Keywords

1 Introduction

Scholarship on social movementsFootnote 1 and media is a heavily populated field. There are two main reasons for this: First, it is essentially interdisciplinary, with researchers coming from at least three main fields: sociology, communication research, and the newest field of Internet studies, which is, in essence, interdisciplinary. The second reason is the fact that both the movements and media landscapes have dramatically changed in the last few decades. When it comes to media, two main developments are key: 1) digital social media use has supplemented or arguably displaced mass media use and 2) social media is used to create organizational networks among populations that lack more conventional formal political organizations (Bennett and Segerberg 2015). In terms of movements, we see that they have changed not only how they organize and communicate but also how they create and spread their frames,Footnote 2 the actions they take, their mobilization dynamics, etc. With digitally enabled content, we have also witnessed the “supersize effect of protest” (Earl and Kimport 2011, p. 13).

Following these developments, digitally produced content has opened a fruitful field of research, starting with comprehensive studies on the most basic and broadest of questions – what can researchers do with all this information (García-Albacete and Theocharis 2014)—and progressing to the most recent research using information from Twitter to answer specific and novel questions (i.e. Cristancho et al. 2020). A corpus of reviews and state-of-the-art pieces discussing research on communication and social movements or digital media and social movements has been made available (Rohlinger and Earl 2017; Bennett and Segerberg 2015; Earl et al. 2015; Earl 2018; Segerberg and Bennett 2011). Similarly, a systematic review of social media studies using qualitative and mixed-methods studies has been provided, for instance, by Snelson (2016).

This chapter aims to provide a brief overview of the use of content analysis in a selection of studies on social movement communication. Specifically, the first search was restricted to research employing the following keywords: “content analysis” and “communication”, and then either “social movement” or “protest.” Given the large number of results found for articles indexed in the Journal Citations Report (JCR), I prioritized state-of-the-art and most-cited pieces.Footnote 3 The final selection consisted in a large number of research articles, books, and volumes with significant variation in terms of objects of study, theoretical perspectives, research questions, channels of communication that provide empirical evidence and, therefore, methods.

2 Common Research Questions and Objects of Study

The emergence of social media – where social movements and participants interact openly – has provided enormous potential for research. The now very populated area of research on social movements and communication, for which content analysis is being used, could be categorized under four very general headings: 1) studies aimed at answering traditional questions in the area of social movements research; 2) research answering traditional questions in the area of communication studies; 3) research examining the use and effects of “new” tools and technology; and finally, 4) studies discussing methodological opportunities and challenges posed by digital media.

The first group includes studies that revisit or continue to explore the traditional questions in social movement research. These traditional questions are concerned with our understanding of the organization of a social movement, its life-cycle, its tactics, how it frames information, its frame dynamics, its identity, or the solidarity built among its members. By analyzing, for instance, content and comments on Facebook, one could answer a comprehensive set of questions regarding how a specific social movement develops and uses specific frames. For instance, this was the purpose of Harlow’s (2012) study on the Guatemalan’s justice movement. An analysis of the information provided on Facebook allowed her to answer a diversity of questions: Who were the organizers of Guatemala’s Facebook justice movement, and what were their motivations and expectations? Were Facebook users more likely to employ diagnostic, prognostic, or motivational frames? Among users employing the three types of frames in the Facebook comments, which thematic frame was most frequently emphasized: An agency frame, a values frame, an adversarial frame, or a reflective frame? What topical and functional subframes emerged among the Facebook comments? How is the frequency of users’ posts related to the frames and subframes of a comment? What kinds of news information did Guatemalan Facebook users post? What kinds of interactive comments and interactivity between the “real” world and “virtual” worlds did Facebook engender?

As discussed earlier, frame analysis is one of the traditional strong lines of research on social movements that can now benefit from large amounts of evidence. Harlow’s study is a good example of the use of content analysis of digital information to explore the internal organization of movements. Other examples focus on less formally organized protest events or movements such as the Gezi Park protest’s use of tweets (Ogan and Varol 2017).

Within the same group, some researchers focused on the question of how movements use media or social media to mobilize and create awareness: A broad category ranged from the analysis of the ads strategy of the Civic Rights Movement in the New York Times (Ross 1998) to the content analysis of movement websites (Stein 2009) and how specific protest movements used Facebook (Katz-Kimchi and Manosevitch 2015) or Twitter data (García-Albacete and Theocharis 2014; see also Theocharis et al. 2015). A good example of digital content analysis in exploring how a component of social movements, as a relevant collective identity, is developed using new communication protocols online is the work of Coretti and Pica (2015) of the Il Popolo Viola anti-Berlusconi protests in Italy between 2009 and 2011.

A classical question within social movement research refers to the life cycle of protest movements, and digitally created content is a useful means through which to address this issue. For example, we can now witness the evolution from a social movement to a new political party, which was the case of Barcelona En Comú (García-Carretero and Díaz-Noci’ 2018), similarly Borge and Santamarna’ (2016) study how movements develop into parties and construct their identities, manifestos, etc.

A major research strand deals with the analysis of activists or sympathizers during protest events or movements, but some studies have also examined the reaction of politicians during concrete events, such as Yavuz et al. (2018) during the Gezi Park protests in the summer of 2013. Alternatively, some studies have focus on how individual citizens get the attention of political elites in their protests during electoral campaigns (Sibinescu 2016).

A second strand of literature addresses traditional questions in the area of political communication. What this group of studies have in common is that they explicitly or implicitly use or refer to how media frame social movements, often employing content analyses of press outlets. The studies in this group vary significantly in their focus. Many examine how protest is framed in the media. Grimm and Andsager (2011), for example, showed that protest frames in the news vary according to the geo-ethnic context. Sneyd et al. (2013) worked on food riots across countries in Africa from 2007 to 2011. Cammaerts (2013) analyzed student protests in London and their symbolic damage in the press. Castro et al. (2018) examined Portuguese protests toward nature laws. Other instances include studies focusing on how the characteristics of the media outlet (i.e., geographical interest) might change how newspapers frame movements (Veneti et al. 2016); whether the characteristics of a movement were reflected in the media or conditioned by the media coverage of the movement (Kowalchuk 2010); or the relationship between media coverage and public support for a movement (Vliegenthart et al. 2005), to name a few.

A third broad group of studies is organized under common questions about how movements use “new” technological tools in light of the emergence and spread of new technology, digital tools, and skills. If, traditionally, one looked at newspaper ads to anderstand how movements used media to mobilize or create awareness (Ross 1998), the question now is how movements use social media, whether it is Facebook, Twitter, YouTube, or blogs (i.e., Avigur-Eshel and Berkovich 2017; Drueeke and Zobl 2016; García-Albacete and Theocharis 2014; Katz-Kimchi and Manosevitch 2015; Smith et al. 2015; Stein 2009).

Furthermore, digital tools bring about new forms of activism for exploration, for instance, in the case of videos and their diffusion. For example, using videos posted on YouTube, Thorson et al. (2010) posed several questions: Are videos circulated during protest movements amateur or professional? Do they report on life events? To what extent are they edited? Do they change as the debate progresses? How do audiences change according to the video’s format? Another fruitful line of inquiry uses and analyzes images. For instance, using images from the Egyptian revolution, Kharroub and Bas (2016) sought to contribute to a theory of user-generated content during political movements. The diffusion of images has become a main tool for protest movements aroand the word and provide the opportunity to explore the most dominant visual themes, the emotions they try to trigger, and how they change as the protest evolves. The macro information provided by social media (such as the number of retweets or images circulated) provides key movement-related information. Similarly, Casas and Williams (2019) analyzed the effect of emotions from pictures and their ability to mobilize people onto the streets in the Black Lives Matter movement.

Finally, a fourth group of studies focus on the methodological opportunities and challenges brought about by digital media for social movement and protest research. In this vein, some of the most-cited studies pioneered the analysis of social movements’ use of websites (Stein 2009) or Twitter (García-Albacete and Theocharis 2014) and, thus, elaborated on the potential of digital content in addressing old and new questions, even if it is from a descriptive or exploratory perspective. The latest work on methodological issues is much more sophisticated and has moved to (and lobby for) the need to integrate information across platforms to have a better understanding of specific movements (Driscoll and Thorson 2015; Von Nordheim et al. 2018).

3 Trends in Methodology

In terms of the communication channel through which the empirical evidence was gathered or produced, there was also considerable variation. Earlier work focused heavily on print newspapers. As Rohlinger and Earl (2017) noted, until digital and social media were added to the toolbox, the study of media and social movements was a subfield primarily focused on newspaper coverage. From the social movement research front, the norm were interviews with activists, professionals in movement organizations, and journalists. Others used internal documents from movements (i.e., Payerhin, 1996). For digital content, earlier work used websites and blogs as they were developed first. Later studies incorporated the analysis of social media data, such as from Facebook or Twitter, or pieces of the information included in a tweet, for instance, media links distributed in social networks (Segerberg and Bennett 2011). Soon after, there was a focus on the distribution of videos (Hermida and Hernandez-Santaolalla 2018; Thorson et al. 2010) and images (Kharroub and Bas 2016; Neumayer and Rossi 2018). Another common strategy was the combination of content distributed through at least one of the social media channels with other data such as media reports of movements and in-depth interviews. If a trend is detectable, it likely involved the combination of empirical evidence coming from different communication channels.

A review of the literature selected in this chapter showed that systematic manual coding was the most commonly used methodology, independently of the channel of communication from which the text originated. The usual strategy behind this methodology was the development of a codebook, with the evidence and research questions at hand, and the systematic manual coding after inter-coder reliability tests had been conducted. Complete codebooks were occasionally reported in publications. Most importantly, few of the studies revised referred to previous study variables and categories when presenting their coding procedure, so it is unclear whether the variables and codes were developed ad hoc or built on previous studies. This commonly used exploratory approach, which is based on developing a codebook as researchers separately analyze information for each protest event or movement, is useful in gaining a better grasp of the characteristics of a specific case of contentious politics, especially since replication might not be the primary goal of many studies. The lack of replication, however, might be a limitation of this subfield, and it is surely a limitation for the original goal of the present review of identifying specific variables and how they are usually coded.

Another clear trend in the existing research was the combination of methodologies. A large number of studies combined qualitative and quantitative (primarily systematic manual coding) approaches of data collection and analysis. A common strategy, for instance, was the above-mentioned combination of in-depth interviews along with the quantitative manual coding of social media text. Others combined two quantitative methodologies, such as a combination of social network analysis and Twitter conversations (Himelboim et al. 2013; Ogan and Varol 2017; Papacharissi and Oliveira 2012; Wonneberger et al. 2020). A trend could also be observed regarding the use of computer assisted or automated methodologies with manual coding. In the political communication studies, an increasingly common combination consisted of a computer assisted search of articles that were then manually coded (Cinalli and Giugni 2016; Veneti et al. 2016). State-of-the-art methodologies were constantly developed as different types of evidence were being incorporated, such as videos and images. In addition, a number of voices have started advocating for greater integration of information gathered across platforms in order to avoid potential biases (Von Nordheim et al. 2018). For instance, Driscoll and Thorson (2015) used the 2011 Occupy Movement protests and the 2013 consumer boycotts to illustrate methods for creating integrated datasets of political event-related social media content by using 1) fixed URLs to link posts across platforms (URL-based integration) and 2) semiautomated text clustering to identify similar posts across social networking services (thematic integration). These approaches helped them identify biases in the way that previous studies characterized political communication practices when focusing on a single platform. For example, by only analyzing a video uploaded to YouTube by a movement, we can gather valuable information to anderstand the movement. However, we might overlook the specific use, the extent of the diffusion of that video, and its effectiveness in, for example, a call for action than if we were to add information from other platforms, such as an analysis of tweets linked to that same video.

4 Research Desiderata

As a flourishing field, content analysis of social movements has changed significantly in the past years with the integration of new types of data, additional protests movements, and the creation of specific subfields, making it an exciting area of study with a great deal of potential. Furthermore, it is an interdisciplinary area with increasing opportunities and avenues for collaboration. However, interdisciplinarity, increasingly available data, and continuously emerging objects of study pose some relevant challenges. To begin with, anything Internet-related is often treated as a new object of study, but as Rohlinger and Earl (2017, p. 8) pointed out, “there was protest before the internet and there is literature about it.” Ignoring this history impedes knowledge accumulation. Furthermore, I would add that ignoring previous research means a failure to use the many opportunities brought about by newly available empirical evidence and information to answer important – and often classic – questions in the discipline.

A second challenge is the disconnect between studies (at least explicitly), which has resulted in a large and eclectic knowledge base regarding different movements, protest events, communication practices, and so on but little integration that would permit a speedier accumulation of knowledge. To illustrate this point, say we gather protest-related information on the type of users that distribute posts or information in social media. By systematically analyzing the evidence at hand, we can identify and organize categories of users (e.g., activists, alternative media, movement organization, politicians, etc.). However, it is only by comparing these categories to those developed in previous research can we identify the unique aspects of that specific movement. If, say, we find a large number of types of actors but we do not compare our findings against other studies, we overlook the fact that formal organizations are only marginally involved in the calls for action circulated by the movement. The informality of the movement might be one of its main characteristics, but without a systematic comparison to other movements, the picture would be incomplete. Further comparative work using content analysis or the same coding schemes for different case studies could provide standardized codes and variables to measure specific concepts and, thus, test theoretical claims in a wider context, with the obvious advantage of theoretical development.

To sum up, content analysis of the role of media in contentious politics is a vibrant research area, with a large number of subfield-specific research questions. In terms of methodology, the systematic manual coding of texts has become a standard approach. Given the expansion of this technique, researchers are increasingly adjusting to best practices in terms of developing coding schemes, assuring high inter-coder reliability standards, and the reporting of coding procedures. However, there is some room for improvement in terms of knowledge accumulation. This review has identified at least two avenues: the combination of classical scholarship with digitally created evidence and an explicit effort to test concepts and measurement instruments across diverse protest movements. Finally, as automated analyses of large datasets (text, geographical, etc.) and big data evolve, research on social movements will most likely develop substantially in the years to come.