HATE SPEECH IN DIGITAL ENVIRONMENTS

Online hate speech practices are a challenge to democracies, particularly because they touch on issues related to one of its pillars, namely freedom of speech. Seeking to understand the dynamics of negotiation between the safety of protected groups and the insurance of freedom of speech, this study investigated the public debate on Twitter and Facebook about hate speech and censorship. For that purpose, we will first present an overview of the definitions and legislations informing the understanding about hate speech, discussing the specificities of the online context and the forms of definition and self-regulation proposed by the platforms. Afterwards, based on a data collection on social networks, specifically Twitter and Facebook, we will analyze the public debate about hate speech and censorship, especially considering the peaks of discussion about these topics and events associated with them. Lastly, we will identify the major actors and interactions happening at the highest peak of debate about hate speech and censorship, using graphs that present a structural analysis of social networks. The study used bibliographic and documental sources as well as data collected on Facebook and Twitter, which were then analyzed both qualitatively and quantitatively.

Keywords: hate speech; digital platforms; censorship; freedom of speech; social networks.

SUMMARY OF RESULTS

• Our theoretical review combined with the analysis of the platforms revealed that, although there are guidelines for sanctions of accounts that publish offensive content, it is very difficult and complex to detect this type of post;
• The research focused on collecting data about hate speech and censorship, gathering around 11 million posts on Twitter and 34 thousand posts on Facebook between November 2020 and January 2021;
• Regarding the debate about hate speech on Twitter and Facebook, the posts focused on topics such as misogyny and racism. The peak of posts about the topic on both platforms happened on November 20, 2020, Black Awareness Day. The number of posts increased due to the discussion about structural racism, with the news that a black customer had been beaten to death in the previous day at a Carrefour supermarket in Porto Alegre (RS);
• The discourse about censorship and freedom of speech increased due to the sanctions imposed by the platforms for accounts such as that of the former president of the United States, Donald Trump, whose posts questioning the results of the presidential elections were associated with the storming of the United States capitol by extremist groups. The suspension of accounts mobilized the networks to discuss the approach taken by the platforms regarding the censorship of opinions expressed by users, both on Twitter and on Facebook.

Debate about hate speech and censorship on Twitter

Between November 16, 2020 and February 6, 2021, hate speech was the topic of approximately 11.6 million posts on Twitter. The debate was driven by the celebration of Black Consciousness Day, the assassination of a black customer by a security guard at a Carrefour supermarket in Porto Alegre (RS), the storming of the United States Capitol, and the ensuing suspension of the accounts of the former president of the United States, Donald Trump, on social networks. It became evident that the two topics driving the debate on Twitter were hate speech (the debate about structural racism) and freedom of speech (the debate about censorship). As indicated by the literature, the debate about online offenses is always accompanied by questions about the violation of individual freedoms and the censorship imposed by the platforms against opinions expressed by users.

In the period analyzed, other news stories also contributed significantly to this debate. Regarding hate speech, the harassment episode of the congresswoman Isa Penna (PSOL-SP) was permeated by a discussion about the vulnerability of women and mysogyny in different spaces. Regarding freedom of expression, there was a publication by the Ministry of Health about early treatment for Covid-19, which was considered by Twitter to be misleading.

As indicated previously, the data collection about online hate speech or offenses is always accompanied by posts about the users’ freedom to express freely. During the period analyzed, there was an intense debate about the suspension of accounts or the removal of content in social network websites, such as Facebook, Instagram or Twitter itself, due to violations of the platforms’ guidelines. Some posts coming from profiles of politicians or bloggers with partisan, ideological or identity alignments questioned the decisions made by the websites, claiming that the content that was blocked did not violate the community standards or that the reasons for the suspension were not rational, and that those users were actually being targeted by defamatory campaigns started by opposing groups. The suspensions are eventually compared to other cases in which users indicate posts that also contained sensitive content, such as offenses or threats against certain politicians or partisan groups, or the dissemination of fake news or information, but those posts were not removed by the platforms or caused the authors’ accounts to be suspended.

The following figure shows the evolution of the debate on Twitter. Based on the peaks indicated in it, notably on November 20 (about hate speech) and on December 15, 2020 and January 19, 2021 (about freedom of speech), we will conduct a qualitative analysis based on the engagement around the topics. We will take into account the number of shares, likes and followers in each account in order to detect the publications that drove the debate.

Debate about hate speech and censorship on Facebook

This section will present the findings of the analysis conducted on Facebook between November 15, 2020 and February 11, 2021. We collected 21,149 posts about hate speech and 12,773 posts about the topics of freedom of speech or censorship. The evolution of both debates was observed in parallel, considering the direct relationship between the two topics which we mentioned previously. Thinking about this articulation is important for this study because there is a growing demand for mechanisms to curtail hate speech and, similarly, impose limits to freedom of speech.

 

Although the offenses, threats and discrimination that are part of hate speeches have subjective effects, considering the great psychological suffering they produce, their investigation cannot dismiss their political impacts. The silencing and invisibility of vulnerable groups weaken their demands and pleas, removing them from democratic decision-making processes. Therefore, identifying and fighting hate speeches is important for democracy.

This study presented a literature review about the topic, aiming to demonstrate the major understanding about what hate speeches are in the academy and in the legal field, their specificities in the digital environment, and how they are defined by digital platforms in their self-regulation protocols. The bibliographic and documentary investigation revealed that while it is possible to conceptualized gate speeches in theory, it is very complex to define them in concrete situations. The challenge becomes harder in online pages with memes, videos, organized communities, lexical variations and other aspects.

The material demonstrated that, in addition the adoption of international treaties to fight the different forms of discrimination, it is necessary that countries formulate legislation combining their social, historical and especially cultural contexts with the specificities of the online universe. In that sense, the challenge was not easier for the platforms. When explaining their perception of the different manifestations of hate speeches in their Terms of Use and Community Guidelines, the platforms left a series of gaps about how to approach and moderate the authors of that content. One hypothesis on that regard is that global platforms need to know the sociocultural particularities which their users belong to in order to develop viable and more productive solutions when fighting hate speeches.

The empirical analysis in this study consisted in mapping the debate of hate speeches and its relationship with freedom of speech on Twitter and Facebook between November 2020 and February 2021. Even though they are both digital social networks with a significant number of users, including in the Brazilian context, we understand that they have limitations in terms of representativeness, both regarding the amount of digital interactions and the diversity of authors they enable. This analysis offered an incomplete view of the complex ecosystem involved in the creation, proliferation and maintenance of online hate speeches.

Regarding our methodological concerns, one challenge related to the development of search syntaxes – used to collect and classify the data to be analyzed – was the presence of subtleties in the conceptualization and semantic-pragmatic scope of the topic targeted by the study, that is, “hate speech”. Considering that it is very difficult to conclusively and indisputably determine what actually is hate speech, or even offensive speech, and especially considering how this discourse happens in the public debate on social networks, finding linguistic and discursive strategies to analyze this phenomenon proved to be an undeniably flawed task.

Based on the data collection and analysis we conducted, we found that the volume of debate about hate speech and censorship increased during a holiday and based on the repercussion of national and international news. This leads us to believe that the existence of a public agenda and the participation of the news media contributed to the intensification of ongoing discussions in the platforms we investigated. In addition to these elements, the posts about the topic made by public figures who have a large network of followers produced a high volume of interactions, which demonstrates the politically strategical role of public figures in the production and reach of the debate.

Considering the evolution of the online debate in the period analyzed, this study indicated the existence of tensions related to the structural and personal understandings of discrimination. This means that, for some of the authors, this debate is critical to the development of democracy and must be present in the public arena, while for other authors, this discourse destabilizes the political field without producing effective change. Lastly, the sanctions applied by the platforms interfered with the volume of publications and with the participation of certain actors, intensifying the debate about freedom of speech and leading to a discussion that does not question their legitimacy, but their limits.

By identifying the arguments and authors that compose this debate, this study lays the groundwork for new investigations about the dissemination of hate speeches on the networks. Among them are the appearance of new platforms in response to the sanctions, the expansion of the lexical diversity associated with the different possibilities of language, the relationship between the self-regulatory procedures of platforms and the sociocultural contexts in which they are applied, and the strategies of social actors to disarticulate online discrimination.

ALAVA, S.; FRAU-MEIGS, D.; HASSAN, G. Youth and violent extremism on social media: mapping the research. Paris: Organização das Nações Unidas para a Educação, a Ciência e a Cultura, 2017.

BARBERÁ, P. How social media reduces mass political polarization: evidence from Germany, Spain, and the US. In: AMERICAN POLITICAL SCIENCE ASSOCIATION ANNUAL MEETING, 2015. Proceedings. American Political Science Association, São Francisco, 2015. Available at http://pablobarbera.com/static/barbera_polarization_APSA.pdf. Accessed on: February 26, 2021.

BARBOSA, B.; MARTINS, H.; VALENTE, J. Fake news: como as plataformas enfrentam a desinformação. São Paulo: Intervozes, 2020.

BENESCH, S. Dangerous speech: A proposal to prevent group violence. Dangerous Speech Project proposal paper. February 23, 2013. Available at: https://dangerousspeech.org/wp-content/uploads/2018/01/Dangerous-Speech-Guidelines-2013.pdf. Accessed on: March 01, 2021.

BOWMAN-GRIEVE, L. Exploring Stormfront: a virtual community of the radical right. Studies in Conflict and Terrorism, v. 11, n. 31, p. 989-1007, 2009.

BRASIL. Decreto-lei nº 2.848, de 7 de dezembro de 1940. Código Penal. Available at: http://www.planalto.gov.br/ccivil_03/decreto-lei/del2848compilado.htm. Accessed on: February 26, 2021.

BRASIL. Decreto nº 65.810, de 8 de dezembro de 1969. Promulga a Convenção Internacional sobre a Eliminação de todas as Formas de Discriminação Racial. Available at: http://www.planalto.gov.br/ccivil_03/decreto/1950-1969/D65810.html. Accessed on: February 26, 2021.

BRASIL. Constituição da República Federativa do Brasil. Brasília, 1988. Available at: http://www.planalto.gov.br/ccivil_03/constituicao/constituicaocompilado.htm. Accessed
on: February 25, 2021.

BRASIL. BRASIL. Law 7,716, Thursday, January 5, 1989. Define os crimes resultantes de preconceito de raça ou de cor. Available at: http://www.planalto.gov.br/ccivil_03/leis/l7716.htm. Accessed on: February 26, 2021.

BRASIL. Decreto nº 592, de 6 de julho de 1992a. Promulga Atos Internacionais. Pacto Internacional sobre Direitos Civis e Políticos. Available at: http://www.planalto.gov.br/ccivil_03/decreto/1990-1994/d0592.htm. Accessed on: February 26, 2021.

BRASIL. Decreto nº 678, de 6 de novembro de 1992b. Promulga a Convenção Americana sobre Direitos Humanos (Pacto de São José da Costa Rica), de 22 de novembro de 1969. Available at: http://www.planalto.gov.br/ccivil_03/decreto/d0678.htm. Accessed on: February 26, 2021.

BRASIL. Declaração e programa de ação [da III Conferência Mundial de Combate ao Racismo, Discriminação Racial, Discriminação Racial, Xenofobia e Intolerância Correlata]. Brasília, DF: Ministério da Cultura, 2001. Available at: http://www.unfpa.org.br/Arquivos/declaracao_durban.pdf. Accessed on: February 26, 2021.

BRASIL. BRASIL. Law 10,741, Wednesday, October 1, 2003. Dispõe sobre o Estatuto do Idoso e dá outras providências. Available at: http://www.planalto.gov.br/ccivil_03/leis/2003/l10.741.htm. Accessed on: February 26, 2021.

BRASIL. BRASIL. Law 12,735, Friday, November 30, 2012. Altera o Decreto-Lei nº 2.848, de 7 de dezembro de 1940 ‒ Código Penal, o Decreto-Lei nº 1.001, de 21 de outubro de 1969 ‒ Código Penal Militar, e a Lei nº 7.716, de 5 de janeiro de 1989, para tipificar condutas realizadas mediante uso de sistema eletrônico, digital ou similares, que sejam praticadas contra sistemas informatizados e similares; e dá outras providências. Available at: http://www.planalto.gov.br/ccivil_03/_ato2011-2014/2012/lei/l12735.htm. Accessed on: February 26, 2021.

BRASIL. BRASIL. Law 12,965, Wednesday, April 23, 2014. Estabelece princípios, garantias, direitos e deveres para o uso da Internet no Brasil. Available at: http://www.planalto.gov.br/ccivil_03/_ato2011-2014/2014/lei/l12965.htm. Accessed on: February 26, 2021.

BROWN, A. What is so special about online (as compared to offline) hate speech? Ethnicities, v. 18, n. 3, p. 297-326, 2018.

BRUGGER, W. Proibição ou proteção do discurso do ódio? algumas observações sobre o direito alemão e o americano. Revista de Direito Público, n. 15, p. 117-136, 2007.

BUYSE, A. Words of violence: “fear speech,” or how violent conflict escalation relates to the freedom of expression. Human Rights Quarterly, v. 36, n. 4, p. 779-797, 2014.

COHEN-ALMAGOR, R. Fighting hate and bigotry on the Internet. Policy and Internet, v. 3, n. 3, 2011. pp.: 1-26.

CONSANI, C. F. Democracia e os discursos de ódio religioso: o debate entre Dworkin e Waldron sobre os limites da tolerância. Etic@, v. 14, n. 2, p. 174-197, 2015.

DWORKIN, R. O direito da liberdade. São Paulo: Martins Fontes, 2006.

DWORKIN, R. Foreword to Extreme Speech and Democracy. In: HARE, I.; WEINSTEIN, J. (eds). Extreme Speech and Democracy. New York: Oxford University Press, 2009, p. v-ix;

FARIS, R.; ASHAR,, A.; GASSER, U.; JOO, D. Understanding harmful speech online. Berkman Klein Center Research Publication, n. 2016-21, 2016. Available at: https://papers.ssrn.com/sol3/papers.cfm?abstract%5Fid=2882824. Accessed on: March 01, 2021.

FERNANDEZ-MATAMORO, A.; FARKAS, J. Racism, hate speech, and social media: a systematic review and critique. Television & New Media, v. 22, n. 2, p. 205-224, 2021.

FORTUNA, P.; SOLER-COMPANY, J.; WANNER, L. Toxic, hateful, offensive or abusive? What are we really classifying? An empirical analysis of hate speech datasets. In: CONFERENCE ON LANGUAGE RESOURCES AND EVALUATION, 12, 2020. Proceedings. European Language Resources Association, Marseille, 2020. p. 6786-6794.

FUNDAÇÃO GETULIO VARGAS. Diretoria de Análise de Políticas Públicas. Monitor de temas. Rio de Janeiro: FGV DAPP, 2015. Available at: http://bibliotecadigital.fgv.br/dspace/handle/10438/15262. Accessed on: February 26, 2021.

GILLESPIE, T. A relevância dos algoritmos. Parágrafo, v. 6, n. 1, p. 95-121, 2018.

JURNO, A. C.; D’ANDRÉA, C. (In)visibilidade algorítmica no “feed de notícias” do Facebook. Revista Contemporânea, v. 15, n. 2, p. 463-484, 2017.

LUCCAS, V. N.; GOMES, F. V.; SALVADOR, J. P. F.. Guia de análise de discurso de ódio. Rio de Janeiro: Fundação Getulio Vargas, 2020. Available at: https://www.conib.org.br/wp-content/uploads/2019/11/Guia-de-An%C3%A1lise-de-Discurso-de-%C3%93dio.pdf. Accessed on: February 26, 2021.

LUNA, N e SANTOS, G. Liberdade de expressão e discurso do ódio no Brasil. Revista Direito e Liberdade, v. 16, n. 3, 2014, p. 227–255. Available at: http://ww2.esmarn.tjrn.jus.br/revistas/index.php/revista_direito_e_liberdade/article/view/780. Accessed on: March 01, 2021.

NANDI, J. O combate ao discurso de ódio nas redes sociais. Trabalho de Conclusão de Curso (graduação). Universidade Federal de Santa Catarina (UFSC), Campus Araranguá, Graduação em Tecnologias da Informação e Comunicação, Araranguá, 2018. Available at: https://repositorio.ufsc.br/handle/123456789/187510. Accessed on: March 01, 2021.

NERIS, N. (coord); VALENTE, M.; CRUZ, F.; OLIVA, T. Outras vozes: gênero, raça, classe e sexualidade nas eleições de 2018. InternetLab, São Paulo, 2019. Available at: https://www.internetlab.org.br/wp-content/uploads/2019/10/OutrasVozes_2018.pdf. Accessed on 3/10/2021.

ONU (ORGANIZAÇÃO DAS NAÇÕES UNIDAS). Carta das Nações Unidas. São Francisco: ONU, 1945. Available at: https://www.un.org/en/charter-united-nations/index.html. Accessed on: February 26, 2021.
ONU (ORGANIZAÇÃO DAS NAÇÕES UNIDAS). Declaração Universal dos Direitos Humanos. Paris: ONU, 1948. Available at: https://www.un.org/en/universal-declaration-human-rights/. Accessed on: February 26, 2021.

PAREKH, B. Is there a case for banning hate speech? In: HERZ, M.; MOLNAR, P. (eds.). The Content and Context of Hate Speech: Rethinking Regulation and Responses. Cambridge: Cambridge University Press, 2012, p. 37-56.

PEARSON, J.; CHILD, J.; DEWITT, L.; KAHL, D. Facing the fear: an analysis of speech-anxiety content in public-speaking textbooks. Communication Research Reports, v. 24, p. 159-168, 2007.

ROGERS, R. Engajados de outra maneira: As Mídias Sociais – Das Métricas de Vaidade à Análise Crítica. In: OMENA, J. J. (ed.). Métodos Digitais: Teoria e Prática. Lisboa: Editora da Universidade Nova de Lisboa, 2019. p, 73-96.

RUEDIGER, M. (coord.). Nem tão #simples assim: o desafio de monitorar políticas públicas nas redes sociais. Rio de Janeiro: FGV DAPP, 2017. Available at: http://hdl.handle.net/10438/19436. Accessed on 3/1/2021.

SANCHES, D. Construindo grafos de conhecimento (knowledge graphs) para análise de redes sociais: uma análise sobre discurso de ódio em Portugal. In: WORKSHOP ON MEDIA, INFORMATION AND DATA SCIENCE, 1, 2020. Anais. Universidade Federal de Goiás, Goiânia, 2020.

SELLARS, Andrew. Defining Hate Speech. Berkman Klein Center Research Publication, n. 2016-20, Boston Univ. School of Law, Public Law Research Paper, p. 16-48, 2016. Available at: https://ssrn.com/abstract=2882244 or http://dx.doi.org/10.2139/ssrn.2882244. Accessed on: March 01, 2021.

SHEPHERD, A. Extremism, free speech and the rule of law: evaluating the compliance of legislation restricting extremist expressions with article 19 ICCPR. Utrecht Journal of International and European Law, v. 33, p. 62-83, 2017.

SIEGEL, A. Online hate speech. In: PERSILY, N.; TUCKER, J. (orgs.). Social media and democracy. Cambridge: Cambridge University Press, 2020. p.: 56-88.

SILVA, L. R.; BOTELHO-FRANCISCO, R. E.; OLIVEIRA, A. A.; PONTES, V. R. A gestão do discurso de ódio nas plataformas de redes sociais digitais: um comparativo entre Facebook, Twitter e Youtube. Revista Ibero-americana de Ciência da Informação, v. 12, n. 2, p. 470-492, 2019.

SILVIA, L.; MONDAL, M.; CORREA, D.; BENEVENUTO, F.; WEBER, I. Analyzing the targets of hate in online social media. In: Proceedings of the Tenth International AAAI conference on Web and Social Media, 2016. Available at: https://arxiv.org/abs/1603.07709v1.

SIMONS, A.; MUELLER, John. The Dynamics of Internal Conflict. International Security, n. 24, v. 4, p. 187-192, 2001.

SLOAN, R. H.; WARNER, R.. Beyond notice and choice: privacy, norms, and consent. Journal of High Technology Law, v. 14, p. 370, 2014.

WALDRON, J. The harm in hate speech. Cambridge: Harvard University Press, 2012.

WEAVER, S. A rhetorical discourse analysis of online anti-Muslim and anti-Semitic jokes. Ethnic and Radical Studies, v. 3, n. 36, p. 483-499, 2013.

WEINSTEIN, J.; HARE, I. General introduction: free speech, democracy, and the suppression of extreme speech past and present. In: HARE, I.; WEINSTEIN, J. (eds.). Extreme speech and democracy. Nova York: Oxford University Press, 2009, p. v-ix.

Research Coordination
Marco Aurelio Ruediger
Amaro Grassi

Researchers
Dalby Hubert
Danielle Sanches
Eurico O. Matos Neto
Luiza C. Santos
Lucas Roberto da Silva
Polyana Sampaio Barbosa
Renata Tomaz

Technical Review
Renata Tomaz

Graphic Project
Luis Gomes
Daniel Cunha