First, my most recent work investigates disinformation campaigns, specifically those emanating from Russian-affiliated media outlets, on Twitter. Using on analysis of big, longitudinal data sets I retrieved from Twitter, I—along with my coauthor—developed the concept of a percolation network—a network of social media accounts held together by common social media sharing practices and ideological perspectives standing at the ready to amplify particular types of information and narrative frames, especially during salient political events. Disinformation campaigns can take advantage of these networks. To do so, many accounts seeking to spread disinformation need to be viewed as trusted brokers of information within the percolation network; if they are not perceived as such, the tweets will vanish in the information deluge. We argue they rely on the mechanism of keying, whereby they deploy claims that resonate with their target audiences. Keying takes on two prevalent forms in the percolation network. First, distribution agents such as RT and Sputnik actively try to “fit in” by deploying cultural cues (e.g., language, images, values, memes) likely to resonate with local audiences. This can help legitimize the content state-affiliated accounts transmit into the target network and thus strengthen the link between the affiliated twitter accounts of RT and Sputnik and local users.
The Twitter data was sampled from accounts that tweeted in French and English about the #MacronLeaks within the first 24 hours of an information dump that occurred in the 2017 French election that pitted centrist Emmanuel Macron against right-wing and Russian dove Marine Le Pen. The idea was that Russian state-affiliated outlets had tried to implement another October Surprise in a different Western democracy in the hopes that their preferred candidate would benefit. Although the #MacronLeaks episode was short-lived and had little impact on the electoral outcome, analyzing the behavior and effects of the entities involved revealed evidence for the existence of the percolation network. A short qualitative case study demonstrated the integral role that RT and Sputnik had played in spreading the #MacronLeaks stories that followed the dump of emails on Pastebin. After conducting Structural Topic Modeling (STM)—a form of Computer Assisted Textual Analysis (CATA)—I found that the percolation network activated by Russian disinformation campaigns are sustained and maintained by the routine, sometimes mundane sharing of Islamophobic, Eurosceptic, and anti-centrist (political party corruption) content. The analysis also revealed a substantial transnational element, as a significant number of the (sometimes unwitting) participants in the MacronLeaks disinformation campaign tweeted in English, French, German, and Dutch. The measure of success in disinformation campaigns lies in the ability to make and foster connections to existing social network enclaves in ways that attract little scrutiny. The study reveals a sophisticated process operating at the level of individual users and localized internet communities. Those interested in studying foreign interference and disinformation should pay closer attention to the microdimensions of power-political competition in the digital age.
This project and its subsequent publication in The Journal of Global Security Studies demonstrated the feasibility of combining the theoretical arguments in my dissertation and the IR theory literature with big data analysis. As a result, I gained competency in computational social science methods and various programming languages such as Python and those in R, and will continue to develop these skills, especially sentiment analysis, machine learning, and social network analysis, as I seek to bring several other papers and projects based on social media data retrieved from Twitter to conclusion and publication. One of these (forthcoming) concerns the threat disinformation poses to liberal democracies, and uses a similar set of data from Twitter to develop a causal, chronological link between Covid-19 disinformation and the ultimate occupation of downtown Ottawa by the so-called Freedom Convoy in Canada in 2022. Moreover, I am working toward improving my data visualization skills in R, and, in doing so, hope to be able to better present the findings to more public-facing audiences. I have presented these most recent projects at conferences and workshops over the past several years.
Other work related to these topics is oriented toward contributions in international relations, as it focuses on the efficacy and practices of digital diplomacy, especially regarding the online behavior of international organizations. I am working on project that analyzes digital diplomacy in times of crisis. It aims to conceptualize a new, emerging form of crisis diplomacy that spans the traditional divide between diplomatic practices and communications between diplomats and state actors and public diplomacy. The project traces the social media activity of the Russian invasion of Ukraine in three phases: 1) the lead up to the crisis (from December 1, 2021), 2) the period of escalation and allegedly unexpected failure for Russian forces to topple Kiev (the first four weeks of conflict), and 3) the period of relative entrenchment and Ukrainian counteroffensive(s) (ongoing). I focus on the social media activity of three categories of actors and institutions: 1) official international organizations; 2) ministries of foreign affairs; 3) official accounts of individual political leaders (and diplomats).
With the collected data, the goal is to reveal if and how the content of the shared information—what they are talking about—changes. If the entities change their approach to digital diplomacy over time, is it correlated with changes in the offline situation (to the extent they can be disentangled)? What public narratives are they attempting to construct and shape? What social media practices are used in which situations? I hypothesize that these accounts resort to trolling practices when the identity of organization is perceived to be under threat or when it perceives itself to be in a far superior position—in terms of social capital/prestige—than those it views as adversaries; this makes them bad faith actors. Spreading disinformation may also be a common tactic. Mundane information sharing and social media posts using traditionally diplomatic language will be most common in periods where the entity benefits from the status quo or from traditional diplomatic approaches to conflict and cooperation; this makes them good faith actors.
I also support this analysis with an analysis of the replies to the information these accounts share. Practitioners routinely claim they read the replies. And so, how do the replies affect digital diplomatic practices? Examining whether people read and respond to the tweets themselves or respond to what they imagine the entity itself to represent gets at the ever-present question of efficacy digital diplomats’ communications. Preliminary analysis reveals that, by and large, these organizations are reaching audiences that tend to agree with them, and thus, the tweets that receive the most engagement are the ones that most forcefully emphasize the mission of the organization. These findings and ongoing research points toward a more comprehensive notion of digital diplomacy, one that focuses on various modes of digital diplomatic performances.
This arm of my research agenda was sparked and inspired by an invitation from a loose affiliation of scholars studying the bourgeoning field of digital diplomacy to contribute to a workshop and edited volume about international organizations and their use of social media. In particular, my co-author and I published a chapter investigating the extent to which international organizations could or even should engage in Twitter diplomacy. In it, I developed a theoretical framework based on my dissertation project that might be used to evaluate and validate various attempts at digital diplomacy. Further exploring digital diplomacy and conceiving it as a hybrid form of public and formal diplomacy, while situating it historically, is of import. There is a need within the larger literature for digital diplomacy to be investigated on its own terms, a task requiring deep conceptual work on cultures of practice and the historical development of diplomatic institutions and other entities entering the diplomatic space. Furthermore, this line of research is well-positioned to produce recommendations to policymakers navigating the fast-paced environment of social media and diplomacy, as well social media companies whose ambitions are vast.
Possible future research questions include:
- How is the conduct of diplomacy, especially diplomatic communication, affected by digital mediation? How do allies and adversaries encounter each other digitally?
- How do non-state actors, such as IOs, engage with digitalizing diplomatic institutions?
- How do states navigate and negotiate various target audiences, both domestic and international, on social media? Why do states and non-state actors turn to Twitter diplomacy?
- What practices and performances constitute digital diplomacy? What is its relationship to traditional diplomacy and public diplomacy?
These research questions are premised upon a careful treatment of the historical development of diplomacy, various unit-types engaged in the diplomatic realm, and the relations between these and the public. Cultures of practice arise over time and are constituted relationally. The digitalization of political processes has disrupted the normal conduct of diplomacy, and adversaries and allies alike must navigate a rapidly shifting environment where common ground based on the Anglo-led liberal order seems to be shifting or disappearing altogether.
Second, my dissertation project that focused on the concept and emergence of democratic subjectivity also used social media practices in momentous and mundane times as an analytical lens. The dissertation makes a contribution to the political theory subfield by distilling how political subjects come to be as a result of everyday (digital) practices. I argue that coping with everyday life is inextricably intertwined with digital practices and modes of being. A close reading of liberal democratic theory, the theoretical assumptions of transnational IR literature, and insights from recent sociological theorizing about identity and community formation, I theorize that information sharing practices—now conducted mostly on social media and other digital communications platforms—exert a mechanistic effect toward disruption. In some power political contexts (e.g. as described in my case study chapters on the Arab Spring and the Occupy Central Movement), the disruption can be liberalizing. In other political contexts, it can be radicalizing. The notion that information sharing is inherently liberating, and that horizontally organized networks and organizations are more democratic, form a kind of ideology found in the literature, reporting, and theorizing emerging from and during the Occupy movement era. However, I argue, the disruptive mechanism is driven by everyday necessity—people use information sharing practice to get through their everyday lives—and by the social rewards associated with increasing extremism (or vapidity).
I focus on the communicative styles and performances of individuals and how they come to shape their identity in relation to the liberal democratic political environment. Along with the case study chapters focusing on the social media practices of the Occupy-style movements, I lean on computer assisted textual analysis—specifically structural topic modeling—of big Twitter data. I conclude that: 1) collective subjectivity can emerge around and be based upon shared practices; 2) people poach performances across borders; 3) collective political subjectivity is disruptive and shallow, breeding mundane extremism; 4) states struggle to comprehend everyday digital practices, for now; and 5) a practice-based conceptualization of contemporary democracy is necessary. I end with a discussion of the epoch between the Egyptian Revolution to January 6th. While I shy away from causal arguments linking emerging technology to political outcomes, my work certainly speaks substantively about these issues. I am well-positioned to contribute research and expertise on questions lying at the intersection of democratic politics and emerging technologies.
Third, my work in voting behavior, especially as it relates to information gathering on the part of voters, colors my interpretation of online behavior. It has informed my interpretation and understanding of the partisan and extremist behavior on the part of the far right online. The concept of the social construction of hard times has been very salient in explaining the rise of right-wing populism and far right online behavior that I have studied in the American, British, French and German contexts. In my research, I ask how identity—in particular, the political psychology of identity construction—interacts with and shapes political and social behavior. I seek to understand how people use identity/identities to navigate an extremely complex information environment, and the consequences of doing so. It seems that in the current era, cross-cutting identities are exerting pressure on citizens in ways that encourage them to perceive the information environment in ways unexpected by long accepted partisan or class-based perspectives on political behavior. The blurring of demarcations between public and private sphere, of online and offline personas, of work and personal life—these trends are contemporaneous with the rise of social media and digitalization of political and social processes the world over—as is the intensification of radicalization and polarization processes. My research agenda addresses these processes by investigating how social media identities are shaped by practices and vice versa.
Fourth, my most recent conference papers indicate a common interest in studying right-wing online extremism, especially its relationship to social movements and their influence on political coalitions, party platforms, and governance. The extant literature still tends to underplay the significance of what regular people are saying and sharing on social media about particular events and topics, particularly the transnational dimension of sense-making in domestic political processes. My investigations into the #Chemnitz protests, the 2021 German elections and devoted MAGA followers show that my current theories, methods, and analytical frames can be applied to questions and cases in other subfields, such as campaigns and elections, comparative political systems (especially Europe), and American politics. In each project, I use Twitter data to explore how political agency and identity are formed in relation to momentous events, such as protests, elections, and domestic crises. I also seek an explanation as to why it has become commonplace to negotiate these processes on social media as an event is ongoing.
As these projects show, I believe it is important to develop a more systematic and comparative approach to analyzing the communicative processes by which individuals engage with politics in the digital age, across contexts, whether that be state, type of social media, language, ideology, or power-political position. Turning to grounded theory is one way to try to avoid privileging the largely Western notions of what is relevant to political science research. Considering evidence from the Global South will enrich this line of research further, which I plan to do in collaboration with other scholars. Continuing to employ a multimethod approach would require better quality social media data, especially from sources other than Twitter. More and better data would allow me to follow the promising work of so many others engaging in digital politics and bolster a robust research agenda. Working with data in languages other than English, French and German would further illuminate such processes. The fact that I have been able to make sense of datasets in multiple languages using my code means such investigations hold promise.
In general, my research program seeks to address how and why individuals are mobilized to spread (dis)information to foreign publics on social media, why social movements organized online fragment or radicalize, and how political identities are performed on social media. With these kinds of research questions, it can be useful to be able to gather data in real time. Having my own tools ready for rapid deployment in the field, such as code, storage, and analytical methods, also allows for a more immediate public-facing contribution and response to ongoing events so as to aid policymakers and the public comprehend ongoing developments. Situating current developments historically can help make legible the digitally-mediated disruption of normal politics, the radicalization of individuals and groups, and times of international crisis.
As such, I seek to study social movements and their communication, especially those that exhibit a transnational dimension, and the effects their communication practices ultimately have on political processes, parties, and governance. I would like to explore also working in experimental research to explore the affective dimensions associated with social media practices in an effort to strengthen external validity. Altogether, the long-term goal is to demonstrate the viability of a multimethod research that integrates with computational social science techniques to answer interesting and salient questions about politics.
This research agenda may fill some gaps in the literature, but, more importantly, it seeks to further develop, enrich and extend the current research on digital diplomacy and the digitalization of political processes. By focusing on what people do from the ground-up, we can better understand the dynamics of digitalization that affect political participation, domestic and international conflict, and the (ability to) conduct foreign policy.
Mike Habegger, Ph.D. 