Pronunciation of Consonants from the Perspective of the Modern Russian Pronunciation Norm

If we are to consider the difficulties experienced by people when pronouncing consonants, we should first of all discuss what problems arise in the course of phonetic realization of voiced and voiceless consonants. This seems to be a simple thing and no problems arise in the case of word-middle consonant clusters. We all know well that the Russian language is governed by the rule of retrograde voicing assimilation and we understand that a voiceless consonant cannot follow a voiced obstruent (hence /próz’ba/ (‘a request’) and the /z’/ and /b/ cluster), while a voiced consonant cannot follow a voiceless consonant, hence we have the cluster /tk/ in the word /lótka/ (‘a boat’). But we also know that there is a group of linguistic sounds called sonorants, which can follow both voiceless and voiced consonants; hence we can think of the word /sl’ot/ (‘a gathering’), where a voiceless consonant is followed by the sonorant /l’/, and the word /zloj/ ('evil'), where /z/ precedes the sonorant /l/; or we may think of the word /sm’éna/ (‘a shift’), with the voiceless consonant preceding /m/, and the voiced /z/ preceding /m/ in the word /zm’ej/ (‘a snake’). We all know that /v/ is a special sound, which can also be classified as a sonorant due to its phonetic behaviour in speech and its phonetic features; the word /sv’et/ (‘light’) has a voiceless consonant preceding /v’/, while the word /zv’er’/ (‘an animal’) has a voiced consonant preceding /v’/. However, difficulties arise at word boundaries; I mean the difficulties arising at the boundaries between a function and a content word or two content words.

Perhaps we should start with the boundary between two content words. We should of course bear in mind that we are now discussing cases when there is no pause between two content words. Let us imagine this word-combination: воз большой ('a large cart'). Certainly, if the second word, a content word, starts with a voiced obstruent, then the preceding consonant will also be voiced - /voz bal’shój/; if the second content word starts with a voiceless consonant, for example, воз сена (‘a cart of hay’), then the first consonant is also voiceless /vos s’éna/.

If the second content word starts with a sonorant, for example, воз листьев ('a cartful of leaves') or воз мяты ('a cartful of mint') or some воз веток ('a cartful of branches'), we should bear in mind that, according to the pronouncing standard, only a voiceless consonant is possible before a sonorant: it would be incorrect to say /voz m’ati/ or /voz v’étak/. Difficulties in pronunciation and errors connected with them here are connected with difference in rules governing pronunciation of voiced and voiceless consonants at the boundaries between two content words and at the boundaries between a function word and a content word; for if we examine the boundary between a function and a content word - let us take something like the phrase через год (‘in a year’) - then just like in the case of two content words, we have a voiced consonant preceding the voiced obstruent /g/, i.e. /ch’ér’ez got/, and before the voiced consonant, for example, in через село мы проезжали (‘we were driving the village’), where the content - and still not a function word - starts with a voiceless consonant, it naturally requires a voiceless consonant /ch’ér’es s’iló/. We should pay attention to the fact that if a function word is followed by a content word starting with a sonorant or a vowel, then the rules are different from those governing the boundaries between two content words; for instance, мы встретимся с вами через век, (‘we’ll meet you in a hundred years’), if this is possible, /ch’ér’ez v’ek/, so /v’/ is a sonorant, but it is preceded by a voiced and not a voiceless consonant; or take, for instance, через лето что-то произойдет (‘something is going to happen after next summer’); here we again see a voiced consonant: not /cher’s l’éta/ but /ch’er’z l’éta/; similarly, мы прыгали через огонь (‘we were jumping over the fire’). Linguists have written a lot about this issue and argue that such cases are most probably governed by the same rule that governs the middle of the word, since usually the stress falls on the content word and a kind of single phonological word is formed; some other rules apply here as well.

It appears that there are different rules governing phonetic realization of voiced and voiceless consonants at the boundaries between function and content words as compared to the boundaries between two content words; besides, a rather strong southern influence leads to a number of errors that appear even in the speech of native speakers. Students conducted an interesting research experiment. They took a trolley-bus along the Nevsky Prospekt and listened to the drivers announcing the name of the stop located across from the Eliseevsky Store. The name of the stop was «Сад отдыха». They counted how many drivers pronounced the phrase incorrectly as /sadód:iha/ and how many pronounced it correctly as /satód:iha/. But this is rather a simple case that had of course been studied before.

Many more differences that used to exist between St. Petersburg and Moscow pronunciation were connected with pronunciation of soft and hard consonants. Pronunciation of hard velars in masculine singular adjectives included hard and soft variants, for example, /v’il’íkij/ and /v’il’ík’ij/. These variants were both equally acceptable in the past. In Moscow people would say /v’il’íkij/, /zvónkij/, /grómkij/, and in St. Petersburg - /v’il’ík’ij/, /zvónk’ij/, /grómk`ij/. It should be mentioned, however, that starting from about the 1960s-70s, these two variants have stopped being equally acceptable; the variant with a hard velar has become archaic and obsolete; the standard of pronunciation, the codified standard, required, according to a serious research, the use of soft consonants, i.e. /v’il’ík’ij/, /zvónk’ij/, /grómk’ij/. This was not the only case when one had to think about pronunciation requiring either a soft or a hard consonant; there were also cases of hard /s/ or soft /s’/ that we have already discussed before. Before, it used to be easy to hear this Moscow phonetic feature, like /uchús/, /b’irús/, or /str’iml’ús/, when seeing a play in the Maly Theatre in Moscow. Older actors were trying not only to preserve this special feature but to emphasize it; but time went by and now we would hardly be able to hear this hard consonant on the stage of the Maly Theatre. St. Petersburg standard with the soft consonant - /uchús’/, /b’irús’/, /str’iml’ús’/ - has won.

Coronals are predominantly hard and 80% of labials are soft: /eff`ékt/, /azb’ést/, /f’en/, /graf’éma/, /maraf`ét/, /prap’éll’er/, /argum’ént/, /dif`ékt/, /mal’b’ért/. Similarly, 80% of velars are soft: /apag’éj/, /lak’éj/, /trah’éj/, /sh’éma/, /júnk’er/, /brók’er/. It turns out that here the influence of the system is also manifest: even though 80% of labials and velars are soft, exceptions also exist, for example, /b’iznesmén/, with a hard labial, or /vebsájt/. It is necessary to stress the dependence of the norm on the system: the system has expanded its possibilities and the norm took advantage of this fact.

Of course we should also mention consonant clusters; in the past it was easy to identify whether the person was from Moscow or St. Petersburg just by the way he or she pronounced the word конфеты ('chocolates'): /kanf`éti/ with a hard /n/ or /kan’f’eti/ with a soft consonant. It is interesting that while the second consonant is soft, the first consonant in most cases is now not soft, as it was the case in Moscow pronunciation, but hard. In Moscow they used to say /d’v’er’/, /kan’f’eti/, /stáf’k’i/, /ár’m’ija/, while in St. Petersburg they said /dv’er’/, /kanf`éti/, /stáfk’i/, /árm’ija/. But if the cluster includes consonants produced by different active articulators, as it is the case in the word /árm’ija/ ('the army'), for instance, when a coronal is combined with a labial, or in the case of /kanf`éti/ where again we see a combination of a coronal with a labial, the modern norm requires the first consonant to be pronounced hard, just like it was pronounced in St. Petersburg. If a consonant belongs to the same articulator group, then in some cases a soft first consonant is possible, for example, in the combination of /s’/ and /t’/, like in the word /s’t’ep’/ (‘steppe’), the combination of /z’/ and /d’/, like in the word /z’d’es’/ (‘here’), or the combination of /n’/ and /t’/, like in the word /v’in’t’ik/ (‘a small screw’). Therefore, if consonants belong to different active articulators then the first is hard; if the consonants are produced by the same active articulator - they are both coronals - then the first consonant is still pronounced soft.

It is also noteworthy that /j/ is a coronal and functions as any other soft consonant; it is also interesting that in the overwhelming majority of cases a consonant preceding /j/ is now pronounced hard: /pjut/ (‘they are drinking’), /abjóm/ (‘volume’), /s’imjá/ (‘family’); and only in those cases when we are dealing with a boundary between the root and the following part of the word, usually the inflexion, or some other part, a soft consonant is still possible, like in the following cases: /sud’já/ (‘a judge’) or /stat’já/ (‘article’), /kamar’jó/ (‘a swarm of mosquitoes’), /varan’jó/ (‘a murder of crows’), /kalós’ja/ (‘wheat ears’), and /palóz’ja/ (‘sledge runners’). In these cases we can hardly expect a replacement of a soft consonant with a hard consonant any time soon; it is hard to imagine that the variant /sudjá/ will become the norm.

In connection with consonant clusters we should mention the remarkably illustrative combinations "чн" and "чт". Indeed, in St. Petersburg they said /bulachnaja/ (‘a bakery’), while in Moscow they said /búlashnaja/, i.e. in St. Petersburg "чн" was pronounced as /chn/ and in Moscow as /shn/. Similarly, they said /jáblachnij/ (‘apple [juice, drink]’) in St. Petersburg and /jáblashnij/ in Moscow. St. Petersburg norm of pronunciation has won, but it is interesting that here we also see a certain morphonologic reason. In the words with the к//ч interchange, which is so well known to everyone speaking the Russian language, we see /chn/ pronounced in such consonant clusters. The modern norm is /jáblachnij/ (яблоко — яблочный), /par’ádachnij/ (порядок — порядочный), /búlachnaja/ (булка — булочная), /malóchnaja/ (молоко — молочная); but in words that do not have the mentioned consonant interchange, the cluster is still pronounced /shn/, for example, /skvar’éshn’ik/ (скворец — скворечник), /jiíshn’ica/ (яйцо — яичница). I suppose it would be hard in this case to explain the victory of /chn/ by the spelling, even though such attempts there have been: it is written “чн”, so how can one pronounce it differently? However, despite the fact that in Moscow people knew how to read, at least the majority of the city population did, they still said /búlashnaja/, /malóshnaja/, and /par’ádashnij/; therefore, it is obvious that the changes that happened were not connected with the influence of spelling but rather had some deep, internal - we can even say inherently morphonologic - causes.

As for the “чт” combination, it is interesting to note that the modern norm requires its pronunciation as /shto/ and /shtóbi/, like it used to be pronounced in Moscow, and not /chto/, /chtóbi/. The fate of separate words like, for example, /skúshnij/ (‘boring’), is of particular interest. Pronunciation of the word /kan`eshna/ (‘of course’) used to be the norm in the past and has remained the dominant variant of the norm now. At the same time, pronunciation with /ch/, /skúсhnij/, has become, unlike /skúshnij/, the dominant variant of the norm. I would like to stress again that here changes are possible. We see the continuous process of a single norm evolving; we started writing about this in the 1960s-70s, about a single norm devoid of local features. Lev Shcherba once wrote in his work on the norms of model Russian pronunciation: "Pronunciation of the future will brush aside everything that is too local, too Moscow or Leningrad, too Orlov or Novgorod, let alone the various characteristic features of other languages, like the Caucasian and Central Asian guttural /g/, the Ukrainian /?/, the Tatar /i/, etc." We see now that the single norm has indeed evolved.

Finally, I would like to draw your attention to pronunciation of the words дождь and дождя. The thing is that as late as in the middle of the 20th century you could still easily determine where the weather forecast you were listening to was broadcast from - Moscow or St. Petersburg. If the radio announcer said that tomorrow /dosht’/ ('rain') was expected, then the broadcast was of course from St. Petersburg, or Leningrad; the same goes for the sentence /dazhd’á/ не ожидается ('is it not expected to rain'). However, if the announcer forecast /dosh’:/ or said that /dazh’:á/ не будет ('it is not going to rain'), then the broadcast was from Moscow. The single norm of pronunciation that borrowed some features from the old St. Petersburg pronunciation and some from the old Moscow pronunciation, in this case retained the feature of St. Petersburg pronunciation; and now Moscow broadcasters also say ожидается /dosht’/ ('rain expected') or /dazh’:á/ не ожидается ('rain not expected'). These are the principal features of modern Russian orpthoepic standard.

Speaking about orthophony, I would like to stress that the modern norm requires a hard /c/, even though it used to be slightly softened in St. Petersburg: /r’ival’úс’ii/, /int’il’ig’énс’ii/. Today, pronunciation /r’ival’uсii/ and /int’il’ig’énсii/ is the norm. At the same time, /ch/ that used to be slightly harder in St. Petersburg pronunciation, in accordance with the modern norm, is to be very soft.