The symbol of social media in contemporary protest: Twitter and the Gezi Park movement

This article explores how Twitter has emerged as a signifier of contemporary protest. Using the concept of ‘social media imaginaries’, a derivative of the broader field of ‘media imaginaries’, our analysis seeks to offer new insights into activists’ relation to and conceptualisation of social media and how it shapes their digital media practices. Extending the concept of media imaginaries to include analysis of protestors’ use of aesthetics, it aims to unpick how a particular ‘social media imaginary’ is constructed and informs their collective identity. Using the Gezi Park protest of 2013 as a case study, it illustrates how social media became a symbolic part of the protest movement by providing the visualised possibility of imagining the movement. In previous research, the main emphasis has been given to the functionality of social media as a means of information sharing and a tool for protest organisation. This article seeks to redress this by directing our attention to the role of visual communication in online protest expressions and thus also illustrates the role of visual analysis in social movement studies.


Introduction
Protest movements that straddle virtual and physical space are on the rise across the world, with activists increasingly able to engage with media technologies to become more visible and to ensure their voice is heard (Mason, 2014). The importance of social media for contemporary protest movements has been theorised from different perspectives. Overall, emphasis has been given to the functionality of social networking sites as a means of information sharing and a tool for organising protests (Bimber et al., 2012). Less attention has been given to the contemporary potential of visual culture as it intersects with social media to effect social change (Neumayer and Rossi, 2018) and how protestors use visual expression to communicate their ideas, identities and emotions across diverse social spaces, both material and virtual ones. Digital artefacts linked to protests and other forms of online activism offer rich troves of visual imagery such as photos, videos, memes, collages and posters, but they are seldom analysed by researchers, who still tend to privilege text over images (Bennett and Segerberg, 2013: 73;Doerr et al., 2013;Highfield and Leaver, 2016). Clark (2016: 235) theorises these as 'artifacts of engagement', emphasising that they are central to how individuals personalise 'expressions of a movement's goals' and signal political involvement on social media. This article seeks to redress the lack of attention given to images in social movement and activist media studies beyond the study of citizen photojournalism by directing our attention to the role of visual communication in digital activism, and as a specific case study it investigates how 'the symbol of social media' is inscribed into the visual discourse of protest. The article puts two lines of enquiry into dialogue: how do protest movements imagine their relationship to the corporate social media technologies they use? And how is this expressed visually? We suggest that how activists imagine the potential of media technologies determines how they use them and that therefore the relationship of protestors to the social media they use can be theorised productively through a social media imaginaries lens. We explore how visual communication informs and sustains such imaginaries. In this article, we thus seek to examine how social media is symbolically implied in protestors' self-image, with the aim of forming an understanding of how Twitter is imagined as a technology of protest both by activists and by other actors.
Using the concept of media imaginaries, a term derived from earlier sociological conceptualisations of 'the social imaginary of information and communication technologies' (Lesage and Rinfret, 2015), to investigate how social media (in this case Twitter) is 'pictured' and visually represented by protestors, their sympathisers and others, we argue, can offer new insights into social movements' engagement with media technologies and how they 'imagine together alternative media appropriations as an ongoing enactment of their social and political engagement' (Treré, 2018: 108). The imaginary, as Bucher (2018: 113) suggests, 'is to be understood in a generative and productive sense as something that enables the identification and engagement with one's lived presence and socio-material surroundings'. To this end, we use the term social media imaginaries, which, drawing on Treré, we take to denote both how protestors imagine themselves as part of a digitally connected mobilisation and how social media platforms take on meanings in such imaginings. As Treré et al. (2017) have shown, collective attitudes to, and employment of, social media vary significantly among contemporary protest movements. It is therefore important to gain more precise and situated knowledge not just about contemporary social protest movements' differing practical use of digital technologies but also about their conceptualisation of, and political and emotional investment in, these technologies. To explore this, the article raises questions about the relationship between visual activism and social media imaginaries.
This research offers a new methodological approach by combining close readings of images utilising visual analysis with thematic analysis of a larger number of images tweeted during the Gezi Park protest in 2013, which is further enriched contextually by drawing on activist interviews and wider media content. It thus fills a gap in the literature where visual political communication intersects with the field of media imaginaries. Methodologically, by bringing together the theories and practices of visual communication studies and social movement studies, the article addresses a gap in social science approaches to visual research: the consideration of aesthetics. We address this by analysing both the content and the form of images in relation to media technologies and situated representational cultures. Researching the cultural and symbolic value accorded to social media requires thinking about what social media signifies on a metaphorical level in the context of contemporary protest and social activism. It is clear that activists have been empowered by new functionalities of digital technology, but what is their symbolic relation to the digital media they use, or, in other words, what does social media stand for in the production of protest imaginaries? We aim to move beyond debates about optimism or reductionism in digital activism by offering a detailed case study of the complexity of the relationship between a protest movement and its media technology of choice as expressed through visual representations. We use a visual semiotics analysis of images collected via a scraping of historical Twitter data using the keywords gezi parki. The data set is contextualised by analysing associated images found via other purposeful online manual searches as well as interviews with protestors. This article does not seek to investigate how activists use digital and social media to express alternative political imaginaries but rather focuses on the way in which they imagine social media to be part of their collective identity and what these media imaginaries can tell us about contemporary social movements, thus addressing a gap in the literature, which has neglected the 'symbolic nature of communicative processes' (Gerbaudo and Treré, 2015: 868). To do so, we bring together the theoretical frameworks of digital activism, visual activism, media imaginaries and visual communication.
Empirically, the project focuses on the Gezi Park protests that took place in Turkey in 2013. The protests initially started on 27 May 2013 as a reaction to plans to reconstruct military barracks and build a new shopping mall in Gezi Park, which is in the centre of Istanbul. The initial aim was to stop the bulldozers and other demolition machinery entering the park, but the protests were also a reaction to neoliberal governance and authoritarian social-political engineering implemented by the ruling Justice and Development Party (AKP). Following heavy-handed police intervention, the protests developed into a broad cross-sectional mobilisation of citizens challenging state violence, media censorship (Coskuntuncel, 2018) and the government's 'neoliberal project [ . . . ] and conservative majoritarianism' (Bilgiç, 2016). Eleven people were killed during the protests and many thousands were injured. The Turkish Interior Ministry has estimated that at least 2.5 million people participated in the protests (House Committee on Foreign Affairs, 2013).
Commentators on the Gezi Park protest movement have predominantly focused on themes such as motives for participation (Erhart, 2014;Varnali and Gorgulu, 2015), communication and organisation (Mercea et al., 2017), and media representations of the protest (Oz, 2016). The Gezi Park protest constitutes a particularly rich case study for enquiries interested in creative and aesthetic aspects of protest. As noted by Aytekin (2017: 191), 'the predominant form of protest in the [Gezi Park] movement was aesthetic political acts', and he goes on to argue that 'artistic practices and cultural symbols employed by protestors' served to bring diverse groups of people together politically. Reflecting this, an emerging body of work concerns itself with various aspects of creative outputs linked to the protest, including street art and graffiti (Seloni and Sarfati, 2017;Taş, 2017), music Bianchi, 2018;Parkinson, 2018;Way, 2016) and photography (McGarry et al., 2019), and it is in this strand of Gezi Park research that we situate our work. Approaching social media imaginaries through visual methods, focusing on visual representations of social media that circulate online, offers an approach to understanding activists' engagement with media technologies that moves away from operative concerns (i.e. how do movements use social media and to what end?) to focus on the meaning-making processes and the aesthetics that underpin them, a perspective that is largely missing from the social movement literature.

Social media and protest media imaginaries
We argue that the concept of 'social media imaginaries', a derivative of the broader field of 'media imaginaries', may help inform an understanding of activists' relation to and conceptualisation of social media and how it shapes their digital media practices. We use the description social media imaginaries to mean imaginaries that are primarily understood as media imaginaries about social media, but we also note that such imaginaries are expressed and circulated (and thus produced) on social media, as exemplified in our data. In this section, we will first explore theories of media imaginaries to develop the theoretical frame for our approach and will then discuss the smaller body of literature that specifically addresses digital protest media imaginaries. Our theorisation of the symbol of social media in contemporary protest is situated at the intersection of these two fields. Treré (2018: 110) argues that 'in Gramscian terms, there is a continuing struggle between the hegemonic imaginaries regarding communication technologies fuelled by corporations and mainstream media, and the counter-hegemonic imaginaries pushed forward by social movements, civil society, and citizen media outlets', but as our analysis will show, such imaginaries may not always be so easily distinguishable in contrasting hegemonic and counter-hegemonic terms. Mansell (2012) also foregrounds in her work on how we experience the Internet the importance of being attentive to the complexities of contemporary media imaginaries and the competing agendas embedded in different visions of the Internet. She highlights that social imaginaries about the Internet often differ regarding ideas about where authority is located. This model is useful for our case in terms of unpicking how different actors engender the Twitter sphere as a powerful actor, either to support a movement or as a threat to state power. Using Gershon's (2010) concept of media ideology, Menke and Schwarzenegger (2019: 657) argue that new media users continuously 'renegotiate the meaning of media for themselves or collectively with others'. Factors such as 'political debates, academic expertise, popular culture, advertising' (2019: 660) all contribute to the formation of collective cultural media imaginaries about new media technologies. What is of particular interest here is that media imaginaries are not just reflections of our ideological investments in, or understandings of, media technologies; they are in themselves performative and have an impact on both media development and social change. Bucher (2018) points out that social media use is just as much affected by users' imaginations of technical infrastructure and affordance as it is by the technical system itself. While her research mainly concerns algorithms, her assertion that 'experience and affective encounters [are] valid forms of knowledge of algorithms' (2018: 94) also applies to models of understandings of particular social media, and we take this forward in our analysis by paying attention to how protestors feel about the social media they use.
Drawing on Mosco's (2005) work on the digital sublime and media technologies as myths, Treré notes that the Internet 'forms the basis for the perfect myth since it tells a story about how communication technologies can help to realise the seemingly impossible dream of democracy and community ' (2018: 112). In practice, the Internet and in particular social media have altered how contemporary social movements and protest are organised (Funke and Wolfson, 2017). This organisation may look different across movements and time. Different movements are characterised by their different techno-politics and media practices (Treré et al., 2017). But what is the relationship between the technological and the political as they manifest in the imaginings of -and in particular the visualisations of -social media?
Social media offers new opportunities for people to engage in different forms of activism, political resistance and protest outside traditional political fora and civic institutions. However, there is little research on how activists perceive social media. This pertains to what Treré (2018: 2) refers to as 'the media/movement dynamic', by which he means the 'mutual shaping of social movements and media technologies'. We need a more nuanced understanding of the relationship between social media and protest movements; this involves questioning the prevailing communicative reductionism within social movement research, which tends to regard social media simply as a tool in the hands of protestors (Treré, 2018), as well as challenging technocratic statements such as the claim that the Arab Spring happened because of social media (Dencik and Leistert, 2015; Poell and van Dijck, 2018). The latter is an example of how technological overdeterminism risks simplifying the relationship between social media and social movements (Morozov, 2011), and it is something we believe needs questioning as the field matures. This is nevertheless a narrative that fuels and, arguably, is fuelled by the visual and discursive examples from our data, as discussed in this article. Although we acknowledge this powerful narrative, we still offer a critique by unpicking how it is produced rhetorically and visually. Treré et al. (2017) have studied digital protest media imaginaries relating to the 2011 wave of anti-austerity protests in Greece, Italy and Spain, noting the radically different imaginaries shaping the three different movements and in some cases differences in how particular social media platforms were perceived to have distinct purposes in relation to online protest activities. They argue that the ways in which digital technologies are imagined, including how specific perceived meanings, values, capabilities, and ideologies are ascribed to them, shape the practices developed to engage with them and can therefore configure distinct types of digital activism, leaving others aside. (Treré et al., 2017: 407) However, they don't conceptualise this in terms of 'visualizations, analogies to more familiar domains, or by the use of metaphors', which is how Bucher (2018: 97) suggests we make sense of media technologies, and we will pursue this in our analysis of primarily visual imaginaries.
Less attention has been given to how the technological affordance and business model of corporate social media have an impact on activist communication. Responding to Poell's (2014: 717) call for a more complex understanding of social media that takes into account not only 'how social media shape activist communication' but how 'these media are, in turn, shaped by intersecting techno-cultural and political economic relations', the discussion that follows later in this article takes into consideration how the Turkish state has sought to control Twitter output at national level and how Twitter's business model both supports and constrains activism.
Researchers have shown the significance of social media in mobilising protestors and consolidating their sense of group belonging (and for the Gezi Park protest specifically, see SMaPP, coherence as shared symbols' for movements, where 'meanings are assembled and conveyed, their identities negotiated and maintained, their imaginaries forged and propelled' (Treré, 2018: 204). Treré usefully highlights here how protest imaginaries are supported by the use of social media, and this supports our theory that, in particular, visual communication, as circulated on social media, informs such imaginaries. But we also seek to expand on this by suggesting that the idea of social media itself is a central part of contemporary protest imaginaries. Focusing on social movements' processes of meaning making, Milan (2015: 890) argues that social media is now the dominant domain for such interaction taking over as 'the process through which the symbolic takes form'. Similarly, as Fotopoulou (2016) has shown, the notion of contemporary 'networked feminism' is produced through a shared social imaginary made up of the discourses, values and ideologies that shape feminist activism as it intersects with digital media. This is important research because it helps us understand how deeply integrated social media is in contemporary movements.
We have yet to explore, however, how social media attains a symbolic value in itself as part of the formation of a protest movement's self-image.

Visual activism
Images and other practices of visuality have been part of political communication and protest communication for a long time. However, as visual media become increasingly central to our everyday communication and experience of the world, the field of visual activism has gained importance in the study of social movements, and perhaps particularly of digital activism. The field of visual activism is not bound by aesthetic regimes of 'art' or formal categorisations of its mediums of expression (expanding across cinema, photography, graffiti, street performance, digital media, etc.) but rather encompasses a wide, and more democratic, use of visuality (Mirzoeff, 2015) that is 'aimed at catalyzing social, political, and economic change' (Demos, 2016: 87). Grassroots visual activism often strategically remediates images that typically don't 'belong' in the realm of politics, thus subverting them, or simply repurposing them, so that they take on new meanings , and as Olesen (2018: 657) has noted about the production of visual injustice symbols online, social media has transformed 'the way photographs are politicized'. Poell (2014: 728) links social media to the prominence of the visual in contemporary protest, arguing that social media 'greatly enhances [the] visual character' of protest communication. He notes that the visual material in his case studies of protest communication across different social media platforms emphasises the spectacular and confrontational, but the study does not encompass close readings of images that would explore how the visual spectacle is produced. However, as Gerbaudo and Treré (2015: 868) point out, the visual communication in activism is important beyond reporting injustices or documenting street protest, and they emphasise its role in the 'identisation' of a movement. This is directly relevant to the types of images empirically explored in this article because they clearly perform identity work but also speak to the relationship between visual activism and the social media imaginaries of the movement. Treré (2018) also suggests that visual activist communication online serves both external and internal communication purposes, nurturing and reinforcing collective identities among activists. This is a point also made by Kharroub and Bas (2016) in their study of Twitter images of the 2011 Egyptian revolution where they note that images eliciting efficacy (e.g. protest activities and crowds, and various national and religious symbols) were as prominent as images that arouse emotion. Approaching the question of how images affect online activism from a different perspective, Casas and Webb Williams (2019) argue that images have a mobilising effect because they evoke emotional reactions in ways that text does not. The amount of literature on visual activism across social media platforms is still relatively small in relation to traditional media, and it predominantly deals with citizen photojournalism. The technologies used for producing, distributing and consuming visual documentation and expression, such as mobile phones with a built-in camera and an Internet connection, in combination with the use of social media, are also closely linked to the emergence of contemporary visual activism, which encompasses a range of activist strategies from citizen camera witnessing to vlogging, selfie activism and political meme culture (Shifman, 2013). These practices have a stake in reclaiming visual representation, often from a marginalised position. But visual activism is not only about ways of representing oneself or gaining visibility. For example, as Mirzoeff (2017: 85) has argued in relation to the online #BlackLivesMatter movement, such visual practices have created a new way of seeing [ . . . ] understood to mean that point of intersection between what we know, what we perceive, and what we feel -using all our senses. Unlike the traditional one-sense visual perspective, it is a collective way to look, visualize, and imagine.
It is perhaps here that we sense most directly the relationship between visual activism and social media imaginaries. In the collective sharing, remixing and recontextualising of images relating to a particular protest or social movement, the networked participatory processes of these acts themselves are implicated in the visual output, as are photographic and graphic design visualisations of the movement itself. A widely published digital visualisation of Gezi Park Twitter activity, 1 for example, visually communicates the 'global' support for the 'local' protest in Istanbul, with lines between nodes indicating 'connections' criss-crossing a globe-shaped background, a visual shorthand for the world.
Rhetorically, visual activism is engendered by, and employs, a range of modes and aesthetics, from the ethics and practice of bearing witness, often according to the convention of documentary realism (Nichols, 1991), to the carnivalesque, evocative, satirical and humorous. All of these were present in the Gezi Park case study. Furthermore, it is not uncommon for visual activism to draw on the rhetorical tools and established language of visual advertising and visual branding, including pictorial metaphors (Barthes, 1977;Foreceville, 1996) and graphic design. An example of such use of metaphors would be the use of images of polar bears or the colour green in climate crisis activist communication. Purposeful visual branding becomes increasingly important in the context of social media platforms, where the currency is attention. Contemporary protest movements and advocacy groups appreciate the value of visual communication that is instant, recognisable and impactful in an increasingly fast-moving and saturated mass media. Recent examples of how protest communication approximates and borrows strategies from the world of commercial marketing and branding include the umbrellas of the Hong Kong student protests (Wong and Liu, 2018) and the ubiquitous Guy Fawkes masks that are popular with the Occupy movement (Riisgaard and Thomassen, 2016). Later in this article, we will explore the visual repertoire of a social movement, which also includes visualisations of the movement itself, both metaphorically and symbolically. We will approach this by looking at the Gezi Park movement's visual shaping and imagining of Twitter as part of its protest. In the following section, we will unpick the visual construction of Twitter as the symbol of freedom and a Gezi Park protest advocate. We will start by looking at the production of protest iconography using the example of the emblematic Gezi Park protest symbol, the penguin.

Researching visual social media
The data collecting for this article involved multiple stages. First, we conducted a keyword query of tweets that were posted during May and June 2013, which was when the protest reached its most intense phase, using the words Gezi Parkı, as it would be written in Turkish. This generated a significant amount of visual data (see Table 1). When we explored the themes of the images found in this data, one of the more noticeable visual tropes was creative memes referencing the bird in the Twitter logo. Second, to further examine the symbolic connotations of Twitter in protest communication, we performed manual keyword searches across multiple social media platforms, generating images referencing the Twitter corporation in the context of protest, outside the initial May-June data-scraping period. Thirdly, to contextualise the visual analysis and explore how activists envisage in their own words the different types of social media they use as tools for their activism, the article draws on 37 interviews with participants who took an active part in the Gezi Park protests. By foregrounding a visual analysis approach, the article makes a contribution to a field in which ethnographic approaches dominate.

Twitter as a visual media
The article draws on the Aesthetics of Protest project's Twitter data set that was created by querying the Twitter API against an existing set of historical tweets (from 27 May to 30 June 2013). Over 20% of the tweets contained images (see Table 1), indicating that Twitter as a social media platform also needs to be conceptualised as a visual media. Images embedded within the downloaded tweets were extracted using a purpose-built PHP script. Retweets and duplicates were eliminated. At the end of this process, we had harvested 714 images for qualitative thematic analysis. Each image typically had 6-12 codes attributed to it and meaning categories were determined and integrated in different stages of increasing abstraction (see Table 2). Codes were derived from categories such as actors, attitudes, and content composition and we linked them to our theory by examining mode, style and aesthetic techniques as well as rhetorical devices. Digital images are extremely 'mutable' (Rose, 2016), meaning that they change and are open to different interpretations, so to mitigate against this, we worked as a group, discursively and collectively coding the first 100 images so that we could agree on and verify the accuracy of the coding.
Several visual themes emerged from the analysis of the sample, and a range of styles, techniques and aesthetics were identified. Key themes included the documentation of police confrontation as well as everyday life in the park (McGarry et al., 2019); photographs of street graffiti and banners; screenshots of text strategically converted into visual file formats; digital artwork and cartoons; and the prevalent use of humour (see Table 2). This allowed us to gain insight into the movement's accumulative visual self-representation. While photography and documentary-style images dominated, there were also noticeable instances of creative digital artworks, mash-ups, cartoons, visual puns and memes. We also identified some recurring strong symbolism associated with the protest, including traditional ones, like the Turkish flag; some highly recognisable ones, like the Guy Fawkes mask, shorthand for political subversion and adopted by Occupy movement and others, visually linking the Gezi Park protest to these other global uprisings; and some new iconic images that emerged from within this protest specifically, like the gas mask, the penguin and the woman in the red dress. These widely circulated symbolic mediations visually supported the construction of a collective identity but also served to fuel further vernacular creativity such as Internet memes that can be appropriated, remixed and recontextualised in different ways by anyone. Such visual practices are typified in the case of the reproduction of the Twitter bird (Figure 2), which this article will explore further in relation to what these images may tell us about movements' digital protest media imaginaries.
The prominent occurrences of the 'Twitter logo bird' used in various protest messages spurred the present exploration of what meaning the brand itself and social media more broadly took on in relation to the protest movement. To follow up on this, a more detailed engagement with the categories relating to 'iconography' and 'graffiti' ensued, involving further visual analysis of a selection of images that include visual references to Twitter (e.g. the Twitter brand, the Twitter logo bird). Using the model of social semiotics (Van Leeuwen, 2000) and Roland Barthes's (1972) approach of denotation and connotation to tease out some of the meanings (in a particular cultural context) of the images selected for close reading, we considered the formal, aesthetic qualities of the images -their composition, colour, the production techniques used -and also conducted an interpretative analysis, noting their semiotic hybridity, symbolic content and connotative references to the protest. In the final iteration, this critical visual analysis thus links the images (content and form) to contexts to understand how meaning is produced via 'codes, aesthetics and rhetoric' (Aiello and Parry, 2020: 4). The analysis was guided by our theory that the media imaginary of a social movement is expressed in its protest-related communication and that the visual communication on social media, with its layers of denotation and connotation, may yield insights into the movement's construction of a self-image.

Mapping an expanding media imaginary
To home in further on the visual theme of the Twitter logo bird, additional manual keyword searches were conducted across Twitter, Instagram and Pinterest, but without the set time-period constraints, to survey what further creative renditions (collages, digital posters, cartoons, graffiti or street art) visually referencing or citing the Twitter logo could be found in Gezi Park-related outputs. The aim of doing this was to get a sense of the expansion of the media imaginary that puts Twitter at the heart of the protest movement as it continues to be memorialised across these digital spheres. For these searches, we used keywords that had appeared in images found in our primary data set, such as occupygezi (see Figure 2) on its own in English, in Turkish (i.e. direngezi), and in combination with the word twitter (i.e. occupy gezi twitter). We did a small number of such manual searches with the purpose of (a) establishing whether the visual implication of Twitter as part of the Gezi Park movement, which we had documented in our Twitter data, had lived on by continually being posted and (b) getting a sense of the aesthetic variety of images referencing this theme. Pinterest in particular stood out as a site for the archiving and memorialisation of Gezi protest digital artefacts -not only through practices of collecting documentary images from the protest following the site's scrapbook aesthetics (Jones, 2016) but also through users continually engaging with digital art and political posters from across the Internet centring on the Gezi Park protests. The image-based social curation site (Lui, 2015) is mainly a site for taste making through the curating of lifestyle images, but it also contains activist content. Users can search for and organise content via hashtags, and numerous tags reference the Gezi Park protest. We used search-based navigation by entering the keywords described above. Lui (2015) notes that the site is dominated by 'repins' of content already on the site, so once an image has been included on the site it is likely to be continually republished, and this characteristic of the platform has been taken into consideration. In other words, while being mindful of Pinterest's very particular protocol of content circulation and its thoroughly commercial logic, we note that some users wilfully repurpose the platform for activist expression. We could conclude from a relatively small sample that the concept of a bird in the style of the Twitter logo wearing a gas mask was redeployed in different ways and that the Twitter logo bird was imagined in various visual narratives, such as cartoons. The fact that the motif occurs outside the historical data set to an extent confirms that the images we chose for close reading are indicative of the imaginary we seek to analyse.

Interviews
The visual and thematic analysis of the images discussed in this article is further contextualised by textual discourse analysis of in-depth interviews conducted in 2017 with 37 individual participants who who were recruited using a snowball technique. 2 The participants were men and women aged 25 to 75. They were all activists who took part in the Gezi Park protests. Some camped in the park and others participated in the protest on a daily basis. In these interviews, the participants reflected on their involvement in the protest and on the role of social media and how they used it themselves. In a similar study on perceptions of social media's role in activism relating to protests in Israel in 2011, focusing on Facebook, Lev-On (2018) found that different actors within or adjacent to the movement held very different views on the function and significance of social media vis-à-vis mainstream media. In the light of these results, the thematic analysis of our interviews focuses for the purposes of this article on the participants' engagement with and perception of social media as well as their views on mainstream media.

Intersecting protest and media imaginaries
A prominent iconic image for the Gezi Park protest movement is a penguin wearing a gas mask (see Figure 1). Using humour and popular culture aesthetics, it references two key components of Gezi Park resistance. Firstly, the national mainstream media's initial refusal to report on the protest, exemplified by CNN Turk's decision to broadcast a documentary about penguins instead of reporting on events as they unfolded in Istanbul, and secondly it draws attention to the police's tear-gas attacks on protestors. The penguin has become emblematic for the movement and appears in digital collages, satirical cartoons and profile pictures, as graffiti, and more recently has been reproduced on T-shirts and other commodities and artefacts, including body tattoos. Initially, the penguin imagery directly and satirically referenced the news reporting strategies of Turkish broadcast media, connoting a press freedom in crisis. It was subsequently mobilised as a symbol of resistance and empowerment, and often adorned with protest paraphernalia or animated to express dissent, as in Figure 1 where the penguin appears to raise a fist in defiance, connoting the determined mood of the movement. During the many processes of reproducing, recirculating and remediating multiple versions of the penguin image across social media, at some point it overlapped with the official Twitter brand icon, the blue bird, and a semantic overlaying or blurring took place (see Figures 2 and 3). The Twitter bird thus took on similar associations with the movement that the penguin already signified. In what follows, the article will discuss the symbolic meaning of Twitter to the protest movement, beginning by investigating the creative engagement with the Twitter brand in the protest's visual output.  We will now turn our attention to unpicking how the Twitter logo bird was appropriated as a symbol of freedom of speech that coincided with the Gezi Park movement's self-image and points to how Twitter came to symbolise the key values of the movement. This is contextualised by the discourses surrounding Twitter that were produced by the protestors and the government. Homing in on creative engagements with the Twitter logo within the Gezi Park movement offers an opportunity to study media imaginaries on a more detailed level and to move away from generalisations about such vast notions as 'the internet' or 'social media' and drill down into what the visual output tells us about the 'actual imaginaries of real people' (Treré, 2018: 112). The analysis demonstrates how the social media technology is not just a communication tool; it is also a significant site for producing social values that in turn shape media practices. How the Twitter logo bird is reimagined in the visuals discussed gives an indication of the social values ascribed to the media corporation.
At first glance, examples like the Twitter bird depicted wearing a gas mask ( Figure 2) may appear to be examples of brandalism or cultural jamming, which typically involves reproducing alternative versions of a targeted brand's logo or advertising campaigns that undermine the intended message, as a 'symbolic form of protest located within a field of anti-corporate activism' (Wettergren, 2009: 2). However, in the case of the Gezi Park activists' interference with the Twitter company logo, while the visual intervention draws on the logic of appropriation inherent in such practices, there is a significant difference: the attitude of the activists towards the brand is not one of criticism or disapproval. Rather, we are looking at an appropriation with different intentions. Looking at the collage in figure two, the Twitter bird has taken to wearing a gas mask, thus symbolically signifying not only that it is in the midst of protest events but also that it is on the side of the protestors and, like them, suffering attacks from the police. The image is performative in that the subject, the bird, becomes a Gezi Park protestor when wearing the gas mask. And because the blue bird is a metonym for Twitter, the corporation becomes a protestor in this image. Another aesthetic component that is important to note in this image is how it is designed to mimick street graffiti by using a photograph of an urban wall surface as the background onto which the blue logo mark is superimposed. The contours of the bird and the mask are rendered in the style of stencilling. This, together with the #occupygezi slogan font, gives the appearance of street graffiti. Mimicking street graffiti aesthetics, that is, simulating the visual language of the streets, in digital protest communication produces a layering of the digital realm and the protests on the ground. In this visual metaphor, the gas attacks on protestors symbolise the battle for press freedom, and in this combat, Twitter is depicted as fighting alongside the activists. Similarly connecting Twitter to the Gezi protest by depicting the logo bird wearing a gas mask, the image in figure three also produces an imagining of Twitter as an actor on the protestors' side, this time further underlining the association by adding the tagline 'Gezi Summer. Revolution in 140 characters', which directly references the specific number of characters which is the (former) maximum length of a tweet, a well-known characteristic of the social media. The Twitter bird is depicted in a more active and confrontational pose here, with its wings spread out, facing the viewer. Animating the Twitter bird in this way speaks to how activists imagine the social media platform as an active agent.
Various versions of images of the Twitter logo bird wearing the gas mask appeared across different digital collages and memes, and by March 2014 when the Turkish authorities temporarily banned Twitter, the logo image was already established as visually representing the Gezi Park movement and appeared on satirical images protesting against the ban across social media. 3 The use of the Twitter bird to visually represent variously the Gezi Park movement (self-image) and its ideals of freedom of speech as well as the movement's ideas about how social media serve the movement involves symbolic diffusion and semiotic layering. The objects of their imaginaries are thus often more abstract such as a broad coalition of government critical voices connected and amplified via social media.
This image exemplifies the many playful and creative engagements with the Twitter bird that produce an alignment of the social media company with the aims and ideals of the protestors. The bird is represented in different ways: as the 'postman' of the movement's message; as a defiant objector to the authorities; and as a hopeful peace symbol carrying an olive branch in its beak. Some noteworthy examples of this include the cartoons by the Brazilian political cartoonist Carlos Latuff, 4 whose satirical depictions of Erdogan were widely circulated on Twitter during the protests. One of his images depicts the Twitter bird fouling on Erdogan's head and is an example of incorporating the brand into a well-established political satirist's style. This particular motif clearly resonated with others, exemplified by the many different renditions that were circulated. Another cartoonist depicts Erdogan trying to shield himself and run away from a flock of Twitter birds aggressively descending on him in an image that evokes Hitchcock's classic 1963 horror film The Birds. Several cartoonists have subsequently been subjected to attempts by the Turkish government to censor their work. 5 However, Twitter has not made any effort to repress or remove activist artwork or cartoons from the point of view of protecting its brand image, despite the fact that any interference with the design of the logo is against Twitter's terms of use. Twitter as a corporation is, like most companies, very protective of its logo design, and as it explains on its website, if it is replicated, it must not be modified in any way. Three of the guidelines are as follows: 'Don't surround the logo with other birds or creatures', 'Don't animate the logo or make it talk, chirp, or fly' and 'Don't anthropomorphize the logo', 6 but all the images discussed here do this. This raises the question of why the corporation decided not to intervene. One way of reading the activists' engagement with the Twitter brand is to think about their visual paraphrasing as a form of reverence or something akin to Internet cultures of creative fan art. In other words, the images connote a strong attachment to what the brand has come to signify to activists. The creative appropriation evidenced in the digital artefacts online is clearly not intent on brand sabotage and as such perhaps not perceived to be worth pursuing as copyright infringements. The Twitter corporation has not commented directly on its position in relation to the Gezi Park movement. However, to further nuance our discussion of the relationship between the social media corporation and the Gezi protest, we can look at two adjacent events: the protest-style slogan T-shirt featuring the Twitter Blackbird logo and the hashtag #StayWoke produced by the company in 2014, which stylistically mimics how the Gezi Park activists used the company logo, and the corporation's response to the Turkish authorities' attempts to control content on Twitter. These will be discussed further below.
Another aspect that clearly strengthened the protestors' embrace of social media, and Twitter particularly, was the government's open condemnation of it. In the next section, we will discuss how such statements by the AKP, in combination with the decision by prominent Turkish news channels to self-censor, contributed to the emerging image of Twitter as the de facto public service news broadcaster.

Social media imagined
The failure of domestic mainstream traditional media to report on the protest created a void that social media (and to an extent citizen media) filled. The protestors and its followers around the world both practically and symbolically turned to Twitter and other social media as an alternative to, or perhaps as a replacement of, public service media. The lack of domestic news coverage in the early days of the protest and the bias of mainstream media regarding the reporting of the events meant that turning to Twitter was in effect turning away from traditional news media. Some of the visual social media memes we collected from Twitter communicated this in no uncertain terms, with calls to boycott particular newspapers, TV channels or media corporations. A similar picture emerges from the statistical research conducted by the Turkish public perception polling organisation KONDA, which in a 2014 report stated that although the main source from which the general population first learned about the Gezi Park protest was television (70%), among those protesting in the park the main source of initial information about the protests was overwhelmingly social media (KONDA, 2014: 25). It is reasonable to conclude from this that the use of social media as a major communication tool for social and activist interactions as well as an important source of news was already established among respondents in this group. However, the way they used social media more actively and with increased intensity changed as the protests evolved. One interviewee described this change: 'Although I wasn't a Twitter user -I only had an account -I became an active Twitter user during that period -very intensely' (Onur). In the interviews we also found support for the phenomena described earlier where the mistrust of the mainstream media increasingly created a space filled by social media, approximating in people's perception an independent news organisation in terms of news reporting: 'I have not been watching the news for a long time. It is because there is no news channel that can be watched. I learn the news from Twitter' (Munevver). And one respondent commented on how the activists would take on the responsibility of reporting on events: 'I mean when you are despairing about TV and other news outlets, you try to be a media [outlet] yourself and try to inform in some way' (Sahin).
Visually, the distrust of mainstream media was most strikingly communicated through the image of the penguin. As mentioned earlier, in late May and early June 2013, very large numbers of people were protesting against the government on the streets in Istanbul and elsewhere, and the police retaliated with increased brutality, while the Turkish mainstream media notably avoided reporting these events. The term 'penguin media', used to describe the self-censoring Turkish mainstream media, was coined by Gezi activists, and it referred to the widely shared meme of two television sets next to each other, one showing CNN international broadcasting live from the Gezi Park protests and the other showing CNN Turk broadcasting a documentary about penguins. It was this image that ignited more widespread criticism of the government's control of broadcast media using political and financial pressure (Tufekci, 2014). The image's aesthetic draws on a domestic yet absurdist juxtaposition -the idea of simultaneously watching two television sets in a home setting is ridiculous -while also communicating through a rhetoric of immediate realism by presenting us with a simple snapshot of two different programmes being broadcast at the same time. Incensed, activists initially turned the penguin into a motif for the hostile mainstream media that was more or less controlled by the AKP. As the protests evolved, in the streets and online, the movement appropriated the image of the penguin (see Figure 1) as one of their mascots. Demonstrators appeared in penguin outfits or wearing DIY penguin-style facemasks, performing their resistance to mainstream media through their playful costumes, and many other visual representations of penguins (e.g. DIY art, graffiti, cartoons) appeared in the streets and on social media, with the aim of drawing attention to the extent of the government's restrictions on the media and censorship of the news. Such satirical responses have become emblematic of the Gezi protests. As many as 19% of the images in our data set used humour as the primary frame. As noted above, over time the image of the penguin became iconic. It is endearing and somewhat comical in its posture. 7 There are more images of penguins in the data set than the otherwise most prominent images associated with the protest such as the woman in red, the standing man or the whirling dervish wearing a gas mask. The 'çapulcu' penguin has thus gone from hiding the truth to exposing it with regard to the relationship between the media and the government.
The Gezi Park activists we interviewed also commented on how 'news in social media and news in normal press media were so different' (Kalan) and stressed the importance of providing a corrective to the mainstream news by telling their own personal stories via social media. One participant stated that they were motivated by 'show[ing]  One of the media imaginaries underpinning the activists' motivation for using social media to inform themselves as well as to report on the protest is the idea of the 'uncontrollable' Internet, alluded to in Erdogan's statements, which stands in contrast to the stark editorial control of traditional media. There appear to be two main concepts that make this an appealing idea; first, that the immediacy of 'direct' or live reporting via a protestor's smartphone provides a first-hand, authentic account with no editorial interference, and second that a plurality of accounts in itself guarantees parity. Both of these concepts can be challenged. Documentary reporting always produces images through a particular editorial lens and aesthetic (see McGarry et al., 2019). And the number of uploaded images or videos does not in itself ensure a diversity of perspectives, or 'verify' integrity. This is not to say that activists' use of mobile media technology and social media to report on human rights infringements and other injustices is not important, valid and effective, but it needs to be understood as shaped by these persistent imaginaries of immediacy and plurality.
In the light of increasingly vocal criticism of mainstream media as defunct, Twitter became a figure of blame for the authorities to attack. The Turkish Prime Minister, Erdogan, blamed Twitter for the unrest, calling it a 'curse', a 'menace' and a 'scourge': 'Now we have a menace that is called Twitter . . . To me, social media is the worst menace to society', he stated (Erdogan cited in NY Times 2 June 2013). While reflecting negatively on Twitter, this statement actually simultaneously inscribes the social media platform as a force to be reckoned with in society. The hyperbolic language used (e.g. 'the worst menace') may be rhetorical but reveals that the authorities feel threatened by Twitter's role in Turkish protest (because authoritarian states want to control media and information) and grappled with the control of information surrounding the event of the Gezi protest. This fear is further exemplified by statements like the one made by Ali Sahin, advisor to the AKP: 'A provocative tweet is a lot more dangerous than a vehicle rigged with a bomb exploding in a crowd' (cited in Yaman, 2014). Again, the commentator is using hyperbolic language and the metaphor of a car bomb explosion to express the idea that the protestors are disruptive and dangerous. However, the statement also reifies the power of social media and clearly confirms that Twitter is a powerful tool in the hands of the protestors. This is similar to how the platform is depicted in several of the digital posters in which Erdogan is being attacked by a 'dangerous' Twitter bird or flock of birds (see e.g. Sofia Mamalinga's cartoon 'The Birds' 8 ). What the politicians' statements quoted here illustrate is their contribution to a shaping of social media imaginaries in public discourse. These form part of a wider, both visual and textual, discourse that works to underline how social media is perceived to be the main domain for criticism of the government and a key place for political mobilisation. Commenting on Turkey, Karakaya and Glazier (2019: 292) note how 'social media's nebulous character and often deft ability to avoid government control, may make it a particularly useful place for political news' in an illiberal democracy and describe how since the Gezi protests, the government has continued 'to play a cat-and-mouse game with both traditional media sources and online critics' (Karakaya and Glazier, 2019: 300), citing the increased restriction on the free press since the 2016 coup attempt and mass investigations of social media accounts. Some interviewees expressed disillusionment with the development of the Turkish digital public sphere in the aftermaths of the protests, referring to the various AKP trolling strategies (Bulut and Yörük, 2017). Bulut's (2016: 607) analysis of the relationship between social media companies and the Turkish state has similarly shown that since the Gezi protest, the state has continually forced social media companies 'to comply with [ . . . ] [their] own political framework'. When Turkey blocked Twitter in March 2014, this attracted international attention and was widely regarded as an authoritarian and clumsy attempt by Erdogan's government to repress the freedom of citizens for the sake of preventing leaks implicating political corruption. The blocking of Twitter was both technically flawed and subject to a legal challenge. However, for the purposes of our argument here we note that it backfired mainly because of the negative publicity it caused the AKP (see Harris, 2015). The many memes and cartoons posted on social media in response to the suspension of Twitter in Turkey reinforced the image of Twitter as a symbol of freedom of speech -freedom visually represented by the Twitter bird in flight -in contrast to the Turkish government, depicted as attempting but failing to variably encage or shoot the bird. Twitter's Public Policy account took a stance by tweeting 'We stand with our users in Turkey who rely on Twitter as a vital communications platform. We hope to have full access returned soon' (21 March 2014), but its transparency reports reveal that the company has complied with a significant number of requests to remove content and accounts since 2014 showing how extensively Turkey makes use of Twitter's 'country withheld content' policy. 9 In the following section, we will look more closely at the company's views and actions in relation to being associated with social uprising.

Twitter and activism
As a company, Twitter is keen to be associated with a commitment to 'freedom of expression' and civic engagement (twitter.com). This in part explains why its attitude to the unauthorised use of the Twitter logo for activist purposes has been an accepting one. 10 However, imagining Twitter as a liberation site that symbolises freedom of speech is not unproblematic. Youmans and York (2012) point to several cases in which social media platforms' design and policies create tensions between the uses by activists and the commercial interests of the corporations. Such tensions include how collective action on social media potentially conflicts with the platform's priority of monetisation, which is linked to 'authentic' identity requirements, but also concern governance and pressure from state power. Bucher (2018: 8), applying a critical political economy perspective notes how the algorithms of social media platforms such as Twitter and its 'programmed sociality' are ostensibly designed to create financial value. This should be a central concern when discussing Twitter and activism. It may help understand why the company's stance on civic engagement varies depending on the context: for example, Twitter has very publicly embraced the #StayWoke campaign, visually exemplified by the company's CEO, Jack Dorsey, wearing a Twitter-branded slogan T-shirt with the hashtag when speaking publicly. While the examples of Gezi Park and the #StayWoke campaign demonstrate how Twitter seeks to align itself with certain types of protest, the company has on other occasions been more ambivalent about being associated with social uprising, such as in the case of the August 2011 riots in London and other British cities (Vis et al., 2013). This ambivalence raises questions about what types of protests are advantageous for the company image and what types are less palatable. Youmans and York (2012) note that avoiding liabilities and bad publicity are key concerns for corporate social media driven by revenue protection. This corresponds with how Jack Dorsey often circumvents being aligned with particular political causes or leanings. Here, for example, he uses the analogy of the public square to describe one of the imaginaries of Twitter: We believe that many people see us as a public square and use us as a public square [ . . . ] they have the same sort of expectations they would have of a public square, like Bryant Park. (available at recode.net; Dorsey in conversation with Jay Rosen 14 July 2018) Both Dorsey's and Erdogan's statements illustrate how the press as well as social media corporations themselves also contribute to the production of social media imaginaries, albeit from different standpoints. The idea of Twitter as the public square, a digital replica of the democratic public square that Taksim Square could not be, may have resonated with protestors. It is, however, more likely that they would have taken a more pragmatic approach. Research by Jenzen (2017) on trans youth digital activism has demonstrated that over time, mainstream, commercial or 'feudal' social media has also allowed for much more diverse activities than it is intended for. Within digital discursive environments, 'the loci of power are much more diffuse and instruments of ideological control and discipline are more subtle and complex' (Jones et al., 2015: 1). Having said that, it should also be noted that there are real risks involved when using commercial social media platforms for activist purposes, such as accounts being closed or data being handed over to the authorities (Neumayer and Svensson, 2016). Vis et al. (2019: 249), commenting on the Black Lives Matter movement, express concerns over 'the superficiality of political relationships that appear to have developed between black protest and the tech industry, and also of the ways that Twitter in particular has benefitted from activists using its platform'. This resonates with a growing worry that businesses in the digital economy, like Twitter, subtly co-opt models of democratic expression by 'implicitly endors[ing] a notion of public collectivism that functions entirely inside commodity culture' (van Dijck and Nieborg, 2009: 855). The paradox of social media in contemporary protest is illustrated here; it provides opportunities and material resources, but it can also be limiting because of its ultimately corporate drivers.

Conclusion
Social imagining is a collective process that brings a movement into being through a narrative shaped by the people involved in the movement itself and by societal institutions and public reactions to the movement. Visual communication is an important part of this narrative and provides key insights into activists' imaginaries and how they negotiate and dramatise their collective identity. Paying particular attention to the intertwinement of the aesthetics of protest and corporate aesthetics, we have sought to highlight how the concept of social media imaginaries elucidates how protest movements imagine the 'we' that their collective identity hinges on. The article has also shown how the concept is useful to understand how protest movements conceptualise their media practices. Through an analysis of images and adjacent discourse, the article has sought to empirically fine-tune how this social media imaginary is constructed through protestors' use of aesthetics and how the visual narrative of the collective identity is built. While we have prioritised visual communication in this article, we have not fully mapped the mediation and remediation practices of those producing and/or distributing the images. Understanding the propagation patterns of images online may be important in terms of better understanding how effectively they operate as political communication across both private and public spheres and across activist and professional media (although they are sometimes appropriated and remediated in new contexts), but our data on its own do not allow for networks structure analysis in relation to the images collected. In our case study, we demonstrate how, as part of widely subscribed social media imaginaries, Twitter has emerged as a signifier of contemporary protest, and we have found that during the Gezi Park protest, Twitter was variably imagined as the extended public space for protest expression -the 'digital public square' -as assuming the role of mainstream broadcast news media, and as the protestors' ally and a symbol of freedom of expression. However, we also note that corporate social media design and policies can inhibit protest and involves calculated risks for activists. The scope of this article was to demonstrate how the aesthetics of protest play an often overlooked but important role in complex processes of a social movement's emergence. Further research is needed into how not only corporate social media brand strategies but the 'inherent logics' of such platforms (Neumayer and Rossi, 2018) shape the expression of protest to fully understand the constraints and possibilities of using social media for contemporary protest movements.