‘Compassion, like many of our other complex emotions, has a heady political life’.
Elisabeth V. Spelman: Fruits of Sorrow
Introduction
Compassion is one of the central emotions of modernity. As Michael Barnett (Reference Barnett2011, 49) has recently observed, ‘[t]he revolution in moral sentiments and the emergence of the culture of compassion is one of the great unheralded developments of the last three centuries’. Today, compassion is seen as an important ‘humanitarian’ (Sznaider Reference Sznaider1998, 128), ‘connecting’ (Clark Reference Clark1997, 5) and ‘caring’ (Porter Reference Porter2006, 108) emotion that extends the boundaries of the self and works for the alleviation of human suffering ‘out there’. Because of this, it has been – and still is – a normative and prescriptive emotion with a positive valence and benevolent character: we are compassionate, and if we are not, we certainly should be!
This high regard of compassion stems from the recognition that emotion and morality are connected. Since contemporary Western societies emphasize individual autonomy and freedom, they face the problem of moral obligation: are we responsible for the welfare of others, and why, in a society that views individuals as free and self-interested agents? As a response, it is increasingly said that our moral sense is grounded in – or at least motivated and influenced by – emotion instead of theoretical moral knowledge (Rorty Reference Rorty1998). ‘Morality’, as Prinz (Reference Prinz2009, 2) puts it, ‘is a human construction that issues from our passions’. And passions, particularly benevolent ones like compassion, are seen as irresistible and compelling in their moral role, relating us to others and motivating us to do the right thing.Footnote 1
This is especially the case with recent liberal political thought that – despite conceptual differences – has taken the emotion of sympathy (Rorty Reference Rorty1998) or compassion (Nussbaum Reference Nussbaum2001) as the moral compass without which our ethical or moral sensibility cannot operate. While critical of the popular idea that compassion is a natural passion, these key liberal thinkers still argue that without fine-tuned capacity for painful fellow-feeling we are blind to the suffering of others and to the potentially cruel implications and consequences of our actions. As Nussbaum (Reference Nussbaum2001, 391–2) has put it, compassion is ‘the eye through which people see the good and ill of others, and its full meaning. Without it, the abstract sight of calculating intellect is value blind’.
Despite its clear importance to liberal modernity, compassion has largely been bypassed as an object of analysis in mainstream International Relations (IR).Footnote 2 Until relatively recently, the same could be said of emotions more generally. However, beginning in earnest with Neta Crawford’s (Reference Crawford2000) analysis of ‘the passions’ in realism and liberalism, the discipline has begun to move, text by text, away from its ‘rationalist’ comfort zone in areas of strategic or purposive action.Footnote 3 This development is important, because it acknowledges the rich totality of human experience – that which is subject to analysis – and broadens the realm of ‘properly political’ modes of action to include those areas where emotion arguably features as a central element. Consequently, for at least some in the discipline, affective aspects of political life are being taken seriously.
This article is part of this scholarly trend. It agrees that emotions matter for explaining international outcomes, and, through an analysis of compassion, it generates an additional insight: different kinds of emotions enable different kinds of politics. Take this article’s substantive area of focus, humanitarianism: actions undertaken to relieve the suffering of distant strangers (Barnett and Weiss Reference Barnett and Weiss2008; Walker and Maxwell Reference Walker and Maxwell2009). In global politics, themes of compassion, care, and moral obligation find their starkest expression in contemporary humanitarian assistance, where, as Peter Redfield (Reference Redfield2008, 196) observes, ‘concern for human life serves as a key value in international moral discourse’. To create moral attention to suffering and crisis and accumulate private funding for help, humanitarian organizations rely on the systematic use of emotionally appealing and ‘ethically informed’ advertising campaigns, an example of which is found in Figure 1.
Save the Children (USA) – 14 January 2010 – http://www.savethechildren.org – REUTERS/Eduardo Muno
Emotionally appealing images of suffering constitute one of the core mechanisms, or technologies, of contemporary humanitarianism. An iconic image of distant suffering is said to promote the humanitarian cause more expediently than purely textual forms of communication (Dauphinée Reference Dauphinée2007; Douzinas Reference Douzinas2008, 76). But how is it that emotions are translated into action? What makes an image, such as that in Figure 1, effective as a communicative device – if indeed it is effective? And why is it, as we find in the third section, that humanitarian campaigns tend to evoke similar themes? These questions call serious attention to the complex role of emotions in driving this domain of practice.
We argue that humanitarian action is incomprehensible outside of a general theory of how emotions – and compassion, in specific – structure the encounter between the suffering object of relief and the caring public. Specifically, our theoretically pragmatist and eclectic approach to compassion points us to the ways in which seemingly internal, even hard-wired, emotions are in practice already enabled by forms of productive power. In other words, when humanitarians fundraise through images of suffering (or even smiling) children, what sets the humanitarian machinery into motion is not simply an innate biological impulse – compassion is already mediated – but the ways that images and messages become intelligible through specific, already constituted frames, frames that are subsequently reconfigured by the images.
This insight leads us to pose several critically oriented framing questions. Could compassion be ontologically malleable, that is ‘educated’ or even ‘manipulated’ (Rorty Reference Rorty1998)? And if so, could it also become ethico-politically problematic, that is ‘calculating’ (Woodward Reference Woodward2004) or sometimes even ‘malevolent’ (Garber Reference Garber2004)? In other words, what if compassion is not the endogenous moral compass we often take it to be, but in fact an object of politics on its own right? What if compassion is an effect of social and political life in addition to being a motivating cause? What if (part of) our capacity to feel compassion is a result of subtle forms of cultivation, power and governance? And, precisely because of this, what if compassion does not always get the world of suffering right? What if the concrete practices of reproducing – both eliciting and expressing – compassion are, themselves, morally suspect? And, stemming from this, can compassion be saved?
To investigate this problématique, the article proceeds in four substantive steps. In the first section, we theorize the emotional encounter between an appealing image and its spectator, and highlight their co-constitutive interplay as well as the role of productive power in governing it. The second section enters the encounter from the ‘spectator side’ and elaborates an eclectic theoretical account of compassion as a conditional and governed embodied experience with narrative logics. The third section then approaches the encounter from the ‘image side’ and empirically illustrates concrete efforts at governing compassion in fundraising and awareness campaigns after the 2010 earthquake in Haiti. Two regulative narratives – ‘youth and gender’ and ‘crisis and urgency’ – are highlighted. Key ethico-political problems of these narratives are also discussed, to the effect that the very benevolent self-image of compassion may in fact become circumspect. Finally, in the fourth section, the article explores potential alternatives for compassion, care and engagement with distant suffering, specifically pluralistic photography and linking compassion and outrage.
Government of compassion: imagery, spectator, and interplay
The humanitarian task is not an easy one. Aid agencies must convince publics to part with funds to assist distant strangers, otherwise out of sight, out of mind and unknowable (Kennedy Reference Kennedy2009). As such, one of the most immediate concerns of humanitarians is to connect the potential recipients and providers of relief, and to do so in a way that is both ethical and efficient. This problem divides into two practical sub-problems for aid agencies. First, how to overcome distance and bridge the gap between sufferers and helpers, both geographically and culturally? And, second, how to motivate and emotionally activate the target audience? In short, the question is: how can compassion and compassionate practices be made possible?
This article starts with the assumption that technology, and especially imagery as a technology, has provided a contemporary solution to the problems of distance and motivation. By technology we mean both technology as ‘high-tech, scientific and mechanical means of fulfilling tasks’ and also technology in a more basic sense, ‘as arts, skills, and crafts – ways of doing things’ (Haskell Reference Haskell1985; Kennedy Reference Kennedy2009). For humanitarianism, efficient transportation and logistical capabilities as well as new forms of information technology have intensified global interconnectedness, making the provision of aid increasingly efficient and timely in times of crises. But it is especially images, in concert with technologies such as the internet, computers, digital cameras, and television, that have shrunk the world geographically and – importantly – culturally, morally, and emotionally. As Kennedy (Reference Kennedy2009) has put it, ‘through the medium of the photograph the viewer is drawn into the position of being witness to these distant events. In this way, suffering becomes real to those who are elsewhere’.
Inasmuch as it has been studied in global politics, the role of imagery has been investigated from two broad theoretical perspectives. For the most part, the emphasis has been on the image itself, and specifically on the ways in which images frame human apprehension of reality. This entails a commitment to photo-constructivism: images do not represent the world as it is, but are constitutive of the very ways in which individuals come to know and experience – or remain ignorant and obtuse of – the world due to aesthetic choices by their producers (Bleiker Reference Bleiker2001, 512; Bleiker and Kay Reference Bleiker and Kay2007, 143). As such, recent photo-constructivist research has tended to focus on how images are framed to ‘either reproduce dominant forms of discourse’ or how they ‘provoke critical analysis’ in and about world politics (Shapiro Reference Shapiro1988, 130). This characterizes recent IR research on imagery in war, genocide, humanitarianism and popular culture.Footnote 4
An alternative approach – and one with which this article has sympathies – has led scholars to look past the image itself, to the receptive audience and/or a particular spectator viewing the image. While recognizing the power of hegemonic discourses, it is equally apparent that images are apprehended in ways that hinge on the identity of the viewer. In other words, if images do help organize the way we register reality, then they are also ‘bound up with the interpretive scene in which we operate’ (Butler Reference Butler2009, 71). Consequently, the causal capacity of imagery must be seen as dependent on the contingent, socially conditioned identity (Rorty Reference Rorty1989) – and here the emotional identity (Oksenberg Rorty Reference Oksenberg Rorty2008) – of the spectator.Footnote 5
This is something that recent studies on imagery in IR have started to notice, though this observation has yet to be extended to emotions, including compassion. For example, Bleiker and Kay (Reference Bleiker and Kay2007, 143) have observed that while framing an image is important, ‘a photograph cannot speak for itself. It has to be viewed and interpreted’. Importantly, the interpretation of an image is intertwined not only with ‘previous experiences’, but also with ‘values and visual traditions that are accepted as common sense by established societal norms’. Similarly, Möller (Reference Möller2010, 116) has suggested that the images of genocide in Rwanda are recognizable ‘if they appeal to the images that the viewers already carry with them as visual memories’, while Campbell (Reference Campbell2002, 159) has argued that the famous 1992 ‘barbed wire picture’ and debate about concentration camps in Bosnia is only comprehensible when read through the ‘collective memory’ of the Holocaust. As these insights suggest, images can be thought to have complex socially produced meanings only insofar as they function to perpetuate the already-existing and internalized ‘preferred meanings’ of the political order (Rose Reference Rose2007, 87) – that is, when there is a collective ideological ethos through which ‘one reads into the photograph what it should be saying’ (Sontag Reference Sontag2003, 29).
This article seeks to integrate these insights about imagery, spectators and context into an approach that focuses on government Footnote 6 – specifically the government of compassion – in and through aesthetic encounters. For this, we first draw on Judith Butler (Reference Butler2004, 141), who has reminded us that images ‘are framed, surely, but they are also playing the frame’. This observation opens a window of opportunity for a political understanding of the emotional encounter between an image and its spectator that builds on the above discussion, but relies on neither the (appealing) image nor the (feeling) spectator alone. Instead, it foregrounds the vital co-constitution and interplay between the image and its spectator, and does that by conceptualizing the efforts to regulate – indirectly through social conditioning – the very forms of the ‘playfulness’ of the encounter that lead up to the emotion itself.
We also turn to Michael Shapiro (Reference Shapiro1988, 128, 149–50), who has discussed the ‘reading of photographs’ as a practice conditioned by a set of ‘social codes’ with which the photograph and the spectator interact. In particular, he argues, there are subtle ‘meaning-engendering norms’ that are always already involved in both the production and reception of imagery and limit the potential for meaning variance. As a result, there is often a ‘congruence between photographing something and looking at it, and thus there is a tendency for the photograph to be reconciled with the social order’. These normative and normalizing codes do not appear in a void, but originate in what scholars of social semiotics theorize as the ‘production regimes’ and ‘reception regimes’, respectively. Both of these regimes, in turn, are embedded in a broader regime that the semioticians define as the ‘logonomic system’, that is as ‘the set of rules prescribing the conditions for production and receptions of meanings’ (Hodge and Kress Reference Hodge and Kress1988, 4).
Now, to be clear, while the analytical attention on imagery and the regimes that regulate their production and reception is important, it is the register of compassion that plays the central role in this article. Thus, our focus is on the production and expression of compassion that takes place in and through aesthetic encounters between appealing NGO imagery and their sympathetic spectator-donors. This means that we need to specify our approach from the ‘aesthetic’ in general to the ‘emotional’ and the ‘compassionate’ in particular. From this perspective, the analytic of production and reception regimes helps define not merely an aesthetic but an emotional regime of compassion Footnote 7 in which the encoded acts of producing and expressing compassion are inculcated and regulated through discourse and action in the social sphere. The practical upshot of this is, of course, that the very ways in which individual capacities to ‘read’ and ‘feel’ images are conditioned end up bearing upon the range of possibilities for moral and humanitarian action: images condition reception that conditions action, the sustenance of which may call for new images, and so on.
Crucially, from a political perspective, this implicit form of regulation – or of government – relies on a subtle mode of power that is ‘productive’ (Foucault Reference Foucault1980). Productive power refers to the ‘constitution of […] social subjects with various social powers through systems of knowledge and discursive practices of broad and general social scope’ (Barnett and Duvall Reference Barnett and Duvall2005, 55). In this article, productive power functions to shape particular kinds of feeling subjects by encoding specific preferred emotional outcomes (e.g. compassion for a suffering woman) with the embodied premises (e.g. norms of women as non-violent and requiring assistance) through aesthetic encounters. As such, productive power does not repress compassionate subjects and their instincts, but seeks to cultivate individuals into certain taken-for-granted – and often hegemonic – ways of feeling and performing compassion.Footnote 8
Governed compassion: an eclectic reading
The government of compassion depends on the ability of humanitarian imagery to interact with the contingent and socially conditioned emotional identities of their spectators in a way that makes the successful performance of compassion possible. To support this claim, we elaborate a theoretical approach to compassion that allows for and makes possible three broad claims: first, that there exist shared compassionate dispositions among the recipient spectators; second, that these dispositions may have contextual origins; and, finally, that experiencing compassion is a narratively structured and conditioned act (i.e. a practice) with onto-epistemic consequences: it provides an affective way of connecting to and seeing the world and thus also helps re-constitute its referent objects (as well as the feeling subject).
Compassion: traditional characteristics
Lauren Berlant (Reference Berlant2004, 1) has claimed that there is ‘nothing clear about compassion except that it implies a social relation between spectators and sufferers, with the emphasis on the spectator’s experience of feeling compassion and its subsequent relation to material practice’. However, even this observation implies a standard view: compassion is an altruistic emotion in which a person is capable of placing herself into the skin of another fellow human being, of experiencing some of the pains and sufferings of that person, and thus becoming motivated to either alleviate (immediate aid) or ameliorate (long-term social transformation) her condition.
In more detail, and to add clarity to Berlant’s framework, compassion has four key characteristics. First, compassion apprehends: not only does it push the boundaries of the self outwards, but it also offers an affective way of understanding – and reconstituting – afflicted human beings. Second, compassion touches: it engages with the pains of (certain) others and thus activates non-sufferers affectively, consciously and unconsciously. Third, compassion promotes action: it entails a disposition to perform beneficent actions to either alleviate or ameliorate suffering unless there are extraordinary obstacles (e.g. distance) or the assistance is clearly inappropriate (e.g. it risks the sufferer’s autonomy). And, finally, compassion connects: it assumes and reproduces a sense of shared humanity in which the sufferer is perceived as a fellow human beingFootnote 9 (Blum Reference Blum1980; Nussbaum Reference Nussbaum2001; Garber Reference Garber2004; Jimenez Reference Jimenez2009).
In everyday use, compassion is often conflated with related emotions or sentiments, most notably with empathy, sympathy and pity. Empathy entails a non-evaluative sharing of any emotional state with another person, regardless of its type, quality or intensity. As such, it may be involved in compassion, but does not necessitate painful suffering, nor does it involve appraisals or interpretations that such suffering is either significant or harmful to someone (Nussbaum Reference Nussbaum2001, 302; Batson Reference Batson2009). Sympathy, on the other hand, comes close to compassion and has frequently been used synonymously (Clark Reference Clark1997; Rorty Reference Rorty1998). If one wants to find some difference between the two, it is perhaps that compassion is ‘more intense’ and ‘suggests greater suffering’ (Nussbaum Reference Nussbaum2001, 302) or that sympathy is more distant, i.e. it has lost the ‘sense of direct engagement’ (Jimenez Reference Jimenez2009).
Finally, in its contemporary usage, pity has come to have negative nuances of condescension, superiority, and even blameworthiness with regard to the sufferer. It is characterized by an attitude of fundamental difference and asymmetry. For the pitier, certain kinds of misfortunes and painful experiences are correlated with certain kinds of people – people different from the pitier herself (Nussbaum Reference Nussbaum2001, 301; Spelman Reference Spelman2001, 120). Pity is also often emotionally shallower than compassion. It is a commiserating sentiment – as in ‘feeling sorry for another’ (Jimenez Reference Jimenez2009) – that lacks a more direct and affective connection with the sufferer. Precisely because it is asymmetric and intellectual, pity tends to be motivationally restricted. It easily leads to the misappropriation of suffering without any genuine attempt to alleviate it, such as in voyeuristic and pornographic consumption of dehumanizing imagery of suffering for curiosity, pleasure or narcissism (Halttunen Reference Halttunen1995). As we highlight in the third section, the dividing line between compassion and other emotions – notably pity – is often quite fragile; consequently, in practice, asymmetry, dehumanization and even paternalism can never fully be escaped.
Compassion as socio-political
Most modern accounts of emotion share an overwhelming tendency to explain emotionality internally and individually, from the perspective of the isolated feeling subject. This bias is especially evident in the on-going debates on whether emotion should be thought of primarily as affective or cognitive – that is as biologically evolved and hard-wired embodied reactions conducive of survival and social life (Zajonc Reference Zajonc1984; LeDoux Reference LeDoux1996; Damasio Reference Damasio2003; Prinz Reference Prinz2004) or as informed responses that involve emotion-typical thoughts, beliefs, appraisals, or evaluative judgements about the object of an emotion (Solomon Reference Solomon1980; Reference Solomon2003; Lazarus Reference Lazarus1984; Roberts Reference Roberts1988; Nussbaum Reference Nussbaum2001). Given the focus on the individual – on her body, brain or cognition – emotions are often conceptualized as asocial and apolitical phenomena insofar as they are thought to be unaffected by the workings of social norms or rules (Hochschild Reference Hochschild1979; Parkinson Reference Parkinson1996). As a result, the study of emotion is often limited to psychology, philosophy or, recently, neuroscience, and it does not, and need not, pay serious attention to the contextual origins of emotional behaviour.
This bias also applies to the emotion of compassion. Advocates of the affective view typically maintain that compassion is a natural and involuntary reaction of co-suffering – a ‘passion of compassion’ (Arendt Reference Arendt1990) or an ‘irresistible compassion’ (Fiering Reference Fiering1976). Some emphasize that it is a basic human emotion the core of which is the ‘non-cognitive element of pain or distress at the pain or distress of others’ and which – in its most primitive form – boils down to the ‘instinctive distress caused by, rather than at, the suffering of others’ (Crisp Reference Crisp2008, 240–1). Others admit the importance of painful feeling, but are more interested in establishing whether and how such experiences recruit non-conscious brain systems for sensing and regulating the body functions of compassion. They highlight the importance of homeostatic regulatory mechanisms that typically characterize and regulate the more ‘basic emotions’. In particular, compassion for other people’s physical pain is said to ‘co-opt neural mechanisms for personally experienced pain most efficiently and directly’, whereas compassion for social pain in psychological or moral situations ‘build[s] on these same mechanisms but may operate less efficiently and directly’ (Immordino-Yang et al. Reference Immordino-Yang, Helen, McColl, Damasio and Damasio2009, 8024).
The cognitive view, on the other hand, focuses on individual cognitive processing and the cognitive structure of compassion that is necessary for co-suffering. From this perspective, compassion is an ‘informed passion’ (Spelman Reference Spelman2001, 85) that is the result of – or emerges with – characteristic beliefs or appraisals that the compassionate person must hold about the object of compassion and her misfortunate condition, often aided by imaginative dwelling on the predicament of the sufferer. Typically, going back to Aristotle, these include three key components: first, the misfortune and suffering of another must be serious rather than trivial, for example serious mental or physical pain rather than irritation, say, at losing a pen. Second, the person is not to be blamed for her misfortune and suffering, or at the very least the suffering is out of proportion to her fault. And, third, the non-suffering person must be able to consider herself subject to a similar fate with the sufferer (Aristotle 1959, 1385b–1386a; Blum Reference Blum1980, 507–11; Nussbaum Reference Nussbaum2001, 306).
For us, compassion is not, strictly speaking, affective, cognitive or even individual, despite the common sense view that we experience it privately, as my feeling in my life for the pains of someone else. Our approach to compassion is eclectic; we appropriate insights from contending scholarly traditions and merge them together for the broader purpose of constructing a socio-political account of compassion (Rorty Reference Rorty1982, 151; Katzenstein and Sil Reference Katzenstein and Sil2010).Footnote 10 Thus, we agree with the affective view that the human body and automatic affective processing play a crucial part in, and are a precondition for, compassionate upheavals. However, we also claim with the cognitivists that there is more going on in compassion than, say, a constantly correlated firing up of certain pain (or mirror) neurons or the activation of specific parts of our pain-related nervous systems. Some form of ‘cognitive activity’ – including affective appraisal – seems necessary if we are to apprehend and bring home someone as an appropriate object of compassion, that is as a person who is in pain and whose well-being is in jeopardy in a way that touches us. In this sense, we should avoid reducing ‘affective’ or ‘irresistible’ to the ‘primitive’, since even automatic neural processing can involve complex symbolic significances condensed into an instant meaning (Lazarus Reference Lazarus1984, 124).
We also maintain that there is more going on in compassion than a mere conscious feeling of pain. We accept that painful sensations arising from neural and somatic activation are frequently associated with compassion and thus characteristic of both lay and scholarly understandings of the emotion. Yet compassion can also be non-conscious, and even when consciously experienced it is not reducible to pain since pain sensations can be related to various other emotions, too (Nussbaum Reference Nussbaum2001, 58–9, 325–6; Jaggar Reference Jaggar2008, 380).
Having said this, we acknowledge that the cognitivists have often cut their approach too thin by overemphasizing the role of conscious beliefs, thoughts or judgements – for example beliefs about ‘size’ and ‘fault’ regarding the suffering. Porter (Reference Porter2006, 102), for example, suggests that a compassionate person is ‘making a reasoned judgment’ about the needs of a specific person or group ‘after careful consideration’. By doing this, cognitivists overlook unconscious or embodied responses (LeDoux Reference LeDoux1996), bodily sensations (Crisp Reference Crisp2008) and/or social and cultural influence (Ahmed Reference Ahmed2004) on the emotion. In this respect, a corrective move towards a view of compassion as a socially inculcated ‘judgment of the body’ (Solomon Reference Solomon2003, 189) is in order. Second, a cognitivist commitment to a strict formal structure of compassion may result in a limited understanding of concrete cases where compassion is ‘irrational’ (or socio-political). For example, compassion for a justly incarcerated inmate would be impossible since it is incompatible with the appraisal of fault (Crisp Reference Crisp2008, 235–8).
On these grounds, we suggest a working definition for a socio-political compassion. Compassion is an embodied yet socially informed response in which an intimate connection between a sufferer and a non-sufferer becomes established through the latter’s acknowledgement and co-experiencing with the former’s pain or misfortune. As such, compassion is often, and in time, experienced as a painful sensation, and the overall emotional attitude may be further embellished or perpetuated by higher cognitive processing, including beliefs or thoughts about the suffering object of emotion and her negative condition. The emergence of the compassionate connection typically motivates acts towards the mitigation of suffering, though the practical form of mitigation – for example immediate aid or eradication of the causes of suffering (including their financial support) – remains subject to social formation; compassion has no essential action tendencies.
To understand compassion as social is, in part, to return to the etymological origins of the word ‘emotion’. The English word ‘emotion’ comes from Latin and French expressions for moving or migrating from one place to another, or for exciting, agitating, or stirring up (Crawford Reference Crawford2000, 123–4). For the emotion of compassion, this implies that it is often seen as a motivational cause of social or political action, such as the alleviation or amelioration of pain. This is the case in humanitarian assistance (Barnett Reference Barnett2011) or human rights promotion (Hunt Reference Hunt2008). However, this is only a part of the story. Compassion itself can also be understood as an effect of socio-political life. From this point of view, to claim that compassion is social means not only that its expression requires a social situation between suffering and non-suffering, but also that the emotion itself is constituted by a range of social resources – we highlight imagery, narratives and norms – that individuals need to (make) sense (of) suffering in social reality and to make themselves emotionally intelligible to each other.
The stronger claim that compassion is political means that emphasis is placed on productive power and the government of compassion. Not only are compassionate reactions facilitated and informed by adopted social resources, but these social resources seek to condition and regulate the specific ways of feeling, expressing and enacting it: they enable appropriate and constrain inappropriate compassionate behaviour, such as in terms of the production of proper and deviant objects, intensities or instantiations of compassion or forms of compassionate action. For example, as we highlight in third section in terms of appropriate objects, compassion is often elicited and regulated through imagery of suffering children or women, but less so of adult men, because of the social construction of the female gender as non-violent, vulnerable and in-need of compassionate male protection. Together, these social and political aspects amount to a claim that our compassion discourse, constitutive of the experience itself, is comprised of ideological ‘models of and models for how one should feel and behave’ (Myers Reference Myers1979, 345).Footnote 11
To sum up, we argue that compassion is not a natural passion, a biologically programmed affect, or a rational appraisal that relies on a formal set of beliefs about the suffering person and her condition. Instead, it can be seen as a conditional and governed socio-political emotion. What is crucial in this view of com-passion is the social production of taken-for-granted familiarity, intimacy and importance through which the co-sufferer becomes instinctively connected to the adjacent or distant sufferer in a way that – as Arendt (Reference Arendt1990, 85) once put it most starkly – she is truly ‘touched in the flesh’ and ‘stricken with the suffering of someone else as though it were contagious’. But unlike Arendt, who endorsed a largely internalist position, we argue that the very capacity to apprehend suffering bodies and to irresistibly co-suffer with them is often calibrated, and governed, through the circulation of social resources that we (have) come to learn as inherent and foundational parts of our embodied being. The bodies in pain out there are apprehended, and felt and cared for, only insofar as they are already within us.
Compassion as narrative
There are two obvious analytical ways to approach compassion as a socio-political emotion. On the one hand, it is possible to focus on the productive power of language and narrative, or discourse, in the relevant political and emotional regime (Abu-Lughod and Lutz Reference Abu-Lughod and Lutz1990; Rorty Reference Rorty1998).Footnote 12 On the other hand, the focus could also be on the productive power of social practices that tend to precede the adoption of language and discursive resources, at least when seen from the perspective of developmental psychology (Allen Reference Allen2008, 166–7). Given our focus on narrative in both compassion and imagery, we highlight the first, discursive alternative. That said, we do acknowledge the interdependency of discourse and practice in the course of our argument.
In our discursive approach, we draw from the work of the late neo-pragmatist Richard Rorty, who emphasized the importance of language and its radical consequences for Western thought on subjectivity and knowledge. In particular, Rorty (Reference Rorty1989, 21) maintained that we have ‘no prelinguistic consciousness to which language needs to be adequate, no deep sense of how things are which it is the duty of philosophers to spell out in language. What is described as such a consciousness is a simply a disposition to use the language of our ancestors, to worship the corpses of their metaphors’. We build on this ontological idea that language – or more broadly discourse – may penetrate who we are and what we know and do not merely at the level of the conscious (beliefs), but also the unconscious and the embodied (emotion).
In more detail, our central assumption is that there is a close connection between feeling selves and stories; that a ‘life as led is inseparable from a life as told’ (Bruner Reference Bruner2004, 708). This narrative view of the feeling self assumes not only that what is told but also how it is told is important, in so far as ‘the ways of telling and the ways of conceptualizing that go with them become so habitual that they finally become recipes for structuring experience itself’ (Bruner Reference Bruner2004). In this sense, the key feeling of compassion – just like other forms of human ‘doings’ ranging from thought to action – may be understood as ‘enacted narratives’ (MacIntyre Reference MacIntyre1985, 211–2). This view entails that we all live out narratives in our lives and tend to understand our own lives, as well as the lives, (painful) experiences and actions of others, in terms of narratives we have acquired and are capable of enacting, and to some limited extent verbalizing.
The role of narratives is particularly important in terms of compassion. From a narrative viewpoint, compassion is a social construct reliant on what Rorty (Reference Rorty1998) has called ‘sentimental education’ or ‘manipulation of sentiment’. This view holds that our embodied moral intuitions about the right thing to do are deep down quotidian: they owe nothing to ‘increased moral knowledge’ and everything to ‘hearing [and seeing] sad and sentimental stories’ (Rorty Reference Rorty1998, 172). From this follows that the answer to the humanitarian moral question ‘Why should I care about a complete stranger?’ relies on socially circulated narrations that work, or have worked, to create familiarity, intimacy and care between distant strangers from the point of view of the person not in pain or distress: ‘Because this is what it is like to be in her situation – to be far from home, among strangers’, ‘Because her mother would grieve her’, and so on (Rorty Reference Rorty1998, 185).
Narratives play a significant part in the conditioning and regulating of our moral sensibilities, or lack thereof (Nussbaum Reference Nussbaum1992, 286–313). This is because we tend to learn in social life how, for whom and when to feel compassion in particular situations, and conversely how, for whom and when not to. These stories express their narrative theme (suffering) and structure (tragedy), implant in us their social and emotional dynamics (passive suffering and active compassion), and highlight key subjects (e.g. archetypical suffering children or women).Footnote 13 The stories we receive, though not of our own making, are ultimately internalized and embodied by us. Once/if embodied, and familiar, they bear upon the way we come to apprehend the world and suffering therein. Reminiscent of the age-old pragmatist premise that beliefs are ‘rules of action’ (Peirce Reference Peirce1997, 33), this article maintains that tragic narratives entail embodied rules for our sentimental and, especially, compassionate behaviour.
Governing compassion: humanitarian narratives in post-earthquake Haiti
It is frequently said that humanitarianism represents the better instincts of humanity. Humanitarians are the ‘last of the just’ (Rieff Reference Rieff2002, 333), acting to relieve human suffering across the globe. These grand ambitions carry correspondingly grand costs; the humanitarian field was funded to the tune of $18.2 billion in 2010, and needs far exceed resources (Stoianova Reference Stoianova2012). The organizations populating this field face acute funding imperatives, both to serve their ethical mission (Walker and Maxwell Reference Walker and Maxwell2009) as well as to sustain their organizational existence (Cooley and Ron Reference Cooley and Ron2002). Consequently, NGOs have embraced private voluntary assistance; today, donations from individuals are the fastest growing segment of the funding pool and the single largest source (57%) of NGO funding.Footnote 14 In 2010, the vast majority of this private funding – $4.9 billion of $5.8 billion – was given to NGOs (Stoianova Reference Stoianova2012, 5).
In order to stimulate this private funding, humanitarian agencies rely heavily on image-centered advertising. As Jonathan Benthall writes, disasters do not exist – save for the victims – unless publicized. In this sense media and humanitarian representations actually construct disasters (Benthall Reference Benthall1993, 27; see also Calhoun Reference Calhoun2008, 82–9). Through the medium of the photograph or video the spectator is drawn into the position of being witness to these distant events.
Existing scholarship has suggested that humanitarian advertising campaigns tend to rely on a fairly consistent set of motifs (Halttunen Reference Halttunen1995; Manzo Reference Manzo2008; Kennedy Reference Kennedy2009). We agree. However, this research has largely failed to connect these general observations to specific empirical evidence from advertising campaigns. In this section, we draw on research collected during the response to the 2010 earthquake in Haiti – by number of organizations, the largest ever mobilization – to analyze the interlinkages among representations, compassion and humanitarian action. We supplement this data with observations made during the Syrian civil war (2013). Although the crisis in Syria differs in fundamental ways from that in Haiti – notably, as a man-made, not (ostensibly) natural, event – NGOs have narrated it in similar ways.
In particular, we are interested in reading these images and messages to tease out the narratives through which aid agencies seek to connect us to distant suffering, and, in so doing, to govern the way we come to feel compassion now and in the future. As such, we investigate: What are the narratives that seek to govern the performance of compassion in the interplay between appealing image and its spectator-donor? Through which narratives do we construct and recognize suffering? What themes make images intelligible, and thus facilitate a reading of them that elicits compassion and compassionate action? We also investigate and discuss critically: What are (some of) the ethico-political problems of such narrations in the context of contemporary humanitarianism?
On January 12, 2010, Haiti was devastated by a 7.0 magnitude earthquake, its epicenter just 25 km west of Port-au-Prince. The earthquake was the second deadliest in the last 100 years; it left 230,000 dead, 300,600 wounded and 2.3 million displaced – in a country of 10 million (IASC 2010; Bhattacharjee and Lossio Reference Bhattacharjee and Lossio2011). Large swaths of the capital’s housing stock, architectural heritage and infrastructure were destroyed. The humanitarian response to the earthquake was unprecedented; it is estimated that 1000–2000 agencies were involved (IASC 2010) and charities raised more than $1.4 billion for relief and recovery (Lieu Reference Lieu2011). This outpouring of support was fueled by media and NGO accounts of the suffering and devastation; the encounter between donor and Haitian was mediated by images and depictions of the earthquake and the compassionate response was channelled through aid agencies.
The universe of potential cases (i.e. images) is quite large indeed. We focus our analysis on the websites (homepages) of a selection of Western NGOs. This bracketing appears justifiable for two reasons: first, humanitarian organizations receive a substantial – and increasing – percentage of their funding through direct online contributions. For instance, during the Haiti earthquake response, Boston-based Partners in Health raised $23.2 million online, 27.6% of its total fundraising. Among others, World Vision (39.4%), Doctors Without Borders USA (21.8%), Catholic Relief Services (10.7%) and Islamic Relief USA (47.0%), also raised significant sums online (The Chronicle 2011). Second, websites provide a fairly direct link between images, messages and action (giving).
Our data consist of homepage screenshots – the very first image(s) and messages a visitor would have seen – for 25 aid agencies, recorded each day for the first 2 weeks (14 days) after the earthquake – 14–26 January 2010 – a total of 350 images.Footnote 15 We returned to these same agencies on 31 August 2013, as discussion of international intervention in Syria reached its zenith. The 25 agencies were selected to be broadly representative of the sector; our study includes the largest organizations, namely CARE, Catholic Relief Services (CRS), Médecins Sans Frontières (MSF), Oxfam, Save the Children and World Vision International (WVI), which together deliver the ‘lion’s share’ of all emergency assistance (Walker and Pepper Reference Walker and Pepper2007, 5), as well as a range of far smaller organizations. We would suggest that it is particularly important to focus on the largest actors, as, by virtue of size and reach, they can be said to define the sector. See Appendix I for a list of agencies and dominant motifs.
While we sampled from a wide range of organizations – varying in size, faith orientation (11 are faith-based and 14 secular) and location (12 are headquartered in the United States and 13 in Europe) – we found a remarkable degree of consistency in humanitarian representational practices. (In fact, in several cases, as we highlight below, different NGOs actually used the same images.) The fact that common themes transcended organization type speaks to the power of collective social understandings and to the strength of narratives about the way suffering is framed and told. Consequently, what is often seen as the result of a spontaneous outpouring of compassion happens in the first place because images resonate with the socially constituted, embodied emotional identities of the target audience. We focus our analysis on two primary themes – understandings of youth (with discussion of gender) and motifs of crisis and urgency.
Narratives of youth and gender (governing compassion)
Scholarship on humanitarianism has highlighted the key role of a familiar – or archetypical – stock of characters in the often-repeated crisis narrative with which the popular understanding and dynamics of emergencies are constructed (Chandler Reference Chandler2001, 690; Douzinas Reference Douzinas2007, 12–4).Footnote 16 In our analysis of aid agency websites, we found that the most common characters depicted in the Haiti crisis narrative were children, who were featured by 21 of 25 agencies (84%) in the 2 weeks after the earthquake, frequently as the dominant imagery. More recently, in Syria, among agencies responding to the crisis, 15 of 19, or 79%, portrayed children. Figure 1, from British charity Save the Children, provides an example of a typical advertisement; it depicts an injured Haitian child, gazing at the camera; also visible are the healing (intervening) hands of the doctor. This image headlined Save the Children’s webpage for 7 days after the earthquake; this exact image was used by CARE International and Caritas Internationalis. This replicates findings made elsewhere; as Kate Manzo (Reference Manzo2006, 10) has argued, ‘the dominant iconography of the majority world as a whole is […] the lone child photographed in close-up’.
Plan Finland (Finland) – 14 January 2010 – http://www.plan.fi
The second archetypical character in the humanitarian narrative is the female. This can be a young girl in which age and gender factors co-exist, but also an adult woman – often accompanied by children – in which gender remains the key constitutive factor. Our analysis supports this. For example, we found that female subjects, including adult women, were far more likely to be depicted than adult males, particularly in the Syria crisis.Footnote 17 Like the children, adult women were also frequently portrayed as inactive victims in need of help. Figure 3, from Samaritan’s Purse (United States), illustratively depicts a woman with her hand extended; this unspoken plea for help is accompanied by a message asking for ‘Help for Haiti Quake Victims’ and a link to donate online.
Samaritan’s Purse (USA) – 15 January 2010 – http://www.samaritanspurse.org
A photo of a child or woman means little in and of itself; it is the context, both suggested (as by the text) and implicit (the normative environment), that fills it with meaning. These advertisements are intelligible – and our emotional reaction made possible – inasmuch as they interact with overarching and paradigmatic narratives of youth and gender. In more detail, these images derive their emotional weight from three related narrative frames that, collectively, constitute the ‘legitimate’ victim character: anonymity, passivity, and innocence.
First, humanitarianism is based on ethical obligations, such as impartiality and humanity, that require transcending particulars, be it kinship, background or even acquaintance; agency images focus on universal symbols – women and children, suffering and destruction – to cut across boundaries of comprehension. The marketed image thus sheds much of the specificity of the locale and situation in favor of universal markers. One scholar has called this ‘anonymous corporeality’: humanitarian images portray ‘generalities of bodies’ (Malkki Reference Malkki1996, 388). Here, though we know that events have occurred in Haiti, it is nonetheless true that the injured child or desperate woman seen on one aid agency website could easily be found (and were found, in the cases of Figures 1 and 4) on almost any other. Generic images ultimately deny the very particulars that make people something other than anonymous bodies. ‘Haitian in need’ is all we (need to) know about them. These images do not dehumanize, as such, but humanize in a particular mode: a mere, bare, naked or minimal humanity is set up (Malkki Reference Malkki1996, 390; Douzinas Reference Douzinas2007; Calhoun Reference Calhoun2008). It is into this minimal humanity – this ultimately empty vessel – that our thoughts, hopes and care are typically poured.
Second, and stemming from this, advertising images, often operating at the register of the tragic, subordinate the self to the physical body and its pains. The images tend to portray bodies – ‘bare life’ – not political beings – or ‘qualified life’. What are the children in Figures 1 and 2 actually doing?Footnote 18 The child of the humanitarian advertisement merely stares at us with pleading eyes. She is filling her role. Whether the advertisement is from Save the Children or Plan Finland, all too frequently the subject of the nonprofit campaign is helpless and forlorn, defined not by agency or ability but rather by vulnerability, deficiency and inactivity. And, while Figures 3 and 4 depict adults, the images are scarcely more empowering. In Figure 3, the woman’s apparently desperate gesture for help is reinforced by the capitalized text ‘Help for Haiti Quake Victims. Donate Online Now’.
As one vocal critic of contemporary humanitarianism has claimed, the ‘subject’ – or archetypical character – of the alleviative humanitarian narrative is often portrayed as a tragic and impotent victim ‘whose dignity and worth has been violated. Powerless, helpless, innocent, her basic nature and needs have been denied’ (Douzinas Reference Douzinas2007, 12). Consequently, the subject is acted on and for; she cannot herself act or even speak because she is not qualified.Footnote 19 Passivity thus works in support of the third key narrative frame – threatened innocence and potentiality. According to the humanitarian narrative, the objects of aid campaigns have great potential, but they cannot fulfil this promise without our help. We can help create ‘lasting change for children in need’, as Save the Children reminds us in Figure 1. In this way, ad campaigns justify action and intervention on ultimately paternalistic grounds. In emphasizing generic figures of children, aid agencies subtly, perhaps even unconsciously, pull from deep and embodied cultural reservoirs to depict large parts of the world as innocent, passive and replete with potential.Footnote 20
The existence of these narratives and their patriarchal logic of masculine protection is something that feminist scholars in particular have discussed. Feminist literature illustrates how women (and particularly young girls) are often thought of, and narrated as, what Elshtain (Reference Elshtain1995, 4) has called ‘Beautiful Souls’. This is a cultural trope that constructs females as ‘nonviolent, offering succor and compassion’ – as representatives of socially sanctioned innocence. Similarly, feminists have pointed out how the discourses of dangerous societal life (Young Reference Young2003) and world politics (Enloe Reference Enloe1990, 12–3; Tickner Reference Tickner1992, 58–9) often tend to construct the ‘real man’ as ‘the protector’ who will suppress his fears and defend the weak, paradigmatically children and women, in emergency situations. Women, according to this logic, are constructed as ‘the protected’; they are expected to (gratefully!) seek protection from males, from their fathers and/or husbands. As such, women and children are seen, importantly, as both the community’s most valuable possession and the most vulnerable part susceptible to defilement and exploitation. Particularly in the case of humanitarianism, the extensive focus on universal images of women and children in positions of passivity plays off of existing gendered narratives to produce the world as a space of threatened innocence requiring external assistance.
Why not focus on active people? As Barbara Harrell-Bond (Reference Harrell-Bond1985, 4) noted nearly three decades ago, ‘Who would give money to refugees to help themselves? Humanitarian agencies are in a straitjacket with little else than human misery upon which to base their appeals’. Our point is not that we should criticize or congratulate nonprofit campaigns for focusing so strongly on children or women. Rather, it is merely to observe that compassion in these cases hinges on a certain implicit understanding of who and what is deserving of compassion – on who is the ‘real refugee’ (Malkki Reference Malkki1996), whose is the paradigmatic ‘real body in pain’. The advertisements are drawing from deep and embodied cultural reservoirs whereby the childhood years are associated with victimhood, passivity and innocence. This is why even though – empirically speaking – the ‘real world’ victims of crisis are not exclusively pre-pubescent or without agency, they are often portrayed as such.
In addition to their undemocratic, reductionist or paternalistic tendencies, our focus on emotions suggests that these representational practices may be ethically problematic and even counterproductive in several additional ways. Perhaps most strikingly, such representations may end up eliciting pity instead of compassion. Unlike compassion, which assumes intuitive identification and co-suffering – and thus also a shared humanity – pity stems from a separation through which the sufferer is not only seen as particularly vulnerable to adversity, but maybe even partly or wholly to blame for it. This is possible, for example, when pre-existing visceral cultural biases such as racism or misogyny dampen the affective identification with the sufferer in an image. At worst, visceral biases may lead to a total disregard and lack of moral attention (Spelman Reference Spelman2001, 34–58; Jaggar Reference Jaggar2008).
Similarly, pity may result when a representation of particular suffering is used to establish attention to a broader negative collective condition that surpasses both the image and the individual suffering (Boltanski 1999, 4). The portrayal of the suffering man in the rubble to draw attention to the ‘Haiti emergency’ in Figure 4 as well as the three distressed women to highlight the cause of the ‘Haiti quake victims’ in Figure 3 are cases in point. When a person no longer co-suffers with the particular sufferer but concerns herself with a discursively established collective – ‘quake victims’ or ‘Haitians’ – she may easily end up neither personally stricken nor bodily engaged, but merely intellectually sorry for the suffering of the generalized collective – ‘them’ – out there (Arendt Reference Arendt1990, 85). Consequently, while pity may lead to some form of action, inasmuch as this action is facilitated by explanatory frames that are premised on the presumption of inequality, this action is likely to be partial; it certainly precludes direct engagement with the (structural) causes of chronic vulnerability.
Action Contra La Faim (France) – 14 January 2010 – http://www.actioncontrelafaim.org – REUTERS/Reuters TV – © ACF
Representations depicting intimate bodily states or afflictions may also become conducive of voyeurism, if not wholesale ‘pornography of pain’ (Halttunen Reference Halttunen1995). In such cases, spectators of sensational imagery are not indifferent to the suffering, but express controversial emotional responses to them; they may be curious, fascinated or even sadistically excited by the spectacle of suffering in distant places. Given the pleasure, voyeuristic spectators may be reluctant to lift a finger to end the suffering; images do not result in any significant action for change. This criticism is often raised with regard to pity. Insofar as pity raises moral concern but produces scant action to change, it may amount to a self-gratifying practice, the core of which is the public declaration of one’s ‘virtue’ through the announcement of one’s feelings about the continuing suffering of others (Spelman Reference Spelman2001, 64–5).
Narratives of crisis and urgency (translating compassion into action)
The sense of urgency is tangible in the humanitarian imaginary. Humanitarian images often shout out: suffering is present, action is needed, and there is no time to lose (Boltanski Reference Boltanski1999, 80). Indeed, increased sensitivity to sudden and urgent crises has become so common that it is today ‘celebrated as an indication of growing cosmopolitanism and conscience’ (Calhoun Reference Calhoun2008, 85). Tragic, mobilizing themes of contingent and severe loss, crisis, and calamity are thus constants in non-profit campaigns. Together with images of children, they constitute two of the omnipresent themes of contemporary humanitarian advertising. Indeed, often, these themes are combined into a master narrative, such as in the specter of ‘childhood lost’. Here, emotions of compassion elicited by images of distressed children fuse with feelings of shock and anxiety.
The image in Figure 4 is a good example of the ways in which crisis and urgency were invoked to advertise the Haiti earthquake. ‘Haiti Emergency’, the headline shouts; ACF ‘is urgently intervening’. The photo is evocative; a man, in the rubble, desperate for a hand up (or out). This exact image was also used by Muslim Hands (United Kingdom), Trócaire (Ireland), Islamic Relief (United Kingdom) and the International Committee of the Red Cross (Switzerland) – five of the 25 organizations we studied – and other NGOs, including American Refugee Committee (United States) and Concern Worldwide (Ireland), used thematically related pictures of Haitians being pulled from rubble. In total, 60% of agencies surveyed, or 15 of 25, emphasized rubble in their homepage images. Other avatars of crisis and urgency included injuries (76% of agencies) and the use of a bull’s-eye to mark the epicenter of the earthquake, as shown in Figure 1 (36% of the agencies). In Syria, too, crisis serves as a common denominator. This is apparent in the Muslim Hands (United Kingdom) advertisement in Figure 5, in which rubble and distraught children illustrate the claim that Syria is in ‘crisis’.
Muslim Hands (UK) – 31 August 2013 – http://muslimhands.org.uk
We have argued that compassion is an essentially ambivalent socio-political emotion. Though it implies the motivation of acts to mitigate suffering, the emotion makes no essential claims about the targets of assistance or the alleviative actions themselves. The production of compassion, then, implies both the elicitation of the emotion – which we have argued occurs through established narrative frames (youth and gender) – as well as its translation into action. Crisis is the glue that connects the emotional spark to the compassionate resolution. It does so in two related, but conceptually distinct, ways.
First, crisis motivates through the parsimony of its narrative. Crisis, in its most visceral sense, signals the obvious: the sudden disruption of social order and the potential for (or realization of) severe loss, misery, and adversity, individually as well as collectively. In short, it is the harbinger of bodily pain, threatened innocence, vanishing lifestyles, and lost social stability. In a sense, then, not only individual sufferers but also the very body politic may become the extended object of worry and compassion; lose social order and you may lose individuals (most notably innocent children and women), too (Aaltola Reference Aaltola2009). In this way, anxiety, worry and compassion can become closely connected in the call for immediate acts of humanitarian assistance. The crisis narrative is one of haste: we must act now, before it is too late.
The crisis narrative is elegant in its simplicity. When, with regards to Syria, WVI states that there is ‘dire need’, UNICEF claims that ‘Syrian children need us today’, or ACF references the ‘critical situation’, all that is conveyed of the situation is its urgency; it has otherwise been stripped of context or attribution of blame. Inasmuch as complexity is the enemy of action, a direct message has motivational power. It may even be smart strategy; though definitions of humanitarianism do not distinguish between natural and man-made crises – given that the impact of a disaster is a function of preexisting vulnerabilities – in practice, donors do make distinctions.Footnote 21 Consequently, ‘crisis’ and its familiar representations – rubble, injuries and camps – attempt to activate donors’ compassion while bypassing messy or inconvenient political realities. However, these generic representations may also be counterproductive as they largely relieve the (Western) spectator from the burden of reflecting on the historical, social and political causes of pain in distant places – including one’s own potential causal connection (Belloni Reference Belloni2007, 455–6). The villain in the Haitian narrative is the earthquake; in the Syrian case, it is an abstract invocation of ‘crisis’ or ‘civil war’. French colonial and American interventionist histories are thus excised. What is more, the ‘act now, reflect later’ mentality arguably contributes to the sorts of inefficiencies, competitive pressures and even harmful outcomes that habitually plague interventions, including in Haiti (e.g. IASC 2010).
Second, urgency cultivates compassion in favor of particular kinds of responses. The language of crisis and compassion evolves rapidly into a discourse of opportunity: distant destruction is seen as creative (Aaltola Reference Aaltola2012, 70). In these cases, the emotion of compassion may be coupled with the emotion of hope. It is precisely because emergencies disrupt the established order that they create potential for action and space for social change. As Kevin Rosario (Reference Rozario2007) and othersFootnote 22 have observed, disasters have historically played a prominent role compelling the rethinking and revising of government and the organization of the economy. Today, Rosario argues, disasters have taken on an even more central role in a world lost to artifice and illusion. In this context, disasters are singularly ‘real’ and deserving of unique attention.Footnote 23 Indeed, with regards to Haiti, former President Bill Clinton suggested that ‘the earthquake opened an opportunity for the international community to remake Haiti into a thriving democracy that would serve as a model for development efforts in other parts of the world’ (quoted in McKinnon Reference McKinnon2010).
As this last comment highlights, modern compassionate humanitarianism – even if not always consciously instrumentalized to promote specific foreign policy goals – has been increasingly ambitious in its efforts to ‘help out’ people amidst crises. What scholars call ‘new humanitarianism’ (Duffield Reference Duffield2001; Fox Reference Fox2001) has mixed immediate assistance with long term goals of promoting human rights and transforming societies into law-abiding, market-based (neo)liberal democracies. This progressive agenda carries the risk, however, that political debate on the future of society (e.g. Haiti) will be subsumed under an ethics of compassion and hope; and furthermore, that local social engineering will effectively be outsourced to assumedly well-meaning international (or private) actors over and beyond the reach of democratic decision-making – in the case of Haiti to the US government and the international humanitarian community.Footnote 24
When the discourse of urgent crisis comes in touch with actual or potential social disorder, it can easily merge with the discourse of security that is dominated by the emotion of fear. This is because specific kinds of crises, for example those involving state failure, pandemic disease or terrorism, are not only about tragic local misery and compassionate responses, but also about potentially mobile sources of danger and disruption, contagious disease or massive refugee flows, that may spread from the crisis zone to the wider regional or international space. These threats raise fear, and call for acts of ‘compassionate containment’ (Aaltola Reference Aaltola2012, 63–5) under the broader rubric of ‘humanitarianism as containment’ (Donini Reference Donini2010, 223), understood here as either the ‘provision of minimum assistance to ensure that crises do not spin out of control’ or the ‘incorporation of humanitarian action in the world ordering and security strategies of the North’. Both are present in the immediate aftermath of Haiti. The first is evident in the overall humanitarian effort, which sought to save individuals in dire straits and functioned to contain the crisis and its symptoms (e.g. refugee flows) in Haiti itself. The second is exemplified by the US military’s provision of tens of thousands of portable radios on which they streamed messages advising Haitians not to leave their devastated homeland for the United States by boat (Aaltola Reference Aaltola2012, 63–4).
Beyond governed compassion: potential for alternatives?
Our discussion of compassion has complicated perceptions of what is often seen as an unequivocally benevolent and virtuous emotion. In the context of humanitarianism, we have highlighted the extent to which compassion is implicated in problematic practices of representation vis-à-vis distant sufferers as well as how compassion – especially when coupled with other emotions (e.g. pity) – may be mobilized in support of certain forms of action and assistance. It may be that some of this is inscribed in the form itself; representations are always partial (Kennedy Reference Kennedy2009). However, insofar as our argument relies on a socio-political understanding of the emotion, including its action-responses, there is the possibility that compassion might be cultivated to evade some of the problems we have identified. In short, we want to consider whether compassion can be saved as a humanitarian emotion, or, alternatively, whether compassion can be linked to other, richer emotional responses in order to enable new kinds of politics.Footnote 25
Saving compassion: pluralizing representation of suffering
The very process by which compassion is traditionally elicited and governed in the humanitarian imaginary is the same process that ultimately silences – or, at least, marginalizes – the voices of distant sufferers. In playing on socially circulated and embedded narrative frames, humanitarians risk perpetuating what Richard Bernstein (Reference Bernstein1998, 71) has called ‘ethical imperialism’, that is a practice in which ‘the language of reciprocal recognition and reconciliation masks the violent reduction of the alterity of “the Other” […] to “more of the same”’. And yet, arguably, these practices also fail to realize the full potential of compassionate action. This is because compassion, as defined, emerges out of an intimate connection between sufferer and non-sufferer such that the former’s pain or misfortunate is co-experienced. Generic representations (may) create knowledge, but they lack the specificity to facilitate intimacy, and their mediation weakens the power of the human connection. As an alternative, we would suggest practices that empower those who suffer, such as through increased sensitivity to local voices, wishes and experiences.
The work of Bleiker and Kay (Reference Bleiker and Kay2007, 151–9) on ‘pluralistic photography’ and compassion is instructive. Pluralistic photography seeks an alternative to the ‘shocking’ and essentially top-down representational practices adopted by the Western mainstream media (including NGO campaigns). Instead, it starts at the local level and puts the camera in the hands of those who are traditionally the objects of care. These specific persons – young or old, male or female – are thus activated as agents; they are able to convey and validate their own diverse, often mundane experiences and social practices even during a time of crisis and pain. For example, Tenanesh Kifyalew, a 12-year-old child living with HIV/AIDS in Ethiopia, could – and did – express what Bleiker and Kay called ‘playful defiance’ (Bleiker and Kay Reference Bleiker and Kay2007, 156) by simply documenting her everyday attempts to live a normal, and recognizable, childhood, despite her disease. In so doing, pluralistic photography seeks to disrupt existing power hierarchies that mediate the cross-cultural humanitarian encounter, especially ‘the ability of western photographers and media representations to frame the suffering of others’ (Bleiker and Kay Reference Bleiker and Kay2007, 151).
However, pluralistic photography is not without its challenges, some which Bleiker and Kay (Reference Bleiker and Kay2007, 157–9) also acknowledge. First, it remains intertwined with certain power relations even if it is stripped of others. Who, for example, provides the technology and aesthetic and political know-how to take photos? It is conceivable that the involvement of Western activists and NGOs in facilitating pluralistic photographic practices places local individuals under the influence of Western assumptions about representation and agency. Consequently, using a local photographer does not guarantee different kinds of images.Footnote 26 Second, pluralistic photography may not have the same impact as traditional imagery. Images could be unintelligible to Western (donor) audiences if they differ too dramatically from the norm. Even beyond this, local imagery may simply face challenges of market penetration given that larger organizations have greater ‘voice’; absent widespread adoption of these practices, indigenous photographs may be marginalized. Third, it remains unclear in what kinds or phases of crises pluralistic photography is viable. It may well be that these practices are possible in the context of persistent, slow-moving events (e.g. a person living with HIV/AIDS), but unavailable or counterproductive during the acute phase of a natural disaster (e.g. in Haiti).
Politicizing the response to suffering: compassion and anger
Transforming representational practices may address some of the problems that the governance of compassion entails, but even pluralistic representations confront the problem of translating compassion into action. As we have outlined, compassion implies a disposition to action, but the emotion is inherently agnostic about the precise nature of the alleviative action required. Traditional compassion emphasizes the urgent crisis, the unfortunate individual and the immediate alleviation of her suffering as a form of beneficence. (Compassion functions for humanitarians as a ‘blank check’ authorizing action – in haste – to respond to crises.) While not without merit, this ‘benevolent’ form of compassion is ethico-politically limited insofar as it does not motivate the spectator to ask critical causal and evaluative questions, such as why did the suffering occur now or who might be responsible. Consequently, though action may result, crisis factors remain misunderstood and often re-emerge in the future.
Compassion can be rehabilitated, but to do so requires rethinking the assumption that compassion can (or should) stand alone. An alternative is to view compassion as a stepping stone that requires further, more critical emotional support to flesh out its action orientation. As Matthew Newcomb (Reference Newcomb2007, 107) has suggested, compassion can lead to imaginative connections between people ‘when it is not a totalizing concept or sole basis for relationship’.
We have already discussed a variant of this, namely the evolution of compassion into hope. When crises are seen as opportunities they create hope for social progress; the so-called ‘root causes’ of crises can be tackled and future calamities avoided, primarily by introducing liberal political and economic order as a beneficent and effective solution. However, we have found that this is done at a certain cost; external crisis dynamics are bracketed out, questions of blameworthiness are at best local and political debate is replaced by self-serving ethics and international paternalism.
A potentially more critical and politicizing alternative involves the evolution of compassion into anger. The work of Spelman (Reference Spelman2001, 78–82) on ‘outrage’ or of Boltanski (1999, 57–76) on ‘denunciation’ is instructive. This alternative is especially relevant for our discussion of compassion because becoming morally outraged requires, and emerges from, a moral attention to distant suffering; if there is no moral attention to suffering in Haiti, there is no reason to become indignant about it in the first place. Compassion implies moral attention to suffering; it can be transformed and empowered by anger when distant suffering is apprehended in its context, as a case of historical and political injustice instead of as a sudden and urgent crisis. The benevolent humanitarian response is replaced by a politics of justice.
The very possibility of anger depends on a distinction between ‘ordinary’ and ‘avoidable’ suffering, that is suffering that has either natural or political origins (Boltanski 1999, 67). To the extent that disasters such as the Haiti earthquake are presented as exclusively ‘natural’ events, they are stripped of any notion of political causality. This much was clear in our image analysis. Avoidable suffering, on the other hand, is not merely distressing; it is unacceptable due to its social and political causes and, thus, also potentially interesting. To enable anger, then, one must first reconsider the ‘ordinariness’ of suffering, even in ostensibly natural events like the Haiti earthquake.
The outrage elicited by ‘avoidable’ distant suffering is of a specific kind. Given the geographical separation, compassion filtered through anger is likely to take the form of public accusation or condemnation as opposed to, say, violent retribution at the site of suffering. The response, in other words, appears in an argumentative and often written form. This means that while the accusation is an emotional response, it is also a response that must be able to ‘tone down’ to provide a composed and detailed case (Boltanski 1999, 65). For instance, as Ashley Smith (Reference Smith2010) asked after the Haiti earthquake:
Why were 60 percent of the buildings in Port-au-Prince shoddily constructed and unsafe in normal circumstances, according to the city's mayor? Why are there no building regulations in a city that sits on a fault line? Why has Port-au-Prince swelled from a small town of 50,000 in the 1950s to a population of 2 million desperately poor people today? Why was the state completely overwhelmed by the disaster?
For Smith, as for others, such as Zanotti (Reference Zanotti2010), the disaster emerged out of the intersection of geological events with geopolitical realities, namely the aggregate effect of US-promoted neoliberal economic plans that exacerbated inequality, contributed to deforestation and the deterioration of infrastructure and left the Haitian state reliant on foreign assistance. ‘The fault line of U.S. imperialism’, concluded Smith (Reference Smith2010), ‘interacted with the geological one to turn the natural disaster into a social catastrophe’.
This illustrates the potential for outrage or anger to combine with compassion to yield more fundamental engagement with political realities. However, outrage also faces challenges. First, there is the danger that the act of accusation is simply a hollow substitute for action. Anger at unjust suffering may lead to mere public denunciation that costs little and helps appease the conscience of the accuser while having little or no effect at reducing the suffering. In other words, anger does not provide the most sound basis for political action. Second, there is the practical problem of establishing the connection between suffering and its cause. The more distant and complex the connection, as in the case of the long, complicated U.S. relationship with Haiti, the more difficult it becomes establish a water-tight case – and the greater the risk, again, of oversimplification. Third, forceful accusations – and the resulting scrutiny of the accuser – may de-legitimize the argument. For example, that Smith’s accusation appears in an avowedly critical left-wing venue may confer a bias (he’s a socialist) that marginalizes his case in the eyes of the broader public, and certainly in the United States (Boltanski 1999, 62, 70). Finally, the step from compassion in the face of distant suffering to anger over the avoidability of this suffering suggests a third step: awareness of one’s implication in the very political order that has enabled this suffering in the first place. After all, those who donate or intervene are able to do so by virtue of their privileged subject positions in overarching economic, political, and social structures. This final step may be a bridge too far; in any case, it requires more concerted action (and more self-critique) than it takes to click a button and ‘donate now’ (Figures 1–5).
Conclusion
This article is framed around a seemingly counterintuitive question: can it be ‘cruel to care’? For many, as we have outlined, compassion is a moral compass that essentially gets the world ‘right’. It is compassion that alerts us to the suffering outside of our own lives and compassion that enables alleviative action in the service of a greater good. Yet, this article has suggested that there is far more to compassion than the prevailing view typically assumes.
This article has argued that compassion is an essentially ambivalent socio-political emotion; what is frequently interpreted as an innate human impulse is ultimately the result of the workings of the productive power of social resources. It is productive power that cultivates and regulates our ‘ordinary’ capacity for co-suffering and, thus also, our potential for relieving distant suffering. By engaging a representative sample of NGO advertising imagery related to the 2010 post-earthquake response in Haiti, the article illustrated not only how specific narrative frames – ‘youth and gender’ and ‘crisis and urgency’ – seek to both elicit and govern the ways of feeling compassion, but also how these aesthetic and emotional practices are ethico-politically problematic in their portrayals of distant sufferers and in the kinds of actions they facilitate. Finally, the article highlighted two potential alternatives to governed compassion, namely pluralistic representation and the response of moral outrage. Neither is a panacea – our analysis identified limitations in both approaches – but collectively they suggest that there need not be one hegemonic response to distant suffering.
To the extent that our argument and critical elaboration is plausible, there is clearly ground for a critical reconsideration of both the standard view of compassion and of the ethico-political implications of responding. This has potentially significant implications for the future study and understanding of compassion and compassionate humanitarianism because it underscores the ubiquity of relations of power and domination, even in areas widely held to be virtuous or beyond reproach. Moreover, it suggests the need for a continued investigation of the role of emotions in International Relations theory writ large; our findings with respect to socio-political embeddedness likely have applicability vis-à-vis other emotions. All this does not imply that compassion is an altogether (or even necessarily) nefarious emotion. What it does suggest, however, is that compassion is – in practical terms – an ambivalent political emotion with both positive and negative consequences. Care and cruelty may not always be that far apart.
Acknowledgements
Drafts of this article were presented at several conferences, including the 2010 annual meeting of the International Studies Association, the SGIR 7th Pan-European Conference on IR (2010), and the 7th European Peace Research Association General Conference (2011). For their constructive and detailed feedback, we would like to thank Chris Reus-Smit, Duncan Snidal, Alexander Wendt, and three anonymous reviewers at International Theory. In addition, we are grateful for financial support from the Academy of Finland and support (intellectual, financial and otherwise) from Mika Aaltola, Michael Barnett and numerous other friends and colleagues.
