Beyond Nations and Nationalities: Discussing the Variety of Migrants’ Identifications in Russian Social Media

This article examines how transnational labor migrants to Russia from the five former Soviet Union countries – Azerbaijan, Kazakhstan, Kyrgyzstan, Tajikistan, and Uzbekistan – identify themselves in social media. The authors combine Rogers Brubaker’s theory of identifications with Randall Collins’ interaction ritual theory to study migrants’ online interactions in the largest Russian social media (VK.com). They observed online interactions in 23 groups. The article illuminates how normative and policy contexts affect the Russian Federation’s migration processes through a detailed discussion of migrants’ everyday online interactions. Results reveal common and country-specific identifications of migrants in their online interactions. Migrants from Kazakhstan and Azerbaijan employ identifications connected to diasporic connections. Migrants from Kyrgyzstan, Uzbekistan, and Tajikistan in their identifications refer to low-skilled labor migration to Russia as a fact, a subject for assessment, and as a unifying category. For these countries, the present and the future of the nation is discussed in the framework of evaluation of mass immigration to Russia.

Introduction 326 Natalia D. Tregubova, Maxim L. Nee These five countries share two important characteristics as the countries of origin. First, they are countries with predominantly Muslim populations. Second, migrants from these countries are predominantly "visible" migrants. Religion is one of the key characteristics used for the identification of migrants in post-Soviet Russia (Sokolov, 2017;Turaeva, 2019). Visibility is highly significant for migrants' perception by the Russian officials and the public in general (Chandler, 2011;Sokolov, 2017).
At the same time, citizens from these countries experience different official regulations and migration policies in Russia. Kazakhstan and Kyrgyzstan are the member states of the Eurasian Economic Union (EEU), so their citizens enjoy a privileged access to Russia's labor market in comparison with migrants from Azerbaijan, Tajikistan and Uzbekistan. Moreover, Kazakhstan is both a sending and a recipient country. It is a place of destination for many labor migrants from Kyrgyzstan, Tajikistan and Uzbekistan (Laruelle, 2013). Finally, Azerbaijan, Kazakhstan, Kyrgyzstan, Tajikistan and Uzbekistan present regional diversity at both inter-state level (Central Asia and Caucasus) and sub-state levels (Fergana and Pamir regions, among others).

Literature Overview
Research materials that have already been published characterize a variety of ways of how migrants identify themselves and how they are identified by others in the fSU countries. Identification could be based on nationality (ethnicity), religion, citizenship, the region inside the state, the larger region (Central Asia, Caucasus), type of work, etc. Some of these characteristics are directly inherited from the Soviet time, or even in the earlier periods, some became relevant only after the USSR's breakdown. Different actors in different post-Soviet states have used different combinations of these characteristics in their claims to construct their identity.
In Russia, labor migrants identify themselves in various ways that extend beyond nationality and citizenship. Sometimes their identifications are related to labor: migrants identify themselves as "hard workers" (Ni & Lisitsyn, 2017). Another identification is religion intersected with nation and region in various ways, for example as "Central Asia Muslims" (Turaeva, 2019) or "Tajiks are the strongest believers" (Roshe, 2018). A person from a local village/mahalla might also be a kind of identification (Urinboyev, 2017). Migrants' identifications are influenced by the local residents' attitudes towards migrants' connections to ethnic hierarchies intertwined with job hierarchies, visibility, Russian language proficiency, affiliation to a broader ethnicity (such as being Slavs), and the region of origin. Kyrgyz migrants perceive their ethnic status as low but superior to Tajik and Uzbek migrants and associate their superiority with proficiency in Russian language (Gerber & Zavisca, 2020), while some Tajik migrants compensate their low status by positioning themselves as pious Muslims (Roshe, 2018).
Migration research generally addresses three main topics: movement, control, and settlement (Kivisto & Faist, 2010). Movement refers to the aims and patterns of migration. Migration to Russia from Azerbaijan, Kazakhstan, Kyrgyzstan, Tajikistan, and Uzbekistan is primarily labor migration, as we have already noticed. Labor migrants in Russia are oriented towards paychecks. They commonly face dequalification, informal employment, and poor labor conditions, while wages received by the Russian and migrant workers are comparable (Mukomel, 2017). However, migrants' aims may eventually be transformed: migrants return to their homelands and then move back to Russia, labor migration entails family migration and so on (Brednikova, 2017). Migration policies developed in the Russian Federation are strict and asymmetric: even a minor break of the legal rules means that a migrant has little chance to be legalized again (Kubal, 2016). Moreover, migrants often do not have a clear understanding of the Russian legislation (Varshaver, Rocheva, & Ivanova, 2017). A gap between formal and informal institutions leads to corruption and proliferation of migration-related businesses (Malakhov, 2014). Struggles for getting the proper documents are an important part of migrants' everyday practices and interactions (Reeves, 2019b). These problems, however, are less significant for migrants from the member countries of the Eurasian Economic Union. For them, the regulations and requirements to stay and work in Russia are more relaxed (Mukomel, 2017).
The issues of settlement concern migrants' incorporation into host societies. Migrants from Azerbaijan, Kazakhstan, Kyrgyzstan, Tajikistan and Uzbekistan are partly transnational migrants. Transnational social formations "consist of combinations of social and symbolic ties and their contents, positions in networks and organizations, and networks of organizations that cut across the borders of at least two nation states" (Faist, 2013, p. 450). Transnational ties are typical for migrants from Kyrgyzstan, Tajikistan and Uzbekistan (Abashin, 2017;Brednikova, 2017;Varshaver & Rocheva, 2014) and occasionally occur for migrants from Azerbaijan (Braux, 2013;Pinchuk & Minyazhev , 2017) and Kazakhstan (Ryazantsev, 2016;Safonova, 2008). Migrants to Russia engage in transnational social formations by staying in touch with their relatives at home and with compatriots abroad, sending remittances, visiting home, presenting gifts to their friends in their country of origin and in Russia, and so on. Interactions with the country of origin and with compatriots in Russia are highly important for starting the migration process as well as for the settlement in a new environment (Abashin, 2017;Lisitsyn & Stepanov, 2019;Pinchuk & Minyazhev, 2017).
Thus, research findings on movement, control and settlement of migrants to Russia reveal three crucial sources for identification. They are class (labor status), citizenship and nation, respectively.
Interactions via social media constitute one of the most important sources of information and emotional support for migrants living in Russia. Social media are characterized by "scalable sociality": "Social media [colonize] the space between traditional broadcast and private dyadic communication, providing people with a scale of group size and degrees of privacy that we have termed scalable sociality" (Miller et al., 2016, p. 9). Migrants to Russia use social media to stay in touch with relatives and friends in their homeland and in Russia (Ruget & Usmanalieva, 2019;Schröder, 2018;Urinboyev, 2017), as well as to exchange information and goods with strangers (Timoshkin, 2019). This scalability, together with the ubiquity of online connections make interactions on the social media of particular interest for studying migrants' identifications. They also present the kind of data that is underappreciated in the studies of migrants' everyday lives in Russia, so far dominated by interviews and observation.

Theoretical Framework and Research Questions
This article looks at how inherited social categories and related symbols constitute resources for migrants' identifications in online interactions. To analyze how ethnic, national, and other categories are used in online interactions of migrants, we combine several conceptual and theoretical sources.
One source is the theory of ethnic and national identifications proposed by Rogers Brubaker (2004): "As a processual, active term, derived from a verb, 'identification' […] invites us to specify the agents that do the identifying" (p. 41). For Brubaker, to study ethnic, national and other identifications means to answer the following questions: How do people identify themselves? How do other people identify them? In what types of situations do ethnic, national and other social categories become resources for identifications? Answering these questions demands knowledge of historical and cultural contexts as well as details of everyday social encounters. Brubaker provides two key distinctions: (a) between self-identification and identification by other actors, and (b) between relational and categorical modes of identification. Relational identifications refer to the participation in a web of social relations; categorical identifications refer to the "membership in a class of persons sharing some categorical attribute (such as race, ethnicity, language, nationality, citizenship, gender, sexual orientation, etc.)" (Brubaker, 2004, p. 42). Our research focuses on migrants' selfidentifications and external identifications of the categorical mode.
We base our argument also on the concept of "everyday nationalism" proposed by Paul Goode and David Stroup (2015). We focus on the "quotidian practices by which ethnic and national identities are elaborated, confirmed, reproduced, or challenged" (Goode & Stroup, 2015, p. 718) in contrast to institutional and discursive identifications in official documents, political texts, mass media, etc. Everyday/"private" aspects of nationalism in contemporary Russia seem to be quite distinct from the institutional ones and thus constitute a subject of special interest (Goode, 2017). However, our unit of analysis is not everyday practice as such, but the situation of interaction in the social media. Thus, the paper has a micro-sociological focus in contrast to the anthropological approach taken by Goode. To study migrants' interactions, we rely on the interaction ritual theory (IRT) by Randall Collins (2004). The IRT provides guiding principles for studying how cultural symbols gain relevance in interaction. The concept of the interaction ritual embraces a whole spectrum of interactions, from everyday talks to ceremonies. Interaction rituals, according to Collins, have four ingredients and four outcomes connected by a specific situational mechanism (Collins, 2004, pp. 47-49). Ritual ingredients are the group assembly, the group boundary to outsiders, the mutual focus of attention, and the shared mood. "As the persons become more tightly focused on their common activity, more aware of what each other is doing and feeling, and more aware of each other's awareness, they experience their shared emotion more intensely, as it comes to dominate their awareness" (Collins, 2004, p. 48). Persons engage in common rhythms; rhythmic entrainment produces intersubjectivity and, for successful rituals, evolves into collective effervescence (Emile Durkheim's term). Ritual outcomes are group solidarity, the emergence of group symbols, the standards of morality towards the group and its symbols, and emotional energy in individuals. The latter is a feeling of enthusiasm about the interaction that pushes individuals to participate in similar interactions in the future. The importance of symbols for individuals tends to fade away gradually; hence, symbols need to be recharged in new interactions. Thus, according to Collins, interactions tend to form patterns in time, interaction ritual chains.
Additionally, we apply the concept of attention space also developed by Collins (1998;. Attention space characterizes symbols and ideas that are relevant for a community and are used in its interactions and discourses. There are two important observations about attention space: it is limited, and it is structured. First, several symbols are at the center of attention while others are employed marginally, in an episodic way. Second, symbols and ideas are interrelated: some are in opposition, some are in affinity. The concept of attention space helps to combine Collins's and Brubaker's theoretical perspectives. Who and what situational mechanisms make limited attention in interaction focused on the ethnic, national, religious or other social categories? This question combines Collins's interactional analysis with Brubaker's interest to mechanisms of group formation. To analyze online interactions of migrants as interaction rituals we rely on conceptual and methodological adaptations of the IRT for studying online interactions provided by Paul DiMaggio and his colleagues (DiMaggio, Bernier, Heckscher, & Mimno, 2019). The authors combine Collins's ideas with a theory of speech genres developed by Mikhail Bakhtin in order to explore how mutual focus of attention, shared mood, and rhythmic entrainment exist in posts, comments, and threads. In particular, we consider national, ethnic and other social categories as potential group symbols.
Research materials that have already been published in combination with the formulated theoretical framework allow us to formulate five research questions for this study: 1) What social categories do migrants to Russia from Azerbaijan, Kazakhstan, Kyrgyzstan, Tajikistan and Uzbekistan employ in their attempts to identify themselves and to be identified by others in online interactions? 2) How are these categories related to each other? Are they used individually, or are there elective affinities between social categories exercised for the purposes of identifications? 3) Do identifications differ for migrants from different countries? 4) Do they differ for self-identifications and external identifications? 5) What are the typical situations when the usage of "migrant" category is relevant?

Data and Methods
The source of empirical data are interactions in a sample of online groups found on the social media. Specifically, we study migrants' interactions in VK.com (VK is short for its original name Vkontakte), Russia'a most popular social network and one of the most popular in the fSU countries (Timoshkin, 2019). The group is a specific form of interaction that enables users to communicate with a wide range of people, from friends to strangers. Our sampling of groups was organized in two steps. First, in January-February 2020, we searched for groups that contain the morphemes migrant*/migrat*, azer*, kazakh*, kyrgyz*, tajik*, uzbek* in their titles. To embrace regional and ethnic diversity of sending countries, we also included such morphemes as badakshan*, dungan*, fergan*, karalakalpak*, pamir*, talysh*, and uyghur*. The search was conducted in the Russian language for all morphemes and then additionally in the national languages for morphemes migrant*/migrat* and the name of the nation 3 . We looked through all the groups displayed in VK's search output and selected groups of three types: (1) groups related to migrant activities; (2) 20 largest groups that provide information about the country of origin for each nation (both for Russian and for the national language); (3) groups that discuss Islam for each nation. More than 2000 groups were selected at this stage.
Second, in March 2020, we applied network analysis for further sampling 4 . We considered common members of two groups as a link between them. Network analysis was conducted for the groups with different morphemes in titles, as well as for all groups together. Based on the results of the network analysis, we selected 23 groups based on (a) centrality; (b) relevance for migration issues; (c) intensity of interactions in a group; (d) diversity of groups (including clusterizations) 5 .
The online groups that were included in our study can be divided into four types. The first type is a community of migrants from one country that settled in a specific Russian city. Three Kyrgyz, one Tajik, and two Kazakh groups belong to this type 6 . Migrants' communities provide a space for casual interactions, mutual assistance, and information exchanges. The second type is an information group. It contains information on various topics about a country of origin, including migration issues. This type is represented by one Kyrgyz group; two Tajiks groups (one of which is focused on the history of Tajikistan); four Uzbeks groups (with different target audiences: Uzbek youth outside the country, immigrants who left the country in the 1990s, mixed audiences); one Kazakh group; and two Azerbaijani groups. The third type consists of the groups of national cultural organizations in Russian cities. It embraces one Azerbaijani and two Kazakh groups. The fourth type consists of groups devoted to legal assistance/mutual assistance for migrants to Russia from different countries.
This type includes four groups selected on the basis of the morpheme migrant*/ migrat*. These groups are of different sizes and orientations, from pragmatic issues to defense of migrants' rights. Interactions in the former three types of groups combine, in different parts, the Russian language and the national languages 7 . Interactions in the fourth type are entirely in Russian.
The main research method of this study was online observation. We investigated online interactions in these groups in April 2020. Specifically, we observed interactions that took place during the period from January 2019 to April 2020 8 . In the study, we adopted the methodology developed by DiMaggio et al. (2019). The unit of observation is a post along with its thread of comments. We focused on the posts that triggered discussions among users: for small and medium groups, we analyzed the posts with 10 or more comments; for large groups, we analyzed the posts with 20 or more comments. Moreover, our observation was focused only on the interactions that comprise migrants' identifications. In total, 578 posts were analyzed, along with their threads.
During the observation, we identified the symbolic focus of the original post, symbolic foci of the thread, as well as self-identifications and external identifications applied by the users. Then prevalent and auxiliary types of identifications were revealed, as well as typical situations of interaction. Our analysis is qualitative, highly dependent on the context of interactions in a particular group. Thus, in the presentation of our results we provide and discuss observations of interactions exemplary for all groups or for a specific type of groups.

Common Prevalent Identifications: Between Market, State and Nation
Five common prevalent identifications characterize migrants' interactions in online groups. These identifications are common as they occur in all types of groups. They are prevalent as they occur regularly and were a topic for focused discussions. Two identifications deal with a migrant as a person who moves from one country to another. Three more identifications cover a migrant's settlement in Russia. For the latter selfidentifications are tightly intertwined with external identifications.
The first prevalent identification is "migration as economic necessity". It relates labor migration to economic or political situation in the country of origin, and image of the nation depends on how migration to Russia is evaluated. This identification could be further specified in two matters. First, the definition of a migrant can be either derogatory (abusive) or neutral/moderately positive, and this issue is debatable by migrants themselves. The illustration is the following interaction from an Uzbek information group: 7 We used Google Translate (https://translate.google.com) to comprehend the statements that are in the national languages.

Comment:
No one chose where he was born, it became easier for us to complain about one's own life and to look for those who are guilty than to fight and to move, to find solutions to problems!!! As long as you have health and brains in place, you can work and earn money not only in Kyrgyzstan but also in Africa, for this you need only confidence and aspiration! The second prevalent identification is "a migrant subject to state regulations". This identification is typically discussed in connection with the EEU and with requirements for obtaining Russian citizenship. Moreover, citizenship is spontaneously associated with ethnicity, in a variety of migrants' interactions, both in national groups and in legal assistance groups. The logic of nationality and the logic of citizenship mismatch. This is articulated in claims to simplify requirements for Russian citizenship for "Russians", in accusations of discrimination against non-Slavs, as well as in assertions that nationality actually does not matter for migration regulations. To illustrate this incongruity let us refer to two posts with opposite logics, both are from groups of legal assistance: Comment: And to other migrants from Central Asia? If you abolish it, do it for all. They again divide people into Slavs and non-Russian, yeah.

Observation 4 10
Post: A letter was sent to me by mail. "[…] May be, we will try to write a collective appeal to Putin and to ask him to supplement the second law, on obtaining a Russian passport in a simplified way by amendments about Ukrainian citizens living in the Russian Federation with a [legal] status (temporary asylum, temporary residence permit, residence permit), but with propiska in Donetsk and Lugansk region? […]" Electronic signatures will not work. But, as far as I remember, is there some kind of official/semi-official website for petitions? Do you believe in success of this idea, in general?

Comment: You want to give citizenship to Ukrainians but what about [ethnic]
Russians? How is it for them? In common order? Justice is off the hook. 10 This observation comes from a group of legal assistance for migrants from different countries. That is why issues of people from Donetsk and Lugansk region are discussed below together with migration from Central Asia.

334
Natalia D. Tregubova, Maxim L. Nee Reply to comment: This is not the point. The fact is that the Donetsk region was divided in into parts, into us and them. The third prevalent identification is "a migrant representing the nation". In national information groups it is presented in the news about migrant heroes and migrant criminals. It constitutes the basis for solidarity or splits in online discussions where self-identifications and external identifications are intertwined (see Observations 7 and 8 below). In migrants' communities, there are typically discussions and criticism (or, less often, praise) of migrants' behavior (i.e. everyday practices) and demands to behave properly. The argument is that the behavior of one Kyrgyz in Russia is a ground for judgment about all Kyrgyz (the same for Tajiks, Uzbeks, etc.). This identification is combined sometimes with characterization of migrants as hillbilly insufficiently imbued with urban culture. Consider, as an example, how this kind of self-identification arises in response to external identification in the Kyrgyz migrant community:

Observation 5
Post: This is to a post about the attitude of [ethnic] Russians towards us! It's our own fault that they treat us like that. Firstly, our country, to put it mildly, is weak. Secondly, our people in Moscow and other Russian cities behave very indecently. They are very arrogant and immediately begin to swear from a scratch [...].

Comment:
In principle, there are people who behave just wildly, but again, not all.

Comment:
Ahahw author is a moron! I will provide arguments about the post in private message. Nothing personal, your worldview is just so funny, I can't call you anything else. The fourth prevalent identification is "a migrant as a native speaker". There are claims on VK groups that migrants should use their native language. These claims are often related to fear of their own culture and language (see Observations 9 and 12 below).
The fifth prevalent identification is "a migrant as inferior to local people". It relates to the following topics: difficulties in employment, racial profiling, biased media, and to a general belief that Russians consider migrants as second-rate people. This self-identification is supported by two kinds of external identifications. The first one is provocations against migrants in online groups that do or do not result in further discussions. The second one is the accusation of migrants in hostility to Russians that may turn into a squabble fueled by mutual diminishment of nationalities. Counterarguments against radical nationalism in such discussions often arise from the Soviet notion of "the friendship of the peoples". The most vivid example that is characteristic for Uzbek and Tajik groups are claims that these nations expelled the Russians in 1990s. Here is a typical interaction from a Tajik information group: Reply to comment: I know you hypocrites. I have a girlfriend from Uzbekistan. She told how they terrified the Russians back in the 90s. After they stabbed her uncle, they had to leave. And before that they had always said, they said "Russians go home". So get out of here, goddamn Nazis.
Reply to comment: How old are you? Probably, you've heard that then everyone had troubles. The Union was disintegrating. There were clashes between people everywhere. According to the stories of elderly people, relatives, Uzbekistan, in particular, Tashkent City was built by the whole country, they sent professionals from all the national republics to rebuilt the City that was destroyed by the earthquake. So, it has become multinational. People lived well, friendly, nobody distinguished who is of what nationality, they learned the Russian language, and Russians learned the Uzbek language. You need to understand, there is no bad nations, all are good, it is just bad persons in every nation […].

Country-Specific Prevalent Identifications: Internal Differentiation
Prevalent country-specific identifications in migrants' online interactions split the countries into two groups. One includes Kyrgyzstan, Tajikistan and Uzbekistan, the other embraces Azerbaijan and Kazakhstan.
The key category for the split is "low-qualified labor migrants from Central Asia". "Central Asia" typically refers to Uzbekistan, Tajikistan and Kyrgyzstan. This category is employed for self-identifications as well as for external identifications and constitutes a basis for internal differentiation. For migrants from Uzbekistan, Comment: I don't know where they decreased in numbers, today I was driving past the migration service, and it was crowded. Five hundred people stood. And all were Uzbeks or Kyrgyz. This is in April, and soon there will be warmer and there will be more of them. It seems to me that though the cost of a patent has been raised and laws are constantly changing against them, those who are in need go here, they are trying their best, work for a penny to somehow feed their family at home. I feel pity for those people.
"A member of the diaspora" is a crucial self-identification for migrants from Kazakhstan and Azerbaijan. This identification shapes a group border that separates "members of the diaspora" from "low-qualified labor migrants", or "gastarbeiters". For Azerbaijani diasporic identification is not specified further. For Kazakhs the division within the diaspora between Russian Kazakhs and Kazakhstan Kazakhs turns out to be salient (see Observation 9).
These two identifications, "low-qualified labor migrants from Central Asia" and "members of the diaspora", imply different uses of ethnic categorizations. On the one hand, discussions of everyday problems of labor migrants amalgamate symbols of class, social status, religion and ethnicity. Strong ethnic categorizations arise here mainly in response to violation of moral boundaries: for example, when someone generalizes one's own negative experience of interaction with "Kyrgyzs", "Tajiks" or "Uzbeks" to the entire nation. On the other hand, ethnic categorizations strengthen transnational diasporic networks. Being "Azerbaijani" or "Kazakh" means learning native language and culture, as well as being aware of what is going on in the Homeland. Such categorizations are constructed through symbols that refer to the imagined community, from national culture to political debates. Comment: "Not to pretend to be someone you don't know?" Why such a neglect for Caucasians. After all, it is clear whom and why they [Tajik migrants in Russia] are pretending to be.

Comment:
You yourself say that the guys are pretending to be Caucasians, but you yourself write in Russian. So, you are pretending to be a Russian?
Both observations refer to change in self-identification related to external identifications in Russian society. However, these changes have different directions. Pamiri migrants in Russia are claimed to reveal their identity, whereas Tajik migrants are claimed to hide it. Observation 11 reveals a conflict between concepts of Tajik nation and Tajik citizenship for Pamiris. Migration to Russia seems to be liberating for them, as, it is argued, it allows for national self-identification that is possible in Tajikistan only in opposition to Tajik majority. Post in Observation 12, in contrast, accuses Tajiks in faking their identity to be more prestigious (Caucasians). Tajiks are considered here as having an inferior position in ethnic hierarchies, and Caucasians better as they have reputation of tough guys and are Russian citizens (Dagestani and Chechens are mentioned). At the same time, Caucasians are migrants from what was called the "Soviet South", as Tajiks are, and here citizenship becomes less important than visibility and ethno-territorial divisions inherited from the Soviet past.
Juxtaposition of these two observations supports the statement about internal differentiation of migrants. Both transformations of identifications help to avoid being classified as "low-qualified labor migrants from Central Asia" that is associated with weakness and low position in Russian society. It is obvious in switch from "a Tajik" to "a Caucasian", while "a Pamiri" seems to be exotic/less-accountable identification in Russia that is not strongly associated with labor migration. These observations also demonstrate that clash between national pride and "migrant" categorization is painful and fuels collective emotions from humiliation to rage.
At the same time, a change of identification could have pragmatic purposes connected with labor migration. Consider an observation from a Kyrgyz migrant community:

Observation 13
Post: Hello, a passport has found more than a week ago. If anyone knows anything about the owner, please contact me [...] [Photo of Kyrgyz passport].

Comment: Who is this Uzbek with a Kyrgyz passport?
Reply to comment: Kyrgyz Uzbek.
Reply to comment: I've heard that Uzbeks and Tajiks make fake [Kyrgyz] passports so as not to pay for a patent and work permit.
Reply to comment: Yes, it is true! And this is the state who allows it.
Here the identification changes only in institutional contexts (for the police, the migration service, etc.) as citizens of Kyrgyzstan are subjects to much simple migration regulations. Thus, the citizenship has been changed, not ethnicity. Note, however, that the change of identification lies within the category "low-qualified labor migrants from Central Asia". Probably, that is why it does not arouse emotional responses connected with ethnic hierarchies: neither "an Uzbek" nor "a Kyrgyz" seems to be better identification in Russian society.

Auxiliary Identifications
In addition to the identifications characterized above, we observed migrants' identifications that are auxiliary: they did not occur regularly, or often, or are not a topic for focused discussions.
There are several auxiliary self-identifications in migrants' online interactions. They are: • "Muslim migrants". This identification is primarily connected to discussion of everyday religious practices. For Tajik and Uzbek groups, it is also concerns a discussion that Islam is easier to practice in Russia than in the country of origin due to strict regulation of Muslim practices by the state (especially in Tajikistan). • "A person from a specific place" -city, town, village, or region (see Observation 11). This self-identification could also be combined with the emphasis on nation, such that "we are all Tajiks". • "Nostalgic migrants" identification arises in various situations, and it is of special relevance for Russophone immigrants from Uzbekistan to Russia who left the country in 1990s. • A type of migration: educational or labor.
• "Too many migrants from my country". • Negative and positive auto-stereotypes. They are: "Your own people cannot be trusted" (observed in Kyrgyz and Kazakh online groups); "Tajiks are passive"; "Azerbaijanis are entrepreneurial, decent and solidary"; "there exists 'warm' Central Asian mentality" (observed in Uzbek groups); "Uzbeks are generous".
Auxiliary external identifications are: • Imputation of ethnicity to the opponent (observed for Russian, Armenian and Uzbek nationalities). • "We-migrants are better than you-migrants". This identification applies to migrants from Central Asia, and the arguments are: "Because we are Russian"/"we know the language"/"our behavior is proper" (see Observation 4 above). This identification occurs spontaneously or in response to provocations, "migrants are not welcomed in Russia". • "A migrant obliged to Russian laws". This identification appears in two forms: as a moral obligation to obey laws and as a righteous anger to unjust laws. • "A hostage of the situation" due to the actions of the authorities or economic policies of Russia and/or the country of origin. It is simultaneously a self-identification and an external identification employed by advocated of migrants' rights.

Contexts of Identifications
Three more remarks about varying contexts of identifications in online interactions. First, discrimination against migrants does not universally function as a collective symbol in migrants' interactions. A significant part of interactions that we observed are pragmatic. Some interactions involve emotions that original posts do not intent to produce. Let us consider a post focused on the insult of a Kyrgyz migrant by a popular Russian talk-show host. Comments to this post do not demonstrate collective offence against Russia: commentators blame the person, discuss his actions in a pragmatic manner, or address their negative emotions to Kyrgyz authorities:

Observation 14
Post: Andrey Malakhov reacted to angry comments of the Kyrgyz people addressed to him. Russian TV-presenter Andrey Malakhov made a statement after the scandal with the comment about the Kyrgyz people […].
Comment: Well, he is partially right. Salary of doctors and teachers in Kyrgyzstan is tiny. Our elections: for 1000 soms 11 you can buy a voice. Maybe those who are offended do not know how people live outside Bishkek.
Reply to comment: I did not hear him speaking. But what I've read -it did not offend me at all. I am more offended by the theft of the state budget, the venality of the authorities and Nazism.
Comment: Malakh, you are an oligarch whore.
Comment: What do you want from a person who "washed" other people's lingerie and digs into it during all his professional life. Those who remember his program "Big wash" will understand what I mean and how it relates to this person.
Second, the Soviet type of identification -"the friendship of the peoples" policy in combination with "nationalities" classification -regularly arises in observed interactions. It is often presented in a truncated form, as a widespread judgment: "There are no bad nations, there are bad people". However, some users directly refer to "the friendship of the peoples" and its implementation in Soviet history, as Observation 6 above reveals.
Finally, the dynamics and the focal topics of interactions also depend on the characteristics of a particular online group: its size, audience, language, who can post, who can comment, and so on. Observations 3 and 4 (see above) are a good illustration. The observations are from two groups, both are legal assistance groups with active moderators. The first group is small. It is characterized by human rights rhetoric applied to low-qualified labor migrants in Russia. Interactions in this group often occur in broken Russian. The second group is large. It focuses on obtaining 11 The som is the national currency of Kyrgyzstan.
Russian permanent residency and citizenship and involves empathy over migrants' troubles with these issues. It aims predominantly at migrants from the fSU countries, mainly at those for whom Russian is a native language. As Observations 3 and 4 demonstrate, in both groups there is a transition from the terms of citizenship to the terms of nationality. However, due to differences in initial goals and the target audience of the groups, this transition moves in the opposite directions: to accusations of discrimination against "non-Slavs", on the one hand, and to a call for privileges for "Russians", on the other hand.

Discussion and Conclusions
Let us begin this section with the answers to the original research questions of this paper.

What social categories do migrants employ in their attempts to identify themselves and to be identified by others in online interactions?
Migrants to Russia from Azerbaijan, Kazakhstan, Kyrgyzstan, Tajikistan and Uzbekistan employ a diverse set of social categories in their identifications. The prevalent categories are citizenship, nationality, language abilities, migrant-ness, as well as the category "low-skilled labor migrants from Central Asia" that amalgamates class, visibility and a broader region. Auxiliary social categories are religion, interstate region, town/village/city of origin, urban/rural origin, tradition/modernity, broader ethnicity (such as Slavs) in connection with race and belonging to the Soviet culture. Minority status, the aim of migration and kinship are employed occasionally.
How are these categories related to each other? Are they used individually, or are there elective affinities between social categories exercised for the purposes of identifications?
There is an elective affinity that manifests itself in the category of "low-skilled labor migrants from Central Asia". It exists in the interactions in all types of groups and provides the basis for internal differentiation between migrants. "Low-skilled labor migrants from Central Asia" are opposed, on the one hand, to visible, but not (necessarily) low-skilled members of the Kazakh and Azerbaijani diasporas and, on the other hand, to non-visible ethnic Russian and Russophone migrants from Central Asia.

Do identifications differ for migrants from different countries?
There are substantial differences in identifications of migrants from different countries. Migrants from Kyrgyzstan, Uzbekistan and Tajikistan in their identifications refer to low-skilled labor migration to Russia as a fact, a subject for assessment, and on occasion as a unifying category. For these countries, the present and the future of the nation is discussed in the framework of evaluation of mass immigration to Russia. Migrants from Kazakhstan and Azerbaijan employ identifications connected to diasporic connections. Furthermore, in Kazakhstan there is a clear division between Russian and Kazakhstan Kazakhs.

Do identifications differ for self-identifications and external identifications?
External identifications and self-identifications are intertwined in typical discussions in VK groups. The perspective of Russian society on migrants is, as it were, built into the perception of migrants themselves.
What are the typical situations when the usage of "migrant" category is relevant?
There are three types of situations associated with the usage of the "migrant" category. They are connected with the market, state and nation, respectively. First, "migrant" is a synonym for "labor migrant" who moves to Russia due to economic necessity. Second, "migrant" is a citizen of the state subject to specific institutional regulations. Third, "migrant" is the Other, a representative of his/her nation in Russian society. The situation of the third type is primarily connected with three contexts: (a) discussions of migrants' behavior in Russia; (b) discussions of discrimination by police and employers; (c) provocations against migrants in social media. We should also add that migrants themselves sometimes perceive migrants from a different country as the Others.
The analysis of migrants' online interactions allows us to formulate the following conclusions: 1. "Low-qualified labor migration from the countries of Central Asia" (Uzbekistan, Tajikistan and Kyrgyzstan) is a key categorization for national and other identifications of residents of these countries (both migrants and non-migrants).
In Uzbekistan, Tajikistan and Kyrgyzstan, the fact of labor migration to Russia is highly relevant for the evaluation of the nation and its further developments.

Migration to Russia from Kazakhstan and Azerbaijan is fused with diasporic
connections. 3. The issue of the language is highly significant for migrants' identifications. The national language is often regarded as an indicator of belonging to a nation. The role of Russian language is ambivalent. It is perceived as a valuable competence for a migrant in Russia, yet as a threat to national culture as well. We believe this ambivalence is rooted in the legacies of the Soviet national politics and policies. 4. Current migration flows to Russia are related with migration processes that took place after the collapse of the Soviet Union. For the countries of Central Asia (primarily, for Tajikistan and Uzbekistan), the exodus of ethnic Russians in the 1990s is still important as a point of reference and evaluation of contemporary mass labor migration to Russia. At the same time, the Soviet type of identifications based on "nationalities" and "the friendship of the peoples" remains a reference point in discussions about migration. How do these results supplement understanding of migration processes in post-Soviet region? Let us return to the similarities and differences between the Azerbaijan, Kazakhstan, Kyrgyzstan, Tajikistan and Uzbekistan that were formulated in the first part of this paper.
The five countries can be structured into two classes: (a) Kyrgyzstan, Tajikistan and Uzbekistan are characterized by huge labor migration to Russia, while (b) Azerbaijan and Kazakhstan provide less migrants coming with diversity of objectives. Our study shows that for Azerbaijan and Kazakhstan migration to Russia is closely connected with diasporic ties. Both Kazakhs and Azerbaijanis detach themselves from "gastarbeiters" who came from Kyrgyzstan, Tajikistan, and Uzbekistan.
Visibility of migrants from Azerbaijan, Kazakhstan, Kyrgyzstan, Tajikistan and Uzbekistan turns out to be important for identifications in online interactions. However, visibility is not equal to race: it acts as a combination of phenotype, fluency in Russian language, and migrants' everyday practices. Visibility is also paired in online interactions to "invisibility" of ethnic Russian (Russophone) migrants from Central Asia.
Islam is present in observed online discourses to a varying degree and in different variations. Islam turns out to be more significant for image of a nation. In different cases it is related to a nation in different ways. Islam is a unifying category, however, for migrants it is typically an auxiliary identification. Two points are important here: (1) belief that Muslim migrants are not welcome in Russia;and (2) conception that Islam is easier to practice in Russia than at home (for migrants from Tajikistan and Uzbekistan).
Differences in the migration regulations among the post-Soviet countries are frequently discussed in social media. However, they are often connected to nationality: from the migrants' point of view, nationality is/should/should not be decisive for facilitation of migration legislation in Russia.
Regarding regional differences, we observe that "Central Asia" as an identification is widespread online while local identifications are not so important. Perhaps, this is due to the specifics of the studied groups that are primarily focused on nations.
Finally, transnationalism manifests itself in migrants' online interactions in two basic ways: as transnational labor migration and as diaspora. Besides, there are traces of transnational citizenship in discussions about dual citizenship and about politics in both Russia and the country of origin. This conclusion complements scholarly literature on transnational migration in Russia that pays attention to transnational practices of labor migrants from Central Asia and considers relations with relatives and friends, not political activities, as the main type of transnational practices.