Persistent Activist Communication in Occupy Gezi

Abstract

We revisit the notion of activist persistence against the backdrop of protest communication on Twitter. We take an event-based approach and examine Occupy Gezi, a series of protests that occurred in Turkey in the early summer of 2013. By cross-referencing survey data with longitudinal Twitter data and in-depth interviews, we investigate the relationship between biographical availability, relational and organisational ties, and social and personal costs to persistent activism online and on location. Contrary to expectations, we find no clear-cut relationship between those factors and sustained commitment to participation in the occupation. We show that persistent activist communication did not feed into enduring organisational structures despite the continuous online activity observed during and beyond the peak of the Gezi occupation. The article concludes with reflections on the organisational ramifications of persistent communication and its significance in a political context posing high risks to participation in dissident politics.

Keywords

activism availability Occupy Gezi organisation persistence Twitter

Introduction

This article examines Twitter communication associated with high-risk protest. We shift the analytical focus away from the prevailing interest in the use of social media for participant mobilisation (Anduiza et al., 2014; Enjolras et al., 2012; Mercea, 2012; Tufekci and Wilson, 2012; Valenzula, 2013) to advance an enquiry into activist persistence. Activist persistence is the sustained commitment over time to one or more activist causes by various means and actions (Downton and Wehr, 1998: 534). Rather than continuing to scrutinise activist organisations as the breeding ground for persistence (Bunnage, 2014; Rohlinger and Bunnage, 2015), we concentrate on Occupy Gezi, an ‘episodic mass mobilisation’ (Císař, 2013) assembling ‘ordinary citizens’ (Flesher Fominaya, 2015) that unfolded in Istanbul in late May and June 2013.

Echoing the Occupy Movement, Occupy Gezi was an outburst of collective action in which activist organisations did not have a steering influence on the events on the ground nor on the Twitter hashtags associated with the protests (Demirhan, 2014: 282). In a climate of residual organisational structures underpinning collective action, we investigate the costs of sustained commitment, the circumstances – both structural and personal – that bear on it and its networked communication on social media. Inquiring into individual decisions to persist or cease to be involved with an activist cause, we outline the aspects that carry particular significance for collective action in a context, as was that of Turkey, where both participation in street demonstrations and their networked reportage risked being met with government repression (Budak and Watts, 2015: 375; Demirhan, 2014: 284).¹

Gratifying to the poster, to cautious observers tweeting about protests is short-lived, low-investment and impulsive affective communication (Halupka, 2014), which at best complements long-standing activist campaigning (Morozov, 2011). Symbolic endorsements of or commentary on contentious political issues on social media variably impact action on the ground (see Agarwal et al., 2014; Margetts et al., 2016). Yet, the publicisation of street actions on social media displays the possible motivations for people at large to enact their collective grievances together (Mercea and Funk, 2016). Moreover, techno-social structures erected with digital and mobile media have helped rally disparate activist groups and individuals involved in collective action (Bennett et al., 2014; Treré, 2015).

Complementing the latter research into the mediation of activism (Neumayer and Svensson, 2016), this investigation delves deeper into the practice of ‘hashtag activism’ undertaken for extensive periods of time (Bastos and Mercea, 2015). To this end, the article embraces the theory of activist persistence. We seek to discern the premises and implications of a commitment to tweeting high-risk protests and take a similar methodological approach to the foregoing study by relying on longitudinal Twitter data to select users who tweeted about Occupy Gezi over comparatively long intervals of time. In in-depth interviews, we collected personal accounts that may illuminate tweeting as a communicative practice distinctive for its endurance beyond the flashpoint of street protests. Most interviews took place in Turkey and were later translated into English. The defining aspects of Occupy Gezi are outlined in the following section. After detailing the rationale of this project and its theoretical underpinnings, we present four research hypotheses that are evaluated with a mixed-methods research design.

Occupy Gezi

The Gezi protests erupted on 28 May 2013 at the site of a public park in central Istanbul that had been slated for redevelopment by the central government. In its place, a gleaming shopping mall would arise. The uprising was initially manned by environmental activists who had gathered to oppose the plans, citing a desire to protect public space from commercial splintering. Protestors appealed for extensive public consultations by the government, increasingly regarded as authoritarian (Göle, 2013: 9) and forceful towards dissident protests (Budak and Watts, 2015).

In the wake of a violent police crackdown the following day, the protest mutated into an occupation, attracting a wide cross-section of Turkish society. Rallies in support of the sit-in were subsequently staged in 60 other urban centres (Göle, 2013). Two weeks of clashes, inflamed rhetoric by the government chastising the protestors and experimentation by the latter with collective organisation and self-expression on various social media outlets came to a head on 15 June when Gezi Park and the famous nearby Taksim Square were violently cleared by the police.² Subsequently, acts of dissent were staged in Istanbul and around the country, gradually winding down by 20 June (Demirhan, 2014: 295).

Turkish pollster Konda surveyed the participants in the Gezi Park occupation (N = 4393) from 6 to 8 June.³ Respondents were predominately young (81% were 35 years old or younger), well-educated (54% held a higher education degree, 34% had a high-school degree) and in paid employment (51%) or studying (36%), with far fewer either performing housework (2%) or in retirement (3%). One in five of those participants were members of a party, civic association or Non-Governmental Organisation (NGO), while 93 per cent declared themselves to be ‘ordinary citizens’ who did not represent any group or organisation at the protests.

A sizeable majority of demonstrators either first caught wind of the events in the Park on social media (69%) or were told of them by friends (15.4%).⁴ The social media prowess of the demonstrators was put into relief by the sizeable number of surveyed participants – 85 per cent – who had posted at least one message about the protests over the course of the month up to the survey (Konda, 2014: 29). Of the social media used by the protestors, Twitter was looked upon with the greatest anxiety by the authorities. Recep Erdoğan, the incumbent prime minister, exclaimed that ‘there is now a menace which is called Twitter’ (Konda, 2014: 28).

We tracked the communication on Twitter from 31 May 2013 to December 2014. In the first month of that period we recorded 1,106,383 unique tweets about Gezi posted by 346,598 users with the hashtags #direngeziparki, #occupygezi, #occupyturkey and #occupytaksim. While we recognise the constrictions inherent to the study of hashtagged communication (Tufekci, 2014) – particularly in regard to its representativeness and completeness – our analysis focuses on a distinct subsample of steadfast posters whose communication is discussed against the question of its significance for activist persistence.

Theoretical Framework

The issue of activist persistence seems topical at a time when there is much pessimism about political participation. Against a continuously eroding interest in institutional politics and the avenues for participation in them – for example, voting, party and associational membership (Dalton, 2008; Van Deth, 2014) – alternative forms of political engagement have caught the attention of scholars. Although proportionately not the most important practice to counterbalance the ostensible decline (Dalton, 2008), protest participation has seen recent prominent surges in liberal democracies and authoritarian regimes alike (Biekart and Fowler, 2013).

Historically, a consistent finding has been that protest participation is galvanised chiefly by social interaction through membership in activist organisations; or, alternatively, by close contact with members of an activist organisation (McAdam, 1986). Structural availability is the term designating one’s embedment in the social environment where mobilisation into activism transpires and is sustained (Schussman and Soule, 2005). While the aforementioned pathways to participation may remain in place, there are indications of at least one parallel mobilisation process combining personal contact with social and alternative activist media usage (Anduiza et al., 2014; Enjolras et al., 2012), ultimately resulting in ‘the mobilisation of people with lower levels of previous political involvement’ (Anduiza et al., 2014: 760).

Mobilisation, however, is a two-step process (Beyerlein and Hipp, 2006; Klandermans and Oegema, 1987). First, one develops a commitment to the goals and tactics of an activist group or movement, principally via membership in germane social networks (Crossley and Ibrahim, 2012; Klandermans and Oegema, 1987). That commitment can subsequently lead to participation if one is biographically available to convert beliefs into meaningful action (Klandermans and Oegema, 1987; McAdam, 1986: 67). Biographical availability refers to the ‘absence of personal constraints that may increase the costs and risks of protest participation, such as full-time employment, marriage and family responsibility’ (Corrigal-Brown, 2011; McAdam, 1986: 70), which are less prevalent among the young and the retired (Beyerlein and Hipp, 2006: 301). In their study, however, Beyerlein and Hipp (2006: 314) claimed that biographical unavailability precludes the formation of one’s commitment to participation rather than participation itself. In this light, the ability to sustain one’s commitment to activism is of vital importance to protest participation.

Scholarship on activist persistence has queried both the individual circumstances that make one available to activism and the social context which nurtures one’s commitment. As an outcome of mobilisation, persistence is the fruit of abiding biographical and structural availability (Downton and Wehr, 1998: 540). Bonds to organisational values and goals, as well as to leaders and the wider community of fellow activists, contribute to one’s readiness to remain biographically available and one’s vision for a positive contribution to the common cause.

Activist persistence entails a considerable expenditure of time and effort. In his landmark analysis of protest participation, McAdam (1986) distinguished participants on two discrete dimensions of the cost and risk attached to involvement in collective action. Costs pertain to the personal disbursement of time, money and energy towards participation (Erickson Nepstad and Smith, 1999). Risk, on the other hand, is the danger that involvement in protest activity exacts an important financial, legal or social toll as well as potentially bringing harm to one’s physical integrity (Erickson Nepstad and Smith, 1999: 25). Activist persistence may incur at least personal costs as one strives to maintain a level of biographical availability and commitment that puts one at odds with any competing demands from kith and kin as well as with an employer or educator (Downton and Wehr, 1998). Those costs may be particularly substantial in high-risk activism (McAdam, 1986: 67).

Participation in high-cost/risk activism further pivots on one’s structural availability and in particular one’s organisational ties (McAdam, 1986). Without ties to an organisation – both direct through membership and indirect through proxies such as friends – potential participants would be more likely to falter in their determination to participate in a protest. In their assessment of the same topic, Erickson Nepstad and Smith (1999: 35) additionally argued that the longer the duration of a protest, the higher the likelihood of indirect or relational ties – rather than organisational ties – becoming more prominent a channel for stimulating participation.

Likewise notable is the observation that various activist organisations of late exhibit a readiness to accept that participants in their actions will self-organise with networking technologies such as social media (Bimber et al., 2012; Karpf, 2010). Indeed, a sizeable contingent of protest returnees has been uncovered who have a track record of protest participation and only loose connections to activist organisations (Saunders et al., 2012). Returnees may revert to social media to compensate for the relative disadvantage of an absence of membership or indirect association with activist organisations (Mercea, 2014). They may thus prime their participation by retrieving pertinent information on social media while at the same time conversing with friends about a joint attendance and displaying a social identity predicated on a sense of fellowship with other participants mobilised in support of the same protest (Mercea, 2014: 403–404). Social identity is a mutual conception of membership within a group that ‘typically includes stereotypes of in- and out-groups’ (Van Zomeren et al., 2008: 505). In the Arab Spring, a social identity that crystallised around a shared opinion of opposition to an out-group, namely the authoritarian government, assembled variegated collections of political protest participants (McGarthy et al., 2014). A shared opinion is the perception that ‘people are the same because they agree with each other’ despite any differences in ascribed collective characteristics such as race, class or ethnicity (McGarthy et al., 2014: 729).

There is, nonetheless, deep-seated concern about a purported feigning of activism on social media. The claim is that key group processes conducive to ‘solidarity, commitment and responsibility towards fellow activists’ are compromised by an atomistic desire for visibility and self-representation on social media (Milan, 2015: 896). Yet, other authors have proposed that involvement in online and loosely organised actions more neutrally termed ‘information activism’ may help sustain communication even beyond the conclusion of a protest (Soon and Cho, 2014: 550). A significant investment of time into activist communication on Twitter can, however, put a strain on relationships with family and friends, diminish performance at work and threaten job stability (Mercea and Bastos, 2016).

Ultimately, Downton and Wehr (1998) used the metaphor of a career to describe activist persistence. Persisters have to juggle multiple responsibilities that fall within as well as beyond the remit of their activism. They have to manage relationships with activist peers as well as with family and friends, striking a balance between the support and chagrin they get from either side, all the while avoiding burnout and cultivating a social identity articulated as opposition to an out-group. Most notably, what distinguishes persisters from other types of participants is a unique capacity to ensure activism in support of their cause endures. As Downton and Wehr (1998: 542) put it: ‘persisters know they are persisters, keeping at it while others come and go’. Most importantly, however, persistence has been expected to transpire in organisational settings. In what follows, we consider whether those users that tweeted continuously about the Gezi Park demonstrations for their entire two-week duration and then beyond that climactic period bore any of the traits that distinguish persistent activists. To that end, we explored the following hypotheses:

Hypothesis 1 (H1): We expected to find a significant level of biographical availability of age, domestic and professional responsibilities and structural availability of organisational ties among individuals who displayed a sustained commitment to tweeting about the Gezi protest.

Hypothesis 2 (H2): Turning to the micro-mobilisation context, we envisioned that more than organisational membership, relational ties to other protest participants and a social identity predicated on a sense of democratic opposition to the government would underpin tweeting about the protest during but also in the wake of the Gezi occupation.

Hypothesis 3 (H3): Thirdly, we posited that to communicate on Twitter, persistent posters circumvented not only apparent legal and physical risks due to the clampdown by the state, but likewise had to overcome social and personal costs.

Hypothesis 4 (H4): Finally, we anticipated that the Twitter posters would be aware of their persistence and the contribution it could make to the endurance of their activist cause.

Data and Methods

Adopting a sequential mixed-methods design (Ivankova et al., 2006), we conducted a binary logistic regression analysis on Konda’s Gezi survey data. The method enabled predictions to be made regarding the membership of survey respondents in the categories of the dependant variable with a set of independent variables (Menard, 2002). This was done to ascertain the bearing that biographical and structural availability,⁵ together with one’s participatory experience and a social identity defined by opposition to the government, had on the odds that participants in the Gezi occupation would return to the protest site one week after its inception. Resultant insights formed a general baseline (Ivankova et al., 2006: 5) for our analysis of H1 and H2 while focusing specifically on the Twitter communicators.

We next turned our attention to the Twitter data. From 2013 to 2014 we monitored a set of four Gezi hashtags (see Figure 1b) by tapping directly into the Twitter Streaming API. The resulting datasets were marked by a skewed, long-tail distribution of contributions that is characteristic of political communication on Twitter (Mustafaraj et al., 2011). The largest number of unique users – almost two-thirds (225,019) – only tweeted once and on a single day; 57,967 (17%) tweeted on two separate days; 63,612 (18%) tweeted on more than two and less than five days; 16,909 (5%) tweeted on between five and 10 days; 3183 (1%) between 10 and 20 days. As to the time distribution of the tweets, a marked peak was recorded on 15 June at the time of the violent evacuation of the demonstrators from the Park (230,009 tweets, see Figure 1a). Further, Figure 1b reveals that the Gezi hashtags remained in use after the June flashpoint, albeit to a more modest degree.

Figure 1a.

Hashtagged Gezi tweets in the period 31 May to 30 June 2013.

Figure 1b.

Hashtagged Gezi tweets from 2013 to 2014.

The longitudinal nature of our Twitter data allowed for the identification and operationalisation of persistent activism across this temporal series. Thereby, we selected users who kept tweeting after the police clampdown and continued to do so for at least 20 days during June 2013. We viewed these posters through the prism of a case study design, with each potential interviewee adding to replication (Yin, 1994: 49–50) in a case series. We aimed to attain both literal and theoretical replication in case selection. We juxtaposed the two procedures to generate a heuristic for participant selection whereby we attained a sufficient number of interviews with persistent communicators (literal replication) for us to model the variability in the participants’ conditions of action (theoretical replication; Yin, 1994: 49–50). Following this selection protocol, between April and September 2015 we interviewed 24 out of 100 people contacted from our target population. Of the potential interviewees, 20 declined to be interviewed; nine accounts had been deleted and three had been suspended by Twitter by 2015. The remaining 44 prospective interviewees did not reply to our invitations. Levels of interviewee activity on Twitter in June 2013 are plotted in Figure 2.

Figure 2.

Tweets by interviewees in the period 31 May to 30 June 2013. A colour version of this figure is available online.

The conditions of action controlled for theoretical replication were the structural availability and micro-mobilisation context as well as the biographical availability of the interview participants. Through this approach to recruitment we were able to find and interpret contrasting conditions among the participants. Lastly, we addressed H3 and H4 by producing a rich account of how persistent communicators contended with their conditions of action, enacted and interpreted their sustained commitment to the Gezi protests and its ramifications. We tapped into these aspects with semi-structured respondent interviews. Participants in respondent interviews are recruited on the expectation that they will speak about themselves, their own situation, and the motivations and interpretations of their circumstances (Lindlof and Taylor, 2002: 178).

We organised the qualitative data with clustered summary tables (Miles and Huberman, 1994) so as to generate an integrated and sensitised description of the aspects of interest across all interviews. We numbered all interviews and used an alpha-numeric indicator to designate individual interview participants (e.g. IP01 for the first interviewee). We identified, coded and grouped together segments from all interviews germane to our research questions. This was an iterative process that we understood as grounded thematic analysis (Cronin et al., 2008: 574). We coded the conditions – both structural and personal – of the participants, their interpretations thereof, as well as the relationship between tweeting practices, involvement in the Gezi Park protests and membership of activist organisations (or absence thereof).

As to the remit of our research and the robustness of the research design, our enquiry developed from an independent measurement of sustained activist communication on Twitter which was further contextualised and explained with survey and interview data. Explanations of the interactions between the conditions of interest and participant actions were validated through a pattern-matching protocol (Yin, 1994). We adjudicated between biographical and structural availability to qualify an emergent model of persistent activist communication, relating it back to the theory and the statistical data on persistent participation in the Gezi occupation. In addition, the replication procedures helped minimise potential bias in interviewee selection. All along, rather than proposing empirical generalisations, our intention was to reconstruct the theory (Lichterman, 2002: 124) of activist persistence to account for sustained activist communication on Twitter. The next section provides a report of our principal findings.

Research Findings

Of the participants surveyed in Gezi Park by Konda, a very sizeable contingent (87% of the protesters) said that they had been returning to the site of the protest following demonstrations on 28–29 May 2013. Table 1 presents the results of the logistic regression we ran on the survey data. An inspection of the adjusted odds ratios in the figure evidences that returnees were more likely to be found among participants aged 36 and above and the organisationally affiliated. Further, the odds of being a returnee were lower had one learnt of the occupation either from friends and acquaintances or, to a lesser extent, on social media. Finally, and surprisingly, one’s educational achievement, employment status or the espousal of a social identity appeared immaterial to an enduring commitment to participation in the occupation.

Table 1.

Predictors of participant return to the site of the Gezi occupation (N = 4361 cases in the final model).

Variable	Logistic regression coefficient (b)	Standard error	Adjusted odds ratio [Exp(B)]
Age 17 or below 18–20 21–25 26–30 31–35 36 and above	–.129 – –.124 –.152 – .588***	.172 – .137 .174 – .144	.879 – .883 1.165 – 1.800***
Education Illiterate Literate without degree Primary school degree Junior school degree High school degree College degree or higher	.049 .354 .222 .701 .284 .262	.888 .570 .442 .392 .367 .364	1.051 1.424 1.249 2.016 1.329 1.299
Employment In paid employment In unpaid employment, unemployment or retirement Student	–.647** –.196 –.391	.283 .301 .284	.524** .822 .676
Organisational ties	.597***	.120	1.817***
Relational ties Friends and acquaintances Social media	–1317* –565*	.184 .111	.268* .568*
Social identity Opposition to PM Erdoğan Opposition to dictatorship and oppression	–.067 .273	.100 .169	.935 1.314
Constant	–1.669***	.581	.188***

Notes: R² = .073 (Nagelkerke), χ² = 146.070, d.f. = 17, p < .001; significance: *.05, **.01, ***.001.

Our expectations were again confounded by the interviewees. Like the sustained commitment showed to physical participation in the occupation, persistent protest communication on Twitter was chiefly the province of individuals aged over 30. The median age of the interviewees was 38. At the time of the protests, they were in professional employment (five in six interviewees) or studying (three full-time students). They exhibited high educational achievement – seven in eight were university graduates; the remaining three held college and/or high-school degrees – and lived in proximity to the demonstrations.

Seventeen interviewees were tweeting out of Istanbul; four were tweeting from other urban centres in Turkey which had seen demonstrations in support of the Gezi occupation. Three interviewees were tweeting from abroad. Among them there was a Turkish émigré (IP12) and two non-native speakers posting and retweeting principally in English, one of whom was bed-ridden at the time of the protest due to medical problems (IP23). Except for the three individuals living outside the country, all interviewees made their way to the protests either in Istanbul or in the other Turkish cities where these unfolded. Notwithstanding his relocation, IP12 organised his own event in support of the Gezi occupation. The other research participant located abroad who was physically fit attended demonstrations regularly.

We interviewed an equal number of female and male interviewees, 11 of whom were in a relationship (six were married), while slightly more of them were or became single around the time of the protests (12). One interviewee declined to disclose his relationship status. Finally, the median number of Twitter followers for the interviewed users was 1865 (min = 355, max = 190,000, x– = 11,898, σ = 38,291). With a single exception, therefore, these posters appeared significantly less prominent in their global Twitter outreach than influential users whose networks surpass 100k followers (Bastos and Mercea, 2015).

Confounding H1, biographical availability was less of a definitive feature than we expected it to be for persistence either at the site of the occupation or in one’s rolling communication about the protests on Twitter. Aside from one young female student and two mature postgraduate students, none of the interviewees had personal situations that were unambiguously favourable to involvement in activism. Thirteen faced especially testing circumstances at work or at home which were aggravated by their involvement with the occupation. For instance, a female participant was a single mother who had to juggle childcare with freelancing and the precariousness of short-term contracts. A single female retiree, IP23, was convalescing following an accident; yet another was the mother of a child with learning difficulties whose rearing had been made particularly hard by an unrelenting experience of discrimination. Another self-employed male interviewee lost an important contract after the protests erupted. All were, however, opinionated individuals ready to express their political convictions on social media (see Quintelier and Theocharis, 2013) and likewise to realise them in the street protests, their health permitting.

The structural availability of the persistent communicators was probed on two levels, namely of membership in formal organisations – chartered associations and political parties – and informal activist groups. Two interviewees were members of the Democratic People’s Party (DPP), a left-leaning opposition party. The DPP was represented in the coordination body of the Gezi occupation, the Taksim Solidarity Platform. Three other interviewees were members of the Alternative Informatics Association (AIA), an Istanbul-based NGO advocating media literacy, freedom of expression and controls over mass surveillance. AIA was also represented in the Gezi occupation. Of these affiliated participants, only one of the DPP members recalled learning of the protests from the internal communications of the organisation. Painting a fuller picture of the moment, she underscored how organisational and relational ties were intertwined with Twitter communication:

I was on the Taksim [Solidarity] Platform, so I was aware of the activities and the events going on. But I saw the tents burnt on Twitter accounts of the friends in the network of Taksim.

The above examples were exceptions among the interviewees. There were 19 out of the 24 interviewees who were not affiliated to either formal or informal organisations or groups. If socialisation incentivising participation seemed very marginally attributable to direct organisational ties, relational ties had catalysed the interest in the protests of seven of the interviewees. Double that number (14), however, first learnt of the occupation on Twitter; of the latter, half got news of it from friends or family’s posts. For two other participants who lived in a neighbouring area, news and the desire to make one’s way to Gezi were catalysed by proximity to the protest site.

Thus, contrary to on-site persisters, the mobilisation of persistent activist communicators was more likely fostered by relational rather than organisational ties. While this finding confirmed H2, tweeting was deeply woven into both the activation and the rolling commitment of the interviewees to the Gezi protests during and more importantly following the dismantlement of the camp site in Istanbul. The commitment shown by the unaffiliated Twitter communicators was the product of a social identity predicated on popular opposition to authoritarian government personified by the prime minister Recep Tayyip Erdoğan and a countervailing promise of democratic rejuvenation unlocked by the Gezi occupation. Those who did not explicitly speak of an aggravating erosion of democracy in Turkey nevertheless invoked a perceived extraordinary assault by the government on cardinal democratic principles such as institutional checks and balances, freedom of expression or assembly as motivating their participation in Occupy Gezi and their persistent Twitter communication.

All but two interviewees viewed information dissemination as their chief contribution to the collective outcry and a sense of outrage which they expected would have been less likely to arise in the absence of the communication on Twitter. The assertion was grounded in personal assessments of the limited mainstream media coverage of the protests. As IP09 expounded: ‘authorities controlled the central media hubs … Twitter challenged the control of the media hubs’. However, IP16 spoke of an enmeshment of media sources that corroborated the outbreak of the protests to her:

I think it [participation] started on Twitter on which I am quite active. At the same time, I came across some timely TV news reports from the Park … but the main source of information was indeed Twitter and some of my friends on Twitter; and not friends but contacts on Twitter.

Next, we ascertained the risks evaded and costs borne by the interviewees so as to remain biographically available for activist communication on Twitter. A perception of legal and physical threats, namely the risk of being arrested for one’s dissident views, existed among all interviewees but acted as a spur rather than a deterrent to one’s activism (cf. McAdam, 1986). Reflecting on the hard line taken by authorities on posting about Gezi, IP15 commented: ‘People were afraid of tweeting before Gezi. Suddenly [when Gezi happened] a freedom environment emerged.’ The intensity of the pressure felt was nevertheless made vivid by the confession of two participants that they overall ‘slowed down’ their tweeting after June 2013 for fear of government repression.

Financial insecurity loomed over several interviewees. Seven among them struggled to keep their jobs, facing the prospect of financial hardship as they descended on the frontlines of the protests, on-site in Gezi Park and on Twitter. Conversely, those in secure employment were at pains to balance their time. IP10 recounted: ‘In the mornings, we were working. In the evening, after taking off our ties, we were going to the protests … [I’d get] out to take photos and write news from the field and tweeted them.’

The personal expenditure of time and energy dedicated to the occupation was significant among them all. The main challenge for the interviewees was to square competing work obligations with their engagement in the communication of the protests and participation in them. There were sleepless nights devoted to posting messages or keeping a close watch on updates from the Park; or vexed sensitivities of family or friends due to the outlay of time to support the Gezi occupation. IP21 recalled how friends started avoiding him because of his political views. IP19 spoke of both a direct cost of fraught filial and intimate relationships and an indirect social cost she paid for her politics when she unfriended contacts on social media, an act associated with negative emotional responses such as anger and sadness (see Bevan et al., 2012):

He was sleeping, I was on Twitter, typing. It bothered my boyfriend very much. My mother was sad because I was very angry. Gezi emotionally affected me so much that … my eczema became really bad. When it comes to personal relationships, I deleted a lot of friends on Facebook because [I thought] people might be embarrassed that I had friends who are close to [the governing] AKP [party]. There were many years of friendship [between us] … There was already a disagreement on miniskirt vs headscarves the summer before Gezi. That incident broke us apart somewhat. I said that I would do anything to defend [the] headscarf but [they] did nothing when people were picking on our miniskirts on the internet.

Interviewee accounts went some way towards substantiating H3. All posters paid some financial, personal or social cost for their activist communication. However, five interviewees spoke of favourable family relationships or workplace conditions that allowed them to combine profuse tweeting with participation in the protests. The main offset for the costs and risks attendant to posting on Twitter came from a combination of the oppositional social identity and a desire to provide an accurate public representation of Occupy Gezi. Sustaining the tweeting of the Turkish interviewees, this cognitive outlook was bolstered by physical participation in the occupation.⁶ Interviewees spoke of the need to bear witness to the struggle, to chronicle the discordant experience of subjection to state repression that could resonate with many other democratic citizens. Evocatively, relating the part she played in the occupation, IP05 characterised herself as ‘walking media’, adding:

We made people get heard. The collectivity lived in the protests … sharing your food, tea. So we communicated this spirit of solidarity to the people who could not be there. They still had the opportunity to follow [the protests on Twitter] and feel this spirit. The visibility of this collective spirit increased.

We further noted that the earliest any interviewee ceased to tweet about Gezi was in late July 2013. One interviewee quoted personal reasons for discontinuing his tweeting, while the foreign posters cited the ebbing coverage both by activists and mainstream media as having driven them to stop tweeting about Gezi. Five others decided to gradually decrease the amount of tweeting about Gezi, citing a marked reduction in public interest in the follow up to the street protests. The practice was nonetheless resumed to mark the one-year anniversary of the occupation (see Figure 1b). Two years on from the uprising, IP07 portrayed her continued tweeting as an act of memorialisation:

Now, for instance, we start to tweet as Remember Gezi as we are getting close to May 31st. I don’t know if we will go out on the streets because we saw what [violence] happened on May 1st but … we will commemorate Gezi on social platforms.

Indeed, 19 of the interviewees did not stop tweeting at the end of July 2013. They reiterated the notion that the practice was a vehicle for remembrance. An unbroken commitment to the articulation of values embodied in Gezi underlay the lasting proclivity to post on Twitter. A milestone for Turkish democracy, the occupation was evoked for epitomising a democratic check on authoritarianism. Illustratively, IP17 said:

I didn’t quit tweeting about Gezi because people can be collectively mobilised, there are other alternatives besides the parliament. We can produce ideas and news and stand up [together]. I mean, this is the moment at which an awareness was created and Gezi still keeps happening, I think.

Although committed to tweeting about Gezi, none of the interviewees spoke of their communication as having fed into an organisational project that would help perpetuate their activist cause (Bimber et al., 2012; Downton and Wehr, 1998). Gezi was an expression of prefigurative politics as a laboratory for political organisation and participatory decision-making.⁷ Research participants aided with the coordination of actions on location via Twitter. They endorsed the on-site participatory decision-making process in Gezi Park, likewise partaking in it with a single exception. Interviewees nevertheless did not place themselves directly at the heart of any organisational projects seeking to continue the legacy of the Gezi protests. Relevant examples were the Gezi Party that was founded in October 2013, or the United June Movement established in February 2015.

Counter to H4 and our supposition that the activist communicators would make a sustained contribution to the organisation of pro-democracy activism, their tweeting did not feed into any organisational projects brewed during or following Gezi. Indeed, close involvement with the occupation generated polarised effects among the interviewees. While in its aftermath, two of the research participants joined formal organisations – namely an opposition party and a trade union – another participant spoke of an overwhelming disillusionment with the political regime which sealed her desire to emigrate. The rest of the interview participants remained unaffiliated to any organisation.

In sum, Twitter was primarily a medium for the documentation and later the memorialisation of protest, which did not stand in contradiction with embodied participation. The Twitter communicators were crowd-sourced activists that helped build the public momentum for the Gezi occupation. In its wake, Twitter was a stage for acts of memorialisation of the 2013 uprisings by the same activists who thereby sought to keep the symbolic legacy of the events alive and relevant to contemporary Turkish democracy.

Discussion and Conclusions

We uncovered composite modes of participation in an episodic mass mobilisation that question the distinction between communication on Twitter and other modalities of sustained engagement in collective action. ‘Asphalt’ activism and activist communication flow into each other, enabling not only mobilisation (Castells, 2012; Mercea, 2014) but also sustained commitment to a protest, its public communication and its cause, following the cessation of the collective action.

The returnees to the physical occupation of Gezi were of a similar age to the persistent activist communicators. The persistence of the two cohorts was set apart by the prominence of a social identity predicated on opposition to authoritarianism that emerged as instrumental to the sustained commitment of the activist communicators. The structural availability of organisational ties was salient to the persistence of on-site participants. Relational ties contributed to the mobilisation of the Twitter communicators but had an inverse relationship to the return of activists on location. For the majority of the communicators who first learnt of the protests on the micro-blogging service, more often than not, relational ties were enmeshed with communication on Twitter. In those cases where friends first alerted interviewees to the protests, these reverted to Twitter for a decisive validation of the rousing information (e.g. that protestors’ tents had been set alight by authorities on 30 May 2013).

Biographical availability did not appear as an unmistakable prerequisite either to physical involvement or the persistence of activist communication. Twitter communication was, for the largest number of the posters, a modality for conveying the values of Occupy Gezi. In addition, it was a vehicle for remote activism for those who were physically unable to attend the protests. Persistent communication was driven by a sense of mission to defend democracy, to bear witness and memorialise the occupation so as to make the experience more vivid and pressing to friends, family and ultimately the general public who may not have been imbued with the same sentiment of moral outrage (see Jasper, 1997) and urgency. The communicators’ commitment exacted personal and social costs but also family sympathy and support for some. A heavy involvement thus carried social costs as well as rewards.

The persistence of activist communication was not twinned with attempts to make an enduring organisational structure emerge from the Gezi occupation. This finding calls into question the organisational implications of persistent activist communication. Ad hoc collective bodies, such as the Taksim Solidarity Platform, together with intense protest communication on social media, may help momentarily arrest a march of hegemonic power on democracy. While the Gezi protests were successful in that they reversed the decision to raze the Park, it remains an open question what their memorialisation on Twitter and perhaps elsewhere might achieve for the wider goal of organising to safeguard a robust democracy in Turkey. In the aftermath of the occupation, the persistent activist communicators did not embrace any of the spin-off organisational projects. Subsequent to the occupation, the same communicators were at best agents of memorialisation.

Twitter communication was a resistance tactic which, in the vein of De Certeau (1984), served to maximise the subordinates’ conditions of action in ways that may wrong-foot hegemonic power. To disrupt discretionary actions by the government, expose police violence and circumvent an apparent media embargo while sensitising others to the urgency to defend democracy, the activist communicators expanded important resources of time, contending with multiple risks while continuing to tweet, an activity closely entwined with their embodied activism.

Lastly, this research is not an exhaustive study of the conditions that enable persistent activist communication, or of the personal investment made to sustain one’s commitment. To continue this investigation, persistent communication may be mapped onto multiplying forms of activist participation that differentially reinforce each other (see Neumayer and Svensson, 2016). We would invite comparative research that marries an in-depth examination of activist cultures galvanised by protest events on social media with time-series analyses of their endurance over time in different socio-political contexts.

Footnotes

Acknowledgements

The authors are grateful to Eylem Yanardağoğlu for her help with data collection and to KONDA for access to the survey data. We thank colleagues at the symposium ‘Times are a Changin’: Temporality, Memory and Social Movements in the Digital Age’, and the anonymous reviewers for their valuable comments.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Notes

Dan Mercea is Senior Lecturer in the Department of Sociology at City, University of London. He holds a PhD in Communication Studies from the Department of Sociology, University of York. He is the author of Civic Participation in Contentious Politics (Palgrave, 2016).

Duygu Karatas is Doctoral Candidate in the Department of Journalism and Mass Communication, University of Westminster, London, UK.

Marco T Bastos is Lecturer in Media and Communication at City, University of London and an affiliate of Duke University’s Network Analysis Center. This research was conducted while he was a postdoc at the University of California at Davis.

References

Agarwal

Barthel

Rost

Borning

Bennett

Johnson

(2014) Grassroots organizing in the digital age: Considering values and technology in Tea Party and Occupy Wall Street. Information, Communication and Society 17(3): 326–341.

Anduiza

Cristancho

Sabucedo

(2014) Mobilization through online social networks: The political protest of the Indignados in Spain. Information, Communication and Society 17(6): 750–764.

Bastos

Mercea

(2015) Serial activists: Political Twitter beyond influentials and the Twittertariat. New Media and Society 18(10): 2359–2378.

Bennett

Segerberg

Walker

(2014) Organization in the crowd: Peer production in large-scale networked protests. Information, Communication and Society 17(2): 232–260.

Bevan

Pfyl

Barclay

(2012) Negative emotional and cognitive responses to being unfriended on Facebook: An exploratory study. Computers in Human Behavior 28(4): 1458–1464.

Beyerlein

Hipp

(2006) A two-stage model for a two-stage process: How biographical availability matters for social movement mobilization. Mobilization 11(3): 299–320.

Biekart

Fowler

(2013) Transforming activism 2010+: Exploring ways and waves. Development and Change 44(3): 527–546.

Bimber

Flanagin

Stohl

(2012) Collective Action in Organization: Interaction and Engagement in an Era of Technological Change. Cambridge: Cambridge University Press.

Budak

Watts

(2015) Dissecting the spirit of Gezi: Influence vs. selection in the Occupy Gezi Movement. Sociological Science 2: 370–397.

10.

Bunnage

(2014) Social movement engagement over the long haul: Understanding activist retention. Sociology Compass 8(4): 433–445.

11.

Castells

(2012) Networks of Outrage and Hope: Social Movements in the Internet Age. Cambridge: Polity.

12.

Císař

(2013) A typology of extra-parliamentary political activism in post-communist settings: The case of the Czech Republic. In: Jacobsson

Saxonber

(eds) Beyond NGO-ization: The Development of Social Movements in Central and Eastern Europe. Farnham: Ashgate, 139–169.

13.

Corrigal-Brown

(2011) Patterns of Protest: Trajectories of Participation in Social Movements. Stanford, CA: Stanford University Press.

14.

Cronin

Alexander

Fielding

Moran-Elli

Thomas

(2008) The analytic integration of qualitative data sources. In: Alasuutari

Bickman

Brannen

(eds) The SAGE Handbook of Social Research Methods. London: SAGE, 572–584.

15.

Crossley

Ibrahim

(2012) Critical mass, social networks and collective action: Exploring student political worlds. Sociology 46(4): 596–612.

16.

Dalton

(2008) Citizenship norms and the expansion of political participation. Political Studies 56(1): 76–98.

17.

De Certeau

(1984) The Practice of Everyday Life. Berkley, CA: University of California Press.

18.

Demirhan

(2014) Social media effects on the Gezi Park movement in Turkey: Politics under hashtags. In: Pătruţ

Pătruţ

(eds) Social Media in Politics: Case Studies on the Political Power of Social Media. Heidelberg, New York, Dordrecht: Springer, 281–314.

19.

Downton

Wehr

(1998) Persistent pacifism: How activist commitment is developed and sustained. Journal of Peace Research 35(5): 531–550.

20.

Enjolras

Steen-Johnsen

Wollebaek

(2012) Social media and mobilization to offline demonstrations: Transcending participatory divides? New Media and Society 15(6): 890–908.

21.

Erickson

Nepstad S

Smith

(1999) Rethinking recruitment to high-risk/cost activism: The case of Nicaragua Exchange. Mobilization 4(1): 25–40.

22.

Flesher Fominaya

(2015) Debunking spontaneity: Spain’s 15M/ Indignados as autonomous movement. Social Movement Studies 14(2): 142–163.

23.

Göle

(2013) Gezi-Anatomy of a Public Square Movement. Insight Turkey 15(3): 7–14.

24.

Goss

(2003) Rethinking the political participation paradigm. Women and Politics 25(4): 83–118.

25.

Halupka

(2014) Clicktivism: A systematic heuristic. Policy and the Internet 6(2): 115–132.

26.

Ivankova

Creswell

Stick

(2006) Using mixed-methods sequential explanatory design: From theory to practice. Field Methods 18(1): 3–20.

27.

Jasper

(1997) The Art of Moral Protest: Culture, Biography and Creativity in Social Movements. Chicago and London: The University of Chicago Press.

28.

Karpf

(2010) Online political mobilization from the advocacy group’s perspective: Looking beyond clicktivism. Policy & The Internet 2(4): 7–41.

29.

Klandermans

Oegema

(1987) Potentials, networks, motivations, and barriers: Steps towards participation in social movements. American Sociological Review 52(4): 519–531.

30.

Konda (2014) Gezi Report: Public perception of the ‘Gezi protests’. Who were the people at Gezi Park? Available at: http://konda.com.tr/en/raporlar/KONDA_Gezi_Report.pdf (accessed 15 June 2015).

31.

Lichterman

(2002) Seeing structure happen: Theory-driven participant observation. In: Klandermans

Staggenborg

(eds) Methods of Social Movement Research. Minneapolis, MN: University of Minnesota Press, 118–145.

32.

Lindlof

Taylor

(2002) Qualitative Communication Research Methods. 2nd ed. Thousand Oaks, London, New Delhi: SAGE.

33.

McAdam

(1986) Recruitment to high-risk activism: The case of Freedom Summer. American Journal of Sociology 92(1): 64–90.

34.

McGarthy

Thomas

Lala

Smith

LGE

Bliuc

(2014) New technologies, new identities, and the growth of mass opposition in the Arab Spring. Political Psychology 35(6): 725–740.

35.

Maeckelbergh

(2011) Doing is believing: Prefiguration as strategic practice in the Alterglobalization movement. Social Movement Studies 10(1): 1–20.

36.

Margetts

John

Hale

Yasseri

(2016) Political Turbulence: How Social Media Shape Collective Action. Princeton, NJ: Princeton University Press.

37.

Menard

(2002) Applied Logistic Regression Analysis. 2nd ed. Thousand Oaks, London, New Delhi: SAGE.

38.

Mercea

(2012) Digital prefigurative participation: The entwinement of online communication and offline participation in protest events. New Media and Society 14(1): 153–169.

39.

Mercea

(2014) Towards a conceptualization of casual protest participation: Parsing a case from the Save Roşia Montană campaign. East European Politics and Societies 28(2): 386–410.

40.

Mercea

Bastos

(2016) Being a serial transnational activism. Journal of Computer-Mediated Communication 21(2): 140–155.

41.

Mercea

Funk

(2016) The social media overture of the pan-European Stop-ACTA protest: An empirical examination of participatory coordination in connective action. Convergence 22(3): 287–312.

42.

Milan

(2015) From social movements to cloud protesting: The evolution of collective identity. Information, Communication and Society 18(8): 887–900.

43.

Miles

Huberman

(1994) Qualitative Data Analysis. 2nd ed. Thousand Oaks, CA: SAGE.

44.

Morozov

(2011) The Net Delusion. London, New York: Allen Lane.

45.

Mustafaraj

Finn

Whitlock

Metaxas

(2011) Vocal minority versus silent majority: Discovering the opinions of the long tail. In: IEEE Third International Conference on Privacy, Security, Risk and Trust, Boston, MA, USA, 9–11 October 2011. DOI: 10.1109/PASSAT/SocialCom.2011.188.

46.

Neumayer

Svensson

(2016) Activism and radical politics in the digital age: Towards a typology. Convergence 22(2): 131–146.

47.

Quintelier

Theocharis

(2013) Online political engagement, Facebook and personality traits. Social Science Computer Review 31(3): 280–290.

48.

Rohlinger

Bunnage

(2015) Connecting people to politics over time? Internet communication technology and retention in Moveon.org and the Florida Tea Party Movement. Information, Communication and Society 18(5): 539–552.

49.

Saunders

Grasso

Olcese

Rainsford

Rootes

(2012) Explaining differential protest participation: Novices, returners, repeaters and stalwarts. Mobilization 17(3): 263–280.

50.

Schussman

Soule

(2005) Process and protest: Accounting for individual protest participation. Social Forces 84(2): 1083–1108.

51.

Soon

Cho

(2014) OMG! Offline-based movement organizations and network mobilization: A case study of political bloggers in Singapore. Information, Communication and Society 17(5): 537–559.

52.

Treré

(2015) Reclaiming, proclaiming and maintaining collective identity in the #YoSoy132 movement in Mexico: An examination of digital frontstage and backstage activism through social media and instant messaging platforms. Information, Communication and Society 18(8): 901–915.

53.

Tufekci

(2014) Big questions for social media big data: Representativeness, validity and other methodological pitfalls. In: ICWSM ‘14: Proceedings of the 8th International AAAI Conference on Weblogs and Social Media, Ann Arbor, MI, USA, 1–4 June 2014, Palo Alto, CA: The AAAI Press, 505–514.

54.

Tufekci

Wilson

(2012) Social media and the decision to participate in political protest: Observations from Tahir Square. Journal of Communication 62(2): 363–379.

55.

Valenzula

(2013) Unpacking the use of social media for protest behavior. American Behavioral Scientist 57(7): 920–942.

56.

Van Deth

(2014) A conceptual map of political participation. Acta Politica 49(3): 349–367.

57.

Van Zomeren

Postmes

Spears

(2008) Toward an integrative social identity model of collective action: A quantitative research synthesis of three socio-psychological perspectives. Psychological Bulletin 134(4): 504–535.

58.

Yin

(1994) Case Study Research Design and Methods. 2nd ed. London: SAGE.