#GiletsJaunes: Exploring the Yellow Vest movement on Social Media.

(1)

Master Thesis

#GiletsJaunes: Exploring the Yellow Vest movement on

Social Media.

Anastasia Bianca Maria Pelini (12360694)

MA New Media & Digital Culture

University of Amsterdam

June 28

th

_{, 2019}

(2)

Abstract

This thesis aims to explore the Yellow Vest movement through their use of the social media platforms Facebook and Twitter. The main research goal relies on understanding in which ways the Yellow Vests utilize these social media platforms. In order to accomplish this goal, relevant literature was explored, and thorough research was conducted. This research consisted on the analysis of 250 Facebook posts on the public page of an important figure of the movement. Moreover, two closed Facebook groups were examined. Finally, the 250 most popular Twitter profiles regarding the movement were observed. This research led to findings regarding the nature of the content the Yellow Vests discuss online, as well as understanding the coherence of the movement. The content found on the two platforms were compared in order to provide a cross-platform analysis. The Yellow Vest movement remains at the heart of France’s political climate, as they have just reached their eighth month of protests. The Yellow Vests, their political tendencies and online behaviours will be presented in this thesis.

(3)

1. Introduction: The Yellow Vests leading protests in the streets

and online

“Macron Démission!” is a popular slogan that Yellow Vests have been chanting during protests, and numerously posting on Facebook; meaning “Macron Resignation!” (Le Cain). As the movement has grown, while being discussed by media outlets, this thesis aims to explore the link between the notions of political movements and social media. More specifically, it will focus on the Yellow Vest movement in France and the protesters’ use of the social networking platforms Facebook and Twitter. In order to understand the relation between protesters and the digital world, the Yellow Vest movement needs to be presented.

1.1 An introduction to the Yellow Vests as a political movement

1.1.1 The birth of the movement

The Yellow Vest movement began in November 2018, revealing “deep and intense anger towards the government and elites among the working classes in France” (Duvoux and Papuchon 1). More specifically, the Yellow Vests are often part of France’s lower middle class, Gobry defines them as being a product of the French political system, they are “people who make just enough money to fail to qualify for most social programs, but not enough to live comfortably, even as the country’s elite keeps doing better” (Gobry 24). Even though the movement does not belong to a specific political party, does not recognize any leader, nor has official demands, it is historically clear when the movement began. The rise of fuel taxes created anger from those living in cities where public transportation is not accessible, and in May 2018 an online petition was created voicing these concerns. Created by Priscilla Ludowsky, who eventually became an important figure in the Yellow Vest movement, the petition received over a million signatures (The Guardian). In October 2018, when the petition was still growing, she was contacted by Eric Drouet, a truck-driver who wanted to promote this petition online. He created a Facebook event, inviting people to protest against the rise of fuel prices; this marked the first protest of the Yellow Vest movement in 2018. As the movement grew and protests began spreading throughout the country, it moved from a protest solely about fuel taxes, to other demands towards the government. This included job insecurity, as President Macron reformed France’s labour code “whose cumulative effect is that it is now easy to hire and fire in France” (Gobry 27). Moreover, the president has come under criticism not only because of

(6)

representative of a specific “caste” which includes “France’s technocratic and pseudo-meritocratic elites” (Gobry 26).

1.1.2 The movement’s demands uncovered through large-scale protests

Shortly after the creation of the Facebook event by Eric Drouet, large-scale protests grew throughout the country. November 17th_{, 2018 “marks the first and biggest day of road blockades} across France, with nearly 290,000 demonstrators donning Yellow Vests” (The Local). As the protests consisted of highway blockades, they had a major impact on the traffic of the whole country. The protests became more violent, and on the 27th_{of November the president called} for a “three-month national consultation” and proposed a mechanism to adjust the fuel tax (The Local). As no concrete plans to lower the taxes were presented by the president, the protests increased in violence and attendance. On December 4th_{, the Prime Minister announced that} “planned tax increases on petrol and diesel on January 1st_{would be suspended for six months”} (The Local). The protesters’ demands were clear at the beginning of the movement, although this changed as the movement progressed. They demanded the increase of the price of fuel to be stopped immediately, and indefinitely, which explains their growing anger at the six-month term governmental proposal. At this point, more and more people had joined the protests with various demands. Coming from drastically different political parties, these demands varied: the left wing demanded “higher minimum wages, lower retirement age”, while the right-wing were claiming “small-business tax cuts, a crackdown on illegal immigration, tougher policing” (Gobrys 25). These very different demands all began from a petition on fuel taxes and grew into a general resentment towards the government’s new policies.

1.1.3 The movement’s important figures

As the movement brings together people from different backgrounds and political parties, it does not have specific leaders. Gobry defines the movement as “entirely organic and leaderless” (Gobry 25). However, certain people who joined the movement have become important figures through various ways. Priscilla Ludowsky is one of them, as she created the petition, attended protests and spoke on behalf of the movement during interviews on national television (The Guardian). Eric Drouet created the first Facebook event leading to a protest and since then has highly been involved with the organization of protests. He created a Facebook page named “La France énervée” which translates into “The angry France”, on this page he regularly broadcasts Facebook lives while he works as a truck driver, where he discusses his political opinions and

(7)

different ways to impact change (Anizon). He also uses Facebook live during protests, and famously streamed a conversation with government officials. Even though Eric Drouet repeatedly claimed not to be a leader, he was asked to attend a “crisis meeting” at the Élysée Palace on December second, thus being recognized as a leader of the movement by the government. Other Yellow Vest representatives were invited and parts of the meeting were streamed on Facebook (The Local). The representatives claimed the meeting was unsuccessful and denied a second invitation by the government, “with some saying they had received death threats from other protesters contesting their claim to speak for the movement” (The Local). Eric Drouet has therefore shared key moments of the movement on Facebook, whether it is government meetings, to protest live-streams; his Facebook page is indeed very active. As the protests became more numerous, they became more and more violent. Eric Drouet used his Facebook lives to show online certain acts of police brutality. Police violence became an important topic amongst protesters, and it was completely ignored by the government until an elderly woman died in Marseille after being hit with a tear gas grenade (France 24). The conversations about police brutality became even more prominent after an important figure of the movement, Jerome Rodrigues was injured during a protest. The forty-year-old construction worker was placed in a coma after being struck in the eye by a rubber-bullet, he eventually lost his eye and his lawyer claimed that he would be “handicapped for life” (France 24). The fact that an important figure of the movement was injured in such a severe manner led to more anger, and more violence during the protests. Overall, a study by Adoxa claims that twelve people have died during the protests, and seventeen thousand have been injured (Adoxa). The various protests have revived conversations about police brutality in the media, while the government refuses to take responsibility for the training of its police forces, or the equipment used to repress protests. Incidents of police violence have been documented on social media, reminding the impact that social networking sites can have during protests (Freelon et al).

1.2. The movement and social networking sites

Describing the birth of the movement showed that the creation of petitions and Facebook events had a major impact on the movement. Nicolas and Candau claim that Facebook “posts are abounding, the number of groups are increasing, and leaders are emerging here and there on Facebook. The social network creates links and not only shares information but also the stress and fake news”; therefore, Facebook facilitates political conversations, but also the spread of false information and frustration among protesters (Nicolas and Candau). Certain

(8)

pages and Facebook events are also a key tool for the government to estimate the numbers of protesters, as the “planned gatherings are a great cause of concern to the French government and the security forces” (Nicolas and Candau). Twitter is another social networking site which is used to discuss the movement, as French news source “La Dépêche” claimed that more than one million tweets under the hashtag #GiletsJaunes were published in forty-eight hours, in November 2018 (La Dépêche). It is still unclear how both Facebook and Twitter are being used by protesters, and what type of content is being posted and shared. This leads to the research which will be conducted in this thesis, to explore the movement as well as its link to social media.

1.3. Research Questions and academic contribution

In order to frame this thesis, one main research question as well as two further questions have been determined:

§ In which ways do the Yellow Vests utilize social media platforms ?

§ What does an analysis of the Yellow Vests on social media say about the coherence of the movement ?

§ What are the differences between Facebook and Twitter in the ways they are used to discuss the Yellow Vest movement ?

This thesis will therefore aim to understand the various ways in which the Yellow Vests use social media. This will be conducted through an analysis of Facebook pages and Twitter profiles. This study will enable comparison of the two social media networking sites, by determining the differences between the content posted, the relation between users and the affordances of the platforms. The key concepts presented in the literature will be used together with the findings of the analysis to understand the relation between social media and the Yellow Vest movement.

In order to provide an academic contribution, this thesis aims to shed light on a very recent political movement. As the Yellow Vest movement began in November 2018, few researchers have had time to expose their findings on the movement. Moreover, this thesis does not contribute to a specific academic discipline, as it crosses different fields such as social

(9)

movements, digital activism, or platforms theory. Using this literature will enable to conduct qualitative research on 250 Facebook posts, two closed Facebook groups, and 250 Twitter profiles. Various media sources have discussed the importance of social media as a communication tool within the movement, but very few have conducted research examining specifically how these platforms are being used by the Yellow Vests, and more specifically what subject matters they are discussing online (Semerdjian).

1.4. Thesis structure

In order to combine the analysis of relevant literature, as well as the presentation of qualitative research, the thesis will be structured in the following way. The literature review will aim to introduce concepts relating to the Yellow Vest movement. This includes social movements theory, digital activism and concepts of collective identity. Moreover, social networking sites will be theoretically examined using theories of platforms. Further on, the methodology chapter will precisely describe the ways in which the online research was conducted, limitations and ethical concerns will be addressed. After the research has been conducted, the findings chapter will present the result obtained from the qualitative analysis. This chapter will lead to a discussion, combining the results from the research, to the concepts which were explored through the literature. Finally, the conclusion will aim to clearly answer the research question, as well as examine the impact of the literature and qualitative analysis. Furthermore, it will explore the need for further research.

(10)

2. Literature Review: A theoretical analysis of social media’s

political use, and its connective abilities

This literature review aims to reflect and analyse relevant academic work, enabling to critically examine the research questions that frame this thesis. Understanding precisely in which ways the Yellow Vests use social media is a topic which will be explored throughout this research, using literature as well as data. In order to understand the specificities of social networking sites, as well as explore what social media says about the movement, examining relevant literature is a necessary step.

2.1 Previous movements paved the way for political use of social media

In recent years, social media has become a tool which has been used to communicate during times of political instability. Various hashtags are created and used frequently to highlight injustice, or political tensions. In order to understand how the Yellow Vests are defined by their use of social media, it is important to look at movements which have paved the way for social media used in politically complex situations. The Arab Spring is a movement that began at the end of 2010 and spread from North Africa to the Middle-East. It can be considered as a leading political movement where social media was used “by protesters to organize themselves and to disseminate footage from rallies” (Bruns, Highfield and Burgess 872). This movement is important because of its long-lasting impact, but it is valuable to note that “protests and unrest in countries from Tunisia to Syria generated a substantial amount of social media activity” (Bruns, Highfield and Burgess 871).

Another group which strongly relied on collective organisation through social media is the Occupy movement. In 2011, crowds gathered in downtown New York, setting up encampments in order to protest against “the privileges accorded by politics to economic elites, who were responsible for the financial crisis of 2008” (Beraldo 77). Their slogan “We are the 99%” was highly circulated on social media, emphasizing their critic of financial markets and corporations. The main social media tool used to bring attention to the Occupy protests was Twitter. Not only used by protesters, a study by Gleason claims that “Twitter supports multiple opportunities for participation in the Occupy movement—from creating, tagging, and sharing content to reading, watching, and following a hashtag—which may facilitate learners becoming more informed, engaged citizens” (Gleason 966). Twitter has therefore facilitated discussion,

(11)

as well as created an “informal learning” setting, where other users can view information about the movement (Gleason 978).

Briefly presenting movements such as the Arab Springs or Occupy is not sufficient to understand their complexity and impact; however, it is valuable to look at a part of history where social media was used as a key tool for political participation. In order to place the Yellow Vest movement on an academic map, various theories will be used to define and frame the movement. Firstly, in order to further explore history, theories surrounding digital activism will be presented. These theories will aim to place the Yellow Vest movement within a framework of digital activism history. Secondly, an analysis of platforms will be necessary. From analysing key theories on the theme of platforms, to focusing specifically on Facebook and Twitter, this chapter will aim to provide essential theory on platforms as a concept, as well as a medium of connection between users.

2.2 Social movements and digital activism

In the following chapter, the notions of social movements and digital activism will be defined. This will be done through an analysis of academic definitions, as well as an examination of the history surrounding these terms. The particular link between activists and technologies will be observed. Using notions of media as practice, and “acting on” one can aim to understand how activists use digital tools to their advantage (Kubitschko).

2.2.1 Social movements and their history

When analysing theories revolving around the digital world, it can seem counterproductive to look at the past for innovative concepts, however, “we tend to equate ‘new’ with unprecedented, and forget that successful innovations are almost always built on the foundations of existing techniques and systems” (Lievrouw 28). Therefore, in order to understand digital activism, it is crucial to understand activism at its core, starting with social movements. Social movements have been defined in various ways, but can be considered as a “network of informal interaction between a plurality of individuals, groups and/or organizations, engaged in a political or cultural conflict on the basis of a shared collective identity” (Diani 13). A shared collective identity is a crucial part of traditional definitions of social movements. When observing the Yellow Vest movement, similar values might unite members, as well as common demands towards the government. Moreover, the symbol of

(12)

as being part of the same movement. This central symbol of the Yellow Vest can be described as a “contentious brand” (Beraldo). It produces the ‘conditions of recognizability’ of a movement, while being surrounded by controversies over its true definition (Beraldo 16). It also links to the notion of branding, meaning that it “can be seen as both a matter of strategy deployed for organizational purposes and a matter of identification reflecting cultural processes” (Beraldo 16); the symbol of the Yellow Vest therefore holds both identification, and strategic purposes.

The Yellow Vest movement contains inconsistencies, with demands that drastically vary within the movement, making it seem incoherent at times. The unity, as well as inconsistencies of a movement are referred to in the theory of media ecology. Introduced by McLuhan and Postman, this concept captures “the coexistence of and interdependence between human actors and technologies while addressing their systemic discontinuities” (Dahlberg-Grundberg 526). Media ecology highly relates to the link between a social movement and technology, it captures the notion that digital tools are “extensions […] through which human agents operate as political subject” as well as a tool that “embeds and precedes us”, meaning that it can come from the past. Digital tools are also engrained in social movements, impacting their ways of expression and construction of their collective identities (Dahlberg-Grundberg 526).

When referring to social movements and digital campaigns, it is valuable to approach the notion of digital activism. Kaun and Uldam describe digital activism as a “hot topic” and highlight the difficulties in defining the term (Kaun and Uldam 2099). They illustrate that certain definitions include activism which “engages both fixed and mobile devices with access to the Internet” such as various forms of hacking, and advocating for users online privacy; as well as broader definitions that “include all digital media used for political purposes” (Kaun and Uldam 2100). Considering this wide-ranging approach, observing the history of digital activism can provide more clarification on this specific notion. Karatzogianni describes four waves of digital activism, the first beginning in 1994 with strong political claims, and political groups or alternative media such as the Zapatistas and Indymedia. The second wave is linked to “anti-iraq war mobilizations”. After 2007, the third wave spread to other parts of the world, with major events such as Barack Obama using “the impact of social media activism to win the U.S. presidential elections. […] It is also one of the first instances where political participation and digital activism blend” making a crucial impact in representative politics (Karatzogianni 4). The final fourth wave relies on discussions surrounding privacy and surveillance, with crucial information revealed by Wikileaks and Snowden. Karatzogianni claims that this fourth

(13)

wave marks the “mainstreaming of digital activism” (Karatzogianni 4). As the history of the concept is presented, it is unproductive to identify the Yellow Vest movement to any specific wave, definition, or theory within digital activism. However, it is important to acknowledge the important history of the term, as the various diverse events of the past might have affected the ways in which the Yellow Vests use digital tools towards political goals.

2.2.2 Activists and their link to technology

Digital technologies pave the way for social movements theory to evolve through time, but more specifically, they can enable political groups to mobilize and organize (Dahlberg-Grundberg 524). This facilitation can lead to what Treré describes as the “coexistence and coevolution of activists and their technologies” (Treré in Dahlberg-Grundberg 525). This coevolution can be assimilated to the importance of social media within the creation of certain movements, more specifically “the role that hashtags, slogans and symbols played in assembling diverse groups, places and issues” (Beraldo 15). In order to understand the magnitude of the link between technology and social movements, the notion of media as practice should be introduced. Kubitschko introduces this notion, while bringing in different perspectives with his concept of “acting on” (Kubitschko). He describes that media as practice focuses on understanding “what people do with media” (Kubitschko 629). More specifically, Mattoni presents media practice as “social practices that combine interactions between media objects and media subjects, how these objects and subjects are perceived and operate in the media environment” (Mattoni in Kubitschko 630). This definition further proves Kubitschko’s claim, which is that almost any form of political engagement relies on media technologies and infrastructures, therefore it is important to expand the concept of media practice (629). This development of the concept is done with the use of the term “acting on”, Kubitschko presents it as a lens, through which one can examine an actor’s practices, and how they link to media infrastructures and technologies (629). His definition of the term particularly relies on the relation of actors and infrastructures for political contribution. Kubitschko further highlights the specificities of the notion of “acting on” (Kubitschko). As previous studies focused on “the use or the impact of media”, “acting on” focuses on “who can and does act on the conditions, formations and affordances of media technologies and infrastructures” (Kubitschko 632-633).

In order to illustrate this point, the case of the #MeToo movement will be linked to Kubitschko’s theory. This movement began on Twitter in 2017 and fights against sexual harassment. It “revealed the prevalence and magnitude of problems with sexual harassment and

(14)

assault, especially in the workplace” (Lee 433). Twitter was the platform that enabled the movement to grow tremendously, as a research led by Pew Research demonstrates ; “the #MeToo hashtag has been used roughly 19 million times on Twitter in the past year, and usage often surges around news events” such as the women’s marches, the Weinstein trials, or the hearings and nomination of supreme court justice Kavanaugh (Pew Research). Twitter is a platform which enables users to share their stories, relate to others or gain information on the movement. As this movement deals with sensitive issues and victims of trauma, Twitter is part of the social media platforms which are “revolutionizing the way users communicate by increasing the exposure to highly stigmatized issues in the society” (Manikonda et al 1).

When observing theory surrounding social movements, it is important to analyse the particular ways in which these groups form an identity. Furthermore, this collective identity can be built using technology as a key component. Milan presents the notion of cloud protesting, “a specific type of mobilization that is grounded on, modelled around, and enabled by social media platforms and mobile devices and the virtual universes they identify” (Milan 1). This notion combines the facts that activists interact using technology, as well as the fact that it is used towards “activists’ sense-making activities” (Milan 1). This notion is particularly relevant when applied to the Yellow Vest movement, as it specifically relies on technology. The Yellow Vest movement brings together people with various demands, and most importantly people who belong to different political parties. This can be seen as a divisive part of the movement, but Milan’s notion of “cloud protesting” proves that the Yellow Vests do not need to be a perfectly aligned political movement, as technology is a crucial part of what brings them together (Milan).

The way activists join movements because of their interests is linked to the formation of a group’s identity, more precisely a collective identity. Melucci defines it as “an interactive and shared definition produced by a number of individuals (or groups at a more complex level) concerning the orientations of their action and the field of opportunities and constraints in which such action is to take place” (Melucci 70). He further describes that collective identity relies on cognitive processes, such as the sharing of a defined language within society or a group, or a set “of rituals, practices, cultural artifacts” (Melucci 70). These allow information to circulate, as well as create a network of active relationships. Melucci claims these active relationships grow “between actors who interact, communicate, influence each other, negotiate, and make decisions”, they also rely on different forms of leadership, communication and

(15)

organization within a movement (Melucci 71). Finally, a key point in the formation of a collective identity is “a degree of emotional investment” which enables individuals to feel part of a united movement (Melucci 71).

Bennet and Segerberg depart from the concept of collective identity to make a distinction between collective and connective action. They present two organizational patterns which define a movement’s action. Firstly, certain actions “are coordinated behind the scenes by networks of established issue advocacy organizations that step back from branding the actions in terms of particular organizations, memberships, or conventional collective action frames” (Bennet and Segerberg 742). This form of action is therefore enabled with the help of specific organizations. The second form of action entails that platforms and applications can take “the role of established political organizations”, grievances are therefore shared using personalized accounts, as opposed to established social media strategies (Bennett and Segerberg 742). Bennett and Segerberg claim that the logic of collective action relies on “getting individuals to contribute to the collective endeavour that typically involves seeking some sort of public good (e.g. democratic reforms) that may be better attained through forging a common cause”, a collective action therefore strongly relies on a community (Bennett and Segerberg 749). On the other hand is the concept of connective action, which at its core contains “the recognition of digital media as organizing agents” (Bennett and Segerberg 752). More precisely, connective action formations present certain specificities. They create political content “in the form of easily personalized ideas”, such as the famous Occupy slogan “we are the 99%”; this political content is inclusive, as it brings together many people who are contesting a similar situation, with potentially different ideas (Bennett and Segerberg 744). Moreover, these formations gain impact because they are shared through “personal communication technologies” such as texts or shares on social media (Bennett and Segerberg 745). Whether it is the political content created, or the way it is personally shared, it leads to connective actions being a part of a group’s collective identity.

Treré and Gerbaudo suggest that the concept of collective identity has been neglected in the previous account; this neglect being theoretical, empirical and methodological (Treré and Gerbaudo 3). They believe that opposing networks to collective identity is problematic, as only two factors are considered decisive “in mobilising protest participants: the presence of network ties, and the presence of a sense of belonging to a certain category of people” (Treré and Gerbaudo 4). The concept of collective identity is therefore more complex than simply

(16)

contrasting networks and collective identity, especially when digital media are included in this concept, “social media have become the key site where protest identities, are created, channelled and contested.” (Treré and Gerbaudo 8). Treré and Gerbaudo distinguish two ways in which collective identities are being interpreted. The first one depicts collective identity as “a central object of concern” where social media platforms have a major impact on the ways in which “new identities are being forged and channelled” (Treré and Gerbaudo 8). On the other side, collective identities developed on social media platforms can be seen as having a “multiple and fragmented nature” (Treré and Gerbaudo). The concept of collective identity is therefore complex, and its unity or fragmentation varies. However, the impact of social media for the creation of this sense of collectivity is clear.

A key component in the formation of social movements and their collective identity is the link between activists and various types of technologies. However, technology and more specifically social media platforms present specific characteristics which enable activists with communication and organizational tools. The notion of platforms and their specificities will be further explored.

2.3 Activism on digital platforms

In the following chapter, platforms will be explored taking into account various points of view. The term will be defined, and notions such as platformization will be applied to social movements. Moreover, emphasis will be put on online connections and friendship-making on social networking sites as well as understanding how a platform’s structure impacts its users’ activity.

2.3.1 The complexity of platforms as tools for activism and online connections

When discussing social movements and digital activism on social media, it is particularly important to observe the concept of platforms. Bratton defines them as a “standards-based technical-economic system that simultaneously distributes interfaces through their remote coordination and centralizes their integrated control through that same coordination” (Bratton 42). He highlights their technical and economic characteristics, but they can also facilitate political exchanges. Bratton claims that platforms “add value both to what is brought into the platform and to the platform itself” (Bratton 41). Therefore, political discussions and content can be valued by the platform, while benefiting it at the same time. The popularity of platforms can be explained using many different factors, but a key one relies on the fact that platforms

(17)

“standardize and consolidate the terms of transaction through decentralized and undetermined interactions” (Bratton 42). A social media platform will present a standardized interface which will appear similar to all its users, therefore facilitating the use of certain tools for political expression.

When analysing platforms through the lens of social movements and political interactions, it is useful to observe the concept of platformization. Nieborg and Poell define it as “the penetration of economic, governmental, and infrastructural extensions of digital platforms into the web and app ecosystems, fundamentally affecting the operations of the cultural industries” (Nieborg and Poell 2). This concept highlights the importance that platforms can have, more specifically on online communication. Social movements and activists can use social media platforms to organize and communicate, however Couldry describes a myth regarding online collectivity. Couldry’s “the myth of us” describes social media users as being encouraged “to believe that our gatherings on social media platforms are a natural form of expressive collectivity, even though it is exactly that belief that is at the basis of such platforms’ creation of economic value” (Couldry 608). Couldry dismantles this myth, by highlighting that social networking sites change the conditions of politics, as they have an impact on the connection of political communications (Couldry 608). He acknowledges the importance of social media networks, and describes their “political uses, especially in mobilizing disparate and previously unconnected groups of people” he further highlights that this importance only represents a fraction of their power, he claims that “the larger question of what is a network society remains obscure” (Couldry 609).

The notion of platformization, as well as Couldry’s “the myth of us” can be further explored when specifically looking at the social networking site Facebook. Facebook is particularly relevant to social movements and online communication, as it is “structured around the principle of sociality, the relations between self and others” (Bucher 1). Bucher’s analysis presents the characteristics of friendships on Facebook, she claims that friendship on Facebook is not only represented by the connection between multiple users, but also by multiple actors, as well as “nonhuman software actors” (Bucher 2). These actors can be interpreted as being bots, algorithms, or other forms of nonhuman online interaction which impacts how a user views digital content. After a user has created and uploaded content on Facebook, other users are “aggregated by posting, promoting, and advertising content on, or via, a wide range of social platforms” (Niebrog and Poell 2). Bucher’s friendship analysis is not entirely positive, as she

(18)

highlights that platforms such as Facebook are “technologically driven and commercially motivated”, therefore leaving human connection outside of their main focus (Bucher 2). Brügger presents a different point of view as he chooses to highlight the affordances of Facebook, which enable activity and communication on the platform. He claims that Facebook provides an “empty structure” to its users. The structure relies on the fact that certain options are provided in terms of semiotics and interaction (profile pages, networks of friends, sharing of messages and photos, banner ads) which allow and promote a particular use and specific users, while making it difficult for others (or excluding them altogether)” (Brügger 3). Brügger therefore puts emphasis on the duality of Facebook as a social networking platform. Its structure enables communication, while it can completely alienate certain users.

As this research focuses also on Twitter, it is relevant to present Twitter as an important platform for online connections. Twitter was launched in 2006 “as an ambient friend-following and message utility” platform (Rogers in Weller et al 3). Rogers describes Twitter as being synonymous with microblogging, a genre which was considered new during the creation of the platform (Rogers in Weller et al 3). Twitter is no longer being used solely for microblogging, and further studies “explore how the tool is used for various purposes, such as politics […] cultural conversation […], and cultural performance” (Rathnayake and Suthers 2). Twitter users have had a tremendous impact on the evolution of the platform and they can be considered as very diverse, as Pushmann and Burgess claim “Twitter’s end users are private citizens, celebrities, journalists, businesses, and organisations; in other words, they can be both individuals and collectives, with aims that are strategic, casual, or a dynamic combination of both” (Puschmann and Burgess in Weller et al 47). With various types of users, and genres in which they use the platform the communicative mechanisms present on Twitter have evolved. Bruns and Moe show that Twitter offers various ways of communicating on the platform, “from the comparatively intimate, one-on-one level of @replies through the narrowcast level of personal publics constituted by follower networks to the collective, diffused, many-to-many level of hashtags” (Bruns & Moe in Weller et al 27). Twitter therefore presents many ways to interact, some more personal than others, but the platform gives various tools to its users in order for them to create the communicating experience they wish on the platform. This also means that Twitter as a social media platform enables the sharing of information, and potentially the creation of political movements and collective identities.

(19)

2.4 The Yellow Vests and their theoretical links to collective identity, digital activism and social media platforms

The complex notions of digital activism and social movements have been explored. Their history has been presented, while their link to technology has enabled the understanding of activist communication online. The importance of the fusion between social movements and social media has been presented. Platforms’ vast power towards users, content and display of information proves why further analysis is necessary when observing movements on social media. Specific arguments towards the social networking platform Facebook were observed, leading to the search for further information. Twitter’s versatility was presented as a way to enhance communication between users on the platform. The various examples of movements and platforms presented highlight the importance of social media when creating a collective identity, which can “be understood as the set of operations by means of which social movements define their collective sense of self, who they are and what they stand for” (Treré & Gerbaudo 1). Collective identity is a key notion when discussing communication within political movements online. As the complexity of social media platforms have been mentioned, the following chapter will present the methodology of the research which will be conducted on social media platforms Facebook and Twitter.

(20)

3. Methodology: The research methods supporting the analysis

of the Yellow Vest movement on social media

After understanding the Yellow Vest movement in its political context, as well as observing numerous theories, the movement’s political inclinations still remain complex. In order to further explore the political movement, and more specifically the Yellow Vests through their social media usage, the methodology chapter will present the ways in which the research has been conducted.

3.1. Research questions and their link to a methodological approach

Examining the research questions is a key point in the definition of a clear methodology. The first question, in which ways do the Yellow Vests utilize social media platforms? Is a question which relies on the analysis of literature, more specifically on the concept of platforms, as well as on the data collected. With the data collected, specific content posted by the Yellow Vests will allow the investigation of the various ways in which they use social media platforms.

The second question, what does the use of social media say about the Yellow Vest movement? Is an interrogation which is strongly linked to the data collected and discussed. With the data analysed, some discussions will enable to question how social media usage shapes the movement.

The final research question relies on the distinctions between Facebook and Twitter. What are the differences between Facebook and Twitter, in the ways they are used to discuss the Yellow Vest movement? This final interrogation demonstrates that the data collected will also be used as part of a cross-platform analysis, where the content found on both social media platforms will be compared. The methodology used to collect, analyse and discuss data was qualitative and ethnographically oriented. I will further describe the notion of digital ethnography, as well as the precise manner in which the research was conducted on both platforms.

(21)

3.2 Understanding Digital Ethnography

In order to understand the specific methods used to conduct this research, the concept of digital ethnography is particularly useful. Pink et al describe it as “an approach to doing ethnography in a contemporary world. It invites researchers to consider how we live and research in a digital, material and sensory environment” (1). This digital world is constantly developing itself, especially on social networking sites where users are free to evolve through the posting of new content. Murthy depicts four ways in which social networking sites can be useful to ethnographers. He claims that “they are virtual ‘gatekeepers’ with chains of ‘friends’ who are potential research respondents” and they contain a vast amount of multimedia material even towards marginal social movements (Murthy 844). Moreover, he claims that “ethnographers can ‘invisibly’ observe the social interactions of page members, gleaning a previously unavailable type of ethnographic data” and with social interactions online comes the observation of relationships and most importantly how they are structured (Murthy 845). Caliandro claims that a researcher exploring online environments shouldn’t only focus on identifying an online community to research, but more so on “mapping the practices through which Internet users and digital devices structure social formations around a focal object” (Caliandro 551). Digital ethnography on social media platforms therefore presents certain unique characteristics, the contact between the researcher and the online users is particular: “In digital ethnography, we are often in mediated contact with participants rather than in direct presence” (Pink et al 3). This distant contact can make the research uncertain; the following texts will therefore present how data will be collected and analysed on Facebook and Twitter.

3.3 Research on Facebook

3.3.1 Data collection on Facebook

Facebook is one of the primary tools used throughout this research. In order to conduct qualitative analysis, it was crucial to focus on a side of Facebook where exchange between members of the Yellow Vest movement could be observed. One Facebook page titled “La France énervée” (Translated into “The angry France”) was chosen for analysis. After reading various news articles on the movement, certain Yellow Vest participants have been qualified as leaders, even though the movement emphasises the absence of hierarchy. Eric Drouet’s name is often heard when the movement is mentioned. He was part of the organization of the first protests and was invited to discuss the Yellow Vest’s demands in front of members of the

(22)

government. He communicates through his own Facebook page, “La France énervée”, and exchanges thoroughly with his audience. His specific ways of communicating are a key point of analysis when examining the use of social media by the Yellow Vests. This Facebook page is public, I therefore used the Digital Methods Initiative (DMI) tool Netvizz. It is a “data collection and extraction application that allows researchers to export data in standard file formats from different sections of the Facebook social networking service” (Rieder 1). The “page posts” module of the app was used to collect the 250 most popular posts on the “La France énervée” page, from December until April. The most popular posts are determined by adding together the number of likes, reactions, comments and shares. The top 50 posts were collected each month, from December 2018 until April 2019, leading to a total of 250 posts.

Two other Facebook groups were chosen to be part of this research: “Gilet Jaune” and “Gilets Jaunes”. These two Facebook groups are closed, meaning that a user has to request access to an administrator in order to join them. These two groups were chosen because of their high number of members and activity. Moreover, even though they present technical similarities, the content and discussion vary between the pages.

In order to join the closed group “Gilet Jaune”, two questions needed to be answered:

- “This group abides to rules of politeness, anti-racism, and no political-exploitation, do you understand that you will be sanctioned, or expelled if you do not respect these rules?”

- “What are your motivations in joining this group?”

I replied “I understand” to the first question, and “I want to observe how the movement functions” to the second question. I was approved into the group soon after. Moreover, this group filters its posts, meaning that an administrator of the page will need to approve a post before it appears in the group. The group therefore announces “opening hours” at the top of its page, when administrators will actively be filtering posts. In order to conduct this research, I therefore joined both closed groups, as well as liked the public Facebook page “La France énervée”.

Because of the qualitative nature of this research, I chose not to use any pieces of software in order to examine the closed Facebook groups. I manually observed the posts, comments, likes, reactions and documented them for analytic purposes. I did not collect a specific number

(23)

of posts, but more so observed and documented which posts were relevant to this research. The criteria for this documentation process will be presented.

3.3.2 Data Analysis on Facebook

Once access had been gained to the relevant pages and groups, the data was collected. For my own categorizing purposes, I chose to collect data from the public Facebook page into an excel file, using the data collected on Netvizz. For the closed Facebook groups, I chose to collect the data by taking screenshots of the posts and comments. The screenshots do not appear in the research, for ethical reasons which will be presented further on. It is important to mention that the content being collected throughout this research is in French. As a native speaker, I translated the posts as accurately as possible. After collecting data from Netvizz and screenshots from the closed groups, I proceeded with the process of coding. Saldaña defines a code as “a word or short phrase that symbolically assigns a summative, salient, essence-capturing, and/or evocative attribute for a portion of language-based or visual data” (Saldaña 3). The process of coding was therefore applied after the data had been collected. Posts from the public page as well as the closed groups were accompanied by their respective date and time of post, furthermore, I coded the posts by attaching specific keywords to them. This enabled the gathering of similar themes when analysing the data, as certain posts presented comparable topics and behaviours. As the research on Facebook focuses on specific posts, only one code was applied per post, referring to the dominant sentiment of the post collected.

The various codes were established by looking through the closed Facebook pages, as well as the data collected on Netvizz. By looking at a small sample of posts, a few coding keywords were established. They were later applied to specific posts when looking at the entirety of the data collected. Various coding keywords have been established, these include: Protest organization, Protest attendance, Police violence, Legal matters, Government criticism and Support. These tags were attached to the screenshot, which is originally in French, and later to their translated version in English in order to determine potential patterns or topics of discussion. “Coding is not a precise science; it’s primarily an interpretive act”, this signifies that coding keywords are subject to change as the analysis progresses (Saldaña 4).

I will further describe the coding keywords, and which types of posts are associated in which categories. “Protest organization” is a keyword which includes any type of Facebook post

(24)

event, or specific instructions on where to meet on a specific date. An example demonstrates the protest organization category, the administrator of the “La France enervée” page wrote on the 8th_{of March: “Tomorrow’s programme: Charles de Gaulle airport from midday to 1pm.} Afterwards, Paris Champs-Elysées for the rest of the day”.

Another category named “protest attendance” describes content showing whether a user has attended a protest. This does not include shares of another user, or news outlets. An example which highlights this category would be a Facebook live led by the administrator of the “La France enervée” page, on the 9th_{of April, he was hosting a live-stream titled “In front of the} senate for the beginning of the protest”.

“Police violence” is a category which includes various types of posts, whether they are created by a user or shared from a news page. The key point of this category is that the posts have to show some form of police violence towards the Yellow Vests during a protest.

“Legal matters” are part of a category which essentially assembles content that deals with legal issues. It includes content such as a user asking for legal help, or posts regarding trials, court dates or fines. An example would be a post by a user in the group “Gilets Jaunes” on the 29th_{March. The caption claims: “Attention! I just received a summoning at the police station} because of content I posted on Facebook, be careful about the posts and comments you write” while attaching a photo of the official police document.

“Support” is a category which applies to specific posts where support is either offered to someone, or when a user asks for support from others. This does not refer to legal or medical support, but more so to emotional encouragement.

Finally, the category “government criticism” includes user-created content, or shared content which criticises the government. This clearly includes content showing disapproval towards the president, government members, as well as laws and policy changes. Criticism of specific candidates or parties for the European elections are also included in this category, an example would be a post from the page “Gilet Jaune”, where a user claims “Macron never stopped being a banker, being elected president has just enabled him to obtain more power in order to help his financial friends organize the reign of money”, this was a post created by a user on the 23rd_of April. Further categories will be described and analysed when presenting the findings of the Facebook research.

(25)

3.4 Research on Twitter

3.4.1 Data Collection on Twitter

As one of the research questions mentions, a part of this research aims to compare and analyse the differences between the use of Twitter and Facebook regarding the Yellow Vest movement. This cross-platform analysis influences the methodology, as similar ways of collecting and analysing data were necessary, in order to proceed to a comparison. For the purpose of data collection on Twitter, the Twitter Capture and Analysis Toolset (TCAT) was used. TCAT is “a toolset for capturing and analysing Twitter data” (Borra and Rieder 1). A data set titled “Yellow Vests”, collected on the Digital Methods Initiative’s servers at the University of Amsterdam, contains millions of tweets collected from December 2018 to today. Using the TCAT tool, and more specifically the “User activity (tweet frequency)” module, I collected the 250 most popular usernames. This module presents a list of the usernames which tweet the most, and therefore enabled me to find the most active users. I collected the top 50 users per month, from December 2018 until April 2019. This collection represents the 50 users who tweeted the most through the course of this determined period.

3.4.2 Data analysis on Twitter

After collecting the usernames of the most active Twitter users under the query “Yellow Vest”, their Twitter profile was examined similarly to the Facebook analysis. The profiles were observed and scrolled through, looking at the content of the tweets and retweets, as well as the biography, profile photo and geo-location. Using this observation, different coding keywords were applied to the Twitter profiles. As opposed to the Facebook research when one post was categorized, the data collected on Twitter referred to an entire profile. I choose to apply multiple codes to one profile, as a user can discuss various subjects on their entire Twitter profile.

The tags were constantly evolving, but some of them existed as: Police violence, government criticism, legal matters, protest organization, protest attendance, geo-location and importance of original tweets. Location is an interesting factor in analysing the tweets, as one can determine whether the most active Twitter accounts are in France or abroad. The gathering of locations was done by using the geo-location indicated on a user’s biography. As most users did not disclose this information, it was not a key point of the analysis. This Twitter analysis uses TCAT to filter through the vast number of tweets, in order to determine the most active users, and therefore proceed to a qualitative analysis of their Twitter profile. Twitter is a crucial part

(26)

of this analysis because of what it enables its users to do, as Caliandro claims: “On the one hand, Twitter provides people with a new means of interaction and materializes new forms of sociality. On the other hand, Twitter provides researchers with the very tools for measuring those new forms of interaction and sociality” (Caliandro 558).

Some of the categories mentioned are similar to the ones used on Facebook; they include similar content which can be found across both platforms. Categories such as geolocation, and importance of original tweets are specific to Twitter. The latter category aimed to identify whether a Twitter user posts original tweets, or if they are only focused on retweeting content from other users. This will enable to analyse if users interact within themselves, or if they prefer to share content created by others.

3.5 Ethics

Ethical considerations are crucial when conducting research on social media users, even though there is no contact with individual users. As Pink et al claim: “In digital ethnography, we are often in mediated contact with participants rather than in direct presence” (Pink et al 3). Townsend and Wallace describe three key arguments regarding ethical considerations with social media research, these rely on private versus public, anonymity and risk of harm. “One of the biggest areas of concern with social media data is the extent to whether such data should be considered public or private data”; a part of this argument relies on the platforms terms and conditions, and the fact that every user has agreed to them upon entering the platform (Townsend and Wallace 5). The terms and conditions of these platforms make the data analysed public, but it does not make it ethical. This is why the concept of anonymity is particularly important when presenting the data: “with traditional forms of research, it is generally straightforward to anonymise data so that research participants cannot be identified”, this will be the case during this research (Townsend and Wallace 7). Moreover, screenshots were used for coding purposes, but did not appear in the final research. Tweets and Facebook posts were quoted, making sure that no information identifying the user could be found. Finally, the risk of harm appears “where a social media user’s privacy and anonymity have been breached”, importance was given to make sure the data is secure, and the users identity remains unknown. This research fits into the requirements established by the Netherlands code of conduct for research integrity, which relies on five main principles: honestly, scrupulousness, transparency, independence and responsibility (Netherlands Code of Conduct for Research Integrity 13). As it was previously mentioned, a set of questions had to be answered in order to join one of the

(27)

closed Facebook groups, these questions were answered truthfully, but there are still ethical considerations when joining these closed groups. In order to abide to the ethical principles mentioned above, all data was anonymised, and most importantly I did not interact, or interfere with any of the users or posts. This means that I was not in touch with any users, or react, like and comment on any of the posts.

3.6 Limitations

The methodology described above presents certain limitations, the first being that the TCAT data set is dated from December 2018 to April 2019, whereas the Facebook pages presented were created around October 2018. This creates a gap of two months between the data analysed on both platforms. Because of the cross-platform analysis side of the research, I chose to only look at data on both platforms starting in December. This means that two months won’t be included in the research in order to make the time frame equal. Unfortunately, as the protests began in November, this omits the first month of the demonstrations. Social media use might have been different during this month; therefore, my results speak for the more advanced stage of the protests.

Another limitation relies on issues which arose during the data collection. On Twitter, a few accounts which were placed in the most popular accounts were bots, meaning that they only retweet content and do not interact with other users. There were only a small number of bots, fourteen in total, and they didn’t particularly hinder the results. The data collected on Twitter also showed that some accounts had been suspended or deleted. The accounts might have breached Twitter’s terms and conditions or have been deleted by the user. This represents a total of 42 accounts and even though it is a significant number, it did not affect the data analysis process.

3.7 A qualitative analysis of 500 Facebook posts and Twitter profiles

The methodology leading to the collection and analysis of data presents various aspects. On Facebook, Netvizz was used to collect data on a public Facebook page. Two other closed groups were analysed manually, collecting posts which were relevant to the research and coding categories. On Twitter, TCAT was used to analyse the profiles of the users which tweet the most. Specific code words were attached to their profile during the data analysis. Through both

(28)

comparison of users, as well as content. Overall, this methodology permitted a qualitative analysis of an important number of tweets and Facebook posts, as over 500 profiles and posts were analysed. Even though pieces of software were used to collect data, the two closed Facebook groups used a different approach, as data was collected manually, starting in December but without following a monthly time frame. The following chapter aims to present the findings obtained through this methodology.

(29)

4. Findings: A presentation of the results of a cross-platform

research on social media

This chapter aims to present the findings from the conducted research. As described in the methodology, the findings will come from four different sources. The first and second being two closed Facebook groups named “Gilet Jaune” and “Gilets Jaunes” as well as Eric Drouet’s Facebook page “La France énervée”. Finally, the Twitter analysis will rely on the 250 most popular accounts regarding the Yellow Vest movement.

4.1 Facebook Findings: exposing the research results of various Facebook groups

This section will present the data collected and analysed on Facebook, through the closed groups and public Facebook page. The findings will be organized thematically for the two closed groups, as no precise amount of posts were collected, yet they were organised using coding categories. The findings of the public page will be presented chronologically, as 250 posts were analysed, it is valuable to see the evolution of the content through the time period analysed.

4.1.1 “Gilet Jaune”

The closed Facebook group “Gilet Jaune” is one of the most popular, as it contains over 160 000 members. The description of the group claims “Yellow Vests at the service of other Yellow Vests! Information, demands, propositions, new ideas. Together for an improvement of our living conditions! Strength and honour.” This description shows the beliefs of the administrators, but ultimately the members are the ones posting content and determining in which direction the group will evolve. For this analysis, a variety of posts were selected, between the months of December 2018 and April 2019. These posts were categorised using the codes described in the methodology. The posts and categories will be presented in order of importance. The coding category which appeared the most was the one revolving around government criticism, the posts belonging to this category varied in the ways in which they shared this message. A few posts used text to describe the incompetency of the government: one of the members created a post urging others not to vote for President Macron’s party during the European elections, they claimed “With the elections coming up, do not let LREM (La République en Marche) fool you! They will only cause more precarity to those like us who are already struggling”. This user shares a sense of community as they use the word “us”, moreover

(30)

the comments showed that other members shared similar feelings. However, as a few members commented that they were voting for the “Rassemblement National”, Marine Le Pen’s right-wing party, tensions and debates started groright-wing in the comment section. Almost all the posts which were related to government criticism experienced some form of tension in the comments. Even though members seem to agree with the original posts, often critiquing President Macron himself, they start arguing once members mention the political parties they support.

Another important coding category is the one titled “police violence”. This category regroups posts which mention police violence and abuse of power. Most of the posts related to police brutality were videos and photos of police forces abusing Yellow Vests during protests. Some of the images were graphic, yet still were approved by the group’s administrators. The comments under these posts are almost all expressing negative sentiment towards police forces. One post was related to police brutality but did not show content of police forces being violent. This post was from a street medic, who wanted to share a message after a protest turned violent in Toulouse; they claimed:

As a street medic in Toulouse, I would like to thank some of you for what you keep doing for us during protests: creating bubbles of protection while we care for someone injured, charging towards the police so we can evacuate someone wounded, chanting and singing for us. Thanks for what you do, and don’t give up!

Street medics are inherently tied to police brutality, as they are often the first respondents to injured protesters, this post was placed in the police brutality category.

Three other posts were considered to be part of the “protest attendance category”, one of them being a video, and two being photos. These posts simply aimed to show the number of attendants during a protest. The video posted by a user had the caption “Many Yellow Vests heading towards Bastille”. Furthermore, three other posts were placed in the “event” category, meaning that there were Facebook events shared in the group, enabling users to click the buttons “interested”, or “participate” in order to join. One of them was published on the 23rd_{of March} and was titled “Acte 35, 14th_{of July 2019, Yellow Vest parade at the Champs-Élysées”. This} post was therefore created in advance of an event planned in July.

The creation of these Facebook events leads to the notion of protest organization. A few posts were placed in this category, as they clearly showed members were aiming to facilitate the planning of protests. One member posted an image representing six French institutions, including the Senate, the Élysée palace and the National Assembly. On top of these images the

#GiletsJaunes: Exploring the Yellow Vest movement on Social Media.

Master Thesis