Overcoming Aggressive Monolingualism: Prejudices and Linguistic Diversity in Russian Megalopolises

The study deals with linguistic prejudices of citizens of the two main Russian cities, Moscow and St. Petersburg, toward speakers of foreign languages. It aims to reveal possible recent changes in the language ideology dominating Russian society. Monolingual and linguistically normative orientations rooted in the Soviet ideological approach are being challenged nowadays by global processes of migration and cultural diversification, which influence the everyday reality of Russian megalopolises. The research is based on the analysis of two sets of data: (1) meta-discourse on language attitudes derived from interviews with labor migrants and native Russian speakers in St. Petersburg and Moscow, and posts and comments on issues of language, migration, and linguistic landscapes, collected from websites and social media and (2) linguistic landscape data collected in 2016–2019, mainly in St. Petersburg, which reflect recent changes in attitude toward linguistic diversity in public space. These data show, on the one hand, that most city dwellers still relate to monolingual speech norms and try to implement control over public space; on the other hand, that the tolerance toward multilingual communication has been increasing over the years. The study suggests that these “first cracks” in monolinguals facades of Russian cities could eventually lead to the establishing of a less rigid language regime.


Introduction
Linguistic prejudices that exist in societies are and should be treated as complex social phenomena both reflecting and constructing social inequality. On the one hand, they are usually closely associated with negative attitudes toward certain ethnic, regional, or social groups speaking in languages or language varieties differing from the standard and/or prestigious ones. On the other hand, they can reveal more general assumptions about norms of language use. For example, when we see people frowning upon foreigners speaking loudly on a bus, we may assume that they do so because of their dislike of foreignspeaking newcomers, maybe from a particular part of the world. Alternatively, it may be because they believe it to be rude to raise one's voice in public places. Or, and this case is especially interesting, maybe they would tolerate loud utterances in their native language but are not ready to accept the invasive clamor of languages they cannot understand. Certainly, in real-life situations we encounter every day, all these reasons may be applicable simultaneously as people tend to be prejudiced in many different ways at the same time. Negative attitudes to speakers of particular languages may also be complemented by strong 2 Theoretical background Linguistic stereotypes, as already mentioned, tend to be complex structures, with several sets of ideas intervening and influencing processes of identifying and evaluating others and oneself. Language differences construct and maintain borders between different groups (ethnic, socioeconomic, professional, etc.) influencing relations between these groups, everyday interactions between their members, and their behavior toward each other (Ross 1975;Fought 2006;Rosenberg et al. 2015;Adamo 2018). In general, people tend to be less positively disposed toward members of the out-group. Thus, in many experiments, respondents have been shown to remember more negative information about out-group members and to be less ready to help them compared to in-group members (Brewer and Brown 1998). In some cases, language can be the main, or even the only, criterion for groups' separation, as e.g., with French-and Englishspeaking Canadians, or Finnish-or Swedish-speaking citizens of Finland. More often though, language differences constitute and are perceived as significant, but secondary, characteristics of a certain group distinguished by other criteria (race, ethnicity, country of origin, religion, etc.).
In the case of migrants, their linguistic performance tends to be evaluated by majority language speakers in terms of success or failure: those who adapt themselves well to the speech norms of the receiving society get better treatment and attract less negative attention than those who stick to the their own language. This traditionalist approach is rooted in the belief in essentialist cultural norms as a static target, a destination point for migrants who are moving in an otherwise static world in a "failed numerical attempt to reproduce this same static condition" (Nail 2019: 63). In reality, modern global urban centers, which attract people from all over the country as well as from abroad, are multilingual, super-diverse spaces (Vertovec 2007) where numerous language variants can be used on an everyday basis by speakers with complex linguistic repertoires (Pennycook 2010;Rindler Schjerve and Vetter 2012;García and Wei 2014;Rampton 2015). However, "monolingualism is woven into modernity's most minute and sophisticated political structures" (Gramling 2016: 3), and both government officials and ordinary citizens tend to follow a modernist approach to migrants. The one-nation-one-language ideology emphasizing uniformity and homogeneity continues to dominate mainstream discourses on language diversity (Piller 2015), making newcomers' ability to speak the national language a crucial condition for their adaptation, a measure of success.
Bilingualism (or multilingualism) of migrants, therefore, is not usually considered as an advantage since their mother tongues lack social prestige in the eyes of linguistic majority. Nonprestigious bilingualism (Deprez 1994: 32-3), as opposed to prestigious bilingualism of the elite mastering foreign languages (traditionally European -English, French, German, etc.), is seen in the so-called deficit paradigm (Castro and Prishker 2019) based on a language-as-problem rather than a language-as-resource approach (Faltis and Smith 2016). In Jan Blommaert's words, migrants' "multilingual repertoires are not recognized and certainly not used as existing and valuable linguistic communicative instruments," and their "linguistic resources are considered to be without value 'to live in our society'" (Blommaert 2010: 174). The dominance of the monolingual ideology does not mean that people do not know that bilinguals exist; it means that they are sure that it is normal to know and speak just one language. The term monolingual bias (or monolingual perspective or monolingual mindset) has been used since the 1980s in second-language acquisition research (Baratt 2018) to challenge the traditional practice of taking monolingualism as a starting point and modeling second-language learners as acquiring some ideal monolingual variety of target language (May 2014b). However, this approach is by no means limited to second-language teaching or even education in general. It exists (and is more obvious) in these spheres because it is inherent for many western societies relying upon the idea of a national state with a homogeneous population speaking one and the same language: "public monolingualism in the national language is simply taken for granted by its citizens" (May 2014a: 372). In this system of beliefs, learning prestigious foreign languages demands formal instructions and conscious efforts while speakers of nonprestigious languages are expected to shift to the national language as soon as possible. On a societal level, monolingual ideology works against all minority language speakers, both local ethnic minorities and migrants.
Monolingual bias, however, is not the only aspect of language ideology contributing to language prejudices. Standard language bias, emphasizing normativity and labeling language variation as an unwelcome deviation from the ideal forms, is extremely important. Standard language cultures are characterized by a firm belief in correctness: "It is taken for granted as common sense that some forms are right and others wrong, and this is so even when there is disagreement as to which is which" (Milroy 2001: 535).

Sociohistorical background
Considering the fact that monolingual and normative perspectives prevail in language ideology in Russia, it is not surprising that many Russian speakers, as we will show in this study, do not approve of any violation of the monolingual regime in public places. At the same time, rapid changes in ethnic and social composition of Russian megalopolises due to immigration from Central Asia, China, and other regions on the one hand, and powerful influence from the international globalized cultures (Hollywood films, rap music, K-pop, etc.) on the other, challenge these attitudes. Are there any visible changes in public discourse and the everyday reality of Russian cities that result from these influences? As Thomas Nail (2019: 57) puts it, reflecting on the image of a migrant and the primacy of mobility in the twenty-first century, "societies are not static places with fixed characteristics and persons;" they are "dynamic processes engaged in continuously directing and circulating social life." Is it possible to trace these dynamic processes, and what data could be relevant for that? In our study, we will try to go beyond the simple presentation of linguistic prejudices of Russian speakers and attempt to outline domains where the inevitable, though slow, erosion of the monolingual ideology can be observed.
Despite being a de jure and de facto multiethnic and multilingual country, Russia can hardly be considered a tolerant place that favors diversity and multiculturalism. Sociological studies expose rather high levels of xenophobia among Russian citizens (Gudkov 2006;Pain 2007;Herrera and Kraus 2013;Bessudnov 2016), with labor migrants from Caucasus, Central Asia, and China as its main target. Hostility to migrants and their behavior (including verbal behavior) can be witnessed both in sociological surveys and interviews and in public media discourse (Hutchings and Tolz 2015), where the terms tolerantnost' "tolerance" and mul'tikul'turalizm "multiculturalism" are often used as symbols representing weak political position of western countries capitulating to Muslim migrants or any other minorities. According to Shnirel'man (2011Shnirel'man ( , 2013, modern racism and xenophobia in Russia are a legacy of the Soviet regime: starting in the 1930s, despite declaring druzhba narodov "friendship of the people" as its main slogan, the USSR official policy and social sciences approached ethnicity from a primordialist perspective, which gave more rights to the indigenous population than to newcomers (Shnirel'man 2013: 102).
Modern Russian culture is notable for its normative orientation. After a relatively short period of what could be called linguistic anarchy in the first years after the October revolution, there was a rather sharp turn to language purism and standardization in the name of "language culture." Michael Gorham (2000) shows how this process was connected with Vladimir Lenin's position and, more generally, with the efforts of the new state to extend its power and control to all spheres of people's lives. This trend dominated for many decades (see, e.g., the detailed description of the total dominance of a normative approach in Soviet lexicography in Perelmuter 1974), and the term "language culture" became one of the milestones of Soviet linguistics. The idea of speakers being characterized by different levels of speech culture or speech competence (high, average, or low) was the closest approximation to social stratification found in a country emphasizing social equality of its citizens. Another short period of relative linguistic relaxation occurred in the 1990s after the fall of the Soviet Union but, starting from the year 2000, the trend reversed again, when the state started to finance many projects aimed at language regulation and propagandizing Russian language and culture (Gorham 2012).

Methods and data
This study aims at two goals: first, to describe the mainstream language ideology that affects language choices made by the Russian speakers of the two Russian main cities, Moscow and St. Petersburg, in their communication with nonnative speakers; second, to discover the evidence of gradual changes in the ideology of monolingualism. Language ideology, as a set of ideas about social values of linguistic facts, or "sociolinguistic imagination" (Milani and Johnson 2010: 4), cannot be directly observed or measured (Kroskrity 2004); however, it can be studied via its manifestations in discourse and practices. Both what people say about language and what they do with it are important; consequently, using different kinds of data and methods of analysis is crucial: "The study of language ideologies benefits from a methodological approach that intentionally integrates multiple expressions of data" (Bacon 2020: 174).
For the purposes of this study, we used two sets of data analyzed separately to reveal different aspects of the process of ideological change. The first type of data provided us with access to what can be called meta-discourse on language attitudes; it reflects the way people discuss and contemplate language matters and communicational practices. Main sources for this meta-discourse were interviews on multilingual language practices and Russian speakers' attitudes toward migrants' speech.
The interviews with migrants and locals were conducted by researchers and students of the Higher School of Economics, St. Petersburg, in 2018. The first group of respondents was composed of 20 migrants from Tajikistan, Kyrgyzstan, Kazakhstan, Armenia, Azerbaijan, and China; another 21 interviews were conducted with native Russian speakers (mostly highly educated middle-aged females) residing in St. Petersburg. To reveal the recent changes in opinions on language expressed by locals and migrants and in reported communicational practices, we compared this recent data with earlier findings based on the interviews with migrants collected in 2009-2010 in Moscow and St. Petersburg in the framework of two research projects on migration conducted by the Sociology of Education and Science Laboratory at the Higher School of Economics, St. Petersburg, and reported in Baranova (2014).
As an additional source of meta-discourse data, we used different Internet publications dealing with the same issues of language attitudes and interethnic communication. To collect a corpus of such texts, we had been regularly monitoring eight news portals and public groups in social networks Facebook and Vkontakte (the most popular social network in Russia and many post-Soviet countries; it has numerous groups for locals to share news, pictures, and opinions such as, e.g., Podslushano Metro Moskvy ("Eavesdropped [on] Moscow Metro")). We aimed for publications and comments discussing the influence of migration on urban linguistic landscapes and soundscapes (i.e., written and oral communication in public places). The resulting corpus (around 150 texts written in 2015-2020), therefore, includes comments, mainly anonymous, provoked by language-related news such as, for example, the appearance of new McDonald's ordering screens in Uzbek, Kyrgyz, and Tajik. Certainly, such small corpus cannot be used for quantitative analysis; however, it is valuable for providing graphic examples of expression of language attitudes and for revealing some trends in related internet discussions. All quotations from the meta-discourse data are given in our translations into English and, in the case of interviews, supplied with information about respondents' gender, age, and ethnic identity or country of origin.
Our second set of dataon linguistic landscapereflects everyday reality of language use in public urban space. Linguistic landscape studies, i.e., analysis of signs, advertisements, announcements, and every other case of written language use in the streets, in shops, restaurants, offices, schools, governmental institutions, etc., proved to be an important instrument not only for evaluating existing language diversity but also for understanding power relations between different languages and groups of their speakers (see Shohamy et al. 2010;Blommaert 2013). We started observing St. Petersburg's linguistic landscape regularly in 2016 and also collected some data during fieldwork and short trips to some other cities, including Moscow. In 2018, in cooperation with the University of Manchester, we launched a crowdsourcing project "Multilingual St. Petersburg." Via a mobile application designed and shared by the team from the "Multilingual Manchester" project and localized by Vyacheslav Barinov at the Higher School of Economics, St. Petersburg, users can now take geotagged photographs of multilingual signs and upload them to the project website (https://linguasnapp.hse.spb.ru). After premoderation, these photographs are put on the interactive map of the city alongside metadata (type of sign, where it was placed, number of languages, number of scripts, translation, whether there is a dominance of one language over other(s), etc.), where they can be browsed and sorted with the help of a search engine. This method of obtaining data has certain limitations since researchers have very little control over the process, and some photos can lack crucial contextual information. On the other hand, crowdsourcing provides researchers with a lot of opportunities, not only to collect much more material but also to cover districts and places difficult to access (for more details, see Gaiser and Matras 2016), something which is especially important in order to collect representative data on megalopolises. The current number of photographs gathered via the LinguaSnapp SPB mobile application exceeds 1,000; for this article, we used around a half of the data set and did not take into account signs in English and other major European languages designed for tourists. We also used around 200 photographs taken manually in St. Petersburg in 2016-2019 (before the launch of the mobile application or when it was unavailable) and 70 photographs taken in 2016-2019 in Moscow by us and our Moscow colleagues. We analyzed the linguistic landscape data in relation to the representation of migrants' languages as evidence of changes in the domineering language regime.

Language attitudes data: double catch of negative stereotypes
In this section, we will try to answer the question: What are the attitudes to communication in languages other than Russian in public places among native Russian speakers in St. Petersburg and Moscow? Interviews with locals in St. Petersburg conducted in the course of research supervised by one of the authors show predominantly negative attitudes to non-Russian languages sounding in public space, especially to those associated with Central Asian countries (Maksimova 2018). When asked about the soundscape in their district, i.e., different sounds and the sources of noise, most respondents often mentioned that migrants speak too loudly in the streets.
But there are cases when you go out or it is coming [from the street] through the window and the conversation is, I don't know, Uzbeks or not Uzbeks or whoever speaks but it is Asian speech. And it's very uncomfortable. Very-very much. And they are very chatty and either on cell phone bla-bla-blah, or together [with other speaker]. (female, 47, Russian) (Maksimova 2018: 43) The respondents were also asked what languages heard in their district that they could recognize (without understanding the conversation content); almost none of them could indicate specific languages (Azeri, Tajik, and Uzbek), and instead often unified them under the label of "Central Asian" or "Asian speech," as in the quotation above. The idea of a soundscape with elements of nonintelligible languages is described as unpleasant and disturbing by locals: Sometimes it feels like we do not live in Russia but in some other country because often you hear such talk. (female, 36, Russian) (Maksimova 2018: 43) Very similar attitudes are expressed in the internet posts and comments related to different issues of migration and its impact upon everyday life in Russian cities: It is so annoying when you walk along the center of the capital [Moscow], and behind your back they mumble in their language and are evidently happy no one can understand them. (Anonymous 1, 2016) After all, it's simply unethical when labor migrants communicate in their language! We can't understand them, and many conflicts and confusions result from that! (Anonymous 2, 2017) In all similar examples, even when resorting to logical arguments about possible miscommunication, speakers are frustrated by the very fact that they have to hear something they cannot understand although they are not actually addressed. It is significant that native speakers of other languages are portrayed as malicious and ill-intentioned: their private communication in their own languages is seen as an aggressive encroachment on Russian speakers' peace of mind.
Another striking feature of such comments is the active use of different devaluating and even insulting expressions when referring to migrants' languages. They are called тapaбapщинa, "mumbo jumbo," and when referring to migrants' act of speaking them, instead of гoвopить, "to speak," verbs such as лoпoтaть, "mumble," or бopмoтaть, "babble" are used in order to accentuate the incomprehensibility. Other verbs defining migrants' mode of speech underline its inadequacy when compared to "civilized" (meaning Russian) norms of public communication: "like savages they bawl in their [language] all over the street" (Anonymous 3, 2016).
These speakers' attitudes to migrants' languages spoken in public, therefore, seem to be overtly negativethey would prefer to live in a purely monolingual space or at least in a place where there are no "undesirable" languages. The desire to control language choices for communication in the public space or in working places of migrants is a clear case of linguistic prejudice. Moreover, there are indeed attempts to implement such control. In interviews with migrants, there are numerous mentions of cases when business owners or managers demand workers to use exclusively Russian, even when communicating with coworkers speaking the same mother tongue. Even when not explicitly required, the respondents felt that communicating in Russian was expected from them. Submitting to these unspoken rules looks like implementing self-control. Thus, a cleaning lady in a student hostel mentions that she speaks Uzbek with her coworkers and friends, but in the presence of their boss and other administrative staff, they try to keep silent and not say anything. Sometimes, however, there is an explicit rule. In personal communication, Valentina Chupik, an activist who provides juridical help for labor migrants, shared an appendix to the employment contract used by some businesses in Moscow. This appendix listed possible misdemeanors and penalties for committing them. In the same way as such actions as being late or absent from work were penalized, "speaking foreign languages in the working place" would result in a fine of 1,000 rubles, a rather significant sum for labor migrants who do not earn much.
It should be noted that communication between migrants and the ethnic majority is hierarchically structured. According to interviews conducted in 2009-2010, not only employers but also doctors in hospitals or schoolteachers could advise migrants to use only Russian, e.g., in communication between parents and children, or in other ways try to control migrants' language choices (Baranova 2014: 217-9). Even strangers in public transport or customers in shops can make negative comments when hearing non-Russian languages.
However, according to our data, the choice of language for communication in working places has become more diverse and pragmatic in the last decade. There are such options as Standard Russian, ethnolects of Russian, native minority languages, or Uzbek as a lingua franca for any Turkic-speaking migrants. Comparing two sets of interviews with migrants, we can see this slight shift. In interviews with migrants conducted in St. Petersburg and Moscow in 2009(analyzed in Baranova 2014, references to negative attitudes of Russian speakers prevailed. Almost 10 years later, in 2018, we returned to the same questions in interviews with migrants in St. Petersburg (there were both newly arrived migrants and people who had been living in Russia since 2000s among them). According to these recent interviews, language control has become less oppressive. The choice of language for communication in the workplace is described by some of respondents as more diverse and optional: Certainly, language choice correlates to the level of proficiency in Russian among migrants. In the last two decades, the number of migrants in St. Petersburg and Moscow has increased. More importantly, newly arrived people either do not know Russian at all or can speak very limited Russian since the younger generation in post-Soviet countries is less exposed to Russian in education and in the mass media. It should also be taken into account that many of the former migrants, who were employees when they first arrived to the new country, have now started businesses of their own and employ other migrants. In many cases, however, such ethnic entrepreneurship does not really lead to full ethnic and linguistic homogeneity. For instance, the speaker of the last quotation, D., works as a cook in a Georgian restaurant whose owner and some administrative staff members are Georgian, but most workers are Uzbek-speaking people like our respondent himself. According to our observations, the same pattern is found in other small Georgian bakeries in St. Petersburg. Another respondent (female, 23, Tajik, a cleaning lady) told us that in her previous workplace, a Georgian café, there were Azerbaijani and Georgian cooks, Georgian managers and both Georgian and Russian waiters. In such a place, there is not one prescribed language determined by the dominant group since the language of both most customers (Russian) and the owners (Georgian) could be considered as equal options. These situations promote more fluid and diverse language choices including other native languages, especially among people who know more than two languages, like D. (native Uzbek born in Kazakhstan) who speaks Uzbek, Russian, and Kazakh. In terms of language, these small ethnic enterprises resemble the fluid multilingual context of metrolingual practices in an Australian bistro, as described by Pennycook and Otsuji (2016). The communication in such a context is not based on monolingual ideology; however, in other contexts monolingual bias still prevails.
Many respondents mentioned negative attitudes to their speech expressed by locals. Moreover, some migrants adopt and share such negative attitudes to nonnative pronunciation or grammatical peculiarities. This may be a manifestation of the common language ideology that is still powerful in post-Soviet countries or an acceptance of external negative categorization. For example, one respondent from Kyrgyzstan describes other Kyrgyz' knowledge of Russian and her attitude to their speech in the following way (her own Russian is perfect and she is very proud of this fact): Only a handful of people know Russian. If they speak, they have an accent, a strong accent. That annoys me very much, frankly speaking. (female,40,Kyrgyz) In some cases, such strict negative attitudes to accented Russian force nonnative speakers to limit their career options. For instance, M., a native Georgian, a refugee from Abkhazia since 1993, was a teacher of physics and mathematics. In St. Petersburg, she had been working at a Georgian school teaching mathematics. The school coexisted with a state school with standard curriculum and provided afternoon classes for Georgian-speaking children focusing on their native language and culture. There were Georgianspeaking teachers using both Russian and Georgian during their classes. In 2003, the school merged with the state school, and its teachers had to teach Russian-speaking children as well. From our point of view, M's level of Russian is absolutely sufficient for teaching her discipline. Still, after 25 years in Russia, she has a recognizable accent. Nevertheless, she clearly remembers her efforts to speak Russian just after coming to Russia: "But when we came here, what use did we have of Georgian, we needed Russian! And I myself was speaking Russian badly but I tried anyway." Probably influenced by these hurtful memories, M. evaluates her Russian very negatively and she is very unsure about her ability to teach in Russian: I was offered to move [to a Russian school from the Georgian one] but I refused immediately since my Russian was not of the highest level, how could I teach classes ? With Georgian [children] it was ok with an accent but what would Russian children make of me? (female, 56, Georgian from Abkhazia) It should be noted that the respondent here refers to possible external evaluations of her Russian by her students and not to some presumed difficulties with making documentation in Russian or using specific terminology. Such stories reveal that Russian with an accent is not just a low-prestige linguistic variety. It can stigmatize people and can be more decisive than their professional qualification. Standard language bias stimulates representation of migrants as incompetent and comic figures.
One aspect of language prejudice widespread among the Russian-speaking population of megalopolises concerns Russian spoken with an accent or other cases of nonstandard varieties usage. Even welleducated local residents tend to mention "correctness" of speech as an important factor for evaluating all service workers. For example, numerous videos of them speaking "bad Russian" were uploaded to www. youtube.com by both Russian speakers and migrants themselves with derogatory comments. At the same time, migrants speaking almost without an accent often receive positive comments directly relating their intelligence with an ability to speak "pure Russian:" Kyrgyz boys are so good. Nice appearance, well-groomed, in clean clothes, and speak without accent. (Anonymous 4, 2018) Kyrgyz are good boys. And they speak Russian well. (Anonymous 5, 2018) Monolingual bias and standard language bias prevailing among Russian-speaking city dwellers, therefore, work like a double bind for newly arrived people preventing them from speaking their native languages, on the one hand, and from trying to improve their Russian through practicing it, on the other hand. As one of our respondents formulated it: "[they tell us] you are not in your own home, so you should speak Russian. But if we speak Russian they do not like it either" (female, 42, Kyrgyz).
However, some traces of changes are found in this ideology among both young locals and migrants themselves. The frustration expressed by the respondent in the last quotation shows that external negative stereotyping is not always acceptable for nonnative speakers, and they may try to oppose it. The fact that the voices of multilingual people living in Russian megalopolises have become increasingly audible in the streets in recent years is mentioned by every respondent in Maksimova (2018). And a similar increase in representation can be found in the linguistic landscape. (1) The real linguistic diversity of the city is underrepresented in its linguistic landscape. Only two foreign languages spoken by labor migrants, Chinese and Uzbek, are relatively regularly used in public written communication. The former is also targeted to Chinese tourists who make up a significant part of St. Petersburg's tourist flow.
Uzbek signs are written mostly in Cyrillic letters while Uzbek in Uzbekistan, after an alphabet reform in 1993, is still slowly undergoing the process of shifting from Cyrillic to Latin script (see Fierman 2009;Landau 2010).
(2) The languages of migrants are mainly used in a rather restricted number of domains, mostly associated with services demanding bodily proximity and intimacy: medicine, cosmetic procedures, and sexual encounters.
(3) The overwhelming majority of signs in migrants' languages belong to in-group communication, i.e., are both produced by and directed toward migrants themselves. Russian speakers address migrants in Russian, expecting them to understand it to a certain level or simply ignoring their possible linguistic problems. This is true for both state officials and private business representatives, even those who deal with migrants on regular basis (agents helping to obtain working permits, money transfer services, etc.). Ethnic shops and restaurants aiming to attract Russian customers tend to limit their use of migrants' languages to purely symbolic functions (e.g., in brand names or via stylizing Cyrillic script to look like Chinese characters or Arabic letters). (4) Signs in migrants' languages are unevenly distributed in the city's space; they are much more likely to be present in marginalized districts in the outskirts of the city currently under development, or around cheap markets such as Apraksin dvor. Even there they tend not to be found in the open but rather in places hidden from the eyes of the locals: inside migrants' shops and cafes, in inner parts of markets, near construction sites, etc.
Moscow's data from 2016 to 2017 (Fedorova and Baranova 2018) demonstrated the same tendencies: underrepresentation of migrants' languages in the linguistic landscape and almost total disregard for migrants' linguistic needs on the part of the Russian-speaking majority.
Analysis of the data obtained in 2018-2019 confirmed these previous findings. However, it also revealed some slight changes in the direction of a greater visual linguistic diversity. In what follows, we will focus mostly on the trends revealed by St. Petersburg data, which is better documented so far due to the launch of the aforementioned crowdsourcing project. Recent Moscow data confirm these trends; we will refer to similar cases where we have concrete evidence.
First of all, there is some increase in the number of languages represented in St. Petersburg's linguistic landscape. Among these new additions are, e.g., Vietnamese, Dari, Korean, Serbian, Tadjik, and Kyrgyz. The cases where the first two languages are used are completely in line with the trends described above. On the signboard outside Vietnamese café in Apraksin dvor (Figure 1), the Vietnamese language is presented on a merely emblematic level, while inside it dominates in-group communication, with menus in Vietnamese and TV broadcasting Vietnamese programs (the same situation can be observed in a Chinese café nearby described in Baranova and Fedorova 2019). Another typical example of in-group communication is a handwritten announcement in Dari (Figure 2) placed on the wall inside the same Apraksin dvor. It informs Afghans about the death of their fellow countryman and the date of his memorial service. However, there are other cases, going beyond in-group communication, and they need to be discussed in detail.
The emergence of Korean in St. Petersburg's linguistic landscape can largely be explained by the growing fascination with Korean culture in Russia due to the popularity of K-pop, Korean TV series, and cosmetic products. Alongside the purely conventional use of Korean (as well as English and Chinese) in the advertisement of Korean restaurants (Figure 3), less obvious examples are available. Figure 4 represents an advertisement of Кopeя мapкeт, "Korea market," a shop specializing in Korean food, cosmetics and home supplies. References to Korean culture (as it is represented in historical TV dramas) are expressed through the images of a wooden palace, a girl in hanbok, a traditional Korean costume, and a red and blue yin and yang symbol inside the letter "o." There are also several interesting aspects of language use in this case. First, the very name of the shop is a result of linguistic hybridization: the English word "market" written in Cyrillic is used instead of the Russian мaгaзин, "shop" or pынoк, "market," and the use of a noun in nominative case (Кopeя, "Korea") in attributive function rather than a combination of an adjective and a noun (кopeйcкий мaгaзин, "Korean shop") breaks the rules of Standard Russian grammar and can be found only in some recent loans from English (e.g., пpoдaкт-мeнeджep or пpoдaкт-мeнeджep, "productmanager"). The same hybridized name is not translated but transliterated into Korean, making it hardly comprehensible to Korean speakers: in Korean, the word "Korea" sounds as 한국 [hanguk], while the word borrowed from English (마트, [maty]) is normally used to mean "a shop." On the other side of the same billboard, there is a Russian version of the same advertisement; however, due to all the hybridization this case looks more complicated than a simple split of information between potential Russian and Korean customers. The website and Vkontakte group of this shop are, undoubtedly, maintained by and addressed to Russian speakers. However, they both regularly use Korean etiquette phrases 안녕하세요, "hello" and 감 사합니다, "thank you." In a similar way, inside the restaurant from Figure 3, menus are provided in Russian  and English with most, but not all, dishes also represented in Chinese and Korean. Translanguaging, we can conclude, becomes more acceptable and even fashionable, and references to language diversity gain marketing value.
This observation is confirmed by other cases where pragmatically unnecessary use of foreign languages indicates more positive attitudes to multilingualism. In the Serbian restaurant Serbish, e.g., clients can entertain themselves with Serbian textbooks or by solving linguistic puzzles: they have to guess Serbian equivalents for Russian words, and there are favorable mentions of this in some reviews on Google Maps and TripAdvisor.
All languages discussed so far are unusual and exotic to Russian speakers, even Serbian, despite its being Slavic and using Cyrillic letters. The use of such languages by Russian speakers for commercial  purposes of ethnic marketing does not necessarily presuppose actual communication with their speakers. Tajik and Kyrgyz, as well as other languages of labor migrants, are another matter. As mentioned above, Russian speakers, especially authorities, have regularly addressed migrants exclusively in Russian, simply ignoring their communicative problems. Rare attempts of businesses to attract foreign clients by using their languages were limited to presenting some words or phrases purely symbolically while all practical information were provided only in Russian (see Baranova and Fedorova 2019: 24-5). Recently though, at least in some cases, Russian speakers have started to use migrants' languages as an actual communication tool.
One of the most prominent examples of such meaningful multilingual communication with migrants is the emergence of migrants' languages in the area of the one-stop center for paperwork on Krasnogo Tekstilshchika street in St. Petersburg. There are several divisions occupying separate buildings, two of which deal with Central Asian migrantsthe official center for registration of international migration (it also issues working patents and residence permits) and the medical center issuing vaccination certificates, providing HIV tests, etc. A home-printed note inside the medical center provides instructions in Uzbek for preparing for an X-ray screening. What is more, the official signs at the entrances to the different buildings are trilingual: in Russian, Tajik (Cyrillic), and Uzbek (in the official Latin script). Interestingly, one such sign ( Figure 5) demonstrates a sincere attempt to achieve understanding: in Russian it says дaктилocкoпия "dactylography," whereas the Tajik and Uzbek variants, instead of simply translating scientific term, possibly incomprehensible for many migrants, read "to register fingerprints." At the same time, the website of the same center is still purely monolingual in Russian. Moreover, governmental websites, e.g., that of the Ministry of Internal Affairs, responsible for the registration of migrants, have only functionally limited versions in English and nothing in other languages.
Other cases of attempts to communicate with migrants more effectively, this time on the part of businesses, include recent advertisements for jobs using migrants' languages. They usually concern unskilled jobs such as house cleaner and warehouse worker. Such signs can use Uzbek (in Cyrillic, Figure 6) or even three languages, namely, Kyrgyz, Uzbek, and Tajik (Figure 7). Another piece of evidence of the growing multilingualism is the addition of interfaces in the abovementioned three languages to some ATMs located near the biggest national bank Sberbank and to McDonald's ordering screens.
Certainly, all such examples of multilingual communication are still mere isolated cases, and most texts addressed by Russian speakers to migrants are in Russian. For example, agencies providing translating services do not use other languages than Russian in their advertisements and on their websites (actually, very few of them have websites, even in English). Even firms specializing in legal services for migrants do not use Uzbek or Kyrgyz. The linguistic innovations by McDonald's in Moscow provoked heated discussions on Facebook under posts with images of screens that showed a choice between six languages: Russian, English, Chinese, Kyrgyz, Uzbek, and Tajik, marked also with the images of their respective national flags. There were many negative and overtly xenophobic comments: some people were furious about the fact that they had to see migrants' languages having the same rights as Russian and "proper" foreign languages. One of the most revealing comments was "it is they who have to learn the language and not we who have to accommodate!" (Anonymous 6, 2020). In other words, seeing migrants' languages reflected in public space is, in this perspective, an equivalent to serious inconveniences or even hostile conditions demanding special efforts to accommodate on the part of local citizens.
However, there are (a few) instances in our material which express a different approach to non-Russian speakers. These examples are important since they attest to an increasing tolerance toward migrants' languages in public space. More specifically, in the same online discussions as to referred to above, there were other comments favoring a multilingual approach, shaming xenophobia, and even demanding more  appropriate representation of migrants' languagesnow, unlike English and Chinese, which are represented as English and 中文, the names of these languages are given in Russian.
Another easily perceptible (and much discussed) change in both Moscow's and St. Petersburg's linguistic landscape is the rapidly growing presence of signs in Chinese meant for tourists from China. Unlike signs and advertisement meant for Chinese labor migrants, these tourist-oriented signs are placed where everyone can see them and often advertise products and services of premium class. Although there is a gap between the luxury brand advertising for Chinese tourists at Sheremetievo or Pulkovo airport ( Figure 8) and monolingual signs in Chinese hair-dressing salons meant for the local Chinese workers and students (Figure 9), the visual presence of Chinese in a way legitimizes its use in public space. Moreover, the emergence of new signs in Chinese can be viewed as evidence of some Chinese migrants becoming wellestablished middle-class citizens: for instance, there are advertisements in Chinese in real-estate agencies ( Figure 10).
Summing up, the recently gathered data on the linguistic landscape reveal certain tendencies. New languages and, consequently, new groups associated with these languages, are becoming more visible in the public space. The linguistic landscape is still far from a fair representation of all languages actually spoken in St. Petersburg, but it is definitely much more diverse now than it was 2 or 3 years ago. Even more important are the changes in agentivity: while migrants' languages were used almost exclusively for ingroup communication previously, there are now cases when Russian speakers attempt to communicate with migrants taking into account their linguistics needs.

Conclusion
In this study, we have addressed the issue of language ideology shared by the citizens of Russian megalopolises through its manifestation in two data sets. On the one hand, we have analyzed meta-discourse on interethnic communication revealed in interviews and language-related internet comments; and on the other hand, representations of multilingual practices in the linguistic landscape. In both spheres, a normative monolingual approach prevails: many people feel that it is appropriate to express very negative feelings toward languages other than Russian when they are spoken or written publicly, and they are eager to implement control over the verbal behavior of other people, especially those who they regard as a threat. Interiorizing this language ideology, vulnerable groups try to be as invisible as possible: they refrain from speaking their languages in public places, restrict their communication with Russian speakers to a necessary minimum, and conduct their written in-group communication, as any other private activities, in closed safe spaces hidden from the eyes of the linguistic majority.
At the same time, no discourse can be entirely monolithic, and language ideologies should not be understood as fixed and stable (Rosa and Burdick 2017). As we have tried to show, there are some traces of an ongoing change in language attitudes and beliefs among Russian speakers in our data. We can assume that the monolingual ideology in Russia, challenged by the growing linguistic diversity and visibility of different groups in the urban space, will inevitably start, though slowly, to give way to multilingualism.  This process is reflected both in new signs appearing in the linguistic landscapes of two Russian main cities and in the turn to the more practical and reasonable language choices for communication in the public sphere and in workplaces. All these tendencies lead to lessening the pressure exerted by locals upon newcomers and to an increase in acceptability of their languages. Voices opposing the monolingual regime grow louder every year. Presumably, at least some of those who enjoy the privileges of being native speakers and belonging to the ethnic majority "begin to view the marginalized as persons who have been treated unjustly, whose voices have been silenced, and who have been cheated by society" (Endo and Reece-Miller 2010: 96). The changes in language attitudes discussed in this study are the first "cracks" in the monolingual facade, the signs of its inevitable defeat in the face of the indisputable diversity of people's lives.