The discourses of journalistic vehicles and the impact of the audience on Twitter about March 15^th 2015 protests in Brazil

Abstract

In this paper, we aim to discuss how news journalistic vehicles described the Brazilian March 15^th demonstrations on Twitter and how these descriptions of the event reverberated in the audiences in the same tool. To that end, we part from a data set composed of tweets and retweets of Carta Capital, UOL Notícias, Estadão, G1 and Folha de S.Paulo. We observed the media discourses using a contingent content analysis, whereas we observed the reverberation by identifying the most replicated messages. The results point to small differences between what the media post and what the public reverberates through retweets. We identified that whether the demonstration was or was not against Dilma’s government may not have been the fact that most interested the audiences at that time. Timely information about the events on the streets received more emphasis through retweets than messages that just announced the demonstrations.

Keywords
Twitter; Journalism; Audience; Social networks; Demonstrations

Resumo

O trabalho tem por objetivo discutir como os veículos jornalísticos descreveram os protestos de 15 de março de 2015 no Twitter e como essas descrições repercutiram em suas audiências na mesma ferramenta. Para tanto, parte de um conjunto de dados composto por tweets e retweets de Carta Capital, UOL Notícias, Estadão, G1 e Folha de S.Paulo. O discurso dos veículos foi observado a partir do emprego de uma análise de conteúdo focada em contingência, ao passo que a repercussão foi observada por meio da identificação das mensagens mais replicadas. Os resultados apontam para pequenas diferenças na agenda entre o que os meios postam e o que o público confere destaque por meio de retweets. Percebe-se que o protesto ter sido ou não contra o governo Dilma pode não ter sido o que mais interessou as audiências naquele momento. Informações pontuais sobre os atos nas ruas receberam mais destaque por meio dos retweets do que outras que apenas anunciavam os protestos.

Palavras chave
Twitter; Jornalismo; Audiência; Redes sociais; Protestos

Resumen

El estudio tiene como objetivo analizar cómo los vehículos periodísticos describieron las manifestaciones del 15 de marzo de 2015 en Twitter y cómo esas descripciones se hizo eco en sus audiencias en la misma herramienta. Para eso, se parte de un conjunto de dados compuesto de tweets y retweets de Carta Capital, UOL Notícias, Estadão, G1 y Folha de S.Paulo. El discurso de los vehículos fue observado a partir de la utilización de una análisis de contenido centrado en contingencia, mientras que el impacto se observó mediante la identificación de los mensajes más replicados. Los resultados muestran pequeñas diferencias entre la agenda de los vehículos y lo que el público destaca a través de retweets. Se observa que el facto de la manifestación ser o no contra el gobierno Dilma puede no haber sido lo que más interesó la audiencia en ese momento. Información específica sobre los actos en las calles recibieron más destaque que otras mensajes solamente anunciando las protestas.

Palabras clave
Twitter; Periodismo; Audiencia; Redes sociales; Manifestaciones

Introduction

In 2014, Dilma Rousseff was reelected as president of Brazil with 51.6% of the total number of valid votes. In January 2015, she announced a package of adjustments that increased the value of several taxes in the country, which, along with the progress of the Operation Car Wash¹, contributed to a significant drop in approval ratings of her government². Opposition groups reacted to these facts through the organization of several protests, including on March 15th, 2015. The Movimento Brasil Livre – MBL (Free Brazil Movement) is seen as the main responsible for rallying up the demonstrations. Founded in São Paulo, the movement claims to have as goals “free and independent press, economic freedom, separation of powers, free and reputable elections, and the end of direct and indirect subsidies to dictatorships”³. The impeachment request for president Dilma, the accountability of PT for the corruption scandal at Petrobras, and the defense of military intervention in the country were some of the topics highlighted in the discourses of those demonstrations. Differently from the June 2013 protests (RECUERO; BASTOS; ZAGO, 2014), the acts were not marked by violence among police and demonstrators. Demonstrators even took selfies with police officers.

In the context of demonstrations, the public agenda does not always coincide with media agenda. While media highlight certain news related to the event, the public gives visibility to other aspects. Thus, the appropriation of websites, applications and platforms of digital communication to the coverage and conversation about this type of news event directly reflects on the political debate that circulates through the media. At the same time that traditional news outlets publish content on Twitter and their publications are retweeted by their followers, the production of meaning around the coverage reflects in the construction of a myriad of discourses about the facts.

Based on this scenario, this paper aims to: a) discuss how news outlets described the March 15 2015 protests on Twitter, and b) understand how those descriptions reverberated within their audiences in the same tool. In order to address the first outlined objective, we used contingency analysis and network analysis of a sample of tweets by or mentioning Carta Capital, UOL Notícias, Estadão, G1 and Folha. We collected this sample between March 14 and 16, 2015. In order to address the second objective, we identified the most replicated messages with mentions to the journalistic profiles from the sample.

The paper is organized as follows: first, we address journalism in the context of Twitter. Next, we focus on its use in the context of protests and political movements. Later, we present methods, results, and discussion. Our conclusions point to a light dissonance between the outlets discourses on Twitter and the most replicated tweets by their audiences.

Journalistic appropriations of Twitter

Studies on journalistic appropriations of Twitter are not something new. The myriad of sites, platforms and other supports for digital communication generate different publication formats that are able to also generate a series of appropriations in the field of journalism (ZAGO, 2009). That’s what happened with Twitter. The website brings together several web and mobile publication capabilities, allowing anyone to publish and promote the circulation of news. Journalistic appropriation of Twitter took place when news outlets entered the website, and with the proliferation of informative content published on the tool. Java et al. (2007) identified that one third of the updates did not respond to Twitter’s original question – “What are you doing?” – thus, it was no wonder that the question was later changed to “What’s happening?”.

Regarding journalistic practices on Twitter, Silva and Christofoletti (2010) highlight the diffusion of news headlines from portals, some of them accompanied by links. They cite Zago’s (2008) analysis, which identified news distribution with headlines and links as one of the most frequent uses performed by journalistic accounts on Twitter. Uses such as news alerts and live coverage were also identified. By appropriating it for journalism, Twitter can have different uses, and among the specific journalistic practices that can be observed on the tool, Zago (2009) lists some that are particularly important for the purposes of this article. This set of characteristics, besides allowing new forms of production and circulation of contents, when thought in the context of protests are able to enhance discourses that come either from news outlets or from their followers.

• Updates via mobile devices, enhancing the ability of sending and receiving messages in real time: the coverage on the streets is facilitated by Twitter, which can be used via mobile devices and require short updates. This allows the live coverage not only through texts, but also using images and audio. In the June 2013 demonstrations, Mídia Ninja⁴ (Ninja Media) intensively used its Twitter profile for the live coverage of the acts on Brazilian streets, combining this coverage with the use of Twitcasting⁵ – a service that allows the live transmission of images in the broadcasters’ Twitter profile.

• Production of specific contents for Twitter: this stimulates the practice of live coverage, because it allows the journalist or citizen to publish content directly from where the event is taking place, providing details using different formats;

• Exploration of hypertext on the updates: most of Twitter updates have links to other online spaces with more information; that is, at the same time that the coverage can happen in the same place as the event, it can also be implemented through tweets with links for publications with more details;

• Mobile access: mobility allows the publication of live content. The access to Wi-Fi networks during demonstrations on the streets⁶ freed users from depending on the signal from carriers;

• Social network characteristic of Twitter, allowing followers to communicate among themselves and the circulation of contents: this feature contributes to a higher visibility of the contents, through replication, and also to an increase in the number of followers in the replicated profiles.

Taken together, these features contribute for understanding Twitter not only as a space for broadcasting, but also for information circulation. In the context of journalism, while news outlets use it to put into circulation tweets related to news events, users can promote the recirculation of these contents, through retweets, comments and mentions (ZAGO, 2014).

Circulation and recirculation of contents can be seen in a context of spreadable media (JENKINS; FORD; GREEN, 2013), in which traditional media no longer holds the exclusive monopoly of circulating contents – other actors can contribute to this circulation. The authors identify an “emerging hybrid model of circulation, where a mix of top-down and bottom-up forces determine how material is shared across and among cultures in far more participatory (and messier) ways” (JENKINS; FORD; GREEN, 2013, p.1). By retweeting and commenting, users contribute to the visibility of certain topics that are being discussed on Twitter (ZAGO; BASTOS, 2013). This spreadability is not the mere reproduction of content. According to Jenkins, Ford and Green (2013, p.27), “As material spreads, it gets remade: either literally, through various forms of sampling and remixing, or figuratively, via its insertion into ongoing conversations and across various platforms”. Thus, even though information circulation and recirculation happen in a specific space, these actions influence, too, other media.

Protests and political debate on Twitter

Perhaps the Arab Spring would not have been the same without the use of social network sites for organizing and promoting the demonstrations in the North of Africa and in the Middle East. Gomes and Maia (2008, p.310) highlight the network architecture as ideal for political participation, because they understand that “a multilateral structure, equipped with multidirectional flows of information and communication, is a symptom for a political structure in which it is possible to recognize that public sphere has something to say and can directly influence political decision”. Facing these appropriations, it is clear that the content production and recirculation resources that Twitter offers for journalism are being appropriated not only by journalists and news outlets in the coverage of protests and demonstrations, but also by other actors, that use the website as an informational and organizational space, publicizing and organizing movements and acts (BENNETT; SEGERBERG, 2011; POEL; BORRA, 2011; TUFEKCI; WILSON, 2012; TORET, 2012; RECUERO; ZAGO; BASTOS, 2014), as well as incorporating the tool into political activities, such as in the successful Barack Obama presidential election campaign (GOMES et al, 2009).

Analyzing recent movements, Castells (2012) points out to reconfigurations in the forms of social mobilization due to the uses and appropriations of mobile devices and social network sites. After dedicating himself to the studies of the transformations on power relations in interaction with changes in the field of communication, the author identifies a new pattern of social movement, the networked social movements: they coexist online and offline, depend on the use of social media for its onset, and constitute a “hybrid space of freedom”, with the participation of both sides involved in the conflict (RECUERO; ZAGO; BASTOS, 2014). The presence of diverse actors on Twitter is part of this understanding of networked social movements. It is an aspect that brings together different analyses about its uses in this context and refers to the lost of centrality of mass media on the narrative of news events (MALINI; ANTOUN, 2013).

In the informational level, Twitter’s appropriation as an information channel occupied by different actors sets up what Malini and Antoun understand as the new multitudinous narratives, capable of make the transition from the informational model of mass media to a communicational model of multimedia, which favors the coordination of collective action in the movements. Penney and Dadas (2013) identify Twitter functions during the Occupy movement, showing that beyond facilitating face-to-face and online actions, the website allows live coverage, information replication and use of links, the exposition of the diversity of opinions, and the facilitation of the discussion about the mobilizations.

The use of Twitter by networked social movements also refers to the concept of technopolitics by Toret (2012), when he analyzes the Spanish movement 15M: “the tactical and strategic usage of technological tools (including social networks) for the organization, communication, and collective action”. His approach is simple and direct, but might suggest that it is basically a relation between internet and digital devices with street mobilization. He explains that technopolitics goes beyond cyberactivism or clickactivism – it manifests itself by taking over the physical, digital and media public space. This triple occupation refers to what the authors calls collective transmedia narrative, which occurs with the use of a multilayered system, from an overlap between the street and the network that generate organizational demonstrations and information exchanges that report the progress of the mobilizations. These organizational demonstrations and information exchanges take place due to the possibility of collaboration for content production and circulation and to the multimedia and transmedia options through the network. Toret highlights the use of Twitter in a context of networked movements, emphasizing his preference for it over Facebook, mostly because of the privacy configurations of the last one. On Twitter, it is less easy to hide contents, in a way that information ends up reaching more people in comparison to the other social network site.

In this sense, Twitter appropriations by social movements and activists appear on the media scenario previously occupied only by the said official communication channels. The opening characteristics of Twitter give space for multiple discourses that gain visibility through the dynamics of replication offered by the site. It is possible to identify a dispute for the protagonism of the narrative of the protests in the quantity of tweets published by different profiles, not only traditional news outlets, but also from movements and citizens, which either reply or produce original content about the events.

Methodological proposal

As we previously explained, our aim is to analyze (1) how news outlets described the March 15th, 2015 protests on Twitter; (2) how those descriptions reverberated within their audiences in the same tool. Therefore, we combined data collection and analysis approaches, discussed below.

Data collection

In order to collect the tweets, we used yTK⁷, a script that allows data collection using Twitter’s API. We included in the searches 18 news outlets among those with the higher number of followers (and, possibly, with the higher impact with Brazilian audience) and the main outlets in each region of the country. Those profiles were monitored between March 14 and 16, 2015. We collected data on (1) number and content of tweets from outlets; (2) number and content of tweets citing outlets; (3) number of following and followers of the outlets by that time. After collecting the data, we selected five journalistic profiles with the higher number of tweets about the protests. We ended up with the outlets listed on Table 1.

Table 1
Data on the five outlets with the higher number of tweets about the protests

	Total tweets in the period	Tweets about the protests	Total RTs	Followers	Followers
Cartacapital	725	64	15157	778.000	106
Uol Noticias	2308	171	22713	1,17 million	7.490
Estadao	1150	66	62315	2,62 million	67.700
G1	2151	83	58402	5,07 million	134
Folha	1720	71	44380	2,82 million	194.000

Data analysis

In order to understand how the news outlets narrated the protests, we worked with contingency analysis (OSGOOD, 1959; BARDIN, 2006) and network analysis (WASSERMAN; FAUST, 1994; SCOTT; CARRINGTON, 2014). Contingency analysis is a type of content analysis. In this proposal, content units (in our case, the tweets) are analyzed in a quantitative way, by verifying the frequency of the most used words. These words are then classified into concepts, by a qualitative analysis of their position on tweets, by similarity. Thus, for example, “Dilma” and “president” were classified in the same category. Finally, textual units are analyzed by co-occurrence of the concepts. We observed which concepts were more likely to appear together in the dataset and how strong their relationship was. In order to visualize these data, we used network analysis. That way, data are plotted as a graph, where nodes are the concepts we built and connections are the cooccurrence relationships among these concepts. The more central the concept, the more it was used in the news outlet coverage. The stronger the relationship between two concepts, the stronger the line joining them. Graph colors (tones) were built using the modularity (tendency to appear together) of the concepts, aiming at building a more complete view of the discourse and narrative forms of the outlets in question.

In order to understand how the outlets reverberated within their audiences, we analyzed the retweets. Data were initially accounted for (that is, we counted the total number of retweets per tweet). Later, we removed from data those retweets that were not related to the protests, and selected for the analysis the top 10 most retweeted posts about the protests mentioning the outlets. In this case, we were less interested in the set of discourses propagated by the outlets, which is the topic of the first analysis, but ultimately we were interested in which discourses received more attention and reverberated more within the audience. Thus, we used a different method. Once tweets were selected, they were classified according to their content, always with a focus in how the protests were described.

Analysis

Our analysis is divided into two parts. In the first, we analyze the outlet’s discourses. In the second, we analyze the retweets, focusing on the top 10 most retweeted tweets and the associated concepts.

Media discourse

Carta Capital

The @cartacapital profile published a total of 64 tweets that directly talked about the protests. The most central concept is “protests”. Among the groups of concepts (marked by differences in tone in Figure 1) that co-occurred more often, we see a group with medium tones on the right, where the main focus is on the debate of tvcarta, which analyzed the protests. In the lower right part of the graph, a second focus is on the live coverage. Featured in this group is the co-occurrence of “coup”, “dictatorship”, “groups”, and “ask”. In the dark group in the top portion of the graph, we can find another narrative of the events. Here, there is a talk of “impeachment” and of the “act” as being “pro-impeachment”. In the left core of the graph, there is a talk of “indignation” that the “organizers” intend to “keep going”, and in the group in light tones in the top portion of the graph, “protests”, “massive”, and “dilma”. Protests here, therefore, are mainly narrated as being in favor of impeachment, in favor of a military coup. Their narrative modes are “live” and in a “debate” discussion. President Dilma is cited as directly associated with the protests, though without a direct qualification. We observe that the only place mentioned is São Paulo.

Figure 1
Co-occurrence of concepts on @cartacapital profile

Estadão

The second news outlet that we analyzed was @estadao, with 66 tweets directly related to the protests. In Figure 2, the graph lists the key discourses related to the events. As in @cartacapital’s case, the most central concept is “protest”. We see here a series of discourses separated by groups. The first of them, in the top central part of the graph, narrates the protests as taking place primarily in São Paulo, Rio de Janeiro, and Federal District, and link the protests to the “government”. The group on the left side of the graph reports the “panelaço”⁸ that occurred after the pronouncement of the ministers and asks for the participation of readers by sending information. In the top left part of the graph, “Porto Alegre” is mentioned. In lighter tones, in the top right, there are talks of “social networks” and “reverberation”. The group in the lower part of the graph mentions “Dilma”, related to “demonstrations”, “impeachment”, “larger”, and “diretas-já”⁹. Here, the protests are once again associated with Dilma, and presented as the largest one since 1984 (“diretas já”). In the same way, they are also associated with “impeachment” requests, and more geographically dispersed than in @cartacapital’s coverage.

Figure 2
Co-occurrences of concepts on @estadao profile

Folha

@folha was one of the outlets with more tweets about the protests in the analyzed period, 67. In Folha data, we also observed a varied set of tweets that built more frequent discourses (marked by the tone differences in the graph on Figure 3). In the top central part of the graph, we see “protests” associated with expressions like “government”, “anti Dilma”, and even with “corrupts”, “corruption”, and “cheering”. Besides that, the core located on the right associates “act” to “impeachment”, “petistas” (people affiliated to PT party in Brazil), “Dilma”, “ministers”, “changes”, “package”, and “anti-corruption”. In these two cores, we see that the protests are associated to the government, and personally against Dilma Rousseff, PT party, and corruption (which consequently is associated to these two actors). In the lower left part of the graph, we see the expression “demonstrations” associated to “photo” and “video”, and an invite to “send”, focusing on the participation request. In the top left part, we see “police”, “protesters”, “bombs” and “selfie”, in one of the only cores that associated (although weakly) the protests to some kind of violence.

Figure 3
Co-occurrence of concepts on @folha profile

G1

G1 published a total of 83 tweets focusing on the protests. On Figure 4, we can see the key concepts present on these tweets and its co-occurrences. There is a big central cluster in light tones, followed by 5 smaller clusters. In the main cluster, we can see protests associated to “corruption” and to “Dilma”. A small cluster in light tones in the top portion of the graph focuses on the live coverage, with the concepts of “real” and “time”. In the low part of the graph, there is a cluster focused on “democracy”, “strong”, “country” and “fight”, concepts from president Dilma’s discourse. In the top part of the graph, there are two small clusters. In one of them, we see “são paulo” as location, and, in the other, “aécio” (in a narrative about the remote participation of the former candidate). It is the first group where the name of the former candidate appears with strength.

Figure 4
Co-occurrences of concepts on @g1 profile

@uolnoticias

In the data referring to @uolnoticias, we see that the outlet published 171 tweets about the protests. In Figure 5, we can see five main groups. The top right portion represents the largest group, in which the most important concepts are the hashtags #protests and #march15, besides locations in São Paulo and Rio (“avenue”, “Copacabana”, #rio, #sp, and #pernambuco), and “datafolha”, “acts”, “demonstrations”, “diretas-já”, “hired”, “two”, “politician”, “multitude”, “court of justice”, “pt”, “images” and “photos”. A second group (located on the left) is focused around “Dilma”, with elements related to its portrait, such as “cowardice” (together with “aécio”, “silence”, “sign”), “tension”, “groups”, “five”, “before demonstrations”, “streets”, and “record”. This group refers to the terms that appeared in more co-occurrence with Dilma. In the lower right, concepts are centralized around “Brazil”, “intervention” and “military”. Around are: “supporters”, “panelaço”, “jihadists”, “war”, “civil”, “#video”, and “impeachment”. In the center of the graph, we can see the concepts “protests” (in the center), surrounded by “government”, “reunite”, “million”, “states”, “country”, “thousands”, “thousand”, “police”, “brasilia”, “people”, “minister”, “march”, “uolnoticias” and “@uol”. In the lower left the graph, in one of the corners, there is another group of strongly related concepts. In this group, almost all concepts in the same degree, we see “old”, “media”, “risks”, “white”, “hate”, “international”, and “forbes”.

Figure 5
Co-occurrence of concepts on @uolnoticias profile

Audience discourse

Carta Capital was the only news outlet that didn’t use expressions that clearly defined the protests as being against the government or that personally target the president. Estadão, G1, Folha, and UOL Notícias used expressions like “against Dilma”; “against the government”, “against the government and against corruption”, and “anti-Dilma”, directly linking, in their tweets, the protests with corruption, PT and Dilma. It is important to understand that, even though there was notably a smaller quantity of demands from the protesters, the translation of the event by the outlets is noticeably focused on the corruption associated with federal government and PT. However, these positions not necessarily reflect in what was more reverberated by the audiences in the same period. Besides collecting tweets by the outlets, we also collected data on the reverberation of tweets from these same outlets. In order to collect the RTs, we identified the total number of replications of tweets reproducing or mentioning the news outlets from the sample. Thus, the total count refers not only to the reproduction of tweets form the outlets (as in “RT @g1: Former player Ronaldo participates in the protests against the government in São Paulo http://t. co/UNLAmNjXiP #G1 http://t.co/eEVVgk8FIk”) but also to retweets to posts mentioning the outlets (as in “RT @ricardodesales: @g1 confirms: government of SP released the ticket gates in the subway for the protests http://t.co/ieOEu4JBgB http://t.co/FNo4LWV”). Thereby, the analysis of the audience reverberation goes beyond the tweets from the outlets.

Thus, the most retweeted message by Carta Capital talks about Lula and Dilma dolls hanging from a overpass in Jundiaí. However, among the top 10 most retweeted, some suggest positioning against the protests: the second, that addresses a criticism against the understanding of democracy by those that suggest the impeachment of Dilma; the third, that joke with those that took baby-sitters to the demonstrations on the 15^th; the eighth, by pointing the misconception of those that say that corruption only occurs within PT. The tenth most retweeted message related to Carta Capital mentions the press itself, accusing it of manufacturing truths, instead of just informing it.

860 retweets were made to a tweet by Estadão mentioning the reaction of a sociologist to the protests. According to her, the demonstrations looked like the Family March, without any type of political support. The second most retweeted message talks about the size of the demonstrations, considering them the largest one since Diretas Já in the 80s. The third and fourth most shared tweets also highlight the number of people on the streets, and the fifth tweet once again compares the protests of March 15th to Diretas Já, and the tenth most shared one highlights the number of 2 million protesters on the 15^th.

A tweet with a protester testimony that is said to be black and poor and asking for Dilma to leave was the most retweeted by Folha de S.Paulo, followed by a tweet that said that the size of the protests had surprised the government, which considered them significant. Shortly thereafter, the profile of the protesters appears as the most retweeted, in a message about the green and yellow clothes of 580 thousand people on Paulista Avenue – which refers to the number of protesters on the 15^th. An understanding about the profile of the protesters appears again as the fourth most shared tweet, that talks about the groupie photos made with police officers from the Shock Battalion, followed by another tweet that talks about the number of people in the protests, but this time in Ribeirão Preto. Clothes are once again mentioned in the sixth most retweeted message, which talks about Aécio dressed in the shirt of the Brazilian soccer team while he follows the demonstrations from his balcony. The seventh most shared tweet can be linked to an understanding of the profile of the protesters, and maybe as against the protests. It talks about the coxinha¹⁰ barbecue in Porto Alegre.

A tweet by G1 about the participation of the former soccer player Ronaldo in the protests in São Paulo received 619 shares. Ronaldo is once again mentioned in another tweet among the top ten. Four tweets, that call followers to follow the coverage and see photos of the protests, are among the top ten most shared from this outlet. The response of the international press also appears among the ten most shared.

How international media saw the protests of March 15^th compared to the protests of 2013 (older, whiter, and richer) appear as the first and fifth most retweeted message from UOL Noticias. The tenth tweet also refers to comments from abroad, about a North American TV program that satirizes the corruption on Petrobras and the protests. The number of people protesting against Dilma (210 thousand) in São Paulo was the second most shared tweet from the outlet, followed by another mentioning the fall of a Folha de S.Paulo drone in the middle of Paulista Avenue, injuring two people. Tweets criticizing Dilma’s government and the mention to panelaço during her pronouncement are shared as well as a tweet that says that it will not make any difference changing PT for another party.

Notes and considerations

The analysis of the tweets published by the outlets and the ten most shared tweets by their audiences aimed at identifying how media described the protests of March 15 and how those descriptions reverberated in the audience on the tool, so then we could evaluate how Twitter appropriations interfere in the political discussion about the protests. Even though the majority of outlets frames the demonstrations as against the government, against the president Dilma, and against corruption, many times the most replicated information do not refer to the nature of the acts. We can understand that the fact that the protest was or was not against Dilma’s government may not have been what caught more the attention of the audiences at that time. Other information, about the facts on the street during the demonstrations, as well as the number of people in the protests, received more prominence through retweets compared to messages that simply announced the protests.

Even though Carta Capital discourse frames the protests as in favor of the impeachment and military coup, in the analysis of the retweets it is possible to perceive a framework division. While the most shared tweet talks about Dilma and Lula dolls being hanged, criticism for lack of political basis also reverberates, against those that defend that the president be dismissed and those that took baby sitters to the protests. The act of sharing, in a criticizing tone, of a tweet that mentions the media as manufacturer of truths calls into question the role of media in the dissemination of the protests. Audience behavior, in this case, reflects a critical tone before the outlet’s publications on Twitter. The situation is faced through the retweets that put into circulation media contents accompanied by contrary positions to the ones undertaken by the vehicles. The connection between how Estadão described the protests and the reverberation of this description by its audience stands out in relation to the number of protestors on the streets. The outlet and its audience highlight the size of the protests considering the number of participants. The most shared tweet, however, puts in evidence an analysis that criticizes the lack of political support in the acts. There is not, however, a significant difference between the discourses of the outlet and the audience.

The relationship between Folha’s discourse and the audience discourse is not so clear. While the outlet links the protests to Dilma’s government, organized against corruption and against the president, the audience appears to consider more relevant the information about the profile of the protesters and the number of people on the streets. It is possible to note a strong position of the audience, however, when the most retweeted message mentions the participation of a black person asking for Dilma to leave.

The mention to the participation of the player Ronaldo in São Paulo appears in G1 graph (Figure 4), but with a smaller representativity compared to the significance conferred by users, that appear twice among the ten most shared tweets. Even though the outlet had linked the protests to Dilma’s figure and to corruption, the message that reverberated more was about the participation of a celebrity in the protests.

UOL Noticias graph is dense, gathering expressions in five groups, which shows how the outlet fragmented the coverage of the protests in different approaches. The variety of terms does not reflect in the audience appropriation of tweets, which ended up retweeting information on the profile of the protests in the first place. There was not a clear definition of discourse appropriation from the outlet to defend or criticize the government, considering that the most shared tweets did not entered this merit.

The outlets and audiences’ Twitter appropriations here analyzed showed uses according to Zago (2009) about taking advantage of mobility for the live coverage, generating specific content for the medium, making calls to the followers to follow the live tweets, and exploring hypertext by adding links to news pieces. These uses, already common nowadays among media outlets on Twitter, define the tool as a space for news circulation (ZAGO, 2014), endorsing the idea of spreadability proposed by Jenkins, Ford and Green (2013), due to the sharing of tweets by the audiences. In a context of social movement, as in the March 15 case, Twitter appropriation not only by outlets, but also by its public, shows a strong tendency of its use for the broadcast of the events. This broadcast through number of protesters and expressions that seek to define the acts meets the arguments of the authors that study the relationship between networked movements and digital communication technologies (BENNETT; SEGERBERG, 2011; POEL; BORRA, 2011; TUFEKCI; WILSON, 2012; TORET, 2012; RECUERO; ZAGO; BASTOS, 2014). Regarding the lost of centrality of the narrative of facts by mass media, supported by Malini and Antoun (2013), we realize that while in some cases the discourse of the vehicle is very close to the discourse by the audience about the protests, in other cases, however, we identify the emergence of a discourse that even by sharing tweets from the media passes a different message about the events.

If we consider that the internet can enhance political discussion and participation, Twitter meets this possibility by serving as a space that houses the discourses of different actors, as well as the activity of the news consumer public in the circulation of the debate about the protests. Our analysis showed, in this case, how the tool appropriations by mass media and its audiences could collaborate for the effectiveness of a communicational process that, even guided by the information distribution without the configuration of an effective dialogue between producers and receivers, houses diverse voices about the events.

• 1
  With its onset on March 2014, the Operation Car Wash investigates a laundering and diversion scheme involving Petrobras, large contractors from Brazil and politicians.
• 2
  Dilma’s popularity drops from 42% to 23%, says DataFolha. Source: http://goo.gl/qh3cjJ
• 3
  Source: http://goo.gl/BzFfK8 Accessed on: 06/19/15.
• 4
  https://ninja.oximity.com/
• 5
  http://us.twitcasting.tv/midianinja
• 6
  In 2013, a poster was largely promoted on the internet requesting that residents opened their Wi-Fi network signal during the demonstrations in Brazil – http://goo.gl/erbaLB.
• 7
  yourTwapperKeeper
• 8
  The term “panelaço” refers to people banging pans in protests.
• 9
  “Diretas-já”, or “Direct Elections Now” was a protest that occurred in Brazil asking for direct elections in 1984.
• 10
  Coxinha is a very popular deep fried chicken snack in the country. The term was used as a form of pejorative reference to the right wing elite by the left parties.

Referências

• BARDIN, L. Análise de conteúdo Lisboa: Edições 70, 2006.
• CASTELLS, M. Networks of outrage and hope New York: Polity Press, 2012.
• BENNETT, W.L.; SEGERBER, A. Social technology and the organization of protests against the global economic crisis. Information, Communication & Society, Londres, v.14, n. 6, p. 770-799, 2011.
• GOMES, W. et al . “Politics 2.0”: a campanha online de Barack Obama em 2008. Revista de Sociologia e Política, Curitiba, UFPR, v.17, n.34, p.29-43, 2009. Disponível em: http://goo.gl/L0SokH Acesso em: 19 abr. 2015.
  » http://goo.gl/L0SokH
• GOMES, W.; MAIA, R. Comunicação e democracia – problemas e perspectivas São Paulo: Paulus, 2008.
• JAVA, A.; SONG, X.; FININ, T.; TSENG, B. Why We Twitter: Understanding Microblogging Usage and Communities. Procedings of the Joint 9th WEBKDD 2007. Disponível em: http://goo.gl/T0ld5N Acesso em: 19 abr. 2015.
  » http://goo.gl/T0ld5N
• JENKINS, H.; GREEN, J.; FORD, S. Spreadable Media New York: NYU Press, 2013.
• MALINI, F.; ANTOUN, H. @ Internet e a Rua Porto Alegre: Sulina, 2013.
• OSGOOD, C. The representational model and relevant research methods. In: POOL, I. (Org). Trends in Content Analysis Urban, Illinois: Univ. of Illinois Press, 1959.
• PENNEY, J.; DADAS, C. (Re)Tweeting in the service of protest: digital composition and circulation in the Occupy Wall Street Movement. New Media & Society, 2013. Disponível em: http://goo.gl/ghNf6u Acesso em: 19 abr. 2015.
  » http://goo.gl/ghNf6u
• POEL, T.; BORRA, E. Twitter, Youtube, and Flickr as platforms of alternative journalism: the social media account of the 2010 Toronto G20 protests. Journalism, v. 13, n. 7, 2011. Disponível em: http://goo.gl/uwHhfW Acesso em: 19 abr. 2015.
  » http://goo.gl/uwHhfW
• RECUERO, R.; BASTOS, M.; ZAGO, G. Narrativa e violência: a cobertura do Outono Brasileiro no Twitter. Matrizes (USP), São Paulo, v. 8, n. 2, p. 191-217, 2014a. Disponível em: http://dx.doi.org/10.11606/issn.19828160.v8i2p191-217 Acesso em: 19 abr. 2015.
  » http://dx.doi.org/10.11606/issn.19828160.v8i2p191-217
• RECUERO, R.; ZAGO, G.; BASTOS, M. O Discurso dos #ProtestosBR: análise de conteúdo do Twitter. Galáxia (PUCSP), São Paulo, v. 14, n. 28, p. 199-216. 2014b. Disponível em: http://goo.gl/eiyi6F Acesso em: 19 abr. 2015.
  » http://goo.gl/eiyi6F
• SCOTT, J.; CARRIGNTON, P. The SAGE Handbook of Social Network Analysis London: Sage, 2014.
• SILVA, F.A.M.; CHRISTOFOLETTI, R. Jornalismo em 140 toques: análise de três contas do Twitter no Brasil. In Texto (UFRGS. Online), Porto Alegre, v. 1, n. 22, p. 65-80, 2010.
• TORET, J. Una Mirada tecnopolítica sobre los primeros días del 15M. In: ALCAZAN et al. Tecnopolitica, Internet y Revoluciones sobre la Centralidade de Redes Digitales en #15M Barcelona: Icaria Editorial, p. 50-69, 2012.
• TUFEKCI, Z.; WILSON, C. Social Media and the Decision to Participate in Political Protest: Observations From Tahir Square. Journal of Communication, v.62, n.2, p. 363-379, 2012. Disponível em: http://goo.gl/MGcqS1 Acesso em: 12 mar. 2014.
  » http://goo.gl/MGcqS1
• WASSERMANN, S.; FAUST, K. Social Network Analysis. Methods and Applications Cambridge: Cambridge University Press, 1994.
• ZAGO, G. Jornalismo em Microblogs: Um Estudo das Apropriações Jornalísticas do Twitter 2008. 140p. Trabalho de Conclusão de Curso (Graduação em Comunicação Social, Habilitação em Jornalismo), Universidade Católica de Pelotas.
• _______. O Twitter como suporte para produção e difusão de conteúdos jornalísticos. Ciberlegenda Niterói, UFF, v. 11, n. 21, p. 1-16, 2009.
• _______. Circulação e Recirculação de Narrativas do Acontecimento no Jornalismo em Rede: A Copa de 2014 no Twitter. 2014. 217f. Tese (Doutorado em Comunicação e Informação), Universidade Federal do Rio Grande do Sul.
• _______.; G.; BASTOS, M. T. Visibilidade de Notícias no Twitter e no Facebook: Análise Comparativa das Notícias mais Repercutidas na Europa e nas Américas. Brazilian Journalism Research, Brasília, v. 9, n. 1, p. 116-133, 2013.

Publication Dates

• Publication in this collection
  Sep-Dec 2016

History

• Received
  26 Sept 2015
• Accepted
  16 Aug 2016