In Virality we Trust! The Quest for Authenticity in Digital Diplomacy
The Studium and the Punctum
Virality in is the new black in digital diplomacy, and rightly so, one may add! For Ministers of Foreign Affairs (MFA) and embassies, being on social media is no longer only about presence and networking, but about standing out through the virality of their messages. Creating content that is shared exponentially on social media, in a very short timeframe, with multiple levels of reactions from a mosaic of audiences is, to put is simply, ‘pure gold’ from a communicational perspective. Virality allows digital diplomats to step out of their immediate ‘bubble’ and reach out to unfamiliar audiences, showcase their position on important policy issues or normative claims, and enhance their social authority in front of their peers or the online public. In the attention-deficit space of the digital medium, virality promises to inject a high-dose of authenticity and engagement, even though, the outcome often has a short-lifespan and generates transient effects. The challenge lies, of course, with the fact that viral content is not that easily to create, especially by MFAs and embassies who generally lack the human, know-how and technical capacity to make their messages ‘go viral’.
As a first step towards addressing this challenge, we need to develop a good theoretical understanding of how virality works in the context of digital diplomacy. Roland Barthes’ reflections on the study of photography may assist us with this task, as they provide some useful clues about how to think analytically about the issue of virality. More specifically, Barthes argues that the way in which we make sense of the meaning of a photograph much depends on the distinction we draw between the studium and the punctum aspects of the image. The studium represents the contextual reading of the image that is, the historical, social or cultural details that the picture makes available to the viewer. The punctum, on the other hand, is the ‘out of place’ aspect of the photo that punctuates the studium and ‘pierces’ the viewer with an unexpected arrow of acuity. Put it differently, while the studium tells the viewer what the image is about in a manner that can be similarly understood by many, the punctum disrupts the studium and establishes a personal connection between the viewer and the image.
For example, Barthes finds that the picture taken by a Dutch reporter of a military unit patrolling a street in an unnamed Nicaraguan city during the uprising in 1978-79 (see Image 1) resembles well the duality and co-presence of the studium and punctum. The studium informs us about the gravity of the political situation, the desolation and destruction produced by the insurrection, the casual display of military force, and the bleakness of the future. The punctum, on the other hand, reveals, at least for Barthes, an unexpected contrast between two elements that do not usually belong together, the nuns and the soldiers, which seems to invite the reader to reflect on questions about war and death, violence and religion, destruction and reconstruction. The studium makes available to viewers a particular narrative about a historical situation with the goal to stimulate his/her interest and take notice of the human tragedy unfolding in Nicaragua at that time. However, it is the punctum that makes the photo transcend its state of general interest and connect it more intimately with the viewer by reaching out to Barthes’ subjectivity and rendering the image personally meaningful to him.
Image 1: Koen Wessing, Nicaragua, 1978
Barthes’ reflections on the art of photography carry good analytical value for the study of online virality as well. It offers us a framework for deconstructing viral content into tangible components by which to capture the interplay between the general and the specific, the common and the personal, the informative and the emotional, the inconsequential and the meaningful. Understanding the studium of viral content can give us a better sense of the themes, compositions, formats and approaches that makes certain messages highly popular. Understanding the punctum can reveal us the “out of place” profile of viral messages and their propensity for personalisation and micro-engagement. Drawing on relevant case studies of viral digital diplomacy, the next two sections will integrate the concepts of studium and the punctum into the discussion of two important aspects of online virality: contextual dimensions of viral dissemination (external vs internal) and rules of operation (information, emotions and personalisation).
External vs Internal Virality
A tweet by the former UK Ambassador to Egypt, John Casson, showing him strolling in Cairo, shortly before ending his tenure in Sept 2018, gathered 1.4k Reactions, 1.7k ReTweets and 11k Likes. By contrast, the tweet of the High Commissioner of Cyprus to the UK, Euripides Evriviades, showing him posing in front of the residence of the British Prime Minister as a memento before his departure from the post, garnered only 28 reactions, 25 ReTweets and 294 Likes (see Image 2). The question this example invites us to address is threefold: a) to what extent the virality of the two tweets is comparable?, b) how well each tweet performs relative to other messages produced by the same author? and c) what characteristics of the two tweets explain their virality? The first part of the question concerns itself with the issue of ‘external virality’ (cross-source comparison), the second with the issue of ‘internal virality’ (same source comparison) and the third part with the application of the stadium/punctum framework.
Image 2: Tweets with asymmetrical viral content
Many would be probably tempted to consider the tweet of Amb. Casson as being decisively more viral in its outlook than that of HC Evriviades given its sizeable lead in quantitative metrics. However, closer scrutiny reveals the two tweets are somewhat similar in terms of online influence once the number of followers is considered (see Table 1). More specifically, Amb. Casson’s tweet has only a small lead in terms of RT, but a stronger presence in terms of likes and reactions. This is so because the number of followers distorts the quality of the virality metrics by amplifying the randomness of the reactions. Put differently, it is clearly impressive when an account with 100 followers generates 1000 RTs, but arguably less so when the same number of RTs come from an account with 1 million followers. Therefore, RT/Likes/Reactions per follower provides a better basis of comparison of the ‘external’ virality of competing accounts.
Table 1: External virality adjusted by the number of followers
At the same time, it is important to observe the ‘internal’ dimension of virality that is, the extent to which a tweet aligns itself or diverge from the average reach of other messages generated by the same source. For illustration purposes, the average number of RT, Likes and Reactions of a sample of the most recent ten tweets produced by Amb. Casson (28 Aug – 5 Sept, 2018) and HC Evriviades (21-23 June 2019) fall well outside the normal distribution, between 2-3 standard deviations in the case of Amb. Casson and even further in the case of HC Evriviades. In other words, while both Tweets have performed extremely well relative to other tweets posted by the same author, the one posted by HC Evriviades is a clear outlier, especially in terms of Likes and Reactions.
Table 2: Internal virality relative to the average tweet reach in each account
This brings us back to Barthes’ distinction between the studium and the punctum and its role as an analytical tool for explaining the performance of these two tweets. From a studium perspective, both tweets speak to traditional themes about what it means to be a diplomat. Engaging with people and cultivating issues of common interest in the case of Amb. Casson and building political relationships in the case of HC Evriviades. However, it is not the conventionalism of the stories that makes the difference in terms of the reception by the audience, but the punctum by which the viewer is invited to interpret the message. The casualness and naturalness by which Amb. Casson mingles with regular Egyptian citizens stands in clear contrast with its official position, while the note of subtle humour that HC Evriviades drops in his message acts as a relaxing counterpoint to the solemnity that the 10 Downing Street door conveys as the centre of political power in UK.
The stadium/punctum framework also adds an interesting reflexive angle to the discussion regarding the influence of external vs internal virality. As the audience gradually becomes familiar with the style of the author, internal validity can sustain itself if the punctum constantly refreshes itself. For example, the casualness shown by Amb. Casson in his public interactions can demonstrate its viral value if it continues to surprise the viewer, by engaging, for instance, with unexpected guests, or changing the dynamic of the interaction with the public. In the same way, the light/solemn punctum adopted by HC Evriviades will require creatively updated formats of expression so as to maintain the attention of the audience. From the perspective of external virality, the studium can offer interesting insights about how certain themes of diplomatic reflection travel across space and time. For example, does the idea of direct engagement with the public resonate better in places where the local relationship between citizens and policy-makers is more hierarchical? Similarly, would humour be able to drive viral content in the same way in places where the reputation of power holders is negative?
As mentioned elsewhere, diplomatic communication has been traditionally embedded in a text-oriented culture that has favoured ‘constructive ambiguity’ over precision, politeness over frankness, reason over passion, and confidentiality over transparency. The arrival of digital technologies has infused the public sphere in which diplomacy operates with a set of new elements that have completely rearranged the ‘grammar rules’ of online engagement. Data and algorithms are now the new syntactic units of the new ‘digital language’ to which various combinations of visuals, emotions and cognitive frames are attached to create semantic meaning. This also means that digital content on social media platforms must tailor itself closely to these rules in order to be able to go viral. If so, what exactly are these rules and how the stadium/punctum framework can help us unpack the scope of application of these rules?
Rule #1: Information overload and limited attention contribute to the degradation of the quality of information that goes viral.
As shown by Weng et al. the combination of social network structure (the denser, the better) and competition for finite user attention provides a sufficient condition for the emergence of a broad diversity of viral content. However, out of the ‘soup’ of contending viral messages, it is more likely that those that come on top expose low-quality information as both the information load and the limited attention lead to low discriminative power. As Qiu et al. point out, viral diversity can coexist with network discriminative power when we have plenty of attention and we are not overloaded with information, conditions that are increasingly difficult to meet in the digital medium. In other words, the network structure of social media platforms favours the formation of viral content, but the attention deficit of the users acts as a filter for the quality of the viral content.
Image 3: Tweets with asymmetrical quality of information
As suggested in Image 3, Rule #1 carries empirical relevance. The tweet posted by the UN account showing the UN Secretary General, Antonio Guterres, inviting Arnold Schwarzenegger to attend the Climate Action Summit in September 2019 went quickly viral (by internal standards). It has swiftly reached roughly three times the average of Likes and RTs received by the UN account despite the scarcity of the information provided, except for a brief reference to the actor’s ‘I’ll be back” famous line. By contrast, the information rich tweet posted by the European External Action Service (EEAS) outlining EU-Asia security priorities, an important topic in the current geopolitical context, has been hardly noticed by the online public. One important implication of Rule # 1 is that the punctum needs to really stand out (via emotional framing of the use of a dynamic format) if the quality of the information reflected by the studium is to stay high and make a significant difference for the audience.
Rule # 2: Emotions rule! Content that evokes intense emotions is more likely to go viral.
One important school of thought on the psychology of emotions links Paul Ekman and Robert Plutchik’s influential theories of basic emotions to the pleasure, arousal and dominance model of environmental perception developed by Mehrabian and Russell (see Image #4). It is thus argued that emotions are associated with different degrees of positive (joy, surprise) or negative feelings (anger, disgust, fear, sadness), that they come with different levels of high (joy, anger, fear) or low energy (sadness, disgust) and that they are connected to feelings of control (anger, joy) or inadequacy (fear, sadness). Building on this model, Rule #2 states that messages reflecting high levels of valence, arousal and dominance, such as joy and anger, are more likely to go viral.
Image 4: Affective space spanned by the Valence-Arousal-Dominance model, together with the position of six Basic Emotions
Rule #2 has received empirical support from a few studies. Fan et al. have found, for instance, that angry emotions could spread more quickly and broadly on social media than joy. Stieglitz & Dang-Xuan have also found that emotionally charged Twitter messages tend to be retweeted more often and more quickly compared to neutral ones. In a controversial experiment conducted on Facebook, Kramer et al. have demonstrated that emotional states can be actually transferred to others via emotional contagion, leading people to experience the same emotions without even their awareness.
Image 5: Tweets with asymmetrical emotional valence
What is interesting in the case of Rule #2 is that the punctum is not necessarily defined by a particular feature of the message, but by the emotion that transpires from the message. The two tweets in Image 5, posted by the Iranian Foreign Minister, Javad Zarif (left) and the UK Ambassador to Argentina, Mark Kent (right), illustrate well this point. FM Zarif’s tweet conveys a pugnacious expression of angry defiance, while Amb. Kent relies on the positive emotion of surprise to ‘pierce’ and establish a connection with the viewer. Both emotions enjoy high levels of energy and dominance, which explains the excellent reception by the audience (several times the average of RTs and Likes normally received by the two diplomats). An interesting implication of Rule #2 is the potential constitutive effect of emotion-driven virality on the formation of online audiences: do emotional punctums provide the anchor around which audiences coalesce and if so, at what stage the studium becomes irrelevant for how messages are received by the emotionally primed audience?
Rule # 3: Content that can be easily personalised is more likely to go viral
In a seminal article, later expanded in a book, Bennett and Segerberg make the argument that unlike the top-down mechanisms of content distribution favored by hierarchical organizations, social networking involves co-production and co-distribution based on personalized expression. According to this connective logic, taking public action becomes less an issue of demonstrating support for some generic goals, as noble as they may be, but an act of personal expression and self-validation achieved by sharing ideas online, negotiating meanings and structuring trusted relationships. For example, the personalized action frame ‘we are the 99 per cent’ that emerged from the US occupy protests in 2011, or the more recent ‘MeToo” movement, quickly turned viral and traveled the world via personal stories, images and videos shared on social networks such as Twitter, Facebook, and Instagram. In short, the easier to personalise a message, as Rule #3 states, the lower the barriers for individual identification with social or political goals, the more opportunities for horizontal engagement, and by extension the more likely for such content to be absorbed, reflected upon, and disseminated through the social networks.
Image 6: Tweets with asymmetrical degree of personalisation
For MFAs and embassies, personalisation is not necessarily an easy task, as often case their online activities are primarily about projecting and emphasising their own set of policy priorities, approaches, and strategies to addressing various issues on the global agenda. Personalisation would imply exactly the opposite: removing oneself from the “digital spotlight” and identifying themes that can connect with as many individuals as possible. The examples in Image 6 aim to achieve this in slightly different ways.
The #WeAreNATO videoclip produced by NATO for its 70th anniversary in April places the member states at the forefront of the story about the historical evolution of the organisation. Personalisation takes place, in this case, via state representatives who come together to share their commitment to the values of freedom and security projected by the organisation. The viral tweet of the German Ministry of Foreign Affairs takes a different approach. It invites viewers to recall the suffering of those persecuted for fighting for justice and freedom and to identify themselves with the courage demonstrated by one of the last survivors of the resistance movement to the Nazi regime.
In contrast to Rule #2, personalisation does not primarily focus on emotions but rather on recognition and self-validation. The studium moves back to the centre stage as the repertoire of themes it proposes for discussion needs to offer points of connection by which individuals can express themselves in their own voice through the sharing of similar stories, images and actions. In the case of Rule #3, the punctum emerges not as an anchor by which the viewer is drawn to absorb the message of the studium via subtle contradictions or surprises, but as an invitation to engage as a co-participant in the production of stories connected to the studium that maximise perceptions of self-worth and social recognition.
To conclude, the dynamic environment in which digital diplomacy operates has increased the pressure on MFAs and embassies to become more conscious of the need to better understand how their messages could excel in terms of engagement. Barthes offers us good analytical tools (the studium and the punctum) for unpacking the contextual dimensions of viral dissemination (external vs internal) as well as the role of information, emotions and personalisation in informing the rules of operation of viral engagement.
This piece was originally published by the Elcano Royal Institute
 Roland. Barthes, Camera Lucida : Reflections on Photography (Hill and Wang, 1981).
 Corneliu Bjola, Jennifer Cassidy, and Ilan Manor, “Public Diplomacy in the Digital Age,” The Hague Journal of Diplomacy 14, no. 1–2 (April 22, 2019): 86.
 L. Weng et al., “Competition among Memes in a World with Limited Attention,” Scientific Reports 2, no. 1 (December 29, 2012): 335.
 Xiaoyan Qiu et al., “Limited Individual Attention and Online Virality of Low-Quality Information,” Nature Human Behaviour 1, no. 7 (July 26, 2017): 5.
 According to Ekman, human emotions can be grouped in six families ( anger, disgust, fear, happiness, sadness, and surprise) while Plutchik eight, which he grouped into four pairs of polar opposites (joy-sadness, anger-fear, trust-distrust, surprise-anticipation). Paul Ekman, “An Argument for Basic Emotions,” Cognition and Emotion 6, no. 3–4 (1992): 169–200; Robert Plutchik, “The Nature of Emotions,” American Scientist 89, no. 4 (2001): 344–50.
 A Mehrabian and J A Russell, An Approach to Environmental Psychology, Cambridge, Mass (M.I.T. Press, 1974).
 Graph adapted from Sven Buechel and Udo Hahn, “Word Emotion Induction for Multiple Languages as a Deep Multi-Task Learning Problem,” 2018, 1908, https://doi.org/10.18653/v1/n18-1173.
 Rui Fan et al., “Anger Is More Influential Than Joy: Sentiment Correlation in Weibo,” accessed June 25, 2019, www.weibo.com.
 Stefan Stieglitz and Linh Dang-Xuan, “Emotions and Information Diffusion in Social Media—Sentiment of Microblogs and Sharing Behavior,” Journal of Management Information Systems 29, no. 4 (April 8, 2013): 217–48.
 Adam D I Kramer, Jamie E Guillory, and Jeffrey T Hancock, “Experimental Evidence of Massive-Scale Emotional Contagion through Social Networks.,” Proceedings of the National Academy of Sciences of the United States of America 111, no. 24 (June 17, 2014): 8788–90.
 W. Lance Bennett and Alexandra Segerberg, “The Logic of Connective Action,” Information, Communication & Society 15, no. 5 (June 2012): 752–54.