Mediatised engagement with the environmental debate on Twitter

ABSTRACT

This article discusses the engagement with civic-political issues from news comments on Twitter and how these users relate to the topics under debate, which can be based on moral values or personal political ideology. This essay deals with participatory culture, a connection between individuals and media content; public engagement and political participation based on media practices; and how moral emotions would affect public engagement – such as participatory culture, a connection between individuals and media content; public engagement and political participation based on media practices; and the concepts of moral emotions. This work analysed 21,327 tweets in Portuguese. As a result, the main narratives in the collected tweets dealt with public and political figures, such as the Brazilian President, the Brazilian Minister of Environment, the activist Greta Thunberg, and the European Union. The results are consistent with previous studies, in which public figures remain relevant to the climate change discourse. Such results justify that public debate on the environment is more focused on politics and political action than ecology.

KEYWORDS:

• Engagement
• climate change
• moral foundations
• public
• media

Introduction

Climate change is a complex and multifaceted issue that has been widely discussed and debated in society over the past few decades. People’s understanding and representation of climate change can vary depending on their level of awareness, education, and exposure to information. In addition, personal beliefs and moral values in shaping how climate change is represented.

Aiming to understand how to generate engagement for the climate change debate by young people, we identify that the debate over the environment occurs in different ways from those that appear regularly in the public debate promoted by the traditional media. This article focuses on discussing the relationship between morale and engagement from twitter comments about environmental news. We analysed publications from Brazil and Portugal through keyword collection.

This discussion is part of a research project, called ‘Youth Engagement with Sustainable Development Goals: The Choice of Moral Arguments in the News for Use in Debate’ which aims to understand how Lusophone youth engaged with news related to climate change based on their innate moral attributes. The project involves three Portuguese-speaking countries, namely Brazil, Mozambique¹ and Portugal. With this purpose, the project analyses posts on Twitter related to environmental news, the application of a psychometric survey about the protruding moral attributes, and focus groups (to be held in 2023 in those countries). The preliminary data collected in the focus groups indicate that the young people questioned want to engage in tackling the climate crisis, but need information from the media to decide how to do so. That is, they reject news that only highlights the climate disaster, and engage with narratives that call to action in the public space.

The main focus of this article concerns the analysis of gathered tweets, which provided a more targeted discourse on the key players involved in the environmental debate, particularly those held accountable for the current climate crisis. Furthermore, our observations revealed significant political polarisation among the comments, subsequently influencing how users assessed the issues being discussed within the tweets. Therefore, this article will present part of the theoretical framework from the original study, such as moral attributes, and the tweets collection methodology.

The analysis of the collected tweets is focusing on the relationship between themes, moral values and personalities most likely to be mentioned by users for the environmental public debate in Portugal and Brazil. Finally, there will be considerations between the concepts and conclusions.² This article is based on a previous paper in Portuguese (Costa, Capoano and Balbé, 2022), where some data from the original project might be duplicated.

Theoretical framework

When environmental public debate migrates to digital environments, like social networks and platforms, it is necessary to highlight the concept of engagement. The engagement is defined such as: maintaining a unique focus of cognitive and visual skills of two or more participants in a situation (Goffman 1966, 8); when participants establish, remain involved and end an intentional connection (Sidner and Dzikovska 2022, 10); the value that a participant in an interaction attributes to the objective of being together with others and maintaining such interaction (Poggi 2007); or shared and coordinated activities in which participants begin, maintain, put together, abandon, suspend, resume or end an interaction (Bohus and Horvitz 2009, 25).

These definitions demonstrate how the concept of engagement supports the reflection upon public debate inserted in communication outlets and digital platforms, as they identify the composition of a collective environment for exposure and exchange of ideas, in person or symbolically, among different individuals willing to do so. Jenkins, who coined the expression ‘Convergence Culture’ (2009) to describe flows of individuals and content through different media supports, speaks of a participatory culture (Jenkins and Carpentier 2013), performed by means of the web tools, for the sake of democratisation and diversification of ideas.

Such a viewpoint is not a consensus, not even to Jenkins, when he points out that the companies which control the platforms capture and price the participatory energies of the public (Jenkins and Carpentier 2013). Although digital networks allow quick mobilisation, action cycles and some new forms of collectivity, it is unclear that they maintain political effort or long-term public debate (Couldry 2015).

As for the investigation that originated this article, about how the young relate to the debate on climate change, the construal of empathic engagement is used (Hall et al. 2005), when fostering emotional involvement with the intention of creating a coherent cognitive system and an emotional experience that results in empathic relations. In this case, the researched emotions are secondary, social, moral or high emotions (Damásio 2018), developed within social organisations and which define what it is like to be human, such as admiration, contentment and morality. Our specific focus, moral attributes (Graham et al. 2013), are emotions related to values, obtained throughout the lives of individuals, but with biological and native origin. Like primary emotions, they are triggers of action and reaction, both of material and symbolic actions, such as production, consumption and sharing of opinions via social networks.

Our study chose Twitter for the data collection because it is identified as a public perception thermometer (Brown 2015), and because it allows communication and large-scale collaboration of users without geographical borders (Lai and Turban 2008), as we deal with tweets in two countries with the same language. Its interface enables search and aggregation of publications by key terms or hashtags, creating an environment segmented per theme. This simplifies the identification of the most valued content within the social network and the engagement of a larger number of people – who normally do not have a previous friendship, but who meet by thematic compatibility. Even considering their caveats (such as bubbles of interest, echo chambers, fake profiles and bots), it can be seen as a digitised environment of public debate.

On Twitter, it is possible to notice the existence of the relation of users’ influence over others within social networks (Cha et al. 2010). This influential position depends on a variety of dynamics and can be built, for example, through tweets on distinct themes or the opposite, an array of topics. Also, the authors proposed a methodology of how to ‘measure’ influence within the social network by comparing three different levels of influence: the number of profile/user followers, the number of retweets, i.e., the number of user messages that are shared by other users, and finally the mentions, i.e., the number of times the profile/user is cited.

Despite the amount of Twitter messages on climate change, the debate ends up focusing on few media, celebrities and digital influencers (Kirilenko and Stepchenkova 2014). The digital public debate has expanded in recent years, including new (and young) representatives of society. Even so, it is possible that mainstream media, profiles of the media company, journalists and information portals will continue to dominate the topics discussed, which can generate a restrictive effect on the participation of young people in the public debate, inside and outside the social media.

If this is the case, Twitter has been substantiated as a space for the reproduction of media discourse and as a means of political dissemination and promotion, in spite of the high participation of citizens currently – as opposed to European legacy media in the European debate on the environment, in which climate change is treated using faint information, based on non-resilient, scientific and political narratives, while neglecting the role of civil society in adapting to climate change (Tavares et al. 2020). It is the same perception we had from the data collected by our study, as it will be shown next.

In the context of communication about climate change in Spanish, for example, the most cited users are actors of the political class, such as politicians, government pages and political parties, followed by the media (Balbé and Carvalho 2017). In the Portuguese language, the most cited profiles are from mainstream media, in most cases, the profiles of the media company, journalists and information portals, followed by citizens’ profiles.

Finally, moral debates are common in socio-technical networks like Facebook, Twitter, Instagram or YouTube. These platforms are full of hate speech, fake news, misinformation dynamics and moral dilemmas (Capoano and Costa 2021; Costa 2020; Máximo Pimenta et al. 2021). However, these platforms are increasingly being used for news consumption, becoming essential for the informative process for some age groups (Capoano 2021). If digital networks have provided the exposure of privacy in a public environment, a larger number of positions on several issues has generated a growing multiplication of moral dilemmas (Bloom 2011). For instance, which side should I take on the subject in discussion? Should I, do it? What if I do it? What are the arguments? This kind of concern appears in all kinds of subjects. In the digital environment, there are debates about morality in virtually every context. As issues associated with the environment and its concomitants (Carvalho 2011) compete with current affairs as the top of interests (Costa and Capoano 2020), it is necessary to consider the moral values of the public to attract it for debate. If their moral values suggest that the public debate on climate change is not contemplated, young people could reject news about it. Otherwise, moral intuitions can be individual guides for youth to select content, think about it and contribute to the debate, composing public opinion inside and outside social media.

The concept of interpretive association comes from the actor-network theory, from Latour (2012). This theory emphasises the interpretive subject as an actor in a tangle of knowledge networks. The interpretive subject is traversed by information networks that conceive pre-interpretations and judgments, then leading to imitation, counter-imitation, or innovation in interpretive processes (Costa 2021b; Tarde 1978). It should be noted that imitating or counter-imitating is part of the same movement: the imitative contagion that sometimes suggests a current of ideas and opinions, sometimes suggests its opposite – a reflection of the same movement but in the opposite direction (Tarde 1978). Thus, a pre-interpretation or an imitated or counter-imitated judgment, whether moral or fact-based, implies the actor's connection to the network from which the pre-interpretations come (Latour 2012). In most cases, the interpretive subject is just an element that follows a path already taken by others, falling into an interpretive web that limits his or her judgments and considerations. The most creative possibility is neither imitation nor counter-imitation, which are parts of the same contagion. It is innovation, if from the assumptions and pre-interpretations, sometimes imitated and sometimes counter-imitated, result an individualised and subjective synthesis that does not follow the majority. Which is complex, since an interpretative association unifies the actor to his closest networks, conditioning him by social symmetry (Latour 2012).

From a neuroscientific perspective, Graham et al. (2013) define five moral foundations (MFT) which aim to express socio-psychological dynamics in five independent domains: Harm, which refers to suffering and includes the notion of caring for others and compassion; Fairness (Justice), which concerns the global notion of fairness and law, as well as cooperation, competition and cheating; Ingroup (Belonging), which relates to commitment to the group, to self-sacrifice, loyalty and vigilance against betrayal; Authority, which refers to obligations related to hierarchy, obedience, respect and fulfilment of duties; and, finally, Purity, which concerns the physical and spiritual influence on chastity, wholesomeness, and the control of desires. Such a theory identifies moral and modular attributes to human beings, which innately influence individual moralities, as well as nativism, cultural learning, the primacy of intuition and plurality.

As mentioned before, our study’s preliminary results indicate that young people who inquired in psychometric questionnaires have moral damage/care and justice/equity as more salient than the others and the attributes purity/sanctity as less salient (Capoano and Costa 2022). In the collection of tweets in this article, the arguments that deal with justice prevail over the others. This indicates that morality can be used to involve young people in the public debate on climate change, similar to what eco-activists and digital influencers already do.

Methodology

The analysis of Twitter posts was performed using Bardin’s (2017) content analysis from a quantitative and qualitative perspective. Namely through the use of thematic analysis and structural analysis (therefore, we left the formal analysis aside). In the thematic analysis, and through the frequencies, we reach the implicit categorizations in the conversations, as well as some of the associated value judgments, as suggested by this technique. In structural analysis, we studied the co-occurrences and associations between ideas and categories, to finally arrive at the models made possible by the Figures (from 5 to 11) and by Tables 1 and 2 (Bardin 2017).

Table 1. Pearson correlation coefficient between Morality and Greta Thunberg.

Table 2. Pearson correlation coefficient between Morality and Bolsonaro.

To analyse the contents of the tweets, we resorted to NVIVO software to formulate the quantitative and descriptive analysis and the qualitative, focused on moral foundations used. In the quantitative methodology, we used the word grouping statistics and calculated associations among terms by country and by a numerical recurrence of each in their relationship with themes and arguments. In the qualitative methodology, we started with the five moral foundations to record social-psychological dynamics in five independent domains. From them, we realised how the collected tweets relate to the subjects they addressed, both using a numerical analysis and an expressed meaning analysis. Finally, we used the NVIVO to crisscross the moral foundations with the topics used in the tweets.

Data gathering and sampling

Regarding the collection of comments on Twitter, we used automated methodology to analyse data from social networking based on what is commonly referred to as Big Data Science. Thus, comments on environmental and climate change-related news in Portuguese were collected remotely from Brazil and Portugal using the geolocation filter. The data were captured and processed with the help of the AP Exata tool, which connects to Twitter's API and exports them to a database. The data capture period took place between October 1, 2021 and April 1, 2022, collecting 21,338 tweets, all data gathered in Portuguese. In Brazil, data from 148 largest cities were collected. As for Portugal, five points of collection were distributed, three covering the northern, southern and central areas, and two covering each of the archipelagos. First, we chose a set of keywords linked to the debate over the environment, in order to identify Twitter profiles interested in this topic. The terms chosen were ‘climate change’, ‘climate crisis’, ‘environment’ and ‘nature’. At that point, we realised that comments on environmental news were more related to political-ideological aspects and to public opinion spokespeople – as in the case of young Swedish activist Greta Thunberg and the Brazilian president Jair Bolsonaro – than the environmental terms originally used for the collection of tweets. Therefore, we added terms such as ‘climate denialism’, ‘Greta Thunberg’, ‘liberal’, ‘capitalism’, ‘progressive’, ‘conservative’, ‘right’, ‘left’ to the environmental terms of the original collection.

After the compilation of tweets, a second module analysed whether the topic environment was relevant to the profiles in the database. If the content on this topic represented 10% of published tweets, the profile in question was sent to a second database, this time associated with a dashboard. In addition to various outputs such as graphs and word clouds, data can be discharged in CSV format for later applications. After sending the environment-related profiles to the dashboard, their publications go through a third algorithm that evaluates the emotions in users’ speech (Plutchik 1984).

Results

First, the sample collected was unbalanced by country: 20,608 tweets from Brazil and 714 tweets from Portugal. The reasons are, fundamentally, the population dimensions of the three countries and the different digital stages of each country (Costa 2021a).

The theme of climate change and the environment instigates strong oppositions, controversies and dilemmas. As such, the polarity generated, over the period of observation of the data, shows a negative connotation, where negative, offensive and accusatory words and expressions circulate, both from the institutional point of view and the one of the positions of digital speakers (Figure 1). Regarding the expressed emotions on the subject, the collection period revealed a considerable oscillation, in an accumulated temporal, between trust (13.41%) and fear (12.03%), and between sadness (10.65%) and anger (10.16%). Also, concerning the emotions involved, it is possible to observe a steady trend in each of the seven analysed primary emotions; and it is worth mentioning a slight growth of fear over the analysed period. The relationship between fear and climate change has recently been explored (i.e., Balbé and Loose 2020).

Figure 1. Relative frequencies of emotions expressed during the observation period.

Brazil

In the results from Brazil’s IP, we found a predominance of the expression ‘environment’, in three out of four tweets, i.e., over 15.000 cases in 20.608 tweets contain this term. Excluding prepositions, adverbs, linguistic tics and programmed keywords that generated our data collection, we detected that there are strong associations and grounds resorting to the terms ‘Brazil’, ‘government’, ‘crisis’, ‘minister’, ‘health’ and ‘Bolsonaro’, denoting a strong national politicisation of the theme (Figure 2).

Figure 2. The 20 most frequent words from Brazil, excluding prepositions, adverbs, linguistic tics and programmed keywords.

Due to the tweets from Brazil, it appears that the theme of climate change is dominated by a discourse in which national political power is the central element. In the Brazilian context, the conversations begin with climate change the terms ‘government’ and ‘minister’ lead the subject to the theme of national politics and their decisions regarding the problem of climate change.

Of the 20 most repeated words per tweet (excluding keywords), there are eight words that relate to politics in Brazil, having Jair Bolsonaro (the president) as a central figure. The name Jair Bolsonaro appears 926 times associated with the theme of climate change.

The theme related to the Amazon region and its ‘destruction’ and/or ‘deforestation’ is a considerably addressed topics as well. However, there is also a constant criticism of previous and left-wing Brazilian governments in a contrast to criticism of the current government.

In Brazil, the average word written per phrase on Twitter about the environment topic during the period observed is 22.7. The strongest associations between terms (excluding technical terms from programming) come from a relationship between the ‘environment’, the verb ‘help’ and the relation with the ‘minister’ (related to a former environment minister Ricardo Sales). Terms such as ‘environmental sustainability’ or ‘ecology’ interrelate with an ‘environment’ defence discourse, whereas the verb ‘help’, especially in the volume of writings and conversations collected, relates to ‘children’, a ‘disposable’ dimension of the phenomenon and to ‘the elderly’, terms that can indicate concern for the future.

Among the most used arguments, we highlight the ‘climate crisis’ association, with 923 repetitions, where there are three axes of argumentation. The first is that ‘the climate crisis is still ignored by many’; the second is that ‘the climate crisis is a crisis of rights’; and the last one is that ‘the climate crisis is real and we need to tackle the problem’. Moreover, we point out the repetition of the idea of ‘environment destruction’, a thought with 294 repetitions, and the combination ‘Bolsonaro Government’ with 167 repetitions that focus, fundamentally, on attacks upon his management (e.g., ‘we regressed 100 years in everything’ and ‘Bolsonaro’s government is an ecological disaster’).

Portugal

Comparing the results from Brazil with those from Portugal, there are some considerable differences in the analysis of the most used terms. The word ‘climate’, associated with terms such as ‘emergency’, ‘transition’ or ‘action’, is the most representative. Although the average number of words written per sentence on Twitter about the environment theme during the period observed is 22.5 (very similar to 22.7 in Brazil). From 714 tweets collected, there is no level of national politicisation of the climate change theme as it happens in Brazil (Figure 3). Nonetheless, there are some direct mentions of the new Mayor of Lisbon (Carlos Moedas) and the fact that he should do more and better in the city regarding this topic. The term ‘Fairness’, one of the five main axes of the theory of moral foundations that guides this investigation, appears more often in tweets from Portugal (0.23%) than in tweets from Brazil (0.03%).

Figure 3. The 20 most frequent words from Portugal, excluding prepositions, adverbs, linguistic tics and programmed keywords.

While in Brazil, among the most cited personalities related to the subject is the Brazilian politician and then President Jair Bolsonaro (0.16%), in Portugal the Swedish activist Greta Thunberg (0.10%) is the most associated with the subject. But like in the Brazilian case, there is an accusatory and ironic tone that occupies the moral foundations invoked around the young activist, considering her, in a great majority of cases, negatively. Among the most frequent terms, we found the expression ‘climate’ (feminine, singular), which relates to ‘emergency’, ‘crisis’, ‘transition’, or ‘action’ (all feminine words in Portuguese). In Portugal, it is to stand out that the focus on ‘climate’ is superior to the focus on ‘environment’.

The word ‘climate’ was mentioned 443 times in publications from Portugal, with a strong relationship between climate emergency and the extreme drought in the country during the data collection period. In addition, several relationships are established between climate urgency denialism and right-wing parties in general. There is not, however, a systematic alignment with the theme – since the arguments follow in various directions and at various angles. The relationship between climate emergency and economic development strays from the debate on energy sources and electric vehicles to government positions regarding companies such as Galp and the like.

The word ‘justice’, which appears 14 times associated with ‘climate justice’, reveals a dynamic movement in this social network that appeals to sectoral and geographical reductions to fight climate change more effectively, but also based on the idea that ‘climate justice is to hold those who pollute responsibly’ (with some recurrence).

The expression ‘climate transition’, on the other hand, is oriented to a ‘more sustainable economy’ discourse, even though examples of irony arise, mixing the various types of agenda for the transition (of gender, of business through energy sources etc.).

Moral positions in tweets from Brazil and Portugal

In the analysis of 20,608 tweets from Brazil, we noticed that the most commonly used moral foundations in the theme of the environment are Fairness (justice) (450) and Harm (damage) (206). The remaining foundations – Authority (51), Purity (17) and Ingroup (14) – have low quantities as grounds used in the moral debate on the environment and climate change (Figure 4). In the analysis of the 714 tweets collected from Portugal, however, we noticed that the most used moral foundations in the theme of the environment are Fairness (justice) (55) and Ingroup (18). The foundations of Harm (11), Purity (3) and Authority (1) emerge with little use in the moral debate on the environment and climate change.

Figure 4. Moral positions in tweets from Brazil [left] and Portugal [right].

The following analysis presents the results of the relationship between the moral attributes, the themes that were most associated with them and the differences between or repetitions among the conversations collected in Brazil and Portugal. The figures reflect the words that were most repeated and most connected in these conversations with the moral attributes under study.

Harm versus Environment. As mentioned earlier, in the theory of moral foundations, Harm refers to suffering, including the notion of care for others and compassion. Among the tweets analysed from Brazil, we found 206 direct mentions of this moral foundation (harm/damage). Regarding Portugal, the most used term to refer to this moral foundation is composed of the expressions from the word family ‘destroy’ (destruction/destroyed/destroys). Among the tweets analysed from Portugal, we found 11 direct mentions of this moral foundation (which we obtained through the word family linked to destroy/mistreat/spoil/damage). As example, we listed five tweets that appeared to US in chronological order (Figure 5).

Figure 5. Arguments that justify harm morality in tweets from Brazil and Portugal.

While comparing, we noted that in tweets from Brazil the use of the moral argument Harm generates a more linear discourse (with the idea that attitudes and aggressions generate climate change that harms nature and humanity). From Portugal’s tweets, on the other hand, it is possible to observe that Harm as a moral argument makes greater interconnection between fossil fuels and capitalism as agents of environmental destruction, even if another portion of the users is sceptical about the existence of the climate crisis.

Fairness versus Environment. In the theory of moral foundations, Fairness (justice) concerns the global notion of fairness and law, relating to the ideas of cooperation, competition and cheating as well. Thus, as we did with Harm (damage), we tried to find the word ‘fairness’ (justice) in tweets from Brazil to try to observe its use as a moral foundation. We found 450 direct mentions. Aditionally, we listed the five tweets that appeared to US in chronological order (Figure 6).

Figure 6. Arguments that justify fairness morality in tweets from Brazil and Portugal.

Concerning the comparison between the moral foundation Fairness, Portugal and Brazil coincide in form and differ very little in content. A speech on fairness in general and climate fairness in particular finds in Bolsonaro’s government (Brazil) and in the right liberalism (Portugal) a common enemy – it is a speech where the right is seen as an obstacle to an idea of fairness (which can be social or ecological, too).

Ingroup versus Environment. Ingroup/belonging relates to the commitment to the group, self-sacrifice, loyalty and surveillance against betrayal. In this case, the commitment to the group, the self-sacrifice, the loyalty and the surveillance against betrayal is in favour of nature. We found 14 direct mentions of the term ‘belonging’ in Brazil. In tweets from Portugal, we found 18 direct references to the idea of belonging or being part (Figure 7).

Figure 7. Arguments that justify ingroup morality in tweets from Brazil and Portugal.

Regarding the moral foundation Ingroup, there is, in both cases, a separation between the feeling of belonging to a ‘collective us’ that is threatened, and a consensus concerning the impact of the economy and large companies on the climate crisis. Among the tweets from Brazil, there is a feeling of strong belonging amid the ‘Brazilian people’ and the natural elements (where the Amazon issue stands out). Among Portugal’s tweets, there is a strong idea of belonging to European Union as a key element in battling climate change and even in new economic models (e.g., green new deal).

Authority versus Environment. With reference to the moral foundation Authority, which refers to obligations related to hierarchy, obedience, respect and compliance with duties, in Brazil, there were 51 direct occurrences. Portugal, on the other hand, recorded only one direct mention of the word ‘obey’ and 20 of the word ‘duty’ (Figure 8).

Figure 8. Arguments that justify authority morality in tweets from Brazil and Portugal.

The energy lobby and profits obtained from exploration (such as lithium) are the main issues that ‘should’, in the view of the tweets from Portugal, be the reason for an insurgence led by Europe and Portugal. From the perspective of tweets from Brazil, the Bolsonaro government is accountable for enabling illegal timber export and illegal mining. While according to Portugal’s tweets the solution drives through a united Europe to fight lobbies and profits from large companies and the nefarious things they do to the environment, in tweets from Brazil, those responsible are governments, namely the current one (2019–2022), which are not acting as a responsible authority.

Purity versus Environment. The moral foundation Purity concerns the physical and spiritual influence on chastity, wholesomeness and the control of desires (Figure 9). In tweets from Brazil, we found two direct and three indirect mentions (related to the issue of desire). Among the tweets from Portugal, three of them relate directly to the question of ‘purity’, and five refer to the idea of ‘clean’. In this case, purity is related to an idea of pro-renewable and pro-ecological behaviour, although there is a clear impression of some criticism of the hypocrisy of the invoked statements (once again, the issue of an attack on the energy lobby).

Figure 9. Arguments that justify purity morality in tweets from Brazil and Portugal.

Regarding the moral foundation Purity in tweets from Brazil, the attitudes toward environmental and nature protection are at the top of the arguments, and they place Bolsonaro's government as impure in that matter. On the other hand, among Bolsonaro's advocates, there is some criticism of a general idea of climate messianism imposed by left-wing policies that do not convince this quota of the population.

In tweets from Portugal, the moral foundation of Purity falls into the Portuguese Constitution, renewable energies and ecologist parties. Yet, there is a wing that considers the existence of hypocrisy in policies and arguments since the energy lobby is referred to as the centre of environmental contradictions – namely in the issue of lithium exploration.

Moral values and the most invoked personalities

By crossing moral values with the most summoned personalities, we noticed different foundations that establish different interpretative associations between Greta Thunberg and Bolsonaro.

Pearson’s correlation coefficient can have a range of values between +1 and −1. In this case, the values are all of positive order. In the previous table, we differentiate, with different colours, the values equal to or above 0.8, between 0.5 and 0.79 and between 0 and 0.49. The values recorded by us are in the range between 0.7 and 0.9, which means a strong positive correlation, between 0.5 and 0.7, moderate correlation, and between 0.3 and 0.5, weak positive correlation.

Moral values about Greta Thunberg

Greta Thunberg’s name is one of the most frequent personality when the subject is the climate issue from Portugal’s tweets. The largest relationship established with this personality in this Twitter sample is with Harm morality, where the terms ‘a combat’ or ‘to fight’ appear associated with the issue of the climate crisis.

Although the correlation is the weak type (r = 0.33), for Portugal, Greta Thunberg appears in Table 1 as a symbol of the battle against ‘climate crisis’. Here, the positive correlation (r = 0.815) between the terms ‘crisis’ and ‘climate’ is strong. Also, it is worth mentioning the moderate correlations between the expression climatic and harm-vice moralities from the (to)-Combat perspective (r = 0.63) and harm-virtue moralities from the fight perspective (r = 0.60). (See also Figure 10).

Figure 10. Cluster links by word similarity (Portugal).

Moral values about Bolsonaro

From the data collected, the most referred personality in Brazil was the President Jair Bolsonaro (2019–2022), and it is possible to perceive the links and connections established, as well as the syntactic and semantic connections (Figure 11). The sets highlight the relationship between Bolsonaro and farmers and indigenous people, in addition to the so-called negationist group with which the president is associated and who usually supports him. In contrast, environmentalists and scientists appear as a force contrary to his intent.

Figure 11. Cluster links by word similarity (Brazil).

On the other hand, contingencies appear as elements that influence the debate on the issue under study through a complex exteriority. The climate problem is heavily petitioned, albeit from an activism perspective and not so much from a conceptual discussion. On the other hand, contingencies such as war, pandemic or tragedies are debated as endogenous and exogenous factors to the issue of climate change, as well as the effects of deforestation and mining with which Jair Bolsonaro appears to be associated.

In relation to the policies issues, education is the most requested, followed by health and economy. The matter of education at a higher level is still relevant at a time when the pandemic (health issue) dominated the contingency.

As for the places that were brought up, if Brazil or the Amazon Forest appear naturally due the relationship with the country of study and the problem of the Amazon in the context of climate change, it is worth remarking mentions of Petrópolis (city in the mountainous region of Rio de Janeiro and which suffers from ever stronger storms) and the island of Fernando de Noronha (the government intended to end its status as a Natural Reserve, on the northeast coast of Brazil).

Regarding the connections between Bolsonaro and other personalities, it is important to mention the strong relationship with Ricardo Salles, a former Minister of Environment, Paulo Guedes (Bolsonaro’s Minister of Economy) and Cármen Lúcia (member of the Supreme Court). In the opposite direction, separately, we have Greta Thunberg and political commentator Gerson Camarotti, a critic of the Brazilian government.

Finally, in respect of action verbs, there is a predominance of the verb ‘approve’, because it functions as a governmental action (approval of legislation), but also the predominance of verbs of negative nature (vice, as we will see below), such as verbs ‘destroy’, ‘cut’, ‘blame’ or ‘avoid’. This relational network is shown in Table 2. Correlations between moral values and the most dominating personality throughout the tweets collected in Brazil allow us to realise the relationship perceived by network users between moral foundations and actions and the dynamics starred by Jair Bolsonaro. The correlation between Bolsonaro’s tie and the Harm-vice moralities stands out., based on the term destroy, at a moderate level (Pearson correlation > 0.5). Also noteworthy are the correlations, albeit of weak intensity, between the Harm-vice morality based on the term deforestation (r = 0.469), and between the Ingroup-vice morality through the term enemy (r = 0.459).

In general, we identify that there is a negative connotation between moderately and weakly correlated moralities since the perspective of vice appears with higher values than the perspective of virtue (refer to the moral foundations theory).

The values in Table 2 reveal a moderate positive correlation (between 0.5 and 0.7) and the remaining between 0.3 and 0.5, are weak positives.

Conclusions

Our results indicate that the public debate on climate change generates more engagement when related to social, political and economic matters. The Portuguese speakers in our sample are more interested in the human causes of the climate crisis than in the atmospheric consequences, which they do not have enough control over or information to discuss. Also, the tweets collected from Portugal and Brazil about environmental news indicate that the climate debate is pessimistic, criticising those responsible for the climate crisis and lamenting the destruction of nature, based on the moral terms used.

Through analysis of the terms used, we perceive patterns of moral sentences that may present the potential to generate engagement with climate news. The general pattern, found in the samples from both analysed countries, relates environment to the Harm and Care domains, and social actors to Justice and Betrayal.

Another pattern emerges between Brazilian and Portuguese tweets, when considering the analysis of emotions used in sentences. In the comments of young Brazilians, a sense of commitment to climate debate on the binomial nature – politics prevails, with emotions of negative bias and whose interpretation of the news accesses the moral attribute of justice. In the comments of Portuguese young people, a sense of commitment to the climate debate on the binomial emergency – action prevails, with emotions of a negative bias and whose interpretation of the news approaches the moral attribute hierarchy.

A third pattern occurs when collected news highlighted public figures, as in the case of the Brazilian president Jair Bolsonaro, in Brazil, and the activist Greta Thunberg, in Portugal (Tables 1 and 2). The news about the activist provoke engagement from Portuguese youngsters by the binomial fight – emergency (Figure 10). In what concerns moral attributes, news about Greta generates more engagement when they stimulate Harm (Table 1). News about the Brazilian president, on the other hand, provokes engagement from young Brazilians by the binomial politics – destruction (Figure 11). Regarding moral attributes, news about Bolsonaro generates more engagement when it stimulates Harm, but also invites the climate debate when it uses the moral attributes Ingroup, Purity, Authority and Fairness (Table 2).

From this thematic analysis, we reaffirm that the climate debate is centred on the action of public actors (e.g., Balbé and Carvalho 2017). This happened, pointing out errors in the protection of the environmental heritage in the Brazilian sample and charging for punishment of those responsible for the current climate crisis in the Portuguese sample.

In fact, the presence of public actors in social media climate debate could improve changing public perceptions (Mavrodieva et al. 2019), or even generate a ‘Greta Thunberg effect’ (Sabherwal et al. 2021), which influences people to have higher intentions of taking collective actions to reduce global warming and that stronger collective efficacy beliefs mediate this relationship.

Given this hierarchical model of news that hovers over networks, instead of being active in the contemporary climate debate, young people are encouraged to be reactive to presented matters. It is still necessary to consider the possibility of greater use of other social networks by younger people, such as Instagram and the emerging TikTok. However, due to their format, these are social networks that require other analysis methods.

New studies can be carried out on engagement with environmental news based on moral attributes, with young people from other countries, other topics besides the climate crisis and with the participation of other social actors in the news. It will be interesting to apply the patterns identified in the use of moral terms for the climate debate in other cultures than the Lusophone. Specifically for journalism, we recommend to allow room for public debate growth (as in the case of news on digital social networks) that includes emotions, values and interpretations on climate change, in addition to scientific and political discourses.

Overall, we hope that our work will contribute to public policies and the ensuing public debate to engage young people through moral queries used to assess the current climate crisis. It is fair to include their values in how climate change is fought; after all, they will be heirs of a planet whose decisions on the subject are being taken today. In a impending work, we will hold focus groups with young Portuguese speakers to analyse environmental news and moral values.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported and financed with national funds from FCT – Fundação para a Ciência e a Tecnologia (Foundation for Science and Technology), I.P. [grant number EXPL/COM-JOR/1534/2021].

Notes

1 The number of publications collected from Mozambican, in Portuguese, in the period was only 12 tweets, which is not representative of the population, thus, it is not part of this analysis.

2 This article is an English version of the paper “Alterações Climáticas e Argumentos Morais no Twitter: Um Estudo Comparativo entre Brasil e Portugal”, published in the journal Interações: Sociedade E As Novas Modernidades, (43), 107–127.