Narratives of migration on Facebook: Belonging and identity among former fellow refugees

Abstract

This article brings together research on migration and identity in translocal and superdiverse contexts, and the recently expanding interest in narratives and interaction in social media, by examining the construction of identities in narratives shared in a private Facebook group message. The participants are former fellow refugees from Poland who reconnected on Facebook after two decades. The article analyzes three narratives produced in response to the researcher's question about ethnic and national affiliations. Using Bucholtz & Hall's (2004) tactics of intersubjectivity framework, this study examines the complex and conflicting ways in which individuals position themselves with respect to various contexts of belonging and difference (Meinhof & Galasiński 2005) that emerge in their narratives. I argue that the narratives show a link between essentialist or nonessentialist views of ethnicity/nationality, and the teller's assumed agency over her identity. The study also discusses new possibilities for discursive practices in social media contexts. (Narrative, migration, social media, identity, belonging)

INTRODUCTION

Studies of migration, the ‘narrative turn’ (De Fina, Schiffrin, & Bamberg 2006; De Fina & Georgakopoulou 2012; Georgakopoulou 2013), and the examination of the new media's impact on interaction and discourse analysis (e.g. Tannen & Trester 2013), have recently been prominent in humanities and social science research. This article speaks to all three of these research directions by presenting an online ethnography (Androutsopoulos 2006, 2008) of a small social network of former fellow refugees from then-communist Poland, currently living in English-speaking countries, sharing narratives on Facebook. The group consists of six women, including myself, and the data comes from the ongoing private group message that was started as this project took flight. The group message contains interview-style questions and discussions prompted by them, interspersed with unrelated conversations.

Narratives have emerged in the group in response to the researcher's questions, and spontaneously through reminiscing and sharing personal stories. The narratives vary greatly in their structure and dynamics. Some resemble longer monologues, while others are fragmented and enmeshed in the surrounding discourse. Some have a single teller, while others are co-authored by multiple participants. They are also characterized by features common to social media discourse, such as high interactivity, variability, audience participation, translocality, intertextuality, and multimodality (Georgakopoulou 2013; De Fina 2016). Accordingly, I adopt a social-interactional (Georgakopoulou 2013), practice and user centered (De Fina 2016), and dimensional (Ochs & Capps 2001) approach to narrative.

In this article, I focus specifically on three narratives produced as longer, uninterrupted monologues in response to my question about the participants’ ethnic and national identity. My analysis and discussion is organized around four themes. First, I demonstrate how the online context—specifically, that of a Facebook private group message—influences the structure and interrelatedness of the narratives. Second, I identify various contexts of belonging (Meinhof & Galasiński 2005) that emerge as relevant in the narratives. Third, using Bucholtz & Hall's (2004) tactics of intersubjectivity framework, I explore how the participants position themselves in complex and sometimes conflicting ways with respect to these contexts of belonging and to identity categories. Finally, I argue that when the narrative's teller constructs herself as passively subjected to life's circumstances, she also tends to describe ethnic and national identities in essentialist terms as inherent qualities of people. By contrast, the construction of a more agentive role in her own life story tends to be accompanied by the teller's more dynamic and fluid depiction of ethnicity and nationality. In these ways, the narratives represent different ways of conceptualizing one's agency over ethnic or national belonging, and over one's experiences.

Below, after introducing the participants and research methods, I discuss the theoretical frameworks adopted in my analysis: Meinhof & Galasiński's (2005) language of belonging, and Bucholtz & Hall's (2004) tactics of intersubjectivity. Then I briefly discuss superdiversity and translocality, concepts that are salient in recent work on migration (Arnaut & Spotti 2014) and relevant to the present data, since both the identities the participants construct for themselves and the space of the group message itself are translocal and superdiverse in character. I also review relevant work on narratives in social media before presenting the analysis of my data.

PARTICIPANTS AND DATA

The women involved in this project, including myself, are former Polish refugees who met as teenagers in Rome, Italy, in 1987. That year saw the peak in defections from then-communist Poland to the refugee camp in Latina, a city fifty miles south of Rome. The refugees were housed in converted military barracks in Latina, and eventually moved to locations in and around Rome. They were provided with basic necessities while waiting for their asylum applications to be processed by third country consulates (cf. Laszkiewicz 2012). Between January and mid-September of 1987, more than 8,000 Poles arrived as refugees in Italy,¹ including my family.

I became friends with the other five women through the junior high school (scuola media) where we were enrolled, and the apartment complex in the Trastevere district of Rome where we lived. I was the first one to leave Italy for the United States, in September 1988. The others followed to the US, Australia, and Canada between 1989 and 1991. Having been mostly out of touch for nearly two decades, the six of us reconnected through Facebook in 2008, and began exchanging frequent posts, comments, and messages. In 2011, I invited the other women to participate in a research project on narratives of migration and the construction of migrant identities.

This article is based on two years of a group message conversation on Facebook, including responses to interview-style questions and spontaneous interactions. The six participants (whose names have been changed) are: Sonia (in Canada since 1990), Majka (in Canada since 1990), Anna (in Canada since 1991), Ada (in Australia since 1989), Ola (in the US since 1989), and myself (in the US since 1988).

PARTICIPANT AGENCY: METHODS AND ETHICS

Sociolinguistic research where the participants are the researcher's family and friends is common (e.g. Coates 1998; Tannen 2005; De Fina 2012; among others). The pre-existing relationship that the researcher-analyst has with the participants has inevitable implications for the research process. In my case, having shared the experience that is the subject of this project with my participants, I am as much a part of the data as its collector and analyst, which presents both methodological and ethical challenges. Methodologically, the challenge is not to let the intimate knowledge of the participants’ lives influence one's interpretation and analysis. Ethically, one of the challenges is to ensure that private interactions not flagged as ‘research’ do not inadvertently make it into the data. Accordingly, I have strived for complete transparency regarding the research process. I always ask for the participants’ approval before using any segment of the group conversation in my data. The participants signed consent forms and read a detailed project description. I explained to them that I looked for spontaneous discussions generated by my questions and invited them to contribute questions of their own.

The initial questions that I presented to the participants were very general, since I was interested in themes that would organically emerge out of group discussions. The specific question that elicited the three narratives presented here was stated rather informally as: ‘Would you describe yourself as Polish, or your current home-ish—Canadian, American, Australian—or something else, and why?’. It came early in the project, as one of the first questions I posed to the group. It generated a lengthy discussion that veered in numerous directions, but at some point each participant answered the question with a longer narrative about her experience with identity.

BELONGING AND DIFFERENCE

As has been recognized by researchers in sociocultural anthropology, cultural studies, and related fields, identities are produced and negotiated in social interaction, enacted in discourse through sociocultural practices, and situated in context (Hall 1992; Antaki & Widdicombe 1998; Holliday, Hyde, & Kullman 2004; De Fina et al. 2006; De Fina 2007; Riley 2007; Baran 2017). Stuart Hall argued for the fluidity, complexity, and changeability of identities: ‘Within us are contradictory identities, pulling in different directions, so that our identifications are constantly being shifted about’ (Hall 1992:277). Identity as something one does rather than has is emphasized in Bucholtz & Hall's observation that ‘identity inheres in actions, not in people’ (2004:376). Similarly, De Fina reminds us that aspects of identity should be regarded ‘as an interactional achievement grounded in concrete social contexts and evolving with them’ (2007:374).

Meinhof & Galasiński (2005) adopt this dynamic notion of identity in their study of three-generational families living on the Polish-German border. They examine identity construction as enacted in what they refer to as ‘the language of belonging’. They see their interviewees as utilizing a range of linguistic resources at the content, grammatical, rhetorical, and interactional levels (Meinhof & Galasiński 2005:41) to position themselves with respect to ‘contexts of belonging’, which they categorize into the contexts of time, place, social relations or ‘in’ and ‘out’ groups, and social encounters. However, the authors recognize that these categories are not discrete, but overlap and interact with one another, and that they do not represent a definitive list, but rather entry points for analysis. The discursive construction of identities relies on making use of linguistic resources within these different contexts. For example, Meinhof & Galasiński show how their interviewees positioned themselves in relation to dominant ideologies surrounding specific time periods (2005:21), including World War II and its immediate aftermath, the period of communist rule in the former German Democratic Republic and Poland, and of German reunification.

The authors argue that many Polish interviewees tended to rely on lexical and grammatical constructions that remove agency from the Poles and depict their lives as the result of events characterizing the different time periods (2005:27). German interviewees, by contrast, tended to employ discursive structures that negotiated their positioning with regards to the German collective responsibility for the aftermath of the war (2005:29). In this article, I show that my participants also use specific linguistic resources and strategies to construct an image of themselves as either passively subjected to life events, or actively shaping them, which in turn correlates with—or perhaps produces—a more or less essentialist conceptualization of identities in the participants’ discourse.

The notion of contexts of belonging as outlined by Meinhof & Galasiński (2005) can be productively applied to the data on which this article is based. As the authors emphasize, the list of contexts they propose is not definitive, but emerges from their particular data set. Here, I have adapted some of Meinhof & Galasiński's proposed contexts, and added others that appear significant in my data. They include time, place, sets of social relations or ‘in’ and ‘out’ groups, language, and religion. I elaborate on the ones relevant for this article below.

• • Time: Stages of the immigration process, specifically time spent in the refugee camp, early days in the final destination country, and the present

• • Place: Poland, Italy, and final destination (US, Canada, Australia)

• • ‘In’ and ‘out’ groups: Several dimensions of ‘in’ and ‘out’ groups emerge from my data; the one relevant for this article is ethnicity/nationality (e.g. Poles vs. non-Poles; Americans vs. non-Americans); significantly, ethnicity/nationality as a category is, in this article, posited in the interview question itself

• • Language: includes Polish, English, and Italian

Meinhof & Galasiński observe that ‘[t]he language of belonging is more often than not a language of not belonging, of lack of agency, of creating an outgroup rather than an ingroup, of negative rather than positive social encounters’ (2005:42; emphasis in original). The language of belonging is thus complemented by the language of difference. This is likewise true in my data, since the participants position themselves both with and against the various aspects of the contexts of belonging that emerge. To examine in greater detail how my participants discursively establish their belonging or nonbelonging, I employ Bucholtz & Hall's (2004) tactics of intersubjectivity framework.

TACTICTS OF INTERSUBJECTIVITY

Bucholtz & Hall (2004) developed the framework of tactics of intersubjectivity as a tool for analyzing the ‘contextually relevant sociopolitical relations’ produced through the semiotic processes involved in identity work (Bucholtz & Hall 2004:382). The tactics involve pairs of contrasting discursive moves that social actors make to position themselves in relation to socially available identity categories and the ideologies surrounding them. Bucholtz & Hall propose three pairs of tactics:

1. (i) Adequation and distinction, which construct similarity or difference between subjects

2. (ii) Authentication and denaturalization, which claim realness or expose falseness of identities

3. (iii) Authorization and illegitimation, which support or negate the ‘legitimate’ expression of identities

A number of studies have employed this framework to theorize how goals are accomplished in identity work through specific, situated practices (e.g. Fitts 2006; Higgins 2007; Cashman & Williams 2008; Cashman 2008; Nylund 2009). In this article, I invoke tactics of intersubjectivity as I examine the discursive moves made by three participants—Ola, Majka, and Ada—as they talk about their immigrant journeys and their ideas surrounding ethnicity and nationality. The most common pair of tactics that each participant employs is adequation/distinction. Similarities are constructed between members of perceived groups (e.g. Americans, Canadians, Poles, immigrants, non-Americans), and differences between groups perceived in opposition to each other are emphasized. Participants also engage in discursive authentication of their own group membership by claiming engagement in practices stereotypically associated with different groups, such as playing particular sports or consuming ‘ethnic’ foods. Conversely, an individual may be portrayed as an ‘inauthentic’ group member if their practices break with the set of similarities that has been constructed as unifying the group.

The three narratives differ in the degree to which the tellers portray themselves as passive or agentive, and to which they construct essentialist representations of identity categories. Not surprisingly, essentialist depictions of ethnic and national groups rely on the tactics of adequation and distinction for the discursive reification of these groups, as well as on authentication for claiming ‘realness’ of membership in them. By contrast, when identities are discussed as fluid constructs, the tellers also use denaturalization, which operates by severing or inverting the expected links between identity categories and their perceived attributes.

SUPERDIVERSITY AND TRANSLOCALITY

The technological innovations and new ways of interacting brought about by globalization have made communication and the exchange of ideas more immediate and accessible than ever before, consequently producing new sociocultural and political phenomena that some researchers in the humanities and social sciences have labeled ‘superdiversity’ (Vertovec 2007; Arnaut & Spotti 2014). Thinking from the perspective of superdiversity, described by Vertovec as the ‘diversification of diversity’ (2007:1025), has led sociolinguists to make conceptual shifts from fixed and discrete notions such as language, native speaker, foreign accent, or ethnolinguistic identity, to fluid and dynamic relations such as translanguaging and transculturations (Orellana 2009; García & Li Wei 2014), translocal and transnational speech communities, and the intersectionality and hybridity of multilingual and migrant identities (Baran 2017).

The data presented in this article exemplifies some of the phenomena that the superdiversity discourse (Arnaut & Spotti 2014) emphasizes. The renewed connections among the former refugee women were formed and are sustained in an online context. With some exceptions, the women have not been together in person since their days in Italy, when their primary language was Polish. Yet, the Facebook conversations take place almost exclusively in English, which all the participants learned as young adults in their eventual countries of settlement. When Polish emerges for a specific reason, such as discussing a recipe, telling a joke, or quoting an older person, the participants switch back to English within a few turns. The participants unanimously chose English when I asked them which language they preferred for our discussions, even though my initial letter introducing my project was in Polish. The choice of English is in itself an act of alignment with postimmigration identities, with the adult stage of life that the participants have experienced in English-speaking countries, and with the global orientation of their present-day lives as manifested through international travel and multicultural social networks.

The choice of English for Facebook discussions underscores the fact that the sociocultural meanings associated with specific language varieties or forms are contextualized, negotiated, and fluid (Cameron 1997; Bailey 2000; De Fina 2012), as are the meanings and boundaries of ‘in’ groups. Although one might expect Polish to become the in-group language for these women because it is linked with a shared past and shared ethnic background, English emerges as more relevant for the people they are today: fairly acculturated in their current countries of residence, middle-aged, middle-class, successful, educated women with families. Polish appears only marginally associated with this set of identities; indeed, all of the women have stated in individual and group interviews that Polish is not central to their current lives, including in their families. While it was the language through which the women formed their initial friendship, their current online community is not a reenactment of the original social network, but rather its reinvention. Research in the area of language and emotion (e.g. Pavlenko & Lantolf 2000; Wierzbicka 2004, 2008; Pavlenko 2005; Baran 2017) has demonstrated that multilingual speakers form emotional connections between language and experience, and between language and the identities that particular experiences produce or enact. Similarly, literature on codeswitching has regularly discussed the sociocultural meanings and contexts invoked by the use of particular languages among multilingual speakers (e.g. Blom & Gumperz 1972; Myers-Scotton 1993; Auer 1998; among others). We can thus argue that English allows the six women to recognize and validate in each other a successful accomplishment of their current, adult stage of life, and to enact their present-day relationships with one another.

English is also the primary language of internet use for all of them, so its choice in the Facebook chat both reflects and emphasizes the fact that this is an online community. This community is also translocal and transnational: first, because, like any online interaction, it brings together in real-time participants living in different countries and regions, and second, because conversations that produce this community move through topics pertaining to different locations and associated with different practices and language forms, as well as different time frames. The translocal and superdiverse nature of the Facebook group has implications for the structure of the emerging narratives, as I discuss below.

NARRATIVES IN SOCIAL MEDIA

As De Fina (2016) points out, research on online storytelling from a discourse analytic perspective across various platforms is still in its beginning stages, and this article contributes to this growing body of research. In this section, I discuss some of the characteristics of online interaction and storytelling that may have a bearing on the analysis of online narratives such as those produced by Ola, Majka, and Ada.

The development of the internet and of computer-mediated communication, which has aided the expansion of superdiversity discussed above, has also produced interactions characterized by translocality, multimodality, multi-authorship (Georgakopoulou 2013:709), fluidity of participation, and complex intertextuality. Consequently, online narratives provide an additional challenge to Labov & Waletzky's (1967) canonical model of narrative, whose structural approach has been contrasted with a social-interactional one (Georgakopoulou 2013:696), as exemplified by Ochs & Capps’ (2001) work. As Ochs & Capps argue, ‘[a]s such, narrative bows to no simple generic blueprint that sets it apart once and for all from other forms of discourse’ (2001:18). Instead of trying to define what narrative is, Ochs & Capps examine a set of ‘narrative dimensions’ (2001:18), which they list as tellership, tellability, embeddedness, linearity, and moral stance. According to the authors, narratives exhibit characteristics that fall at various points along the continua that the dimensions represent: one teller or multiple co-tellers, a highly or moderately tellable account, relative detachment or embeddedness in surrounding talk, linear or nonlinear organization, and constant or fluid moral stance.

In online interactions such as posts, comments, and messenger conversations, new aspects of the narrative dimensions emerge as salient. For example, Facebook posts and comments have a public audience whose membership is fluid and never fully known, as participants can join or leave without being noticed. In private-message group conversations such as the ones in my data, all participants are assumed to be always present, because even when they are away and neither reading nor contributing to an interaction, they can later go back and reread the thread. These factors as well as the norms of Facebook interactions produce the potential for new and sometimes unexpected co-tellers to emerge and reinterpret or revise the narrative (Georgakopoulou 2013:709), and possibly redirect the entire conversation (cf. De Fina & Georgakopoulou 2012:93). Embeddedness also takes on a new form in a Facebook group message because several conversations can be smoothly proceeding at once. Unlike in face-to-face interactions, it is possible to keep track of different threads by reviewing the conversation, and speakers develop innovative strategies for keeping track of overlapping threads and managing the incoherence of turn adjacency (Herring 1999; Berglund 2009).

De Fina & Georgakopoulou also make the case against ‘functionally restrictive schemes about the various structural components’ of narrative structure (2012:47). They argue that structure should be viewed as emergent and ‘sequentially unfolding’, rather than postulated a priori (2012:46), thus allowing for flexibility and variability depending on sociocultural context, tellership, audience and participants, type of story, and power relations among the participants. In social media interactions, all of these factors are affected in ways specific to the online context. For example, the authors point out that ‘the audience of a telling may be composed of knowing vs. unknowing recipients, principal recipients, ratified recipients, recipients who may be promoting the teller's view or who may be delegitimating or undercutting the telling activity, offering side comments, introducing other topics, etc.’ (De Fina & Georgakopoulou 2012:92). A Facebook group message creates potential for an ongoing rearrangement of audience composition and of audience members’ participatory roles over time.

Researchers have frequently observed the ‘dialogicity and openness of storytelling practices’ (De Fina 2016:477) in online narratives (Page 2012; Georgakopoulou 2013; De Fina 2016). Narratives in social media encourage extensive audience participation and collaboration, and emerge over time in an open-ended fashion (Page 2012, cited by De Fina 2016:477). De Fina suggests that perhaps these features of online narratives make them especially suitable for practice- and user-oriented analytical approaches (2016:475). She also argues against viewing various social media entities—blogs, messenger conversations, public Facebook posts, and so on—as separate and distinct genres, and in favor of seeing them instead as ‘spaces that have in common only very general characteristics’ (2016:476). Thus, blogs may share one set of characteristics with each other, and messenger conversations a different one, but there is also variability within and overlap between these settings, which again renders a focus on practices rather than structure more productive. The notion of an online context—blog, Facebook status update, messenger conversation, discussion board—as a space also emphasizes its spatial dimension, since online interactions are not time-bound, but recorded and available for reviewing. This feature distinguishes them from face-to-face interactions, and impacts the range of practices that they make possible.

THE NARRATIVES

My question, ‘Would you describe yourself as Polish, or your current home-ish—Canadian, American, Australian—or something else, and why?’, triggered responses in the form of relatively long autobiographical narratives spanning large portions of a lifetime (Ochs & Capps 2001:40), excerpted in the texts in (1)–(3). I did not explicitly ask for stories; rather, they emerged organically and spontaneously. Instead of responding, for example, with introspective reflections or analyses of the meaning of ethnic labels, the participants chose to discuss their self-identification through synopses of their immigrant journey. Perhaps, they took this opportunity to give shape to this journey: as has frequently been observed, narratives emerge as attempts at organizing and making sense of past events and experiences (Ochs & Capps 2001; Mishler 2006; De Fina & Georgakopoulou 2012; cf. Polkinghorne 1988; Pavlenko & Lantolf 2000). Similarly, Meinhof & Galasiński show not only the significance of ‘narrativization in the construction of people's sense of identity and belonging’ and attempts ‘to create cohesion in their lives in the telling’, but also that stories ‘revealed ambiguities, contradictions and unresolved tensions in people's lives’ (2005:112)—both of which points are relevant to my data.

The narratives in (1)–(3) demonstrate different approaches to linearity, one of the many dimensions along which narratives vary (Ochs & Capps 2001). Thus, Ola in (1) and Ada in (3) begin in the future (“I will die feeling Polish” in (1)) and the present, respectively, establishing from the start where the narrative is going in what might qualify as an abstract (Labov & Waletzky 1967). Majka in (2), by contrast, proceeds in a linear order, organizing her narrative around the juxtaposition between the past and the present. Nonetheless, in each narrative the teller attempts to construct a coherent story, which appears to lead up logically to the teller's present-day feelings and to motivate her claims of ethnic and national self-identification.

The fact that all of the responses take the form of narratives may suggest that the other participants are following Ola's lead, since, as discussed above, this online space is by necessity a shared one that is always available for review. To be sure, such interdependence of narratives is not limited to online contexts: as Ochs & Capps remind us, ‘[t]he extent to which a personal narrative is an entity unto itself, separate from prior, concurrent, and subsequent discourse, is related to turn organization, thematic content, and rhetorical structuring’ (2001:36). An online group conversation precludes the narratives’ separateness, however. Each one has to be viewed in the context of the previous ones because its author has access to what has been said before.

Thus, while the answers resemble uninterrupted monologues that might appear as separate entities, the participants clearly address their answers to the entire group, not just the researcher (e.g. Majka's “Hello girlz:)” in (2) or Ada's “Very interesting reading! I guess it's my turn now…” in (3)). The answers respond to previous answers and comments as much as to the original question, in contrast to other online spaces, such as public and anonymous comment sections in blogs, where, as De Fina shows, few comments are in dialog with each other (2016:492).

In the present example, Ola's response comes first, at 10:48 am on a Tuesday, immediately after my question. Majka's response follows closely at 11:32 am. Then come two days of silence. I initiate interaction again by thanking Ola and Majka for their responses. Sonia reacts within half an hour by promising that she is considering her answer. Four hours later comes Anna's response, and then a long exchange between Ola and Majka regarding Anna's daughter and former boyfriends, on which Anna does not comment. Eventually, Sonia provides her response to my original question, as do I, and more discussion ensues. Finally, Ada joins the conversation, comments on what she has just caught up with (“Very interesting reading!”), and becomes the last one to provide her answer.

Ada's entrance into the conversation emphasizes the point that because online interactions are not time-bound, a participant can be absent for some stretch of the conversation, and upon rejoining later, experience it verbatim as it took place. Ada, who lives in Australia in a time zone very distant from the others, often misses chunks of conversations when everyone except her is awake, but rejoins, reads, and contributes later. Since everyone can easily scroll back and catch up on anything they missed, the conversation is both an activity, and a space that can be revisited and rearranged. This space is a joint enterprise, co-created by all of the participants.

The ability to look back, reintroduce topics, respond to earlier ideas, and revise one's earlier statements, gives online interactions like this one the potential for greater intentionality and performativity on the part of the participants than is possible face-to-face. Contributions can be as spontaneous or as curated as their author decides, which allows them to communicate directness and sincerity by appearing unedited, but also lets the author showcase their writing style and participate actively in shaping the conversational space. For example, Majka, responding directly after Ola just forty-four minutes later, structures her answer according to some of the same parameters set by Ola, specifically, the immigration time frame and the juxtaposition of ethnic groups—points that I examine below. At the same time, Majka's response appears both unedited and curated for literary effect. Her text contains minor typos and stylistic errors: we see the lack of space between the smiley face and the next word in “Hello girlz:)When I first…”, and the absence of apostrophes in “my parents and their friends attitude towards Canadians”. But we also see the juxtaposition of images of Polishness and Canadianness that appears purposeful. The passage begins with “I had Polish friends, drunk Polish wodka at a Polish club called Zgoda, I spoke Polish and I ate Polish foods”, and ends with assertions of Majka's love for Canada, followed by “Well I could go on but I gotta take my kids to school and then meet my Canadian friends for coffee;)”. Here, drinking Polish vodka with Polish friends in the past is contrasted with drinking coffee with Canadian friends today. This stylistic move helps Majka to construct her identity as changeable and something she has agency over—a further point that I develop below.

In the next sections, I first discuss the contexts of belonging that emerge in the three narratives and the participants’ alignment with respect to these, including the way their answers respond to and interact with each other. Then I discuss the relationship between essentialist and nonessentialist conceptualizations of identity, and the authors’ construction of passive or agentive roles in their own life story. I focus on each narrative in some detail, referring to specific tactics of intersubjectivity and linguistic resources employed by the authors.

CONSTRUCTING BELONGING AND DIFFERENCE

The contexts of belonging that emerge as salient in my data, and that appear in the three narratives discussed here, include time, place, ‘in’ and ‘out’ groups, and language. Time is primarily divided into that spent in the refugee camp in Italy, and in the final destination country. The latter is the most relevant in the answers below, and is further subdivided into stages of immigration. The context of place includes Poland, Italy, and final destination. There are multiple ‘in’ and ‘out’ groups that emerge in the data, but in the three narratives below they are mostly constructed around ethnic and national affiliations. At times these are challenged and reinterpreted, which we see to an extent in Majka's narrative, and much more extensively in Ada's. Overall, the three narratives are in a dialogue opened by the first response, and can be seen responding to each other in both structure and content.

1. (1) Ola's response

Ola's answer, coming first, establishes some of the parameters for how the contexts of belonging are structured in the subsequent responses. In her story, the context of time postimmigration consists of three salient stages: the first immediately after arriving, the transitional second stage while still living with her parents but in a new location, and the final stage of establishing herself as an independent individual. In (1), Ola sets up these three stages when she says, “I lived in Detroit for 3 years when I came to the States”, and then “After 3 years in Detroit I moved to Georgia where I lived in the deep suburbs at first and then moved downtown for college”. Majka appears to accept and respond to this time-frame structure in her account. Majka opens with, “When I first came to Canada”, then introduces her time in college as a transitional stage, and ends by explicitly fast-forwarding in time and shifting to the present tense: “Now 16 years later I have 2 Canadian kids and my best friends are Canadian”.

Ola does not appear to differentiate between the three stages of immigration in terms of her own attitude towards them or alignment with them. Rather, each stage serves as an example of Ola's alienation from her environment in the United States. Majka, however, appears to align herself most closely with her ‘current self’ (Weldeyesus 2007), as opposed to the early stage of immigration during which she did not care for friendships with Canadian people.

1. (2) Majka's response

In addition to the time-frame, Ola also proposes a set of ‘in’ and ‘out’ groups by juxtaposing ‘Americans’ to everyone else including ‘Africans’ and “American people that have the ‘European mentality’ ”. Significantly, ethnic and national categories are invoked here in response to my question, in which I set up a distinction between Polish, American/Canadian/Australian, and ‘something else’ identities. The specific categories, however, emerge from the participants’ narratives, as does the way in which the participants construct the meaning of these categories.

Majka accepts the framework of juxtaposed ethnic groups as she contrasts ‘Poles’ with ‘Canadians’, and comments on how the former in her community viewed the latter (“my parents and their friends attitude towards Canadians was not very good. They made me feel like we Poles were superior to them”). This response highlights the point that identities are constructed contextually, and shift depending on the reference points with which they align or disalign. Majka sets up the opposition between Poles and Canadians, but then separates out her parents’ generation from the overall Polish group, and positions herself against them in the phrase “they made me feel”. At the same time, she describes engaging in practices that identified her as a member of the Polish group, and set her apart from Canadians.

The fact that Majka's account strongly parallels Ola's in its structure may be a function of their spatial proximity in the conversation. Ada, by contrast, who joins the conversation much later, does not adopt the time frame and the ‘in’ and ‘out’ group structure of the first two responses. Instead, she begins by discussing her current feelings and thoughts regarding ethnic categories. She starts with, “I feel somewhere between Polish and Australian, however I think I feel a citizen of the planet earth first and foremost, and only then I associate with a particular country”, and continues, “I guess what I'm trying to say is that I do not really experience extremely strong feelings of patriotism towards a particular country. I try to take what's best from Poland and Australia, and Italy as well”. The one time that Ada refers to the context of time, she introduces an alternative time frame, contrasting time spent in Poland, in Italy, and in Australia. In a way, it appears that Ada responds more directly to my initial question, which was framed as ‘Would you describe yourself as (ethnic/national category)?’.

Ada also introduces the context of place, as can be seen in the above quote. Unlike Ola and Majka, she moves from talking about ethnicity/nationality as defining groups of people, to discussing geographical places and their relationship to her life story. She continues: “When I think about childhood I definitely associate with Poland, teenage years with Italy… and… working and becoming a parent with Australia”.

1. (3) Ada's response

Furthermore, Ada does not set up distinctions between ethnic groups, and does not establish her Polishness as competing or contrasting with her Australianness. Instead, she focuses on the cross-cultural and cross-ethnic interactions that she portrays as characterizing life in Australia. She recounts, “My early friends in Aus were Polish mostly, but also Italian or of Italian descent, South American or from other countries. I suppose I felt that I had a lot in common with other migrants or kids of migrants”. Then she reflects, “Multiculturalism is very much celebrated in Australia. I guess you could say I am Australian because I feel partly Polish”.

ESSENTIALISM, NONESSENTIALISM, AND AGENCY

Ada articulates a perspective on her ethnic/national identity that seems to align with nonessentialist approaches that view identities as complex, fluid, relational, and enacted through sociocultural practices, in contrast to essentialist perspectives that see members of identity categories as ‘fundamentally similar to one another and fundamentally different from members of other groups’ (Bucholtz & Hall 2004:374), and the groups’ perceived characteristics as ‘natural’ and inevitable attributes of their members. As Bucholtz & Hall (2004:376) point out, however, essentialist views of identity are often very real and relevant to members of particular communities. Essentialism frequently informs how individuals make sense of their social reality, and how they position themselves with respect to identities and ideologies that are salient in their lives. Accordingly, Ada's nonessentialist analysis in this narrative is only one way in which the participants envision their social groups and relationships.² Indeed, the three narratives presented here illustrate different degrees to which essentialism becomes an explanatory tool for the participants. The women also claim different degrees of agency, which is expressed in word choices and grammatical constructions; in particular, by choosing state or action verbs, that is, ones that assign thematic roles of either experiencer or agent to the subject. It appears that the essentialist view of ethnic and national groups as fixed and of their characteristics as inherent tends to be accompanied by the use of verbs and constructions that frame the subject as passively acted upon by inevitable circumstances. This is similar to Meinhof & Galasiński's observation that their Polish interviewees tended to use impersonal constructions such as stało się ‘it happened’, which remove agency from the subjects (2005:27). In contrast, the teller's claim of agency in shaping her own experience, expressed through verbs and grammatical constructions that frame the subject as an active ‘doer’, co-occurs with less essentialist portrayals of ethnic and national groups.

Ola

In Ola's response in (1), people are constructed as belonging to national and supranational units that are connected with distinct, separate cultures (Holliday et al. 2004). Ola sets up a dichotomy between American people and culture on one hand, and the rest of the world on the other. She opens by saying, “I will die feeling Polish or at least European or even latin American, but never American”. This juxtaposition is strengthened by Ola's punctuation: the only comma in the sentence visually separates American from the other national-cultural designations. Ola elaborates: “American people have a totally different culture. It's not bad just different than the rest of the world it seems. I even have more in common with Africans than I do with Americans”. When an American, in this case Ola's husband, behaves in ways that Ola can perhaps identify with, she describes him as having “the ‘European mentality’ ”, thereby denying authenticity to his American identity. American culture is thus reified as a fixed, uniform category. At the same time, Ola says little about specific characteristics constituting American culture or ‘mentality’, other than mentioning football, baseball, and Greek organizations on college campuses (“I even went through the rush experience”) as aspects of American culture that she cannot relate to. In the middle of her narrative, she describes qualities that she observed among the Polish community in Detroit, and that she perceives as un-Polish and resulting from American influence: “People were all for themselves and seemed isolated. Everyone cared about how much they had and if the other had anything more (or even different than them) it seemed to be a problem”. What Ola seems to describe is a lack of community, and perhaps a sense of competitiveness over material possessions or socioeconomic status. These are qualities found among many human groupings, but Ola constructs them as products of Americanization. She contrasts them—along with American sports and Greek organizations—generally with attributes of all non-Americans (Europeans, Latin Americans, Africans), and more specifically with particular examples of what she portrays as Polish culture in the last sentence of the narrative: “I missed my Polish friends (which there were none), cooking Polish food together, playing instruments, singing around the fire”. These practices, again, are not necessarily exclusively Polish, since many communities around the world gather around camp fires with music and songs. Nonetheless, in Ola's narrative they become symbols of Polishness, which is presented as lost in the past in the phrase I missed.

Ola establishes a distinction between herself and those she identifies as representing American culture, and simultaneously she erases differences among non-Americans. By using the first-person I missed, she identifies with the cultural practices she ascribes to the Polish ethnic/national group as ones she recognizes and has taken part in. She also establishes similarity between herself and all non-Americans, including Europeans, Latin Americans, and Africans. In Bucholtz & Hall's (2004) terms, Ola employs tactics of adequation with non-American ethnic/cultural groups, while emphasizing distinction between them and Americans. These essentialist representations of people as members of ethnic, cultural, and national groups are accompanied in Ola's discourse by the image of isolation, and by verb choices and grammatical constructions that convey Ola's passive role in her personal story. Americanness is linked to isolation three times in the short narrative. Also three times, Ola uses the verb felt: she “felt a culture shock”, she “felt the difference”, and “even living downtown amongst people felt isolating”. Since the stative verb feel assigns the thematic role of experiencer—‘the animate being affected inwardly by a state or action’ (Brinton & Brinton 2000:299)—to the subject, its use in Ola's narrative positions her as passively responding to external circumstances or events, which are the cause of her feelings. In the third example, Ola is not even the subject of the sentence; rather, she is the implied indirect object (we can assume that she means “felt isolating to me”), which, if it had been explicitly included, would take the same role of experiencer. This is akin to Ochs & Capps’ observations regarding the verb got in statements such as “I got very upset and I must have got a black-out”: ‘This verb positions the protagonist (who is also the teller) as one who passively experiences two unpleasant conditions: being ‘very upset’ and blacking out. The verb ‘got’ is typically used to make a protagonist look like a victim in a situation that calls for accountability’ (2001:49). The feelings experienced by Ola are also unpleasant: a culture shock, a negative difference between Poles in Detroit and (presumably) in Poland, and isolation. At one point, Ola also presents ‘the adjustment to being isolated’, rather than herself, as the subject of the verb, thus emphasizing her role as a passive experiencer of unpleasant feelings: “The adjustment to being isolated never came”.

Majka

In (2), Majka structures her answer around the time frame and the kinds of ethnic/cultural/national groupings introduced by Ola. The context of time for her consists of three phases: (i) “When I first came to Canada”, (ii) “I started college and started meeting some Canadian people”, and (iii) “Now 16 years later… in my heart I am Canadian”. This follows Ola's framework, which sets up the same three stages: life with her family in Detroit, immediately postimmigration; life in Atlanta with her mother; and her current life as her own individual, married to her American husband.

In Majka's account, the first stage is annotated by very specific Polish cultural markers: Polish friends, Polish vodka, Polish club, Polish language, and Polish food. Majka opens with the same verb as Ola, feel, perhaps echoing Ola's word choice: “I was feeling very much Polish”. However, this is followed by a list of action verbs that illustrate how Majka enacted her Polishness (“drunk Polish wodka”, “spoke Polish”, “ate Polish foods”), all of which assign the thematic role of agent to the subject, Majka, thus positioning her as a ‘do-er’ and suggesting that she is claiming agency over her feelings. While Majka follows Ola's structuring of the context of time, and of ‘in’ and ‘out’ groups (Poles in opposition to Americans or Canadians), she frames her account around her own actions and her own transformation. Indeed, even though Polishness and Canadianness are each linked to a set of stereotypical cultural markers such as vodka or hockey, the main theme in Majka's account is changeability. She moves from feeling and acting “very Polish” in the early days of her life in Canada, when she was influenced by her community's negative opinion of Canadian people, through recognizing that “Canadians are not so bad after all” once she made Canadian friends, to embracing herself as Canadian following her marriage. Things do not just happen to Majka in these stages, but rather she interacts with her changing environment, actively making new friends and associating with Canadians.

Majka's narrative breaks with a fully essentialist view of identity at several points. First, she allows for a more complex, hybridized ethnic/national identity for herself, both in her story of morphing from a Pole into a Canadian, and in her explicit statement: “I don't deny my roots and never will. I will always be Polish but in my heart I am Canadian and very much proud of it”. Majka is not positioning herself against Canadians; rather, she is describing herself as learning how to embrace both, and to accept that her identity may accommodate both. At the same time, she sets up a distinction between her former and present selves, and evaluates the former as somewhat closed-minded and misguided. She recounts that in her first years in Canada, when she “was feeling very much Polish”, she had no interest in making friends with Canadians because her Polish community otherized them as culturally inferior. She is critical of this disdainful attitude (“my parents and their friends attitude towards Canadians was not very good”), and goes on to describe Canadians as “amazing people” with whom she identifies, since she has “2 Canadian kids” with her “as Canadian as they come” husband. In a study of narratives of Ethiopian immigrants in the US, Weldeyesus (2007) observes that his participants construct their present selves in opposition to earlier, less acculturated, and less sophisticated selves. Similarly, Majka juxtaposes her former self, who aligned herself with the Polish community and its practices and rejected Canadians, with her present-day, English-speaking self who is more open-minded and able to appreciate multiculturalism and Canadianness. This juxtaposition forms a tactic of distinction, not just between Majka's former and present selves, but also between successfully acculturated immigrants such as Majka, and those who do not integrate and continue to position themselves against local Canadians or Americans.

In addition, Majka employs the tactic of denaturalization when she severs the ‘common-sense’ connection between ethnicity and language: “Honestly the only person I still speak Polish to is my mom. I don't deny my roots and never will”. In this later stage, following her transition towards embracing both Polishness and Canadianness, Majka claims identification with her Polish roots even as she admits to only speaking Polish with her mother.

Majka's narrative emphasizes agency and departs from essentialism by allowing for the fluidity of ethnic and national identities, and in this way contrasts with Ola's. In addition, in Majka's narrative there are no references to isolation, which features prominently in Ola's account. Instead, Majka's story refers to friendships and socializing. Nonetheless, Majka's time frame and the construction of ethnicity and nationality as important for categorizing people responds directly to the parameters set out by Ola.

Ada

Ada's narrative in (3) comes days later, following not just everyone else's responses, but also unrelated conversations among participants. As already discussed, it breaks with the time-frame and ethnic/national category scheme set up in the first narrative (Ola's) and reaffirmed in the second (Majka's). Ada's account also constructs a complex, hybridized migrant identity that Ada affiliates herself with. She tells us, “I try to take what's best from Poland and Australia, and Italy as well”, and later offers the insight, “I suppose I felt I had a lot in common with other migrants or kids of migrants”. Ethnic, national, and regional identifiers (cf. De Fina 2000) are used not as primary tools for grouping people, but as relevant mentions in a story about multiculturalism, and there is no evaluation of any ‘inherent’ qualities associated with ethnic/national groups. Instead, Ada rejects the idea of belonging to one: “I think I feel a citizen of the planet earth first and foremost”. At the same time, her account does not erase differences among groups, or construct “citizen of the planet earth” as a uniform or supra-national category. Rather, her identification with “the planet earth” emerges out of her complex and blended experience with multiple cultures, places, nationalities, and languages. This complexity is enacted in particular when Ada claims ownership of her Polish accent, rejecting the idea that it may be incompatible with her Australian identity. Here, at the end of her narrative, she sets up a juxtaposition between Australians who identify with multiculturalism (her ‘in’ group) and those who “do not appreciate this concept”. Her alignment with multiculturalism and with complex identities finds expression in her exaggerated representation of /r/ as a prominent trill in her accent: “I'll keep my accent, thank you verrrrrry much!”. This claiming of Polish-accented English as part of her Australianness is accompanied by many similar verb choices to Majka's, namely ones that position Ada as a ‘do-er’. For example, she uses verbs such as discover (in the sense of explore) and choose when discussing her experience: “I loved discovering Australia”, “I chose Australian History”, and “I chose my friends for the kind of people they were, not their nationality”.

Like Majka, Ada presents herself as taking agency over her immigrant experience, emphasizes actively making friends with non-Poles, and expresses appreciation for cultural diversity in her new homeland. By contrast, Ola's account focuses on the passive experience of feeling isolated, and this isolation is presented as stemming from the incompatible distinction that Ola repeatedly constructs between Americans and non-Americans.

CONCLUSION

The Facebook group message shared by Ola, Majka, Ada, Anna, Sonia, and me, constitutes a space in which the present-day friendship among the women is being built. The particular space of the Facebook group message structures the interaction in specific ways. Participants can always review the conversation before contributing, even if they were absent for many turns, and engage with comments made much earlier. As we saw in the narratives in (1)–(3), participants tell their stories in response to an interviewer's questions, but also to previous contributions. The framing of online interactive contexts as spaces (De Fina 2016) highlights the idea that they are not only types of interactions, but also spaces that can be visited, reviewed, and reorganized.

Reconstructions of the past through the sharing of memories are opportunities for individuals to make sense of their life experiences, and we see this in (1)–(3) as the participants respond to the question ‘Would you describe yourself as (ethnic/national group)?’ with personal narratives as opposed to other answer formats. As Mishler points out, we constantly revise and re-revise our life stories ‘in terms of ever-widening contexts of what happened later’, and in this process continuously reconstruct our multiple identities, ‘each rooted in a different set of relationships that form the matrix of our lives’ (2006:41).

The intertextuality, multimodality, and translocality that characterize online interaction are also features of the superdiverse reality produced by increased movements of people and global intercultural connections made possible in part by new technology and new media. The very possibility for former fellow refugees to reconnect decades later while living in vastly different regions of the world, and to engage in ongoing real-time interactions through which their friendship can be reconstructed, is afforded by the existence of the internet and of social media platforms such as Facebook. The analysis in this article contributes to the still small but growing body of research on narratives in social media, and also to the study of migration and identity in the globalizing, superdiverse world more generally.

By exploring how participants construct their life stories in personal narratives of migration, this study highlights the contextualization, fluidity, and translocality of identities, and of migrant identities in particular. Meinhof & Galasiński's (2005) framework of contexts of belonging proves useful in examining the various dimensions of identity that emerge in the narratives, and with which the participants establish similarity or difference. The analysis of examples in (1)–(3) is particularly instructive for understanding different ways in which people make sense of ethnicity and nationality, and how their personal models of these are constructed through the use of specific linguistic resources. As we saw, the notion that belonging to an ethnic or national group is an inherent state of being rather than a choice tends to occur with state verbs presenting the speaker as an experiencer rather than an actor, and with an emphasis on the experience of isolation and alienation. By contrast, when belonging is framed as a choice, and ethnicity and nationality as fluid, the chosen verbs tend to position the subject as an agent, highlighting her proactive role in her own story. The emphasis also tends to be on breaking out of isolation by making friends, dating, and raising a family outside of the Polish context. This brings us to my final point, which is that the narratives discussed here demonstrate the complex ways in which people respond to and make sense of the process and consequences of migration. As I have argued elsewhere (Baran 2017), an emotional experience of dislocation and of a fragmented or ‘lost’ sense of self is common among migrants who are faced with a life in a new cultural context and in a second language (cf. Pavlenko & Lantolf 2000). Part of the migrant experience is situating oneself within this new context. In narratives of migration, this situating of one's migrant identity is accomplished discursively through recounting, revising, and reinterpreting one's migrant journey.