All Slavic Dialects: Non-Russian akanye

arn00b

Senior Member
English
The Wikipedia page for "akanye" mentions the phenomenon in Ukrainian, Slovene, Serbo-Croatian and Bulgarian dialects.

Are there any speakers on the forum who speak a dialect with akanye? How common is it? Is it growing/shrinking? How standard is it? Is it considered archaic or is it now developing? Is it youths using it in urban slang or is it an aging rural population?

What are the characteristics of it? Is it similar to Russian? When does it occur? What are the rules (or so to speak) or conditions of this phenomenon?

Does your dialect have ikanye or yakanye as well?

Are there any audio samples we can listen to to get a sense of it?

Thanks, everyone. I really appreciate any help.
 
  • Panceltic

    Senior Member
    Slovenščina
    Not my dialect, but akanye is widespread in South-East dialects of Slovenian (Dolenjska). I think it's perceived as a rural trait and is avoided by the speakers when communicating with anyone outside their local community.
     
    As a sidenote, ikan'ye and akan'ye are most probably not the phenomena of the same league in Russian. Akan'ye lies in the core of the language: every unstressed o is pronounced as an a-like sound, even when a word is spoken in syllables. In contrast, ikan'ye is a phenomenon of the allegro speech: it is absent in the careful pronunciation and often in songs, i. e. it is considerably younger. Moreover, in my idiolect (St. Petersburg, which we consider the standard language), the final unstressed -я, -е and are normally pronounced differently (i. e. Ваня, Ване and Вани are not homonyms, unless pronounced fast and careless), though for speakers from other places it is not necessarily so; on the other side, сено and сена are full homophones (I know that speakers of the northern dialect pronounce the unstressed o, but I have never heard anyone speaking this way in the real life).
     
    Last edited:

    marco_2

    Senior Member
    Polish
    You haven't mentioned Belorussian here, where akanye is very consecutive and even reflected in spelling.
     
    Last edited:

    bibax

    Senior Member
    Czech (Prague)
    ... on the other side, сено and сена are full homophones (I know that speakers of the northern dialect pronounce the unstressed o, but I have never heard anyone speaking this way in the real life).
    In school we pronounced сéно/сéна, лéто/лéта, опоздáла/опоздáло, etc. differently, no homophones (same for Ваня/Ване/Вани, pronounced -a/-e/-i). I think it would be too confusing for small children to pronounce the endings in лéто/лéта, etc. the same way (in Czech léto/léta, seno/sena, opozdila/opozdilo - no homophones). The rule was: every unstressed o before the stressed syllable is pronounced as an a-like sound (e.g. malakó, aná apazdála, but ľéto, anó apazdálo). So in the upshot the endings retained their meanings even in the spoken language. After all we understood Russian quite well in comparison with the German (in the DDR) or Hungarian children.

    To the original question: no "akanie" in Czech (including dialects).
     
    Perhaps not a topic of this thread, but one of the two prevalent theories relates akan'ye to the substrate influence of the Baltic-speaking population of the middle latitudes of pre-Slavic Eastern Europe, compare Russian dialects - Wikipedia (+ Belarusian to the west) with Balts - Wikipedia (presumably Baltic-speaking archeological cultures in violet). If it is true, akan'ye in other Slavic areas should have a different origin.
     

    Michalko

    Member
    Slovak - Slovakia
    No akanye in Slovak, but in some dialects, ä, pronounced as e by most people today, is instead pronounced as a. So mäso (mæso in the traditional pronounciation, meso by most speakers) gets pronounced as maso (like in Czech) by some people.
     
    Concerning the origin of the East Slavic akan'ye. The major obstacle is that there are no ancient documents reflecting the actual speech of the areas where this phenomenon appears today, so it is impossible to decide whether the late medieval texts with a in place of o reflect (a) the initial development of this phenomenon, or (b) its penetration into the dialects of the scribes from a narrower area, or simply (c) the drop in the level of education. For example, the Old Novgorod vernacular speech (known from the birch bark manuscripts) has turned out to be very different from the standard written language of the first centuries of the last millennium, but only faint signs of this could be found in the official texts (laws, chronicles, church literature) written in Novgorod.

    Also, it is important that akany'e (and yakan'ye and ikan'ye) are not related directly to the vowel reduction: for example in the old Belarusian standard language (e. g. in the radio records of the 20–50's), the unstressed o and e become a and ʲa but don't undergo any serious reduction in quantity (modern Belarusians mostly pronounce unstressed a like modern Russians do).

    Also, there are several types of akan'ye in the dialects, e. g. in some of them it depends on the surrounding vowels, in others doesn't, and there is disagreement which one should be considered original.

    I personally think that the merger of unstressed a and o must postdate the East Slavic development ъ>o, because I have never read that in any dialect the unstressed o resulting from ъ has not merged with the basic unstressed o, i. e. that поставити>поставить and съставити>составить have different vowels in the prefix. The Baltic substrate theory in its most straightforward variant implies that, unlike in other Slavic areas, the unstressed Late Common Slavic *a didn't become o and, when the old shortened in unstressed position, both a's (the originally short and the newly shortened) simply merged (i. e. Nom. Sg. *sēna and Gen. Sg. *sēnā merged directly into the modern Nom./Gen. Sg. [се́на]). This, however, leaves unexplained why ъ>о merged with a instead of remaining on the o stage and filling the gap in the unstressed vocalism. In the more moderate substrate scenario, the Baltic heritage may have induced the open pronunciation of o (and e), which later developed into akan'ye and yakan'ye as we know it.
     

    arn00b

    Senior Member
    English
    Thanks, @ahvalj. This has been very informative.

    Also, it is important that akany'e (and yakan'ye and ikan'ye) are not related directly to the vowel reduction: for example in the old Belarusian standard language (e. g. in the radio records of the 20–50's), the unstressed o and e become a and ʲa but don't undergo any serious reduction in quantity (modern Belarusians mostly pronounce unstressed a like modern Russians do).

    This is interesting. So does that mean that 20's - 50's Belarusian made no difference between, for example, an etymological a and an unstressed o? How would something like "горад" be pronounced? (Is this a good example?)

    What happened after the 50's? Did the "shifted" vowels become reduced through "normal" language evolution or was it through indirect (neighbors) or direct (government reforms) Russian influence?

    Could this phenomenon (akan'ye without reduction in quality) be described as a vowel shift or is that too simplistic?

    If I understand this correctly, then the akan'ye (and ikan'ye) and the vowel reduction are two separate phenomena that occurred at two different time periods. Could the same have occurred in Russian?
     
    Well, as far as I understand, no akany'e idiolect makes difference between unstressed a and o of any origin.

    The lack of reduction in Belarusian is prescribed by the normative grammar, though I never could figure out why real people I heard in the media didn't speak this way. I have finally discovered this Belarusian pronunciation when listening this series of programs: На хвалі Свабоды (the anniversary programs of the Belarusian service of the American propagandistic radio "Freedom/Free Europe" dedicated to fifty years of its history). Of course, there is this artificial articulation of that time, but when comparing the speech of the emigrants vs. that of their Belarusian contemporaries (there are examples of both), or (e. g. in the program of 1954) the voice of one of their female announcers then and in the early 2000's vs. the speech of their modern hosts, the difference becomes rather obvious. Some of their announcers were emigrants from the western part of Belarus that was under Polish control in the interwar period (annexed in 1920–1921 by Poland from the Soviets and then reacquired in 1939 by the USSR), so this could explain that their Belarusian was subject of different standardization than in the Soviet part (my Russian ear hears a noticeable foreign accent in them but no foreign accent in the fragments of the Soviet Belarusian broadcast of that time). Modern Belarusians from both former parts of the country (cp. the modern voices in the same files) still may pronounce the unstressed a clearer than many Russians do, but it is no more a longer, emphatic sound as in those records. This radio, however, has two modern announcers, one native of Poland [Максімюк], other native of the USA [Данчык], who still have this clear unstressed a and the overall foreign accent.

    So, I think that (a) the Polish Belarusian pronunciation was based on more remote dialects that may have had a different flavor of the unstressed a and (b) the Soviet Belarusian was influenced by Russian (that is much clearer in Ukrainian, whose phonetics is much more different from Russian and where therefore the influence is more obvious).

    Akan'ye (opening of the unstressed o) can be compared with closing of the unstressed o (towards u) and the unstressed a (towards ə) found in pre-war Ukrainian (again described in the old books but virtually eliminated now) and in Bulgarian (I usually fail to discern the unstressed a vs. ъ and o vs. u in this language) [the schwa in Russian and Bulgarian are acoustically two very different sounds]. Don't know if the evolution of the unstressed vocalism can be regarded as vowel shifts: that's probably a matter of taste. At least this is not the chain reaction like the early Ukrainian ě>i pushing the old i>ɪ.

    Among standard Slavic languages, the pronounced vowel reduction is found in Slovene (with loss of unstressed vowels), Bulgarian, modern Belarusian and Russian (without such a loss in normal circumstances); of them Slovene (as far as I understand) simply shortens the vowels (only losing the distinction between open and closed e and o), whereas in the other three languages they initially merge in pairs when unstressed (Russian also merges three vowels, a, e, and i, after palatalized consonants in most positions). This, along with the above phenomena in Belarusian and Ukrainian, makes me think that this merge doesn't necessarily need to be related to the quantitative reduction: it may, however, slacken the unstressed vocalism favoring the development of the proper vowel reduction as a consequence. In any case, if the vowels are not dropped, there is no way to identify in the old texts when the unstressed vowels cross the boundary between simply short and reduced. For the reasons described in #3 I am sure akan'ye in Russian is older than ikan'ye/yakan'ye and the reduction itself.
     
    Last edited:

    Christo Tamarin

    Senior Member
    Bulgarian
    http://www.promacedonia.org/jchorb/st/st_2_b_izt_3.htm#smoljanski

    5. Характерна и занимлива фонетична особеност на смолянския говор е и т. нар. а к а н е — изговор на неударено o като а: вадѝца, гòрạ—гạрѝца, кạбѝлạ, дạшòл, гạлềм, ạфчềри, ạтѝде, мạмà. Тая особеност се среща в речта на помашкото население в южната част на говора и много напомня южноруското акане. В смолянския говор обаче това акане се явява само пред ударена сричка. След ударена сричка o се редуцира в у (мàслу, жèлну), т. е. изговаря се, както в повечето български говори. Интересно е, че наред с акавия изговор на неударено o ce среща в няколко случая акав изговор и вместо у: ạхàпạ (ухапа), стạдèнạ, кạрбàн (курбан).

    6. У помаците, и то най-вече в южната част на смолянския говор, пред плавните съгласни р, л, когато са след съгласна, се изговаря слабо тъмно а, та се получават форми, напомнящи руското пълногласие: вạрêтèно (вретено), сạрềдạ (сряда), сạрèбру, бạрàдвạ (брадва), сạтạрѝжену (стрижено), тạрồн (трън); пạлềвạ (плява), мạл’ềку (мляко), хạлềп, кạлềште (клещи).

    I will do a translation upon request.

    I think Stoykov was wrong. Both these features are not related to East Slavic. Rather, these are Turkish influence. Note that both these features can be found only in dialects spoken by Muslim people.
     
    Last edited:

    arn00b

    Senior Member
    English
    Not discounting the idea of Turkish influence, but what part of the Turkish would have influenced this phenomenon? I'm only familiar with modern Turkish and this language features what could arguably called the "opposite" of akan'ye - vowel harmony.

    The slightest shift in vowels, such as Kurds struggling with ö/ü/ı and using o/u/i instead, often leads to complete confusion.

    Not using vowel harmony is considered funny and is used as a joke in some comedy shows.

    Turkish is probably the most vowel-strict language in the region, with Arabic on the other end of the spectrum (not just regionally.)

    Having that said, I don't know much about Ottoman Turkish or other Turkic languages that could have contributed to that in Bulgaria. But I don't know of an equivalent phenomenon in Bosnia.
     

    Christo Tamarin

    Senior Member
    Bulgarian
    Exactly, the strict vowel harmony was the factor which caused the changed in p.5: changing O to A in order to achieve a more frontal vowel, as those that follow. In the case дạшòл, before applying "the vowel harmony" tendency, it probably was дạшъл in the past, same as in the standard Bulgarian and in most of the eastern Bulgarian dialects.
     

    arn00b

    Senior Member
    English
    That makes a lot of sense, actually. I didn't think of it that way.

    What makes the Turkic influence more obvious now is the breaking up of word-initial consonant clusters.
     

    Panceltic

    Senior Member
    Slovenščina
    Among standard Slavic languages, the pronounced vowel reduction is found in Slovene (with loss of unstressed vowels), Bulgarian, modern Belarusian and Russian (without such a loss in normal circumstances); of them Slovene (as far as I understand) simply shortens the vowels (only losing the distinction between open and closed e and o)

    In the Gorenjska dialect, all reduced vowels generally merge into a schwa. This is also present in my dialect though I'm not exactly from Gorenjska.

    So brat, pes, miš, kup are all pronounced with the same vowel.

    In oblique cases, the vowels are pronounced clearly because they are not reduced any more (not being in the last syllables).
     
    Concerning Ukrainian, here is a relevant paper: Назарова ГВ · 1977 · Аканье в украинских говорахНазарова ГВ · 1977 · Аканье в украинских говорах.pdf

    P. S. Ukrainian (including the standard language) also knows assimilative change of the unstressed o>a before a stressed syllable with a: багатий, гаразд, гарячий, кажан, калач, качан, хазяïн, чабан etc. (Жовтобрюх МА, Русанівський ВМ, Скляренко ВГ · 1979 · Історія української мови. Фонетика: 288–291).
     
    Last edited:
    In the Gorenjska dialect, all reduced vowels generally merge into a schwa. This is also present in my dialect though I'm not exactly from Gorenjska.

    So brat, pes, miš, kup are all pronounced with the same vowel.

    In oblique cases, the vowels are pronounced clearly because they are not reduced any more (not being in the last syllables).
    But brat, pes, miš and kup are stressed, or?
     

    Christo Tamarin

    Senior Member
    Bulgarian
    When referring just to akanye, there is no akanye of east-Slavic type in Bulgarian dialects, despite other claims. Wikipedia used the same same source (Stoykov's dialectology) as that of the quotes I provided above (points 5 and 6 about the Smolyan dialect). Despite of the mentioned east-Slavic connection there, both those features are influenced by Turkish - Turkish is spoken in neighborhood, those two features are not common to the Smolyan dialect - rather, they are applied by Muslim people only.

    However, if we consider a broader subject, the vowel reduction, then there is a vowel reduction in Bulgarian pretty similar to that in Russian. Please note that east and standard Bulgarian shares with Russian the lexically-important and free and unpredictable and mobile stress position.

    The vowel reduction means that:
    • the number of vowels recognized in unstressed positions (3 or 4 in Bulgarian) is less than the number of vowels recognized in stressed positions (6 in Bulgarian: A,Ъ,О,U-У,Е,I-И).
    • a vowel in an unstressed position may have another quality.
    Especially, we can observe the following phenomena in Bulgarian:
    • The vowels A and Ъ are distinguished in stressed positions only. All speakers, all listeners, always. An unstressed vowel A or Ъ can be implemented at any phonetic position between A and Ъ including the bounds A and Ъ. This is a similar phenomenon as the akanye in Russian. However, no standard implementation is prescribed in Bulgarian. Examples: In the following words, A and Ъ are not distinguished, despite of the orthography: Петър, Димитър, Искър, Вардар, лекар, палав, мъртъв, садѝ (Садѝ го пиперът по-нарядко!), съдѝ (<=сѫдѝ, Съдѝ според доказателствата!).
    • The vowels O and U (Cyrillic У) are distinguished in stressed positions only. An unstressed vowel O or U (Cyrillic У) can be implemented at any phonetic position between O and U (Cyrillic У) including the bounds O and U (Cyrillic У). Valid for more than one half of speakers and more than one half of listeners. However, some people do distinguish unstressed O and U (Cyrillic У) - TV and radio speakers are hired among them.
    • The vowel I (Cyrillic И) never changes in unstressed position.
    • The vowel E in some unstressed cases changes to I (Cyrillic И), in other unstressed cases it changes to Ъ with preserving the softness of the preceding consonant. Valid for less than one half of speakers and less than one half of listeners. Surely, both cases cannot be mixed. However, it is hard to express the rule when each of the cases applies. Anyway, every word-final E can go into I (Cyrillic И). Examples: Both Бèлене and белени can be heard as Бèльъни (definitely not бèлини).
     

    metaphrastes

    Senior Member
    Portuguese - Portugal
    You haven't mentioned Belorussian here, where akanye is very consecutive and even reflected in spelling.
    Now a question: are there in Belorussian cases of unwritten akanye, that is, not reflected in spelling?
    Otherwise said: if all akanye is reflected in spelling, it would in practice eliminate every possibility of writing unstressed "o", since all of them would have been written as "a". A consequence would be that every written "o" by need is stressed.
    Another consequence would be that the poll of options to inflect words would be much lesser, since even in writing there would be no possibility of changing the ending, for example, from unstressed "o" to unstressed "a".
    Since I have nothing but a rudimentar knowledge on Russian, I wonder how actually works the written akanye in Belorussian.
     

    metaphrastes

    Senior Member
    Portuguese - Portugal
    And, not related with the last post, if it is of any interest, Old Church Slavonic, theoretically and so far I know, would suffer no vowel reduction, nor akanye, ekanye or ikanye. However, in many (Russian) recordings of divine Liturgy, it is possible to hear clearly the akanye, mostly from the priests who sing alone the ektenias and exclamations. Any decent choir will have to work to pronounce homogeneously the vowels, consciously, and syllables in general are produced one-by-one, with regular duration (all of this avoids or strongly reduces akanye), while the recitative used by priests favor a more spontaneous diction, and then akanye often appears unhindered, though probably unwillingly or unaware of itself. Probably some would perceive akanye, in this case, as an error.
     

    Милан

    Senior Member
    Serbian (Србија)
    In Serbian every O is pronounced as O, the same goes for E.
    There was one video from Bosnia that was popular 5 or 6 years ago when a girl says 'kisla mi je gloooooooova' (instead of glava). It was really funny.
    I remember when we started learning Russian in high school and the teacher said DamadEdava and wrote Domodedovo. We were like WTF, why are you writing four Os?
     
    Now a question: are there in Belorussian cases of unwritten akanye, that is, not reflected in spelling?
    Otherwise said: if all akanye is reflected in spelling, it would in practice eliminate every possibility of writing unstressed "o", since all of them would have been written as "a". A consequence would be that every written "o" by need is stressed.
    Another consequence would be that the poll of options to inflect words would be much lesser, since even in writing there would be no possibility of changing the ending, for example, from unstressed "o" to unstressed "a".
    Since I have nothing but a rudimentar knowledge on Russian, I wonder how actually works the written akanye in Belorussian.
    The Belarusian orthography works exactly as you have written (except in some loanwords). See Marchant C · 2004 · Fundamentals of modern BelarusianMarchant C · 2004 · Fundamentals of modern Belarusian.pdf The only thing that every foreigner tends to forget when discussing in the Internet the consequences of the Russian and Belarusian vowel reduction is that East Slavic has many word forms with stressed endings.
     
    And, not related with the last post, if it is of any interest, Old Church Slavonic, theoretically and so far I know, would suffer no vowel reduction, nor akanye, ekanye or ikanye. However, in many (Russian) recordings of divine Liturgy, it is possible to hear clearly the akanye, mostly from the priests who sing alone the ektenias and exclamations. Any decent choir will have to work to pronounce homogeneously the vowels, consciously, and syllables in general are produced one-by-one, with regular duration (all of this avoids or strongly reduces akanye), while the recitative used by priests favor a more spontaneous diction, and then akanye often appears unhindered, though probably unwillingly or unaware of itself. Probably some would perceive akanye, in this case, as an error.
    Yes that's true. On the other hand, liturgical languages like Latin or Greek or Church Slavonic are inevitable victims of such modernizations and accents. As a person familiar with languages, I feel pain when hearing Church Slavonic or Latin liturgy or, actually, people singing in foreign languages, be it opera or pop music. And, to be honest, Church Slavonic itself is the result of rather massive phonetic russification (fate of the yers and nasal vowels, palatalization etc.), so akan'ye and ikan'ye is just one more vice.
     
    Last edited:
    I remember when we started learning Russian in high school and the teacher said DamadEdava and wrote Domodedovo. We were like WTF, why are you writing four Os?
    I had a contrary experience when our French teacher jokingly pronounced the surnames Protopópov and Bogomólov with a French accent — with the end stress and emphasizing all the o's. Honestly, I think that a and i are much nicer sounding vowels than the general Slavic o and e (whose abundance is one of the main reasons why the Slavic languages sound so dull) and if not the harmful homophony it causes, this shift could be welcomed.
     

    metaphrastes

    Senior Member
    Portuguese - Portugal
    The only thing that every foreigner tends to forget when discussing in the Internet the consequences of the Russian and Belarusian vowel reduction is that East Slavic has many word forms with stressed endings.
    Thanks for the answer. Would you say that the stress strongly tends to shift to the ending whenever it is needed to (audibly) distinguish between two inflections of the same word? I had this impression about a few cases in Russian, but Russian stress is so varied and capricious that I cannot say I have some kind of panoramic overview on the matter.
     
    Thanks for the answer. Would you say that the stress strongly tends to shift to the ending whenever it is needed to (audibly) distinguish between two inflections of the same word? I had this impression about a few cases in Russian, but Russian stress is so varied and capricious that I cannot say I have some kind of panoramic overview on the matter.
    No, the current stress position is partly inherited (the choice of stress position in Proto-Slavic was automatic and depended on the interplay of prosodic properties of the vowels in the word or combination of words) and partly modified by processes occurring within the paradigms and word-formational patterns, but it doesn't have the purpose to compensate the damage caused by the vowel reduction or the older homonymy (the same processes happen in Ukrainian and Northern Russian dialects where there is no vowel reduction). For example, Old Russian didn't distinguish between the Gen. Sg. stáda "of the herd" and Nom./Acc. Pl. stáda "herds", whereas the modern language opposes stáda for the first and stadá for the second, but this has developed as an extension of the opposition found in other words and is probably meant to contrast the stress of the Sg. and Pl. in general, not in these particular homonymic forms. There are also instances of stress shift from the ending, e. g. while standard Russian and eastern Russian dialects preserve the end-stress in many Past Active Participles (fem. pʲerʲevʲedʲená, neut. pʲerʲevʲedʲenó, Pl. pʲerʲevʲedʲený "transferred, translated"), Ukrainian and Belarusian (and western Russian dialects) move the stress to the syllable before the Participle suffix: it is not harmful in Ukrainian (perevédena, perevédeno, perevédenı), but confuses the feminine and neuter in western Russian dialects (fem. and neut. pʲerʲevʲédʲena) and in Belarusian (fem. and neut. pʲaravʲéʣʲanaja).
     

    DarkChild

    Senior Member
    Bulgarian
    Yes that's true. On the other hand, liturgical languages like Latin or Greek or Church Slavonic are inevitable victims of such modernizations and accents. As a person familiar with languages, I feel pain when hearing Church Slavonic or Latin liturgy or, actually, people singing in foreign languages, be it opera or pop music. And, to be honest, Church Slavonic itself is the result of rather massive phonetic russification (fate of the yers and nasal vowels, palatalization etc.), so akan'ye and ikan'ye is just one more vice.
    Russian accent is quite difficult to hide. In opera it is quite obvious. Especially the typical Russian stressed O which to me sounds like уо.
    I had a contrary experience when our French teacher jokingly pronounced the surnames Protopópov and Bogomólov with a French accent — with the end stress and emphasizing all the o's. Honestly, I think that a and i are much nicer sounding vowels than the general Slavic o and e (whose abundance is one of the main reasons why the Slavic languages sound so dull) and if not the harmful homophony it causes, this shift could be welcomed.
    First time I ever hear of such abundance. Well, Russian has lots of a's.
     

    Awwal12

    Senior Member
    Russian
    Perhaps not a topic of this thread, but one of the two prevalent theories relates akan'ye to the substrate influence of the Baltic-speaking population of the middle latitudes of pre-Slavic Eastern Europe, compare Russian dialects - Wikipedia (+ Belarusian to the west) with Balts - Wikipedia (presumably Baltic-speaking archeological cultures in violet). If it is true, akan'ye in other Slavic areas should have a different origin
    Sorry to disappoint you, but I have to point out several important things.
    1. Russian akanye is not just "a instead of o" or "a instead of o in certain positions". It is, at least in the general sense, a much more fundamental thing: an indistinction between all non-close vowels in all non-stressed syllables. Note that by that definition it also must include (and always does include!) some form of such indistinction after soft consonants as well; in the standard language and many Middle Russian dialects it's ikanye and/or yekanye (the latter being a standard at least for the St.Petersburg pronunciation), in most South Russian dialects it's yakanye.
    2. There is pretty strong evidence that akanye originated in the south-eastern area of the Old Russian language around the early XIV century. In more western areas (including the Grand Principality of Lithuania) it's attested later (the XV century at least) and, more importantly, not in any of its archaic forms.
    3. There were many (quite natural) attempts to explain akanye as a result of some non-Slavic substrate. The problem is, none attested language group in the area could simply produce such results (that is, archaic forms of akanye) after being combined with the Old Russian phonology. Note that another feature of the archaic variants of akanye is at least some form of dissimilation by closeness between the vowels in the syllables, and that's hardly a coincidence.

    The similar (but fundamentally different) phenomena outside of the East Slavic area are very likely unrelated in origin.
     
    The substrate theory is not the one I find fully convincing: as I have written, it fails to explain the merger of ъ>о with a (at least, we should have possessed some intermediate stages when unstressed o and a had already merged while ъ>o still didn't, which are absent as far as I can judge). Yet, the more open pronunciation characteristic of the Baltic speech may have triggered the opening of unstressed o (and e), which, as this thread shows, is so unusual in the Slavic context and seems to be absent or occasional in the areas where Balts were not attested (including south-western and north-eastern East Slavic).

    How many non-close vowels in non-stressed syllables are there? Leaving aside the dialects that distinguish between open and closed e and o (and I can't recall right now whether in case of o this distinction ever occurs in unstressed syllables), there are actually three: a, o and e, and, since the development after hard and soft consonants isn't always parallel (it isn't in Standard Russian, for example), when speaking about akan'ye we only deal with two underlying phonemes, a and o.

    I don't agree with the term "indistinction" since nowhere in Russian and Belarusian the unstressed a becomes labialized: what happens is always the loss of labialization and opening of o, not the opposite way. Hence the very term "akan'ye".

    I am aware of the existence of various flavors of akan'ye (e. g. I wrote about cases when it depends on the neighbor vowels and when it doesn't), but we actually don't know which is archaic and which is not: this terminology depends on the theories the scholar follows. The nuanced akan'ye is not necessarily the original one: it may have perfectly been the result of interference of straightforward akan'ye from a prestigious dialect with the local way of speaking (e. g. many Ukrainians and Lithuanians who are not very fluent in Russian may pronounce the unstressed Russian a differently from the unstressed o: it is a rather clean a in the former case and a more closed and labialized vowel in the latter: not exactly the case of dissimilative akan'ye, but a typologically similar situation).

    The pretty strong evidence you mention is (again as I have written) the appearance of akan'ye in the texts. The question is (see #8), whether these texts reflect the very emergence of this phenomenon or something else. I have no opinion: we simply don't know and currently have no ways to evaluate this. I am waiting for the discovery of birch bark manuscripts in the future akan'ye areas (there are some from ancient Polotsk, if I am not mistaken, but I have never seen them published and commented).

    The substrate seldom transfers the phonetics of the original population to the new language: it may be so in the first generation, but with time the continuous influence of the prestigious clean speech leads to a compromise when only some aspects of the substrate pronunciation remain (e. g. if Russian in Ukraine stops being influenced by the metropolitan speech, it will remain Russian but with an obvious substrate accent, which is the result of interference of both languages and not the clean Ukrainian pronunciation; if at the same time Ukrainian disappears and this Ukrainian Russian remains the only language in Ukraine, it will simply continue to evolve with these h's, sing-song intonations, not quite open unstressed a/o's etc., nothing of which has roots in the standard Russian phonetics).

    Yet, do we have similar (if even unrelated) phenomena outside East Slavic or not? We have read here that the Bulgarian akan'ye is different. The literature also mentions akan'ye in Slovene dialects, what about them?

    P. S. How would you explain the faster rate of the spread of akan'ye comparing with ikan'ye (i. e. the unstressed a and o have merged many centuries ago, whereas in case of ʲa/ʲe and ʲi it is still not completed)?
     
    Last edited:

    Awwal12

    Senior Member
    Russian
    Leaving aside the dialects that distinguish between open and closed e and o
    Why should we leave them aside? By now they're exotic indeed (and are almost exclusively North Russian). But judging on the patterns of dissimilation in archaic types of akanye, such contrasts were typical for the time and place where akanye has originated. And the very Moscow speech consistently retained the contrast between yest' and yat' till the XVIII century (not in all positions, of course, but still).
    I am aware of the existence of various flavors of akan'ye (e. g. I wrote about cases when it depends on the neighbor vowels and when it doesn't), but we actually don't know which is archaic and which is not
    We don't know anything for sure, but we have pretty many reasons for such assumptions. The processes which take place in idioms may be unpredictable yet they are anything but chaotic. Say, if the Belarusian akanye (we speak about akanye in the most broad sense here) is attested later, is basically identical to akanye of the Kursk region, the latter can be produced by simplifying more complex patterns of akanye and the cluster of such more complex patterns lies further eastwards, it already looks like a pretty simple crime scene to me. Other arguments cover the possible development of various types of yakanye, etc.
    The question is (see #8), whether these texts reflect the very emergence of this phenomenon or something else.
    They reflect the presence of the phenomenon (other factors being filtered out, of course). No more, no less.
    But regarding the "falling level of education" I must point out that even if it's possible (albeit unlikely) that scribes in some areas made no mistakes (in regard to some local system of writing - since the written language wasn't globally standartized anyway and that is obvious), it's very hard to believe that they made a lot of mistakes in other places but were too well educated to make some precise possible mistake. And we may not have too many Old Russian texts (or rather a not too representative set of them) , but those we have undergo all applicable statistical analyses.
     
    To illustrate the substrate theory, let's recall the Slavic loans to Baltic-Finnic.

    Slavs penetrated to the modern north-west of Russia (Pskov, Novgorod and Leningrad oblast's) after the 5th century, probably in two waves: first came Slovenes, then Krivichians. In the 11th century, Slovenes lived in this area approximately east of the Volkhov river and Krivichians west of it. Pskov was inhabited by Krivichians; Novgorod by both tribes. The contacts with Finnics were extensive, e. g. Rus from overseas were invited by чюдь, словѣне and кривичи; to capture Smolensk and Kiev, Oleg took варѧгы, чюдь, словѣни, мѣрю, весь and кривичи (in this order according to the "Primary Chronicle").

    Baltic Finnic languages have initial stress and clearly distinguish a and o in both stressed and unstressed syllables: while the unstressed o is a much newer sound than the unstressed a (which goes back to Common Uralic), it already existed in the second half of the first millennium.

    The earliest layer of Slavic loans to Baltic-Finnic can be illustrated by the following examples (Finnic substitutes voiceless stops with long voiceless and voiced stops with short voiceless):
    • Finnish pappi — Old Novrogod попе (-e is the Krivichian trait, elsewhere we find -ъ, попъ; Finnic didn't possess Nominative Singular on -e, hence -i)
    • Votic kaputtaкопыто (the unstressed long vowels didn't exist in Baltic Finnic at that period, they emerged later after the fall of intervocalic consonants and monophthongization of the diphthongs, hence not the expected **kapuutta; p instead of the expected **pp is the result of the Baltic-Finnic lenition of stops in the originally closed syllables, cp. the Illative Sg. kaputassa, where tt>t after the addition of the Illative suffix -ssa)
    • Finnish palttinaполотьно (notice the lack of pleophony)
    • Finnish talkkunaтолoкъно
    • Finnish akkunaокъно
    • Finnish talttaдолото
    • Finnish varp-unenвороб-ии
    • Finnish raamattu — грамота (notice the preserved long vowel in the first syllable; -u is mysterious, perhaps borrowed in the Accusative form? the attested 11th century daju ti gramotu śu < the 10th century *dājū tī grāmatū sū)
    • Finnish pakana — поганъ
    • Finnish piirakkaпирогъ (again the Slavic long vowel; -kka is a substitution with the widespread Finnish suffix, otherwise we would expect **piiraa<*piiraka)
    • Finnish pirta — бьрдо
    • Finnish vapaa — свободь (the long vowel from the lenited intervocalic *t)
    • Finnish tapparaтопоръ
    • Estonian sahkсоха
    • Izhorian kassaкоса
    • Izhorian koominaгѹмьно
    • Finnish dialectal siivatta — животъ (again, the etymologically long vowel in the first syllable is preserved)
    • Finnish saapa-sсапо-гъ (the long a is preserved; pp>p in the closed syllable, cp. the Gen. Sg. saappaan<*saappasen where it reappears)
    (also tuska — тъска, lusikka — лъжька, risti — крьсте, sirppi — сьрпе, turku — търгъ (the u-stem, cp. Lithuanian turgus), läävä — хлѣвъ, määrä — мѣра, kuoseli — кужель [*ō>uo is a later Finnish development], niitti — нить, Votic/Izhorian värttänä — веретено, Votic koontala — кѹдѣль, lookka — лѹка, Izhorian kooma — кѹмъ: notice the stable preservation of the Common Slavic long vowels in the first syllable; the original Slavic *i and *u; the Common Slavic *au̯ is still *ō, not the later u; *an is also borrowed at the stage when it was heard as ō [lookka]; is an open vowel as in Krivichian, Lechitic and Bulgarian; -u is different from -i/-a in the thematic declension).

    All this, as well as the evidence from loans from and to other neighbor languages, from Norse to Greek, suggests that the shift *a>o occurred in Slavic very late, during the last centuries of the 1st millennium. In particular, notice the Christian words raamattu, pappi and pakana. If in the 10th century the newly introduced Christian terminology was still heard by Finnics with a, and the first signs of akan'ye appear in the texts of the 14th century, i. e. only four centuries later (and in the areas from where we have simply no old texts and thus know nothing about the actual pronunciation of this vowel in 10–13th centuries), this makes the theory explaining the merger of unstressed a and o as the result of the incomplete shift *a>o (caused either by the substrate or by something else) at least worth being considered.
     
    Last edited:
    Why should we leave them aside? By now they're exotic indeed (and are almost exclusively North Russian). But judging on the patterns of dissimilation in archaic types of akanye, such contrasts were typical for the time and place where akanye has originated. And the very Moscow speech consistently retained the contrast between yest' and yat' till the XVIII century (not in all positions, of course, but still).
    But since ǒ never existed in unstressed syllables (the distinction between o and ǒ arose only under the stress), we have at most four non-close unstressed vowels: a (dalá), o (volá), e (vʲelá) and ě (dʲělá), of which anyway only two merge as the result of akan'ye: a and o.

    We don't know anything for sure, but we have pretty many reasons for such assumptions. The processes which take place in idioms may be unpredictable yet they are anything but chaotic. Say, if the Belarusian akanye (we speak about akanye in the most broad sense here) is attested later, is basically identical to akanye of the Kursk region, the latter can be produced by simplifying more complex patterns of akanye and the cluster of such more complex patterns lies further eastwards, it already looks like a pretty simple crime scene to me. Other arguments cover the possible development of various types of yakanye, etc.
    I started to write a reply, but thought that perhaps I don't understand something crucial, so could you explain me how the existence of various types of akan'ye suggests that one of them is older and how it is relevant to the origin of akan'ye in general? As far as I understand, the unstressed a and o in any case merge in the same vowel and are never distinguished: when the dissimilative akan'ye creates variously open unstressed vowels, outcomes of a and o are always the same (e. g. in трава/сова, травы/совы, травѣ/совѣ, траву/сову, травою/совою). So, the question is how this has developed?

    Also, again, I am not convinced that akan'ye and yakan'ye/yekan'ye/ikan'ye are two necessarily parallel processes, since, as I have written, I, in 2016, pronounce се́но and се́на in exactly the same way, while до́ля, до́ле and до́ли in three different ways: I would expect at least долꙗ and долѣ to have merged back in the same 14th century (OK, ě was a higher vowel, then consider the merger of полѥ and полꙗ that still hasn't happened in St. Petersburg as of November 6, 2016, i. e. 650 years later). From what I know about how languages evolve, I conclude that akan'ye is older.

    They reflect the presence of the phenomenon (other factors being filtered out, of course). No more, no less.
    But regarding the "falling level of education" I must point out that even if it's possible (albeit unlikely) that scribes in some areas made no mistakes (in regard to some local system of writing - since the written language wasn't globally standartized anyway and that is obvious), it's very hard to believe that they made a lot of mistakes in other places but were too well educated to make some precise possible mistake. And we may not have too many Old Russian texts (or rather a not too representative set of them) , but those we have undergo all applicable statistical analyses.
    Зализняк АА · 1987 · О языковой ситуации в древнем Новгороде: 117 writes the following:
    Что касается стандартного древнерусского, то, как это ни парадоксально, для раннего периода (XI – рубеж XII и XIII вв.), мы располагаем здесь гораздо меньшим объёмом материала, дошедшего до нас в подлинниках, чем для древненовгородского диалекта. Если не считать совсем мелких надписей и временно оставить в стороне вопрос об отражении данной формы языка в части новгородских берестяных грамот, мы можем отнести сюда только Мстиславову грамоту (ок. 1130 г., 156 слов) и запись о покупке Бояновой земли на стене Киевского Софийского собора (XII в., 64 слова). Таким образом, мы фактически знакомы с данной формой древнерусского языка в основном по памятникам XIII–XV вв. — подлинникам или спискам с документов более раннего периода. Соответственно, ниже необходимо учитывать, что для XI–XII вв. наши сведения о стандартном древнерусском в значительной степени носят характер реконструкции, тогда как сведения о древненовгородском диалекте этой эпохи получены более непосредственно.

    Thus, the Muscovite texts of the 14th century, where akan'ye first appears (Moscow Gospel of 1340 etc.) seem to be simply the earliest preserved documents written in these areas. Do we have Church Slavonic texts from the belt Grodno — Moscow older than the 14th century? How many vernacular traits do they have?
     
    Last edited:
    For the Slovene phenomenon enigmatically called akanje see Lencek RL · 1982 · The structure and history of the Slovene language: 149 (Lencek RL · 1982 · The structure and history of the Slovene language.pdf). Here akan'ye is mentioned for the Lower Carniolan (Dolenjskan) dialect. There seem to be three major differences from the East Slavic situation:
    • this change affects short unstressed and stressed vowels;
    • when some dialects change o>a (otrok>atrȁk), their neighbors may change o>u (oko>ukȗ);
    • the degree of perturbations that characterizes the Slovene vocalism (including prosody) very far exceeds everything occurring in the akan'ye dialects of East Slavic.
    Дуличенко АД · 1998 · Словенский язык: 88 mentions akan'ye also for the Upper Carniolan (Gorenjskan) dialect: here the shift o>a occurs in posttonic closed syllables (starost>stárast).
     

    Awwal12

    Senior Member
    Russian
    Regarding the (hypothetical) phonetic quality of the Old Russian /o/ - I just don't see how relevant it is. The whole matter is mostly not about actual phonetic quality but about contrasting certain phonemes in certain positions. And with all the sources we have, we are positively sure that:
    - the Old Russian dialects, till the XII century at least, had from 8 to 10 vowel phonemes (if I remember correctly, some scholars speculate that Old Novgorodian might have originally had 7);
    - they contrasted all full-fledged vowels in all positions for the said period (with the reduced vowels the situation may be more complex); whatever sound /o/ actually represented, it was never mixed with /a/.
    Зализняк АА · 1987
    But since the 1987 the situation somewhat changed. The number of written records is slowly increasing (Old Novgorodian is also not an exception, of course). For instance, in the 1988, 2007 and 2015 four birch bark manuscripts were found in Moscow alone (two short fragments, a register and a business letter).
    By the way, you shouldn't discard the Church Slavonic texts from consideration. Although they may very poorly reflect the actual state of the spoken Old Russian dialects, the scribes made mistakes often enough to make it clear which phonemes they had not in their speech, for example (the opposite is not true, mind you), or which phonemes merged in certain positions.
    But since ǒ never existed in unstressed syllables (the distinction between o and ǒ arose only under the stress), we have at most four non-close unstressed vowels: a (dalá), o (volá), e (vʲelá) and ě (dʲělá), of which anyway only two merge as the result of akan'ye: a and o.
    Ahem.
    After the hard consonants in unstressed syllables 2 close (/u/, /y/) and 2 non-close (/a/, /o/) phonemes were possible. The latter two get unavoidably merged, possibly together with some of the close phonemes (and in the weakest positions the total merger is not uncommon). The maximal number of distinct vowel phonemes after the hard consonants with akanye is, to put it simple, 3-3-4-3, and the minimal is 1-2-4-1.
    But you forget that akanye never limits itself with vowels after the hard consonants.
    After the soft ones, 2 close (/i/, /u/) and 4 non-close (/a/, /e/, /ě/, /o/) phonemes are/were possible, and all the four non-close phonemes get unavoidably merged as well, with various results; in case of ikanye, they are altogether merged with /i/ in addition.

    Sorry, I just had no time to reply to everything. :)
     
    Last edited:
    We seem to be discussing different things. I understand that akan'ye in its present form is part of the broader set of phenomena, but what was the starting point of this process? How did it emerge? Was akan'ye the first shift that triggered the vowel indistinction and later reduction, or did several mergers originate at the same time? Why do unstressed front vowels merge slower (standard Russian and standard Bulgarian)?

    I am still unaware of any works presenting and commenting the attestations of the vernacular speech of the future akan'ye areas. What do these Moscow birch bark manuscripts tell us about the pronunciation of the unstressed o before the 14th century? (By the way, some scholars believe that middle Russian dialects, to which the Muscovite dialect belongs, acquired akan'ye later, so strictly speaking we should evaluate vernacular texts from South Russian and Belarusian lands).

    The phonetic quality of the early Old East Slavic o is important in the context mentioned in #8. Slavic in the turn of the 1st and 2nd millennia experienced the shortening of some etymological lengths (it happened differently in different dialects, hence it must have occurred after the split of the common language), and the unstressed acute mostly became short (e. g. Czech dělat(i); the old length in Czech was preserved in unstressed non-acute non-final syllables, e. g. dělán(ъ); the new length immediately emerged from the contraction, e. g. dělá). If the unstressed Common Slavic *a in some areas failed to become o to the 11th century (e. g. *sḗna failed to become sěno), the shortening unstressed simply merged with this old sound (e. g. when the Gen. Sg. *sḗnā became sěna). This merger must have created great pressure to the vocalism since it produced a considerable number of homonymic forms and therefore may have triggered further changes towards merger and reduction. This scenario can explain two additional moments: (1) that akan'ye (for the reasons outlined twice in previous posts) seems to be older than the phenomena involving the front vowels and (2) that akan'ye is uniform (it begins with that the unstressed o loses labialization), whereas yakan'ye vs. ikan'ye develop into two contrary directions: e either may lower towards a or may close towards i.

    P. S. The Common Slavic *i normally developed to ь and further either disappeared or became e in modern Russian, however, word-initially (and occasionally word-internally) after j it retained its quality and eventually merged with the outcome of the Common Slavic *ī, e. g. *ikrā>jĭkrā>икра (Lithuanian ikra) vs. *īskātī>jiskati>искать (Lithuanian ieškoti) or *ājiķ->jajĭcь>яиц vs. *wajīn->vojinъ>воин. Since the occurrence of the opposition *ji/jī was minimal, this merger didn't cause any phonological consequences.
     
    Last edited:
    One more note concerning the Common Slavic *a. In every manual we can read that the Common Slavic groups *CoRC (consonant+o+r/l+consonant) metathesized in South Slavic, Czech and Slovak to CRaC. There are two problems with this interpretation. First, the foreign sources seem to lack examples with oR. In contrast, e. g. in Byzantine chronicles we find Slavic names like Βαλδίμερ (>Vladimirъ~Vladiměrъ~Vladimerъ) and Δαργαμηρός (>Dragomirъ). Second, how could the metathesis of *oR have produced Ra, e. g. *korwa > Czech kráva? [Á here was acute and hence preserved its length; in zlato it was non-acute and hence shortened; in Serbo-Croatian the distribution is contrary: the acute lengths shortened, krȁva, and the non-acute ones persisted: zlȃto]. A couple of centuries later the falling yers lengthened o into ō, as one can expect, cp. stolъ > Czech stůl, Polish stół: why then did this *o lengthen into ā during the metathesis? It is more parsimonious to suppose that this metathesis occurred when Slavs still had *a in the place of the later o, and that the difference between Lechitic on the one side and future Czech, Slovak and South Slavic on the other is that the metathesized vowel remained short in the former (*karwā>*krawā>krowa) but simply lengthened in the latter (*karwā>*krāwā>kráva), and the development *a>o occurred immediately after.
     

    arn00b

    Senior Member
    English
    Thanks, @ahvalj for this last post.

    Let me try to understand this correctly, do you propose an intermediate step between Proto-Balto-Slavic (whatever the form may be) and Proto-Slavic's *kőrva, a form which had *a of some type?

    Or have I misunderstood?

    (As a side note: Your explanations are clear, it just goes a little over my head sometimes)
     
    I am not original in this respect, simply authors of the manuals usually read only other manuals, so such hypotheses become known to the wide audience with difficulty. The idea is even more radical — that the so-called "Proto-Slavic" *korva simply never existed: until approximately the 8th century CE it was *karwā (cp. the above Finnic loans from early East Slavic without metathesis or pleophony like palttina, talkkuna, taltta, varp-unen from *paltina, *talkuna, *dalta, *warb-iji attested 2–3 centuries later as polotьno, tolokъno, doloto, vorob-ьjь), which then independently developed to *krawā>krowa in Lechitic (Polish and Cassubian), *krāwā in future Czech, Slovak and in South Slavic and *karəwā>*korŏwa in East Slavic (the second o was not identical to the first one since it often didn't lengthen in Ukrainian, cp. stolъ>стіл vs. wolosъ>волос). The very idea that Common Slavic had *o/*or was never based on any evidence: the forms reconstructed since the end of the 19th century as "Proto-Slavic" are simply Old Church Slavonic words rolled back to one evolutionary step when other languages show divergent forms (e. g. strana/strona/storona according to this logic must go back to *storna). However, the evidence of loanwords from and to prehistoric Slavic suggests that o, ъ, ь and ɨ were new sounds that emerged right before the first Slavic written records, and, 1–3 centuries before, these vowels had the shape a, u, i and ū, or at least they were reasonably close to these sounds for foreigners to perceive them as such.

    The evidence and the discussion can be found e. g. in Shevelov GY · 1964 · A prehistory of Slavic∶ the historical phonology of Common SlavicShevelov GY · 1964 · A prehistory of Slavic∶ the historical phonology of Common Slavic.pdf and Тохтасьев СР · 1998 · Древнейшие свидетельства славянского языка на БалканахТохтасьев СР · 1998 · Древнейшие свидетельства славянского языка на Балканах.pdf

    The above Lencek RL · 1982 · The structure and history of the Slovene languageLencek RL · 1982 · The structure and history of the Slovene language.pdf mentions Wallucus~Walducus (p. 39 — Valuk (duke) - Wikipedia) possibly reflecting the Common Slavic *waldūkā (> OCS vladyka) in the "Chronicle of Fredegar" (7th century) and Dabramuzli (p. 80; in Bavarian records at around 750), Common Slavic *Dabramūsli, later known as Dobromyslь.
     
    Last edited:

    arn00b

    Senior Member
    English
    Thank you so much for this, @ahvalj. I'm thoroughly convinced. I wonder why the internet, from serious to casual sources still persist with the *o based etymologies.
     
    Thank you so much for this, @ahvalj. I'm thoroughly convinced. I wonder why the internet, from serious to casual sources still persist with the *o based etymologies.
    That persists not only in the Internet, the academic literature follows the old conventions as well. I think conventionality is the answer — in most cases the user isn't interested in how the Slavic words looked like in, say, the 7th century: people need a convenient citation form able to explain the Slavic words a person will be dealing with. In this respect gostь, directly attested in Old Church Slavonic and Old East Slavic and allowing to deduce forms of later languages, is more informative than something like *gasti. Only discovery of actual pre-Cyrillian Slavic texts may change that. I would compare it with the Indo-European notation: almost everybody in the last 70 years agreed that *bʰ, *dʰ, *gʰ, *gʲʰ and *gʷʰ were not voiced aspirated stops, yet this 19th century convention is both familiar and allowing to predict the shape of the outcomes.

    P. S. Two amendments to the list of Finnic loans. Along with piirakka "cake" (with the analogical -kka), Finnish has the word piiras, which has the form piiraa- in most Singular oblique cases (e. g. Gen. Sg. piiraan). Contrary to Wiktionary's opinion (piiras - Wiktionary), this piiraa- is, as I had written, the regular descendant of the expected Finnic *piiraka- borrowed from the early East Slavic *pīragV-. Since -s-, -t- and -k- in trisyllabic and longer words lenited and disappeared between vowels in Finnish and most other Baltic-Finnic languages, the type with the Nom. -s embraced many old words of all the three stem types. This also explains saapas (Gen. Sg. saappaan) "boot" from early East Slavic *sāpagV-.

    P. P. S. And despite all these kilobytes of posts, I still don't think that akan'ye is directly related to the preservation of Common Slavic *a in unstressed syllables: I simply tried to illustrate the arguments of the supporters of that theory. The second predominant theory that *a>o (поставити) and ъ>o (съставити) first merged in o (поставити, составити) and only then in unstressed position merged with a (паставить, саставить, заставить) seems to be more justified by the textual and dialectal evidence than the assumption that *паставити and заставити merged first, and then съставити evolved through составити (not attested in akan'ye dialects) to the modern саставить.
     
    Last edited:

    Zec

    Senior Member
    Croatian
    To contribute something to the original post:

    Russian-style vowel reduction does exist in some dialects of Croatian, especially the so called prigorski dialect of the kajkavian macrodialect, which I can describe pretty decently since, while not being an active speaker, I live there.

    There, the stressed vowel system consists of six or seven vowels vowels: i, e, ɛ, (ə), a, o, u, some of which merge in unstressed syllables. We may distinguish ikanye akanye and ukanye. Regularly, the schwa merges with a (which, paradoxically, means it is found only in stressed syllables!), and closed e merges with i (ikanye). Unstressed o can, depending on the dialect, merge either with u (ukanye), which is more common, or with a (akanye), which is rarer. The result is always a system of four unstressed vowels: i, ɛ, a, u, which are all lax and somewhat centralized, that is, they are also phonetically reduced.

    This system is found at it's fullest extend only among the elderly. Among the young, only ikanye remains, and unstressed vowels are not phonetically reduced.

    As in Russian, mobile accent can cause vowel alternations within paradigms.

    Concerning other dialects, I don't know of any čakavian or štokavian dialect with vowel reduction, though there are some that wildly syncopate unstressed i's and u's.
     
    The discussions of the origin of akan'ye usually imply that okan'ye dialects oppose o and a. However, it seems to be not always so.

    The book Русская диалектология · 1972: 97–98 writes that [Russian] okan'ye dialects clearly split into three groups: (1) proper North Russian with the very open unstressed [oᵃ] and two intermediate (Middle Russian), namely (2) Novgorodian with a clear [oᵒ] and (3) Vladimir-Volgan with a very close [oᵘ] and the reduction in the first pretonic syllable. The book considers (1) to be the oldest type of okan'ye as it reflects the Old Russian pronunciation of o with its indistinct labial articulation.

    The same can be read in Колесов ВВ, Ивашко ЛА, Капорулина ЛВ, … · 1990 · Русская диалектология: 62 — "very open [oᵃ] (resembling [a])" and "the oldest type […] with indistinct labial articulation of the vowel".

    This division is also mentioned in Аванесов РИ · 1949 · Очерки русской диалектологии. Часть первая: 63–64, "open shades of o, very weakly labialized, i. e. sounds that can be marked with the sign oᵃ" and "this pronunciation is more characteristic of the dialects of the north-west, e. g. on the territory of the Karelo-Finnish republic".

    If I understand correctly, it turns out that in the proper North Russian dialects okan'ye means the opposition not of o : a but of something like å ["very open", "very weakly labialized", "resembling a"] : a, i. e. the condition I considered extinct and only reconstructible from old loanwords. If it is indeed ancient, it seems to be supportive of the hypothesis that the East Slavic akan'ye arose not as the secondary opening of the unstressed o, but as merger of either (I) directly the Common Slavic *a with the shortened ā (discussed in previous posts) or (II) later rather weakly distinct å and a (based on the North Russian situation presented here). This latter scenario may explain the absence of o>a in the oldest texts: if the scribe distinguished å and a in the unstressed syllables, he would not confuse the letters o and a (cp. the American dot and dart).

    In any case I wonder what was the evolution of the unstressed ъ>о since in North Russian dialects of the abovementioned type it is also pronounced very open, as any modern o.
     
    Last edited:

    Awwal12

    Senior Member
    Russian
    If I understand correctly, it turns out that in the proper North Russian dialects okan'ye means the opposition not of o : a but of something like å ["very open", "very weakly labialized"] : a
    .
    I still cannot get why you pay so much attention to the actual phonetic quality of these Old Russian phonemes. From phonological point of view, the only relevant thing is the fact that they are opposed (or not) and, less importantly, what are their distinctive features (closeness, labialization, etc.). Obviously enough, the sounds representing certain phonemes may vary greatly from dialect to dialect. /ě/ and /ô/ have at least two possible realizations each, realization of /o/ also may vary (also depending on the actual position), etc - and you cannot be sure it wasn't the same in the Old Russian language already.

    What is actually the relevant difference between "merging of poorly distinct å and a (surely in unstressed positions only as well?)" and "secondary opening of the unstressed o"? To merge, previously distinct å and a in unstressed positions have to change their articulation anyway - at least one of them, and likely it was "å". It essentially looks like speaking about the same thing, just with minor disagreements about the actual quality of the Old Russian /o/ (mind you, /o/ is nothing but a conventional sign anyway).
     
    We look at the situation from the perspective of our different interests: I from the history of language, you (as far as I can tell) from phonology. I am trying to understand why akan'ye in Russian/Belarusian is different from the similar phenomena in other Slavic languages mentioned in this thread — in particular, (1) why Russian doesn't know the merger of o with u (in Slovene, Serbo-Croatian and Bulgarian it occurs in parallel with o>a, in neighbor dialects, though always more often), which, by the way, is grammatically safer since it creates much less homonymy, and (2) why changes after soft vowels are not always parallel (e. g. akan'ye vs. ikan'ye). I understand that in terms of phonological oppositions it is irrelevant which pairs of vowels merge to form a simpler unstressed vocalism, o+a or o+u, and whether the unstressed ʲe develops towards ʲa or ʲi, but it is intriguing to me as a person interested in the "flesh" of the language.

    The first question can be solved if indeed the unstressed North Russian å continues the originally open pronunciation: the unstressed o across much of the Russian territory then simply would have never reached the proper fully labialized o stage and thus couldn't develop further towards u.
     
    Last edited:

    Hachi25

    Member
    Serbo-Croatian
    Russian doesn't know the merger of o with u (in Slovene, Serbo-Croatian and Bulgarian it occurs in parallel with o>a, in neighbor dialects, though always more often)

    Can you give me an example of those mentioned mergers in Serbo-Croatian?
     
    Can you give me an example of those mentioned mergers in Serbo-Croatian?
    Russian-style vowel reduction does exist in some dialects of Croatian, especially the so called prigorski dialect of the kajkavian macrodialect, which I can describe pretty decently since, while not being an active speaker, I live there.

    There, the stressed vowel system consists of six or seven vowels vowels: i, e, ɛ, (ə), a, o, u, some of which merge in unstressed syllables. We may distinguish ikanye akanye and ukanye. Regularly, the schwa merges with a (which, paradoxically, means it is found only in stressed syllables!), and closed e merges with i (ikanye). Unstressed o can, depending on the dialect, merge either with u (ukanye), which is more common, or with a (akanye), which is rarer. The result is always a system of four unstressed vowels: i, ɛ, a, u, which are all lax and somewhat centralized, that is, they are also phonetically reduced.

    This system is found at it's fullest extend only among the elderly. Among the young, only ikanye remains, and unstressed vowels are not phonetically reduced.
     

    Hachi25

    Member
    Serbo-Croatian
    That's not an example, but let's say I got the point because I expected to see something connected to the Shtokavian variant. I missed that part you quoted because when you just say 'Serbo-Croatian', I don't really have our dialects in mind, probably because Kajkavian, Chakavian and Shtokavian very often don't even look like they belong to the same language.

    I can't (and shouldn't) really debate about the prigorski dijalekt without knowing which particular town/village the author described here. Those things I know is that it shouldn't have the ə-sound and that unstressed o doesn't merge with a (and can merge with u, but only if it was a closed o before). However, it's impossible to make a generalization here as almost every populated place speaks their own variant of the dialect, and something like a standardized variant of it doesn't exist.
     
    Top