Bolsonaro laughed at it, and so did I: sarcasm as bonding rhetoric in online conversations about cutbacks in education

Abstract

In light of deliberative theory and based on the concepts of bonding rhetoric and bridging rhetoric (Dryzek, 2010), this article aims to investigate how sarcasm was employed in online conversations on Facebook and YouTube regarding budgetary cutbacks in education made in 2019 by the Ministry of Education (MEC). Sarcasm was the most used argumentative device and was often associated with insults, contributing to a polarized discussion. Sarcastic comments also had low rates of justification, were predominantly declaratory, and largely had a generic or specific target, generating reaction and conflict while strengthening a bond with those who understood the sarcastic meaning behind the message and agreed with the subject who employed it.

Keywords
Online conversation; Deliberation; Rhetoric; Sarcasm; Social Media

Resumo

À luz da teoria deliberativa e partindo dos conceitos de retórica de vínculo e retórica de conciliação (Dryzek, 2010), este artigo tem como proposta investigar de que maneiras o sarcasmo foi instrumentalizado em conversações on-line no Facebook e YouTube sobre cortes orçamentários da educação promovidos em 2019 pelo Ministério da Educação (MEC). Neles, o sarcasmo foi o recurso argumentativo mais empregado, frequentemente associado ao insulto, contribuindo com o acirramento de uma discussão polarizada. Os comentários sarcásticos apresentam baixos índices de justificação, são majoritariamente declaratórios e, em maior grau, possuem um alvo genérico ou específico, gerando um efeito de oposição e conflito, e criando vínculo com indivíduos que compreendem o sentido da mensagem e concordam com o interlocutor.

Palavras-chave
Conversações on-line; Deliberação; Retórica; Sarcasmo; Redes sociais digitais

Resumen

Basado en la teoría deliberativa y partiendo de los conceptos de retórica de vínculo y retórica de conciliación (Dryzek, 2010), este artículo tiene como objetivo investigar de qué forma el sarcasmo ha sido instrumento de conversaciones en línea en el Facebook y YouTube sobre cortes en el presupuestos de la Educación promovidas en 2019 por el Ministerio de la Educación (MEC). En ellos, el sarcasmo fue el recurso argumentativo más utilizado con frecuencia relacionado al insulto, contribuyendo para una fuerte discusión. Las opiniones sarcásticas presentan bajos índices de justificación, son en la gran mayoría simples declaraciones y, muchas veces tienen un objetivo genérico o específico, generando un efecto de oposición y conflicto, y creando vínculo con individuos que comprenden el sentido del mensaje que concuerdan con el interlocutor.

Palabras clave
Conversaciones en línea; Deliberación; Retórica; Sarcasmo; Redes sociales digitales

Introduction

Dryzek (2010) argues that rhetoric is a vital tool to ensure that diverse groups are represented when deliberative democracy is not restricted to a single forum but extended to various formal and informal arenas, including online conversations. In this sense, rhetoric would have the potential to “enable effective communication between differently situated actors and can both establish and maintain deliberative systems”¹ (DRYZEK, 2010, p. 320).

The author also points out two distinct types of rhetoric: bridging and bonding. The first one seeks to reconcile opposing groups, finding shared points where the parties can reach some consensus; the second aims at strengthening bonds that had already been established between members from the same group. Although bonding rhetoric can be used in favor of minority groups to “generate a degree of solidarity that will enable [the oppressed group] subsequently to enter with confidence into a larger public sphere” (DRYZEK, 2010, p. 331), it also carries a negative aspect: widening the gaps between groups, reinforcing extremist views. Thus, bonding rhetoric can act as a way of recognizing and reaffirming similarities while deepening differences between groups.

Following Perelman and Olbrechts-Tyteca’s New Rhetoric (2014, p. 189), one of the ways to employ rhetoric discursively is through figures of speech: “forms of expression that do not fit into the common” or a “style feature that allows a person to express oneself in a simultaneously free and coded way” (REBOUL, 2004). Among these figures of speech, this article focuses on sarcasm, a type of irony that carries a high level of injury (CHERUBIM, 1989).

Sarcasm is peculiar for having a double role: it offends while also causing laughter among those who agree with the injury². Therefore, also based on Dryzek’s definition of bonding rhetoric (2010), sarcasm is a figure of speech doubly effective in creating bonds – it distances targets through injury and brings peers closer through its implicit ironic humor.

Irony is a key element of sarcasm that results from the overlap between connotative and denotative meanings: “Irony is a discursive strategy that cannot be understood separately from its embodiment in contexts, [...] it is not solely about who may use irony (and where, when, and how) but who may interpret it”. (HUTCHEON, 2000, p.135).

Sarcasm thus strengthens bonds between people who hold the same point of view, given that its ironic aspect is dependent on a pre-established perspective within a group. Hence, its members can understand the irony based on their shared worldviews.

With this in mind, this article explores how sarcasm was instrumentalized in online conversations about the budgetary cutbacks in education by the Ministry of Education (MEC, Ministério da Educação) in 2019.

On April 30, 2019, Brazilian public universities were notified about a 30% cut in their operating budgets. There were reactions from students, teachers, and public servants, causing a fierce debate regarding the importance of public education. It was widely broadcasted in the media, which covered presidential statements, speeches from experts and government leaders, as well as the protests organized by students and organizations in the field.

This debate, especially the one that took place on Facebook and YouTube, is the focus of the research this article is a part of, which is carried out by the Communication and Political Participation Research Group (Compa, Grupo de Pesquisa Comunicação e Participação Política), from the Federal University of Paraná (UFPR, Universidade Federal do Paraná). Its main goal is to investigate the characteristics of conversations surrounding controversial topics in digital networks.

The first topic of the article describes the data collection and analysis methodology and techniques used, the corpus, and the results. It was identified that sarcasm was the most used rhetorical device by users against or in favor of the cutbacks, and 25% of the time it also included some insult.

In the later sections, the study analyzes and discusses these results based on simple and cross rates of the variables to investigate how this rhetorical device highlights or interacts with other form and content-related aspects of these conversations. In agreement with Dryzek (2010), bonding rhetoric emerged as a symptom or expression of the intensification and polarization of opinions in the corpus.

Corpus and methods

The technique chosen to organize, code, and investigate the comments is Content Analysis (CA). This rule-based set of operations supports large amounts of data and enables us to achieve the goals set for this project based on descriptions and inferences. As Krippendorf (2004) details, the set of procedures and tools that comprise CA enables researchers to understand a particular phenomenon. This systematic and objective analysis involves careful inspection of interactions and can summarize the content of messages in a scientifically rigorous manner (NEUENDORF, 2012). In the interpretive phase, this statistical treatment transforms the original text “to create new information” (BAUER, 2008, p. 191).

In conversation studies, this technique can be employed to understand the underlying structure of natural conversations in different contexts between two or more participants – preserving communication details and characteristics to avoid the loss of information from the source material (KRIPPENDORF, 2004, p. 67). The procedures described herein were also based on the principles described by Sampaio and Lycarião (2018) regarding three essential characteristics of a CA study: validity, replicability, and reliability.

Validity is the “epistemological adequacy between the research goals and the tools used to identify the investigated phenomenon” (SAMPAIO; LYCARIÃO, 2018, p. 33) and relies on the connection between the theories that support the analysis and the choice of variables that compose the codebook.³ The present study is based on the precepts of everyday conversations in Stromer-Galley (2002, 2007) and the criticisms of deliberation and rational argumentations in Young (2001). In addition, the variables concerning persuasive and argumentative strategies and figures of speech like sarcasm are grounded on the works of Cherubim (1989), Hutcheon (2000), Reboul (2004), and Dryzek (2010).

While it can be perceived as a critical component of reliability, replicability involves ensuring that scientific peers can wholly or partially reproduce the study based on a transparent description of the steps of the process and the materials provided by the researchers. Another component that helped ensure this was the codebook developed by Compa and adapted for this study. It includes the variables that allow researchers to make inferences based on the comments, with emphasis on the form and content of the text and the presence of conversation or interaction, in addition to the use of persuasive devices and rhetorical strategies.

Finally, the steps set by Neuendorf (2012) and reinforced by Sampaio and Lycarião (2018) were followed to further ensure reliability. After extensive training from database samples in December 2019, five coders passed the test with the Kappa (k) reliability index advocated by Landis and Koch (1977) for data analysis in multivariate categories⁴.

In the coding involving all the study variables, a Free Kappa of 0.71 was achieved, equivalent to 78.4% of agreement – a substantial (LANDIS; KOSH, 1977) and suitable level to continue the research. In isolation, the variable on the presence or absence of sarcasm also had substantial reliability, with an agreement index based on a Free Kappa of 0.62.

The database encompasses 14,533 social network comments – 7,924 from Facebook and 6,609 from YouTube. All of them included text, i.e., comments that consisted solely of emojis or emoticons, laughter, punctuation marks, or onomatopoeias were excluded.

The publications involved the same topic: the cutbacks in the public university budgets signed off by the Federal Government at the end of April 2019, when the Ministry of Education, under the command of Abraham Weintraub, determined that 30% of the total budget destined to all federal educational institutions (universities and institutes) should be blocked. The comments were collected from 33 news items posted on official newspaper profiles on Facebook⁵ between April 29 and May 13, 2019: O Globo (12 posts), Estadão (8 posts), Gazeta do Povo (8 posts, including opinion articles), and G1 (5 posts).

In addition, the repercussion also yielded analyses and opinion pieces from content creators with a diverse range of political perspectives. The YouTube comments were collected from six videos, three in favor and three against the budgetary cutbacks.⁶ Those in favor of the cutbacks in Education were Nando Moura (3.14 million subscribers at the time of data collection in September 2020), Mamãefalei (2.67 million subscribers), and Diego Rox Oficial (1.2 million subscribers). The videos with arguments against the cutbacks were Meteoro Brasil (850 thousand subscribers), Henry Bugalho (556 thousand subscribers), and O Historiador (250 thousand subscribers). All content creators were chosen based on the topic of the videos when the subject was highly publicized and by their relevance in the platform, according to the number of subscribers. As shown, there was a clear discrepancy in popularity between those who approved and disapproved of the government measure.

The present article only encompasses user comments in which the binary variable regarding the argumentative device “sarcasm” was coded as “present”. The final corpus comprises 4,411 posts, 30% of the entire database.

Sarcasm was the most used argumentative device on both Facebook and YouTube. Second place is the insult (2,759), followed by analogy (623), narration (458), and threat (377), as shown in Graph 1.

Graph 1
Most used argumentative devices

Given that each argumentative device was represented by a binary variable (i.e., “present” or “absent”), the coders were able to identify more than one of these elements in the same comment, regardless of which was the dominant one.

In this regard, the most relevant crossing was the one with the two most frequent devices, that is, sarcasm and insult. Among these, it can be seen that 1,119 comments that used one also employed the other. In other words, 25% of the times in which sarcasm was present, it also included a high level of aggression and use of expletives, indicating the quality of the debate in the two social networks.

Of the 4,411 comments that employed sarcasm, 2,857 (65%) were on Facebook, while 1,554 (35%) were on YouTube. Although the initial corpus of 14,533 units included more comments from Facebook (55%), sarcasm is indeed more present on that network than on YouTube. Regarding the commentator’s position as for cutbacks in education funds, individuals against the contingency showed more sarcasm than those in favor, as shown in Graph 2⁷.

Graph 2
Commentator positioning sarcasm

It was also possible to notice that most sarcastic comments do not justify their position based on an explicit source. Only 975 sarcastic comments inform where they got their arguments from, as seen below:

so I suggest you get to know more about what is done with your (our) money; see, for example, the FAPESP magazine, which always brings high-level results financed with state resources. http://revistapesquisa.fapesp.br/ spoiler: there are no performances with naked people.

The form of the comment concerns how the debater supports his argument. A significant portion of the comments was construed to establish conflict (Table 1)⁸, but most were predominantly declaratory, i.e., they only stated something, with no other clear intention. The lack of other purposes in sarcastic comments is in line with the very semantic nature of sarcasm, which is a type of insult disguised as a statement, a type of indirect argumentation (PERELMAN; OLBRECHTS-TYTECA, 2014). Its double meaning is at first merely declarative but has an underlying meaning aimed at an attentive audience.

Table 1
Sarcasm versus Form

Form	Sarcasm
Declaration/statement	2615
Opposite point of view	78
Clarification	159
Questioning	485
Solution proposal	77
Call to action	28
Establishing conflict	943
Does not apply	25
Total	4411

• Source: author’s own elaboration.

The relationship between sarcasm and the type of argument used in the analyzed comments (Graph 3) and the rhetorical strategy employed to support the arguments (Graph 4) was also investigated.

Graph 3
Arguments used

Graph 4
Sarcasm versus Dominant persuasive strategy⁹

Out of the total number of sarcastic comments, most do not use a central argument. Among the 1,543 that have a central argument, those who employ sarcasm most frequently are commentators against the cutbacks, who base their arguments on the inability or incompetence of the government concerning education. On the other hand, debaters who justify their arguments in favor of the cutbacks by saying that resources should be directed to other levels of education (rather than universities) are the ones who least make sarcastic comments.

Since they generally question opponents and are directed at a specific target, sarcastic comments are more frequent among subjects who employ the persuasive strategy of appealing to source credibility. In all, 1,737 sarcastic comments (40%) also had a common adversary or enemy. For most of the comments that mention an adversary (1,554), this enemy was specific, not generic, that is, the injury was directed at someone, be it the interlocutor or a public agent or body. That will be shown in the examples of the following section, which deals with the main research results.

Discussion of results

The first cross-analysis concerns the deliberative degree of the discussion based on rationality and reciprocity, and how it relates to sarcasm. Argumentations rely on rationality, i.e., the capability of justifying one’s claims, and reciprocity, which indicates that the participants are listening to other people and engaging with their arguments (Habermas, 1996). The majority of sarcastic comments are not justified (77.9%) and are mostly comprised of declarations/statements (59.2%), followed by comments that seek to establish conflict (21.3%). These comments illustrate this relationship:

What previous governments called education was simply raising cattle.

Look, even my dog is aware that Philosophy is important to society... lol

And the stupid bolsominions think Bozo will send this money to their accounts. Pack of donkeys

Let’s not criticize, people. Soon someone will remember that Lula is barely literate.

Let’s wait for the expert to say how much better than Harvard she is.

The examples above correspond precisely to the structures that go against reflexivity (marked by an opposing view and questioning), decision-making (indicated by a solution proposal or call to action), and engagement, which encompasses indications of an opposite point of view and clarification (STROMER-GALLEY, 2002, 2007).

Graph 2 shows that sarcasm is slightly more frequent among commentators against the cutbacks, identified as progressive in this debate. Historically, the left is characterized by a “truncated” discourse, which would prevent the establishment of a dialogue with a broader audience. In this case, paradoxically, the progressive field often abandons complexity and instead resorts to a culture of “wokeness”. In both cases, they also employ devices that hinder dialogue. On the other hand, it is worth mentioning that the strategies leveraged by conservatives are likewise not aimed at enriching the discussion. By observing the corpus, conservatives justify their positions only 5.48% of the time, whereas progressives do it 6.67% of the time; moreover, conservatives are more prone to insulting their opponents (3.83% against 2.46% for progressives). The most commonly used form for both groups is the declaration/statement (11% for both).

Sarcasm is also used in online discussions to deny dialogue with opponents, creating a barrier to those who do not understand the meanings shared by the interlocutors or those who oppose the commentator. Graph 4 shows what the interlocutors appeal to when writing their comments. It points towards a convergence of sarcasm and critical rhetoric/appeal to source credibility, the one that focuses on criticizing one’s opponents while also indicating the low rate of seductive/threatening rhetoric, which seeks to convince the interlocutor using persuasion.

The negative character of sarcasm embodies the bonding rhetoric described by Dryzek (2010). In this case, the goal is to: widen the gap between opposing groups by insulting a real interlocutor present in the discussion (as in “The university is so good that you didn’t even learn how to read... the article is saying that the universities are the ones that don’t even provide minimum conditions... do you understand now?); target agents that are absent from the debate (as in “Bolsonaro’s organized crime is destroying education, stealing public money, long live Bolsonaro’s organized crime with the support of Globo”); or when employed by a commentator who is in favor of the cutbacks (as in “Students? Professors? No, no. Militants!”). In the cases, sarcasm disqualifies opponents, excluding the possibility of argumentative exchanges centered around education. Not surprisingly, sarcasm and insult overlap in a considerable portion of the comments (25%), as shown in Graph 1.

Moreover, as mentioned, sarcasm strengthens the bond between people with the same political views and beliefs. In these cases, the argumentative device works to both cause laughter among allies and declare one’s approval of specific ideas, as in “For people who believe that Olavo de Carvalho is a philosopher and an intellectual, that the Earth is flat, that vaccines are bad, that 35% of 100 are 3.5, that the bozos are honest and that SpongeBob is dating Patrick, it’s normal to slander common sense!”

Thus, the difference between the two roles that bonding rhetoric plays in the use of sarcasm lies, on the one hand, in the use of insults and, on the other hand, in irony disguised as mockery. It is also possible to find both in the same comment, as seen below, in a comment criticizing the National Union of Students (UNE, União Nacional dos Estudantes): “Is UNE an institution? Bolsonaro laughed at this and so did I”.

Another noteworthy fact refers to the 1,554 comments containing a specific common enemy, pointing out the target of sarcasm while also distancing the commentator from it. Among commentators favorable to the cutbacks, sarcasm is often directed at previous governments or professors with opposing political views (in relation to the commentator) and who would supposedly encourage the so-called “balbúrdia” (lit. “ruckus”, an expression that was widely used at the time to refer to reckless behavior that was supposedly encouraged in universities). However, sarcasm is more prevalent among commentators against the cutbacks, who commonly link the education crisis to the 2018 presidential race, which elected Jair Bolsonaro (at the time a member of the PSL party): “Did any dumb fuck believe he was going to invest in primary education? He said he would invest in the base... his own political base.” This opposition, with specific targets highlighted in the comments, reflects the polarization of the Brazilian political debate.

Final considerations

After analyzing comments from publications on YouTube and Facebook about the education budgetary cutbacks in 2019, results showed that sarcasm was the primary argumentative device used and was accompanied 25% of the time by some insult. These sarcastic comments were mainly posted on Facebook and mostly by individuals against cutbacks in education. However, this discrepancy is insignificant compared to people with neutral and favorable views.

These data and the other aspects presented in the discussion of the results of this article suggest that sarcasm acted as an essential argumentative tool for social network users during the analyzed period. The relationship between sarcasm and the different rhetoric operators is a promising niche in studies on online conversations and deliberation, given its potential to unveil how these processes take place in polarized contexts – as was the case with the entire discussion about cutbacks in education in Brazil in 2019.

More specifically, the concept of bonding rhetoric conceived by Dryzek (2010) proved to be a fruitful perspective to analyze everyday conversations, especially in the online debate around controversial topics. In concordance with what the author argues, bonding rhetoric (exemplified here by the use of sarcasm) was an indication of the intensification of polarized opinions.

Future analyses that bridge the literature on deliberation/online conversations and recent discussions on rhetoric could focus on responses and reactions to sarcastic comments, assessing whether this resource encourages aggression between people with opposing opinions or serves as “bait” to establish other types of bonding among the like-minded, in addition to confirming a self-centered humor that echoes their own perspectives. It would be possible to verify, for example, if sarcasm can create poles of interaction and what their characteristics are, given that it could indicate a safe starting point for a discussion depending on how other commentators engage with it. In the rhetorical sphere, other points to be investigated are whether and how rhetoric can be used to improve deliberation, as theorized by Dryzek (2010), either through the presence of bridging rhetoric or the use of bonding rhetoric by non-majority groups as an instrument that can strengthen arguments to be presented in public debates.

• 1
  All the references written in Portuguese or other languages have been freely translated to English.
• 2
  In this article, the term “injury” is distinguished from the term “insult”; “injury” represents the offensive content of sarcasm, while “insult” is any type of offense and represents a distinct variable in the research codebook this article is a part of.
• 3
  Codebook available at: https://bdc.c3sl.ufpr.br/handle/123456789/75. Accessed on: 7 nov. 2022.
• 4
  Based on their own experiences, the authors defined an agreement standard to support statistical descriptions that was also adopted for this study. It can be summarized as follows: “Poor” (no agreement index), “Slight” (0.00 to 0.20), “Fair” (0.21 to 0.40), “Moderate” (0.41 to 0.60), “Substantial” (0.61 to 0.80) and “Almost Perfect” (0.81 to 1.00).
• 5
  Grupo Globo is the leading news outlet regarding digital audience according to Comscore as of May 2020, including platforms (O GLOBO, 2019). According to the same ranking, Gazeta do Povo, a news outlet from the state of Paraná, was the most read newspaper in Brazil in October 2018 (the month of the presidential elections for that year) and is also known nationwide for its transition from print to digital-only. The leading outlet as per the number of visits, Folha de S.Paulo, was not included due to its decision to not have a Facebook page since February 2018. Instead, O Estado de S. Paulo was chosen because it is the main news outlet in terms of mobile users in the country (ESTADÃO, 2020).
• 6
  The list of news and videos where the data was collected is available at: <https://bit.ly/3iFXkF3>. Access on: 17 sep. 2020.
• 7
  Pearson’s chi-square and Spearman’s correlation tests were performed, indicating that there is no relationship statistically significant between data regarding position and the use of sarcasm. The data - p-value of 0.004 and significance of 38% - are not completely independent, but do not present a linear correlation, i.e., one does not increase nor decrease the other. However, the purpose of this graph is only to present the frequency of use of sarcasm in this dataset, rather than the statistical relationship between them.
• 8
  Chi-square test of 597.100, with 6 degrees of freedom and statistical significance of 99.9%.
• 9
  Chi-square test of 891.450, with 3 degrees of freedom and statistical significance of 99.9%.

Referências

• BAUER, M. W. Análise de conteúdo clássica: uma revisão. In: BAUER, M.; GASKELL, G. (ed.). Pesquisa qualitativa com texto, imagem e som: um manual prático. Petrópolis: Vozes, 2008.
• CHERUBIM, S. Dicionário de figuras de linguagem São Paulo: Editora Pioneira, 1989.
• DRYZEK, J. S. Rhetoric in democracy: A systemic appreciation. Political Theory, v. 38, n. 3, p. 319-339, 2010.
• ESTADÃO. Quem Somos. Estadão, 2020. Disponível em: <http://patrocinados.estadao.com.br/medialab/about-me/>. Acesso em: 19 de set. 2022.
  » http://patrocinados.estadao.com.br/medialab/about-me/
• HABERMAS, J. Between facts and norms Cambridge, MA: MIT Press, 1996.
• HUTCHEON, L. Teoria e política da ironia Belo Horizonte: Ed. UFMG, 2000.
• KRIPPENDORF, K. Content analysis: an introduction to its methodology, 2. Thousand Oaks: Sage Publications, 2004.
• LANDIS, R. J.; KOCH, G. G. The Measurement of Observer Agreement for Categorical Data. Biometrics, v. 33, n. 1, p. 159-174, 1977.
• NEUENDORF, K. A. The content analysis guidebook California: Sage Publications, 2012.
• GLOBO, O. O GLOBO termina 2019 como líder no país. O Globo, 2020. Disponível em: <https://oglobo.globo.com/economia/o-globo-termina-2019-como-lider-no-pais-1-24205934>. Acesso em: 19 de set. 2022.
  » https://oglobo.globo.com/economia/o-globo-termina-2019-como-lider-no-pais-1-24205934
• PERELMAN, C., OLBRECHTS-TYTECA, L. Tratado da argumentação: a nova retórica. 3. ed. São Paulo: Ed. Martins Fontes, 2004.
• REBOUL, O. Introdução à retórica São Paulo: Ed. Martins Fontes, 2004.
• SAMPAIO, R.C; LYCARIÃO, D. Eu quero acreditar! Da importância, formas de uso e limites dos testes de confiabilidade na Análise de Conteúdo. Revista de Sociologia Política, v. 26, n. 66, p. 31-47, 2018.
• STROMER-GALLEY, J. New voices in the public sphere: a comparative analysis of interpersonal and online political talk. The Public, v. 9, n. 2, p. 23-42, 2002.
• STROMER-GALLEY, J. Measuring deliberation’s content: a coding scheme. Journal of Public Deliberation, v. 3, n. 1, 2007.
• YOUNG, I. M. Comunicação e o Outro: Além da Democracia Deliberativa. In: SOUZA, J. (ed.). Democracia hoje: novos desafios para a teoria democrática contemporânea. Brasília: Universidade de Brasília, 2001.

Edited by

Editor: Maria Ataide Malcher

Editorial assistant: Weverton Raiol

Publication Dates

• Publication in this collection
  05 Dec 2022
• Date of issue
  2022

History

• Received
  22 Dec 2020
• Accepted
  11 May 2022