Ethics in rational cosmopolitanism

For rationalist cosmopolitans, ethics are defined in terms of three principles: individualism, universality, and egalitarianism (Atack, Reference Atack2005: 42; Marchetti, Reference Marchetti2008: 36). Thus, in order to be considered ethical, a rule must conform, at the very least, to the ideas that the individual is the fundamental unit of analysis to whom all ethical principles apply, that all moral rules must be universally applicable, and that all individuals in this universal moral realm must be afforded equal respect. Drawing on the enlightenment commitment to individual reflective autonomy, Kant, following the Stoics, argued that universal laws find their origins in the wills of all rational beings (1948: 93; Nussbaum, Reference Nussbaum1997: 30; Hayden, Reference Hayden2005: 13). In doing so, he identified the individual autonomy that is central to rationalist cosmopolitan ethics with ‘the individual exercise of reason’ (Frazer, Reference Frazer2007: 758) and argued that the ‘universal supremacy of rational beings in themselves as ends’ is derived from ‘pure reason’ alone (Kant, Reference Kant1948: 93).

In rationalist cosmopolitan thought, however, reason does not simply stand in close relationship to the principles of individualism and universality, but provides the means by which egalitarianism is achieved. In particular, the relationship between reason and the principle of impartiality, according to which egalitarianism is operationalized, is central to contemporary rationalist cosmopolitan understandings of the nature of ethics itself. ‘[To] reason ethically, to consider things from the moral point of view’, rationalist cosmopolitans argue, ‘is to adopt an impartial standpoint’ (Cottingham, Reference Cottingham1983: 83). The ‘impartiality thesis’ that Singer, Rawls, and others advocate, argues that ‘when we are making moral decisions…we ought not to give any special weight to our own desires and interests; instead of giving preferential treatment to ourselves, or to members of our own particular group, we should try to adopt a neutral standpoint, detaching ourselves as far as possible from our own special desires and involvements’ (Cottingham, Reference Cottingham1983: 83).

Impartiality is thus the guiding principle that ensures that cosmopolitan ethics are indeed universal and egalitarian. In this vein, Singer argues, from a utilitarian perspective, that ‘an ethical principle cannot be justified in relation to any partial or sectional group’ because ‘[e]thics takes a universal point of view’ (Reference Singer1993: 11). In his famous and often repeated analogy, Singer drew on the principle of impartiality to argue that those of us living in relative affluence have an obligation to assist strangers living in poverty in other parts of the world. In particular, he was concerned with the lack of assistance being afforded to the several million Bengali refugees who were, at the time, living ‘on the edge of starvation’ in refugee camps in India (Reference Singer2002a: 156). By way of analogy, Singer argued that ‘if I am walking past a shallow pond and see a child drowning in it, I ought to wade in and pull the child out’ (Reference Singer1972: 231). To do so, he reasoned that the negative consequences associated with rescuing the child, getting my clothes muddy, are insignificant in comparison to the negative consequences associated with leaving the child to drown (1972: 231). He then argued that the vast majority of people living in the developed world have the same obligation to, at a relatively small cost to themselves, assist the Bengali refugees. In this case, the principle of impartiality dictates that ‘[i]t makes no moral difference whether the person I can help is a neighbor's child ten yards from me or a Bengali whose name I shall never know, 10,000 miles away’ (1972: 231–232). We will return to this analogy later and consider the role that emotions may play in it, but for now it suffices to say that adherence to the principle of impartiality is what facilitates universalism and egalitarianism in consequentialist variants of rationalist cosmopolitanism.

John Rawls’ deontological theory of justice as fairness is no less reliant on the concept of impartiality (Reference Rawls1971: 30). His ‘veil of ignorance’, a hypothetical scenario in which individuals are required to determine the principle of justice without reference to their own personal circumstances, enforces impartiality by moderating selfish rational choices with ‘reasonableness’ (Reference Rawls2005: 490). Defending what is often portrayed as the ‘extreme’ rationalism of Rawls’ theory of justice, Barry argues that ‘justice as impartiality’ can be conceived as ‘[a] theory of justice which makes it turn on the terms of reasonable agreement’ (Reference Barry1996: 7). ‘Reason, in this context’, Barry explains, ‘means reasoned argument, from premises that are in principle open to everyone to accept’ (Reference Barry1996: 7). From this starting point Barry thus attempts to formulate an impartial ethic that can cope with the widely differentiated understandings of the good that mark international politics (Reference Barry1996: 119).Footnote 6

Emotion in rationalist cosmopolitanism

Within theories that conceive ethics as impartial and understand impartiality as the product of reasoned argument, the relationship between reason and emotion is one of master and slave. Kant thus argued that for an action to have ‘genuine moral worth’ it must be done ‘not from inclination but from duty,…the necessity to act out of reverence for the law’ (Reference Kant1948: 64, 66). Although he reluctantly acknowledged that ‘reverence is a feeling’, Kant defined it in terms of reason as being ‘self-produced by a rational concept’ rather than ‘received through outside influence’ (Reference Kant1948: 66). Thus, for Kant, our duty to revere the law which formed the basis of ‘moral goodness’ can only consist in the thought of ‘a rational being’ (Reference Kant1948: 66). It is guided not by the emotions, inclinations, or selfish interests, but by the demands of reason.

Similarly, for rationalist consequentialist cosmopolitans, the demands of reason over-ride judgements driven by emotions. Thus, classical utilitarians argue that where feelings contradict reasoned calculations of utility, they ought to be discarded on the grounds of being ‘irrational’ (Williams, Reference Williams2004: 212). In this vein, Singer argues that if we simply ‘accept our feelings’ without engaging in reasoned reflection, ‘we would not be able to decide which of our intuitive inclinations to endorse and support and which to oppose’ (Reference Singer2002a: 163–164). In doing so, he ultimately suggests that our feelings and intuitive inclinations are only valid insofar as they accord with the reasoned reflections that constitute ethical judgements.

The relationship between reason and emotion in rationalist theories of ethics can be summarized in terms of three prevailing assumptions about the nature of emotion. The first assumption is that emotions are inescapably selfish and is derived, at least in part, from their conceptualization as mere feelings in much rationalist cosmopolitan thought. According to Kant, ‘[t]he appeal to the principle of moral feeling’ is not only ‘superficial’ but is incapable ‘of providing a uniform measure of good and evil’ (1948: 103). Emotions, or ‘moral feelings’, differ from ‘one another by an infinity of degrees’ and are thus a function of individual inclination rather than a source of universal moral judgement (1948: 103). As mere feelings, emotions can be nothing other than subjective inclinations with a tendency to self-interest, although Kant did acknowledge, with Francis Hutcheson, the moral feeling of ‘sympathy for the happiness of others’ (1948: 103).

The second assumption is implicit in the explication of the three fundamental principles of rationalist cosmopolitan thought and conceives emotions as being wholly distinct from reason. In rationalist cosmopolitan thought, emotion is consistently conceived as an inappropriate alternative to reason in processes of ethical deliberation. Thus, Rawls’ explicit Kantianism leads him to emphasize ‘autonomy and rationality as the defining characteristics of moral subjects’, to pursue a ‘rigid separation of reason from feeling’, and to, at least outwardly, refuse ‘to allow feeling any place in the formulation of moral principles’ (Okin, Reference Okin1989: 231).Footnote 7 As such, rationalist cosmopolitans maintain that the emotions ought to be controlled by their master, reason, lest they contaminate the formulation of moral judgements.

On the basis of the first two assumptions, the third contends that emotions, being inescapably selfish and fundamentally distinct from reason, are incapable of contributing to impartial ethical deliberation. Considered together, rationalist cosmopolitan assumptions about the nature of ethics and the nature of emotions thus lead to the conclusion that ethics ought to be based on reason and not emotion. However, even if we are to accept the, admittedly contested, claims that ethics ought to be individualist, universal, egalitarian and, by extension, impartial, it does not necessarily follow that emotions ought to be omitted from processes of ethical deliberation. In particular, the assumptions about the nature of emotions and their relationship to reason that, as discussed above, mark rationalist cosmopolitan thought, are unfounded. As suggested earlier, emotions may not be synonymous with mere feelings but include cognitive components. This not only renders them capable of contributing to ethical deliberations, but establishes a basic connection between emotions and reason.

From this starting point, the following two sections challenge the three basic assumptions that lead rationalist cosmopolitans to reject the place of emotion in processes of ethical deliberation. They demonstrate that emotions are not the dichotomous opponent of reason but are, in fact, central to processes of reasoning and rational thought, that emotions are capable of reflection that is not selfish or self-centered, and that emotions are capable of rendering impartial judgements. To do so, they draw on the relationship between reason and emotion posited by the moral sentiment theorists of the 18th century and recent advances in neuroscience and psychology to argue that emotions play and ought to play a role in processes of ethical deliberation.

Reason and sentiment

Following Shaftesbury and his defender, Francis Hutcheson (1694–1746), Hume argued that ‘morality is determined by sentiment’ (1998: App.1.10, emphasis added). According to Hume, the emotional responses we have to the impressions we experience determine whether their constitutive actions are deemed moral or immoral. While enacting moral principles ‘call[s] forth…favourable and affectionate sentiments’, immorality elicits favourable ones (Hume, Reference Hume1998: 2.5). On that basis, our sentiments determine whether particular actions are judged moral or immoral.

The relationship between emotion and reason in Hume's moral thought is thus one of master and slave. Reason, he argued, ‘is and ought only to be the slave of the passions, and can never pretend to any other office than to serve and obey’ (2000: 2.3.3.4). However, Hume did not view reason and emotion as dichotomous opponents but explicitly conceived them as being ‘uncompounded and inseparable’ (Reference Hume2000: 3.2.2.14). Defined as ‘the discovery of truth or falsehood’ (Hume, Reference Hume2000: III.I.I), reason functions as a ‘calculator, an instrument of analysis’ (Wolin, Reference Wolin1954: 1002) that assists us in understanding our passions. Reason and emotions or passions, as he referred to them interchangeably (Hume, Reference Hume2000: 2.1.1.3), thus work in concert with one another in making moral judgements.

However, Hume was not simply concerned with the role that emotions play in making moral judgements but recognized that action constitutes an essential part of ethics. Thus, Hume's often repeated phrase about the enslavement of reason reflected his observation that reason, ‘being cool and disengaged, is no motive to action’ (Reference Hume1998: App.1.21). Reason, he argued, can only help us to reflect upon our emotions but cannot, by itself, motivate ethical action in response to the moral judgements we make (1998: App.1.21).

Selfishness

In accordance with the passions the central authoritative role in making moral judgements and motivating ethical action, the moral sentiment theorists challenged both the ‘selfish theories’ of John Locke and Thomas Hobbes, and the claim that emotions, and hence moral judgements derived from them, are inescapably selfish (Shaftesbury, Reference Shaftesbury1897; Garrett, Reference Garrett2002: 8). Hutcheson, in particular, argued that human nature is marked not only by self-love but by benevolent affections towards others (Hutcheson, Reference Hutcheson2002: II.i). Although our moral sense may be ‘counterbalanced by Interest’, legitimate moral judgements are not derived from the exclusive consideration of our own advantage (Hutcheson, Reference Hutcheson2002: II.I; 2004: II.I.V). Thus, the central guiding principle of Hutcheson's moral sense theory was the sentiment of benevolence, pursued not with one's own interests in mind, but on a disinterested basis (2004: 92.III).

Hume confronted the claim that moral sentiment theory is inherently self-interested by distinguishing between two different species of sentiment, those derived from selfish considerations of utility, and universal sentiments drawn from the ‘principles of humanity’ (1998: 5.39; 9.7). This second form of ‘sentiment common to all mankind’ is central to the ‘very notion of morals’ itself and forms the basis of moral judgement (1998: 9.5). Hume thus recognized that not all emotional responses elicit legitimate moral judgements. Those borne of purely selfish interests stand outside the bounds of morality. In contrast, sentiments common to all mankind produce legitimate universal moral principles.Footnote 8

Hume also refuted claims that the passions are inherently selfish by reference to the concepts of sympathy and benevolence. For Hume, sympathy could be defined as a passion akin to benevolence that ‘makes me concerned for the present sorrows of a stranger’ or as ‘a faculty of mind with an informational function’ (Hume, Reference Hume2000: 2.2.9.13; Krause, Reference Krause2008: 79). In this second sense, sympathy is not a passion itself but a function charged with communicating the passions of others and provoking in us similar passions and sentiments (Hume, Reference Hume2000: 3.3.2.2). It is not the basis on which we judge the actions of others, but provides a conduit for the sentiments of disapproval and approval that constitute moral judgements.

In part, it is the very functioning of the faculty of sympathy in Hume's theory that renders moral sentiments capable of contributing to unselfish judgements. For Hume, sympathy is not a function of unitary independent human actors but like all affections ‘readily pass[es] from one person to another and beget[s] correspondent movements in every human creature’ (Hume in Krause, Reference Krause2008: 81). Thus, sentiments borne of sympathy that provide moral judgements about the extent to which the actions of others are acceptable or unacceptable, are not the subjective view of a single individual but are derived from the reverberation of sympathy ‘within and between persons in complex ways’ (Krause, Reference Krause2008: 81).

Impartiality

By approaching the moral sentiments from a generalized perspective, Hume made impartiality possible within his approach. He argued that ‘Tis only when a character is consider'd in general, without reference to our particular interest, that it causes such a feeling or sentiment, as denominates it morally good or evil’ (Hume, Reference Hume2000: 3.1.2.4). Thus, where a response is derived purely from self-interest and does not accord with a generalized and hence impartial moral sentiment, it is an illegitimate basis on which to make a moral judgement.

However, critics of Hume's theory argued that in some circumstances discerning precisely what the generalized perspective is might be difficult as some actions may genuinely invoke conflicting sentiments, such as gratitude and resentment, in different people. The question then becomes one of how to choose between these sentiments when making a judgement. At the same time, other critics argued that Hume's theory remained inherently selfish on the grounds that it is not possible to judge oneself using benevolence.Footnote 9 Adam Smith sought to overcome both of these problems through his discussion of sympathy and by introducing the impartial spectator to moral sentiment theory.

According to Smith, sympathy between individuals is one of the most conspicuous features of human existence. ‘That we often derive sorrow from the sorrow of others’, he argued, ‘is a matter of fact too obvious to require any instances to prove it; for this sentiment, like all the other original passions of human nature, is by no means confined to the virtuous and humane, though they perhaps may feel it with the most exquisite sensibility’ (1984: I.I.I.1, p. 60). In particular, Smith observed that witnessing the pain or distress of another ‘excites’ in the observer ‘some degree of the same emotion’ (Reference Smith1984: I.I.I.2). Hypothetically changing ‘places in fancy with the sufferer’, our ‘fellow-feeling for the misery of others’ thereby elicits some degree of the sufferer's response in ourselves (1984: I.I.I.3). On this basis, Smith argued, we determine the wrongfulness of hurtful actions inflicted on others.

Like Hume, however, Smith also acknowledged that not all emotional responses are acceptable. For example, rather than provoking a sympathetic response, ‘[t]he furious behaviour of an angry man is more likely to exasperate us against himself than against his enemies’ (1984: I.i.I.7). Thus, the act of sympathizing with another entails a judgement about the propriety of the original passion being observed. As Smith explained: ‘When the original passions of the person principally concerned are in perfect concord with the sympathetic emotions of the spectator, they necessarily appear to be just and proper, and suitable to their objects’ (I.I.iii.1). Conversely, we are ‘disgusted with that clamorous grief, which, without any delicacy, calls upon our compassion with sighs and tears and importunate lamentations’ and detest anger when its fury is indulged ‘without check or restraint’ (1984: I.I.V.3; I.I.V.4). To choose between different responses thus requires a judgement about the propriety of those responses to be made. To explain the basis on which this judgement is made, Smith introduced the concept of the impartial spectator.

The impartial spectator is an ‘imagined ideal’ that Smith ‘deployed to establish the existence of an objective perspective which can be learned and then used’ (Haakonssen, Reference Haakonssen2002: xv), not only to make impartial judgements about the actions of others but to ‘temper one's own tendency towards selfish anti-social behaviour’ (Young, Reference Young1986: 366). According to Smith, the impartial spectator is a ‘judge within’ that allows us to view our own actions from a disinterested perspective and properly compare our interests to those of others (1984: III.3.1). It tells us whether our behavior is ‘something which other men can go along with’ (Smith, Reference Smith1984: II.ii.2.1). If it is not, we lose ‘the indulgence of the spectators…[who] readily…sympathize with the natural resentment of the injured’, and become the ‘object of their hatred and indignation’ (Smith, Reference Smith1984: II.ii.2.1). With this, our sentiments are tempered, internally by the impartial spectator, and externally by other human observers, thereby limiting their selfishness and making impartiality a possibility.

In reality, however, the notion of impartiality presented in Smith's work operates in conjunction with what might be called ‘indiscriminate partiality’.Footnote 10 Sympathy, by its very nature, implies partiality with the sentiments of others, be the expressions of joy or manifestations of suffering. That we sympathize with another suggests that we are partial to their emotional response to an action or event. When we sympathize with another, we make two judgements. The first expresses that the object of our sympathy was right in their initial judgement and the second deems the emotional expression of that judgement to be appropriate. Thus, to return to Smith's example, we have no sympathy for the anger of one who has not been wronged or where the wrong is so slight as not to warrant a furious response. However, the principle of impartiality dictates that the judgements we make are undertaken on a disinterested basis. Our sympathy must thus be afforded to all others without recourse to our own interests; that is, it must be expressed on an indiscriminate basis. Smith's ‘impartial spectator’ ensures that our expressions of partiality are made on an impartial basis. It allows us to take on the suffering of another, as if it was our own and thereby to be motivated to alleviate that suffering, but without our judgements and their associated actions being fundamentally self-interested.

In sum, the moral sentiment theorists of the 18th century presented two main arguments defending the role that emotions play and ought to play in processes of ethical deliberation. First, Hume argued that emotions play an important role in making moral judgements and motivating ethical actions and, further to this, that emotion is central to reason. Underpinning this argument is the explicit contention that reason and emotion are inseparable and uncompounded. This is in direct contrast to the rationalist view that emotions ought to be omitted from processes of ethical deliberation, which assumes that reason and emotion are separable forms of reflective thought. Second, Hume and, in particular Smith, demonstrated that the basis on which emotions are ordinarily rejected in theories of ethics is false by establishing that emotions are capable of rendering impartial judgements.

The legacy of moral sentiment theory

Although it attracted considerable criticism from Thomas Reid, Adam Ferguson, and Lord Kames, Smith's ‘theory of sympathy’ remained popular into the early 19th century. Later in the 19th and 20th centuries, however, three factors contributed to its demise. The first was the ‘imaginary and exaggerated’ ‘Adam Smith Problem’ which fixated on the apparent incompatibility of Smith's moral sentiment theory with the ‘self-interested, profit-maximizing ethic which supposedly underlies The Wealth of Nations’ (Young, Reference Young1986: 365; Reeder, Reference Reeder1997: viii). Second, the moral sentiment theories of Smith and, in particular, Hume, became subsumed by utilitarianism (Sprague, Reference Sprague1954: 794). With its emphasis on rational calculation, utilitarianism, ironically a central element of both Hume and Smith's moral theories, pushed sentiment-based theories to the outer margins of scholarship where they have largely remained. Finally, with the rise to popularity of Kantian cosmopolitanism, the subjugation of moral sentiment theory and, with it, emotion-based theories of ethical deliberation, was complete.

Despite its general marginalization, moral sentiment theory has enjoyed three minor revivals in subsequent scholarship. In the late 19th century, the German philosopher and psychologist Franz Brentano's work, The Origin of Our Knowledge of Right and Wrong (1889), sought to find an ethic on emotions and, following the moral sentiment theorists, drew a ‘strong analogy between emotion and judgement’. In the late 20th century, feminist writers such as Carol Gilligan and Fiona Robinson drew on elements of moral sentiment theory in developing their ‘ethics of care’ approaches. Thus, Gilligan noted that ‘[t]he age-old split between thinking and feeling underlies many of the clichés and stereotypes concerning the differences between the sexes’ (Reference Gilligan1982: 69). Rather than disputing this split in its entirety however, Gilligan argued that they reflect the existence of ‘two modes of judging, two different constructions of the moral domain’, masculine rational ethics and feminine ethics founded on sympathy and care (Reference Gilligan1982: 69). As an advocate of the ethics of care she, unsurprisingly, endorsed the latter mode of judgement over the former. Making the connection between the emotionally derived ethics of care and moral sentiment theory explicit, Annette Baier argued, ironically, given his reputation for misogyny, that Hume ‘should be given the status of [an] honorary’ woman on account of his feminine account of the role that emotions play in ethical deliberation (Reference Baier1992: 2).

However, there are two major objections to be leveled at feminist accounts of emotional ethics exemplified by proponents of the ‘ethics of care’. The first is that an essentialist connection between masculinity and rationality, and femininity and emotion, is highly disputed. At the height of debate on this issue in the late 1980s and 1990s, Joan Tronto argued that there is little empirical evidence to support Gilligan's account of gender differentiation as a ‘definitive description’ (Reference Tronto1993: 84). In doing so, she echoed Susan Moller Okin's earlier claim that ‘the evidence for differences in women's and men's ways of thinking is not (at least yet) very clear; neither is the evidence about the source of whatever differences there might be’ (1989: 15). More recently, however, psychological studies by Nancy Clopton and Gwendolyn Sorrell (Reference Clopton and Sorrell2006) and Michael Pratt et al. have found that gender differences in moral reasoning ‘were not as pervasive as Gilligan argues’ (2006: 373).

Second, feminist accounts of the ethics of care place too much emphasis on the emotions to the detriment of reason and rationality. As mentioned earlier, it is not the intention of this article to argue for the replacement of reason with emotion or even to establish the superiority of the emotions over reason, but to demonstrate that both play essential roles in making moral judgements and motivating ethical actions. This second objection is borne out in recent developments in the fields of neuropsychology and political psychology that demonstrate that the relationship between reason and emotion is as close as Hume proposed. As such, they provide substantial evidence against the rationalist conception of reason and emotion as separable forms of reflective thought and establish the essential role of emotions in processes of ethical deliberation.

Most recently, James Brassett (Reference Brassett2010) has sought to develop a critical cosmopolitan response to trauma that relies on the sentiments, particularly those shared through the function of empathy. In doing so, he draws on Richard Rorty's pragmatic argument in favour of sentimentalist education. According to Rorty, a pragmatic approach to questions of morality and, in particular, human rights, must start ‘from the fact that the emergence of the human rights culture seems to own nothing to increased moral knowledge, and everything to hearing sad and sentimental stories’ (1998: 172). For him, the sorts of foundationalist arguments about morality most commonly associated with Kantian notions of truth and rationality ought to be put aside to allow us to ‘concentrate our energies on manipulating sentiments, or sentimental education’ (Rorty, Reference Rorty1998: 176). According to Rorty, this ‘flexible sentimentality’ is not only capable of being cultivated but, unlike ‘rigorous rationality’, is what unites human beings and, when educated effectively, can drive them to be nicer to each other (Reference Rorty1998: 183, 177).

Drawing on Rorty, Brassett argues that cultivating understandings of our mutual vulnerability, derived not simply from the identification of subjects of trauma or even sympathy with them, but from a ‘politics of empathy’ might transform ‘processes of global governance’ to better address traumatic events (2010: 14–15). In doing so, he goes beyond Rorty's emphasis on sympathy to suggest that ‘the existential nature of vulnerability ‘might’ allow for the identification and nurturing of empathy as a more appropriate/powerful ethical sentiment’ (Brassett, Reference Brassett2010: 24).

Although the type of sentimental cosmopolitanism advocated in this article is broadly sympathetic to the agendas pursued by Rorty and Brassett, it is not merely pragmatic in motivation. That is, it does not simply argue that the emotions ought to be afforded a central role in making moral judgements and motivating ethical action because doing so produces the most effective outcomes. Rather, by drawing on advances in psychology and neuroscience in the following section, it suggests that the role played by the emotions in making moral judgements and motivating ethical action has a natural basis. The sentiments are not merely manipulable or limitlessly flexible phenomena, but, as the products of particular neurological and psychological processes, have a material foundation. This is not to suggest that emotions are wholly natural phenomena and thus lack a social dimension. On the contrary, the fact that emotions can be manipulated, cultivated, or controlled without directly altering their physiological aspects suggests that, at the very least, their expression is, in part, socially constructed (Griffiths, Reference Griffiths1997: 139–140). However, the natural foundations of emotions delimit the ways in which they may be effectively manipulated, demarcate the bounds within which emotions may be cultivated, educated, and restrained and, as we will see in the following section, reveal how emotions interact with reason in making moral judgements.

Reason and emotion in decision neuroscience

Like the standard separation of reason and emotion in International Relations, for much of its history, the study of decision neuroscience has been dominated by the assumption that cognition and affect constitute independent mechanisms for reflective thought (Simon, Reference Simon1979).Footnote 12 Although some contemporary scholars maintain this distinction (Marcus, Reference Marcus2000), others dispute their separation (Bechara et al., Reference Bechara, Damasio, Tranel and Damasio1997). In particular, Damasio's (Reference Damasio1994) study of the relationship between reason and emotion in patients with lesions of the ventro-medial prefrontal cortex of the brain demonstrated that emotion is central to reason.

To establish a relationship between reason and emotion, Damasio drew on several cases of patients who suffered damage to the ventro-medial prefrontal cortexes of their brains, including the famous case of Phineas GageFootnote 13 and the more recent case of ‘Elliot’ (1994: 34–51). Elliot was a patient with brain tumour who, in the course of having his tumour removed sustained damage to his prefrontal cortex. After the operation, the once successful businessman found that he was unable to make effective, reasoned judgements, or choose appropriate actions. As a result, he lost his job, started several ill-considered businesses, and ended up bankrupt and divorced.

Standardized psychological and neurological tests performed after the operation revealed not only that Elliot possessed ‘a superior intellect’ but that his ‘perceptual ability, past memory, short-term memory, new learning, language, and…ability to do arithmetic were intact’ (Damasio, Reference Damasio1994: 41). What Elliot had lost was his ability to respond emotionally to stimuli that had once evoked strong emotional reactions (Damasio, Reference Damasio1994: 45). What is more, while Elliot remained able to make choices when presented with hypothetical test scenarios, in ‘real life’ he was unable to make effective decisions (Damasio, Reference Damasio1994: 48–49).Footnote 14

Damasio thus concluded that ‘[c]ertain aspects of the process of emotion and feeling are indispensable for rationality’ (Reference Damasio1994: 53). Emotion, he argued, ‘was in the loop of reason, and…could assist the reasoning process rather than necessarily disturb it, as was commonly assumed’ (2005: x–xi). Contrary to some interpretations of his work, in doing so Damasio did not ‘suggest that emotion was a substitute for reasoning’ or ‘set emotion against reason’ but rather confirmed Hume's argument that emotion ought to be seen ‘as at least assisting reason and at best holding a dialogue with it’ (Reference Damasio2005: xi, xiii). Descartes’ error, as the title of Damasio's work hints, was therefore to accept a Cartesian dualist division of mind and body, reason, and emotion.Footnote 15

The relationship between reason and emotion suggested by the cases of Phineas Gage, Elliot, and others have been further investigated using neuroimaging techniques including functional magnetic resonance imaging. Neuroimaging techniques offer unique advantages in the study of emotion by complementing neurological and psychological studies of patients who have suffered brain injuries and enabling researchers to examine ‘brain function in the intact human’ (Wager et al., Reference Wager, Barrett, Bliss-Moreau, Lindquist, Duncan, Kober, Joseph, Davidson and Mize2008: 250). Identifying those regions of the brain responsible for particular functions no longer relies on correlating instances of brain damage with particular types of functional impairment, as occurred in the case of Phineas Gage. fMRI technology has thus allowed researchers to identify the particular brain regions and structures activated when we engage in reasoning and experience emotion (Phan et al., Reference Phan, Wager, Taylor and Liberzon2002).

fMRI technology has contributed, in particular, to several important studies of the neural activity involved in processes of moral deliberation. For example, Jorge Moll et al. used fMRI techniques to identify the brain regions activated by the experience of basic and moral emotions, those ‘readily evoked by the perception of moral violations’ and found that emotional experience is linked to moral appraisal (Reference Moll, de Oliveira-Souza, Eslinger, Bramati, Mourão-Miranda, Andreiuelo and Pessoa2002: 2730, 2736). Similarly, Joshua Greene et al. (Reference Greene, Brian Sommerville, Nystrom, Darby and Cohen2001) also used neuroimaging technology to study the nature and extent of the role that emotions play in making moral judgements. In doing so, they demonstrated not only that there is a correlation between neural activity associated with emotions and making moral judgements, but that emotional responses ‘have an influence on and are not merely incidental to moral judgements’ (2001: 2107).

In a separate study, again using fMRI technology, Greene (Reference Greene2008) found that both deontological and consequentialist forms of rationalist cosmopolitan decision-making include emotions in their deliberations. Surprisingly, given the close relationship between moral sentiment theory and utilitarianism, he found that ‘deontological patterns of moral judgement are driven by emotional responses while consequentialist judgements are driven by ‘cognitive’ processes’ but also include emotions in their calculations (Greene, Reference Greene2008: 59). By also considering the time subjects took to make a decision in response to each of the dilemmas, Greene concluded that deontological judgements, as immediate responses, are fundamentally driven by our emotions, while consequentialist judgements involve the weighing of emotional and cognitive responses over a longer period of time.

Greene's findings have a number of significant implications for how we understand processes of cosmopolitan moral reasoning. First, they pose a serious challenge to the historical stereotype that assumes that ‘consequentialism is more emotional’ due to its close association with moral sentiment theory, and ‘deontology is more ‘cognitive’ (encompassing Kant's ‘rationalist’ tradition)’ (Greene, Reference Greene2008: 41). Second, by demonstrating that deontological reasoning is driven by emotions, they bring into question the assumed rationalist underpinnings and status of deontological cosmopolitanism. They show deontological cosmopolitanism, as presented in purely rationalist terms, to be ‘a kind of moral confabulation’ (Greene, Reference Greene2008: 63). Thus, rather than being rationally driven, deontology provides a rational justification for emotionally driven judgements; it is, according to Greene, ‘a natural “cognitive” expression of our deepest moral emotions’ (Reference Greene2008: 63).Footnote 16 Finally, these findings not only support the general argument that processes of moral deliberation incorporate both reason and emotion but demonstrate, more specifically, that both play key roles in the two dominant forms of rationalist cosmopolitan thought, consequentialism, and deontology.Footnote 17 In doing so, they accord with Hume's view that all forms of moral judgement, by their very nature, include emotions (Hume, Reference Hume1998: App.1.2; Greene, Reference Greene2008: 41, 64).

A normative argument?

Advances in decision neuroscience, including the findings of studies by Damasio (Reference Damasio1994), Bechara (Reference Bechara2004), Moll et al. (Reference Moll, de Oliveira-Souza, Eslinger, Bramati, Mourão-Miranda, Andreiuelo and Pessoa2002), Greene et al. (Reference Greene, Brian Sommerville, Nystrom, Darby and Cohen2001), and Greene (Reference Greene2008) have all lent support to Hume's claim that emotion is central to reason and demonstrated that reason and emotion are engaged in both deontological and consequentialist variants of cosmopolitan moral deliberation. As phenomena inseparable and indivisible from processes of rational deliberation, it thus makes little sense to argue, as rationalists conventionally do, that moral deliberation ought to take place on the basis of reason alone. Extending that sort of rationalist logic to its natural end, moral deliberation would simply not be possible as emotions cannot be excised from the deliberative process and, more than that, play an indispensable role in making rational judgements. That is, without emotion, there can be no reason.

The question of whether emotions ought to play a role in processes of ethical deliberation is thus null and void. This is not to breach Hume's law that an ‘ought’ cannot be derived from an ‘is’, but is to recognize that ‘ought’ may be constrained by ‘is’. That is, in order to be meaningful, ‘oughts’ must reflect actions that can, in fact, be enacted. A moral principle that requires individuals to do something that is functionally impossible for human beings to perform is effectively meaningless.Footnote 18 The ‘is’ in this case, that emotions are inseparable from reason in processes of ethical deliberation, thus presents normative arguments about the place of emotion in ethics with a practical constraint.

However, two legitimate questions about the contribution of emotions to moral judgements and ethical actions remain and will be addressed in the next section. The first concerns the nature of the contribution made by emotions to processes of ethical deliberation and, in particular, whether emotions hinder our ability to make effective, reasoned judgements. Acknowledging that emotions play a role in making moral judgements is one thing, but it is quite another to argue that contribution is a positive one. The second concerns the extent of that contribution and, in particular, the role that emotions play in motivating ethical actions in response to moral judgements.

Cognition and affect in political psychology

It is commonly assumed that emotions have an essentially negative impact on processes of ethical deliberation. According to the prevailing assumption, emotions exert a disruptive influence over decision-making processes and, in doing so, render judgements ‘irrational’, and hence less reliable than they otherwise might have been (Isen, Reference Isen2008: 548). Some evidence seems to support this claim. For example, Baron has demonstrated that moral decisions driven by intuitions, defined as ‘blind feelings’ and ‘reflective beliefs’, do not lead to the most optimal public policy outcomes (Reference Baron1998: 5, 2). He argues, instead, from a utilitarian perspective, that moral judgements must be based on quantitative analysis of the possible outcomes achieved by pursuing one course of action over another. Although intuitions ought not to be banished from the decision-making process entirely, he argues, they ought to play a secondary role to reason (1998: 19).

However, several psychological studies of emotions in politics contradict this finding. In particular, Marcus et al.'s examination of how emotions affect moral judgements about war and political leadership demonstrated that emotion is not ‘a secondary, derivative phenomenon’ but drives rational decision-making (Reference Marcus, Neuman and MacKuen2000: 105). Their study hypothesized that ‘people who experienced anxiety during the war – about the troops or about the possible expansion of the war – would think more deeply about policies and about the president's leadership’ and ‘[t]hose who were unmoved would not do so’ (Marcus et al., Reference Marcus, Neuman and MacKuen2000: 106). Their findings not only confirmed this hypothesis, but demonstrated that anxiety ‘encourages people to make rational judgements’ (2000: 108). Marcus, Neuman, and MacKuen's affective intelligence theory thus argues that ‘anxiety leads individuals to play closer attention to the environment and bring more information to task in processing information’ (Cassino and Lodge, Reference Cassino and Lodge2007: 105). In doing so, it also demonstrates that ‘when stimulated by their emotional systems’, voters who ordinarily vote a particular way out of habit, ‘think through their decisions and act as relatively well-informed rational voters’ (Marcus et al., Reference Marcus, Neuman and MacKuen2000: 129). That is, emotions do not hamper the ability to make rational decisions but rather drive rational calculations.

As Marcus, Neuman, and MacKuen focused on negative emotions, such as anxiety, sadness, fear, and anger, Isen (Reference Isen2008) considers the role that positive emotions have on decision-making processes. Contrary to previous studies, which found that positive emotions lead to careless thinking (Mackie and Worth, Reference Mackie and Worth1989, Reference Mackie and Worth1991; Bless et al., Reference Bless, Bohner, Schwarz and Strack1990; Schwarz, Reference Schwarz2002), Isen argues that positive affect can benefit thinking and decision-making (2008: 548). This is not to say that all emotions, at all times, have a positive impact on processes of ethical deliberation but rather that they are capable of doing so.

Having established that reason and emotion are dissociable and that emotions so not necessarily have a detrimental effect on processes of ethical deliberation, what remains to be considered is the extent of the role accorded emotions in ethics. At one end of the spectrum, a minimalist argument maintains that emotions help us to identify injustices. This argument is most prominently articulated by Amartya Sen who acknowledges the importance of emotions in diagnosing injustices, but argues that they must be coupled with ‘reasoned scrutiny’, lest they become uncritical and disengaged (Reference Sen2009: vii, viii, x). In thus arguing, Sen sees ‘no irreducible conflict’ between reason and emotion but rather maintains that they complement one another in processes of reflective thought (2009: 1, 39). What is more, he explicitly evokes the moral sentiment theories of Hume and, in particular, Smith, when he argues that our ‘sentiments and psychological concerns’ must be subjected to reasoned reflection (2009: 50). In particular, Sen draws on Smith's notion of the impartial spectator, although not by name, when he insists that we ‘view our sentiments from “a certain distance from us” ’ to ensure that our judgements are impartial (2009: 45).

For Sen, the process of ethical deliberation thus entails two steps. First, our emotions help us to identify injustices. We respond with anger, horror, resentment, or indignation to the injustices we experience ourselves and witness in others. These negative emotions, all of which signify different forms of disapprobation, provide provisional judgements about the occurrence of an injustice. Second, the judgements entailed by our emotional designation of a particular action as unjust are either justified or discarded in rational terms in a manner akin to consulting the impartial spectator. As the earlier discussion of Greene's work revealed, for consequentialists, such as Sen, this means incorporating both emotional responses and rational reflection in a weighing process, while for deontological cosmopolitans this entails formulating a rational justification for an essentially emotional judgement.Footnote 19

Combining this minimalist position with an understanding of the indivisible relationship between reason and emotion established earlier, an intermediary argument contends that emotions help us to identify injustices and then contribute to the formulation of reasoned justifications, both deontological and consequentialist, of those judgements. However, if we are to accept Hume's idea that ethics requires action (1998: App. 1.2), that it is not, as Onora O'Neill (Reference O'Neill1986a, Reference O'Neillb) put it, ‘a spectator sport’, then we must consider whether emotions not only help us to identify injustices, and contribute to the formulation of reasoned justifications of those judgements, but also guide actions undertaken in response to those injustices.

Evidence of a link between moral emotions and moral actions can be found in studies of moral judgement in psychopaths and, again, in patients with damage to the ventromedial area of the prefrontal cortex which demonstrate that emotions constrain immoral behavior and motivate moral actions (Haidt, Reference Haidt2001: 823). As Hervey Cleckly's landmark study, The Mask of Sanity (Reference Cleckly1941) revealed, psychopathic behavior can be explained by the absence or attenuation of the ‘major emotional accompaniments’ that ordinarily restrain immoral behavior (1976: 371). Further studies have suggested that the disconnection between cognition and affect that this entails occurs at the neurological level as psychopaths and other aggressive criminals exhibit different neural activity in the prefrontal cortex to non-criminal and non-psychopathic subjects (Raine, Reference Raine1997).

This finding echoes Damasio's work on patients with damage to the ventromedial prefrontal cortex. As discussed earlier with reference to the cases of Phineas Gage and Elliot, such patients experience ‘loss of emotional responsiveness to the world in general and to one's behavioural choices in particular’, what Damasio calls ‘acquired sociopathy’ (Damasio, Reference Damasio1994; Haidt, Reference Haidt2001: 824). Like the psychopaths of Cleckly's study, ‘[w]ith no moral sentiments to motivate and constrain them’, patients with damage to the ventromedial prefrontal cortex ‘simply do not care about the pain they cause and the lives they ruin’ (2001: 824). In the case of Elliot, this was seen in his inability to understand the impact that his reckless behaviour had on his family and resulted in divorce (Damasio, Reference Damasio1994: 37). Emotions, it thus seems, at the very least help to constrain unjust or unacceptable behaviour.

However, others argue that rather than simply constraining unjust behaviour, emotions help to motivate ethical action. As Carver et al. note, ‘[t]he notion that motivation and emotion are linked has long been a staple of emotion theories’ (Reference Carver, Sutton and Scheier2000: 741). For some, that connection is absolute, with the emotions system constituting ‘the primary motivational system for human behavior’ (Tomkins, Reference Tomkins1962; Izard, Reference Izard1971; Izard and Ackerman, Reference Izard and Ackerman2000: 253), while for others, emotions are just one factor that motivates behaviour (Frijda, Reference Frijda1986; Lazarus, Reference Lazarus1991a, Reference Lazarusb). Either way, a connection between emotion and motivation is posited.

Where specifically ethical behaviour is concerned, Batson and Shaw's (Reference Batson and Shaw1991) ‘empathy-altruism hypothesis’ posits the existence of a connection between emotions and helping others. In particular, their study found that ‘the more empathy felt for a person in need, the more motivation to have that need reduced’ (Batson and Shaw, Reference Batson and Shaw1991: 114). As Cialdini explains, ‘when we pay attention to another's suffering, we hurt in a way that frequently leads to helping’ (Reference Cialdini1991: 124). That hurt, he continues, in a manner akin to Hume and Smith's understanding of sympathy, ‘is vicarious and it is emotional.’ It is, in short, ‘a second-order, affective response to another's pain’ (1991: 124). While debate has been waged over whether or not in this context help is a function of altruism or egoism, directed at alleviating suffering or need for the primary benefit of the victim or the person witnessing, and vicariously experiencing their suffering (see exchange between Batson and Shaw, Reference Batson and Shaw1991 and their critics), it remains the case that emotion motivates ethical action. As we will see in the following section, this connection is of particular pertinence when we consider the problem of motivating action to alleviate world poverty.