Performing leadership: international politics through the lens of visual narrative analysis

ABSTRACT Since visuality in the (self-)representation of politicians and other influential figures has become an important part of political storytelling, we propose to use visual narrative analysis (VNA) as a systematic approach for its better understanding. VNA is particularly suited for this performative strand of interpretive analysis, as it does not study images in isolation but in the broader context of political narratives. By analysing different layers of communication (images, narratives, competing narratives) VNA enables us to identify internal contradictions that undermine political efforts of self-representation in contexts of global governance (e.g. multilateral diplomacy) and render them unstable and contestable. By analysing competing (self-)representations at a G7 meeting in 2018, we show how VNA can be applied fruitfully to the study of international politics and, second, how VNA can explain some of the reasons why one image became iconic (Angela Merkel as female leader of the liberal world), i.e. appealed to a wider audience, and others (focusing on Emmanuel Macron or Donald Trump) did not. While our article is primarily a demonstration of the methodological benefits of VNA for various research contexts in world politics, it also contributes to conceptual debates on the combination of visuality, narratives and emotions in changing practices of political storytelling.


Introduction
World politics and its representation for an international public has always been closely connected to the power of images and visual storytelling to (re-)produce and (re-)stabilise the political authority and legitimacy of international institutions and their political leaders as representatives (Williams 2018). The global public has long become familiar with a similar set of images, such as the ritualised pictures of global summits when diplomats or political leaders are grouped in varying, often rather monochromatic rows and made to hold hands for pictures, often posing in front of flags. The images produced during summits are part of diplomacy, which partly builds on states seeking to broker which we see as central in the self-representations of politicians and other actors of (international) politics. Scholars in the tradition of symbolic interactionism and the work of Erving Goffman, for instance, have shown that diplomacy can be understood as impression management, in which the actors represent national interests, but are mainly involved in strategic facework and thereby create a social world of its own (Adler-Nissen 2013; also see Nair 2019). This form of status signalling is also observable in virtual summitry, where political actors draw on a shared understanding of symbols and resources that have social value in the interaction order of summit diplomacy (Danielson and Hedling 2022).
We aim to show how to analytically carve out the contestability of self-representations that always threatens their strategic content. VNA is a multi-modal methodological approach, which we propose here as an interpretive perspective that has potential for further exploration. VNA is particularly suited for reconstructing this performative mode of interpretive analysis, as it does not study images in isolation but in the broader context of political narratives. By analysing different layers of communication (images, narratives, competing narratives), VNA enables us to identify internal contradictions that undermine public diplomacy efforts and render them unstable and contestable. In our analysis of competing (self-)representations at a G7 meeting in 2018, we illustrate how visual narrative analysis can be applied fruitfully to the study of international politics and, second, how VNA can explain some of the reasons why one image became iconic, i.e. appealed to a wider audience, and others (e.g. of Trump or Macron) did not. Even though we cannot 'measure' the success or resonance of a distinct visual narrative, our interpretive perspective allows us to analyse the performative effects of images as analytical elements of visual narratives that have become one important means of political communication in diplomacy, and to relativise the assumption of their strategic uses. While our article is primarily a demonstration of the methodological benefits that VNA offers for the study of (international) politics in other research contexts, it also contributes to recent conceptual debates on visuality, political storytelling and performativity in world politics.
In the following contribution, we concentrate on a widely discussed image of Angela Merkel and other state leaders at the G7 summit in 2018 to explore the different ways and means employed in diplomacy and international politics in such situations of high attention in political communication and self-representation. While Merkel's representation was widely perceived as an indicator of her status as the new 'leader of the free world', with Donald Trump leaving a gap in the traditional leadership role of the US, not all of the represented (male) leaders were happy about this image and the implied narrative. Some contested the interpretations, and others even tried to produce different versions that would boast their own self-representations (including Donald Trump and Emmanuel Macron) and scratch at the image of Merkel as a symbol of (female) leadership of the liberal world, or draw attention to them. In the end, the most iconic image that prevailed since the summit remains the picture with a focus on Merkel. Even though this particular image is highly iconic and rather provocative, doing VNA does neither presuppose that images need to be iconic nor does it work best when they are (see Unrau 2022 for an adoption with ordinary images). At the same time, the case illustration we chose for this paper has been discussed in many outlets and thus demonstrates how a (rather schematically conducted) analysis can uncover the interplay between different layers of meaning that even characterise such seemingly obvious cases.
The article proceeds as follows. First, we carve out the unstable nature of self-representations in political communication efforts. Second, we explain the promise of VNA for political science research. We outline some conceptual thoughts on narrative and visuality and explain how the method of VNA can be adopted in empirical research. Third, we illustrate our argument by means of a short visual analysis of competing images around the 2018 G7 meeting. We conclude by reflecting on the role of images and visual storytelling in world politics.

The contestability of self-representation in political communication efforts in global affairs
While self-representation enable demonstrations of the ordering element with which those practices of international politics normatively stabilise the authority and legitimacy of international institutions, they always remain contestable. For that reason, they provide an excellent methodological entry-point for laying open power relations, normative contestation as well as critique and resistance. These are important dimensions, which are analytically lost if our focus turns exclusively to strategic considerations when it comes to the self-representation of leadership or status competition. Therefore, our approach addresses contingency as a key feature of political storytelling as it has always been emphasised by discourse-oriented scholars (e.g. Doty 1993) who have shown the performative effects of political language. As such scholarship has shown, this narrative, or discursive power is a process, however, that is neither linear nor controllable process by political actors in the instrumental sense of rationalist actor models (also see Adler-Nissen, Andersen, and Hansen 2020; Hansen, Adler-Nissen, and Andersen 2021).
Think of former US-President Donald Trump's handshake with Japanese Prime Minister Shinzo Abe on the sidelines of the G20 meeting in 2019, which received a lot of (negative) attention, due to Trump's blatant break with diplomatic protocol. When Trump and Abe joined hands for a photo-op handshake, Trump used the handshake to pull Abe over to his side. One version of the images derived from the photo-op shows how Abe's hand has been drawn over Trump's right armrest, across the gap between them, while Trump is firmly gripping Abe's right hand from below and further holding the hand from above with his left. While handshakes in general are meant to demonstrate mutuality in (symbolically) equal settings, the diplomatic convention is compromised here. The patronising gesture and awkward moment were widely interpreted in the media as an example that demonstrated President Trump's contempt for diplomatic rules and preference for gestures of dominance, catering to certain audiences at home and abroad. From a political science perspective, the instance can also be used as an example to illustrate how Trump's gesture seemed to be aimed at changing the logic of multilateral politics and their established practices to instead treating them as a matter of business and competition .
The handshake can further be interpreted as an attempt by Trump to humiliate Abe (as representative of former enemy and economic competitor Japan, with whom he personally maintained cordial relations, however) and elevate himself to a position of superior international standing. Moreover, since handshakes primarily emerged in the West, whereas bowing is more traditional in Japan, the breach of etiquette is particularly objectionable. It seems to indicate Trump's underlying need to demonstrate a form of rather basic masculinity, in which competition and physical strength are seen as key assets of a leader. His action resonated with ideas of 'America First' and 'Make America Great Again' and through a visual 'moment of truth' revealed the underlying rationales behind Trump's move of discarding multilateralism.
This case of the 'handshake' exemplarily reminds us how political leaders make use of self-representations and visual storytelling as strategic instruments of communication towards different audiences, yet can never entirely control them. With regard to his domestic (and some international) supporters, Trump may actually have reached his political goal, as he was successful in being perceived as a strong (i.e. masculine) leader by these supporters. At the same time, his open break with diplomatic principles was not well received by more liberal domestic public and state leaders within the G20. The relationship between the self-representation and an audience can be described in performative terms, pointing to the unstable, contingent nature of performances in many political contexts, ranging from multilateral diplomacy to public statements and election campaigning (Alexander 2011).
Many examples of status signalling or public diplomacy (Danielson and Hedling 2022;Hellmann and Oppermann 2022) are ambiguous and underline the performative power of visual storytelling in politics, which despite all efforts can neither be strategically controlled nor perfectly predicted by an analyst. For instance, José Manuel Barroso's numerous attempts, as president of the EU Commission, to improve the public reputation of the European organisation never came to fruition; even though PR agencies and many others worked hard to come up with a narrative that would change the perception of the EU among the European people, his strategy would simply not succeed (Kaiser 2017). Recent, much more urgent public performances by Ukrainian president Volodymyr Zelenskyy, however, have certainly mobilised a large public not only because of the massive breach of international law by the Russian attack on Ukraine, but also because of his versatile, credible and extremely powerful personal appeal. For instance, Zelenskyy's virtual appearance in national parliaments may even have created a new genre of political communication that relies heavily on the employment of cultural repertoires in terms of storytelling and evoking emotions. His appeals managed to reach a wider audience and also surprised people because of his undiplomatic, highly authentic way of talking. However, the effects of this form of visual storytelling in different contexts differed and remain beyond perfect control. In the German case, for example, the members of parliament as well as chancellor Olaf Scholz simply could not respond to Zelenskyy due to adherence to a priorly agreed upon agenda in parliament, which weakened the appeal of the speech (and caused irritation in the German and general public). Therefore, even though the performance was very fittingly catered to the specific context of German history and responsibility, Zelenskyy's appeal resonated less with his audience due to reasons beyond his control. This is indicative of not only the contestability of self-representations but also their generally unstable nature.
Recent interpretive studies in IR and comparative politics have become more interested in the performativity of international politics and consider the role of practices, narratives, visuality and material objects in diplomatic activities in the everyday (see Neumann 2012 for a first overview). Examples of such diplomatic practices range from writing a speech as a joint activity (Neumann 2007) to using track change as a powerful technology in writing common documents (Adler-Nissen and Drieschova 2019) and developing new visual communication practices or 'visual diplomacy' in 'screen worlds' in times of Covid 19 (Danielson and Hedling 2022;Eggeling and Adler-Nissen 2021). We link up to this latter aspect and argue that the role of images and visual storytelling in the field of international politics and diplomacy should be further explored, particularly with regard to their methodological foundation and interpretive potential (Crilley, Manor, and Bjola 2020;Möller, Bellmer, and Saugmann 2022;Schlag and Heck 2020). Based on previous work Gadinger 2020, 2022;Unrau 2020, 2022), we use the methodological tool of VNA to explore the changing ways of doing and representing diplomacy by means of visual storytelling. Since VNA combines the analysis of images with the analysis of narrative context (such as cultural repertoires and competing narratives), we incorporate both visual narratives and their performative dimension, which also means paying attention to their potential receptionas the example of the handshake illustrated. We want to show how visual storytelling is used to stabilise fragmented identities of certain audiences, for instance through evoking affective responses; again, the handshake did resonate positively with domestic and international publics that did not see dominance as more valuable than adhering to diplomatic conventions. These kinds of analyses allow us to formulate substantiated criticism of underlying logics or ideologies in the political stories of specific sets of actors such as right-wing populist movements and their political self-representations that proved to be rather contestable.
As IR scholars and political scientists, we are inclined to take images as illustrative instances of broader themes such as cooperative practices and power relations, but less so to interpret them with regard to their subconscious associations and intertextual references. Simply put, we often ignore the narratives in which they are embedded and thereby miss their potential for critical analysis. In order to capture the interplay between these elements, a VNA builds on a multi-modal proceeding, using a technique of layered interpretation.

The promise of visual narrative analysis
Our conceptual and methodological ideas around the notion of VNA are mainly inspired by recent work on narratives and visuality in political science and IR (see e.g. Bleiker 2018a; Callahan 2020; Ravecca and Dauphinee 2018). Both research streams are closely connected and share many interests, in particular in their interdisciplinary conceptual thinking and combination of different interpretive research methods, such as image analysis, visual autoethnography and other rather novel research techniques in IR (see Bleiker 2019; Harman 2019; Möller, Bellmer, and Saugmann 2022). As we base our conceptual understanding of visual narratives mainly on narrative studies, we start with some key conceptual premises, relate them to visuality and develop our methodological framework of VNA.
Narrative approaches have experienced a renaissance in political science after a short period of hesitant interest some decades ago (e.g. Patterson and Monroe 1998). While the term narrative is sometimes narrowly used in terms of strategic action, we support a processual and relational notion of narrative as it has been developed in narratology (Koschorke 2018) and is established in practice-oriented perspectives in IR (Bueger and Gadinger 2018). Narrative analysis draws on insights from discourse analysis and practice-oriented approaches but is geared more towards paying close attention to aesthetics, fiction and emotions than the former, particularly with regard to traditions in political science (Park-Kang 2015). That is, while in strategic thinking narratives are often understood as instrumental tools that can be controlled by political actors in their use, for example in earlier understandings of public diplomacy; practiceand discourse-oriented scholars emphasise that narratives are never fixed, but instead permanently produced and negotiated under conditions of polyphony.

Making sense of the world through narratives
Since narratives are a form of configuration device by which actors seek to make sense of the world and order it in a specific way (Czarniawska 2004, 11), IR has been concerned with how notions of a global order can fruitfully integrate the concept. As narratologist Jerome Bruner (1991, 4) famously remarked, 'we organize our experiences and our memory of human happenings mainly in the form of narrativestories, excuses, myths, reasons for doing and not doing, and so on'. Since narrative theory rejects a strict separation between 'the real' and 'the fictional' or the mythos and the logos, as it has been pervasive both in everyday as well as academic understandings. Drawing on narratology can be extremely helpful for understanding (international) politics epitomised by such controversial figures as Donald Trump. Our own mundane experience underlines that the boundaries are blurred, for instance, if we relate stories of our daily lives that encompass motivations, attempts at explanations and the like, so that they vary according to addressees, temporal distance to the event and new insights gained in each re-telling. It is not a coincidence that current political events (e.g. the Brexit vote, Donald Trump's electoral victory, Erdogan's rise) are often evaluated through the lens of fiction, such as poetry (e.g. Amanda Gorman's poem at Joe Biden's inauguration), TV series (e.g. The Handmaid's Tale), and films (e.g. The Joker); or that politicians, in turn, seem to take cues from such fictional representations to orchestrate their performances, such as Erdogan, for example, who uses the national sentiment evoked by the popular TV series Ertuğrul for pushing his own 'New Turkey' narrative (Khan 2020). Narrative approaches thus also move social science research closer to the research agenda of cultural, film and literary studies by reconsidering the creative and strategic capacities in storytelling activities of all kinds (Park-Kang 2022).
Narrative constructions of the world can be seen as attempts to make sense of reality, both as introspection (ontological narratives) and with regard to audiences of prospective voters or audiences (political narratives). Storytelling is subjective and linked to practical judgments of selective interpretation, personal experiences and sequencing of events (Somers 1994, 616), which are not necessarily real (Patterson and Monroe 1998, 316). Narratives are organised in particular configurations, or plots' which 'weave together a complex of events to make a single story' (Polkinghorne 1988, 19). The success of narratives, referring to the resonance they produce, depends on established sociocultural narrative conventions that build on cultural repertoires, including iconic images or shared (often implicit) stories that do not necessarily make sense in other communities, for instance national security narratives (Krebs 2015) or fairy tale plots. Rather than depending on the material capacities, i.e. authority, of the storyteller, storytelling is thus embedded in cultural practices of communication, which means stories are successful only if they resonate with their specific audience. Accordingly, for researchers, knowledge of repertoires and context is a prerequisite to capture different dimensions, similar to any meaningful interpretive analysis, and thus also benefits from joint work that can test different views and perceptions. The choice of plots and characters is thus (culturally) limited, as they need to be linked to (implicit) understandings in particular contexts. The classic plot genres such as romance and tragedy do not only work in literature and films; they also shape our practices of political storytelling in policy issues (Stone 2002) and international politics (Kuusisto 2019).
How narratives are configured and how knowledge is selectively appropriated is primarily a matter of claims to power and authority, not of the authors but the stories. Selecting the beginning of a story, for instance in processes of transitional justice, is already an intrinsically power-imbued action because it determines which information disappears, and which events are kept alive (Koschorke 2018, 196). Since transformations of actors into characters in the course of storytelling imply moral judgments, it makes a difference, for example, whether migrants are described as human beings seeking protection or as potential criminals (Freistein and Gadinger 2020, 222)such as in Donald Trump's portrayal of Mexican men as 'rapists'for the sake of justifying his project of a border wall. In the context of Trump's pet project of a border wall to Mexico, we find a narrative that established a simple inside/outside plot and equated potential migrants with many different threats such as drug trafficking, rape or even (Muslim) terrorism. In doing so, it also posited Trump himself as the very masculine hero who would protect innocent US citizens by building a massive fence (persistently referred to as a 'wall'), even though most of the narratively constructed external threats have identifiably been domestic, such as an endemic drug crisis and home-grown terrorism.
Political storytelling has its own logic, but is not completely detached from the larger cultural repository of familiar stories. As noted above, not all stories are equally successfulin the sense of their receptionwhich can be in part explained by their general appeal. This has to do with the way in which people can relate to stories, i.e. the resonance stories invite, either because they have heard similar ones before or because they enjoy them. Self-representations in particular are geared towards appealing to a larger public; they are thus both part of broader political narratives and specific incidences of condensed ontological narratives, i.e. narratives that make sense of one self (Patterson and Monroe 1998).

Visuality and the power of images
As earlier studies have shown (Adler-Nissen, Andersen, and Hansen 2020; Bleiker 2018a; Bleiker et al. 2013), images, in addition to roles and plots, serve as constitutive elements of narratives that can be analysed in their complementarity and offer focal points for analysis. Following art history, images therefore do not play a passive role, but can be understood as an 'iconic act' (Bredekamp 2015) that underlines the performative dimensions of showing and seeing (Schlag and Heck 2020), similar to the ideas of speech act theory. As elements that become entry points to narratives, images enable the reconstruction of contexts in the same way that they function in political storytelling to structure collective sense-making. In our understanding, images are elements of narratives but not narratives in themselves. They provide access to interpretations that are part of cultural repertoires, i.e. meaning that a given cultural community considers as acceptable interpretations; images thus evoke associations implicitly known to people of the same community, e.g. shared generational, gender or national contexts. This dimension turns every narrative analysis 'into an activity of cultural analysis' (Bal 2009, 12). Similar to metaphors, imagesincluding videos, memes or similar pictorial itemsare shortcuts to identifying narratives, since they incorporate different layers of information, ranging from the obvious surface of what is depicted to allusions and intertextual references that create (and possibly limit) horizons of interpretation (Yanow 1997, 132). These layers are interrelated but each bring their own meaning and surplus of meaning (such as associations and intertextual references), which means that recipients may not each have the same responses to each layer but can be partly addressed by the overall impression or parts of a visual narrative. As said above, in our understanding, images in political narratives are not by themselves visual narratives but part of storytelling and as such never detached from a broader narrative. Our analysis builds on different methodological suggestions, including iconological approaches (Heck and Schlag 2013) in the tradition of Erwin Panofsky (see also Mitchell 2006) and the study of images with regard to their emotional appeal (Bleiker and Hutchison 2008). Recent studies in IR that incorporate images (Adler-Nissen, Andersen, and Hansen 2020; Bleiker et al. 2013; Crilley and Chatterje-Doody 2020; Freistein and Gadinger 2020; Shim 2017) instruct our visual narrative analysis that follows a process of layered interpretation, which often starts with images or moving images in different forms of media as methodological entry-points for exploring further layers of meaning in their narrative context. The pictorial, aesthetic or visual turn in IR has inspired scholars to pay attention to previously sidelined objects and visual artefacts such as images or videos as sources of political narration, both as material for analysis and as a means of communication (Harman 2019; also Callahan 2020). Visuality can be seen as a distinct form of communication (more on this: Vuori and Saugmann Andersen 2018), as artefacts such as monuments or mosaics are often more immediately evocative of emotional responses than textually mediated material.
Following Bill Callahan's (2020, 19) idea of sensible politics that deconstructs 'visual images in order to lay bare the ideologies that they illustrate', we agree that 'visual artefacts are more than illustrations of ideology, because they can have their own agency: visuals can actually 'do' things, and 'make' things'. For these 'doings' and 'makings', emotions play a pivotal role: 'Part of what makes images unique is that they often evoke, appeal to and generate emotions' (Bleiker 2018b, 9). However, images need to be translated in political storytelling to adequately address potential audiences. Therefore, '(s)tudying the polities of visuality involves understanding not only the role of imagesstill and moving onesbut also how visual artefacts and performances take on political significance' (Bleiker 2019, 118); here, the multi-modal approach provides further context and links them to broader notions of politics. Our multi-modal analysis combines these conceptual and methodological considerations of visual and non-visual aspects in political storytelling as interlinked but nevertheless analytically distinct. Accordingly, our interpretation analyses different layers of representation (imagery, written/spoken text, inter-visual and intertextual linkages) first separately and then in conjunction with each other.

Layering as an interpretive tool of exploring underlying narratives
As stated, our research technique can be understood as layered interpretation, which we have developed and empirically applied to analysing explicit and implicit messages of right-wing populism in election campaigning in our earlier interpretive work . The steps of the analysis follow from the material at hand and can vary in terms of the depth of description but generally revolve around the same analytical procedure. Starting with the in-depth study of an image, our reconstruction of narratives then continues by contextualising images to add layer upon layer of interpretation, first zooming in and then out again from individual elements to a broader picture. Since symbolism used in images can mean different things in different contexts, contextualisation is key. The interpretive moves of analysis are geared towards understanding the underlying presuppositions of the imagesmostly in terms of attempts at narrative closureby reconstructing their cultural (not only in a narrow, national sense) context, which can, for instance, refer to the situatedness of recipients such as their economic status or broader worldviews. Our focus on the interlinkages between images and affective content within certain narratives and understanding the potential effects of these narratives is part of this interpretive step (see also Bleiker and Hutchison 2008).
Finally, identifying the contestation between narrativesor polyphony, as narratology terms it (Koschorke 2018)is the last step. Contestation may be aimed at undermining the validity of a certain interpretation and situate narratives in the context of political competition. Particularly since our analytical interest always remains wedded to our own disciplinary context of political science, we mediate between a more in-depth interpretive analysis and broader contextualisation that is interested in the political landscape into which such narratives are embedded. This also means that the interpretation may never be fully explored in all detail, since the interplay of layers is being prioritised over the exhaustive exploration of aesthetics. As we show in our study of right-wing populism (see also Unrau 2020, 2022), visual narrative analysis allows us to reconstruct how different layers of meaning can be combined, work together or develop 'productive friction' (Freistein and Gadinger 2020, 235). Images and their associated narratives operate through a dual strategy of attempting interpretive closure and simultaneously leaving room for a variety of interpretations, which produces the appeal of narratives aimed at reaching a broader audience.
To put it in terms of a concrete research approach, VNA can be structured as a threestep interpretive process, which resonates with similar accounts interpreting visuality and narratives (Bleiker et al. 2013;Heck and Schlag 2013;Wagenaar 2011;Wiesner, Palonen, and Haapala 2017;Yanow 1997). In a first step, we describe and analyse the image; composition, atmosphere, symbolism etc. can be studied in isolation from all other contextual information. Furthermore, we then take into account the cultural allusions of images, providing an inter-visual context of the presentation, depending on our own interpretive horizon and its limitations. In a second step, we add the narrative context, in terms of plots, roles and relations. The context refers to inter-textual (or inter-image) relations, narrators and potential audiences that are (implicitly) addressed and more immediate context such as concrete occasions (e.g. summits or official meetings) or even content (e.g. who or what is represented in an image). Here, different images can be related to each other to show how a political narrative is created around metaphors and images. Identifying larger, polyphonic narrative structures adds a third layer, which here refers to the reception and challenge of the narratives in the context of domestic and/or international politics. This third layer is most explicitly in line with political science research in general, as it draws on the context of political contestation, both of the individual self-representation and the broader narratives.
How Merkel became 'the leader of the free world': a visual narrative analysis of competing images Our reconstruction of the following visual narrative illustrates how our approach can combine different interpretive layers into a comprehensive analysis of international politics and diplomacy.

Visual layer
The following image, taken by German photographer Jesco Denzel and issued by Angela Merkel's team on her official Instagram account as well as by her spokesman Steffen Seibert on Twitter during a meeting break at the G7 summit in 2018, has become iconic. Even right after it was posted, many news agencies took up the image and commented both on the aesthetics and the content; a rare success in going viral in the global public. The image was issued with a caption that read: 'Day two of the G7 summit in Canada: spontaneous meeting between two working sessions. #G7Charlevoix #g7 #g7summit #multilateralism #multilateralismus', a very matter of fact, descriptive text oddly in contrast with the beautifully composed, politically punchy image. Appraisal of the composition as well as satirical re-interpretations soon could be found on the internet. Most observers agreed that the aesthetics resonate with a Baroque painting (whereas others felt they evoked a Renaissance feeling), underlining the iconic quality of the image and its different intervisual trajectories. For that reason, an innocent, untarnished interpretation of the image and its narrative context is not our goal here. Instead, against the background of studying the image as part of a public diplomacy strategy, in this case of the German delegation, and as part of one among many competing narratives, i.e. other countries' diplomatic strategic interventions, our visual narrative analysis builds on earlier studies to drive home our methodological and analytical main arguments. We will start by going through the different layers we described above.
The image of state leaders meeting during the G7 summit is kept in a muted, bluish tone. In fact, blue is the dominant colourranging from a powder blue silk jacket worn by German chancellor Angela Merkel to the navy blue of suits worn by Japanese prime minister Shinzo Abe, French president Emmanuel Macron and others. The mise-en-scène is what sets the image apart from many other such iterations of summit pictures. Superficially, we see a diverse group of peoplepredominantly mengrouped around a table that cuts through the foreground of the image. With the exception of Donald Trump, all characters are standing, while Trump is seated behind the table. Both with regard to the lighting and placement, Angela Merkel is positioned in the optical centre of the image. She is shown with her hands pressed onto the table, leaning slightly towards the seated Trump. On her left, nearly hidden, are UK prime minister Theresa May and Macron, whose ringed right hand is also lightly placed on the table. Next to Merkel we see Shinzo Abe and next to him John Bolton, at that time US National Security Adviser. Other people, presumably aides to the heads of states are placed in the second or third row, standing or coming into the shot while it is taken. From other angles of the same image, we know that Canadian prime minister Justin Trudeau is positioned behind Trump's chair and that Italian prime minister Giuseppe Conte was also present.
The table is an important element for the visual composition of the image, as it marks a line between a bigger group of people with Merkel in the centre, and Trump as a single figure on the other side. The dividing line is further emphasised by the chosen central perspective of the image with two vanishing points: One running from the diagonal table behind Trump's head, another one formed through the views of Abe, Bolton and Trump, who look at Macron on the left side. Through this perspective and the sharp contrast of Merkel and Trump, the viewer gets the impression that the interplay between these two leaders is the most relevant aspect and at the core of the image. This impression is underlined through the scenic choreography of Abe, Bolton and Trump acting in a rather passive mode, underlined by their body posture and their folded arms. Trump does not sit near the table, he is moving back with his chair, which can be perceived as an attempt at creating maximum distance to the others. Merkel, in contrast, is perceived in a clearly active mode by pressing her hands on the table and looking resolute. Trump's facial expression conveys not only passivity, but also indifference, defiance and stubbornness. Finally, the variation in the illumination of the image transports the effect of Merkel in the centre, as she stands on the brighter side, whereas Trump's seating position lies in the dark corner.
The image is a rare specimen in that it manages to be both aesthetically pleasing and meaningful, thus standing out from the routine and often rather generic depictions of summitry. The clear alignment of aesthetics and symbolism, which are for instance reflected in Manichean characterisations (i.e. protagonists vs antagonists) and the inherent tensions displayed (i.e. in the powerplay between standing and seated characters), renders the image perfectly suitable to transport a surplus of meaning that invites fantasynamely seeing Angela Merkel, who is put centre stage, as the shining heroine of world politics.

Narrative layer
As said before, the image has been described and interpreted many times before, 1 which attests to its immediately iconic status, i.e. the resonance it had with a global public. At the same time, interpretations have differed and often focused on one element rather than many at the same time. Beyond the laconic or even somewhat caustic caption 'spontaneous meeting between two working sessions', which created the 'productive friction' mentioned above by ironically downplaying the iconic character of the image, what seemed to hit a nerve with observers is the implied power constellation that comes across in the image. This leads us to the narrative context in which the image is embedded. While some saw Trump, who is seated and looks truculently up at Merkel, as the dominant figure because he is the only one sitting, 2 most observers clearly identified Merkel as the person in charge. Her stance and facial expressionseen as stoic, as neutral as possible or strict, among othersseemed to indicate what many in the world thought: they were fed up with Donald Trump's antics.
The most influential narrative that emerged from the German PR efforts (which managed to be conveniently forgotten, as the image spread, and became secondary to the performative power of the image itself) was thus indeed the confrontation between her and Trump, from which Merkel was seen to emerge victorious. Their gesturing is therefore symbolic. Merkel's hands are placed flatly onto the table and underline her strong determination, whereas Trump crosses his arms and thereby hides his hands in a rather defensive mode of body language. Other images taken from alternative angles and only moments earlier show a much lighter mood with the participants laughing; again others put different heads of state in the centre, such as Macron, Trudeau or Trump, focusing on their contributions to the scene. And yet, due to not only the beautiful composition of the image (which certainly put it ahead of most of the other shots) but also the predominant anti-Trump message of the narrative, we could observe a public preference for the image selected here. The narrative it resonates with both connects to earlier visualisations of Merkel's relationship with US presidents, among them another somewhat iconic image that showed Merkel (standing) with US president Barack Obama (seated on a bench) in an idyllic countryside scene in German Bavaria, 3 and to the overall fatigue of a liberal global public (as opposed to other political camps) with Trump. The mood captured in the image suggests a constellation in which Merkel chides Trump for his inappropriate behaviour, while he sulks. The asymmetry of positions and lighting puts Merkel at a different level from Trump and recalls claims that she, not (as was traditionally their due) the US president, was the true leader of the Free Worldpartly from the liberal camp inside the US, who preferred her over Trump. 4 The scenic choreography of Merkel in the central position, concentrated and determined to act in turbulent times, and Trump, looking like a petulant child, underlines the visual establishment of a master-pupil relationship. In the context of the rise of populism, the image can be interpreted as a symbolic expression of the ongoing controversy between 'rationality', a belief in multilateralism and cooperative leadership style on the one hand, and 'stubbornness' and a belief in unilateralism (America first) and a narcissistic leadership style on the other. The visual authority of Merkel in this antagonistic relationship also refers to her nickname 'Mummy' ('Mutti' in German) in the German context, which is partly perceived as conveying blind trust in her integrity of doing politics in a rational and competent way. This perception of an asymmetric relationship also shows that the constellation of an image is more relevant than the bodily posture of a leader, as Trump likes sitting in such situations to provoke the impression of being powerful in reminiscence of the kings on thrones, something that does, however, not work in this image.
The gendered dimension of this encounter further adds to the clout of this imagined dynamics, as Merkel famously made the cover of Time magazine as Time Person of the Year (2015), one year ahead of Trump, and also challenged Trump (and others like Vladimir Putin), who had proven his misogynist leanings many times, by her mere presence. Merkel repeatedly responded to (male) leaders in a fashion that many women (and others) can empathise with; she rolled her eyes at Putin 5 or pointed Bulgaria's Prime Minister Boyko Borissov to the right way of wearing a face mask (over his nose) during the Covid pandemic. 6 At the same time, she was notoriously charmed by Barack Obama and got along well with other women such as the International Monetary Fund's Christine Lagarde or EU Commission's president Ursula von der Leyen. Trump, on the other hand, had a long history of treating women like female reporters or politicians with contempt.
Therefore, the fact that a woman between so many men is Trump's direct antagonist here may have evoked feelings of malicious glee in many observers, who were aware of Trump's attempts to project masculinity (and who had just been forced into a painful handshake with Macron, which allegedly left an imprint on his hand), 7 but may have interpreted the scene as leaving him feeling slightly humiliated. While Merkel is clearly looking into Trump's eyes to calculate the situation, Trump's facial expression, the bodily gesture and the ignorance of Merkel's view signal a general disinterest, which can be interpreted as a form of disrespect against Merkel as a powerful female leader as well as a general indifference towards the sense of the current meeting and multilateral diplomacy. The composition of the image and the symbolic line that runs through the table lead the viewer to the impression that all (male) political leaders form a united front behind Merkel, accept her authority and are happy with her strong leadership and ability to encounter the new egoistic leadership style by Trump.

Competing narratives
John Bolton tried to correct the impression by turning the story around to claim that the US had had enough of Europe (particularly Germany) wiggling out of paying their dues to NATO, tweeting: 'Just another #G7 where other countries expect America will always be their bank. The President made it clear today. No more', 8 but the main interpretation of Merkel in charge prevailed. Debate ensued about the body language of the two leaders, sometimes also looking at Abe whose stance mirrors Trump's to some extent. The crossed arms and vacant facial expression that Trump presents have been interpreted as either very much displaying confidence (added to his spread knees) or as defensive. Since our interpretive take puts the performativity of images first, either could be true, but the image would still be part of different narratives that compete. The competing images by Macron and Trump, which did not have the same performative effect on the public audience, demonstrate that visual storytelling is not as simple as it is sometimes believed.
The French and US attempts at telling a different story about that situation work with the simple move of using images, in which their leaders respectively were seen in the centre of the image, in a talking position in the middle of the crowd of political leaders. Obviously, such a choice of images can mobilise emotions such as pride or even superiority, celebrating strong leadership in a masculine tradition, such as Bolton's tweet that claimed that all figures on the left side of the image could be seen as petitioners who needed America's help to survive. Trump's supporters might have liked this description, which signalled that the lonely man in the right corner was being excluded, but still resisted in the interest of the nation.
However, those images of 'strong leadership' developed no performative power beyond already supportive groups. This evidences the limitations of strategic self-representations. Both competitor images could not deliver the same aesthetically pleasing, fantasy-inducing visuality nor did they provide an opportunity for the viewer to discover more meaning behind the depicted visual moment. In the image provided by Trump's people, he is seated and looks upwards, but more in a helpless, superior expression of responding to excessive demands in a chaotic situation. Macron, in contrast, clearly comes across as a strong political leader, yet cannot overcome the lack of appeal to the viewers' fantasy. The challenge of creating a visual appeal lies in the unstructured and dark composition of the image, which creates a perception that Macron is rarely visible and has to draw strings in the background, for instance through exerting pressure. His natural charisma of leadership, which he often aims at representing in the French presidential tradition of glamourous, stylish grandeur, seems to be easier to develop when he is the clear centre or focus in strategically chosen images without other (taller) political leaders. The genre of representing leadership within a group seems less suited, and Macron is visually juxtaposed by Merkel's calm and collected representation, underlining her seemingly natural authority. The allusions and inter-visual layers, in these cases, were thus much less powerful than in the above case. Furthermore, they did not resonate with a similarly appealing narrative context, as neither Macron nor Trump, to select only these two, were seen as embodiments of the hero character, i.e. as relevant or much-needed leaders in issues of multilateral diplomacy. The gendered dimension of Angela Merkel, a womanof all possible politiciansstanding up to the well-known bully Donald Trump may further have fuelled a collective liberal fantasy.
Timing, visual aesthetics as well as content, and the broader narrative (political) context were thus serendipitously aligned in rendering one image very appealing, whereas others did not resonate equally with a global audience. This global audience, of course, presupposes the idea of an imagined global public that does not as such exist. While we could discuss the Western-centric implications of such an assumption and address the problem of inter-cultural interpretive translation or potential conflicts, our goal is not to take for granted that we could distinguish the respective cultures or interpretive communities. Instead, we assume that the performativity of images allows for a large variety of interpretations that may or may not resonate with cultural repertoires that provide meaning to what observers see. Whether spectators focus on the political aspects of leadership depending on political or national preferences, whether they interpret the scene primarily through a lens of gendered power relations, whether they object to all of the decisions taken in the exclusive forum of the G7 on principle or simply ignore politics, these and many interpretations all depend on very different contexts that would be difficult to simply grasp by assuming a narrow understanding of geographically, ethnically or ideologically bound 'culture'. A cultural repertoire as we understand it provides sources of narrative interpretation such as folk and fairy tales, generational memories of cultural artefacts or (increasingly globalised) entertainment channels such as Tiktok or Netflix, which may or may not homogenise our visual habits. The medium of summit imagery at least has long become a genre in itself that rarely produces interesting or even controversial material.

Conclusion
It is a truth commonly acknowledged that an image says more than a thousand words; the saying, almost a truism, reminds us that we often underestimate the power of images and their performativity, also in contrast to arguments and text in general. Particularly with regard to the emotional dimensions of political storytelling, images can be instrumental. Politicians have always been aware of the power of images and their role in political storytelling, both in creating public personas and in fearing that a misleading image in the wrong situation can cause a political disaster, for instance in campaigning (e.g. when politicians that advocate Christian conservative values are photographed with their illicit lover or when the prime minister and a group of staff members celebrate Christmas parties while everyone else suffers a general lockdown). Social scientists have recently started to focus more systematically on the crucial role of visuality and storytelling in (world) politics, both in interpretivist camps and as big data approaches, since the complexity conveyed in images is an asset of politics that deserves our attention.
In this article, we suggest the methodological tool of VNA for exploring the interrelated practices of political storytelling, i.e. the complex relationship between image and narrative, and the performative dimension that concerns an (imagined) audience. Against seeing either narratives or the use of images as engineered we argue that the connection between different layers of meaning, which we reconstruct in the analysis, creates overlaps, frictions and a surplus of meaning that can produce cumulative but also contradictory effects in the audiences. Beyond the immediate case, a visual narrative analysis can help uncover underlying assumptions that are not made explicit in speeches or texts but that come to the fore when we peel away each layer of meaning. The infamous handshake, for instance, that did not undermine the male friendship between Abe and Trump, nevertheless hinted to (racialised) hierarchies between Japan and the US, which remained implicit but created a fantasy of a superior US that appealed to parts of a domestic US audience.
Similarly, in our example of storytelling with an iconic image of former German chancellor Angela Merkel at the G7 meeting and counter-storytelling by other Western political leaders (Macron, Trump), we show how the challenge to or even reversal of gendered hierarchies, in which masculinity is key, was created by the implied power play performed through the image. In line with our argument that visual storytelling in political issues such as international politics can neither be predicted nor instrumentally controlled, we see this as one reason why our interpretive approach can support critical research that renders underlying power structures more explicit. At the same time, the analysis shows how the interplay between the different layers, including aesthetics, content, associations and political context transported the image to a much greater success than any blatant stance of power by Angela Merkel or official narrative of portraying her as a world leader could have achieved. Observers agreed that Merkel, as representative of Germany, benefited from this ascribed role of being the female leader of the liberal world, who countered an overly masculinity obsessed Donald Trump and challenged his political authority. However, even though Merkel's team had previously produced some other iconic images, it would not have been possible to predict or perfectly engineer the resonance this visual storytelling evoked. Furthermore, no lasting benefit for a post-Merkel Germany could be gained; current German chancellor Olaf Scholz certainly lacks status beyond being the leader of a wealthy Western country.
The iconic status of the G7 image was not merely the result of a perfect shot but further depended on the consistent linkages between the image and its narrative context. The photo was taken in a moment when the world seemed to be increasingly exhausted by the question of how to deal with an egomaniac figure like Trump. The large international camp opposing Trump was looking for a rational and powerful alternative voicewhich Merkel managed to embody in a fitting and new way of female leadership. Thus, strategies and means of visual storytelling, which are fundamental elements in public diplomacy, should not be overestimated in their strategic qualities, because the appeal of visual storytelling presupposes a relationship between image and narrative context that evokes a fantasy of consistency and credibility. Our suggestion of a visual narrative analysis thus promises to be a method for conceptualising the complex relationship between images, narratives and emotions as well as for guiding empirical research in methodological terms. The study of international politics would possibly benefit from more narrativeand visuality-based research to show the current changes in diplomatic activities across the world, as would many different fields of political science beyond that.