consonants is distinguished by an exceptionally wide oscillation in the pronunciation of the vowel inside the same morpheme, the indifferentiability of [и] and [ь], and the very week differentiability of [ь] and [ъ].”
  • page 123: “... unstressed [и] with a higher level of reduction and fading articulation, with a decreased slope of the tongue, and is the sound that in phonetic transcription is denoted by the letter [ь].”
  • page 125: “... all four described weak phonemes of a posttonic final open syllable are differentiated only in distinct, chiseled pronunciation. In fast speech, [и] and [и^е]... may not be differentiated: instead of them an open type of [и] may be pronounced...”

--WikiTiki89 15:52, 18 December 2015 (UTC)

@Wikitiki89 Thanks. I had a big night last night. Couldn't do it. --Anatoli T. ^{(обсудить}/^вклад) 23:08, 18 December 2015 (UTC)

Avanesov quote pp. 123-125

@Atitarev, Cinemantique, Wikitiki89 The critical section starts near the bottom of p. 123:

Рассмотрим теперь вокализм заударного конечного открытого слога. Как и в других заударных слогах, здесь имеются различия в произношении гласных, обусловленные положением после мягких согласных, заднеязнычных, твёрдых шипящих и парных твёрдых согласных. Общим для вокализма заударного конечного открытого слога является различение в отчетливом стиле произношения трех степеней подъема при различении отсутствия или наличия лабилизации только в верхнем ряду.

В положении после мягких согласных вокализм заударного конечного открытого слога может быть представлен в следующей схеме:

и	у
ие
ъ

(various examples, e.g. жи́тели (nom pl) [жы́·т'ьл'и], жи́теле (prep sg) [жы́·т'ьл'и^е], жи́теля (gen sg) [жы́·т'ьл'ъ], жи́телю [жы́·т'ьл'у]; similarly for о́куни [о́кун'и], в о́куне [в-о́кун'и^е], о́куня [о́кун'ъ], о́куню [о́кун'у]; идите [ид'·и́т'и^е], знайте [зна·i̯т'и^е]; по́ле (nom sg) [по́·л'ъ] or [по́·л'ь], по́ля (gen sg) [по́·л'ъ], по́лю [по́·л'у]; similarly for мо́ре, мо́ря, мо́рю.

В положении полсе заднеязычных различается:

и	у
ие
ъ

В положении перед [и] и [и^е] заднеязычные позиционно снягчаются.

(examples, e.g. услу́ги (nom pl) [услу́·г'и], услу́ге (dat and prep sg) [услу́·г'и^е], услу́га (nom sg) [услу́гъ], услу́гу (nom sg) [услу́гу]; similarly for му́хи, му́хе, му́ха, му́ху; поро́ки (nom pl) [пʌро́·к'и], поро́ке (prep sg) [пʌро́·к'и^е], поро́ка (gen sg) [пʌро́къ], поро́ку (dat sg) [пʌро́ку].

В положении после твердых шипящих система гласных заударного конечного открытого слога имеет следующий вид:

ы	у
ыэ
ъ

(examples, e.g. лу́жи (nom pl, gen sg?) [лу́жы], в луже [в-лу́жы^э], луже (nom sg) [лу́жъ], лу́жу (dat sg?) [лу́жу]; similarly for гру́ши (nom pl, gen sg?) [гру́шы], на гру́ше [нʌ-гру́шы^э], гру́ше [гру́шъ] (nom sg), гру́шу [гру́шу].

После твёрдых парных согласных в заударном конечном открытом слоге представлена та же система гласных, с той однако особенностью, что перед [и^е] имеет место позиционное смягчение согласной:

ы	у
ие
ъ

(examples, e.g. нака́зу (dat sg) [нʌка́зу], нака́зы (nom pl) [нʌка́зы], нака́за (gen sg) [нʌка́зъ], в нака́зе [в-нʌка́·з'и^е]; складу (dat sg) [скла́ду], скла́ды (nom pl) [скла́ды], склада (gen sg) [скла́дъ], в скла́де [ф-скла́·д'и^е]; similarly for кана́ву, кана́вы, кана́ва, в кана́ве.

Гласный [и^е] после мягких согласных можно бы было обозначить буквой [ь] (речь идет о случаях типа [ʌб-уч'и̇'т'ьл'и^е], [нʌ-ба́·з'и^е]). Но в этом случае соответствующий гласный после твердых шипящих (и [н]) пришлось бы обозначить буквой [ъ] (в случаях типа [в-лу́жы^э], [нʌ-гру́шы^э]). Это представило бы большие неудобства, так как при помощи буквы [ъ] в положении после твердых шипящих (и [ц]] обозначается более открытый гласный и притом не смягчающий предшествующий согласный (ср. смягчение перед [и^е] предшествующего заднеязычного и парного твёрдого согласного). Ср. [в-лу́жы^э] и [лужъ] (им. п. ед. ч.), [нʌ-гру́шы^э] и [гру́шъ] (им. п. ед. ч.). Поэтому гласный переднего образования в заударном конечном открытом слоге, по подъемы языка средний между [и] и [ь], обозначаем знаком [и^е], а его отодвинутый назад вариант (в положении после твердых шипящих и [ц]) знаком [ы^э].

Следует отметить, что все четыре описанные слабые фонемы заударного конечного открытого слога различаются лишь в отчётливом, чеканном произношении. В беглой речи [и] и [и^е] (а также соответственно [ы] и [ы^э] могут не различаться: вместо них может произноситься открытого типа [и] (и соответственно [ы]): [жы́·т'ьл'и] (им. п. мн. ч.) и [ʌ-жы́·т'ьл'и]; [ко́жы] (им. п. мн. ч.) и [ф-ко́жы]. Ср. просторечные [нъ-рʌбо́·т'и], [нʌ-скла́·д'и] и пр.

При таком произношении система гласных заударного конечного открытого слога не отличается от вокализма других заударных слогов, так как она принимает следующий вид:

после мягких согласных и задненебных:		после твердых:
и	y	и	у
ъ		ъ

Comments on Avanesov pp. 123-125

First, can someone help translate the above text? My sense of it comes from Google Translate, which is imperfect (e.g. it doesn't have the word заударный, which I assume means "post-tonic" aka "post-stressed"; maybe it means "unstressed"?). But what I'm getting is:

1. There are three sounds in final unstressed open syllables, where -я, -е and -и are all pronounced differently, at least in careful speech. In less careful speech (I think that's right), -е and -и merge.
2. Avanesov uses [и] for both stressed and unstressed syllables, and similarly for [у], not necessarily clearly distinguishing the stressed and unstressed variants. This suggests to me that unstressed [и] should be IPA [ɪ]. Now, [и^е] is clearly different from [и], so it probably should be IPA [e]. Meanwhile, [ъ] is clearly IPA [ə], and [ʌ] is clearly IPA [ɐ] (this is clear from their distribution pre-stress when standing in for orthographic о and а). If we follow Cinemantique in using IPA [ɪ] for [и^е], then we have to use [i] for unstressed [и], which seems wrong.
3. For [ь] I'm not sure what to use; if we follow ruwiki in equating [и^е] and [ь] then we should use IPA [e]. Otherwise, it should be IPA [ɪ].
4. Complicating this is the non-final post-tonic use of [ь] to represent unstressed е in жи́тели/жи́теле/жи́теля, and the pre-tonic use of [и^е] in медве́дь (с медведем [с-м'и^ед'в'е̂ʹд'ъм]).
5. I also don't know what the center-dot and prime (ʹ) are that follow vowels that are followed by a soft consonant.
6. In conclusion, I think we should follow essentially what I've proposed before: Use [-e] for final -е at least in the prepositional singular, feminine dative singular, and verbal ending -те, which per Avanesov have the same sound. Benwing2 (talk) 04:54, 19 December 2015 (UTC)

Avanesov p. 666

Section 52 says this (can someone translate?):

В заударных слогах после мягких согласных, кроме гласных [и] и [у] (о них см. 5-13), произносятся гласные [ь] и [ъ]. Гласный [ь] обозначается обычно буквой e, но в некоторых случаях также и буквой я (а после [ч] и [щ] буквой а). Гласный [ъ] обозначается я (а после [ч] и [щ] буквой а), но в некоторых случаях также буквой е.

Примечание. Гласные [ь] и [ъ] отличаются друг от друга прежде всего местом образования, которое является более передним у [ь] и более задним у [ъ]. Кроме того, они отличаются и степенью подъема языка -- более высоким у [ь] (близким к подъему языка при [и]) и более низким у [ъ] (при произношении гласного [ъ] подъем языка заметно ниже, чем при [ы]). Ср. ка́пля и по ка́пле (произносится [капл^ьъ] и [па-ка́пль]).

I'm going to paraphrase what I think the rest of it says.

It uses [ь] to represent final -е and [ъ] to represent final -я, but says there are exceptions in both directions. It appears to use [и] to represent final -и, as in вы́тяни, spelled [вы́тьни], and uses the same [ь] to represent non-final post-stressed я and е, as in вы́тяни just cited, вы́беру [вы́бьру], краси́вее [краси́вьйь] (note pre-stressed [а] to represent IPA [ɐ]). If we take this literally, then final [ь] represents [ɪ] and final [и] represents [i], but I don't think I believe this.

BTW it says that [ь] is the pronunciation of final -е in вы́чет, па́лец; оле́ней, сосе́дей (gen pl?); ка́плей, ба́ней, ту́чей, ро́щей (gen pl?); армя́не, горожа́не (nom pl); в до́ме, в по́ле, на се́не (pre sg); к рабо́те, в ба́не, в ро́ще (fem dat/pre sg?); краси́вее, угрю́мее (comparatives?); деше́вле, бога́че (adverbs); ста́нешь, бу́дешь, пла́чешь (2nd sing verbs); ста́нет, бу́дет, пла́чет (3rd sing verbs); ста́нем, бу́дем, пла́чем (1st pl verbs); стане́те, бу́дете, пла́чете (2nd pl verbs, both е's); ста́ньте, бу́дьте, пла́чьте, купи́те, клади́те (imp pl).

It says that the following can have [ъ] ("old norm") or [ь] (more modern): зло́е, друго́е; ста́рое, до́брое; бара́нье, пти́чье; дво́е, тро́е; ка́мнем, учи́телем, пла́чем; бра́тьев, сту́льев, су́чьев; мо́ре, по́ле, сча́стье (nom sg). Benwing2 (talk) 05:55, 19 December 2015 (UTC)

Action taken

@Atitarev, Cinemantique, Wikitiki89 I got tired of waiting for any of you to comment, and so I went ahead and enabled my final-е code. This renders most final е's as [e]. Per Avanesov, final -е and final -и are different, and we seem to have two choices: Render final -и as [ɪ] and final -е as [e], or render final -и as [i] and final -е as [ɪ]. Rendering them both the same isn't an option. I chose the former option because it seems to match up with what I hear spoken, and Anatoli and Wikitiki seem to agree. But if the consensus is to choose the latter option, I'll go with that instead. I'm going to shortly run a script to change pronunciations where -и or -е̂ is used to force a particular pronunciation back to using -е, and add pos= params based on the actual part of speech of the word in question, which should allow for fine-grained control of the pronunciation of final -е if/when we need it. Benwing2 (talk) 05:24, 22 December 2015 (UTC)

fronting of я, ю only between soft consonants

@Atitarev, Cinemantique, Wikitiki89

1. Currently we front /a/ to [æ], /u/ to [ʉ], /o/ to [ɵ] whenever following a soft consonant. Wikipedia claims that this is accurate for /o/, but /a/ and /u/ get fronted only when between soft consonants, e.g. in пять and чуть. Should we implement this?
2. Wikipedia also claims that xkʲ assimilates to xʲkʲ, which is why мягкий has [æ] in it. Should we implement this?
3. Wikipedia also claims that /a/ before l and not after a soft consonant becomes [ɑ], e.g. in палка. Should we implement this?

Benwing2 (talk) 15:39, 22 November 2015 (UTC)

I am OK with 1 and 2, not sure about 3 but won't object if it's confirmed and no other objections. --Anatoli T. ^{(обсудить}/^вклад) 22:00, 22 November 2015 (UTC)

@Atitarev, Cinemantique, Wikitiki89 I am going to implement 1 and 2. Some questions (all of the following concern /a/ but I imagine it's the same for /u/):

1. Does fronting of /a/ to [æ] occur when followed by hard consonant and then immediately by a soft consonant? E.g. is в поря́дке pronounced [f‿pɐˈrʲatkʲɪ] or [f‿pɐˈrʲætkʲɪ], and is дегтя́рня pronounced [dʲɪkˈtʲarnʲə] or [dʲɪkˈtʲærnʲə]?
2. Does fronting of /a/ to [æ] occur when followed by an optionally soft consonant? E.g. is сча́стливо pronounced [ˈɕːas⁽ʲ⁾lʲɪvə] or [ˈɕːæs⁽ʲ⁾lʲɪvə], or is it [ˈɕːaslʲɪvə] / [ˈɕːæsʲlʲɪvə]?
3. Does fronting of /a/ to [æ] occur when preceded by a soft consonant at the end of a word? E.g. is Амударья́ pronounced [ɐmʊdɐrʲˈja] or [ɐmʊdɐrʲˈjæ] (or maybe [ɐmʊdɐˈrʲja] etc. with stress before cons+ь)?

Thanks! Benwing2 (talk) 05:38, 23 November 2015 (UTC)

• In fact, all vowels differ in these positions (for example [a]):

1. normal articulation: at the beginning of word in front of hard consonant / between hard consonants / at the end of word after hard consonant - [aaaaa]
2. in front of soft consonant - [aaaaaⁱⁱ]
3. (difference is stronger) after soft consonant - [ⁱⁱⁱⁱaaaaa]
4. (stronger) between soft consonants - [ⁱⁱⁱⁱaaaaaⁱⁱ].

• I really don't know how to show the difference and if we should do that.--Cinemantique (talk) 10:43, 23 November 2015 (UTC)

@Cinemantique I'm not really sure what your symbols mean here. There are various IPA ways to indicate different degrees of fronting, e.g. you can use the IPA "advanced" symbol (a small + under the vowel) and "retracted" symbol (an underline, probably meant to bring to mind a minus symbol), and write e.g. [ä] for a centralized a, [ä̟] for a fronted centralized a, [a̱] for a retracted front a, [a] for a regular front a, [æ] for an even more front, and slightly raised, a, etc. But when listening to the difference between этап (hard-hard) and радио (hard-soft) and мясо (soft-hard) and пять (soft-soft), the first three sound pretty similar while the last while is clearly a lot more fronted. этап is a bit more back than радио and мясо (which sound indistinguishable to me in their /a/ sounds), but I don't know how much that's a function of the speaker and the following labial. So I think we should follow what Wikipedia says in w:Russian phonology and use [æ] only between soft consonants.

This again brings up the issue of what to do with /a/ + optionally palatalized sound as in my example of сча́стливо above. What I'm planning on doing is have the module generate two different pronunciations in this case, so it writes [ˈɕːaslʲɪvə], [ˈɕːæsʲlʲɪvə]. If you think this is wrong please let me know. Benwing2 (talk) 07:52, 24 November 2015 (UTC)

OK, I agree.--Cinemantique (talk) 09:31, 24 November 2015 (UTC)

Cool. Please check out Module:User:Benwing2/ru-pron/testcases. This is after I enabled restricted fronting of a/u. It increased the number of test failures from 8 to 45; we'll have to change 37 of them. It looks good to me, and I checked all the failures to make sure all the new changes were good; let me know how it seems to you. Benwing2 (talk) 10:05, 24 November 2015 (UTC)

код морзе has /...d m.../ per Anatoli, does this also happen with д л and д р?

@Atitarev, Cinemantique, Wikitiki89 See subject line. Benwing2 (talk) 02:31, 23 November 2015 (UTC)

Yes. Voicing and devoicing doesn't happen but consonants don't lose their qualities in front of sounds m, n, l and r. --Anatoli T. ^{(обсудить}/^вклад) 02:53, 23 November 2015 (UTC)

How about д..я and д..в? Does the д remain voiced? Benwing2 (talk) 03:00, 23 November 2015 (UTC)

And д...vowels, like д..о? Benwing2 (talk) 03:01, 23 November 2015 (UTC)

Yes, if it's "в" + voiced consonant or a vowel (над водой, над облаком), otherwise, "t" - над вторым.

Yes. --Anatoli T. ^{(обсудить}/^вклад) 03:05, 23 November 2015 (UTC)

OK, those examples with над are in close junction with the following word, does град идёт also have [grad ɪ...]? Benwing2 (talk) 03:20, 23 November 2015 (UTC)

@Atitarev Please check the new examples in Module:ru-pron/testcases, thanks! Benwing2 (talk) 03:27, 23 November 2015 (UTC)

Yes and "град идёт" has another feature where "и" turns to "ы". It's either [ɡrat ɪˈdʲɵt] (slow) or [ɡrad‿ɨˈdʲɵt] (fast) cf. "в Ита́лии" - IPA^(key): [v‿ɨˈtalʲɪɪ]. --Anatoli T. ^{(обсудить}/^вклад) 03:33, 23 November 2015 (UTC)

Hmmm, what should we do, slow or fast? в Ита́лии works the way it does because the preposition в triggers close junction, which is indicated by ‿. It's not clear to me we want to do this for all words, although maybe we should. Benwing2 (talk) 03:40, 23 November 2015 (UTC)

Unfortunately, there are always variations when we talk about full sentences or phrases. I won't insist on this change ("град идёт") as long as cases "над Ита́лией" - IPA^(key): [nəd‿ɨˈtalʲɪ(j)ɪj] work, which they do (no devoicing and [ɨ]). Assume "град идёт" is working as expected, can be taken out. --Anatoli T. ^{(обсудить}/^вклад) 03:46, 23 November 2015 (UTC)

@Atitarev OK, what about the various other examples, including the ones before m and l and n, should they also be taken out? Benwing2 (talk) 03:48, 23 November 2015 (UTC)

• @Atitarev, Benwing2, please don't do it. Read Avanesov, §9.1.5.--Cinemantique (talk) 12:53, 23 November 2015 (UTC)

Do you think we went too far with consecutive assimilations? Both ways are correct, IMO, but change the test cases, if you insist. It'll make things simpler. --Anatoli T. ^{(обсудить}/^вклад) 22:46, 23 November 2015 (UTC)

@Atitarev, Cinemantique, Wikitiki89 The section from Avanosev says:

вместо звонкого согласного на конце слова произносится соответствующий глухой не только тогда, когда конец слова является в то же врем концом фрезы или части фразы, отделенной от следующей части паузой, но о многих случаях также и при слитном произношении данного слова с последующим, т. е. при отсутствии между ними паузы.

При этом глухой согласный вместо звонкого в конце данного слова обычно произносится не только перед глухим согласным следующего слова (например, хле́[п-х]оро́ший), но и в тех случаях, когда следующее слово начинается с гласного звука или одного из сонорный согласных -- [р], [л], [м], [н], [j], а также [в].

With help from Google Translate it's clear that Avanesov wants voiceless consonants even before sonorants, so I'll fix up the test cases accordingly.

BTW You can force close juncture by inserting the ‿ symbol between words. Hence град идёт [ɡrat ɪˈdʲɵt] but гра̀д‿идёт [ˌɡrad‿ɨˈdʲɵt]. Note how the д gets voiced and the и gets backed (and the first syllable will show up as [grəd] if you don't put some sort of accent on the word).

Also, fastpic.ru has rather interesting ads ... I got an interactive one with an Asian-looking woman (a "prisoner") in a red bikini and bondage wear, with the option to "strip" or "free" her. The US must be more prudish, or something, because you wouldn't see something like this on any mainstream photo-sharing site here. Benwing2 (talk) 07:21, 24 November 2015 (UTC)

Sorry, I've got into the habit of using this site. I don't look at such pictures.--Cinemantique (talk) 09:22, 24 November 2015 (UTC)

Moving the stress mark before consonant + soft sign

@Atitarev, Cinemantique, Wikitiki89 I implemented this change in my sandbox module. It changes the pronunciation of the following four test cases:

• Асунсьо́н: new pronun. [ɐsʊn⁽ʲ⁾ˈsʲjɵn]
• Амударья́: new pronun. [ɐmʊdɐˈrʲjæ]
• Вьентья́н: new pronun. [vʲjɪnʲˈtʲjæn]
• варьи́ровать: new pronun. [vɐˈrʲjirəvətʲ]

I think the new pronunciation looks OK for Асунсьо́н and Вьентья́н, where a consonant precedes the stress mark, but I'm not so sure about the other two, where a vowel precedes. What do you all think? Benwing2 (talk) 04:13, 23 November 2015 (UTC)

@Atitarev, Cinemantique, Wikitiki89 Any comments on this? Should it be [vɐˈrʲjirəvətʲ] or [vɐrʲˈjirəvətʲ]? Benwing2 (talk) 07:30, 24 November 2015 (UTC)

The first is good for me.--Cinemantique (talk) 09:24, 24 November 2015 (UTC)

Syllabification is not my forte but according to ru-wikt's syllable breakup, it's [ɐsʊn⁽ʲ⁾sʲˈjɵn], [ɐmʊdɐrʲˈjæ], [vʲjɪnʲˈtʲjæn] and [vɐrʲˈjirəvətʲ]. I know it's inconsistent.--Anatoli T. ^{(обсудить}/^вклад) 10:07, 24 November 2015 (UTC)

Hmmm. I considered moving the stress before consonant + soft sign only when a consonant precedes, but Wikitiki and Cinemantique seem to be OK with it coming before consonant + soft sign when a vowel precedes as well. It can always be changed later if we decide to do it differently. Benwing2 (talk) 10:12, 24 November 2015 (UTC)

I won't object if you do. --Anatoli T. ^{(обсудить}/^вклад) 10:31, 24 November 2015 (UTC)

I looked up Асунсьон and it actually says А·сунсь-о́н, which if taken literally would imply [ɐsʊn⁽ʲ⁾sʲjˈɵn], which surely can't be meant. Another data point: the 5th test case that was changed by this syllabification change was ничья́, and for this one ruwikt has ни-чья́, with pronunciation given of [nʲɪˈʨjæ], which is inconsistent with all their other results, so who knows. Benwing2 (talk) 10:46, 24 November 2015 (UTC)

Syllable-boundary rules

• There are many theories, the most recognized is Avanesov's theory, according to which Russian syllables (except initial and final) usually have "a rising sonority". What does it mean?

1. Unvoiced obstruents (к, х, п, ф, т, с, ш, ч, ц) have sonority level 1.
2. Voiced obstruents (г, б, в, д, з, ж) have sonority level 2.
3. Sonorants (й, л, м, н, р) have sonority level 3.
4. Vowels have sonority level 4.

• Sonority level rises in syllable. For example, коро́ва [kɐ.ˈro.və]: 14-34-24.
• Consonants usually moves to the next syllable (if sonority rises). For example, буква́льно [bʊ.ˈkva.lʲnə]: 24-124-334.
• Sonorants in front of obstruents don't move to the next syllable. For example, конве́рт [kɐn.ˈvʲert]: 14(3)-24(31).--Cinemantique (talk) 14:16, 26 November 2015 (UTC)

@Cinemantique Thanks for posting this. The principle of rising sonority that you mention is pretty standard cross-linguistically, and tends to determine which clusters are possible at the beginning and end of syllables, generally with falling sonority moving from the vowel towards both the beginning and end (Russian is unusual in allowing words like льда that violate the sonority principle). However, we need more information to properly determine where syllable boundaries fall, e.g. we might want to make a difference between ка-ста and ках-та even though all the sounds in both words are unvoiced obstruents. Furthermore, I think a word like буква́льно should be divided бук-ва́ль-но instead of бу-ква-льно. The rule currently used by the module is to put the syllable boundary before the last consonant but move it leftwards to keep certain clusters (e.g. пл, ст) together. I added soft cons + /j/ to the list of such clusters. Benwing2 (talk) 08:18, 30 November 2015 (UTC)

This is why a preceding word that ends in a vowel will steal the [lʲ-] from льда. Thus, ѝзо льда́ = [ˌi.zɐlʲˈda] and not [ˌi.zɐˈlʲda]. --WikiTiki89 18:39, 30 November 2015 (UTC)

Vocalization in front of в

Unvoiced consonant at the of word becomes voiced if the next word begins with в + voiced obstruent (вг, вб, вд, вз, вж). See tests 'ваш взор', 'от взгля́дов'.--Cinemantique (talk) 13:25, 23 November 2015 (UTC)

I have implemented this in my sandbox (it will be pushed soon). Benwing2 (talk) 07:30, 24 November 2015 (UTC)

It is pushed. Benwing2 (talk) 10:13, 24 November 2015 (UTC)

@Benwing2 "муж Во́вы" (Vova's husband) is a funny example, since "Вова" is a boy :) "муж Ва́ли" would be better but I am not homophobic. In any case, while "ш" doesn't become voiced in such positions but "ж" would retain its voiced quality. --Anatoli T. ^{(обсудить}/^вклад) 04:40, 26 November 2015 (UTC)

OK, Cinemantique seems to disagree because he put this test example in and so I changed the module accordingly. Avanesov seems to say that ж becomes unvoiced before в; is this another case where it depends on how fast it's pronounced? Benwing2 (talk) 04:56, 26 November 2015 (UTC)

Yes but that can be said about в + voiced obstruents as well. Devoicing will happen in a slower speed. (Муж вдовы - widow's husband is strange as well).--Anatoli T. ^{(обсудить}/^вклад) 04:59, 26 November 2015 (UTC)

I would pronounce муж Вали with [-ʂ-]. --WikiTiki89 18:41, 30 November 2015 (UTC)

Both pronunciations are acceptable. The voiced pronunciation is for the fast speech but it's natural as well. --Anatoli T. ^{(обсудить}/^вклад) 23:53, 30 November 2015 (UTC)

comments from Wanjuscha

@Atitarev, Cinemantique, Wikitiki89 @Wanjuscha I've moved your comment here because this is where most discussion of Russian pronunciation is happening.

On the page cтанок-качалка pronunciation looks like [t͡stɐˈnok kɐˈt͡ɕalkə]. Why there's ts in the beginning? On the page подзатыльник pronunciation looks like [pəd͡zzɐˈtɨlʲnʲɪk]. Why there's double z? --Wanjuscha (talk) 06:24, 27 November 2015 (UTC)

The issue with cтанок-качалка is that there is a mistake in the title and elsewhere -- a Latin c in place of Cyrillic с, which gets rendered as [t͡s]. I've fixed it. @KoreanQuoter Please be careful with Latin vs Cyrillic. Also, check out the proper usage of {{ru-decl-noun-see}} on the corrected page станок-качалка (stanok-kačalka).

The [d͡zz] is intentional, where д and з are pronounced separately because of the morpheme boundary between под- and затыльник, and д assimilates to the following з, producing [d͡z]. This is per Cinemantique. Benwing2 (talk) 06:46, 27 November 2015 (UTC)

(edit conflict) @Wanjuscha The first entry was misspelled, moved to станок-качалка. In the 2nd entry there's an assimilation happening. Do you disagree?--Anatoli T. ^{(обсудить}/^вклад) 06:44, 27 November 2015 (UTC)

компа̀кт-ди́ск

@Atitarev What is the rule here that causes there not to be voicing assimilation in the final кт before д? Benwing2 (talk) 04:49, 1 December 2015 (UTC)

Unfortunately, I don't know but [-gd-] sounds weird. Maybe because there are too many consecutive consonants or maybe User:Cinemantique can dig something out. --Anatoli T. ^{(обсудить}/^вклад) 04:59, 1 December 2015 (UTC)

I think [gd] is OK here, if we pronounce it quickly, as one word. You can here it on RNC.--Cinemantique (talk) 06:40, 1 December 2015 (UTC)

Thanks. I'll leave things as-is, then. Benwing2 (talk) 00:40, 2 December 2015 (UTC)

As a side comment, English has no assimilation in consonant clusters. duck-pack, duck-back, dug-pack and dug-back all sound different. Benwing2 (talk) 00:42, 2 December 2015 (UTC)

What you mean is English has no voicing assimilation across word boundaries (except in a few rare cases that have essentially become lexicalized, like have to and used to). It does have voicing assimilation in other places, and it does have other types of assimilation across word boundaries. --WikiTiki89 00:55, 2 December 2015 (UTC)

Right. There's of course voicing assimilation involving final -s, but this is arguably morphological more than phonological. And e.g. there's a tendency to assimilate [s],[z] to a following [ʃ], so is sure tends to sound like [ɪʒ ʃɚ] when spoken casually; but AFAIK when you do X-ray scans of assimilations like this, you see that the assimilation isn't complete -- the tongue still makes the [s] shape, but this is overlaid by the dome of the [ʃ] shape. So this isn't really phonological assimilation either. Benwing2 (talk) 03:34, 2 December 2015 (UTC)

How about "Yesh you do?" That's pretty complete. Also, Don't forget that there are many dialects that do things differently. Anyway, incomplete assimilation is still phonological, what else would it be? --WikiTiki89 16:07, 2 December 2015 (UTC)

You're right about "Yesh you do", similarly "gotcha", etc. By "phonological" I mean it's not just the inevitable result of articulatory overlap. Benwing2 (talk) 07:08, 3 December 2015 (UTC)

No result is "inevitable". It's perfectly possible for these partial assimilations not to happen and in some languages they don't. --WikiTiki89 15:39, 3 December 2015 (UTC)

Well, some languages separately articulate the two consonants in a cluster. I think Russian does that for example in a number of word-initial clusters -- I recall hearing what sounded like a short schwa between consonants that were written without any vowel between them. Benwing2 (talk) 04:12, 4 December 2015 (UTC)

syllabification changes

@Atitarev, Cinemantique, Wikitiki89 The syllabification mechanism distinguishes between clusters like kt, rn, etc., which are broken ak/ta, ar/na, etc. and pl, str, etc., which are broken a/stra, a/pla, etc. Formerly it was somewhat inconsistent, e.g. it grouped together str but not spr or skr, and gl but not gr, etc. I fixed this. A few questions:

1. I removed tl and dl, so that e.g. what was formerly а/тла́с is now ат/ла́с. Is that reasonable?
2. I kept ml and mn from before, so we get e.g. зе/мля́, о/мле́т, се/мна́дцатый. This seems a bit strange to me, do you think I should remove them? If they should be kept, should we add мр (hence изу/мру́д) or вн (hence призы/вни́к)?
3. I kept sc from before, so we get e.g. аб/сци́зовый, ква/сцы́. This also seems strange, should I remove it?
4. I kept xr and xl from before, so we get e.g. со/хра́нность. This seems reasonable, do you agree?
5. I added vr and vl to parallel fr and fl, so we get e.g. я/вля́ть, та/вро́. This seems reasonable, do you agree?

Benwing2 (talk) 01:36, 8 December 2015 (UTC)

Palatalization

@Benwing2, [s] and [z] in front of [nʲ] or [tʲ] should be palatal (снег, стирать). Palatalization is optional if they're preceded by [l] or [r] (толстяк, оползни). See Avanesov's orthoepic dictionary, §88-90. @Atitarev, could you confirm?--Cinemantique (talk) 23:18, 29 December 2015 (UTC)

I think the prepositions and prefixes, such as с and с-, can optionally remain unpalatalized before [nʲ] and [tʲ]. Does Avanesov mention this? --WikiTiki89 23:22, 29 December 2015 (UTC)

According to the orthoepic dictionary, [s] in снести, снизу, стекать, стемнеть, etc. should be palatal. It's fun that palatalization in сникерс is optional.--Cinemantique (talk) 23:40, 29 December 2015 (UTC)

Hmmm... Out of those, I pronounce снизу only unpalatalized (with the meaning "from below", not necessarily so with "I will lower"), and the other three both ways (with стекать almost always palatalized, but not necessarily so with стёк). --WikiTiki89 00:48, 30 December 2015 (UTC)

(Before E/C) Confirmed. (BTW, Belarusian features the same, also in front of palatal "ць" and the older Tarashkevitsa spelling, which is closer to pronunciation than the modern, russified one, uses "ь" in words such as сьнег, радасьць - cf. with modern Belarusian снег (snjeh), ра́дасць (rádascʹ)). Russian ра́дость (rádostʹ) is handled correctly - [ˈradəsʲtʲ] but снег (sneg) needs to make [ʲ] mandatory. --Anatoli T. ^{(обсудить}/^вклад) 23:29, 29 December 2015 (UTC)

@Atitarev, Cinemantique, Wikitiki89 I've implemented this, and added some test cases (and fixed up the test cases for final -е and syllabification fixes). Benwing2 (talk) 01:04, 30 December 2015 (UTC)

@Atitarev, Cinemantique, Wikitiki89 BTW, I just added the ability to force optional palatalization by writing (ь), as in с(ь)ни́керс. To force no palatalization (and in general to suppress assimilation of all sorts), put an underscore between the consonants. Benwing2 (talk) 01:17, 30 December 2015 (UTC)

@Benwing2, thank you! Can you make it for пьянка, скамья, etc.? I mean "скам(ь)я".--Cinemantique (talk) 05:14, 31 December 2015 (UTC)

I think you want скам(ь)я to have optional palatalization. I thought this already worked, but I will get it working. Benwing2 (talk) 05:20, 31 December 2015 (UTC)

Final -е when followed by another word

@Atitarev, Cinemantique, Wikitiki89 I added fixes for final -е when followed by another word and not by a pause, so that it generates [ɪ] instead of [e]. A couple of questions:

1. Does this happen always, or only for the adjectival endings -ие and -ые? I made it happen always, so it also happens in e.g. по́ле есть.
2. Why is there no optional (j) in this case ([sinɪɪ] not [sinɪ(j)ɪ])? Under what circumstances does this lack of optional (j) exist? It seems to occur in -ие and -ые; does it also occur in -ее (e.g. бо́лее всего́), in -е́е (e.g. скоре́е всего́), in -ое (кое-кто), in -те (да́йте мне), ...?

Benwing2 (talk) 02:34, 30 December 2015 (UTC)

1. It doesn't happen with -ее/-ое (neuter) adjective endings, IMO, but with plural. I think (with neuters) it's either a clear [-e] or [-ə], as in абстрактное существительное in both final positions.
2. In adjectival endings -ие and -ые before other words it's [-ɪ] without [j] or a clear [-je]. In the beginning and in the middle of the word we can leave the optional [j], also in кое- words. In бо́лее and скоре́е the final "-е" is [-je] or (less commonly) [-jə], [j] is pronounced.

Some notes about words like Япония, Европа, японский, европейский and others :

в Япо́нии, в Евро́пе - [j] is not optional, always pronounced

о Япо́нии, о Евро́пе - [j] is optional

по-япо́нски, по-европейски - [j] is optional

за́яц, ро́ет (рыть), но́ет (ныть) - [j] is optional. --Anatoli T. ^{(обсудить}/^вклад) 03:11, 30 December 2015 (UTC)

@Atitarev OK, so [-ɪ] in non-final words doesn't occur in neuter -ее/-ое, but I'm not completely clear about whether you get [-ɪ] in по́ле есть, бо́лее всего́, скоре́е всего́, or да́йте мне. Benwing2 (talk) 03:23, 30 December 2015 (UTC)

"по́ле бо́я" is pronounced "по́ли бо́я", "да́йте мне" -> "да́йти мне" but in "по́ле е́сть" there's probably a micropause but otherwise yes, also "по́ли". --Anatoli T. ^{(обсудить}/^вклад) 03:32, 30 December 2015 (UTC)

@Atitarev From what I remember of your pronunciation, you do get [-ɪ] in Четыре всадника Апокалипсиса, right? Benwing2 (talk) 03:27, 30 December 2015 (UTC)

Yes. --Anatoli T. ^{(обсудить}/^вклад) 03:32, 30 December 2015 (UTC)

BTW the module currently gets в Япо́нии, в Евро́пе wrong, will have to fix. Benwing2 (talk) 03:29, 30 December 2015 (UTC)

I knew it doesn't cover such cases. Thanks! --Anatoli T. ^{(обсудить}/^вклад) 03:32, 30 December 2015 (UTC)

@Atitarev OK, it looks like we need to special-case this. Can you help me by letting me know, for each of the following words, whether the pronunciation [-ɪ] occurs when followed by another word without a pause? I filled in some of them per the above discussion. (Sorry I can't come up with examples for all of them, my Russian isn't good enough ...) Benwing2 (talk) 03:48, 30 December 2015 (UTC)

1. моро́женое (no)
2. собра́ние (no)
3. всле́дствие (no)
4. сча́стье (no)
5. со́лнце (no)
6. се́рдце (no)
7. ло́же (no)
8. по́ле (yes)
9. жили́ще (no)
10. ве́че (no)
11. а́вторские (yes)
12. да́нные (yes, as pl adjective prob not as a noun) (?) Not 100% sure

1. да́нные ли́ца - yes (these, given persons)
2. да́нные студе́нтов - no (students' data)

1. си́нее (no)
2. ли́сье (no) (?)
3. си́не (short neuter adj)
4. си́не (prefix, сине-зелёный - yes)
5. дурне́е (no)
6. ле́гче - yes
7. гу́ще- yes
8. досе́ле - yes
9. ина́че - yes
10. злове́ще - yes
11. не пла́чьте (yes)
12. дава́йте (yes)
13. дво́е (yes?)?
14. четы́ре (yes)
15. о го́де - yes
16. о пла́че - yes
17. о бо́е - yes
18. об и́нее (no?)?
19. ста́рче ? (vocative, hardly followed by other words)
20. сы́не ? (vocative, hardly followed by other words)

Done. I did what I could. I've added some question marks where I am not so sure. @Cinemantique I could use some help here. --Anatoli T. ^{(обсудить}/^вклад) 04:31, 30 December 2015 (UTC)

I've messaged Cinemantique on ru:wikt. I'm not entirely sure with my answers and I only used my intuition. --Anatoli T. ^{(обсудить}/^вклад) 04:39, 30 December 2015 (UTC)

@Atitarev Thanks very much! си́не is intended to be a short neuter adjective form. I guess these normally occur in the predicate and probably aren't followed by other words. сы́не is intended as a vocative, although you're right in this case it hardly is followed by other words. Benwing2 (talk) 05:46, 30 December 2015 (UTC)

• I don't know about such effect. -ие and -ые in adjectives (also substantivized) are pronounced [ɪɪ] (or [ˈiɪ]) and [ɨɪ] (синие like синии, большие like большыи). See Avanesov, §16.4, page 198.--Cinemantique (talk) 05:53, 30 December 2015 (UTC)

Thanks for the link. This is all very confusing. On top of that, Avanesov seems to use different vowel symbols in each of his books! Note that most of the examples Avanesov gives are in adjectives preceding other words. Anatoli, in your pronunciations of plural adjectives standing alone or substantivized, do you have [-e]? Benwing2 (talk) 06:14, 30 December 2015 (UTC)

The examples Avanesov gives, are all collocations. Yes, they are pronounced as -ыи/-ии in front of other words. You can listen to those songs I linked you, the collocations are repeated several times. As stand-alones or in front of a pause, I pronounce these endings without a reduction. If you check our recordings at Wiktionary you may find confirmation. (Russians tend to pronounce "-е" endings without a reduction when asked to pronounce them but it's different in collocations, in fast, natural speech). Not sure if Avanesov's confusing notation means the same thing, he mentions that (casually), "BTW, in front of a pause ...". Also, he mentions that -ыя/-ия (jə) pronunciations are now dated. Yes, that's my understanding too.

Further he mentions that "новая" and "новое" are pronounced identically - [-jə]. I agree only partially. Yes, in fast casual speech. In slow, careful speech "новое" will have /-je/ (I think he mentioned that as well in some other places, no longer sure. --Anatoli T. ^{(обсудить}/^вклад) 09:10, 30 December 2015 (UTC)

Somewhere else Avanesov does say that -ое has the [-je] pronunciation and that [-jə] is "old Moscow". This may be incorrect. For the moment I'm going to keep it at [-jə] but at some point soon I'll implement multiple styles for final -е and then I'll try to put in "old Moscow" pronunciations ("slightly dated" is how you've rendered it), and I'll also put in that -ое is [-je] in slow speech but [-jə] in fast speech, if you think this is reasonable. Benwing2 (talk) 09:29, 30 December 2015 (UTC)

If you're going to work on variant pronunciations, it's going to be a tremendous task. E.g. there are some noticeable pronunciation differences in South Russia (also Ukraine, Belarus) where "г" is pronounced "the Ukrainian way" (or similar), people with this pronunciation live all over Russia, perhaps 20% of speakers or more. Some northern oblasts, mainly rural almost lack "akanye". The percentage of these is smaller and dwindling. Both these accents are considered dialectal/non-standard but happen in movies as well. There's also "stage accent" where adj ending like кий is pronounced [-əj], высокий - [vɨˈsokəj]. Some opera singers practice it. Priests have their own peculiarities. Although Russian is surprisingly homogeneous for its size and the territory, there are nuances. --Anatoli T. ^{(обсудить}/^вклад) 09:42, 30 December 2015 (UTC)

We can definitely keep "новая" and "новое" identically for collocations (when something follows after). As you have noticed and implemented, stand-alones are pronounced with clearer endings. --Anatoli T. ^{(обсудить}/^вклад) 09:47, 30 December 2015 (UTC)

I think we can leave [-je] as it is in between words, there's no need to overdo every detail. Also, I really don't think that this so-called "old Moscow" pronunciation is actually dated, at least with regard to the [-jə] (on the other hand, pronouncing -кий as [-kəj] is certainly no longer normal in Moscow). --WikiTiki89 16:26, 30 December 2015 (UTC)

@Atitarev My intention isn't to add non-standard or dialectal pronunciations but to try to get standard pronunciation, maybe including fast/slow variants if such things exist and are really part of the standard. As for "Old Moscow", you already included some examples of that e.g. in поле, and Wikitiki thinks this is still pretty standard, and Avanesov mentions it, so it may be includable. Benwing2 (talk) 18:01, 30 December 2015 (UTC)

@Wikitiki89 The pronunciation between words seems like something that's important enough to include; at least, I heard it clearly in Anatoli's word samples. Benwing2 (talk) 18:01, 30 December 2015 (UTC)

You heard it because it is exists, but its existence doesn't necessarily make it important enough to include. We don't need to include every phonetic detail. --WikiTiki89 18:17, 30 December 2015 (UTC)

In this case it seems wrong not to include it because it's a case where you have phonemic merging. Benwing2 (talk) 19:04, 30 December 2015 (UTC)

Same with latter and ladder. The problem is the phonemic merger does not always happen, or can even go in different directions depending on the accent. --WikiTiki89 19:30, 30 December 2015 (UTC)

If Benwing2 has some strong interest, we should not prevent him doing it, even if it's sometimes not easy to answer all his questions. Now, with long adjectival endings ending in -е, we can conclude:

1. For -ые/-ие: 1. stand-alone end in [-je] (careful and slow) or [-ɪ] (fast and casual), 2. when something follows without a pause - [-ɪ]. In a very careful speech [-je] is also correct for collocations. I think it's OK to default to [-ɪ]
2. For -ое/-ее: 1. stand-alone end in [-je] (careful and slow) or [-ə] (fast and casual), 2. when something follows without a pause - [-ə]. In a very careful speech [-je] is also correct for collocations. I think it's OK to default to [-ə]

Regarding the adjectival -кий/-гий endings pronounced as [-kəj]/[-gəj]. I didn't mean it belonged to old or current Moscow pronunciation. I just mentioned it as a former semi-standard for formal singers. You can also hear this pronunciation in folklore read out loud in the media. --Anatoli T. ^{(обсудить}/^вклад) 11:38, 1 January 2016 (UTC)

But it did belong to the Old Moscow pronunciation. I wasn't contradicting you, but just making a separate point that that part of the Old Moscow pronunciation has mostly died out, while the [-ə] pronunciations have not. --WikiTiki89 16:21, 4 January 2016 (UTC)

Adjectival and что transformations are now by default, and adj= param is instead noadj= and noshto=

(see Module talk:ru-translit for discussions leading up to this)

@Atitarev, Cinemantique, Wikitiki89 Adjectival -го transformations and что transformations are turned on by default in transliteration in general as well as in {{ru-IPA}}. The former {{ru-IPA}} adj= parameter is gone, replaced by noadj= and noshto= to turn off these transformations. Both of these transformations, as well as the лёгкий/мягкий transformation, are displayed using phonetic respelling. Benwing2 (talk) 09:35, 16 January 2016 (UTC)

IPA^(key): [nʲɪʂˈto] (phonetic respelling: ништо́)

IPA^(key): [ˈʂto ˌlʲibə] (phonetic respelling: што́-либо)

IPA^(key): [nʲɪkɐˈvo] (phonetic respelling: никово́)

IPA^(key): [sʲɪˈvodʲnʲə] (phonetic respelling: сево́дня)

IPA^(key): [ˈlʲeɡə]

IPA^(key): [ˈmnoɡə]

IPA^(key): [ˈsoɡə]

IPA^(key): [ˈnʲet͡ɕtə]

IPA^(key): [ˈʂtobɨ] (phonetic respelling: што́бы)

IPA^(key): [nʲɪt͡ɕˈtoʐnɨj]

PASS. Thank you!--Anatoli T. ^{(обсудить}/^вклад) 09:42, 16 January 2016 (UTC)

You're welcome! Benwing2 (talk) 09:54, 16 January 2016 (UTC)

One question: Should words that have transformations applied to them (-го words, что and friends, лёгкий/мягкий and friends) go into Category:Russian terms with irregular pronunciations? Benwing2 (talk) 10:09, 16 January 2016 (UTC)

Yes. All words with е=э, г=в, ч=ш, г=х, and г=[ɣ] belong there. --Anatoli T. ^{(обсудить}/^вклад) 10:40, 16 January 2016 (UTC)

молодца and дц/тц not post-stress

(moved from Talk:молодца)

• @Benwing2, Wikitiki89, Cinemantique I don't know the rule here but I know it's geminated. Still a bit concerned that when a bot starts adding pronunciations to inflected forms, words like отцы́ (otcý) will get an incorrect IPA. Is Cinemantique still here with Avanesov? Not sure about Avanesov but my gut feeling tells me (again) that тц and дц should always be geminated. --Anatoli T. ^{(обсудить}/^вклад) 06:51, 22 January 2016 (UTC)

• I agree, see Avanesov's dictionary, §125. It's strange but отца, отцы, отцепить, подцепить, двадцать look good.--Cinemantique (talk) 10:15, 22 January 2016 (UTC)

• The third and the fourth have prefixes, the last one immediately follows a stressed syllable. Both rules are implemented. Thanks for the link. My assumption was right. Дц and тц are always geminated. --Anatoli T. ^{(обсудить}/^вклад) 10:22, 22 January 2016 (UTC)

• Yes, "отцы́" gives [ɐˈt͡sːɨ], which is right but not sure why "молодца́" doesn't produce a gemination. --Anatoli T. ^{(обсудить}/^вклад) 11:38, 22 January 2016 (UTC)

• It may sound contradicting but endings -тся/-ться should have optional geminations when not immediately following a stress. Currently they are ungeminated.--Anatoli T. ^{(обсудить}/^вклад) 06:55, 22 January 2016 (UTC)

• Avanesov says (§126 above) -ться, -тся are always geminated. I'm not sure.--Cinemantique (talk) 10:38, 22 January 2016 (UTC)
  • I'm OK with that. The geminations might be very light sometimes, so it's hard to be sure 100%. This could be implemented as well. --Anatoli T. ^{(обсудить}/^вклад) 11:38, 22 January 2016 (UTC)

@Atitarev, Cinemantique, Wikitiki89 I fixed the module so молодца́ is geminated by default. отцы́ was geminated because от- looks like a prefix. Note that this change also changes 11-ый, четырнадцать, четырнадцатый, which may need gem=n. (What about тридцать, двадцать, etc., these currently are geminated, is this correct?) Benwing2 (talk) 18:46, 22 January 2016 (UTC)

Also, the change I made also makes тч, дч remain geminated everywhere by default. Is this correct? And what about дьт or тьд? Should these be geminated everywhere? Benwing2 (talk) 18:46, 22 January 2016 (UTC)

BTW do you want me to make -ться, -тся not following a stress be optionally geminated or always geminated? Benwing2 (talk) 19:05, 22 January 2016 (UTC)

@Cinemantique Could you double-check if numerals like четырнадцать, even тридцать should have a mandatory gemination? To me, it seems it should be at least optional.

OK to always geminate тч, дч.

дьт or тьд - Not 100% sure about дьт or тьд, maybe no, if you mean пятьдесят.

As for -ться, -тся not following a stress I thought of making them optional but according to Avanesov, they are always geminated. As a minimum, they should be optional, currently not geminated. If Cinemantique insists, we can make them mandatory.

It seems safer to make дц and тц always geminate immediately before a stress (but optional for тридцать and двадцать) or in any position after a stress (молодца́ ) and optional after a stress but not immediately, including endings -ться and -тся. --Anatoli T. ^{(обсудить}/^вклад) 00:24, 23 January 2016 (UTC)

Clarified something with Avanesov reference. He considers ungeminated pronunciation of numerals with дц like двадцать non-standard. I would say, even if dropping geminations here is non-standard, it's quite common. --Anatoli T. ^{(обсудить}/^вклад) 02:06, 23 January 2016 (UTC)

Is it optional in мо́лодца?--Cinemantique (talk) 08:05, 23 January 2016 (UTC)

No. --Anatoli T. ^{(обсудить}/^вклад) 09:13, 23 January 2016 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ @Atitarev Re "It seems safer" ...: I'm confused about what you say. Keep in mind first that дц/тц are distinct cases from -т(ь)ся, no need to combine them. As for дц/тц, can you restate what you want? What you say above seems to have "before" and "after" mixed up. Benwing2 (talk) 12:00, 23 January 2016 (UTC)

Yes, I sorry, I have mixed them up.

You have already implemented mandatory geminations for дц and тц. It's fine but IMO, gem=opt and gem=n should still work, e.g. двадцать, тридцать should be optional. I might mark these as colloquial, if Avanesov thinks that the lack of gemination is non-standard here.

Please make -ться and -тся geminated as well, in any position. (It seems geminations are very light when the ending is further away from the stressed syllable.)--Anatoli T. ^{(обсудить}/^вклад) 12:13, 23 January 2016 (UTC)

@Atitarev gem=opt and gem=n do work with дц, тц. I'll make -ться, -тся geminated everywhere. (If the gemination is very light, maybe we should mark it as optional?) Benwing2 (talk) 20:05, 23 January 2016 (UTC)

@Atitarev I've made двадцать, тридцать and derived numbers have optional gemination. What about двадцатый, тридцатый and forms of those adjectives? What about одиннадцать, двенадцать, четырнадцать etc. and derived ordinals четырнадцатый etc.? Currently одиннадцать and одиннадцатый have gem=n, двенадцать and others don't so they're currently marked with mandatory gemination. Benwing2 (talk) 22:41, 23 January 2016 (UTC)

@Benwing2 Re: -ться, -тся - optional is fine when it's not immediately after a stress.

The optional gemination would apply to ALL numerals on -дцать and -дцатый (also одиннадцать, двадцатый, etc.), also to inflected forms when stress positions change, e.g. "двадцати́". --Anatoli T. ^{(обсудить}/^вклад) 07:00, 24 January 2016 (UTC)

@Atitarev I'm thinking of adding a rule to make дц optionally geminated in -адцат- and –идцат- (or maybe even just -дцат-). This should take care of all the numerals plus their ordinals, non-lemma forms and ordinal non-lemma forms, without the need to do all this manually. What do you think? I notice you edited many of them manually but there are still many more, e.g. a bunch of non-lemma forms of пятнадцатый, семнадцатый and тринадцатый, plus полдвенадцатого, тридцатьчетвёрка, etc. A grep for дцат through the list of words produces only numbers and number-related words; can you think of any other words with дцат, адцат or идцат? Benwing2 (talk) 19:59, 25 January 2016 (UTC)

"дцат" only occurs in numerals, you can use that. Please note that this is my observation without a good support by references. Avanesov called these pronunciations non-standard but I disagree.--Anatoli T. ^{(обсудить}/^вклад) 20:39, 25 January 2016 (UTC)

-э in То́гане et al. is pronounced clearly

(moved from Talk:Тогане)

@Benwing2 My gut feeling tells me that final unstressed "-э" is usually unreduced as well, just like "е", especially in isolation, in a clear and slow speech. --Anatoli T. ^{(обсудить}/^вклад) 09:14, 20 January 2016 (UTC)

@Atitarev, Cinemantique, Wikitiki89 Do you agree, Wikitiki and Cinemantique? Benwing2 (talk) 17:00, 21 January 2016 (UTC)

Since it's always foreign, I guess it might behave like the "о" in ра́дио (rádio), but I don't know if I can make a generalization. --WikiTiki89 17:47, 21 January 2016 (UTC)

Yes, in the final position after consonants with palatal/unpalatal pairs, they are always foreign and I thought about other words as well, which are pronounced the same way, eg Кобе, Хакодате. They words like солнце, ложе are in a different category. --Anatoli T. ^{(обсудить}/^вклад) 20:44, 21 January 2016 (UTC)

I can distinguish the two because ц and ж are unpaired hard consonants. Benwing2 (talk) 20:55, 21 January 2016 (UTC)

OK, thanks. the case for unreduced "-э" is stronger than "е", since they are, indeed, all loanwords and don't easily make pairs (e.g. "поле боя" - "battlefield"), where the pronunciation is reduced. --Anatoli T. ^{(обсудить}/^вклад) 21:49, 21 January 2016 (UTC)

@Atitarev, Benwing2 I believe there must be a standardized guideline of pronuncing Russian words that came from Japanese. --KoreanQuoter (talk) 08:09, 22 January 2016 (UTC)

The only Japanese thing that may affect the Russian pronunciation in some cases is the pitch accent, the way it's perceived by Russians. Otherwise, the pronunciation just follows basic rules for loanwords.--Anatoli T. ^{(обсудить}/^вклад) 08:16, 22 January 2016 (UTC)

ɪ̯

What do you think about replacement of [j] with [ɪ̯] in some positions? According to Avanesov (Russian Standard Pronunciation, 1984, §7.8, p. 121), [j] is pronounced in front of stressed vowel: я́корь, пою́, чья; [ɪ̯] is pronounced (instead of [j]):

• in the beginning of words in front of unstressed vowel: Юпи́тер, ясна́, едва́;
• after vowel in front of unstressed vowel: мо́ю, зна́ю, уме́ю, поясню́;
• after consonant in front of unstressed vowel ([j] could be pronounced in the clear speech): бра́тья, коло́сьев;
• after vowel in front of consonant: ма́йка, война́;
• at the end of words: мой, чай, кий.--Cinemantique (talk) 10:18, 1 February 2016 (UTC)

It's up to Benwing2 but I'm fine with the way it is, using allophones, otherwise we start using more symbols - [ə̥, ., o̞, ɫ] and similar notation, which will make maintenance for IPA more complex. --Anatoli T. ^{(обсудить}/^вклад) 10:29, 1 February 2016 (UTC)

I'm leaning towards keeping [j]. [ɪ̯] is an uncommon IPA symbol and I doubt very many people will have any idea how to pronounce it, esp. before a vowel (for that matter, I wouldn't know what the difference is, either). Benwing2 (talk) 12:07, 1 February 2016 (UTC)

[ɪ̯] is not so uncommon, we use it in English after all (as the second part of diphthongs). The difference is essentially how close your tongue gets to your palate. But I agree with Anatoli that there is no need to complicate our transcription any further. --WikiTiki89 16:26, 8 February 2016 (UTC)

Right, I get how [ɪ̯] is pronounced in diphthongs like [aɪ̯], but what does it mean to say it's pronounced in words like Юпи́тер or бра́тья? Benwing2 (talk) 16:43, 8 February 2016 (UTC)

(@Wikitiki89) Benwing2 (talk) 16:43, 8 February 2016 (UTC)

It just means the flat part of the tongue does not get as close to the palate for [ɪ̯] as it would for [j]. The same difference as between the vowels [ɪ] and [i]. --WikiTiki89 16:50, 8 February 2016 (UTC)

Geminates in non-lemma forms

@Atitarev, Cinemantique, Wikitiki89 I'm implementing the following rules in my bot:

1. If a lemma form has a final geminate, add gem=n to non-lemma forms, which typically have the geminate in the middle of a word; otherwise it will typically be pronounced geminated. For example, lemmas like алема́нн and non-lemma forms алема́нна, алема́нны, etc, or lemma абсце́сс and non-lemma forms абсце́сса, абсце́ссы, etc.
2. If a lemma has gem=opt or gem=y with a geminate in the middle of a word and a non-lemma form causes that geminate to end up at the end of a word, remove gem=opt or gem=y. For example, lemma=а́белева гру́ппа, non-lemma form а́белевых гру́пп. Otherwise the pronunciation of the non-lemma form will have a final geminate indicated in it.

Are these both correct? I'm especially unsure about #1. Benwing2 (talk) 21:20, 7 February 2016 (UTC)

No gemination in both cases. There are some exceptions, like вилла, вилл - both geminated.--Anatoli T. ^{(обсудить}/^вклад) 21:26, 7 February 2016 (UTC)

Hmm, in the second case, other examples in лл with gemination are булла/булл, горилла/горилл, мулла/мулл, новелла/новелл, шиншилла/шиншилл. Do any of these have final geminates? Benwing2 (talk) 21:46, 7 February 2016 (UTC)

And what about words in мм, e.g. анаграмма/анаграмм and others in -грамма, аналемма/аналемм, гамма/гамм, лемма/лемм, сумма/сумм? Benwing2 (talk) 21:52, 7 February 2016 (UTC)

No, вилла, вилл is a special case. Final consonants are seldom (almost never) geminated.--Anatoli T. ^{(обсудить}/^вклад)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ @Atitarev By the above rules, пре́ссы as a form of пре́сса should be geminated but пре́ссы as a form of пресс should not be geminated. Is this correct? Benwing2 (talk) 22:55, 7 February 2016 (UTC)

It's possibly a different rule here but yes, inflected forms of пресс (an other such words) are ungeminated. --Anatoli T. ^{(обсудить}/^вклад) 23:06, 7 February 2016 (UTC)

дождь, words in жж/зж and ӂӂ pronunciations

(moved from Talk:дождь)

@Atitarev, Cinemantique, Wikitiki89 The pronunciation дощ is possible; do the non-lemma forms also have possible pronunciations like доӂӂя́? I also notice that дождик has a pronunciation до́ӂӂик, do дождевик, дождевой, дождевой червь, дождливый have similar pronunciations with ӂӂ? Benwing2 (talk) 18:07, 9 February 2016 (UTC)

Yes, the traditional Moscow pronunciation was до́щ and доӂӂа́, and my grandmother still pronounces it that way. I don't know about до́ӂӂик, or the other words you list, but we don't necessarily have to include this pronunciation for all of them, especially if it cannot be verified. --WikiTiki89 18:18, 9 February 2016 (UTC)

@Benwing2 Correct for inflected forms and до́ӂӂик but not for any derivations you've listed.--Anatoli T. ^{(обсудить}/^вклад) 19:38, 9 February 2016 (UTC)

All derivations have a variant with [ʑː], too.--Cinemantique (talk) 11:13, 10 February 2016 (UTC)

@Cinemantique Thanks! This is Avanesov? Do you have a digital copy by any chance? Or is it available online somewhere? Benwing2 (talk) 16:44, 10 February 2016 (UTC)

Yes, it's Avanesov's orthoepic dictionary. Its digital copy is here.--Cinemantique (talk) 05:34, 11 February 2016 (UTC)

If you don't get it, please remind me, I've got a copy. I think I already sent you, no? Apologies for confusion. I don't object to [ʑː] to those. I don't use this pronunciation on дождевой, etc. and it's even less common. --Anatoli T. ^{(обсудить}/^вклад) 22:01, 10 February 2016 (UTC)

You sent me a copy of Zaliznyak, I don't remember getting Avanesov's book. Benwing2 (talk) 22:05, 10 February 2016 (UTC)

OK, I'll email it to you. --Anatoli T. ^{(обсудить}/^вклад) 22:18, 10 February 2016 (UTC)

@Cinemantique, Atitarev Also:

• How are бзжу, набзжу, пизжу, спизжу pronounced? Currently they're indicated as if бжу, набзжу, пижжу, спижжу. Do they have a ӂӂ pronunciation?
• Could you indicate for the following words: (1) do they have a ӂӂ pronunciation? (2) should it go first or second?

1. брызжу, брызжешь etc. of брызгать - y
2. изжога and non-lemma forms - n
3. можжевельник and non-lemma forms - y
4. мозжечок and non-lemma forms - y
5. мозжечко́вая минда́лина and non-lemma forms- y

• Thanks! Benwing2 (talk) 19:40, 10 February 2016 (UTC)

IMO, again but let Avanesov override if he found that these pronunciations exist. --Anatoli T. ^{(обсудить}/^вклад) 22:01, 10 February 2016 (UTC)

Thanks again. Benwing2 (talk) 22:05, 10 February 2016 (UTC)

BTW, I'm going to put the ӂӂ pronunciations of брызжу, брызжешь etc. second. What about бзжу, набзжу, пизжу, спизжу? Benwing2 (talk) 22:08, 10 February 2016 (UTC)

Yes for пизжу, спизжу but not the others (бзжу). Note that eg пизжу is for пиздеть and пиздить. --Anatoli T. ^{(обсудить}/^вклад) 22:18, 10 February 2016 (UTC)

Does Avanesov mention things like [pədɐˈʑːi] for подожди́ (podoždí)? --WikiTiki89 01:07, 11 February 2016 (UTC)

Good pickup. We should include it, even if it's not, it's common. This pronunciation is casual and colloquial, the same person may use either version.--Anatoli T. ^{(обсудить}/^вклад) 01:27, 11 February 2016 (UTC)

He doesn't include this for дождаться so I assume not for подождать. Does this colloquial pronunciation apply to all forms of these two verbs? Benwing2 (talk) 01:38, 11 February 2016 (UTC)

In this case, it's only for подождать, and I think only for the imperatives (which have an idiomatic meaning similar to "hold on a sec" in the sentence "Hold on a sec, what's going on?"), but it might also apply to future forms that have the vowel ё, I'm not sure. There may be other verbs that have a similar situation, but I can't think of any off the top of my head. --WikiTiki89 01:45, 11 February 2016 (UTC)--WikiTiki89 01:45, 11 February 2016 (UTC)

Yes, correct. Future forms as well. Basically, when "жд" in -ждать verbs (only for now) follows vowels and is palatal. I'd label them "colloquial", I don't think it's part of the old Moscow accent. And this doesn't apply to words like "рождение", etc. --Anatoli T. ^{(обсудить}/^вклад) 02:01, 11 February 2016 (UTC)

Wiki is probably right and it's only for подождать, not so for other related verbs (or not common enough).--Anatoli T. ^{(обсудить}/^вклад) 02:08, 11 February 2016 (UTC)

нео-, термо-, тетра- palatal or non-palatal

(moved from Talk:неонацизм)

@Atitarev Avanesov says this word along with other неонац- and неофаш- words have phon=нэ̀о- (and maybe also phon=нэ̀о̂-, I'm not quite sure about his notation). You created these pages without this, is this an optional pronunciation? Benwing2 (talk) 19:50, 12 February 2016 (UTC)

Palatal and reduced pronunciations are more common. (You can check with tons of anti-Ukrainian propaganda videos on YouTube). You can add the alternatives but my original is more common.--Anatoli T. ^{(обсудить}/^вклад) 21:33, 12 February 2016 (UTC)

@Atitarev, Cinemantique, Wikitiki89 A couple more questions:

1. What about термо-? Avanesov marks термодинамика as being either те- or тэ- and that's what we have indicated currently, but he marks термометр as having only те-. You indicated термопауза and термосфера as only те-. Should we change термодинамика to similarly have only те-? Or should we add тэ- as an alternant to термопауза and термосфера? Or neither?
2. What about тетра-? Per Avanesov we have тетраэдр marked as having only тэ- pronunciation, but you indicated тетраграмматон as having only те- pronunciation. Should we change either of these words to match the other?

Benwing2 (talk) 23:07, 12 February 2016 (UTC)

@Benwing2 If Avanesov says so, you can add those alternatives. I am only providing my personal experience on pronunciation, which is considered standard. To me, if someone says "нэ̀о̂-наци́зм", they are trying to sound smart or foreign or some other effect but if it's referenced, by all means. Same with other words you mentioned. Words with термо-, тетра- терро-, нео- are now normally palatalised (Avanesov's тэррори́зм, added by Cinemantique sounds quite dated to me, тэрмо́метр - even more so (but not тэ́рмос!) but the unpalatalised pronunciations of these must be standard as well. --Anatoli T. ^{(обсудить}/^вклад) 01:07, 13 February 2016 (UTC)

@Atitarev OK thanks. Note that Avanesov doesn't indicate a pronunciation тэрмо́метр but he does for other термо- words. Benwing2 (talk) 01:13, 13 February 2016 (UTC)

• I've looked in a new edition (2015) of this dictionary. Unpalatal pronunciation of нео- is still more standard, palatal pronunciation is marked as "admissible"; unreduced [o] is optional. Most of words with термо- have two standard variants (термодинамика, термодинамический, термозащита, термоизоляционный, термоизоляция, термообработка, терморегулятор, терморегуляция, термоскоп, термостат, термостатика, термостатический, термостойкий, термотерапия, термоустойчивый, термоэлемент, термояд, термоядерный); термограф, термопсис, термос have unpalatal variant only; термометр has palatal variant only. All words with тетра- have unpalatal [t].--Cinemantique (talk) 07:05, 13 February 2016 (UTC)
  • Thanks. Sounds like the 2015 dictionary is pretty similar to the 1988 one, and maybe is behind the times. Cinemantique, do you have a copy of this that you could email me? Benwing2 (talk) 07:42, 13 February 2016 (UTC)
    • You're welcome. I have only printed book.--Cinemantique (talk) 15:15, 13 February 2016 (UTC)

more palatal assimilation, per Avanesov

@Atitarev, Cinemantique, Wikitiki89 What do you all think of the new test cases I added, based on Avanesov? These propose additional palatal assimilations:

1. optional in тв and дв (and in ств + palatal, you get either sʲtʲvʲ or stvʲ, which I write s⁽ʲ⁾t⁽ʲ⁾vʲ)
2. optional in зб (and maybe сп?)
3. optional in св (and maybe зв?)
4. optional in см (and maybe зм?)
5. mandatory in дм (and maybe тм?)
6. optional in тл (and maybe дл?); Avanesov also puts a tie bar over it, but that's awkward with the optional palatalization of the т
7. tie bar over дн and тн per Avanesov

Cinemantique, could you verify what Avanesov says about these, and whether there are additional conditions (e.g. maybe it's optional in рдм/ртм/лдм/лтм)? Anatoli, does this sound right to you? Benwing2 (talk) 07:50, 13 February 2016 (UTC)

• They look right and some I can confirm straight away. Cinemantique's case (лучший) makes sense too. --Anatoli T. ^{(обсудить}/^вклад) 09:12, 13 February 2016 (UTC)
• All of this is dated a little. You can use more modern dictionaries, New Russian Orthoepic Dictionary by Ivanova (2005) or Russian Orthoepic Dictionary by Reznichenko (2003). For example,

твёрдый

Reznichenko: “! произн. [т^ьв^ь] имеет архаич. оттенок” = pronunciation [tʲvʲ] has archaic shade.

Ivanova: “[допуст. устар. т^ьв^ь]” = admissible obsolescent [tʲvʲ].

дверь

The same.

And so on.--Cinemantique (talk) 15:13, 13 February 2016 (UTC)

Thanks. What do the other dictionaries say about the other assimilations? Benwing2 (talk) 20:41, 13 February 2016 (UTC)

избить (Ivanova: no marks), сбить (Ivanova: “admissible [zʲbʲ]”), свежий (Ivanova: “admissible [sʲ]”; Reznichenko: no marks), смелый (Reznichenko, Ivanova: no marks), впечатляющий (Reznichenko, Ivanova: no marks). Only one word with дм, седмица, should have palatal [d], according to the new edition of Avanesov's dictionary (Reznichenko says it's archaic); the others haven't it (администрация, адмирал, бадминтон, кадмий, предмет, надменный...).--Cinemantique (talk) 02:38, 14 February 2016 (UTC)

@Cinemantique BTW Reznichenko sometimes marks ств as optionally palatal, sometimes only archaicly. E.g.

• in смертоуби́йственный as [stvʲ] and [sʲtʲvʲ], and similarly сле́дственный as [t͡stvʲ] and [t͡sʲtʲvʲ], no indication that either is archaic.
• On the other hand, in о славя́нстве, it says [nstvʲ] with the note that [nsʲtʲvʲ] имеет архаич. оттенок and [nʲsʲtʲvʲ] устар.
• Two lines above this, о славянофи́льстве is again marked simply [stvʲ] and [sʲtʲvʲ], no indication it's archaic.
• Likewise о скря́жничестве is marked simply [stvʲ] and [sʲtʲvʲ].
• On the other hand, о ско́тстве is marked [t͡sʲtʲvʲ] имеет архаич. оттенок, similarly for о скотово́дстве and о скопидо́мстве and о скоморо́шестве.

Do you think this could differ depending on whether there's a palatal vowel or consonant preceding? Benwing2 (talk) 20:07, 15 February 2016 (UTC)

Avanesov (dictionary, 1988, p. 673) says that palatalization in -ств- occurs more often when it's after [e] and [i] (especially stressed) or after [j]. Maybe that's the reason. I think all these palatal variants seem dated for young people.--Cinemantique (talk) 08:11, 16 February 2016 (UTC)

palatalization in imperative forms

(moved from Talk:высыпьте)

@Atitarev Is this actually pronounced as written with -пьть-, or more like -пть-? Similarly, are words like метьте actually pronounced with two distinct тьть sounds, or more like a single long тть sound? Benwing2 (talk) 19:55, 29 June 2016 (UTC)

1 - optional palatalisation, 2 - a single long sound. --Anatoli T. ^{(обсудить}/^вклад) 20:39, 29 June 2016 (UTC)

@Atitarev OK. What about other such combinations? I'm guessing:

• дьте, тьте: a single long sound
• пьте, бьте, мьте, вьте, фьте: optional
• льте, ньте, щьте, чьте: palatalized
• шьте, жьте: not palatalized
• сьте, зьте: palatalized
• рьте: palatalized
• цьте: doesn't seem to occur

Benwing2 (talk) 21:07, 29 June 2016 (UTC)

Good guesses. I've answered "рьте" and removed "?". цьте shouldn't occur. ць probably only occurs in imitations of Ukrainian and Belarusian accents or Russian dialects sounding like Belarusian (пець, взяць). Etymologically palatalised ц are very different in these languages (uk/be). --Anatoli T. ^{(обсудить}/^вклад) 22:38, 29 June 2016 (UTC)

@Atitarev I'll implement this. The only thing that needs to be handled is дьте, тьте and пьте, бьте, мьте, вьте, фьте. What about cases like сьс (e.g. in бросься) and other cases of palatalized sounds before ся, e.g. рассыпься? Is it similar? If so, бросься should be pronounced something like [ˈbros⁽ʲ⁾ːə] and рассыпься should be pronounced something like [rɐˈsːɨp⁽ʲ⁾s⁽ʲ⁾ə]. Benwing2 (talk) 23:50, 29 June 2016 (UTC)

@Atitarev, Cinemantique, Wikitiki89, Wanjuscha, KoreanQuoter Pinging various people. Benwing2 (talk) 20:25, 30 June 2016 (UTC)

Are you talking about the Old Moscow pronunciation? Otherwise why would -ся be depalatalized when it is not following a т(ь)? --WikiTiki89 20:34, 30 June 2016 (UTC)

(edit conflict) BTW is there a general rule about palatalized labials being optional before other consonants? What about восемьдесят, восемьсот, семьдесят, семьсот, and what about cases of palatalized labial + /j/, e.g. копьё, интервьюер? (NOTE: Ivanova claims that восемьдесят, восемьсот, семьсот have no-palatal м, but семьдесят has palatal м, and in дистрибьютер, the б is optionally palatalized, likewise in интервьюер, but doesn't say anything in the копьё entry. So maybe there's no rule. The only obvious thing that семьдесят has against the other three is the stress on the syllable preceding мь, so that might be a trigger.) Benwing2 (talk) 20:36, 30 June 2016 (UTC)

@Wikitiki89 We currently indicate final -ся as [s⁽ʲ⁾ə], with optional palatalization. I don't know what's normal and what's old-Moscow. Benwing2 (talk) 20:37, 30 June 2016 (UTC)

BTW we currently indicate imperative тьс as non-palatal, e.g. отметься [ɐtˈmʲet͡sːə], like in infinitives. On the other hand, imperative дьс like in обидься is e.g. [ɐˈbʲitʲ͡sʲs⁽ʲ⁾ə]. One of them is likely wrong. Benwing2 (talk) 20:40, 30 June 2016 (UTC)

Normal is palatalized (except after ть of infinitive or т of 3rd person), Old Moscow is depalatalized (almost?) everywhere. --WikiTiki89 20:43, 30 June 2016 (UTC)

I'm actually not sure about отметься, but I would pronounce it [-tʲsʲə]. --WikiTiki89 20:44, 30 June 2016 (UTC)

I think you're right about the depalatalization of labials. But it's something most people don't notice. Note also that there is a regional pronunciation of семь as сем. --WikiTiki89 20:54, 30 June 2016 (UTC)

Depalatalisation of -ся is an "old Moscow" feature, it's quite spread but is now less common than in the past. -т(ь)ся ending is always depalatalised and this standard, -лся is depalatalised frequently, more than other cases. I'd find it difficult to depalatalise forms like отметься, бросься, обидься. They should be all palatalised. "отметься" is a special case, since it looks like an infinitive ending.

Scholars consider dropping palatalisations in восемьдесят, восемьсот, семьдесят, семьсот non-standard (Ushakov). Depalatalised семь and восемь are even less common (and more noticeable) or dialectal - Ukraine, Belarus, southern Russia. --Anatoli T. ^{(обсудить}/^вклад) 00:14, 1 July 2016 (UTC)

@Atitarev This means we need to distinguish ться occurring in infinitives and in imperatives, right? I can do this using the pos= argument that we already use for final -е. Benwing2 (talk) 00:29, 1 July 2016 (UTC)

Yes, a good idea. I can't comment on optional palatalization rules for labials + j + vowel, we need to find a rule. Leave it as is for now. Ivanova also seems to treat words like восемьдесят differently from Ushakov. --Anatoli T. ^{(обсудить}/^вклад) 00:33, 1 July 2016 (UTC)

@Atitarev I have implemented various fixes for imperative pronunciation. As of yet it's only in my test version of ru-pron, but soon will be pushed live. Can you check the first seven entries in Module:ru-pron/testcases? Benwing2 (talk) 05:26, 1 July 2016 (UTC)

Looks OK. BTW, the "Зимба́бве" case must be simple but important. BTW, Московское произношение has important pronunciation points. --Anatoli T. ^{(обсудить}/^вклад) 05:34, 1 July 2016 (UTC)

Agreed that Зимба́бве should be simple; I'll get to it shortly. Benwing2 (talk) 06:11, 1 July 2016 (UTC)

pronunciation of стл

(moved from Talk:стлать)

@Atitarev Is it correct that this is pronounced the same as слать? Benwing2 (talk) 10:38, 29 June 2016 (UTC)

[t] should be optional, IMO.--Anatoli T. ^{(обсудить}/^вклад) 10:57, 29 June 2016 (UTC)

@Atitarev Is there a rule for this? For example, does optional [t] apply whenever the following syllable is stressed (in which case, what about вы́стлать, по́стланный, etc.) or simply in all forms of this verb and derivatives? Benwing2 (talk) 11:03, 29 June 2016 (UTC)

Hmm,according to this dictionary [3], [t] should be pronounced. I assume it will apply to derivatives. I still find it slightly awkward. Perhaps silent t should be marked as colloquial. I don't know the rule. Perhaps you can help me find more examples with such clusters. --Anatoli T. ^{(обсудить}/^вклад) 11:08, 29 June 2016 (UTC)

@Benwing2 On a 2nd thought, I now think we can follow that dictionary. A little bit of difficulty is with the pronunciation of standalone "слать". "Растление" is in common use and [t] is pronounced.--Anatoli T. ^{(обсудить}/^вклад) 11:14, 29 June 2016 (UTC)

And here is the rule!--Anatoli T. ^{(обсудить}/^вклад) 11:17, 29 June 2016 (UTC)

@Atitarev OK. That rule only mentions костлявый and постлать. Presumably it applies to all forms of all derivatives of стлать? Benwing2 (talk) 11:25, 29 June 2016 (UTC)

BTW here is the list of words with стл in them:

• Page 1588 армрестлинг: Processing - opt silent
• Page 9432 завистливый: Processing - silent
• Page 10207 застлать: Processing - pronounced
• Page 13549 костлявый: Processing - pronounced
• Page 18290 несчастливый: Processing - silent
• Page 20366 осчастливить: Processing - silent
• Page 24078 постлать: Processing - pronounced
• Page 24121 посчастливиться: Processing - silent
• Page 26969 разостлать: Processing - pronounced
• Page 31075 стлать: Processing - pronounced
• Page 31585 счастливец: Processing - silent
• Page 31586 счастливо: Processing - silent
• Page 31587 счастливого пути: Processing - silent
• Page 31588 Счастливого Рождества: Processing - silent
• Page 31589 счастливчик: Processing - silent
• Page 31590 счастливый: Processing - silent
• Page 34085 участливый: Processing - pronounced silent
• Page 34712 хвастливо: Processing - silent

Benwing2 (talk) 11:28, 29 June 2016 (UTC)

Marked above. --Anatoli T. ^{(обсудить}/^вклад) 20:44, 29 June 2016 (UTC)

@Atitarev I wonder if we shouldn't remove the rule that auto-reduces стл -> сл and instead handle these manually (maybe with a special rule to auto-reduce счастл and Счастл). Benwing2 (talk) 20:59, 29 June 2016 (UTC)

According to the link above, [t] is only pronounced in "костлявый" and the verb "-стла-" (all derivations). It's OK to leave it silent in "армрестлинг" as well. --Anatoli T. ^{(обсудить}/^вклад) 23:06, 29 June 2016 (UTC)

@Atitarev What about участливый? Benwing2 (talk) 23:41, 29 June 2016 (UTC)

Silent too, sorry! --Anatoli T. ^{(обсудить}/^вклад) 23:44, 29 June 2016 (UTC)

Interestingly, the [t] is also silent in English armwrestling, and in nestling it depends on the meaning: with [t] (and two syllables) when meaning "a small bird" = nest + -ling, without [t] (and three syllables) when the present participle of "nestle" (= nestle + -ing). Benwing2 (talk) 23:55, 29 June 2016 (UTC)

@Atitarev I fixed костлявый and all the стлать words. Benwing2 (talk) 18:59, 30 June 2016 (UTC)

@Atitarev, Cinemantique, Wikitiki89, Wanjuscha, KoreanQuoter Other words I just added that have стл with likely pronounced т: истлеть, истлевать. Can you review? Benwing2 (talk) 17:29, 23 July 2016 (UTC)

@Benwing2 Yes, these words have [t] pronounced. Please add as exceptions. --Anatoli T. ^{(обсудить}/^вклад) 12:23, 24 July 2016 (UTC)

• It seems that [t] is silent only in -стлив- (жалостливый, завистливый, пакостливый, совестливый, счастливый, участливый, хвастливый, and all derivatives) with exception for растливший (not silent). Рестлинг and derivatives are exceptions, too (silent).--Cinemantique (talk) 18:55, 24 July 2016 (UTC)

@Atitarev Do you agree? If so we can simplify things by making changing the стл -> сл rule to apply only to -стлив-. Benwing2 (talk) 21:08, 24 July 2016 (UTC)

It won't work for растливший, past active participle of растлить, impf - растлевать. Please note that in растлевать, истлевать, etc. [t] is pronounced because it's on the prefix-stem boundary. --Anatoli T. ^{(обсудить}/^вклад) 21:21, 24 July 2016 (UTC)

@Atitarev Right, but currently we have to make exceptions for костлявый, all words in -стлать, истлеть/истлевать, растлевать etc. What I'm suggesting is to make стлив -> слив be a rule, and then use manual respelling to handle растливший (in one direction) and армрестлинг (in the other direction). It's possible to add some additional logic involving stem boundaries so that растливший is automatically handled correctly, but it may not be worth it if it's only one word. Benwing2 (talk) 21:28, 24 July 2016 (UTC)

OK, if you think it's easier to implement that way.--Anatoli T. ^{(обсудить}/^вклад) 21:34, 24 July 2016 (UTC)

Some important points from the Russian Wikipedia's Moscow pronunciation article

@Benwing2 Московское произношение I'd like to mention some points, which have spread to the rest of Russia, or at least, as a variant.

1. -чн-, -чт- : [шт]о (что), [шт]обы (чтобы), коне[шн]о, «яи[шн]ица», наро[шн]о, скуч[шн]о, пустя[шн]ый, скворе[шн]ик, деви[шн]ик и т. д., а также в женских отчествах на -чна: Савви[шн]а, Ильини[шн]а и др.

   Most are already implemented but not all, especially feminine patronymics

2. на месте буквы а в первом предударном слоге после твёрдых фрикативных ш, ж и звука ц[источник не указан 1081 день] по старым московским нормам произносился звук /ɨ/, то есть говорили: [Шы]ляпин (Шаляпин), [шы]мпанское (шампанское), [шы]ги (шаги), [жы]ра (жара), [жы]ндарм (жандарм), [цы]ризм (царизм). Следы этого сохранились в современном произношении, например, в некоторых формах числительных: двадцати́ [dvə.t͡sɨˈtʲi], в слове жасмин ([жы]сми́н, [ʐɨˈsmʲin]), ржаной (р[жы]но́й, [rʐɨˈnoj]), в производных словах от глагола «жалеть» и производных от него: ж[ы]леть, к сож[ы]лению, пож[ы]лей, а также в формах слова «лошадь»: лош[ы]дей, лош[ы]дям, на лош[ы]дях.

   Some of these need to be verified. I agree with a few of them (bolded).

3. Дмитрий Ушаков писал, что окончание прилагательных -гий, -кий, -хий, например: «долгий, широкий, тихий» по-старомосковски произносятся так, как если бы было написано -гой, -кой, -хой, то есть как /-əj/.

   This is now dated but interesting. It was required by some actors and singers to use this pronunciation in the past.

4. также существуют особенности произношения определённых слов: целовать как «цаловать», танцевать как «танцавать».

   This is now dated. «цаловать» and «танцавать» are still heard in some movies. They actually contradict other features of the Moscow accent.

--Anatoli T. ^{(обсудить}/^вклад) 05:46, 1 July 2016 (UTC)

Thanks. I'll look into this. A few questions:

• For the ones in #1, we list что, чтобы, конечно as having only the [ш] pronunciation, but for the others we list both pronunciations, with [ш] first, and [ч] second, sometimes identifying them as Moscow and St. Petersburg respectively. Should we do that consistently? (i.e. except for что, чтобы, конечно, if I e.g. create пустячный, should it list both pronunciations and identify them as Moscow and St. Petersburg?)
• For the boldfaced ones in #2, should the [ы] pronunciation be given as the only one, or should we list both "Moscow" and "St. Petersburg" variants as in #1?
• For #3 and #4 should we include these dated variants at all? And if so should we identify them as simply "dated" or as e.g. "dated, Old Moscow"?

Benwing2 (talk) 08:03, 2 July 2016 (UTC)

@Atitarev A ping. Benwing2 (talk) 08:04, 2 July 2016 (UTC)

@Benwing2 I went ahead and made some changes as per above and added some some sample entries. I've got an error on пустячный (already pinged you on it.). I didn't consider #3 important but I've added variants for #4. It's not always easy or correct to flag something as "(old) Moscow" these features are common for Russian, so I don't always add "Moscow" tag. --Anatoli T. ^{(обсудить}/^вклад) 11:43, 2 July 2016 (UTC)

Optional palatalization in Зося, Люся, бабуся, etc.

What should we do here? The module currently makes all words ending in -ся have optional palatalization. This clearly isn't correct for words where the -ся isn't a reflexive ending. My first thought was to conditionalize on whether a vowel precedes, but this doesn't work because participial adjectives take -ся regardless of what precedes, e.g. учащегося, пресмыкающееся (in addition to occasional cases like Нинся). What I've done so far is respell words like Люся as Люсьа, which works but is kind of hacky. We have three possible solutions as far as I can see:

1. Require words like Зося, Нинся, Люся, бабуся, порося etc. to be respelled somehow or other, e.g. with -сьа, or otherwise have a special flag added. Perhaps a rule can be added to check for an initial capital letter; this would require only бабуся, порося, клюся and хибакуся to be respelled. See the list below.
2. Require words that need optional palatalization in -ся to be respelled or otherwise have a special flag added. This would require a lot more words to be respelled than the previous option: 61 lemmas currently (40 if we add a rule that covers infinitives in -чься), and a whole lot of non-lemmas.
3. Just remove the optional palatalization entirely and don't worry about it.

For reference, here are the 69 lemmas in -ся once those in -ться and -тся are excluded (I boldfaced the 8 words that aren't verbs or verb forms):

• Page 1999 -ающийся: Processing
• Page 2031 бабуся: Processing
• Page 2605 беречься: Processing
• Page 3223 борющийся: Processing
• Page 3315 брачущиеся: Processing
• Page 3316 брачующиеся: Processing
• Page 3904 Вася: Processing
• Page 4650 влечься: Processing
• Page 4654 вливающийся: Processing
• Page 5140 вооружившийся: Processing
• Page 5741 выдающийся: Processing
• Page 6237 вьющийся: Processing
• Page 8572 достичься: Processing
• Page 8801 дурак родился: Processing
• Page 9737 зажечься: Processing
• Page 10106 запечься: Processing
• Page 10807 Зося: Processing
• Page 11704 испечься: Processing
• Page 11966 кажущийся: Processing
• Page 12879 клюся: Processing
• Page 15169 льющийся: Processing
• Page 15222 Люся: Processing
• Page 15839 мент родился: Processing
• Page 16247 много будешь знать — скоро состаришься: Processing
• Page 17400 напрячься: Processing
• Page 17832 невзрывающийся: Processing
• Page 18384 неразорвавшийся: Processing
• Page 18457 несостоявшийся: Processing
• Page 18519 неудавшийся: Processing
• Page 18695 Нинся: Processing
• Page 19157 облечься: Processing
• Page 20415 осечься: Processing
• Page 20515 оставшийся: Processing
• Page 20532 остеречься: Processing
• Page 20672 отвлечься: Processing
• Page 20792 отклоняющийся: Processing
• Page 20982 отречься: Processing
• Page 22050 пересечься: Processing
• Page 22265 печься: Processing
• Page 23048 поддающийся: Processing
• Page 23793 полуразвалившийся: Processing
• Page 24117 порося: Processing
• Page 25056 прерывающийся: Processing
• Page 25069 пресмыкающееся: Processing
• Page 25202 привлечься: Processing
• Page 26301 проснувшийся: Processing
• Page 27041 развивающийся: Processing
• Page 27053 развлечься: Processing
• Page 27687 растечься: Processing
• Page 28576 с кем поведёшься, от того и наберёшься: Processing
• Page 28903 сбившийся: Processing
• Page 29019 светящийся: Processing
• Page 30074 сложившийся: Processing
• Page 30118 случившееся: Processing
• Page 30192 смешавшийся: Processing
• Page 30696 соотносящийся: Processing
• Page 30718 соприкасающийся: Processing
• Page 31392 стечься: Processing
• Page 32035 -ся: Processing
• Page 32036 ся: Processing
• Page 32688 толочься: Processing
• Page 33152 трудящийся: Processing
• Page 33326 ты мне нравишься: Processing
• Page 33516 увидимся: Processing
• Page 33532 увлечься: Processing
• Page 33887 улечься: Processing
• Page 34467 учащийся: Processing
• Page 35127 хибакуся: Processing
• Page 35804 чередующийся: Processing

Benwing2 (talk) 15:54, 10 July 2016 (UTC)

@Atitarev, Cinemantique, Wikitiki89, Wanjuscha, KoreanQuoter Ping. Benwing2 (talk) 15:55, 10 July 2016 (UTC)

According to this book, palatalization is optional only after [l]. In the other cases [s] is usually palatal.--Cinemantique (talk) 16:38, 10 July 2016 (UTC)

In the old Moscow pronunciation, the optional palatalisation was much broader for verb forms and participles and included the ending -сь: сержус(ь), горжус(ь). I think it's okey if we make it optional only for the ending -"лся", which is both modern and common. This will save us a lot of trouble.--Anatoli T. ^{(обсудить}/^вклад) 20:31, 10 July 2016 (UTC)

OK, implemented. Benwing2 (talk) 22:43, 10 July 2016 (UTC)

I agree. Just two follow up questions: (1) Are there any corner cases where -лся occurs as a form other than the masculine singular past? (2) Is optional palatalization still modern/common in the masculine singular past when it does not end in -лся (such as испёкся (ispjóksja))? Also, I just want to point out that I think this "optional palatalization" is at least sometimes realized as partial palatalization. I've begun to notice this in my grandfather's speech and in Vladimir Vysotsky's songs (I personally fully palatalize -ся in all cases other than infinitives in -ться and 3rd person forms in -тся). --WikiTiki89 14:37, 11 July 2016 (UTC)

1) No. I'm not aware of any such cases. 2) It may be a bit awkward to preserve the palatalisation in such positions - испёкся, обжёгся or увлечься. I'm OK to make it optional in such positions as well but I won't insist. We're making it mandatory to avoid problems. These clusters shouldn't cause problems, IMO (no ambiguity). --Anatoli T. ^{(обсудить}/^вклад) 04:31, 12 July 2016 (UTC)

Are you saying that увлечься can be depalatalized in the infinitive, or did you mean увлёкся? Anyway, I think we can make a simple rule that the palatalization is optional in the masculine singular past. By the way, do you agree with my observation about partial palatalization? --WikiTiki89 13:50, 12 July 2016 (UTC)

I meant увлечься but увлёкся can also be made optional, of course. The above endings will only occur in verbs, so they are safe. Yes, partial or light palatalisation can be heard. --Anatoli T. ^{(обсудить}/^вклад) 21:01, 12 July 2016 (UTC)

If you both agree I'll make кся, гся and чься have optional palatalization. Are you sure there aren't any names like e.g. Алекся that end in кся? Benwing2 (talk) 04:24, 13 July 2016 (UTC)

I think we should do it grammatically rather than solely by string matching. We also need to make sure we do it right, rather than whatever is easier. We should indicate optional palatalization when it is still modern. Here's a chart:

• Infinitive:
  • сади́ться — mandatory depalatalization
  • бере́чься — dated? or modern? optional depalatalization
  • найти́сь — dated optional depalatalization
• Participle:
  • сади́вшийся — dated? or modern? optional depalatalization or no depalatalization
  • сади́вшимся — dated? or modern? optional depalatalization
  • сади́вшихся — dated? or modern? optional depalatalization
  • сади́вшегося — dated optional depalatalization
  • садя́сь — dated optional depalatalization
  • сади́вшись — dated optional depalatalization
• 1st person singular:
  • сда́мся — dated? or modern? optional depalatalization
  • сдаю́сь — dated optional depalatalization
• 1st person plural:
  • сади́мся — dated? or modern? optional depalatalization
• 1st person plural imperative with -те:
  • сади́мтесь — dated optional depalatalization
• 2nd person singular:
  • сади́шься — dated? or modern? optional depalatalization
• 2nd person singular imperative:
  • отме́ться — dated? or modern? optional depalatalization or no depalatalization
  • оби́дься — dated? or modern? optional depalatalization or no depalatalization
  • бро́сься — dated? or modern? optional depalatalization or no depalatalization
  • ле́зься — dated? or modern? optional depalatalization or no depalatalization
  • сы́пься — dated? or modern? optional depalatalization or no depalatalization
  • понадо́бься — dated? or modern? optional depalatalization or no depalatalization
  • напра́вься — dated? or modern? optional depalatalization or no depalatalization
  • знако́мься — dated? or modern? optional depalatalization or no depalatalization
  • це́лься — dated? or modern? optional depalatalization or no depalatalization
  • ста́нься — dated? or modern? optional depalatalization or no depalatalization
  • уско́рься — dated? or modern? optional depalatalization or no depalatalization
  • пря́чься — dated? or modern? optional depalatalization or no depalatalization
  • тара́щься — dated? or modern? optional depalatalization or no depalatalization
  • нае́шься — dated? or modern? optional depalatalization or no depalatalization
  • вре́жься — dated? or modern? optional depalatalization or no depalatalization
  • ля́гся — dated? or modern? optional depalatalization or no depalatalization
  • сда́йся — dated? or modern? optional depalatalization or no depalatalization
  • сади́сь — dated optional depalatalization
  • Any others?
• 2nd person plural:
  • сади́тесь — dated optional depalatalization
• 2nd person plural imperative:
  • сади́тесь — dated optional depalatalization
• 3rd person singular:
  • сади́тся — mandatory depalatalization
  • нае́стся — mandatory depalatalization
• 3rd person plural:
  • садя́тся — mandatory depalatalization
• Masculine singular past:
  • нёсся — dated? or modern? optional depalatalization
  • вёзся — dated? or modern? optional depalatalization
  • расползся — dated? or modern? optional depalatalization
  • скрёбся — dated? or modern? optional depalatalization
  • сади́лся — modern optional depalatalization
  • испёкся — dated? or modern? optional depalatalization
  • обжёгся — dated? or modern? optional depalatalization
  • Any others?
• Other past:
  • сади́лись — dated optional depalatalization

@Atitarev: Can you check the above and answer the questions in orange? --WikiTiki89 16:22, 13 July 2016 (UTC)

Note that Cinemantique above claims that depalatalization is optional only after л. If we do try to "do it right" we run into the same issue I mentioned above, i.e. we have to distinguish verb forms from non-verb forms, and we have to choose whether to make the depalatalization optional by default or only with a respelling or an additional arg, and we have to choose whether to use respelling or an additional arg. Benwing2 (talk) 01:37, 14 July 2016 (UTC)

Often when a claim is that simple, it is because it wasn't investigated thoroughly enough. I just want to make sure that is really the case. We were previously operating under the claim that -ться is always depalatalized, but then it turned out that wasn't true in imperatives. --WikiTiki89 01:43, 14 July 2016 (UTC)

@Wikitiki89, Benwing2: Sorry for not getting back to you earlier. The old Moscow pronunciation featured optional palatalisations in ALL the cases above. It's like simply replacing particles -ся = -са, -сь = -с in all cases for speakers who used it. And they can be observed in all cases in movies, songs, recording, etc. Imitations in movies may not be consistent enough. (Someone must have considered this accent posh or aristocratic, so it's common in movies with Russian aristocrats.) The frequency will differ. I still think it's better to simplify this and not worry about missing some variants. We don't cover all dialectal, regional or dated pronunciations, anyway. For simplicity, I think it's only worth keeping optional -лся as it is considered modern and common with Russian speakers but I'll let you decide. Sorry if it sounds like a change of heart. A good list, Tiki. Thanks for the efforts. --Anatoli T. ^{(обсудить}/^вклад) 09:14, 15 July 2016 (UTC)

This is how things are currently implemented and I'm fine leaving it as-is. Benwing2 (talk) 16:32, 15 July 2016 (UTC)

Pronunciation of unstressed я when not directly before a syllable

@Atitarev, Cinemantique, Wikitiki89, Wanjuscha, KoreanQuoter I have a dictionary that I haven't looked at much before (Oxford Russian Dictionary). It claims that unstressed я has two pronunciations with the same distribution as unstressed а: like unstressed и directly before the stress, with a schwa sound elsewhere. They specifically give the example of genitive singular языка, which they claim should be pronounced [jǝzɨˈka]. Is this simply wrong? Benwing2 (talk) 05:08, 1 August 2016 (UTC)

It's not always easy to define, what is right and what is wrong in the pronunciation. Some speakers are affected by the way they write and they use much less reduction of vowels. IMO, Moscow, St Petersburg and Russian Central European regions normally reduce "я" as is currently implemented by our module. Speakers in South Russia, North Russia have different reduction rules. South Russians also pronounce "г" differently. Besides, in a very slow speech, in songs, the reduction of "я" and "е" may be less expressed but we are targeting the natural speech. I'm unconcerned and happy to continue reducing "я" as we normally do with your help. In verb endings like ходят it's important to use a shwa, which you have implemented. --Anatoli T. ^{(обсудить}/^вклад) 07:14, 1 August 2016 (UTC)

special-casing счит -> щит

@Atitarev, Cinemantique, Wikitiki89, Wanjuscha, KoreanQuoter I am thinking of adding a rule that says that the sequence счит should be pronounced as щит rather than щчит. It looks like all our existing words with счит in them are pronounced as щит. Note that we already have a similar rule affecting the sequence счёт. Any objections? Benwing2 (talk) 14:07, 20 August 2016 (UTC)

BTW here is the full list:

• Page 6238 высчитать: Processing
• Page 6239 высчитывать: Processing
• Page 8773 досчитать: Processing
• Page 8774 досчитаться: Processing
• Page 8775 досчитывать: Processing
• Page 8776 досчитываться: Processing
• Page 10650 засчитать: Processing
• Page 10651 засчитывать: Processing
• Page 18074 насчитать: Processing
• Page 18075 насчитаться: Processing
• Page 18076 насчитывать: Processing
• Page 18077 насчитываться: Processing
• Page 18445 недосчитать: Processing
• Page 18446 недосчитаться: Processing
• Page 18447 недосчитывать: Processing
• Page 18448 недосчитываться: Processing
• Page 19964 обсчитать: Processing
• Page 19965 обсчитывать: Processing
• Page 21678 отсчитать: Processing
• Page 21679 отсчитаться: Processing
• Page 21680 отсчитывать: Processing
• Page 21681 отсчитываться: Processing
• Page 22729 пересчитать: Processing
• Page 22730 пересчитывать: Processing
• Page 24019 подсчитать: Processing
• Page 24020 подсчитывать: Processing
• Page 25118 посчитать: Processing
• Page 25119 посчитаться: Processing
• Page 25120 посчитывать: Processing
• Page 26457 присчитать: Processing
• Page 26458 присчитывать: Processing
• Page 27186 просчитать: Processing
• Page 27187 просчитаться: Processing
• Page 27188 просчитывать: Processing
• Page 27189 просчитываться: Processing
• Page 28555 рассчитать: Processing
• Page 28556 рассчитаться: Processing
• Page 28557 рассчитывать: Processing
• Page 28558 рассчитываться: Processing
• Page 31818 сосчитать: Processing
• Page 31819 сосчитаться: Processing
• Page 31820 сосчитывать: Processing
• Page 32995 считанный: Processing
• Page 32996 считать: Processing
• Page 32997 считать галок: Processing
• Page 32998 считаться: Processing
• Page 32999 считывать: Processing
• Page 36767 цыплят по осени считают: Processing
• Page 37285 чуть-чуть не считается: Processing

Benwing2 (talk) 14:07, 20 August 2016 (UTC)

Sounds reasonable. Although note that I pronounce считывать and the second verb section at считать as с+ч rather than щ. --WikiTiki89 15:17, 20 August 2016 (UTC)

Hmm. Ivanova under считать(ся) says "[щи и щчи] (во всех знач.)". So I'd assume считывать can go both ways as well... OTOH считанный is only [щи] (both as adj and participle) and likewise считаный, whereas считка and считывание are only [щчи]. считывать itself isn't listed. @Atitarev What do you think? Benwing2 (talk) 17:42, 20 August 2016 (UTC)

I checked Avanesov and he makes the same distinction I do between the different meanings of считать ([щ] for counting and [ш^ьч] for reading off) and a corresponding distinction in the participle считанный, and does include считывать (as [ш^ьч]). I personally don't assimilate the с at all (it's not even palatalized), but I guess that's phonemically the same thing. --WikiTiki89 19:15, 20 August 2016 (UTC)

Maybe we should split the two meanings etymologically? считать "to read off" is clearly с- + читать; считать "to count" maybe has the same origin but it is strange that it's imperfective. Benwing2 (talk) 20:08, 20 August 2016 (UTC)

BTW ruwikt splits the two like you do. Benwing2 (talk) 20:11, 20 August 2016 (UTC)

They definitely feel like different etymologies. Ru.wikt sometimes splits things up too much, but in this case I think it's the right thing to do. --WikiTiki89 20:34, 20 August 2016 (UTC)

OK, I split up the etyms. Benwing2 (talk) 00:02, 21 August 2016 (UTC)

Thanks. Yes, it should be "щч" when the root is from "читать" (to read). --Anatoli T. ^{(обсудить}/^вклад) 01:48, 21 August 2016 (UTC)

Some расч- words

@Atitarev, Cinemantique, Wikitiki89, Wanjuscha, KoreanQuoter Avanesov, Ivanova and Reznichenko don't completely agree about various words in расч-.

• Avanesov says расчистить, расчленить, расчленять, расчесть, расчесать, расчёска take щч only.
• Ivanova agrees that расчистить, расчесть, расчесать take щч only, but says that расчёска takes щч or щ, and says that расчленить/расчленять is "[допуст. щч]", which I think means that щ is preferred but щч is also possible.
• Reznichenko says расчистить and расчесть take щч or щ, and is silent on the rest.

Who do we trust the most? Currently our pronunciations for these words are entirely щч, in agreement with Avanesov. Benwing2 (talk) 00:56, 21 August 2016 (UTC)

@Atitarev, Cinemantique, Wikitiki89, Wanjuscha, KoreanQuoter in case the last ping didn't go through. (And note that Wanjuscha thinks that расчесать, расчёсывать, расчёска take щ; I changed them all to щч.) Benwing2 (talk) 00:57, 21 August 2016 (UTC)

I pronounce расчесать, расчисывать, расчистить, and расчищать with щч, and расчёска with щ. I'm not sure about the other ones because I don't hear or use them very often. I honestly think that any instance of щч can be pronounced щ by some people. --WikiTiki89 01:19, 21 August 2016 (UTC)

I agree with Wikitiki89 about these words' pronunciation. You can follow Ivanova, who says "расчёска" can be either щ or щч (the former is more frequent, IMO) and "расчесать" can take both щ or щч (the latter is more frequent, IMO). I don't agree with Reznichenko - only щч for those terms. Avanesov's statement is right, except for расчесать and расчёска. --Anatoli T. ^{(обсудить}/^вклад) 01:46, 21 August 2016 (UTC)

Ok, implemented. I made расчёсывать follow расчесать, hopefully that is right. Benwing2 (talk) 05:22, 21 August 2016 (UTC)

@Atitarev, Cinemantique, Wikitiki89, Wanjuscha, KoreanQuoter What about words in исчисл- (исчислить, исчислять, исчисляться, исчисление, исчисляемый, исчисляемое существительное)? Wanjuscha thinks the first three are pronounced ищисл-. Avanesov says all are щч only, while Ivanova says "щ or щч". Benwing2 (talk) 15:39, 21 August 2016 (UTC)

recent change by Erutuon

@Atitarev, Cinemantique, Wikitiki89, Wanjuscha, KoreanQuoter Can one of you review this change? Benwing2 (talk) 03:56, 1 September 2016 (UTC)

I saw it. Probably OK. --Anatoli T. ^{(обсудить}/^вклад) 03:58, 1 September 2016 (UTC)

In this book (p. 56) I see symbols д’з͡’ and д’ж͡’ (apostrophe means palatalization). So I think Erutuon's change isn't correct.--Cinemantique (talk) 04:18, 1 September 2016 (UTC)

It is technically correct to write [tʲsʲ], but as I understand it, that is not the usual convention when using the International Phonetic Alphabet. Although both the stop and the fricative portion of the affricate have the same place of articulation, the IPA only marks the place of articulation on the fricative portion. Hence (though I am omitting tie bars here), we write [tɕ], not [ȶɕ] (though [ȶ] is not a standard IPA symbol); [tʂ], not [ʈʂ]. [t] has a different place of articulation in all of these, but the same symbol is used nevertheless. Similarly, we write [tsʲ], not [tʲsʲ], even though both the stop portion [t] and the fricative portion [s] of the fricative are actually palatalized. The affricate is a single unit: the stop and the fricative portion have the same place of articulation even if it is not marked on the stop. — Eru·tuon 06:28, 1 September 2016 (UTC)

@Atitarev, Cinemantique, Wikitiki89, Wanjuscha, KoreanQuoter, Erutuon OK, should I revert the change or leave it alone? If we leave it alone we need to fix a number of the test cases in Module:ru-pron/testcases which assume the [tʲsʲ] representation. Benwing2 (talk) 19:58, 3 September 2016 (UTC)

I would say leave it. I've actually mentioned this before, that we should either be consistent and duplicate all the details onto the t (like [ʈʂ]) or always use a plain t as the first element of affricates. --WikiTiki89 20:08, 3 September 2016 (UTC)

станция

• станция:

  (file)

Hi, I came upon this word while adding the pronunciation of w:Vostok Station to Wikipedia. I wanted to ask about it; {{ru-IPA}} generates the pronunciation [ˈstant͡sɨjə], but the soundfile above sounds more like [ˈstant͡sɨə] or [ˈstant͡sɨː]. I don't know a great deal about Russian; is the soundfile wrong or is the module generating an incorrect transcription for this word? — Eru·tuon 20:10, 12 September 2016 (UTC)

Your ears are just not adjusted to hearing Russian. The /j/ is very central and the /ə/ is very short, so they might be difficult to distinguish if you're not used to it. --WikiTiki89 20:13, 12 September 2016 (UTC)

Huh, I have never heard of a centralized [j]. I suppose that would make it a post-palatal or pre-velar approximant, [j̠] or [ɣ̞˖]. — Eru·tuon 20:54, 12 September 2016 (UTC)

Umm... Ever heard of [ɪ̯]? --WikiTiki89 21:05, 12 September 2016 (UTC)

Well, that would be equivalent to a lowered and backed j̠˕. But that would be a weird transcription: [ɨ] is supposed to be close central, [ɪ̯] near-close near-front, and [ə] mid central, so the sequence [t͡sɨɪ̯ə] has a close vowel opening to near-close and finally opening to mid. That is odd; I would expect the semivowel to be at least slightly closer than the vowels surrounding it... A table on Russian Wikipedia, which someone just translated indicates that pronunciation is [t͡sᵻɪ̯ɪ], but that doesn't sound right either.

• Аркадия (Arkadija):

  (file)

• акация (akacija):

  (file)

Maybe I'm overanalyzing, but the transcription just isn't very accurate or illuminating: I feel that the very marked difference between unstressed -ия after a soft consonant and unstressed -ия after a hard consonant should be transcribed somehow. Аркадия clearly has a much closer vowel and a more marked semivowel in -ия than станция or акация. It's not a difference of the first vowel being near-front near-close and second being central close vowel, as the transcriptions [ɪjə, ɨjə] imply. — Eru·tuon 22:18, 12 September 2016 (UTC)

To clarify, by "central" above I really meant "closer to mid-central". A few things: the chart you linked to ignores that there are a few exceptions to the reduction rules in inflectional suffixes, which is why -я here reduces to /-jə/ and not to /-jɪ/. As for the difference between ия after hard and soft consonants, there probably is no difference in the closeness of the first vowel, but only in the frontness. Now maybe you could figure out better than I do what the exact articulation of the /j/ is, but regardless of what it is, it is there. Also listen to the audio at моро́женое (moróženoje) for another example of an almost inaudible /j/. --WikiTiki89 22:32, 12 September 2016 (UTC)

possible secondary stress on само-

@Atitarev, Cinemantique, Wikitiki89, Wanjuscha, KoreanQuoter Avanesov, Ivanova and Zaliznyak are consistent in indicating that само- never has secondary stress, but Cinemantique claims it often does, according to the following rule:

• Always са̀мо- if three or more syllables intervene between сам- and the stressed syllable;
• Never са̀мо- if one syllable intervenes between сам- and the stressed syllable;
• Sometimes са̀мо- if two syllables intervene between сам- and the stressed syllable.

We generally don't indicate secondary stress in само- words, including words like самоудовлетворение with tons of intervening syllables. Is Cinemantique correct, and if so, which words need fixing? If needed, I can compile a list of words with two syllables intervening for someone like Anatoli to mark up, and fix up the non-lemma forms appropriately. Benwing2 (talk) 01:52, 25 May 2017 (UTC)

Very long words like "самоудовлетворение" may have an alternative secondary stress but perhaps we need to see more examples, like you suggested.--Anatoli T. ^{(обсудить}/^вклад)

OK, here is the list:

With 3 or more intervening syllables:

• самовыражение
• самообладание
• самооборона
• самооговор
• самоопределение (note, has alternative secondary stress indicated)
• самоорганизация
• самоотречение
• самоощущение
• самопоедание
• самопровозглашённый
• самосохранение
• самоуважение
• самоудовлетворение
• самоуправление
• самоустраняться (note, has mandatory secondary stress indicated)
• самоутверждение

With 2 intervening syllables:

• самодержавие
• самодержавный
• самодовольный
• самодовольство
• самодостаточность
• самодостаточный
• самозабвение
• самозабвенно
• самозабвенный
• самозащита
• самонадеянный
• самоназвание
• самообман
• самообслуживание
• самоотверженность
• самоотверженный
• самооценка
• самопожертвование
• самосознание
• самостоятельно
• самостоятельность
• самостоятельный
• самоубийство
• самоубийца
• самоуверенность
• самоуверенный

Benwing2 (talk) 03:28, 25 May 2017 (UTC)

@Atitarev Benwing2 (talk) 03:30, 25 May 2017 (UTC)

@Benwing2 You can add an alternative secondary stress on ALL terms in the first group. The 2nd group (with 2 intervening syllables) don't have a secondary stress. --Anatoli T. ^{(обсудить}/^вклад) 22:20, 25 May 2017 (UTC)

@Atitarev Done. Benwing2 (talk) 03:03, 26 May 2017 (UTC)

профессор, профессура, профессиональный, etc.

Some questions.

• Based on the gem=opt in профессор, I added it to профессорский. Is this correct?
• ruwikt has gem=opt on профессура, so I added it here and fixed the non-lemma forms. Is this correct?
• Do профессионал, профессиональный, профессионально, профессионализм have gem=opt?
• Do the plural forms of профе́ссор (профессора́, профессора́ми, etc.) still have gem=opt although the stress is no longer preceding the сс? My bot assumes so, and adds it.

Thanks! Benwing2 (talk) 14:54, 27 May 2017 (UTC)

@Atitarev, Cinemantique, Wikitiki89, Wanjuscha, KoreanQuoter Benwing2 (talk) 14:54, 27 May 2017 (UTC)

профессорский - OK. IMO, профессионал, профессиональный, профессионально, профессионализм have no geminations. Same with forms of профе́ссор where the stress is on the ending. I wouldn't geminate профессура either. --Anatoli T. ^{(обсудить}/^вклад) 15:03, 27 May 2017 (UTC)

None of these has gemination, according to orthoepic dictionaries.--Cinemantique (talk) 15:57, 27 May 2017 (UTC)

I don't mind if the gem. is removed. Perhaps the note on the "профессор" audio file should be made. (It can be heard). --Anatoli T. ^{(обсудить}/^вклад) 22:00, 27 May 2017 (UTC)

Let me think about this, I'll leave them for now. Benwing2 (talk) 22:04, 27 May 2017 (UTC)

@Atitarev, Cinemantique, Wikitiki89, Wanjuscha, KoreanQuoter I removed the optional gemination from профессия, профессор, профессорский. Benwing2 (talk) 07:22, 28 May 2017 (UTC)

@Atitarev, Cinemantique, Wikitiki89, Wanjuscha, KoreanQuoter Also, what about the end-stressed forms like глиссера́ (glisserá) of гли́ссер (glísser)? Do they have optional gemination? Ivanova reports optional gemination of гли́ссер, but only lists stem-stressed plural forms like гли́ссеры (glíssery). Benwing2 (talk) 07:29, 28 May 2017 (UTC)

Remove gem. if before the stress in this case.--Anatoli T. ^{(обсудить}/^вклад) 09:45, 28 May 2017 (UTC)

видео-

Most words in видео- are given with secondary stress вѝдео-, and with two pronunciations, one with reduced о and the other with unreduced о, e.g. видеоигра, видеокассета, видеоклип, видеокамера. However, видеоблог, видеозапись, видеомагнитофон, видеоплёнка are given with no secondary stress and one pronunciation only, with reduced о, and видеоинформация is given with secondary stress but still one pronunciation only, with reduced о. Should these words follow видеоигра видеокассета etc.? What about видеофильм, which I'm going to add? Same? Also, while we're at it, should we add видеотека and видеодисплей, which are currently found only in other languages but also exist in Russian? Do these follow the same rules as видеоигра, видеокассета, etc.? Benwing2 (talk) 01:21, 31 May 2017 (UTC)

видеоте́ка is the only exception (we haven't a separate word *тека in such sense).--Cinemantique (talk) 06:46, 31 May 2017 (UTC)

Agreed. Words with аудио- (audio-) can be handled the same way. --Anatoli T. ^{(обсудить}/^вклад) 06:50, 31 May 2017 (UTC)

When you say видеоте́ка is the only exception, do you mean it lacks secondary stress on видео, or has only reduced о, or both? Benwing2 (talk) 06:51, 31 May 2017 (UTC)

@Atitarev, Cinemantique, Wikitiki89, Wanjuscha, KoreanQuoter Benwing2 (talk) 06:51, 31 May 2017 (UTC)

It has neither secondary stress nor unreduction.--Cinemantique (talk) 06:55, 31 May 2017 (UTC

@Atitarev, Cinemantique, Wikitiki89, Wanjuscha, KoreanQuoter OK. I fixed up аудиокнига, аудиоплёнка and аудиовизуальный to have both reduced and unreduced о. Not sure about аудиовизуальный. I didn't this to аудиология, which I assume has no secondary stress and only reduced о. Benwing2 (talk) 07:15, 31 May 2017 (UTC)

радио- words

Anatoli says words in радио- such as радиопередача can be pronounced with unreduced о or with reduced о, similar to аудио- and видео- words. I want to fix up these words but I have a feeling that not all such words allow unreduced о. Can you help me by putting a "y" or "n" by each of the following indicating whether it can be pronounced with unreduced о?

• Page 1346 антенна радиолокационного дальномера
• Page 3533 Британская телерадиовещательная корпорация
• Page 28784 радиоактивный
• Page 28785 радиоаппарат
• Page 28786 радиоаппаратура
• Page 28787 радиовещание
• Page 28788 радиограмма-- No.--Cinemantique (talk) 04:32, 17 June 2017 (UTC)
• Page 28789 радиозонд
• Page 28790 радиокоммуникационный
• Page 28791 радиолокационный
• Page 28792 радиолокационный дальномер
• Page 28793 радиомолчание
• Page 28794 радиопередатчик
• Page 28795 радиопередача
• Page 28796 радиоприёмник
• Page 28797 радиопромышленность
• Page 28798 радиосвязь
• Page 28799 радиостанция -- No.--Cinemantique (talk) 04:32, 17 June 2017 (UTC)
• Page 28800 радиотехнический
• Page 28801 радиоэлектронный
• Page 30854 сарафанное радио -- Only unreduced.--Cinemantique (talk) 04:32, 17 June 2017 (UTC)
• Page 34587 телерадиовещательный
• Page 36281 ультракоротковолновая радиостанция -- No.--Cinemantique (talk) 04:32, 17 June 2017 (UTC)

Thanks! Benwing2 (talk) 04:23, 17 June 2017 (UTC)

@Atitarev, Cinemantique, Wikitiki89, Wanjuscha, KoreanQuoter Benwing2 (talk) 04:24, 17 June 2017 (UTC)

• Words with secondary stress can be pronounced with unreduced о.--Cinemantique (talk) 04:32, 17 June 2017 (UTC)

@Atitarev, Cinemantique, Wikitiki89, Wanjuscha, KoreanQuoter Thanks! For the "no" words above, is there no secondary stress? Currently both радиограмма and радиостанция are listed with secondary stress. Benwing2 (talk) 05:22, 17 June 2017 (UTC)

As per Cinemantique, no sec. stress on those words.--Anatoli T. ^{(обсудить}/^вклад) 05:32, 17 June 2017 (UTC)

Back /l/

@Atitarev, Cinemantique, Wikitiki89, Wanjuscha I'm thinking we should represent non-palatal /l/ as [ł]. The darkness of non-palatal Russian /l/ is one of the most notable characteristics of Russian pronunciation; palatal /lʲ/ is definitely not just the palatalized equivalent of non-palatal /l/, unlike e.g. for /s/, /p/, /r/ or /n/. It seems strange not to note this highly salient feature of Russian pronunciation when we note so many other things (e.g. the different qualities of unstressed /a/, the assimilation of /t/ to a following /s/, /š/ or /č/, etc.). This would be an easy change to make in the module, and it's already made in the ruwikt equivalent module. Benwing2 (talk) 23:19, 30 July 2017 (UTC)

Another possibility, which I feel less strongly about, is to note the velar offglide of label consonants before /ɨ/; a word like мыть sounds to me almost like [mʷɨtʲ], although perhaps [mˠɨtʲ] is more accurate (although probably also more misleading). Something else I noticed right off the bat when I first went through the Pimsleur Russian tapes is the tendency of many speakers to pronounce /tʲ/ as [tˢʲ] and similarly for /dʲ/, although Anatoli informs me that this is not found in all speakers.) Benwing2 (talk) 23:26, 30 July 2017 (UTC)

@Benwing2 I never objected implementing [ł]. If I'm not mistaken it was User:Wikitiki89 but I don't remember exactly why and where. I think the argument was that we are using a phonemic transcription, e.g. exact realisations of [r], [j], etc. are not shown or something. As for your other questions, I don't agree and yes, [tˢʲ] (and its voiced equivalent) are not used in standard Russian. --Anatoli T. ^{(обсудить}/^вклад) 08:01, 31 July 2017 (UTC)

Regarding the /l/. If we don't show the distinction between the dental /t/, /d/, /n/ and the nearly dental /s/, /z/, /ts/ vs. the totally alveolar /tʲ/, /dʲ/, /nʲ/, /sʲ/, /zʲ/, and the distinction between alveolar /r/ and nearly-dental /rʲ/, and the heavy labialization and rounding of all consonants before stressed о and ё, etc., then why should we show the darkness of the /l/? Just because it is "one of the most notable characteristics" in your mind, doesn't make it more important than other characteristics. As far as phonemes are concerned, the pair /l/ and /lʲ/ is not much different from any of the others, so there is no reason to doubly differentiated it when we don't do so for the others. Regarding [tˢʲ], I would probably disagree with Anatoli and say that this is within the boundaries of standard Russian (note that [tˢʲ] is not the same as [tsʲ]), although it is not necessarily "the standard" realization. --WikiTiki89 15:46, 31 July 2017 (UTC)