Abstract
The Gezi Park demonstrations across Turkey in the early summer of 2013 offered another opportunity to examine the role played by social media in a social movement. This survey of 967 ethnic (Turkish or Kurdish) minorities living in Belgium, the Netherlands, and Germany focuses on attitudes and behaviors alongside uses of offline and online networks to make connections with others during and after Gezi. We investigate whether the respondents living in the diaspora experienced communication-generated social capital. We also examine whether the social capital already built through lives spent in Europe, where connections to majority populations had been forged, was at least temporarily reversed through a process of re-bonding, as ethnic minorities turned their attention and loyalty to the social movement in Turkey.
Introduction
Understanding the media’s role in building or failing to build social capital has been a topic of discussion even before Robert Putnam (2000) wrote Bowling Alone, where he made the case for increased mass media as a cause of reduced social capital among Americans. Putnam (2000) defined social capital as the “connections among individuals—social networks and the norms of reciprocity and trustworthiness that arise from them” (p. 19). The concept was thought to be the key to social movement mobilization by many scholars in the years before and since Putnam’s publication, notwithstanding the criticisms regarding its conceptual clarity, or lack thereof, and reliable and valid ways to measure it.
Seen from the perspectives of women and cultural minorities, Barbara Arneil (2006) has articulated several differences between Putnam’s positive view of social capital and that of other scholars (e.g. Bourdieu, 1986). Bourdieu (1986) did not see social capital as a group of individuals deciding to connect with one another, but rather as a concept considering the economic, cultural, and social forces that “limit the range of possibilities that certain individuals or groups have for creating or drawing on resources inherent in them” (Arneil, 2006: 201). Focusing more on the inequalities, Bourdieu examined the nature of the connections between people in their historical context rather than the number of connections, as Putnam did (Arneil, 2006). James Coleman (1988) sees both positive and negative aspects of social capital but in his research on high school dropouts he concludes that the public good property of social capital differentiates it from other forms of capital: “the actor or actors who generate social capital ordinarily capture only a small part of its benefits, a fact that leads to underinvestment in social capital” (p. 119). Although we accept the critiques and modifications to Putnam’s conceptualizations, we focus here on his notions of bridging and bonding, first articulated in his book written at the turn of the century.
More recent scholarship examines the ability of individuals to build networks, at least partially through online connections in social media (Ellison et al., 2011). Putnam (2000) mentioned that network relationships with close friends and family are maintained through a process of “bonding,” while relationships with those less well known or formerly unknown to us are established through “bridging.” As Granovetter (1973) put it in his earlier work on individuals seeking employment, strong ties are those that represent the bonding relationships, while weak ties, which may ultimately work best when trying to reach a goal, such as finding a job, are established through bridging to others with whom we have less frequent contact.
In this study, we examine changes in social capital that may already have been built by Turkish ethnic minorities through the years of contact with the majority populations in Europe when a disruptive event reconnects them with their families’ cultural roots. We ask whether social capital was being maintained or demolished among the Turkish diaspora in Europe by supporters and opponents of the Gezi anti-government protests in Turkey in the summer of 2013.
We measure bridging and bonding social capital both online and offline, noting as Putnam (2000) did, that they are not “either-or categories into which social networks can be neatly divided, but more or less dimensions along which we can compare different forms of social capital” (p. 23).
Based on a survey of 967 ethnic minorities (Turkish/Kurdish 1 ) in Belgium, the Netherlands, and Germany, we studied attitudes and behaviors related to those protests alongside the use of offline and online networks (i.e. temporary use of strong and weak ties or social capital) before and during the demonstrations. Since a great number of the respondents were born and raised in Europe or spent most of their adult life there, it is likely that they had already become substantially integrated into the several country-based networks, thus relying on social capital previously built through bridging with members of the majority population as they engaged with the Gezi social movement.
This research operationalizes the concept of social capital in the context of social and traditional media used for political mobilization. We define mobilization as online and/or offline participation in the protests. Both supporters and opponents of the Gezi demonstrations consumed information from traditional media in Europe and in Turkey. Social media were used to communicate with others to encourage their participation in demonstrations and forums and to access information about developments in Turkey. We asked questions related to Gezi supporters’ and opponents’ bridging and bonding activities. As scholars have established (Ellison et al., 2011), the weak ties established through social media make it possible to connect to others with whom one has little or no direct contact. We ask whether the connections made through the Gezi protests allowed respondents to bridge to those who had information or expressed feelings of support for the positions regarding Gezi. Alternatively, respondents might also have forged stronger bonds with close friends in the diaspora or in Turkey as they stayed current with the evolving situation on the ground (Ellison et al., 2011). We investigate whether the respondents experienced communication-generated social capital. Furthermore, we examine whether the social capital already built through lives spent in the diaspora (in school, at work and in the community), where connections to the majority population had been forged, was at least temporarily reversed, as ethnic minorities turned their attention and loyalty to the social movement in Turkey.
When Putnam examined the state of social capital, he concluded that an important factor for the decline was the increased time spent in television viewing, which is largely a private or individualized experience, leaving less time for civic engagement. Putnam (1995) found that watching television “dwarfed all other changes in the way Americans passed their days and nights” (p. 72). But that was before the introduction of social media. Now we can interact with others online and use those connections for civic engagement perhaps making up for the time watching television.
Putnam (2000: 175) was also skeptical about the value of communication via the Internet because of the lack of nonverbal information. More recent work by Sajuria et al. (2015: 733) contradicts this assertion, finding that bonding social capital online is greater than assumed in theoretical models. The authors also dispute Putnam’s (2000) point that online communication only increases bonding as a form of “cyberbalkanization” (or the confining of Internet conversation to “people who precisely share our interests”) (p. 177). However, Sajuria et al.’s (2015) analysis of tweets posted in several social movements found that “the presence of organizations and professional brokers is key to the formation of bridging social capital” in the online environment (p. 710).
Because the respondents in our study have strong ties to both the community of their family origins and that of the society in which they live, we see the two processes of bridging and bonding as somewhat more fluid than did Putnam or Sajuria and his co-authors. Although many of the individuals in our study were born and educated in Europe, they remain rooted in their ethnic communities where cultural, financial, medical, and social services are offered in their parents’ native languages, and where bakeries, restaurants, and a range of shops that specialize in the food and clothing from Turkey are available. This blend of cultures in the lives of our respondents as it relates to bridging and bonding reinforces the idea of a conceptual continuum. Where Putnam might have assumed a more or less homogeneous society in his study, we examine a group of people who live in Europe but have a mixture of identities. For both those born in Europe or Turkey, we argue that they first bonded with their families’ culture particularly in those families where Turkish was the language of the household (66.4% of the respondents still spoke Turkish or Kurdish at home at the time of the survey). Once they entered school or began to interact with children outside their neighborhoods in Europe, they began to bridge to the dominant culture.
We argue that “cyberbalkanization” took place among the ethnic Turkish minorities, but refer to it as “re-bonding” to others with connections to Turkey. Both participants in the protests and those who opposed them were drawn into the sphere of their Turkish roots more intensely when Gezi exploded and caused the ethnic minorities to focus outside of their European homes to the upheaval in Turkey. Media, both traditional and online, also played a role in this process, just as they did for Putnam, as vehicles for accessing information, making contacts with others who held similar opinions and organizing activities in support of the movement or building solidarirty with the government in Turkey (Ogan et al., 2016).
This study also examines the European context for respondents’ attitudes and behaviors. In Turkey, protestors were met with tear gas and water cannons during their protests. In the more liberal European environment, supporters and opponents of the events in Turkey were free to express their opinions face-to-face with one another in the streets and at publicly organized forums in the European countries where they lived. As a result, the social movement continues somewhat unabated in Europe, while it has been silenced or physically crushed in Turkey. Although many studies have examined the building of social capital by migrants and ethnic minorities, we believe that this is the first to study such a group becoming engaged with a political and social movement occurring outside the country where they live and also the first to examine social media use within the social movement as experienced by those in the diaspora. Putnam (2000: 152–153) believed that social capital plays an essential role in social movements and that social movements can also create social capital by fostering and extending social networks. Therefore, it seems key to our study that we explore the nature of social capital in the Gezi movement.
Backdrop for the study
In the morning of the 30 May 2013, people in Turkey witnessed the beginning of the largest public uprising in Turkey’s social movement history. What started on 28 May as a group of around 50 activists gathering in the Taksim area of Istanbul to contest the urban development plan for the transformation of Gezi Park into a shopping mall and reconstruction of Ottoman-era military barracks soon escalated into a country-wide social movement. It was quickly transformed from a narrowly focused environmental protest to a referendum on the authoritarian policies of the government led by Prime Minister Recep Tayyip Erdoğan. The demonstrators took to the streets in multiple cities, only to face water cannons, tear gas canisters, and plastic bullets.
Mainstream national public and private television channels either refused to cover the demonstrations or under-reported the scale of the activity. Under extreme government pressure, print and broadcast media imposed self-censorship that has increased over time. For those outlets daring to report the actual situation, the government forced the resignations of the offending journalists and sanctioned their media companies. Consequently, protesters encouraged Turks to turn off their television sets in a Twitter campaign with the hashtag #BugünTelevizyonlarıKapat (Turn off Televisions today), which was used in more than 50,000 tweets (Devitt, 2013).
Mainstream media censorship provided the impetus for new online media platforms that were used as the primary source of information about the demonstrations. Abandoning traditional media, the protestors preferred these online sources because they were more trusting of the accuracy of the information (Tunç, 2014).
Role of social media in social movements
The social movement on which we focus was located in Turkey, but also took place in the diaspora in several European countries where support for the Turkish nation remains strong across the three generations of ethnic Turks and Kurds who live there. According to many newspaper accounts, when Gezi broke out, hundreds to thousands of people gathered to protest in the streets of Brussels, Amsterdam, Stuttgart, Berlin, Hamburg, and many other European cities (International Occupy, 4 June 2013). An analysis of tweets conducted during the protests indicated intense online participation from people across Europe (Varol et al., 2014). Street protests continued into 2014 in Europe and Gezi-based forums attended by the researchers are still being used by movement activists in several European countries.
Also, scholars studying the Arab Spring claimed a powerful role for social media (Howard and Hussain, 2013) despite the on-the-ground causes that provoked the demonstrators in the first place. Writing for The Guardian, Beaumont (2011) provided a balanced view of the situation. “Often, the contribution of social networks to the Arab uprisings has been as important as it also has been complex, contradictory and misunderstood.” Systematic research of various social movements has supported that position (see Gerbaudo, 2013; Lim, 2012; Theocharis et al., 2015; Tufekci and Wilson, 2012).
Based on Bakardjieva’s (2009: 4) concept of subactivism, or the participation in inconspicuous processes like identity construction related to social and political discourses, Valenzuela et al. (2012) argue that “social network sites therefore emerge as resources that may create the kinds of collective experiences that are necessary conditions for successful protest movements” (p. 303). Valenzuela et al. (2012) examined the relationship between Facebook use by 18- to 29-year-olds and participation in the Chilean protests of 2011, finding that “having a Facebook account and using it frequently were positively and significantly related to participation in protests, even after taking into account other known sources of this type of political action” (p. 308). When youth used Facebook for news and for socializing with other youth, they were more likely to participate in protests, but when they used it for self-expression, no such relationship emerged (Valenzuela et al., 2012: 310).
Breuer (2012: 25–26) also found that social media played a significant role in the Tunisian revolution. Based on expert interviews and a web survey of Facebook users, she found that Facebook use allowed a digital elite to form personal networks who were able to bypass the censored national media and report events and other information online as well as directly to the international mainstream media. In Facebook conversations and other Internet-based communication, networks were formed that combined the intellectual elite, the rural poor and the urban middle class.
The missing piece: social capital building by ethnic minorities in social movements
The aforementioned research builds a case for the importance of social networks in enabling offline networks to function more efficiently in a variety of social movements. We will also examine that relationship in our study of the supportive networks in the Turkish diaspora. However, as discussed, we were equally interested in examining the use of these newly formed or recently reinforced networks of weak and strong ties in building or realigning social capital. To reiterate then, bonding occurs within homogeneous networks of like-minded people (such as the Turkish ethnic minorities in Europe) and is likely to result in emotional support for the individual, while bridging facilitates the dissemination of information and solidarity among rather weak ties and is likely to lead to social and political participation within the majority society. We believe that this could have occurred in our study if the Turkish diaspora was joined by non-Turkish Europeans in their organizations and activities related to Gezi. Whatever the case, it is important to obtain a clear picture of the nature of social effects of online communication via social networks.
Recent research examines the relationship between social capital and social media use in a protest environment (see, for example, Gil de Zúñiga et al., 2012; Park et al., 2009; Valenzuela, 2013). As best summarized by Theocharis et al. (2015), the positive relationship between the use of social media and the building or maintaining of social capital in these and other studies is based on several factors including their capacity to reach a critical mass, to promote personal and group identities relevant for political behavior, to function as information hubs allowing to create and join groups based on common interests, to enable people to encounter more opportunities to engage in political activities and to build trusting relationships among members enhancing the “protest potential” (p. 211).
Social capital has been cited as a key variable in the integration process for ethnic minorities in various countries around the world but little of the research has focused on social media use as a way to bridge or bond social capital in the diaspora. Some research studied the use of the Internet (but not social media) by migrants and ethnic minorities for bridging and bonding social capital (d’Haenens et al., 2007; Ogan and d’Haenens, 2012). Ogan and d’Haenens’ study of ethnic Turkish women in Belgium and the Netherlands found that the use of the Internet and the mass media had limited advantage for women who lacked education and skills in trying to bridge to the majority population. Nevertheless, d’Haenens et al. (2007) found that migrant Moroccan and Turkish youth in Flanders demonstrated bonding activity with family and friends in their families’ countries of origin, but also bridging activity to majority population youth in the countries where they lived.
The theoretical orientation to this study brings together three streams of research: social media use and social movement participation, social media use and its relationship to social capital and social capital building through social media use by diasporic populations. Previous work on communication in social movements has been based on demonstrations supporting change that occurred in the same country where the goal of the movement was targeted, while our study is of sympathizers who live far from the impact of the government policies that triggered the protests.
Drawing on the findings from previous research, we hypothesize the following:
H1a. A reduced amount of bridging to members of the majority population (through offline and online networks) as compared with pre-Gezi levels will be reported by the Turkish diaspora who either supported or opposed the Gezi movement in Belgium, Germany, and the Netherlands during the time of Gezi.
H1b. Increased online and offline bonding (or re-bonding) to other Turkish/Kurdish minorities in the diaspora or in Turkey as compared with pre-Gezi levels will be reported during the social movement.
We ask the following research questions:
What is the profile of the pro-Gezi supporters of the Turkish diaspora in Belgium, Germany, and the Netherlands?
What is the profile of the anti-Gezi members of the Turkish diaspora in Belgium, Germany, and the Netherlands?
Do those who use social media more frequently increase their bonding activity offline? What differences in the building/failure to build social capital during Gezi can be found between those who supported and those who did not support the movement?
Methodology
In November 2013, we fielded a survey of Turkish and Kurdish ethnic minorities whose families emigrated to Europe from Turkey. Participants in the survey were either first generation immigrants or children/grandchildren of workers who left Turkey to make their homes in Europe. A large number of the respondents, who live in Belgium, Germany, and the Netherlands, were participants in some form of protest—in street demonstrations, in forums, in Facebook-based organizations created for protests and in other social media such as Twitter. But just as in Turkey, about an equally large number opposed the protests. The opposition was not publicly active, but did post Tweets, made comments on Facebook and in other social media and discussed the events in offline religious and civic organizations.
Our division of respondents into the pro- and anti-Gezi groups was not an arbitrary decision. In the months leading up to Gezi (and increasing since that time), Turkish society has become overwhelmingly polarized. Many attribute this political and social division to the current government, the Justice and Development (AK) Party-run majority. Researchers and analysts had been reporting on the divisions in Turkish society brought about by the directions taken by then Prime Minister Recep Tayyip Erdoğan even before the Gezi protests (Taşpınar, 2012; Tepe, 2013). In a 2015 Pew survey of Turkish citizens, respondents were split evenly between those who believed democracy was working in the country and those who did not. Those who were satisfied included 68% of those who pray five times a day and 79% of those identifying with the ruling party (AKP). Those who were dissatisfied included 67% of respondents who said they hardly ever pray, and 77% of those who identified with the major opposition party (CHP).
Respondents were recruited to complete the questionnaire online through Facebook and ethnic organizations we located in social media, and offline in predominantly Turkish neighborhoods where potential respondents lived, shopped, and attended meetings of community organizations. Random sample drawing was not possible because of the nature and location of the population, and no publicly available list of names and telephone numbers exist. Because those opposed to the Gezi protests were more reluctant to complete the survey (and even viewed it as biased given the number of questions related to activist behavior), 31.6% of the final group of respondents was in opposition to the movement while 68.4% were either active or inactive supporters. It would have been difficult to prevent the perceived bias in the questionnaire. To our knowledge, there were no scheduled events, forums or street protests in Europe that supported the Turkish government’s position. Of the respondents who participated in offline protests, 76.6% said they were active in the countries where they lived, while 46.2% attended forums offline and 64.5% participated in such forums online. Respondents who supported the movement encouraged others to become active protestors (84.7%). They also reported increased use of Facebook (86.2%) and Twitter (59.3%) during the protests.
The respondents were nearly evenly split among the three countries (34.3% in Belgium, 32.6% in the Netherlands, and 33.1% in Germany). The survey collection period went from November 2013 through the end of May 2014. Although this long period was not desirable, the difficulties experienced in recruitment made it necessary. Many ethnic Turks were fearful that representatives from the Turkish government might hold them accountable for answering such questions, while others were angry about the demonstrations and viewed the researchers as supportive of the actions of the protestors.
The questionnaire 2 consisted of 65 questions. Respondents could complete the survey in Qualtrics online in multiple languages (English, Dutch, Turkish or German) or offline in the language of their choice. Offline questionnaires were distributed, and often retrieved at a later date, at meetings of organizations, in headquarters of organizations, and at social gatherings in neighborhoods. Offline responses to the survey constituted 11% (N = 107), while 89% (N = 860) completed the survey online. Online, some respondents would begin the study, stop, and then restart at a later time. We believe that we did not have any duplicate completions because Qualtrics assigned a unique identification number that corresponded to IP numbers from respondent computers and Internet connections.
Social capital was measured through several variables in the study. All were related to activity before, during, and even after the Gezi protests. We asked about the numbers and nature of the organizations to which respondents belonged including whether they increased their activity with these organizations, the number of close friends they had from the European majority populations versus those of Turkish or Kurdish ethnicity and whether those connections had increased, and their feelings about belonging to the country where they lived or whether they felt at home elsewhere.
Findings
An overview of the respondent demographics is provided before we report the results of hypothesis testing. Pro-Gezi supporters in Turkey were popularly thought to be left-leaning youth who were not religious and opposed the policies of the AKP government. The prime minister referred to them as “marjinal” or those on the fringes of society rather than in the mainstream and also “çapulcu,” or marauders/looters/riffraff (Kuyumlu, 2013). Those labels were taken up as a mantra by the protestors who added the terms to their Twitter names or used them in reference to themselves on Facebook, thus becoming important memes of the movement.
Konda Research and Consultancy (2013), a respected polling company in Turkey, surveyed those who demonstrated within Gezi Park. Its profile of the typical protestor, formed after analyzing data from 4401 demonstrators, was as follows: the average age of the Gezi demonstrator was 28 years with about an even gender split, about half were college graduates, one out of three was a student and more than half were employed. No party affiliation was declared by nearly 80% of the respondents (Konda Arastirma ve Danismanlik, 2014).
We were interested in the differences between those who were active or passive Gezi supporters and those who opposed the activity. In many respects, the demographics of the groups resemble one another, with a few striking differences (see Table 1 for details). Our respondents (both pro- and anti-Gezi) resembled those gathered in Turkish cities in age (32.6 and 28.2) and gender (with more males in the anti-Gezi group—56.6–51.3%). They were also well educated, with the pro-Gezi respondents having an average of a bit more than an undergraduate degree (and overall a significantly higher level of attainment (V = .18; p = .000) and the anti-Gezi group reporting a bit less than that level. More of the anti-Gezi respondents were students (68.7%) than the pro-Gezi group (55.6%).
Profile of pro- and anti-Gezi respondents.
The Dutch sample included a higher percentage of anti-Gezi respondents (42.0%) than Germany (28.8%) or Belgium (29.2%). Almost twice as many (70.6% vs 37.6%) of the anti-Gezi respondents were born in Europe. The length of time spent in Europe by the two groups was more similar (18.2 years on average for the pro-Gezi group and 23.5 years for the anti-Gezi respondents). Perhaps as a result of a higher number of anti-Gezi respondents being born in Europe, most of them said that the country where they live is their homeland, rather than Turkey (r = .15; p = .000).
Among the largest differences between the groups is their expression of belonging to a faith. Only about half (51.4%) of the pro-Gezi supporters answered that they felt a belonging to a faith, while 94.5% of anti-Gezi respondents answered positively to that question (Tau b = .44; p = .000).
A second large difference between the groups was reflected in their media use during the protests. In a combined variable taking into account use of several social media, the pro-Gezi group emerged as greater users of such media than the anti-Gezi group (53.4% of the pro-Gezi group use it “most of the time” vs 35.1% of the anti-Gezi group (V = .204; p = .000). Traditional media were reported being used “most of the time” by most of the anti-Gezi respondents but there was no significant difference between them (43.1% vs 37.4%; V = .10; p = .06; see Table 2).
Media use by pro- and anti-Gezi respondents in Belgium, the Netherlands and Germany.
What best describes opinion of pro- and anti-Gezi respondents related to media coverage of Gezi in Turkey and Europe.
Finally, a third large difference was in the language reportedly spoken in the home. Most of the pro-Gezi respondents reported speaking only Turkish or Kurdish in their homes (72.0% vs 52.2%; see Table 1). This is a partial outcome of the place of birth of the respondents (Tau C=.27; p = .000) and the number of years spent in Europe. However, most of the pro-Gezi respondents spoke only English or the language of the country where they lived at home (4.4% vs .05%). We first viewed the language spoken at home as an indicator of integration and a sign that linguistic social capital was a result of more bridging activity. About double the amount of anti-Gezi respondents said they speak both languages (i.e. Dutch, French or German in combination with Turkish) in their homes (47.3% vs 23.3%). Later, when we examined the location of the birthplace of our respondents compared with the language spoken at home, it became clear that a much larger percentage of the pro-Gezi respondents were born in Turkey and therefore preferred speaking the language of their birthplace. Out of all the respondents born in Turkey, 86.4% said they spoke Turkish or Kurdish at home. However, even among those born in Europe, 48.6% reported speaking Turkish or Kurdish in their homes. In all three countries, these ethnic minorities tend to live in neighborhoods where shops and services are run by others of Turkish/Kurdish ethnicity creating less need to learn the languages of the countries where they live, a housing preference being coined “voluntary segregation” (see Entzinger, 2014).
We measured mobilization or engagement through their use of social media and other online activities, such as recruiting others through social media, participating in activist forums online and by increases in their use of Facebook and Twitter during Gezi. Offline participation was measured by demonstrating in the countries where they lived or in other European countries, or by going to Turkey to demonstrate. We also asked about face-to-face recruitment, participation in offline forums and participation in other street demonstration related to Turkish anti-democratic policies (Ogan et al., 2017). The comparison between the pro-Gezi and anti-Gezi group as to engagement in offline and online activities shows significant differences. The pro-Gezi group (71.5%) encouraged people to take part in the protests through social media, while the anti-Gezi group hardly showed such engagement (4%). Unsurprisingly, Gezi opponents did not attempt to recruit people to take part in the protests either in person or by phone. The Gezi supporters consisted of people who actively sent out invitations (43.3%), and of people (25.2%) who were hesitant to invite people to participate in the protests in person or by phone. In the pro-Gezi group, a significant number (38.2%) engaged in offline forums, increased their level of Twitter (53.6%) and Facebook usage (77.1%), became members of activist discussion forums (37%) during the Gezi Park movement and were somehow involved in other street demonstrations related to Turkish anti-democratic policies before Gezi (52.3%).
Media use profile
Again, dividing the respondents into pro- and anti-Gezi groups, we find some differences in specific media use. Gezi opponents were slightly heavier Internet users (r = .08; p = .01) and also watched more television from Europe (r = .20; p = .000) and from Turkey (r = .28; p = .01; see Table 2).
But Gezi supporters made much greater use of social media, as hypothesized in (H1b) relying on Facebook (r = .21; p = .000) and Twitter (r = .20; p = .000) more often for their information about Gezi despite the similar percentage claiming to have a Twitter account. Supporters also preferred online and offline newspapers more than the anti-Gezi group (r = .08; p = .04) to obtain news about the demonstrations while the anti-Gezi group preferred television (r = .19; p = .000; see Table 2).
We asked about the bias of the Turkish media, as many people had complained about the censorship of both print and broadcast news. Not surprisingly the anti-Gezi group saw less bias in the Turkish media (35.2% vs 58.4%), and fewer pro-Gezi respondents viewed the Turkish media as reliable and objective (2.2% vs 25.2%). When it came to the European media coverage, the anti-Gezi group claimed more bias (78.5%) than the pro-Gezi group (13.7%). Given that the pro-Gezi group were less rooted in Europe and given their Turkish birthplace, it is somewhat unexpected that European media were trusted more by them than the anti-Gezi group, the majority of whom were born in Europe and who regularly consume the European media at a higher rate.
We asked whether the respondents’ use of social media had any relationship to their offline bonding. Selecting only those who supported Gezi, we found a positive relationship between use of Facebook (rs = .14; p = .001) and Twitter (rs = .10; p = .000) and increased connections with Turkey. A positive relationship was also found between the use of Facebook (rs=.13; p=.001) and Twitter (rs = .10; p = .051) and connections with those in Europe who come from Turkey. The question whether the respondents had increased meetings with those of Turkish origin over concerns or problems related to the country they are living in was positively related to their use of Facebook (rs=.13; p=.001) and Twitter (rs = .12; p = .01), but also related to use of traditional media—newspapers (rs = .11; p = .01) and television (rs = .16; p = .000). As for actual participation in protests, when the protest activity occurred in European countries other than where they lived, only viewing of television was related (rs = .12; p = .01), but when the protest was in their own country in Europe, overall use of social media was positively related (rs = .12; p = .02). We also asked about participation in public forums, which was positively related to Twitter use (rs=.11; p=.05), to the reading of newspapers (rs = .14; p = .01) and viewing of television (rs = .14; p = .006).
The argument is often made that during the social movements, social media are important and directly linked to on-the-ground participation. Although there is some evidence for this in our analysis, no particularly strong relationship was revealed. The lack of such a pattern may possibly be due to the smaller number of respondents who actually took to the streets to demonstrate as compared to using their social media connections alone to express themselves (n = 389 or 61% vs 598 or 94% of the pro-Gezi respondents). So the answer to our second research question is that bonding activity offline was in some ways connected to use of social media but also somewhat connected to use of traditional media.
Connections to Europe and Turkey
In our study, bridging is meant to signify the connections made to the majority population by the group of Turkish/Kurdish ethnic minorities living in Europe who responded to our Gezi survey, covering the period before and during Gezi. For this study bonding refers to the connections made or strengthened as a result of attitudes and behaviors in the Gezi protests by our respondents. The several measures of bridging and bonding in this study can be divided into attitudinal and behavioral measures.
Gezi prompted the development of a range of new organizations in the three countries. We asked about membership in organizations including civic, cultural, professional, feminist, religious, and political. Then we summed the number of organizations to which they belonged. On average, the respondents belonged to 1.4 organizations (with a range from 1 to 12), indicating some degree of social capital building; there was no significant difference between the pro-Gezi and anti-Gezi groups. However, when we asked about whether their membership in organizations had increased since the time of Gezi, or whether they had joined a new organization as a result of Gezi, 13.5% of the pro-Gezi respondents said they had increased attendance during Gezi and an additional 3.9% had joined new organizations compared to only 3.8% of anti-Gezi respondents’ increasing attendance and 1.4% joining an organization during Gezi. Since the new organizations were frequently based on the social movement (as indicated by their descriptions on Facebook), it is likely that more of the Gezi supporters were bonding with others of their ethnicity through these new or increased contacts. In terms of hypothesis H1a, we argue that while the supporters of Gezi have increased their bonding as a result of the events in Turkey, the anti-Gezi group was less likely to make alterations in their organizational connections, therefore maintaining the ties already established. We test that further in a regression analysis (Tables 4).
Predictors of increased connections to Turkey and people from Turkey living in Europe since the Gezi protests.
The negative sign means that they are more likely to be female.
That finding is supported by responses to the questions about increased connections with Turkey and with people from Turkey living in Europe during the time of the protests. Increased connections with Turkey were made by 54.1% of supporters versus 18.8% of anti-Gezi respondents (r = .39; p = .000), while increased connections with friends from Turkey living in Europe were made by 49.5% of supporters and only 13.9% of the opponents (r = .39; p = .000). Similar results were found when questions about increases in face-to-face interactions with those from Turkey in Europe (pro-Gezi = 37.0%; anti-Gezi = 27.3%; r = .14; p = .000). However, the pro-Gezi respondents were also having more face-to-face contact with those from the majority population in Europe (pro-Gezi = 42.5%; anti-Gezi = 22.2; r = .26; p = .000), which signals some increased bridging behavior, probably with those sympathetic to the Gezi movement.
We also asked about the numbers of close friends of the respondents who represented either the European majority or the Turkish/Kurdish ethnicity. More European friends should signal more bridging because the anti-Gezi group had lived in Europe a longer time and had more ties to Europe in other ways; the larger significance would be that they were well integrated. Pro-Gezi respondents listed a smaller number of both Turkish/Kurdish friends (r = .12; p = .01) and European friends (r = .17; p = .000).
Attitudinal measures included the feeling of belonging to the country where they live, the interest in life/politics in Turkey after Gezi, and their attachment to Turkey. Anti-Gezi respondents had by far less feelings of attachment to Turkey than pro-Gezi respondents (r = .67; p = .000), consistent with the lack of close friends in Europe and even those of Turkish/Kurdish ethnicity living in Europe on the part of the pro-Gezi group. Of the anti-Gezi respondents, 36.1% said they considered the country where they lived to be their home, while an additional 43.9% said it was their second homeland. Fewer pro-Gezi respondents viewed Europe as their home (27.1%) or even their second home (40.3%) (r = .13; p = .000).
In the correlational analysis then, it appears that the anti-Gezi residents of European countries were less likely to feel renewed bonding to Turkey during the time of the protests and are settled into their European homes such that they were not as disrupted in their feelings of identity with Europe as were the pro-Gezi supporters who had weaker ties to Europe at the time. That finding is also borne out by the use of European languages in their homes and the citizenship claimed by the respondents—71.6% of the anti-Gezi and 59.9% of pro-Gezi group held, and citizenship in a country other than Turkey or in addition to Turkey. 3 This means that the hypothesis H1a held for pro-Gezi, but not anti-Gezi respondents.
When we conducted a linear regression analysis attempting to predict the connections to Turkey and to those people who came from Turkey (or who claimed Turkish or Kurdish ethnicity) now living in Europe, we found that the significant predictors of that bonding activity were two media use variables (the degree of use of Facebook and Twitter since the protests began) and two Gezi-related variables (their support of Gezi and whether they demonstrated in the streets; adjusted R2 = .31; see Table 4). The results of this regression analysis support other studies that find that social media use during protests is linked to on-the-ground participation in the movement. Both of these led to bonding activity with others who had ethnic ties to the respondents.
Discussion
This study began as an inquiry into the possible changes in social capital that could occur in an ethnic minority as a result of a social movement occurring far away from the lived experience of that minority. Several things prevented a complete understanding of the shifts that occurred. First, as we were unable to obtain a random sample, the group that completed the survey, though large, was likely not representative of the population of the Turkish/Kurdish diaspora in the three countries. Over-representation of younger, more educated, and more active participants in the Gezi protests likely led to greater reporting of the use of social media, actual participation in demonstrations, and several other behaviors. As to measuring changes in social capital, we had no measures of the types of bonds the respondents had formed prior to the study, rather only the respondents’ assessment of change in their connections. The self-reported responses lead us to believe that such changes actually did occur, however temporarily, but more so for the pro-Gezi than the anti-Gezi respondents.
However, if our indicators are accurate, it means that those who may be the best integrated into European societies are those who have lived longer in Europe, speak the language more frequently at home, feel more connection to Europe as their home, and have many close friends who are European, but who also hold strong Muslim beliefs and retain an interest in both the Turkish and European media. We would argue that the respondents’ feelings that Europe is their home are not incompatible with being religious and perhaps also holding conservative beliefs. In other words, they have not become assimilated into European culture and seemingly do not have a problem with that.
Those respondents who supported the Gezi protest movement engaged in more bonding or re-bonding with those living in Turkey or those of Turkish ethnicity living in Europe and the social media, especially Facebook, but also Twitter, allowed them to make or strengthen those relationships. The face-to-face contacts they made while demonstrating in the streets were equally important in that bonding activity.
We hope that future research on this diaspora will be able to obtain wider participation by Turkish/Kurdish residents of Belgium, the Netherlands and Germany in order to obtain a more complete picture of their ties to Europe. This research is part of a larger study that includes network analysis of the respondents’ use of Twitter during the protests.
Footnotes
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This research is funded by the KU Leuven Research Council.