Issue Mapping for Brazilian Politics
by
Daniela Demarchi
Supervisor: drs. E.J.T. Weltevrede Second Reader: dhr. dr. M.D. Tuters
MA New Media & Digital Culture
Graduate School of Humanities | University of Amsterdam
24 June, 2016 Academic Year 2015/2016
Dedication
To my mom,
The most good hearted, brave and honest person anyone will ever meet. Hi, mom! *waves* I’ve finished it! Can I go play outside now?
To my dad,
In a Batman vs. Dad fight, we would have bat meat on Sunday’s family barbecue. Did I mention he is Argentinian?
My Dad is my hero and he has the best sense of humour.
Actually, Dad doesn’t know a word in English, so if nobody translates this for him, he’ll never know.
To my sis,
The Arnold Schwarzenegger of my Deni De Vito.
She always tells me when I am being an idiot. Which is quite often. I would trust her with my life.
To Balu, my best friend.
An honourable soldier with whom I would fight beside in the unholiest of wars. He also has smelly feet.
Abstract
This research paper aims to investigate the presence of extremist discourses resulting from the emergence of a new right-wing movement in Brazil. It shows that the type of content shared by Facebook pages played a vital role in steering polarised reactions from users. This argumentation is developed through a detailed case study, which contemplated pages involved in the organisation of the movement as well as news media pages. The analysis focuses on the period from March 1st to March 15th, the two weeks preceding and two days after the largest ring-wing protest in the country on March 13th, 2016. It draws from a total of six pages, 300 posts and 600 user comments, extracted via the web application Netvizz. For each page, the top 50 most engaged with posts, as well as the top 10 most engaged with comments per post, have been translated and categorised, making it possible to examine the dynamics in which the right-wing movement is organised.
The research also presents a new approach to detect extremist ideologies on social media platforms, looking in particular at Facebook. In order to do so, the paper defines extremist discourse, looks at Facebook as a medium for political debate and describes a mechanism for categorising and organising textual data. The main research objective, along with mapping a political debate, is to contribute with a new approach to examine online extremism on social media platforms, and therefore open up the possibility for methodological discussion within the field of political communication.
Key Words
Brazilian politics | Online extremism | Issue mapping | Content analysis | Comment analysis | Research methods
Table of Contents
1 Introduction ... 6
2 Researching online extremism... 9
2.1 History of the term extremism ... 12
2.2 Forms of extremism and democratic societies ... 13
2.3 Operationalising extremism ... 15
3 Facebook, politics and extremism ... 19
3.1 Facebook’s technicity ... 19
3.2 Facebook’s flagging mechanism... 21
3.3 The Echo Chamber and polarized discourses ... 23
4 Actor-Network-Theory ... 25
4.1 How to map the social ... 25
5 Method ... 29
5.1 Page Selection ... 29
5.2 Tools ... 32
5.2.1 Netvizz ... 32
5.2.2 INFOGR.AM and Sankeymatic ... 32
5.3 Exploratory Approach ... 33
5.3.1 Collective Action Pages ... 33
5.3.2 Connective Action Pages ... 34
5.4 Emergent Coding ... 35
6 Case Study I – Collective Action Pages ...40
6.1 Estadao Newspaper ...40
6.2 Folha Newspaper ... 44
6.3 Veja Magazine ... 48
7 Case Study II - Connective Action Pages ... 55
7.1 Free Movement Brazil Page ... 55
7.1.1 Facebook Page profile ... 55
7.1.2 Page Posts ... 57
7.1.3 Posts Comments ...60
7.2 Take the Street ... 63
7.2.1 Facebook Page Profile ... 63
7.2.2 Page Posts ... 65
7.2.3 Posts Comments ... 69
7.3 Online Revolted ... 73
7.3.1 Facebook Page Profile ... 73
7.3.2 Page Posts ... 75
7.3.3 Post Comments ... 79
7.4 Findings Connective Action Pages ... 82
8 Discussion ... 86
9 Conclusions ... 89
9.1 Opportunities and Limitations of the Research Method ...90
9.2 Suggestions for further research ... 91
6
1 Introduction
Brazil is experiencing the height of its gravest political crisis since the return to democracy in 1985, after two decades of military dictatorship rule. The climate of tension began making strides in 2013, in what became known as "June Journeys" (Jornadas de Junho): a popular movement originally born to oppose the increase in public transport but which, with the support of social media platforms1, ended up with thousands of Brazilians taking the streets to demand changes in the political landscape. A social movement of this nature was considered unprecedented and is often viewed as the “awakening2” of the Brazilian people. Nevertheless, with a political maturity still in embryonic stage and in the absence of a concrete political agenda, the movement soon lost strength, but managed to leave behind the first seeds3 of a new extreme-right wave4 that is now besetting the country. Although rooted in the demonstrations of 2013, the new digital activism expanded and consolidated a year later, during a hard-fought Presidential campaign between the Brazilian Social Democracy Party and the Worker’s Party, both strong symbols of right and left-wing ideologies in Brazil5. In an unexpected outcome, the left triumphed for the fourth time consecutively - after Worker’s Party leader Lula da Silva first became president in 2003 – and re-elected President Dilma Rousseff with 51% of the votes. Since then, the discontent with the election results, combined with a faltering economy and the blatant support of the traditional media6, further contributed to the rise of a new right with increasingly radical frames.
This polarisation process was reflected on the streets. On March 13th, in one of the biggest demonstrations in Brazil’s history7, right wing protesters took over the streets in favour of President Dilma Rousseff’s impeachment. The left’s response a week later, against the president’s deposition, only demonstrated a responsiveness that intensified the degree of social polarisation. Regardless of the legitimacy of the
1 Sakamoto, Leonardo. “And, in Sao Paulo, Facebook and Twitter went to the streets. Literally”. Folha UOL. 2013 2 Hanati, Yuri. “What is left from the June Journeys”. Gazeta do Povo. 2013
3 Monteiro, Mauricio. “And the ‘Giant’ moves forward”. Rolling Stones. 2013
4 Weissheimer, Marco. “There is no "movement" in dispute, but a crowd hijacked by fascists”. Carta Maior. 2013 5 Ramos, Mailson. “The Social Media factor in the 2014 elections”. Pragmatismo Politico. 2014
6 Zanin, Cesar. “The press and the role of media in Brazil”. Pragmatismo Politico. 2015.
7 “The largest march in the history of the country increases pressure in favour of Dilma’s deposition”.Estadao.
7
request, the March 13th act was characterised for reuniting actors with strong undemocratic agendas; between cries against corruption, it was possible to see signs asking for military intervention, hate speeches directed to the Worker’s Party and physical attacks on bystanders wearing red t-shirts8. Self-proclaimed a non-partisan movement, protesters also expelled right wing politicians - including former opposition presidential candidate, Aécio Neves – while, at the same time, allowing far-right politicians to take the stage9. Furthermore, there was also general praise for Judge Moro - acclaimed by the right as the nation’s hero in the fight against corruption, despite accusations of him acting arbitrarily in the “Operation Car Wash10” (Operação Lava Jato).
While left-leaning newspapers committed themselves in condemning protesters actions, accusing the demonstration of being fascist in nature11; right wing papers treated the events as isolated cases, detached from the reality of the movement12. In light of this controversy, the objective of this study is to engage in a critical assessment of right wing discourses on Social Media platforms in order to understand their relationship with extremist ideologies. For this end, the research is interested in analysing the frames in which the movement was organised, as well as the way users’ utilise digital platforms to convey their political views. By pursuing a more contextualised approach to political discourse, this study intends to capture the dynamics in which the rise of this “new right” is manifested. Thus, the research will revolve around the following questions:
To what extent are extremist discourses part of the new right wing movement in Brazil? What are their triggers and how they are organised?
In order to answer these questions, this study will rely on Social Media platforms, specifically Facebook, as a space for where political debate is articulated. This is because the so called awakening of the Brazilian people has everything to do
8 Araújo, Glauco et al. “Teenager is assaulted in protest against the government”. 2016 9 Rovai, Renato. “Moro and Bolsonaro, the big winners of today’s protest”. 2016
10 Operation Car Wash is currently investigating a corruption scheme connected to Petrobras, the most
important State company in the country. It is the largest corruption and money laundering investigation Brazil has ever had. More information in the Public Prosecution Service’s website: http://lavajato.mpf.mp.br/entenda-o-caso
11 Nogueira, Kiko. “Fascist salute in the protest: who will put the ghost back in the closet”.DCM. 2016 12 Carvalho, Marco Antonio. “Analysis: the protests of March and the Brazilian political crises”. Estadao. 2016
8
with its growing familiarity with the internet. According to the Brazilian Research Media of 201513, the internet is, by far, the fastest spreading medium in the country. Half of the Brazilian population have access to the internet, spending an average of five hours a day connected. Furthermore, the appetite for social change has roots on digital platforms and Brazilians adopted these spaces as a sort of national agora, a gathering place where they are able to share their interests, spread information and construct their political identity. Thus, this research will operate from the assumption that the web today represents one crucial arena of political and social action.
Brazilian extremism is not easy to identify. Unlike the German Nazism, the Italian fascism or the new American authoritarianism of Donald Trump, which have distinct characteristics such as xenophobia and ultra-nationalism, the Brazilian extremism is more fragmented, formed by a vast and diffuse mass. It is endorsed by a media that acts in self-interest, a religious movement with authority in Congress, and an elite which claims to operate in defense of liberalism, but in practice acts to maintain their inherited privileges. Thus, in order to investigate the idiosyncrasies of the ‘tropical14’ extreme right, the study will be divided into sections that seek not only to understand the movement, but also to identify their agenda. Furthermore, this research aims to provide different forms of analysing online extremism, usually limited to studies on rhetoric strategies of radical organizations or religious extremism. Instead, this research will focus on the rise of a movement created and organized by citizens, which relies on social media as their main platform.
13 Secretary of Social Communication of the Republic. “Brazilian Research Media of 2015”. 14 Reference to the Brazilian climate.
9
2 Researching online extremism
“Political extremism involves two prime ingredients: an excessively simple diagnosis of the world’s ills, and a conviction that there are identifiable villains back of it all”. (John W. Gardner)
The study of extremism began shortly after World War II, as political scientists and psychologists in the US and Europe tried to figure out how the Nazis had managed to win such wide public support for such extreme ideology (Taub, 2016). Around that time, theorist Theodor Adorno, developed what he called the “F-scale” (Adorno, 7), which sought to measure “fascist” tendencies, that is, anti-democratic tendencies at the personality level. However, the test was soon discarded under the pretence that people were capable of assessing the measurement of the F-scale, giving them the chance to react in a more “suitable” manner. In this fashion, the internet has provided people with the opportunity to openly deliberate their views of the world without the need to dilute their opinions with the same intensity. The anonymity of the internet allows individuals to disentangle themselves from their own words, and speak their minds freely. This becomes more evident in what Caiani and Parenti call, “the dark side of the web” – the emergence of far right activities on the internet – which showed that “distributing media contents and new methods of mass communication empower actors with explicit undemocratic agendas” (Ekman, 56). Some scholars argue that social media activity “cannot be a substantive form of ‘grass root activism’” (Conway, 4), that is, the kind that is exercised on the “real world”. They believe that online platforms only work for donations or propaganda purposes, and that this “virtual ‘venting’ satisfies the desire to act” (Conway, 4). Social theorist Maura Conway refutes these views, arguing that if a free platform can create fundraising then it has a financial aspect to it; if it can attract so-called ‘foreign fighters’, then it has a recruitment function; all this while it is also used for communication, logistics and propaganda. Henceforth, the use of internet for and by extremists should not be viewed as a secondary tool. Today’s digital platforms do not only work on the dissemination of extremist material, but on the social interaction surrounding it. It has the potential to formulate a collective identity, organise contacts and mobilise citizens. Moreover, studies on violent radicalisation have shown that isolated individuals find a common identity through extreme right platforms, convincing themselves that there are not alone, but instead part of a community, if only a virtual one (Caiani and Parenti, 274);
10
something that can deeply influence the way they connect to the real world. There is no doubt that the internet plays a significant role in “contemporary radicalisation processes” (Conway, 4), but up until now scarce scientific attention has been devoted to extreme right wing in the internet (Caiani and Parenti, 274). Research has predominately focused on the configurations of jihadism, terrorist recruitment processes and on the rhetoric of extreme organisations. Furthermore, the existing studies typically focus on North American or European cases of extremism.
In this study, focusing on Brazilian right wing debate environment and combining a content analysis with an engagement (comment) analysis, I aim to explore the structure and characteristics of the right, investigating to what extent extremist discourses are part of the everyday online debate and when they are most “activated”, that is, what are the triggers of extremist ideologies/behaviours. Studying the Brazilian right wing milieu is of particular interest since extremist ideologies seem to be expanding15 between the increasing anomie of international actors and the overall unbalance on the global stage. In this destabilised and tense scenario, extremism tracks its way and consolidates as a player in the global balance, with it being necessary to understand all its aspects and dimensions. On the query of determining whether radicalisation is occurring on the internet, Conway suggests that research should “extend to inquire the whole range of contemporary violent extremists and terrorists and their interactions” (6). This research looks to contribute to widening the field of extremism in order to “allow for comparative research” (Conway, 7), understand the similarities between all variants of extremism and come up with explanations for these, especially around online radicalisation strategies. Moreover, Conway also underlined that “there has been little attention to the individual users’ experience online and usage of the internet in the process of radicalisation” (Conway, 9). In this sense, the research looks to contribute on the development of new methods to study extremism on an individual-level, in order to gain a deeper and more contextualised understanding of the varieties of extremists interactions online, what their agendas are and what triggers this kind of behaviour.
Extremism can take many forms. It represents the polarisation of different actors in the heterogeneous dimensions in which politics is shaped. Unlike the type of
11
extremism portrayed by Hobsbawm in The Age of Extremes, the current political scenario is not defined by the dualism of the Cold War, or the tensions between two actors and their world views; but the existence of various concepts of reality which collide with one another. The line that defines what is democracy and autocracy is becoming increasingly blurred as is the political and ideological views of actors. Even the word extremism is now used as a synonym for terrorism, ignoring the original Platonic-Aristotelian political theory. Thus, since the aim of this study is to contextualise online political discourse in a more “subtle” form of extremism; in the sense that it will be less focused on violent radicalisation and more on polarised/authoritarian discourses; it is important, first, to give a short introduction on the political sociology of extremism, grasping the root meaning of the term, its overall structure inside democratic societies and establishing a few traits to assist the investigation of such social phenomena.
As mentioned, the constitution and discernment of extremism can take many forms: radicalisation, hate speech, hard-right conservatism, violent acts and so forth. In this view, before outlining the grounds of political extremism, it is important make a note on the meaning of the terms related to it and how this research will apply them. Sometimes, for example, concepts such as radicalism and terrorism become conflated. In this thesis, the term radicalism is used to refer to the “process of developing extremist ideologies and beliefs” (Borum, 8). When something is referred to as radical, it means it has an extremist view. Secondly, the term ‘hate speech’ will be used in reference to a disparagement of others, a hatred against groups (or the perception of a group in the body of an individual) characterised by the use of foul language or stereotypical/discriminatory words (Warner and Hirschberg, 19). Hard-right conservatism will not be considered extremist ideology since this concept does not necessarily rest in the notion of political opponents as “enemies” (Vega, 2009).
With the vocabulary surrounding extremism having been taken into account, what will follow is a description of the terminology history, a discussion around its structural characteristics and an outline into how the concept will be operationalised in this research.
12
2.1 History of the term extremism
The idea of the political extreme is mounted on writings of Plato which connected the ethics of moderation – the notion that in every action there is a halfway (mesotes) between the excess and the scanty - to constitutional doctrine (Backes, 242). He took government forms such as aristocracy and legal democracy and located them between despotia/tyrannis and anomia/lawless democracy (mob rule) which, in his view, were extremes. Later on, Aristotle framed Plato’s terms into a comprehensive and realistic political system. According to Aristotle, in a society conducted by the middle class, the interest of the upper and the lower classes should be balanced “by means of a composition of politically institutional elements from different constitutional forms” (Backes, 243). Aristotle proposed a “politeia”, a combination of democracy and oligarchy, as the best form of government, because it would simultaneously guarantee stability and the liberty of citizens (ibid). This Aristotelian idea of politics shaped the history of what we now call the constitutional state. Extremes were viewed as “aberrant human behavior” (ibid) which stood for depluralisation and the focus on violence, whereas the mean (mesotes) symbolised pluralism and organisation.
The Platonic-Aristotelian view of mesotes and mixed constitutional state offered the political physiology of post-French Revolution in which, on the new parliamentary model, the differentiation between extreme and reasonable found new connections with the terms “right” and “left” (Backes, 244). Extremism found its entry into political language in numerous ways, especially on times of political polarisation such as the Russian Revolution in 1917, in which the term permanently established itself in western states (Backes, 244). In 2011, the British government defined extremism as opposition to fundamental British values, including democracy, the rule of law, individual liberty and mutual respect and tolerance of different faiths and beliefs (Prevent Strategy, 107). In the case of the internet, the concept of online radicalisation is here conceived as a process whereby individuals, through their online interactions and exposure to various types of internet content, come to view “politics as warfare” (Vega, 2009), in which violence and intolerance are legitimate methods of solving social and political conflicts (Bermingham et all, 1).
Defining the history and meaning of a term is an essential part of establishing a debate. Especially in the case of controversial topics such as extremism, which its
13
many-faceted components opens room for interpretations, depending on the perspective from which it is approached. Thus, this clarification is the first step towards understanding the basis and structure of the field.
2.2 Forms of extremism and democratic societies
Another important step towards an appropriate examination of the subject is understanding how extremism is manifested in democratic societies. Etymologically, “extreme” or “extremism” determines something which is the “farthest out” - the opposing poles of a middle (Backes, 248). Evidently, the history of the terminology proves that the term possessed different meanings at different times and the controversies connected to a view of a “midpoint” have been frequently subjected to change. Nevertheless, taking from the Aristotelian tradition of politics it is possible to delineate an essence of what is to be defined as radical ideology. Politics should be about “the confluence of multiple social forces” (Backes, 243), where heterogeneous interests are institutionally coordinated and power does not suppress nor obstruct the autonomy of citizens. Extremism is, thus, the antithesis of the constitutional state.
Political theorist Carl J. Friedrich described the notion of modern constitutional state as a process of the merging of democracy and constitutionalism. Democracy in the sense of a state which recognises human equality and the people’s sovereignty and constitutional in the sense of plurality, ensuring and power- controlled institutional structure (Backes, 251). The constitutional state represents a legitimacy of all groups’ interests, which are taken into account and discussed in a “marketplace of political ideas” (Backes, 243). Contrastingly, extremist discourses disentangle a bipolar perspective of the world. It is an ideology based on the “belief that political opponents are literally ‘enemies’” (Vega, 2009) and “one’s own side is good and the side of the enemy is evil” (qtd. in Borum, 28). In this Manichaean, us-them paradigm, it is important to identify, differentiate and define the other: he does not look or think like me, therefore he does not belong. “Extremists have assume the position – existentially and practically – of society’s defenders as they define society” (Guiora, 32). In short, extremist speech pushes towards a demand for incontestable truths, prohibiting any form of criticism and “gives grounds for the impossibility of coexistence” (Backes, 249).
14
As discussed, the constitutional state, or in Rousseau terms, the “social contract” (Guiora, 21) is predicated under the understanding that respect for individual rights, pluralism and control of power is essential to a thriving civil democratic society. However, for the analytical differentiation of political discourses it is important to answer the following; in order for a speech to be classified as extreme, is the combination of anti-constitutionalism and anti-democracy a necessary requirement? Political scientist Uwe Backes tries to answer this question with a scheme (see Figure 1) that combined the two dimensions into four forms:
Figure 1 – (Anti-)democracy and (anti-)constitutionalism scheme. Source: Meaning and Forms of Political Extremism in Past and Present by Uwe Backes.
Form (1) stands for a movement which attends to the equality of citizenship, but rejects the authority of the social contract. According to Backes, a good example of this model is the communist and anarchists doctrines, which preach for human equality at the same time as they move towards a “power controlling design of the constitutional state”. (Backes, 252). Form (2) applies to the original Aristotelian idea of politics, which is a “constitutional state in the basis of slavery” (ibid). This model can also be found in the constitutional basis of the Apartheid in South Africa and in the North American republic on its founding days (ibid). Form (3) corresponds to the dictatorship regimes that took place in a number of South American countries during the mid-1900’s: negation of individual’s sovereignty connected to a totalitarian state. The differentiation of the two dimensions is vitally important for this analysis, because by classifying only one of those dimensions as radical would contradict the
15
current understanding of what is to be a democratic constitutional state. Thus, a definition of extremism should be “drawn up in such a way that the negation of at least one dimension is required” (Backes, 253). What is more, in order to create a better understanding of the application of this concept in the study, the dimensions of anti-democratism and anti-constitutionalism can be further subdivided into any domain that contain speeches such as: anti-pluralism (“anti-party affect and interest-group prudishness” (Backes, 253)), hate speech (or expressions which convey hateful intent), illegitimacy of political opposition, anti-egalitarian in respect of individual liberties (e.g. advocacy for military intervention in politics), negation of power-balancing institutional structure (e.g. independent judiciary, legitimacy of elected government). A further explanation of the traits of extremism will be addressed in the next section.
2.3 Operationalising extremism
Detecting extremist discourses is not an easy task. There are numerous issues which need to be resolved in order to develop a consistent language model. First, a mere mention of an organisation associated with hate speech or the application of stigma words by itself does not necessarily constitute extremism. The name Al-Qaeda alone does not express extremism, since it appears in history articles, legal documents and several other contexts of communication (Vega, 2009). One should also be careful not to confuse criticism of the status quo and enmity towards the system with extreme tendencies (Backes, 258). In the same fashion, it would be inappropriate to only view the use or incitement of violence as a defining characteristic of radicalism. Warner and Hirschberg research on “Detecting hate speech in the world wide web” showed that data contains all sorts of examples of offensive language that is sometimes hateful and sometimes not, leading to the hypothesis adopted in this study which states that “hate speech resembles a word sense disambiguation task, since, a single word may appear quite frequently in hate and non- hate texts” (Warner and Hirschberg, 19). With this approach it will be possible to identify the contextual polarity of the phrase in which a word may express a different intention from the word’s prior polarity (Wilson et al, 347). Nevertheless, the previous explanation of the constitution of extremism in conjunction with a useful set of starting guidelines, will allow for a more meaningful progress studying online political discourses.
16
In order to examine how extremism is manifested on the internet, I will address some of the key social and political characteristics of extremist tendencies. As Tehmina Basit points out, raw data can be very interesting to look at, yet it does not help the reader to understand the social world under scrutiny, and the way the participants view it, unless such data have been systematically analysed to illuminate an existent situation (144). Along these lines, Miles and Huberman suggested two methods for creating codes, one of which – and preferred by the theorists – creates a provisional “start list” to fieldwork. This list has no intention in framing the data into pre-established categories. The job of defining the actors will rely on the actors themselves. Instead, the list sets a conceptual framework, a list of “hypothesis, problems, areas and/or key variables that the researcher brings to the study”. (Basit, 145). Undoubtedly, the extreme right harbours ideological positions, cultural signifiers which are manifested in different versions amongst societal groups (Ekman, 83). At the same time, the analysis of qualitative data is a continuous job and not a “separate self- contained phase” (Basit, 144). This means that although an understanding of the field is an important part of the outcome, it is only the first step of a much deeper analysis. Thus, in order to contextualise the data, I will outline an analytic approach to it, identifying general elements that will be useful in finding out to what extent extremist discourses are part of the political debate. In analysing the rhetoric of several social groups across the political spectrum, political researcher Laird Wilcox distinguished a number of specific “traits or behaviors that tend to represent the extremist ‘style’” (Wilcox, 110). He analysed more than 320 staged hate crimes in the US since 1994 and found characteristics that are particular to fundamentally anti-democratic and anti-constitutional ideals. Henceforth, what follows is a conflated list of extremists traits this research will look for on the quest to answer the inquiry:
a. Demonisation of the other: “Extremists have a tendency of viewing the world in terms of absolutes of good and evil, for them or against them, with no middle ground or intermediate positions” (Wilcox, 109). There is an orientation toward an egotistical execution of interests where a conflicting or contradictory approach to a matter of concern is illegitimate and discarded under the claim that it is dishonest, immoral, unscrupulous or hateful. Issues tend to have a moral stand and because of that “extremists often attack the character of an opponent instead of the issues raised” (Wilcox, 108). They are inclined to personalise hostility, directing their hatred towards individuals themselves
17
instead of ideas; addressing them as enemies and questioning their motives, personality, looks or past associations as a way to divert from the issues under consideration (Wilcox, 108). Essentially, this trait of extremism arises from the application of broad statements, sloppy use of analogies and harsh judgments based on little or no evidence (Wilcox, 108). Furthermore, this type of dualism gives room to an apocalyptic view of the world and the belief in the inevitability of a “Final Battle that will sweep away unbelievers and usher in an era of world-wide unity under the truth of one view” (Lurie, 2011). This characteristic is not solely related to religious extremism (i.e. jihadism). On the contrary, it seems that most radicals tend to embody some form of “mythical components” (Ekman, 84) to their political views: extremists often deify leaders or public figures to the conviction that they have been chosen to save the people from the evilness of the opposition.
b. Naming strategy: the application of negative terms to serve as labels for political opinions. The “naming strategy” (Backes, 246), in political argumentation, constitutes a mean of employing a derogative characterisation of the opponent and make use of “stereotypes to express an ideology of hate” (Warner and Hirschberg, 19). It is a way for the author of the message to invoke well-known, stigma words, used to “mark the boundaries of political legitimacy, to judge other unworthy and to designate dangers” (Backes, 246). The circulation of hate speech rapidly resorts to epithets, to designate and criticise opponents in order to divert attention away from their arguments and discourage others from hearing what they have to say (Wilcox, 108). Extremist discourse use this trait to reinforce their enmities “in a manner that bolsters their confidence and sense of self- righteousness” (Wilcox, 110)
c. Advocacy for censorship/violence: the goal of this extremist trait is to have some kind of control over what is being said, done or even thought by their opponents or critics. “Extremists would prefer that you listen only to them. They feel threatened when someone talks back or challenges their views” (Wilcox, 109). Inside the “politics as a warfare” (Vega, 2009) belief system to which extremists tend to attach themselves, there is a pursuit for “absolute security in an environment that is naturally unpredictable or perhaps populated by people with interests opposed to their own” (ibid). Thus, in order to achieve their view of ideal society, extremists engage in censorship, repression and/or
18
violence with the objective of defeating the enemy and guaranteeing the permanence of their values. According to Backes, the experiences of totalitarian regimes in the 20th century showed how this kind of idealistic perspective of politics gives grounds to totalitarian practices, “since the utopia is able to deliver a foundation for the justification of a rigorous transformation, ‘Gleichschaltung’ (forcing into line) and ‘cleaning’ of society” (Backes, 256). d. Groupthink: the internet has a strong capacity of generating collective identities (Caiani and Parenti, 274), that is, a cohesiveness and a commitment to a movement or group. In Victims of Groupthink, Irving Janis, explored the dynamics of group formation and highlighted elements applicable to the study of social phenomena. Janis suggests that groupthink involves a tendency to conform to group norms, preserve solidarity and claim unity. Moreover, there is a “concurrence at the expense of distorting member’ observation of facts, conflicting evidence, and disquieting observation that would call into question the shard assumptions and beliefs of the group” (Wilcox, 110). Extremists have a strong belief in the morality of their group, combined with a decidedly evil picture of the group’s opponents (Hart, 247).
19
3 Facebook, politics and extremism
Social network sites have revolutionised the way people communicate and share information. Platforms such as Facebook have quickly grown in popularity; celebrated for providing people with the ability to express themselves, disseminate knowledge, engage in political debate and participate in the overall democratic “commons of the cyberspace” (Gillespie, 351). Its rapid spread among users has also attracted researchers from a wide range of disciplines interested in studying these platforms and the social phenomena they are entangled with (Rieder et al, 1). Thus, in order to align this research with existing studies and, at the same time, discuss the role social platforms play in political debate, this section will set its argument in three steps: (1) first it will examine the limitations and opportunities of data-driven Facebook research; (2) outline Facebook’s guidelines on the type of content it allows or bans on its service; (3) consider claims made by many authors that personalisation algorithms, like Facebook’s News Feed, act like echo chambers, which not only coarsen online debate, but also leads to the polarisation of opinion across communities.
3.1 Facebook’s technicity
When analysing social phenomena such as online extremism, a researcher cannot ignore the technicity and policies of the host of the data. The technicity of a digital platform is the technology analysed through its “operative functioning” (Rieder et al, 2). Fundamentally, an analysis based on the forms, features and overall infrastructure adopted by a platform. This will be important, not only to understand Facebook’s role as an actor on the making of the social phenomena, but also to work as a valuable resource for developing “analytical strategies and interpreting the findings” (ibid). Since its creation in 2004, Facebook rapidly grew to become one of the most important communication services in the world. As of 2016, Alexa, a company specialized in tracking web traffic, ranked the site as the third most visited both in the world and in Brazil16. Facebook is used for a variety of practices: to stay connected with friends and family, upload personal pictures, share news about the world, and much more. These practices are enabled and structured by the so called
20
grammas of action (ibid). The grammas of action are a set of features a platform grant users so they can interact and produce content.
From a methodological perspective these built-in structures offer both opportunities and limitations for empirical research. The 0pportunities are due to two main reasons: firstly, the possibility to collect and analyse large amounts of human data is unprecedented and very appealing to any social researcher; secondly, engagement measures work as “markers of user’s intentional behavior” (Rieder, 4), allowing the researcher to systematically interpret particular aspects of the social. On the other hand, grammas of action also has its limitations in the sense that the interfaces and functionalities of the platform establish a controlled environment for users to act in (Rieder et al, 3). This means that users are subjected to “artificial conditions” (ibid) of behaviour and, thus, the data extracted will be formatted according to the field previously specified by the platform. It is true that in February of this year, Facebook redesigned the “Like” button, giving users the possibility to “React17” to content by accessing five additional emotive icons available: “love”, “haha”, “wow”, “sad” and “angry”. The “Reaction” button was not available for analysis on the time of this research. Regardless, the new possibilities of interaction are far from representing the multiple nuances of human behavior. Once inside the platform, users are forced to channel and simplify their emotions using features Facebook established in order to optimise engagement, “making it measurable and profitable” (Rierder et al, 3). Taking this under consideration, this research will mainly focus on investigating textual data, such as posts and comments. Deconstructing discourses rather than simply measuring engagement through clicks will provide both a qualitative and quantitative approach to the following data. Qualitative in the sense of understanding the frames in which these discourses are developed, and quantitative in relation to the large amounts of empirical data available for collection. This approach will not only help reconcile the limitations previously mentioned, but also provide a deeper perspective on understanding and measuring extremist discourses.
21
3.2 Facebook’s flagging mechanism
Since this research aims to identify extremist content on Facebook, it is important to understand what the platform’s policies on the issue are. When it comes to abusive/offensive speech, most sites that host user-generated content, rely on a system known as Flagging - an “ubiquitous mechanism of governance” (Crawford and Gillespie, 411), where users and platforms are responsible for surveilling, reporting and regulating content that infringe a predetermined set of community guidelines. According to the “Community Standards Guidelines18”, Facebook will “remove content, disable accounts, and work with law enforcement when we believe there is a genuine risk of physical harms or direct threats to public safety”. Nudity, hate speech, self-injury, attacks on public figures, dangerous organisations, criminal activity, violence, bullying and graphic content are among the areas covered by the guidelines. Facebook’s definition of hate speech covers content that explicitly attacks individuals based on their race, ethnicity, nationality, sexual orientation, gender or disabilities. Facebook’s flagging system is intricate, partly because of the plethora of content it handles daily, and partly because if its history of taken a more interventionist approach to moderation (Crawford and Gillespie, 416). Currently, users have the option to monitor the flags they have reported. The “support inbox” (See figure 2) shows a list of the users’ complaints, along with the current status of the review in process and a description of why an action was or was not taken.
22 Figure 2 - Facebook support dashboard. Source: Facebook
Flagging plays an important role in the “techno-cultural construct of social media platforms” (Crawford and Gillespie, 411) because they allow users to participate in how content is organised, ranked, valued and presented to others. Nevertheless, it is also important to note that while this mechanism does represent a “willingness to listen to users” (Crawford and Gillespie, 413) it also obscures the fact that the final decision is made by the platform. The flag is only a superficial step among the multiple variables of content regulation processing and end up being subjected to interpretation and judgment of the host. On a blog post about the Community Standards, Monika Bickert, Facebook’s Head of Global Policy Management, and Chris Sonderby, Deputy General Counsel; admitted that the hate speech policy remains “particularly challenging...we know that our policies won’t perfectly address every piece of content, especially where we have limited context, but we evaluate reported content seriously and do our best to get it right”. Evidently, disagreements about what is offensive or unacceptable are inevitable when a diverse audience encounters shared cultural objects (Crawford and Gillespie, 413). But the fact is that as “keepers of public discourse” (Crawford and Gillespie, 413), social media platforms should focus more towards becoming mere facilitators of social interaction and less towards active (and fallible) curators of information. Achieving this would mean to be more open to “public negotiations about the contours of public discourse itself”, allowing users to deliberate on how to act in regards to content deemed harmful, offensive or controversial.
23
It seems Facebook’s policies play an insignificant role on filtering the type of content examined in this research. Firstly, because Facebook’s community guidelines are based on international standards and although it does eventually comply to local laws, the report must come directly from governments or local courts. Secondly, because as mentioned before, this study seeks to examine a more nuanced form of extremism, diffuse in nature and with characteristics that are often overlooked by policymakers and the public in general. Finally, because although the research does scrutinise page posts, it gives special attention to user’s reactions through comments, which are exclusively moderated by page’s curators. Page’s policies on content will be addressed on the following sections.
3.3 The Echo Chamber and polarized discourses
Online platforms are often celebrated for giving users the opportunity to consume and diffuse knowledge, “supporting conversation, dialogue and debate” (O’Hara and Stevens, 401). However, this characterisation has been challenged by scholars such as Cass Sustein who claims that social networks act like echo chambers in which people only consume and share information from like-minded close friends, stifling the spread of diverse points-of-view and therefore fomenting extremism. Filtering and recommendation have been pointed as the responsible - Eli Pariser, the former executive director of MoveOn.org, argued in his book “The Filter Bubble” that web personalisation algorithms, such as Facebook’s News Feed, forces users to consume a dangerously narrow range of news creating “a unique universe of information for each of us” (Pariser, 9). Sustein goes further with his concerns with “the growing power of consumers to filter what they see” (Sustein, 8) by saying that the Web may be incompatible with democracy itself: “people should be exposed to materials that they would not have chosen in advance. Unplanned, unanticipated encounters are central to democracy… Without shared experiences, a heterogeneous society will have a much more difficult time in addressing social problems. People may find it hard to understand one another” (Sustein, 8-9). The echo chamber hypothesis has been greatly supported, especially as an argument for online radicalisation (O’Hara and Stevens, 402). In order to build up a consistent case for answering the research question, it is important to examine to what extent online platforms support the formation of echo chambers and its implications in extremist ideologies.
24
There is no question that social media sites are not mere facilitators of public interaction and, once inside a platform, users are subjected to the normative and technical structures of its environment. In addition to that, users tend to seek out information perceived consistent with their own points of views (O’ Hara and Stevens, 410). Nevertheless, a rigorous study conducted on Facebook by Eytan Bakshy showed that this does not mean that users are spared of clashes of opinions. First, Bakshy found that the closer people are on Facebook - that is, the more they interact with each other on the platform – the greater the probability of sharing each other’s links: “one of the most robust findings is that of homophily, the tendency individuals with similar characteristics to associate with one another” (Bakshy, 2012). At first glance, it looked like a confirmation of the echo chamber: users are more likely to echo their close friends. But Bakshy’s most crucial findings show that a user’s weak ties are collectively responsible for a vast majority of information spread; while the links shared by an individual’s close ties is more likely to contain information he/she would have seen elsewhere, links posted by the weak ties are more likely to point to novel information. Baykshy’s research reveals that most people’s relationships on Facebook are composed of weak ties. A calculation of the ties strengthens distribution combined with their corresponding probabilities, further supports the claim that users tend to receive more (and new) information from distant sources. Instead of serving as a “breeding ground for extremism” (O’Hara and Stevens, 405), it seems that Facebook may actually increase the spread of heterogeneous points of views.
Another important point to be made is that individuals’ use of social media platforms have more to do with their personal identity construction, through the promotion of one’s own views, than with other types of political reflection (O’Hara and Stevens, 407). Political positioning is an important part of identity and the internet helps enhance most people’s social experience by providing a space where they can express their opinions to a wider audience. The echo chamber argument seems to suggest that the internet has a homogeneous influence on an individual whose social context is particularly monodimensional (O’Hara and Stevens, 412). Nevertheless, as “mediated individuals” (Koukal, 190-240), users have “partial membership of multiple networks” (O’Hara and Stevens, 413), resulting in an inevitable exposure to disagreeing positions. Even if social media platforms do create echo chambers, it also creates opportunities to learn “through those networks of contacts, and the latter can ultimately outweigh the former” (O’Hara and Stevens, 413).
25
4 Actor-Network-Theory
The relevance of this research relies on adopting social media platforms, in particular Facebook, to investigate political discourses that could be otherwise unattainable. While traditional empirical methods, such as interviews and social experiments, often are manipulated by the participant, online platforms allow for what social scientists refer to as “data crawling”: “gleaning information about users from their profiles without their active participation” (Rieder, 1). Henceforth, with the growing importance of social media platforms in political debate, a methodology which contemplates such environments becomes increasingly relevant. For this reason, in this exploration of the social realm, an actor-network-theory (ANT) approach will provide the necessary resources for situating different political stands online. ANT is a field of knowledge departed from Science and Technology Studies (STS) that defends the idea that there should be no separation between human and non-human actors in social sciences. In other words, that the objects of science and the objects of technology should be seen as socially compatible (Latour, 10). This notion has been an important contribution for social studies since it shows how non-human actors such as microbes, rocks, ships or, in the case of this research, online platforms, also play a significant role in the construction of the social. The previous chapter elaborated on Facebook as a medium. In the current chapter I will focus on the methodological layer of the analysis. I will do this using the guidance of Bruno Latour’s social cartography and ANT and borrowing some of Venturini’s commandments for mapping controversies.
4.1 How to map the social
“The world is not a solid continent of facts sprinkled by a few lakes of uncertainties, but a vast ocean of uncertainties speckled by a few islands of calibrated and stabilized forms” (Bruno Latour)
In Reassembling the Social: an Introduction to Actor-Network-Theory, sociologist Bruno Latour introduces ANT as a means to redefine the concept of “social” in social sciences. He criticises what he calls the “sociology of the social”, which places the “social” as something material, a “specific type of ingredient that is supposed to differ from other elements” (Latour, 1). Instead, society, if it exists at all, has to be traced not as a place, a thing, a domain, or a kind of stuff, but as a "provisional
26
movement of new associations" (Latour, 238). There are no social explanations, only connections between things that are not themselves social (Latour, 5). Thus, the author seeks to introduce the concept of “sociology of associations” (Latour, 9)as a means to replace the aforementioned and give an alternative framework to the social sciences field. Essentially, the sociology of associations “reverses and generalises the sociology of the social” (Latour, 12). It reverses in the sense that it tries to explain, as an effect, what the sociology of the social understands as a cause; and generalises because it offers a relativist and broader approach to theory (Collier, 81). This society proposed by Latour is most visible in situations where the social is not stabilised. That is, when controversies dominated the debate and “the most heterogeneous relationships are formed” (Venturini, 261). In this fashion, ANT reveals itself as a fundamental approach in the pursuit of mapping Brazil’s current political crisis.
In its pursuit to analyse the social, the ANT strategy does not take into account a priori categories, nor limits the actor to a role. Instead it focuses on associations and entrusts the construction of the groups to the actors themselves. As explained, extremist traits listed in the previous section are not established with the intention of defining the actors, but only to frame the conceptual grounds in which this research will work on. Hence, to best understand how networks are forged it is necessary to “follow the actors” (Latour, 12) and travel wherever new heterogeneous associations are being made. This research will reassemble the social in the situation of Brazil by listening to actors’ comments on posts, seeking for groupings, relations and connections from their traces and locating what binds them. Following the actors will help this research discover aspects of societies that preconceived notions could overshadow.
On the exercise of “looking at controversies and describing what you see” (Venturini, 259), the researcher is given tools to make him more aware of what should not be taken for granted in the course of the inquiry. Latour proposes five steps, or as he calls, “five sources of uncertainty” that can be followed. The first states that “there is no group, but rather only group formation” (Rogers, 16). He explains that groups do not exist, only assemblages in continual development. The researcher is therefore encouraged to follow traces: an active spokesperson, a set of opposing discourses, an attempt to redefine existing formations. Here Latour introduces the difference between an intermediary and a mediator, which serves as a compass for the purpose
27
of mapping. An intermediary “transports meaning or force without transformation” (Latour, 39). It is a black box of sorts, which does not have any influence on what it’s carrying. A mediator, on the other hand, transforms, translates, distorts and modifies the meaning or the elements they are supposed to carry (Latour, 39). To understand the distinction between intermediaries and mediators is to understand, for example, how a medium (mediator) modifies information (intermediary).
The second source of uncertainty pertains to the nature of action. Here the task of the researcher is to map the driven forces of an action, looking through its various formats and manners in which an agency brings about effects. In particular, this question encourages the researcher to look for “how issues are made into matters of concern” (Rogers, 17). The third set of instructions concerns non-humans: “objects, too, have agency” (Latour, 63). Non-human actors are also actively involved in interaction and should be viewed as equally important. Fourthly, uncertainty is concerning the nature of facts, calling for the researcher to consider “the difference between a matter of fact and a matter of concern” (Rogers, 17). That is, facts are neither veridical (positivist empiricism) nor unreachable (subjective idealism), but “constructed through certain complex assemblages that end up prevailing” (Gubert, 605). Finally, the fifth guideline relates to something called second-degree subjectivity, which states that facts are always in service of a matter of concern, working in favour or against a specific actor. Taking these areas of uncertainty as part of the methodological outline, will allow this research to discover how political extremism came into being, who are their sources of authority and in what configuration do they present themselves. More specifically, the relevance of understanding how the “matters of concern” are constructed will be especially useful when looking to what the “pulling factors” of extremism are and what facts are most paramount to such discourses.
In ANT the researcher is invited to follow the traces that prompts an actor to act. In social cartography, actions take place in moments whereby actors find themselves in conflict with one another, creating a controversy. In this sense, the current political turmoil in Brazil presents itself as a controversy; if all actors agreed there would be no reason for the protests, no discussion over the legality of impeachment and, perhaps, less room for the rise of extremism. Thus, ANT will serve as the exercise of tracing the actors as they form networks inside the controversy, and
28
tracing the networks as they become actors. Furthermore, actors “partake in political work through engaging, mobilising and translating other actors’ interests, and finally enrolling actors to take on the same cause” (Durepos and Mills, 708). It is through this translation and alignment of actors that networks are formed. And it is this oscillation between actors and networks that controversies are constructed, associations become visible and issues become traceable. The political extremism in Brazil is being represented and assembled in form of statements, comments on news articles, violent activities and protests. This study aims to grasp the instances that allow for it to happen. Consequently, the research will rely on online platform’s capacity to function as a space for political engagement in the pursuit of describing controversies and of contributing to the “solidification of some portions of the social magma” (Venturini qtd in Rogers, 19).
29
5 Method
This section will introduce the steps taken to select Facebook spaces this research will study, explain the methods used to create a dataset suitable for the analysis and present the tools adopted to assist on the visualisation of the outcomes. To answer my primary research question both quantitative and qualitative analysis will be necessary. The decision comes upon the fact that mixed methods can provide a variety of analytical directions on complex social phenomena such as extremism. The quantitative approach will work with statistical methods of analysis, documenting, for instance, how many individuals expressed a particular view or the number of times a particular event occurred. The qualitative research objective is to look for connections within the data, “determine categories, relationships and assumptions that inform the respondents’ view of the world in general, and of the topic in particular” (Basit, 1). The limitations of this category of research should also be noted: the fact that the categorisation and analysis of what is to be considered extremist discourse will rely on the perspective of only one researcher. On the article “Diving on Magma”, Venturini says that when mapping controversies it is important to multiply the points of observation: “the more numerous and partial are the perspectives from which a phenomenon is considered, the more objective and impartial will be its observation” (260). Therefore, this limitation should be considered, since it can affect the findings and outcomes of this research. With this in mind, it will be important to detail the data collection process and keep an open mind when describing and interpreting the outcomes. In this context, ANT shows itself as a valuable approach to method, because it relies more on the actors and less on the researchers of the interpretation of the social. What follows is a step-by-step explanation of page selection, visualization procedures, analytical directions and coding scheme.
5.1 Page Selection
Before initiating the analysis it was necessary to decide upon a collection method suited to the research demands. Facebook has the potential not only to work as a platform where Brazilian politics is staged, but also as a place in which people and institutions use the subject to build their personal/political identity. As a starting
30
point, this research examined the study19 carried by professors Marcio Ribeiro and Pablo Ortellado at Sao Paulo University (USP), Brazil, which traced the digital profile of both right and left wing protesters. To conduct the study, Ribeiro and Ortellado followed the actors: they collected a unique user ID from each participant who publicly announced - by confirming presence on the Facebook events - his/her intention to participate on the ring-wing act of March 13th (about 410.000 users) and the left-wing act of March 18th (about 24.000). The participants were then compared to a second database of participants which listed people who, between February 13th and March 13th, had liked any publication from a series of pages selected from the monitoring of the political debate. A digital profile of the protesters was then created, pointing to which pages interested each group, according to categories such as news media, communities, political party and so forth. Since the objective of this research is to find out the resonance and configuration of extremists discourses, it was important to know which digital spaces are most relevant to users in articulation of political views. Thus, from USP’s digital behaviour report were selected the six most important Facebook pages in terms of popularity (number of likes) and overall engagement (sum of likes, comments and shares per post), that will serve as the basis for analysis. It was found that these pages match the concepts of collective and connective action proposed by social movement theorists Bennett and Segerberg.
According to Bennet and Segerberg, social movements navigate across different kinds of phenomena and action. The first, and more ‘familiar’ one, is the logic of “Collective Action”(5), which emphasises the need for formal organisations to hold individuals together on the search for a public good. In this view, organisations - such as news media corporations - are essential to “harnessing and coordinating individuals in common action” (Bennett and Segerberg, 12). This kind of social movement requires a stronger commitment from people and is viewed by the authors as a model of social activism that is getting in disuse. Nevertheless, I argue that the digital world and more specifically social media platforms, not only help reduce some of this partnership costs; but also allows users to create an identity through those organisational ideologies, without the need of being entirely devoted to them. Users often digitally engage with such institutions as a way of forging their own political
31
views. In this fashion, Facebook members share content, comment on posts, write messages and curate their profile as a way to represent who they are (or want to be) in real life. It is true that one of the most notable aspects of the right-wing supporters is the fact that they positioned themselves as being completely independent from any organisation or political party. However, organized spaces such as new media pages played an important role in helping create the fundamental ideologies that would lead to action. Undoubtedly, organised institutions played a secondary, if not, background role in the ring wing movement. It was an important role nonetheless. As channels of support, they served as an endorsement to users’ political views. Moreover, as suggested by Latour, a network is formed through the assemblage of multiple associations. This means that the coordination of action was not exclusively made by the Facebook pages responsible for the organisation of the protests, but was influenced by several actors. Three of the most relevant right-wing pages from the aforementioned study comply to the collective action notion: Veja Magazine, Estadão News and Folha News.
The second form of action present in social theory is “Connective Action” (Bennett and Segerberg, 6). Different from Collective Action Networks, this concept represents a type of self-motivated movement that grants the possibility for large crowds to “survive, deliberate, coordinate and expand” (Castells, 229) without the need of any formal social organisation. Connective Action Networks are “far more individual based and technologically organised sets of processes that result in action without the requirement of collective identity framing or the level of organisational resources required to respond effectively to opportunities” (Bennett and Segerberg, 14). In simple terms, the logic of Connective Action use the facilities provided by digital platforms to organised their “actions”. It is noticeable the surprising success of political messages that are being informally transmitted through digital technologies such as Facebook. Bennett and Segerberg suggest that these so called “Digitally Networked Action” (DNA) emerged from a shift from group-based to individualised technological societies, which depends of three main elements to survive: (1) “political content in the form of easily personalised ideas” (Bennett and Segerberg, 4); (2) concepts which rely more on feeling than on reason; (3) technologies that enables sharing these themes” (Bennett and Segerberg, 4). The Connective Action pages concept summarizes the very essence of the right-wing movement, as they were the main organizers of the protests. Around this thought, the three remaining Facebook
32
pages from the selection can accurately represent the so-called “Connective Action” idea: Vem Pra Rua (Take the Street), Movimento Brasil Livre (Free Brazil Movement) and Revoltados Online (Online revolted). These platforms were not only among the most popular between users, but were also responsible for “framing protest activity, gathering previously separate individuals in common action” (Poell et. al., 4). It has to be noted that “Connective Action pages” will inevitably carry the views and marketing strategies of the curators - an aspect that will be taken into account on the categorisation and analysis of the content – but their main function is to coordinate the organisation of social action through connecting people, spreading information and working as an autonomous environment for debate.
5.2 Tools
5.2.1 Netvizz
The Digital Methods Initiative tool, Netvizz, was used to collect the data. Netvizz is a data extraction application that provides information from different sections of Facebook’s service, which is essential to investigate complex social phenomena like extremism. The extraction mode used on Netvizz was the Page Data, which includes general statistics about the page, number of engagement, post link, post text, date and time of postage, list of comments per post and so forth. This feature was used to collect information from all six pages. Firstly, I established a time period. Since the analysis focus on right-wing extremism, it made sense to choose a period that would be relevant for both the group and the overall political discussion. Analysing the timeline of events I decided to set up the period from March 1st to March 15th. This decision is made for two reasons: (1) most important political happenings took place on the beginning of March, (2) the main march against the government happened on March 13th. These data files were then opened on excel and analysed using statistical and emergent coding techniques.
5.2.2 INFOGR.AM and Sankeymatic
Infogr.am is a web application that provides tools for the creation of interactive and downloadable infographics on the web. This resource was used for the visualisation design of content analysis for news pages. Using the bubble graph option, it was possible to build a visual relationship between news subject, comment category
