Jump to content

Talk:Russian phonology/Archive 1

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 1Archive 2Archive 3Archive 4

"The hard х /x/ is a guttural similar to the German [ch] in ach. The soft хь /ç/ is a soft sibilant similar to the German soft [ch] in ich."

I'm pretty sure there's a significant difference between хь and /ç/. The articulation of хь is predominantly velar, and while the tongue is raised at the palate, it's not raised to the level one would expect in a palatal fricative.

As a native speaker of German who has learned Russian in school, I can confirm that the difference is audible, although pretty small (keep in mind that German /ç/ is a whole range of allophones; the one in Milch, although still palatal and not palatalized, is closer to [xʲ] than the one in ich is, the one in Mönch is intermediate, and so on).
David Marjanović david.marjanovic_at_gmx.at 00:33 CET-summertime 2005/8/6

Ч?

In the kind of Russian I learned, this was the usual Slavic [tʂ-ligature]. Do they palatalize it in St. Petersburg (resulting in Serbocroatian ć/ћ), or what? ~:-| David Marjanović 00:40 2005/8/6

I'm from Ukraine and as far as I know they always palatalize it in Ukraine while speaking Russian. It's considered a soft sound. In Ukrainian it's a hard sound, though, so I can tell the difference. -Iopq 22:21, 5 December 2005 (UTC)
It's quite palatalised. Zbihniew 14:03, 11 March 2006 (UTC)

/č'/ is palatalised (both phonetically and phonologically) as /[t'ʃ']/. Benchik 10:33, 14 June 2006 (UTC)

Conventional wisdom says that as ч is by default a palatal sound, it cannot easially be palatalised. Matusevich suggests:

...как, например, русские аффрикаты [ц] и [ч'], которые стоят особняком. Во-первых, они не входят в коррелятивный ряд глухих и звонких, так как у них нет звонких парраллелей, и вторых, [ц] не имеет мягкой парраллели, а [ч'] не имеет твредой параллели.

It's a palatalized alveolar voiceless affricate while the other one is a non-palatalized alveolar unvoiced affriacte. -Iopq 20:07, 17 July 2006 (UTC)
No, it's alveolopalatal, as the article indicates. AEuSoes1 23:08, 17 July 2006 (UTC)
Refer to IPA: ɕ Voiceless palatalized postalveolar (alveolo-palatal) fricative. -Iopq 01:21, 8 September 2006 (UTC)
Hmm I am not sure what you mean by including the IPA graph ɕ. This is not a sound in Russian at all. AND a palatalised postalveolar fricative is not the same as an alveolo-palatal fricative —Preceding unsigned comment added by 68.21.79.234 (talkcontribs) .
Well according to the kwami and literature I've read on Russian, that's simply untrue. Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 22:49, 5 December 2006 (UTC)

/ɔ/ vs /o/

Not only is the article inconsistant as to which o Russians use, but other sources of Russian vowel inventory that I've read have uniformly described it as the close-mid o not the open-mid one. Can anyone verify the open-o usage or should we change all the open o's back to regular ones?

It should be just o everywhere. -Iopq 22:33, 5 December 2005 (UTC)

soft consonants

Not all the 'soft' consonants are phonetically palatalized, though they may be phonologically palatalized. The alveolars at least are simply laminal.

Also, the transcription should be cleaned up to follow the IPA. kwami 10:22, 21 November 2005 (UTC)

You might need to point out what parts of the article aren't following the IPA. There's a lot and most of it seems okay. AEuSoes1 23:16, 21 November 2005 (UTC)
Alveolars? Which ones? Some people describe t, d, and s as denti-alveolar. -Iopq 22:33, 5 December 2005 (UTC)
My understanding is that /t/ and /d/ are dental when hard and alveolar when soft (in addition to velarized/plain and palatalized respectively). I don't know about /n/, /s/, or /z/ though. AEuSoes1 00:19, 6 December 2005 (UTC)

Ladefoged and Maddieson on Russian

Here's the main bits on Russian scattered throughout The Sounds of the Word's Languages:

  • p128 "Palatalized" /tʲ nʲ/ (their scare quotes) are laminal postalveolars. The scare quotes imply that palatalization is not the defining feature of these sounds (see below. I always thought they sounded palatalized myself).
  • p187-88 /l lʲ/ are shown as apical and laminal. Cine x-ray sketches show that apical /l/ has tongue backing and is uvularized or pharyngealized (not actually velarized, but still "dark"), while laminal /lʲ/ is palatalized. "The laminal articulation is the phonologically palatalized counterpart of the apical lateral" (implying that the difference isn't merely a matter of palatalization) and "These are normally treated as differentiated by presence or absence of palatalization, but in our view, the difference is really in the primary articulation." They contrast this with Serbo-Croatian, where for " the occlusion is palato-alveolar". In a broad phonetic transcription, Ladefoged & Maddieson transcribe /tʲ lʲ/ as [t̻ l̻] - that is, the essential difference in their view is the sounds being laminal rather than palatalized. A narrower transcription would presumably be [t̻ʲ d̻ʲ n̻ʲ l̻ʲ], but you'll have to decide how much palatal off-glide there is to these sounds.
  • p221 "In Russian, Salozub (1963) shows a post-alveolar trill as typical for r, but a dental contact for the palatalized trill ... greater lateral contact for the palatalized trill." (Tht is, more contact with the hard palate, p225.) "Skalozub reports that the post-alveolar trill had typically 3-4 contacts, whereas often has only one. The raising of the blade and front of the tongue that is required for the palatalization may make it more difficult to maintain the aerodynamic conditions for trilling." (Note that this doesn't make it a flap.)
  • p361 "The set of Russian palatalized ('soft') sounds is often said to be opposed to a set of velarized consonants, but a study of the available x-rays [most presumably of Moscow dialect?] ... suggests that the term velarized may be appropriate only for the laterals."
  • p364-65 Comment that palatalization doesn't interfer with bilabials, but causes a shift in the articulation of coronals and velars. Spectrograms of p vs. vs. pj (all contrastive). has more of an off-glide than an on-glide. "The situation is more complex with Coronal primary articulations in which specific aspects fo the tongue configuration are essential for the primary consonant articulation. In this situation palatalization consists of a displacement of the surface of the tongue front from the position that it would assume in the non-palatalized counterpart, when its role is to support the movement of the tongue tip or blade. A palatalized articulation can be viewed as the summation of two movements, with the displacement of the tongue front often producing a slightly different primary constriction location."

A couple points on the sibilants. Unfortunately, only Polish is discussed in detail, but the symbols ʒ] are not appropriate for Russian, except as a dumbed-down transcription for those only familiar with English. Those symbols are for semi-palatalized postalveolars, and I think everyone agrees the Russian sounds are not at all palatalized. They can be transcribed as retroflex ʐ], especially with laminal diacritics, [ʂ̻ ʐ̻], to show that these aren't apical as in Hindi or sub-apical as in Tamil, or as simple postalveolars, [s̠ z̠] (note that retroflex simply means non-palatalized postalveolar, so these are equivalent to ʐ] without diacritics). When possible, separate symbols are usually chosen for separate phonemes, with diacritics used for phonetic detail, so [ʂ̻ ʐ̻] might be preferred over [s̠ z̠] on those grounds, but it doesn't really matter.

The symbols ʑ] are defined as (fully) palatalized postalveolars, so there's no need for [ʃʲ ʒʲ]. In a phonological description, /ʃʲ ʒʲ/ may be preferred as it makes the palatalization feature explict, parallel to the other consonants, but in a phonetic transcription, normal IPA usage would have ʑ]. Those symbols are also more legible: In some fonts, <ʃʲ> is really difficult to read.

kwami 23:01, 16 December 2005 (UTC)

I thank you for providing notes on your source; my local university library is closed for the holidays.
It sounds to me more like palatalization in general makes coronal consonants more laminal. It's still palatalization, much in the same way that standing on one foot is still standing on one foot even though I have to wave my arms in the air to help my balance, but some people might see me as trying to fly. So I would say that [t̻ʲ d̻ʲ n̻ʲ l̻ʲ] is correct, but probably redundant.
Unless I misunderstand you, I think you might be misunderstanding what "retroflex" means and what ʒ] represent. When ʒ] represent something other than simple postalveolars it is likely due to broad transcription or some other convention that generally glosses over non-contrastive features. Retroflex doesn't mean "non-palatalized postalveolar" it means the tip of the tongue is curled back. Thus, using a retroflex symbol for a sound that is nowhere described as being retroflex and simply because it's different from English seems awkward and inappropriate.
According to the IPA page, "Curly-tail esh" was withdrawn in favor of [ʃʲ]. If [ʃʲ] and [ɕ] are the same thing, then the IPA page may need some changes.
Does SOWL elaborate on their methods of acquiring data? I'm curious as to how inclusive they were in representing more than just one dialect of Russian.

AEuSoes1 13:23, 17 December 2005 (UTC)

I seriously doubt they included more than one dialect. They're looking at precise enough detail that that would blur their results. In the preliminary research I've seen on endangered languages, they've separated men from women in a single dialect. Also, their approach is looking at many languages, rather than any one in depth, although of course much of the book is based the results of decades of work on those individual languages. Peter Ladefoged recently came out with a book on field recording for phonetic analysis. You should be able to find it on his web site.
Their point with the precise detail of the coronals was to distinguish this from other languages, such as Serbo-Croatian. The way I read it, it was L&M's opinion after looking at many languages that, phonetically, the important feature in SC was palatalization, but in Russian it was lamination. But you're right, lamination is a common feature of palatalized sounds.
ʒ] represent palato-alveolars, a kind of postalveolar. L&M called them "domed" postalveolars:
"By domed we mean to denote the raising of the front of the tongue that occurs, irrespective of whether an apical or laminal articulation is used. This doming is equivalent to a small amount of palatalization. We will regard the phrase palato-alveolar sibilant as an exactly equivalent specification, denoting a comparatively wide constriction in the post-alveolar region near the center of the alveolar protruberance, with concomitant raising of the front of the tongue. We will distinguish between palato-alveolar and alveolo-palatal sibilants, using the latter term as an alternative specification for the post-alveolar palatalized sibilants that we will describe in Standard Chinese." (That is, [ʃʲ] is [ɕ].)
(Remember, Ladefoged is on the board of the IPA and the editor of the IPA journal, while Maddieson is vice-president of the IPA.)
English and French are often given as prototypical examples of palato-alveolar ʒ]. Chinese, Polish, and Russian, however, don't have the "domed" aspect to their post-alveolars. That is, they're pure singly articulated post-alveolars (unless they have something else going on, of course).
True, retroflex consonants prototypically have the tongue curled back, as in Tamil. That's the European conception of them. However, in Hindi the tongue isn't curled back; rather, it's apical postalveolar. These are both traditionally called "retroflex" regardless. In Mandarin, the "retroflex" consonants are "flat" post-alveolars, without the doming of English. In other words, all the non-palatalized post-alveolars are often called "retroflex". L&M like to distinguish these possibilities, and restrict the symbols ʐ] to the sub-apicals, as in Tamil. For the apicals, as in Hindi, they've resurrected the under-dot diacritic, placed on the alveolars, which after all came from the Indologist tradition. For flat postalveolars, they just use the underbar diacritic for retraction. Of course, that's ambiguous on its own, indicating only postalveolar, but in context it means "other" postalveolar. (Of course, this is their personal notation. Actually, they've reversed the two ad hoc notations between their treatment of stops and fricatives, which is annoying.)
In publications that adhere to the inventory of symbols provided by the IPA, all "retroflex" articulations are typically denoted by ʐ], whether they're Tamil, Hindi, Chinese, or Russian, just as all "alveolar" articulations are typically written [s z], whether they're apical alveolar as in English or laminal (denti-)alveolar as in French.
There is some rather lengthy discussion of all this, so I'll just call attention to tables 5.3 and 5.4. For both Mandarin and Polish, L&M list the columns as "Flat post-alveolar (retroflex)" and "Palatalized post-alveolar (alveolo-palatal)" In the Latin alphabet, these are sha, xia in Mandarin, and kasza, Basia in Polish. The palatalized ones, of course, have their dedicated symbols, ʑ]. The "flat" retroflex ones could be indicated with a laminal diacritic, if you wish to be precise. (The reason L&M call them "flat" rather than laminal is because ʒ] can also be laminal; by the ad hoc description "flat" they specifically mean that the articulation is not domed as in English and French.) —kwami 20:37, 17 December 2005 (UTC)
I find it hard to accept that a series of consonants called "retroflex" don't always have retroflexing; That's not just the prototype, that's the defining feature! Wouldn't it be more the case in instances where "retroflex" sounds aren't are either mislabled or that the language has changed since the original analysis? As for the "postalveolars;" I think it would make more sense that they would have a symbol that meant just postalveolar and if a language has special features (like palatalization etc) that there would be diacritics to add to that in narrow transcriptions or when it becomes important because of contrasts.
In looking things up, I found this paper [1] that states that retroflexing often also retracts (displaces the tongue dorsum or root towards the pharynx), “traditional” retroflexes can turn out to be contrastive more on their retraction than retroflexing, and because of retraction retroflex consonants are never palatalized unless they don’t have the feature retraction (which is rare). This may explain, as the phonology article details, why ш is always hard but not why ж has a palatal. AEuSoes1 09:41, 18 December 2005 (UTC)
The term 'retroflex' does not predicate retroflection because most of the languages identified as having retroflexes, such as the Indic family, do not have retroflection. Words evolve as people use them. I think L&M would agree with your impression of the IPA, which is why they've extended it to their liking in their book.
The IPA would have turned out differently if it had been designed from scratch with prior knowledge of the world's phonemic inventories. But it was based on Western European languages and developed as a hodgepodge from there.
As for the distinction between ш and ж, that might simply be a historical accident. ш does have a palatalized version, but only as an affricate or geminate. kwami 11:06, 18 December 2005 (UTC)
So do L & M say anything about Й being /i̯/ instead of /j/? AEuSoes1 01:16, 19 December 2005 (UTC)
No, but [j] is a consonant, and if I'm not mistaken Russian Й is purely vocalic, an element of diphthongs. Sometimes [kai̯] may be written [kaj] for simplicity in a broad transcription, but in more precise notation [kaj] is a closed CVC syllable, whereas [kai̯] is an open CV syllable. kwami 02:14, 19 December 2005 (UTC)

All right, I got a hold of SOWL and I’ve looked myself at what you’ve pointed out.

According to the authors, on page 148, /ʃ/ is indeed present in Russian, as they note that lip rounding "is a feature of 'ʃ' in some languages, such as English and French, but it is not found in many other languages, such as Russian."

There is nothing to indicate in the book that retroflex consonants do not have retroflexing. There is even an image on page 22 showing exactly where the tongue touches the roof of the mouth. Granted, in languages like Hindi, there may be less retroflexing but it is still there. They even say on p153 that Mandarin sh is inappropriately described as retroflex when it is clearly a laminal flat post-alveolar sibilant (although it is strange that they continue, perhaps for continuity's sake, to represent it by an underdot s). AEuSoes1 20:49, 22 December 2005 (UTC)

Common usage is to call Hindi and Mandarin 'retroflex'. The retroflex articles discuss this ambiguity. As for Russian, these talk pages have gone over this: Russian ш is like Polish sz, not like English sh. Unless the several people who contributed to that discussion are wrong, the symbol [ʃ] is inappropriate for Russian. (L&M mentioned Russian ш only in passing, and addressed labialization, not palatalization. It's somewhat questionable to draw conclusions from so little.) Whether we use [ʂ] or not depends on how narrow a definition of that symbol you wish to use. I'll change the Russian article to [s̠] as a compromise. kwami 22:54, 26 December 2005 (UTC)
I'm sorry I made my last revert before replying here; I tend to read top-down in my watchlist. According to L&M, retroflex is always retroflexed unless it's inappropriately labled as such. If you look at the examples of on page 67 and 91, you'll see that Affricates such as d̠ʒ are labled with the stop part underscored. This tells me that there is little to no difference between and ʃ; I could be wrong about that, but I don't know why else the affricates would be labled as such and I would favor {ʃ over for the same reason that I'd favor [t] over []. In addition, if L&M thought that Russian ш was not ʃ, they would not have made the statement on p 148, they had many other languages to choose from that had ʃ and could have picked them. Also considering the excruciating detail that they laid out, I doubt that they would gloss over something as significant as retroflexing or doming, even in passing.
On a side note, I noticed that the 'hard sign' is marked as indicating velarization but that would favor the notion that all non-palatalized obstruents are velarized. I'm willing to go with the only-lateral explanation in our representation of phonemes so we can probably change that. AEuSoes1 01:21, 27 December 2005 (UTC)
I don't know what I'm talking about with Russian, so I'll call in someone who does. As for L&M being sloppy, well, they are occasionally sloppy. For instance, with stops an underdot is apical and a retraction sign laminal, but for fricatives it's the reverse. The difference between and ʃ is quite clearly spelled out: is (presumably plain) postalveolar, and ʃ is domed/semi-palatalized postalveolar. It's the difference between Polish sz and English sh (both of which are semi-labialized, BTW). AFAIK, the Russian postalveolars, like the Polish, are not palatalized.
The hard sign note was a compromise for the purpose of simplification. You might be able to come up with something better, but perhaps an explanatory note would be best. kwami 02:28, 27 December 2005 (UTC)
Bolla 1981 is a standard reference used by L&M and many others. It states that the Russian frics are retroflex (in the broad sense). See the PDF here for work based on Bolla. 'Polish and Russian have retroflex fricatives (i.e., /ʂ/ and /ʐ/), whereas Bulgarian has a laminal palatoalveolar fricative (/ʃ/).' (Note that this paper tries to distinguish [ʃʲ] from [ɕ], which is not supported elsewhere. Since Russian contrasts [pʲ] and [pj], there could easily be something similar going on with the frics.) Note that in another publication (here), Hamann states that 'A retroflex fricative with a curling backwards of the tongue tip, comparable to the Tamil stop in figure 1b, does not seem to occur in any language.'
On page 22 they contrast dental, alveolar and retroflex with alveolar having the underscore, so apparently they aren’t using it for retraction. How confusing.
The first paper you cite claims that the Russian sound is the same as that in Mandarin, thus it is a laminal flat post-alveolar, not a retroflex. Notice that they call it retroflex because the Mandarin sound is retroflex and then use that to redefine retroflex: “Hence we do not employ the bending backwards of the tongue tip or the apicality of a sound as the criterion for retroflexion… but post that velarization and flat tongue shape are the crucial parameters.” It struck me as interesting that they consider the view of Russian having /ʃ/ as the “prevailing view”; I’m assuming that this is prevailing even among linguists.
As you recall, L&M contrasted "sub-apical palatal" with "apical post-alveolar... with the part of the tongue behind the tip hollowed to some extent" we can call these tamil-retroflex and hindi-retroflex. Your second link didn’t work, but from the quote it sounds like Hamann is saying that there aren’t any tamil-retroflex fricatives but it doesn’t sound like he’s discussing the hindi-retroflexes. AEuSoes1 08:00, 27 December 2005 (UTC)


Prehistory

The modern phonological system of Russian is inherited from Common Slavonic, but underwent considerable modification in the early historical period, before being largely settled by about 1400.

How on Earth one can be so sure without any tape recorders dating by 1400? Whatever it was, this is hardly a reasonable intro, which is supposed to be a summary of the artricle.

Also, isn't it upside down: "Common Slavonic" is a hypothetical reconstruction derived from modern languages rather than vice versa? mikka (t) 02:53, 21 December 2005 (UTC)

Also such a big article is ridiculously void of references. mikka (t) 02:53, 21 December 2005 (UTC)

I agree that there should be more references. I don't doubt that the information is accurate, it's just nice for verification and further research purposes. If you have a problem with the introductoriness of the introduction, be bold.
As for Common Slavonic, even without the use of writing systems, the methods used to reconstruct language histories has proven to be very accurate, especially with phonology. Saying that x inherits a phonological system from y is not "upside down," it is a description indicating the chronological order.AEuSoes1 08:04, 21 December 2005 (UTC)
Academics can tell about the status of the historical phonology of a language from texts, e.g. by "mistakes" made in the orthography and the creation of new symbols for poorly represented sounds. Benchik 10:31, 14 June 2006 (UTC)
The occurence of such mistakes and new graphemes are so rare that you certainly couldn't determine a language's complete phonology from them. The comparative method is the main way linguists reconstruct past forms and proto-languages. Writing systems simply help verify the accuracy of the comparative method. AEuSoes1 05:21, 15 June 2006 (UTC)

ŋ?

I'm wondering if the consonant ŋ is pronounced in the combination of the letters нг, or if the two sounds are pronounced separetly.

Never mind - I have already gotten my answer. This discussion page is a barren wasteland.

BirdValiant 04:35, 3 February 2006 (UTC)

AFAIK they should be pronounced seperately. Or so my Russian phonetics handbook claims. Zbihniew 14:01, 11 March 2006 (UTC)

Retroflex sounds in Russian?

As a student of both Phonetics and Slavonic linguistics, I would like to point out that the use of the retroflex symbol for what is normally represented in the orthography as 'ш' and 'ж' is not a convention used in advanced Slavonic linguistics. Retroflex is defined as follows:

"Retroflex: articulated with the tip of the tongue or the underside of the tip against the back of the alveolar ridge" (from Peter Matthews' Oxford Concise Dictionary of Linguistics, 1997:320)

That means that a retroflex sound, at the very least, has the tongue tip on the last bit of the alveolar ridge, and most of the time involve a noticeable curving-back of the tongue tip (as with many Indian dialects). The sounds in question in Russian simply cannot be termed 'retroflex' because the tongue is not in the position, as is certainly not curved back. In fact, my Russian-speaking friends tell me that the tongue tip does not come into contact with the alveolar ridge, especially in fast speech, and curving-back of the tongue (the MAIN feature of retroflexing, hence the name) is completely out of the question.

The /ʃ/ and /ʒ/ symbols are admittedly not perfect - and are used to represent slightly varying sounds from language to language, but are a far better compromise than the retroflex ones. For Slavonic linguistics, using /š/ and /ž/ is the norm - and if you're well read in the area, you'll know this. The IPA is not ideal, and the late Ladefoged would have argued that.

I've actually seen x-ray tracings of the manner and place of articulation for 'ш' and 'ж'. They are curled back enough to be considered retroflex. Check out The Phonetics of Russian by Jones and Ward (for the tracing, they use <ʃ ʒ> and simply call them "harder" than English postalveolars). There was an article in the Journal of the IPA a few years back that very explicitely demonstrated that they are retroflex in Russian and similar languages while in other Slavic languages they are simple postalveolar. AEuSoes1 00:43, 26 May 2006 (UTC)
In fact, some people do actually curl back the tip of the tongue while pronouncing ш and ж (mostly women), but such articulation is traditionally considered to be a speech defect in Russian. Normally, these sounds seem to be sub-apical. —Preceding unsigned comment added by 62.141.84.150 (talk) 09:14, 29 January 2008 (UTC)

a or ɐ? + connection with AKANIE

This page used to claim, and I think this was correct, that Russian /a/ is usually pronounced as ɐ. Perhaps there is some regional and personal variation in this, but certainly Russian /a/ is further back and more closed than most /a/ phonemes. What does the literature say about this? --platypeanArchcow 08:00, 13 June 2006 (UTC)

That in a stressed position that it's [a] (front and low and whatnot), there's varying allophony depending on surrounding consonants but the closest allophone to [ɐ] is in the pretonic syllable where it's described as an advanced and lowered [ʌ]. AEuSoes1 05:08, 14 June 2006 (UTC)
"an advanced and lowered [ʌ]" is essentially [ɐ], not [ʌ]. I think it's misleading to describe the unstressed a/o with the [ʌ] symbol which most people would associate with English "sun, love" etc. By the term "akanie" we can see that the linguistically unsophisticated Russian view is that the unstressed vowel is a kind of [ a ]. Cf. "Phonetics and Phonology in Russian Unstressed Vowel Reduction: A Study in Hyperarticulation" by Jonathan Barnes (Boston University)p.4 (http://www.bu.edu/linguistics/UG/barnes/Barnes%20VR%20final%20draft.pdf) : "the vowel usually rendered as [a] in Western sources is in fact said to be realized somewhat higher, closer to IPA [ɐ]. (The traditional transcription in Russian sources is [ʌ], which can be somewhat misleading, given this symbol’s IPA value.)"Jakob37 16:58, 29 August 2007 (UTC)
The main resource I've used for putting <ʌ> is pretty old and I'm not sure about the history of the <ɐ> symbol. I'll look at the pdf you mentioned, but something to consider is that <ʌ> is arguably an "a" sound; also, for numerous dialects (including RP, the vowel in cup is technically [ɐ] but the symbol for the back vowel is used anyway (for tradition, I suppose). If we decide that <ɐ> is the better symbol then there's quite a few Russian phonetic transcriptions that we'll have change. Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 20:31, 29 August 2007 (UTC)
Why does it look like the pdf is of an unpublished paper? Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 18:05, 30 August 2007 (UTC)
It's quite true what you say about the "new" southern British pronunciation of the vowel in "cup" -- so we could just broaden our concept of <ʌ> to include <ɐ>; my worry is just that many readers (keeping in mind, for example, that there are many more speakers of American English than of British) will see the the <ʌ> sign, check it on some chart, see words like "cup", "love" etc., and come up with a sound which is not quite appropriate for Muscovite Russian.Jakob37 08:43, 2 September 2007 (UTC)
I can't think of any dialect of English (although I'm sure there are some) that has cardinal [ʌ]. Even in American English it's a bit advanced. I recall talking to a phonologist about the difference between [ʌ] and [ɐ] and I think they're two sounds that are acoustically similar even if they're articulatorily different. The phonetic transcriptions of Russian used around Wikipedia aren't perfectly accurate anyway; unstressed /u/ and [ɨ] aren't transcribed as near-close, the vowel /e/ has some varying height allophony that isn't indicated, and /r/ as a postalveolar sound is not indicated. Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 19:14, 2 September 2007 (UTC)

Semi-vowel

There is no semi-vowel in Russian, it is consonant. It was considered semi-vowel until 1950s.--Nixer 06:50, 20 June 2006 (UTC)

The 1969 book The Phonetics of Russian describes Russian /j/ as having two principle allophones. In diphthongs (where it is represented by й) it is [ĭ] while "consonantal" /j/ occurs before a vowel. Thus Russian does have a semi-vowel represented by the letter й. AEuSoes1 07:43, 20 June 2006 (UTC)
This is rather old book. If the book is not Russian, then it may be based on even older Russian sources. In modern books it is clear stated й is consonsnt. For example my parents learned й was semi-vowel, but now any book states it is consonant. --Nixer 10:44, 20 June 2006 (UTC)
Whatever it is described as in Russian sources, according to the terminology used by linguists [ĭ] is a semivowel when it is part of a dipthong. Popular and technical usage often differ - see the hard/soft popular description of English consonants vs the technical description involving voicing and aspiration. Ergative rlt 16:34, 20 June 2006 (UTC)
Or for a closer example, consider the sounds denoted by English y, and their role in the syllable. Ergative rlt 16:41, 20 June 2006 (UTC)
The й letter never creates syllable, so it is never a vowel.--Nixer 21:23, 20 June 2006 (UTC)
Right, it's a semivowel. It may be an old source, but it partly backs up your claim that Russian /j/ is considered a consonant. However, when it's represented by <й> then it is almost always a semivowel and part of a diphthong (only foreign borrowings that put it word-initially are exceptions). AEuSoes1 02:45, 21 June 2006 (UTC)
Now й considered consonant. And you cannot cite any contemporary Russian source which states otherwise. It is pure clear consonant. And it never creates syllable. In foreighn borrowings it pronouced exactly the same way as in Russian native words. Look here for example:[2]. Й is soft, voiced, non-paired consonant. There are 36 consonant phonemes and 6 vowel phonemes in Russian. --Nixer 06:54, 21 June 2006 (UTC)
I'm sorry, but you have the burden of proof. I've provided a source and it didn't merely indicate a theoretical difference but an actual phonetic non-consonantal allophone of /j/. The website you provided is confusing because it uses a cyrillic-based notation system and uses Й word-initially, which Russian doesn't do (I also don't speak Russian and doubt my translator is very good). You need to provide a source that clearly states that Russian had a semivowel in the past but the language (not the scholarly discussion on the language) has changed so that/j/ a consonant in all environments. I don't think you'll find one, but good luck. Until then, do not alter the article's reference to semi-vowels. AEuSoes1 14:49, 21 June 2006 (UTC)
The words starting with й is the transcription. The source clearly states that Russian has 36 consonants, 6 vowels and no semi-vowels. Please do not insert wrong info in the article. I am native Russian speaker and studied Russian at school for 10 years.--Nixer 08:57, 22 June 2006 (UTC)
I understand the transcription, but it's not a good source. It's great that you're a native speaker. Perhaps in your ten years of study you would have encountered some sources that explicitely back up your claim that the language has changed or even ones that explain that the earlier descriptions were incorrect. AEuSoes1 19:20, 22 June 2006 (UTC)
It is a very good source, be sure. And the site is specially itentend to help with questions with Russian grammar.--Nixer 14:55, 25 June 2006 (UTC)
This is not a question of Russian grammar. It's a question of Russian phonetics. Considering your confusion with IPA, I'd venture to guess that you aren't an expert in phonetics. Please provide a better source. Multiple ones, even. AEuSoes1 19:18, 25 June 2006 (UTC)
This is a good academic source. You can open any Russian schoolbook and see the same. And I know Russian phonetics better then you anyway.--Nixer 07:16, 26 June 2006 (UTC)

This argument seems silly to me. As far as I was aware, semivowels are consonants -- just approximants that happen to be articulated similarly to a vowel. Unless Russian й is actually a fricative [ʝ], I don't think it's incorrect to call it a semivowel. --Ptcamn 19:48, 22 June 2006 (UTC)

It is not approximant.--Nixer 14:55, 25 June 2006 (UTC)
I don't know if this debate can really go on. Not only do you have difficulty with finding sources, but you claim knowledge of phonetics on one hand and then say that /j/ is not an approximant? It would actually have to be an approximant to be a consonant.
In addition, I've asked you to not change the article until we build consensus and instead you prefer to blindly revert the article after every comment you make in the talk page. That is not consensus, you have not provided sources ("any Russian schoolbook" doesn't cut it, I need examples), and you have not demonstrated your knowledge of phonology so you can't hide behind your claims of being some sort of expert. AEuSoes1 17:06, 26 June 2006 (UTC)
Look here: [3]. The question to expert No.881. "Q: It said that й is a consonant. Is it true? A: Й - is soft voiced non-paired consonant. There are six vowel sounds in Russian: а, о, у, э, и, ы. Vowel letters are 10: а, о, у, э, и, ы, е, ё, ю, я. The last four are not sounds, but letters."--Nixer 05:45, 27 June 2006 (UTC)
Um... how about a book? AEuSoes1 05:51, 27 June 2006 (UTC)

[nʲ] or [ɲ]?

Are these equals and can they be both written for Russian, or is one more correct? BirdValiant 23:50, 22 June 2006 (UTC)

They're not the same. From my understanding, [nʲ] is like an alveolar /n/ with a coarticulated /j/. AEuSoes1 02:45, 23 June 2006 (UTC)

So basically [nʲ] is more "forward" in the mouth than [ɲ]? BirdValiant 06:47, 23 June 2006 (UTC)

Yes, but so is [n]. The main difference betwen [n] and [nʲ] would be a difference in the shape of the tongue. I could be wrong, though. AEuSoes1 15:47, 23 June 2006 (UTC)

Wait... you know I'm talking about ɲ, not n, right? BirdValiant 18:04, 23 June 2006 (UTC)

Yeah. It's just easier for me to compare it to [n].
What in IPA written as [nʲ] does not contain j sound at all. It is just soft n.--Nixer 15:06, 25 June 2006 (UTC)
You seem to misunderstand IPA: <ʲ> marks palatalization ("softening") of a consonant. AEuSoes1 19:16, 25 June 2006 (UTC)
Well, it is written in the Palatal nasal article that [ ɲ ] is used in Russian, so maybe somebody who truly knows that there is no real difference between [nʲ] and [ ɲ ]... or it could be somebody who doesn't know what they're talking about. But just a few sentences up in Spanish, it says:
"However, in Mexico <ñ> is a palatalized alveolar nasal, [nʲ]. The difference is this: a true palatal is pronounced with contact between the middle of the tongue and the palate. The front of the tongue is not involved. In a palatalized alveolar (or dental), it is the front of the tongue that makes the contact, as in [n], but the middle of the tongue is simultaneously raised toward the palate, as in [j]. That is, [nʲ] is pronounced like a simultaneous [n] and [j], while [ɲ] or perhaps [ɲj] is palatal, though it may have a [j]-like offglide."
So end of case I guess. They're close enough (I probably couldn't tell the difference), but if you want to be precise, as with IPA, you probably should use the correct one then. So you win Aeusoes1 (for now...)
Woah! When did they update the symbol selection down there in the edit page?! I can actually see the IPA symbols!! (Until they get on this editing panel... but oh well!) BirdValiant 01:11, 26 June 2006 (UTC)

[ɑ]??

"The default pronunciation of а is /a/; я differs from this merely by indicating the palatalization of the preceding consonant (or iotation at the beginning of the word). In both cases, /a/ is fronted to [æ] between palatalized consonants (see consonants below). So мать is realised as [matʲ], whereas пять is realised as [pʲætʲ]. When not following a palatalized consonant, /a/ is retracted to [ɑ̟] as in палка ['pɑlkə]."

Does this not contradict itself? m is not palatized, so мать should be [mɑtʲ]?

Wouldn't [ɑ] be the default pronunciation then? BirdValiant 01:20, 26 June 2006 (UTC)

Good catch. I've fixed the discrepancy. AEuSoes1 05:59, 26 June 2006 (UTC)

Ы

Is [ɨ] a phoneme or not? The article lists it in phonemic slashes with the other vowels, but then subsequently says it's an allophone of /i/. I've heard that there is disagreement among linguists over this issue; if that's the case, it'd probably be best to have a subsection listing arguments for an against each analysis. --Ptcamn 07:54, 5 July 2006 (UTC)

Or simply to make note that there is a controversy. Michael Z. 2006-07-05 14:31 Z
Yeah, we've discussed it off and on in talk pages. If it an allophone, it's one of those rarer instances where you have a phonemic orthography with one letter for an allophone. AEuSoes1 20:01, 5 July 2006 (UTC)
Are Ш and Щ allophones? Since Ш is the hard version of Щ, and they have separate letters for each other, is it the same situation for Ы and И, in which Ы is the "hard" version of И? (I have no knowledge on this... I'm just making a wild guess in the dark...) BirdValiant 02:03, 6 July 2006 (UTC)
No. The soft consonants aren't allophones of hard consonants, they're separate phonemes. There are some cases of assimilation and whatnot. In addition, Щ is geminated and has an alternate pronunciation with /t/ sandwiched in. AEuSoes1 04:21, 7 July 2006 (UTC)
Ok... whatever. Does it really matter if those two sounds and symbols are allophones anyway? Who cares? BirdValiant 19:36, 9 July 2006 (UTC)
There have been wikipedia flame wars over less. To those it matters to, here are some sources that argue for the one-phoneme analysis.
  • Trubetzkoy, Nikolai. 1969. Principles of Phonology. Berkeley: University of California Press.
  • Avanesov, Ruben Ivanovich, and Vladimir Nikolaevich Sidorov. 1945. Ocherk Grammatiki Russkogo Literaturnaia Iazyka, Chast' 1. Fonetika i Morfologia. Moscow: Gosudarstvennoe uchebno-pedagogicheskaia Izdatel'stvo.
  • Halle, Morris. 1959. The Sound Pattern of Russian. Mouton, The Hague.
  • Hamilton, William S. 1980. Introduction to Russian Phonology and Word Structure. Columbus, OH: Slavica Publishers
  • Sussex, Roland. 1992. Russian. In International Encyclopedia of linguistics, edited by W. Bright. New York: Oxford University Press, pp 350-358.

I haven't read any of them so I don't know how good they are. AEuSoes1 03:44, 10 July 2006 (UTC)

Why don't we just ask somebody who speaks native Russian? BirdValiant 22:43, 10 July 2006 (UTC)
Because native speakers can be wrong. That would also constitute as original research. AEuSoes1 22:48, 10 July 2006 (UTC)
It'd be better than nothing. Just ask somebody "Hey, do Ы and И sound the same to you? (That is what we're talking about, right?)" "No." "Alright, thanks." "Sure, no problem." BirdValiant 04:12, 12 July 2006 (UTC)
"Sound the same" is not equal to "are allophones". Sometimes people are aware of allophones. Probably more significantly, people's thinking about language is often quite colored by the orthography. I've heard stories of Spanish speakers who insist that there's a difference between v and b, for example. --Ptcamn 04:19, 12 July 2006 (UTC)
As a native speaker of the language I can testify that Ы and И do not sound the same, but that has little to do with whether they are different phonemes or not, I think. Jslllnghmck 13:57, 16 July 2006 (UTC)
I believe there have been Wikipedia discussions where two native speakers actually disagreed on the issue. See Talk:Russian language#number of vowels. AEuSoes1 08:21, 12 July 2006 (UTC)
Unless I'm imagining it, my Spanish teacher pronounces a slight difference between b and v, but maybe that's because he doesn't live in a native Spanish-speaking area, and has an influence from English. Or it could be all in my mind. In any case, the difference between his b's and v's are much less distinct than the differences between ы and и. But that doesn't really matter since we're not talking about Spanish here. BirdValiant 19:22, 12 July 2006 (UTC)
I bet your Spanish teacher also speaks English. Language contact can do those sorts of things. AEuSoes1 20:28, 12 July 2006 (UTC)
This is why a lot of linguists reject the idea of a phoneme... In Japanese, /e/ and /i/ have complimentary distribution, /i/ palatalizes. It is the same exact case, but somehow most people don't say /i/ is just a pronounciation of /e/ following a palatal phoneme. -Iopq 14:43, 16 July 2006 (UTC)
If Ы is not a phoneme, and if Russian has a whole large extra set of consonantal phonemes for expressing palatisation, that's news to the hundreds of millions of average Russians who have had to "put up with" a much simpler phonemic analysis, i.e. the Russian alphabet, for so many centuries. The standard arguments for putting i and Ы into one phoneme are largely based on morpheme-boundary or even word-boundary phenomena --- not the ideal place to look for basic minimal pairs. Otherwise it's based on the existence of a few foreign loans which have not yet accomodated themselves to the normal Russian pattern. Such loan words are marked forms within the traditional Russian phonological system, and the alphabet recognises this by using the marked ( = less commonly used) ь (softening sign) as in полька. As Y.R. Chao pointed out long ago in "The Non-Uniqueness of Phonemic Solutions of Phonetic Systems", the way that a phonemic system is designed depends on one's purpose or goal; the traditional Russian orthographic analysis embodies convenience and accuracy (well, it would be nice for foreign learners if the stress were marked...). Every now and then one hears calls for an English spelling reform; I have never heard any calls for replacing the Russian orthography with a newer, more "accurate" system containing a whole extra set of palatalised consonants. There seems to be a lack of imagination among the many modern phonologists who can only conceive of a diphthong as falling, not as rising. Treating the material between the two consonants in люб (pleasant) as a single vowel is analogous to the traditional(but now outmoded) analysis of English cute as having just one vowel, the "long-U" in between the two consonants.Jakob37 06:54, 16 April 2007 (UTC)
It may be news to speakers, but it's not news to linguists. The problems cited with English orthography are the predictability of spelling and of pronunciation, not on the phoneme to grapheme correspondance. Russian gets the job done of indicating whether a consonant is hard or soft very effectively so no spelling reform is needed in this regard. As for comparing [lʲup] with [kjut], I'm not sure what you are saying about the latter. My understanding is that (in English) whenever [j] is in the onset of a syllable, it is a consonant, making cute a word that starts with a consonant cluster. One test for this is to ask whether yule rhymes with fool (it does).
Why are morpheme-boundary and word-boundary phenomena not good places to look for determining the phonemic status of a sound? In English, morphemic analysis can show that [ɾ] represents /t/ (writer=write + er), as well as /d/ (rider=ride + er) and word-boundary examples reinforce this (it is [ɪɾɪz]). Turning to Russian, the adjectival morpheme |ij| is realized as [ɪj] when added to a root that ends in a soft consonant (широкий) and [ɨ̞j] when added to a root that ends in a hard one (действенный). Now, this in itself could be seen as allomorphy if you'd like, but then you'd have to admit that at least at some point the two sounds were allophones of one phoneme. Then, when you add the fact that the the first name of Владимир Ильич Ульянов ends with a soft rhotic when it is followed by the second name but a hard one when it is spoken in isolation or with a significant pause, you've really got to scratch your head and wonder why that is if, gosh darn it, [ɨ] doesn't strictly follow hard consonants.
Iopq is right in noting that there are examples where thinking in terms of an underlying phoneme can be problematic, but the Russian и/ы distinction is not a troublesome enough example to throw out the phoneme just yet. Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 03:51, 17 April 2007 (UTC)
In pre-structural, pre-generative analysis, i.e. the kind you will find in older dictionaries, a word like "cute" is/was analysed as having 3 phonemes. The popular trend now is to take the palatal element before the u as a consonant --- but then one should ask oneself, why are there a huge number of j + u combinations, and hardly any j + o, j + a, j + e etc. We have coot, cot, caught, Kate, cat, coat, cut, but there is no **/kjat/, no **/kjet/, no */kjot/ etc. Only the so-called /kju:t/ (which I write as three phonemes: /kyt/). For such structural reasons a few linguists maintain that this /y/ is a phoneme; of course we would also phonemicise "time" and "house" as having 3 phonemes too. - The flapping examples you gave can also be found within single morphemes, e.g. "butter, little, Frodo". If we only looked at your examples, we might think that the flapping was not due to purely phonological reasons, but was limited to certain morpheme-boundary situations. Some phonologists(including myself) regard "ŋ" as a non-phoneme in English since it's seemingly contrastive occurrence is limited to morpheme-boundaries sich as "singer" (contrast "finger"). Phenomena which mainly occur at such boundaries may be predictable by rule, thus non-phonemic. - The vowels in "cot" and "caught" are different in my English, but there are many Americans and Canadians who say them the same. Also, few speakers contrast them before an "r". Gosh darn it, are they really just one phoneme? Well, maybe some day, but for now, all things considered, it's simpler to keep them distinct. Because non-recognition of ы is one of the good excuses for saying that Russian actually has a whole extra set of palatalised consonant phonemes, it also keeps matters a lot simpler by not throwing out ы , at least not yet. Jakob37 05:21, 17 April 2007 (UTC)
The reasons for a paucity in words like that are largely historical (yes, /ju/ comes from /eu/, which was a diphthong) and phonotactical. With that logic, /kj/ is a phoneme itself because /tj/ and /pj/ don't occur (although I recognize that there are other things in the way of this analysis; I'm just indicating that for your argument to be valid, other arguments would have to be made).
Yes, flapping occurs in single morphemes, but the morpheme and word boundaries are how we know that the flap is not simply a third phoneme and, moreover, what phoneme(s) to attach it to.
I'm not sure what your point about the velar nasal or about the cot-caught merger has to do with what we're talking about. Sure, you can analize the velar nasal like that and if one wants to create an analysis of English that encompasses different dialects, some of which have merged two vowels, well so be it.
I've cited several works above that argue that the two Russian letters denote one phoneme and I'm fairly certain that the fact that Russian contrasts /pʲj/ and /pʲ/ (and /pj/ in careful speech so I've heard) shows that even the separation of the two sounds as distinct phonemes would not make the analysis of consonantal inventory collapse. Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 05:57, 17 April 2007 (UTC)
1)**/kj/ should not be a phoneme because its environment is limited to _u: (going by your analysis). 2) There are many other ways of determining that the flap is not a phoneme, e.g. its non-occurrence initially or finally, or even internally except intervocalically, and its alternation with /t/ and with /d/ in monitored speech. No appeals to monitored speech, because it may not be natural? Then you can't cite "careful speech" for Russian either.... 3) the point with the velar nasal was that occurrence in certain morpheme-boundary environments does not necessarily stand as evidence of a sound being a phoneme -- single word minimal pairs are the standard expected. 4) diminishing o/au contrast mentioned: just an echo of your previous statement "the Russian и/ы distinction is not a troublesome enough example to throw out the phoneme just yet". Jakob37 11:05, 17 April 2007 (UTC)
After thinking about it, I realized that I've made a mistake. The example with Vladimir Lenin should have read as this: In the name Владимир Ильич Ульянов, Ильич in isolation is pronounced [ɪˈlʲitɕ]. However, the preceding unpalatalized /r/ of Владимир makes the first two names pronounced together as [vlʌˈdʲimʲɪr ɨ̞ˈlʲitɕ]. The vowel frontness is deteremined by the phonological environment of a preceding consonant.
Anyway, the strongest reinforcement of the analysis of the one phoneme is word pairs like бить (bang) быть (be). Phonetically, these are [bʲitʲ] and [bɨtʲ] respectively and under this one phoneme system the phonemic transcription would be /bʲitʲ/ and /bitʲ/ respectively, making it a minimal pair for the palatalized and unpalatalized consonant. Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 02:22, 18 April 2007 (UTC)
Of course your /bʲitʲ/ and /bitʲ/ are the "standard analysis" among linguists, but /bitj/ and /bïtj/ is not only simpler (one symbol fewer for the first word) but also corresponds directly to the practical and widely used native script. By the way, not trying to be too much of a nag, but your statement "the strongest reinforcement of the analysis of the one phoneme is..." is an example of circular reasoning wherein you are using the contents of an argument as part of its proof -- well, it would be like saying, "Santa Claus exists because we know he comes to visit every Christmas" In your case, you are saying "Proof of this analysis is this analysis". Jakob37 11:10, 18 April 2007 (UTC)
The word pairs are the reinforcement, the analysis expresses this. I believe that you're advocating what I'd call the non-phonetic phoneme. What I mean by this is that you're advocating phonemes or groups of phonemes that will always appear a certain way that is phonetically different than the symbols themselves. However, phonological systems are guided by Occam's Razor. Schane (1968) advocates a phonological system of French wherein there are no front rounded vowels and, for example, [y] comes from phonological fronting of a back vowel that is never back. Unfortunately, this relies on a complicated and circular reliance on a stress system that causes changes that precede a sweeping phonological reduction of stress where utterance final syllables are generally stressed. Now if a French baby hears [y] in all contexts, why would they think [u]? Likewise, if a Russian baby hears [tʲ] (or [tsʲ]), why would they think [tj]?
It is perfectly legitimate to have /tʲ/ as a phoneme and if the writing system leads you to conclusions different than the linguistic analysis then the writing system is wrong because writing systems never trump phonetics in determining a language's phonological aspects. Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 00:41, 19 April 2007 (UTC)
I'm getting tired!(agree to disagree?), but once again: there is no one, single correct phonological analysis for any given language: it depends on motives, convenience and other factors. Japanese ɯ, one of its 5 main vowels, doesn't sound anything like u, but for convenience that is how it's transcribed and/or analysed. The izaafa (attributive particle) in Persian is in terms of syntax a part of the attributive phrase after the head noun (kitaab i-man, my book), but in the Persian script the izaafa is written as an appendage of the head noun(kitaab-i man)due to a carry-over of Arabic script practices, and in published linguistic analyses of Persian I have never seen the izaafa attached to the attribute, only to the head. How about the final vowel in "happy" - is it in the tense series, or the lax series? The orthography doesn't give a clue...if someone from, say northern England who says hapɪ hears someone from London say æpi , would they think it's a different word, that 3 of the 4 phonemes are NOT the same, or would they regard it as essentially the same thing, with acceptable variation -- in which case, what are the TRUE phonemes of the word? I'm sorry, although a phonologist, I teach sociolinguistics and find that that perspective keeps us closer to the real world, where writing systems may and do trump not only phonological analysis but even the phonology itself (cf. spelling pronunciation). Jakob37 02:19, 19 April 2007 (UTC)
I wasn't saying that writing systems are irrelevant to pronunciation, just the phonology itself. Yes, we can agree to disagree and this was a fun conversation to have. Regards! Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 02:27, 19 April 2007 (UTC)

Drawl

It is somewhat strange to read that Russian stressed vowels are "somewhat drawled", while the article on "Drawl" defines it as "a perceived feature of some varieties of spoken English." Russian is definitely not a variety of spoken English. In Russian sources, e.g. Great Soviet Encyclopedia 3rd edition, stress in Russian is described as dynamic-quantitative, with quantity of the vowel being the primary feature. Perhaps the link should be removed or the "Drawl" article edited accordingly? Jslllnghmck 13:07, 16 July 2006 (UTC)

That's a good idea. We could also find a better word (I find "Drawl" to be a loaded term). I've read that many vowels have offglides after palatalized and phonetically velarized or rounded consonants. AEuSoes1 18:20, 16 July 2006 (UTC)

Щ + я, ё, etc. = [ɕːʲ]?

Can [ɕ] be palatalized? Isn't it already? BirdValiant 04:28, 20 July 2006 (UTC)

No. It's as palatalized as it can be. That is simply an orthographic choice. There are some unpaired consonants that are often followed by an iotating vowel when the choice of either is really arbitrary. These are: Ж, Ш, Щ, Ц and Ч. AEuSoes1 06:58, 20 July 2006 (UTC)
So should I remove any ʲ's after ɕː, ʂ, etc. in pronunciation whenever I find them?
And what about g, k, and x? Should those be written as palatalized if followed by я, е, ь, и, etc.? If I understand this correctly, their palatalized versions are allophones, right? Do you need to write the ʲ in those cases? BirdValiant 18:50, 20 July 2006 (UTC)
Yeah, in both cases it's pretty much impossible to have a palatalized version. Be careful about palatalization markers after [ʂ], though, because it could be that [ɕː] is the correct fix.
Palatalized /g/ /k/ /x/ do only occur allophonically, but the current trend in Wikipedia, as far as Russian is concerned, has been a narrower transcription. So keep those in. AEuSoes1 19:58, 20 July 2006 (UTC)

/g/, /gʲ/, /k/, /kʲ/, /x/, /xʲ/ all are separate phonemes. Minimal pairs are marginal and almost necessarily include loanwords but they are. For example, /kʲet/ (кет = Ket) vs. /ket/ (Кэт, like in Китти-Кэт = Kittie Cat), /gʲu'go/ (Гюго = Hugo) vs. /gugəl/ (Гугл = Google), /gʲandʐa/ (Гянджа = Ganja) vs. /gandʐa/ (ганджа = marihuana), /xʲot/ (Хёд = Höðr) vs. /xot/ (ход = way) and so on. And we really have no troubles to distinguish them in pronounciation. --Koryakov Yuri 21:54, 5 September 2006 (UTC)

Reorganization

Just to let everyone know, I'm about to start reworking the article. As it stands now it is very orthography-centric. As I'm working it in my head, it seems as though this will be a major reworking and it might be unrecognizable. Any new information that I put in will come from either:

  • Jones, Daniel & Ward, Dennis (1969). The Phonetics of Russian. Cambridge University Press. {{cite book}}: Cite has empty unknown parameter: |1= (help)CS1 maint: multiple names: authors list (link)
  • Halle, Morris (1959). Sound Pattern of Russian. MIT Press. {{cite book}}: Cite has empty unknown parameter: |1= (help)

Notice that they are both several decades old. That shouldn't make much difference but if anyone feels like checking then by all means do so. AEuSoes1 01:07, 18 August 2006 (UTC)

"in unstressed positions, retracted [ɨ̠] is diphthongized to [ɯ̟ɨ̟]" Are you telling me that выплыть is pronounced ['vɨ.plɯ̟ɨ̟tʲ]? -Iopq 10:09, 19 August 2006 (UTC)
SO MANY MISTAKES. I'm passing out, someone take over. -Iopq 10:27, 19 August 2006 (UTC)
These are my notes from Jones & Ward: regarding [ɨ]: page 33 "After dental, alveolar, postalveolar, and velar consonants it’s [ɨ]. When it is both preceded and followed by one of such consonants, it is fronted ([ɨ̟]) (especially when both consonants are dental or alveolar). The other variety occurs after labial and labio-dental consonants and /l/. This way it is retracted ([ɨ̠]). It is also slightly diphthongized, beginning even further back at [ɯ̟]. When it occurs at the end of a word, the tongue moves even further forward. This fronting effect is also produced by a following soft consonant." I'm not sure where I got unstressed from that.
Thanks for correcting some of my mistakes. I don't speak Russian so it's harder for me to know if translations are accurate or to spot errors. Would you say that the Russian velar nasal occurs only/primarily in loan words? The example you give seems to be one. AEuSoes1 21:03, 19 August 2006 (UTC)
The velar nasal is very rare, I only know very few examples. -Iopq 00:26, 20 August 2006 (UTC)

Aww, man... where'd the vowel diagram go? That took all of ten minutes! BirdValiant 06:44, 28 August 2006 (UTC)

ɣ in Russian language?

i've read that ɣ (not sure about the IPA) is also pronounced in some Russian words (marked by letter г). E.g in words like благо and благославить. Can someone verify it?--Constanz - Talk 13:11, 1 September 2006 (UTC)

Not in standard Russian, but possibly in other dialects. In southern dialects, г represents /ɦ/.
I don't have any examples (if I did I'd put some in the article), but /x/ can certainly voice as a result of regressive voicing assimilation. AEuSoes1 20:53, 1 September 2006 (UTC)
In standard Russian ɣ should be prononced in word Бога (Gen.sg. of Бог = God, which in turn is pronounced /box/ not /bok/). It also stands in place of /g/ in all South Russian (and Belorusian) dialects (and not /ɦ/ like in Ukrainian) and thus heard quite often even in Moscow. But it's quite easy to switch from ɣ to g (like it was done by me 10 years ago). --Koryakov Yuri 22:01, 5 September 2006 (UTC)
Wow, you're gonna need a source on that; all my sources for Standard Russian say nothing about that, although they are kinda old. Are you sure this isn't simply a non-Standard Moscow dialect? AEuSoes1 04:49, 6 September 2006 (UTC)
Some dialects have ɣ in words like его instead of v. But South Russian dialects pronounce /g/ as Ukrainian does, ɦ due to the Ukrainian influence. -Iopq 09:31, 14 September 2006 (UTC)
I can confirm that бог (god) is pronounced /box/ to avoid confusion with бок /bok/. Normally the final "г" becomes k. This is an exception. Two more exceptions where "г" is reduced to x in standard Russian instead of k are "мягкий" and "лёгкий'. Nobody pronounces k in these 3 words. As for "бог", "богом", etc. it is very common to say ɣ here but I can't say it's standard. People say both /'boga/ and /'boɣa/ It simply follows бог -> бог. I can try and find references. Three more cases I can think of where "г" is commonly pronounced as ɣ in Russian are colloquial words "ага" (yeap), "ого" (wow) and "и-го-го" (neighing of a horse). However, /g/ is also OKin these words! --Atitarev (talk) 21:55, 20 September 2008 (UTC)
In my opionion, ɣ is marginal and always optional in standard Russian. Although, some textbooks use the above examples where ɣ should be pronounced, referring to the sound as Ukrainian "г" (украинское "г"). --Atitarev (talk) 20:17, 21 September 2008 (UTC)

[ʌ.ʌ] = [ʌː] or [ʌʔʌ]?

As in the letter combinations оо, аа, etc. I think I asked this before, but I don't remember. And it isn't clear in the article either.

And the same with ии; is it pronounced ['iː], ['iʔi]], or ['i.ɪ] as in Росси´и and everything else that changes to ии due to grammar? What about if they're both unstressed? BirdValiant 20:30, 3 September 2006 (UTC)

It's [ʌ.ʌ]. As for ии, I don't think it's in any native Russian words so it's probably gonna be just like и or possibly ий: either /i/ or /ij/. AEuSoes1 20:49, 3 September 2006 (UTC)
And uh... how exactly do you say [ʌ.ʌ]? Does the period mean that you stop your vocal cords for a split second or uh... I don't know. And what do you mean by ии isn't in any native Russian words, since it happens all the time due to noun cases. Wikibooks says "Never write ие, instead write ии (Russians pronounce both, like "ee-ee")." but I don't know what that means. In English, the - usually means ʔ, as in uh-oh, but again, I don't know. BirdValiant 03:14, 4 September 2006 (UTC)
In the Phonetics of Russian by Jones and Ward (1969), it states that doubled vowels [ʌʌ] are distinct from long vowels [ʌː] by "diminishing the force of the breath in the middle of it." The difference is that "Doubled vowels constitute two syllables..." (p 213)
I'm really not sure about ии. It could be similar, a disyllabic /i.i/ that is either [ɪ.i] or [i.ɪ] depending on stress. Or it could be purely orthographic and is exactly the same as и. We'll have to wait for a native speaker to enlighten us. AEuSoes1 04:06, 4 September 2006 (UTC)
Alright, thanks. :) I thought that it sounded much more like a long vowel than the - in "uh-oh" (too lazy to look up the name of ʔ...), but wasn't exactly just a long vowel. I think I hear that in ии , but I might be wrong. BirdValiant 05:38, 4 September 2006 (UTC)
ии is a genitive case ending (and, also dative and locative ending ) for feminine words enging with ya. e.g. armiya, liniya, Rossiya, Estoniya. This ya --> ii declination is a sort of irregular, compare usual declination lampa - lampu - lampy - lampe.Constanz - Talk 17:20, 5 September 2006 (UTC)
What's irregular? Осени is the genitive of осень. Тёти is the genitive of тётя. Бестии is the genitive of бестия. -Iopq 01:16, 8 September 2006 (UTC)
Yeah, that's in the [[[Russian grammar]] page, but there's no clarity on how exactly it's pronounced. AEuSoes1 20:51, 5 September 2006 (UTC)
My untrained ear receives the word Rossiy as simply [i:] or perhaps [i.ɪ].--Constanz - Talk 10:52, 7 September 2006 (UTC)
I hear a tone change, actually. Must be dialectal. -Iopq 01:16, 8 September 2006 (UTC)

Just so that I don't start a new section for something small... so unstressed ы is basically the same as stressed ы? Where's that said in here? I miss the old article... :(

it's in the vowel reduction section regarding high vowels. Unstressed [ɨ] is [ɨ̞]. In the IPA transcriptions that I've put on Wikipedia, I haven't represented this partly because it's a cumbersome diacritic. I'll put an example in the article. AEuSoes1 05:26, 7 September 2006 (UTC)

æ in Russian

As a surprise to me, this article suggests 'ya' is pronounced as æ between soft consonants (e.g мягкий ['mʲæ.xʲkʲɪj]). I'd never noted it before, and Soviet-era books on Russian (for Estonians) do not really mention any æ sounds. My native language has ä [æ], but I have never noted the same sound being spoken in Russian.Constanz - Talk 07:25, 8 September 2006 (UTC)

Both sources that I read were very clear about this as an allophone. I group it together with the advancing of /o/ and /u/ in the same context. What is unclear to me right now is whether Russian /a/ is a phonetically front or a central vowel. If it's front, then it raises to /æ/ but if it's central then it advances to /æ/. AEuSoes1 08:38, 8 September 2006 (UTC)
I'll look it up in the library (æ business and a front/central). as for pronunciation, a native user might clarify it.--Constanz - Talk 08:43, 8 September 2006 (UTC)
I'm not native, but I can confirm that my native girlfriend clearly pronounces [æ] in the position between two soft consonants. It seems to me that she also pronounces [æ] in word-final position after a soft consonant, e.g. in the short form of her name: [ulʲæ]. Marcoscramer 07:30, 28 September 2006 (UTC)
It's front after palatalized consonants and it's advanced if the next consonant is also palatalized. If it's after a non-palatalized consonant it's back. -iopq 23:33, 27 March 2007 (UTC)
It's [ɑ] after non-palatalized consonants? or do you mean [ä] Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 02:30, 28 March 2007 (UTC)
Probably "central" is more correct. -iopq 09:30, 4 November 2007 (UTC)

Add Russian letters?

Reading that article, I very much missed any reference to actual letters. Couldn't you just include the letters that the different sounds stand for (even if there may be two sounds represented by E, etc.)? BTW, I wonder if it would get too long/ confusing to simply include a short description of the phonological signs for those signs that aren't as commonly used (e.g., [ɨ]). At least to me, such information would have been quite valuable... Thanks.

Wrong brackets in consonant table?

Is there any reason for phonetic rather than phonemic brackets around and in the consontant table? to me it just seems wrong. Marcoscramer 07:32, 28 September 2006 (UTC)

Yeah, they're allophones. AEuSoes1 20:14, 28 September 2006 (UTC)

Too many commas?

Under "Vowels" → "/i/" (close to the top of the article, section 1.1 per the table of contents) we read:

When a preceding consonant is hard, /i/ is retracted to [ɨ]. While this is phonetically central, for phonological purposes, it is considered back.

That second sentence is confusing. Apparently there should only be one comma, not two.

Which of the following was intended?

(1) While this is phonetically central, for phonological purposes it is considered back.

or:

(2) While this is phonetically central for phonological purposes, it is considered back.

Similar phrasing in the next section ("/a/") indicates that (1) above is the correct punctuation.

I have therefore deleted the second comma.

:It should be the latter for both. AEuSoes1 23:10, 1 October 2006 (UTC)

"for phonological purposes" is an independent clause so it should be surrounded by commas. -Iopq 09:07, 2 October 2006 (UTC)
Right, what he said. If you take "for phonological purposes" out it would be: "while this is phonetically central, it is considered back." Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 09:19, 2 October 2006 (UTC)

Exceptions in Vowel reduction

Exceptions to this are across certain word-final suffixes and in non-final post-tonic (after the stressed syllable) positions. E.g. память [ˈpa.mʲɪtʲ] (memory) and выглянул, [ˈvɨ.glʲɪnul] (has looked out).

I can't see where is exception? --Koryakov Yuri 10:25, 25 November 2006 (UTC)

Apparantly these words are mistranscribed. I guess they're supposed to be [ˈpa.mʲətʲ] and [ˈvɨ.glʲənul]. That's my guess anyway. Technically the "exception" is unsourced. Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 10:35, 25 November 2006 (UTC)
Well, they really can be pronounced both with [ə] and [ɪ]. --Koryakov Yuri 17:28, 26 November 2006 (UTC)

Another example: "Thirdly, when the perfectivizing prefix /vi/ is added to a verb form, post-tonic /a/ and /o/ become /i/ after soft consonants where they otherwise would not: вытянет [ˈvɨtʲɪnʲɪt] (he will pull out)." Again, I can't see the point. --Koryakov Yuri 17:28, 26 November 2006 (UTC)

Me, too: I was struck by these two things when reading. I think it should be stated in more detail what is meant there, with links to some data. Otherwise these statements do not add anything to the article. I can't imagine an [ə] sounding distinctly from [ɪ] in those positions. Yes, in pre-tonic positions, there might be variants: [ɪ] or dialectal [ʌ] (which i strange for the standard). Native speaker of Russian. --Imz 18:14, 28 October 2007 (UTC)

I can think of cases where [ɪ] would be wrong indeed in unstressed positions after soft consonants. E.g., тётя `aunt' (all such singular nominatives), тётям `to aunts' (all such plural datives). But the examples in the article do not illustrate the points, I accept [ˈpa.mʲɪtʲ], [ˈvɨ.glʲɪnul] and, say, тяну [tʲɪˈnu] `I pull' (contrary to where they otherwise would not as said in the article).

So, the first statement could be valid in some reformulated form; and I do not yet know valid illustrations for the second dubious statement.--Imz 18:58, 28 October 2007 (UTC)


According to Jones and Ward (1969):

"The phonemes /o/ and /a/ are also replaced by the phoneme /i/, but only in pretonic positions and in the case of /o/ only after soft consonants, /j/ and the hard consonants [/ʂ/ and /ʐ/], and in the case of /a/ only after soft consonants and /j/." (p 194)

They also say that "the replacement of /a/ and /o/ by /i/ occurs also in post-tonic positions in certain grammatical forms." These include:

  • when the perfectivizing prefix вы- is added to a verb form, the replacement of /a/ and /o/ by /i/ after a soft consonant, etc, still occurs
  • /o/ is replaced by /i/ after soft consonants in post-tonic positions in the following grammatical forms:
    • Nom. Sing. ending of neuter nouns, adjectives, and pronouns
    • Instr. Sing. ending of masculine and neuter nouns
    • Gen., dat., and prep. sing of masculine and neuter adjectives
    • Instr. sing. of feminine nouns and adjectives
    • Second and third person sing., first and second person pl., of first conjugation verbs

Halle (1959) says that unaccented /o/ and /e/ become /a/ and that unaccented /a/ then becomes /i/ after all palatalized segments (but not necessarily palatal consonants) but not after what he calls "Certain word final suffixes" that he doesn't seem to describe. Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 20:56, 28 October 2007 (UTC)

Thank you for the edits. Yes, I think that the current formulation for the first point we discussed is good: it is simple and doesn't contradict anything (Posttonically, /o/ and /a/ reduce to [ə]). The phonetic details that this [ə] can sound similar to/indistinguishable from [ɪ] in some positions after soft consonants would just add more confusion to this article; these details would need a good research paper that deals with sound characteristics of native pronunciations.
But as for the second point, it is still confusing: what is meant by where they otherwise would not? Well, according to current description of the processes, they would otherwise reduce to [ə]. It is again about that subtle point: whether we aim at precisely describing all the cases where [ə] is practically indistinguishable from [ɪ]. If we don't, then the default conclusion is OK for the article and the "exception" need not be formulated.
In other words (perhaps, I used some bad notions above), there are many cases when you can't tell post-tonically the difference between |a| and |i|. память is one example, pronounced the same as памить would be pronounced. What to write in the transcription? I do not object to describing such cases as just a reduction to [ə], as it is done now. But вы- is just one of these cases! I do not see why it needs an exception statement in the article. Either describe all the cases precisely or do not describe any.--Imz 22:22, 28 October 2007 (UTC)
All right, I changed "otherwise would not" to "otherwise would reduce to [ə]. I'm not sure how to account for the merging of |a| and |i| posttonically unless we perhaps discard the account in Jones and Ward and go with Halle. Halle's is an older resource but he is a pretty reputable linguist. Or maybe we can find something published more recent than forty years ago, eh?Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 23:07, 28 October 2007 (UTC)
When you were writing your reply, I was just thinking about this issue further:
It is not good not to give a hint about this typical merger of |a|, |i|, |e| (orthogaphical а, и, е) after soft consonants in unstressed positions in the article. From the above quotations, the view of Halle (1959) seems to me to be good. The default situation is that their pronunciations are indistinguishable. (Which sign to use in the phonetic transcription?..) But in some suffixes, there is a distinction: e.g., when pronounced carefully, ˈно.сит `he carries' is different from ˈно.сят, ˈтё.тя `aunt (sg.nom)' is different from ˈтё.ти `aunts (pl.nom)'. Which two different symbols to use in the transcriptions here?..
The approach of Jones and Ward (1969) leads to complex description and the cited formulation doesn't seem to be complete (and neither completely correct in the cases of the considered suffixes).
So, I would adopt Halle's approach for the description in the article. (So, no вы- "exception".) But I have no idea which symbols in the transcriptions would be the best from the view of physical phonetics.--Imz 23:14, 28 October 2007 (UTC)
I felt it is pertinent to mention here in the discussion that Halle's approach matches well some more recent explanatory work: there has been an idea that optimality theory that works on paradigms should be applied to the reduction in Russian; it is able not to produce the merger where it would violate a constraint for distinctness of different forms inside one paradigm (as in my examples above, in conjugation or inflection), but, obviously, there is no such restriction for the positions not in the suffixes that would become homophones. I'll look for references.
Yes, I think, this what I told must stem from the work by Crosswhite. But I'm not 100% sure.--Imz 23:38, 28 October 2007 (UTC)
Vowel Reduction in Russian: A Unified Account of Standard, Dialectal, and 'Dissimilative' Patterns, Intra-Paradigmatic Homophony Avoidance in Two Dialects of Slavic (and other works).--Imz 23:45, 28 October 2007 (UTC)
But those references won't help much in choosing the appropriate phonetic symbols, I guess.--Imz 23:29, 28 October 2007 (UTC)
Padgett & Tabain 2005 seem to not consider Jones & Ward (1969) as a good account. I'll use their paper to edit the section. Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 23:27, 28 October 2007 (UTC)
Interesting work. Thanks!--Imz 23:38, 28 October 2007 (UTC)

Thank you for cleaning it up! I added the Halle's formulation about the non-merge. Unfortunately, I do know all the data (of course, I do, but unconsciously): if it is definitely known that those are only flexions, where keeping the distinction between /a/ and /i/ is favored, then it would be better to write "flexions" instead of "word-final suffixes". (I also have a feeling that the distinction between /e/ and /i/ might be kept in flexions. Well, Jones and Ward probably mentioned it in what you quoted, but they called it /o/ (morphonologically, /e/ and /o/ are the same here, e.g., neuter nominative flexion).--Imz 02:22, 29 October 2007 (UTC)

I"m not sure what a flexion is. Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 02:31, 29 October 2007 (UTC)
What changes in declension or conjugation or similar. That's not very important, but would be nice if it were known. (Sorry, I'm going away, will return later.)--Imz 03:35, 29 October 2007 (UTC)

[ɘ]

In Close-mid central unrounded vowel it's written that that sound occurs in Russian: "for some speakers after /ʦ/ солнце [ˈsonʦɘ], "sun"" It's interesting and I have never heard about it before. Any ideas? --Koryakov Yuri 19:12, 26 November 2006 (UTC)

I got that from Russian Phonetics by Jones & Ward (1969). Granted it's an old source and it doesn't explain who exactly pronounces it as such, but it's an unstressed allophone of /i/ after /ts/ The example from солнце goes something like this:
/ˈsolntse/ underlying form
[ˈsolntsi] conversion of /e/ to /i/ in unstressed syllables
[ˈsolntsɨ] retraction of /i/ to [ɨ] after hard consonants
[ˈso̞lntsɨ] specifying phonetic detail of /o/
[ˈso̞lntsɨ̞] lowering of unstressed high vowels to near-close values in unstressed syllables.
So, I certainly know about lowering of unstressed high vowels but not so low! It's usually transcribed like [ɪ] or [ᵻ]. But [ɘ] - that's what amazed me. And why only after /ts/? Lowering is common place for any unstressed [ɪ] and [ᵻ]. --Koryakov Yuri 23:15, 26 November 2006 (UTC)
I noticed your edit removing the [l] and I appeal to your status as a native speaker on the truth of that. I just don't know where it would go in the process (probably near the beginning). I hope it's not too much of an original research thing on your part. Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 21:59, 26 November 2006 (UTC)
Yes, it's very amazing that they didn't mention omitting of /l/. It's a standard for Russian language (but not in солнечный). So no original research from my side. --Koryakov Yuri 23:15, 26 November 2006 (UTC)
Oh! I just noticed that you put an example for a labialized soft consonant before a back rounded vowel. My source says it's just hard consonants.
Yes, I have got sources but in Russian language. Eg. Акишина А.А., Барановская С.А. Русская фонетика. Москва, 1980.. --Koryakov Yuri 10:58, 28 November 2006 (UTC)
And my source describes /s/ and /z/ as laminal alveolar fricatives. You got a source for those things or should I change it back? Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 00:48, 27 November 2006 (UTC)
They or the same Polish sounds are described as such in several Wiki pages, eg. Sibilant consonant, Dental consonant, etc. But myself, I'm not sure: as I pronounce them they don't look like other dentals, eg. t & d. So we have change either here or there. --Koryakov Yuri 10:58, 28 November 2006 (UTC)

scornful case?

This is odd:

деньжонки [dʲɪnʲˈʐonkʲɪ] (money (scornful)).

How is money scornful? the parentheses make it seems like it's some sort of case of Russian. Would (scornful money) be a better way of putting this? Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 00:32, 30 November 2006 (UTC)

No, it means the word is said with contempt. Like девка dʲevka ("girl" (disrespectful)) would mean not that the girl would be disrespectful, but that the attitude towards the girl is disrespectful hence the speaking style. On another note I've never heard this word in my life. -Iopq 12:39, 30 November 2006 (UTC)
You've never heard деньжонки before? It's probably not a good example then. Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 22:24, 30 November 2006 (UTC)
Since Russian is an synthetic language even if the word doesn't really exist, one could just make it up at any time using деньг + онк + и = деньжонки through palatalization. But it wouldn't make it into any kind of dictionary. -Iopq
But it's not our case. I have found this word in at least 6 dictionaries with among others a quote from A.N.Tolstoy. --Koryakov Yuri 18:10, 27 December 2006 (UTC)
Well I didn't find it in the two dictionaries I looked in so I assumed I wouldn't find it anywhere else. Oh well. -Iopq 00:10, 28 December 2006 (UTC)

/i/ allophones

It is clear from transcriptions on this page and on Vowel reduction in Russian that /i/ is sometimes realised as [ɪ]. But neither article includes this sound among the allophones of /i/. Under what circumstances does this realisation happen? Widsith 14:56, 13 December 2006 (UTC)

When it's unstressed and either word initially or following a soft consonant. It's in the article under vowel reduction. Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 19:39, 13 December 2006 (UTC)

Sorry, but I still can't find this in the article. All it says under Vowel Reduction is:

In addition to this, the unstressed high vowels /i/ and /u/ become lax (or near-close) as in ютиться [jʉ̞ˈtʲitʲsə] (to huddle), этап [ɪˈtap] (stage), дышать [dɨ̞ˈʂatʲ] (to breathe), and мужчина [mʊˈɕːinə] (man).

That is rather unclear. The examples show three different values of /i/, but no explanation as to when each of them is produced. Widsith 14:19, 15 December 2006 (UTC)

It's [ɨ] when stressed and after a nonpalatalized consonant.
It's [i] when stressed and in all other positions.
When either of those become unstressed, they become near-close so that [i] becomes [ɪ] and [ɨ] becomes [ɨ̞].
If you don't get that from the article, do you think you could suggest a way for me to change it so it's clearer? Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 21:27, 15 December 2006 (UTC)

That's brilliant, thanks. I get it now. You need to add that to the article. Widsith 12:03, 17 December 2006 (UTC)

All right, thanks for your input. Sometimes my head is stuck too far up my ass for me to tell what's clear and what isn't. Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 21:31, 17 December 2006 (UTC)

The article suggests a consensus on [ɨ] and [i] being allophones. At the first sight, sounds completely wrong, because there are plenty of word-pairs with different meaning, e.g. мыла and мила, and the only difference is [ɨ] vs [i]. However, the Russian version of this article states that even Russian linguists do not agree on this topic. The Moscow school considers them allophones, where Saint-Peterburg school considers them to be different phonems. So as, this consensus is not a good idea. What do you think about reflecting both hypothesis in this article also mentioning positions of Moscow and Saint-Petersburg schools? User:itman

I've only read American texts on the subject, one of which referred to several other texts that also adopted the [ɨ]-allophone analysis; so I'm not familiar with the specifics of the dispute. A detailing of the different interpretations would be nice, but I do take issue with your example: мыла and мила contrast in the consonants /mila/ and /mʲila/, not just the vowels. Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 22:13, 23 July 2007 (UTC)
Yet another argument to call [ɨ] and [i] allophones: there are some exotic geographical names, e.g. Ытык-Кюёль that starts from Ы. Also, there is a rather complicated discussion on this topic (http://rus.1september.ru/2004/03/3.htm). It is not trivial, one of the arguments is as follows: the Moscow school insists that [k] and [kʲ] are allophones. Yet, the name "Кире" /kʲirʲe/ and the phrase "к Ире" (to Ira) /kɨrʲe/ sound differently. Remember now that [k] and [kʲ] are allophones. The only difference is between [ɨ] and [i]! Hence, these are two different phonemes. You see :-) even the Russian academia has not agreed on this topic. User:itman
From what I've heard, the "exotic geographical names" that start with Ы are pronounced as if they start with И by most Russian speakers. All that the Кире/к Ире pair works to indicate is that [k] and [kʲ] are separate phonemes (something the article mentions). I put my money on the allophone interpretation, but like I said we can and should mention in greater detail the two interpretations. I can't read Russian so I can't digest the link you've provided. Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 05:40, 6 September 2007 (UTC)

Opaque writing

An article on phonology will of course have technical content, but sentences such as "Phonetically, /a/ is a front vowel, but phonologically it is back" are a bit much. Some cleanup required. —Preceding unsigned comment added by 59.100.74.99 (talkcontribs)


Help wanted

I am the creator and main contributor of the Portal about Russia in the greek wikipedia. My knowledge of russian language is poor and I have the following question about accent: When the last syllable in patronymic names is "-OB", is it pronounced as "ov" or as "of"? I know that there are differences between north and south and one can hear it in both forms, but which is the right one according to the "rules"? If someone can answer, please do it here. --Panagiotis botsis 17:59, 11 August 2007 (UTC)

In standard Russian, if "-OB" is not followed by a vowel, it is pronounced as "of".Jakob37 01:02, 30 October 2007 (UTC)

unstressed <е> after soft consonants = /o/ or /e/?

It seems, two different analyses got mixed in the article. Both unstressed /o/ and /e/ after soft consonants are mentioned.

Whichever will be chosen, this issue needs to be explained in the article.--Imz 03:27, 29 October 2007 (UTC)


What's wrong with there being both? While there is morphosyntactic correlation between unstressed /o/ and unstressed /e/ after soft consonants, this is because the distinction is neutralized. нес [nʲos] ('was taking' masc.) and несла [nʲɪˈsla] ('was taking' fem.) show an alternation between /o/ and /i/ for /o/ (or |o|) while many words with a stressed /e/ (I can't think of any) show variation between /e/ and /i/. Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 03:51, 29 October 2007 (UTC)
For the examples, look at the roots at ru:Ять в дореволюционной русской орфографии.
Yes, I agree, there can be both. (The definition of a phoneme for Russian phonology is a thing full of controversies. Different schools have different views.)
So, if we have both phonemes in unstressed positions, and the distinction is neutralized, in all the applicable rules they should go side by side. To be correct, we should write something like {/e/, /o/} everywhere. --Imz 04:26, 29 October 2007 (UTC)
It would be quite controversial to say that you can establish phonemic distinctions in vowels by using unstressed syllables. Can unstressed vowels really be used as minimal pairs? Generally, their phonemic representation is established by referring to related words or morphemes that are stressed. If no such related forms exist, the unstressed vowel should be represented as a neutral unstressed vowel (I usually use Unicode 1D4A). In the case of нес and несла, the traditional (pre-structuralist, where the vowels, not the consonants, carry the palatalising force) analysis regards these as two separate vowels: ё and е , thus nothing to do with o. Since both ё and е are palatalising, in unstressed environments they are the same, therefore it seems more logical, despite conventional orthography, to consider несла as containing an unstressed ё since the word clearly is paired with нёс.Jakob37 01:34, 30 October 2007 (UTC)
Yeah, I think that what is the case here is that orthography indicates the reduction of /o/ after soft consonants. Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 05:33, 30 October 2007 (UTC)

CONGRATULATIONS TO ALL!

This talk-page has become quite extensive, and full of controversy, with many back-and-forth exchanges, but it is remarkably free of rancor and nasty ad hominem remarks seen on many other talk pages. Keep up the good work!Jakob37 02:03, 30 October 2007 (UTC)

Velar consonants are palatalized when following /i/; ...

within words, this means that velar consonants are never followed by [ɨ].

May be I don't understand this implication and it's my own problem. but I don't see how "Velar consonants are palatalized when following /i/" ( e.g. миг [mʲik] - k velar non-palatalized ) and "within words, this means that velar consonants are never followed by [ɨ]" - what about гыканье and кыш. Linefeed (talk) —Preceding comment was added at 18:14, 19 January 2008 (UTC)

My bad, that should be "velar consonants are palatalized when preceding /i/" — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 22:44, 19 January 2008 (UTC)

I think I've found more proof of /kʲ/!

So the only minimal pair for /k/~/kʲ/ ever given is "этот кот" /k/ ~ "это ткёт" /kʲ/.

What do you say to "к эфиру" ~ "кефиру"?? I've quizzed native speakers, and it appears to be another pair supporting /kʲ/ as a phoneme in its own right. —Preceding unsigned comment added by Benchik (talkcontribs) 00:16, 28 January 2008 (UTC)

Neat. The phonemic status of /kʲ/ is the result of borrowing. Keep in mind, though, that [kʲ] still acts as an allophone of /k/ before front vowels. It's sort of like English [tʃ]. It's definitely a phoneme, but for some people (including myself) it's also an allophone of /t/ before /r/: /triː/[tʃɻɪi]Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 00:31, 28 January 2008 (UTC)
  • It's an old discussion but you seem both confused.
  • Russian letter э never palatalises consonants, it is pronounced as /e/ and "к эфиру" is pronounced as /kefʲiru/. Have you mixed up Russian э and е? I am saying Russian, not Cyrillic because the Ukrainian е has a different value - /e/. The Ukrainian є (looks similar to the Russian "э") is actually /je/.
  • However, letters я, ю, е, ё, и and ь palatalise preceding consonants. There are no words with with final or pre-consonant кь/гь but there are plenty of ке/ги, ки/ги in native words and кю/гю, кя/гя, кё/гё in borrowings. In all cases /k/ and /g/ are palatalised to /kʲ/ and /gʲ/ with some exceptions where "е" may be pronounced as "э" (loanwords).
  • Benchik, your quest for palatalised /kʲ/ is valid, they do exist in Russian but not in the example you gave. There are many native and borrowed words with /kʲ/ and /gʲ/. --Atitarev (talk) 13:07, 18 September 2008 (UTC)

EDIT:

  • It's also worth mentioning that although different words are pronounced jointly when the 2nd word starts with a vowel, e.g. "в окне" is read as if it's one word "вокне", it does not apply to the iotised vowels (я, ю, е and ё), so "к ели" and "к ёлке" are pronounced as if they are written "къели" "къёлке", preserving the initial /j/ sound and not palatalising the preceding consonant. --Atitarev (talk) 21:40, 18 September 2008 (UTC)
I'm not sure what it is I've said that indicates I'm confused. Other than wording that suggests pronunciation depends on orthography (I'd say that я, ю, е, ё, и and ь indicate palatalization) I can't find anything to contradict in what you've said. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 19:27, 21 September 2008 (UTC)
You seemed to be agreeing that "к эфиру" was pronounced "кефиру" (/kʲ/)? That's why I said you seemed both confused. --Atitarev (talk) 20:05, 21 September 2008 (UTC)
I think Benchik was saying that к эфиру and кефиру form a minimal pair. That is, the former is [kɨ'fʲiru] and the latter [kʲɪ'fʲiru]. That's how I understood it anyway. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 20:39, 21 September 2008 (UTC)
I see. Sorry, I misunderstood. Perhaps, another one: кури - Кюри. --Atitarev (talk) 21:43, 21 September 2008 (UTC)

Velarization?

I read in the article on velarization that the hard consonants of Russian are velarized, like in Irish, but this article states that they are plain. Which is it? —Preceding unsigned comment added by 76.185.21.163 (talk) 03:26, 4 March 2008 (UTC)

The only verifiable thing I've seen says that "hard" consonants are only velarized before back vowels. Looking at velarization, it seems to be sort of glossing over phonetic nuances and caveats (in both Russian and Irish). In certain environments there is contrast between velarized and palatalized consonants, as with нес [nʲos] ('was taking' masc.) vs нос [nˠos] ('nose'). The velarization isn't transcribed because it's redundant. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 03:41, 4 March 2008 (UTC)

Russian borrowing from Finnish? (or vice-versa)

It must be like this! Just read this bit of the article: синего [ˈsʲi.nʲɪ.və] ('blue' gen. sg.). Hah! And "sini" means "blue" in Finnish! Hence these two languages appear to have borrowed more from each other than it might pop into the eye at first glance. -andy 78.51.104.41 (talk) 19:44, 24 March 2008 (UTC)

Yeah, that wouldn't be surprising at all. But what does that have to do with phonology? — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 20:02, 24 March 2008 (UTC)
This is off topic. But it might be so or might not be (and be a coincidence, or a borrowing from Iranian): Uralic etymology, Vasmer's Russian etymology, the last one saying: Similarly sounding Finno-ugric words: Finnish sini "синий цвет", Mordvin Moksha śeń, Erzya seń "синий", Mari šinžiš -- same meaning; cf. Сетэлэ, FUF 2, 250; Миккола, Berühr. 162; ÄВ 75. Mikkola courageously supposed a borrowing from Iranian.--Imz (talk) 22:55, 24 March 2008 (UTC)
I should have added also this link above--the Indo-European context for the above Vasmer's etymology according to starling: Proto-IE: *k'ēy(w)-, *k'yē(w)- (which links among others to Old English: hǟwen `blue, azure, purple, discoloured, green' with similar suffix and similar meaning). So both Finnish and Russian have certain possible etymologies in their families and it is probably not a borrowing between Russian and Finnish. Perhaps, an earlier borrowing between some earlier Finno-Ugric dialect and an earlier Indo-European dialect, or just independent words.--Imz (talk) 02:11, 25 March 2008 (UTC)
Since we're talking about borrowing, it would be interesting to see the article elaborate on loanword phonology. Information on the phonological history of Russian would probably delve up loanwords that came before or after certain phonological changes. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 23:58, 24 March 2008 (UTC)
Yes (but I'm not the one who is ready to write about this).--Imz (talk) 02:11, 25 March 2008 (UTC)

tvʲ and dvʲ are standard

Reference: [4] (section 3.3); also listen look at wikt:ru:твёрдый, wikt:en:твёрдый, listen there.--Imz (talk) 23:36, 1 June 2008 (UTC)

I don't speak Russian, could you quote and translate the part that you're referring to? Also, I don't know the credentials of this website and I'm really hoping for a scholarly source. Can we find any of those? Currently, what we have is Halle (1959, page 68) saying "Before palatalized labial and dental consonants and before ||: paired dental consonants" are palatalized in the Moscow Standard. He details some complicated differences in "literary pronunciation" which tells me that Russian was/is in a state of change and/or variation in this regard, which makes the age of this source all the more problematic. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 02:05, 2 June 2008 (UTC)

Translation: A fragment Variants of Russian literary pronunciation of a writing by Liudmila Verbitskaya (doctor of philology; until recently the rector of Saint-Petersburg State University) Давайте говорить правильно! (1993):

3.3. Characterization of the modern Leningrad-Petersburg and Moscow pronunciation
[...]
Some orthoepic properties of the Leningrad variant are now the common literary norm. Thus, [...] the number of cases of assimilative softening of consonants in sequences C1C'2 has declined (дверь is /dv'er'/, твердый is /tv'órdыj/).

This fact is too obvious to me for me to become excited about searching for other sources. It is not a surprise that Halle tells about a change going on because it's actually true. :)

Also, another thing. I agree that the content of Wikipedia should be verifiable, and I would be glad to try to provide references for the information I added, if you asked me to because you thought there were not enough of them. But I disapprove of your behavior: you presupposed that you were more correct than others, and instead of paying attention to the information others and I added/supported, and instead of considering it to be a common task to find better references for it, you were reverting to your variant and putting on others the burden of proof (that their variant is not garbage). It would be nicer for you to put a sign of doubt at these variants, because there were strong reasons to think the not-yours variant (tvʲ) is not random garbage: first, it has stayed in Wikipedia for quite a long time and probably it has been seen and approved by many readers, and, second, after you put your variant, I, a native speaker of Russian, changed it back. Is it OK for a linguist, if a native speaker gives data contradicting the description the linguist has, to almost ignore and hide this data? I suppose the linguist should--in contrary--expose it and ask for comments.

And, if you or someone are curious about these variants and the discussed change, here is a little personal comment from me: I don't perceive the difference between the two variants (I mean: dʲvʲ doesn't strike me as an incorrect pronunciation, I simply don't notice it, unless I'm consciously listening for this peculiarity). I'm even not quite sure about my spontaneous pronunciation. I don't know how people with different backgrounds than mine perceive these variants, perhaps they could feel the difference. (My background is that my grandfather was also born in Moscow and might have pronounced dʲvʲ, so I might have learned it from him that this variant is also OK.)--Imz (talk) 17:33, 6 June 2008 (UTC)

Okay, so we have one source saying that it's [tʲvʲ] and another source saying that the occurrence of [tvʲ] has increased. You're right that this is a minor feature, but it would be nice to see a more current scholarly source going into such minor details. — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 04:42, 7 June 2008 (UTC)
I'm also a source, and people I know (and whose pronunciation I can record) are a source, the person who reads it aloud for Wiktionary is a source, the TV and radio are sources, the announcements in the underground in Moscow, St.-Petersburg are also sources ("doors are closing" record is played at every station, двери "doors" is an example of the discussed combination). This kind of contrast can be heard by any native speaker (it's not something becoming meaningful only under a certain understanding of phonetic theory like the issue of retroflexion). Scholarly works would not be made out of the air, they would be also based on the real pronunciation. Providing recorded pronunciation would make an article verifiable.
So, my point is that this kind of facts has a right to stay in the Wikipedia article even before scholarly sources are found for them. A record of a native speaker is a sufficient source for a transcription.
I condemn you for deleting meaningful content from Wikipedia articles. (I'd rather leave it and just mark it as lacking enough enough sources.)--Imz (talk) 10:13, 8 June 2008 (UTC)
Actually, read WP:V and WP:NOR. You are not a source. Most of the things you listed are primary sources and we rely on reliable, third party sources. The contributions of native speakers such as yourself are greatly appreciated but we can get things like this native speaker who says that к х г are "never soft" when they most certainly can be. In this instance, your status as a native speaker has brought the questionable nature to the assertion that dental consonants are palatalized before palatalized labial ones; what we can do is transcribe such words with tvʲ/dvʲ (with the pronunciation you prefer), keep the statement in the article that says "Before palatalized labial and dental consonants or /lʲ/, dental consonants (other than /t͡s/) are palatalized" but mark it with {{dubious}} which can then link to this conversation.
Now, what exactly have I deleted? — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 19:04, 8 June 2008 (UTC)
I know quite well about verifiability and "no original research" principles, agree with them and like them. (Perhaps, I need to elaborate more on the issue of "primary sources".) And I believe that in the case of the description of a language providing a record of a native speaker is just the limit of verifiability: any other research will give less information and less exact information about the language. A transcription without a record is not as verifiable as a transcription supplied with a record. For theoretical interpretations, I'd say, only cited statements from scholarly sources are acceptable. So my ideal is: for "individual" facts resp. languages, there is a justifying record of a native speaker (or, if the language is dead, some written source), and for "generic" statements, there are references to scholarly works.
Funny record. Well, I wouldn't argue the same thing as I did here for tvʲ/dvʲ for that kind of statements. "к г х are never soft" is a generalization (a little theory), it needs to be sourced in accepted published work. Also, another part of my argument was against deleting; and as for "к г х are never soft", I would agree it must have been deleted because it can easily be proven to be wrong (by demonstrating a recorded counterexample). So nothing of what I argued about would apply to that case.
As for the "deletions" by you: I meant that by hiding/removing the variant given or agreed upon by native speakers (tvʲ) you were actually deleting some bits of meaningful information about the described language. Input from native speakers is information.--Imz (talk) 20:34, 8 June 2008 (UTC)
Well, I'd hardly call that "meaningful content" and hardly worth condemning. It's a minor phonetic note. Are we good for now? — Ƶ§œš¹ [aɪm ˈfɻɛ̃ⁿdˡi] 21:14, 8 June 2008 (UTC)
Ok, sorry for that word. I hope you could see my reasoning, and if not agree, but just be more careful in such deletions of what I would call "meaningful bits of information".--Imz (talk) 21:38, 8 June 2008 (UTC)

Just one note: as someone who has to do with Russian a lot, I agree with Imz that "твердый" normally has a non-palatalized "t". I'm not sure I've ever heard anything like the other version. But I agree with Aeusoes1 about the "primary sources" issue. There can be no such thing as an "indivudual fact" as opposed to "theoretical interpretation" here; anything beyond the sound recording, any verbal claim about the nature of a sound would be, not a mere fact, but an interpretation of the facts. No matter how obvious that interpretation can seem to a native speaker or even an amateur phonetician, it can still be incorrect. So only a scholarly work is an acceptable source for any statement on phonology or phonetics. But personal knowledge such as native speaker competence can be used as a basis for marking an unsourced claim as dubious and for removing it. As usual, the burden of proof is on those wanting to keep a claim. --Anonymous44 (talk) 18:43, 10 June 2008 (UTC)

Thanks for the comment. I agree with your view. Some additional thoughts/notes from me:
  1. I put it differently above because I wanted to emphasize the different levels of complexity of interpretations (but anyway, I agree that ultimately all the transcriptions must be sourced), and immediately deleting a trivial interpretation is more weird than deleting a more complex unsourced theory. It feels like being told "I'm going to ignore you", not like "I'm going to cooperate with you".
  2. Ultimately, the burden of proof is on those who want to keep a statement, but if I'm not assumed to be crazy or malicious, and I'm contributing a quite simple native speaker's interpretation, I'd like to expect more collaboration, and not a decision made by one person to delete the information, so that the work starts feeling like a fight between two sides. I would be glad if the attitude is rather like "Ok, let's find good sources then for your statement" and not immediate deletion. The whole community could help me find good sources, that's why I suggest first to mark dubious statements and request for sources.
  3. As for the particular issue of the status of the statement about the softness of [t]. As I said, basically, I agree with you that it's a theoretical interpretation, not just a mere fact, and I oversimplified things in my reply above. But, since we are talking about this, I want to add one more curious note about this: there is a sense in which the native speaker could be considered the source of such a judgment on softness/hardness of the [t]: the contrast between [t] and [t'] is phonemic for Russian, so it can be told by a native speaker (it's almost the definition of a phonemic contrast); and actually a physical analysis of the sound by a phonetician wouldn't help, because the border between the phonemes is primarily available only as the native speaker's competence. (Talking about phonemes is always a complex controversial thing, so I suggest not to go into long discussions about the definition of a phoneme here.)--Imz (talk) 20:58, 10 June 2008 (UTC)
No comment concerning the particular conflict between you and Aeusoes1 - I only read the last couple of paragraphs on this talk. As for your last, theoretical point - well... I don't think it's a sufficient reason to change the principle. There can be thousands of issues that mislead the native speaker. The first thing I think of is that in our case, there happens to be no phonemic contrast between palatalized and non-palatalized [t] in this particular position, so the speaker need not be able to distinguish between them in this position. But there's something more important - the native speaker may have problems with the difference between the phoneme and the allophone. He may always think he hears/pronounces the "typical", "classic" realization of the phoneme, even though he is actually hearing/pronouncing a fairly "aberrant" allophone - to the point of coinciding with the "typical" realization of another phoneme. How many people are aware of, say, all the reduction of unstressed vowels they are making? We think "plod" and pronounce "plot", we think "ljebjedj" and pronounce something like "ljebjitj" (sorry, don't have the patience to copy proper transcriptions). In Bulgarian, we will often think "l" even though we actually hear or pronounce non-syllabic "u" - and even though we also have "real" non-syllabic "u" in many words! Analyzing oneself to find out what one really pronounces can be quite hard - even for someone with some background in phonetics. To just rely on native speaker competence would be quite unrealistic, and one must always be very suspicious. It would be quite conceivable that both you and I think we hear "tvjordyj" even though we actually hear "tjvjordyj" - just because we know that that it's tvjordyj phonemically, or even simply orthographically. I would be just as suspicious towards a native speaker of, say, Bengali, who contradicts sourced scholarly material about Bengali. And I certainly have encountered many similar mistakes regarding Bulgarian. --Anonymous44 (talk) 00:21, 11 June 2008 (UTC)