Pluralist Methods for Visual Global Politics

Introduction

Images play an increasingly important role in global politics, so much so that some speak of a ‘pictorial turn’. ¹ Our understanding of terrorism, for instance, is inevitably intertwined with how images dramatically depict the events and actors in question and with how politicians and the public respond to these depictions. Images are, of course, not new, nor have they necessarily replaced words as the main means of communication. But images are now produced and circulated in ever faster and more complex ways and in the context of a rapidly changing global media economy. Understanding the political nature and impact of images has thus become more challenging too. Several methodological problems stand out.

Images work at numerous overlapping levels: across national boundaries and between the physical and the mental world. Images come in still and moving versions, each posing unique methodological challenges. ² Images work differently from words. They are of a non-verbal nature but we, as scholars, need words to assess their political significance. Something inevitably gets lost in this process. The meaning of images is always dependent on context and interpretation. This is why there is always a certain excess to images, a kind of ‘surplus value’ that escapes our attempts to define them definitively. ³ Add to this that images often work through emotions, which have traditionally been seen as personal and internal phenomena that pose similarly thorny methodological challenges.

A paradoxical situation emerges: while international relations scholars increasingly recognise the importance of images, few if any of them contemplate the methodological issues at stake. Those who work on images tend to come from post-structural backgrounds and the perception here is, as Lene Hansen put it, that methods ‘don’t mix’ with post-structuralism. ⁴ W.J.T. Mitchell, perhaps the leading proponent of the pictorial turn, simply says that ‘no method is being offered here’. ⁵ This is not to say that there is no work on visual methods. There are, indeed, numerous methods textbooks that either contain visual components ⁶ or focus exclusively on visual methods, ⁷ but they rarely engage political and international themes.

The purpose of this article is to engage this gap and to offer a methodological framework for the study of images in global politics. One point is immediately clear: the politics of images is far too complex to be assessed through a single method.

I thus begin by outlining the need to draw on a wide range of methods that are not usually used in combination with each other, including ethnography, semiotics, discourse analysis, content analysis and experimental surveys. Building on Gillian Rose, I show how multiple methods can assess images across a range of sites and modalities, from the production, content and impact of images to their technological, compositional and social dimensions. ⁸ Such a move seems commonsensical and uncontroversial at first sight, but it is actually riddled with obstacles. For one, only a few researchers possess the methodological skills to navigate across such a wide range of methods. Their work is rendered even more difficult by an antagonistic opposition that still divides quantitative and qualitative approaches to international relations research.

The biggest obstacle to a truly multi-method approach to the political study of images is, however, not of a practical but of an epistemological nature. Using methods as diverse as discourse analysis and quantitative surveys can only be done if each of these methods is given the chance to work according to its own logic.

I thus argue for a heterogeneous combination of seemingly incompatible methods. Expressed in other words, multiple methods should be used even if, or precisely because, they are not compatible with each other. I draw on assemblage theory and the concept of rhizomes ⁹ to defend an approach I believe is both necessary and controversial: necessary because it is the most convincing way to understand the complex links between images and politics; controversial because such a strategy breaks with deeply entrenched social scientific conventions that require each methodological component to behave according to the same coherent overall logic. ¹⁰

I illustrate the issues at stake by engaging a particularly important challenge: the question of impact. Images clearly matter in global politics but how exactly do we know? Prevailing social scientific models of cause and effect are of limited use. Only in rare instances do images directly cause political events. In most cases the impact of images is more diffuse. There are, for instance, clear links between the dramatic images of the terrorist attacks on 11 September 2001, the highly emotional rhetoric of good versus evil that emerged in response, and the ensuing war on terror. But these links would be very difficult – if not impossible – to assess with cause–effect models. An assemblage-inspired approach relies on multiple methods to appreciate how images perform the political in more indirect ways. Images establish what William Connolly called ‘the conditions of possibility’. ¹¹ This is to say that they frame what can be seen, thought and said. In doing so, they delineate what is and is not politically possible. Expressed in other words: how we visualise the political shapes the very nature of politics.

The final part of the article addresses the charge of relativism that inevitably comes up in the context of a pluralist approach that abandons the notion of a single standard of evidence. The fear here is that once we give up a fixed set of criteria to judge the world, we will no longer be able to advance a coherent and effective scientific or political position. I show that this fear is misplaced. I demonstrate that genuine multidisciplinarity not only offers more accurate and self-reflective insights into visual world politics but also serves to check what is one of the biggest problems in international relations research, that is, unchecked reification: the tendency to uphold and rehearse one subjective position to the point where it becomes so widely and seemingly objectively accepted that its subjective origins become erased.

A brief disclaimer is in order before I start. My prime focus lies less on the politics of images than on the methodological challenge of understanding them. This is the case because an appropriate methodological framework offers the very precondition for understanding how images work politically. Given the elusive nature of images, a pluralist approach is particularly essential. But the so gained methodological insights have implications that go beyond the study of visual politics. The world is and always has been far too complex to be understood through social scientific methods alone, no matter how sophisticated they are. We need the full spectrum of knowledge to understand and face the challenges that make up global politics. It is in this sense that my inquiry into images contributes to broader ongoing debates – played out in the pages of this Millennium special issue – about how to widen the methodological purview of international relations research.

Methodological Challenges in the Study of Images

There is something unique about images. They have a special status. They generate excitement and anxieties. ‘Why is it’, Mitchell asks, ‘that people have such strange attitudes towards images?’ ¹² Why is it that audiences are given a stern warning before they see shocking images of, say, war or terror or bodily mutilation? Why not the same warning with verbal depictions? ¹³ Consider how news outlets that published images of the bombing of the Boston marathon in 2013 felt compelled to add notes that read ‘Warning: This image may contain graphic or objectionable content’. ¹⁴ No such warning was given with language-based articles of the same event, even though they described the horror of the attack in equally great detail. What makes images seemingly more dangerous and powerful than words?

There is, indeed, something unique about images. They work differently from words. They are, by definition, different from words. Whether they be photographs, films or visual art, images always contain a certain excess, a part of them that escapes our understanding. Images do not speak for themselves. They need to be interpreted. And this interpretation contains values that inevitably have as much to do with the values of the interpreter than the content of the image itself. Roland Barthes writes of a ‘connoted message’, of how an image is read and how it fits into existing practices of knowledge and communication. ¹⁵ Consider how an early modern icon of the Madonna and child would have been perceived differently at the time it was painted than today, when the very same image is seen in a museum in St Petersburg.

Part of what makes images unique is that they evoke, appeal to and generate emotions. Pictures of traumatic events, such as terrorist attacks, natural catastrophes or aeroplane crashes, are seemingly able to capture the unimaginable. This is why news coverage of such traumas is frequently accompanied by images, as if they could provide readers/viewers with a type of emotional insight that words cannot convey. But emotions are notoriously difficult to assess. They have traditionally been seen as internal and private phenomena: you can, by definition, not know how I feel. All you can know is how I tell you how I feel. This is why Neta Crawford notes that emotions are ‘deeply internal’, making it very difficult to isolate or distinguish what a ‘genuine’ emotion may indeed be. ¹⁶ Quantifying emotions is difficult. Labelling and measuring them, even in qualitative terms, is also a delicate process. That these problems exist at the level of the individual is significant enough, yet the difficulties are magnified when it comes to examining the role of images – and the emotion they engender – in the context of global politics.

These conceptual challenges are exacerbated by the ever increasing speed and complexity through which images circulate. Images seem to be everywhere. It is not just that global media networks now cover news events 24 hours a day. The issue goes well beyond the influential CNN-effect. ¹⁷ The circulation of news is not just global but transcends global media networks. Even traditional newspapers – from the New York Times to Le Monde, Der Spiegel and the Guardian – are meanwhile multimedia organisations with a substantial internet presence. ¹⁸ They cater to an audience that consumes news increasingly through smartphones, tablets and other mobile devices. Add to this that the visual field has become increasingly democratised. Anyone with a mobile device can now take images and distribute them via Facebook, Twitter, Instagram, Tumblr and other social media tools. The results are fundamentally new interactive dynamics that are rooted in various networks and webs of relations. ¹⁹ Look at the recent Arab spring uprising in Egypt. One of the most remarkable episodes occurred when a young woman blogger called Aliaa Elmahdy posted a nude photograph of herself on her blog. She did so to protest against gender discrimination in Egypt and to call for more personal freedom, including sexual autonomy. Her private-cum-public photograph circulated around the world. It caused extensive public protests at home and created a wave of feminist solidarity around the world.

There are significant methodological challenges entailed in understanding this rapidly changing world of visual global politics. Consider, as an example, one of the most widely practised visual methods: content analysis. With traditional media outlets a content analysis is fairly straightforward. One can, for instance, identify politically significant visual patterns by systematically analysing front-page newspaper coverage of particular issues over an extended period of time. ²⁰ Doing so will become more and more difficult. Newspapers are now as much online as print media: website headlines often change by the hour or even minute, so much so that it is no longer possible to identify images that visually symbolise the day’s key news. Retracing this online coverage from past years is far more difficult than retrieving front pages of past print editions. Methodological issues get exponentially more difficult when it comes to applying content analysis to moving images and to internet sites, where there is a limitless number of images. Finding the equivalent units of analysis in such a fluid multimedia environment is thus enormously difficult. ²¹

Towards a Pluralistic Methodological Framework

So far I have pointed out that the elusive nature of images and the speed and complexity through which they circulate pose significant methodological challenges. The world of visual politics is, indeed, so complex that there is only one logical conclusion: to recognise that there is no one method, no matter how thorough or systematic, that can provide us with authentic insights into what images are or how they function.

The beginning of my argument, then, is simple: I advocate drawing on multiple, diverse and even incompatible methods. In doing so, I employ Gillian Rose’s insightful work on visual methods. Rose distinguishes between three different sites: the production of an image, the image itself and how it is seen by various audiences. ²² Overlapping with this triptych are three modalities: technological, compositional and social. Each of these sites and modalities requires entirely different methods, which is why I now discuss each of them briefly.

First is the production of images. Methods here are linked to technical as much as interpretative tools. Part of the task involves understanding how images – still or moving – are taken: not only what kind of technical processes and choices are involved, but also what political and ethical consequences follow. ²³ This matters greatly, for instance, when it comes to photojournalists taking pictures of war or of victims of famine. Epitomising the dilemmas at stake here is the much discussed suicide of Kevin Carter, who struggled with the ethical implications of having won a Pulitzer Prize for his photograph of the Sudan famine in 1994. ²⁴ The photograph, which depicted a vulture watching a starving child, created a major public reaction as well as numerous discussions about the ethics of photojournalism. Carter’s iconic photo is an extreme case of a confrontational image, but every picture of a famine inevitably raises questions. What kind of politics and ethics are entailed in the depiction of misery? Who views and who is being looked at? Who decides what is visually newsworthy and what are the consequences? To what extent is the global circulation of images linked to and further entrenches prevailing power relations? Challenges here involve understanding the processes through which images are produced, selected and, finally, make it into the news – from front-page newspaper coverage to television film coverage to new media sources that go ‘viral’ on the web. The methods needed to understand the issues at stake include, amongst others, interviews and ethnographic inquiries.

Second is the challenge of understanding the images themselves – that is, their actual content. The methods required here are very different. They range from semiotics (which explores how images work through symbols and signs) to discourse analyses (which examines the power relations involved) and content analysis (which empirically measures patterns of how images depict the world). Numerous complexities are involved in the respective inquiries. For one, it is imperative to investigate how images interact with other mediums and objects. There are macro and micro dimensions to this task. At the macro level there is the intertwinement of images with numerous material, cultural, symbolic and other factors, including media environments and, in a more general sense, the entire sociopolitical context within which images gain meaning. ²⁵ At the more micro level there is the need to investigate the link between visual and verbal representations. Mitchell speaks of ‘image–text’ constellations and goes as far as arguing that ‘all media are mixed media’, ²⁶ that there is nothing that is either purely verbal or purely visual. Examples of scholarly works that examine these complex visual dimensions of global politics are far too numerous to list here. Recent examples include work by David Campbell on famine, Lene Hansen on torture, Emma Hutchison on trauma, Frank Möller on peace, David Shim on North Korea and Sharon Sliwinski on human rights. ²⁷

The third site relates to how audiences receive images, or as I prefer to put it: the actual impact of images. And here, too, a different set of methods is required. They include audience interviews and observations. Other options are quantitative surveys that assess attitudinal reactions to images or neuroscientific lab experiments that measure physical responses to visual stimuli. Given the importance of understanding the impact of images, I will return to this topic in more detail later in the article. For now, let me give just one brief example of impact studies: investigations that assess the political perception of and response to images of suffering that circulate in global media networks. ²⁸

At first sight it seems commonsensical to rely on such a broad set of methods to understand the construction, content and impact of images. Many method scholars acknowledge the need for pluralism and recognise that, by extension, their own approach is a ‘necessary but not sufficient methodology’. ²⁹ A content analysis, for instance, can identify important visual patterns but say nothing about the impact of images, just as a survey experiment can gauge impact but offer no knowledge of the origin or content of images. This is why Sarah Pink called for more collaborative and interdisciplinary approaches to visual research. ³⁰ She too advocates a form of collaboration ‘whereby disciplines might learn from each other without seeking narrative folds to assert the supremacy of their own discipline at the expense of others’. ³¹ Numerous other scholars agree and argue against analytically separating sensory domains. ³² David Silverman stresses that because the visual works on numerous levels none of the data generated by specific methods ‘are more real or more true than the others’. ³³ Sarah Elwood argues for a ‘diverse range of visualization practices and visual methods’. ³⁴ Mitchell advises ‘against putting all eggs in one disciplinary basket’ and advocates criss-crossing links between fields such as art history, philosophy, media studies, psychology and anthropology. ³⁵

While commonsensical in principle, the actual application of a pluralist methodological approach to visual global politics is far more complex and difficult. Few international relations scholars try to combine the types of methods required to assess the comprehensive dimensions of visual politics. There is, for one, the practical challenge of acquiring a highly diverse set of methodological skills. Scholars who employ, say, discourse analysis rarely have the skills to conduct large-scale quantitative surveys. Likewise, researchers who, say, do lab experiments are not usually equipped to conduct a semiology. The breadth of methodological skill necessary to conduct interdisciplinary visual research is, as Luc Pauwels notes, truly daunting. ³⁶ But there is more at stake than the challenge of acquiring methodological skills.

Those who venture further and engage in genuine multidisciplinary research face significant obstacles. They might be considered ‘thin’ since they disperse their efforts across a range of complex bodies of knowledge and thus seem to lack the kind of detailed insights that only specialists can provide. As a result, they risk not only being seen as ‘amateurs’ but also missing out on the complexities of disciplinary debates and insights. ³⁷ In addition, such scholars face publishing practices by academic journals that still largely run along – and evaluate according to – disciplinary standards. It is symptomatic, then, that one of the most prominent scholars in visual culture – James Elkins – stresses that ‘the existence of borders, and the competencies they enclose, are what gives sense to our peregrine scholarship’. ³⁸

Most significantly, multidisciplinary scholars in international relations are met with a deeply entrenched antagonistic dualism that continues to separate those advocating qualitative from those advocating quantitative methods. The divide between these traditions is enforced not only by different methodological trainings, but also by a range of epistemological assumptions that seem to make genuine cross-method inquiries difficult. One of the most influential methods textbooks reduces social science to the task of learning ‘facts about the real world’. Both quantitative and qualitative methods are meant to operate according to the same logic: they have to be based on hypotheses that ‘need to be evaluated empirically before they can make a contribution to knowledge’. ³⁹

Such methodological approaches have meanwhile been widely critiqued for their problematic positivism. But their impact remains remarkably strong. In much research in the social sciences there is still a deeply held dualism between positivist and post-positivist approaches, a dualism that mistakenly either validates or discredits methods according to certain epistemological positions. Consider two brief examples. Content analysis often remains wedded to a scientific ethos that stresses the method’s ‘objective, systematic and quantitative’ qualities, even though the actual set-up of the experiments is inevitably contingent on numerous highly subjective decisions. ⁴⁰ Such an ethos of scientific objectivity makes it difficult to embrace more interpretative methods at the same time. But resistance comes from other sides too. More interpretative scholars can be just as suspicious of science as scientists are of interpretation. Consider my own experience. I identify with post-structural and related methods, but recently employed quantitative tools for work on images, including survey experiments and content analysis. ⁴¹ The reaction from like-mined post-structural colleagues was all too often one of deep concern that by embracing quantitative methods I would buy into a positivist epistemology and lose my ability to critically analyse political phenomena. I was accused of ‘selling out to the enemy’.

Method, Methodology, Assemblages

The practical challenges to a pluralistic methodological approach are significant but can be overcome. Extra training can provide scholars with the skills needed to employ a wider range of methods. The hostility between scholars advocating quantitative and those advocating qualitative approaches can, equally, be overcome through a mutual learning process. The more significant obstacles, however, are of a more substantive nature.

A genuinely interdisciplinary and pluralistic approach to the study of images in global politics can only be employed once one abandons the idea that all methods have to operate according to the same rules and standards of evidence; that is, once one abandons the notion of an overarching framework that can regulate all the various inquiries. As opposed to prevailing assumptions, there is no logical link between certain methods and certain epistemological positions. Methods are tools to understand the world and epistemologies are assumptions about the values that should be attached to the knowledge that these tools produce.

This is why methods cannot be employed or understood without a proper engagement with methodologies – a point that is being made increasingly by international relations scholars working on critical approaches to methods. John Law and John Urry, Lene Hansen, Linda Tuhiwai Smith, Patrick Jackson, Michael Shapiro, Claudia Aradau and Jef Huysmans all stress that methods are inevitably intertwined with the strategies that these methods employ and the context within which they are carried out. ⁴² The task of methodologies is to challenge the idea of methods as neutral techniques and to reflect upon the choices and implications that they embody. Implied – and at times explicitly articulated – in these positions is the idea that one can embark on systematic and rigorous research even while one accepts that there are several, and at times even incompatible, models of doing so. Although still fairly controversial in international relations, such critical positions on methodology are not new. They have long been debated in the philosophy of science or in quantum and complexity theory. Consider, just as an example, how Paul Feyerabend argued 40 years ago that the numerous procedures that make up the sciences have no common structure and that, as a result, ‘successful research does not obey general standards; it relies now on one trick, now on another’. ⁴³ He presents the violation of existing basic rules as the very process through which science progresses – not towards a new and better paradigm, but towards a recognition that science, and the methods it applies, is always incomplete and bound by its social context.

The concept of assemblage thinking provides a particularly useful way of anchoring a pluralist approach to the study of visual politics. ⁴⁴ It offers the kind of framework that can assess how images work in intertwined ways across their construction, content and impact. While ideally suited to capturing the complex politics of images, an assemblage approach has wider implications for the study of global politics.

Assemblages can be defined in their opposition to totalities. The latter are systems of thought based on relations of interiority. Manuel DeLanda stresses that in such systems each component has to behave according to a central logic that structures the movement of parts. ⁴⁵ The above positivist methods textbook is a key example of such a coherent and clearly delineated system: it is structured according to an overall logic, that of social science as a science. To make sense and fit in, each methodological component of this system has to operate according to the same principles: those of testable hypotheses. Methods that do not fit these criteria are seen as unscientific and illegitimate.

Assemblages offer a clear alternative to totalities and thus a conceptual base for a pluralist approach to the study of images. This is the case because an assemblage, according to DeLanda, is structured by relations of exteriority: the properties and behaviour of its components neither have to explain the whole nor fit into its overall logic. ⁴⁶ Heterogeneity is a key feature here, for each component is both linked and autonomous. Law and Urry as well as Aradau and Huysmans speak of ‘messy’ methods, but this does not mean that individual inquiries cannot be, at the same time, meticulous, thorough and systematic. ⁴⁷ The key is that the criteria by which they operate are seen as being independent of their specific purpose. An embrace of a post-positivist epistemology is an inevitable side-product of assemblages: an attempt to refuse totalities and embrace life and the political as a decentred, heterogeneous alignment of emerging and constantly moving parts. ⁴⁸ Epistemological positions are then no longer linked to particular methods, but to the value claims that are attached to them. A discourse analysis can be part of a positivist totality, just as a quantitative survey can become integrated into a post-positivist approach – as long as the respective claims to knowledge are seen as contingent and are not advanced in reference to an allegedly value-free and overall frame of reference.

The contours of a genuine multidisciplinary framework now become visible. Once the logic of totality is forgone, it becomes possible to combine seemingly incompatible methods, from ethnographies to semiologies and experiential surveys. The logics according to which they operate do not necessarily have to be the same, nor do they have to add up to one coherent whole. My position here differs from interdisciplinary approaches in visual culture that promote multiple methods but then seek to unite them through a synthesis, whether this is an ‘integrated framework’, ⁴⁹ a ‘systematic framework’ ⁵⁰ or a ‘deeper holistic understanding of the world’. ⁵¹

The approach I argue for understands method not as an internally consistent system that can be united through a synthesis, but more like Deleuze and Guattari describe assemblages: as a rhizome, a type of loose network of methodological connections that has no central regulatory core but, instead, operates at various interconnected levels, each moving and expanding simultaneously in different directions. ⁵² Deleuze and Guattari juxtapose rhizomes to roots or trees: hierarchical systems in which one becomes two, in which everything can be traced back to the same origin. Roots and radicles may shatter the linear unity of knowledge, but they hold on to a contrived system of thought, to an image of the world in which the multiple always goes back to a centred and higher unity. A rhizome works as an assemblage: it is not rooted, does not strive for a central point. It grows sideways, has multiple entryways and exits. It has no beginning or end, only a middle, from where it expands and overspills.

Understanding the Impact of Images beyond Causal Models

I have taken a slight detour from visual politics and outlined the foundations of a pluralist framework that has the potential to understand how images work in complex ways across their construction, content and impact. I now tackle a particularly challenging task that opens up with such a move beyond an exclusive reliance on the type of social scientific methods that have dominated international relations scholarship: the challenge of demonstrating how exactly images matter politically.

We all know that images matter, but how do we actually know that what we know is accurate? For instance, what is the exact political impact of an image – say, a photograph of a tsunami victim on the front page of the New York Times? People around the world are inevitably influenced by seeing a humanitarian tragedy depicted through the photograph of a suffering individual. But what is the exact impact of this image and how are we to assess it methodologically?

Prevailing social scientific methods assess impact through causal models. They revolve around ‘a logic of stability and linear causality’. ⁵³ But there are only rare instances where causality can be attributed to images. Consider one example: the debate on the use of torture in the war against terror. As early as the summer of 2003 it was publicly known – in part through reports from Amnesty International – that US troops were using torture techniques when interrogating prisoners in Iraq. There was, however, little public interest in or discussion about the issue. Domestic and international outrage only emerged in the spring of 2004, in direct response to graphic photographs of US torture at the Abu Ghraib prison facilities. The intensely emotional images of torture managed to trigger major public discussions in a way that ‘mere’ words could not.

In most instances the power of images is far more difficult to identify through causal mechanisms. This is even the case when the influence of images is obvious and uncontested. It would be difficult, for instance, to retrace causal or even constitutive links between the dramatic visual representations of the 9/11 terrorist attacks, the emergence of a discourse of evil and the ensuing war on terror. And yet, hardly anybody would question that images were a key part of the nature and impact of 9/11 or, for that matter, almost any political event.

Images work in complex ways, criss-crossing a range of geographical and temporal boundaries – all the more since new technologies, from global media networks to new media sources, now allow for an ever faster and easier circulation of images. To understand the political dimensions of this process we need to supplement social scientific models of causality with methodological strategies that acknowledge the multidirectional and multifaceted dimension of political events. ⁵⁴

Images often work more indirectly, by performing the political, by setting the ‘conditions of possibility’ through which politics takes place. ⁵⁵ They have the potential to shape what can and cannot be seen, and thus also what can and cannot be thought, said and done in politics. Consider, as an example, how a mixture of discourse and content analysis can facilitate understanding of how media images have framed Australia’s approach to refugees. ⁵⁶ Over a decade asylum seekers have primarily been represented as medium/large groups and through a focus on boats. This visual framing, and in particular the relative absence of images that depict individual asylum seekers with recognisable facial features, associates refugees not with a humanitarian challenge, but with threats to sovereignty and security. But asserting a direct causal link to specific policy outcomes would be impossible. And yet, a pluralist approach that draws on multiple methods can reveal how these dehumanising visual patterns played a key political role by framing the parameters of debates. In doing so, they reinforced a politics of fear that explains why refugees are publicly perceived as people whose plight, dire as it is, nevertheless does not generate a compassionate political response.

In situations where direct causality is impossible to ascertain one could perhaps speak of ‘discursive causality’ ⁵⁷ or ‘discursive agency’. ⁵⁸ Such an approach would retain a notion of impact but acknowledge that images work gradually and across time and space: their influence crosses numerous borders – spatial, linguistic, psychological and other ones – and unfolds only gradually. Doing so illuminates how images work inaudibly but powerfully: by slowly entrenching – or challenging – how we view, think of, and thus also how we conduct, politics. Only a multitude of methods, qualitative and quantitative ones, can attempt to stitch together the intricate and non-linear processes through which visual factors shape the political.

Relativism, Reification and Situatedness

By drawing on but reaching beyond causal models, a pluralist method can offer key insights into the impact of images. But this is only one example of how the complex political role of images cannot be understood by one method alone or even by one methodological standpoint. Only a combination of heterogeneous methods can hope to capture how images intersect across their construction, content and impact. But advancing such a pluralistic methodological framework is neither easy nor uncontroversial, for it breaks with deeply held assumptions in social scientific research and in international relations scholarship in particular.

For many scholars the danger of relativism opens up as soon as one abandons a fixed and universal standard of evidence. The assumption here is that social science proper requires an internally coherent framework that can establish ‘facts about the real world’. ⁵⁹ Abandoning such a framework is said to undermine both scientific rigour and political judgement.

Not so. Fears of relativism are misplaced. Several reasons stand out. To address them it is necessary to momentarily move away from the specific study of images and address larger issues that deal with how a pluralist methodological framework can assess the complexity of world politics in a way that remains rigorous and politically insightful.

First, and most simply, there is no choice. We do not have access to methods that somehow generate neutral and value-free knowledge. There is an overwhelming consensus among scholars who engage critically with methodology: they all point out that methods are not and should not be seen as tools that reflect and represent realities. Methods are inevitably embedded in social practices. While describing reality, they are also, at the very same time, enacting and constituting this very reality. Methods inevitably select and present the world from a particular perspective, one that conceals as much as it reveals. ⁶⁰ This does not mean that realities do not exist and we can just make up whatever we like. Rather, it requires recognising that making sense of the world inevitably entails both material and social dimensions and that, as a result, ‘different methods or practices tend to produce different realities’. ⁶¹ This very recognition lies at the heart of my suggestion to approach visual politics through a framework that allows multiple and at times incompatible methods to coexist without having to answer to the same central regulatory rules.

Second, rather than being doomed, a pluralist framework that accommodates multiple and at times incompatible methods can serve as an antidote to one of the biggest challenges in international relations scholarship: ‘unchecked reification’. In making this point, I draw on a recent and highly compelling study by Daniel Levine. ⁶² Reification refers to the widespread and dangerous process of forgetting the distinction between concepts and the real-world phenomena they seek to depict. The dangers are real, Levine stresses, for international relations scholars deal with some of the most difficult issues, from genocides to war. Upholding one subjective position without critical scrutiny can thus have far-reaching consequences. Assuming that the world cannot be known outside of our human perceptions and the values that are inevitably intertwined with them, Levine presents reification not as a flaw that can be expunged, but as a priority condition for scholarship. The challenge then is not to let it go unchecked. Reification can be countered through what Levine calls a form of self-reflection, a kind of ongoing, inward-oriented sensitivity to the ‘limitations of thought itself’. ⁶³

The benefits of drawing on a range of approaches, even incompatible ones, go far beyond the opportunity to bring out nuances and new perspectives. Once the false hope of a smooth synthesis has been abandoned, the very incompatibility of the respective perspectives can then be used to identify the reifying tendencies in each of them. This is, for Levine, how reification is being ‘checked at the source’, and this is how a ‘critically reflexive moment might thus be rendered sustainable’. ⁶⁴ It is in this sense that Levine’s approach is not really post-foundational but, rather, an attempt to ‘balance foundationalism’s against one another’. ⁶⁵ A scholar oscillates back and forth between different methods and paradigms, trying to understand the event in question from multiple perspectives. No single method can ever adequately represent the event or should gain the upper hand. But each should, in a way, recognise and capture details or perspectives that the others cannot. ⁶⁶

Third, and finally, if each method performs and enacts the social world, then we need as much awareness of this process as possible. Required, then, is an explicit engagement with the situatedness of knowledge. A researcher cannot pretend that she or he had nothing to do with the choice of method or the creation of data; that somehow all the information had been there already and the research simply consisted of unearthing hidden gems that are then presented in their original authenticity to a reader.

The issues at stake here are particularly pertinent when it comes to the study of visual politics. Images always need to be interpreted. They have no meaning on their own. Their meaning is contingent on other images and on the verbal context in which they are embedded. There is, thus, always a leap of meaning that forces a scholar to offer a particular interpretation. This interpretation is never definitive. It is always linked to particular methodological choices.

Rather than hiding the choices made in the interpretation of images, scholars should expose them and lay bare the paths taken and forgone. This is why Sarah Pink stresses that for visually oriented researchers ‘reflexivity should be integrated fully into processes of fieldwork and visual or written representation in ways that do not simply explain the researchers’ approach but reveal the very processes by which the positionality of researcher and informant were constituted and through which knowledge was produced during fieldwork’. ⁶⁷ Such forms of reflexivity are not only pertinent to field research but apply to all methods, whether field-, archive- or text-based. ⁶⁸ But in all of these and other cases, self-reflectivity will always remain incomplete. The very nature of discourse is that there is no outside: we are inevitably caught in a web of meanings and we can only be aware of part of them.

For some scholars such forms of reflective situatedness must be supplemented with an additional component. Drawing on Jackson, ⁶⁹ Cecilie Basberg Neumann and Iver Neumann stress the need to distinguish between the above-described reflexivist situatedness and a second one they call ‘analyticist’. These two are located at the ‘opposite ends of the researcher/informant relationship’: the former is about awareness of how the researcher is influenced by his or her environment, while the latter focuses on how ‘the researcher affects the informants’ and the research environment in general. ⁷⁰ As Smith puts it in the context of indigenous research methods, whose interest does the research serve? How and to whom will it be distributed and with what consequences? ⁷¹ Law and Urry had already drawn attention to this distinction, though they depict it as a move from epistemology (how the researcher is shaped by context) to ontology (how a researcher enacts and shapes a world ⁷² ). I see these two approaches as not as tension-fraught as others do. For me the former is part of the process that leads to the latter, with both of them highlighting different aspects and consequences of situated knowledge.

Conclusion

The main objective of this article has been to outline the contours of a methodological framework that can assess the increasingly complex role that images play in global politics. I started with a simple proposition: that a range of different methods are necessary to understand the construction, nature and impact of images. Although seemingly uncontroversial, such a proposal faces significant practical obstacles. This is in part why scholars in visual culture have long debated whether the study of images should be a disciplinary or an interdisciplinary scholarly endeavour. ⁷³ Multidisciplinarity might be in academic fashion, but international relations scholars who embark on related research face numerous difficulties. They range from disciplinary-bound publication practices of key journals to the perception that a scholar who navigates multiple bodies of knowledge cannot possibly acquire the level of expertise of a specialist. Add to this a persistent antagonistic divide between scholars who advocate quantitative and those who advocate qualitative methods. They often live in different scholarly worlds – worlds that are separated by the training scholars receive, the outlet where they publish and the kinds of scholarly projects they embark on. Crossing the lines between these two traditions is not easy, in part because it requires a broad methodological training that most scholars do not have, in part because it triggers a misleading epistemological battle between positivists and post-positivists. I have shown that these epistemological issues are not linked to the methods scholars employ but to the values that they attach to the results of their research.

I then defended my pluralist methodological framework as the most plausible way to understand how images work across three overlapping realms: their production, their content and their impact. The complex realm of visual politics cannot be appropriately understood through a single method or even a methodological framework that revolves around an internally coherent and closed logic.

The key argument I have made is that different methods need to be given the chance to work based on their own logic, even if they are not compatible with an overall set of rules. I then illustrated the issues at stake with regard to the political impact of images. Images clearly play a key role in global politics. Images are absolutely essential to understanding, for instance, the impact of a terrorist attack, the solidarity that emerges after a natural disaster or the dynamics that unfold during an election campaign. These and countless other political phenomena cannot be separated from the manner in which they are visualised in both old and new media sources. But images only rarely cause political events directly. Prevailing cause–effect models are thus only of limited use. Images work indirectly, across space and time. It is only through a combination of multiple and at times incompatible methods that we can understand how images frame the conditions of possibility; how they influence what can and cannot be seen, thought and discussed; in short, how they delineate and shape the political.

My methodological reflections on visual global politics have inevitably been long on epistemology and short on ontology. This is to say that I have spent more time on identifying methodological challenges and offering solutions to them than I have on exploring the actual political content and consequences of images. This is in part due to the limits set by a short article, in part the result of recognising that questions of method are the precondition for understanding the politics of images. Methods are not just technical issues. They are about how we know the world and how this knowledge is part and parcel of the very political phenomena we seek to understand.

I devoted the final part of the article to addressing what is the biggest possible objection to the type of multidisciplinary approach I have outlined: the fear of relativism that opens up once one abandons a foundational framework from where everything can be judged in a standard, internally coherent manner. This fear is not only misplaced but also misses out on how using a multitude of methods, even incompatible ones, can play a key role in countering a far greater problem: the reifying tendency to forget how we, as scholars, inevitably impose our subjective position upon a far more complex political world. The hubris of thinking that one can possess definitive and indisputable knowledge is significantly more dangerous than a clash of different perspectives. Indeed, the very combination of incompatible methods makes us constantly aware of our own contingent standpoints, so much so that we can gain the kind of scholarly humbleness and self-reflectiveness required to approach the world of visual politics in all of its nuances and complexities.