Help talk:IPA/Russian/Archive 1

Archive 1

Archive 2

Archive 3

IPA-rus

Just as there is a template {{IPAEng|}} there could be an {{IPAru|}} that links to here. This would be a nice place to set up procedures for transcribing Russian in a consistant manner across Wikipedia. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 15:58, 6 March 2008 (UTC)

Actually, a place to set up procedures for transcribing Russian should not be located in mainspace. Cf. Romanization of Russian and Wikipedia:Romanization of Russian.—Ëzhiki (Igels Hérissonovich Ïzhakoff-Amursky) • (yo?); 16:53, 6 March 2008 (UTC)

Huh? We're talking about IPA. Maybe Help:IPA might be a better place. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 18:47, 7 March 2008 (UTC)

I know we are talking about IPA :) I used romanization as an example of the encyclopedic article being separate from Wikipedia's guidelines. Procedures (of romanization, trascription, etc.) should not be set up in mainspace (a suggestion which I, perhaps mistakenly, derived from your original comment). Help space is, of course, another possibility. Cheers,—Ëzhiki (Igels Hérissonovich Ïzhakoff-Amursky) • (yo?); 20:31, 7 March 2008 (UTC)

Phonemic vs phonetic

Although vowel allophony is an important aspect of Russian pronunciation, which phone a phoneme will be realized as is predictable from its environment. The table should probably make clear the distinction between the relatively small number of vowel phonemes that should be used in a phonemic transcription of Russian (e.g., жест /ʒɛst/, цель /t͡sɛlʲ/, пять /pʲatʲ/, чуть /t͡ʃʲutʲ/) and the much larger number of vowel phones to be used in a phonetic transcription (e.g., жест [ʐɛst], цель [t͡selʲ], пять [pʲætʲ], чуть [t͡ɕʉtʲ]). Also, /o/ is often realized as a clearly audible [oə], which is not indicated in the current table. Furthermore, if you're going to provide such a narrow phonetic transcription for the vowels, you might also want to note that phonemically nonpalatalized (or "plain") consonants are often affirmatively velarized/pharyngealized, so that нос and нёс are phonemically /nos/ and /nʲos/, but phonetically [n̴oəs̴] and [nʲoəs̴] (or [nˠoəsˠ] and [nʲoəsˠ]), respectively. And what about the labial offglide in, e.g., вы [v̴ʷɨ]? My point is that not everyone always wants to show the fine details of Russian phonetics; sometimes, a phonemic transcription is enough, and the table as it currently stands gives the impression that Russian has a lot more phonemes than it actually does. —Preceding unsigned comment added by Hizkuntzalari (talk • contribs) 11:07, 7 June 2008 (UTC)

If Russian is transcribed phonemically, then the link should go to Russian phonology (which is where one ought to get information about Russian phonology anyway). It's true that there can be even greater specificity in the phonetic transcription, but there are practical considerations. I happen to think that it's a good balance as it is fairly specific but doesn't require any diacritics but for the superscript j (which is unavoidable). I have a suspicion that you were being sarcastic in your suggestions, though if you have some serious ones I've got open ears. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 22:04, 27 June 2008 (UTC)

Stress

One important thing is not mentioned here, that is, how to mark stress in IPA representation of Russian words?--92.39.161.219 (talk) 21:11, 11 September 2008 (UTC)

Very true. I've added it with an example. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 03:36, 12 September 2008 (UTC)

Stressed vowels

There is no difference in pronunciation of stressed vowels in "жест" and "цель", "трава" and "палка", "пуля" and "чуть". At least any differences are absolutely incomprehensible by a native speaker (not a linguistics professional). As such, I believe there shouldn't be any difference in pronunciation charts for these sounds in general audience encyclopedia. Conversely, there are quite pronounced differences between unstressed vowels in "тяжелый" and "этап"; "дышать", "жена" and "сердце". Bottom line: the table should be fixed (not being familiar with IPA, I can't do this myself). --Dp074 (talk) 01:48, 2 May 2008 (UTC)

Allophony usually means that speakers don't notice. Besides, vowel allophony is one of its more noticeable aspects of Russian (that and palatalization). In addition, a number of the allophones for the vowels are those that English speakers make contrasts in (the difference between [ɛ] and [e], [æ] and [ɑ]). — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 04:47, 2 May 2008 (UTC)

If speakers don't notice, the reader shouldn't either. Wikipedia is not a linguistic source, but rather targets a broad audience. I am not sure I understand your arguments about English, as this is about Russian pronunciation. To make it look like these sounds are different in Russian is definitely wrong. Can anybody offer a reasonable solution? --Dp074 (talk) 01:57, 4 February 2009 (UTC)

That's a bit draconian. The transcriptions are to give English speakers (Wikipedia's target audience) a phonetic rendering of Russian pronunciation without being unnecessarily technical. If we were to transcribe phonemes only, then readers would be misled on Russian pronunciation because, for example, жена is /ʐeˈna/ phonemically but ʐɨ̞ˈna] or [ʐɨˈna]. Do you have any suggestions for reasonable solutions? — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 18:41, 4 February 2009 (UTC)

Added links to consonantic sounds

Hi there, I forgot to add a note in the comment. I added links to the IPA sound articles where applicable. Other "IPA for" languages are doing the same, and I think it's very useful and for some letters more precise than the English approximation examples. Thanks. 123.225.195.98 (talk) 09:05, 3 June 2010 (UTC)

Good job. Two things:

• [ɛ] is used for both an open mid and mid vowel. The link is then too specific. If no one has a problem with this (it's not like English speakers notice the difference) then keeping the link should be fine.
• Russian [l] is technically a velarized (or dark) [ɫ], though we're using [l] because it's non-contrastive and easier to type. So I'm thinking that the link should go to velarized alveolar lateral approximant. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 16:22, 3 June 2010 (UTC)

Hello. I feel I am not competent enough, so you may want to wait for other comments or just go ahead with what you think is best. From my perspective, we should use the correct sound ([ɫ]) and link it as you suggest. As for [ɛ], it sounds to me like we should have two separate rows for two separate sounds.

Also, I was thinking of creating specific IPA templates (if they are not there) for each sound, so that for example {{IPA-ɫ}} expands to something like

[[Velarized alveolar lateral approximant|{{IPA|ɫ}}]]

It could then be reused by all these "WP:IPA for X" pages more easily. What do you think? 123.222.215.125 (talk) 23:54, 3 June 2010 (UTC)

OK, I have created a parametric Template:IPAlink, which is more maintainable. I have fixed up this article accordingly, please review. Thanks. 123.218.168.102 (talk) 12:16, 12 June 2010 (UTC)

Cluster palatalization

Should we be marking the palatalization of earlier consonants in clusters, or not? (i.e., should сть be [stʲ] or [sʲtʲ]) From what I can tell, this is an allophonic process, not a phonemic one, but it's comparable to final-devoicing. Some articles mark it; some do not. — ˈzɪzɨvə (talk) 18:48, 6 April 2010 (UTC)

There's a complex set of rules of when it's palatalized and when it's not. Because it's so complex, we should mark it whenever it occurs. What are some pages that don't? — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 23:57, 6 April 2010 (UTC)

I'm curious: what are these rules? Russian phonology doesn't seem to include them. Of the articles I've edited in the last day or so, these feature clusters with palatalization: Glasnost (both palatalized), Ptichka (spacecraft) (I wasn't aware of this process at the time), and Lyudmila Alexeyeva (I didn't palatalize the whole cluster, assuming the syllable break blocked it). Oooh, just I found the section of Russian phonology regarding this. If transcriptions should include this, perhaps we can boil that section down to include as a note on this page? I don't think I understand well enough, or I'd try it. — ˈzɪzɨvə (talk) 01:55, 7 April 2010 (UTC)

Technically, Russian phonology doesn't break down the rules as they are laid out in Sound Pattern of Russian. I think I can whip something up to put here. Your transcription is accurate, though AFAIK syllabification incorporates a maximized onset in Russian. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 05:58, 7 April 2010 (UTC)

Okay, it's late. I've started a table that incorporates these rules, though there are a few ambiguous cells:

		labial (p b f v m)		dental (t d s z n)		r		l		velar (k g x)
		soft	hard	soft	hard	soft	hard	soft	hard	soft	hard
labial (p b f v m)	soft	x		x	-	-	x	?	?	-	x
	hard	-	x	-	x	-	x	-	x	-	x
dental (t d s z n)	soft	x	-	x	-	-	x	? льн	?	-	x
	hard	-	x	-	x	-	x	-	x	-	x
r	soft	-	x	-?	x?	x	-	-		-	x
	hard	-	x	-	x	-	x			-	x
l	soft	-	x	x	-	-?	x?	x	-	-	x
	hard	-	x	-	x	-	x	-	x	-	x
ch sch	soft	-	x	?	?	-	x	?	?	-	x
sh zh	hard	-	x	-н	xн	-	x	?	?	-	x
velar (k g x)	soft	-	x	? ньг	?	-	x	л(ь)к		x	-
	hard	x	x	-	x	-	x			-	x

• ^льн больница
• ^ньг деньги
• ^н before retroflexes, [ʂ ʐ], both [n] and [nʲ] appear (e.g. деньжонки 'money [dimunitive]' and Анжелика ('Angelica'))
• ^л(ь)к сколько vs. палка; сколькие vs палки (assuming these are pronounced as they are spelled)

This is more-or-less the information I've been operating under. I suppose we can assume that, for the question marked cells, palatalization is non-distinctive and not present but that's a questionable assumption given the example words I've provided. The grayed out box means that that sequence never occurs — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 08:46, 7 April 2010 (UTC)

Navigating through Halle's dense and obscure wording may be difficult to the Google books previewer, so here's the relevant phonological rules (which start on page 63 for anyone interested) boiled down:

Palatalization of [r(ʲ) t(ʲ) d(ʲ) s(ʲ) z(ʲ) p(ʲ) b(ʲ) f(ʲ) v(ʲ) n(ʲ) m(ʲ)] is distinctive only before vowels other than |e| and at the end of lexical morphemes and non-final suffixes.

Before [tɕ ɕɕ], only soft [mʲ pʲ bʲ fʲ vʲ] appear. Before [ʂ ʐ] only hard coronal consonants (non-liquids) appear.

Before hard [t d n s z l r], only hard [r p b f v m t d s z n] appear

Before soft [fʲ vʲ pʲ bʲ mʲ tʲ dʲ sʲ zʲ nʲ lʲ], only soft [tʲ dʲ sʲ zʲ nʲ] appear

Before soft (non-liquid) consonants, only hard [r p b f v m k g x] appear. Exceptions:

• [xʲ] appears before [kʲ gʲ]
• soft [pʲ bʲ fʲ vʲ mʲ] may appear before [pʲ bʲ fʲ vʲ mʲ]
• before [lʲ tʲ dʲ sʲ zʲ nʲ tɕ], only [p b f v m k g x] appear

I may be missing some deductive reasoning somewhere. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 09:03, 7 April 2010 (UTC)

I can't claim that [zʲ] before [vʲ] does not appear in any Russian word, maybe it does somewhere, but in all the examples I can remember it's pronounced [zvʲ]. Just like the name of the Russian city Tver is pronouced [tvʲerʲ]. If any book states otherwise, then the book is wrong. Hellerick (talk) 07:03, 22 August 2010 (UTC)

I don’t know which sources was used, but in standart russian pronunciation before zʲ (sʲ), v (fʲ) must be hard. For examples: зверь, звезда, сфинкс, связь, свекла and many-many others. Normal trancsription of the word зверь is given even in russian wiktionary. Tat1642 (talk) 08:53, 22 August 2010 (UTC)

The source used is Sound Pattern of Russian (linked above), which is a quite old source. Perhaps the language has changed since then, but I have yet to find a more recent source that goes into such great detail about the phonetic nuances of consonant clusters in Russian. Most clusters are clearly nondistinctive in regards to palatalization, which may (along with the orthography) lead native speakers to perceive them as hard even if they're actually phonetically soft. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 16:41, 22 August 2010 (UTC)

In fact, first consonant in palatalised св and зв may be both soft and hard in Russian, but the latter pronunciation is more common nowadays. E.g., in Розенталь Д. Э., Джанджакова Е. В., Кабанова Н. П. Справочник по правописанию, произношению, литературному редактированию it's said that “в сочетаниях зв и св могут смягчаться з и с: зверь, звенеть [з’в’] и [зв’]; свет, свеча, свидетель, святой [с’в] и [св’], а также в слове змея [з’м’] и [зм’]” (italised by me), thus they may be soft, but normally are hard. — Glebchik (talk) 20:20, 25 August 2010 (UTC)

My Russian isn't up to snuff. Is your quote saying that the hard pronunciation is more common or is it saying that there's free variation between the two forms? — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 23:22, 25 August 2010 (UTC)

In English it reads as “в) In combinations зм and св, з and с may be softened: зверь, звенеть [з’в’] and [зв’]; свет, свеча, свидетель, святой [с’в] and [св’], as well as in the word змея [з’м’] and [зм’]”. The same is said about тв and дв. To the contrary, it's said said that “н must be softened (обязательно смягчается) before soft з and s”, and “n before soft т and д are softened (смягчается)”. (§236. Произношение некоторых согласных) Thus in our case c and з just may be softened. — Glebchik (talk) 12:34, 26 August 2010 (UTC)

I couldn't find this rule on the website of Russian phonetics, but in pronunciation samples the hard з and с are used, e. g., звездный [зв'`ознъj], звездчатка - [зв'иш'ч'`атка], сверстник [св'`ерс'н'ьк]. — Glebchik (talk) 12:51, 26 August 2010 (UTC)

Russian pronunciation

Even though the Russian letter <Ж> sounds like a Voiced retroflex fricative, in many dialects in major cities (especially in Moscow, St. Petersburg, etc.) it is a Voiced postalveolar fricative. Here are the features of a Voiced postalveolar fricative:

• Its manner of articulation is fricative, vvhich means it is produced by constricting air flovv through a narrovv channel at the place of articulation, causing turbulence.
• Its place of articulation is postalveolar vvhich means it is articulated vvith the tip of the tongue betvveen the alveolar ridge and the palate, but closer to the alveolar ridge than for alveolo-palatal consonants.
• Its phonation type is voiced, vvhich means the vocal cords are vibrating during the articulation.
• It is an oral consonant, vvhich means air is allovved to escape through the mouth.
• It is a central consonant, vvhich means it is produced by allovving the airstream to flovv over the middle of the tongue, rather than the sides.
• The airstream mechanism is pulmonic egressive, vvhich means it is articulated by pushing air out of the lungs and through the vocal tract, rather than from the glottis or the mouth.

This includes many features in the letter Ж. So, vvhy is the Russian letter Ж classified ONLY in the phonology category of a Voiced retroflex fricative and not also in the category of a Voiced postalveolar fricative? There is a vvebsite that allovvs native speakers from their native countries to pronounce words (letters in this case) in fluent Russian, and here is "Zhe" (transliterised as either <žɛ>/<ʒɛ>, or in the case of a Voiced retroflex fricative: <ʐɛ>):

| Forvo Ж

Here is another vvebsite, shovving all the Russian letters pronounce by Russian natives: http://listen2russian.com/lesson01/a/

Please, If it is possible to consider BOTH VVIDELY KNOVVN DIALECTS OF RUSSIAN (Both vvhich are considered official to the USSR [Soviet Union] novv knovvn as the Russian Federation.) Thank you. :D

序名三「Jyonasan」 ^Talk_Stalk 01:07, 5 September 2010 (UTC)

Long consonants

Shouldn't the sounds for [щ'] and [ж'ж'] be written as [ɕː] and [ʑː] instead of [ɕɕ] and [ʑʑ]? Hellerick (talk) 15:42, 21 January 2011 (UTC)

What's the difference? — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 00:03, 22 January 2011 (UTC)

No difference, but I thought that's how long consonants are supposed to be designated in the IPA. Hellerick (talk) 05:23, 22 January 2011 (UTC)

AFAIK, they're normally equally correct. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 16:41, 22 January 2011 (UTC)

Transcription [щ'] and [ж'ж'] are influenced by spelling. In the majority of serious works on the Russian phonology [ш̅’], [ж̅’] are used. [ɕɕ] and [ʑʑ] look like two identical sounds are put together. [ɕː] and [ʑː] are preferable (ignoring the fact that [ʃː] and [ʒː] are more correct, Polish ś and Russian щ have not the same articulation, though…)--Luboslov Yezykin (talk) 05:51, 21 October 2011 (UTC)

Devoicing of /ɡ/.

According to note 8, When /ɡ/ loses its voicing, it is also lenited.

In the standard dialect, this is only true for a handful of words as far as I know. Among them are the word <Бог> "god," pronounced [box] and the adjectives <лёгкий> "light" and <мягкий> "soft" which are pronounced [ˈlʲɵxʲkʲɪj] and [ˈmʲæxʲkʲɪj] respectively. Otherwise, it's just pronounced [k] as one would expect. AlexanderKaras (talk) 05:15, 27 September 2010 (UTC)

It's not lenition but dissimilation before plosives, it happens only in words with -мягк-/-мягч- and -легк-/-легч-, and rarely in ногти, когти. The note 8 is not properly correct. Бог, Господь with [ɣ/x] is an ecclesiastic pronunciation, there is also no lenition, but rather a tradition. Luboslov Yezykin (talk) 08:23, 16 October 2010 (UTC)

It sounds like it's both lenition and dissimilation. So is [ɣ] a notable enough pronunciation that we should feature it here or is it too precise? Does [ɣʲ] ever occur?— Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 16:19, 16 October 2010 (UTC)

I was under the impression that [ɣ]/[ɣʲ] was a dialectal pronunciation of standard [ɡ]/[ɡʲ]. But that isn't the point. I'm talking about whether [ɡ] is always lenited to [x] when devoiced, or if this is a feature specific to a very few words. All the material I've read about Russian suggests the latter.

N.B. for dialects that realize /ɡ/ as [ɣ], this probably is the case. AlexanderKaras (talk) 14:05, 3 November 2010 (UTC)

Ooh, you have sources? Do you think you could share them? (either here or at Russian phonology) — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 14:08, 3 November 2010 (UTC)

Yes, [ɣ]/[x] instead of [g]/[k] is the last South-Russian dialectical feature, which still remains and is widely used, and it is well persistent to the normalization. If you live some time in the South Russia, you will surely listen this even from well-educated people, though they will pronounce other sounds quite properly according to the standard. Other South Russian pronunciation features such as yakanie (ему:[jʌ'mu]), not [ji'mu]), wekanie (бровь: [-w], not [-fʲ]), hwekanie (фартук: [xw-], not [f-]) etc. have nearly disappeared; only older (>60 yo.) less-educated people from the countryside speak in such a manner.--Luboslov Yezykin (talk) 05:43, 21 October 2011 (UTC)

Though I insist on dissimilation. /g/ in fact dissimilates to the next plosive becoming a fricative. During lenition (Latin vita > Spanish vida) a consonant assimilates becoming closer in articulation to the nearest sounds.--Luboslov Yezykin (talk) 05:43, 21 October 2011 (UTC)

In my pronunciation шаг and шах sound the same, but I am also a speaker of the Southern Russian dialect. -iopq (talk) 22:03, 30 May 2011 (UTC)

---

Note that only-[g]-dialects and only-[ɣ]-dialects are not all Russian language. Some mid-southern accents have them allophonic, using ɣ in final positions (where it devoices to [x]), but preferring [g] before vowels. Incnis Mrsi (talk) 11:24, 24 July 2012 (UTC)

СЧ and Щ

While some speakers pronounce words with <щ> as [ɕɕ] and some as [ɕtɕ], none contrast the two pronunciations, even in words where this sound is spelled with other letters. — It's generally but not always true. In some rare cases the prefix "с-" is still perceived as a separate morpheme and is pronounced as a separate sound, which is affected by the next sound, but does not merge with it. Thus, while a person generally pronounces Щ as [ɕɕ], he still would pronounce "считывать" with [ɕtɕ]. Curiously, the word "считать" may have two pronunciations depending on its meaning: "считать на пальцах" — with [ɕɕ] (because "счит" is treated as an indivisibile stem), and "считать с дискеты" — with [ɕtɕ] (because "с-" is a perfectivizing suffix speakers still aware of). Hellerick (talk) 11:35, 12 April 2011 (UTC)

So what's a good reword? — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 13:08, 12 April 2011 (UTC)

Dunno... Maybe something like this: "Generally the pronunciation of <сч> and <щ> is the same, either [ɕɕ] or [ɕtɕ], and does not depend on orthography. However the sounds [ɕɕ] and [ɕtɕ] may contrast in some morphological conditions." Hellerick (talk) 13:02, 13 April 2011 (UTC)

How about, "While some speakers pronounce words with <щ> as [ɕɕ] and some as [ɕtɕ], none contrast the two pronunciations. This generally includes words spelled with other letters, though speakers with the [ɕɕ] pronunciation may still pronounce words like считывать with [ɕtɕ] because of with a morpheme boundary between с and ч." — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 19:04, 13 April 2011 (UTC)

Well, sounds good. Strictly speaking though, the combination <сч> always has a morpheme boundary between the letters, it's just sometimes this boundary is "dead" and does not affect pronunciation anymore, and sometimes it's still "alive". Hellerick (talk) 11:14, 14 April 2011 (UTC)

I don't believe in dead morpheme boundaries. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 11:46, 14 April 2011 (UTC)

Hellerick is right, except that "it does not depend on orthography". A speaker uniformly pronouncing ⟨щ⟩ with only [ɕ]s (without [t]), may nevertheless distinguish between [ɕɕ] or [ɕtɕ] in ⟨сч⟩ on the grounds explained above. Incnis Mrsi (talk) 11:24, 24 July 2012 (UTC)

Reduced vowels /ɨ/ and soft /u/

Are they really not lowered from their stressed pronunciations, namely [ɨ] and [ʉ]? Maybe, near-close vowels [ɪ̈] and [ʊ̈] respectively are more appropriate, just like for [ɪ] and [ʊ]? Incnis Mrsi (talk) 11:24, 24 July 2012 (UTC)

They are, but it's not necessary IMHO to transcribe every vowel with absolute phonetic precision. We basically stop when being phonetically precise would require diacritics. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 13:11, 24 July 2012 (UTC)

Errors

I found many errors in example tables. Some of them mentioned above, in addition four top lines of unstressed vowel examples are incorrect (sounds in different words given are definitely not the same (i.e. this is the situation opposite to the aforementioned problems). Has this article undergone any professional review? What are credentials of the original contributors? --Dp074 (talk) 02:05, 4 February 2009 (UTC)

Can you be more explicit about what the errors are? I'm pretty much the major original contributor. Although I don't speak Russian, I've done a lot of research in editing Russian phonology and almost everything here is sourced or has its basing in sources. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 18:18, 4 February 2009 (UTC)

I gave quite specific examples above. I do not know what sources you have used, but you definitely should have double-checked them with Russian experts on the subject. I am not a trained linguist or philologist myself, but the problems with the page are obvious and amount to the inapplicability of this information. I can email you the most detailed information on the errors on this page - however I can fix not all of them myself --Dp074 (talk) 03:38, 22 March 2009 (UTC)

ɣ

Remove that "Господь" there: г can only "glottalise" in such a way when starting a stressed syllable! You'd better show the word "бухгалтер" there: /bu'ɣalt^jer/. JLincoln (talk) 13:49, 4 August 2012 (UTC)

Thank you, Ƶ§œš:) JLincoln (talk) 15:40, 4 August 2012 (UTC)

æ пять tan  ????!!!

English example, right... Maybe Japanese, but sure as hell not English! Пять >> pyat, tan >> тэн - no similarity anywhere. UNLESS this is a wisecrack from the Russian anime and internet community, which refers to women/girls/females as -тян (japanese: -chan, -tan). Aadieu (talk) 19:12, 14 January 2010 (UTC)

пять is pronounced [pʲætʲ] and English tan is pronounced [tʰæn]. I suppose we could change the English approximation to pat, since that doesn't feature nasalization of the vowel. Convinced? — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 22:52, 14 January 2010 (UTC)

I don't hear in пять (pronounced as пьать) tan, bad or sad sounds. It seems to me much more similar are sounds in car, but, dove. Russian readers in Russian Wikipedia express same doubts. But Russian linguists state that tan-sound is more exact. Common people and linguists are two completely different kind of people. From Russia with loveNice big guy (talk) 15:58, 28 May 2010 (UTC)

Since it's an allophone, native speakers of Russian aren't as likely to perceive the distinction. — Ƶ§œš¹ _{[aɪm ˈfɹ̠ˤʷɛ̃ɾ̃ˡi]} 20:20, 28 May 2010 (UTC)

Indeed. Knowing that this sound exists in the word "пять" is a good trick for Russians learn how to pronounce the sound /æ/ correctly. Hellerick (talk) 15:18, 29 May 2010 (UTC)

Hellerick, I don't think it's for Russian learners here at all. And Ƶ§œš was absolutely right. (I hope you don't forget it's all approximation, for it's hard to find identical sounds in different languages.) JLincoln (talk) 14:12, 4 August 2012 (UTC)

Pat can not be approximation for 'пять, because this word doesnt contain "æ" sound. I'm pretty sure, because i'm native russian speaker. Пять- [pʲatʲ] p'at' -pyat' — Preceding unsigned comment added by Eldavidok (talk • contribs) 13:45:26

@Eldavidok: To the right is the recording of пять from Wiktionary compared with папа. To me, it's clear the word have different vowels, [æ a]. Is this how you would say the words, and can you hear the difference? — Eru·tuon 19:43, 27 January 2016 (UTC)

bad comparison "папа" and "пять", because of "ь" sound , shortly "пять"="пьать", in russian p+y = p', therefore it sounds as p'a't'

as i said Пять- [pʲatʲ] p'at' -pyat' and no "æ" sound there. Im very sure of that. — Preceding unsigned comment added by Eldavidok (talk • contribs) 16:43, 19 February 2016 (UTC)

It's not a bad comparison. That's the point. When in the context of soft vowels, /a/ is fronted to [æ]. Being a native speaker means you're less likely to hear this. If it helps, what is often transcribed in English as /æ/ isn't exactly the same as this sound in Russian. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 18:46, 19 February 2016 (UTC)

Russian Pronounciation audio

I with my Russian students can create audio files for each line of text in this table, e.g.

• audio file 1 with text "бок; небо"
• audio file 2 with text "дом; деда"
• etc.

So I have two questions:

1. Does it help readers of this article to understand pronounciation?
2. Is it possible to add audio files at the end of each text line in the table? -- Andrew Krizhanovsky (talk) 17:00, 9 October 2013 (UTC)

I'm definitely for number 1.--2.245.187.254 (talk) 03:30, 31 December 2013 (UTC)

I'm not sure what the anon means, but it'd be a very good idea. You're welcome to do that if you're still interested. Peter238 (talk) 00:30, 27 February 2016 (UTC)

Allophony of /u/

Russian phonology says [u] becomes [ʊ] when unstressed. Shouldn't this be included?AlexanderKaras (talk) 07:58, 31 December 2009 (UTC)

I have asked about this issue on a linguistic forum and they told me that since the quality of the unstressed vowel is pretty much different from the stressed one it would be better to use a distinct character for it, i.e. [ʊ] for unstressed <у>. Hellerick (talk) 15:37, 1 January 2010 (UTC)

Yes, it's true that /u/ becomes [ʊ] when unstressed. I just thought it might not be a necessary detail. But if you guys think otherwise, it's not too difficult to go through articles that use {{IPA-ru}} and adjust it. — Ƶ§œš¹ _{[aɪm ˈfɻɛ̃ⁿdˡi]} 20:33, 1 January 2010 (UTC)

Sorry but you're all wrong. Save you've been talking about phonetics not phonology. Phonologically, Russian has only one /у/. Only one.
Phonetically? Oh, we could find thousands of "sounds" in Russian pronunciation — especially not being restricted to the Moscow region:)

Nice to neet you, Josh, linguist — JLincoln (talk) 14:01, 4 August 2012 (UTC)

I'm not sure how exactly we're wrong. This is indeed a common allophone of unstressed /u/. — Ƶ§œš¹ _{[ãːɱ ˈfɹ̠ˤʷɪ̃ə̃nlɪ]} 16:42, 4 August 2012 (UTC)

O'k, I see, the article's about IPA — phonetical alphabet; but the author of this section started with "Russian phonology says...":) Here's the incompatibility within our discussion:D JLincoln (talk) 10:57, 8 August 2012 (UTC)

Our "X phonology" articles are pretty much all about both phonology and phonetics. Some of them are even only about phonetics (e.g. Danish phonology), at least to a large extent. Mr KEBAB (talk) 11:17, 12 June 2016 (UTC)

May'2016 edits

(Note after rv-ing me:) Biggest mistake I've noticed before editing was «шея» (along with «пень») is used as an example for «e» phoneme. It's а vocal equivalent of «шэя»: Ш is always-hard consonant, so it's hardening a vowel. I proposed to replace it with «се́я», which is a better example of non-yotted stressed «e» — ['seja]/['sejæ]. So, why this is wrong? Tacit Murky (talk) 23:26, 1 May 2016 (UTC)

That I haven't reverted. Remember to compare the revisions. Mr KEBAB (talk) 23:36, 1 May 2016 (UTC)

All right, fair. But how come «pay» is better than «yes» ? The vowel is nowhere near Russian equivalent — «пей», but more like [пэй]. Moreover, below «choose» is used as an example for initial (therefore yotted) Ю, that is like «jʉ»=[you]. My proposal «лю́ди» shows only (non-yotted) ʉ, as required. Tacit Murky (talk) 00:29, 2 May 2016 (UTC)

It's not, and it is an exaggeration to say that it is "nowhere near" Russian [e]. English /ɛ/ is most commonly an open-mid-to-mid front unrounded vowel, with some dialects realizing it as close-mid. The starting vowel of English /eɪ/ is also variable, most commonly realized as open-mid-to-mid front unrounded, but some dialects pronounce it even more open ([æ]), or monophthongize this diphthong altogether ([eː]). So Russian [e] and [ɛ] should have only one English example word.

Changed to «лю́ди». Mr KEBAB (talk) 00:47, 2 May 2016 (UTC)

We also need a better example word for the unstressed [ʉ] (that is, with no <ю> in spelling). Any ideas? Mr KEBAB (talk) 00:52, 2 May 2016 (UTC)

According to https://en.wiktionary.org/wiki/yes , both UK and US English recommend [jɛs] as correct IPA transcript. So, I'd prefer it for ɛ, otherwise it wouldn't be useful for readers to distinguish ɛ and e, which is important for correct Russian pronounce. Otherwise, how would you demonstrate the difference between «сэр» and «сер»? Tacit Murky (talk) 01:26, 2 May 2016 (UTC)

«Ура́!», «усы́». To me (as native): not much difference between «у» in «ус» and «усы». Tacit Murky (talk) 01:15, 2 May 2016 (UTC)

(Please do not reply inside my message). The Wiktionary transcription is not [jɛs], but /jɛs/, which is a very important difference. You're confusing phonemic and phonetic transcription (and there are levels of depth of the latter) - see phonetic transcription. The important distinction is not between [ɛ] and [e], but between /s/ and /sʲ/, as only the latter is phonemic. Besides, we do transcribe both «сэр» and «сер» with the same vowel: [sɛr], [sʲɛr], as the close-mid allophone [e] appears only before soft consonants, and the mid allophone [ɛ̝] we transcribe the same as the open-mid allophone [ɛ] for simplicity.

But that's not unstressed [ʉ] but [ʊ] - the former appears only between two soft consonants. Mr KEBAB (talk) 01:36, 2 May 2016 (UTC)

I'd agree, but [e] must appear not before, but after soft consonants. Moreover, it's easier to explain, that soft vowel is affecting the softable consonant before itself (retrograde morphoponology), not that soft consonant is changing next vowel (progradely). (However, the later seems more scientific for linguists.)

If you want [ʉ] to appear between two soft consonants, that's not normally possible without [Ю]. «Люде́й» = [ль-у-дь-е-й]. Except if using always-soft consonants й/ч/щ: «чуде́са». But is there a difference for [у] with «чу́до» ? Tacit Murky (talk) 02:03, 2 May 2016 (UTC)

According to Jones & Ward it appears only before and between soft consonants.

Maybe that's an acceptable analysis for some scholars, I don't know. That's not how it is explained on Russian phonology.

There is, we wouldn't be using different symbols otherwise. <чудеса́> will do, thanks.

I suggest that you read Russian phonology, Jones & Ward (1969) - The Phonology of Russian and Yanushevskaya & Bunčić (2015) - Russian (the last one appeared in Journal of the International Phonetic Association 45 (2)). Mr KEBAB (talk) 02:11, 2 May 2016 (UTC)

Thanks for reading suggestions, I have a lot of it in Russian originals. And I noticed a lot of English-speaking scholars use outdated information about pronounces and dialects. Let's switch to consonants table here, shell we? Unlike 1960-s, Ж (as Ш and Ц) is considered to be always-hard now (i.e. 20xx urban Russian), and so recommends Russian literary norm (of Moscow school). So [ʑː] should be considered as outdated and rarely occurring. Even «дрожжи» is now ['droʐːɨ] — «дро́жжы».

Щ as [ɕtɕ] is obsolete, too (as noted in recent reference to Yanushevskaya & Bunčić). However, apart from [ɕː] or [ɕɕ], there is also non-elongated [ɕ], if Щ is sided by another consonant: «Мощный», «кровельщик». Not sure, but all other cases seem to extend [ɕ] to [ɕː]. Tacit Murky (talk) 02:45, 2 May 2016 (UTC)

Last suggestions for now: Note 7 is missing Ё, along with example cell for [j]. Then, yotting is default for Е-Я-Ё-Ю, except for most common case (vowel after consonant); other 4 cases are: initial letter, after vowel (including same one), after dash and after a sign letter (Ь or Ъ). (The term «Yer» in not used in Russian linguistics anymore.) Finally, И (non-yotted by default, in contrast to other soft vowels) may also became yotted, but only after Ь or Ъ: «судьи» = ['sudʲjɪ], «ручьи»; but no Slavic examples for «ъи» combination. Tacit Murky (talk) 04:03, 2 May 2016 (UTC)

(Next time, don't hurry me after 20-something hours of no response, right?)

I find it extremely dubious that any scholar would consider the sounds [ɛ] and [e] to be phonemic (/ɛ, e/) in Russian. The only Slavic languages that have these as separate phonemes are Upper Sorbian, Lower Sorbian and Slovene, and maybe some regional dialects (Kajkavian?). So you probably misunderstood someone, or maybe read (past tense) someone who is a fan of a very unusual analysis, an analysis which wouldn't be considered "correct" by many.

Yep, Yanushevskaya & Buncic (2015:224) confirm that [ʑː] is "somewhat obsolete", so we can replace it with [ʐː]. Will do it in a second.

Yanushevskaya & Buncic (2015:223) mention short [ɕ], which is an allophone of /s/ in e.g. <с чаеm>, so we can add it.

However, I'm not sure about [ɕ] occuring in e.g. «мощный», «кровельщик» - we'd need a source for that. Russian Wiktionary transcribes the former as long, whereas on Forvo, «мощный» has a short [ɕ], whereas «кровельщик» clearly has a long [ɕː]. So, unfortunately, native speaker's opinion won't suffice in this case, as it may be a mere free variation.

The same goes for yotting И, which we also need a source for. On Forvo, «судьи» is clearly pronounced [ˈsudʲɪ], not [ˈsudʲjɪ], but «ручьи» is pronounced [rʊˈt͡ɕji], maybe because it is stressed. So, as above, your words won't suffice, unfortunately.

I can add Ё, but I'm not sure if we need the rest. It's not Russian alphabet, we don't need detailed explanation on the relationship between Russian sounds and the spelling.

Oh, and the relevant thing is not whether the term "yer" is used in Russian writings, but whether it is used in English writings. This is English WP. Mr KEBAB (talk) 14:13, 3 May 2016 (UTC)

The note that you added ё to says "When these vowels are unstressed, the /j/ may be deleted." If we really want to add it there, that might need to be reworded. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 14:58, 3 May 2016 (UTC)

Better? I hope that's what you meant. If I only had Jones & Ward (1969) with me now... :P Mr KEBAB (talk) 15:14, 3 May 2016 (UTC)

I agree with «dubious that any scholar would consider the sounds [ɛ] and [e] to be phonemic». I did confuse [*] with /*/, and Moscow school is using „morphoneme“ category to describe their statement «only 5 vowels in Russian». Denoting allophones would be redundant, but without it marking «сер» as [sʲɛr] (in Cyrillic phonetic — [с'эр]) may be misleading for non-linguists like me. Traditional Russian lessons in schools (including «Russian as foreign») insist that vowels are in charge of consonants (by softing — palatalization), therefore (presumably) there must be a difference between phonemes for (non-yotted) Е and Э letters. However, I'm still confused about subtle details.

Is there a point to «replace it with [ʐː]», instead of removing from the table? What is the point of keeping examples of long consonants, especially since they are marked by double letters («жж» in this case)? [nː] is much more common, but it's not in the table.

Example of «short [ɕ], which is an allophone of /s/ in e.g. <с чаеm>» is correct, but rare and obscure. Unlike „normal“ consonant morphing («счастье» ⇒ «щчастье» with plain Щ ⇒ «щастье» with long Щ), cross-word morphing is restricted to only 2 versions here: pure phonetic «с чаем» ([s't͡ɕa(j)em] or [s't͡ɕaæm]) and morphed «щ чаем» (with [ɕt͡ɕ+]), but no «щаем» (with [ɕː+]).

I don't see phonetic transcription for «кровельщик» here. But here are more words (and cases) present in Forvo: крановщик, точильщик, хищник, сообщник (CCC cluster), умерщвление (CCCC !), поощрение (somewhat long here). However — «вещмешок» maybe pronounces with [ɕː], because it's a colloquial slang for «вещевой мешок» (haversack, knapsack, kit-bag); and boundary between 2 stems allow to violate some morphophonology (just like «ИнЯз» is not [inʲ'as], but [in'jas] without Ъ).

«The same goes for yotting И» — much easier for this case. Use this page (this site helps to find words for crossword puzzles), enter letter mask *ьи* and you'll see a lot of examples with many of them present in Forvo: Ильин («Иван Ильин» is very clear), Ильич (massive of them), оладьи, соловьи, Ильичёвск (toponim, but Slavic one). «On Forvo, «судьи» is clearly pronounced [ˈsudʲɪ]» — right, and that is an obvious mispronounce.

As a matter of fact, it 'is' possible to leave Ё unstressed (without converting it to Е), but only in an extreme case: «ёфик'ация», rare new word, meaning «converting Е's то Ё's were needed in a text».

Maybe we should add a notion, that Ц-Ж-Ш are considered to be always-hard and non-palatalizable, even in presence of soft vowels. (Except old-style [ʑ] for Ж.)

(Isn't this message too long to keep it whole?) —Tacit Murky (talk) 01:27, 4 May 2016 (UTC)

No, it's not too long, but we seem not to understand each other that well.

The Moscow school states that there are five phonemes: /i, u, e, o, a/. You must be confusing phonology with orthography, or you simply misunderstood words of some scholar. If somebody says that there are /e, ɛ/ phonemes in Russian, that's not Moscow school - that's his own interpretation, and an unpopular one.

I replaced because of the former pronunciation with [ʑː]. The presence of [ʐː] allows for a note explaining that [ʑː] is somewhat obsolete. EDIT: I think you're right, I've just removed [ʐː] from the table. A note next to the short [ʐ] should be enough.

We're talking about allophones here, not morphology. I'm pretty sure that both of your transcriptions are wrong; phonemically, it is /s ˈt͡ɕajem/, phonetically - [ɕ ˈt͡ɕæ(j)ɪm]. [æ] appears only between soft consonants.

I said "Russian Wiktionary transcribes the former as long", so I was referring to «мощный», not «кровельщик» (which is not transcribed in IPA on Russian Wiktionary, and which I would have to call "the latter", as it was the second word that I mentioned).

Let's judge the recordings:

- крановщик - two out of three pronunciations are long

- точильщик - long

- хищник - long

- сообщник - sounds short to me

- умерщвление - this may be short (it also sounds abnormally hard to me, but that's probably just my non-native hearing)

- поощрение - long, as you say

- вещмешок - long, as you say

- Иван Ильин - does have [j], as you say

Slightly elongated [ɕˑ] still counts as long, not short. I'm afraid that you're just mishearing some of these, so, until you present reliable sources to back up your claims, nothing can be done in this regard - especially since what we're doing is WP:OR judgement of Forvo recordings which, as you yourself admit ("«On Forvo, «судьи» is clearly pronounced [ˈsudʲɪ]» — right, and that is an obvious mispronounce."), are also unreliable themselves!

I'm not sure about yotting И - maybe other editors would want to comment on that. I'm not sure whether this is the appropriate place for such information.

What we were talking about is [j]-deletion in unstressed <ë>, not whether <ë> can be unstressed... or at least that's my understanding. Maybe I'm wrong. EDIT: Apparently I am wrong.

Added. Mr KEBAB (talk) 11:03, 4 May 2016 (UTC)

So, in case «We're talking about allophones here, not morphology», here's something I find strange: English Wiktionary shows «change» as /tʃeɪndʒ/, despite clear sounding (especially for UK-En) with soft /nʲ/. Likewise «million» is also with hard /l/. So how came Russian version show this word as ['ʧænʤ] (still hard n) and then «есть» as [jesʲtʲ] (with proper palatalization marks)? So, does square brackets have different meaning?

OK, let's use «phonetically - [ɕ ˈt͡ɕæ(j)ɪm]» as one case, but my point was that it's not the only one. Preposition still may sound as [s], which is how I and my surroundings pronounce this.

You say «not sure whether this is the appropriate place for such information», but other yotting examples are there, so we should make it complete with «ьи»/«ъи» (even that later is not used anywhere).

Looks like «лёгкий» may not be perfect example for lenition + palatalization of Г into [xʲ]. Listen to Forvo's clips for «лёгкий» (лёгкие, лёгким, лёгких and «воспаление лёгких» with 2 indicative cases) and «мягкий» (plus more cases with «мягк-»), and you'd find some 50-50 distribution between [x] and [xʲ]. I suggest to replace with «легко́» — lenition Г-Х applies without palatalization. Tacit Murky (talk) 02:09, 5 May 2016 (UTC)

But we've already talked about this. I told you that "The Wiktionary transcription is not [jɛs], but /jɛs/, which is a very important difference. You're confusing phonemic and phonetic transcription (and there are levels of depth of the latter) - see phonetic transcription. The important distinction is not between [ɛ] and [e], but between /s/ and /sʲ/, as only the latter is phonemic."

Anyway:

- English does not have the /nʲ/ phoneme, but Russian does. You can't transcribe change as /tʃeɪnʲdʒ/, because in English, [nʲ] (really, it's just laminal postalveolar [n̠] with optional palatalization [n̠ʲ]. Some speakers pronounce English /tʃ, dʒ, ʃ, ʒ/ quite hard (similar to Russian and Polish), and so there probably is little to no palatalization ([n̠]) for them.) is merely an allophone of /n/, not a phoneme.

- There are levels of depth of the phonetic transcription. The General American pronunciation of change is, in narrow transcription, [t͡ʃʲʰe̞ˑɪ̯n̠ʲˑd̥͡ʒ̥ʲ ~ t͡ʃʰe̞ˑɪ̯n̠ˑd̥͡ʒ̥], with the first affricate aspirated, the first element of the diphthong and the nasal somewhat lenghtened, both affricates and nasal somewhat palatalized (or not - this depends on the speaker, I think weak palatalization is, however, pretty normal), and the final affricate strongly devoiced, unless the next word begins with a voiced sound. Normally, however, we're happy with [t͡ʃeˑɪ̯nˑd͡ʒ] (if we want to show the lack of pre-fortis clipping) or [t͡ʃeɪ̯nd̥͡ʒ̥] (if we want to show the final devoicing).

- The Russian IPA was totally incorrect, so I fixed it.

- Once again, please read phoneme, allophone and phonetic transcription. The difference between /this type of transcription/ and [this type of transcription] is dramatic.

Ok, fair enough. Looks like we don't really need to include the short [ɕ] after all...

I'd rather let other editors comment on that. Mr KEBAB (talk) 13:56, 5 May 2016 (UTC)

Thanks, much more clear now with this explanation. But should we add a statement in the intro about level of allophonic depth and coverage in transcriptions in this article?

Maybe we should make a link to «Iotation». Moreover, I think Note 1 better be placed above consonant table as a plain text, as it is most important. And we should update Russian alphabet article as well — after all we did here :) Tacit Murky (talk) 23:21, 5 May 2016 (UTC)

Should we replace «fish ship» with «fish sheep» ? Because later has [i] instead of [ɪ], which is a better palatalizer. Tacit Murky (talk) 23:45, 6 May 2016 (UTC)

What's a fish sheep? Our examples have to be in proper English. Mr KEBAB (talk) 10:02, 7 May 2016 (UTC)

OK, but Russian_alphabet article has «sheer» as an example for Щ. Should it be consistent across pages? Also see other cells in the „Alphabet“ table. Tacit Murky (talk) 20:46, 7 May 2016 (UTC)

Russian alphabet should not have such half-assed approximations at all. We should reduce their amount to what we have on this page.

I don't see much of a point in consistency, as long as the examples are correct.

Changed to "wish sheep", as in e.g. "I wish sheep could fly". Mr KEBAB (talk) 15:25, 8 May 2016 (UTC)

Why «yes» and «pay» are «too variable, should have only one example word», if we are denoting allophone details? «Yes» is most appropriate for «пень», as it has close-mid [e], not open-mid [ɛ] of «met». Tacit Murky (talk) 17:02, 11 May 2016 (UTC)

"«Yes» is most appropriate for «пень», as it has close-mid [e], not open-mid [ɛ] of «met»" - This is the third time you're writting the same, incorrect stuff. The answers are above. Please pay closer attention next time. Unless you're trying to tell me that Russian vowel allophony of /e/ applies to English? That, obviously, is not true. Mr KEBAB (talk) 17:38, 11 May 2016 (UTC)

So, notion of allophony is different in English and Russian. It seems awkward: if we are making English article about Russian, what is the point of using English allophony, if it can't represent correct phone? I mean the difference between correct [sʲer] and [sʲɛr], assumed by «met». If I'd tech someone to pronounce unyotted /e/ correctly, vowel of «yes» seems to be best example. «Yemen» in both Rus and Eng is supposed to be pronounced the same, however, both «Йемен» and *«Емен» give that result. Is there a better word for „Russian-style [e]“, where E is not follows Y? Tacit Murky (talk) 02:31, 12 May 2016 (UTC)

It's called "doing our best". English is not Russian and Russian is not English. The rest, as I said, has already been answered. Look above please, there's no point in me repeating myself. Mr KEBAB (talk) 02:42, 12 May 2016 (UTC)

All right, I suppose, the answer to my last question about „Russian-style [e]“ is «none». Now, was there a debate about adding a glottal stop? Sometimes it appears at the starting vowel and between words or hyphened word-parts as a vowel delimiter. It may have some phonemic weight, marking the difference in «a-a» interjection between [aː] („got it“, „oh well“) and ['aʔa] („no!“ = «не-а», also with [ʔ]). GS is mentioned here. Tacit Murky (talk) 02:58, 12 May 2016 (UTC)

I have no idea why you won't just scroll up and re-read my messages again (hint: one of them is my second response to you), rather than write me a third message based on your suppositions. After all, it's the facts we want, no?

I don't know about the glottal stop, I'll let other editors comment on that. Mr KEBAB (talk) 03:01, 12 May 2016 (UTC)

Yes, I did re-read your explanations, but one thing still unclear. Usually, morphphonological rules of a certain language still apply to some degree to extreme cases: voice mimicking (onomatopoeia), interjections and/or exclamations, recently loaned (not-yet-adapted) words and transliterations. This later case is most interesting: we can type «йес» in Russian, expecting it to sound as close as possible to original. But what phonemic and phonetic transcriptions would look like for this „word“? Despite the alternative allophony, one might expect minimal or no difference with the original, right? Tacit Murky (talk) 05:40, 12 May 2016 (UTC)

Russian <йес> would be [jɛ̝s] phonetically (note a mid [ɛ̝], which you can also write [e̞]) and /jes/ phonemically. It'd sound almost completely native, yes.

English /ɛ/ is typically an open-mid/mid vowel, as I already said, so the typical English pronunciation is [jɛs ~ jɛ̝s], with the open-mid/mid vowel height pretty much in free variation (especially in Received Pronunciation). This is not the case in Russian, in which the open-mid, mid and close-mid allophones appear only when certain conditions are met (unlike in English). Mr KEBAB (talk) 08:55, 12 May 2016 (UTC)

Not sure, but I think I've finally found an alternative for «met» to show actual (little varying) [ɛ̝] as «е» in English: Cheyne or Cheney; if diphthongs shouldn't be allowed, then there is Chechnya or Chechen. Reason: /ɛ/ is surrounded by (somewhat) palatalized /tʃ/ and /j/ (if later is not diphthongized to /ɪ/). That'll make English /ɛ/ very close to Russian /ɛ/ in «чей». Is it better than «met»? Tacit Murky (talk) 22:42, 16 September 2016 (UTC)

It's not. Diphthongs are not allowed, and English palato-alveolars do not raise English /ɛ/ like Russian soft consonants. Mr KEBAB (talk) 07:45, 17 September 2016 (UTC)

я

Im not linguist or russian language specialist, but wouldn't a better english approximation for "я" in пять be more like "piet" or something like that — Preceding unsigned comment added by Petergstrom (talk • contribs) 21:59, 18 December 2016 (UTC)

«Я» is a letter denoting 1 or 2 sounds: (optional) /j/ and /a/ (with some allophones for it); «piet» is not a proper word in English (unless you mean some of that: Piet ) and it encodes several vowel sounds (including possible diphthong). That's a bad choice for an explanation. Tacit Murky (talk) 22:38, 18 December 2016 (UTC)

Reference check needed

Could someone check Lightner (1972) "Problems in the Theory of Phonology, I: Russian phonology and Turkish phonology" and see whether on page 67 he actually says that Russian has retroflex (hard postalveolar) affricates? No other sources say that AFAIK. Mr KEBAB (talk) 15:16, 11 June 2016 (UTC)

I just sent a request to researchgate.net to get a full text for checking; need to wait. Meanwhile, since we are not shy to add loans, maybe «soft Ц» would do as well? As explained [[1]], there are some words with apparent [tsʲ] for some dialects (especially southern): Пацюк, Цюрупа (Ukrainian family names), Цюрих (city), хуацяо (Chinese word). See Forvo's examples for these. Tacit Murky (talk) 17:37, 11 June 2016 (UTC)

I've read Lightner (1972) (and am the contributing editor to the claim in question). I would find the claim of hard affricates a little surprising as well, but the claim is not that Russian has such an affricate. At Russian phonology (and, if my past competence surpasses my recall from 8 years ago, in Lightner 1972), /dʐ/ is described as a sequence, not an affricate. For many languages, there is no distinction between the two. But, for Russian (as well as its sister language, Polish), there is indeed a contrast between a sequence of a stop and a sibilant and an affricate. The Russian phonology article doesn't articulate this difference, but it would be easy to do so. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 05:12, 12 June 2016 (UTC)

So the source doesn't claim that. Looks like our anon misread something... or simply lied. Who knows. Thank you, Aeusoes. Mr KEBAB (talk) 11:03, 12 June 2016 (UTC)

This difference «between a sequence of a stop and a sibilant and an affricate» is described in quite a detail in Russian article for Affricates, a cite from Н.Трубецкой (1960). These rules are claimed to be universal, not just Russian-specific. Tacit Murky (talk) 11:11, 12 June 2016 (UTC)

Отсек [ɐtʲ͡sʲˈsʲek] is an example I've found for the allophonic [tʲ͡sʲ], a non-sequential affricate. It should be a genuine soft counterpart for a nominal «ц» [t͡s]. However, other people may pronounce this cluster as [t͡ssʲ] or sequential [tsʲ] (with a clear stop). Not sure about adding this… — Tacit Murky (talk) 09:03, 11 July 2017 (UTC)

Move discussion in progress

There is a move discussion in progress on Help talk:IPA which affects this page. Please participate on that page and not in this talk page section. Thank you. —RMCD bot 16:18, 15 July 2017 (UTC)

Affricates

Thanks for explanations about some consonants, but I still don't understand why [d͡ʐ] and [t͡ʂ] have been removed. They actually are separate phones. Moreover, they appears only within one morpheme. For example, поджидать [pədʐɨˈdatʲ] but джем [d͡ʐɛm], артшкола [ɐrtˈʂkolə] but лучший [ˈɫut͡ʂɨj] (it can be listened) and cognates, музыкантша [mʊzɨˈkant͡ʂə]. [d͡zʲ] appears in same cases: подземелье [pədzʲɪˈmʲelʲjə] but дзинькать [ˈd͡zʲinʲkətʲ]. Maybe they should be in the table? 176.103.81.104 (talk) 17:16, 11 June 2016 (UTC)

Some of them are 2 consonants («отцы» = [ɐtː'sɨ]; Rus.Wiktionary propose [ɐɗ̥ˈt͡sɨ] , but I disagree with voiced ɗ̥ in there — it should be devoiced by both t and s); others are true affricates («джинсы» = ['d͡ʐɨnsɨ], but Eng.Wiktionary has it [ˈd͡ʐʐɨnsɨ], go figure; BTW, in both cases unstressed Ы is still [ɨ]). «Музыкантша» has a boundary between suffixes -ант (action performer) and -ш (noun feminizer), so it's [tʂ] here, too. Unfortunately, «лучший» falls out of the rule here; for my ears, „чш“ for this case can be an affricate only as a combination with a standalone consonant: [ˈɫut͡ʂʂɨj]. Tacit Murky (talk) 17:50, 11 June 2016 (UTC)

English Wiktionary transcribes automatically. I can call to mind only one word's root with «чш». It's «лучш». However, «чш» can appear at the junction of morphemes, for example, начштаба that pronounced [nat͡ɕ ˈʂtabə]. I have to apologize, I've written that affricates (that not allophones) appears only within one morpheme. But in mentioned музыкантша «тш» is pronounced [t͡ʂ], because it is at the junction of one type of morphemes. Thus [t͡ʂ] should be in the table as separate phone, I think. In similar article in Russian Wikipedia it is designated as [ʧ]. That's incorrect, however. [d͡ʐ] and [d͡zʲ] are forming in same cases. They appears pretty rarely, but I think that it's enough already. 176.103.81.104 (talk) 19:14, 11 June 2016 (UTC)

Looks like there is no other root with «чш». But there is more complex case: Маньчжур+ (-ия, -ский; and so on) — «Ч» is voiced within presence of 2 voiced consonants. Possibly, other Chinese or Eastern toponims might have «чш» combination. Good cases with «цс»: соцстрах (-ование), соцсоревнование, спецслужба, спецсубъект. And a curious case: «танцзал»; almost [nd͡ʐː]. About «музыкантша»: [t͡ʂ] in these examples is very well fused, unlike «тш» here with clear [ʂ]; here's Forvo's example. Moreover, casual fast speech elides [t] in such clusters to form [nʂ]: here. Tacit Murky (talk) 22:34, 11 June 2016 (UTC)

A few comments. As I mentioned in the above section, Russian instances of a stop followed by a hard sibilant are likely stop+fricative sequences, rather than actual affricates.

Also, [ɗ̥] is devoiced. That's what the little circle at the bottom means. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 05:31, 12 June 2016 (UTC)

This case is harder than it seems. Most verbs with reflexive suffix -ся/-сь have a combination of -тся/-ться in the end. Despite an explicit digraph and morpheme boundary, morphophonologically it's a case of Ц (which is always hard): both Т and С become hardened, so both cases are [-tsa], which is close to -ца in pronounce. This is the reason why it can be hard for scholars (even native Russians) to distinguish between these cases: -ться is used only with infinitives, and [s] can be soft even in the same verb — одевать (inf; normal [tʲ]), одевайся (imperative+refl; normal [sʲ]), одеваться (inf+refl; hard [ts]), одевается (singular+refl; hard [ts]). This may be a kind of exception, as other combinations of -тс- can have phonetically clean [tsʲ] pronounce (which is not possible for true or „simulated“ affricate): отсидеть, отсев. But: «отсю́да» (= from here) can be respelled as «отсу́да» (see here), which can be another case of implicit morphophonology for Ц, but more likely is a consequence of an alternative spelling for «сюда» as «суда́» (= (to) here). Tacit Murky (talk) 12:25, 12 June 2016 (UTC)

Look, 176.103.81.104. While your comments are appreciated, you still haven't provided a source to back up your claims. Moreover, you lied about (or misread?) Lightner (1972) confirming the existence of the voiced retroflex affricate in Russian (his claim is that it is a stop-fricative sequence). I think we should end this discussion, as the affricate retroflexes won't be included in the table unless you (or anyone else) provide a source, and someone else confirms its reliability (which is reasonable to ask for in this situation - see above). Mr KEBAB (talk) 11:10, 12 June 2016 (UTC)

I am not going to deceive somebody. This source I've found in corresponding article. Because it has been reviewed I've thought that verification was not necessary. However, I have some sources. Soviet linguist Mikhail Panov in “Russian phonetics” (Русская фонетика, 1967, page 35) has mentioned about a phone that he named «[ч] твёрдый» or “hard [ч]”. He has given three examples with cyrillic transcription. It is лучше [лу́чшъ], Ницше [нʼи́чшъ] and Ротшильд [ро́чшылʼт]. According to traditional Russian transcription, phone [t͡ɕ] designates as [чʼ] (che with an apostrophe). For example, человек [чʼилавʼэ́к] in IPA is [t͡ɕɪɫɐˈvʲek]. [ч] (che) signifies a phone that in IPA designates as [t͡ʂ] today. Transcriptions of example words in IPA looks [ˈɫut͡ʂʂə], [ˈnʲit͡ʂʂə] and [ˈrot͡ʂʂɨlʲt]. See page scan (Russian). Is it a reliable source? I shall try to find sources for other cases. But I think that it can be only in modern literature. 176.103.81.104 (talk) 16:27, 12 June 2016 (UTC)

Then we need to decide whether [t͡ʂʂ] is a long type of affricate, or it's just [tʂː]. First case is more viable, since [t] in these combinations is not laminal dento-alveolar [t̪] (i.e. standard Russian Т), but more like normal English [t], which is the case with „short“ affricate [t͡ʂ]. However, all of these cases also present clear [ʂ] just as a separate letter would produce. Same with other „long affricates“. Tacit Murky (talk) 23:22, 12 June 2016 (UTC)

For my ears it's just [t͡ʂ] at all. I agree that it phone is not simple [t] + [ʂ], because articulation is completely different (idem book, page 34 as a source). In Russian language phone [t] is laminal denti-alveolar. [t] in [t͡ʂ] is alveolar rather, you're right. I think that [t͡ʂʂ] is more correct variant. But I repeat that it is rather just [t͡ʂ] for me. 176.103.81.104 (talk) 07:52, 13 June 2016 (UTC)

The articulation is not completely different, it's only slightly different. In [t̪], the blade of the tongue touches the alveolar ridge and the tip of the tongue touches the back of the upper front teeth. In [t], there's no tongue-tip contact with the teeth. That's the only difference. Mr KEBAB (talk) 01:15, 13 July 2017 (UTC)

I've made experimental addition with Лучше and Отсек per this section. 176.103.81.104's arguments sound legit. Since we have notes 1 and 2 about allophonic appearances of [dz] and [dʑ], it should be extended to other hard-soft pairs of allophonic affricates. [d͡zʲ] and [d͡ʐ] are awaiting. — Tacit Murky (talk) 10:34, 12 July 2017 (UTC)

@Mr KEBAB, is M.Panov's work „Русская фонетика“ (mentioned in the comment above — 16:27, 12 June 2016) not reliable as a source for [t͡ʂ]? Russian biorgaphy says: «В 1968 году получил степень доктора наук за исследование „Русская фонетика“.» — «In 1968 [he] received doctor's degree for „Russian phonetics“.» The book can be DL-ed here. Section 61 (pp. 32-34) describe affricate [t͡ʂ] (очумел) and the difference with cluster [tʂ] (отшумел); a table of consonants mentions [t͡s t͡ʲsʲ d͡z] dentals and [t͡ʂ t͡ʲɕ d͡ʲʑ] DA's. Section 62 (pp. 34-35) is devoted to allophonic affricates with many examples: [d͡z] in Шпицберген, спецзадание; [d͡ʲʑ] in начбазы, испечь бы; dialectical [t͡ʲsʲ] in цвет; and [t͡ʂ] in лучше, Ницше, Ротшильд. That should be enough.

About two sources for [ɖʐ͡]: yes, Lightner is claiming about this being a sequence, not affricate; interestingly, Hamann (2004:65) also don't mention any «jam» words neither in Polish nor Russian, investigating only fricatives, as paper's name implies. This may be a call to question this source too in «d͡z» article. — Tacit Murky (talk) 12:47, 12 July 2017 (UTC)

@Tacit Murky: (Writing @Mr KEBAB doesn't notify me, you need to use the ping template instead). Panov only provides examples of the alleged affricate [tʂ] followed by [ʂ]. My Russian is bad, but I think he considers it to be a mere allophone of /t/ before /ʂ/, which is a valid analysis. However, it's not the only possible one. Stops followed by homorganic fricatives tend to form affricates with a lengthened fricative component no matter the language (there may be some exceptions though), so [tʂ] seems like a good enough transcription to me. The fact that the stop is laminal alveolar in that sequence is irrelevant - it's just a slight allophonic retraction from the default laminal denti-alveolar articulation.

There's a good argument not to include the phonetic affricates [tʂ, dʐ, tsʲ, dzʲ] in this guide. The JIPA article on Russian states explicitly that these consonants are non-standard and do not occur in standard Russian, only in local dialects. It's only 2 years old and so we should consider it a more reliable source than the ones you mention.

I see that you also brought up the case of the word очумел, which you say contrasts with отшумел. But according to Russian Wiktionary, the former word is pronounced with a soft [tɕ]. Forvo confirms that. I'm afraid that Panov's paper may be simply outdated in some aspects. Mr KEBAB (talk) 01:15, 13 July 2017 (UTC)

Yes, idea was that fluent /tʂ/ must be [t͡ʂʂ] with apical [t̺], not laminal [t̪] (produced by /t/ in isolation). But other allophones are in this article's table, so at least it's worth noting below the table.

I'm curious about what are these dialects with above-mentioned 4 allophones? Because modern literary Russian is almost identical to Moscow dialect, widely known by TV/movie speech and recommended in schools for both natives and foreigners. So, we can hear a variety of «лучше» samples, including Moscow's pronounce. Do the sound non-standard?

Yes, «очумел» involves [tʲ͡ɕ] affricate. Panov should have been used «лучше» for [tʂ͡] example as a contrast. Later in that paragraph he made a better pair with «отсадит»/«оценит» for /ts/. — Tacit Murky (talk) 13:11, 13 July 2017 (UTC)

@Tacit Murky: I really doubt that the difference is apical vs. laminal. Russian /ʂ/ is pretty much the same as Polish /ʂ/, which is laminal postalveolar. Therefore, the difference between /t/ before /ʂ/ and the default allophone of /t/ is (post)alveolar vs. dental. They're both articulated with the blade of the tongue, so they're both laminal. I don't find it that notable at all, as I said this affrication happens in pretty much any language I'm aware of and it's automatic.

The source doesn't say, but what they mean in the case of [tʂ, dʐ] are hard realizations of /tɕ/, so it's about something else.

The очумел - отшумел contrast would be truly phonemic if the former word was pronounced with an affricate [tʂ]. That's what I meant by 'contrast'. I'm pretty sure that the correct phonemic transcription of лучше is /ˈlut.ʂa/ (or perhaps /ˈlut.ʂi ~ ˈlut.ʂɨ/), with a syllable boundary after /t/. Mr KEBAB (talk) 06:46, 16 July 2017 (UTC)

I'm not sure how to distinguish laminal and apical fricative (especially in case of [ʂ]). But this «They're both articulated with the blade of the tongue» is clearly not always true, because /tʂ/ may be pronounced like this on the morpheme boundary (like отшумел with от- prefix), but not within (as for лучше). And Russian consonants are not dental, they are denti-alveolar.

IMO, correct phonemic transcription of лучше is /ˈlut.ʂe/ (because |-е| is a productive suffix, see here). According to Russian rules of syllabification, word-break between lines should be луч-ше, but syllable breaks can be both луч-ше and лу-чше, and later is preferred (§2). Tacit Murky (talk) 17:36, 16 July 2017 (UTC)

@Tacit Murky: How is it not true? I'd like to see some proof of that. You said it yourself that you can't distinguish laminal and apical fricatives, so I'm not sure why your statement is so definitive.

Morphology/grammar in general may or may not be taken into account when discussing phonology. Clearly, there's not one 'correct' phonemic analysis of this word but several. Mr KEBAB (talk) 18:00, 16 July 2017 (UTC)

Your quoted statement was about /t/, for which there is a clear difference for laminal and apical sounds. I'm arguing that in fluent speech laminal /t/ becomes apical, when followed by /ʂ/; otherwise it wouldn't be possible do hear and feel the difference between stop+fricative and a true affricate, like «Джон» and «John» (assuming voiced pair acts the same). Tacit Murky (talk) 19:27, 16 July 2017 (UTC)

@Tacit Murky: That's almost completely wrong. The difference between the stop-fricative sequence /tʂ/ and the dialectal affricate /tʂ/ (which doesn't occur in standard Russian as a phoneme) is mostly the length of the fricative (an affricate is a stop released into a fricative). And again, prove that /t/ is apical before /ʂ/ which, as far as I know, is not apical itself. Mr KEBAB (talk) 20:17, 16 July 2017 (UTC)

Yes, it is mostly the length of the fricative; but stop also sounds differently. As for /ʂ/, I've yet to hear any difference between «apical /ʂ/» and «laminal /ʂ/», even not from the same language. I'm comparing sounds and tongue positions for typical Romance /t/ (apical alveolar) with East-Slavic /t/ in isolation (laminal) and before /ʂ/. This later case sounds and feels much more like the first one. Panov mentions this effect on page 35 (§63): «Comparing pronounces of „от сора“ and „от шума“, it's easy to gear different „т“'s. First one is pronounced by tongue touching the teeth, second — by touching alveolae above the teeth. We'll call this one [т́]. It's like [ш], front-tongued front-palated.» (In Latin terms this means coronal palato-alveolar. Obviously, Panov translates «ш» into [ʃ], not retroflex [ʂ].) «Also, in words like вонжу, ханжа we can hear nasal [н́] with similar place of articulation as for [ш, ж]. How frequent this pronounce is and is it the only norm for these words — it's hard to say. Same place of articulation has plosive element of [ч'].» (I find this last statement very questionable.) And a note for this paragraph: «Л.И.Жирков (Лингвистический словарь / Linguistic dictionary; Moscow, 1946, p. 130) also noted that [т, д, н] after [щ, ж] are also front-palated: ш[т́]ук/с[т]ук, ж[д́]ал/с[д]ал, ж[н́]у/з[н]ал. It's not clear, if this is obligatory for literary Russian, or it's just one of possible cases.» Well, for me there is no difference in Zhirkov's 3 pairs. But [n] in вонжу and ханжа does sounds distinct. Curiously, this trick does not work with /lʂ/ and /lʐ/: волшебный generates regular (velarized lamial) /l/, despite the same place of articulation. Tacit Murky (talk) 00:00, 17 July 2017 (UTC)

@Tacit Murky: A typical Romance /t/ is laminal denti-alveolar, not apical alveolar. Please do your research. The citation you provided doesn't prove that /t/ is apical before /ʂ/, it proves that it's alveolar which we already know.

Even if he means palato-alveolar, he means a palato-alveolar fricative that is not palatalized, i.e. 'flat postalveolar' or 'laminal retroflex'. The way it's called varies from scholar to scholar. Read Russian phonology#Consonants.

I find this last statement very questionable. So you're wrong, what else can I say. You need to learn more about possible places of articulation and stop basing so much on your suppositions or your ears (you can't hear everything). You get too many things wrong and it's confusing people, especially when they're said with definitive language.

Yes, all of that is quite obvious, so I'm not surprised. Mr KEBAB (talk) 06:41, 17 July 2017 (UTC)

Oops; yes, yesterday I confused /t/ with description of /l/, which (for Romance) is apical alveolar. However, Panov citation shows, that (as he thinks) there still is a difference in articulation of /t/ and /n/ before /ʂ/ or /ʐ/ (whatever it is). So, is it indicated here?

This — It's like [ш], front-tongued front-palated. — was about /t/, not /ʂ/. I replaced parts of that proposition for translation. Originally it's like this: It's front-tongued front-palated, like [ш]. Tacit Murky (talk) 17:44, 17 July 2017 (UTC)

@Tacit Murky: Are you asking if we're transcribing those allophones here? No, and we shouldn't. Such details are appropriate for Russian phonology, not this guide.

Fair enough, but what I wrote also applies to /t/. Mr KEBAB (talk) 09:03, 19 July 2017 (UTC)

Shouldn't we use this ͡ symbol for all affricates in a table? Why it's «tɕ», not «t͡ɕ» ? Tacit Murky (talk) 00:38, 3 November 2016 (UTC)

@Tacit Murky: Laziness, perhaps. The tie is omitted in Help:IPA for English too. But it is probably more important in Russian than in English. — Eru·tuon 19:48, 11 July 2017 (UTC)

I browsed the page history and discovered that @Aeusoes1: removed the ties in this edit. Could you give your reasoning? I think the ties should be restored. — Eru·tuon 20:03, 11 July 2017 (UTC)

If I remember right, it was because The tie-bars do not display correctly in all browsers. That's what the note at Help:IPA for Polish says, which shares Russian's distinction between affricates and stop+fricative clusters. — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 00:24, 12 July 2017 (UTC)

Oh, okay. (So not laziness.) I noticed that Help:IPA for Polish has a note saying that it omits ties for the same reason. — Eru·tuon 05:01, 12 July 2017 (UTC)

@Erutuon: A similar note should be placed here. Mr KEBAB (talk) 01:15, 13 July 2017 (UTC)

Unstressed е and о

Erutron, could you provide a source for this edit? Everything I've seen points to them not appearing in unstressed position (at least in Standard Russian). — Ƶ§œš¹ _{[lɛts b̥iː pʰəˈlaɪˀt]} 14:00, 18 July 2017 (UTC)

1) The lack of reduction for unstressed /e/ and /o/ in the careful, if not somewhat affected, pronunciation, is recommended by some orthoepists (e.g. Avanesov). I believe it is not as widespread in everyday speech but may still be worth to be noted.

2.1) In word-initial and word-final positions certain vowels are affected by less reduction than expected. This is foremost true for word-initial /o/ (here we do not show the lesser degree of its reduction, but whatever) and less so for word-final /o/.

2.2) This is also somewhat true for /e/ and /a/ from etymological/orthographic «е» and «я», but this case is more obscure, and the available sources are not quite so clear on this. E.g. Avanesov in his Fonetika (1956) gives жители vs (о) жителе vs жителя, for the former he uses [и], for the second [и^э], and for the third [ъ]. Considering we do not distinguish between unstressed [и] and [ь] (but he does) and write both as [ɪ], the second sign may be interpreted as [ə] or even better as [ɘ]. Though further he says that in fast speech this distinction disappear and both merge to [и], or our [ɪ]. So I'm less sure if we need any new signs for our already over-detailed table like [ɘ]. For now I've put this case to [ə].--Lüboslóv Yęzýkin (talk) 18:47, 18 July 2017 (UTC)

My edit might have been overly bold, as I don't have access to any sources.

Regarding final -е, it was based on the automatic transcription module on Wiktionary (wikt:Module:ru-pron), which encodes a bunch of rules for the pronunciation of final -е according various categories, including part of speech. I do not have the sources used to create that module. Wiktionary doesn't have the same requirement of no OR, so that module may have been partly derived from some of the editors' own native-speaker intuitions. Atitarev was one of the creators of that module. I wonder, from what Любослов Езыкин says, if the "enunciation" of final -е in endings (pronunciation as /e/ rather than /i/) is somewhat old-fashioned, and if the phonetic value would be similar to pretonic е pronounced by speakers from St Petersburg, as described in Russian phonology: [ɘ]. (In that case, the Wiktionary module is somewhat inexact in its transcription.)

As for о in ра́дио, it's transcribed with unreduced /o/ on Wiktionary and is mentioned in the Russian article (ru:Русская фонетика#/o/ (/о/)) as an exception to reduction of historical /o/, along with some other words. (Our article, Russian phonology, mentions lack of reduction in a similar loanword, ве́то.) — Eru·tuon 22:39, 18 July 2017 (UTC)

@Erutuon: Делаете or similar with the unreduced final vowel [e] seems for me incorrect. I rather say and hear it with [ɪ]. I've no idea where they got this.

On the other hand we indeed have certain morphemes where expected /Cʲe#/ → [Cʲɪ#] is [Cʲə#] instead. This is covered in Timberlake's Grammar (p.48-51). He says that there is a contradiction in the sources: one claim that the shift/merger is complete, others, based on experimental studies, claim the opposite that the shift moves backward. My anecdotic evidence: I believe that I sometimes may pronounce it without the merger. Though, maybe it is the othography that makes me believe this. However, I've seen the claims that the word-final vowels tend to be less reduced than it is expected: Нужно при этом заметить, что заударный гласный в абсолютном конце слова не подвергается сильной количественной редукции и его длительность может быть очень значительной, хотя артикулируется он с ослабленной напряженностью и имеет малую громкость.[2]

Finally we must not confuse the above mentioned with ekanye, typical to St.-Petersburg[3], but which now seems to be largely extinct. This applies only to the pre-tonic vowels. In Russian tradition it is written as [э^и] to contrast with modern normative [и^э], e.g. веселить is [в’ьс’э^или́т’] in old ekanye, [в’ьс’и^эли́т’] in the modern normative, but I believe in the modern colloquial the latter even goes further to [в’ьс’или́т’] or [в’ьс’ьли́т’]. I know that [ь] is interpreted as [ɪ], but I cannot get how we might convert [э^и] and [и^э] into the IPA. All the slots seem to be filled. If [э^и] = [ë]/[ɘ], what then [и^э]? Here the author has simply decided to ignore the distinction. We may do and in fact have done this already.--Lüboslóv Yęzýkin (talk) 22:12, 19 July 2017 (UTC)

According to allophony table in Russian version of R.phonology article, [ə] should be used only for шея and кожа, as it denotes word-final unstressed vowels. Other cases («ы-и, э-е, о») are [ɪ] (that is [ь] in Cyrillic notation), and поле is not an exception. Tacit Murky (talk) 23:45, 19 July 2017 (UTC)

@Tacit Murky: It hardly matters, as WP cannot be a reliable source for itself. While there are enough sources which suggest the opposite: I've given one English source which in turn refers to many Russian sources where they clearly write [ъ], that is our [ə], after soft consonants. If anything your table, which, by the way, is also needed to be checked, gives but an overview of the reduction and does not cover peculiarities.--Lüboslóv Yęzýkin (talk) 15:10, 23 July 2017 (UTC)

Tables (in both Russian and English versions of the article) «does not cover peculiarities» intentionally — it's a set of typical cases. And what source say clearly that «[ъ] is our [ə]»? It can only be seen, that unstressed vowel can be [ь] and [ъ] in Russian notation, but only [ɪ] in non-Russian works. So, Cyrillic notation is somewhat closer to explain those «peculiarities». Tacit Murky (talk) 17:50, 23 July 2017 (UTC)

@Tacit Murky: I've given one of Timberlake's above, can't you see? And Panov, which you cited above yourself, dedicates a whole chapter on this matter and he clearly states that [ъ] is central which can only be [ə].--Lüboslóv Yęzýkin (talk) 19:40, 23 July 2017 (UTC)

Types or /r/

There is a difference with alveolar tap [ɾ] (more frequent in Russian, but not mentioned in the table) and alveolar trill [r], which (within morphemes) may happen only in loans with «рр» (and even there trill is not obligatory): кора́ — корро́зия (hard); гори́ — корри́да (soft). Heteromorphemic «рр» generates trill most of the time: контрразве́дка (even if morphemes are loans by themselves). Should we add or change anything? Tacit Murky (talk) 20:26, 17 July 2017 (UTC)

@Tacit Murky: I'm not sure. Are there any minimal pairs? Is the ⟨рр⟩ mandatorily trilled in words like контрразве́дка? Mr KEBAB (talk) 17:55, 18 July 2017 (UTC)

The only MP I can envision is: контроль (control) / контрроль (counter-role); while later word is synthesized, there is a rock-band called «КонтрРоль». Trill is inevitable, when there is morpheme boundary (incl. word boundary: «тигр рычит» — tiger roars), and it's possible to use one /r/-containing morpheme with another non-/r/-containing (or without another one): in above example prefix контр- (with [ɾ]) can be used with roots not starting with /r/; and разве́дка is a standalone word (also with [ɾ]). Of course, forced trills may happen in mimicking (e.g. animal roar) and emotional exclamations («Ур-р-ра!») etc. Tacit Murky (talk) 18:56, 18 July 2017 (UTC)

@Tacit Murky: I think we would transcribe these as [kɐnˈtrolʲ, ˌkontrˈrolʲ, ˈtʲiɡr rɨˈtɕit], all with [rr]. контроль - контрроль is only a near-minimal pair, the first vowel is different because the syllable containing it has secondary stress in the latter word. Mr KEBAB (talk) 19:04, 18 July 2017 (UTC)

But if you mean by [r] particularly trilling /r/, then it's not correct for контроль. Other 2 cases require at least 2-tap /r/=[ɾɾ], so it should be considered a trill, right? And what exactly is [rr] then? Analogue of [rː], i.e. very-multi-tap trill? Tacit Murky (talk) 19:19, 18 July 2017 (UTC)

@Tacit Murky: No, [r, rʲ] denote rhotics with undefined amounts of alveolar contact (one or many, i.e. taps or trills). [rr, rʲrʲ] denote trills explicitly. Mr KEBAB (talk)

Well, then we should note that tap is most common «р» in Russian. So, it's a question about level of detalization in this table. Tacit Murky (talk) 20:04, 18 July 2017 (UTC)

@Tacit Murky: We already say that in Russian phonology, I'm not sure whether we need to repeat that here. It's a phonetic detail. Mr KEBAB (talk) 20:08, 18 July 2017 (UTC)

But the table cell here say «trilled r, like in Spanish». That's not the main case! Tacit Murky (talk) 20:30, 18 July 2017 (UTC)

@Mr KEBAB: Actually Russian phonology says "Hard /r/ is postalveolar, typically a trill [r̠]", and the table row for the r's is marked trill. Only the soft /rʲ/ item mentions a tap at all. --Ørjan (talk) 07:06, 27 July 2017 (UTC)

Then, I'm afraid, Skalozub's (1963) interpretation needs to be checked by newer sources. Because now-days trill is not typical. This can be easily heard by any speech sample. Tacit Murky (talk) 14:09, 27 July 2017 (UTC)