Why De-Russification Isn’t On The Cards

My post last week on the increasing visibility of the Russian language on the Internet provoked a heated counter-attack from commentator Ildar Adi, who asserted (without much in the way of proof) that it is actually in significant retreat in Europe, the Near Abroad, and even Russia itself.

He believes that whereas there were almost 500 million Russian speakers in 1990, there will be just 150 million of them in 2030. If that were to be true, it would imply that practically nobody outside Russia would still speak Russian in 20 years time. Does this sound like a very likely prospect?


Not if the above map, linked to by commentator Glossy, is anything to go by. In this case, one image really is worth a thousand words. Quoting Glossy:

It seems that the only languages of the former Soviet Union that are used enough on Twitter to have merited their own colors on that map are Russian, Latvian and Lithuanian. It looks like Estonia isn’t tweeting much in Russian either though. The rest of the post-Soviet space is. On the map Catalans are tweeting in Catalan, but Ukrainians are tweeting in Russian. And that’s the young generation, the future. Who else is going to use Twitter? Kiev looks like the third-brightest Russian-tweeting city in the world, right after Moscow and St. Petersburg. Slovenian, Slovak and Albanian have their own colors, but Ukrainian doesn’t.

The original w3techs report confirms that the ex-USSR Internet is, for all intents and purposes, the Runet:

Russian is also the most used language in several countries that belonged to the Soviet Union: 79.0% in Ukraine, 86.9% in Belarus, 84.0% in Kazakhstan, 79.6% in Uzbekistan, 75.9% in Kyrgyzstan and 81.8% in Tajikistan.

A 2012 study showed that:

  • On the radio, 3.4% of songs are in Ukrainian while 60% are in Russian.
  • Over 60% of newspapers, 83% of journals and 87% of books are in Russian.
  • 28% of TV programs are in Ukrainian, even on state-owned channels.

The Russian Wikipedia consistently gets about 70% of all hits from IP’s located within Ukraine. This figure hasn’t budged since 2009 when they started gathering data on this. The Ukrainian language Wikipedia, getting 15%, is only about twice as popular as the English language Wikipedia.

In Belarus, this figure is close to 90%. In Kazakhstan, it’s 80%, and is likewise dominant in the rest of Central Asia. Even in Azerbaijan (!), its 40%: That’s about as much as English and Azeri combined. The only countries in the ex-USSR where the Russian Wikipedia isn’t dominant are Georgia and the three Baltic states.

Despite the official efforts to De-Russify, the Russian language has if anything grown in prevalence in Ukraine since the end of the USSR. The percentage of those who consider it their “native language” went from 35% in 1995 to 40% in 2013, despite the substantial outflows of Russians in that period. In any case, many Ukrainians who answer “Ukrainian” do so for sentimental reasons, not practical ones. Here are the results of a 2004 survey of high school students in Kiev to questions about their usage of the Ukrainian and Russian languages:

  Ukrainian Russian Both
Speak at home 13 61 25
Speak at school with friends 4 65 29
Watch TV 16 26 57
Read literature 12 30 57

This is the youngest generation, which lived at a time of state-backed efforts to De-Russify the schools and everyday life (whereas 54% of Ukrainian students studied primarily in Ukrainian in 1991, by 2002 this had increased to 74%). But these same efforts, however – contrary to Ildar Adi’s assertions – more an expression of the perceived weakness of the weakness of the indigenous languages than anything else.

So what explains this? Ironically, the likely answer is IT and modern technology, and increasing, globalization. While the USSR promoted Russian as a language of inter-ethnic communication, in practice for most of that period many ethnic minorities were taught in their own language. For instance, a little known fact is that there were riots in Tbilisi in the 1970’s when it was proposed to make Russian a language equal to Georgian in education. In the Warsaw Pact countries, Russian was taught as a foreign language, i.e. most people didn’t really end up mastering it. So outside of the top bureaucratic echelons, where Russian was the lingua franca, it was not nearly as prevalent as it is sometimes made out to be.

Conversely, its “retreat” hasn’t been as universal. The ex-socialist bloc countries stopped teaching it as a foreign language, replacing it with English and German, and aside from the unemployed Russian language teachers it didn’t have a big effect. However, in the ex-USSR countries, barring Georgia and the Baltics, Russian remains pretty much universally known. And even if there are efforts to support indigenous languages at the level of government documents and schools, it is hard to make it supplant Russian because of (1) the self-reinforcing fact that everyone already knows Russian and (2) the much vaster “gravitational weight” of Russian on the Internet. The “weight” of English is far vaster than Russian, of course, but the Ukrainians and Kyrgyz aren’t drawn to it because so few of them know English in the first place.

Consider your typical Kievan Ukrainian patriot. He votes for Vitaly Klychko, against making Russian the second official language, and wants to join the EU if not NATO. On the other hand, the people he communicates with in everyday life usually speak Russian, the new movies he devours are in Russian or English with Russian subtitles, the literature he reads (be it Dostoevsky or Boris Akunin) is typically Russian, etc. When he wants to buy a new car and Googles (or more likely, Yandexs) “авто”, about 90% of the results will be in Russian. There is now too big a critical mass of people who know Russian and use it to communicate across the CIS for it to ever “vanish”, if anything it is likely to further grow in prominence as education levels, economic inter-connectedness, and Internet penetration increase. All this doesn’t make him a hypocrite or not a “true” patriot – consider the relationship of Ireland or India to the English language. That, approximately and with a caveats, is the true relation of most of the ex-USSR countries to the Russian language.


  1. A very good article. One quibble: I suspect that the map simply does not list the Ukrainian language as an option, not that the Ukrainian language is nonexistent on twitter. For whatever reason, sometimes on surveys or elsewhere this language is ignored (the first example I googled):


    There are about 9.5 million people living in the 7 almost totally Ukrainian-speaking oblasts that were annexed to the USSR after World War II. Lviv has two major universities (including a technical one) and is Ukraine’s third largest city in terms of having IT specialists:


    The map shows western Ukraine as a largely black area with almost no tweets. Moreover, on the map Kiev is only about as bright as Minsk. Either as many people use twitter in Minsk as in Kiev (which is a much larger city) or for whatever reason the Ukrainian language isn’t included, so Kiev is as bright as Minsk and Western Ukraine is big black spot.

    Otherwise, your article is pretty much correct. It would be interesting to see surveys about students later than 2004 to see if anything changed after the Orange Revolution (I make no claims that it has, but am curious).

    • Thanks for the comment.

      About Kiev and Minsk, while the former is larger than the latter, it’s not by an absolutely huge margin (Minsk has about 2mn, Kiev 3mn). Also Internet penetration is higher in Belarus, though in fairness the capitals are likely to be about equal in that respect.

      But overall you make some very good points. It’s quite possible that Ukrainian was simply left out. But I’m not absolutely sure. For instance, the point of light that denotes Lviv appears to have something of a greenish hue (in addition to cyan) while the areas that are overwhelmingly Russian speaking are cyan.

      It would indeed be interesting to see if the Orange Revolution had any obvious statistical effects on Ukrainianization (and its end, on the converse).

    • SFReader says:

      Lvov used to be Polish-Jewish city before the WWII. The Holocaust and repatriation of Poles led to total replacement of its native population, primarily with ethnic Russians and Ukrainians from eastern Ukraine.

      As a result, post-war Lvov started as a Russian-speaking city. Only by 1980s did the influx of Ukrainians from rural Western Ukraine change ethno-linguistic composition of the city.

      Now Lvov is an Ukrainian speaking city with significant Russian-speaking minority (about a quarter of population, last time I checked). This led to emergence of local variant of surzhik – mixed Russian-Ukrainian speech so dominant elsewhere in the Ukraine.

      In contrast, the rest of Western Ukraine has almost no Russians and speaks entirely in Ukrainian (though in different dialects from standard Ukrainian)

      • One set of my grandparents lived in Lviv before the war (right next to market square). It was historically slightly under 20% ethnic Ukrainian, 30% Jewish and slightly under 50% Polish (almost half of the ethnic Ukrainians in the city, including even my ideologically Russophile relatives, spoke Polish). The city was surrounded by ethnic Ukrainian villages and was headquarters to all the most important Ukrainian cultural institutions (the Church, university departments, newspapers, etc.) Polish policies between the wars brought the Ukrainian population down to 15% and boosted the Polish population. While Ukrainians were a minority in the city, Polish and Russian nationalists both sometimes pretend that they were nonexistent and that this was totally a Polish or a Polish-Jewish city.

        In mid-century most (but not all) of the local Ukrainian intelligentsia fled or were killed, most Jews were killed, and most Poles expelled so the city was indeed largely Russian-speaking immediately after the war. Some of my relatives, academics, stayed behind and were mostly mixing with incoming Russians or Russian-speaking people from Eastern Ukraine after the war (my uncle married a Russian girl). A lot of Soviets were desperate to move there: an opportunity to live in a Western city.

        However the Russian period was quite short-lived and Lviv became Ukrainian-speaking much sooner than the 1980’s. By language, Lviv was never more than 50% Russian-speaking and by 1970 the city was already 65% Ukrainian speaking and 31% Russian speaking. The Russian share was down to about 25% in 1979, and less than 20% in 1989. I suspect it’s no more than 10% now; even many ethnic Russians have switched to Ukrainian as their primary language. In contrast Vilnius is about 14% ethnic Russian.

        You might find this article (in Russian) about ethnic Russians in Lviv, written from a Russian nationalist perspective, to be interesting:


        • SFReader says:

          I’ve been to Lvov a few times in 2000s and definitely can say that Russian language was very much alive there. In fact, I was told that real Ukrainian is spoken only in the countryside!

          What passes for Ukrainian in the city is heavily influenced by Russian (and I am sorry to report, by Russian obscenities). And I noticed that Lvov citizens have almost similar difficulties understanding official TV variety of Ukrainian as people from eastern Ukraine.

          • I’ve been there a few times also, and am in regular contact with people from that city. My half-Russian cousins speak perfectly normal Ukrainian (this is their preferred language); Russian is heard in the city but not often. I don’t recall hearing surzhyk there. I suspect that surzhyk is mostly used by lower class people and that the Russians that settled in Lviv after the war were, as you mentioned, typically educated ones. At any rate, the stats seem to support my anecdotal observations of Russians being relatively uncommon in Lviv vs. yours of the Ukrainian language being spoken only in the countryside. A very detailed academic article (in Ukrainian, so googletranslate) written by a Russian from Lviv:


            According to the article, in 2000 12% of the city’s residents considered themselves to be Russians and 20% of Lviv’s population preferred the Russian language. I suspect these numbers have dropped further in the last 13 years, as more of the older generation of Russians and Russian-speaking eastern Ukrainians who came post-war has passed away.

            In terms of speech, the biggest difference between the city and the surrounding countryside is that city-dwellers tend to speak literary Ukrainian while the Galician dialect (familiar to diaspora Ukrainians) is prominent in the countryside.

            • SFReader says:

              Of course, Lvov is predominantly Ukrainian-speaking city, I don’t dispute this. I was just pointing out that this Ukrainian is quite different from being standard literary Ukrainian being influenced both by Russian and by local Galician dialects

              Thanks for quote on Lvov being 20% Russian-speaking in 2000. I think this is the figure I remembered.

              This number well may be lower today, but still Lvov remains a city with substantial Russian language presence, certainly much higher than the rest of Western Ukraine.

              PS. By the way, I hear that there is still a widespread adoption of Russian language by western Ukrainians who move to predominantly Russian-speaking cities, including capital Kiev.

              • Yes, if today 10% or 15% of the city’s residents speak Russian this is much higher than virtually 0% who speak this language outside the city. But it is still uncommon (btw, this figures makes Lviv much less Russian speaking than any of the Baltic capitals).

                One thing that is quite noticeable is that Lviv’s numerous bars and restaurants (which mostly didn’t exist in the early 2000s and only mushroomed at the end of that decade) are filled with Ukrainian-speaking people. Unless they are Kievan visitors who switched over while visiting the city, this means that there aren’t many tourists and therefore that the locals are doing well economically – at least well enough to go out a lot.

                And the Ukrainian they speak is certainly not heavily Russian-influenced. It is simply standard literary Ukrainian, unlike the Galician dialect spoken in the villages. Your claims seem rather bizarre….

                Here is an article, in English, from the Moscow News, about the Russian language in Lviv. It’s mostly anecdotes, but still:


                Apparently, Polish, German and English are more popular as second languages in secondary school than is Russian. The newest generation of Lviv residents often doesn’t even speak Russian, unless they come from Russian families.

                A fascinating academic study about Russian natives of Lviv and their own complex self-identification:


                (scroll down to pg. 795)

                I don’t know about Galicians switching to Russian when they move to, for example, Kiev. However some have claimed that the only reason the Ukrainian language is more widespread in Kiev now is because a lot of western Ukrainians moved there. So which is it?

  2. I’d think that mass migration flows from Central Asia to Russia also contribute to Russian’s popularity and usefulness. I already mentioned this in the last thread, but migrants usually speak at least some Russian so that they can function. That’s been my experience anyway.

    • Little Pig says:

      In the Asian CIS countries people care a lot about their children being able to speak Russian, and not simply to speak, but be close to native speakers. They often choose to speak Russian to their children even if they aren’t really good at it. At least it’s true for many immigrants. And I talked to some of them – in case of going back they would give their children to kindergartens and schools where Russians study so that they would be able to practice it.

  3. Scowspi says:

    “All this doesn’t make him a hypocrite or not a “true” patriot – consider the relationship of Ireland or India to the English language”

    Actually, this analogy doesn’t hold up very well, for a simple reason: English does have official status in both Ireland and India, but Russian has no official status in Ukraine. The Ukrainian situation is unlike any other I know, in that a universally spoken (and for most people native) language gets no official recognition.

    • Israel has Hebrew and Arabic as official languages. Was this the case at its founding, when the early generations of settlers were mostly Yiddish or English-speaking and Hebrew was “reborn” as a native language?

      • Scowspi says:

        Don’t know about Arabic, but Hebrew must have been official from the beginning. There was actually a campaign to suppress Yiddish in favor of Hebrew when the state was founded.

      • Both Hebrew and Arabic were official languages from the beginning. The use of Hebrew and Arabic in an official capacity (and English in a semi-official capacity) was a hold over from the Mandate era since from 1922 when the British authorities made all three languages the official languages of government (including that of local authorities and municipalities).

  4. Erik Phiipsen says:

    Sorry to let you know that to day there is NO secondary schools where Russian is taught. There used to be some 20 in the late 80s. This is in my view due to a russophobia which stems from a bacground radiation from the cold war, enhanced by the cayalyctic influence in western media of Boris Berezowsky & consorts. The murders of Politkovskya and Litvinenko had an enormeoius nfluence. It is going to be very interesting to follow what happens in the new investigations of the Litvinenko murder.
    iNCIDENTALLY: What has transpired from the Arafat 210- allegations and investigations – and conclusive results ?

  5. There is something odd about the map. I notice that Belgrade is visible but does not seem to be in the Croatian colour. Is it that Belgrade is coloured in a Serbian colour but that the Serbian colour is not listed?

    And Lvov/Lviv is visible and the colour most prevalent there does not seem to be the same hue as that used for Russian.

    Is it possible that the key for the map is incomplete? After all it would seem weird that Afrikaans and Croatian are on the key for the global map but not Serbian and Ukrainian.

    If the colour in Lviv is indeed a Ukrainian color and the colour in Belgrade is a Serbian colour then what does that suggest about the use of twitter in Ukraine, including among the IT centre that is Lviv? Could it be that the IT specialists in Lviv just don’t use Twitter? (which would not be a surprising outcome to me, since it might be that they consider it a waste of time and of limited practicality).

    • You’re right. I initially didn’t notice the fact that Lviv/Lvov was in a distinct color. I now subscribe to your theory that the key in the upper right is incomplete. Besides Lviv/Lvov and Belgrade, Estonia looks suspicious. There are specks of a reddish color there that’s probably Estonian, a language that’s not listed in the key either.

  6. Ildar Adi says:

    I’m sorry to hear that providing hard data in a form of statistical report is considered to be “a heated counter attack” here. I provided statistics to prove that people in Europe proper think Russian language as of zero usefulness (except in the Baltics). The situation was quite different just a quarter of century ago.

    Map of languages used in Twitter is interesting, but misleading. That particular map was created in 2011 by using Google Chrome’s language detector to extract the languages written in tweets then correlates them with GPS location data sent to the website by those users. Detecting languages algorithmically is not an exact science, and when language detector is given the short and abbreviated texts, like tweets, it can give inaccurate results. This is especially true with languages that are somewhat similar, like Russian, Ukranian and Belarusian. It is also very well known that geotagging feature can be unreliable. For example, even if you want to believe that nobody tweets in Ukrainian even in the Western Ukraine, you certainly shouldn’t believe that so many tweeters in Wallonia use Dutch, like the map indicates. In conclusion, and just like in your previous post on this subject, proof and data you provided of prevalence of Russian language in Little and White Russia is very likely to be incomplete and inaccurate.

    This is very interesting topic, and I shall comment your other claims when I have time.

    • Detecting languages algorithmically is not an exact science, and when language detector is given the short and abbreviated texts, like tweets, it can give inaccurate results. This is especially true with languages that are somewhat similar, like Russian, Ukranian and Belarusian.

      Swedish and Norwegian are at least equally similar, and yet they are clearly delineated on the map.

      • In any case it does appear that Ukrainian and Serbian are on the map unless the points of light representing Lviv and Belgrade are anomalies.

    • For example, even if you want to believe that nobody tweets in Ukrainian even in the Western Ukraine, you certainly shouldn’t believe that so many tweeters in Wallonia use Dutch, like the map indicates.

      Umm….why wouldn’t one expect some tweeters in Wallonia to use Dutch? Is it that inconceivable that some of the Flemings in Belgium might actually have reason to be in Wallonia (whether for business or pleasure)? Belgium allows freedom of movement within the country after all and freedom of expression. So it’s quite possible for a Fleming to go to Wallonia and tweet in Flemish.

      And you need to look more closely on the map. The colour for Lviv in western Ukraine does not seem to be the same colour as the cyan used to indicate Russian (likewise the colour in Belgrade is not the same colour as is used for Croatian). So unless the map is indicating that people in Lviv tweet in Slovak or Czech and that people in Belgrade tweet in English or some such then it seems likely that the key for the map is incomplete.

      • I think you’re right. The map key appears to be incomplete and glossy (a generally good contributor not prone to making mistakes) erroneously concluded that nobody in Ukraine tweets in Ukrainian.

        • And it would seem that the author of the map had erroneously had “Latvian” in the key twice and left out “Lithuanian”; see the comments on October 27, 2011: http://flowingdata.com/2011/10/27/language-communities-of-twitter/

          The author also admits that the map does not represent total twitter volume but only the twitter volume that is geotagged.

          And here we have the author admitting the caption is incomplete: http://www.flickr.com/photos/walkingsf/6277163176/in/photostream/

          “Eric Fisher (17 months ago)
          Yes, it’s Peters projection because equal-area seemed important for showing density. I know it looks lousy in some places—sorry about that.

          The software claims to detect Irish, Welsh, and Basque, but the numbers are very small compared to other languages. I don’t know whether it is misidentifying them as something else or if they are just really very rarely used on Twitter. I did cut off the caption at 10,000 tweets because it seemed like anything lower than that was such a small fraction that they would be more confusing than useful to list. The dots are still drawn on the map, they just aren’t labeled.

          So it seems that Serbian, Ukrainian, Welsh, Basque and Irish are probably all on the map (probably Belarusian as well), but that geo-tagged tweets in all of those languages are most likely below 10,000.

          I’m pretty certain the software the author is referring to CAN detect Ukrainian separately from Russian since Google can detect Ukrainian and offer to automatically translate it for you.

          • That all seems about right.

            I know for a fact, having been to both places, that Welsh and Gaelic are hardly ever used in both Wales and Eire, respectively. Except perhaps for 2 hours or so of your weekly high school curriculum, and the “Garda” on Irish police cars, they are fully Anglicized.

            As regards Europe, perhaps the only surprise is that Catalan is considered a separate language to Spanish, and that most Catalans use it on Twitter.

            • Scowspi says:

              Maybe this is nitpicking, but I can’t resist. Welsh is definitely used in Wales, particularly in the south of the country. I’ve known several Welsh people who were fluent in it and had passed through the Welsh-medium school system. And Catalan really is a separate language; it’s about as different from Spanish as Portuguese is.

        • Thanks for the kind words and yes, I now see that Lviv/Lvov is shown in a distinct color on that map. The key is incomplete.

  7. Oops. AK could you put in a quotation mark after the bolded “aren’t labeled” please? I also intended to underline the the entire section of the quote starting from “I did cut off the caption at 10,000 tweets” to “just aren’t labeled” but it seems that didn’t work. Any way you can change the underline tags to italics tags so that relevant section can be highlighted in some way?

  8. Ildar Adi says:

    Look, everyone, the language detection engine in question is just not awfully good at detecting Ukrainian. You can test it your self. Ask Google Translator to detect in what language is this:

    Дякую тобі, боже, що я не москаль

    My quess is that there are several reasons for this:
    1) Ukrainian and Russian text tokenizes similarily into quadgrams, but Russian gets detected because of it’s general prevalence
    2) Training data that was used for Ukrainian was of poor quality and quantity compared to Russian (for example, Ukrainian web pages are more likely to contain also Russian than vice versa)
    3) Some Twitter clients at the time simply did not have Ukrainian language support.

    Why Norwegian, Danish and Swedish get detected much better? Because none of the points above apply to these languages.

    • So what colour is that in Lviv on the map then? It definitely does not seem to be the same colour used for Russian……

    • Ildar, that’s certainly an… interesting choice for your example Ukrainian sentence.

      • Ildar Adi says:

        Well, it definitely is not Russian in any sense, is it? Besides, last time I was in Kyiv I saw T-shirts sold with print everywhere.

        • It’s not Russian in any sense?

          Looks like a Russian dialect to me. I can’t speak Ukrainian, but I’m pretty sure I know what it

          4 words in the expression are identical in meaning to Russian: боже, я, не, москаль;

          two words have slight pronunciation differences to their Russian equivalents (тебе, что), resulting in slight changes in their orthography when compared to their Russian cognates, including the letter (i) no longer used in the Russian Cyrillic alphabet: тобі, що;

          the remaining word I do not know but recognize as a verb, as it has an identical 1st person singular ending of a Russian conjugation: дякую.

          Oh, it’s Ukrainian!

          So it IS a Russian dialect.

          So is this, but it’s a dialect of Russian:

          Че те надо?


          • Ildar Adi says:

            I’m not saying that Ukrainian and Russian are not mutually intelligible. I’d even say that there are some grounds to claim them being the same language (but then you would have to say that from a scientific, linguistic point of view Russian is actually a derivative and dialect of Ukrainian).

            I’m saying that Google language detection obviously fails to detect Ukrainian in a sentence that for a human knowing both Russian and Ukrainian is glaringly obvious; therefore it is extremely likely that the Twitter language map has inaccurate and incomplete information about the usage of Ukrainian; therefore the case for disproving linguistic de-Russification in Ukraine by that map collapses.

            • Yes, I should have added that one could argue that Russian is a dialect of Ukrainian, though I should think that it would be more accurate to say that both languages originated as dialects of a common proto-East Slavic one.

              I checked out “дяковать”, the verb that I could not recognize, by the way, and from which one arrives at the first person singular “дякую”. It means “to thank” (благодарить) and evolves from the Polish “dzięk”, giving “dziękować” (“благодарность), Polish loan words not being uncommon in the Ukrainian lexis.

              So the expression translates into English as:

              “I thank Thee, O God, for my not being a Muscovite!”

              More colloquially and in Modern English:

              “Thank God I’m not Russian!”

              Which is what I thought it meant at first glance.

  9. Ildar Adi says:

    So Anatoly, now even the Russian government acknowledges that the Russian language is in massive retreat all over the world. When will you admit you were wrong?


  10. Another point to consider is weight of Russian language is high in Cyrillic script is very high, compared to english in Latin script.