According to a multi-million-views talk of John Mearsheimer (2015), the "ethno-linguistic map of Ukraine" looked like this.

![enter image description here

There's no source given in the talk for the map and I've not seen this distinction made before between (e.g.) "Ethnic Russians" and "mostly Russian-speaking Ethnic Ukrainians". How does Mearsheimer or whomever he sourced that map from (assuming someone here can suss that) make that kind of distinction?

  • 3
    Ethnicity tells nothing about the political views, nothing says it cannot be ethnic Ukrainian with totally pro-Russian thinking and in reverse. Just some subtle demographic science. Hence this may be off topic.
    – Stančikas
    Sep 23, 2022 at 14:27
  • 1
    Unless the source of the map is identified, any answer will be supposition. There are many ways of specifying/identifying ethnic identity, ranging from asking, to looking at factors like place of origin, language, religion, customs, even genetics.
    – Stuart F
    Sep 23, 2022 at 15:42
  • 5
    You ask "Вы русский?". If the answer is "Да", you've found an ethnic Russian.
    – Mark
    Sep 23, 2022 at 23:45
  • 4
    These discussions about ethnicity and language are rather meaningless: people are free to identify themselves as Russians, Ukrainians or anything else, and their self-identification may even change, depending on who asks and what the political situation is. And let us not forget that there have been a lot of intermixing. Sep 26, 2022 at 10:47
  • 1
    @alamar Depends on how you compile the map. Like if you'd ask people around the country concerning their primary language and ethnicity then the result is rarely equivalent to regional boundaries (unless they were established very recently with respect to that fact). Though if you go the other way around and use data from these specific regions with regards to what language is spoken there, then such a map is exactly what you' d expect. Whether it's any useful to do it like that is a different question.
    – haxor789
    May 18, 2023 at 11:56

3 Answers 3


This is usually done by asking the people of their ethnicity and writing down their answers.

Now the frame challenge here is that the same state who is asking its people about their ethnicity is also projecting its own definitions of nationality and its history on the population, so this measurement tool set is influencing what it is measuring all the time and it is actually its prime function.

Still, some of these (striped) regions were just not seeing themself as ukrainian, given they were industrial, coal-mining new settlements populated with newcomers, compared to south ukrainian countryside populated with peasants in somewhat older (still very young by European or even American standards) settlements.

Ukrainian SSR and then Ukrainian state were trying to explain the population of those regions that they are Ukrainians now, but as you can see the process did not complete. These regions had legacy of Donetsk–Krivoy Rog Soviet Republic and were appended to Ukraine by a decree, much like Crimea which was also handed over by a decree.

  • Not my DV, but are you saying those are two different Qs (ethnicity, language) normally asked in Ukraine in sociodemographic polls, maybe also in the (rather ancient by now) census etc.? Or is any of these two factors determined by an external/"objective" method, to get results like that map? Sep 23, 2022 at 15:18
  • I don't know how these polls were conducted in Ukraine, but the language is not a singular question either: in places with complex ethnic situation it makes sense to ask about language of birth, language of study, language used at home and language used at work - these might be all different and it makes sense to make them multi-select.
    – alamar
    Sep 23, 2022 at 15:23
  • N.B. I recall from some other Q here that "natsional’nost’" is actually bit ambiguous in the [post-]Soviet space. Is that more likely what we're seeing here, i.e. answers to "natsional’nost’" question being more skewed towards Ukrainian, despite the answer given to language? Or to get the divergence one these, one needs to use a method for language deamination other than just asking the respondents about it? I guess some concrete polls would be useful to look at, but for me there's the language barrier in finding them. For Mearsheimer to have found them, i suspect some may have been in E Sep 23, 2022 at 15:33
  • 2
    I'm not sure why you would expect ethnicity (or self-perception of that) to match the language. There's a huge number of Rissian-speaking kazakhs in Kazakhstan, for example. They look (and feel) unambiguously kazakh but often can't speak kazakh language comfortably.
    – alamar
    Sep 23, 2022 at 15:37
  • @alamar "language used at home": there may also be more than one language used at home, though I have no idea how common that might be in Ukraine.
    – phoog
    May 18, 2023 at 9:48

It depends on how old the map is.

A possible answer to the methodology of "how" vis-a-vis ethnicity.

The USSR recorded everyone's ethnicity in their internal IDs. They were also recorded in all internal school and work records. I don't believe these records are still used.

However, it obviously made statistical determinations about ethnic breakdown much more trivial. While there is flux over the years, most people don't move that much. The amount of moving can also be estimated based on birth/death records and population numbers. Even college students usually study in the cities in which they grow up.

A possible answer to the methodology of "how" vis-a-vis language is that there are a lot of statistical markers for language consumption.

Newspaper, book and other media sales are good indicators.

Just as a side note, the map is also somewhat lacking. The West Ukrainian population is a mix of ethnic Poles and ethnic Ukrainians. And the language is a mix of both. Native Ukrainian speakers often report having a difficult time understand the Western Ukrainian dialect. The few times that I personally had to encounter it, I found it easier to understand than the proper Ukrainian. It could be because my Ukrainian is somewhat weak and the mountain Ukrainians (because the West is mostly mountains) are forced to use a lot of English-based neologisms. I really can't tell why, to be honest.

The map also doesn't include ethnic Crimean Tatars who were force-removed from Crimea to the Asian regions of the USSR during Stalin's rule. Ukraine, after becoming an independent state, invited them to return. They have their own language and they are ethnically closer to Turks than to any Slavic ethnicity. In Ukraine, they are considered to be the native population of Crimea. Tragically, after the Russian Federation occupied Crimea in 2014, it compounded Stalin's legacy by forcefully removing the Crimean Tatars again. Crimea is not that big a region (it's roughly a 70mix70mi square). So influx or expulsion of the native population could change the demographic quite a bit.

  • Yeah, from what I've read elsewhere, chucking Rusyns as ethnic Ukrainians in the 2001 census was quite controversial. The map in question here makes the same hierarchy, so I suspect is based on either the census itself or some other poll that used the same official categories. Sep 23, 2022 at 16:55
  • 1
    Soviet Union had significant internal mobility. Such as by making ukrainians, second largest ethnicity of the Soviet Union, the second large ethnic group in 36 out of 85 regions of Russian Federation (number taken from the internets) as an obvious result of that mobility.
    – alamar
    Sep 23, 2022 at 16:58
  • @alamar but how would those numbers be determined if the mobility is so great? The formal records of everyone's ethnicity stopped being kept in '91. So unless the ethnic composition stayed mostly the same (a reasonable assumption), the ethnic compositions of regions of the Russian Federation would not be determinable at a level more precise than race. But your "36 out of 85" statement wasn't about the racial decomposition. It was about the ethnic decomposition.
    – wrod
    May 18, 2023 at 5:44
  • @wrod People self-report their ethnicity on polls and during census. However the quality of the latest census is so abysmal it comes into question what did they measure exactly.
    – alamar
    May 18, 2023 at 8:40

In cases so subtle, likely just by the opinion of that person. Some Ukrainians actually used to use Russian more even in the family (source). However they identify themselves as Ukrainians and now started to learn the Ukrainian language much more seriously. Language is important part of the nation, but not the only and not mandatory. Switzerland also uses mostly three languages for communication, none unique only for them (German, French and Italian).

  • Switzerland has four official langauges: the three you mentioned + Rätoromanisch. ALso worth noting that the official Standard German is not the same as Swiss German, which is spoken at home and varies significantly from one canton to another. In fact, Swiss broadcasts and films in native dialect are dubbed when shown in Germany. (The official Swiss German is however very similar to the High German used in Germany.) Sep 26, 2022 at 10:25
  • Rätoromanisch is currently very little used. Swiss German can indeed sound so that no German would understand but it has no official status. It is often used as the mark of "one of us" in German speaking part because due the lack of any formal and written rules can only be learned in community, but French and Italian speaking parts do not know it either. There is no single language or dialect that majority of Swiss citizens would share.
    – Stančikas
    Sep 26, 2022 at 10:55
  • There is no single language or dialect that majority of Swiss citizens would share. - English ;) Sep 26, 2022 at 12:21
  • 1
    @phoog what is your point exactly? Anyhow, Swiss policy in terms of languages is an example to follow, IMHO. May 18, 2023 at 10:01
  • 1
    @RogerVadim my initial point is that we should be careful to remember what "majority" means (not directed at you; I tagged you only because it followed from your comment). The more substantial second point is that majority is often not a good way of evaluating political questions. Rereading this answer and the comments, I think we agree that language questions are complex, but I get the impression that Stančikas introduced the example of Switzerland less for its policy and more to show that national identity need not depend on language or ethnicity.
    – phoog
    May 18, 2023 at 10:45

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .